Ho EVs – how empe a u e a ec s ex acellula esicles
Eilien Schulz 1,2, Anna Ka agianni 1, Ma cus Koch 3, G ego Fuh mann 1,2*
1 Biogenic Nano he apeu ics G oup (BION), Helmhol z Cen e o In ec ion Resea ch (HZI), Helmhol z
Ins i u e o Pha maceu ical Resea ch Saa land (HIPS), Campus E8.1, Saa b ücken 66123, Ge many
2 Depa men o Pha macy, Saa land Uni e si y, Campus E8.1, Saa b ücken 66123, Ge many
3 INM – Leibniz Ins i u e o New Ma e ials, Campus D2.2, Saa b ücken 66123, Ge many
*Co esponding au ho , phone: +49 68198806 1500, Email: [email p o ec ed]
ORCID IDs: Eilien Schulz: 0000-0002-9769-8980, Anna Ka agianni: 0000-0002-0831-1247, G ego
Fuh mann: 0000-0002-6688-5126
Keywo ds: ex acellula esicles, ou e memb ane esicles, lymphoblas oid cells, myxobac e ia, d ug
ca ie s, low cy ome y, hea s abili y, au ocla ing
ABSTRACT
In ecen yea s, ex acellula esicles (EVs) and ou e memb ane esicles (OMVs) ha e become an
ex ensi e and di e se ield o esea ch. They hold po en ial as diagnos ic ma ke s, he apeu ics and
o undamen al biological unde s anding. Despi e ongoing s udies, nume ous in o ma ion ega ding
unc ion, con en and s abili y o EVs emains unclea . I EVs and OMVs ough o be used as
he apeu ics and in clinical en i onmen s, hei s abili y is one o he mos impo an ac o s o be
conside ed. Especially o o mula ion de elopmen , EVs and OMVs need o be s able a highe
empe a u es. To he bes o ou knowledge, e y li le wo k has been published ega ding hea
s abili y o nei he EVs no OMVs. In he p esen s udy, we in es iga ed B lymphoblas oid cell-
de i ed EVs and OMVs de i ed om myxobac e ial species So angiineae as model esicles. We
exposed he esicles o 37 °C, 50 °C, 70 °C and 100 °C o 1 h, 6 h and 24 h, and also au ocla ed hem.
Physico-chemical analysis such as size, pa icle concen a ion and p o ein concen a ion showed
in e es ingly mino al e a ions, pa icula ly a 37 °C. Flow cy ome y analysis emphasised hese
esul s sugges ing ha a e hea impac , EVs and OMVs we e s ill able o be aken up by
mac ophage-like dTHP-1 cells. These da a indica e ha bo h mammalian and bac e ial esicles show
in insic s abili y a physiological empe a u e. Ou indings a e impo an o conside o esicle
o mula ion and o ad anced bioenginee ing app oaches.
1. In oduc ion
Ou e memb ane esicles (OMVs) we e i s men ioned 50 yea s ago in 1967, as Cha e jee e al
de ec ed pa icles in p oximi y o Vib io chole a memb anes in elec on mic oscopy images (1). A ew
yea s la e , in 1983 Pan e al. we e he i s ones o obse e ex acellula esicles (EVs), while hey
moni o ed a ans e in ecep o in sheep (2). Since hen, his a ea o esea ch has expanded
ex ensi ely. Bo h, EVs and OMVs a e nano-sized phospholipid bilaye ed assemblies and se e as
anspo ehicles o cell-cell communica ion (3). Thei s uc u e and su ace is, in mos cases,
compa able o hei cellula o igin and consis s o ecep o s, p o eins o o OMVs
lipopolysaccha ides (4, 5). Con en s may a y, bu gene ally imply nucleic acid, p o eins and
seconda y me aboli es, such as oxins o compounds ha a e o en unique o hei o igin (4-6). EVs
ha e been isola ed om a la ge a ie y o cells de i ed om he immune sys em, di e en issues o
he ne ous sys em (3). OMVs, on he o he hand ha e been isola ed om almos all known g am-
nega i e bac e ia (7). Bo h, EVs and OMVs hold po en ial o he de elopmen o new he apeu ics.
They ha e al eady been applied in issue epai , neu odegene a i e diso de s and cance he apy (8).
Fo example, EVs de i ed om mesenchymal s em cells ha e p e iously eached clinical ials as
no el he apeu ics o unc ional eco e y a e ischemic s okes (9). OMVs, on he o he hand ha e
been s udied o accina ion, as d ug deli e y sys ems agains cance (10) o in ec ions (11). Di e en
ou es o EV applica ion ha e been es ablished, ei he as in a enous suspensions o inco po a ed
in o hyd ogels, o example o enzyme p od ug he apy (12, 13). Al hough many p o ocols ha e
been es ablished conce ning esicle isola ion, he empe a u e s abili y o EVs and OMVs has no
been s udied comp ehensi ely. Some publica ions analysed s o age condi ions o EV suspensions,
while we and o he s ha e ecen ly in oduced a eeze-d ying me hod using c yo-p o ec an s (14-16).
To ou knowledge, e y li le is known o da e conce ning he s abili y o EVs and OMVs a inc easing
empe a u es. Lee e al. s udied he s abili y o EVs om Kaposi's sa coma-associa ed he pes i us-
in ec ed human endo helial cells and ound ha e en a e 4 days he pa icle concen a ion o
samples incuba ed a 37 °C did no al e (17). Mo eo e , Cheng e al de e mined he pa icle
concen a ion o EVs a 37 °C and a 60 °C, esul ing only in mino physical changes (18). Howe e , i
is essen ial o es ima e he hea s abili y o EVs and OMVs in a mo e comp ehensi e manne , as i
will help o e alua e hei clinical and pha maceu ical applicabili y. In e ms o pha maceu ical
applicabili y, hea will also play a ole du ing sp ay d ying (19) o in he chemical modi ica ion o
a ach a ge ing moie ies (20). In addi ion, o unde s and he biological ole o EVs, highe
empe a u es could be necessa y. As ex acellula esicles a e o en comp ised o phospholipids,
such as phospha idylcholine, sphingomyelin and phospha idylse ine, hey a e simila in s uc u e
compa ed o liposomes (21). Hea -induced usion o EVs wi h liposomes may also be bene icial o
o m new biocompa ible nanoca ie s (22). Fo in a enous injec ion, EVs need o be s e ile as an
asep ic p oduc ion is no always possible. S e ilisa ion me hods such as il a ion may esul in a loss
o sample, leading o concen a ion issues (23). P essu ised sa u a ed s eam s e ilisa ion, au ocla ing
may be con enien and quick, bu po en ially ha sh me hod o ob ain s e ile esicles o clinical
applica ions.
He e, we used EVs de i ed om B lymphoblas oid cells (RO cells). They a e a well-de ined and
comme cially a ailable cell line ob ained om a pa ien wi h se e e combined immunode iciency,
no exp essing MHC class II complexes (24, 25). Thus, RO cell de i ed EVs may be low in
immunogenici y. The cells can be g own in suspension and, he e o e can be easily cul i a ed e en in
la ge quan i ies. We also es ablished a me hod o cul i a e RO cells in a low space consuming
manne . As a second model esicle we s udied OMVs de i ed om he myxobac e ial s ain SBS 073.
Myxobac e ia a e g am-nega i e, soil li ing bac e ia, ha a e p oduce o a la ge a ie y o seconda y
me aboli es (26). We ecen ly showed ha SBS 073 OMVs a e non- oxic o human cells and ha e he
po en ial o be u he de eloped as a d ug ca ie sys em (11). As myxobac e ia a e ound in a ious
en i onmen s including dese s (27), we hypo hesised ha he OMVs hey p oduce may be mo e
esis an o high empe a u es compa ed o human EVs, which a e adap ed o a body empe a u e o
37 °C.
2. Ma e ials and Me hods
2.1. Cell cul u e
B lymphoblas oid cells (RO cells) (DSMZ, ACC 452, B aunschweig, Ge many) we e cul u ed in T75
lasks wi h an ini ial seeding densi y o 0.75*106 cells/mL. A e hawing cells, hey we e g own in
RPMI (Gibco) wi h 15% ( / ) oe al cal e se um (FCS) (25). Be o e EV isola ion, RPMI wi h 10% ( / )
insulin- ans e in-selenium-e hanolamine (The mo Fische ) was used. Cul u es s a ed wi h a
olume o 45 mL in an up igh posi ion. A e 3 days, 25 mL supe na an was emo ed and eplaced
wi h 50 mL esh medium. Ano he subsequen 4 days la e , 50 mL supe na an was emo ed o EV
isola ion. Supe na an s we e s o ed a -80 °C up o 2 mon hs. Cells we e used un il passage 40. THP-1
cells we e cul i a ed in RPMI wi h 10 % ( / ) FCS. Al e na ing, cells we e seeded wi h an ini ial
densi y o 2 o 3 million cells and cul i a ed o 3 o 4 days.
2.2. Mic obial cul u e
SBS 073 myxobac e ia (kindly p o ided by Rol Mülle , Depa men o Mic obial Na u al P oduc s,
Helmhol z Ins i u e o Pha maceu ical Resea ch, Saa b ücken) we e cul i a ed as desc ibed
p e iously (11), in 2SWT medium (0.3% (m/ ) bac o yp one, 0.1% (m/ ) soy one, 0.2% (m/ )
glucose, 0.2% (m/ ) soluble s a ch, 0.1% (m/ ) mal ose monohyd a e, 0.2% (m/ ) cellobiose, 0.05%
(m/ ) CaCl2*2H2O, 0.1% (m/ ) MgSO4*7H2O and 10mM HEPES, pH 7.0 adjus ed wi h KOH). The
bac e ial suspension was cul i a ed a 30 °C and shaken a 180 pm (Eco on, In o s HT, Bo mingen,
Swi ze land) o one week un il OMV isola ion. As his s ain o ms agg ega es, i was no possible o
de e mine a g ow h cu e based nei he on op ical densi y measu emen s, no on colony o ming
uni coun ing (11). OMVs we e isola ed om cul u es cul i a ed un il passage 6.
2.3. Isola ion and pu i ica ion o EVs and OMVs
Fi y millili es o RO supe na an we e i s cen i uged a 300 × g o 8 min o emo e cells. Fo y
millili e we e hen ans e ed o a new ube and cen i uged a 9,500 × g o 15 min. SBS 073
supe na an was i s cen i uged a 9,500 × g o 10 min o emo e bac e ia (28). Fo y millili e o
his supe na an we e cen i uged, a 9,500 × g o 15 min. Bo h samples we e hen ul acen i uged
a 100,000 × g o 2 h a 4 °C (Ro o SW 32 Ti, Beckman Coul e , B ea, USA) (29). Pelle s we e
esuspended in ei he 400 µL phospha e bu e ed saline (PBS, Gibco PBS able s wi hou calcium,
magnesium and phenol ed) o , in case o RO EVs, in 500 µL cell cul u e supe na an . Vesicles we e
pu i ied by size exclusion ch oma og aphy, using a 30 mL (SBS 073 OMVs) o a 10 mL (RO EVs)
sepha ose CL-2B (GE Li e Science, Uni ed Kingdom) column, collec ing 1 mL ac ions wi h PBS as
elu ion bu e .
2.4. Physico-chemical cha ac e isa ion o EVs and OMVs
Nanopa icle T acking Analysis (NTA LM-10, Mal e n, Mal e n, Uni ed Kingdom) was used o
de e mine pa icle concen a ions and hei hyd odynamic diame e (30). Samples we e dilu ed up o
1000 old in o de o ha e a concen a ion o 10 o 100 pa icles pe ame. A 30 s ideo a a came a
le el o 14 o 15 was eco ded 3 imes be o e he pa icle concen a ion was calcula ed by NanoSigh
3.3 so wa e wi h a de ec ion h eshold o 5. A bicinchoninic assay ki (Sigma Ald ich) was used o
quan i y p o ein concen a ions o each sample in duplica es, acco ding o he manu ac u e ’s
speci ica ions be o e and a e hea ea men . To quan i y he o al p o ein con en in esicles, 25 µL
o RIPA bu e (50 mM T ic-HCl, 150 mM NaCl, 0,5% deoxycholic acid, 1% NP-40, 0,1% sodium-
dodecyl-sul a e) we e incuba ed wi h 75 µL o sample o 5 min be o e ano he bicinchoninic assay
was pe o med.
2.5. Hea es ing
The wo SEC ac ions wi h he highes pa icle concen a ion we e pooled ( inal olume 2 mL),
ans e ed o glass con aine s, ai igh sealed wi h caps and used o hea expe imen s. Samples
we e incuba ed in an incuba o (Memme UN 75, Schwabach, Ge many) wi h a cons an
empe a u e o 37 °C, 50 °C, 70 °C o 100 °C. A e 1 h, 6 h and 24 h e apo a ed wa e was measu ed
by weigh di e ence, eplaced and samples we e used o u he expe imen s. To au ocla e esicles,
hey we e injec ed in o b own glass con aine s wi h ubbe plugs and sealed wi h me al caps
(Zscheile & Klinge GmbH, Hambu g, Ge many) and hea ed up o 121 °C o 20 min a 2 ba . The
empe a u e o a wa e con ol wi h he same olume was used. Acco ding o he Eu opean
Pha macopoeia, his me hod is one o he ecommended me hods o s e ilisa ion (31). To es
s e ili y, 100 µL o au ocla ed samples we e incuba ed on lysogeny b o h aga pla es (Sigma Ald ich)
o 4 days a 37 °C.
2.6. Elec on mic oscopy
To pe o m c yogen elec on c yomic oscopy, esicles we e concen a ed using cen i ugal il e s
(Ul acel YM – 30) un il one-hund ed h o olume was le . Th ee mic oli e s o his solu ion was
placed on o a holey ca bon ilm ( ype S147-4, Plano, We zla , Ge many) and plo ed o 2s wi h a
Ga an ((Pleasan on, CA, US) c yoplunge model CP3, be o e plunging in o liquid e hane a T = 108 K.
Unde liquid ni ogen, esicles we e ans e ed o a Ga an model 914 c yo-TEM sample holde . A T
= 100 K samples we e imaged ia b igh ield TEM (JEM-2100 LaB6, Jeol, Akishima, Tokio, Japan)
unde low-dose condi ions.
2.7. FACS analysis o THP-1 wi h esicles
OMVs and EVs we e s ained wi h 2 µL o DiI (Vyb an DiI Cell-labelling solu ion 1 mM) o 15 min a
37 °C. A size exclusion ch oma og aphy wi h sepha ose CL-2B was pe o med o emo e non-
inco po a ed dye. Fluo escence in ensi y (λEx/ λEm 490/570 nm) was measu ed o each sample. The
ac ion wi h he highes in ensi y was used o u he expe imen s. I is impo an o men ion, ha
au ocla ed samples had o be cen i uged o 2.5 min a 9,500 × g in o de o emo e dye agg ega es
induced by hea and p essu e. THP-1 (DSMZ, ACC16, B aunschweig, Ge many) we e seeded in o 48
well pla es wi h a densi y o 200,000 cells pe well. THP-1 we e s imula ed wi h 7.5 ng/mL pho bol
12-my is a e 13-ace a e (PMA) (Sigma Ald ich) o 24 h, o s imula e he di e en ia ion o
mac ophage like dTHP-1. A e wa ds cells we e incuba ed wi h 100 µL o each esicle sample o 24
h, esul ing in a a io 2.5 × 105 RO EVs and 1 × 106 SBS 073 OMVs pe cell. Two washing s eps wi h
PBS we e ca ied ou be o e cells we e incuba ed wi h accu ase solu ion (Sigma Ald ich) o 20 min a
RT o de ach he cells. Fou wells we e pooled o pe o m one low cy ome y (FACS) (LSRFo essa X-
20, BD) analysis. Cells incuba ed wi h 100 µL PBS se ed as con ol. A ed lase a 561 nm (PE
phycoe y h in) was used o de ec DiI labelling. Ten housand e en s pe sample acqui ed by BD
FACSDi a 8.0.2, we e analysed wi h FlowJo so wa e e sion 7.6.5.
2.8. Con ocal imaging

S imula ed THP-1 (see 2.8.) wi h a densi y o 200,000 cells pe well, we e incuba ed wi h 100 µL o
each DiI labelled EV o OMV sample in an 8 well chambe pla e (SPL Li e Science) o 24 h. A e
emo ing he supe na an , cells we e s ained wi h luo escein labelled whea ge m agglu inin (Vec o
labo a o ies) o 15 min a 37 °C. Subsequen ly, cells we e ixed wi h 3.7 % ( / ) pa a o maldehyde
o 20 min a oom empe a u e (RT). Nucleus s aining was pe o med using a 1 µg/mL 4′,6-
Diamidino-2-phenylindole dihyd ochlo ide (DAPI) (Sigma Ald ich) solu ion (32). Fo con ocal imaging
(Leica TCS SB8) a 488 nm lase was used o isualise luo escein, a 405 nm lase o DAPI and a 561
nm lase o DiI. Leica Applica ion Sui e X so wa e was used o p ocess he images.
2.9. S a is ical analysis
All da a is epo ed in mean (x) and s anda d de ia ion (SD), whe e n indica es he numbe o
independen expe imen s. S a is ical analysis was pe o med by SigmaPlo 14.0 using One-way
ANOVA , ollowed by a Tukey pos -hoc es , o compa e he di e en g oups. Signi ican p- alues
we e s a ed as * o p < 0.05, ** o p < 0.01 o wi h he exac p- alue.
3. Resul s and Discussion
3.1. Hea -induced physico-chemical al e a ion o esicles
In his wo k, we analysed he hea s abili y o wo di e en ypes o esicles, EVs de i ed om B
lymphoblas oid RO cells and OMVs de i ed om he myxobac e ial s ain SBS 073. Fo his, he
esicles we e isola ed, pu i ied and incuba ed a 37 °C, 50 °C, 70 °C and 100 °C o 1 h, 6 h and 24 h.
The size o EVs and OMVs emained cons an a 37 °C, e en a e 24 h. Ye , a highe empe a u es o
50°C, 70 °C and 100 °C small changes we e de ec ed (Fig. 1 a,d). The pa icle concen a ion o RO EVs
dec eased he highe and longe he samples we e incuba ed. The mos d as ically change was
de ec ed a 100 °C, when pa icle concen a ions dec eased almos en old (Fig. 1 b). SBS 073 OMVs
also showed a s ong pa icle concen a ion dec ease a e 24 h a 100 °C wi h a emaining 36% (Fig.
1 e). Con ibu ing o his, he size dis ibu ion o esicles ea ed a 37 °C o 24 h showed simila
ends compa ed o hei con ols (Fig. 2 a,c), whe eas he 100 °C samples showed s onge
a ia ions (Fig. 2 b, d). I one looks a he p o ein concen a ion, he e again, high empe a u es und
longe incuba ions imes esul ed in physico-chemical al e a ion. Especially RO EVs showed an
inc ease o ca. 140 % o p o ein concen a ion a e 24 h a 100 °C (Fig. 1 c). In con as , he p o ein
concen a ion o SBS 073 samples emained ela i ely cons an , e en a high empe a u es. We
hypo hesise, ha he dec ease o he pa icle concen a ion is due o a dis up ion o he esicles
hemsel es, leading o a elease o encapsula ed p o eins and hus an inc ease o p o ein
concen a ion. The highe he empe a u e, he less s able he esicles we e and he leakie hese
nanos uc u es became. Ne e heless, o bo h, pa icle and p o ein concen a ion, RO EVs al e ed
mo e d as ically compa ed o SBS 073 OMVs, sugges ing a highe physico-chemical s abili y o
bac e ia de i ed esicles. Con ibu ing o his e ec is he common en i onmen bo h ypes o
esicles can be ound in. As he o igin o RO EVs is he human body, which is main ained a 37 °C,
hey a e mo e likely s able a 37 °C. SBS 073 OMVs, howe e , a e de i ed om myxobac e ia, a
g am-nega i e popula ion ha has been adap ed o a ious en i onmen s, om he sou h pole and
opical ain o es s o dese s (27). In o de o su i e, one o hei communica ion ools, hei OMVs
need o be s able in hose ha sh condi ions. I is he e o e likely, ha he OMVs a e mo e s able han
he EVs.
Fig 1 Physico-chemical al e a ion a e incuba ion a 37 °C, 50 °C, 70 °C, 100 °C o 1 h, 6 h and 24 h. a) RO EV size
dis ibu ion be o e and a e 24 h a 37 °C b) RO EV size dis ibu ion be o e and a e 24 h a 100 °C c) RO EV no malised
pa icle concen a ion measu ed by NTA d) RO EV no malised p o ein concen a ion de e mined ia BCA e) SBS 073 OMV
dis ibu ion be o e and a e 24 h a 37 °C ) SBS 073 OMV size dis ibu ion be o e and a e 24 h a 100 °C g) SBS 073 OMV
no malised pa icle concen a ion h) SBS 073 OMV no malised p o ein concen a ion. Mean ± SD, n = 3, *p < 0.05 (ANOVA
ollowed by Tukey pos -hoc es ). Samples we e no malised o pa icle and p o ein concen a ions be o e he hea
ea men . The dashed line indica es he a e age alue o all samples a ime poin 0 h.
Fig 2 Rep esen a i e size dis ibu ion be o e and a e incuba ion a 37 °C, 100 °C o 24 h. a) RO EV size dis ibu ion a e
24 h a 37 °C b) RO EV size dis ibu ion a e 24 h a 100 °C c) SBS 073 OMV size dis ibu ion a e 24 h 37 °C d) SBS 073
OMV size dis ibu ion a e 24 h a 100 °C.
3.2. Impac o au ocla ing on esicles
Au ocla ed samples showed compa able esul s o samples incuba ed a 100 °C. As seen in Figu e 3,
pa icle concen a ions dec eased while p o ein concen a ions inc eased. The mean size inc eased
sligh ly, and he size dis ibu ion became b oade (Fig. 3 a, b). Howe e , he e again, he elease o
p o eins o OMV samples was no as s ong as he p o ein elease o EVs (Fig. 3 c, ).
Fig 3 Physico-chemical al e a ion o RO EVs and SBS 073 OMVs a e au ocla ing. a) RO EV size dis ibu ion b) RO EV
no malised pa icle concen a ion c) RO EV no malised p o ein concen a ion d) SBS 073 size dis ibu ion e) SBS 073 OMV
no malised pa icle concen a ion ) SBS 073 OMV no malised p o ein concen a ion. Samples we e no malised o pa icle
and p o ein concen a ions be o e au ocla ing. Mean ± SD, n = 3
The al e a ion o he mo phology o he esicles due o he physico-chemical changes we e also
in es iga ed by c yogenic elec on mic oscopy (c yo-EM). By using c yo-EM we also wan ed o e i y
he bulk size measu emen s de e mined by NTA. As shown in Figu e 4 he lipid bilaye o he esicles
was clea ly isible and in ac (whi e a ows). Sizes de e mined by c yo-EM indica ed smalle
diame e s compa ed o da a collec ed by NTA. Indeed, NTA u ilises B ownian mo ion o calcula e he
hyd odynamic diame e o pa icles and has a de ec ion limi o 10 nm wi h low sensi i i y (33). This
may lead o enla ged pa icle sizes compa ed o c yo-EM imaging, whe e samples a e mos likely
isualised in hei na i e s a e. As pu i ying wi h a size exclusion s ep dilu ed samples, we
concen a ed hem using cen i ugal il e s and compa ed hem wi h esuspended pelle s a e
ul acen i uga ion. We no iced ha su icien concen a ions o esicles a e c ucial in o de o
isualise hem and ha esul s conduc ed by NTA o en leads o an o e es ima ion o sample