scieee Science in your language
[en] (orig)

Regulation of TIA-1 Condensates: Zn2+ and RGG Motifs Promote Nucleic Acid Driven LLPS and Inhibit Irreversible Aggregation

Abstract

Stress granules are non-membrane bound RNA-protein granules essential for survival during acute cellular stress. TIA-1 is a key protein in the formation of stress granules that undergoes liquid-liquid phase separation by association with specific RNAs and protein-protein interactions. However, the fundamental properties of the TIA-1 protein that enable phase-separation also render TIA-1 susceptible to the formation of irreversible fibrillar aggregates. Despite this, within physiological stress granules, TIA-1 is not present as fibrils, pointing to additional factors within the cell that prevent TIA-1 aggregation. Here we show that heterotypic interactions with stress granule co-factors Zn2+ and RGG-rich regions from FUS each act together with nucleic acid to induce the liquid-liquid phase separation of TIA-1. In contrast, these co-factors do not enhance nucleic acid induced fibril formation of TIA-1, but rather robustly inhibit the process. NMR titration experiments revealed specific interactions between Zn2+ and H94 and H96 in RRM2 of TIA-1. Strikingly, this interaction promotes multimerization of TIA-1 independently of the prion-like domain. Thus, through different molecular mechanisms, these stress granule co-factors promote TIA-1 liquid-liquid phase separation and suppress fibrillar aggregates, potentially contributing to the dynamic nature of stress granules and the cellular protection that they provide.

Read accessible full text

Regulation of TIA-1 Condensates: Zn2+ and RGG Motifs Promote Nucleic Acid Driven LLPS and Inhibit Irreversible Aggregation

Author: West, Danella L.; Loughlin, Fionna E.; Rivero Rodríguez, Francisco; Vankadari, Naveen; Velázquez Cruz, Alejandro; Corrales Guerrero, Laura; Díaz Moreno, Irene; Wilce, Jacqueline A.
Publisher: Frontiers Media S.A.
Year: 2022
DOI: 10.3389/fmolb.2022.960806
Source: https://idus.us.es/bitstreams/c0487725-0be0-420a-b028-d8871ebbedbd/download
Regula ion o TIA-1 Condensa es: Zn
2+
and RGG Mo i s P omo e Nucleic Acid
D i en LLPS and Inhibi I e e sible
Agg ega ion
Danella L. Wes
1
†
, Fionna E. Loughlin
1
†
, F ancisco Ri e o-Rod íguez
2
, Na een Vankada i
1
,
Alejand o Velázquez-C uz
2
, Lau a Co ales-Gue e o
2
, I ene Díaz-Mo eno
2
* and
Jacqueline A. Wilce
1
*
1
Monash Biomedicine Disco e y Ins i u e and Depa men o Biochemis y and Molecula Biology, Monash Uni e si y, Clay on,
VIC, Aus alia,
2
Ins i u e o Chemical Resea ch, Uni e si y o Se ille—CSIC, Se ille, Spain
S ess g anules a e non-memb ane bound RNA-p o ein g anules essen ial o su i al
du ing acu e cellula s ess. TIA-1 is a key p o ein in he o ma ion o s ess g anules ha
unde goes liquid-liquid phase sepa a ion by associa ion wi h specific RNAs and p o ein-
p o ein in e ac ions. Howe e , he undamen al p ope ies o he TIA-1 p o ein ha enable
phase-sepa a ion also ende TIA-1 suscep ible o he o ma ion o i e e sible fib illa
agg ega es. Despi e his, wi hin physiological s ess g anules, TIA-1 is no p esen as
fib ils, poin ing o addi ional ac o s wi hin he cell ha p e en TIA-1 agg ega ion. He e we
show ha he e o ypic in e ac ions wi h s ess g anule co- ac o s Zn
2+
and RGG- ich
egions om FUS each ac oge he wi h nucleic acid o induce he liquid-liquid phase
sepa a ion o TIA-1. In con as , hese co- ac o s do no enhance nucleic acid induced fib il
o ma ion o TIA-1, bu a he obus ly inhibi he p ocess. NMR i a ion expe imen s
e ealed specific in e ac ions be ween Zn
2+
and H94 and H96 in RRM2 o TIA-1. S ikingly,
his in e ac ion p omo es mul ime iza ion o TIA-1 independen ly o he p ion-like domain.
Thus, h ough di e en molecula mechanisms, hese s ess g anule co- ac o s p omo e
TIA-1 liquid-liquid phase sepa a ion and supp ess fib illa agg ega es, po en ially
con ibu ing o he dynamic na u e o s ess g anules and he cellula p o ec ion ha
hey p o ide.
Keywo ds: TIA1, RNA binding p o ein, liquid-liquid phase sepa a ion, p ion-like domain, RRM, amyloid fib il, zinc,
RGG mo i
INTRODUCTION
S ess g anules p o ec he cell du ing imes o acu e s ess by seques e ing mRNAs in dynamic RNA-
p o ein compa men s whe e hey a e p o ec ed om deg ada ion (Riggs e al., 2020). Fo ma ion o s ess
g anules occu s h ough a summa ion o mul i alen p o ein and RNA in e ac ions esul ing in de-mixing
om he cy osol h ough he physical p ocess o liquid-liquid phase sepa a ion (LLPS) (Ho mann e al.,
2021). The g anules o m wi h a dense co e and a liquid-like shell, and emain in a dynamic assembled
s a e o se e al hou s un il disassembly occu s (Wheele e al., 2016). Abe an s ess g anule dynamics is
associa ed wi h neu odegene a i e diseases, including amyo ophic la e al scle osis (ALS), on o-
empo al loba degene a ion (FTLD) as well as auopa hies (Albe i and Do mann, 2019;Wolozin
and I ano , 2019;Albe i and Hyman, 2021). Specifically, delayed disassembly and educed dynamics
Edi ed by:
Tomohi o Yamazaki,
Osaka Uni e si y, Japan
Re iewed by:
Masa o Ka o,
Uni e si y o Texas Sou hwes e n
Medical Cen e , Uni ed S a es
Joseph B. Rayman,
Columbia Uni e si y, Uni ed S a es
*Co espondence:
I ene Díaz-Mo eno
[email p o ec ed]
Jacqueline A. Wilce
[email p o ec ed]u
†
These au ho s sha e fi s au ho ship
Special y sec ion:
This a icle was submi ed o
RNA Ne wo ks and Biology,
a sec ion o he jou nal
F on ie s in Molecula Biosciences
Recei ed: 03 June 2022
Accep ed: 24 June 2022
Published: 14 July 2022
Ci a ion:
Wes DL, Loughlin FE,
Ri e o-Rod íguez F, Vankada i N,
Velázquez-C uz A,
Co ales-Gue e o L, Díaz-Mo eno I
and Wilce JA (2022) Regula ion o TIA-
1 Condensa es: Zn
2+
and RGG Mo i s
P omo e Nucleic Acid D i en LLPS and
Inhibi I e e sible Agg ega ion.
F on . Mol. Biosci. 9:960806.
doi: 10.3389/ molb.2022.960806
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608061
ORIGINAL RESEARCH
published: 14 July 2022
doi: 10.3389/ molb.2022.960806
wi hin s ess g anules is widely hypo hesized o acili a e he
agg ega ion o s ess g anule-associa ed p o eins, esul ing in
pa hologic inclusions ha a e a hallma k o hese
neu odegene a i e diseases (Zhang e al., 2019).
TIA-1 is a cha ac e is ic s ess g anule p o ein ha binds
mRNA and con ibu es o s ess g anule o ma ion ia i s
C- e minal P ion-like domain (P LD) (Gilks e al., 2004). The
N- e minus o TIA-1 comp ises h ee RRM domains, o which
RRM2 and 3 (“RRM2,3”) coope a i ely bind a ge RNA wi h
high a fini y (Dembe e al., 1996;C uz-Galla do e al., 2014;
Wa is e al., 2017). In i o s udies o TIA-1 iden ified i s in insic
abili y o unde go LLPS ei he alone (Mackenzie e al., 2017)o as
enhanced by mul i-si e a ge nucleic acid (Loughlin e al., 2021).
These same s udies also showed ha , o e ime, TIA-1 p og esses
o o m i e e sible ß-shee ich agg ega es and ha his is also
enhanced in he p esence o a ge nucleic acid. Thus, he sel -
associa ing p ope ies ha unde lie TIA-1 unc ion also
p edispose i o i e e sible agg ega ion, po en ially unde lying
disease. In ac , mu a ions in TIA-1 associa ed wi h Welande
dis al myopa hy (WDM) and ALS con e a g ea e p opensi y o
agg ega ion and delayed s ess g anule disassembly, consis en
wi h his mechanism o disease (Mackenzie e al., 2017).
Howe e , ex ensi e amyloid agg ega es o TIA-1 a e no a
dis inc i e ea u e o s ess g anules in heal hy cells, sugges ing
ha in e ac ions p e en ing agg ega ion o TIA-1 p o ein exis
wi hin he cell. He ein we in es iga e he e ec o s ess g anule
co- ac o s on LLPS and fib il o ma ion o TIA-1 in i o.
S ess g anules a e made up o hund eds o p o eins and co-
ac o s, a subse o which specifically con ibu e o he assembly
and main enance o he liquid s a e (Jain e al., 2016;Taube e al.,
2020). Zn
2+
has been epo ed o be a second messenge ,
enhancing he o ma ion o TIA-1 posi i e s ess g anules
(Rayman e al., 2018). The elease o in acellula Zn
2+
was
obse ed in cells and issues exposed o a seni e ea men ,
consis en wi h Zn
2+
playing a ole in s ess g anule o ma ion
in esponse o oxida i e s ess. When Zn
2+
was made una ailable
in i o by chela ion, he o ma ion o TIA-1 posi i e s ess
g anules was educed (Rayman e al., 2018;Ca ascoso e al.,
2019). S udies o pu ified TIA-1 p o ein showed Zn
2+
dependen
mul ime iza ion using a Fö s e esonance ene gy ans e
(FRET) coinciding wi h o ma ion o LLPS wi hou ex ensi e
ß-shee o ma ion (Rayman e al., 2018). This s udy hus
sugges ed ha a di ec in e ac ion be ween Zn
2+
and TIA-1,
and subsequen change in mul ime iza ion leading o LLPS,
could influence s ess g anule o ma ion. This aises ques ions
o how Zn
2+
di e en ially a ec s TIA-1 LLPS s fib illa
agg ega ion, and he molecula basis o he Zn
2+
-TIA-
1 in e ac ion. In he cu en s udy, he e o e, we analysed he
e ec o Zn
2+
on LLPS and fib illa agg ega ion o TIA-1, he
nucleic acid-induced TIA-1 condensa es and also in es iga ed he
molecula basis o he Zn
2+
:TIA-1 in e ac ion.
S ess g anules a e pa icula ly en iched wi h RNA-binding
p o eins wi h low complexi y egions, including P LDs and also
ano he ype o in insically diso de ed egion (IDR) ich in A g-
Gly-Gly epea s (RGG) (Chong e al., 2018;Youn e al., 2019).
RGG-con aining p o eins include he co e s ess g anule p o ein
G3BP1/2, in which he RGG egion media es an RNA induced
swi ch o ins iga e condensa ion (Guillen-Boixe e al., 2020;
Sande s e al., 2020;Yang e al., 2020), and he FUS p o ein,
ha is ec ui ed o s ess g anules by i s RGG-zinc finge domain
in disease (Ben mann e al., 2012). I has also been shown ha he
A g- ich epea pep ides de i ed om pa hogenic
C9o 72 expansions associa e wi h s ess g anules, a ec ing
hei dynamics (Boeynaems e al., 2017). While well known
o hei abili y o in e ac di ec ly wi h RNA, RGG- ich IDRs
can also in e ac wi h P LDs, enhancing LLPS (Kau e al., 2021).
Mul i alen ansien in e ac ions be ween he P LD and RGG
IDRs o FUS enhance phase sepa a ion (Qama e al., 2018;Wang
e al., 2018). Fu he mo e, ensuing condensa es can co- ec ui
o he RNA-binding p o eins, including TIA-1, po en ially also
ia in e molecula he e o ypic RGG:P LD in e ac ions (Qama
e al., 2018;Wang e al., 2018). Thus, in e ac ions be ween P LD
and RGG- ich IDRs a e likely o undamen al impo ance in
s ess g anule biology and disease. We ha e hence examined he
in e ac ions o he na i e RGG- ich RNA-binding domain o FUS
(FUS-RBD) wi h TIA-1 as a model o he e ec o he e o ypic
in e ac ions in p o ein-RNA condensa es.
The cu en s udy in es iga es he way in which s ess g anule
p o eins and co- ac o s influence LLPS and fib illa agg ega ion
o TIA-1. We epo he impac s o Zn
2+
and FUS-RBD on TIA-1
in he absence and p esence o nucleic acid. Ou in es iga ion
u ilises in i o assays o TIA-1 LLPS and fib illa agg ega e
o ma ion as p e iously used o p obe he e ec o nucleic
acid (Loughlin e al., 2021) as well as NMR spec oscopy and
si e di ec ed mu agenesis o u he iden i y Zn
2+
binding si es.
These expe imen s e ealed ha he me al ion Zn
2+
no only
enhances LLPS o TIA-1 alone, bu also u he enhances nucleic
acid-induced TIA-1 LLPS. NMR spec oscopy and mu a ional
s udies defined a zinc binding si e on RRM2 leading o he
iden ifica ion o H94 and H96 as Zn
2+
ligands. Su p isingly,
Zn
2+
was ound o induce mul ime iza ion o he isola ed
TIA-1 RRM2,3 domains e en in he absence o he P LD,
po en ially eflec ing an addi ional mode o TIA-1 sel -
in e ac ion ia Zn
2+
liga ion. Zn
2+
-enhanced LLPS o ull
leng h TIA-1, howe e , did no esul in fib illa agg ega e
o ma ion o TIA-1 and Zn
2+
s ikingly inhibi ed nucleic acid-
induced fib illa agg ega ion o TIA-1 e en a sub-s oichiome ic
le els. FUS-RBD also u he enhanced nucleic acid-induced
LLPS o TIA-1, bu e ec i ely inhibi ed fib illa agg ega ion,
p esumably h ough a di ec compe i ion be ween P LD:P LD
and P LD:RGG in e ac ions. Toge he , his s udy showcases wo
ypes o he e o ypic in e ac ion o TIA-1 ha p omo e LLPS, bu
supp ess he agg ega ion capaci y o TIA-1 in i o, po en ially
con ibu ing o he main enance o he e e sible and dynamic
s a e o TIA-1 wi hin s ess g anules.
MATERIALS AND METHODS
Cloning o TIA-1 RRM2,3 His Mu an s
To cons uc he plasmid o exp ession o TIA-1 RRM2,3 (93-
274), pGEX-4T-2 (GE Heal hca e) was modified o bea a 6xHis
ag and a h ombin clea age si e. TIA-1 RRM2,3 was inse ed
di ec ly a e he h ombin si e using oligonucleo ides TIA-1
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608062
Wes e al. Regula ion o TIA-1 Condensa es
RRM2,3-Fo (GCGTGGATCCCCAGGAATTCCCAATCATTT
CCATGTCTTTGTTGGTG) and TIA-1 RRM2,3- e (CAG
TCACGATGCGGCCGCTTATTTGCCCCAATAGCATTTCAC
AAC). To cons uc he plasmid TIA-1 RRM2,3 bea ing
mu a ions H94A, H96 A and H94A H96A, si e-di ec ed
mu agenesis by in e se PCR (Liu and Naismi h, 2008) was
pe o med using he oligonucleo ide pai s:
TIA-1 RRM2,3-H94A-Fo : CCAATgc TTCCATGTCTTTG
TTGGTGATCTCAGCCCAG
TIA-1 RRM2,3-H94A-Re : ATGGAAagcATTGGGAATTCC
TGGGGATCCACG
TIA-1 RRM2,3-H96A-Fo : TTTCgc GTCTTTGTTGGTGA
TCTCAGCCCAG
TIA-1 RRM2,3-H96A-Re : AAAGACagcGAAATGATTGGG
AATTCCTGGGGATCCACG
TIA-1 RRM2,3-H94A H96A-Fo : CCAATgc TTCgcTGTCT
TTGTTGGTGATCTCAGCCCAG
TIA-1 RRM2,3-H94A H96A-Re : AAAGACagcGAAagcATT
GGGAATTCCTGGGGATCCACGCGG
Full-Leng h TIA-1 P o ein P epa a ion
TIA-1 p o ein sho iso o m (P31483-2) was exp essed om
plasmid pETM11 as a 6xHis- agged p o ein in BL21 pLysS
E. coli cells and pu ified as p e iously epo ed (Loughlin
e al., 2021). In b ie , his in ol ed ini ial pu ifica ion using
Co-TALON me al a fini y esin (Taka a), ollowed by clea age
o he 6xHis- ag by using TEV p o ease and i s emo al using Ni-
NTA esin. The final pu ifica ion was by size exclusion
ch oma og aphy in a bu e es ablished o main ain TIA-1
solubili y (20 mM sodium phospha e, 60 mM KCl, 0.5 M
a ginine-HCl, 1 mM MgCl
2
, 2 mM DTT, 0.5 mM EDTA,
pH 7.0). Aliquo s we e concen a ed o a maximum o
100 μM, fil e ed and s o ed a −80°C in size exclusion bu e .
TIA-1 p o ein used in Zn
2+
expe imen s was p epa ed in he same
manne , wi h he excep ion ha EDTA was no included in he
final size exclusion s ep. The A
280/260
a io o each p epa a ion
was <0.65 AU and p o ein was quan ified using a heo e ical
mola ex inc ion coe ficien o 80,330 M
−1
cm
−1
.
TIA-1 RRM1 and RRM2,3 P epa a ion
GST-TIA-1 RRM1 (amino acids 1-81; in plasmid pGEX-4T-2)
and 6xHis-TIA-1 RRM2,3 (amino acids 93-27; in plasmid
pET x-1a) a ian s (Wa is e al., 2017) we e p oduced in
E. coli BL21 (DE3) cells g own a 37°C oanOD
600
o 0.6–0.8.
LB medium was used i p o eins we e in ended o u bidi y
measu emen s, whe eas
15
N-labelled TIA-1 cons uc s o
NMR i a ions we e p oduced in M9 minimal medium
wi h
15
NH
4
Cl as a ni ogen sou ce. Exp ession was induced
by he addi ion o 1 mM IPTG and cul u es we e incuba ed a
30°C o 16–18 h. Cell pelle s we e esuspended in lysis bu e
(20 mM po assium phospha e, 500 mM KCl, pH 7.4) wi h
20 μg/ml DNase I, 1 mM phenylme hanesul onyl fluo ide
(PMSF) and 100 μg/ml lysozyme. Cells we e dis up ed by
sonica ion (cycles o 30 s a 40% o ampli ude, 60 s o es ,
6 min o al ime, on ice), and cell deb is was emo ed by
cen i uga ion a 28,000 × g (4°C o 45 min). The supe na an
om GST-TIA-1 RRM1 was loaded on o a 5 ml P ofini y GST
column (BioRad) and elu ed in a single-s ep wi h lysis bu e
con aining 20 mM GSH in a NGC Ch oma og aphy Sys em
Ques 10 (BioRad). The supe na an om 6xHis-TIA-
1 RRM2,3 WT o i s mu an a ian s was loaded on o a Ni-
NTA ma ix (The moFishe ), p e iously equilib a ed wi h
lysis bu e con aining 5 mM imidazole, and incuba ed o
1h a 4
°C. Recombinan p o eins we e elu ed using a non-
con inuous imidazole g adien and pu i y was u he checked
by SDS-PAGE. The GST- and 6xHis- ag we e emo ed by
o e nigh incuba ion (4°C) o he p o ein wi h 2.5 U/mg o
h ombin p o ease (Cy i a). Clea ed TIA-1 cons uc s we e
hen isola ed wi h Ni-NTA esin ( o RRM2,3) o a wo-s ep
ch oma og aphy wi h EN ich SEC 650 (BioRad) and P ofini y
GST columns ( o RRM1). Pu ified p o eins we e dialyzed
wice agains 5 L o ei he HEPES bu e (20 mM HEPES,
50 mM NaCl, pH 6.9) o ci a e bu e (20 mM ci a e,
50 mM NaCl, pH 5.5) a 4°C o e nigh . Fo comple e
disulphide bond educ ion, bu e ed TCEP was added o
dialyzed p o eins up o a concen a ion o 1 o 5 mM o
u bidi y assays and NMR i a ions, espec i ely. Then,
samples we e concen a ed using Amicon Ul a-15
cen i ugal fil e s (Me ck-Millipo e). P o eins we e
quan ified by spec opho ome y a 280 nm using ex inc ion
coe ficien s o 8,480 M
−1
cm
−1
o TIA-1 RRM1 and
30,940 M
−1
cm
−1
o TIA-1 RRM2,3.
FUS-RBD P o ein P epa a ion
Two cons uc s o FUS we e exp essed and pu ified as pe
Loughlin e al. (Loughlin e al., 2019). FUS-RBD (amino acids
242-526) comp ising FUS RGG1-RRM-RGG2-ZnF-
RGG3 domains was p oduced om pET24b consis ing o a
N- e minal GB1 solubili y ag ollowed by a 6xHis- ag. FUS-
RBD-ΔRGG1/3 (amino acids 269-454) was p oduced om
pET28a consis ing o a N- e minal 6xHis ag. Each
cons uc was exp essed in EcoliBL21 Rose a2 cells g own
a 37°C, ollowed by p o ein exp ession induc ion wi h 0.5 mM
IPTG and o e nigh exp ession a a educed empe a u e o
25°C. Cell pelle s we e esuspended in lysis bu e (50 mM
T is-HCl, 1 M NaCl, 0.5% T i on-X, 5 mM imidazole, 0.5 mM
ß-me cap oe hanol, pH 8.0) and lysed by sonica ion. Lysa e
was cla ified a 48,000 × g o 30 min, loaded on o Ni-NTA
beads, washed wi h 20 and 40 mM imidazole in wash bu e
(50mM T is-HCl, 1M NaCl, 0.5mM ß-me cap oe hanol,
pH 8.0) and elu ed wi h 200 mM imidazole in wash bu e .
Fo FUS-RBD sho , he 6xHis- ag was clea ed wi h TEV
p o ease in 50 mM T is-HCl, 1 M NaCl, 10 mM imidazole,
0.5 mM ß-me cap oe hanol, pH 8.0, and emo ed wi h Ni-
NTA beads. P o eins we e dialysed in o s o age bu e (50 mM
T is-HCl, 1 M NaCl, 20 mM imidazole, 0.5 mM ß-
me cap oe hanol, pH 8), concen a ed o 0.4 mM and s o ed
a −80°C. P o eins we e quan ified using ex inc ion coe ficien s
o 36,883 M
−1
cm
−1
o GB1-6xHis-FUS-RBD usion and
20,970 M
−1
cm
−1
o FUS-RBD-ΔRGG1/3.
Oligonucleo ides
Syn he ic single s anded DNA oligonucleo ides con aining h ee
o fi e TC- ich TIA-1 binding si es we e syn hesized and HPLC
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608063
Wes e al. Regula ion o TIA-1 Condensa es
pu ified comme cially (IDT, Aus alia) and quan ified using
mola ex inc ion coe ficien s supplied by IDT.
TC3 (34 n ): TTTTTACTCCAATTTTTACTCCAATTTTTA
CTCC
TC5 (58 n ): TTTTTACTCCAATTTTTACTCCAATTTTTA
CTCCAATTTTTACTCCAATTTTTACTCC
Tu bidi y Measu emen s
Tu bidi y measu emen s o ull leng h TIA-1 we e used o
measu e he ela i e amoun s o LLPS unde se condi ions.
Measu emen s we e aken wi hin 20 min o ins iga ing LLPS, a
ime ame in which we do no obse e agg ega ion ei he by
mic oscopy o ThT fluo escence. TIA-1 p o ein in size
exclusion bu e was dilu ed ei he alone, o in he p esence
o nucleic acids in agg ega ion bu e . In gene al, TIA-1
p o ein was dilu ed o 2.5 μM in he p esence o 0.5 μM
oligonucleo ides unless o he wise s a ed. Final agg ega ion
bu e was20mMHEPES,50mMNaCl,pH7.2,including
esidual 15 mM a ginine-HCl. As con ols, nucleic acid o
ZnCl
2
alone we e dilu ed in agg ega ion bu e . T iplica e
samples o 150 µL we e se -up a oom empe a u e and
incuba ed o 10 min a 25°C in 96-well clea bo om non-
binding black pla es (G eine ), hen analysed a 385 nm using a
CLARIOs a pla e eade (BMG Lab ech). Assays included
3 eplica es wi h e o ba s ep esen ing S.D.
Tu bidi y measu emen s o TIA-1 RRM2,3 WT, H94A, H96A
and H94A H96A we e pe o med a a final concen a ion o
20 μM in ei he HEPES bu e (20 mM HEPES, 50 mM NaCl,
pH 6.9) o ci a e bu e (20 mM ci a e, 50 mM NaCl, pH 5.5).
Each cons uc was measu ed a 25°C e e y min o 30 min a e
he addi ion o ZnCl
2
o TPEN a he indica ed a ios.
Independen samples o 100 µL we e p epa ed a oom
empe a u e and measu ed a 385 nm in a 96-well pla e wi h a
Va ioskan LUX mic owell pla e eade (The moFishe ). G aphs
ep esen he a e age o he las 10 alues o u bidi y
measu emen s, a e signal s abiliza ion. Assays included
3 eplica es wi h e o ba s ep esen ing S.D. Boxplo s we e
gene a ed using R e sion 4.0.5 (h p://www. -p ojec .o g).
DIC Mic oscopy
Samples o ull leng h TIA-1 wi h o wi hou ZnCl
2
o
oligonucleo ides we e p epa ed esh as pe samples used in
u bidi y measu emen s and imaged wi hin 20 min o mixing
unless o he wise s a ed. A 10 μL-aliquo o each sample was
spo ed on o a glass mic oscope slide wi h double sided ape
and co e ed wi h 0.17 mm HP glass co e slip (Zeiss).
Solu ions we e imaged a oom empe a u e wi h
di e en ial in e e ence con as (DIC) on an In e ed
Olympus IX81 x 2UCB mic oscope (Olympus, Tokyo,
Japan) wi h a 60 × objec i e. Images we e p ocessed using
FIJI (Schindelin e al., 2012).
Thiofla in T Assay
Thiofla in-T (ThT) fluo escence assays we e used o moni o
ThT posi i e agg ega e o ma ion o ull leng h TIA-1 alone
and in he p esence o nucleic acids. Samples we e p epa ed as
pe u bidi y assays in agg ega ion bu e . A e u bidi y
measu emen s, a 1 mM ThT s ock solu ion was dilu ed o a
final concen a ion o 5 μM and equilib a ed o 30
o
C o
se e al minu es p io o measu emen s commencing.
Samples we e hen agi a ed h ough o bi al shaking a
500 .p.m. and 30°Co e 12–16 h, and ThT fluo escence
was moni o ed wi h an exci a ion wa eleng h o 425 nm
andanemissionwa eleng ho 485nm,usingaCLARIOs a
pla e eade (BMG Lab ech). Measu emen s we e baseline
co ec ed using ThT fluo escence o agg ega ion bu e /ThT
alone. Assays included 3 eplica es wi h e o ba s
ep esen ing S.D.
T ansmission Elec on Mic oscopy
To analye he mo phology o he agg ega es p esen a he
endpoin o he ThT assay, samples we e imaged by
T ansmission Elec on Mic oscopy (TEM). A 5 µL o endpoin
sample was applied o he su ace o glow-discha ged con inues
ca bon g ids and s ained wi h 2% u anyl ace a e solu ion. The
excess s ain was emo ed using fil e pape and g ids we e ai -
d ied. All samples we e imaged on an FEI Tecnai F2 F20 TWIN
elec on mic oscope wi h a Ga an 4k x 4k CCD, unde a wo king
ol age o 200 kV a he Ramacio i Cen e o C yo-Elec on
Mic oscopy o Monash Uni e si y.
NMR Measu emen s
Nuclea Magne ic Resonance (NMR) i a ions in ol ing TIA-
1 cons uc s we e eco ded and moni o ed a 25°Cby1D
1
H
and 2D [
1
H-
15
N] He e onuclea Single Quan um Co ela ion
(HSQC) spec a in a B uke A ance-III 700 and 500 MHz
equipped wi h a 5 mm TCI c yop obe. Samples we e dialyzed
in 20 mM ci a e, 50 mM NaCl, 5 mM TCEP, pH 5.5, o 20 mM
HEPES, 50 mM NaCl, 5 mM TCEP, pH 6.9, o NMR
i a ions. 5% D
2
O was added o all samples o adjus he
lock signal o he NMR spec ome e . Samples we e p epa ed
a 200 o 300 μM o TIA-1 RRM1 and TIA-1 RRM2,3,
espec i ely, in a final olume o 350 μL. Samples we e
loaded in o Shigemi ubes (Shigemi Inc.). Da a we e
acqui ed and p ocessed using TopSpin 3.5pL7 so wa e
(B uke ). Linewid h b oadening and chemical-shi
pe u ba ion analysis we e pe o med using NMRFAM-
SPARKY so wa e dis ibu ion (Na ional Magne ic
Resonance Facili y, Madison). The NMR assignmen o
TIA-1 RRM1 and TIA-1 RRM2,3 was al eady a ailable
(Biological Magne ic Resonance Bank [BMRB] accession
numbe s 34144 and 19,735, espec i ely) (Wang e al., 2014;
Sonn ag e al., 2017). Chemical-shi pe u ba ions (Δδ
AVG
)
we e calcula ed as p e iously desc ibed (Ri e o-Rod íguez
e al., 2021).
CD Spec opola ime y
Ci cula Dich oism (CD) spec a we e eco ded in he a -UV
ange (195–250 nm) a 20°C on a Jasco J-815 CD
spec opola ime e equipped wi h a Pel ie empe a u e
con ol sys em. 10 µM o each TIA-1 RRM2,3 cons uc in
phospha e bu e (10 mM sodium phospha e, 1 mM TCEP,
pH 6.9) was measu ed in a 1 mm qua z cu e e. The final
spec a we e an a e age o 20 scans.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608064
Wes e al. Regula ion o TIA-1 Condensa es
RESULTS
Zn
2+
Enhances LLPS o TIA-1, Especially in
he P esence o Nucleic Acid
To explo e he e ec o Zn
2+
on TIA-1 LLPS and po en ial
subsequen agg ega ion, we fi s assessed he e ec o Zn
2+
on
he LLPS o TIA-1 p o ein in he absence o nucleic acid. In his
s udy, pu ified monome ic TIA-1 p o ein was p epa ed in he
p esence o 500 mM a ginine, hen dilu ed in o a low sal
bu e o a concen a ion jus below i s sa u a ion (Csa ), a
which TIA-1 shows minimal LLPS (Loughlin e al., 2021). The
e ec o ZnCl
2
on he LLPS o TIA-1 p o ein was analysed by
DIC mic oscopy and quan ified using u bidi y (abso bance a
385 nm). TIA-1 p o ein alone a 2.5 μM showed minimal LLPS
when imaged by DIC (Figu e 1A). Addi ion o 0.1–50 μM
ZnCl
2
esul ed in he appea ance o sphe ical d ople s ha
escala ed in numbe on inc easing Zn
2+
concen a ion,
demons a ing ZnCl
2
enhanced LLPS o TIA-1 (Figu e 1A).
The u bidi y o TIA-1 also inc eased in he p esence o ZnCl
2
in a concen a ion-dependen manne om 0.2 o 0.8,
eflec ing inc eased LLPS (Figu e 1B). Taken oge he , hese
esul s show ha ZnCl
2
alone enhances LLPS o TIA-1, in
ag eemen wi h p e ious s udies o TIA-1 usion p o ein
(Rayman e al., 2018).
Ou p e ious wo k showed ha nucleic acid ha bou ing
mul iple binding si es enhances LLPS o TIA-1, lowe ing he
concen a ion a which he p o ein spon aneously phase sepa a es
(Loughlin e al., 2021). To de e mine whe he nucleic acid and
Zn
2+
enhance LLPS o TIA-1 coope a i ely o ac in a compe i i e
manne , we analysed he e ec o ZnCl
2
on ssDNA-induced LLPS
o TIA-1. TIA-1 (a 2.5 μM) in he p esence o ssDNA ha bou ing
fi e TIA-1 binding si es (“TC5”) was assessed o LLPS using DIC
mic oscopy and quan ified by u bidi y. DIC showed he p esence
o LLPS d ople s (~2–5μm in diame e ) ha we e sligh ly la ge
han hose ins iga ed only by ZnCl
2
and ga e ise o u bidi y
measu emen s o ~0.3 (Figu es 1A,C). Addi ion o 0.1 μM ZnCl
2
had minimal e ec on he u bidi y o TIA-1:ssDNA; howe e ,
1–10 μM ZnCl
2
enhanced he numbe o d ople s and inc eased
u bidi y o 0.4-0.75, showing ha Zn
2+
and ssDNA oge he
enhance LLPS o TIA-1. Addi ion o 50 μM ZnCl
2
esul ed in a
FIGURE 1 | Zinc enhances LLPS o TIA-1 and TIA-1:ssDNA in a concen a ion dependan manne . (A) DIC mic oscopy o 2.5 μM TIA-1 (uppe ) and 0.5 μM ssDNA
TC5 + 2.5 μM TIA-1 (lowe ) in he p esence o 0–50 μM ZnCl
2
.(B) Tu bidi y (A
385 nm
)o 2.5μM TIA-1 + 0–50 μM ZnCl
2
.(C) Tu bidi y o 2.5 μM TIA-1 + 0.5 μM ssDNA
TC5 + 0–50 μM ZnCl
2
and (D) Tu bidi y o 0–50 μM ZnCl
2
alone. Condi ions: 20 mM HEPES, 50 mM NaCl, 12 mM A ginine, pH 7.2, a 25°C and wi hin 20 min o
incuba ion. 4 eplica es. E o ba s ep esen S.D.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608065
Wes e al. Regula ion o TIA-1 Condensa es

FIGURE 2 | Zn2+ specifically binds o TIA-1 RRM2,3 and media es p o ein mul ime iza ion. (A) De ailed iew o ep esen a i e amide esonances o unpe u bed
(le panel), ß-shee (middle panel)o Zn
2+
binding si e (le panel) TIA-1 RRM2,3 esidues om he supe imposed [
1
H-
15
N] HSQC spec a o
15
N-labeled TIA-1
RRM2,3 cons uc ei he ee (blue) o upon incuba ion wi h ZnCl
2
( ed) o ZnCl
2
+ TPEN (da k g een). Whole iews o 2D NMR spec a a e included in Supplemen a y
Figu es S2A,E.(B) Map o
1
H-linewid hs on he TIA-1 RRM2,3 ibbon upon addi ion o an equimola a io o ZnCl
2
. Residues wi h
1
H line b oadening la ge han
he a e age (µ) plus wo s anda d de ia ions (σ) a e colo ed using a whi e- o-blue scale. P olines and unassigned esidues a e colo ed in g ey. TIA-1 PDB ID: 2MJN
(Wang e al., 2014). Condi ions: 20 mM HEPES, 50 mM NaCl, 5 mM TCEP, pH 6.9, a 25°C and upon 30 min incuba ion wi h ZnCl
2
o ZnCl
2
+TPEN. (C) Compa ison o
u bidi y (A
385 nm
) p oduced by TIA-1 RRM2,3 WT and di e en His- o-Ala mu an s a inc easing concen a ions o ZnCl
2
. All measu emen s we e pe o med by using
TIA-1 cons uc s a 20 µM concen a ion. Condi ions: 20 mM HEPES, 50 mM NaCl, 1 mM TCEP, pH 6.9, a 25°C and wi hin 30 min incuba ion unde shaking.
3 eplica es. E o ba s ep esen S.D.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608066
Wes e al. Regula ion o TIA-1 Condensa es
u he inc ease in u bidi y o ~1.0, and DIC showed he
p esence o a la ge numbe o smalle d ople s, sugges ing ha
a hese concen a ions Zn
2+
-induced TIA-1 LLPS domina es he
phase sepa a ion o hese samples. I was also confi med ha
ZnCl
2
alone did no gi e ise o sample u bidi y (Figu e 1D).
These esul s show ha Zn
2+
u he enhances nucleic acid-
induced LLPS o TIA-1.
Zn
2+
In e ac s Specifically Wi h TIA-1
RRM2 Residues
In o de o explo e he molecula basis o Zn
2+
-induced LLPS o
TIA-1, NMR
1
H-
15
N-HSQC i a ion expe imen s we e
unde aken o RRM1 (amino acids 1-81) and RRM2,3 (amino
acids 93-274) domains. ZnCl
2
was i a ed in o
15
N-labelled
RRM1 o RRM2,3 a 200 and 300 μM, espec i ely, a pH 6.9.
Addi ion o Zn
2+
in o ei he sample esul ed in some u bidi y
obse ed by eye, sugges ing a dec ease in solubili y a hese
concen a ions. The RRM1 amide esonances showed no
significan changes in ei he chemical-shi pe u ba ions o
linewid hs; he e o e, no specificZn
2+
binding si e was
iden ified (Supplemen a y Figu e S1). In con as , when Zn
2+
was added o a sample o RRM2,3, he a e age b oadening was
enhanced om 25.5 ± 4.8 Hz ( ee TIA-1 RRM2,3) o 37.4 ±
18.3 Hz (Zn
2+
:p o ein a io o 1:1), indica i e o mul ime iza ion
o TIA-1 (Supplemen a y Figu e S2). Fu he mo e, ou analysis
also e ealed a subse o amide esonances unde going specific
line b oadening (>250 Hz). In e es ingly, some o he a ec ed
esidues—namely N93 (ß
1
), F95 (ß
1
), H96 (ß
1
), V97 (ß
1
), S122
(ß
2
), A124 (ß
2
) and F143 (ß
3
)—a e posi ioned in a clus e
adjacen o and including RRM2 ß-shee esidues o he
canonical RNA binding si e (Figu es 2A,B and
Supplemen a y Figu e S2A). Mo eo e , o he esidues a
RRM2 α
2
-helix (K146 and D148) and hose loca ed a a
flexible RRM2 loop and he in e domain linke (M130 and
A171, espec i ely) we e also subs an ially b oadened
(>250 Hz; Figu es 2A,B). Mode a e, bu s ill significan ,
b oadening (>mean + 2σHz) was also obse ed o
esonances H94 (ß
1
), F98 (ß
1
), V99 (ß
1
), I111 (α
1
), R125 (ß
2
),
S135 (ß
3
), S142 (ß
3
), A149 (α
2
), N151 (α
2
) and A152 (α
2
)in
RRM2 (Figu es 2A,B,Supplemen a y Figu e S2). Se e al
esidues om he linke and RRM3 also b oadened abo e he
h eshold, bu no o he ex en o ha obse ed in RRM2.
Rema kably, bo h subs an ial line b oadening o RRM2 in
RRM2,3 and sample u bidi y we e ully e e sible upon
addi ion o he Zn
2+
-chela o TPEN (a e age linewid h o
26 ± 4 Hz), which confi ms ha hese e ec s a e
unequi ocally induced by Zn
2+
(Figu e 2A,Supplemen a y
Figu es S2E,G). I was also confi med ha TPEN alone had
no e ec on TIA-1 RRM2,3 esonances (Supplemen a y Figu es
S2F,G). Toge he , hese esul s sugges ha Zn
2+
media es he
mul ime iza ion o TIA-1 RRM2,3 domains independen ly o he
P LD a high p o ein concen a ion (300 µM).
F om hese da a, H94 and H96 in RRM2 s and ou as po en ial
Zn
2+
ligands o TIA-1 as his idine esidues a e o en in ol ed in Zn
2+
coo dina ion (Kocyla e al., 2021). Based on his p emise, he NMR
i a ion was epea ed unde lowe pH condi ions a which his idine
esidues could no in e ac wi h Zn
2+
due o p o ona ion (C uz-
Galla do e al., 2013;C uz-Galla do e al., 2015). Ti a ing TIA-1
RRM2,3 wi h Zn
2+
a pH 5.5 esul ed in nei he enhanced line
b oadening no isible u bidi y, consis en wi h he lack o his idine
coo dina ion o Zn
2+
in TIA-1 RRM2,3 (Supplemen a y Figu es
S2B–D). I should be no ed ha Zn
2+
addi ion o a solu ion o TIA-1
RRM1 a pH 5.5 also did no p oduce any isible u bidi y and he
linewid h o he signals emained unal e ed (Supplemen a y
Figu es S1B–D). Thus, his idines in RRM1, ha occu a
posi ions 54, 56 and 58, could also ansien ly coo dina e Zn
2+
,
bu wi hou o ming s able complexes. Al oge he , NMR i a ions
iden ified TIA-1 RRM2 H94 and H96 esidues as po en ial
candida es o specificZn
2+
binding, possibly media ing he
o ma ion o Zn
2+
-induced mul ime s unde lying he obse ed
solu ion u bidi y o TIA-1 RRM2,3.
To es his hypo hesis we fi s quan ified he e ec o Zn
2+
i a ionin o20µMTIA-1RRM2,3onsolu ion u bidi y—as
measu ed by abso bance a 385 nm. Inc easing he Zn
2+
enhanced he u bidi y o RRM2,3 in a concen a ion dependen
manne (Supplemen a y Figu e S3A). Fu he mo e, unde hese
condi ions, he supp essing e ec s o TPEN and pH 5.5 in educing
TIA-1 RRM2,3 u bidi y we e also obse ed, as p e iously seen in
he mo e concen a ed TIA-1 RRM2,3 NMR solu ions
(Supplemen a y Figu es S3B,C). Thus, in o de o confi m he
oleo H94andH96inZn
2+
-induced mul ime iza ion, TIA-1
RRM2,3 mu an s we e designed so ha one o bo h his idine
esidues we e subs i u ed by alanine. The co ec p o ein olding
was confi med using CD spec opola ime y, e ealing he same
seconda y s uc u e con en o TIA-1 RRM2,3 wild ype and
mu an s (Supplemen a y Figu e S4). While addi ion o Zn
2+
o
a a ioo 8:1inc eased he u bidi yo TIA-1RRM2,3(20µM),
minimal o negligible e ec s we e obse ed o H94A, H96A o
H94A H96A mu an s (Figu e 2C). This finding sugges s a s uc u al
ole o Zn
2+
in specifically coo dina ing H94/H96 esidues in
RRM2, unde lying Zn
2+
-induced mul ime iza ion o TIA-1 RRMs.
Zn
2+
Supp esses Nucleic Acid Induced
TIA-1 Fib illa Agg ega ion
Whils LLPS o TIA-1 con ibu es o s ess g anule o ma ion,
abe an LLPS o disease-associa ed TIA-1 a ian s and
enhanced p opensi y o o m fib illa agg ega es a e
associa ed wi h impai ed s ess g anule dynamics
acili a ing pa hological inclusions linked o ALS
(Mackenzie e al., 2017;Zhang e al., 2019). Thus, we nex
p obed he e ec o Zn
2+
on fib illa agg ega ion o TIA-1 o e
ime, in and ou o he p esence o nucleic acid, as moni o ed
by Thiofla in T (ThT) fluo escence and TEM. A
concen a ions a which TIA-1 shows minimal LLPS
(2.5 μM), no significan TIA-1 agg ega ion was de ec ed
(Supplemen a y Figu e S5A). Addi ion o 10 μMZnCl
2
,
ha obus ly induces LLPS o TIA-1, did no enhance
agg ega ion o TIA-1 and e en appea ed o show a sligh
supp ession o ThT fluo escence (Figu e 3A). TEM
measu emen s did no de ec any fib illa agg ega ion
(Figu e 3B), showing ha Zn
2+
-induced LLPS o TIA-1
does no p oduce fib illa agg ega es.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608067
Wes e al. Regula ion o TIA-1 Condensa es
We hen unde ook he same expe imen in he p esence o he
TC- ich dsDNA ha bou ing 5 TIA-1 binding si es (“TC5”), ha
enhances TIA-1 LLPS and fib il o ma ion (Loughlin e al., 2021). As
p e iously obse ed, addi ion o TC5 o TIA-1 (2.5 μM) esul ed in
fib illa agg ega ion as de ec ed by ThT fluo escence o e ime, and
fib ils as obse ed by TEM (Figu es 3C,D). To es he e ec o Zn
2+
on nucleic acid-induced fib iliza ion o TIA-1, we moni o ed
agg ega ion o TIA-1 in he p esence o TC5 and 10 μMZnCl
2
.
No enhancemen in ThT fluo escence was obse ed o e ime and
TEM o final samples did no show nume ous fib ils, bu small
agg ega es wi h po en ial oligome ic species (Figu es 3E,F).
In e es ingly, we also obse ed inhibi ion o ThT fluo escence a
subs oichiome ic concen a ions o ZnCl
2
(Supplemen a y
Figu e 5B). These esul s sugges ha he p esence o Zn
2+
impedes fib iliza ion o TIA-1, main aining i wi hin a e e sible
LLPS s a e.
RGG-Rich RBD o FUS Enhances Nucleic
Acid Induced TIA-1 LLPS and P e en s
Ma u a ion o I e e sible Agg ega es
As pa o ou s udy o s ess g anule co- ac o s, we also analysed
he e ec s o he RGG- ich RNA-binding domain (RBD) o
FUS—as a ep esen a i e RGG- ich p o ein and po en ial
molecula modula o o TIA-1 LLPS. FUS localises o s ess
g anules ia i s RGG- ich in insically diso de ed egion
RGG3 (Ben mann e al., 2012). Mo eo e , in i o TIA-1
p o ein pa i ions in o phase sepa a ed FUS p o ein media ed
in pa by he RGG ich RBD (Wang e al., 2018). In e es ingly,
al hough RGG IDPs can d i e homo ypic LLPS, unlike P LD-
media ed LLPS, hey do no eadily ma u e in o i e e sible
fib illa agg ega es (Gui e al., 2019). Thus, in e ac ions o
RGG in insically diso de ed egions may also influence TIA-1
LLPS and po en ial fib illa agg ega ion. We he e o e es ed he
e ec o RGG- ich RBD o FUS on TIA-1 LLPS and agg ega ion,
by DIC and u bidi y measu emen s, bo h alone and in he
p esence o nucleic acids. FUS-RBD (amino acids 242-526)
comp ising RGG1-RRM-RGG2-ZnF-RGG3 (Figu e 4A) was
pu ified in high sal o homogenei y, ee om RNA
con amina ion. TIA-1 and FUS-RBD dilu ed o 2.5 µM in
agg ega ion bu e showed minimal LLPS in DIC mic oscopy,
wi h no appa en inc ease upon hei addi ion (Figu e 4B), and
only a e y small inc ease in u bidi y was obse ed upon hei
addi ion (Figu e 4C). Toge he , his sugges s ha FUS-RBD does
no g ea ly enhance LLPS o TIA-1 a hese concen a ions.
Nex , we es ed he e ec o FUS-RBD on TIA-1 LLPS in he
p esence o a ssDNA comp ising h ee andem TIA-1 binding si es
(“TC3”). The addi ion o sub-s oichiome ic amoun s o TC3 o
TIA-1 d ama ically enhanced LLPS, as measu ed by inc eased
solu ion u bidi y (Figu e 4C) and by d ople s obse ed by DIC
FIGURE 3 | Zinc inhibi s fib illa agg ega ion o TIA-1. Thiofla in T (ThT) fluo escence assay moni o ing he o ma ion o ThT posi i e agg ega es ( op) and TEM o
final samples (bo om)(A,B) 2.5 μM TIA-1, 10 μM ZnCl
2
(C,D) 2.5 μM TIA-1, 0.5 μM ssDNA TC5 (E,F) 2.5 μM TIA-1, 0.5 μM ssDNA TC5, 10 μM ZnCl
2
. Condi ions:
20 mM HEPES, 50 mM NaCl, 15 mM A ginine, pH 7.2, a 30°C wi h shaking. 3 eplica es. E o s ep esen S.D. Scale ba ep esen s 200 nm.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608068
Wes e al. Regula ion o TIA-1 Condensa es
ha we e s able o e 2 h (Figu e 4D). Addi ion o equimola FUS-
RBD o TIA-1 and TC3 u he enhanced LLPS, esul ing in g ea ly
inc eased u bidi y and subs an ially la ge LLPS d ople s (Figu es
4C,D). Unlike he s able TIA-1:TC3 samples, TIA-1:TC3:FUS-RBD
condensa es used oge he o e 2 h. To es whe he hese changes
we e due o RGG egions o FUS, a sho e sion o FUS RBD
(amino acids 269-454) wi hou RGG1 o RGG3 egions (FUS-RBD-
ΔRGG1/3) was mixed wi h TIA-1:TC3. On addi ion o FUS-RBD-
ΔRGG1/3, no change in u bidi y o appea ance o LLPS was
obse ed, wi h d ople s appea ing ela i ely s able o e a 2 h ime
pe iod (Figu es 4C,D). Al hough TC3 was designed wi h TIA-1
binding si es, addi ion o TC3 ssDNA o FUS-RBD esul ed in
inc eased u bidi y, wi h a small inc ease in he numbe o LLPS
d ople s (Figu es 4C,E). In con as , he addi ion o TC3 o FUS-
RBD-ΔRGG1/3 did no inc ease u bidi y o d ople o ma ion
(Figu es 4C,E). In o de o check whe he FUS-RBD could be
o ming p omiscuous in e ac ions by displacing TIA-1 om he
TC3 ssDNA, we de e mined i s di ec binding a fini y o a TC DNA
sequence in compa ison o ha o TIA-1 RRM2,3. This showed ha
he FUS-RBD a fini y (K
d
=30nM)wassignifican ly lowe han ha
o TIA-1 RRM2,3 (K
d
=2nM)(Supplemen a y Figu e S6)andis
unlikely o displace TIA-1 unde he condi ions o he LLPS assay.
These esul s show ha he addi ion o an RGG- ich p o ein can
modula e he nucleic acid-induced LLPS o TIA-1, po en ially ia
in e ac ions wi h RRM-bound nucleic acid and/o wi h TIA-1 P LD.
Following he obse a ion o changes in LLPS o TIA-1:
TC3 condensa es, we nex examined he fib il o ma ion
capaci y o TIA-1 in he p esence o RGG- ich FUS-RBD as
moni o ed by ThT fluo escence and TEM. Whe eas 2.5 µM TIA-
1 showed minimal agg ega ion wi h only a small inc ease in ThT
fluo escence a e ~8 h, TC3 ssDNA e ficien ly enhanced ThT
fluo escence o e ime, esul ing in amyloid-like fib ils, as
obse ed unde TEM (Figu es 5A,B). In con as , he addi ion
o FUS-RBD o TIA-1:TC3 esul ed in a d ama ic educ ion in
ThT fluo escence, sugges ing a subs an ial dec ease in fib il
o ma ion, confi med by TEM imaging (Figu es 5C,D).
In e es ingly, he small inc ease in ThT fluo escence o TIA-1
alone is also inhibi ed in he p esence o FUS-RBD (Figu e 5C).
FIGURE 4 | FUS-RBD enhances nucleic acid induced LLPS o TIA-1. (A) Domain s uc u es o TIA-1 (wi h P LD sequence alongside highligh ing Q, N and Y amino
acids) and FUS cons uc s: FUS-RBD and FUS-RBD-ΔRGG1/3 cons uc s (wi h RGG sequences shown unde nea h highligh ing R, F and Y amino acids). (B) DIC
mic oscopy o 2.5 μM TIA-1; 0.5 μM ssDNA TC3; 2.5 μM FUS-RBD-RGG; 2.5 μM FUS-RBD-sho a oom empe a u e (C) Tu bidi y assay o 2.5 μM TIA-1; 0.5 μM
ssDNA TC3; 2.5 μM FUS-RBD-RGG; 2.5 μM FUS-RBD-sho . 20 min incuba ion a 25°C. 3 eplica es. E o ba s ep esen S.D. (D) DIC mic oscopy o 2.5 μM
TIA-1 combined wi h 0.5 μM ssDNA TC3 alone o wi h 2.5 μM FUS-RBD-RGG o 2.5 μM FUS-RBD-sho a 30 and 120 min incuba ion a oom empe a u e. (E) 0.5 μM
ssDNA TC3 alone o wi h 2.5 μM FUS-RBD-RGG o 2.5 μM FUS-RBD-sho . Scale ba s depic 10 μm. Condi ions: 20 mM HEPES, 50 mM NaCl, 12 mM a ginine,
pH 7.2, a 25°C.
F on ie s in Molecula Biosciences | www. on ie sin.o g July 2022 | Volume 9 | A icle 9608069
Wes e al. Regula ion o TIA-1 Condensa es