A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Allos e ic con ol o dynamin- ela ed p o ein
1 h ough a diso de ed C- e minal Sho
Linea Mo i
Isabel Pé ez-Jo e
1,2,9
, K is y Rochon
3,9
,DiHu
4,9
, Mukesh Mahajan
4
,
Pooja Madan Mohan
4
, Isaac San os-Pé ez
5
, Julene O mae xea Gisasola
1,2
,
Juan Manuel Ma inez Gal ez
1,2
, Jon Agi e
6
,XinQi
4,7
,JasonA.Mea s
3,7,8
,
Anna V. Shny o a
1,2
& Rajesh Ramachand an
4,8
The mechanochemical GTPase dynamin- ela ed p o ein 1 (D p1) ca alyzes
mi ochond ial and pe oxisomal fission, bu he egula o y mechanisms emain
ambiguous. He e we find ha a conse ed, in insically diso de ed, six- esidue
Sho Linea Mo i a he ex eme D p1 C- e minus, named CT-SLiM, con-
s i u es a c i ical allos e ic si e ha con ols D p1 s uc u e and unc ion
in i o and in i o. Ex ension o he CT-SLiM by non-na i e esidues, o i s
in e ac ion wi h he p o ein pa ne GIPC-1, cons ains D p1 subuni con-
o ma ional dynamics, al e s sel -assembly p ope ies, and limi s coope a i e
GTP hyd olysis, su p isingly leading o he fission o model memb anes
in i o. In i o, he in ol emen o he na i e CT-SLiM is c i ical o p oduc i e
mi ochond ial and pe oxisomal fission, as bo h dele ion and non-na i e
ex ension o he CT-SLiM se e ely impai hei p og ession. Thus, con a y o
p e ailing models, D p1-ca alyzed memb ane fission eliesonallos e iccom-
munica ion media ed by he CT-SLiM, decele a ion o GTPase ac i i y, and
coupled changes in subuni a chi ec u e and assembly-disassembly dynamics.
In insically diso de ed p o eins (IDPs) and s uc u ed p o eins ha
con ain in insically diso de ed egions (IDRs) a e ubiqui ous com-
p ising nea ly hal o he human p o eome1–3. In con as o he ela-
i ely s a iona y loops and u ns ha connec seconda y s uc u e
elemen s in compac ly olded p o ein domains, IDRs pe sis as a highly
dynamic con o ma ional ensemble, which in many ins ances unde -
goes a localized diso de - o-o de s uc u al ansi ion upon pa ne
in e ac ions (wi hp o ein, lipid, o nucleo ide) and/o ia a ious pos -
ansla ional modifica ions (PTMs)4–6. Consequen ly, IDRs unc ion as
egula o y nodes o hubs ha go e n hos p o ein unc ion by an-
sc ibing biological in o ma ion om mul iple in e ac ions and mod-
ifica ions in o disce nible al e a ions in local p o ein old, dynamics,
and mac omolecula assembly, including p o ein condensa ion ia
liquid-liquid phase sepa a ion (LLPS)7–10.
This gene alized desc ip ion o IDR s uc u e and unc ion also
pe ains o dynamin- ela ed p o ein 1 (D p1), a sel -assembling, mul i-
domain GTPase ha mechanochemically cons ic s ubula memb ane
in e media es en ou e o mi ochond ial fission11,12. D p1 con ains
Recei ed: 2 July 2023
Accep ed: 7 Decembe 2023
Check o upda es
1
Depa men o Biochemis y and Molecula Biology, Uni e si y o he Basque Coun y, 48940 Leioa, Spain.
2
Ins i u o Biofisika, CSIC, UPV/EHU, 48940
Leioa, Spain.
3
Depa men o Pha macology, Case Wes e n Rese e Uni e si y School o Medicine, Cle eland, OH 44106, USA.
4
Depa men o Physiology and
Biophysics, Case Wes e n Rese e Uni e si y School o Medicine, Cle eland, OH 44106, USA.
5
Elec on Mic oscopy and C ys allog aphy Cen e o Coop-
e a i e Resea ch in Biosciences (CIC bioGUNE), Bizkaia Science and Technology, Pa k Bld 800, 48160-De io Bizkaia, Spain.
6
Yo k S uc u al Biology
Labo a o y, Depa men o Chemis y, Uni e si y o Yo k, Hesling on, YO10 5DD Yo k, UK.
7
Cen e o Mi ochond ial Diseases, Case Wes e n Rese e Uni e si y
School o Medicine, Cle eland, OH 44106, USA.
8
Cle eland Cen e o Memb ane and S uc u al Biology, Case Wes e n Rese e Uni e si y School o
Medicine, Cle eland, OH 44106, USA.
9
These au ho s con ibu ed equally: Isabel Pé ez-Jo e , K is y Rochon, Di Hu. e-mail: [email p o ec ed];
[email p o ec ed]
Na u e Communica ions | (2024) 15:52 1
1234567890():,;
1234567890():,;
mul iple IDRs, anging up o 134 amino acid (aa) esidues in leng h,
which make up >20% o i s o e all sequence11,13,14 (Supplemen a y
Fig. 1a). Ye , many o hese IDRs a e ei he absen o un esol ed in any
a ailable D p1 X- ay13 o c yo-EM s uc u e15,16 o da e (Fig. 1aand
Supplemen a y Fig. 1b), obscu ing u he unc ional cha ac e iza ion.
These include: (i) Molecula Recogni ion Fea u es (MoRFs)8,10 o
~10–25 aa esidues, such as MoRFs-1 and −2 embedded wi hin he
la gely diso de ed a iable domain (VD), enabling di ec D p1-
memb ane in e ac ions17 and (ii) Sho Linea Mo i s (SLiMs)8,10 o
~3–12 aa esidues nes ed wi hin highly s uc u ed domains, such as he
G-domain ‘80-loop’and s alk ‘L1N’loop ha di ec p o ein-p o ein
in e ac ions du ing D p1 sel -assembly14. One such IDR is a unique
s e cho ~6 aa esidues a he D p1 ex eme C- e minus, which wecall
he CT-SLiM (Fig. 1a and Supplemen a y Fig. 1a, b) ha , unlike o he
Volume (mL)
Mola mass (kDa)
0
100
200
300
400
500
11 12 13 14 15 16 17
WT
CT+
'CT4
'CT6
4-me
2-me
SEC-MALS
5 PM p o ein a injec ion
~0.5 PM peak p o ein
a he de ec o
0%
10%
20%
30%
40%
50%
60%
Helical leng h (nm)
WT 118.6
± 39.0 (3.5) n=124
CT+ 192.3
± 104.9 (9.1) n=134
Mean
± SD (SEM)
b
cd
e
+ GMP-PCP
WT CT+
'CT4 'CT6
200 nm
50 nm
a
M1
(N- e m) I693
(C- e m)
BSE
GTPase (G) domain
D p1'VD c ys al s uc u e
(PDB ID: 4BEJ)
“CT-SLiM”
D p1 (iso3; 699 aa)
BSE GTPase BSE Middle GED BSE
NC
VD
RETHLW
699694
“CT-SLiM”
Hinge 1
Hinge 1
R403
% o o al
0%
10%
20%
30%
40%
WT 46.9
± 7.0 (0.7) n=84
CT+ 49.5
± 7.0 (0.7) n=91
Mean
± SD (SEM)
Ring diame e (nm)
0%
10%
20%
30%
40%
50%
% o o al
'CT4 29.2
± 6.0 (0.5) n=110
'CT6 32.4
± 8.0 (0.4) n=317
Ring diame e (nm)
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 2
IDRs in D p1, is highly conse ed among me azoans (Supplemen a y
Fig. 1a). Howe e , i s unc ion(s) emain la gely unexplo ed.
Recen s udies ha e indica ed ha his CT-SLiM cons i u es an
a ypical PDZ domain binding mo i (PBM) ha specifically in e ac s
wi h he PDZ domain-con aining adap o p o ein GIPC-1 (GAIP in e -
ac ing p o ein, C- e minus 1)18,19. CT-SLiM-bound GIPC-1, in u n,
associa es wi h he F-ac in minus-end-di ec ed mo o myosin VI
(MYO6) o guide D p1 p esumably o F-ac in-p e-cons ic ed mi o-
chond ial di ision si es18–21. Howe e , whe he o how di ec GIPC-1-
D p1 in e ac ions ia he CT-SLiM influence D p1 s uc u e and/o
unc ion emains unknown.
He e, using a comp ehensi e oolki o s uc u al, cell biological,
and in i o econs i u ion app oaches, we show ha a dele ion (ΔCT)
o a non-na i e ex ension (CT + ) o he CT-SLiM dis inc ly al e s D p1
con o ma ional dynamics, oligome iza ion p opensi y, sel -assembly
geome y, andcoope a i e GTPaseac i i y,inaddi ion o di e en ially
a ec ing D p1 capaci y o emodel a ge memb anes. We demon-
s a e ha whe eas he ΔCT a ian s exhibi a p edic able loss-o -
unc ion by ei he al e ing o inhibi ing memb ane fission bo h in i o
and in i o, he CT+ a ian s display an abe an gain-o - unc ion by
obus ly ca alyzing memb ane fission in i o, while emaining
ep essed in media ing mi ochond ial fission in i o. By con as , WT
D p1, which is limi ed o cons ic ing memb anes on i s own in i o,
ema kably p og esses owa d memb ane fissionuponna i eCT-
SLiM-e ec ed GIPC-1 in e ac ions. Taken oge he , ou da a indica e a
c i ical ole o he na i e CT-SLiM in go e ning D p1 s uc u e, con-
o ma ional dynamics, andmechanoenzyma ic memb ane emodeling
ac i i y. Fu he mo e, key p o ein pa ne in e ac ions o D p1, such as
ha o he CT-SLiM, eme ge as an essen ial egula o y elemen in he
allos e ic con ol o D p1 unc ion du ing mi ochond ial fission.
Resul s
CT-SLiM modifica ions al e D p1 sel -assembly p opensi y and
geome y
To disce n he ole o he CT-SLiM, we gene a ed a hos o D p1 a -
ian s wi h ei he unca ed o ex ended C- e mini (Supplemen a y
Fig. 1c). The unca ed a ian s had he las ou (ΔCT4) o six (ΔCT6)
esidues o he na i e CT-SLiM emo ed in o de o sepa a e he
po en ial elec os a ic (R694E695) and hyd ophobic (L698W699)con ibu-
ions o his segmen o D p1-pa ne p o ein in e ac ions18.Con-
e sely, he ex ended a ian s (CT + ) had non-na i e sequences o
di e en leng hs and composi ion, including a fini y and epi ope ags,
appended o he CT-SLiM. These non-na i e ex ensions we e in o-
duced o isola e he backbone ca boxyla e moie y o he C- e minal
esidue, a equi emen o high-a fini y PDZ domain binding22, om
he p edic ed PDZ domain ecogni ion sequence in D p1
(696THLW699)18. In addi ion, we sough o de e mine he influence o
non-na i e C- e minal ex ensions on D p1 s uc u e and unc ion,
which in pas s udies ha e p oduced con ounding and conflic ing
esul s23–25.Toenablepu ifica ion, WT D p1 andselec CT a ian s we e
modified a he N- e minus wi h a His
6
a fini y ag (see Me hods),
which as p e iously shown26,27 did no a ec D p1 sel -assembly o
GTPase ac i i y in i o. Besides, N- e minally epi ope (Myc)- agged
D p1 e ec i ely es o ed mi ochond ial fission in D p1 knockou (KO)
cells17,26 indica ing ha hese N- e minal modifica ions nei he a ec
D p1 unc ion in i o.
Nega i e-s ain elec on mic oscopy (NS-EM) analysis e ealed
conside able al e a ions in he D p1 oligome s uc u e due o he CT-
SLiM modifica ions. In he p esence o he non-hyd olyzable GTP
analogue, GMP-PCP, which mimics GTP binding and p omo es D p1
helical sel -assembly in solu ion26, WT D p1 cha ac e is ically o med a
mix u e o oligome ic ings and highe -o de spi als o a consis en
diame e and leng h (Fig. 1b–d). In con as , he ΔCT4 and ΔCT6 a -
ian s ailed o assemble in o any such egula highe -o de s uc u es.
Ins ead, he ΔCT4 and ΔCT6 a ian s p edominan ly cons i u ed i-
angula ly shaped nubs o much smalle dimensions wi h li le o no
indica ion o u he highe -o de sel -assembly (Fig. 1b, c). Con-
e sely, he CT+ a ian o med consis en ly longe sup amolecula
helical assemblies, al hough simila in o e all helical diame e o WT
(Fig. 1b–d).
Size-exclusion ch oma og aphy-coupled mul i-angle ligh sca -
e ing (SEC-MALS) analyses o hese a ian s in he nucleo ide- ee
apo s a e a physiologically ele an concen a ions in solu ion28
(~0.5µM a peak de ec ion upon ~10- old SEC dilu ion) e ealed u -
he di e ences in hei oligome iza ion p ope ies ela i e o WT
(Fig. 1e). The ΔCT4 and ΔCT6 a ian s exhibi ed a sha p dime -
e ame equilib ium simila o WT, albei ending ma ginally owa d
minimal dime s unde he condi ions. In con as , he ex ended CT+
a ian la gely a o ed highe -o de oligome s consis en wi h i s
enhanced helical sel -assembly in he p esence o GMP-PCP (Fig. 1b,
Supplemen a y Fig. 2a, b). This g ea e oligome iza ion p opensi y o
he CT+ a ian ela i e o WT was e iden o e a wide ange o
p o ein concen a ions (Supplemen a y Fig. 2c). Besides, i was
independen o he non-na i e CT+ sequence as his endency was
also mani es in a CT+* a ian con aining an ex ension o a di e en
leng h (14 aa esidues) and composi ion (Supplemen a y Fig. 2d).
Sho ening he non-na i e CT sequence o he CT+ a ian om 24 o
9 aa esidues by p o eoly ic clea age (CT+sh) educed i s highe -o de
oligome iza ion p opensi y (Supplemen a y Fig. 2e), al hough his
emained no iceably g ea e han ha o WT D p1. On he o he hand,
sho ening he N- e minal ag sequence om 36 o 7 aa esidues had
no palpable e ec on D p1 oligome iza ion (Supplemen a y Fig. 2 ).
These da a indica ed ha he diso de ed D p1 CT-SLiM is a c i ical
de e minan o D p1 sel -assembly and helical p opaga ion.
CT-SLiM modifica ions al e D p1 con o ma ional dynamics and
s uc u e
To gain insigh in o he molecula mechanisms unde lying CT-SLiM
unc ion, we used NS-EM and pe o med 2D image classifica ion o
assess he impac o he a ious CT modifica ions on D p1 subuni
Fig. 1 | CT-SLiM modifica ions a ec D p1 oligome iza ion p opensi y and
helical geome y. a The loca ion and polypep ide sequence o he CT-SLiM in D p1
iso o m 3 p ima y s uc u e is shown. The c ys als uc u e benea h co esponds o
he D p1ΔVD dime wi h a colo -coded ep esen a ion o domain a angemen in a
monome . BSE (pu ple) is he bundle signaling elemen . The s alk comp ises a ou -
helical bundle composed o discon inuous middle (blue) and GED (GTPase e ec o
domain; o ange) egions. GTPase (G) domain is shown in g een. Connec ing black
lines ep esen a ew p ominen IDRs in D p1. The VD connec s he middle and GED
egions, whe eas he 80-loop and LIN loops a e nes ed wi hin he G and s alk
(middle) domains, espec i ely. The inse is a zoomed-in iew o he BSE showing
he well- esol ed D p1 N- e minal BSE helix (beginning om aa esidue 1). The las
six esidues o he D p1 C- e minus (R694-W699), an IDR which we call he CT-SLiM
and ep esen ed he e by a cu ed black line, emain diso de ed. I693, he las
esol ed esidue o he C- e minal BSE helix is highligh ed.bRep esen a i e NS-EM
images o WT D p1 and CT a ian s in he p esence o he non-hyd olyzable GTP
analogue, GMP-PCP. Scale ba , 200 nm. The le inse unde each panel shows a
zoomed-in iew o he boxed egion in he abo e mic og aph, whe eas 2D class
a e ages o he p edominan oligome ( ing) mo phology a e shown o hei igh .
Inse s scale ba , 50 nm. Da a shown he e a e o mouse CT+ D p1. Human CT+ D p1
da a a e shown in Supplemen a y Fig. 2a. His og ams showing he dis ibu ion o
assessed ing diame e (c) and helical polyme leng h (d) o WT D p1 and CT
a ian s. ΔCT4/6 D p1 do no o m helical polyme s. Mean ± SD (SEM) is indica ed.
n is he numbe o pa icles. eSEC-MALS elu ion and mola mass p ofiles o human
WT D p1 and CT a ian s sie ed h ough a Supe ose 6 10/300 GL column. When
injec ed a 5 µM, D p1 is dilu ed o ~0.5 µM peak concen a ion on a i al a he LS
and dRI de ec o s. Ho izon al lines indica e he heo e ical masses o a D p1 dime
(2-me ) and e ame (4-me ).
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 3
con o ma ional dynamics and sel -assembly unde di e en
nucleo ide-bound s a es (Fig. 2a, Supplemen a y Fig. 3a, b).
In he apo s a e, we de ec ed wo di e en o ien a ions o he
WT D p1 dime —an S-shaped op-down (o bo om-up) o ien a ion
and a V-shaped side-on o ien a ion wi h p ominen densi ies e iden
o he dime ic s alk and he wo indi idual GTPase (G) domains
(Fig. 2a, Supplemen a y Fig. 3b, c). No ably, in he S-shaped o ien-
a ion, he G domains we e se a apa , whe eas in he V-shaped
o ien a ion, he G domains appea ed o be posi ioned in close
p oximi y. Rema kably, he S-shaped s uc u e was ne e ound o
he CT+ a ian in he apo s a e and was ela i ely poo ly sampled by
he ΔCT4/6 a ian s (Fig. 2a, Supplemen a y Fig. 3b, c). As di e en ial
g id deposi ion o p e e ed o ien a ions we e unlikely o be influ-
en ial ac o s owing o he minimal na u e o he CT modifica ions
in ol ed (CT-SLiM dele ion o a sho non-na i e ex ension), we
easoned ha hese wo o ien a ions likely co espond o wo di -
e en solu ion con o ma ions o he D p1 dime ha in e con e
dynamically (Fig. 2a, Supplemen a y Fig. 3c).
W90 only 330
W552 only 332
W699 only 337
WT 337
Omax (T p)
0
0.2
0.4
0.6
0.8
1
320 340 360 380 400
Wa eleng h (nm)
No malized T p emission in ensi y
W699F 329
W552F 332
W90F 341
WT 337
Omax (T p)
0
0.2
0.4
0.6
0.8
1
320 340 360 380 400
Wa eleng h (nm)
CT+ 332
WT 337
Omax (T p)
0
0.2
0.4
0.6
0.8
1
320 340 360 380 400
Wa eleng h (nm)
a
b
d
c
WT ΔCT6
G
G
S
G
G
S
G
G
S
S
GG
G
SAXS-de i ed en elope o a
D p1 dime (R403A)
90o
90o180o
Side-on iew Top-down iew Bo om-up iew
Top
Bo om
G
S
G
G
S
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Open
Compac
Closed
Compac
0%
20%
40%
60%
80%
100%
Wide
Rings
Collapsed
Rings
0%
20%
40%
60%
80%
100%
Wide
Rings
Collapsed
Rings
0%
20%
40%
60%
80%
100%
Wide
Rings
Collapsed
Rings
0%
20%
40%
60%
80%
100%
Wide
Rings
Collapsed
Rings
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Ex ended Compac
0%
20%
40%
60%
80%
100%
Ex ended Compac
*
Ex ended Compac
*
Ex ended Compac
*
Ex ended Compac
+GTP
*
10 nm
Ex ended Compac
Dominan species
WT
*
Ex ended Compac
*
Ex ended Compac
*
Open
Compac
Closed
Compac
10 nm
*
Ex ended Compac
apo
ΔCT4
ΔCT6
CT+
Wide
Ring
Collapsed
Ring
*
Wide
Ring
Collapsed
Ring
*
Wide
Ring
Collapsed
Ring
Wide
Ring
Collapsed
Ring
20 nm
+GMP-PCP
~22 nm
G
G
S
Ex ended
S
GG
Compac
*
*
GG
S
GG
S
CT-SLiM
Side-on iew
Side-on iew
*
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 4
Fo he apo WT D p1 dime , he ex ended S-shaped con o me
was de ec ed a a ~4- old g ea e incidence han he compac V-
shaped con o me , indica ing a g ea e esidence ime o he na i e
dime in he ex ended con o ma ion (Fig. 2a). These da a indica ed
ha he na i e CT-SLiM es ic s D p1 con o ma ional dynamics in
solu ion and e ains D p1 p edominan ly in he ex ended con o ma-
ion, whe eas i s absence o non-na i e ex ension in he ΔCT4/6 and
CT+ a ian s, espec i ely, di e en ially elie es his au o-inhibi ion
a o ing hei con e sion, o a ying ex en s, o he al e na e compac
con o ma ion. Mo eo e , in he p esence o GMP-PCP, he oligome ic
ings o med by he ΔCT4/6 and CT+ a ian s we e la gely i egula o
poo ly o de ed (Fig. 2a) sugges ing ha he na i e CT-SLiM also
unc ions as a space ha se s he egis e and geome y o in e -
subuni in e ac ions du ing nucleo ide-dependen helical sel -
assembly. Fu he mo e, unlike WT, which e e ed o he ex ended
dime con o ma ion upon GTP hyd olysis, he CT+ and ΔCT4/6 a -
ian s la gely emained in he compac con o ma ion (Fig. 2a, Supple-
men a y Fig. 3d). These da a indica ed ha he compac CT+ D p1
con o me likely mimics an assembly-p imed s a e based on i sg ea e
highe -o de oligome iza ion p opensi y ela i e o WT bo h in
absence and p esence o nucleo ide.
3D econs uc ion om 2D class a e ages o he ex ended WT and
compac ΔCT6 D p1 con o ma ions u he allowed us o dock he
a ailable c ys al s uc u e o he ΔVD D p1 dime and examine he
na u e o he con o ma ional ea angemen s (Fig. 2b, Supplemen a y
Fig. 3c). Wi h he ex ended con o ma ion, he G domains o he docked
ΔVD D p1dime s e chedbeyond he compu ed edgedensi ies.These
da a sugges ed ha ei he he ΔVD a ian is in an al e na e con-
o ma ion compa ed o WT14, wi h he WT G domains ucked in owa d
he s alk as ecen ly indica ed16, o ha ou econs uc ed s uc u e
emainspa iallyun esol ed owing oinhe en dynamicsa oundhinge
1 a he BSE-s alk in e sec ion (Fig. 1a). Fi ing o he ΔVD D p1 dime
s uc u e in o he compu ed 3D olume o he compac con o ma ion,
howe e , equi ed a la ge-scale eposi ioning o he G domains a ound
hinge 1 (Fig. 2b). Modeling o hese wo con o ma ions using he
c ys al s uc u e (Fig. 2b) and back p ojec ion o 2D class a e ages
om he compu ed 3D olumes (Fig. 2b) e ealed ha he compac
con o ma ion sampled by ΔCT6 D p1 is no obse ed in any o he
p ojec ed 2D class a e ages o he ex ended WT D p1 dime . These
da a indica ed ha he obse ed ex ended and compac D p1 o ms
a e indeed con o ma ionally dis inc .
Toconfi m he la ge-scale flexibili y o he D p1 dime as indica ed
by he EM da a, we mapped he con o ma ional landscape o a minimal
D p1dime insolu ion using small-angle X- aysca e ing(SAXS)29 as an
o hogonal app oach (Fig. 2c and Supplemen a y Fig. 4). The he e o-
genous mix o dime s, e ame s, and highe -o de oligome s p esen
in dynamic equilib ium o WT D p1 and he CT a ian s is incompa-
ible wi h SAXS and canno be analyzed. The e o e, we employed a
R403A mu a ion in D p1 (R399A in Dyn130)(Fig.1a) ha es ic s D p1
p edominan ly o a minimal dime in solu ion (Supplemen a y Fig. 4a),
simila o he minimal Dyn1 dime p e iously assessed by SAXS31.
Rema kably, ab ini io econs uc ion o he mos p obable low-
esolu ion molecula en elope o he R403A D p1 dime e ealed an
o e all shape ha was compa ible wi h bo h he ex ended and com-
pac con o ma ions (Fig. 2c), wi h su ficien olume p esen be ween
he G domains and below he s alk o he o e laid ΔVD D p1 dime
c ys als uc u e oaccommoda e bo h shapes.Thus, he minimal D p1
dime in solu ion is highly dynamic and capable o in e con e sion
be ween ex ended and compac s a es.
To unde s and he molecula basis o he CT+ D p1 a ian ’s dis-
inc i ely compac con o ma ion and gain-o - unc ion in sel -assem-
bly, we used AlphaFold32 o p edic he influence o he CT+ sequence
ex ension on D p1 s uc u e. Rema kably, he compu a ional da a
sugges ed ha whe eas he N- e minal His
6
a fini y ag in ou WT D p1
was mos ly diso de ed, he non-na i e CT ex ension in CT+ D p1
p opaga ed as a α-helixincloseapposi ion o he opo heGdomain,
po en ially cons aining dynamics a he adjacen nucleo ide-sensi i e
Gdomain-BSE in e ace (Supplemen a y Fig.5a).Consis en wi h his,a
di ec compa ison o he op- anked s uc u es in isola ion (Supple-
men a y Fig. 5b) and upon supe posi ion in o he a ailable D p1
polyme c yo-EM s uc u e (Supplemen a y Fig. 5c), e ealed a sligh
inwa d buckling o he G domain owa d he BSE in CT+ D p1 com-
pa ed o WT D p1. In addi ion, gi en he p oximi y o he CT-SLiM o
he s alk o he adjacen monome in he D p1 polyme (Supplemen-
a y Fig. 5c), he modeling da a u he indica ed ha he CT+ ex en-
sion may influence D p1 subuni -subuni in e ac ions du ing highe -
o de helical sel -assembly.
We used in insic T yp ophan (T p) Fluo escence Spec oscopy
(iTFS)33,34 o expe imen ally alida e hese in silico p edic ions
(Fig. 2d). T p emission is highly sensi i e o i s mic oen i onmen and
he e o e se es as an accu a e p obe o p o ein con o ma ion o
con o ma ional changes25,26. When exci ed selec i ely a λ=295nm,
he T p emission spec um is blue-shi ed (peaking a sho e wa e-
leng hs) when p esen in a nonpola en i onmen , and ed-shi ed
(peaking a longe wa eleng hs) when exposed o a pola o aqueous
milieu. D p1 con ains h ee na i e T p esidues a posi ions 90, 552 and
699 (ubiqui ous iso o m 3 numbe ing; Fig. 1a). O hese, only W90
p esen in he G domain is s uc u ally esol ed13 (Supplemen a y
Fig. 5d), whe eas W552 and W699 a e loca ed in he diso de ed VD and
CT-SLiM, espec i ely (Supplemen a y Fig. 1a). Using si e-di ec ed
D p1 mu an s ha e ained only one o he h ee na i e T p o
ha con ained only a single na i e T p- o-Phe subs i u ion, we asce -
ained ha D p1 T p emission p ima ily o igina es om W699, he
e minal esidue o he CT-SLiM.
Consis en wi h he pa ial bu ial o W90 in he D p1 G domain
s uc u e13 (Supplemen a y Fig. 5d), he W90-only mu an displayed a
p onounced blue shi in T p emission ela i e o WT (Fig. 2d). Simi-
la ly, heW552-onlymu an also exhibi edasignifican blueshi , albei
less han ha o he W90-only mu an , indica ing ha W552 is also
pa ially occluded om sol en in he VD con o ma ional ensemble
(Fig. 2d). By con as , he W699-only mu an was p onouncedly ed-
shi ed and was iden ical o WT in emission spec a (Fig. 2d). These
da a indica ed ha W699 in WT D p1 is sol en accessible, and is he
p ima y emi e la gely owing o i s loca ion wi hin T p-T p homo-
Fig. 2 | CT-SLiM modifica ions al e D p1 con o ma ional dynamics. a NS-EM 2D
class a e ages o dime s and oligome ic ings in he apo, GMP-PCP-bound,andGTP
hyd olysis s a es o human WT D p1 and CT a ian s. D p1 dime con o ma ion in
he apo and GTP hyd olysis s a es is classified as ei he ex ended o compac , wi h
he la e classified u he in o open-compac o closed-compac s a es. Oligo-
me ic ing mo phology in he p esence o GMP-PCP is classified in o wide and
collapsed ing s a es. Top-down and side-on iews o he ΔVD D p1 dime c ys al
s uc u e (PDB ID: 4BEJ) as well as colo -coded ca oon illus a ions o domain
ea angemen s be ween he ex ended and compac s a es a e shown a he a -
igh co ne . G e e s o he G domain, whe eas S e e s o he s alk. The numbe o
pa icles in he ex ended (E) and compac (C) con o ma ions in he apo and +GTP
s a es, and displaying Wide Ring (WR) and Collapsed Ring(CR) mo phologies in he
+GMP-PCP s a e a e shown in Me hods. bi) WT and ΔCT6 D p1 3D densi ies en-
de ed om c yoSPARC homogeneous efinemen p ocessing, ii) WT and ΔCT6
D p1 idealized models gene a ed om he ΔVD D p1 dime c ys al s uc u e (PDB
ID: 4BEJ) by h esholding he s uc u e o 40Å o WT and by eposi ioning he G
domains and h esholding o 40 Å o ΔCT6 D p1, and iii) 2D class a e ages gen-
e a ed om he WT and ΔCT6 D p1 idealized models a e shown in ows. cViews o
he SAXS-de i ed ab ini io en elope o dime ic R403A D p1 o e laid wi h he ΔVD
D p1 dime c ys al s uc u e (PDB ID: 4BEJ). dNo malized T p emission spec a o
human WT D p1 in compa ison o single W-only mu an s, single W- o-F mu an s,
and he CT+ D p1 a ian a e shown, espec i ely, om le o igh . The wa eleng h
o maximum emission (λ
max
)isindica ed.
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 5
FRET dis ance (~24 Å)33 o he high-ene gy FRET dono , W90 (Supple-
men a y Fig. 5d). Consis en wi h his assessmen , he W90F mu an
was ed-shi ed (by 11 nm) compa ed o he W90-only mu an , whe eas
he W699F mu an was blue-shi ed (by 8 nm) ela i e o he W699-
only mu an (Fig. 2d). The W552F mu an , on he o he hand, did no
expe ience any such change (Fig. 2d). These da a confi med ha W699
in he WT D p1 CT-SLiM is exposed o wa e and is highly esponsi e o
i s local en i onmen . No ably, by con as o WT D p1, CT+ D p1
emission was blue-shi ed (by 5nm) indica ing ha W699 in CT+ D p1
is ins ead bu ied and ela i ely sol en inaccessible (Fig. 2d). No such
change in T p emission was obse ed when he 36 aa- esidue N-
e minal His
6
ag o WT D p1 was eplaced by a ela i ely sho 7-aa
esidue o e hang (Supplemen a y Fig. 5e), indica ing ha he di e -
ence in he en i onmen o CT-SLiM in CT+ D p1 is p ima ily due o he
non-na i e CT ex ension. Thus, oge he wi h he c yo-EM da a and
AlphaFold p edic ions, he iTFS da a demons a ed ha non-na i e CT
ex ension o CT+ D p1 al e s CT-SLiM mic oen i onmen and o e all
D p1 con o ma ion. WT and CT+ D p1 hus popula e dis inc con-
o ma ional s a es.
CT-SLiM modifica ions di e en ially a ec D p1 coope a i e
GTP hyd olysis
We nex de e mined he impac o he a ious CT-SLiM modifica ions
on D p1 GTPase ac i i y unde basal condi ions in solu ion and upon
helical sel -assembly on CL-con aining memb anes. In he apo s a e,
he CT a ian s e ained he cha ac e is ic capaci y o WT D p1 o sel -
assemble on, and ubula e, la ge CL-con aining liposomes o na ow
diame e s (Supplemen a y Fig. 6a, b). Simila ly, he CT a ian s also
assembled on highly cu ed and p e o med galac osylce amide-laden
CL-con aining lipid nano ubes (GalCe CL-NTs) iden ically o WT
(Supplemen a y Fig. 6c). These da a indica ed ha he a ious CT
modifica ions do no ad e sely a ec s alk-media ed D p1 sel -
assembly on memb anes.
Su p isingly, howe e , he ΔCT4 and ΔCT6 a ian s bo h exhib-
i ed a ~3- old g ea e a e o GTP hyd olysis in solu ion compa ed o
he CT+ a ian and WT, which we e simila in basal GTPase ac i i y
(Fig. 3a). By con as , he CT+ a ian displayed a ~2- old lowe ac i i y
compa ed o he ΔCT4/6 a ian s and WT when assayed on CL-
con aining liposomes (Fig. 3b). Analysis o he p e-s eady s a e ‘bu s ’
kine ics e ealed ha he ΔCT4/6 a ian s hyd olyzed GTP a a sig-
nifican ly as e a e han WT unde bo h condi ions (Fig. 3c). Simila
ends also held up o he CT a ian s on GalCe CL-NTs on which
ΔCT4/6D p1hyd olyzedGTP a asignifican ly as e a e hanbo hWT
and CT+ D p1 in he o de : ΔCT6 > ΔCT4 > WT > CT+ (Supplemen a y
Fig. 6d, e). Co-sedimen a ion analysis o he D p1 a ian s on GalCe
CL-NTs, in he absence and p esence o GTP, demons a ed a g ea e
s eady-s a e associa ion o he as e GTP-hyd olyzing ΔCT4/6 a ian s
wi h he lipid empla es han WT o CT+ D p1 (Supplemen a y Fig. 6 ).
Consis en wi h his, NS-EM on GalCe CL-NTs in he p esence o GTP
e ealed pe sis en sel -assembly o he ΔCT6 a ian on he lipid
empla es. Such pheno ype was absen o bo h WT and CT+ D p1,
which showed widesp ead disassembly and memb ane dissocia ion
wi h GTP (Supplemen a y Fig.7). In e es ingly,in hep esenceo GMP-
PCP, he ΔCT6 a ian , ela i e o WT and CT+ D p1, o med highly
p ocessi e helical polyme s, which in many ins ances ex ended
beyond he lipid empla e (Supplemen a y Fig. 7).
Collec i ely, hese da a indica ed ha in he absence o he CT-
SLiM, ansi ion s a e-dependen in e -subuni G-domain dime iza ion,
coope a i e GTP hyd olysis, GDP/Pi elease, and G-domain dime dis-
assembly c i ical o p og essi e ounds o GTP binding and hyd olysis
a e all significan ly accele a ed o he ΔCT4/6 a ian s, mani es ed in
as e ecycling and g ea e s eady-s a e associa ion wi h memb anes.
Con e sely, o he same easons, in he p esence o a non-na i e CT
ex ension ha non-physiologically s abilizes in e -subuni D p1 in e -
ac ions and exagge a edly p omo es helical sel -assembly in CT+ D p1,
GTP u no e and ecycling on memb anes appea s o be dec eased.
Thus, as e dynamics in he absence o he CT-SLiM, and al e ed,
slowe dynamics in he p esence o a non-na i e CT ex ension dis-
inc i ely a ec D p1 coope a i e GTPase ac i i y ela i e o WT.
Toge he wi h i s impac on D p1 dime s uc u e, hese da a
he e o e aised he in iguing p ospec ha he na i e CT-SLiM
unc ionsasa‘kine ic ime ’o D p1’s GTP hyd olysis a e and coupled
memb ane emodeling ac i i y.
CT-SLiM con ols D p1-ca alyzed memb ane fission in i o
We he e o e add essed whe he he di e en ial GTPase ac i i y, and
al e ed con o ma ional and sel -assembly dynamics o he CT a ian s
ela i e o WT ansla ed o dis inc memb ane emodeling
abc
0
20
40
60
80
100
120
140
02468
10 12
Time (min)
P eleased (PM)
i
WT 21.0 ± 0.5
CT+ 13.0 ± 0.6
'CT4 22.5 ± 0.1
'CT6 25.2 ± 1.1
A e age
(mink
ca
-1
)
CL-s imula ed GTPase ac i i y
0
20
40
60
80
100
120
140
0 102030405060
Basal GTPase ac i i y
'CT4 5.1 ± 0.1
'CT6 5.2 ± 0.1
WT 2.0 ± 0.1
CT+ 1.8 ± 0.1
A e age
(mink
ca
-1
)
P eleased (PM)
i
Time (min)
0
10
20
30
40
50
WT CT+'CT4'CT6
(mink
ca
-1
)
Bu s phase kine ics
WT 2.4 ± 0.1 32.2 ± 0.5
CT+ 1.9 ± 0.0 18.8 ± 0.8
'CT4 7.3 ± 0.1 36.9 ± 0.3
'CT6 7.9 ± 0.1 47.2 ± 1.8
Basal CL-s imula ed
Fig. 3 | CT-SLiM modifica ions di e en ially a ec D p1 GTPase ac i i y. Rela i e
basal (a) and CL-s imula ed GTPase ac i i ies (b) o WT D p1 and CT a ian s a
0.5 µM each in he absence and p esence o CL-con aining liposomes (150µM o al
lipid). GTPase ac i i ies o he CT a ian s we e measu ed in pa allel in compa ison
o WT om wo independen p o ein alone o p o ein-lipid mix u e samples (n=2
measu emen s o each a ian unde ei he condi ion; n= 3 and 5 o WT basal and
CL-s imula ed ac i i ies, espec i ely). The concen a ion o ino ganic phospha e
(P
i
) eleased is plo ed agains ime. Indi idual da a poin s and bes fi aces o
each ime cou se a e shown. In aand b, he a e age u no e numbe (k
ca
) ± SEM
de i ed om a linea eg ession analysis o bo h p e-s eady s a e (bu s phase;
boxed egions) and s eady-s a e kine ics da a poin s is indica ed abo e. ck
ca
o
he bu s -phase p e-s eady-s a e kine ics only om each measu emen . Indi idual
da a poin s a e o e laid on ba plo s ep esen ing he a e age. Bu s -phase
k
ca
± SEM is indica ed abo e.
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 6
pheno ypes. To his end, we es ed he e ficacy o ou CT a ian s in
di ec ing he scission o suspended lipid nano ubes (NTs) mimicking
he mi ochond ial ou e memb ane a p e-cons ic ed mi ochond ial
di ision si es. NTs anging om ens o hund eds o nanome e s in
diame e we e o med be ween polyme mic opilla s in a mic ofluidic
chambe 35 (see Me hods). WT D p1 and CT a ian s a 0.5 µMfinal
concen a ion, co esponding o he es ima ed cy osolic concen a-
ion o D p128, as well as he concen a ion a which he ca aly ic
ac i i y (k
ca
) o WT D p1 nea s sa u a ion26, we e hen in used in o he
chambe in he p esence o 1 mM GTP, while NT cons ic ion and/o
scission was moni o ed in eal- ime by fluo escence mic oscopy.
As p e iously shown17,25, WT D p1 did no e ec i ely ca alyze NT
fission on i s own (Fig. 4a, b). Ins ead, WT sca old assembly on he
NT esul ed in NT cons ic ion o a s able final adius o 14 ± 2 nm
(measu ed a he memb ane midplane) independen o he ini ial NT
adii (Fig. 4a, Supplemen a y Mo ie 1). Su p isingly, bo h ΔCT4 and
ΔCT6 D p1 selec i ely media ed he fission o NTs wi h he highes
ini ial cu a u es (i.e., wi h adii < ~30 nm), wi h he ΔCT4 a ian
exhibi ing he g ea e fission e ficiency o he wo (Fig. 4a, b, Sup-
plemen a y Mo ies 2, 3). Fo bo h a ian s, he a ea o memb ane
cons ic ion p io o fission appea ed o be highly limi ed and na -
ow, being ba ely de ec able by fluo escence mic oscopy (Fig. 4a).
Thus, a pa ial o comple e dele ion o he CT-SLiM limi s
D p1 sca olding on memb anes, p obably due o he g ea e GTP
hyd olysis a e o he ΔCT4/6 a ian s (Fig. 3and Supplemen a y
Fig. 6d, e), causing apid oligome disassembly and ecycling on
memb anes (Supplemen a y Fig. 7).
In s a k con as o he ΔCT4/6 dele ion a ian s, he CT+ ex en-
sion a ian s elici ed a obus cons ic ion and fission o a b oad ange
o ini ial NT cu a u es (Fig. 4a, b, Supplemen a y Fig. 8a, Supple-
men a y Mo ie 4). No ably, he scission e ficiencies o hese a ian s
di ec ly co esponded o hei highe -o de oligome iza ion
p opensi ies wi h CT+ ≈CT + * > CT+sh (Supplemen a y Figs. 8a, 2d, e).
Thus, he ma kedly imp o ed s abili y o he CT+ a ian s on mem-
b anes di ec ly co ela ed wi h hei imp o ed memb ane fission
ac i i ies. Rema kably, fission e ficiency was di ec ly p opo ional o
he p eponde ance o he assembly-p imed, compac dime con-
o me in solu ion in he p esence o GTP, sampled almos exclusi ely
by he CT+ a ian s, bu no WT (Fig. 2a). Impo an ly, con a y o
p e ailing models, memb ane fission ac i i y was in e sely co ela ed
wi h he assembly-s imula ed GTP hyd olysis a e on memb anes, wi h
CT+ a ian s o lowe GTPase ac i i y being mo e e ficien in fission
(Supplemen a y Fig. 8a, b).
To u he assess he impac o he CT-SLiM modifica ions and
imposed s uc u al al e a ions on memb ane emodeling, we used
c yo-EM o analyze he sel -assembly o WT D p1 and CT a ian s on
p e o med memb ane NTs in he cons an p esence o GTP (Fig. 4c). In
ag eemen wi h he eal- ime fluo escence measu emen s, he c yo-EM
da a e ealed ha WT D p1 o med o ganized helical polyme s ha
cons ic ed heNTs o a adius o ~15 nm.Bycon as , CT+D p1 o med
diso ganized, uzzy coa s ha u he cons ic ed he memb anes o
c i ical adii o <7 nm, equen ly esul ing in fission and consequen
e ac ion o he cu NTs o he memb ane ese oi s loca ed on he
EM g id. In e es ingly, ΔCT4 D p1 displayed helical polyme s wi h
highly a iable diame e (Fig. 4c) consis en wi h a nea comple e loss
o CT-SLiM-imposed in e -subuni helical egis e and polyme geo-
me y. No ably, unde hese condi ions, WT D p1 polyme s we e
obse ed on bo h highly cu ed and ela i ely fla memb ane egions,
whe eas he CT+ a ian was cu a u e-selec i e wi h an acu e p e-
e ence o binding he cu ed NTs (Fig. 4c).
Toge he , hese da a indica e ha he CT-SLiM go e ns bo h D p1
polyme geome y and dynamics on memb anes, and ha CT mod-
ifica ions di e en ially a ec memb ane cu a u e selec i i y and fis-
sion ac i i y.
c
ab
Fig. 4 | CT-SLiM modifica ions di e en ially a ec memb ane emodeling and
fission. a Rep esen a i e kymog aphsshowing he emodeling o eely suspended
NTs upon addi ion o 0.5 μM o WT D p1 o CT a ian s in he p esence o 1 mM
GTP. NT memb ane fluo escence is displayed in cyan pseudocolo o cla i y. Le
images co espond o he ini ial ame o he kymog aphs. Righ image sequences
co espond o he amed egion o he kymog aphs. Pu ple a ows indica e NT
fission. bDis ibu ions o he adiio ee-s anding NTs ha unde wen fission (yes)
o we eonlycons ic ed(no) upon addi ion o 0.5μM WT D p1 o CT a ian s in he
p esenceo 1 mM GTP. The numbe s on op o each box ep esen he o al numbe
o NTs o each condi ion. E o ba s a e SD, n= 3 independen expe imen s. Box
plo s indica e median (middle line), 25 h and 75 h pe cen ile (box) and ou lie s
(whiske s). cC yo-EM images showing WT D p1, CT+ D p1, and ΔCT4 D p1
assembled on p e o med NTs in he p esence o 1mM GTP. Whi e a owheads
indica e D p1 sca olds on highly cu ed NT memb anes. Black a owhead shows
cu a u e-adap able assembly o WT D p1 also on ela i ely fla (low cu a u e)
memb ane egions, no obse ed wi h CT+ D p1.
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 7
CT-SLiM in e ac ions wi h GIPC-1 egula e D p1-media ed
memb ane fission
Nex , we de e mined how D p1-GIPC-1 in e ac ions ia he CT-SLiM
a ec D p1 s uc u e, assembly, dynamics, and unc ion.
GIPC-1 con ains an N- e minal IDR in addi ion o a cen ally loca-
ed PDZ domain flanked by wo unique GIPC homology domains (GH1
and GH2)36 (Supplemen a y Fig. 9a). The GH1 and PDZ domains a e
in ol ed in GIPC-1 mul ime iza ion37, whe eas GH2 binds MYO6. In he
absence o he N- e minal IDR and a PBM (PDZ domain ligand), GIPC-1
o ms an au o-inhibi ed, PDZ domain-swapped dime ha occludes
bo h he PBM and MYO6 binding si es36 (Supplemen a y Fig. 9b).
Howe e , consis en wi h a p e ious epo 37, we ound ha in he
unliganded s a e, and a ela i ely low concen a ions in solu ion,
pu ified ull-leng h GIPC-1 exis s in a as , dynamic monome -dime
equilib ium ha la gely a o s monome s (Supplemen a y Fig. 9c, d).
Con e sely, a e y high concen a ions expe ienced du ing pu ifica-
ion, GIPC-1 also o med long filamen ous sedimen able polyme s in
solu ion (Supplemen a y Fig. 10) indica i e o i s capaci y o mul i-
me ize when seques e ed locally. Pu ified GIPC-1 a low bulk con-
cen a ions in solu ion howe e emains soluble (Supplemen a y
Fig. 10) and does no associa e wi h, o emodel, memb anes (Sup-
plemen a y Fig. 10). Toge he , hese da a indica ed ha D p1 CT-SLiM
binding may unc ion o elie e GIPC-1 au o-inhibi ion, elici GIPC-1
mul ime iza ion, and p omo e coope a i e D p1-GIPC-1 co-assembly
on memb anes. Consis en wi h his no ion, mul ime ic GIPC-1 has
p e iously been localized o memb anes37 indica ing a ole o he
GIPC-1 mul ime iza ion in ligand p o ein (e.g. D p1) confinemen a
a ge memb ane si es.
GIPC-1 obus ly inhibi ed he assembly-s imula ed GTPase ac i i y
o WT D p1 on CL-con aining memb anes in a concen a ion-
dependen manne (Fig. 5a, b). ΔCT6 and CT+ D p1, by con as , we e
modes ly inhibi ed, indica ing weakened binding. The modes , bu
conside able, inhibi ion o hese a ian s also indica ed he p esence
o addi ional GIPC-1 in e ac ion si es besides he CT-SLiM (Fig. 5a, b).
GIPC-1 also po en ly inhibi ed he GMP-PCP-induced sel -
assembly o WT D p1 in o ings and spi als in solu ion in NS-EM
expe imen s, indica ing ha GIPC-1 binding hinde s he helical p o-
paga ion o D p1 (Fig. 5c). As expec ed, GIPC-1 did no conside ably
a ec he GMP-PCP-induced helical sel -assembly o he CT+ a ian
(Fig. 5c). Ne e heless, GIPC-1 s ill educed he GMP-PCP-induced o -
ma ion o iangula nubs by he ΔCT6 a ian p obably owing o he
p esence o addi ional binding si es (Fig. 5c).
Likewise, GIPC-1 po en ly inhibi ed he WT D p1-media ed ubu-
la ion o CL-con aining liposomes (Supplemen a y Fig. 10b). ΔCT6
ab
10 nm WT (apo) + GIPC-1 (1:4)
0%
20%
40%
60%
80%
100%
Open
compac
Closed
compac
ed
6 Pm
WT
WT + GIPC-1
+GMP-PCP
+GMP-PCP
+GIPC-1 (1:4)
WT 'CT6 CT+
c
WT
WT + GIPC-1
NT ission p obabili y, %
** g
4 Pm
30 s
WT + GIPC-1 + GTP
1
10
(mink
ca
-1
)
D p1:GIPC-1 mola a io
1:0 1:1 1:2 1:4
'CT6
WT
CL-s imula ed GTPase ac i i y
CT+
D p1:GIPC-1 mola a io
% o maximum
CL-s imula ed GTPase ac i i y
1
10
100
1:0 1:1 1:2 1:4
'CT6
WT
CT+
Fig. 5 | CT-SLiM in e ac ions wi h GIPC-1 po en ia e memb ane fission in i o.
aCL-s imula ed GTPase ac i i ies o WT and ΔCT6 D p1 wi h inc easing con-
cen a ions o GIPC-1. k
ca
om wo independen measu emen s a e plo ed e sus
D p1:GIPC-1 mola a io. bDa a in a eplo ed as % o maximum ac i i y o each
a ian . cRep esen a i e NS-EMimages o WT D p1 and CT a ian s incuba ed wi h
GMP-PCP in he absence and p esence o a 1:4 mola a io o GIPC-1. Scale ba ,
100 nm. dNS-EM 2D class a e ages o WT D p1 in he apo s a e in he p esence o a
1:4 mola a io o GIPC-1. Only he compac con o me s (open and closed) we e
obse ed o WT-D p1 in he p esence o GIPC-1. eKymog aph showing NT con-
s ic ion and fission (a owhead) byWT D p1 in he p esence o GIPC-1 a a1:1 mola
a io (0.5 μM each) in he p esence o 1 mM GTP. NT memb ane fluo escence is
displayed in cyan pseudocolo o cla i y. Pe cen age o NTs ha unde wen
fission upon addi ion o ei he 0.5 μM WT D p1 alone o a 1:1 mix u e o WT
D p1:GIPC-1 a 0.5 μM each in he p esence o 1mM GTP. Each poin ep esen s a
eplica e. The numbe on op o each column ep esen s he o al numbe o NTs
o each condi ion. Mean ± SD a e shown. ** S a is ically di e en a he 0.01 le el
(unpai ed wo sample - es , equal a iance assumed). gImages showing he igid
polyme iza ion o WT D p1alone e sus he o ma ion o much sho e sca olds in
he equimola p esence o WT D p1 and GIPC-1 on NTs. Images shown we e
acqui ed app oxima ely 2minu es a e he addi ion o he p o eins a 2 μMfinal
concen a ion each in he p esence o 1mM GTP. RhPE channel is shown. Pseu-
docolo is used o cla i y.
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 8
D p1 memb ane emodeling ac i i y, con e sely, was no significan ly
a ec ed (Supplemen a y Fig. 10c). In e es ingly, a equimola con-
cen a ions (1:1) unde hese condi ions, WT D p1 and GIPC-1 o med
amo phous assemblies in solu ion, whe eas a highe GIPC-1 a ios
(1:4), linea and bundled filamen s o assembled p o ein in solu ion,
eminiscen o D p1 co-assembly wi h he adap o mi ochond ial
dynamics p o ein o 49 kDa o MiD4938 we e e iden (Supplemen a y
Fig. 10b). Toge he , hese da a indica ed ha GIPC-1 in e ac ions ia
he CT-SLiM al e s D p1 sel -assembly geome y, wi h p onounced
e ec s on memb ane emodeling as de e mined by he lack o o de ed
helical sel -assembly and esul an memb ane ubula ion. A simila
inhibi ion o CL-s imula ed GTPase ac i i y and helical sel -assembly
was obse ed o WT D p1 in he p esence o GIPC-1 on GalCe CL-NTs
indica ing ha GIPC-1 egula ion o D p1 ac i i y does no a y wi h
memb ane cu a u e (Supplemen a y Fig. 11a, b).
Su p isingly, NS-EM2Dclassifica iono apoWT D p1dime s in he
p esence o GIPC-1 e ealed he p esence o he assembly-p imed,
compac D p1 con o me in solu ion (Fig. 5c), in con as o he au o-
inhibi ed, ex ended D p1 con o me p edominan ly ound in GIPC-1’s
absence (Fig. 2a). Addi ional densi y ep esen ing GIPC-1, howe e ,
was no eadily e iden eflec ing ei he a dynamic in e ac ion o GIPC-
1 wi h WT D p1 in he apo s a e o a subs an ial o e lap o GIPC-1
densi y wi h he closely spaced G domains o he compac WT D p1
con o me . In he case o ΔCT6 and CT+ D p1, howe e , a ious ex a
densi ies and al e ed subuni a angemen s we e obse ed a es ing o
he p esence o addi ional GIPC-1 binding si es and al e na e D p1-
GIPC-1 in e ac ions (Supplemen a y Fig. 11c, d).
The educed assembly-s imula ed GTPase ac i i y obse ed o
D p1 in he p esence o GIPC-1 seemingly po en ia es he memb ane
emodeling e en s leading o fission, as WT D p1 in he p esence o
GIPC-1 ca alyzed fission in >30% o ee-s anding NTs o e a wide ange
o memb ane cu a u es (Fig. 5e, , Supplemen a y Mo ie 5). In e -
es ingly, a 0.5 μM D p1 concen a ion and a 1:1 GIPC-1:D p1 a io, we
de ec ed he o ma iono sho D p1 sca olds on he NTs immedia ely
p io o fission. Besides, hese sca olds we e highly mobile on he NT
su ace (as obse ed in he kymog aph in Fig. 5e, Supplemen a y
Mo ie 5), sugges ing ha in he p esence o GIPC-1, D p1 ini ially
assembles in o small p e-cu ed uni s, and no in o comple e ings, on
he NT su ace. Thus, GIPC-1 associa ion wi h WT D p1 ia he CT-SLiM
appea s o disengage D p1 in e -subuni in e ac ions ha p omo e
highe -o de D p1 sel -assembly.
This di e en ial beha io o WT D p1 in he p esence o GIPC-1
was be e e idenced a highe p o ein concen a ions (2 µM, Fig. 5g).
Whe eas WT D p1 alone apidly polyme ized in o long and igid sca -
olds, ende ing kinks in he NTs ha p ecluded memb ane fission, in
GIPC´s p esence he g ow h o he WT D p1 sca olds on he NTs was
compa able o ha o CT+ D p1 (compa e Figs. 5gand4a), and
esul ed in NT fission a a simila fission a e o ha de ec ed a 0.5µM
p o ein concen a ion (Fig. 5 ). Thus, WT-D p1 in he p esence o GIPC-
1 pa ially mimics CT+ D p1, which exhibi s educed GTPase ac i i y
and al e ed CT-SLiM in e ac ions.
CT-SLiM egula ion o D p1 is c i ical o mi ochond ial and
pe oxisomal fission in i o
D p1 ca alyzes bo h mi ochond ial and pe oxisomal fission12,39–41.Ye ,
GIPC-1 colocalizes wi h he mi ochond ia bu no conside ably wi h
pe oxisomes18. To de e mine whe he he CT-SLiM is he e o e di -
e en ially equi ed o D p1-ca alyzed mi ochond ial and pe oxisomal
fission in i o, we examined and compa ed mi ochond ial and pe -
oxisomal mo phology in D p1 KO mouse emb yonic fib oblas s (MEFs)
o e exp essing N- e minally Myc- agged WT and ΔCT4/6 a ian s, and
he C- e minally Myc/FLAG- agged CT+ a ian (Fig. 6a, Supplemen-
a y Figs. 12–14). Emp y ec o - ans ec ed D p1 KO MEFs displayed
ex ensi ely hype used mi ochond ia o highly elonga ed pe oxi-
somes in he absence o D p1-ca alyzed mi ochond ial and
pe oxisomal fission (Fig. 6a, Supplemen a y Figs. 12–14). As expec ed,
exogenous Myc-WT D p1 o e exp ession e ec i ely escued and
es o ed bo h mi ochond ial and pe oxisomal fission leading o he
o ma ion o highly agmen ed mi ochond ia and dis inc ly punc i-
o m (sphe ical) pe oxisomes (Fig. 6a–d, Supplemen a y Figs. 12–14).
By con as , howe e , he o e exp ession o he ΔCT4 and ΔCT6 a -
ian s had no palpable e ec on he ini ial mo phology o ei he o ga-
nelle (Fig. 6a–d), wi h he g ea majo i y o cells displaying a
p onounced pe inuclea clus e ing o hype used mi ochond ia and
e aining highly elonga ed pe oxisomes (Fig. 6a–d, Supplemen a y
Figs. 12–14). Mo e su p isingly, he CT+ a ian con aining he na i e
CT-SLiM sequence was also significan ly impai ed in he fission o bo h
o ganelles, albei o a lesse deg ee han he ΔCT4/6 a ian s
(Fig. 6a–d, Supplemen a y Figs. 12–S14). Thus, in spi e o hei appa -
en gain-o - unc ion in e ec ing he fission o model memb anes
in i o, he CT a ian s appea ed ne e heless pe u bed in e ec ing
o ganella fission in i o. These da a u he ei e a ed ha he al e ed
sel -assembly p ope ies and dynamics o he CT a ian s as mani-
es ed in i o, and consequen impai men s in e ec o (e.g. GIPC-1)
in e ac ions downs eam a e likely esponsible o hei o ganella
fission de ec s in i o. Consis en wi h his in e p e a ion, when
o e exp essed in D p1 KO MEFs, he CT+ a ian , which o ms sup a-
molecula assemblies in i o (Fig. 1b), cons i u ed la ge g anula
punc a in he cy osol indica i e o agg ega ion. In con as , he WT and
ΔCT a ian s exhibi ed a mo e di use and homogeneous dis ibu ion
(Supplemen a y Fig. 15a, b). In co-immunop ecipi a ion expe imen s,
nei he o e exp essed WT D p1 no he CT a ian s co-p ecipi a ed
wi hendogenous GIPC-1(Supplemen a yFig. 15c, d)indica ing a highly
dynamic in e ac ion ha could no di e en ia e he CT a ian s om
WT D p1 in GIPC-1 binding. These da a u he sugges ed ha pe u -
ba ions in GIPC-1 in e ac ions expec ed o he CT a ian s a e likely
seconda y o hei p ima y de ec s/al e a ions in sel -assembly and
con o ma ional dynamics.
F om hese collec i e da a, we conclude ha he na i e CT-SLiM is
a c i ical s uc u al and unc ional de e minan o D p1-ca alyzed
mi ochond ial and pe oxisomal fission, and ha i s pe u ba ions
influence D p1 unc ion bo h in i o and in i o.
Discussion
S uc u al and unc ional plas ici y a e wo in e linked cha ac e is ics
o IDRs10,42,43.Thisisbes exemplified by he longes and bes -
ecognized IDR in D p1, he VD, which is in ol ed in mul iple
p o ein-p o ein and p o ein-lipid in e ac ions ia a ious iden ified
MoRFs and SLiMs14. Rema kably, he VD is au o-inhibi o y o p e-
ma u e D p1 sel -assembly in solu ion14,44,45, while con e sely p omo -
ing D p1 sel -assembly and unc ion upon pa ne in e ac ions,
specifically wi h a ge lipids on mi ochond ial memb anes14,17,26,46,
hus eflec ing he VD’s duali y and unc ional di e si y. Howe e , he
VD and a ious o he IDRs in D p1 (e.g. he 80-loop) a e ela i ely
poo ly conse ed (Supplemen a y Fig. 1a) and a e subjec o ex ensi e
issue- and o ganism-specifical e na i esplicing
47, indica ing ha
some o hei asc ibed unc ions may no be en i ely uni e sal. He e,
we demons a e ha he highly conse ed CT-SLiM, p e iously impli-
ca edinD p1 anspo 18,isye ano he c i ical,mul i unc ional‘ oggle’
ha no only go e ns D p1 con o ma ional s abili y and dynamics, bu
also di ec s D p1 sel -assembly, assembly geome y, and coope a i e
GTP hyd olysis o acili a e pa ne p o ein-guided memb ane con-
s ic ion and fission.
Ou findings ha e majo implica ions o he unde s anding o
D p1 unc ion and egula ion. We show ha he CT-SLiM is c i ical o
o de ed D p1 sel -assembly, as oligome iza ion o he CT a ian s
p opaga es ou -o - egis e and e en ually becomes diso de ed. The
CT-SLiM also di ec ly impac s D p1 dime con o ma ional dynamics in
solu ion, wi h he CT a ian s mo e eadily sampling he ‘assembly-
p imed’compac con o ma ion in con as o he CT-SLiM-imposed,
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 9
60. Ba ylko, B. e al. The p oline/a ginine- ich domain is a majo de e -
minan o dynamin sel -ac i a ion. Biochemis y 49,
10592–10594 (2010).
61. Fe guson,S.M.&DeCamilli,P.Dynamin,amemb ane- emodelling
GTPase. Na . Re . Mol. Cell Biol. 13,75–88 (2012).
62. Zhang, R. e al. Dynamin egula es he dynamics and mechanical
s eng h o he ac in cy oskele on as a mul ifilamen ac in-bundling
p o ein. Na . Cell Biol. 22,674–688 (2020).
63. Ma ila,J.P.e al.Ahemi-fission in e media e links wo mechan-
is ically dis inc s ages o memb ane fission. Na u e 524,
109–113 (2015).
64. Bos, J. L., Rehmann, H. & Wi ingho e , A. GEFs and GAPs: c i ical
elemen s in he con ol o small G p o eins. Cell 129,
865–877 (2007).
65. Gaspe , R., Meye , S., Go ha d , K., Si ajuddin, M. & Wi ingho e , A.
I akes wo o ango: egula ion o G p o eins by dime iza ion. Na .
Re . Mol. Cell Biol. 10,423–429 (2009).
66. Ramachand an, R. & Schmid, S. L. The dynamin supe amily. Cu .
Biol. 28,R411–R416 (2018).
67. C osby, D. e al. Recons i u ion o human a las in usion ac i i y
e eals au oinhibi ion by he C e minus. JCellBiol221,
e202107070 (2022).
68. K ishna,S.&Fo d,M.G.J.Thea las inpa alogs:Thecomplexi yin
he ails. J. Cell Biol. 222, e202305116 (2023).
69. Osellame, L. D. e al. Coope a i e and independen oles o he D p1
adap o s M , MiD49 and MiD51 in mi ochond ial fission. J. Cell Sci.
129,2170–2181 (2016).
70. Kalia, R. & F os , A. Open and cu : allos e ic mo ion and memb ane
fission by dynamin supe amily p o eins. Mol. Biol. Cell 30,
2097–2104 (2019).
71. S epanyan s, N., Macdonald, P. J., Madan Mohan, P. & Ramachan-
d an, R. A Single Common P o ocol o he Exp ession and Pu -
ifica ion o Soluble Mammalian DSPs om Esche ichia coli.
Me hods Mol. Biol. 2159,31–40 (2020).
72. Liu, R. & Chan, D. C. The mi ochond ial fission ecep o M selec-
i ely ec ui s oligome ized D p1. Mol. Biol. Cell 26,
4466–4477 (2015).
73. Clin on, R. W., Baue , B. L. & Mea s, J. A. A fini y Pu ifica ion and
Func ional Cha ac e iza ion o Dynamin-Rela ed P o ein 1. Me hods
Mol. Biol. 2159,41–53 (2020).
74. Wakabayashi, J. e al. The dynamin- ela ed GTPase D p1 is equi ed
o emb yonic and b ain de elopmen in mice. J. Cell Biol. 186,
805–816 (2009).
75. Leona d,M.,Song,B.D.,Ramachand an,R.&Schmid,S.L.Robus
colo ime ic assays o dynamin’s basal and s imula ed GTPase
ac i i ies. Me hods Enzymol. 404, 490–503 (2005).
76. Punjani, A., Rubins ein, J. L., Flee , D. J. & B ubake , M. A. c yoS-
PARC: algo i hms o apid unsupe ised c yo-EM s uc u e de e -
mina ion. Na . Me hods 14,290–296 (2017).
77. Mon ecinos-F anjola, F., Baue , B. L., Mea s, J. A. & Ramachand an,
R. GFP fluo escence agging al e s dynamin- ela ed p o ein 1 oli-
gome iza ion dynamics and c ea es disassembly- e ac o y punc a
o media e mi ochond ial fission. Sci. Rep. 10,14777(2020).
78. Malaby, A. W. e al. Me hods o analysis o size-exclusion
ch oma og aphy-small-angle X- ay sca e ing and econs uc ion
o p o ein sca e ing. J. Appl. C ys allog . 48, 1102–1113 (2015).
79. Nielsen,S.S.e al.BioXTASRAW,aso wa ep og am o high-
h oughpu au oma ed small-angle X- ay sca e ing da a educ ion
and p elimina y analysis. J. Appl. C ys allog . 42,959–964 (2009).
80. Manalas as-Can os, K. e al. ATSAS 3.0: expanded unc ionali y and
new ools o small-angle sca e ing da a analysis. J. Appl. C ys al-
log . 54,343–355 (2021).
81. Hopkins,J.B.,Gillilan,R.E.&Skou,S.BioXTASRAW:imp o emen s
o a ee open-sou ce p og am o small-angle X- ay sca e ing da a
educ ion and analysis. J. Appl. C ys allog . 50,1545–1553 (2017).
82. Pe e sen,E.F.e al.UCSFChime a–a isualiza ion sys em o
explo a o y esea ch and analysis. J. Compu Chem. 25,
1605–1612 (2004).
83. Godda d, T. D., Huang, C. C. & Fe in, T. E. Visualizing densi y maps
wi h UCSF Chime a. J. S uc . Biol. 157,281–287 (2007).
84. Jumpe , J. & Hassabis, D. P o ein s uc u e p edic ions o a omic
accu acy wi h AlphaFold. Na . Me hods 19,11–12 (2022).
85. K issinel, E. Enhanced old ecogni ion using e ficien sho ag-
men clus e ing. J. Mol. Biochem. 1,76–85 (2012).
86. Agi e, J. e al. The CCP4 sui e: in eg a i e so wa e o mac o-
molecula c ys allog aphy. Ac a C ys allog . D. S uc . Biol. 79,
449–461 (2023).
87. Schindelin, J. e al. Fiji: an open-sou ce pla o m o biological-
image analysis. Na . Me hods 9,676–682 (2012).
88. Da ,S.,Kame ka ,S.C.&Pucadyil,T.J.Useo hesuppo ed
memb ane ube assay sys em o eal- ime analysis o memb ane
fission eac ions. Na . P o oc. 12,390–400 (2017).
89. A ganda-Ca e as, I., Fe nandez-Gonzalez, R., Munoz-Ba u ia, A. &
O iz-De-Solo zano, C. 3D econs uc ion o his ological sec ions:
Applica ion o mamma y gland issue. Mic osc Res. Tech. 73,
1019–1029 (2010).
Acknowledgemen s
We hank Ashu osh P ince and Shane Wybo ny (bo h o CWRU) o
echnical assis ance in p o ein p oduc ion. NIH R01 g an s GM121583
and GM125844 suppo ed wo k in he R.R. and J.A.M. labo a o ies,
espec i ely. Wo k in he A.V.S. labo a o y was suppo ed by he
PGC2018-099971-B-I00 and PID2021-127844NB-I00 g an s unded by
MCIN/AEI/10.13039/501100011033/ and by “ERDF A way o making
Eu ope”and by he Basque Go e nmen G an IT1625-22. Jon Agi e is
a Royal Socie y Uni e si y Resea ch Fellow (awa d codes UF160039
and URF R 221006). I.P.J. acknowledges he p edoc o al ellowship
om he Uni e si y o he Basque Coun y. We a e g a e ul o com-
pu a ional suppo om he Uni e si y o Yo k High Pe o mance
Compu ing se ice, Viking and he Resea ch Compu ing eam, no ably
Jaspe G imm and Emma Ba nes. The au ho s a e g a e ul o he
Elec on Mic oscopy and C ys allog aphy pla o m o he CIC bioGUNE
and he Basque Resou ce o Elec on Mic oscopy (BREM) o he Bio-
fisika Ins i u e o p o iding access o c yo EM sample p epa a ion and
analysis equipmen . Molecula g aphics images we e p oduced using
he UCSF Chime a package om he Resou ce o Biocompu ing,
Visualiza ion, and In o ma ics a he Uni e si y o Cali o nia, San
F ancisco (suppo ed by NIH P41 RR-01081). This esea ch used
esou ces o he Ad anced Pho on Sou ce, a U.S. Depa men o
Ene gy (DOE) O fice o Science Use Facili y ope a ed o he DOE
O fice o Science by A gonne Na ional Labo a o y unde Con ac No.
DE-AC02-06CH11357. BioCAT was suppo ed by g an P30 GM138395
om he Na ional Ins i u e o Gene al Medical Sciences o he Na ional
Ins i u es o Heal h.
Au ho con ibu ions
Concep ualiza ion: A.V.S, and R.R.; Expe imen al me hodology: A.V.S,
R.R., J.A.M.; Expe imen al in es iga ion and analysis: all au ho s; W i ing
—o iginal d a : A.V.S, and R.R; W i ing— e iew and edi ing: all au ho s;
Funding acquisi ion: A.V.S and R.R.; Resou ces: J.A.M., A.V.S. and R.R.;
P ojec supe ision: A.V.S. and R.R.; Nega i e-s ain EM da a acquisi ion
and D p1 s uc u e econs uc ions: K.R. and J.A.M.; Cell biology
expe imen s and pe oxisomal ci cula i y analysis: D.H. and X.Q.; SEC-
SAXS expe imen s and da a analyses: M.M.; P oduc ion o DNA con-
s uc s: P.M. and R.R.; P o ein pu ifica ion and GTPase assays: P.M.,
J.O.G., and R.R.; Alpha old p edic ions and s uc u al o e lays: J.A.; SEC-
MALS expe imen s: R.R.; Analysis o mi ochond ial agmen a ion: R.R.
and A.V.S; P elimina y da a on D p1-media ed NT fission: J.M.M.G.; NT
fission assays and analyses: I.P.J. and A.V.S.; c yoEM analyses o p e-
o med NT cons ic ion: I.S.P. and A.V.S.
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 16
Compe ing in e es s
The au ho s decla e no compe ing in e es s
Addi ional in o ma ion
Supplemen a y in o ma ion The online e sion con ains
supplemen a y ma e ial a ailable a
h ps://doi.o g/10.1038/s41467-023-44413-6.
Co espondence and eques s o ma e ials should be add essed o
Anna V. Shny o a o Rajesh Ramachand an.
Pee e iew in o ma ion Na u e Communica ions hanks he anon-
ymous e iewe s o hei con ibu ion o he pee e iew o his wo k. A
pee e iew file is a ailable.
Rep in s and pe missions in o ma ion is a ailable a
h p://www.na u e.com/ ep in s
Publishe ’s no e Sp inge Na u e emains neu al wi h ega d o ju -
isdic ional claims in published maps and ins i u ional a filia ions.
Open Access This a icle is licensed unde a C ea i e Commons
A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing,
adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as
long as you gi e app op ia e c edi o he o iginal au ho (s) and he
sou ce, p o ide a link o he C ea i e Commons license, and indica e i
changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons license, unless
indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no
included in he a icle’s C ea i e Commons license and you in ended
use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed
use, you will need o ob ain pe mission di ec ly om he copy igh
holde . To iew a copy o his license, isi h p://c ea i ecommons.o g/
licenses/by/4.0/.
© The Au ho (s) 2024
A icle h ps://doi.o g/10.1038/s41467-023-44413-6
Na u e Communica ions | (2024) 15:52 17
