Endocannabinoid receptors, dopamine system and brain volumetry in amyloid plaque producing APP/PS1 mice

(1)

Dissertations | eLisa neVaLainen | enDocannabinoiD receptors, Dopamine system anD... | no 337

uef.fi

pubLications of

the uniVersity of eastern finLanD Dissertations in Health Sciences

ISBN 978-952-61-2061-4 ISSN 1798-5706

Dissertations in Health Sciences

the uniVersity of eastern finLanD

ELISA NEVALAINEN

enDocannabinoiD receptors, Dopamine system anD brain VoLumetry in amyLoiD pLaque

proDucing app/ps1 mice

ELISA NEVALAINEN

This thesis examines how progressive brain amyloid accumulation affects neurotransmitters and brain volume in APP/PS1 transgenic mice modeling Al- zheimer’s disease (AD). Special emphasis is on the

endocannabinoid and dopamine systems that have received much less attention in AD pathology than the cholinergic system. Missing neurotrophic support has been suggested as one cause for degenerative changes in AD. This thesis assesses the role of brain-derived

neurotrophic factor on the integrity of neurons by crossing APP/PS1 mice with partially BDNF deficient

mice and comparing brain volumes and morphology of dopaminergic neurons between the four derived genotypes. The thesis combines functional autora- diography, in vivo and ex vivo magnetic resonance imaging, in vivo voltammetry and post-mortem im- munohistochemistry. The results reveal independent roles of amyloid-β accumulation and BDNF deficiency

in the amyloid-related neuropathology.

(2)

(3)

ELISA NEVALAINEN (NEE KÄRKKÄINEN)

Endocannabinoid receptors, dopamine system and brain volumetry in amyloid

plaque producing APP/PS1 mice

To be presented by permission of the Faculty of Health Sciences, University of Eastern Finland for public examination in Tietoteknia auditorium, Kuopio, on Friday, May 13^th 2016, at 12 noon

Publications of the University of Eastern Finland Dissertations in Health Sciences

Number 337

Department of Neurobiology, A.I. Virtanen Institute, Faculty of Health Sciences University of Eastern Finland

Kuopio 2016

(4)

Juvenes Print Tampere, 2016

Series Editors:

Professor Veli-Matti Kosma, M.D., Ph.D.

Institute of Clinical Medicine, Pathology Faculty of Health Sciences

Professor Hannele Turunen, Ph.D.

Department of Nursing Science Faculty of Health Sciences

Professor Olli Gröhn, Ph.D.

A.I. Virtanen Institute for Molecular Sciences Faculty of Health Sciences

Professor Kai Kaarniranta, M.D., Ph.D.

Institute of Clinical Medicine, Ophthalmology Faculty of Health Sciences

Lecturer Veli-Pekka Ranta, Ph.D. (pharmacy) School of Pharmacy

Faculty of Health Sciences

Distributor:

University of Eastern Finland Kuopio Campus Library

P.O.Box 1627 FI-70211 Kuopio, Finland http://www.uef.fi/kirjasto

ISBN (print): 978-952-61-2061-4 ISBN (pdf): 978-952-61-2062-1

ISSN (print): 1798-5706 ISSN (pdf): 1798-5714

ISSN-L: 1798-5706

(5)

Author’s address: Department of Neurobiology/A.I. Virtanen Institute University of Eastern Finland

KUOPIO FINLAND

Supervisors: Professor Heikki Tanila, MD, Ph.D.

Department of Neurobiology/A.I. Virtanen Institute University of Eastern Finland

KUOPIO FINLAND

Docent Jarmo T. Laitinen, Ph.D.

Institute of Biomedicine/School of Medicine University of Eastern Finland

KUOPIO FINLAND

Reviewers: Professor Raimo K. Tuominen, MD, Ph.D.

Faculty of Pharmacy University of Helsinki HELSINKI

FINLAND

Francisco Rafael López Picón, Ph.D.

Preclinical Imaging and Drug Research Turku PET Centre

TURKU FINLAND

Opponent: Professor Amanda Kiliaan, Ph.D.

Department of Anatomy and Department of Cognitive Neuroscience Faculty of Medicine/Donders Centre for Neuroscience

Radboud University Nijmegen Medical Centre NIJMEGEN

THE NETHERLANDS

(6)

(7)

Nevalainen (nee Kärkkäinen), Elisa

Endocannabinoid receptors, dopamine system and brain volumetry in amyloid plaque producing APP/PS1 mice

University of Eastern Finland, Faculty of Health Sciences

Publications of the University of Eastern Finland. Dissertations in Health Sciences 337. 2016. 73 p.

ISBN (print): 978-952-61-2061-4 ISBN (pdf): 978-952-61-2062-1 ISSN (print): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

ABSTRACT:

Alzheimer´s disease (AD) is the most prevalent disorder of the brain causing dementia. The key neuropathological features of AD include not only extracellular amyloid-β plaques, intraneuronal neurofibrillary tangles and cerebrovascular amyloid accumulation, but also the loss of synapses and neurons in vulnerable cell populations. These losses affect many of the brain signalling systems. In the studies included in this thesis we wanted to investigate the impact of progressive brain amyloidosis in APP/PS1 transgenic AD model mice on endocannabinoid (eCB) and dopamine (DA) systems and brain volumetry. In addition, we assessed the role of compromised brain derived neurotrophic factor (BDNF) support on the latter two outcome measures.

The cannabinoid CB1-receptor (CB1R) is among the most abundant G-protein-coupled receptors in the mammalian brain, and the eCB system is known to participate in the regulation of multiple important tasks, such as neuroprotection and synaptic plasticity.

Reported findings on CB1R levels and activation in both AD patients and animal models have proved ambiguous. To clarify the issue, we employed [³⁵S]GTPγS autoradiography in brain section pairs of amyloid plaque producing transgenic APP/PS1 and wild-type littermate control mice to assess functional CB1R activity in brain regions most severely affected by neuropathological processes in AD, namely the hippocampus, entorhinal cortex and medial frontal cortex. The autoradiography protocol was completed both in the absence and presence of dithiotreitol (DTT) to reveal possible redox-dependent changes in CB1R function. Five treatments were used: baseline, incubation with 10 μM GTPγS to assess nonspecific binding, and the nonselective CB1 and CB2 receptor agonist CP55940 in three concentrations. We found no significant alterations in CB1R functional activity in the studied brain regions of APP/PS1 transgenic mice and no significant redox-dependent changes.

Collectively with earlier published studies, these results indicate that at least in the early stage of AD there is no major loss of functional CB1R signalling and these patients should show normal responses to CB1R-active drug compounds.

There is a growing interest in using magnetic resonance imaging (MRI) on mouse models of AD-related brain amyloidosis to find a noninvasive means to monitor disease progression in preclinical intervention studies. Yet only few MRI volumetry studies have been conducted in AD model mice with ambiguous results. In attempt to clarify this and to compare the developmental volumetric changes due to BDNF deficiency with progressive changes due to amyloid accumulation we performed in vivo and ex vivo MRI in transgenic mice derived from crossing APP/PS1 mice with BDNF^+/‒ heterozygote knockout mice. We found decreased whole brain volume at 3 months and decreased cortical volume at both 3 and 8 months in vivo in BDNF^+/‒ mice but increased whole brain and cortical volumes at 8 months in APP/PS1 mice. Consistent with this, the post-mortem histological analysis showed decreased brain parenchymal area in BDNF^+/‒ mice but an increase in APP/PS1 mice. Importantly, we found an age × APP/PS1 interaction, and saw a relative cortical volume increase with age in APP/PS1 mice. In order to understand the cellular pathology underlying these changes, we performed extensive histological analysis focusing on neurogenesis, amyloid and astroglial

(8)

load. BDNF deficiency did not affect brain amyloid load or astrogliosis, but led to decreased dentate gyrus length, whereas APP/PS1 mice had significantly increased amyloid load, astrogliosis and decreased neurogenesis. Distinct and layer-specific effects of these genotypes were also found in the hippocampus. In conclusion, in contrast to AD patients, brain atrophy in amyloid producing mice appears to be masked by volume increase due to amyloid accumulation and especially accompanying astrogliosis. Our results indicate that cortical MRI volumetry can be used to some extent as a proxy to progressive brain amyloidosis in preclinical studies.

BDNF signalling disturbances have been frequently reported in AD. Specifically, BDNF levels fall but the ratio between its truncated, nonsignalling, TrkB.T1 and full-length TrkB.TK receptors increases in brains of AD patients and APP/PS1 mice. Dopaminergic (DAergic) system disturbances in AD and detrimental effects of BDNF signalling deficits on DAergic system functions have also been indicated. We have earlier revealed an interesting difference in spontaneous activity of these mice, such that APP/PS1 mice are hyperactive, whereas TrkB.T1 overexpressing mice are hypoactive. Since changes in DAergic signalling can influence spontaneous activity, we set out to find out the separate and combined effects of APP/PS1 and TrkB.T1 genotypes on the function and neuropathology of the nigrostriatal and mesolimbic DA systems. Employing in vivo voltammetry, we found normal short-term DA release in caudate-putamen (CPu) of mice carrying APP/PS1 or TrkB.T1 transgenes but impaired capacity to recruit more DA upon prolonged stimulation. However, mice carrying both transgenes did not differ from wild-type controls. Immunohistochemistry revealed normal density of tyrosine hydroxylase (TH) positive axon terminals in CPu and nucleus accumbens in all genotypes and intact presynaptic machinery for DA release and reuptake, as shown by unchanged levels of DAT, SNAP-25 and α-syn, and high correlations between levels of these proteins. However, we observed an increased number of DAergic neurons in substantia nigra of TrkB.T1 mice but unchanged density of DAergic neurons and fibres in the midbrain of APP/PS1 mice. This resulted in decreased TH levels per axon terminal in TrkB.T1 mice but unchanged levels in APP/PS1 mice. The observed increase in midbrain DA neurons in TrkB.T1 mice is a novel finding. We suggest that both APP/PS1 and TrkB.T1 genotypes disrupt DAergic signalling, but via separate mechanisms, and that this difference may explain the earlier findings on spontaneous activity.

National Library of Medicine Classification: WT 155, WL 104, WL 141.5.M2, WL 307, WL 314, WD 205.5.A6, QS 525, QT 21, QV 126, QV 504.5, QY 58, QY 60.R6

Medical Subject Headings: Alzheimer Disease; Amyloidosis; Atrophy; Autoradiography; Brain; Brain-Derived Neurotrophic Factor; Cannabinoids; Corpus Striatum; Disease Models, Animal; Dopamine; Electrophysiology;

Endocannabinoids; Entorhinal Cortex; Frontal Lobe; Hippocampus; Histology; Immunohistochemistry;

Magnetic Resonance Imaging; Mice, Transgenic; Neurogenesis; Neuronal Plasticity; Neuropathology;

Neuroprotection; Plaque, Amyloid; Presynaptic Terminals; Receptor, Cannabinoid, CB1; Receptor, trkB

(9)

Nevalainen (o.s. Kärkkäinen), Elisa

Endokannabinoidireseptorit, dopamiinijärjestelmä ja aivojen tilavuusmittaukset amyloidiplakkia tuottavilla APP/PS1 hiirillä

Itä-Suomen yliopisto, terveystieteiden tiedekunta

Publications of the University of Eastern Finland. Dissertations in Health Sciences 337. 2016. 73 s.

ISBN (print): 978-952-61-2061-4 ISBN (pdf): 978-952-61-2062-1 ISSN (print): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

TIIVISTELMÄ:

Alzheimerin tauti (AT) on yleisin dementiaa aiheuttava aivosairaus. AT:n keskeiset neuropatologiset tunnusmerkit ovat paitsi solunulkoiset amyloidi-β-plakit, hermosolujen sisäiset hermosäievyyhdet ja aivoverisuonten amyloidikertymät, myös synapsien ja hermosolujen kato haavoittuvissa solupopulaatioissa. Tämä kato puolestaan johtaa aivoissa monien signalointijärjestelmien häiriöihin. Väitöskirjan osatöissä halusimme tarkastella aivojen etenevän amyloidoosin vaikutuksia endokannabinoidi- (eCB) ja dopamiini- (DA) järjestelmiin sekä aivojen osien tilavuuteen APP/PS1-siirtogeenisellä AT:n hiirimallilla.

Lisäksi selvitimme aivoperäisen hermokasvutekijän (BDNF) vaikutusta DA-järjestelmään ja aivotilavuuksiin.

Kannabinoidireseptori CB1 (CB1R) on yksi yleisimmistä G-proteiinikytkentäisistä reseptoreista nisäkäsaivoissa, ja eCB-järjestelmän tiedetään säätelevän mm.

hermoyhteyksien muovautumista ja suojaavan hermosoluja yliärtyvyydeltä. AT-potilailla ja eläinmalleilla saadut tulokset aivojen CB1R-pitoisuuksista ja aktiivisuudesta ovat olleet ristiriitaisia. Tämän vuoksi tutkimme [³⁵S]GTPγS-autoradiografiamenetelmällä siirtogeenisten APP/PS1-hiirten ja villityyppisten hiirten aivoleikepareja määrittääksemme toiminnallisten CB1-reseptorien aktiivisuuden AT:lle herkillä aivoalueilla, hippokampuksessa sekä entorinaalisella ja mediaalisella etuaivokuorella. Saadaksemme esiin mahdolliset hapetus-pelkistysreaktioiden aiheuttamat muutokset CB1R-välitteisessä signaloinnissa teimme kokeet sekä pelkistävän dithiotreitolin kanssa että ilman sitä.

Käytimme viittä eri käsittelyä: perus- eli vertailutaso, inkubaatio 10 μM GTPγS kanssa epäspesifisen sitoutumisen arvioimiseksi ja kolme eri epäselektiivisen CB1- ja CB2- reseptoriagonisti CP55940:n pitoisuutta. Emme löytäneet merkitseviä muutoksia CB1R:n toiminnallisessa aktiivisuudessa tutkituilla aivoalueilla APP/PS1-siirtogeenisillä hiirillämme. Myöskään merkitseviä hapetus-pelkistysreaktoiden aiheuttamia muutoksia ei löytynyt. Yhdessä aiemmin julkaistujen tutkimusten kanssa tuloksemme viittaavat siihen, että ainakaan AT:n varhaisessa vaiheessa CB1R-signalointi ei ole merkittävästi heikentynyt, joten näiden potilaiden vasteiden CB1-reseptoriin vaikuttaville lääkeyhdisteille pitäisi olla tavanomaisia.

Kiinnostus magneettikuvauksen (MRI) käyttöä kohtaan AT:n amyloidoosia mallintavilla hiirimalleilla kasvaa, koska halutaan löytää kajoamaton keino, jolla seurata sairausprosessin etenemistä prekliinisissä interventiotutkimuksissa. Silti vasta harvoja MRI- tilavuusmittaustutkimuksia on tehty AT:a mallintavilla hiirillä, eivätkä tulokset ole olleet yhteneviä. Selventääksemme tätä ja verrataksemme BDNF-vajauksen aiheuttamia kehityksellisiä tilavuusmuutoksia eteneviin amyloidin kertymisen aiheuttamiin muutoksiin suoritimme in vivo ja ex vivo MRI-tutkimukset siirtogeenisille hiirille, joita saimme risteyttämällä APP/PS1-hiiriä BDNF^+/‒ -osapoistogeenisten hiirten kanssa. Havaitsimme, että BDNF^+/‒ -hiirillä kokoaivotilavuus oli pienentynyt 3 kk iässä ja aivokuoritilavuus sekä 3 että 8 kk iässä in vivo, kun taas APP/PS1-hiirten kokoaivo- ja aivokuoritilavuudet olivat suurentuneet 8 kk iässä. Yhdenmukaisesti edellämainittujen havaintojen kanssa aivojen histologinen analyysi osoitti aivokudoksen pienentyneeksi BDNF^+/‒ -hiirillä mutta

(10)

suurentuneeksi APP/PS1-hiirillä. Löysimme merkitsevän iän ja APP/PS1-genotyypin yhdysvaikutuksen ja havaitsimme aivokuoren suhteellisen tilavuuden kasvua iän myötä APP/PS1-hiirillä. Ymmärtääksemme näiden tilavuusmuutosten solutason mekanismeja teimme laajat histologiset tutkimukset, joissa analysoimme hermosolujen uudismuodostusta, aivojen amyloidikuormaa ja aktiivisten astrosyyttien määrän muutoksia.

BDNF-vajaus ei vaikuttanut aivojen amyloidi- tai astrogliamäärään, mutta johti hippokampuksen pykäläpoimun (gyrus dentatus) lyhenemiseen, kun taas APP/PS1-hiirillä nähtiin merkittävästi suurentuneet amyloidi- ja astrogliatilavuudet ja vähentynyt hermosolujen uudismuodostus. Näiden kahden genotyypin aiheuttamat muutokset hippokampuksessa olivat selkeästi erilaisia ja kerroskohtaisia. Yhteenvetona voidaan todeta, että toisin kuin AT-potilailla, amyloidia tuottavilla hiirillä aivoatrofia näyttää peittyvän amyloidikertymien ja erityisesti amyloidoosia seuraavan astroglioosin aiheuttaman aivojen tilavuuden kasvun alle. Tuloksiemme perusteella aivokuoren MRI-tilavuusmittausta voi tietyin rajoittein käyttää epäsuorana etenevän aivoamyloidoosin mittarina prekliinisissä tutkimuksissa.

AT:ssa on havaittu useita BDNF-signaloinnin häiriöitä. Aivojen BDNF-tasot laskevat, ja BDNF:n lyhyen, ei-signaloivan, TrkB.T1-reseptorin ja täysimittaisen TrkB.TK-reseptorin suhde kasvaa AT-potilailla ja APP/PS1-hiirillä. AT:ssa on havaittu myös dopaminergisen järjestelmän häiriöitä ja BDNF-signalointipuutosten aiheuttamia ongelmia DA- signaloinnissa. Havaitsimme aiemmin kiinnostavia muutoksia hiirilinjojemme spontaanissa liikeaktiivisuudessa: APP/PS1-hiiret ovat hyperaktiivisia, kun TrkB.T1-hiiret ovat hypoaktiivisia. Koska muutokset DA-signaloinnissa voivat vaikuttaa spontaaniin aktiivisuuteen, tarkastelimme APP/PS1- ja TrkB.T1-genotyyppien erillis- ja yhteisvaikutuksia nigrostriataalisen ja mesolimbisen DA-järjestelmän toimintaan ja neuropatologisiin muutoksiin. In vivo voltammetrialla havaitsimme välittömän DA- vapautumisen striatumissa (caudatus-putamen, CPu) olevan normaalia APP/PS1- tai TrkB.T1-siirtogeeniä kantavilla hiirillä, mutta stimulaation pitkittyessä näiden hiirten kyky vapauttaa lisää dopamiinia oli selkeästi heikentynyt. Kuitenkaan molempia siirtogeenejä ilmentävät hiiret eivät poikenneet villityyppisistä hiiristä. Kaikilla genotyypeillä tyrosiinihydroksylaasi (TH) –positiivisten aksonipäätteiden tiheys CPu:ssa ja nucleus accumbensissa oli normaali, ja DAT-, SNAP-25- sekä α-synukleiinitasot muuttumattomat.

Lisäksi näiden proteiinitasojen hyvät keskinäiset korrelaatiot osoittivat dopamiinin vapautumiseen ja takaisinottoon vaadittavan presynaptisen koneiston toimivaksi.

Havaitsimme kuitenkin TrkB.T1-hiirillä suurentuneen dopaminergisten hermosolujen lukumäärän mustatumakkeessa, kun APP/PS1-hiirten dopaminergisten säikeiden ja hermosolujen tiheys keskiaivoissa oli muuttumaton. Tästä seurasi aksonipäätettä kohti pienentynyt TH-taso TrkB.T1-hiirillä, kun vastaava määre oli muuttumaton APP/PS1-hiirillä.

Havaittu keskiaivojen DA-hermosolujen määrän kasvu TrkB.T1-hiirillä on uusi löydös.

Esitämme, että sekä APP/PS1- että TrkB.T1-genotyypit aiheuttavat DA-signaloinnin häiriöitä, mutta eri mekanismeilla, ja että tämä genotyyppiero voi myös selittää aiemmat löydökset hiirten spontaanissa liikeaktiivisuudessa.

Luokitus:WT 155, WL 104, WL 141.5.M2, WL 307, WL 314, WD 205.5.A6, QS 525, QT 21, QV 126, QV 504.5, QY 58, QY 60.R6,

Yleinen Suomalainen asiasanasto: aivot; aivokuori; Alzheimerin tauti; amyloidoosi; atrofia; autoradiografia;

dopamiini; elektrofysiologia; endokannabinoidit; hermosolut; hiiret; hippokampus; histologia;

immunohistokemia; kasvutekijät; keskushermosto; koe-eläinmallit; magneettitutkimus; muuntogeeniset eliöt;

neurobiologia; neuroplastisuus; reseptorit; tyvitumakkeet; välittäjäaineet

(11)

Muista että ne kaikki suurimmat kauhut on sun toiveittes peilikuvat, käännä ne

ja kädestäs löydät niihin avaimet – Toni Wirtanen

Jarmolle ja Viljalle

(12)

(13)

Acknowledgements

The present study was carried out in the A.I. Virtanen Institute, University of Eastern Finland, Kuopio.

I wish to greatly thank my principal supervisor, Professor Heikki Tanila, for taking me in his research group already a decade ago, and since then guiding me through not only my advanced special studies in medicine, but also the immense effort of planning and undertaking research projects, writing manuscripts and publishing articles that finally made up my doctoral thesis. The thesis has been “under construction” for many years alongside my clinical work, and I have been impressed by Heikki’s patience, understanding and positive attitude even when things did not look so bright to me. His encouragement, fresh ideas, knowledge and love of science have been the true driving force behind all of my projects. I would also like to thank Docent Jarmo T. Laitinen, my second supervisor, who guided my way through the first project included in this thesis and made an indelible impression on me with his knowledge of endocannabinoids among many other matters.

I am grateful to Professor Raimo Tuominen and Francisco López, Ph.D., the official reviewers of my thesis, for their helpful comments and guidance.

I am honoured to have Professor Amanda Kiliaan as my official opponent.

I owe a sincere thank-you to co-authors Professor Olli Gröhn, Johanna Närväinen, Ph.D.

and Leonid Yavich, Ph.D. for invaluable work and advice on the MRI and dopamine projects.

Similarly I want to thank Docent Riitta Miettinen for her guidance at the very beginning of my doctoral studies.

I would like to warmly thank other researchers in our research group, Irina Gureviciene, Ph.D., Kestutis Gurevicius, Ph.D., Susanna Kemppainen, M.Sc., Henri Leinonen, M.Sc. and Sofya Ziyatdinova, M.Sc. and also those that are working elsewhere as postdocs, Lakshman Puli, Ph.D. and Arto Lipponen, Ph.D. You all have supported me throughout the years in this group and provided plenty of food for thought in our group meetings, TYKY-days and conversations inside and outside of the laboratory. I want to say special thanks to Susanna for travelling with me and for being not only a great colleague, but also a good friend.

I am very grateful for the excellent technical assistance provided by Pasi Miettinen, Nanna Huuskonen, Hennariikka Koivisto, Esa Koivisto, Hanna-Maija Lahtinen, Niina Aaltonen, Juha-Pekka Niskanen, Maarit Pulkkinen, Sunna Lappalainen, Jonna Sirkkiä, Saara Staven, Saara Kainulainen, Elina Hämäläinen and any others who have lent a hand in the laboratory along the years. Special thanks go to Pasi, without whom I doubt this thesis would ever have seen daylight and who has been a true friend through good and not-so-good times.

I wish to thank the legendary “neuro coffee room” folk Pasi, Nanna, Anne-Mari, Jukka J, Mari, Anu, Lakku, Saaras K, S and P, Jonna, Laura, Hanna-Maija, Susanna, Sonja, Sunna, Elina, Joonas, Henna and Esa and all others that did not spend that many coffee breaks with us. You all are to thank for the many laughs and tears, parties and heated discussions, great friendships, the “women’s nights” and the great overall atmosphere. You guys made this all worthwhile and fun! Many thank-yous also go out to our long-time floorball team members such as Tsega, Timo, Jouni, Giedrius, Juhana and Jukka P for providing fun, sweaty moments on the field.

I want to thank my loving parents whose support and love mean the world to me. My mother Riitta, who has always been ever so positive and encouraging, and my father Sakari, whose example in writing a doctoral thesis I wanted to follow already from a young age, and who has always told me that I can do well at anything I put my mind into. Many thanks also go to my parents-in-law Irma and Matti and brother-in-law Jarkko for always being there for our family.

Most loving thanks go to my husband Jarmo for all his love and for always standing by my side and enduring all the long nights and weekends I have worked on my projects, and

(14)

to our sweet, beautiful daughter Vilja for being a constant source of joy and such a good sleeper (so that mommy can write her thesis).

This study was financially supported by the European Union 7th Framework Programme HEALTH-2007-201159, the Sigrid Juselius Foundation, the North-Savonia Regional Fund, the A.A. Laaksonen Fund via the Finnish Cultural Foundation, the Emil Aaltonen Foundation, the Finnish Medical Society Duodecim, the UEF-Brain University of Eastern Finland strategic funding, the Kuopio University Foundation and the Finnish Medical Foundation.

Kuopio, March 2016 Elisa Nevalainen

(15)

List of the original publications

This dissertation is based on the following original publications:

I Kärkkäinen E, Tanila H and Laitinen J T. Functional autoradiography shows unaltered cannabinoid CB1 Receptor signalling in hippocampus and cortex of APP/PS1 transgenic mice. CNS Neurol Disord Drug Targets. 2012 Dec;11(8):1038-44.

II Kärkkäinen E, Lahtinen H-M, Närväinen J, Gröhn O, Tanila H. Brain amyloidosis and BDNF deficiency have opposite effects on brain volumes in AβPP/PS1 mice both in vivo and ex vivo. J Alzheimers Dis. 2015 Jun 26;46(4):929-46.

III Kärkkäinen E, Yavich L, Miettinen P O, Tanila H. Opposing effects of APP/PS1 and TrkB.T1 genotypes on midbrain dopamine neurons and stimulated dopamine release in vivo. Brain Res. 2015 Oct 5;1622:452-65.

The publications were adapted with the permission of the copyright owners.

(16)

(17)

Abbreviations

2-AG 2-arachidonoylglycerol 3D three-dimensional

AA arachidonic acid

Aβ amyloid beta

ACh acetylcholine

AD Alzheimer´s disease

AEA N-arachidonoylethanolamide

a.k.a anandamide

AG 2-acylglycerol ANOVA analysis of variance ANOVA ANOVA for repeated -RM measurements ApoE apolipoprotein E APP amyloid beta precursor protein

ABHD alpha/beta-hydrolase domain

-containing protein

ALDH aldehyde dehydrogenase AwBw wt × wt genotype

AwB‒ wt × BDNF^+/− genotype A+Bw APP/PS1 × wt genotype A+B‒ APP/PS1 × BDNF^+/− genotype AwTw wt × wt genotype

AwT1 wt × TrkB.T1 genotype A+Tw APP/PS1 × wt genotype A+T1 APP/PS1 × TrkB.T1 genotype BDNF brain-derived neurotrophic factor

CA1 cornu ammonis area 1

CAMK Ca2+/calmodulin-dependent

protein kinase

cAMP cyclic adenosine monophosphate CB1R cannabinoid receptor CB1 CB2R cannabinoid receptor CB2

cDNA complementary DNA

CNS central nervous system

COMT catechol-O-methyltransferase CPu caudate-putamen

CSF cerebrospinal fluid

CTX cerebral cortex

DA dopamine

DAB 3,3'-Diaminobenzidine DAG diacylglycerol

DAGL diacylglycerol lipase DAT dopamine transporter DCX doublecortin

DG dentate gyrus

DOPAC 3,4-Dihydroxyphenylacetic acid

DTT dithiotreitol eCB endocannabinoid

EPS extrapyramidal symptom

ER endoplasmic reticulum

ERK extracellular signal regulated kinase

FAAH fatty acid amide hydrolase FAD familial Alzheimer´s disease

(20)

GAB1 Grb2-associated binder-1 GABA gamma-aminobutyric acid GDP guanosine diphosphate GFAP glial fibrillary acidic protein Gi inhibitory G protein

G protein guanine nucleotide –binding protein

GPCR G-protein coupled receptor Grb2 growth factor receptor-bound

protein 2

GTPγS guanosine-5’-O-(3-thio)- triphosphate [³⁵S]GTPγS guanosine-5’-O-(3-[³⁵S]thio)- triphosphate HC hippocampus

HPLC high-performance liquid chromatography HVA homovanillic acid IP3 inositol triphosphate LBD Lewy body dementia L-DOPA L-3,4-dihydroxy-

phenylalanine LNGFR low-affinity nerve growth

factor receptor a.k.a p75NTR LPA1 lysophosphatidic acid

receptor 1

LTD long-term depression

LTP long-term potentiation

LV lateral ventricle

MAGL monoacylglycerol lipase

MAO monoamine oxidase

MAPK mitogen-activated protein kinase

MCI mild cognitive impairment

MEK MAP/ERK kinase

MFB medial forebrain bundle MKP1 MAP kinase phosphatase 1 MRI magnetic resonance imaging mRNA messenger ribonucleic acid

NAc nucleus accumbens

NAE N-arachidonoylethanolamine NAPE N-arachidonoyl-

phosphatidylethanolamine NAPE-PLD NAPE-specific phospholipase

D

NEM N-ethylmaleimide NFT neurofibrillary tangle NMR nuclear magnetic resonance NSF N-ethylmaleimide sensitive

fusion protein

OD optic density

P2Y12 a purinergic G protein coupled receptor

p75NTR p75 neurotrophin receptor

a.k.a LNGFR

PD Parkinson´s disease

PDK1 3-phosphoinositide-

dependent protein kinase 1

PET positron emission

tomography

PIP² phosphoinositide bisphosphate

(21)

PKA protein kinase A PKC protein kinase C

PLC phospholipase C

PrP prion protein

PS1 and 2 presenilin 1 and 2 PTEN phosphatase and tensin homolog

R pearson correlation coefficient Raf raf-kinase

Ras ras-GTPase protein

ROI region of interest

SEM standard error of the mean Shc Src homology 2 domain

containing adaptor protein

SN substantia nigra

SNAP25 synaptosome-associated protein of 25 kDa

SNARE soluble NSF attachment

protein receptor

SNpc substantia nigra pars compacta Sos son of sevenless α-syn alpha-synuclein T1 spin-lattice relaxation time T2 spin-spin relaxation time tg transgenic TH tyrosine hydroxylase

TrkB tropomyosin-related kinase B TrkB.T1/T2 truncated isoforms of the

TrkB receptor

TrkB.TK functional, full-length isoform of the TrkB receptor

TRPV1 transient receptor potential cation channel subfamily V member 1 a.k.a vanilloid receptor 1

VGCC voltage-gated calcium channel

VTA ventral tegmental area W02 anti-amyloid beta antibody

WB whole brain

wt wild type

(22)

(23)

1 Preface

This thesis discusses three projects that were finished during the years working part-time in the Neurobiology of Memory laboratory alongside my clinical work; one exploring cannabinoid CB1 receptor signalling in APP/PS1 transgenic mice by functional autoradiography, one looking at the effects of APP/PS1 and BDNF^+/‒ transgenes on brain volumes and at the underlying neuropathological changes by means of MR imaging and histochemical analyses, respectively, and one investigating dopamine signalling in APP/PS1 and/or TrkB.T1 transgenic mice by using in vivo voltammetry and histochemical analyses. Why did we, then, choose these topics and especially some of the more unconventional methods? In general it may be said that it all boils down to special expertise found in the Faculty of Health Sciences in the University of Eastern Finland, but let us specify this regarding each project.

There is firm knowledge on the metabolism of bioactive lipids and their role in health and disease in the group led by Docent Jarmo T Laitinen at the Institute of Biomedicine. Their know- how on using functional autoradiography as an approach to study G-protein-coupled lipid receptor signalling offered a very interesting means to explore changes occurring in the brains of the APP/PS1 transgenic mouse model of amyloidosis that is studied in our group focusing on the neurobiology of memory. We know that one of the characteristic neuropathological changes in Alzheimer´s disease (AD) is neuronal loss in vulnerable cell populations, which, in turn, results in disruption of brain signalling systems such as the cholinergic and monoaminergic systems. Far less is known about the possible changes of the endocannabinoid (eCB) system in brains of AD patients and animal models, although the eCBs and their receptors are highly widespread in the mammalian brain and the eCB system is known to participate in the regulation of multiple important tasks. To this end, we employed [³⁵S]GTPγS autoradiography in APP/PS1 mice to assess functional CB1-receptor activity in the brain regions most severely affected by neuropathological processes in AD.

The Biomedical NMR –group led by Professor Olli Gröhn at A. I. Virtanen Insitute has the latest knowledge on novel magnetic resonance imaging (MRI) techniques that aim to visualize surrogate markers for processes associated with neurodegenerative diseases. With this expertise at hand, and knowing that there is a growing interest to find a noninvasive means to follow the progression of neuropathological changes in AD animal models, we used in vivo and ex vivo MRI volumetry to determine the possible brain volume changes occurring with the progression of amyloid pathology in our APP/PS1 transgenic mice. However, since the levels of brain-derived neurotrophic factor (BDNF) have been shown to be decreased in AD, we wanted to find out the contribution of BDNF deficiency to possible volumetric changes, and crossed APP/PS1 mice with BDNF heterozygote knock-out (BDNF^+/‒) mice. We then performed post-mortem histological analysis of the underlying neuropathological changes to better explain the obtained volumetric data.

As mentioned above, there is disruption of the brain monoaminergic system in AD, including the dopaminergic signalling system. This makes dopamine (DA) signalling a very appealing target for research also in our APP/PS1 transgenic mice. The laboratory of Dr. Leonid Yavich at the School of Pharmacy is one of the few in Europe that masters the field of in vivo voltammetry.

His group studies the pharmacological, neurochemical and neurophysiological aspects of the monoaminergic, mainly DAergic, regulation of behavior in health and disease, using intricate equipment to record the release and reuptake of neurotransmitters in vivo. Due to this special expertise found at our university, we used in vivo voltammetry to assess DA release dynamics upon prolonged stimulation in APP/PS1 mice. Due to BDNF-signalling changes found in AD, we wanted to see the contribution of disrupted BDNF-signalling to DA metabolism, and crossed APP/PS1 mice with TrkB.T1 mice that overexpress the truncated, defunct TrkB receptor for

(24)

BDNF. Finally, we used a morphological approach in addition to the functional one, and used histochemical methods to take a look at not only the number of midbrain DAergic neurons and their striatal terminals, but also the density of key proteins for DA release and reuptake presynaptically.

(25)

2 Review of the literature

2.1 NEUROPATHOLOGICAL PROCESSES IN ALZHEIMER´S DISEASE

Alzheimer´s disease (AD) is the most prevalent disorder of the brain causing dementia. Clinically the diagnosis of AD is based on detecting the progressive memory impairment whereby recalling recent events deteriorates early on in the disease process (Soininen and Scheltens, 1998). The key neuropathological features of AD include extracellular amyloid-β (Aβ) plaques and intraneuronal neurofibrillary tangles (NFTs) formed from hyperphosphorylated, microtubule- associated tau protein (Selkoe, 2001). Additionally, characteristic changes include cerebrovascular amyloid accumulation (Selkoe, 1998) and loss of synapses and neurons in vulnerable cell populations (for review, see Price and Sisodia, 1998 and references therein).

AD is a multifactorial neurodegenerative disease, with ~ 99 % of cases being sporadic, and a small portion of cases being caused by specific autosomal dominant mutations in the transmembrane proteins amyloid-β precursor protein (APP), presenilin-1 (PS1) and presenilin-2 (PS2) (Campion et al., 1999; Price and Sisodia, 1998). The exact causes behind sporadic AD are not yet established, but known risk factors include age and the presence of ApoE ε4 allele (Corder et al., 1993; Hyman et al., 1996; Roses, 1996; Strittmatter et al., 1993). Also many susceptibility genes and environmental and lifestyle related risk factors have been suggested to influence AD onset and progression (Ríos et al., 2014).

2.1.1 Loss of cholinergic and monoaminergic neurons in AD

As mentioned above, a key neuropathological feature of AD is the loss of synapses and neurons in some vulnerable cell populations in the brain. This neuronal degeneration leads to disruption of the brain´s signalling systems, of which the cholinergic system is very widely documented (Bartus et al., 1982; Harkany et al., 2000, 1998, 1995; Lyness et al., 2003; Parvizi et al., 2001).

Degenerative changes have also been revealed in Aβ producing transgenic mice: the very same APP/PS1 transgenic mice as used in our studies showed impairment of muscarinic neurotransmission already before manifestation of cognitive deficits (Machová et al., 2008), and decreased density and size of cholinergic synapses were found in another APP/PS1 mouse model (Bell et al., 2006; Wong et al., 1999). However, the loss of cholinergic neuronal bodies themselves in APP/PS1 mice has not been reported.

In fact, the oldest hypothesis attempting to explain AD aetiology is the cholinergic hypothesis of memory impairment, based on findings of substantial deficits in acetylcholine (ACh) synthesis machinery, loss of neurons in the cholinergic nucleus basalis in the AD brain and the relation of cholinergic signalling to learning and memory (Davies and Maloney, 1976; Perry et al., 1977; for reviews, see Bartus et al., 1982; Francis et al., 1999; Lyness et al., 2003 and references therein).

Originally this hypothesis drove AD drug development, and with the exception of memantine, the currently commercially available medications for AD are cholinesterase inhibitors that aim at increasing ACh concentration in synapses (Schneider et al., 2014).

The loss of synapses and neurons in the monoaminergic system is also extensively documented (Lyness et al., 2003; Palmer and DeKosky, 1993; Parvizi et al., 2001; Rüb et al., 2000; Zweig et al., 1988), and these degenerative changes have been shown to be recapitulated by the same transgenic APP/PS1 mouse model of AD as used in our studies (Y. Liu et al., 2008; Perez et al., 2005).

(26)

2.1.2 Aberrant amyloid processing in AD

Research has shown that the accumulation of abnormally folded Aβ in the brain is one of the cardinal metabolic disturbances in AD. All known mutations behind familial AD have been thought to affect either Aβ production or aggregation (Price and Sisodia, 1998), but more recently it has been found that many FAD presenilin mutations do not result in increased Aβ42 or Aβ42/40 ratio (Shioi et al., 2007; Walker et al., 2005) and FAD mutation induced increases in Aβ42 may not always correlate with age of disease onset (Citron et al., 1997; Mehta et al., 1998), suggesting that the mechanism behind neurodegeneration is independent of the effects of these mutations on Aβ production. However, Aβ accumulation contributes to the development of other neuropathological features of AD, such as tau hyperphosphorylation (Götz et al., 2001) and loss of cholinergic neurons (Harkany et al., 2000). It has also been found that immunization with Aβ antibodies reduces Aβ deposition in the brain (Das et al., 2001; Schenk et al., 1999; Wilcock et al., 2001) and can prevent (Janus et al., 2000; Morgan et al., 2000) or even reverse (Dodart et al., 2002) memory deficits in an AD mouse model. Anti-Aβ immunization in humans, however, has produced some unwanted off-target immune effects and research is underway to solve these problems (for review, see Spencer and Masliah, 2014 and references therein).

Recent studies on PET imaging of brain amyloid in vivo have changed our view of the time course of AD. The present view is that amyloid accumulation takes place decades before the onset of memory problems, since amyloid plaques are commonly found in cognitively normal elderly controls (Aizenstein et al., 2008) and amyloid load in PET imaging does not further increase in diagnosed AD patient during follow-up (Kadir et al., 2012).

It seems, in fact, that impaired memory in AD patients is not directly connected to the brain Aβ plaque load, but instead the amount of NFTs (Guillozet et al., 2003), soluble Aβ species (Mucke et al., 2000; Näslund et al., 2000; Walsh et al., 2002; Walsh and Selkoe, 2004) and synaptic loss (Selkoe, 2002). In most transgenic mouse models, problems in synaptic transmission actually seem to precede the formation of Aβ plaques and NFTs (Arendt, 2009; Blanchard et al., 2010), and these problems may be caused by increased amounts of soluble Aβ oligomers (Haass and Selkoe, 2007; Rowan et al., 2007). However, whether same is true for the human disease await for techniques to directly address synaptic function in living human brain.

2.2 THE BRAIN ENDOCANNABINOID SYSTEM

The section above addressed the disruption occurring in the cholinergic and monoaminergic signalling systems in the AD brain. Far less is known about the possible changes that may occur in lipid signalling systems of AD brains. A prominent endogenous lipid signalling system in the CNS is the endocannabinoid (eCB) system, which consists of two G-protein coupled receptors (GPCRs) CB1R and CB2R, their endogenous activating ligands (eCBs), and metabolic enzymes involved in the biosynthesis and degradation of these ligands. The eCBs are lipid mediators that bind to CB1R and CB2R and consequently regulate the activity of ion channels and release of neurotransmitters and thus synaptic functions (Piomelli, 2003). The most common eCBs are 2- arachidonoylglycerol (2-AG) (Mechoulam et al., 1995) and N-arachidonoylethanolamide (anandamide, AEA) (Devane et al., 1992). 2-AG is more potent of these two and acts as a full agonist at both receptors, whereas AEA acts as a partial agonist (Sugiura et al., 2006). Both eCBs are known to participate in the regulation of neurotransmission, appetite, and analgesia as well as in the control of the immune system and neuroprotection during injury (Eljaschewitsch et al., 2006; Kano et al., 2009; Piomelli, 2003; Sugiura et al., 2006). There is evidence that via CB1R at least 2-AG also participates in more complex processes like regulation of synaptic plasticity, memory and emotions (Carlson et al., 2002; Marsicano et al., 2002; Martin et al., 2002; Piomelli, 2003). In addition, a recent study suggests that CB1R activity on hippocampal GABAergic neurons protects

(27)

against age dependent cognitive decline by reducing pyramidal cell degeneration and neuroinflammation (Albayram et al., 2011).

The eCBs are hydrophobic lipids that are biosynthesized and released on demand (Marsicano et al., 2003) from post-synaptic neurons to activate presynaptically situated CB1Rs. This so-called retrograde signalling results in both transient and long-lasting reductions in the release of other neurotransmitters, thus fine-tuning synaptic efficacy and neural activity (Alger, 2002; Savinainen et al., 2012). AEA is produced by a stimulus-dependent cleavage of N-arachidonoyl- phosphatidylethanolamine (NAPE) (Di Marzo et al., 1994) via alternative pathways (J. Liu et al., 2008). 2-AG is produced from the hydrolysis of diacylglycerol (DAG) by two diacylglycerol lipases (DAGLs), DAGLα or DAGLβ (Bisogno et al., 2003), the first of which is responsible for nearly all of 2-AG acting as an eCB in the adult brain (Tanimura et al., 2010).

Also the machinery responsible for terminating eCB signalling is tightly regulated, and both 2-AG and AEA are very labile, indicating sensitivity to enzymatic activity. eCBs are taken up by cells through a still controversial mechanism (Fowler, 2013) and then inactivated by enzymatic hydrolysis. The hydrolysis of eCBs is mainly done by four enzymes belonging to the serine hydrolase family. AEA is principally hydrolyzed by the membrane enzyme fatty acid amide hydrolase (FAAH) (Ahn et al., 2009; Cravatt et al., 1996). Many enzymes have been implicated in 2-AG hydrolysis, but the serine hydrolase monoacylglycerol lipase (MAGL) (Dinh et al., 2002;

Saario et al., 2004) is responsible for ~85 % of brain 2-AG hydrolase activity, while the enzymes ABHD6, ABHD12 and FAAH hydrolyze the remaining 15 % (Blankman et al., 2007). The main biosynthetic and deactivating enzymes in eCB signalling are presented in Figure 1 (adapted from Di Marzo et al., 2014).

(28)

Figure 1.Main biosynthetic and deactivating enzymes in endocannabinoid signalling, including role of eCBs in retrograde (mainly for 2-AG), anterograde and intracellular (for AEA) signalling (adapted from Di Marzo et al., 2014). 2-AG is produced from hydrolysis of DAG mainly by DAGLα and hydrolyzed mainly by MAGL. AEA is produced by cleavage of NAPE via alternative pathways, one of which is catalyzed by NAPE-PLD, and hydrolyzed mainly by FAAH. Solid arrows denote transformation into active metabolites or activation, dashed arrows denote transformation into metabolites inactive at cannabinoid receptors, the blunt arrow denotes inhibition.

AA = arachidonic acid, AEA = anandamide, 2-AG = 2-arachidonoylglycerol, AGs = 2-acylglycerols, CB1R = CB1 receptor, DAGs

= diacylglycerols, DAGLα = diacylglycerol lipase-α, ER = endoplasmic reticulum, FAAH = fatty acid amide hydrolase 1, MAGL = monoacylglycerol lipase, MAPK = mitogen-activated protein kinases, NAEs = N-arachidonoylethanolamines, NAPE = N- arachidonoyl-phosphatidyl-ethanolamine, NAPE-PLD = NAPE-specific phospholipase D, PIP2 = phosphoinositide bisphosphate, PKA = protein kinase A, PLCβ = phospholipase Cβ, VGCCs = voltage-gated calcium channels

CB1R is among the most abundant GPCRs in the mammalian brain (~1 pmol/mg protein) (Herkenham et al., 1990), and is most abundant in regions such as the hippocampus, cortex, basal ganglia and the cerebellum (Herkenham et al., 1990; Tsou et al., 1998). CB1R is expressed in both excitatory and inhibitory presynaptic terminals, although also some postsynaptic distribution has been reported (Mackie, 2005; Onaivi et al., 2012). Earlier it was believed that CB1Rs were

(29)

predominantly situated in the brain, and CB2Rs restricted to immune tissues and cells, but later CB1Rs have been found in many peripheral tissues and CB2Rs in the CNS (for reviews, see Campillo and Paez, 2009, and Onaivi et al., 2012, and references therein).

2.2.1 Changes in the brain endocannabinoid system in AD

As mentioned previously, there is abundant evidence that cannabinoids participate in neuroprotection (for reviews, see Bisogno and Di Marzo, 2008; Kano et al., 2009, and Micale et al., 2007 and references therein). In AD eCBs seem to have a dual role depending on the disease stage.

The initial neuroprotective effects of eCBs may derive from CB1Rs activation, whereas CB2R activation at the initial disease phase can be harmful due to blockade of the early repair process mediated by microglial cells. Later on in the disease, activation of CB2Rs might diminish the excessive inflammatory response (Bisogno and Di Marzo, 2008; Di Marzo, 2008). In line with this, the memory deficits observed in mice after brain Aβ peptide administration have been found to be alleviated by late (but not early) administration of CB1R antagonist after Aβ administration (Mazzola et al., 2003) or early (but not late) administration of eCB reuptake inhibitor after Aβ administration (van der Stelt et al., 2006). In any case, many studies have shown that cannabinoids (Chen et al., 2011; Ramírez et al., 2005; van der Stelt et al., 2006), CB1R and/or CB2R agonists (Ehrhart et al., 2005; Esposito et al., 2007, 2006; Milton, 2002; Ramírez et al., 2005) and the early (but not late) administration of an eCB reuptake inhibitor (van der Stelt et al., 2006) can protect from Aβ-induced toxic and inflammatory effects both in vitro and in vivo. A more recent study found that cortical CB1R immunoreactivity correlated to the patients´ cognitive abilities prior to death (Lee et al., 2010), suggesting that signalling via CB1R plays a role in preserving cognitive skills in AD. This is consistent with another recent study on ageing which implicated a protective role of CB1R activity on hippocampal GABAergic neurons against age-dependent cognitive decline by reducing pyramidal cell degeneration and neuroinflammation (Albayram et al., 2011).

Studies have been performed in human AD patients and AD model mice to find out whether pathophysiological changes in AD cause alterations in eCB signalling via CB1R. The first study in AD brains employing autoradiography (Westlake et al., 1994) reported decreased in vitro cannabinoid binding in the hippocampus (but not in the neocortex) of AD brains compared to age-matched controls, with no change in the CB1R mRNA levels. However, CB1R expression was found to be unchanged in another study performed on post-mortem AD brain entorhinal cortex and hippocampal immunohistological sections (Benito et al., 2003). Two more recent studies also found no difference between AD and control brains in CB1R immunoreactivity, using Western blotting and CB1R binding with the selective antagonist [³H]-SR141716A (Lee et al., 2010) or immunohistochemistry and Western blotting (Mulder et al., 2011). Yet another study reported decreased CB1R density by immunolabeling but unchanged CB1R density and binding affinity in the AD frontal cortex using the CB1R selective agonist [³H]-WIN55,212-2, and significantly decreased [³H]-WIN55,212-2-stimulated [³⁵S]GTPγS-binding, which measures CB1R activation (Ramírez et al., 2005). Finally, a recent study reported reduced CB1R expression in post-mortem AD brains by Western blotting (Solas et al., 2013). In conclusion, although the results on CB1R density in AD are discrepant, it appears that at least the G-protein coupling of CB1Rs is impaired in AD. On the other hand, CB2R levels seem to increase in AD (Benito et al., 2003; Halleskog et al., 2011; Ramírez et al., 2005) and rats treated with Aβ1-42 (van der Stelt et al., 2006).

2.2.2 Changes in the brain endocannabinoid system in mouse models of AD

So far, only a few studies have investigated CB1R densities in transgenic AD model mice, all using the same common APPswe/PS1dE9 mouse model as our group, with similarly ambiguous results as in humans. The first immunohistochemical study found lower hippocampal density of CB1R-positive neurons in 10- to 12-month-old transgenic mice than in their wild-type littermates (Kalifa et al., 2011). This was associated with astrogliosis and increased pro-inflammatory cytokine levels, possibly indicating an anti-inflammatory role of CB1Rs in the brain. Another

(30)

study using Western blotting found no difference between APP/PS1 transgenic and wild-type mice at 7 months of age, but elevated CB1R protein levels in APP/PS1 mice at the age of 14 months (Mulder et al., 2011; Supplementary Figure 9E). A more recent study assessed the cannabinoid agonist HU210-stimulated [³⁵S]GTPγS-binding in APP/PS1 mice and found increased cannabinoid receptor coupling compared to wild-type mice in the frontal cortex and striatum, but not in the hippocampus (Maroof et al., 2014). These findings were associated with reduced concentrations of striatal AEA at 6 months of age and 2-AG at 8 months, but unchanged concentrations of these eCBs at 4 months and in the other brain areas, as measured by mass spectroscopy. The group speculated that the enhanced striatal cannabinoid receptor activation may represent an attempt to compensate for the reduced eCB levels.

2.2.3 Endocannabinoid study – rationale and overview

Studies on this topic in amyloid plaque producing transgenic AD model mice are thus so far scarce and the results ambiguous. The purpose of the eCB study was to add to the knowledge on possible changes in the eCB system in APP/PS1 mice. At the time of the study, earlier studies had only assessed CB1R quantities in the brain, but we decided to take a look at the functional CB1R activity using [³⁵S]GTPγS-autoradiography (Laitinen, 2004), which shows the CB1R-dependent Gi protein activation, and at the time this method offered a new point of view to the topic. The functional CB1R activity was assessed in APP/PS1 mice in the brain areas most severely affected by neuropathological changes, at the age of 13-14 months when amyloid plaque pathology and neuroinflammation around the plaques are fully developed (Garcia-Alloza et al., 2006;

Jankowsky et al., 2004), and the mice show robust spatial memory impairment (Minkeviciene et al., 2008). We also used the same CB1R ligand (CP55940) as the first CB1R binding study in post- mortem AD brains (Westlake et al., 1994). Our previous studies have indicated that thiol-reactive agents can have profound effects on Gi-mediated receptor signalling, as revealed by functional autoradiography (Kokkola et al., 2005; Laitinen, 2004; Lewandowicz et al., 2006). This is why we decided to replicate the study protocol in the presence of dithiotreitol (DTT; a strong reducing agent) to find out more about the possible redox-dependent changes in the functional CB1R activity.

2.3 THE BRAIN DOPAMINERGIC SYSTEM

The earlier section about AD addressed neuronal degeneration occurring in some vulnerable cell populations in the AD brain and mentioned that the disruption of the monoaminergic signalling systems is well documented. Monoaminergic signalling systems in the brain use monoamine neurotransmitters, i.e. dopamine, serotonin, noradrenaline and adrenaline, the first of which is of special interest to us.

Dopamine (DA) biosynthesis starts with L-phenylalanine, which is converted to L-tyrosine by phenylalanine hydroxylase. L-tyrosine is converted into L-DOPA by tyrosine hydroxylase (TH), which is the rate-limiting enzyme in the DA biosynthesis pathway. L-DOPA is finally converted into DA by DOPA decarboxylase, and DA itself is a precursor for the synthesis of noradrenaline and adrenaline. (Daubner et al., 2011; Elsworth and Roth, 1997; Sourkes, 1963) After synthesis, DA is transported from the cytosol into synaptic storage vesicles by the vesicular monoamine transporter 2 (Lohr and Miller, 2014). DA is released from vesicles by exocytosis into the synaptic cleft upon the arrival of an action potential, the extent of DA release depending on the rate and pattern of neuronal firing (Schultz, 2007). DA reuptake from the synaptic cleft occurs via the high- affinity dopamine transporter (DAT) or other low-affinity transporters (for review, see Daws, 2009 and references therein), and DA degradation takes place by the enzymes monoamine oxidase (MAO), aldehyde dehydrogenase (ALDH) and catechol-O-methyltransferase (COMT) acting in sequence. (Elsworth and Roth, 1997)

(31)

In the synapse DA binds to and activates DA receptors, which are membrane receptors belonging to a family of seven transmembrane domain G-protein-coupled receptors. DA receptors can be located either on postsynaptic cells or on the presynaptic DA-releasing cell itself (D2 receptors) – in the latter case they are called autoreceptors, which decrease neuronal excitability when located somatodendritically but cause decreases in DA synthesis, packaging and release when located on the nerve terminals. DA autoreceptors seem to regulate dopaminergic (DAergic) neurotransmission in this manner. Five subtypes of DA receptor have been described, and they can be subdivided into D1-like receptors, which stimulate cAMP activity and include D1 and D5 receptors, and D2-like receptors, which inhibit cAMP activity and include D2, D3 and D4 receptors. D1 and D2 receptors are most abundant, and the D2 autoreceptor has two isoforms called D2^Large and D2^Small. (Baik, 2013; Elsworth and Roth, 1997) These receptors have distinct localizations in the brain – D1 and D2 receptors are abundant in the striatum whereas D4 and D5 receptors are mostly located out of the striatum, and D3 receptors are abundant in limbic areas (Leggio et al., 2013).

Dopamine and DAergic cell groups in the brain were first identified and described in the early 1960s among other catecholaminergic cell groups and later this description has been supplemented by others. These cell groups were designated “A” cells (A for aminergic), the DAergic ones being subdivided into groups A8 through A17 (See Table 1 and Fig. 2b) (Albanese et al., 1986; Björklund and Dunnett, 2007). Four major DAergic pathways in the brain have been described. The neurons from substantia nigra (SN; A8-9) mainly project to the basal ganglia, forming the nigrostriatal pathway (Albanese et al., 1986), which is part of the extrapyramidal system and primarily involved in the control of movements (Gerfen, 1992; Leggio et al., 2013) but also in goal-directed behaviours (Baik, 2013). The nigrostriatal pathway has also been implicated in hippocampus-independent spatial-relational learning and memory (Da Cunha et al., 2003). The neurons from ventral tegmental area (VTA; A10) project via the mesolimbic pathway to limbic structures such as nucleus accumbens (NAc), hippocampus (HC), amygdala and the medial prefrontal cortex (Albanese et al., 1986) – this pathway plays an important role in emotional behaviours and the reward system (Leggio et al., 2013). VTA neurons also project via the mesocortical pathway to the cerebral cortex, particularly the frontal cortex, and is thought to play an important role in memory, motivation and emotional responses (Scarr et al., 2013). Another group of cells make up the tuberoinfundibular pathway, originating in the hypothalamic arcuate (A12) and periventricular (A14) nuclei, projecting to the median eminence, the infundibular and the pituitary and inhibiting prolactin secretion (Albanese et al., 1986). These major DAergic pathways are illustrated in Fig. 2a (adapted from Scarr et al., 2013).

(32)

Table 1. Dopaminergic cell groups in the brain (adapted from Albanese et al., 1986 and Sidhu et al., 2003)

Cell group Location

A8 Retrorubral area (caudal part of substantia nigra in the lateral midbrain tegmentum)

A9 Pars compacta of substantia nigra

A10 Ventral tegmental area (ventral midbrain tegmentum) A11 Caudal thalamus, dorsal and posterior hypothalamus (close to

the 3^rd ventricle in the diencephalon)

A12 Arcuate hypothalamic nucleus

A13 Zona incerta and adjacent subthalamic area A14 Rostral hypothalamus, periventricular nucleus (close to the 3^rd

ventricle in the diencephalon)

A15 Periventricular nucleus and adjacent anterior hypothalamus (not present in all species, continuous caudally with A14)

A16 Olfactory bulb

A17 Retina

(33)

Figure 2.Major dopaminergic pathways and cell groups (a) in humans (adapted from Scarr et al., 2013), (b) in rodents (adapted from Björklund and Dunnett, 2007).

The division of the brain DAergic system into these four major DAergic pathways, however, is an oversimplification. SN neurons project not only to the basal ganglia, but also to cortical and limbic areas. On the other hand, VTA neurons not only project to cortical and limbic areas, but also to basal ganglia (Albanese et al., 1986; Björklund and Dunnett, 2007). There are differences between primates and rodents in their DAergic systems – in primates and humans the DAergic system is more complex and includes at least 10-20 times more TH+ cells than in mice or rats. The expansion is most evident in SN, and the innervation territory is also considerably expanded, especially in the neocortex (see Björklund and Dunnett, 2007 and references therein).

Since our interest lies in the disruption of multiple brain signalling systems, primarily the eCB and DAergic systems, it is interesting that cannabinoids seem to play a modulatory role in DAergic neuronal pathways, modulating VTA DAergic neuron firing (and likely NAc DA release as a consequence) (Cheer et al., 2003; Melis et al., 2004a, 2004b; Riegel and Lupica, 2004; Szabo et al., 2002, for reviews, see Gardner, 2005 and Wang and Lupica, 2014 and references therein) and striatal DA release (Cadogan et al., 1997). As it comes to CB1R levels in central DAergic brain

(34)

areas, average to high CB1R concentrations are found in the striatum and the SN, but lower concentrations in areas of VTA and NAc (Herkenham et al., 1990; Tsou et al., 1998).

2.3.1 Changes in the brain dopaminergic system in AD

In the following text on DAergic system changes in AD I will leave studies performed before 1996 out of inspection, since a consensus on diagnostic criteria for frontotemporal dementias and dementia with Lewy bodies (LBD) was reached as late as in 1994 and 1996, respectively (Pasquier and Delacourte, 1998), and prior to that these dementia subtypes were often grouped under the common terms of “AD” or “AD-type dementia”. LBD was only started to be diagnosed in the mid-1990s after the discovery of α-synuclein in Lewy bodies (Spillantini et al., 1997). LBD patients exhibit clear losses of DAergic neurons and thus the inclusion of LBD patients in studies aimed at looking for DAergic changes in the AD brain may result in skewed outcomes.

In AD the DAergic system has been shown to suffer to varying extent from typical AD neuropathological changes such as NFTs and Aβ plaques (Burns et al., 2005; Parvizi et al., 2001;

Reyes et al., 1996), and neuron loss in the SN (Burns et al., 2005; Lyness et al., 2003; Reyes et al., 1996). Over a third of AD patients present extrapyramidal symptoms (EPSs) like bradykinesia and rigidity (Lopez et al., 1997; Soininen et al., 1992), and brain DA (Storga et al., 1996) as well as striatal D1 and D2 (Kemppainen et al., 2000; Pizzolato et al., 1996) and frontal cortex D1, D2, D3 and D4 (Kumar and Patel, 2007) receptor levels seem to be decreased in AD. In fact, reduced hippocampal D2 receptor availability has been found to correlate with memory test performance in AD patients (Kemppainen et al., 2003). All this suggests the involvement of the DAergic system disruption in AD pathophysiology, but this subject is still under debate, especially concerning nigral vs. extranigral changes (for recent reviews, see Martorana and Koch, 2014 and Trillo et al., 2013 and references therein). However, recent studies support the idea of DAergic disruption in AD. One study showed that patients with pure AD (no Lewy bodies at autopsy) exhibiting EPSs had neuronal loss both in the SN and the putamen (Horvath et al., 2014), while others found that administration of DA agonists to AD patients has positive effects on cortical neurotransmission and synaptic plasticity mechanisms (Koch et al., 2014; Martorana et al., 2013).

2.3.2 Changes in the brain dopaminergic system in mouse models of AD

Experimental data examining possible DAergic system changes in APP/PS1 mice is still quite scarce. Two studies have reported unchanged levels of DA (Manaye et al., 2007; Szapacs et al., 2004) and its metabolites DOPAC and HVA (Szapacs et al., 2004) as assessed by HPLC, while a third study found an age-dependent decrease in striatal DOPAC levels in APP/PS1 mice, although DA and HVA levels were unchanged (Perez et al., 2005). Interestingly, in the same study DOPAC levels were higher in young APP/PS1 mice than in wild-type mice. Perez et al. (Perez et al., 2005) also reported a close association between DAergic pathology and amyloid deposition in these transgenic mice by revealing dystrophic DAergic neurites adjacent to Aβ plaques, and showed an age-dependent increase in TH protein levels in the SN but not in the striatum. No decrease in SN DAergic neurons has been found in APP/PS1 mice (Y. Liu et al., 2008; O’Neil et al., 2007; Perez et al., 2005). Liu et al. (Y. Liu et al., 2008) found DAergic neuron loss and atrophy in the VTA, while another group reported unchanged VTA neuron numbers (O’Neil et al., 2007).

The attenuation of memory impairments by the restoration of DA transmission has been found in transgenic mice carrying APP (TgCRND8; Ambrée et al., 2009) or APP, PS1 and tau (3xTg-AD;

Guzmán-Ramos et al., 2012; Himeno et al., 2011) mutations. The study on 3xTg-AD mice also showed that DA receptor agonist treatment caused significant decreases in the levels of intraneuronal Aβ, hyperphosphorylated tau and indicators of oxidative stress, suggesting a protective role of DA on neurons (Himeno et al., 2011). Additionally, another recent study reported SN pars compacta (SNpc) DAergic neuron losses in 3xTg-AD mice (Sun et al., 2012).

Endocannabinoid receptors, dopamine system and brain volumetry in amyloid plaque producing APP/PS1 mice

uef.fi

Dissertations in Health Sciences

ELISA NEVALAINEN

enDocannabinoiD receptors, Dopamine system anD brain VoLumetry in amyLoiD pLaque

proDucing app/ps1 mice

ELISA NEVALAINEN

Endocannabinoid receptors, dopamine system and brain volumetry in amyloid

plaque producing APP/PS1 mice

Acknowledgements

List of the original publications

Contents

Abbreviations

1 Preface

2 Review of the literature