DISSERTATIONS | KAI HÄRKÖNEN | CD44 AND EXTRACELLULAR VESICLES | No 548
Dissertations in Health Sciences
PUBLICATIONS OF
THE UNIVERSITY OF EASTERN FINLAND
KAI HÄRKÖNEN
CD44 AND EXTRACELLULAR VESICLES
Studies on the role of CD44 in vesicle biogenesis, detection and cellular interactions
uef.fi
PUBLICATIONS OF
THE UNIVERSITY OF EASTERN FINLAND Dissertations in Health Sciences
ISBN 978-952-61-3286-0
Secretion of extracellular vesicles is an important mechanism in the intercellular communication in normal physiology and in the development of different pathological
conditions, such as cancer. Better understanding of interactions between EVs
and cells is an essential step towards the clinical use of EVs. This thesis gives new insights into the role of CD44 receptor and
hyaluronan in EV biology in cancer.
KAI HÄRKÖNEN
CD44 AND EXTRACELLULAR VESICLES
STUDIES ON THE ROLE OF CD44 IN VESICLE BIOGENESIS, DETECTION AND CELLULAR INTERACTIONS
Kai Härkönen
CD44 AND EXTRACELLULAR VESICLES
STUDIES ON THE ROLE OF CD44 IN VESICLE BIOGENESIS, DETECTION AND CELLULAR INTERACTIONS
To be presented by permission of the Faculty of Health Sciences, University of Eastern Finland for public examination in SN200 Auditorium, Kuopio on Saturday,
January 18th 2020, at 12 o’clock noon
Publications of the University of Eastern Finland Dissertations in Health Sciences
No 548
University of Eastern Finland Kuopio
2019
Series Editors
Professor Tomi Laitinen, M.D., Ph.D.
Institute of Clinical Medicine, Clinical Physiology and Nuclear Medicine Faculty of Health Sciences
Associate professor (Tenure Track) Tarja Kvist, 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
Professor Tarja Malm, Ph.D.
A.I. Virtanen Institute for Molecular Sciences Faculty of Health Sciences
Lecturer Veli-Pekka Ranta, Ph.D.
School of Pharmacy Faculty of Health Sciences
Distributor:
University of Eastern Finland Kuopio Campus Library
P.O.Box 1627 FI-70211 Kuopio, Finland
www.uef.fi/kirjasto
Grano Oy, 2019
ISBN: 978-952-61-3286-0 (print/nid.) ISBN: 978-952-61-3287-7 (PDF)
ISSNL: 1798-5706
Author’s address: Institute of Biomedicine/School of Medicine/Faculty of Health Sciences
University of Eastern Finland KUOPIO
FINLAND
Doctoral programme: Doctoral Program of Molecular Medicine Supervisors: Docent Kirsi Rilla, Ph.D.
Institute of Biomedicine/School of medicine University of Eastern Finland
KUOPIO FINLAND
Docent Arto Koistinen, Ph.D.
SIB Labs
University of Eastern Finland KUOPIO
FINLAND
Reviewers: Suneel Apte, Ph.D., MBBS.
Department of Biomedical Engineering Lerner Research Institute, Cleveland Clinic CLEVELAND
USA
Associate Professor Ralf Richter, Ph.D.
Faculty of Mathematics and Physical Sciences University of Leeds
LEEDS UNITED KINGDOM
Opponent: Rienk Nieuwland, Ph.D.
Laboratory of Experimental Clinical Chemistry Vesicle Observation Centre
University of Amsterdam AMSTERDAM
NETHERLANDS
Härkönen, Kai
CD44 and extracellular vesicles, Studies on the role of CD44 in vesicle biogenesis, detection and cellular interactions
Kuopio: University of Eastern Finland
Publications of the University of Eastern Finland Dissertations in Health Sciences 548. 2019, 107 p.
ISBN: 978-952-61-3286-0 (print) ISSNL: 1798-5706
ISSN: 1798-5706
ISBN: 978-952-61-3287-7 (PDF) ISSN: 1798-5714 (PDF)
ABSTRACT
Extracellular vesicles (EVs) are small lipid bilayer membrane bound vesicles, which are secreted to the extracellular space by all cell types. These vesicles were thought to be just the waste management system of the cells, when they were first detected in 1940s, but since then their significance as an important mechanism in intercellular communication has been revealed. Nowadays it seems obvious that these nanosized particles participate in a wide variety of physiological and pathological processes.
Hyaluronic acid, also called as hyaluronan (HA) is a glycosaminoglycan produced at the plasma membrane by three HA synthases HAS1-3. It is one of the major components of the extracellular matrix. Because of its negative charge HA is very hydrophilic molecule, which means that it can bind high amounts of water. In a solution HA forms a viscous gel. This property makes it an important space filler and lubricant in the body. However, HA has many other important functions in the body such as cell migration and proliferation. Importantly it has a role in cancer development.
CD44 receptor is a multifunctional transmembrane receptor, which has roles in regulating cell adhesion and migration. CD44 has an ability to interact with multiple ligands of which HA is the principal one. Even though CD44 is the product of a single gene, due to alternative splicing there are multiple isoforms of CD44. Expression of CD44 standard form (CD44s) and many of its variants is suggested to have a role in epithelial-to-mesenchymal transition, which is a key step in initiation of metastasis.
In addition, CD44 expression seems to correlate with poor prognosis in some cancers.
The aim of this thesis was to study how CD44 expression and HA influence EV secretion and to determine whether they are carried by EVs. CD44-HA interaction can affect EV-recipient cell communication in multiple ways. HA carried by EVs can contribute to EV binding on the cells which express CD44. On the other hand, CD44- positive vesicles may adhere to the HA-coat around the cells, respectively. CD44 regulated EV-recipient cell interaction is anything but an unambiguous process.
Previously it has been shown that overexpression of HA synthesis induces the secretion of EVs, and HA can be carried by EVs. Our results show that CD44 receptor is also carried by EVs. We have also detected that CD44 and EV-recipient cell interaction has a possible connection, since EVs given to recipient cells were shown to colocalize with areas on the plasma membrane with CD44. Interestingly CD44s expression itself in gastric cancer cells does not affect EV secretion. Epidermal growth factor (EGF) or wounding treatment of rat primary mesothelial cell cultures could induce epithelial-to-mesenchymal -like changes in cells. Treatments were also able to induce CD44 expression and EV secretion.
Based on our observations, CD44 has a role in EV-recipient cell interaction, but does not regulate the EV secretion rate. Our results also support the earlier findings on the role of CD44 on EMT and tumorigenic properties of cells. In addition, based on our study, CLEM serves as a powerful tool to perform EV-research, combining the best features of light and electron microscopy.
National Library of Medicine Classification: QU 55.7, QU 83, QU 375, QU 475, QZ 203
Medical Subject Headings: Hyaluronan Receptors; Hyaluronic Acid; Extracellular Vesicles;
Exosomes; Cell Communication; Gene Expression; Epithelial-Mesenchymal Transition;
Carcinogenesis; Biomarkers, Tumor; Neoplasms; Neoplasm Invasiveness; Cells, Cultured
Härkönen, Kai
CD44 ja solunulkoiset vesikkelit. Tutkimuksia CD44 -reseptorin roolista solunulkoisten vesikkelien erityksessä, havaitsemisessa sekä vuorovaikutuksesta solujen kanssa.
Kuopio: Itä-Suomen yliopisto
Publications of the University of Eastern Finland Dissertations in Health Sciences 548. 2019, 107 s.
ISBN: 978-952-61-3286-0 (nid.) ISSNL: 1798-5706
ISSN: 1798-5706
ISBN: 978-952-61-3287-7 (PDF) ISSN: 1798-5714 (PDF)
TIIVISTELMÄ
Solunulkoiset vesikkelit ovat pieniä kaksoislipidikalvon ympäröimiä rakkuloita, joita solut tuottavat solunulkoiseen tilaan. Kun solunulkoiset vesikkelit alun perin havaittiin 1940-luvulla, niiden ajateltiin osallistuvan ainoastaan solujen jätehuoltoon, mutta tähän päivään mennessä niiden on huomattu olevan tärkeä osa solujen välistä viestintää. Nykytiedon valossa on selvää, että nämä nanokokoiset rakkularakenteet osallistuvat elimistössä sekä useisiin fysiologisiin, että patologisiin prosesseihin.
Hyaluronihappo, toiselta nimeltään hyaluronaani (HA) on glykosaminoglykaani, jota hyaluronaanisyntaasit HAS1-3 tuottavat solukalvolla. Se on yksi tärkeimmistä soluväliaineen rakenneosista. HA on hydrofiilinen molekyyli, joka voi sitoa itseensä valtavan määrän vettä ja muodostaa vesiliuoksessa viskoosin geelin.
Hydrofiilisyytensä vuoksi HA toimii elimistössä tärkeänä täyte- sekä voiteluaineena.
Sillä on myös monia muita tehtäviä, kuten solujen jakaantumisen ja liikkumisen säätelyyn osallistuminen. Lisäksi HA metabolian on todetty olevan yhteydessä syövän kehitykseen.
CD44-reseptori on kalvoproteiini, jolla on useita tehtäviä, kuten solujen kiinnittymisen ja liikkumisen säätely. CD44-reseptorilla on useita ligandeja, joista HA on tärkein. Vaikka CD44 on yhden geenin tuote, vaihtoehtoisen silmukoinnin tuloksena se ilmenee useina eri muotoina. Useilla eri CD44-reseptorin muodoilla on ehdotettu olevan yhteys epiteeli-mesenkymaaliseen transititioon ja etäpesäkkeiden lähettämiseen.
Tämän väitöskirjatyön tarkoituksena oli tutkia CD44:n ja HA:n yhteyttä solunulkoisten vesikkelien eritykseen sekä niiden vesikkelivälitteistä eritystä soluista. CD44-HA -vuorovaikutus voi useilla tavoilla vaikuttaa solunulkoisten vesikkelien ja kohdesolujen väliseen kanssakäymiseen. Solunulkoisten vesikkelien mukanaan kuljettama HA voi sitoutua solujen pinnalla olevaan CD44:aan tai CD44 solunulkoisen vesikkelin pinnalla voi edesauttaa sitoutumisessa solun pinnalla olevaan HA-kerrokseen.
Aiemmissa tutkimuksissa on todettu, että HA:n yliekspressio johtaa solunulkoisten vesikkelien erityksen lisääntymiseen. Lisäksi on todettu, että HA:ta esiintyy solunulkoisten vesikkelien pinnalla. Tutkimuksissamme näimme, että myös CD44:aa esiintyy solunulkoisten vesikkelien pinnalla. Totesimme myös, että CD44 - reseptorilla on mahdollisesti yhteys solunulkoisten vesikkelien sitoutumisessa kohdesoluihin. CD44-reseptorin ilmentäminen ei kuitenkaan vaikuta suolistosyöpäsolujen vesikkelierityksen aktiivisuuteen. Rotan mesoteelisolujen käsittely epidermaalisella kasvutekijällä (EGF) sekä soluviljelmälle suoritettu haavakoe saivat aikaan solulla epiteeli-mesenkymaaliseen transitioon viittaavia piirteitä. CD44:n ilmentäminen sekä solunulkoisten vesikkelien erittäminen kasvoivat kokeen seurauksena.
Yhteenvetona voidaan sanoa, että CD44-reseptorilla on rooli solunulkoisten vesikkelien ja solujen välisessä vuorovaikutuksessa, mutta se ei vaikuta solunulkoisten vesikkelien eritykseen. Tuloksemme tukevat aiempia havaintoja CD44-reseptorin osuudesta epiteeli-mesenkymaaliseen transitioon sekä solujen muuntumiseen syöpäsolujen kaltaisiksi. Lisäksi voidaan todeta korrelatiivisen mikroskopian toimivan tehokkaana työkaluna solunulkoisten vesikkelien tutkimisessa, sillä se yhdistää sekä valo- että elektronimikroskopian parhaat puolet.
Luokitus: QU 55.7, QU 83, QU 375, QU 475, QZ 203
Yleinen suomalainen ontologia: hyaluronaani; solut; eritteet; vuorovaikutus; soluviestintä;
geeniekspressio; markkerit; syöpätaudit; syöpäsolut; soluviljely
Tästä tulee hyvä päivä.
ACKNOWLEDGEMENTS
This doctoral thesis was carried out in the Institute of Biomedicine, School of Medi- cine at the University of Eastern Finland during the years 2015-2019. The end of such an era has come, and finally it is the moment to breathe and look to the past. It is time to thank all those lovely people who have supported me through these years.
Special thanks go to by principal supervisor Docent Kirsi Rilla. It has been a pleasure to work with you all these years. With this a relatively new and challenging research topic you were a supportive cornerstone and your office door were always open. You were a great boss with your scientific, down to earth personality combined to your great sense of humor. I will be eternally grateful to you for this opportunity to work as a part of your research group!
Support from my second supervisor, Docent Arto Koistinen, has been priceless and I am grateful to you for all the scientific and practical support, especially with elec- tron microscopy. You are a warm-hearted hard-core physicist who gave valuable perspectives to our biological questions. The importance of cooperation between dif- ferent science fields cannot be overemphasized.
Warm thanks to belongs to Doctor Ville Koistinen and Uma Arasu M.Sc., who made this dissertation possible. It was great to have you as my colleaques and I wish you all the best.
In these times of change, when the academic world in Finland has become more and more unstable, and constant uncertainty about continuity inevitably casts its shadow to the mind of a young researcher, it has been awesome to see that it is really possible to create a lifelong career in this fascinating field of science. It has been a privilege to begin my own scientific path with our lovely senior scientists Professor Emeritus Markku Tammi and Professor Emerita Raija Tammi. In addition, it should be men- tioned that Markku has been at least as tough player in the floorball field as he is in the science.
Many thanks belong to the official pre-examiners Dr. Suneel Apte and Dr. Ralf Rich- ter. I really appreciate all the valuable comments from you which helped me to im- prove this thesis!
I am deeply grateful for Docent and head of the Institute of Biomedicine Anitta Ma- honen and professor Mikko Hiltunen, the director of the Doctoral program in Molec- ular Medicine, for the possibility to make my doctoral studies here in the Institute of Biomedicine!
Sometimes you must admit that you shouldn’t trust the first impression. This hap- pened a couple of times during my doctoral studies. The biggest surprise was when I met Tommi Paakkonen. Even though my first thought of you, when you came to our office, was that this guy is definitely so boring that he will only play with his Excel files and never going to say a word, you became a close friend to me pretty fast.
I can’t highlight enough the meaning you as a friend and collegue. Another misinter- pretation happened with Johanna Matilainen, “the girl who never smiles” or that was what I thought when you appeared to our research group. A warm and cheerful per- sonality was revealed from behind this “stone-faced” first impression. I really appre- ciate all the scientific and not so scientific discussions and moments with you during these years.
People come, people go, but memories never die. It has been a privilege to spend these years in the middle of many magnificent persons and feel great togetherness.
For all these years I want to thel you my warmest thanks Raquel Melero, Piia Takabe, Kirsi Kainulainen, Janne Capra, Ashik Deen, Sanna Pasonen-Seppänen, Sanna Oi- kari, Leena Rauhala, Virpi Tiitu, Kari Törrönen, Lasse Hämäläinen, Kirsi Ketola, Pet- teri Nieminen, Sanjeev Ranjan, Sini Hakkola, Heikki Kyykallio and Hannamari Luk- kari.
Any laboratory wouldn’t work for sure without professional staff. I want to deeply thank Riikka Kärnä, Silja Pyysalo, Eija Rahunen, Taija Hukkanen, Kari Kotikumpu and Virpi Miettinen for all your help during my doctoral studies not to mention all the wonderful moments we have experienced together.
I am also thankful to our great secretaries who have kept this academic ship sailing on the course. Thank you for all the precious help Eija Vartiainen, Karoliina Tenka- nen, Tellervo Rajanto, Leena Saario, Marjut Nenonen, Teemu Laitinen and Arja Af- flekt. I really hope that I didn’t cause too much gray hairs for any of you!
I also want to thank the rest of my co-authors Otto Jokelainen, Reijo Sironen, Jaana Hartikainen, Carla Oliveira, George Daaboul and Andrew Malloy. I am grateful for the technical assistance provided by Helena Kemiläinen, Aija Kekkonen and Joonas Malinen. I really appreciate your contribution to the publications I am privileged to use in my thesis.
In addition to my dear colleagues, I am fortunate to have a group of magnificent
Finally, I would like to thank my family members who have always supported me through my life in many ways. I am grateful to my mother and father, Päivi and Timo, my real-life role models, who have always quaranteed a safe environment to grow. I also want to express great thanks my two big brothers Aki and Tomi for your brotherly guidance in my life.
This work was financially supported by Academy of Finland, University of Eastern Finland, Spearhead Funds from the University of Eastern Finland (Cancer center of Eastern Finland), Jane ja Aatos Erkko Foundation, Sigrid Juselius Foundation, North- ern Savo Cancer Foundation, Paavo Koistinen Foundation, Otto A. Malm foundation, the Centre for International Mobility. Research has also been supported by EATRIS, the European Infrastructure for Translational Medicine. I also want to thank for the opportunity to use facilities of the SIB labs and Biocenter Oulu Electron Core facility.
Kuopio, December 2019
Kai Härkönen
LIST OF ORIGINAL PUBLICATIONS
This dissertation is based on the following original publications:
I Härkönen K, Oikari S, Kyykallio H, Capra J, Hakkola S, Ketola K, Arasu UT, Daaboul G, Malloy A, Oliveira C, Jokelainen O, Sironen R, Hartikainen J, Rilla K CD44s assembles hyaluronan coat on filopodia and extracellular vesicles and in- duces tumorigenicity of MKN74 gastric carcinoma cells.
Cells 22;8(3), 2019.
II Arasu UT*, Härkönen K*, Koistinen A, Rilla K
Correlative light and electron microscopy is a powerful tool to study interac- tions of extracellular vesicles with recipient cells.
Exp. Cell Res. 376: 149-158, 2019.
III Koistinen V, Härkönen K, Kärnä R, Arasu UT, Oikari S, Rilla K
EMT induced by EGF and wounding activates hyaluronan synthesis machin- ery and EV shedding in rat primary mesothelial cells.
Matrix Biol. 63: 38-54, 2017
* Equally contributed.
The publications were adapted with the permission of the copyright owners.
CONTENTS
ABSTRACT ... 7
TIIVISTELMÄ ... 9
ACKNOWLEDGEMENTS ...11
1 INTRODUCTION ...27
2 REVIEW OF THE LITERATURE ...29
2.1 What are extracellular vesicles? ...29
2.1.1 A brief history of EVs ...29
2.1.2 Different types of EVs ...30
2.1.3 Nomenclature of extracellular vesicles ...31
2.2 Cargo of EVs ...32
2.2.1 EVs ship biomolecules between cells ...32
2.2.2 Lipids ...32
2.2.3 RNA ...33
2.2.4 DNA ...34
2.2.5 Proteins ...35
2.2.6 Sugars ...36
2.3 Biogenesis of exosomes ...36
2.3.1 Exosomes are formed via endosomal route ...36
2.3.2 ESCRT-dependent route ...37
2.3.3 ESCRT-independent route(s) ...37
2.4 Secretion of exosomes ...38
2.5 Biogenesis and secretion of microvesicles ...39
2.6 EV-mediated interactions ...39
2.6.1 EV-uptake ...39
2.7 Extracellular vesicles as biomarkers ...41
2.7.1 A new minimally invasive liquid biopsy ...41
2.7.2 Exosomal proteins as biomarkers ...42
2.7.2.1Lung cancer ...42
2.7.2.2Breast cancer ...42
2.7.2.3Prostate and other cancers ...43
2.7.3 Exosomal-RNAs as biomarkers ...43
2.7.3.1Lung cancer ...43
2.7.3.2Breast cancer ...44
2.7.3.3Prostate and other cancers ...44
2.8 Hyaluronan ...44
2.8.1 Structure and properties ...44
2.8.2 Production of HA ...45
2.8.3 Degradation of HA ...45
2.8.4 Physiological role of HA ...46
2.9 HA-receptors and HA-mediated signaling ...47
2.9.1 CD44 – Cluster of differentiation 44 ... 47 2.9.1.1CD44 in epithelial to mesenchymal transition (EMT) and cancer ... 49 2.9.2 RHAMM – A receptor for hyaluronan mediated motility ... 52 2.10Isolation methods of extracellular vesicles ... 53 2.10.1Ultracentrifugation/Differential centrifugation ... 53 2.10.2Size exclusion chromatography ... 53 2.10.3Ultrafiltration ... 53 2.10.4Immuno-affinity purification ... 54 2.10.5Precipitation ... 54 2.11Methods for characterization of evs ... 54 2.11.1Confocal microscopy ... 54 2.11.2Transmission electron microscopy ... 55 2.11.3Scanning electron microscopy ... 55 2.11.4Cryoelectron micropscopy... 55 2.11.5Correlative light and electron microscopy ... 55 2.11.6Atomic force microscopy ... 56 2.11.7Nanoparticle tracking analysis ... 56 2.11.8Dynamic light scattering ... 56 2.11.9Tunable resistive pulse sensing ... 56 3 AIMS OF THE STUDY ... 59 4 MATERIALS AND METHODS ... 61 4.1 Materials ... 61 4.1.1 Cell lines ans cell cultures ... 61 4.1.2 QPCR primers ... 62 4.1.3 Methods ... 63 5 RESULTS ... 67
5.1 CD44s assembles hyaluronan coat on filopodia and extracellular vesicles and induces tumorigenicity of MKN74 gastric carcinoma cells (I) ... 67 5.1.1 CD44 has a role in regulation of HA metabolism ... 67 5.1.2 Overexpressed CD44s is accumulated at membranes of filopodia and
assembles a hyaluronan coat ... 67 5.1.3 CD44s receptor is transported to EVs but it does not change the EV
secretion ... 68 5.1.4 CD44s induces growth and invasion of MKN74 cells in 3D collagen
matrices. ... 68 5.1.5 HA content of chick chorioallantoic membrane tumors is increased
due to CD44s expression ... 68 5.2 Correlative light and electron microscopy is a powerful tool to study
interactions of extracellular vesicles with recipient cells (II) ... 69 5.2.1 Combining the confocal imaging and scanning electron microscopy
5.3 EMT induced by EGF and wounding activates hyaluronan synthesis
machinery and EV shedding in rat primary mesothelial cells (III) ...71 5.3.1 EMT is induced by EGF and wounding in mesothelial cells ...71 5.3.2 EGF and wounding lead to CD44 overexpression and increased
hyaluronan synthesis in rat primary mesothelial cells ...71 5.3.3 EGF and wounding cause HAS2 overexpression in mesothelial cell
cultures...72 5.3.4 Secretion of EVs is enhanced due to EGF and wounding in rat
primary mesothelial cells ...72 6 DISCUSSION ...73 6.1 The role of CD44 in EMT and cancer progression...73
6.1.1 EGF and wounding possibly induce EMT via upregulated CD44 expression ...73 6.1.2 CD44 expression alters HA metabolism and increases cell associated
HA ...74 6.1.3 The role of CD44 in association with tumorigenic properties is
conflicting ...76 6.2 EVs carry CD44 and HA and possibly mediate interactions between cells .76 6.3 Correlative light and electron microscopy is an efficient tool to study EVs ..78 7 CONCLUSIONS ...81 REFERENCES ...83 ORIGINAL PUBLICATIONS (I – III) ...109
ABBREVIATIONS
AARDC1 Arresting containing protein 1
A.t.-DNA Arabidopsis thaliana DNA
CAV1 Caveolin 1
CD44 Cluster of differentiation 44
CDE Caveolin-
dependent endocytosis
CLEM Correlative light and electron microscopy CME Clathrin-mediated
endocytosis
CEMIP Cell migration-
inducing and hyaluronan-
binding protein circRNA Circular RNA Cryo-EM Cryoelectron
microscopy
DLS Dynamic
light scattering dsDNA Double-stranded-
DNA
ECM Extracellular matrix
EGF Epidermal growth
factor
ERK Extracellular
signal-regulated kinase
ESCRT Endosomal sorting complex required for transport
ESPR1 Epithelial splicing regulatory protein 1
EM Electron
microscopy
EMT Epithelial-to-
mesenchymal transition
EV Extracellular
vesicle
GFP Green fluorescent
protein
Gluc Gaussia
luciferasetosis GlcUA Glucuronic acid GlcNac N-acetyl
glucosamine
GPI Glycosyl-
phosphatidylinosit ol
HA Hyaluronic
acid/Hyaluronan HARE The hyaluronan re-
ceptor for endo- cytosis
HAS Hyaluronan
synthase
HBMVEC Human brain
microvascular endothelial cell
HMW High molecular weight
hnRNPA2B1 Heterogenous nuclear ribonucleo- protein A2B1 HSP Heat shock protein
HYAL Hyaluronidase
ILV Intraluminal vesicle ISEV International
Society of Extra- cellular vesicles LBPA Lysobiphosphatidic
acid
LMW Low molecular
weight
lncRNA Long noncoding
RNA
LYVE-1 Lymphatic vessel
endothelial hyaluronan receptor
MHC Major
histocompatibility complex
MLCK Myosin light chain kinase
mRNA Messenger RNA
miRNA Micro RNA
MVB Multivesicular
body
NSCLC Non-small cell lung
PEG Polyethylene glycol PHYAL1 Hyaluronidase
pseudogene PI(3)P Phosphatidyl 3-phos-
phate
piRNA Piwi-interacting RNA PS Phosphatidylserine RHAMM Receptor for hyalu-
ronan-mediated motil- ity
rRNA Ribosomal RNA SEC Size exclusion chro-
matography
SEM Scanning electron mi- croscopy
SNARE N-ethylmaleimide- sensitive factor attach- ment protein receptor snoRNA Small nuclear RNA snRNA Small nuclear RNA SPAM1 Sperm adhesion mole- cule 1
TAT5 Phosphatidylenol- amine-translocase TEM Transmission electron
microscopy
TMEM2 Transmembrane pro- tein 2
TRPS Tunable resistive pulse sensing
UDP Uridine diphosphate UF Ultrafiltration
1 INTRODUCTION
Extracellular vesicles are tiny membrane bubbles, which cells secrete into the extra- cellular space (El Andaloussi et al., 2013). Secretion of EVs is an universal process, which occurs also in eukaryotes, such as bacteria, archaea and plants (Deatherage and Cookson, 2012; Lotvall et al., 2014; Lawson et al., 2017; Rome, 2019). Even though the first observation of EVs was made already in 1946 by Erwin Chargaff and Ran- dolph West (Chargaff and West, 1946), the importance of EVs has truly been realized over the last two decades. EVs are important key players in cell-cell communication, and they have been shown to carry wide variety of biomolecules such as RNA (Valadi et al., 2007), DNA (Balaj et al., 2011), proteins (Thery, Zitvogel and Amigorena, 2002) and lipids (Llorente et al., 2013) with them. Intercellular communication is an important part of maintaining homeostasis of tissues, but also has a key role in the development of different pathological conditions such as cancer. Inside the lipid bi- layer membrane of EVs cargo is protected from degradation in extracellular space and can travel long distances in the body. EVs have been shown to successfully de- liver functional biomolecules to target cells, and thereby modify their behavior (Dourado et al., 2019).
EV is an umbrella term for different types of cell secreted vesicles. The two main types of EVs are exosomes and microvesicles (Heijnen et al., 1999). The biogenesis of exosomes occurs via the endocytic route. During maturation of endosomes, as a re- sult of inward budding of endosomal membrane, intraluminal vesicles (ILV) are formed into the lumen of an endosome (Harding et al.; B. T. Pan et al.; Johnstone, Adam, et al.). After formation of ILVs, the endosome is called a multivesicular body (MVB) (Johnstone et al., 1987). The MVB may fuse with a lysosome, when its cargo is degraded, or with the plasma membrane, when ILVs are released into the extracel- lular space (Futter et al., 1996). In the extracellular space ILVs are termed as exosomes (Trams et al.; Harding et al.). Microvesicles, in turn, are formed via direct budding from the plasma membrane (Heijnen et al., 1999).
CD44 is a multifunctional class 1 transmembrane glycoprotein cell surface adhe- sion molecule (Underhill et al., 1987). CD44 receptor is encoded by a single gene, but as a result of alternative splicing multiple variant isoforms exist (Gao et al., 1997).
There are multiple ligands for CD44, such as hyaluronan (HA), osteopontin, colla- gens and matrix metalloproteinases. HA is the main ligand of CD44, and all the isoforms have a HA-binding region (Goodison, Urquidi and Tarin, 1999). CD44 has an important role in cancer progression because it can promote invasion, intravasa- tion and metastasis of tumor cells (McFarlane et al., 2015; Senbanjo and Chellaiah, 2017). CD44 overexpression has also shown to be a mark of stemness of cancer cells (Yan, Zuo and Wei, 2015).
HA is a simplest glycosaminoglycan, which consist of repeating units of D-glucu- ronic acid and N-acetyl-D-glucosamine linked together by glycosidic bonds
(Weissmann and Meyer, 1954; Weissmann et al., 1954). HA is produced by three hy- aluronan synthases (HASes) on the plasma membrane (Lee and Spicer, 2000). HA has high water binding capacity and it forms a viscous gel with water, which makes it an important space filler and lubricant (Valachova et al., 2016). HA has also many other important functions in the body. High molecular weight HA has been associated with anti-angiogenic and immunosuppressive functions (Tian et al., 2013) while low molecular weight HA fragments seem to have connection to inflammation (Termeer et al., 2000; D’Agostino et al., 2017). HA-signaling occurs via multiple receptors which are, in addition to CD44, RHAMM (Misra et al., 2015), LYVE-1 (Banerji et al., 1999), layilin (Bono et al., 2001), HARE (Zhou et al., 2000) and TLR-4 (Termeer et al., 2002).
Epithelial-to-mesenchymal transition (EMT) refers to an event where epithelial cells gain migratory and invasive properties, and change towards mesenchymal state (Greenburg and Hay, 1982). EMT has been proposed to have an important role in the initiation of metastasis, which is a process where tumor cells travel from original tu- mor site to other locations in a body (Gaianigo, Melisi and Carbone, 2017). Different isoforms of CD44 have been shown to have a role in EMT even though there are still a lot of open questions.
EVs are nanosized particles of which majority are smaller than resolution limit of light. This means that light microscopy, such as confocal microscopy, alone has lim- itations to detect EVs. Electron microscopy (EM) can overcome the problem of reso- lution limit of light and nanometer resolution can be obtained. EM, in turn, has limi- tation in protein labeling compared to light microscopy. Correlative light and elec- tron microscopy (CLEM) combines the strengths of both light microscopy and EM. It is a technique where the same specimen is first imaged with light microscopy and then with EM.
In this work we show that CLEM offers an efficient tool to study EV-cell interac- tions. This thesis also sheds light on the role of CD44 in EV secretion and EV-cell interactions. CD44 in addition to HA is carried by EVs, and it can regulate the inter- action between EVs and cells. Our results also support earlier findings that CD44 has a role in EMT.
2 REVIEW OF THE LITERATURE
2.1 WHAT ARE EXTRACELLULAR VESICLES?
2.1.1 A brief history of EVs
The first EV-related observation was done by Erwin Chargaff and Randolph West in 1946 while studying blood coagulation properties. They found out that the high- speed centrifugation of platelet free blood plasma at 31,000 x g increased the blood clotting time significantly compared to low-speed (5000 x g) centrifugation (Chargaff and West, 1946). It has been assumed that they managed to remove EVs from plasma, which has been at least a partial reason for altered clotting properties of blood plasma.
A couple of decades later, in 1967, a British physicist Peter Wolf managed to image these lipid-rich particles that were derived from platelets, which he called “platelet dust” (Wolf, 1967). In 1981 Dvorak and colleagues showed that vesicles with pro- coagulant activity (PCA) were shed from tumor cells (Dvorak et al., 1981). Another cornerstone in the history of EV research was the moment when Mary Johnstone’s research group demonstrated the existence of exosomes. They found out in electron micrograph study performed with reticulocytes that after endocytosis of transferrin receptor from plasma membrane the receptors were internalized in small ~ 50 nm buds into the endosome. These small vesicles were released into the extracellular space in the process of endosome fusion with the plasma membrane (Johnstone et al., 1987).
Despite the relatively slow awakening of the scientific community, since the first detection, these lipid bilayer enclosed particles have been detected in all body fluids such as amniotic fluid (Keller et al., 2007), ascites fluid (Andre et al., 2002), bile (Masyuk et al., 2010), blood (Caby et al., 2005), breast milk (Admyre et al., 2007), bron- choalveolar lavage (Admyre et al., 2003), cerebrospinal fluid (Harrington et al., 2009), nasal secretions (Lasser et al., 2011), saliva (Ogawa et al., 2008), synovial fluid (Mustonen et al., 2016), semen (Ronquist and Brody, 1985) and urine (Pisitkun, Shen and Knepper, 2004).
Awareness of the functions of extracellular vesicles has drastically increased dur- ing the last two decades. Interestingly, in the light of present knowledge, EVs seem to be important key players in cell signaling even though they were thought to func- tion only as a cell waste management system. Shedding of EVs seems to be universal process, which occurs in addition to eukaryotic cells also among the bacteria, archaea and plants (Deatherage and Cookson, 2012; Lotvall et al., 2014; Lawson et al., 2017;
Rome, 2019).
2.1.2 Different types of EVs
Extracellular vesicles (EV) are heterogenous, spherical, membrane bound particles secreted by virtually all cells into the extracellular space. The diameter of EVs varies from ̴ 30 nm to several micrometers. (El Andaloussi et al., 2013; Mathieu et al., 2019).
The structure of the lipid bilayer membrane surrounding EVs is similar to that of the cell membrane including the orientation of the lipid leaflets (Thery, Ostrowski and Segura, 2009). The morphology is variable and at least spherical and tubular EVs have been detected (Arraud et al., 2014). In transmission electron microscopic images EVs are seen as cup-shaped structures because of the collapsing due to dehydration during the sample processing (Raposo and Stoorvogel, 2013). The current classifica- tion based on their biogenesis divides EVs into the several subgroups as exosomes, microvesicles, and apoptotic bodies (Lasser, Jang and Lotvall, 2018). In addition to these, a term “oncosomes” is in use for special form of cancer-derived EVs (Al- Nedawi et al., 2008; Minciacchi, Freeman and Di Vizio, 2015). In this thesis the focus will be in exosomes and microvesicles.
Exosomes are generated through the endocytic route. Endocytic vesicles fuse with the endosome and ubiquitinated transmembrane proteins are internalized into the endosome as intraluminal vesicles (ILVs) (Harding et al.; B. T. Pan et al.; Johnstone, Adam, et al.). Multivesicular endosomes may end up to lysosomal degradation (Futter et al., 1996) or to fuse with the plasma membrane, when ILVs are released to extracellular space (Figure 1) (Johnstone et al., 1991). After entering the extracellular region ILVs are called exosomes (Johnstone et al., 1987). Exosomes are the smallest EVs.
Microvesicles, also called as ectosomes, microparticles and exovesicles in some contexts, are secreted directly from the plasma membrane (Figure 1) (Heijnen et al., 1999; Cocucci and Meldolesi, 2015). The shedding of microvesicles seems to be result of activation of the cells by various stimuli, which leads to increased calcium levels in cells. As exosomes, microvesicles can also contain cytoplasmic components of the cell and carry receptors on their surface (VanWijk et al., 2003).
Third group of EVs, apoptotic bodies, are released during programmed cell death, apoptosis (Figure 1). They are the largest type of EVs, and their diameter can be sev- eral micrometers. The formation of apoptotic bodies is a tightly regulated process.
Apoptotic bodies could easily be thought to be just cellular trash, but they are recog- nized to have roles in immune reactions, infection and cancer. As other EVs also apoptotic bodies can transfer biomolecules between cells (Atkin-Smith and Poon, 2017; Caruso and Poon, 2018).
Figure 1. Cells secrete different types of extracellular vesicles. Exosomes are produced via the endocytic route and released extracellularly when multivesicular body fuses with the plasma membrane. Microvesicles are formed via direct budding from the plasma membrane.
Apoptotic bodies are formed when cells undergo reprogrammed cell death, apoptosis. EVs mediate cell-cell communication. Interaction between EVs and target cells can occur via endocytosis of EVs, fusion of EVs with plasma membrane or via ligand-receptor interaction.
2.1.3 Nomenclature of extracellular vesicles
In a relatively short period of time, during which EVs have been actively investi- gated, researchers have given them a tremendous number of names. Various names are based on the way of biogenesis, the cellular or tissue origin, or the biological func- tion of particular type of EV (Gould and Raposo, 2013; van der Pol et al., 2016).
Currently the nomenclature of EVs has been standardized as a result of the work of International Society of Extracellular Vesicles (ISEV). Classification of EVs based on their biogenesis is in common use today. On the basis of biogenesis, EVs are clas- sified to three subgroups: Exosomes (Trams et al., 1981; Johnstone et al., 1987), micro- vesicles (Holme et al., 1994) and apoptotic bodies (Kerr, Wyllie and Currie, 1972).
Exosomes are the smallest in diameter (30-120 nm) and they are secreted via the en- dosomal pathway. Microvesicles, which are about 50-1000 nm in diameter, are bud- ded directly from the plasma membrane. Apoptotic bodies, in turn, are the biggest
particles (500-2000 nm) and they are produced during the process of cell apoptosis (El Andaloussi et al., 2013).
The multitude of names, the most of which are lead from the origin, biological functions, or the properties of the particles, include for example argosomes (Greco, Hannus and Eaton, 2001), prostasomes (Brody, Ronquist and Gottfries, 1983), promi- nosomes (Florek et al., 2007), oncosomes (Morello et al., 2013), cardiosomes (Waldenstrom et al., 2012), tolerosomes (Karlsson et al., 2001), t-exosomes (Xiang et al., 2009), ectosomes (Stein and Luzio, 1991), deteriosomes (Yao et al., 1993), micro- particles (Mackman, 2009), nanoparticles, sebosomes (Nagai et al., 2005), blebbing vesicles, budding vesicles, shedding vesicles, matrix vesicles and outer membrane vesicles (Srisatjaluk, Doyle and Justus, 1999).
However, after EVs have been secreted to the extracellular space, the separation of different subtypes is difficult or even impossible with the current methods. There- fore, the use of an umbrella term extracellular vesicles is coming more and more com- mon. It should also be mentioned that the terminology used to define extracellular vesicles may in some contexts refer also to completely different particles. For exam- ple, the term exosome is also used to designate an RNA-processing complex (Januszyk and Lima, 2014). Micro- and nanoparticles are general terms to describe any small particles, such as artificial polymers, in that size range.
2.2 CARGO OF EVS
2.2.1 EVs ship biomolecules between cells
In the literature EVs have been reported to be able to carry a variable assortment of biomolecules within them. In addition to lipids EVs have been shown to carry pro- teins, different forms of RNA, DNA and sugar molecules. Vesiclepedia database (www.microvesicles.org) gathers molecular research data of EVs. At the moment Vesiclepedia contains data of 1254 studies performed with 41 species and has 349,988 protein entries, 27,646 mRNA entries, 10,520 miRNA entries and 639 lipid entries.
During the maturation of early endosome cytosolic proteins, lipids and nucleic acids are packed into intraluminal vesicles (ILVs) that constitute the contents of exo- somes. During microvesicle budding, plasma membrane lipids and proteins as well as cytosolic molecules end up as the building material of the vesicle formed. It has been suggested that EV formation is a selective process, where certain molecules are accumulated into EVs, resulting in different composition of EVs compared to the cell of origin.
According to Llorente et al. exosomes seem to have also increased proportion of cho- lesterol (Llorente et al., 2013). It has also been described that exosomes have a higher amount of phosphatidylserine in the outer leaflet than the plasma membrane of orig- inal cell (Fitzner et al., 2011). In turn exosomes have less phosphatidylcholine and diacylglycerol than the plasma membrane (Laulagnier et al., 2004).
Exosome membranes are more rigid than the cell membrane because of higher amounts of sphingomyelin and desaturated lipids. Different studies have shown that exosome membrane rigidity may be pH dependent. The biogenesis of exosomes via MVBs with lower pH could explain the finding that EVs from basophils are less rigid in acidic solution (Laulagnier et al., 2004). On the other hand, melanoma cells have been shown to produce more rigid EVs when cultured in acidic conditions. It has been shown that in the tumor microenvironment, where pH is lower, the uptake of EVs is higher (Parolini et al., 2009). In low pH EVs fuse with plasma membrane easily because their fluidity is then comparable (Laulagnier et al., 2004).
In the study of Bicalho et al. authors detected highly similar lipid structure be- tween banked red blood cells and microvesicles. The most notable difference be- tween the membrane of microvesicles and the cell membrane is the higher proportion of polyunsaturated glycerophosphoserine (38:4) in microvesicles (Bicalho, Holovati and Acker, 2013). Llorente et al. noticed that EVs from PC3 prostate cancer had more ceramide and phosphatitidic acid compared to cells of origin (Llorente et al., 2013).
The exchange of lipids between inner and outer leaflets occur more often in EV mem- branes than in the cell membrane (Laulagnier et al., 2004).
2.2.3 RNA
EVs has been reported to carry multiple classes of RNAs and transfer them to recip- ient cells from the cells of origin (Valadi et al., 2007). RNA transferred between cells via EVs is called EV-RNA. Most of the EV-RNA seems to be shorter in size than av- erage size of RNA in cells (X. Huang et al., 2013; Eirin et al., 2014).
Both coding and noncoding RNAs exists among the types of RNA carried by EVs.
These different types of RNA include messenger RNAs (mRNAs), transfer RNAs (tRNAs), long noncoding RNAs (lncRNAs), microRNAs (miRNAs), circular RNAs (circRNAs), small nucleolar RNAs (snoRNAs), small nuclear RNAs (snRNAs), ribo- somal RNAs (rRNAs) and piwi-interacting RNAs (piRNAs) (Valadi et al., 2007).
Several attempts have been made to analyse the RNA content of EVs. In a micro- array analysis with EVs from glioblastoma cells researchers revealed 27,000 mRNAs of which 4700 mRNAs were detected only in EVs. There were also 1,118 mRNAs shown to be mainly excluded from EVs although they were present in parent cells (Skog et al., 2008). In another microarray study performed with EVs from MC/9 mouse mast cell line 270 mRNAs of total 1,300 detected in EVs were not seen in pa- rental cells (Valadi et al., 2007).
Several studies have shown that mRNAs from parent cells are translated in recip- ient cells. Research group of Tannous et al. managed to show that Gaussia luciferase
(Gluc) mRNAs were transferred via EVs from glioblastoma cells to human brain mi- crovascular endothelial cells (HBMVECs). They saw that fluorescence coming from luciferase protein encoded by transferred Gluc mRNA increased over 24 hours, which indicates that Gluc mRNA was translated successfully (Tannous et al., 2005).
Valadi et al in turn managed to show that EV-carried mRNAs from MC/9 cells were successfully translated in rabbit reticulocyte lysate which they used as an in vitro translation system Transfer of cre recombinase RNA has also been detected in vivo from hematopoietic cells to Purkinje cells in mice. The mechanism of RNA transfer was shown in the further studies to be mediated by EVs. The findings of that study also suggest that recombination events rarely occur in healthy animals but are more likely to be seen in injury models. The presence of inflammation raises the likelihood of recombination events (Ridder et al., 2014).
MicroRNAs are a type of noncoding RNAs, which have been shown to mediate post-transcriptional silencing of gene expression. The size of microRNAs is small, about 22-nt. MicroRNAs delivered via EVs can be used to regulate gene expression in the recipient cells (Iftikhar and Carney, 2016). It has also been assumed that noncoding miRNAs could regulate translation in recipient cells (Valadi et al., 2007).
Long noncoding RNA (lncRNA) is another type of noncoding RNAs. It is longer than microRNA, >200 nt long. LncRNAs have multiple roles in controlling chromatin or- ganization, gene transcription, mRNA turnover and protein translation. Also lncRNAs have been observed in EVs in multiple studies (Martianov et al., 2007; Kino et al., 2010; Ng, Johnson and Stanton, 2012). Another type of noncoding RNA is cir- cular RNA (circRNA). CircRNAs have no free ends, which protects circRNAs from exonuclease degradation and makes it stable. Like miRNA and lncRNA, circRNA also has regulatory roles in cellular processes, and they have been identified in EVs (Lasda and Parker, 2016).
Loading of RNA into EVs seems to be systematic at some level because enrich- ment of certain RNAs in EVs has been shown (Valadi et al., 2007; Skog et al., 2008;
Eirin et al., 2014). The Bolukbasi research group found that the presence of sequence motif CUGCC in the 3’UTR region of mRNA, where binding site to miR-1289 is pre- sent, causes the enrichment of mRNAs in microvesicles (Bolukbasi et al., 2012). Ac- cording to study of Villarroya-Beltri et al. heterogenous nuclear ribonucleoprotein A2B1 (hnRNPA2B1) regulates miRNA loading into exosomes where hnRNPA2B1 recognizes specific motifs in miRNAs via SUMOylation controlled process (Villarroya-Beltri et al., 2013). Posttranslational modifications such as 3’end adenyla- tion and uridylation may have a role in sorting ncRNAs to EVs (Koppers-Lalic et al., 2014). The levels of EV-RNA seem to vary depending on parent cells even though
of EV-DNA is less studied, but a few attempts have been made to demonstrate the exchange of DNA between the donor and recipient cells. Fischer with her team was able to observe the EV-mediated transport of DNA between bone marrow derived mesenchymal stem cells lentivirally transduced to carry Arabidopsis thaliana DNA (A.t-DNA) and human mesenchymal cells without A.t-DNA. Interestingly the results suggest that A.t. -DNA was stably integrated to recipient cell DNA (Fischer et al., 2016). The current findings in the literature suggest that DNA in the EVs varies be- tween 100 bases to 2.5 kb in size. An EV pellet can contain even longer double- stranded DNA (dsDNA), but DNase treatment will enzymatically degrade these longer DNAs. This means that dsDNA longer than 2,5 kb exists on the outside of EVs in the EV pellets and thus is not protected from enzymatic cleavage by EV membrane (Thakur et al., 2014).
2.2.5 Proteins
In addition to lipids, RNA and DNA, EVs also carry a variety of proteins. The small amount of proteins and the lack of amplification method for proteins set the require- ments for sensitivity of analytical methods high. Different proteins which have been detected in EVs include cytosolic proteins (tubulin, actin, actin binding proteins), proteins needed in membrane transport, proteins that facilitate membrane fusion and transport (annexins and RAB proteins, heat shock proteins for example HSP70 and HSP90, tetraspanins as CD9, CD63, CD81 and CD82, phoshpolipases, MHC proteins and metabolic enzymes (Thery, Zitvogel and Amigorena, 2002).
The mechanism of biogenesis and the cell type seem to have an effect to protein cargo of certain type of EVs to some extent, although there is some overlap in protein composition between EV types (Heijnen et al., 1999; Palmisano et al., 2012; Tauro et al., 2012). Characterization of the protein cargo of EVs thus may give information about origin of certain population of EVs. The biogenesis of exosomes via endosomal pathway leads to enrichment of proteins related to the endosomal pathway, such as MHC class II proteins, tetraspanins CD37, CD53, CD63, CD81 and CD82, ESCRT pro- teins and accessory proteins needed in ESCRT pathway such as Alix and TSG101 (Heijnen et al., 1999; Thery et al., 2001; Tauro et al., 2012).
In turn, microvesicles which are formed via direct budding from the plasma mem- brane are enriched with different set of proteins. Unlike exosomes, microvesicles con- tain relatively high amount of integrins, glycoprotein Ib, P-selectin and arrestin con- taining protein 1 (AARDC1) (Heijnen et al., 1999; Nabhan et al., 2012). The amount of posttranslationally modified proteins, for instance glycoproteins and phosphopro- teins, is higher in microvesicles than in exosomes (Palmisano et al., 2012). Proteins suggested for microvesicle markers include KIF23, RACGAP, CSE1L, ARF6 and EMMPRIN (Greening et al., 2017).
2.2.6 Sugars
As compared to nucleic acids, lipids and proteins, carbohydrate content of EVs has been less extensively studied. It is known that EVs are coated with a glycocalyx that is assembled from polysaccharides and oligosaccharides on the EV surface (Gerlach and Griffin, 2016), reflecting the sugar composition of the original cell. EV carry both non-sulfated glycosaminoglycans such as HA (Rilla et al., 2013, 2014; Schmidt et al., 2016) and sulfated glycosaminoglycans such as chondroitin sulfate (Schmidt et al., 2016) and heparan sulfate (Christianson et al., 2013; Bandari et al., 2018). Glypican-1 is a proteoglycan, which is enriched on cancer cell-derived EV (Melo et al., 2015).
Some studies also suggest an impact of polysaccharides on EV functions. Heparan sulfate proteoglycans on the surface of target cells act as receptors for EV uptake (Christianson et al., 2013) and heparanase, an enzyme that cleaves heparan sulfate, enhances secretion of EVs (Thompson et al., 2013). Glycosaminoglycans, such as chondroitin sulfate and hyaluronan, regulate matrix vesicle activity and bone miner- alization (Schmidt et al., 2016). Because high hyaluronan content in many tumors acts as a barrier for drug delivery, oligo-hyaluronan-loaded nanoparticles have been used to disrupt the hyaluronan coat in order to enhance drug delivery into cancer cells (Yang et al., 2013). On the other hand, hyaluronan coating has been utilized in na- nosized vesicles as a tool for targeted cancer therapy (Wickens et al., 2017). It has been suggested that surface glycans increase the negative charge of EVs (Gerlach and Griffin, 2016), which could be utilized as a cancer-specific indicator. A more specific example of sugar-rich molecules on EVs is the proteoglycan glypican, that has poten- tial as EV marker in the early diagnostics of pancreatic cancer (Melo et al., 2015).
2.3 BIOGENESIS OF EXOSOMES
2.3.1 Exosomes are formed via endosomal route
Exosomes are subpopulation of EVs which are formed via endosomal pathway. Dur- ing the maturation process of an endosome, intraluminal vesicles (ILVs) bud into the endosome. Endosome with internal vesicles is called a multivesicular body (MVB) (Johnstone et al., 1987). MVB can end up in the lysosome for degradation or fuse with the plasma membrane when ILVs are released (Futter et al., 1996). After entering the extracellular space ILVs are called exosomes. (Trams et al.; Harding et al.)
The mechanism behind the cargo sorting to ILVs is thought to consist of two steps.
First step is the gathering the selected proteins together at the outer membrane of the late endosome which is followed by budding of the ILVs into the late endosome
2.3.2 ESCRT-dependent route
The most common way of exosome biogenesis is assumed to be dependent of the ESCRT machinery which is required for sorting of membrane proteins into mul- tivesicular bodies/endosomes and also in formation of intraluminal vesicles (Buschow et al., 2009). ESCRT machinery consists of four ESCRT protein complexes (0-III) and AAA ATPase vacuolar protein-sorting -associated protein 4 (Vps4) (Henne, Buchkovich and Emr, 2011).
In the beginning of the formation of ILVs, tetraspanins such as CD9 and CD63, cluster together on the endosomal membrane, and form tetraspanin enriched micro- domains (TEMs) (Pols and Klumperman, 2009). First ESCRT-O binds with FYVE do- main of HRS subunit to endosome specific phosphatidyl 3-phosphate (PI(3)P) lipid followed by binding to ubiquitinated proteins and clathrin as a result of the cooper- ation of another HRS subunit STAM and accessory proteins. During the next step, heterotetrameric ESCRT-I-complex (TSG101, VPS28, VPS37 and MWB12) is incorpo- rated via binding of ESCRT-I subunit TSG101 by PSAP motif of ESCRT-O subunit HRS. This leads to the binding of ESCRT-I to ubiquitinated cargo, which is followed by recruitment of ESCRT-II. ESCRT-II is composed of two EAP20 subunits, one EAP30 subunit and one EAP45 subunit via which ESCRT-I and ESCRT-II interaction presumably occurs. ESCRT-III assembly is then activated through the interaction of both EAP20 subunits of ESCRT-II with ESCRT-III subunit CMPH6. Another possibil- ity of recruitment of ESCRT-III is the direct interaction between ESCRT-I TSG101 subunit and ESCRT-III CHMP6 subunit. ESCRT-III complex finally causes the mem- brane scission and formation of ILVs which is followed by deubiquitination of the cargo and removing of ESCRT-III from endosome and thus allowing the recycling of ESCRT machinery by AAA ATPase VPS4 with the promotion by VTA1 (Hanson and Cashikar, 2012; Colombo, Raposo and Thery, 2014; Villarroya-Beltri et al., 2014; Abels and Breakefield, 2016; Kalra, Drummen and Mathivanan, 2016; Christ et al., 2017;
Hessvik and Llorente, 2018; Anand et al., 2019). To conclude in the light of recent knowledge ESCRT 0-II are responsible for binding of ubiquitinated cargo and initi- ating of budding of intraluminal membrane which is brought to an end by ESCRT- III.
2.3.3 ESCRT-independent route(s)
Even though the ESCRT-dependent route of exosome formation seems to be the main way of exosome biogenesis, there are also mechanisms that do not require ESCRTs.
Depletion of ESCRTs does not prevent the formation of MVBs, which suggests that there should be another way of budding of intraluminal vesicles.
Sphingolipid ceramide has been suggested to provide an alternate way of ILV formation to MVBs. Ceramide-driven merging of microdomains may lead to do-
main-induced budding on ILVs. Neutral sphingomyelinase 2 in turn regulates bio- synthesis of ceramide (Trajkovic et al., 2008; Kosaka et al., 2010). Another lipid sug- gested to drive ILV formation is lysobisphosphatidic acid (LBPA). Additionally, LBPA-triggered budding of ILVs is thought to be result of inward bending caused by local lipid composition of endosomal membrane.
In addition to ceramide-driven pathway, also another ESCRT-independent mech- anism seems to exist. In this mechanism, which is called luminal domain -dependent pathway, packing of cargo to ILVs occurs as a result of clustering of the cargo to be packed on the surface of an endosome. Raft formation of cholesterol and tetraspanins are involved also in this process (de Gassart et al., 2003; Theos et al., 2006).
Tetraspanin CD63 has been shown to participate in ESCRT-independent route of cargo sorting to ILVs by van Niel and colleagues (van Niel et al., 2011). CD63-regu- lated competing mechanism to ESCRT-dependent route has also been studied by Ed- gar et al. They were able to show CD63 -dependent ILV formation to MVB upon de- pletion of HRS which is the subunit of ESCRT-O. In the same paper they also sug- gested that the different size of ILVs is due to cargo and the mechanism of formation (Edgar, Eden and Futter, 2014).
2.4 SECRETION OF EXOSOMES
The whole picture of exosome secretion mechanisms remains partially obscure, but several mechanisms affecting the exosome secretion are known. Many members of Rab family of small GTPase proteins seem to play an important role in exosome se- cretion. Rab proteins have been shown to take part in the process where the MVB is transported to the plasma membrane (Stenmark, 2009).
Actin and microtubule cytoskeleton is needed also in the transportation of MVB to the plasma membrane. In addition, cortactin protein, which binds to actin, is needed in the process of transport and fusion of MVB. N-ethylmaleimide-sensitive factor attachment protein receptors (SNAREs) play an important role in the process of fusion of MVB with the plasma membrane (Zylbersztejn and Galli, 2011). In the fusion process many other proteins such as SNAP-23 (Castle, Guo and Liu, 2002), VAMP-7 (Hirashima, 2000) and VAMP-8 (Luzio et al., 2005) are needed.
Besides the molecular machinery needed to MVB transportation and its fusion with the plasma membrane, it is important to understand factors which lead to deg- radation of the cargo of MVBs. ISGylation, which is a posttranslational ubiquitin-like modification, controls exosome secretion by promoting lysosomal degradation of MVB proteins (Villarroya-Beltri et al., 2016).
2.5 BIOGENESIS AND SECRETION OF MICROVESICLES
Microvesicles are secreted via direct budding from the plasma membrane. The pro- cess in which microvesicles are secreted starts with a stimulus, which causes an in- crease of Ca2+ concentration in the cytosol. This is followed by enzymatic activity of calpains, gelsolins, scramblases and kinases, which results in inhibition of trans- locases and phosphatases. Finally with the contribution of cytoskeleton, microvesicle formation occurs (Wiedmer and Sims, 1991; VanWijk et al., 2003). The contraction of the cytoskeleton is a process which starts when ARF&-GTP (ADP ribosylation factor 6 that binds to GTP) activates phospholipase D. This is followed by the attachment of ERK (extracellular signal-regulated kinase) to the plasma membrane. Phosphory- lation of MLCK (Myosin light chain kinase) by ERK, in turn, leads to phosphorylation of the light chain of myosin which seems to be needed in ERK regulated microvesicle formation (Muralidharan-Chari et al., 2009).
Besides to above mentioned molecular process, there are other factors which are linked to microvesicle formation. The lipid structure of plasma membrane seems to affect the microvesicle formation. Phosphatidylserine accumulation to the outer leaf- let of the plasma membrane has been shown to promote budding of microvesicles.
Aminophosphotranslocases are the regulators of structure of the plasma membrane, and their function is to regulate the transfer of phospholipids between outer and in- ner leaflet (Hugel et al., 2005).
Also, the distribution of phosphatidylethanolamine between outer and inner leaf- let of plasma membrane plays a role in microvesicle formation. Normally almost all the phosphatidylenolamine exists on the inner leaflet of the plasma membrane. Phos- phatidylethanolamine-translocase (TAT5) type IV P-type ATPase has been shown to regulate microvesicle secretion in C. elegans. In mutants without activity of TAT-5, phosphatidylenolamine is transferred to outer leaflet of plasma membrane, which leads to microvesicle formation because of the attachment of ESCRT-complexes to the inner leaflet of plasma membrane (Tuck, 2011).
2.6 EV-MEDIATED INTERACTIONS
2.6.1 EV-uptake
It is nowadays commonly known that EVs play an important role in cell-to-cell mes- saging. What are the delivery mechanisms behind the EV-mediated messaging? To convey the message from the donor cell to target cell EVs must be able to interact with the target cell. Possible ways to this interaction are receptor-mediated signaling, direct fusion of the EV with the cell membrane of the target cell, and endocytosis of the EV (Figure 1). To understand EV-mediated cell-to-cell communication it is im- portant to perceive how EVs are internalized by the target cells. There are two possi- ble mechanisms for the internalization: Endocytosis or membrane fusion of the EVs to the target cell plasma membrane. Endocytosis is an umbrella term, which covers
pinocytosis, receptor mediated endocytosis, phagocytosis and lipid raft mediated en- docytosis. Pinocytosis is the common mechanism of endocytosis, which occurs in all cell types. It is an unspecific mechanism compared to receptor mediated endocytosis.
Different mechanisms of pinocytosis are clathrin-mediated endocytosis and clathrin- independent endocytosis, which can be further divided to caveolin-mediated endo- cytosis and macropinocytosis. Receptor mediated endocytosis is a mechanism of en- docytosis, which is triggered by a specific molecular signal. Phagocytosis is a special- ized form of endocytosis, which is done by phagocytes (Doherty and McMahon, 2009). There is evidence in the literature suggesting that EV-uptake is not a passive process. Lowering the incubation temperature to 4˚C leads to reduction of EV-uptake (Morelli et al., 2004; Escrevente et al., 2011). EVs are neither internalized after para- formaldehyde fixation (Fitzner et al., 2011; Pan et al., 2012).
Clathrin-mediated endocytosis (CME) is a mechanism of endocytosis, where re- ceptors are endocytosed through inward budding of the plasma membrane via clath- rin-coated pits. The role of CME as a mechanism of EV-uptake is supported in the literature. The use of chlorpromazine inhibits the formation of clathrin-coated pits.
Chlorpromazine has been shown to reduce EV uptake on ovarian cancer cells and mildly on phagocytic cells (Feng et al., 2010; Escrevente et al., 2011). Inhibition of dy- namin 2, which is required in CME, was able to efficiently reduce internalization of EVs on phagocytic cells (Feng et al., 2010).
Caveolin-dependent endocytosis (CDE) is another mechanism shown possibly to participate in EV uptake. As in CME also in CDE plasma membrane invaginations are formed. In the case of CDE, these structures are called caveolae. Caveolin1 is re- quired in the formation of caveolae. As CME, also CDE requires dynamin-2, which can be inhibited with dynasore. Inhibition of dynamin-2 leads to reduced uptake of EVs, which can be results of impaired CDE (Menck et al., 2013; Nanbo et al., 2013). In the literature there are interesting findings that knockdown of CAV1 gene may re- sults in decreased or increased EV internalization in different contexts (Nanbo et al., 2013; Svensson et al., 2013).
Macropinocytosis is another unspecific mechanism of endocytosis, where the cell engulfs extracellular material with membrane ruffles through actin-dependent pro- cess. Evidence for the participation of micropinocytosis as a mechanism of EV uptake has also shown. NA+/H+ exchanger inhibition was demonstrated to reduce EV inter- nalization by microglia. Also blocking Rac1, which is GTPase needed in macropino- cytosis, with NSC23766 resulted in reduction of EV uptake (Fitzner et al., 2011).
Phagocytosis is the way to internalize particles, such as bacteria, by macrophages.
Even though phagocytosis is mainly used to uptake larger particles into cells it could
In the light of a recent study it seems the clathrin-independent endocytosis and macropinocytosis are the most prevalent mechanisms which cells utilize for EV up- take. In the study performed with HeLa cells treated with chemical inhibitors or siRNA they observed internalization of fluorescently labeled EVs in 2D and 3D cul- tures. As the results they found that EV uptake was dependent of cholesterol and tyrosine kinase activity which are important in clathrin-independent endocytosis.
Also macropinocytosis seems to be the way of EV uptake because blocking of Na+/H+
exchange and PI3K restricts EV internalization (Costa Verdera et al., 2017). EV-uptake seems to be a complex process which occurs via multiple ways.
2.7 EXTRACELLULAR VESICLES AS BIOMARKERS
2.7.1 A new minimally invasive liquid biopsy
The role of EVs as important players in the intercellular communication and part of physiological and pathological processes in addition to their feature to reflect the properties of parent cell makes them potential candidates for biomarker studies. EVs have been shown to carry proteins, RNA and DNA. The analysis of the change of these cargo molecules has been in the context of interest of biomarkers studies in the EV field since mRNA and miRNA were seen to be transferred between cells via exo- somes (Valadi et al., 2007). The presence of EVs in wide variety of body fluids enables of the development of minimally invasive tools for early diagnosis and prognosis without the need for traditional tissue biopsy.
Several studies have already tried to identify possible biomarkers and have cov- ered multiple types of cancers, neurodegenerative diseases, cardiovascular diseases, type 1 diabetes mellitus, hematologic malignancies, fatty liver disease and many other disease states. The focus in the following chapters will be in the EV-protein and -RNA biomarkers for different types of cancer – especially in lung, breast and pros- tate cancers because of their high prevalence worldwide.
At first glance, low commonality in EV-biomarker candidates for cancers acquired by different research groups, may raise some questions. A closer look to the results reveals that varying results may be due to the different research questions that have been tried to solve. For example, biomarkers to distinguish between two types of lung cancer and between cancer patients and healthy individuals, may differ sub- stanstially. Another point to consider when searching for cancer biomarkers is the dynamically changing RNA and protein repertoire that cancer cells produce depend- ing on the stage of cancer development.
2.7.2 Exosomal proteins as biomarkers 2.7.2.1 Lung cancer
Lung cancer is the most common cancer worldwide and one of the leading causes of cancer-related death. Multiple proteomic studies have been performed to study exo- somal biomarker proteins in lung cancers. Epidermal growth factor (EGFR) carried by exosomes has been suggested to act as a biomarker for lung cancer by several studies (S.-H. Huang et al., 2013; Yamashita et al., 2013; Clark et al., 2016).
To examine exosomal protein markers for non-small cell lung carcinoma (NSCLC) Jakobsen et al. performed a study in 2015 with plasma with 109 NSCLC patients and 110 matched controls. With the multimarker model with 30 different markers they were able to classify NSCLC patients with 75.3% accuracy (Jakobsen et al., 2015).
Sandfeld-Paulsen with colleagues analyzed exosomes from plasma of 276 NSCLC patients in 2016. As a result of phenotyping EVs with an array containing 49 antibod- ies they saw that NY-ESO-1, EGFR, PLAP, EpCam and Alix were significantly af- fected to overall survival in of NSCLC patients in concentration dependent manner.
Only one of these markers which had a significant impact on inferior survival after Bonferroni correction was NY-ESO-1 (Sandfeld-Paulsen, Aggerholm-Pedersen, et al., 2016). In another paper they performed a vast study with 581 patients to search for exosomal protein biomarkers for all stages and different histological types of lung cancer. As a result of multimarker model, which was evaluated by area under the curve and random forest analysis, they found that the biggest difference between cancer and control group was in expression of CD151, CD171 and tetraspanin 8 (Sandfeld-Paulsen, Jakobsen, et al., 2016).
In the study of Vykoukal et al. possible EV-related biomarkers for lung adenocar- cinoma were screened from human blood samples with proteomic approach. They were able to identify several proteins which could act as biomarkers for lung adeno- carcinoma of which SGRN, TPM3, THBS1 and HUWE1 were the most promising ones (Vykoukal et al., 2017).
In a recent study performed with Chinese participants by Niu et al. in 2019, serum exosomes were isolated by ultracentrifugation from 125 NSCLC patients and 46 healthy donors. They identified 43 differentially expressed proteins. Based on the further results AHSG and ECM1 were suggested to act as biomarker for NSCLC (Niu et al., 2019).
2.7.2.2 Breast cancer
Many exosomal marker proteins have been identified also for breast cancer, which is
