Characterization of blood markers and their implication in human aging.

Characterization of blood markers and their implication in human aging Ainhoa Alberro Garitano To obtain a doctoral degree in Molecular Biology and Biomedicine from the University of the Basque Country Supervisors: David Otaegui Bichot Matías Sáenz Cuesta January 2020 (cc)2020 AINHOA ALBERRO GARITANO (cc by.nc-nd 4.0) This thesis has been carried out at the Biodonostia Health Research Institute. During this work, I was supported by a fellowship from the Department of Education of the Basque Government (PRE_2015_1_0341, 2016-2019), by a fellowship from the Jesús de Gangoiti Barrera Foundation (2016) and by an EMBO Short-Term Fellowship (STF_7624, 2018). Izan zirenei, direnei, izango direnei. Laburpena Zahartzea definitzean, gure gorputzak denboran zehar jasaten duen aldaketa multzoa dela esaten dugu, gaitasun funtzionalaren galera eta baita gaixotasunekiko eta kanpoko erasoekiko zaurgarritasun handiagoa dakarrena. Bizi kalitatearen hobekuntzei eta azken hamarkadetan medikuntzan eman diren aurrerapen handiei esker, adinekoen kopurua nabarmen igo da. Batez besteko bizi-itxaropenaren hazkuntza honen eta jaiotza-tasaren beherakadaren ondorioz, biztanleria zentsu gero eta desorekatuagoa daukagu. Gainera, bizi-itxaropena igo den arren, osasuntsu bizi garen urte kopurua ez da hazi eta, beraz, luzeago bizi gara, baina ez hobeto. Horrenbestez, zahartzea orokorrean, eta menpekotasuna bereziki, gure gizartearen erronka bihurtu dira, eta gero eta baliabide gehiago bideratzen dira hauetara. Gauzak horrela, esparru honetan burutzen diren ikerketa medikuen helburu nagusia menpekotasuna murriztea izan behar da, zahartze osasuntsu bat lortzeko. Zahartze osasuntsuaren bidean, hauskortasun kontzeptuarekin egiten dugu lan. Hauskortasuna adinarekin erlazionatutako sindrome medikoa da, eta bere ezaugarri nagusiak erreserba funtzionalaren galera, sistema fisiologikoen egokitze gaitasun murriztua eta zaurgarritasunaren areagotzea dira. Hauskortasuna hainbat ondorio negatibo garatzeko arriskuarekin erlazionatuta dago, hala nola, erorketak, hausturak, infekzioak, desgaitasuna, ospitalizazioa, menpekotasuna eta heriotza. Beraz, hauskortasuna daukaten pertsonen identifikazioa beharrezkoa da neurriak hartu ahal izateko eta etorkizuneko arazoak saihesten saiatzeko. Baina, orain arte ezin izan da hauskortasuna detektatzeko metodo eraginkorrik garatu, eta hainbat test eta proba funtzional badauden arren, hauek ez dira gai pertsona hauskor guztiak identifikatzeko. Gaur egun eskuragarri dauden tresnak osatu eta emaitza hobeak lortzeko asmoz, hauskortasunaren biomarkatzaileak aurkitzeko ikerketak burutzen ari dira, aurrerago eztabaidatuko dugun moduan. Biomarkatzaileen bilaketa azaldu aurretik, ordea, zahartze prozesuaren oinarri biologikoa laburtuko dugu. Zahartzea prozesua unibertsala, progresiboa eta heterogeneoa da, eta beraz, modu oso desberdinetan garatzen da. Zahartzearen jatorri molekularrak eta zelularrak ulertzea biologiaren erronka nagusietako bat da, eta helburu horrekin lan ugari egin dira. Ikerketa horiei esker, orain dela urte batzuk zahartzearen 9 ezaugarri nagusiak zehaztu ziren, ondorengoak direlarik: ezegonkortasun genomikoa, aldaketa epigenetikoak, telomeroen laburtzea, disfuntzio mitokondrialak, zelula amen agortzea, zelulen seneszentzia, zelulen arteko komunikazio desberdina, proteostasiaren galera eta mantenugaien hautemate 18 | Laburpena okerragoa. Hauetariko ezaugarri bakoitza sakonki ikertu da komunitate zientifikoan, baina, jarraian tesi honetan landu direnak azalduko ditugu laburki. Zahartzaroarekin gertatzen den ezegonkortasun genomikoa, neurri batean, DNAren egituran aldaketak metatzeari zor zaio, hala nola mutazioak, delezioak edo DNAren hausturak harizpi batean edo bietan. Zelulek badauzkate akats horiek kontrolatzeko eta konpontzeko mekanismoak, baina hauek ere adinarekin huts egiten dutela ikusi da. Bestalde, geneen adierazpena aldatu egiten da zahartzean, eta hau beste gene batzuen adierazpena erregulatzen duten transkripzio faktoreetan gertatzen bada, eragina are handiagoa izango da. Gene ez kodifikatzaileen adierazpena ere aldatu egin daiteke, eta transkripzio osteko erregulazioan eragina eduki, mikroRNA molekulen kasuan bezala. Seneszentziak zelulen proliferazioaren galera dakar, jarduera metabolikoa, bideragarritasuna eta haien berezko funtzio batzuk mantentzen dituzten bitartean. Zelula kaltetu eta potentzialki arriskutsuak izan daitezkeenak seneszentzian sartzea defentsa mekanismoa dela deskribatu da, hauen proliferazioa ekiditen baita. Era berean, badirudi zelula seneszenteek jariatutako seinaleek ehunen birsorkuntzan laguntzen dutela: zelula seneszenteen digestioa eta zelula amen diferentziazioa sustatzen dute, zelula berri helduek ehuna berritzen dutelarik. Adinean aurrera egin ahala, ordea, prozesu honek ere huts egiteko joera dauka, zelula seneszenteak pilatu egiten dira eta kalteak eragin ditzakete. Lan honetan bereziki immunitate-sistemaren seneszentzian (immunoseneszentzian) zentratu gara, adineko pertsonen zaurgarritasunean eragina baitauka. Aztertu dugun beste ezaugarri bat zelula amen agortzea da. Hau oso lotuta dago seneszentziarekin, izan ere, ehunen birsorkuntza ezin da osatu zelula amarik ez badago edo behar bezala diferentziatzen ez badira. Zelula amak kieszentzia egoeran mantentzen dira eta beharrezkoa denean bakarrik aktibatu, proliferatu eta diferentziatzen dira. Gainera, zelula amek auto-berrikuntza gaitasuna dute, zelula alabetako bat zelula ama moduan mantendu eta bestea bakarrik diferentziatzen bada. Baina zelula amen agortzea zelulak simetrikoki banatzen direnean eman daiteke, edo zahartzearekin lotutako arazoak direla medio kieszentzia egoeratik irten eta aktibatzea lortzen ez dutenean. Izan ere, kieszentziatik irtetzeko prozesua konplexua da eta oso erregulatuta dago, eta zenbait egoerek, hala nola, telomeroen laburtzeak, estres oxidatzaileak edota DNAn eman diren kalteek prozesua eragotzi dezakete. Azkenik, zahartzaroan ematen den zelulen arteko komunikazio desberdinaren inguruan ere ikertu dugu. Zahartzaroko ezaugarri hau bereziki garrantzitsua da guretzat bi arrazoirengatik. Alde batetik, zahartzerakoan ematen diren aldaketek inflamazio basal kronikoa garatzera eramaten baitute, inflammaging izenez ezagutzen dena. Inflamazioa sustatzen duten molekula Laburpena | 19 hauen jariatzea hainbat zelula motek burutzen dute, eta adinarekin ematen den immunitatesistemaren funtzionamendu txarra areagotzen dute. Bestalde, zelulaz kanpoko besikulak (EVak, ingeleseko laburdurarekin) dauzkagu, zelula mota gehienek ekoiztu eta jariatzen dituzten partikulak eta zelulen arteko komunikaziorako bide bat direnak. EVak duela 50 urte baino gehiago aurkitu ziren, baina hasieran ez zitzaien garrantzirik eman eta zelulen “zaborra” ateratzeko” modu bat zela pentsatu zen. Azken hamarkadetan hauen funtzioak sakonago aztertzen hasi zen, eta oraindik ere, urtero EVen funtzio berriren bat deskribatzen da. Hala ere, zahartzearen ezaugarri nagusien artean zelulen arteko komunikazio desberdina izendatu zenean, ez ziren EVak kontuan hartu, eta horregatik zahartze prozesuan daukaten inplikazioa erakusten jarraitu beharra daukagu. Hauskortasun sindromearen biomarkatzaileen gaiari berriro eutsita, lan asko egin izan dira zahartzearen ezaugarri orokorrak hauskortasunarekin erlazionatzen saiatzeko. Ikerketa hauen hipotesia zera da: adinarekin aldaketa edo prozesu kaltegarriak gertatzen dira, eta hauek neurri handiago batetan eman daitezke hauskortasuna edo menpekotasuna pairatzen duten pertsonetan, gaitasun funtzionala mantentzen dutenetan (sendoetan) baino. Ildo horretatik, inflamazioa edo estres oxidatzailearekin lotutako molekulen igoerak, eta hormonen edo metabolismoaren erregulazio galerak hauskortasunarekin izan ditzaketen loturak aztertu dira, adibidez. DNA sekuentzia espezifikoek, geneen erregulazioak edo geneen adierazpen aldaketek hauskortasuna garatzeko joerarekin izan ditzaketen erlazioak ere ikertu izan dira. Biomarkatzaile hauei eta bestelako batzuei buruzko artikulu zientifiko ugari kaleratu dira. Horietako batzuek hauskortasunarekin edo hau garatzeko arriskuarekin harremanak aurkitu dituzte, baina badaude ere proposatutako biomarkatzaileak balioztatu ez dituzten ikerketak, eta horregatik, gaur egun oraindik ez da hauskortasunaren identifikazioan lagundu dezakeen biomarkatzailerik aplikatzen praktika klinikoan. Gauzak horrela, tesi honen lehenengo kapituluan hauskortasun biomarkatzaileak aurkitzen saiatu gara. Horretarako, adineko pertsonen odol laginak eskuratu ditugu, eta hauek sendo eta hauskor taldeetan banatu ditugu test desberdinetan lortutako emaitzen arabera. Lehenik, inflamazioaren ezaugarri diren molekulak neurtu ditugu. Hauekin emaitza positiboak lortu izan dira aurreko ikerketa batzuetan, baina badaude hauskortasunarekin erlaziorik aurkitu ez duten lanak ere, eta horien antzera, gure laginetan ez dugu inflamazio molekulen kontzentrazio handiagorik topatu pertsona hauskorretan. Gure bigarren estrategia azterketa transkriptomikoa egitea izan da. Horren bidez, pertsona sendoak eta hauskorrak bereizten dituzten 35 gene identifikatu ahal izan ditugu. Hasteko, gene horietako 3ren adierazpena neurtu dugu kohorte zabalago batetan eta EGR1 genearen gainadierazpena balioztatu dugu pertsona hauskorretan. EGR1 prozesu zelular garrantzitsuetan parte hartzen duen 20 | Laburpena transkripzio faktorea da, hala nola mitogenoekiko erantzunean, proliferazioan, apoptosian eta hainbat zelula moten diferentziazioan, eta estimulu desberdinen ondorioz aktibatua izan daiteke. Gainera, adineko pertsonen hauskortasun egoerari buelta eman edo murrizten saiatzeko esku-hartze bat burutu da, eta 3 hilabetez ariketa fisikoa egin dute partaideek. Eskuhartzearen ondoren, 12 partaideetatik 9k EGR1 genearen adierazpena jaitsi dutela aurkitu dugu. Emaitza nabarmen hauekin, EGR1en adierazpen altua hauskortasun biomarkatzaile bezala proposatzen dugu, eta etorkizuneko ikerketetan kontuan hartu beharrekoa dela deritzogu. Hauskortasun biomarkatzaileen ikerketan, EVen kontzentrazio plasmatikoa izan da gure azken hurbilketa. Izan ere, zenbait inflamazio prozesuetan, minbizia eta gaixotasun autoimmuneak esaterako, EVen kontzentrazio altua aurkitu da, eta gure helburua hau adinean ematen den inflammaging-arekin ere gertatzen ote den aztertzea izan da. Gure emaitzek erakutsi dutenez, nahiz eta inflamazio basala adineko pertsonetan konfirmatu den, hauek ez dute EV kontzentrazio plasmatiko handiagorik, eta ez dago ezta ere sendo eta hauskorren arteko diferentziarik. Emaitza hauek adierazten dutenez, beraz, ez da EVen kontzentrazio igoera ematen inflamazio prozesu guztietan. Lan honen bigarren kapituluan EVek zelulen diferentziazioan duten eragina aztertu dugu. Osteogenesi eta miogenesian zentratu gara, zahartzaroan kaltetutako prozesuak baitira eta osteoporosian eta sarkopenian eragina baitute, hurrenez hurren. Adinarekin ohikoak diren bi prozesu hauek ikuspegi desberdinetatik sakonki aztertu izan dira, baina oso gutxik ikertu dute EVen inplikazioa. Gantz ehunetik eratorritako zelula ama mesenkimalekin burututako gure lanak erakusten duenez, plasmako EVek osteogenesi prozesuan laguntzen dute, eta eragin hau nabarmenagoa da helduetatik eskuratutako EVekin, adineko emaileekin alderatuta. Era berean, plasmako EVek miogenesian duten eragina aztertzeko protokolo bat garatzen saiatu gara, eta lehen emaitzek adierazten dutenez, mioblastoen diferentziazio prozesua bultzatzen dute EVek, berriro ere adineko pertsonetatik eskuratuko EVek efektu murriztuagoa dutelarik. Hirugarren kapituluan immunoseneszentziaren fenomenoari jarri diogu arreta, eta helduak (20-49 urte) eta pertsona adindunak (70-104 urte) aztertu ditugu. Lehenik, seneszentzia zelula mailan neurtu da, eta ikusi dugu CD8 T linfozito seneszenteen proportzioa hazi egiten dela adinarekin. Zahartzaroaren lehen hamarkadetan berdina gertatzen da CD4 T linfozitoetan, baina 90. urtetik gorako pertsonetan zelula seneszente mota honen kantitatea murriztuagoa dela aurkitu dugu. Ondoren, EVek linfozito seneszenteen ezaugarri berdinak erakusten ote dituzten aztertu dugu. Zelula hauen markatzaile berdinak EVen mintzean aurkitu ditugu, baina ez, ordea, “EV seneszenteen” igoerarik adinak aurrera egin ahala. Azkenik, linfozitoen eta EVen in vitro esperimentuak burutu ditugu, EVek zelula hauen aktibazioan eragina duten Laburpena | 21 ebaluatzeko. Emaitzek erakutsi dute EVek ez dela erantzun immunologikorik eragiten, eta are gehiago, zelulen bideragarritasuna hobetzen dutela. Baina, estimulu immunogeniko baten aurrean erantzuteko orduan, EVek T linfozitoen aktibazioa areagotzen dute. Gainera, aktibazioa areagotzeko gaitasun hau ahulagoa da EVak emaile zaharretatik isolatuak badira. Horrela, gure emaitzek adierazten dute, nahiz eta EVen mintzeko molekuletan desberdintasunik ez egon, EVek gaitasun funtzional desberdinak dauzkatela isolatu izan diren emailearen adinaren arabera. Tesi honen laugarren eta azken kapituluan, immunoseneszentzia aztertu dugu esklerosi anizkoitza duten gaixoetan. Izan ere, proposatu izan da gaixotasun autoimmuneetan ematen den etengabeko immunitate-sistemaren aktibazioa dela eta azken honen agortzea eman daitekeela, eta ondorioz, immunoseneszentzia goiztiarra. Hipotesi hau ebaluatzeko, esklerosi anizkoitza duten pertsonen eta heldu osasuntsuen laginak alderatu ditugu. Esklerosi anizkoitza daukaten gaixoek, tratamendu immunomodulatzaileak hartzen eta erremisioan egon arren, pertsona osasuntsuek baino inflamazio markatzaile kontzentrazio altuagoak dauzkatela aurkitu dugu. Honek inflamazio basala daukatela adierazten du, inflammagingarekin erlazionatuta egon litekeena. Bestalde, T linfozitoen analisian, esklerosi anizkoitza daukaten gaixoek zelula seneszente gehiagorik ez dutela ikusi dugu, baina gaixo hauen T linfozitoek erantzun desberdina daukate in vitro jasotako estimulu baten aurrean, aktibazio murriztuagoa erakusten dutelarik pertsona osasuntsuen zelulekin alderatuta. Bukatzeko, laburpen modura, doktoretza tesi honek zahartze prozesuaren hainbat alderdiri buruzko ezagutza aurreratzen lagundu duela esan dezakegu. Hauskortasun biomarkatzaileak ikertu ditugu, EGR1 genearen adierazpena etorkizunean kontuan hartu beharreko markatzaile bezala proposatuz. Plasmatik isolatutako EVen ezaugarriak ere aztertu ditugu, osteogenesia eta miogenesia bezalako zelulen diferentziazio prozesuetan lagundu dezaketela erakutsiz, baita T linfozitoen aktibazioa sustatzen dutela estimulu immunogenikoen aurrean, eta EVen funtzio hauek zahartzearekin ahuldu egiten direla. Emaitza hauek guztiek erakusten digutenez, EVek paper garrantzitsuak betetzen dituzte hainbat prozesuetan, eta aintzat hartu beharko lirateke zahartzearen ezaugarri nagusietan. Azkenik, esklerosi anizkoitzaren inguruan burututako esperimentuek adierazten digute gaixotasun hau daukaten pertsonek inflamazio kronikoa badaukatela, eta immunoseneszentzia goiztiarra esklerosi anizkoitzean eta bestelako gaixotasun autoimmuneetan garatzen ote den ikertzen jarraitu behar dugula. Table of contents Abbreviations......................................................................................................................................................... 31 INTRODUCTION ......................................................................................................................................................... 33 1. Aging ..................................................................................................................................................................... 35 1.1. Frailty and other relevant age-related health concepts .......................................................... 37 1.1.1. Frailty and robustness .................................................................................................................. 37 1.1.2. Frailty scales ..................................................................................................................................... 38 1.1.3. Resilience ........................................................................................................................................... 40 1.1.4. Intrinsic capacity ............................................................................................................................ 41 1.1.5. Healthy aging, successful aging and related concepts..................................................... 42 1.2. The biology of aging ............................................................................................................................... 43 1.2.1. Genomic instability and gene expression ............................................................................. 45 1.2.2. Telomere attrition and epigenetic alterations ................................................................... 46 1.2.3. Loss of proteostasis ....................................................................................................................... 49 1.2.4. Deregulated nutrient sensing .................................................................................................... 50 1.2.5. Mitochondrial dysfunction ......................................................................................................... 51 1.2.6. Cellular senescence ........................................................................................................................ 52 1.2.6.1. Immunosenescence ...................................................................................................................... 53 1.2.7. Stem cell exhaustion ...................................................................................................................... 55 1.2.8. Altered intercellular communication ..................................................................................... 58 1.2.8.1 Inflammaging ................................................................................................................................. 59 1.3. Molecular biomarkers of frailty ........................................................................................................ 60 2. Extracellular vesicles ..................................................................................................................................... 66 2.1. Biological characteristics of extracellular vesicles .................................................................... 66 2.2. Isolation and characterization of extracellular vesicles ......................................................... 68 2.3. Extracellular vesicles in physiological and pathological processes ................................... 71 2.4. Potential clinical applications of extracellular vesicles........................................................... 73 3. Multiple sclerosis ............................................................................................................................................ 76 3.1. Etiopathology of multiple sclerosis ................................................................................................. 76 3.2. Multiple sclerosis and premature aging ........................................................................................ 78 JUSTIFICATION .......................................................................................................................................................... 81 HYPOTHESIS AND OBJECTIVES .......................................................................................................................... 85 CHAPTER ONE Biomarkers of frailty ................................................................................................................ 89 Inflammation ......................................................................................................................................................... 95 Transcriptomics ................................................................................................................................................ 103 Extracellular vesicles ....................................................................................................................................... 113 CHAPTER TWO Influence of extracellular vesicles and age on cell differentiation ................... 119 Osteogenesis ....................................................................................................................................................... 123 Myogenesis .......................................................................................................................................................... 133 CHAPTER THREE Immunosenescence and the role of extracellular vesicles .............................. 143 CHAPTER FOUR Multiple sclerosis and premature aging..................................................................... 163 GENERAL DISCUSSION, PERSONAL OPINION AND FUTURE PERSPECTIVE .............................. 179 CONCLUSIONS ......................................................................................................................................................... 187 PUBLICATIONS ....................................................................................................................................................... 191 REFERENCES ........................................................................................................................................................... 197 | 31 Abbreviations 7-AAD 7-aminoactinomycin D ADL Activities of daily living ASC Adipose tissue-derived stem cells BBB Blood-brain barrier BSA Bovine serum albumin BDNF Brain-derived neurotrophic factor circRNA circular RNA CNS Central nervous system CRP C-reactive protein CSHA Canadian Study of Health and Aging DHEAS Dehydroepiandrosterone sulfate EM Electron microscopy EV Extracellular vesicle FMO Fluorescence minus one GFST Gerontopole Frailty Screening Tool GS Gait speed HC Healthy control HSC Hematopoietic stem cell HSP Heat shock protein IGF-1 Insulin-like growth factor-1 IIS Insulin and insulin-like growth factor 1 IL Interleukin MD Myogenic differentiation MHC Major histocompatibility complex miRNA microRNA MS Multiple sclerosis MSC Mesenchymal stem cell mtDNA mitochondrial DNA MuSC Muscle stem cell NSC Neural stem cell NTA Nanoparticle tracking analysis OD Osteogenic differentiation PBMC Peripheral blood mononuclear cell PCA Principal Component Analysis PHA Phytohemagglutinin PRP Platelet-rich plasma ROC Receiver Operating Characteristic ROS Reactive oxygen species RT-qPCR Quantitative real-time PCR SASP Senescence-associated secretory phenotype sncRNA small non-coding RNA SNP Single nucleotide polymorphism SPPB Short Physical Performance Battery TAC Transcriptome Analysis Console TFF Tangential flow filtration TFI Tilburg Frailty Indicator TNF-α Tumour necrosis factor alpha TUG Timed up-and-go UPS Ubiquitin-proteasome system INTRODUCTION Introduction | 35 1. Aging The process of getting old affects each of us and our society as a whole. Aging is defined as the accumulation of time-related modifications that lead to decreased functional capacity, as well as increased susceptibility and vulnerability to disease or external insults [1]. It is a universal, complex and heterogeneous process. The present work is focused on biological aspects of aging, but we consider we should first outline some general characteristics to better understand the implications and motivations of our research. Regarding demographics, a dramatic change in the proportions of young and elder people is observed in most of the countries. On one side, the improvements in social, medical and economic conditions have resulted in reduced mortality, better quality of life and consequent increase in life expectancy. On the other side, the reduced birth rate contributes to the overall increase of the population age. In Figure 1 the population pyramids of the European Union in 2003 and 2018 are depicted, showing a clear drop of the inhabitants younger than 44 years and a rise of the ones over 45 years. Figure 1. Population pyramids of the 28 member states of the European Union in the years 2003 and 2018. The percentage of people of all age ranges up to 40-44 years has decreased, while all the age ranges over 45-40 years have increased, demonstrating the aging of the population. Source: Eurostat. “Population structure and ageing”. Available at: https://ec.europa.eu/eurostat/ 36 | Introduction Moreover, the current picture will continue evolving and the population is projected to age notably more. In Figure 2 the present population structure and future projections are shown. Interestingly, we can observe that the proportion of people aged more than 80 years is expected to increase from 5.6% to 14.6% in the year 2100. Figure 2. Population structure by major age groups of the 28 member states of the European Union in the year 2018 and the projections for the next decades. The proportion of children (0-14 years) will have only a minor decrease, while a marked decrease of the adults (15-64 years) and increase of elders (6579 years and 80+ years) is expected. Source: Eurostat. “Population structure and ageing”. Available at: https://ec.europa.eu/eurostat/ However, the reports from the European Union also show that life expectancy at birth is increasing, while the healthy life years at birth (also called disability-free life expectancy) is not rising. For instance, in 2016 life expectancy at birth was 83.6 years for women and 78.2 years for men and healthy life years were 64.2 and 63.5 years respectively. This indicates that a woman born in 2016 will live the approximately 77% of her life without disability, while it would be an 81% of his life for a man (https://ec.europa.eu/eurostat/). Furthermore, these percentages would continue to decline as long as life expectancy increases, but no reduction of disability is achieved. Introduction | 37 In light of the commented demographic changes and disability data, the importance of taking appropriate and effective actions becomes evident. Moreover, as stated by the World Health Organization, the aging of our societies is one of the major challenges of the 21st Century [2], as it reaches not only sanitary but also many socioeconomic aspects. In consequence, decisions coordinated by experts of different fields should be taken, aiming to achieve the well-being and healthy aging of the population while maintaining financial sustainability. One of the key actions to face the aging challenge is research. Research on the underlying mechanisms, the age-associated diseases and loss of functions, the interventions and the outcomes are essential to better understand the aging process and to be able to implement innovative treatments and/or interventions. Besides, the primary objective of biomedical research on aging should not be focused on the extension of life, it should aim to improve the quality of life of the elderly, reducing disability and prolonging healthy aging. 1.1. Frailty and other relevant age-related health concepts Many works have previously studied the loss of functions associated with aging. Notably, there have been different approaches to investigate the age-related dysfunctions, and consequently, multiple terms have also been proposed. In the next lines, the main concepts are presented, and their principal characteristics explained. 1.1.1. Frailty and robustness Frailty is a common age-related medical syndrome, characterized by a reduced functional reserve, impaired adaptive capacity across multiple physiological systems and increased vulnerability to stressors [3]. The accentuated vulnerability results in high risk of negative outcomes, such as falls, fractures, infections, disability, hospitalization, dependency and death [4]. Frailty syndrome has been widely studied for decades, but still, no consensus has been reached on its definition and identification tools. Regarding the concept of frailty, a work by Rodríguez-Mañas and colleagues gathered the definitions of experts in the field and presented a list of accepted statements that define frailty [5]. This list included aspects of physical performance, nutritional status, mental health, and cognition. However, they concluded that, even if some concepts of frailty are widely agreed, there is no consensus on an operational definition of frailty. Despite the lack of a complete definition, as mentioned before, frailty implies a reduced functional capacity of an individual that results in an increased risk of developing dependence. The opposite situation to frailty is most of the times termed robustness. An elder is classified as robust when her/his functional capacity is conserved, and besides, phenotypic stability is 44 | Introduction A few years ago, in 2013, a noteworthy classification of the cellular and molecular aspects of aging was proposed by López-Otín and collaborators [33]. They distinguished 9 hallmarks of aging, and further categorized them as primary, antagonistic and integrative hallmarks (Figure 6). It is important to mention that the authors proposed this categorization because the hallmarks are tightly interconnected and they cannot be understood individually. In the next sections, the main characteristics of the proposed hallmarks of aging are explained, with a special focus on the 4 hallmarks that have been investigated in this thesis. Besides, we complemented these sections with information that was not included in the original description of the hallmarks, but that we consider relevant both in the context of this work and for the general understanding of the biology of aging. Finally, we should bear in mind that thanks to new findings the knowledge about aging is rapidly evolving. The hallmarks proposed in 2013 represented the current state of the art, but probably, modifications such as the inclusion of new categories or further explanations on their connexions and implications will be made in the near future. Figure 6. The hallmarks of aging and their interconnections. Genomic instability, telomere attrition, epigenetic alterations and loss of proteostasis are considered the primary hallmarks, the primary causes of cellular damage. Deregulated nutrient sensing, mitochondrial dysfunction and cellular senescence are part of compensatory or antagonistic responses to the damage. These responses are proposed to initially mitigate the damage, but eventually, they become deleterious themselves. Stem cell exhaustion and altered intercellular communication compose the integrative hallmarks, they are the end result of the previous two groups and are ultimately responsible for the functional decline associated with aging. The black arrows indicate the hallmarks that have been investigated in the present work. Adapted from [33]. Introduction | 45 1.2.1. Genomic instability and gene expression There is extensive evidence showing that genomic damage accompanies aging. The accumulation of DNA damage is caused by both endogenous and exogenous threats, such as reactive oxygen species (ROS) and ultraviolet radiation respectively [34]. Examples of DNA damage include mutations, singleand double-strand breaks or interstrand crosslinks [33]. The cell has multiple mechanisms for genome maintenance and error repair, which illustrate the importance of genome stability. Whether the accumulation of DNA damage is a cause or a consequence of aging was long debated, but the description of genome instability in diseases of accelerated aging demonstrated the causality, as extensively reviewed by Niedernhofer et al. [35]. Indeed, in the same publication, they also reviewed the current data showing that mutations increase and DNA repair capacity decreases with age. Apart from genomic instability, but directly related to it, gene expression modifications have a relevant role in aging. It has been widely shown that the expression of protein-coding genes changes with age [36–38]. Importantly, transcription factors and related signalling pathways have also been found to affect cell senescence and aging. The insulin and insulin-like growth factor 1 (IIS) pathway is the best studied one, and it would be commented in the section of deregulated nutrient sensing, along with other relevant systems. The interconnections between the proposed hallmarks of aging become evident, as the modification of transcripts leads to deregulated nutrient sensing. Besides, not only the protein-coding transcripts are affected by aging. Several works have investigated the post-transcriptional regulation of gene expression by microRNAs (miRNAs) and other small non-coding RNAs (sncRNAs). Differentially expressed miRNAs have been found in various organs as well as in the circulatory system of both humans and model animals [39,40]. We also studied the expression of sncRNAs in human leukocytes with age and identified a subset of 69 sncRNAs that gradually increase or decrease. Interestingly, we reported an accelerated change in sncRNA expression between 47-54 years, suggesting that at this age relevant gene expression modifications occur [41]. Notably, works by Borrás and colleagues recently showed that both mRNA and miRNA expression are different when octogenarians and centenarians are compared, and moreover, the data from centenarians are more similar to the ones obtained from adults [26,27]. Furthermore, a longitudinal study performed by Smith-Vikos et al. evaluated the expression of serum miRNAs in 16 subjects and identified differentially expressed miRNAs between the short-lived and long-lived subgroups [42]. There is still a long way to get to understand the function of all sncRNAs in aging, but the available data highlight their implication in the process and their potential use as biomarkers of age-related modifications. 46 | Introduction In addition, in the last years, a new player has entered the game: circular RNA (circRNA). circRNAs are covalently closed transcripts formed through an RNA back-splicing event and characterized by the presence of a back-splicing junction that makes them distinguishable from their linear counterparts [43]. Although the function of most of the circRNAs remains unknown, it has been found that they can act as miRNA sponges and that they are also involved in gene expression regulation, as circRNAs can regulate the transcription of their parental genes. Moreover, ribosome profiling studies have recently shown that circRNAs can be translated both in vitro and in vivo, which challenge the stereotypic view of circRNAs as noncoding RNAs [44]. With regard to aging, several studies have investigated these molecules and differential expression of a large number of circRNAs during aging has been found in a wide range of organisms. In humans, for instance, they have been proposed to play a role in Alzheimer’s disease and immunosenescence [45]. Even if there are still few reports on the functions of circRNAs, this is an emerging field that will continue developing and circRNAs have to be considered as another piece of the complex puzzle of aging. Finally, it is worth mentioning that several investigations have been carried out in the last decades to try to find genetic variants related to healthy aging. Studies conducted on exceptionally long-lived individuals and genome-wide association studies revealed many candidate loci and single nucleotide polymorphisms (SNPs) that could be linked to healthy aging and longevity. However, there are controversial results and most of the candidates identified in some works have not been confirmed in other studies. These differences could be due to different study designs, phenotype definitions and inter-ethnic characteristics, as well as by the effects of epigenetics, environmental factors and lifestyle differences [46]. 1.2.2. Telomere attrition and epigenetic alterations As mentioned before, alterations in the sequence of genomic DNA are common in aging, but there are other relevant modifications that affect the structure and transcription of DNA, which include telomere shortening (or attrition) and epigenetic alterations (Figure 7). The shortening of telomeres is caused by the incomplete replication of the terminal ends of genomic DNA. The replication of chromosomes is conducted by replicative DNA polymerases that lack the capacity to replicate completely the ends of DNA molecules. This function is carried out by a specialized DNA polymerase called telomerase. However, most somatic cells do not express telomerase, leading to a progressive and cumulative loss of telomeres in each replication cycle, and consequently, with increasing age [47]. Telomeres are implicated in essential biological functions: they protect chromosomes from recombination, end-to-end fusion, and recognition as damaged DNA, contribute to the Introduction | 47 functional organization of chromosomes within the nucleus, participate in the regulation of gene expression, and serve as molecular clocks that control the replicative capacity of human cells and their entry into replicative senescence [48]. Defects in telomerase, telomeres or shelterin (the protein complex that protects telomeres), are linked to diverse problems, including pulmonary fibrosis, premature aging and cancer [49]. Besides, it has been found that telomerase-deficient mice exhibit premature aging, which can be reverted by genetically reactivating telomerase [50]. Figure 7. Schematic representation of age-related modifications in DNA structure. Aging affects DNA organization at the chromosome level (telomere attrition) and at the chromatin level (disorganized heterochromatin). In addition, epigenetic alterations, such as methylations or deacetylations can affect both DNA sequences and histones. Adapted from [51]. In humans, the telomere length of leukocytes has been widely studied, and it has been proposed that longer leukocyte telomeres are linked to longevity [52]. However, several authors have investigated the rate of leukocyte telomere attrition and found that telomere length is highly variable at birth, and besides, their shortening is very high during the first years of life, while it slows down considerably during adulthood [53,54]. Moreover, a publication that evaluated the ranking of leukocyte telomere length of four longitudinal studies demonstrated that, despite the interindividual differences in telomere attrition per year, most subjects maintain their classification with respect to their age-matched pairs, meaning that the ones that have shorter telomeres at the age of 30 are the ones that have shorter telomeres one decade later [55]. These results indicate that telomere length is mostly predetermined and environmental or lifestyle changes have only minor effects on telomere attrition. In consequence, the measurement of telomere length early in life is useful for the identification of telomeropathies and as an age-related risk factor, while its utility for intervention monitoring in the elderly is not promising. 48 | Introduction On the other hand, the age-associated epigenetic alterations have been found to be partially reversible. The best described epigenetic modification that occurs during aging are methylations, histone modifications and chromatin remodelling [33]. Regarding DNA methylation, only a small fraction of the CpG sites have shown age-related modifications (around 2%), but this fraction represents between 2 and 3 million cytosines in the genome, denoting the complexity of the system. In addition, both hypoand hypermethylation of the CpG sites happen with age, and the modifications can occur in certain tissues or cell types, or even affect only one part of a specific cell population, adding further complexity [56]. Despite this, robust mathematical methods have been developed and there are reliable algorithms that interpret the methylation pattern of selected CpG sites of an individual and predict chronological age with high accuracy [57]. These tools are called epigenetic clocks, and recent publications indicate that they could also be useful for the detection of accelerated epigenetic aging related to several problems or diseases, including cancer, Alzheimer’s disease, frailty and the prediction of mortality risk [56]. Besides, modifications in histones affect the organization of the DNA, as well as gene transcription. Histones can also be methylated, but the best studied characteristic of these structures is the deacetylation performed mainly by sirtuins. Sirtuins are a family of NADdependent enzymes able to post-translationally deacetylate histones, which is associated with transcription repression. The increased expression of sirtuins has been related to longevity in humans and model organisms, and moreover, overexpression studies in these animal models resulted in elongated lifespan and healthier aging, while the downregulations of sirtuins increased senescence and accelerated aging [58]. The organization of histones influences in a more general view, the packaging of chromatin into heterochromatin and euchromatin. This organization is coordinated by sirtuins and many other DNAand histonemodifying enzymes. The proper assembly of histones ensures a packaged heterochromatin, which is, however, partially lost and redistributed with aging [33]. Interestingly, the age-associated chromatin remodelling deeply influences the transcription of multiple genes, as the coding sequence of generally repressed genes can then be accessible for the binding of transcription factors, or the other way around. Indeed, as explained in the previous section, the gene expression modifications that occur with aging are diverse, and chromatin reorganization is just one of the changes that affect it. Interestingly, epigenetic modifications have been found to play a role in inflammaging and immunosenescence [51,59], two of the problems linked to aging that will be addressed later in this work. Introduction | 49 Finally, it has to be mentioned that, in contrast to telomere attrition, epigenetic changes are reversible. For instance, several investigations have demonstrated that caloric restriction affects DNA methylation, attenuating some age-related CpG modifications and showing deaccelerated epigenetic aging in mice. Besides, the mTOR inhibitor rapamycin also affects methylation and reduce epigenetic age in treated mice [56]. Similarly, other studies evaluating caloric restriction reported sirtuin mediated slower aging and extended lifespan [60]. Again, compounds that mimic the positive effect of caloric restriction regarding sirtuin modulation are being tested, such as resveratrol and curcumin [58]. In summary, the data from animal models indicate that epigenetic modifications are promising targets, and nutritional or pharmacological interventions could potentially be applied to attenuate age-related changes in humans. 1.2.3. Loss of proteostasis Proteostasis is defined as the proper control of proteins, including their biogenesis, folding, trafficking, function and degradation. All the mentioned steps are essential for maintaining the correct functioning of each cell and the organism as a whole. Therefore, the processes implicated in proteostasis are tightly regulated, but they can suffer modifications that lead to misfunctioning during aging [61]. For instance, the previously described genomic instability or epigenetic alterations have a direct impact on protein biosynthesis, as coding sequences can be inaccessible, damaged or inappropriately copied. Besides, even if the biogenesis is completed, many proteins require a specific folding to be functional. Chaperones are a class of heat shock proteins (HSPs) implicated in protein folding and stabilization. The HSPs, and specifically chaperones, have been widely studied in the field of aging. Indeed, chaperones are involved in the response mechanisms against stressors, and have been shown to fail in elders [62]. Furthermore, many studies have been conducted in model organisms, and reports from worms, flies or mice among others have demonstrated the accelerated aging in chaperone deficient animals, while their overexpression elongated lifespan and reduced the accumulation of protein aggregates [62]. The degradation of proteins is another important step to maintain proteostasis. In fact, misfolded, aggregated or non-functional proteins must be removed from the system to prevent the accumulation of toxicity. There are two major pathways for protein degradation: the ubiquitin-proteasome system (UPS) and autophagy. Both pathways have hundreds of different components, including chaperones, and this complexity demonstrates the investment of the cell on proper protein degradation [61]. However, the function of UPS and autophagy declines with age, and despite the defect of one of the pathways can be partially compensated by the other, many cell types accumulate defective proteins [63]. The protein degradation, in 50 | Introduction combination with the previously mentioned defects in biosynthesis and folding, are responsible for the age-related loss of proteostasis. Indeed, the deficits of each step contribute to the final failure of the system. 1.2.4. Deregulated nutrient sensing As presented before, the hallmark describing deregulated nutrient sensing in aging is tightly connected to the gene expression modifications. In this sense, the genetic polymorphisms or mutations that reduce the function of the IIS pathway have been associated with longevity. These include the growth hormone, insulin-like growth factor-1 (IGF-1) receptor, or downstream effectors such as AKT, mTOR and FOXO [64–66]. However, the components of the IIS pathway are multiple and its regulation and interconnections with other signalling pathways are complex. Indeed, apart from being related to longevity, reduced levels of the IIS components are also reported during normal aging and in animal models of premature aging [67]. This could seem contradictory, but it has been proposed that depending on the duration and extent of downregulation the elicited results could be beneficial or deleterious. Thus, the constitutively decreased IIS functioning implies lower cell growth and metabolism, and consequently reduced rates of cellular damage, while acute decreases or extremely low levels lead to premature aging [33]. This process is comparable to other defensive responses that can become deleterious when not properly controlled, as in the case of inflammatory responses. Other nutrient sensing systems tightly connected to IIS and also associated with aging include mTOR and sirtuins. The mTOR kinases are implicated in anabolic metabolism and the genetic as well as pharmacologic attenuation with rapamycin of this system have been linked to increased longevity in distinct animal models [68,69]. On the other hand, as described in the section of epigenetic alterations, sirtuins are enzymes implicated in the organization of histones and they play a role in the age-associated transcriptional regulation. Notably, sirtuins also respond to nutrient availability, and they get activated under nutrient scarcity. We previously commented the relation between the elevated expression of sirtuins and longevity, and the positive results obtained with caloric restriction or treatments with resveratrol or curcumin and associated with sirtuins [58,60]. Similarly, apart from the pharmacological interventions, the beneficial results of caloric restriction have also been found to be mediated, at least in part, by the reduction of mTOR activity. This link has been demonstrated in animal models under caloric restriction, in which the downregulation or deletion of mTOR genes prevented the otherwise observed extension of lifespan [70]. Introduction | 51 1.2.5. Mitochondrial dysfunction The wellbeing of mitochondria is essential for the appropriate functioning of cells. Despite this, there are hundreds to thousands of mitochondria in a single cell (the number depends on the organism, tissue and cell type), and thus, single or small numbers of mutations or deficits can be managed by a cell and maintain proper functioning [71]. However, as we age, the increased damage, reduced respiratory chain functioning, imbalanced fusion and fission, and defective clearance of mitochondria (mitophagy) contribute to cellular and organismal aging [72]. In the last years, the potentially beneficial effects of mild deficiencies in mitochondria have been proposed. Indeed, the elicited mitochondrial defensive response and the low energy state have been found to induce beneficial compensatory responses, and even to extend lifespan in model organisms [33]. This could seem paradoxical, but it is in line with other age-related characteristics, such as cell senescence or inflammation, that could be beneficial when effectively controlled, but detrimental when maintained or abnormally increased. Mitochondrial dysfunction is closely related to other characteristics of aging. The case of genomic instability becomes evident, as mutations in mitochondrial DNA (mtDNA) are one of the main causes of the severe impairment on energy conversion. Even if only around 1% of the mitochondrial proteome is encoded by mtDNA, these include critical components of the oxidative phosphorylation complexes [71]. And besides, apart from the mutations or deletions in mtDNA, the accumulation of changes in nuclear sequences, also affect the components and dynamics of mitochondria. In addition to the defects in mtDNA that accumulate during the organismal life, it has been demonstrated that single SNPs and mitochondrial haplogroups can influence the aging process. For instance, SNPs in mitochondrial uncoupling protein genes have been related to healthy aging [73], the individuals with mitochondrial H haplogroup showed distinct mitochondrial dynamics [74,75] and the D4, D5 and J haplogroups have been associated with longevity. Moreover, studies performed in mice have shown that the mtDNA haplotype profoundly influences mitochondrial proteostasis and function, as well as ROS generation, insulin signalling and telomere length, resulting in differences in the aging process and median lifespan between conplastic strains [76]. With regard to senescence, the issue of whether mitochondrial dysfunction is causative or just part of the consequences of the entrance of a cell in senescence is still debated. However, the bidirectional link between senescence and mitochondria could be too simple to explain the complex underlying processes, and this interplay could be best outlined as a vicious circle, involving a number of feedback loops between the players. In spite of the triggering mechanisms, it has been widely described that senescent cells have mitochondrial deficits both in genome and proteome maintenance, and consequently present dynamic changes and 52 | Introduction dysfunctions [77]. On the other hand, the age-associated problems in mitochondria can also affect stem cells. The accumulation of somatic mtDNA mutations alter stem cell homeostasis and can induce defects such as imbalanced biogenesis, abnormal mitophagy or increased ROS production, which can eventually accelerate stem cell senescence [78,79]. There is also a tight connexion between mitochondrial deficits and the chronic inflammatory state (inflammaging) with aging. The defective or dysfunctional mitochondria can release molecules that promote the activation of the immune system, including mtDNA and ROS among others. The mitochondrial components can boost the immune response through different pathways, as the activation of the inflammasome, the recognition by the cytosolic sensor of dsDNA cyclic GMP-AMP synthase, or the activation of immune cells with secreted metabolites like succinate, as comprehensively reviewed by Jang and collaborators [72]. Remarkably, the contribution of mitochondrial component to inflammaging can result in further injury, as in the case of inflammasome and caspase-1 mediated mitochondrial damage [80]. 1.2.6. Cellular senescence More than half a century ago, the senescence of human cells was described in in vitro cultures of fibroblasts [81]. Hayflick and Moorehead found that after a certain number of passages cultured cells lose proliferative capacity. Besides, it was long described that cellular senescence happens also in vivo [82]. Furthermore, thanks to all the investigations that have been conducted, nowadays we know that apart from the replicative arrest, senescent cells show many other features distinct from non-senescent cells. Some of the features associated with senescent cells, and commonly used as senescence biomarkers, include the increased activity of lysosomal -galactosidase, as well as the elevated expression of p53 and p16INK4a [83,84]. Interestingly, senescence can be induced as a controlled mechanism to prevent the proliferation of damaged cells, before they lose the replicative control and develop tumorigenic features. Similarly, the entrance of cells into senescence can be triggered by telomere attrition, accumulation of DNA damage, or ROS, which induce the activation of the DNA damage response mechanism and lead to senescence mainly through the p53 pathway [85,86]. Thus, these processes are linked to the mentioned characteristic expression of p53 and p16INK4a by senescent cells. Indeed, they are tumour suppressor proteins, part of complex signalling pathways that respond to the expression of oncogenes, and consequently induce cellular senescence or apoptosis [87]. This indicates the protective role to prevent the formation of tumours, even if contributing to the accumulation of senescent cells. In consequence, the positive or negative impact of senescent cells is still discussed. Introduction | 53 The induction of senescence in damaged and potentially hazardous cells is undoubtedly beneficial. In addition, senescence has been proposed to trigger tissue renewal, but in contrast, this process may not be efficiently completed in aged tissues or pathological contexts, resulting in the accumulation of senescent cells (Figure 8) [86]. Therefore, increasing evidence indicates that both pro-senescent and antisenescent therapies can be favourable. For example, in cancer, during active tissue repair and even to prevent age-related damage, controlled pro-senescent therapies could be beneficial, limiting proliferation, accumulation of defective cells and fibrosis [88]. Conversely, antisenescent therapies may help to eliminate the already accumulated senescent cells and to recover tissue function in aged individuals [89]. Figure 8. Proposed model of senescence. Senescence initiates a tissue remodelling process by recruiting immune cells through the senescence-associated secretory phenotype (SASP, explained in the section 1.2.8). Macrophages clear the senescent cells, and progenitor cells regenerate the damaged tissue. This sequence is impaired under persistent damage, pathological states or aging. In these cases, senescent cells are not efficiently cleared, the tissue is not fully regenerated, and its functionality diminishes. Resolution of the damage in these cases involves a fibrotic scar with senescent cells, inflammatory cells and fibrotic tissue. Adapted from [86]. Moreover, even if some molecular features –like replication arrest or elevated -galactosidase activity– are reproduced in most senescent cells, senescence affects in a different manner to distinct animals, individuals, tissues and cell types [89–91]. In this work, we have focused on the senescence of the human immune system, and consequently, in the next lines the alterations of this system are presented. 1.2.6.1. Immunosenescence The term immunosenescence is used to refer to all the changes that occur to the immune system during aging leading to its dysfunction. However, immunosenescence is not only caused 60 | Introduction mediators such as tumour necrosis factor alpha (TNF-) and C-reactive protein (CRP) have also been investigated by many authors, and their concentration have also been found to be elevated in elders [143–145]. Figure 11. The benefits of maintaining balance. The importance of keeping a functional immune system and a balanced inflammatory state has been shown be related with healthy aging and longevity. On the other hand, when the proinflammatory molecules accumulate and the balance is lost during aging, chronic low-grade inflammation or inflammaging develops. However, with regard to the roles of cytokines, we should always bear in main the complexity of their functions. Actually, molecules such as the mentioned IL-6 and TNF- have proinflammatory but also anti-inflammatory functions [146–150]. Their effect depends on the surrounding environment, on the membrane molecules expressed in receptor cells and on the signalling cascaded they elicit. Thus, even if IL-6 and TNF- are in most of the cases indicators of inflammaging among the old individuals, in some cases there could be other underlying processes and the presence of these molecules could be beneficial [95]. 1.3. Molecular biomarkers of frailty When defining frailty and the main characteristics of the people affected by this syndrome (introduction section 1.1), we commented about the complexity of its identification. Due to the lack of consensus on an operational definition of frailty, many different tests are employed nowadays at primary care services. The physical, cognitive or psychological characteristics that the frailty scales measure, are considered functional biomarkers [151,152]. In this sense, the aim of molecular biomarkers is to complement the already applied tests to help the identification of frail individuals. Introduction | 61 Many efforts are being made to try to identify and validate molecular biomarkers of frailty, and even if some have been proposed, to our knowledge none of them are applied in the clinic. In the search of frailty biomarkers, several studies have investigated molecules implicated in the biology of aging and, therefore, included in the hallmarks of aging, aiming to find differences between healthy aging and frailty. In the next, lines we will comment some of the biomarkers that have been suggested to be implicated. A graphical representation of the biological characteristics and biomarkers of frailty is shown in Figure 12. Figure 12. Schematic representation of the main biological processes, their interconnections, the implicated molecules and their relation to frailty syndrome [153]. The current knowledge points, for instance, to a role of oxidative stress in the development of frailty. In this sense, most works have shown increases in oxidative damage indicators, while reduced levels in antioxidant micronutrients, in frail individuals, which gave rise to the recently proposed free radical theory of frailty [154]. This theory suggests a change in the focus of oxidative stress, as diverse studies showed that oxidative damage does not correlate with chronological age, but rather with frailty. One of the oxidative stress biomarkers of frailty is the elevated levels of circulating protein carbonyls, a well-established indicator of protein oxidative damage. For example, high protein carbonylation correlated to poor grip strength, particularly among older women [155]. Besides, low levels of circulating antioxidants like vitamin E have been found to be associated with frailty [156,157]. However, other authors did not find differences between vitamin E levels and frailty [158], and the fact that vitamin levels can be easily altered with diet or during 62 | Introduction disease, complicates even more the interpretation of results. Another vitamin that has been studied in many diseases and processes, including frailty, is vitamin D. Lower levels of this hormone have been linked to frailty, but the normal ranges of vitamin D are highly variable depending on the geographical area and among seasons, and the effects of long-term supplementation with vitamin D are still controversial [3,159,160]. The levels of other hormones have also been investigated as potential biomarkers. Most of the endocrine markers proposed in the context of frailty are those related to the decline in muscle mass and function. During aging, there is a progressive switch from anabolic to catabolic metabolism that affects muscle proteostasis, which has been related to variations in certain hormone levels. Indeed, dehydroepiandrosterone sulfate (DHEAS) is an important regulator of muscle mass and strength that decreases with age and it has been related to sarcopenia [161]. Besides, DHEAS stimulates the production of IGF-1, which is required to muscle regeneration. Some publications have reported lower levels of DHEAS in frail subjects and improved physical function with DHEAS supplementation and exercise [162,163]. In contrast, other authors did not find significant correlations between frailty and the levels of testosterone, DHEAS or IGF-1 individually, while the accumulation of multiple anabolic deficiencies was a good predictor if frailty [164]. Another characteristic generally associated with aging and widely investigated in frailty is metabolic imbalance, and specially glucose and insulin dysregulation. Elevated basal levels of glucose and insulin, insulin resistance and abnormal insulin-glucose dynamics have been related to higher rates of baseline frailty and greater odds of frailty onset [153,165–167]. Notably, it should be mentioned that even if most of the works found some alterations, not all of them obtained the same results. For example, the basal levels of glucose were reported to be elevated in frail subject in some works, while they were not significantly different from nonfrails in other publications. In addition, as epidemiological studies indicate that diabetes is a risk factor for developing frailty, and some frail elders without diabetes have elevated levels of glucose, the question of whether this imbalance could be a cause or a consequence of frailty remains open. The link between chronic inflammation and frailty has been extensively investigated. The concentration of inflammatory mediators in circulation has been measured in many different cohorts aiming to test whether proinflammatory molecules are specially increased in frail individuals when compared to robusts. Certainly, an elevated concentration of IL-6, TNF- and CRP, among others, have been reported in most of the cohorts in frail elders [144,168–172], but there are also some studies that did not find significant differences between robust and frail individuals [159,173]. Moreover, these three molecules are increased in a vast range of Introduction | 63 inflammatory or infectious conditions, so they could not be used as a single measure, and should be combined with other biomarkers that provide information about additional variables related to frailty, such as muscle loss or bone degeneration [174]. Also, the coagulation activity is related to inflammation. Indeed, hypercoagulability both reflects and contributes to enhanced inflammation [153]. Hypercoagulability is generally observed during aging, and elevated levels of coagulation markers, such as fibrinogen, factor VIII, D-dimer and tissue plasminogen activator have also been linked to higher rates of frailty [165,173,175]. However, similar to the previously mentioned biomarkers that have been proposed for frailty, not all the authors obtained the same results. For instance, elevated fibrinogen was related to a higher risk of frailty in women and men by Walston and collaborators [165], while it was only associated with frailty risk in women by Gale and coworkers [175], and the study performed only in women by Reiner et al. found no associations between fibrinogen and the risk of incident frailty [173]. Another remarkable source of frailty biomarkers is linked to brain changes. This organ is markedly affected by aging and indeed, the incidence of many brain diseases increases notably in elders. A study by Buchman and collaborators followed nearly 800 aged people and showed that frailty progresses with age, and an accelerated decline was reported in the participants that were found to have brain pathologies in the postmortem evaluation [176]. In an attempt to identify easily measurable brain biomarkers, reduced cerebellar grey matter volume assessed by magnetic resonance imaging have been found in frail elders when compared to robusts [177]. Similarly, the neuroprotective brain-derived neurotrophic factor (BDNF), which protects adult neurons from death during stress and promotes the development of immature neurons, can be measured in plasma, and decreased levels of this protein were linked to higher rates of frailty in women. Moreover, a physical intervention elevated the concentration of BDNF both in robust and pre-frail participants, suggesting its implication in the pathophysiology of frailty [178]. In a different approach, alterations at the genetic level are evaluated. This is the case of telomere length, epigenetic changes and gene expression modifications or even posttranscriptional regulation. With regard to telomere length, many authors have investigated whether there are associations between shorter telomers and frailty syndrome. A systematic review and meta-analysis that was published recently, identified 155 publications on this topic [179]. Interestingly, they selected 9 studies that measured telomere length in leukocytes and concluded that, in accordance with previous reports, telomere length might not be a meaningful biomarker of frailty. In fact, they reported no significant differences in 8 of the selected studies and found only shorter telomeres in the study that was performed in Hispanic 64 | Introduction individuals. As discussed in one of the works that investigated the association between telomere length and frailty, even if some aspects that are related to frailty, such as oxidative stress or inflammation contribute to telomere shortening, they may not represent the predominant factors influencing the complex and multicomponent syndrome of frailty [180]. The regulation of gene expression by epigenetics was already commented to be implicated in the biology of aging. Some authors are also working on the question of whether epigenetic modifications, such as DNA methylation influences the incidence of frailty. The investigations performed on the DNA methylation patterns indicate that frailty could be related to accelerated epigenetic aging [181,182], and even specific differences could be observed between twins with a distinct frailty index [183]. However, the cost and complexity of DNA methylation pattern studies in comparison to directed gene expression analyses should be taken into account to evaluate the applicability of these methods in the clinics. Another study approach that focuses on the factors that influence gene expression and function is the identification of SNPs associated with frailty. In this sense, polymorphisms in genes involved in inflammation, muscle biogenesis or apoptosis regulation among others have been related to frailty [7,184]. Notably, the association between gene expression and frailty is probably one of the most widely investigated features, as it englobes all the cellular pathways as well as the studies that measure the expression of thousands of transcripts or only a single one. In consequence, there are hundreds of publications that investigated the expression of certain genes or pathways, and now also ‘omics’-based approaches are being developed [185,186]. Recently, a comprehensive review of the biomarkers of frailty was published, in which genes but also proteins and secreted factors related to aging were included [187]. They differentiated the biomarkers in seven categories: inflammation, mitochondria and apoptosis, calcium homeostasis, fibrosis, neuromuscular junction and neurons, cytoskeleton and hormones, and other principles. The authors also classified the biomarkers depending on their priority, with highest scores being attributed to the factors associated with frailty and with more than one hallmark of aging, and with a considerable amount of evidence that the marker is not equally expressed in frail versus non-frail individuals. Finally, they proposed a panel of frailty biomarkers composed of 19 high priority, plus 22 medium priority and 3 low priority markers. Most were proteins or genes, but other emerging biomarker candidates such as miRNAs and microparticles were also included (Figure 13). Importantly, the investigation of the emerging biomarkers of frailty continues to develop, as illustrated by the publication of a work focused on miRNAs nearly at the same time that the commented review [188]. Besides, as mentioned before, the case of microparticles (or extracellular vesicles) in age-related processes and also Introduction | 65 in frailty is of central interest in our work, and therefore, the following section is dedicated to these particles (introduction section 2). Figure 13. The proposed biomarkers of frailty. The panel is composed of a core of high priority factors and complemented by medium and low priority markers [187]. 66 | Introduction 2. Extracellular vesicles EVs are membrane-coated particles secreted by almost all cell types. Their first identification was already reported in 1946, as procoagulant platelet-derived particles in normal plasma [189] and more than 20 years later, in 1967, they were referred as “platelet-dust” [190]. Since then, several publications started to report novel particle sources and functions and by the end of the 20th century they were already known to play a role in relevant processes, such as antigen presentation [191]. Importantly, at the beginning of the present century, the research on EVs gained interest among the scientific community, as they were also found to the implicated in other central issues, including the immune system mediated antitumor response [192], and due to the discovery that EVs transfer mRNAs and miRNAs from the donor cell that can induce functional changes in recipient cells [193]. In the last decade, thousands of works have continued describing the characteristics, functions and implications of EVs in intercellular communication. Thanks to all of them, we can now state that EVs are important players in most biological processes. However, as it usually happens in scientific research, the more we know, the more complex the picture is, and the more we need to investigate to understand the molecular processes that govern ourselves and the rest of living organisms. 2.1. Biological characteristics of extracellular vesicles The term EVs is used to refer to all the particles that cells secrete to the extracellular media. There are two main categories of EVs: exosomes and microvesicles. Besides, apoptotic bodies are also considered EVs. Indeed, apoptotic bodies play an essential role in the proper clearance of the dying cell as well as for the signalling of this programmed cell death to surrounding cells and for the regeneration of the tissue [194]. However, most of the works studying EVs are focused on exosomes and microvesicles, due to their multiple functions and implications. Exosomes are secreted particles originated by the fusion of a multivesicular body and the plasma membrane, while microvesicles are formed by the direct budding and fission of the plasma membrane. Moreover, apart from their distinct biogenesis, exosomes and microvesicles have also classically been differentiated based on their size. Exosomes were defined to be around 50-100 nm in diameter, and microvesicles from 100 nm up to 1 m [195]. However, even if this classification was formerly accepted, nowadays we know that there are larger particles originated at multivesicular bodies, as well as smaller vesicles that evaginate from the plasma membrane. Consequently, the International Society of Extracellular Vesicles (ISEV) discourages the use of these terms if the biogenesis pathway of the vesicles is not Introduction | 67 known, and recommends the use of EVs or just small, medium or large EVs if we want to refer to their size [196]. With regard to the molecules carried by EVs, we have to consider both their membrane and inner cargo. The membrane of EVs consists mainly of proteins and lipids, but each EV has distinct types of proteins and lipids depending on their origin and function. Furthermore, the composition of the EV membrane influences the fate and internalization by recipient cells [197]. The components of the EV lumen are even more diverse and include proteins and many different nucleic acids. Apart from the above-mentioned mRNA and miRNAs, EVs carry other types of small and long ncRNAs, circRNAs and dsDNA fragments [198,199] (Figure 14). Importantly, the investigations about EV secretion and their cargo revealed more than a decade ago, that the sorting of components into a forming particle is a controlled mechanism and not a random packaging of the available molecules in the secreting cell [193,200]. Similarly, the uptake of EVs is thought to be a controlled process. Many authors have studied the binding and internalization of EVs by recipient cells and multiple molecules, such as tetraspanins, integrins, lipids and lectins, have been identified to mediate the uptake. Besides, the integration of EVs can be performed by the fusion of the EV and cellular membranes, or by distinct endocytic pathways (Figure 14). An extensive and complete review on the biogenesis, release and targeting mechanisms of EVs was recently published by Niel and co-authors, and it is a recommended read to go into this subject more in depth [197]. Figure 14. Simplified representation of the secretion (microvesicles and exosomes) and uptake of EVs. For the internalization, EVs can (1) dock and (2) fuse with the plasma membrane, or (3) get endocytosed and (4) eventually fuse with a membrane of the endocytic compartment [195]. 68 | Introduction Due to the focus of this work, we present here the main characteristics and functions of EVs from human cells, but the intercellular communication mediated by EVs is also present in many other organisms. Furthermore, it has been found that unicellular organisms like bacteria also secrete EVs. Interestingly, the bacteria produce EVs for multiple purposes, including horizontal transfer between intra-species cells, stress response, or biofilm formation. In addition, in mammals, the communication between host and bacterial cells of the microbiota is mediated, at least in part, by EVs [201]. 2.2. Isolation and characterization of extracellular vesicles The first step to take into consideration to obtain EVs is to decide the sample from which we want to isolate the particles, and to perform a proper collection, handling and storage. EVs can be isolated from biological fluids, including blood, urine or cerebrospinal fluid and from cell culture media [202–205]. Depending on the selected source, specific recommendations have been proposed [206]. However, there are many variables that can influence EV secretion and that cannot be completely controlled. For example, when taking blood samples, it has been described that age, sex, diet, infections, treatments or even circadian variations can affect the EVs in circulation [196]. Besides, as we will describe below, there are plenty of EV isolation methods and the choice would depend on the sample characteristics, study objective and available techniques. Thus, it is essential to collect and report all the possible information about the donors, samples and applied steps so that we can take into consideration all our variables, and also to enable the potential replication by other authors. Regarding EV isolation, it is important to mention that in most of the cases, if not in all of them, it is not possible to achieve a complete separation of the vesicles of interest. Therefore, we have to consider that even if the term isolation is commonly applied, we are probably enriching our samples for EVs. Moreover, this issue is not exclusive for EVs, as other techniques also used for cells, such as sorting, precipitation or immunocapture present good but not perfect yields. The methods for EV separation are diverse, and besides, each technique can have distinct settings depending on the subtype of EVs aimed to enrich. For instance, differential centrifugation is one of the most widely applied methods, but the centrifugation sequences, forces and times vary among studies. The first steps are usually similar, with centrifugations at low centrifugal forces (< 10,000 g) to pellet cells and debris. Then, some investigators apply middle force centrifugations (15,000-30,000 g) and recover the EVs from the pellet, while the ones that focus on small EVs take the supernatant and perform high speed centrifugations, or ultracentrifugations (usually 100,000-200,000 g). Besides, there are authors that complement differential centrifugations with density gradient centrifugations [206,207]. Introduction | 69 Ultracentrifugation has been one of the most used methods, but some authors have reported that it can coprecipitate protein aggregates or viruses and can even induce EV clumping and damage [206,208]. There are other classically applied techniques for EV isolation that include size exclusion chromatography, precipitation, filtration and immunocapture. For size exclusion and filtration, the pore size of the matrix and of the membrane, respectively, can be selected. In addition, in the last years, a different filtration method has been introduced: tangential flow filtration (TFF). In contrast to the common filtrations that pressure the sample perpendicularly to the filter, TFF consists on the application of a tangential force, which minimizes pressure and enables the recirculation of the sample into the system. Besides, the pore size of the membranes applied for TFF can also be chosen depending on the desired EVs. The use of TFF is particularly beneficial when large volumes of samples are handled, as cell culture media or urine [209,210]. However, TFF can only separate the EVs based on their diameter. On the other hand, the immunocapture methods are attractive when a specific subpopulation of EVs want to be separated. This system is based on the use of immobilized antibodies that recognize and bind EV-specific molecules, usually proteins exposed at their membrane. The selected antibodies can be immobilized on a plate, a chip or a magnetic bead, and there are many commercial kits available [207]. Nevertheless, when using immunocapture protocols unwanted soluble ligands can also be recovered, or part of the desired EVs lost if there are more ligands than antibodies available. In addition, immunocapture will always separate a subpopulation of EVs, as no universal marker has been found. For this reason, as an example, we cannot claim to isolate all exosomes from a complex sample when applying an anti-CD63 antibody because not all the exosomes are positive for this tetraspanin. To illustrate the complexity of EVs, a representation of some of the most common molecules identified to be carried by EVs can be seen in Figure 15. Interestingly, new methods are being developed for the isolation of EVs. The microfluidic and acoustic settings are promising techniques, and even the combination of both of them have been shown to be effective to isolate EVs from whole blood [211]. In any case, the election of the EV separation method (or combination of methods) is strongly influenced by the objective of the study, as well as by the required time, costs and applicability, if it is directed for a potential clinical application [202]. With regard to the characterization of EVs, and despite their small size, there are multiple methods available: the ones that could be applied to characterize cells that have been adjusted for EVs, and the ones that have been specially developed for EVs. On one side, we can describe 76 | Introduction 3. Multiple sclerosis MS is a chronic autoimmune disease of the CNS, characterized by pathologic demyelination of axons and subsequent neurodegeneration. It is a heterogeneous disease and, clinically, it can follow relapsing-remitting or progressive forms. The relapsing-remitting forms are characterized by outbreaks of neurological disability symptoms lasting at least 24h (relapses) followed by recovery periods (remissions) in which symptoms improve partially or completely. This is the most common disease course at the time of diagnosis, with approximately 85% of patients initially diagnosed with a relapsing-remitting form of MS. In the progressive forms, the disease develops steadily and results in a rapid accumulation of disability. Approximately 50% of patients with relapsing-remitting forms, convert to a secondary progressive phase within 10 years of disease onset [266]. 3.1. Etiopathology of multiple sclerosis MS is a complex disease, and its etiology is not completely understood. It has been found that a combination of genetic, epigenetic and environmental factors, increase the risk of developing MS. Among the genetic factors, the HLA-DRB1*15:01 allele in the MHC class II is the earliest, and most dominant risk factor identified, while cigarette smoking, higher latitudes, low sun exposure, low vitamin D levels and Epstein-Barr virus infection are the principal environmental risk factors. Besides, in the last years, the implication of the microbiome is being investigated, as several studies have shown its influence on the immune system regulation, and some differences in the gut microbiome between MS patients and healthy controls have also been reported [267]. The pathological process of MS is initiated by an inflammatory process mediated by autoreactive T cells. The trigger of the autoimmune attack is thought to be an autoantigen, but it has not been identified yet. The autoreactive T cells get activated in the periphery, start to produce proinflammatory molecules and to express adhesion molecules that favour their attachment to endothelial cells of the BBB. In addition, the BBB is usually damaged in MS patients and, as a result, the autoreactive cells are able to first firmly adhere to endothelial cells, and then migrate through the BBB into the CNS [268]. Once in the CNS, T cells are reactivated by astrocytes or microglia, inflammation spreads and finally, effector T cells damage the myelin sheath of axons, and macrophages and glial cells participate in the digestion process. As a result, there is an impaired isolation of axons and abnormally slow action potential transmission [269]. An illustration of this process is presented in Figure 17. Introduction | 77 Figure 17. Proposed mechanisms and implicated cells and molecules in demyelination and remyelination processes [269]. In the first stages of MS, the neurologic function is partially or completely restored after a demyelinating event. This process is mediated by oligodendrocyte precursor cells that get activated, migrate to the lesion, proliferate and differentiate to mature myelin-producing oligodendrocytes. However, the newly produced myelin sheath is usually thinner than the original, and besides, the regenerative process becomes less efficient with increasing age. Consequently, the pathologic autoimmune attacks can result in a axonal degeneration and subsequent neurodegeneration, which affects the neurologic function of MS patients and increases disability [270]. 78 | Introduction 3.2. Multiple sclerosis and premature aging Most of MS patients experience the first symptoms at their 20s or 30s, but it should be mentioned that there are also paediatric or juvenile [271] and late-onset MS cases [272]. Importantly, in the last decades, effective disease-modifying treatments that slow the progression of MS have been developed [267]. Thanks to the beneficial effects of these treatments, MS patients have a slower rate of disability accumulation and thus, a better quality of life than MS patients of previous generations [273]. In consequence, and following the same trend as the general population, the life expectancy of MS patients is increasing, and with it, the mean age of MS patients is also getting higher. The reports of the MSBase registry [274] show that already more than 20% of MS patients are aged ≥60 years (msbase.org). Therefore, an elevated number of patients suffer from the interactions between the MS disease and the aging process. However, in most of the cases, it is not possible to distinguish between the characteristics of MS and aging in a patient of advanced age. This is due to similarities between the typical features of the two processes, which include cognitive and cardiovascular problems, bowel and bladder dysfunction, or reduced mobility, among others. Of course, not all the MS patients or elders present these problems, but they are common in the two cases. The similarities between MS and aging are also reported at the biological level, as immune system exhaustion and chronic inflammation occur in both processes [275]. Furthermore, it would not the possible to discriminate between the causes of each feature, as the organism has to be understood as a whole entity, in which the dysfunctions accumulate and can influence the other systems (like explained for the biologic hallmarks of aging). In a different approach, the possible development of premature aging in patients with autoimmune diseases like MS, type 1 diabetes or rheumatoid arthritis has been proposed [276– 279]. Particularly for MS patients, the chronic and intense implication of the immune system during MS pathology, as well as the effects of immunomodulatory drugs prescribed, have been suggested to promote the premature exhaustion of the immune system [279]. Indeed, some of these works also alluded to the possibility of an inverse relation, with increased risk of developing autoimmune diseases under premature immunosenescence. With regard to immunosenescence, some works reported increased levels of CD4+CD28T cells [280,281], thymic involution [282], altered T cell homeostasis [283] and disturbed regulatory T cell development and function [284] in adult MS patients. In relation to inflammation, elevated levels of TNF- and IL-6 among other inflammatory markers have been found in the cerebrospinal fluid and serum of MS patients during remission, indicating that Introduction | 79 some signs of chronic inflammation, similar to the ones observed the age-associated inflammaging could be present [285,286]. These publications point to the possible premature immune decline in MS. In contrast, there are other works that did not find immunosenescent features in MS patients, or that reported differences in immune characteristics depending on the immunomodulatory drug received by the MS patient [277,287,288]. In any case, the link between premature aging and MS should be further investigated to elucidate whether there is a causative relation, and if such, which steps could be taken to prevent or appropriately treat this situation. JUSTIFICATION Justification | 83 One of the major concerns of our society is the aging of the population. The rapid increase of life expectancy and the consequent rising incidence of age-associated diseases and dependency have made us aware of the interdisciplinary challenge we face. Many efforts are conducted in numerous fields to manage the socioeconomic impact of aging and we, as part of the scientific community, are working to describe and understand this complex process from the biological, biochemical and biomedical point of view. In this context, the present project was outlined in 2015, aiming to advance knowledge and contribute to the ultimate goal of improving the quality of life during the natural process of aging. HYPOTHESIS AND OBJECTIVES 92 | Chapter one Common methods Participants and Frailty classification For the present study, samples from 3 different cohorts of elder donors and samples from healthy adults were used. We obtained the samples of the elder cohorts in collaboration with the Primary Care Unit of Biodonostia Health Research Institute and the Neurology department of Donostia University Hospital. Participants are from the province of Gipuzkoa (Basque Country, Spain) and meet the criteria shown in Table 1. Table 1. Inclusion criteria of the cohorts. All participants completed a questionnaire and donors with acute illness were excluded. The study was approved by the hospital’s ethics committee and all participants provided written informed consent before blood sampling. Frailty status of elder participants was assessed by primary care services. A battery of tests was conducted. The translated version of frailty tests was applied. A short description of frailty tests is shown in Table 2 and the main characteristics of each cohort in Table 3. Cohort Description Participants from Errenteria and Pasaia Survivors of a previous study were contacted and invited to participate No further inclusion criteria Samples obtained July 2014 – May 2015 Participants form Errenteria, Irun and Hondarribia Aged 70 or over, community-dwelling and autonomous (Barthel > 90) Samples obtained May 2015 – July 2016 Participants from Getaria, Urnieta, Zumaia and Zestoa Aged 70 or over, community-dwelling, including autonmomus and non-autonomous Samples obtained August 2016 – May 2017 Cohort 1 Cohort 2 Cohort 3 Chapter one | 93 Table 2. Short description of the tests applied to measure frailty. Frailty assessment test Description Tilburg Frailty Indicator (TFI) [15] A user-friendly questionnaire based on a multidimensional approach. It is composed of a physical, a psychological and a social domain. Gait speed (GS) [17] Expressed in meters per second (m/sec). Participants were asked to walk at their usual pace. The test was performed twice and GS was calculated based on the shorter time. Timed up-and-go (TUG) [18] The time needed to stand up from a chair, walk 3 meters, turn around, walk back and sit down, with the help of their usual walking aid, if any. Short Physical Performance Battery (SPPB) [19] A functional capacity test composed of gait speed, test of balance and time needed to stand up from a chair 5 consecutive times. Gerontopole Frailty Screening Tool (GFST) [16] Based on clinical judgement. 6 yes/no questions that help the physician to evaluate the existence of frailty. Barthel Index (Barthel) [291] A multiparametric test measuring the performance in activities of daily living and mobility. Table 3. Information of study participants. Adults Cohort 1 Cohort 2 Cohort 3 - Participants 53 295 91 57 Female/Male 30/23 153/142 56/32 30/27 Age (mean) 79-92 (82.51) 71-91 (79.83) 70-96 (76.98) 24-46 (33.51) Frailty assesment - TFI X X GS X X X TUG X X X SPPB X X GFST X Barthel X X X Elders 94 | Chapter one Blood sampling Peripheral blood was collected by experienced nurses by venipuncture with a 21-gage needle in 8 ml serum separator tubes and 4 ml EDTA tubes (Vacutainer, BD Biosciences) and directly deposited in the Basque Biobank for their processing and storage. Serum separator tubes were allowed to clot for 30 min and centrifuged at 1258 g for 20 min to recover serum from the supernatant. EDTA tubes were kept upright and centrifuged at 1258 g for 20 min to recover plasma. To obtain RNA, samples from EDTA tubes were incubated with Buffer EL (Qiagen) for erythrocyte lysis and then RNA from leukocytes was extracted with QIAamp RNA Blood Mini Kit (Qiagen) following manufacturer’s instructions. For DNA, a second EDTA tube was used and the extraction was performed with FlexiGene DNA Kit (Qiagen) following the manufacturer’s instructions. RNA and DNA quantity and quality were assessed with a Nanodrop 1000 spectrophotometer (Thermo Fisher). The obtained serum, plasma, RNA and DNA samples were aliquoted and stored at -80 °C. When needed, corresponding request forms were fulfilled to obtain the samples from the Basque Biobank. Chapter one | 95 Inflammation Introduction Chronic low-grade inflammation is one of the best described characteristics of aging. It has been widely shown that elders present elevated levels of inflammatory markers in circulating blood, in the absence of overt infection [137,138,292]. This accumulation of proinflammatory molecules, termed inflammaging, has also been proposed as a potential biomarker of frailty. Previous studies have measured the concentration of proinflammatory markers such as IL-6, TNF- and CRP among others in donors with different degrees of dependency. However, the obtained results are controversial, and even if some researchers have found an increased concentration in frail and non-autonomous elders, other studies did not report significant differences [159,169,171,174]. Other molecule related to inflammation and proposed as a frailty biomarker is albumin. The rate of albumin synthesis is affected by both nutrition and inflammation, and inflammation alone is associated with a greater catabolic rate of albumin. Decreased albumin levels have been proposed as a risk factor for frailty, but similar to the above-mentioned inflammatory markers, there is no consensus on its validity [293,294]. The aim of our study is to investigate the validity of inflammatory mediators as biomarkers that could complement the functional and clinical evaluation of elders for the identification of frailty. To that end, we first compared the concentration of the above cited molecules between adults and elder donors, and then, based on the frailty classification of elders, evaluated whether these molecules show different levels with frailty and dependency in our cohorts. Materials and methods CRP ELISA and TNF- Luminex Plasma samples from elders (cohort 2, n=111) and adults (n=39) were thawed on ice. CRP concentration was measured with Quantikine ELISA (R&D) following the manufacturer’s instructions. Plasma samples were diluted 1:150 to fit the standard curve of the kit. On the other side, a panel of 6 interleukins was designed for luminex measurement: IL-6, IL-10, IL-2, IL-1, IL-1Ra and TNF-. The Milliplex Map #HCYTOMAG-60K kit (Merck) was used. Manufacturer’s instructions were followed and plasma samples were assayed undiluted, but only the measurements of TNF- were above the lowest point of the standard curve. We performed a second trial with the same kit and obtained similar results. In order to solve this issue, we repeated the assays using the high sensitivity kit #HSTCMAG-28SK (Merck) provided 96 | Chapter one by the manufacturer, but most of the samples were still non-detectable. Lastly, we also tried a high sensitivity luminex kit from another brand, #FCSTM09-04 (R&D) for IL-6, IL-10, IL-2 and IL-1. With this kit the measurement of analytes was also non-detectable in many samples (65/160). After the obtained negative results, we decided not to measure more plasma samples with the luminex technique and we analysed only the results from TNF- the only analyte that obtained detectable and reliable results (elders n=37 and adults n= 39). CRP, TNF-, IL-6 and albumin ELISA With the objective to test serum samples and to strengthen our results, samples from elders (cohort 3, n=91) and adults (n=18) were used. Samples were thawed on ice, CRP, TNF- and IL-6 were measured with Quantikine ELISAs (R&D) and albumin with an ELISA kit (Invitrogen, Thermo) following the manufacturer’s instructions. Serum samples were diluted to fit the standard curves of each kit: diluted 1:100 for CRP, undiluted for TNF-, undiluted for IL-6 and diluted 1:500000 for albumin. Statistical analysis Statistically significant differences between the study groups and correlations between variables were tested with GraphPad Prism version 6.01 for Windows (GraphPad Software, www.graphpad.com). D’Agostino-Pearson normality test was applied and non-Gaussian distribution was confirmed for all samples. Consequently, Mann-Whitney tests were applied to evaluate differences between two study groups. For correlation analysis, Spearman coefficient was calculated. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001. Results Inflammatory markers in plasma To evaluate the inflammatory status differences between the study groups, CRP and TNF- were measured in plasma samples from cohort 2. First, results from elders were compared to healthy adults, and we confirmed an increased concentration of both CRP and TNF- in aged individuals (Figure 18A and Figure 19A). Then, the correlation between age and inflammatory markers was evaluated, but no significant correlations were found (Figure 18B and Figure 19B). Similarly, no differences were found based on gender (Figure 18C and Figure 19C). Considering other possible confounder factors and taking advantage of the available data about the drug number prescribed to the donors in this cohort, we also evaluated the correlation between the number of drugs each participant takes and the concentration of CRP and TNF-. We found a significant positive correlation between CRP and the drug number – which was corrected by linear regression before further analysis –, while no correlation was found for TNF- (Figure 18D and Figure 19D). Chapter one | 97 Finally, the concentration of both inflammatory markers was compared between robust and frail individuals. The recorded frailty scales for cohort 2 were: TFI, GS, TUG, SPPB and GFST. The classification of each frailty scale was considered, and we did not find any significant differences (Figure 18E-I and Figure 19E-I). To perform a more robust comparison, the participants that are classified as robust or frail for all the tested scales (n=40) were compared, but no differences were reported (Figure 18J and Figure 19J). In the last approach, the elders with the same classification for the 3 scales that evaluate the functional status (GS, TUG, SPPB, n=63) were brought into comparison, and as for the previous analyses, no differences were found (Figure 18K and Figure 19K). Figure 18. Concentration of CRP in plasma. (A) There is elevated CRP in elders compared to adults. (B) Among elders, CRP concentration has no correlation to age and (C) there is no significant difference between females and males. (D) A positive correlation between CRP concentration and drug number was found and corrected. (E-I) No differences in CRP levels between robust and frail individuals were found for the 5 analysed frailty scales. (J) We also compared the individuals classified as robust or frail with all the available tests or (K) with the 3 functional scales (GS, TUG, SPPB), but no differences were found. 98 | Chapter one Figure 19. Concentration of TNF- in plasma. (A) There is elevated TNF- in elders compared to adults. (B) Among elders, CRP concentration has no correlation to age, (C) there is no significant difference between females and males and (D) no correlation between TNF- concentration and drug number was found. (E-I) No differences in TNF- levels between robust and frail individuals were found for the 5 analysed frailty scales. (J) We also compared the individuals classified as robust or frail with all the available tests or (K) with the 3 functional scales (GS, TUG, SPPB), but no differences were reported. Inflammatory markers in serum For the characterization of inflammatory markers in serum, samples from cohort 3 were used. We studied the previously measured CRP and TNF-, as well as IL-6 and albumin. We obtained the same results as in plasma, confirming that there is an elevated chronic inflammation in elders when compared to adults: increased CRP, TNF- and IL-6, while reduced albumin (Figure 20A, Figure 21A, Figure 22A and Figure 23A). Moreover, CRP, TNF- and IL-6 showed a positive correlation to age among elders (Figure 20B, Figure 21B and Figure 22B), which was corrected by linear regression for each analyte. On the other hand, no correlation with age was found for albumin (Figure 23B). Regarding gender, no differences were found Chapter one | 99 for any of the analytes (Figure 20C, Figure 21C, Figure 22C and Figure 23C). At last, we made use of the available data of Barthel and TUG scales, performing the comparison between the different dependency statuses of participants in cohort 3: robust, frail and non-autonomous. We found no significant differences for any of the molecules between the analysed groups (Figure 20D, Figure 21D, Figure 22D and Figure 23D), following the same trend as the analyses in plasma samples. Figure 20. Concentration of CRP in serum. (A) There is elevated CRP in elders compared to adults. (B) Among elders, serum CRP concentration has a positive correlation to age and (C) there is no significant difference between females and males. (D) When compared based on Barthel and TUG scales, no differences in CRP levels between robust, frail and non-autonomous individuals were found. Figure 21. Concentration of TNF- in serum. (A) There is elevated TNF- in elders compared to adults. (B) Among elders, serum TNF- concentration has a positive correlation to age and (C) there is no significant difference between females and males. (D) When compared based on Barthel and TUG scales, no differences in TNF- levels between robust, frail and non-autonomous individuals were found. 100 | Chapter one Figure 22. Concentration of IL-6 in serum. (A) There is elevated IL-6 in elders compared to adults. (B) Among elders, serum IL-6 concentration has a positive correlation to age and (C) there is no significant difference between females and males. (D) When compared based on Barthel and TUG scales, no differences in IL-6 levels between robust, frail and non-autonomous individuals were found. Figure 23. Concentration of albumin in serum. (A) There are reduced albumin levels in elders compared to adults. (B) Among elders, serum albumin concentration has no correlation to age and (C) there is no significant difference between females and males. (D) When compared based on Barthel and TUG scales, no differences in albumin levels between robust, frail and non-autonomous individuals were found. Discussion Inflammaging is one of the main biological characteristics of human aging. This term was proposed in 2000 by Franceschi et al. [136], although a work showing the accumulation of inflammation with age and its relation to mortality was already published in 1991 by Mooradian et al. [295]. Since this term was introduced, many works have investigated the relationship between inflammatory markers and aging, dependency and mortality. However, the obtained results are diverse and many times discordant, so no consensus has been reached [138,144]. Chapter one | 101 The studies investigating the potential role of molecules linked to inflammation as frailty biomarkers encounter the same problem. Many works have been carried out, but no clear association has been found, as some found increased concentrations of proinflammatory markers in frail subjects, whereas others did not report any significant differences [170,173,296]. It should also be considered that the study designs, techniques and the characteristics of included participants are distinct in each investigation. Moreover, the published studies have been carried out in different countries, hence distinctive genetic and environmental aspects should be considered. In addition, the tests employed for frailty assessment evaluate the status of participants based on different aspects, and therefore, the same person can be considered frail based on one scale and robust based on another one. For example, there are tests that focus on clinical aspects [16], while others give more importance to the psychologic domain [15] or the functional performance [19]. This diversity of tests shows the heterogeneity of the “frailty” term, which makes it even more difficult to identify a biological marker of the syndrome. Regarding the techniques available for the quantification of the molecules of interest, we decided to apply ELISA and Luminex. We obtained good results with the ELISA kits for plasma and serum samples, while we encountered detection problems with Luminex kits. We chose Luminex because it is a technique that enables the detection of a panel of analytes in the same experiment, saving time and reducing the amount of sample needed. Moreover, we have applied this protocol in previous experiments for other projects, and we obtained good results with cell culture supernatants. However, when plasma samples were tested, all the selected analytes, except TNF-, did not reach detectable levels in most of the samples. We thought that the concentration of our analytes could be lower than the detection limit of the kits, but we dismissed this possibility at least for IL-6, as we have previously measured it with ELISA in samples of the same characteristics and the obtained concentrations are notably higher than the detection limit of the Luminex kits used. Due to this, we decided to continue analysing our plasma and serum samples with ELISA. We reliably measured CRP, TNF-, IL-6 and albumin in our cohorts, that include aged subjects from the region of Gipuzkoa (Basque Country, Spain) who have been evaluated with several frailty scales. In these cohorts, our results confirmed the presence of inflammaging by the increased low-grade inflammation in elders when compared to adults. But regarding frailty, which was the main objective of this work, none of the analysed molecules showed significant differences based on the dependency status. Moreover, to try to overcome the above-mentioned issue of the heterogeneity between frailty tests, and taking advantage of the available data in cohort 2, we also compared the 108 | Chapter one Table 4. The 35 genes identified to be differentially expressed between robust and frail elders. Gene symbol Entrez Gene ID Frail Avg (log2) Robust Avg (log2) Fold Change p-value EGR1 1958 8.08 6.66 2.66 0.0014 DDX11L1 100287102 7.73 9.32 -3.01 0.0039 MIR454 768216 3.13 3.94 -1.63 0.0004 CISH 1154 6.82 7.99 -2.25 0.0033 DDX11L10 100287029 6.25 8.28 -4.11 0.0013 LOC101929775 101929775 5.82 7.22 -2.62 0.0004 LOC644172 644172 5.82 7.22 -2.62 0.0004 NSF 4905 5.58 7.43 -3.59 0.0002 TRAJ17 28738 7.18 8.57 -2.62 0.0017 TRAJ19 28736 6.52 7.6 -2.11 0.0037 TRAV8-3 28683 7.42 8.86 -2.71 0.0004 CD40LG 959 7.14 8.15 -2.01 0.0098 CLDN12 9069 3.13 4.51 -2.6 0.0151 CNTNAP3 79937 6.8 8.18 -2.6 0.0479 CNTNAP3B 728577 6.95 8.21 -2.39 0.0310 CSRNP1 64651 8.01 6.94 2.1 0.0307 CTSLP8 1518 2.14 3.25 -2.16 0.0260 CXCL8 3576 10.17 9.06 2.15 0.0190 G0S2 50486 8.89 7.46 2.69 0.0318 GCNT4 51301 5.8 6.82 -2.02 0.0217 GJB6 10804 5.18 4.09 2.12 0.0057 IGHV2-26 28455 3.63 4.88 -2.38 0.0291 LOC100505530 100505530 5.35 6.64 -2.45 0.0224 LOC105378916 105378916 6.36 7.72 -2.56 0.0333 MIR3941 100500866 3.16 4.43 -2.4 0.0059 MIR487A 619555 2.42 3.62 -2.29 0.0275 MIR626 693211 2.96 4.04 -2.11 0.0244 MTRNR2L2 100462981 5.58 6.84 -2.39 0.0334 RLN1 6013 4.65 5.71 -2.08 0.0092 TIA1 7072 7.46 8.49 -2.05 0.0462 TRAJ14 28741 8.14 9.17 -2.04 0.0128 TRAJ16 28739 7.77 8.78 -2.01 0.0300 TRAJ48 28707 7.83 8.9 -2.09 0.0185 TRAV16 28667 5 6.58 -2.98 0.0172 TRBV3-1 28619 7.21 8.39 -2.26 0.0058 Chapter one | 109 Next, we evaluated whether the expression of the selected candidates is altered in a larger cohort. For this purpose, we selected 120 RNA samples from cohort 1 and cohort 2. To maintain the previously set criteria, only the participants that obtained the same classification in the TFI, GS and TUG tests were selected (robusts n=103, and frails n=17). The RT-qPCR analysis of these 120 samples confirmed the increased expression of EGR1 in frail elders, while no differences were found for DDX11L1 and MIR454 (Figure 27). Figure 27. Expression of 3 of the candidates in a validation cohort. (A) The expression of DDX11L1 is not different in robust and frail participants. (B) EGR1 is significantly upregulated in frailty. (C) MIR454 is downregulated in frailty, but statistical significance was not reached. Physical intervention and EGR1 expression Some of the donors of cohort 2 were invited to participate in a physical intervention study for 3 months. Blood samples were collected before and after the intervention, and the expression of DDX11L1, EGR1 and MIR454 were determined by RT-qPCR. No clear trends were observed for DDX11L1 and MIR454 (Figure 28A and 28C). In contrast, EGR1 levels were reduced in 9 out of the 12 donors with a p-value of 0.06, reinforcing the potential of this gene as a biomarker of frailty (Figure 28B). Furthermore, when the changes in EGR1 expression and TUG performance were compared, we observed that 8 out of the 12 participants obtained concordant results: a reduction of EGR1 accompanied by a better TUG score, or increased EGR1 accompanied by a worse TUG score (Figure 28D). 110 | Chapter one Figure 28. Physical intervention, gene expression and physical performance. (A) No differences were reported for DDX11L1 expression. (B) EGR1 expression is reduced in 9 out of the 12 participants and a p-value of 0.06 was obtained with the paired-samples statistical analysis. (C) No differences were reported for MIR454 expression. (D) When the expression of EGR1 and TUG scores were compared, 8 out of the 12 participants obtained concordant results: reduction in EGR1 expression and TUG performance time, or increased EGR1 expression and TUG performance time. Discussion In the present study, we have performed a transcriptomic analysis of community-dwelling individuals from the Basque Country. A set of 35 differentially expressed transcripts was found between robust and frail elders. Among them, there were genes linked to inflammation and hypoxia-related pathways, immune response, apoptosis and several members of the T cell receptor alpha locus. These processes have also been related to frailty in several previous studies, but however, our set of genes was different from a recently proposed panel of potential frailty biomarkers [187]. Within the 35 transcripts, we selected 3 for the first validation approach: DDX11L1, EGR1 and MIR454. The technical validation confirmed the microarray results, while when we measured them in a larger cohort, only the increased expression of EGR1 in frail subjects was confirmed. Moreover, we determined the levels of EGR1 in 12 donors before and after a 3-month physical Chapter one | 111 intervention study, and 9 of the participants reduced the expression of EGR1. Again, no differences were reported for DDX11L1 and MIR454. Our results indicate that EGR1 is a promising biomarker of frailty that should be further investigated. Indeed, EGR1 is a transcription factor activated in response to a broad range of stimuli that affects directly or indirectly the expression of multiple signalling pathways and tumour suppressors, and it modulates and participates in multiple cellular processes such as mitogen response, growth, proliferation, apoptosis or differentiation of several cell types [300–302]. Besides, some studies also found EGR1 changes to be associated with aging and age-related phenomena, like senescence or immune response regulation [303–306]. Thus, the role of EGR1 in frailty is probably complex and depending on the tissue and context its target genes and the elicited functions may differ. Moreover, as discussed before, the identification of biomarkers of frailty is challenging due to the lack of consensus its definition and the multiple frailty screening tools available. Thanks to the work performed at primary care services, were had the data of 3 different tests (TFI, GS and TUG) for each of the participants in our study, and after the microarray analysis and the results of the PCA, we decided to focus on the subjects that obtained the same classification in all of the tests, reducing heterogeneity. In addition, it should be mentioned that despite this, the expression levels reported by RT-qPCR were highly variable. In this sense, one of the main advantages of longitudinal studies is that the expression of single participants can be measured over time and the evolution of each of them determined. Indeed, our results from the physical intervention pilot study show that even if the expression of EGR1 was different, most of the subjects reduced the levels of EGR1 after only 3 months. This is an interesting point also from the perspective of the reversibility of frailty. It is generally accepted that frailty is a reversible state [7,307], and the expression of EGR1 could be used as a biomarker of this process. Finally, we have to point out that the present study was the only first step and we focused on the validation of 3 of the transcripts identified. We are now analyzing more candidates from the presented list of 35 differentially expressed genes between our robust and frail individuals. Our aim is to continue evaluating the expression differences and besides, to try to understand the functions or effects that these changes could be inducing in frail elders. 112 | Chapter one Chapter one | 113 Extracellular vesicles Introduction EVs are membrane-coated particles of endosomal or plasma membrane origin that are secreted to the extracellular environment. They play an essential role in indirect intercellular communication as their membrane and cytosolic proteins, lipids and genetic material can be transferred between cells [195]. Moreover, almost all cell types release EVs and they can be isolated from plasma and other body fluids. EVs are released both in physiological and pathological conditions and they are implicated in many cellular processes. In particular, EVs play a role in various stages of the immune response and they have been related to inflammatory, autoimmune and infectious disease pathology. EVs can carry and display antigenic material and are able to trigger antigen presentation and modulate immune responses [192]. It has also been reported that there are increased concentrations of EVs in plasma during inflammatory processes, such as in cancer or autoimmune diseases [234,242]. One of the hallmarks of human aging is the chronic low-grade inflammation, the so called inflammaging [136], a phenomenon that modulates intercellular communication. The ageassociated immune dysfunction and accumulation of senescent cells promote inflammatory signals, such as elevated secretion of proinflammatory cytokines and activation of NF-κB transcription factor. Among inflammaging, the most widely studied feature is the circulating concentration of IL-6. The concentration of this interleukin is normally low (or non-detectable) in healthy adults, while elevated levels of IL-6 have been reported in the elderly, with increasing concentrations in the very old [138,140]. Moreover, elevated IL-6 has also been associated with mortality in the elderly [142]. Despite all this knowledge, there are many aspects of inflammaging that have not been elucidated, as the implication of EVs in the process. In the present study, and based on the previously mentioned increase of circulating EVs during inflammatory episodes, we proposed that this could also be observed in aged individuals as a result of inflammaging. Furthermore, we also designed an approach to evaluate if the concentration of EVs in plasma could be related to the frailty status of old people—frailty status as defined by the Barthel Index [308] and the Tilburg Frailty Index [15]. These tests are applied to evaluate and measure the frailty and dependence status of the elderly, which could also be related to an increased chronic proinflammatory condition. 114 | Chapter one Materials and methods Study participants Samples from 19 aged individuals (from Cohort 1: 8 males and 11 females, mean age 83.73 years) and 18 adults – classified in three age ranges: 21–30, 31–40, and 41–50 years (3 males and 3 females in each group) were used. Plasma and serum samples were obtained as described above. EV Isolation EVs were isolated as described before by our group [202]. Briefly, plasma was centrifuged at 13,000 g for 2 min and supernatant centrifuged again at 20,000 g for 20 min to pellet EVs. The pellet was resuspended with 100 µL of filtered DPBS (GIBCO, Thermo Fisher Scientific), filtered twice through a 0.22 µm-pore filter. Resuspended EVs and serum samples were stored at -80 °C. Serum IL-6 ELISA Assay IL-6 concentration was analysed by ELISA (BD Biosciences) following the manufacturer’s instructions. Samples were measured in duplicate and results obtained with a microplate reader (Thermo Scientific Appliskan, Thermo Fisher Scientific). IL-6 concentrations were calculated and values above the first standard (>4.7 pg/mL) were considered detectable. Nanoparticle Tracking Analysis (NTA) The size distribution and concentration of EVs were measured using a NanoSight LM10 device (Malvern) as described elsewhere [212]. Samples were diluted to appropriated levels to get accurate acquisitions (200–900 recorded tracks) [212] and camera settings were fixed and maintained for all samples. Filtered DPBS was tested and no background signal was detected. For each sample, two videos of 1 min were recorded and analysed with NanoSight NTA software 2.2 (Malvern). Data are shown as the average count of the two duplicates. Statistical Analysis Statistical analysis was performed with R version 3.2.2 (R Core Team (2015) [309] in RStudio v0.99.486 (RStudio Team (2015) [309]). A Shapiro-Wilk test was applied to assess normality. As samples did not follow a normal distribution, Wilcoxon signed-rank test and nonparametric Kruskal–Wallis one-way analysis of variance were conducted to evaluate IL-6 and EV concentration differences between groups. Chapter one | 115 Results Frailty Status Classification of Aged Individuals For the present study, adults of different age ranges and elder people were enrolled. Participants were classified based on their age. Additionally, aged individuals (79–92 years) were asked to complete the Barthel and Tilburg Frailty Index questionnaires and were further classified as Robust, Frail or Non-autonomous, as shown in Table 5. Table 5. Classification of enrolled individuals based on their age and frailty status. Samples of a total number of 18 adults and 19 elders were analysed. IL-6 Concentration Is Increased in the Elderly The level of IL-6 in serum was measured and, obtained results demonstrated a very low, nearly non-detectable concentration in adults of different ages, while an elevated concentration in the elderly (p<0.001) (Figure 29A). This result confirms the low-grade inflammatory condition of aged individuals. Additionally, when IL-6 levels of the elderly were compared depending on their frailty status, an increasing tendency with dependence was found (Figure 29B). Figure 29. An elevated concentration of IL-6 is observed in aged subjects. IL-6 levels were measured by ELISA and concentration values above 4.7 pg/mL were considered detectable. (A) Elderly individuals have a higher concentration of IL-6 than adults (***p<0.001); and (B) there is a high variability among Robust, Frail and Non-autonomous elderly, but an increasing concentration with dependency can be observed. 116 | Chapter one The Concentration of EVs Is Not Affected by Age and Frailty Status To assess the size profile and concentration of circulating EVs, NTA was conducted for all samples. Results showed that, regardless of particle concentration, all samples followed a similar EV size distribution, with most vesicles ranging between 50 and 300 nm in all instances (Figure 30A). This result demonstrated that our EV isolation protocol efficiently isolates small EVs, removing larger particles and platelets that can be found in plasma samples. When comparing the EV number, no significant differences were found between groups (p = 0.505), indicating that EV concentration is not increased with age (Figure 30B). Moreover, the concentration of EVs is also not affected by the frailty status of elder donors (p = 0.424), as shown in Figure 30C. Figure 30. Particle size and EV concentration were measured by NTA. (A) Size distribution of EVs. Each line represents one sample. Despite the particle concentration difference, all samples have a similar size distribution—they are enriched in small EVs (50–300 nm); (B) EV concentration of different age ranges were compared and samples from elder people (79–92 years) do not show an increased EV number; and (C) among elder individuals, the frailty status does also not alter EV concentration. Chapter one | 117 Discussion During human aging, a chronic low-grade inflammatory state called inflammaging has been reported [138,140,142], and to our knowledge, this is the first report investigating, specifically, EV concentration in this process. The results presented in this work demonstrate that there are elevated IL-6 levels in the elderly, confirming the basal inflammaging. In contrast to what we hypothesized, and despite inflammaging, EV concentration in circulation is not affected by human aging. Moreover, frailty or dependence did also not alter the EV number. Many authors have previously studied the implication of EVs in diverse inflammatory processes, including cellular senescence, neurodegenerative diseases and cancer, indicating that both the total number of EVs in circulation and also EVs from specific cell origins can be increased [192,230,310]. Our results present a chronic inflammatory process—inflammaging—in which circulating EV levels are not affected. In this work, and when studying EVs, there are several factors that should be taken into consideration. In our cohort, a high inter-individual EV concentration variability has been found within the same group. Similarly, previous experiments have demonstrated that the protein concentration and content of EVs differ depending on the donor [311]. On the other hand, specific medications may also affect EVs, as there are compounds that can modify EV production and release. For instance, immunomodulatory treatments can affect EV production by immune cells and modulate EV concentration in circulation [241,253,312]. This kind of effects should be considered when measuring EV levels specially in aged people, because nearly all of them have chronic medications. In this study, a representative sample of community-dwelling aged people was analysed and, as expected, all were under chronic treatment. It was ethically not possible to ask the participants in the study to interrupt their medications. Furthermore, the aim of our study was to evaluate whether the low-grade proinflammatory status was sufficient to alter EV concentration in the elderly, despite their medication. Moreover, even if the total number of EVs is not altered, EVs secreted from specific cell types could be affected, both in their concentration and cargo, modulating their function and effect in target cells, as described for other biological processes [230]. Finally, the limited number of samples in the study must be taken into account and results should be validated in a larger cohort. In brief, these results represent a first report and demonstrate that there is no correlation between inflammaging and EV concentration in circulation. More extensive experiments are required to study the specific changes that occur to EVs in regard to human aging, and to further elucidate their role in the process. 124 | Chapter two Materials and methods Obtention and isolation of plasma EVs Peripheral blood was collected by experienced nurses by venipuncture with a 21-gage needle in 4 ml EDTA tubes (Vacutainer, BD Biosciences). Samples from 5 adults (mean age 37.8 years, 2 females and 3 males) and 5 elders (mean age 85.4 years, 3 females and 2 males) were obtained. Tubes were kept upright and centrifuged at 1258 g for 20 min to recover plasma. To isolate EVs, plasma was centrifuged at 13,000 g for 2 min and obtained supernatant (1 ml platelet free plasma) was centrifuged again at 20,000 g for 20 min to pellet EVs. 900 l of supernatant were transferred to another tube and the bottom 100 l with the EV pellet were resuspended with 100 µl of filtered DPBS (GIBCO, Thermo Fisher Scientific, filtered twice through a 0.22 µm-pore filter). The 200 l of resuspended EVs were stored at -80 °C and thawed on ice when needed. Obtention and culture of ASCs Abdominal liposuction was performed in a female donor aged 49 years. With the patient’s informed consent, subcutaneous adipose tissue was obtained by outpatient tumescence liposuction under local anesthesia by an experienced physician. ASCs were isolated according to Wolbank et al. [318] and cultured in control medium consisting of DMEM-low glucose/HAM´s F-12 (GE-Healthcare) supplemented with 4mM L-glutamine (Sigma-Aldrich) and 10% fetal calf serum (Sigma-Aldrich) at 37°C, 5% CO2 and 95% air humidity. Culture medium was changed three times a week and cells were passaged once a week at a split ratio of 1:2 to 1:6 according to the growth characteristics. Coculture of ASCs with plasma EVs and induction of osteogenic differentiation ASCs were seeded in 24-well culture dish wells. 3 days after seeding, osteogenic differentiation was induced by switching the medium to osteogenic differentiation medium consisting of DMEM-low glucose (GE-Healthcare), 10% fetal calf serum (Sigma-Aldrich), 4mM L-glutamine (Sigma-Aldrich), 10nM dexamethasone (Sigma-Aldrich), 150μM ascorbate-2-phosphate (Sigma-Aldrich), 10mM β-glycerolphosphate (Sigma-Aldrich) and 10nM 1.25 Dihydroxyvitamine D3 (Sigma-Aldrich) in a final volume of 1 ml/well. Before performing coculture experiments to investigate the effect of EVs, culture conditions were optimized. Three different cell concentrations were assayed (4,000/9,000/14,000) and 14,000 ASC cells/well was chosen as the best for osteogenic differentiation measurement. Regarding osteogenic differentiation duration, 10 and 17 days were tested, and 10 days were chosen for subsequent experiments. When investigating the effect of EVs, 4 IU heparin/ml were Chapter two | 125 added to prevent culture medium jellification. Taken together, the final experimental setup to test the effect of plasma EVs on osteogenic differentiation is presented below and schematically represented in Figure 31. Besides, all the experiments performed and obtained results for the optimization of culture conditions are presented in the results section. ▪ 14,000 ASCs/well were seeded in 24-well culture dishes with 950 l control medium supplemented with 4 IU heparin. ▪ 3 hours after seeding the cells, 50 l of EVs (or 50 l of DPBS) were added to corresponding wells, reaching a final volume of 1 ml/well, and carefully mixed to ensure a homogeneous EV distribution. ▪ 3 days after seeding the cells, the culture medium was changed. Osteogenic differentiation (OD) medium or control medium was added to corresponding wells. ▪ Cells were maintained in culture for 10 days more with media changes every 3 days. ▪ Alizarin Red or Alkaline phosphatase staining was performed to measure osteogenesis. Figure 31. Cell culture conditions to test the effect of EVs from plasma on osteogenic differentiation of ASCs. (A) Schematic representation of the culture protocol. ASCs were seeded in 24-well culture dishes in 950 l control medium and 3 hours later, when cells were attached 50 l of EVs or DPBS were added. 69 hours later (3 days after seeding) culture media were changed, switched to OD medium or to fresh control medium in corresponding wells. Cells were maintained in culture for 10 days more (13 days after seeding) with media changes every three days. Finally, osteogenic differentiation was evaluated by Alizarin Red or Alkaline phosphatase (ALP) staining. (B) Schematic representation of the 6 different study conditions. EV samples from 5 adults and 5 elders were tested, and all conditions were assayed in duplicate. 126 | Chapter two Alizarin Red staining For quantification of calcified structures, cells were washed 3 times with PBS and then, fixed for 2 h in 70% ethanol (500 l/well) at -20°C. Then, cells were washed 3 times with dH2O and stained for 10 minutes with 40mM Alizarin Red S solution (pH 4.2, 500 l/well, Sigma-Aldrich) in an orbital shaker at room temperature. Subsequently, the remaining dye was removed by rinsing the cells with PBS. Finally, the residual dye was extracted by 0.1M HCL/0.5% SDS solution (200 l/well) for 30 min. The dye signal was quantified in a microplate reader by determining the absorbance at 425 nm. Alkaline phosphatase (ALP) staining To determine the activity of ALP, cells were washed 3 times with PBS and then, lysed by incubating with lysis buffer for 1 h (0.25% Triton X-100, 100 l/well) at room temperature. Next, samples were transferred to 1.5 ml tubes, centrifuged at 13,000 rpm 10 min at 4°C and 90 l of supernatant transferred to a new tube. Subsequently, a buffer containing 20mM 4nitrophenyl phosphate disodium salt hexahydrate, 0.5 M 2-amino-2-methyl-1-propanol and 0.2 mM MgCl2 (pH 10.3, 50 l/tube) was added to each cell lysate and incubated for 20 minutes at room temperature in the dark. The reaction was stopped by adding 50 µl of 0.2M NaOH and ALP activity quantified by determining the absorbance at 405nm (620nm ref). Statistical analysis Statistically significant differences between the study groups were tested with GraphPad Prism version 6.01 for Windows (GraphPad Software, www.graphpad.com). Mann-Whitney test was applied to evaluate differences between EVs from adults and elders. **p<0.01. Results Osteogenic differentiation settings The first experiments were directed to establish the best culture conditions to investigate the effect of EVs on osteogenesis. To this end, we tested 3 cell densities (4,000/9,000/14,000 ASCs) and two different end points (10/17 days after osteogenic differentiation induction). Triplicates were performed for all the conditions. Calcification was measured by Alizarin Red staining and the obtained results are presented in Figure 32. We confirmed the induction of osteogenesis in all the tested conditions, as calcium deposition was higher under OD medium than under control medium (basal calcium deposition by ASCs). Moreover, a higher concentration of cultured cells resulted in an increased staining and a higher difference between control and OD medium. Regarding the two tested end points, an elevated differentiation was shown with the prolongation of osteogenesis (17 days). Chapter two | 127 With these results, we decided to perform the following experiments with 14,000 ASCs/well and to measure osteogenesis 10 days after the induction. We based our decision about the cell density on the prominent differences observed between control and OD medium. On the other hand, for the osteogenesis duration, we took into consideration the possible effects of plasma EVs: previous studies have reported an enhanced differentiation in the presence of EVs, so to be able to see the influence of EVs, we should measure osteogenesis at an intermediate point when an increase of calcification or ALP activity could be reported by the absorbance measurement. Figure 32. Osteogenic differentiation settings assessed by Alizarin Red staining. 3 different cell densities were seeded and calcium deposition measured 10 days (A) or 17 days (B) after osteogenic differentiation induction. Wells with control medium in which osteogenesis was not induced were maintained and assayed to measure basal calcium deposition by ASCs. Increasing staining was observed with more cells and with prolonged culture times. Coculture of ASCs with plasma EVs results in the formation of jelly structures After establishing the cell culture conditions for ASCs, in the next step we performed the first experiments in which EVs from plasma were added to ASCs. Cells were seeded in 24-well dishes and 3 hours after seeding the cells, 50 l of thawed EVs were added to each well. 3 days after seeding the cells, culture media were changed and control or OD medium added to corresponding wells. Unexpectedly, at the bottom of the wells where ASCs and EVs were cocultured a jelly layer had formed. Medium change was done by carefully pipetting, but even so, when aspirating the medium, the jelly layer was partially detached in some wells. We still decided to continue with the osteogenesis protocol and try to measure the effect of EVs on calcium deposition. 10 days after osteogenic differentiation induction, Alizarin Red staining was performed. Despite the careful pipetting, we observed that the jelly-like layers were detached and it was not possible to rinse the wells after Alizarin staining (Figure 33), so no result was obtained from this assay. 128 | Chapter two Figure 33. Formation of jelly structures and cell detachment. ASCs were cocultured with EVs isolated from plasma and a jelly layer had formed at the bottom of the wells. (A) Representative image of a well where the jelly layer was detached and flipped on top of other cells. (B) A picture taken from 2 wells cocultured with EVs. Alizarin Red staining was performed, but no concluding results could be obtained. After observing this phenomenon, we performed an experiment to investigate the formation of jelly structures and whether heparin could prevent them in our cocultures with plasma EVs. We prepared a 24-well dish as shown in Figure 34: EVs alone, ASCs + EVs, EVs in heparin containing medium and ASCs + EVs in heparin containing medium were tested. In all cases, 14,000 ASCs and 4 U of heparin/ml were used, while different EV volumes were added. Figure 34. Schematic representation of the 24-well culture dish plan. In row A control medium and EVs were mixed, while in row B 14,000 ASCs/well were also added. In row C control medium with 4 IU heparin and EVs were mixed, while in row D 14,000 ASCs/well were also added. 3 days later the formation of gels was evaluated. Culture medium was carefully pipetted and we observed no jelly layers in row A and C, confirming that the interaction between control medium and EVs do not produce the jelly structures. When medium in row B was aspirated, jelly layers were found in all wells, with thicker structures in the wells cocultured with higher volumes of EVs. In contrast, medium pipetting was performed normally in row D, demonstrating that the addition of 4 IU heparin/ml prevents the formation of the gel for all the EV volumes tested. Consequently, we decided to incorporate heparin to the control medium for coculture experiments. Chapter two | 129 Effect of plasma EVs on osteogenesis Finally, we tested the effect of EVs isolated from plasma of adult and elder donors on osteogenic differentiation. EV samples from 5 adults and 5 elders were assayed, each of them in duplicate. Our results show that, in all cases, the coculture of ASCs with plasma EVs enhance osteogenesis, and this effect is stronger with EVs from adults (Figure 35). Figure 35. Osteogenesis enhancement by plasma EVs. ASCs were cocultured with EVs for 3 days and then, osteogenic differentiation was induced. 10 days after induction ALP activity was measured. Results are presented in fold change versus the control condition in which no EVs were cocultured. (A) Graph showing the results obtained for each EV donor. In all cases, ALP activity was higher than the control. (B) Box plot representation of results obtained for EVs from adult and elder donors. EVs from adults enhance ALP activity significantly more than EVs from elders. A = adult and E = elder. Besides, we also tested whether EVs alone were able to induce osteogenesis. To this end, we cultured ASCs in control medium. The coculture with EVs was performed as before, but instead of inducing osteogenesis with OD medium, control medium was maintained. The obtained results demonstrate that the presence of EVs alone does not induce osteogenesis (Figure 36). Figure 36. Effect of EVs under control medium. ASCs were cocultured with EVs for 3 days and then, cultured for 10 days more with control medium. Two wells without EVs were cultured with OD medium for 10 days as a positive control of osteogenesis. ALP activity was measured and no differences were reported, indicating that EVs alone do not induce osteogenic differentiation of ASCs. A=adult and E=elder. 130 | Chapter two Discussion In this project, we have set the cell culture conditions and investigated the effect of EVs from plasma on the osteogenic differentiation of ASCs. The starting point of this work was based on a publication by the group of Grillari [245], in which they reported an increased osteogenesis enhancement with EVs isolated from plasma of young donors (less than 25 years) when compared to older ones (more than 55 years). Besides, they focused on galectin-3 protein and showed the important role of this molecule during the osteogenic process. Regarding the EV experiments performed by the group, to isolate EVs from plasma they filtered the samples through 0.22 m pore filters and then applied ultracentrifugation at 100,000 g for 1 hour. With this protocol, they isolated small EVs and discarded bigger EVs. Moreover, when considering potential future applications, their protocol could be difficult to implement in the clinic, as most of the hospitals do not have ultracentrifuges. Taken together, the main objectives of our work were to test the effect of EVs isolated with an easily applicable protocol [202] and to compare the effect of samples coming from adult and elder donors. First, we obtained ASCs from a healthy donor and conducted experiments to establish the appropriate cell culture settings. In our hands, seeding 14,000 cells/well and maintaining osteogenic induction for 10 days resulted in adequate differentiation. When adding EVs to cultured ASCs, a jelly layer was formed. The addition of only 5 l of plasma EVs was enough to induce the formation of this structure. On the other hand, a culture medium prepared with 4 IU heparin/ml prevented jellification. The use of human plasma for cell culture has been widely investigated before, and medium clotting was also reported in many cases [319]. Here we applied EVs isolated from plasma, and even if we have demonstrated that our protocol efficiently isolates EVs, when handling a complex fluid as plasma, other small components are probably coprecipitated. In accordance with previous reports, the addition of a low concentration of heparin to our cultures prevented media clotting. It should be mentioned that some authors reported impaired cell proliferation and differentiation of ASCs under high doses of heparin [320], but we applied only 4 IU heparin/ml for 3 days and then, osteogenic differentiation was induced and no issues were reported. We performed coculture experiments with EVs from 10 different donors: 5 adults and 5 elders. Interestingly, all EV samples boosted osteogenic differentiation, but none of them induced differentiation if OD medium was not used, demonstrating that plasma EVs alone do not induce the differentiation of ASCs to osteoblasts. Moreover, the positive effect of EVs was more prominent when samples from adults were applied. These results indicate that EVs favour osteogenesis, but an age-related exhaustion could be present. Our results are in accordance with the previous study by the group of Grillari, and importantly, we tested more samples, Chapter two | 131 coming also from older donors (> 80 years, instead > 55 years), and with EVs isolated with a different protocol (final pelleting at 20,000 g, instead 100,000 g). This work reinforces previous studies and demonstrate the potential application of EVs for osteogenesis enhancement. Furthermore, other authors showed that EVs secreted by ASCs [260,313], or by a specific subset of plasma EVs promote osteogenesis [245], but their application would need in vitro cultures of ASCs to produce EVs or complicated protocols to isolate EVs from plasma, respectively. In contrast, our results indicate that EVs easily isolated from plasma, alone or in combination with cell therapies, could help osteogenesis. Furthermore, other authors that applied PRP and ASCs for osteogenesis with positive results were probably also administering EVs, as they would be present in PRP samples [316,317]. Regarding the possible clinical applications to promote osteogenesis, it should be mentioned that, similar to our results with EVs, ASCs from elders have a reduced osteogenic potential [125]. We hypothesize that to overcome the age-associated dysfunction and to avoid allogeneic cell transplantations, regenerative therapies in aged patients could be carried out with autologous ASCs and plasma EVs from a young donor. This combination could boost osteogenesis while preventing potential problems associated with allogeneic cells. In summary, ASCs are an easily accessible source of MSCs and they can be differentiated into different cell types, including osteoblasts. Due to their therapeutic potential, many efforts are being made to understand the underlying mechanisms of osteogenesis [321]. In parallel, the implication of EVs in osteogenic differentiation is still starting to be investigated. The first reports, including ours, indicate that EVs play an important role and help ASC differentiation, and we consider that they should be taken into consideration for future clinical applications. 132 | Chapter two Chapter two | 133 Myogenesis Introduction The skeletal muscle is the largest organ in the human body. It is a highly adaptable tissue that responds to environmental conditions and physiological challenges by changing fibre size and composition. However, the incidence of skeletal muscle injuries as a consequence of trauma, inherited genetic diseases, pathology or aging is very high and represents relevant socioeconomic costs. In the case of aging, muscle wasting, defined by marked muscle mass loss and weakening, is one of the major problems leading to increased risk of falls and development of physical disability [123,322]. It is well known that the skeletal muscle has regenerative potential, which, however, becomes compromised in the case of severe or extended damage as well as with aging [123,126]. In this context, several methods are nowadays applied in the clinic to promote muscle repair and regeneration and, besides, many investigations are being conducted to improve the present techniques or implement new and more effective methods [126]. Among these potential new methods, we are specially interested in the ones investigating the role of EVs in myogenesis. Indeed, in the last years, many authors have investigated the composition and functions of EVs secreted by myoblasts and myotubes [323]. Regarding their functions, a work by Forterre et al. reported that EVs secreted by myotubes reduce proliferation and induce myoblast differentiation [324], while a study by Guescini et al. did not obtain the same results [325]. However, the EV concentrations and differentiation endpoint were distinct, which could account for the observed differences. Besides, it should be mentioned that most of the works focusing on the effects of EVs on myogenesis were carried out in the C2C12 immortalized mouse myoblast line, which is an interesting and useful model for studying many processes, but it also presents some differences when compared to humans. To our knowledge, no works have studied the influence of EVs from human plasma on myogenesis. Nevertheless, many researchers have indirectly applied EVs in their investigations, as EVs are part of the components of PRP and platelet-poor plasma (PPP). These preparations have been widely tested for muscle regeneration, as nicely reviewed by Chellini and colleagues [326]. In any case, they did not consider the presence of EVs and their effects as part of PRP and PPP are still unknown. 140 | Chapter two We satisfactorily isolated RNA from all samples and performed cDNA synthesis and qPCR as usual. However, we did not obtain cDNA amplification for some of the samples, specifically for the ones corresponding to controls and to EVs from adults under differentiation medium. Consequently, no results could be obtained from these samples and, besides, the samples cocultured with EVs from elders with differentiation medium were also not analysed, as there was not any control to compare them with. The results from the rest of the conditions, which included all the samples maintained in proliferation medium, were analysed and the results are shown in Figure 42. Similar to the results obtained in the first experiment, we saw that myoblast cocultured with EVs from adults have elevated levels of myogenic differentiation markers when compared to EVs from elders (significant for MYOD1 and tendency for MYOG and DES). In contrast, in this second experiment, we did not observe a significant increase of the myogenic markers between the control wells and the ones cocultured with EVs. Figure 42. Expression of myogenic differentiation markers. Myoblasts were maintained in proliferation medium and they were cocultured with EVs from adult or elder donors. The samples with EVs from adults showed a higher expression of MYOD1, and the same tendency for MYOG and DES, while these two did not reach statistical significance. Discussion The objective of this section was to investigate the effect of EVs isolated from plasma on myogenesis. This is a field that has not been investigated, but it could have potential benefits for muscle regeneration, which is of particular interest for the age-associated sarcopenia. We decided to perform the experiments based on our previous results on osteogenesis, as well as on the literature about myogenic differentiation. Interestingly, a work by Nakamura et al. studied the effects of EVs secreted by MSCs [259] and, on the other hand, many investigations have evaluated the effect of PRP and PPP (that contain EVs) on myogenesis [326]. However, none of them specifically evaluated the potential role of EVs from plasma, which we consider that could also be playing a role. Chapter two | 141 With this in mind, we collaborated with a group from Biodonostia that works with primary myoblasts and performed two experiments with different setups. We tested different cell concentrations, coculture times and endpoints. As in our previous experiments, we evaluated the effect of EVs not only under differentiation medium, but also under proliferation medium, to investigate whether EVs could influence in both conditions. Notably, in the first experiment, we reported an elevated expression of MYOG, MYOD1 and DES myogenic differentiation markers under proliferation and differentiation medium. Besides, when EVs isolated from adult and elder donors were compared in our two experiments, a higher expression of the differentiation markers was observed with EVs from adults, with statistically significant differences for MYOD1 in the second experiment. Our results are a first report indicating that EVs from plasma could play a role in myogenesis and, moreover, that EVs from adults could have a more robust effect than the ones isolated from elders. Further, we performed the experiments on human primary myoblasts, while most works are performed in murine immortalized myoblasts [323–325]. To our knowledge, one publication investigated before the influence of EVs on myogenesis with human primary myoblasts, but they had a completely different objective, as they studied the effect of EVs present on the foetal bovine serum used for cell culture [328]. In any case, regarding the effects of EVs, we should keep in mind the complexity of biological processes and the limitations of the systems that we and all researchers apply. For instance, we try to model and study myogenic differentiation in vitro, by plating myoblasts, coculturing them with EVs from plasma and evaluating the expression of certain genes. We consider that this is a good approach to investigate whether EVs have an effect on the process, which is a novel field that is still in its infancy. However, we do not replicate the microenvironment present when a skeletal muscle of an individual is regenerating, and we would probably never be able to reproduce it exactly. With regard to the potential use EVs, the advantage of EVs with respect to cells is that we can first evaluate their effects in in vitro models, and then, test their potential efficacy in vivo with less safety concerns. As discussed in the previous section for osteogenesis, we hypothesize that plasma EVs could be applied in the future alone or in combination with autologous stem cells to enhance myogenesis. However, and even if some treatments with PRP have already been applied to patients with injured skeletal muscles with positive results [326], we are still far from understanding the implications of EVs. Our approach was just the first step and many more should be taken to describe the roles of plasma EVs on myogenesis and whether aging affects them. CHAPTER THREE Immunosenescence and the role of extracellular vesicles Chapter three | 145 Introduction Human aging is a complex and heterogenic process, in which several cellular mechanisms are affected and modulated, leading to functional decline [33]. One of the most determining consequences of aging is the dysfunction of the immune system, and the subsequent poor response to vaccination, increased susceptibility to infections and age-related diseases observed in the elderly [329]. The molecular and cellular changes that lead to immune dysfunction have been extensively investigated and are generally referred to as immunosenescence [92]. T cells are the most dramatically affected immune components, with a decrease in naïve T cells and an accumulation of terminally differentiated T cells with age. Terminally differentiated T cells exhibit features of replicative senescence and lose the expression of the costimulatory molecule CD28 from their membrane [98–101,330]. CD28 plays an essential role in T cell function, taking part in activation, proliferation and survival processes. Hence, CD28 negative T cells present altered molecular features, as well as distinct cytokine production and effector molecules [102]. The loss of CD28 affects earlier and primarily CD8 T cells, but it has also been described to reach CD4 T cells later in life [103,104]. In consequence, T lymphocytes have a reduced capacity to react against new stimuli, contributing to the aforementioned immune dysfunction. Another feature found in immunosenescent T cells is the enhanced cytotoxicity. Expression of NK cell characteristic receptors such as CD56 and CD57 membrane molecules have been widely reported in these cells, which promote their cytotoxic capacity [105–108]. Additionally, many authors have found a higher prevalence of an inverted CD4/CD8 ratio in the elderly, a feature known as immune risk phenotype, that predicts shorter survival [113– 115]. The immunosenescent process and the changes that occur in other cell types during aging result in an altered secretion of molecules by cells, termed SASP [131]. The SASP components have been classically divided into three groups: i) soluble signalling factors (ILs, chemokines, and growth factors), ii) secreted proteases, and iii) secreted insoluble proteins/extracellular matrix components [132]. One of the consequences of SASP is the chronic low-grade inflammation seen in the elderly, the so called inflammaging [136]. The age-associated immune dysfunction and accumulation of senescent cells promote inflammatory signals, such as elevated secretion of proinflammatory cytokines like IL-6 [138,140,331]. Another remarkable aspect that is affected by the SASP is intercellular communication. Apart from the three classical SASP components mentioned before, in the last decades EVs have been shown to play a central role in intercellular communication and immune system function [234]. 146 | Chapter three EVs are membrane-coated particles that are secreted by almost all cell types and are present in most body fluids, including plasma. They can be of endosomal or plasma membrane origin and they carry proteins, lipids and genetic material that can be incorporated by the target cell. EVs are released in physiologic and pathologic conditions and are implicated in many cellular processes [195]. As stated before, EVs are also implicated in the immune system function, as they can carry antigenic material and modulate immune responses [234]. Regarding EVs in aging and senescence, the expression of p53 transcription factor have been related to increased EV production [332]. However, some works have studied the concentration of plasma EVs with age, with contradictory results [331,333]. One of these works also examined the EV protein cargo and internalization by immune cells and showed proteins differentially expressed with age and that EVs from older donors are more readily internalized by B cells [333]. In spite of that, there are still many aspects of EVs that have not been elucidated. Similarly, even if immunosenescence at a cellular level has been widely investigated, only a few works have analysed samples from nonagenarians and centenarians [98,334]. It is important to note that only a small percentage of people reach these advanced ages, making it even more difficult to include their samples in study cohorts. Works that studied nonagenarians and centenarians showed that their PBMCs have distinct features at transcriptional and functional levels when compared to septuagenarians and octogenarians [26,27,335]. Taking all this into account, the aims of the present work were to characterize the immunosenescence status of our cohort (donors of 20-49 and 70-104 years), comparing different age ranges at the cellular and EV level and to try to describe the possible immune functions of plasma EVs. Materials and methods Participants and blood sampling For the present study, donors of different age ranges were enrolled. Healthy adults between 20-49 years and elders of 70-104 years were included. Elders were assessed at primary care services and by an experienced neurologist. Both community-dwelling and institutionalized participants and with distinct functional capacities were enrolled, aiming to have a representative sample of age-related heterogeneity. All participants completed a questionnaire and donors with acute illness or immunological disorders were excluded. Samples from 51 donors (29 females and 22 males), 18 healthy adults and 33 aged individuals were collected at Donostia University Hospital. Participants were classified based on their age range: 20-29 (n=6), 30-39 (n=5), 40-49 (n=7), 70-79 (n=6), 80-89 (n=10), 90-99 (n=13) and ≥100 (n=4) Chapter three | 147 years. The study was approved by the hospital’s ethics committee and all participants provided written informed consent before blood sampling. Peripheral blood was collected by venipuncture with a 21-gage needle. The first millilitre was discarded and then blood collected in a 2.8 ml citrate tube and 4 heparin tubes of 4 ml (Vacutainer, BD Biosciences). PBMC isolation and storage Within 1 hour of sampling, peripheral blood collected in heparin tubes (16 ml) was processed. PBMCs were isolated by density gradient centrifugation with LymphoprepTM (Abbott), following the manufacturer’s instructions. Cells were frozen in RPMI medium 1640 with LGlutamine (Gibco, Thermo Fisher) supplemented with 10% foetal bovine serum, 10,000 U/ml penicillin, 10,000 μg/ml streptomycin and 10% DMSO and stored in liquid nitrogen until used. For flow cytometry and cell culture experiments PBMCs were thawed and immediately washed and resuspended in the fresh RPMI medium to remove DMSO. EV isolation Citrate tubes were immediately processed after blood collection. EVs were isolated as previously described by our group [202]. Briefly, tubes were centrifuged at 2,500 g for 15 min, and the obtained plasma was then centrifuged at 13,000 g for 2 min and this supernatant centrifuged again at 20,000 g for 20 min to pellet EVs. The pellet was resuspended with filtered DPBS (GIBCO, Thermo Fisher), filtered twice through a 0.22 m-pore filter. Resuspended EVs were stored at -80 ºC. Cryo-electron microscopy (cryoEM) EVs were vitrified following standard protocols [336]. Glow-discharged Quantifoil holey carbon film grids (Orthogonal Array of 2µm Diameter Holes - 2µm Separation, mounted on a 300M Cu grid, #657-300-CU, Ted Pella) were vitrified in liquid ethane in Vitrobot after deposition of 3 µL of the sample. Cryo-transfer sample holders of the type GATAN Model 626 were used to keep the sample vitrified during electron microscopy analysis. The sample was observed in a JEM-2100F UHR (80-200kV, JEOL) field emission gun transmission electron microscope at different magnifications. Micrographs were recorded on a state of the art TVIPS F216 CMOS camera (2k x 2k). Nanoparticle tracking analysis The size distribution and concentration of isolated plasma EVs were measured using a ZetaView (Particle Metrix) instrument following manufacturer instructions. Samples were thawed on ice and diluted with filtered DPBS to get accurate acquisitions. Settings were fixed and maintained for all samples. Filtered DPBS was tested and no background signal was 148 | Chapter three detected. For each sample, two cycles of analysis at 11 positions were performed and results were analysed with ZetaView 8.04.02 software (Particle Metrix). PBMC and EV culture Thawed cells were cultured in 96-well flat-bottom plates in RPMI medium supplemented with 10% exosome-depleted FBS (Gibco, Thermo Fisher), 10,000 U/ml penicillin and 10,000 μg/ml streptomycin. 105 cells were plated in each well and immediately after, 100 μg of thawed EVs (measured by protein quantification with Bio-Rad Protein Assay) were added to the corresponding wells. Cells were cultured in 200 μl medium, at a final density of 106 cells per ml and incubated for 3 h at 37 °C and 5% CO2. Then, activation of cells was induced by adding 10 μg/ml phytohemagglutinin (PHA) (Sigma-Aldrich) in corresponding wells. All cultured cells were incubated for 72 h at 37 °C and 5% CO2. A schematic representation of the coculture protocol is presented in Figure 43. PHA was chosen to induce a polyclonal, nonspecific and significant lymphocyte activation, similar to the one produced against infection agents [337]. The 10 μg/ml concentration of PHA was established after titration. In a sample from a healthy adult 8 different concentrations of PHA (1.25-50 μg/ml) were tested and 10 μg/ml was chosen as the best stimulation (Figure 44). Figure 43. Cell culture protocol to test the influence of EVs on T cell activation. (A) 105 PBMCs were plated in 96-well dishes and then, 100 μg of EVs were added, while DPBS was added in control wells. 3 hours later, 10 μg/ml PHA were added to induce T cell activation in half of the wells, while the rest was maintained with no stimulation. 3 days after plating, T cell activation was evaluated by flow cytometry. (B) Schematic representation of the different study conditions. Chapter three | 149 Figure 44. T cell activation under PHA stimulation, measured by flow cytometry. PHA was titrated with a healthy adult lymphocyte sample and the 10 μg/ml concentration was chosen as the best stimulation. Flow cytometry For the flow cytometric analysis of PBMCs, the following fluorochrome-conjugated anti-human monoclonal antibodies were used: anti-CD3 APC-Fire750, and anti-CD56 APC from Biolegend; Anti-CD8 FITC, anti-CD28 PE, anti-CD4 PE-Cy7 and anti-CD25 PE from BD Biosciences; for cell viability assessment 7-aminoactinomycin D (7-AAD) dye (Thermo Fisher). Different antibody panels were designed. To assess the T cell population percentages and the senescence state of T cells, the combination of anti-CD3 APC-Fire750, anti-CD56 APC, anti-CD8 FITC, anti-CD28 PE, anti-CD4 PE-Cy7 and 7-AAD was used. T cells were identified by CD3+ staining, NK cells by CD3-/CD56+ staining and B cells as double negative CD3-/CD56-. The same panel without 7AAD was applied for the flow cytometry of plasma EVs. For cultured PBMC activation measurement anti-CD8 FITC, anti-CD4 PE-Cy7, anti-CD25 PE and 7-AAD were combined. Directly thawed PBMCs and PBMCs from cell culture were stained following the same protocol. Cells were washed and resuspended in DPBS with 5 % bovine serum albumin (BSA) (SigmaAldrich) to block Fc receptor before staining. Corresponding antibodies were added and samples incubated for 20 min at room temperature in the dark. Then, cells were washed to remove unbound antibodies and acquired in a FACS Canto II flow cytometer (BD Biosciences) or in a Guava EasyCyte 8HT flow cytometer (Millipore, Merck). Single staining and fluorescence minus one (FMO) control tubes were used to adjust compensations and set the gating strategy. After gating for singlets, lymphocytes were gated based on FSC and SSC and 20,000 lymphocytes were acquired for each sample. Then, lymphocyte populations were distinguished based on fluorescence and analysis of obtained results was performed with FACS Diva 8.0.1 (BD Biosciences) and InCyte 3.1 (Millipore, Merck) software respectively. The gating strategy for senescent T cells and representative dot plots are presented in Figure 45. 156 | Chapter three The coculture of PBMCs and EVs improves viability and influences cytokine secretion Next, coculture experiments of PBMCs and plasma EVs were performed. Cell and EV samples of all ages were tested. Four different conditions were assayed: PBMCs alone, PBMCs + EVs, PBMCs + PHA and PBMCs + EVs + PHA. The PHA was applied to stimulate T cell activation, and to test the effect of EVs both under non-stimulated and stimulated conditions. To test whether the addition of EVs affects cell viability, we analysed cells by flow cytometry and compared the 7-AAD negative events between groups after three days in culture. The different conditions of each PBMC donor were normalized to the control wells where only cells were plated. Interestingly, we observed that cell viability improves when EVs are present (Figure 50A). Moreover, PHA stimulation significantly reduces viability and this effect is partially rescued when EVs are added (Figure 50A). In a further analysis of these results, we compared the effect of plasma EVs on cells of adult (20-49 years) and aged (>80 years) donors. The positive effect of EVs is stronger in PBMCs from adults for all conditions tested (Figure 50B). In order to check whether the coculture with EVs could also affect cytokine production in vitro, we performed a luminex assay for TNF-α, IL-6, IL-10, IL-1β and IL-2. Cell conditioned media from all conditions of 2 different individuals (one adult and one aged cell donor, cocultured with EVs from adults and elders) were tested. Importantly, cytokine concentrations were nondetectable in the two conditions where PHA was not added, demonstrating that the only addition of EVs does not induce cytokine production. When compared to PHA stimulation alone, we observed that EV addition influences cytokine production. The secretion of the proinflammatory TNF-α, IL-6 and IL-1β cytokines was reduced, while anti-inflammatory IL-10 was increased and IL-2 not significantly affected (Figure 50C). Chapter three | 157 Figure 50. (Previous page) Effect of extracellular vesicles from plasma on PBMC viability and cytokine secretion in vitro. PBMCs from donors of all age ranges were cultured for 72h in the presence or not of PHA or/and plasma EVs and then analysed by flow cytometry. (A) Cell viability is reduced after stimulation with PHA, while the coculture with EVs improves viability. (B) The positive effect of plasma EVs on cell viability is stronger in cells from adults (20-49 years) than aged (80-101 years) individuals. (C) The analysis of conditioned media by luminex showed a reduced secretion of proinflammatory cytokines TNF-α, IL-6 and IL-1β and an increased secretion of anti-inflammatory IL-10 by stimulated cells cocultured with EVs compared to stimulated cells without EVs. T cell activation under PHA stimulation is affected by the coculture of plasma EVs and depends on the age of the EV donor Finally, the effect of plasma EVs on lymphocyte activation was assessed. Polyclonal activation of T cells was induced with PHA and measured by CD25 expression by flow cytometry. The coculture of lymphocytes with plasma EVs for 72h did not induce T cell activation (Figure 51CD) demonstrating that plasma EVs alone are not immunogenic for non-stimulated cells. PBMC samples of 22 individuals (12 adults and 10 elders) were tested, and each cell donor was assayed with different EV donors (up to 12 different EVs for one PBMC donor, each one in different wells and always in duplicate). The percentage of T cells activated under the same PHA stimulation (and without EVs) was highly heterogeneous for each PBMC donor (33-92 % of CD25+, Figure 51A-B). For normalization, control wells with PBMCs + PHA without EVs were used and fold change was calculated. Figure 51. T cell activation measured by flow cytometry. (A-B) PBMCs were stimulated with 10 μg/ml PHA and 72h later CD25+ cells measured. The percentage of activated CD4 and CD8 cells is heterogeneous and not correlated to age. (C-D) PBMCs were cocultured with plasma EVs. The coculture of EVs alone do not induce T cell activation. 158 | Chapter three Our results show that EVs modulate T cell activation, but the effect is very heterogeneous and is influenced by the age of the EV donor (Figure 52A-D). Taking this into consideration, we performed a separated analysis for plasma EVs from each age range. EVs from adults significantly increase CD4 cell activation, while the ones from nonagenarians and centenarians reduce the activation (Figure 53A). In the case of CD8 cells, EVs from adults also enhance cell activation (Figure 53B). Importantly, when the tendency of the whole cohort was analysed, we saw that the activation enhancement capacity of EVs significantly decreases with age (Figure 53). Figure 52. Analysis of activated lymphocytes under PHA stimulation and the influence of the EV donor age. (A-B) The coculture of PBMCs with EVs under PHA stimulation affects cell activation in a heterogeneous manner. For each cell donor, wells without EVs were taken as reference for fold change calculation. (C-D) Both CD4 and CD8 cells get more activated in the presence of EVs from adult donors when compared to EVs from elder donors. Adults 20-49 and elders 70-104 years. Chapter three | 159 Figure 53. T cell activation under PHA stimulation and the effect of plasma extracellular vesicles. PBMCs from donors of all age ranges were cultured for 72h in the presence of PHA and plasma EVs and then analysed by flow cytometry. Wells without EVs were taken as reference for fold change calculation and Wilcoxon tests. (A) The presence of EVs from adult donors resulted in the promotion of CD4 cell activation, an effect that decreases gradually with EV donor age (in red, Jonckheere test ****). (B) In a similar way, CD8 cells cocultured with EVs from adults get more activated, but this effect decreases with EV age (in red, Jonckheere test ****). Age range in years. Discussion The present study analysed peripheral blood samples from adults and elders of different age ranges, 20-49 and 70-104 years. First, the lymphocyte subsets were compared and an increased T cell and decreased B cell and NK cell proportions were found with age. Several works have studied the lymphocyte subsets with aging, both at the total number and percentage level. Distinct aging patterns have been found between countries and populations, pointing out the complexity and heterogeneity of the immune system and immunosenescence [334,338–342]. To our knowledge, our work is the first investigating lymphocyte populations with age in the Basque Country (in the north of Spain). On the other hand, some aging-related features have been widely reported, such as the loss of the costimulatory molecule CD28 from T cell membrane and a subsequent gain of NK characteristic markers [100,102,107,343]. This process has been shown to affect both CD8 and CD4 T cells, but earlier and to a greater extent to CD8 cells [104,344,345]. Our results are in accordance with previous reports. Nevertheless, previous studies reported a gradual accumulation of CD4 CD28cells with age [98,103,346] and here we showed a higher senescent CD4 cell percentage in the 80-89 age range, and interestingly a lower percentage in nonagenarians and centenarians, following a quadratic effect. We hypothesize that rather than CD28 expression recovery, individuals reaching >90 years could be the ones that presented lower senescent cell proportions also earlier in life. However, a longitudinal study with a larger 160 | Chapter three sample size would be needed to confirm this progression. As mentioned above, other works previously described the loss of CD28 expression in CD4 cells in elders, but to our knowledge, our study is the first one analysing this effect in nonagenarians and centenarians and demonstrates that they follow a distinct aging progression in some aspects. To further characterize immunosenescence, we focussed on plasma EVs. They are known to carry many molecules in their membrane and among them, EVs can also bear markers of the secreting cell [347]. To test whether plasma EVs resemble the senescence status of T cells, we measured T cell membrane markers on EVs. Our results showed that plasma EVs carry T cell specific molecules, while there is not an increased proportion of “senescent-like EVs” with age. Even if no significant differences were found between age ranges, a higher percentage of CD28EVs was observed among CD8 EVs when compared to CD4 EVs, which could be linked to the increased CD28CD8 T cells. Importantly, we also identified the characteristic tetraspanins CD9, CD63 and CD81 of EVs by flow cytometry. A small percentage of circulating EVs in plasma carry these molecules, but it should be noted that observed numbers could be underestimated by other co-isolated particles and that EVs expressing a single or few copies of the surface antigen of interest cannot be detected, as described in previous studies [218,348]. Moreover, it should be mentioned that the detection of EV proteins by flow cytometry is a direct measure that identifies proteins at their physiologic state – at the EV membrane in this case –, in contrast to techniques such as western blotting or proteomics approaches where vesicles are lysed, and the numbers of positive particles for each protein cannot be measured. Regarding coculture experiments of PBMCs and plasma EVs, we showed that EVs from a different donor are not immunogenic for receptor lymphocytes, and in contrast, they affect cell viability and cytokine secretion. Specifically, plasma EVs enhance cell viability and partially rescue the deleterious effect of PHA, the well-known activation-induced cell death [349]. This positive effect is stronger on cells from adults, when compared to elders. Our results indicate for the first time that the presence of plasma EVs in culture can partially rescue the activationinduced cell death and moreover, that EVs enhance T cell viability when compared to the culture of cells alone. Further, plasma EVs reduce the secretion of TNF-α, IL-6 and IL-1β proinflammatory cytokines and increase anti-inflammatory IL-10 in PHA stimulated cells, but EVs alone do not alter cytokine secretion of PBMCs. It has been widely described that PHA stimulates cytokine production [350], but the effect of EVs is still not understood. A previous study reported a similar effect of mesenchymal cell-derived EVs on IL-10 production [351], and some authors have also studied the effect of EVs on lymphocytes and ILs [333,352]. However, they worked with EVs from other tissues or produced in culture, which can lead to distinct outputs. Chapter three | 161 And even if plasma EVs are not immunogenic, they influence T cell activation under PHA stimulation, and this effect is different depending on the age of the EV donor. EVs from adults promote T cell activation and this effect decreases with age. These results highlight the influence of circulating EVs on T cells and interestingly, also demonstrate the distinct effects of plasma EV and T cell interactions depending on age. The coculture experiments enable us to more closely resemble the interaction between circulating cells and EVs. Much work is still needed to elucidate the complex pool of particles present in plasma and the triggers of observed effects, but the present work gives a first description of the role that EVs from plasma have on T cells during aging. In short, our work describes the reduced CD28 loss of CD4 cells in nonagenarians and centenarians, the presence but no accumulation of senescent markers on plasma EVs and the distinct interactions between T cells and plasma EVs with age. CHAPTER FOUR Multiple sclerosis and premature aging Chapter four | 165 Introduction Aging is a universal process. It affects both healthy individuals as well as the ones that present other syndromes and/or diseases. Importantly, aging and the concomitant health problems interact and influence each other. These interactions are observed in diseases that develop in elder people, such as cancer, but the influence of aging also reaches chronic diseases as patients age, like in the case of MS [275]. MS is a chronic autoimmune disease of the CNS characterised by pathologic demyelination of axons and subsequent neurodegeneration. It is a heterogeneous disease and, clinically, it can follow relapsing-remitting or progressive forms [353]. Most of MS patients experience the first symptoms at their 20s or 30s, but there are also paediatric or juvenile [271] and late-onset MS cases [272]. In the last decades, effective disease-modifying treatments that slow the progression of MS have been developed [267], and thanks to them, patients present increasing age at disability milestones [273]. Consequently, the mean age of MS patients is increasing, with already more than 20% of them aged 60 years or over (msbase.org [274]). Therefore, an elevated number of patients suffer from the aforementioned interactions between MS and aging processes. It should be noted, that the characteristic features of MS are very similar to the ones observed during aging, as inflammation, and immune alterations. Moreover, the consequences such as mobility and cognitive problems are also found in both processes, which makes it very complicated to separate the effects of MS and aging [275]. In addition, some authors have suggested that several autoimmune diseases, including MS, show premature aging, specially with regards to the immune system. They observed reduced numbers of naïve CD4 T cells in pediatric MS [354] and increased levels of CD4+CD28T cells in adult MS patients [280,281]. Moreover, the characteristic loss of the costimulatory molecule CD28 from CD4 T cells due to the repeated stimulation and activation is proposed as a sign of senescence and terminal differentiation, but it has been shown that in MS patients these cells remain functional and show increased cytotoxicity [355]. Furthermore, CD4+CD28T cells have also been found in MS lesions in the CNS, suggesting that they could be implicated in MS pathogenesis [356]. Regarding inflammation, elevated levels of TNF- and IL-6 among other inflammatory markers have been found in the cerebrospinal fluid and serum of MS patients during remission, indicating that some signs of chronic inflammation are present [285,286]. The aim of this work was to perform a pilot study to evaluate whether differences are found in age-related features due to the presence of MS disease. We tested the above-mentioned inflammaging (by measuring TNF-, IL-6 and CRP concentrations) and T cell senescence (by