Neurogliavascular remodeling and therapeutic intervention in ischemia and inflammation

DISSERTATIONS | MIKKO HUUSKONEN | NEUROGLIAVASCULAR REMODELING AND THERAPEUTIC... | No 402

uef.fi

PUBLICATIONS OF

THE UNIVERSITY OF EASTERN FINLAND Dissertations in Health Sciences

ISBN 978-952-61-2417-9 ISSN 1798-5706

Dissertations in Health Sciences

THE UNIVERSITY OF EASTERN FINLAND

MIKKO HUUSKONEN

NEUROGLIAVASCULAR REMODELING AND THERAPEUTIC INTERVENTION IN ISCHEMIA AND INFLAMMATION

Ischemic stroke is one of the leading causes of death and disabilities worldwide. In this thesis,

we tested the hypothesis that treatment of ischemic stroke could be improved by testing new treatment strategies in clinically relevant

models and by optimization of the actions of pharmacological thrombolysis. Specifically, the studies improved our understanding about

the mechanisms of action and therapeutic potential of bexarotene, copper complexes and

recombinant tissue plasminogen activator.

MIKKO HUUSKONEN

Neurogliavascular remodeling and therapeutic intervention in ischemia and

inflammation

MIKKO HUUSKONEN

Neurogliavascular remodeling and therapeutic intervention in ischemia and

inflammation

To be presented by permission of the Faculty of Health Sciences, University of Eastern Finland for public examination in Tietoteknia auditorium, Kuopio, on Friday, February 10^th 2017, at 12 noon

Publications of the University of Eastern Finland Dissertations in Health Sciences

Number 402

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

University of Eastern Finland Kuopio

2017

Grano Oy Jyväskylä, 2017

Series Editors:

Professor Tomi Laitinen, M.D., Ph.D.

Institute of Clinical Medicine, Clinical Physiology and Nuclear Medicine Faculty of Health Sciences

Professor Hannele Turunen, Ph.D.

Department of Nursing Science Faculty of Health Sciences Professor Kai Kaarniranta, M.D., Ph.D.

Institute of Clinical Medicine, Ophthalmology Faculty of Health Sciences

Associate Professor (Tenure Track) Tarja Malm, Ph.D.

A.I. Virtanen Institute for Molecular Sciences 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): 978-952-61-2417-9

ISBN (pdf): 978-952-61-2418-6 ISSN (print): 1798-5706

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

Author’s address: A.I. Virtanen Institute for Molecular Sciences KUOPIO

FINLAND

Supervisors: Docent Katja Kanninen, Ph.D.

A.I. Virtanen Institute for Molecular Sciences University of Eastern Finland

KUOPIO FINLAND

Professor Jari Koistinaho, M.D., Ph.D

A.I. Virtanen Institute for Molecular Sciences University of Eastern Finland

KUOPIO FINLAND Tarja Malm, Ph.D.

A.I. Virtanen Institute for Molecular Sciences University of Eastern Finland

KUOPIO FINLAND

Reviewers: Associate Professor Tiina M. Kauppinen, Ph.D.

Kleysen Institute for Advanced Medicine University of Manitoba

Winnipeg Canada

Docent Mikko Airavaara, Ph.D.

Institute of Biotechnology University of Helsinki Helsinki

Finland

Opponent: Docent Jouni Sirviö, Ph.D.

Sauloner Oy Ltd Kuopio

Finland

Huuskonen, Mikko

Neurogliavascular remodeling and therapeutic intervention in ischemia and inflammation University of Eastern Finland, Faculty of Health Sciences

Publications of the University of Eastern Finland. Dissertations in Health Sciences Number 402. 2017. 87 p.

ISBN (print): 978-952-61-2417-9 ISBN (pdf): 978-952-61-2418-6 ISSN (print): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

ABSTRACT

Stroke is one of the leading causes of death and disabilities worldwide. Current treatment options for ischemic stroke are based on pharmacological or mechanical recanalization of the blocked artery. However, these treatments can be utilized for only a minority of stroke patients because of increased risk for hemorrhage and a limited therapeutic time window.

One of the problems in the search for new stroke remedies is that preclinically successful therapeutics do not work in human patients. The main reson for this might be that preclinical stroke models do not capture the main aspects of human stroke. Preclinical drug testing is usually carried out in using young, healthy male animals, whereas stroke patients are typically old, both male and female, and suffering from other diseases as well. Another reason might be that classically stroke research has concentrated on maintaining the viability of a single cell type, even though the brain tissue works as a network, both in normal physiological conditions and during ischemia and inflammation.

In this thesis three treatment options for ischemic stroke and neuroinflammation were assessed. In the first study the effect of bexarotene was tested on old mice expressing human P301L-Tau, and subjected to thromboembolic stroke. The results of this study suggested that using old animals, instead of young ones, might significantly alter results. In this study, bexarotene was protective in mice expressing P301L-Tau by modifying autophagy. In the second study the therapeutic potential of two copper complexes, (Cu^II(gtsm) and Cu^II(atsm)), were tested in acute and chronic models of neuroinflammation. The results of this study indicated that these compounds have anti-inflammatory effects, and the capability to shift neuroinflammation-meditating glial cells into a regenerative direction. In the third study, the efficacy of the copper complex Cu^II(atsm) was investigated in two mouse models of ischemic stroke. The results demonstrated that the compound was protective in both stroke models, and provided immunomodulatory effects similar to those described in the second study.

Finally, a study was carried out utilizing collagen XV deficient and wildtype animals that underwent thromboembolic stroke to gain further understanding about the role of collagen XV in ischemic stroke and during treatment with recombinant tPA (rtPA). The collagen XV deficient mice were protected against ischemic stroke, and the impact of the knockout was comparable to that of rtPA treatment. In addition, collagen XV deficient animals did not benefit from additional rtPA treatment. Overall, the results gained from these studies demonstrated that using old animals instead of young animals might significantly alter results in preclinical drug trials. Moreover, these studies demonstrate novel molecules and pathways that could possibly be utilized in the future in the treatment of cerebral ischemia and inflammation.

National Library of Medicine Classification: WL 356, WT 155, QZ 150, QV 247, QY 58, QY 60.R6 Medical Subject Headings: Brain Ischemia/therapy; Stroke/therapy; Thromboembolism/therapy;

Inflammation/therapy; Neuroglia; Vascular Remodeling; Anti-Inflammatory Agents; Neuroprotective Agents;

tau Proteins; Alzheimer Disease; Retinoid X Receptors/agonists; Copper/therapeutic use; Collagen;

Metallothionein; Age Factors; Aging; Disease Models, Animal; Mice

Huuskonen, Mikko

Neurogliavaskulaarinen uusiutuminen ja terapeuttinen interventio iskemiassa ja tulehduksessa Itä-Suomen yliopisto, terveystieteiden tiedekunta

Publications of the University of Eastern Finland. Dissertations in Health Sciences 402. 2017. 87 s.

ISBN (nid.): 978-952-61-2417-9 ISBN (pdf): 978-952-61-2418-6 ISSN (nid.): 1798-5706 ISSN (pdf): 1798-5714 ISSN-L: 1798-5706

TIIVISTELMÄ

Aivohalvaus on yksi yleisimmistä kuolemaan ja pitkäaikaiseen vammautumiseen johtavista syistä maailmanlaajuisesti. Nykyiset iskeemisen aivohavauksen hoidot perustuvat pääasiassa tukkeutuneen verisuonen avaamiseen joko lääkkein tai mekaanisesti. Nämä hoitovaihtoehdot soveltuvat kuitenkin vain pienelle osalle potilaista verenvuotoriskin ja rajoitetun aikaikkunan vuoksi. Aivohalvaustutkimusten ongelmana on solu- ja eläinkokeissa tehokkaiksi havaittujen lääkeiaineiden huono menestyminen potilastutkimuksissa. Syynä tähän pidetään yleisimmin käytettävissä olevien mallien huonoa kykyä mallintaa ihmisen aivohalvausta. Potentiaalisten lääkeaineiden eläintestaus suoritetaan yleensä käyttäen nuoria ja terveitä uroksia, kun taas aivohalvaukseen sairastuvat ihmiset ovat yleensä erilaisista sairauksista kärsiviä vanhoja miehiä ja naisia. Toisena syynä saattaa olla aikaisempien lääketutkimusten liiallinen keskittyminen tietyn solutyypin elinkelpoisuuden säilyttämiseen, vaikka tosiasiassa aivokudos toimii sekä normaalissa tilanteessa että aivohalvauksen ja tulehduksen aikana verkostona.

Tässä tutkimuksessa keskityttiin kolmeen lähestymistapaan iskeemisen aivohalvauksen ja neuroinflamaation hoidossa. Ensimmäisessä osatutkimuksessa testattiin beksaroteenin tehoa vanhoilla ihmisen P301L-Tau-proteiinia ilmentävillä ja villityypin hiirillä tromboembolisen aivohalvauksen hoidossa. Tutkimuksessa huomattiin, että vanhojen hiirien käyttäminen johti osin poikkeaviin tuloksiin aiempiin tutkimuksiin verrattuna ja beksaroteeni osoittautui tehokkaaksi vain P301L-Tau siirtogeenisillä hiirillä muokkaamalla autofagiaa. Toisessa osatutkimuksessa selvitettiin kuparia vapauttavien molekyylien (Cu^II(gtsm) ja Cu^II(atsm)) tehoa hiiren akuutissa ja kroonisessa neuroinflammaatiomallissa. Tulosten perusteella kupariyhdisteet olivat anti-inflammatorisia ja muuttivat aivojen tulehdusta välittäviä gliasoluja vähemmän tulehduksellisiksi. Kolmannessa osatutkimuksissa selvitettiin Cu^II(atsm)-lääkeaineen tehoa kahdessa erilaisessa hiiren iskeemisessä aivohalvausmallissa.

Lääkeaine osoittautui tehokkaaksi molemmissa ja aiheutti toisessa osatutkimuksessa kuvailtujen kaltaisia anti-inflammatorisia vaikutuksia. Viimeisessä osatutkimuksessa keskityttiin selvittämään kollageeni XV:n roolia hiiren tromboembolisessa aivohalvauksessa ja liuotushoidon tehon kannalta käyttämällä tämän proteiinin suhteen poistogeenisiä eläimiä. Havaittiin, että kollageeni XV:n poisto suojasi aivohalvaukselta yhtä tehokkaasti kuin liuotushoito, eikä liuotushoito aiheuttanut näissä eläimissä hoitovaikutusta.

Yhteenvetona voidaan todeta, että ikääntyminen ja liitännäissairaudet saattavat vaikuttaa merkittävästikin prekliinisten lääketutkimusten tuloksiin. Lisäksi tutkimuksissa löydettiin potentiaalisia uusia lääkeaineita ja vaikutuskohteita aivojen iskemian ja tulehduksen hoitoon.

Luokitus: WL 356, WT 155, QZ 150, QV 247, QY 58, QY 60.R6

Yleinen suomalainen asiasanasto: aivohalvaus; aivoinfarkti; iskemia; tulehdus; Alzheimerin tauti;

liitännäistaudit; hoitomenetelmät; kokeelliset menetelmät; koe-eläinmallit; eläinkokeet; koe-eläimet; hiiret

Acknowledgements

This study was performed in the Molecular Brain Research Group, A.I.Virtanen Institute for Molecular Sciences, University of Eastern Finland, during the years 2013-2017. This study was supported by European commission, Sigrid Juselius Foundation, Academy of Finland, Emil Aaltonen Foundation, Orion Research Foundation and Kuopio University Foundation.

I would like to express my thanks to my main supervisor, Docent Katja Kanninen, for her excellent guidance, support and knowledge during these years. I would also like to thank Professor Jari Koistinaho for the possibility to carry out my research work in the Molecular Brain Research Group. His scientific expertise and experience became invaluable to complete these studies. I would also express my gratitude to my third supervisor, Associate Professor Tarja Malm, for her encouragement, mentoring and enthusiasm during the projects.

I am grateful to Associate professor Tiina Kauppinen and Docent Mikko Airavaara for reviewing the thesis and giving constructive comments and feedback. I would also like to thank Docent Jouni Sirviö for acting as an opponent during the defense.

I would like to express my thanks to all the co-authors who contributed to this study:

Hiramani Dhungana, Sanna Loppi, Paula Korhonen, Sara Wojciechowski, Sighild Lemarchant, Yuriy Pomeshchik, Velta Keksa-Goldsteine, Gundars Goldsteins, Piia Valonen, Eveliina Pollari, Katarína Lejavová,Laura Periviita, Lotta Kosonen, Suvi Vähätalo, Juho Koponen, Taina Pihlajaniemi, Ritva Heljasvaara, Denis Vivien, Qing-zhang Tuo, Paul S.

Donnelly, Alexandra Grubman, Frederick R. Walker, Rong Liu, Ashley I. Bush, Peng Lei, Anthony R. White, Xin Yi Choo, Alexandra Grubman, Diane Moujalled, Jessica Roberts, Jeffrey R. Liddell, Clare Duncan, Simon James, Martin de Jonge, Lachlan E. McInnes, David J. Hayne, Mikko Kettunen, Akihiko Takashima and Gary Landreth.

I am especially grateful to Mirka Tikkanen and Laila Kaskela for their help and experience in practical lab work and maintenance of the facilities. I would also like to thank all the current and past members of the Molecular Brain Research and Neuroinflammation Research Groups, especially Meike Keuters, Natalia Kolosowska, Henna Konttinen, Eila Korhonen, Marja Koskuvi, Sarka Lehtonen, Riikka Martikainen, Minna Oksanen, Yajuvinder Singh, Merja Jaronen and Ekaterina Savchenko.

I would also like to thank the personnel of the administration at A.I.Virtanen -Intitute:

Jaakko Hellén, Hanne Tanskanen, Joanna Huttunen, Docent Riikka Pellinen and Katja Pesonen. In addition, I am thankful for Jari Nissinen and Jouko Mäkäräinen for their technical support.

Finally, I wish to express my gratitude to my family, friends and everyone who supported me during these years.

Kuopio, 2017

Mikko Huuskonen

List of the original publications

This dissertation is based on the following original publications:

I Mikko T. Huuskonen, Sanna Loppi, Hiramani Dhungana, Velta Keksa- Goldsteine, Sighild Lemarchant, Paula Korhonen, Sara Wojciechowski, Eveliina Pollari, Piia Valonen, Juho Koponen, Akihiko Takashima, Gary Landreth, Gundars Goldsteins, Tarja Malm, Jari Koistinaho, Katja M. Kanninen. Bexarotene targets autophagy and is protective against thromboembolic stroke in aged mice with tauopathy. Scientific Reports 2016 Sep 14;6:33176. doi: 10.1038/srep33176.

II Xin Yi Choo*, Alexandra Grubman*, Mikko T. Huuskonen, Diane Moujalled, Jessica Roberts, Jeffrey R. Liddell, Clare Duncan, Simon James, Martin de Jonge, Lachlan E. McInnes, David J. Hayne, Paul S. Donnelly, Eveliina Pollari, Suvi Vähätalo, Katarína Lejavová, Mikko Kettunen, Tarja Malm, Jari Koistinaho, Anthony R. White, Katja M. Kanninen. Copper bis(thiosemicarbazone) complexes attenuate neuroinflammation through modulation of metallothionein 1.

Manuscript

III Mikko T. Huuskonen*, Qing-zhang Tuo*, Sanna Loppi, Hiramani Dhungana, Paula Korhonen, Lachlan E. McInnes, Paul S Donnelly, Alexandra Grubman, Sara Wojciechowski, Katarína Lejavová, Yuriy Pomeshchik, Laura Periviita, Lotta Kosonen, Frederick R. Walker, Rong Liu, Ashley I. Bush, Jari Koistinaho, Tarja Malm, Anthony R. White, Peng Lei, Katja M. Kanninen. The copper bis(thiosemicarbazone) complex Cu^II(atsm) is protective against cerebral ischemia through modulation of the inflammatory milieu. Neurotherapeutics, 2017 Jan 3. doi:

10.1007/s13311-016-0504-9

IV Hiramani Dhungana*, Mikko T. Huuskonen*, Taina Pihlajaniemi, Ritva Heljasvaara, Denis Vivien, Katja M. Kanninen, Tarja Malm, Jari Koistinaho, Sighild Lemarchant. Lack of collagen XV is protective after ischemic stroke in mice. Cell Death and Disease, 2017 Jan 12;8(1):e2541. doi: 10.1038/cddis.2016.456

*Equal contribution

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

Contents

1 INTRODUCTION ………...….………..……...1

2 REVIEW OF LITERATURE………..………….………....………...3

2.1 Ischemic stroke……….………...3

2.1.1 Epidemiology and etiology……….…………..…………...…………...3

2.1.2 Economical burden .……….……….………..3

2.1.3 Treatment…….……….……….………..3

2.1.4 Outcome………...4

2.2 Modelling ischemic stroke………...5

2.2.1 In vivo models………….…...…….……….…….5

2.2.1.1 Global ischemia models….……….………...………...….5

2.2.1.2 Focal ischemia models………...…....………...…5

2.2.1.3 Spontaneous ischemia models………...…...………...………..……...7

2.1.2 In vitro models……….……….………...7

2.3 Translational roadblock……….…..……….8

2.4 Neurogliavascular network in the context of ischemic stroke………..……….………...9

2.4.1 Components of neural tissue……….………9

2.4.1.1 Vascular cells……….………..…9

2.4.1.1.1 TPA and neurovascular uncoupling after cerebral ischemia…...10

2.4.1.2 Oligodendrocytes……...………...………...…11

2.4.1.3 Astrocytes………...………..…….12

2.4.1.4 Microglia ………...…....……….………..….13

2.4.1.4.1 Phenotypes and stroke-induced activation of microglia………..13

2.4.1.3.2 Role of microglia in stroke...………...….………..15

2.4.1.5 Neurons in ischemic stroke……...……...………...13

2.4.2 Mitochondrial function in stroke………..……...…...17

2.4.3 Oxidative and nitrosative stress in the neurogliavascular network…...17

2.4.4 Cell death following ischemic stroke………..………...18

2.4.5 Autophagy ………...………...……….………..20

2.4.5.1 Autophagy in ischemic stroke………...…………...………..21

2.5 Neurogliavascular network in Alzheimer´s disease………..………22

2.5.1 Pathophysiology of AD………22

2.5.2 AD as a neuroinflammatory disease………...22

2.5.3 AD and cerebrovascular dysfunction……….24

2.6 Therapeutic approaches targeting the cerebral network in inflammatory and ischemic conditions………26

2.6.1 Bexarotene...………...………...…………..……....…...……..26

2.6.2 CuII(atsm) ...………...……….……...26

2.6.3 TPA and chondroitin sulphate proteoglycans...28

3 AIM OF THE STUDY………….……….………..…...…..……...29

4 MATERIALS AND METHODS………...…….………...…………...30

4.1 In vivo models……...……..…………...………...…...30

4.1.1 Animals ………..………..………...30

4.1.2 Cerebral ischemia…………..………..……….…...30

4.1.3 Systemic inflammation……….………31

4.2 In vitro models………….……….………..…………..………...31

4.2.1 Primary neuron cultures………...………...31

4.2.2 N2a cell line……….………...31

4.2.3 Astrocyte cultures ………...31

4.2.4 Microglia cultures ……….32

4.3 Drug treatment ……….………...………....32

4.3.1 Bexarotene……….…….32

4.3.2 CuII(atsm) and CuII(gtsm)………...……....32

4.3.3 RtPA………...…….………...33

4.4 Measurement of outcome………..……….………..……..33

4.4.1 MRI……...……….……...33

4.4.1.1 Volumetry ………...33

4.4.1.2 In vivo cerebrovascular inflammation………...33

4.4.2 Behavioral outcome ……...………..………..………...34

4.4.2.1 Catwalk gait analysis………...………...……...34

4.4.2.2 Neurologic examination………..…………...…...34

4.4.2.3 Latency to move………..………...…………..………...34

4.4.3 Histology………….……...………..………...34

4.4.3.1 Lesion volume………...…...………...34

4.4.3.2 Immunohistochemistry..………..………...35

4.4.3.3 Vascular anatomy……...………..………...36

4.4.4 Western blotting….……...………..………..………...36

4.4.5 Cytometric bead array….……...………..………...37

4.4.6 ELISA ………....….……...………...………...37

4.4.7 Brain cell isolation and FACS………...………...38

4.4.8 Inductively coupled mass spectrometry………...…………...…...38

4.4.9 Cell viability……….………..………….………...38

4.4.10 Autophagic flux……….………...39

4.4.11 Mitochondrial function………...……….………...39

4.4.12 NO measurement………….………...39

4.4.13 X-ray fluorecence microscopy……….…………...39

4.4.14 Quantitative real-time polymerase chain reaction (Rt-PCR)...40

4.5 Statistical analyses and exclusion criteria….………...………..…………...40

5 RESULTS………...…….………..………...41

5.1 Bexarotene is protective in a mouse model combining aging, tau-pathology and thromboembolic stroke by modulating autophagy………...……….…...41

5.1.1 Bexarotene protects olf P301L-Tau transgenic animals against stroke….…...41

5.1.2 Neuroprotective effect of bexarotene against excitotoxicity in vitro is not genotype-specifc …………...……….……..………41

5.1.3 Bexarotene has no effect on tau, Aβ, MMP-9, gliosis and the peripheral immune response in aged ischemic mice…………..………...41

5.1.4 Lc3b levels are reduced in ischemic P301L-Tau mice and this drop in protein expression is reversed by bexarotene treatment…….……...………...42

5.1.5 Bexarotene-induced increases in autophagy markers in N2a cells carrying the P301L-Tau mutation is caused by inhibition of autophagy flux……….…...42

5.1.6 P301L-Tau mutation alters bioenergetic functions of neurons and these alterations are alleviated by bexarotene treatment ………...42

5.2 Copper bis(thiosemicarbazonata) complexes modulater experimental neuroinflammation in vitro and in vivo by increasing metallothionein 1………..…...43

5.2.1 Copper bis(thiosemicarbazonato) complexes scavenge acute and chronic neuroinflammation...43

5.2.2 Both microglia and astrocytes gain a less pro-inflammatory phenotype in the presence of CuII(atsm) in vitro...44

5.2.3 CuII(atsm) increases the copper content of cytokine-stimulated cells...44

5.2.4 Anti-inflammatory effects of CuII(atsm) are mediated by metallothionein 1….. 44

5.3 CuII(atsm) improves outcome in permanent and transient models of cerebral ischemia and modifies microglial activation………...44

5.3.1 CuII(atsm) protects cortical neurons against excitotoxicity and N2a cells against OGD in vitro...45

5.3.2 CuII(atsm) is protective in a mouse model of transient ischemia...45

5.3.3 CuII(atsm) reduces ischemic damage in a permanent model of stroke...45

5.3.4 CuII(atsm) reduces CD45 expression in cells located in the ischemic core...45

5.3.5 Iba1 expression is reduced and the morphology of Iba-1 positive cells is altered by CuII(atsm) treatment after ischemia...45

5.3.6 The proportion of resident microglia is increased in CuII(atsm) treated ischemic brains...46

5.3.7 CuII(atsm) has beneficial effects on further inflammatory markers after ischemia...46

5.4 Collagen XV deficiency ameliorates ischemic damage in mice….………..46

5.4.1 Α1-Collagen XV deficiency protects against thromboembolic stroke...46

5.4.2 RtPA increases α1-Collagen XV levels in the plasma of WT mice but not in α1- Collagen XV KO mice...46

5.4.3 RtPA treatment does not reduce post-stroke cytokine production in α1-Collagen XV KO mice...47

5.4.4 Α1-Collagen XV deficiency and rtPa treatment results in a similar pattern of increased VEGF-A expression during ischemia………...……..……..47

6 DISCUSSION……….………..………...48

6.1 Modulation of autophagy by bexarotene protects old mice expressing P301L-Tau against stroke………..……….…...48

6.2 Copper delivery is protective in stroke and alleviates neuroinflammation……….50

6.3 Collagen XV deficiency and rtPA administration are equally protective in a mouse model of thromboembolic stroke………..…...……54

6.4 Significance with relation to stroke and neuroinflammation………..……...……55

6.5 Future perspectives…………...………..………...……56

7 CONCLUSIONS………..………...57

REFERENCES ….………..………...………..………...59

ORIGINAL PUBLICATIONS …….………..………...89

Abbreviations

Aβ Amyloid beta AD Alzheimer´s disease

ALS Amyotrophic lateral sclerosis AMPA α-amino-3-hydroxy-5-methyl-

4-isoxazole propionic acid APOE Apolipoprotein E

APP Amyloid precursor protein AQoL Assessment of quality of life AQP-4 Aquaporin-4

ASC Apoptosis speck-like adaptor Atg Autophagy-related protein ATP Adenosine triphosphate BBB Blood brain barrier

BDNF Brain-derived neurotrophic factor

CBA Cytometric bead array CBF Cerebral blood flow

CCL2 Chemokine C-C motif ligand 2, also known as MCP-1 CCR2 C-C chemokine receptor type

2

CD Cluster of differentiation CNS Central nervous system COX2 Cyclo-oxygenase 2

CSPG Chondroitin sulphate

proteoglycan

Cu^II(atsm) Diacetylbis(N(4)-

methylthiosemicarbazone) copper(II)

CXCL10 Chemokine C-X-C motif ligand 10

CX3CR1 C-X3-C motif chemokine receptor 1

DAMP Damage-associated molecular pattern

DMEM Dulbecco´s modified Eagle´s medium

ELISA Enzyme-linked immunosorbent assay ETC Electron transport chain ERKs Extracellular signal-regulated

kinases

FACS Fluorescence-activated cell- sorting

FBS Fetal bovine serum

GDNF Glial cell line-derived neurotrophic factor

GFAP Glial fibrillary acidic protein HIF-1α Hypoxia-inducible factor-1α HT High-throughput

IFN-γ Interferon-gamma

IGF-1 Insulin like growth factor 1 iNOS Inducible nitric oxide

synthase IL Interleukin

IMDM Iscove´s modified Dulbecco´s medium

LC3B Microtubule-associated protein 1 light chain 3β LPS Lipopolysaccharide

MCA(O) Middle cerebral artery (occlusion)

MCP-1 Monocyte chemotactic protein-1, also known as CCL2 M-CSF Macrophage-colony

stimulating factor

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

MRI Magnetic resonance imaging

mTOR Mechanistic target of

rapamycin

MTT 3-(4, 5-dimethylthiazolyl-2)-2, 5-diphenyltetrazolium

bromide

NADH Nicotinamide adenine

dinucleotide , reduced form NMDA N-methyl-D-aspartic acid NO Nitrous oxide

NOX Nicotinamide adenine

dinucleotide phosphate oxidase

nNOS Neuronal nitric oxide

synthase

OGD Oxygen and glucose

deprivation OR Odds ratio

PAMP Pathogen-associated molecular patterns PBS Phosphate-buffered saline PD Parkinson´s disease PFA Paraformaldehyde

PIK3C3 Phosphatidylinositol 3-kinase catalytic subunit 3

PSEN Presenilin

RAGE Receptor for advanced

glycosylation endproducts RNS Reactive nitrogen species ROS Reactive oxygen species RXR Retinoid X-receptor rtPA Recombinant tissue

plasminogen activator RT-PCR Reverse transcription

polymerase chain reaction SHSPR Spontaneously hypertensive

stroke-prone rat

SOCS3 Suppressor of cytokine signaling 3

SOD1 Superoxide dismutase [Cu- Zn]

TGF-β Transforming growth factor β TLR Toll like receptor

TNFα Tumor necrosis factor alpha TNFR TNF receptor subfamily

member receptor Th2 Type 2 T helper cell

TREM2 Microglial triggering receptor on myeloid cells 2

VCAM-1 Vascular cell adhesion molecule 1

VEGF-A Type A vascular endothelial growth factor

1 Introduction

Stroke is one of the leading causes of death and disabilities worldwide. The prevalence of stroke in the Northern American population is 2.8% and more than 800 000 strokes are registered each year (Mozaffarian, Benjamin et al. 2015). The corresponding numbers are 1.5% and 21 000 in Finnish population (Meretoja, Roine et al. 2010). Most of the strokes, approximately 85 % in Western countries, are caused by blockage in a cerebral artery and subsequent reduction in blood flow to brain (Mozaffarian, Benjamin et al. 2015). The rest of the cases are caused by a rupture in a cerebral artery leading to a hemorrhagic stroke.

Treatment of ischemic stroke relies mostly on recombinant tissue plasminogen activator (rtPA, alteplase) which initiates the cascade leading to breakdown of a blood clot. Based on meta-analyses, the effective time window for rtPA treatment is limited to 4.5 h (Emberson, Lees et al. 2014). However, about 25% of stroke patients arrive at the hospital within a 4.5 h timeframe and only 6% of the patients are eligible for intravenous rtPA treatment due to other contraindications (de Los Rios la Rosa,F., Khoury et al. 2012). Furthermore, thrombolytic use of rtPA is associated with serious side effects such as intracerebral hemorrhage (Graham 2003). Endovascular recanalization techniques have evolved fast during recent years and can sometimes be used if rtPA treatment has been ineffective or contraindicated. Regardless of the improved survival rates after stroke by current treatment options, survivors often suffer from long-term disabilities, cognitive impairment and reduced quality of life (Wolfe, Crichton et al. 2011, Paul, Sturm et al. 2005). Thus, there is a need to discover novel, more effective treatments for the disease that would improve the poor outcome of stroke survivors.

Despite the rigorous reseach in the field of stroke, rtPA still remains the only pharmacological treatment option for ischemic stroke that is approved by the FDA. In fact, the field of stroke research is especially well-known for the poor translatability of preclinically successful treatments into clinics: none of the over 1000 treatment strategies tested in stroke models or more than 100 tested in human patients has led to regulatory approval (O'Collins, Macleod et al. 2006). One of the main reasons for this is the inability to capture all the features of human stroke in to the preclinical models. Even though human stroke patients are typically old, both male and female, and have various comorbid diseases such as diabetes and cardiovascular diseases, most of the preclinical stroke studies are still performed using young and healthy male animals (Dirnagl, Hakim et al. 2013).

Historically, preclinical stroke research has focused on preserving viability of neurons which may also be one of the reasons for the poor clinical success of preclinically neuroprotective compounds. However, the brain tissue should be seen as a network of neuronal, glial and vascular cells that work as a network to maintain homeostasis both during normal conditions and during cerebral ischemia. In this concept, the loss of the neuronal-glial-vascular network integrity is a key link between the phenomena occurring in the ischemic brain that lead to the evolution of a vascular event to a neurological condition.

Future therapeutic strategies should try to preserve the whole network integrity and target pathologic cascades beyond the restoration of blood flow into the tissue.

This thesis aims to address the issues mentioned above by testing three novel treatment options in animal and cell culture models of ischemic stroke, inflammation and Alzheimer’s disease (AD). In the first study, the efficacy and mechanism of action of an RXR agonist, bexarotene, were tested in a mouse model combining high age, tau pathology and ischemic stroke. Next, the effects of copper complexes, the bis(thiosemicarbazones) were assessed in in vivo and in vitro models of inflammation and AD. Third, the therapeutic potential of a certain copper bis(thiosemicarbazone) complex, Cu^II(atsm), was tested in two in vivo models of ischemic stroke. Finally, we provided new insight into the role of a basal membrane protein, collagen XV, in thromboembolic stroke.

2 Review of literature

2.1 ISCHEMIC STROKE 2.1.1 Epidemiology and etiology

Stroke is a major cause of death and disabilities worldwide. The prevalence of stroke in the Northern American population is 2.8% with more than 800 000 new or recurrent strokes each year (Mozaffarian, Benjamin et al. 2015). In Finnish population the corresponding numbers are 1.5% and 21 000 (Meretoja, Roine et al. 2010). In Western countries approximately 85% of strokes are ischemic - caused by a blockage in a cerebral artery leading to reduced blood flow in a certain brain area. In the rest of the cases, a rupture in a cerebral artery leads to a hemorrhagic stroke. Based on their etiology, ischemic strokes can be further divided into cardioembolic, arteroembolic, lacunar and cryptogenic strokes, each accounting for approximately 25% of cases (Ay, Furie et al. 2005). Cardiac emboli is usually associated with atrial fibrillation leading to the formation of thrombi in atrial appendages, whereas atherosclerosis of the aortic arch or extracranial cervical arteries leads to arterial embolism.

Lacunar strokes are small infarcts caused by blockage of penetrating arteries in the deeper parts of the brain. In cryptogenic stroke the initial findings are similar to embolic stroke, but the cause remains unclear.

2.1.2 Economical burden

In addition to hospital-associated costs, stroke causes substantial long-term direct and indirect costs in the forms of rehabilitation and loss of productivity. The mean lifetime costs of ischemic stroke in the United States is estimated to be over $140 000 (Mozaffarian, Benjamin et al. 2015). This means a total of $65.5 billion expenditures in 2008 with indirect costs accounting for more than half of the costs (Demaerschalk, Hwang et al. 2010, Mozaffarian, Benjamin et al. 2015, Mozaffarian, Benjamin et al. 2015). In Finland, annually

€1.6 billion is used for the treatment of stroke patients (approximately 7% of national healthcare costs) (Meretoja, Kaste et al. 2011). Even though the incidence of ischemic stroke is decreasing in high income countries, the increasing prevalence due to aging populations is expected to increase total expenditures dramatically during the next decades (Feigin, Forouzanfar et al. 2014).

2.1.3 Treatment

Traditionally emergency treatment of ischemic stroke has relied on intravenous pharmacological thrombolysis with recombinant tissue plasminogen activator (rtPA alteplase). The efficacy of rtPA treatment is highly dependent on the occlusion location and delay in administration after onset of stroke (Emberson, Lees et al. 2014, del Zoppo, Poeck et

al. 1992, del Zoppo, Poeck et al. 1992). The highest recanalization rates are observed in distal occlusions while proximal occlusions are less likely to undergo recanalization (rates go from 40% to 26% as the location changes from middle cerebral artery (MCA) branches towards internal carotid artery) (del Zoppo, Poeck et al. 1992). Approximately one third of ischemic strokes are proximal, thus reducing the probability of recanalization after rtPA treatment(Heldner, Zubler et al. 2013).

The effective time window for rtPA treatment has been confirmed in a meta-analysis combining data from randomized trials (Emberson, Lees et al. 2014). Results show that administration of tPA within 3 hours of stroke leads to a good outcome versus placebo with an odds ratio (OR) of 1.75, but the efficacy quickly decreases if treatment starts within 3-4.5 h of stroke onset (OR 1.25) and is comparable to placebo if the administration is done later than 4.5 h after stroke onset (OR 1.15). According to studies conducted in the United States, about 25% of stroke patients arrive into the hospital within a 4.5 h timeframe and only 6% of the patients are eligible for intravenous rtPA treatment (de Los Rios la Rosa,F., Khoury et al.

2012). In addition, rtPA treatment can be associated with serious side-effects such as intracerebral hemorrhage (Graham 2003)

In some cases where rtPA treatment has been ineffective or contraindicated, endovascular recanalization techniques can be utilized. The equipment needed for the operation have evolved during the last decade, and recent clinical trials have proven the efficacy of this procedure (Holodinsky, Yu et al. 2016). In certain cases endovascular therapies have advantages over rtPA: recanalization can be done even 8 hours after the occlusion and this method suits well for removal of proximal occlusions where rtPA treatment is less effective.

However, patients for thrombectomy should be carefully chosen based on advanced imaging techniques and treatment should be done in specialized stroke units (Badhiwala, Nassiri et al. 2015).

2.1.4 Outcome

Due to improved stroke care and management during the last decades, mortality rates of stroke patients have declined, especially in high income countries (Feigin, Forouzanfar et al.

2014). Nowadays 64% of stroke patients live for one year, 43% of patients for five years and 24% of patients for 10 years following stroke (Wolfe, Crichton et al. 2011). 10-20% of stroke patients suffer from moderate or severe disability even though rehabilitation interventions are in place. Cognitive impairment and depression are relatively common in stroke survivors, both affecting roughly one third of patients after three years. Finally, the quality of life of stroke survivors is dramatically reduced: approximately 20% of stroke survivors have very poor assessed quality of life 5 years after stroke (AQoL instrument score from -0.04 to 0.10 meaning life quality worse than death or close to death), while 3% of the general population have this low score (Paul, Sturm et al. 2005).

2.2 MODELLING ISCHEMIC STROKE 2.2.1 In vivo models

Rodents are most commonly used to model ischemic stroke in vivo, although rabbits, gerbils, cats, dogs, pigs and monkeys have also been used to some extent. The cerebrovascular anatomy and physiology of rodents is very similar to humans, except for the lack of gyrencephalic brains (Yamori, Horie et al. 1976). This can be modelled by using non-human primates when approaching clinical trials (Stroke Therapy Academic Industry Roundtable (STAIR) 1999). In vivo models of ischemic stroke can be divided into induced and spontaneous models (Krafft, Bailey et al. 2012). Induced models can be further divided into global and focal ischemia models.

2.2.1.1 Global ischemia models

Global ischemia refers to cerebral hypoperfusion caused in humans by reduced blood flow to all parts of the brain due to heart failure from cardiac arrest or arrhythmias or from reduced cardiac output. Experimental global ischemia can be achieved by decapitation, cardiac arrest, induction of simultaneous systemic hypotension and hypoxia, or by using two and four vessel occlusion models (Kumar, Aakriti et al. 2016). Nowadays the decapitation model is rarely used because the assessment of recovery is not possible (Abe, Yoshida et al.

1983). Cardiac arrest can be induced by pharmacological or surgical procedures and it can be considered clinically relevant as heart dysfunction is a relatively common cause of stroke in humans (Safar, Stezoski et al. 1976). However, increased mortality, systemic effects and differences in experimental conditions increase variability between studies and complicate the use of this model. Similarly, a combination of systemic hypotension and hypoxia can be used to produce cerebral infarcts, yet these are difficult to reproduce (Yatsu, Lindquist et al.

1974). In the four vessel occlusion model vertebral arteries are first cauterized and ischemia is next induced by the occlusion of common carotid arteries (Pulsinelli, Duffy 1983).

Reperfusion after the ischemic period leads to changes in cerebral blood flow (CBF) distribution to specific brain areas, which is considered the most interesting feature of this model although the effect is dependent on the rodent and strain used (Pulsinelli, Levy et al.

1982). The two-vessel occlusion model is surgically a more simple method, which combines bilateral common carotid artery occlusion to systemic hypotension (Smith, Bendek et al.

1984). In addition, reperfusion occurs faster in this model when compared to the four vessel occlusion model, and leads to selective injury in the hippocampus, the caudoputamen and the neocortex.

2.2.1.2 Focal ischemia models

Most focal ischemia models involve the occlusion of one major cerebral blood vessel either permanently or transiently. Depending on the site and duration of the occlusion, this results

in a reduction or arrest in CBF to a specific brain area (ischemic core), which is surrounded by an area of some blood flow through collateral circulation (penumbra) (Traystman 2003).

Proximal or distal MCA occlusion (MCAO) are most widely used for their relevance to human stroke (del Zoppo, Poeck et al. 1992). MCAO in rodents leads to long-term sensorimotor and cognitive deficits accompanied by postural and sensory reflex impairments (Bouet, Freret et al. 2007).

Permanent proximal MCAO in rodents has been a standard model for ischemic damage in the cortical and caudoputamen areas since it´s development (Tamura, Graham et al. 1981).

Nowadays several variations of the model exist, one example being the more distal MCAO combined with temporary ipsilateral common carotid artery occlusion (Chen, Hsu et al.

1986). This modification provides a surgically simplified method to achieve ischemic damage that is limited to the frontoparietal cortex. Use of electrocoagulation instead of sutures limits CBF reduction closer to the actual area of injury. This can be further improved using photochemical MCAO, but as a disadvantage, the clinically interesting penumbral area is lost(Watson, Dietrich et al. 1985). Ischemic lesions can also be achieved by injection of vasoconstrictors such as endothelin-1 to the MCA region, but due to the effect on several microvessels, this substance is best suited for the modelling of lacunar infarcts (Bailey, McCulloch et al. 2009, Sharkey, Ritchie et al. 1993).

The clinical relevance of permanent or transient MCAO can be increased by using thromboembolic or non-clot embolic models (Kumar, Aakriti et al. 2016). The thromboembolic model mimics the characteristics of human stroke most closely and is considered the gold standard for mechanistic studies of ischemic stroke. Lesions can be produced using autologous thrombi, human blood clots or fragments as well as by injecting thrombin into the MCA bifurcation (Kudo, Aoyama et al. 1982, Chen, Zhu et al. 2015, Orset, Macrez et al. 2007). In non-clot embolic models artificial material like silicone, collagen, polyvinylsiloxane, retractable silver ball and surfactant sodium dodecyl sulphate (SDS) can be used to induce embolism (Lauer, Shen et al. 2002, Molnar, Hegedus et al. 1988, Purdy, Devous MD et al. 1989, Yang, Yang et al. 2002, Toshima, Satoh et al. 2000). Even though thromboembolic models lead to high variability in lesion size and location, are prone to spontaneous recanalization and increased risk for hemorrhages, they are still the only models that can be used to study the combined effect of thrombolytic and neuroprotective drugs (Zhang, Zhang et al. 2004).

Transient MCA occlusion can be simply achieved by using ligature or clip around the MCA and releasing it after a certain time (Shigeno, Teasdale et al. 1985). Probably the most common method utilizes an intraluminal filament to block the MCA (Longa, Weinstein et al.

1989). Changing intraluminal filament and occlusion properties affects the collateral circulation and nature of the ischemic lesion (Abraham, Somogyvari-Vigh et al. 2002).

Advantages of this model include ischemic lesion and penumbra relevant to human stroke (under laser Doppler guidance), as well as controlled reperfusion of the vessel.

2.2.1.3 Spontaneous ischemia models

In addition to induced ischemic stroke models, spontaneous animal models such as the spontaneously hypertensive stroke-prone rat (SHRSP) exist (Saito, Matsunaga et al. 1975).

This rat strain develops malignant hypertension by 12 weeks of age and stroke-like symptoms by the age of 20 weeks, but it has mostly been used in MCAO studies to produce large infarcts (Krafft, Bailey et al. 2012). Nonetheless, careful examination of the less extensively studied spontaneous infarcts have suggested interesting facts about stroke pathophysiology. Contrary to earlier assumptions, elevated blood pressure may not be the initial cause for infarcts in these animals, but instead, endothelial dysfunction, blood brain barrier (BBB) leakage and inflammation precede the hypertension and stroke lesions as they do in human lacunar stroke (Bailey, Wardlaw et al. 2011). Thus, SHRSP is considered the best animal model for lacunar infarcts (Bailey, Smith et al. 2011).

2.2.2 In vitro models

Cerebral ischemia can be modeled in cell cultures in various ways, but comparative studies between the methods have not been made (Holloway, Gavins 2016). Typically, ischemic conditions are achieved by enzymatic or chemical inhibitors of metabolism, oxygen and glucose deprivation (OGD) or by using excitotoxic compounds. Chemical methods utilize rotenone, antimycin or sodium azide to inhibit the electron transport chain (ETC) and adenosine triphosphate (ATP) production (Kurian, Pemaih 2014). The glucose oxidase/catalase system with 2-deoxyglucose can be used to remove oxygen from the media, leading to ischemic damage of the cultured cells in a relatively short time(Kurian, Pemaih 2014). The OGD model mimics the diminished blood flow in ischemic stroke most closely, and reperfusion can be included in the model. This causes long-term neuronal degeneration and increases extracellular glutamate concentrations when performed in mouse primary cortical neuron cultures (Goldberg, Choi 1993). Glutamate or N-methyl-D-aspartatic acid (NMDA) can also be added for the cell cultures alone to model the excitotoxic component of stroke (von Engelhardt, Coserea et al. 2007).

Most of the in vivo experiments of ischemic stroke are performed using primary cultures obtained from neonatal animals. These cultures have been used to study neuronal cell death, signaling, myelination, synaptic plasticity, potential stroke therapies, and they can be used as a basis for high-throughput (HT) screening assays (Holloway, Gavins 2016). However, these models do not provide the possibility to inspect systemic effects (i.e. cardiovascular system, peripheral immune response or glymphatic system). Primary glial and cortical neuron cultures are widely used and allow the inspection of the role of each cell type on disease pathology, even though it is still questionable how relevant neonatal rodent cerebral tissue is to the adult human brain (Abney, Bartlett et al. 1981, Berger, Di Porzio et al. 1982).

Immortalized cell lines provide a stable supply of human cells and options include immortalized human microglial cells, teratoma derived neuronal cells, astrocytes and oligodendrocytes and neuroblastoma cells as well as endothelial cell lines (Mathews, Johnson

et al. 1976, Nagai, Nakagawa et al. 2001, Haile, Fu et al. 2014). An obvious drawback of these cell lines is the varying expression of cell proliferation and adhesion molecules leading to growth with different morphologies. In addition, differences in the production of molecules, enzymes and transporters causes functional differences between primary and immortalized cells. In the future, the use of induced pluripotent stem cell (iPSC) derived cells will offer an attractive and relevant opportunity for screening of stroke medicines as well as producing patient specific cellsto inspect genetic factors of cerebral ischemia. Moreover, iPSC cell can theoretically be used for regenerative purposes, replacing the original tissue lost by cerebral ischemia.

Another way to study stroke in vitro is to use brain tissue slices. The oldest methods involved the maintenance of rodent brain slices in artificial cerebrospinal fluid and exposure of the sections to focal (by limiting the condition to a certain region of interest) or global OGD (Dong, Schurr et al. 1988). This system can also be applied to human tissue, thus allowing inspection of true neuronal circuits (Werth, Park et al. 1998). Unfortunately, the availability of human samples is very limited and many samples are collected from young epileptic patients, thus having excessive neuronal excitability (Holloway, Gavins 2016). In addition, sectioning of brain tissue always causes trauma to the slices so the experiments are always started from a post-traumatic state.

2.3 Translational roadblock

Translational roadblock is a term referring to the poor clinical success of preclinically successful drugs. This phenomenon is especially evident in the field of stroke: over 1000 treatment strategies have been tested in stroke models and more than 100 in human patients but still rtPA remains the only pharmacological treatment option for ischemic stroke approved by the emea? vai kokonainen or FDA (O'Collins, Macleod et al. 2006). This has led to various meetings and guidelines to improve the translation of preclinical research for stroke therapies (Lapchak, Zhang et al. 2013). One of the most commonly discussed issues is the modelling of human stroke in vivo and in vitro.

Most of the preclinical stroke studies are performed using young and healthy male animals (Dirnagl, Hakim et al. 2013). On the contrary, human patients are typically old, both male and female, and have various comorbid diseases such as diabetes and cardiovascular diseases. Furthermore, cerebral ischemia is typically modelled with MCAo resulting in consistent lesions, whereas in human stroke there is a great variability in the site, duration and extent of ischemia. The method to induce ischemia in preclinical studies should be chosen carefully. The thromboembolic model mimics most closely human stroke pathophysiology, and allows tPA treatment to be included as a positive control and to study protective effects after recanalization. In addition, the treatment regimen in stroke studies involving animals is sometimes impossible to execute in a clinical set-up (especially in studies where a pre-treatment is required). Furthermore, the complexity of stroke pathophysiology

is often underestimated and systemic effects are commonly neglected in preclinical animal studies.

Even though there will be very advanced methods to model the complex structures of brain tissue in the future, some basic things can be done to better mimic physiological condition in the cell cultures (Holloway, Gavins 2016). Cells are typically grown and studied in 5% CO2, but the oxygen concentration is close to atmospheric conditions (20%). However, such high oxygen levels are not achieved in the tissues, as circulating blood contains 10.5- 13% of oxygen, and tissues typically contain 2-8% of oxygen (Tiede, Cook et al. 2011).

Similarly, plasma glucose levels are 5.5-7.8 mmol/L and corresponding levels in brain 0.82- 2.4 mmol/L (0.16-4.5 mmol/L in extreme cases) (Kleman, Yuan et al. 2008). It has been shown that this kind of difference in the availability of oxygen leads both to altered differentiation of the cells as well as changes in response to external challenges (Studer, Csete et al. 2000, Tiede, Cook et al. 2011). Thus, OGD studies should be performed with oxygen below 2% and the cells should be returned to normoxia (2-5% of oxygen) instead of atmospheric levels (Holloway, Gavins 2016). In addition, regular cell culture media can contain more than 20 mmol/L glucose meaning that cells are contantly exposed to ten times higher glucose levels than in vivo. This in addition to the levels of other energy sources has direct effects on the functions of cells (Kleman, Yuan et al. 2008).

2.4 Neurogliavascular network in the context of ischemic stroke

One reason for the poor success of preclinically neuroprotective compounds in clinical trials might be the exaggerated focus on the observed neuronal effects. The brain tissue consists not only of neurons, but a network of neuronal, glial and vascular cells that work together to maintain homeostasis both during normal conditions and during cerebral ischemia. In fact, ischemic stroke can be described as a vascular disease with neurological consequences, and loss of the neuronal-glial-vascular network integrity is a key link between the phenomena occurring in the ischemic brain.

2.4.1 Components of neural tissue 2.4.1.1 Vascular cells

Vascular cells form the surface between the bloodstream and cerebral tissue, as well as the starting point for stroke-induced trauma. The cerebrovascular tree formed by large interconnected arteries, pial arteries, penetrating intracerebral arteries, arterioles and capillaries ensures that the vast energy expenditure of brain (20-25% of total body oxygen and glucose consumption vs. 2% of bodyweight) is constantly satisfied (Raichle, Mintun 2006). Contrary to earlier assumptions, the cerebrovascular system has a more complex and important role in maintaining cerebral homeostasis, especially in the context of stroke.

The BBB is formed by cerebral endothelial cells, astrocytes and pericytes, and constitutes an anatomical, physicochemical and biochemical barrier for controlling exchange of various compounds between the blood, brain and cerebrospinal fluid (Maki, Hayakawa et al. 2013).

Cerebral blood vessels also interact directly with neuronal and glial cells through neurovascular coupling, which means the interaction between different cell types to adjust cerebral blood flow based on cerebral activity (Attwell, Buchan et al. 2010). Interestingly, this signaling is not directly connected to the negative feedback of decreasing oxygen or glucose concentrations in brain tissue, but instead shows a more sophisticated interplay between neurons, astrocytes and blood vessels that is based on synaptic transmission (Yang, Zhang et al. 2003, Takano, Tian et al. 2006, Lindauer, Leithner et al. 2010). Moreover, vascular endothelial cells sustain neighboring neurons via the neurovascular niche: endothelial cells guide developing axons, protect neurons from stress factors and support neural stem cells (Makita, Sucov et al. 2008, Guo, Kim et al. 2008, Ohab, Fleming et al. 2006). Importantly, considering human stroke that often affects the white matter, an analogous niche is found between the endothelium and oligodendrocyte lineage cells (Arai, Lo 2009).

2.4.1.1.1 TPA and neurovascular uncoupling after cerebral ischemia

Neurovascular coupling undergoes a dramatic change after ischemia-induced damage.

Increased BBB permeability is a common marker of brain damage that follows stroke, as are other neurological pathologies which lead to edema and hemorrhage that worsen clinical outcome after stroke (Zlokovic 2008). Numerous factors that increase BBB permeability have been identified. For example, hypoxia has been shown to affect capillary endothelial cell tight junction permeability through increased interleukin-1 (IL-1) beta and nitric oxide (NO) levels (Yamagata, Tagami et al. 2004). Ischemia may also alter signaling between endothelial cells and astrocytes and activation of toll like receptors (TLRs) in astrocytes leads release of matrix metalloproteinases (MMPs) that damage BBB intergrity (Min, Hong et al. 2015).

Furthermore, loss of BBB function may result from inflammatory reactions associated with stroke(Stamova, Xu et al. 2010).

Another common phenomenon linked to stroke-related impairments in cerebral circulation is the so-called no-reflow phenomenon (Ames, Wright et al. 1968). This event is characterized by a short period of hyperemia followed by hours of decreased blood flow to the ischemic area even though the occlusion in a cerebral artery is removed. This neurovascular uncoupling may also be one of the main reasons leading to unsuccessful reperfusion of ischemic tissue despite successful recanalization (Soares, Tong et al. 2010). It has been shown that pericytes have an important role in the long-lasting decrease in blood flow after ischemia (Yemisci, Gursoy-Ozdemir et al. 2009). The mechanism contributing to this is the lack of ATP leading to an imbalance in Ca²⁺, which causes contraction of pericytes and diminished vessel diameter. It is also possible that reactive oxygen and nitrogen species (ROS and RNS) impair the release of vasodilating compounds from endothelial cells (Zou, Leist et al. 1999, Sun, Druhan et al. 2008).

Thrombolysis with rtPA is usually the life-saving action in ischemic stroke patients. The mechanism of action of rtPA is based on enzymatic conversion of plasminogen into active plasmin, which in turn cleaves fibrin thereby leading to resolution of a thrombus. However, thrombolysis with rtPA is occasionally associated with serious complications such as cerebral edema and hemorrhage that are direct consequences of the enzymatic actions of rtPA.

Activation of MMP-9 is the most common side effect of rtPA, especially if the beginning of the treatment is delayed (Sumii, Lo 2002, Ning, Furie et al. 2006). In addition, nonspecific MMP inhibition has been reported to reduce the extent of rtPA induced hemorrhagic transformations (Sumii, Lo 2002). This data was further supported by the finding that MMP- 9 activation is induced by both exogenous and endogenous tPA (Tsuji, Aoki et al. 2005). TPA may also degrade the BBB directly through activation of signaling cascades involving platelet-derived growth factor and low density lipoprotein receptors (Yepes, Sandkvist et al.

2003, Su, Fredriksson et al. 2008). Finally, rtPA treatment-induced MMP level elevations are associated with a worse clinical outcome in stroke patients (Castellanos, Sobrino et al. 2007).

It is possible that rtPA treatment itself causes neurovascular uncoupling, as it can enter the brain parenchyma through the damaged BBB. This statement is based on the finding that both endogenous neuronal and rtPA can cleave the NR1 subunit of the NMDA receptor and increase NMDA receptor signaling (Nicole, Docagne et al. 2001). Furthermore, endogenous tPA has an important role in neurovascular coupling through increased neuronal nitric oxide synthase (nNOS) phosphorylation leading to reduced NO synthesis and vasodilatation (Park, Gallo et al. 2008). Therefore, altered cerebral artery reactivity may be caused by changes in NMDA receptor and vasodilatory function of NO that are caused by rtPA (Armstead, Cines et al. 2004, Armstead, Kiessling et al. 2011).

2.4.1.2 Oligodendrocytes

As described above, stroke in humans regularly affects the brain white matter in which oligodendrocytes are one of the most common cell types. Their main role in the central nervous system (CNS) is to produce a myelin sheath around axons to enforce conduction of electrical impulses. This myelination is connected to brain plasticity even in the adult brain (Scholz, Klein et al. 2009). It has been shown that the interaction between neurons and oligodendrocytes is not limited to myelin-based axonal support, but oligodendrocytes also participate in lactate shuttle to axons with astrocytes, thereby supporting neurons metabolically (Funfschilling, Supplie et al. 2012, Lee, Morrison et al. 2012). In addition, oligodendrocytes and neurons exhibit bidirectional trophic signaling: oligodendrocytes support neuronal growth through glial cell-line derived neurotrophic factor (GDNF), and neuronal signals control oligodendrocyte differentiation via the Notch pathway (Wang, Sdrulla et al. 1998, Wilkins, Majed et al. 2003).

Oligodendrocytes are highly sensitive to stroke-induced factors such as hypoxia, oxidative stress, trophic factor deprivation and excitotoxicity. These factors result in oligodendrocyte-

mediated loss of axonal support to neurons (Pantoni, Garcia et al. 1996). On the other hand, oligodendrocyte progenitors have the capability to remyelinate demyelinated axons in the adult brain and this process is initiated by pathological signaling (for example tumor necrosis factor alpha, TNFα) characteristic for stroke (Arnett, Mason et al. 2001, Gensert, Goldman 1997). Although it is well known that oligodendrocytes promote neuronal survival in normal conditions, little is known about their role in post-stroke neurogenesis. However, the role of the aforementioned oligovascular niche in post-stroke vascular remodeling is generally accepted and indicates that oligodendrocyte targeting drugs might improve both axonal myelination and angiogenesis (Pham, Hayakawa et al. 2012).

2.4.1.3 Astrocytes

Astrocytes are abundant cells in neural tissue, counting for up to approximately one half of the cells in brain. Classically they are known to participate in maintaining extracellular ionic balance, but recent research has revealed that they have other important functions in maintaining homeostasis in brain, as well as in practically all of the diseases affecting the CNS (Barres 2008). Each astrocyte can interact with various cell types and functions through thousands of cellular processes reaching synapses and blood vessels (Eroglu, Barres 2010, Iadecola, Nedergaard 2007).

Astrocytes have an important role in the arrangement and function of synapses (Barres 2008). Thrombospondins, cholesterol and glypicans 4 and 6 secreted by astrocytes can each have distinct effects on synaptic formation and pre- or postsynaptic function (Christopherson, Ullian et al. 2005, Mauch, Nagler et al. 2001, Allen, Bennett et al. 2012).

Astrocytes can also sense synaptic transmission with partly similar receptors as neurons have, and in turn modify both neuronal activity and blood flow based on synaptic activity (Eroglu, Barres 2010, Iadecola, Nedergaard 2007). In addition to exracellular signaling, gliovascular and neuroglial transferring of nutrients, metabolites, ions and messengers can also take place through gap junctions formed by the connexins (Chew, Johnson et al. 2010).

In this way the astrocyte-derived trophic factors regulate neurons, cells of the oligodendrocyte lineage, and endothelial progenitor cells, especially following an insult (Ricci, Volpi et al. 2009, Arai, Lo 2010, Hayakawa, Pham et al. 2012).

Astrocytes can maintain their metabolic activity for hours after stroke, past reperfusion, which makes them a key player during the ischemia-reperfusion damage to the brain (Thoren, Helps et al. 2005). During this period, astrocytes serve as the main source of antioxidant defenses to prevent neuronal death, and especially the loss of astrocytic glutathione (GSH) has been shown to be detrimental for neurons (Chen, Vartiainen et al.

2001). Astrocytes can also provide trophic support for neurons after stroke (Xin, Li et al.

2010). The beneficial role of astrocytes has also been demonstrated with a reactive astrocyte knock-down animal, which have larger stroke-induced lesions and more severe inflammation than wildtype animals (Myer, Gurkoff et al. 2006). However, the actions of astrocytes are not always beneficial for neural network integrity. Most importantly,

astrocytes can become over activated for example through increased secretion of pro- inflammatory cytokines by microglia (Rostworowski, Balasingam et al. 1997). The reason for the distinct effects of astrocyte activation may be the surprisingly vast heterogeneity of astrocytes, leading to different responses depending on the site and type of injury (Takata, Hirase 2008, Zamanian, Xu et al. 2012). Because it is generally recognized that astrocytes from different brain regions can be distinguished based on their morphology (protoplasmic, velate and fibrous) and molecular expression profiles (glial fibrillary acidic protein (GFAP), vimentin, glutamate transporters), it would be important to perform studies with similar damage applied to different brain regions to fully understand how differences in reactivity of astrocytes affect disease outcome (Ben Haim, David 2016).

One of the emerging concepts in CNS disorders is the glymphatic dysfunction. Normally, clearance of waste products is carried out by a system analogous to blood circulation consisting of cerebrospinal fluid influx into the brain parenchyma via para-arterial spaces and exchange of the products with interstitial fluids and clearance along para-venous spaces (Venkat, Chopp et al. 2016). Astrocytes are essential regulators in this system by regulating water movement in the brain by expressing Aquaporin-4 (AQP-4) in their endfeet near capillaries creating a para-vascular channel for cerebrospinal fluid flow (Iliff, Wang et al.

2012). During ischemic stroke, this system becomes impaired (Gaberel, Gakuba et al. 2014).

It has been shown that the impairment after transient MCAO is caused by loss of astrocyte polarization and dysclocalisation of AQP-4, leading to edema (Steiner, Enzmann et al. 2012).

Thus, medicines preserving AQP-4 localization might help to resolve edema after ischemia.

2.4.1.4 Microglia

In the human brain, microglial cells account for 0.5-16.6% of the total cell mass depending on the brain region assessed (Mittelbronn, Dietz et al. 2001). They are usually referred to as the resident immune cells of the CNS, being the first responders to CNS damage. Microglia originate from primitive myeloid progenitors located in the yolk sac during embryonic development, distinguishing them from bone-marrow-derived macrophages (Cuadros, Martin et al. 1993). Contrary to the aforementioned neuroglial cells (astrocytes and oligodendrocytes), microglia are not electrically coupled to the neural network with current sensing gap junctions. However, despite earlier assumptions, resting microglia are constantly forming and withdrawing processes to monitor their territory and quickly move in the direction of observed damage (Nimmerjahn, Kirchhoff et al. 2005). In addition to monitoring the immunological status of the CNS, microglia have an essential role in normal brain development through synaptic pruning and regulation of the neurogenic niche (Solano Fonseca, Mahesula et al. 2016, Paolicelli, Bolasco et al. 2011).

2.4.1.4.1 Phenotypes and stroke-induced activation of microglia

Even though overt activation of microglia associated with ischemic stroke is generally thought to be detrimental, studies with knock-out animals have demonstrated that complete inhibition of microglial activation leads to increased stroke-induced damage (Lalancette-