Un eiling he ac i a ion dynamics o a old-swi ch bac e ial
glycosyl ans e ase by
19
F NMR
Recei ed o publica ion, May 1, 2020, and in e ised o m, May 19, 2020 Published, Pape s in P ess, May 20, 2020, DOI 10.1074/jbc.RA120.014162
Jobs Liebau
1,‡
,Mon se Te sa
2,‡
,Bea iz T as oy
2
,Joan Pa ick
1
,Ane Rod igo-Unzue a
3
,F ancisco Co zana
4
,
Tobias Spa man
5
,Ma celo E. Gue in
2,3,6,
*, and Lena Mäle
1,5,
*
F om he
1
Depa men o Biochemis y and Biophysics, S ockholm Uni e si y, S ockholm, Sweden,
2
S uc u al Biology Uni ,
Cen e o Coope a i e Resea ch in Biosciences (CIC bioGUNE), Basque Resea ch and Technology Alliance (BRTA), Bizkaia
Technology Pa k, De io, Spain,
3
Depa amen o de Bioquímica and Ins i u o Bio isika, Consejo Supe io de In es igaciones
Cien í icas–Uni e sidad del País Vasco/Euskal He iko Unibe si a ea (CSIC, UPV/EHU), Bizkaia, Spain,
4
Depa amen o de
Química, Cen o de In es igación en Sín esis Química, Uni e sidad de La Rioja, Log oño, Spain,
5
Depa men o Chemis y,
Umeå Uni e si y, Umeå, Sweden, and
6
IKERBASQUE, Basque Founda ion o Science, Bilbao, Spain
Edi ed by Wol gang Pe i
Fold-swi ch pa hways emodel he seconda y s uc u e
opology o p o eins in esponse o he cellula en i onmen . I
is a majo challenge o unde s and he dynamics o hese olding
p ocesses. He e, we conduc ed an in-dep h analysis o he
a
-he-
lix– o–
b
-s and and
b
-s and– o–
a
-helix ansi ions and do-
main mo ions displayed by he essen ial mannosyl ans e ase
PimA om mycobac e ia. Using
19
FNMR,weiden i ied ou
unc ionally ele an s a es o PimA ha coexis in dynamic
equilib ia on millisecond- o-second imescales in solu ion. We
disco e ed ha old-swi ching is a slow p ocess, on he o de o
seconds, whe eas domain mo ions occu simul aneously bu a e
subs an ially as e , on he o de o milliseconds. S ikingly, he
addi ion o subs a e accele a ed he old-swi ching dynamics o
PimA. We p opose a model in which he old-swi ching dynam-
ics cons i u e a mechanism o PimA ac i a ion.
P o eins do no jus occupy a single s a e bu a e in dynamic
exchange be ween a se o con o ma ions, and some o hese
s a es may a imes be popula ed a e y low le els, so-called in-
isible o da k s a es (1–4). Mo ions occu on a b oad ange o
imescales, and g owing e idence sugges s ha dynamics a e
ins umen al o he unc ion o a mul i ude o p o eins. In
enzyma ically media ed eac ions, p o ein dynamics enable he
ec ui men and binding o subs a es, s abiliza ion o ansi-
ion s a es, and adap a ion o a dynamic memb ane en i on-
men o binding pa ne s (5,6). Fold-swi ching, i.e. he in e -
con e sion o seconda y s uc u e elemen s, is a unc ionally
essen ial dynamic e en o a ce ain class o p o eins. Such
mo ions modi y he opology o a p o ein, including
a
-helix–
o–
b
-s and and/o
b
-s and– o–
a
-helix ansi ions (7–9).
Al e a ions o such dynamic p ocesses a e essen ial o and
causa i e o human disease, as in he case o he p ion p o ein
P P and a a ie y o p o eins associa ed wi h neu odegene a i e
diseases, including
a
-synuclein, amyloid
b
-pep ide, and hun-
ing in (10,11). Al hough i has been p oposed ha such majo
s uc u al changes a e widesp ead in p o eins, s uc u es o
old-swi ching p o eins a e unde ep esen ed in he PDB (12),
and s udies o associa ed dynamics a e sca ce, emaining a
majo challenge (13–15).
He e, we epo on he dynamics o he phospha idyl-myo-
inosi ol mannosyl ans e ase PimA, an essen ial enzyme in
Mycobac e ium ube culosis (16,17). PimA ca alyzes he i s
s ep in he biosyn hesis o phospha idyl-myo-inosi ol manno-
sides (PIMs) by adding a mannose esidue dona ed by he nu-
cleo ide suga GDP-mannose (GDP-Man) o a phospha idyl-
myo-inosi ol lipid, which is ancho ed in o he cy oplasmic
phase o he plasma memb ane (16,18). PIMs a e c i ical s uc-
u al componen s o he mycobac e ial cell en elope and p e-
cu so s o wo cell en elope lipoglycans, lipomannan and
lipoa abinomannan (19,20), bo h o which a e i ulence ac o s
du ing ube culosis (21), one o he deadlies in ec ions wo ld-
wide (22). C ys al s uc u es we e p e iously ob ained o PimA
in he unliganded s a e (23) and in he p esence o GDP o
GDP-Man (Fig. 1A)(24). In he c ys al s uc u e o unliganded
PimA, wo dis inc con o ma ions o a egion in he N- e minal
domain, e med he eshu ling egion ( esidues 118–163), a e
seen, sugges ing ha his egion can unde go mo ions be ween
a compac and an ex ended con o ma ion, e en in he absence
o subs a e (Fig. 1A)(23,25). Mo eo e , he c ys al s uc u es
o PimA in he p esence o GDP o GDP-Man suppo he hy-
po hesis ha he eshu ling egion displays old-swi ching o
seconda y s uc u e elemen s wi h bo h
a
-helix– o–
b
-s and
and
b
-s and– o–
a
-helix ansi ions upon addi ion o sub-
s a e (Fig. 1A)(23). Fold-swi ching ac i a es PimA, since PimA
mu an s locked in he unliganded con o ma ion a e inac i e,
while mu an s locked in he subs a e-bound con o ma ion a e
ac i e (23). Finally, he occu ence o an open- o-closed mo ion
be ween he N- and C- e minal Rossmann old domains has
been p edic ed and expe imen ally demons a ed o occu in
PimA and o he GT-B glycosyl ans e ases (24–28).
He e, we demons a e by using
19
FNMRme hods ha
mo ions o PimA occu on a second- o-millisecond imescale
and include in e con e sions be ween se e al con o ma-
ional s a es. Ou obse a ions p o ide a basis o sugges an
ac i a ion mechanism o PimA in which he addi ion o
This a icle con ains suppo ing in o ma ion.
Au ho 's Choice—Final e sion open access unde he e ms o he C ea i e
Commons CC-BY license.
‡
These au ho s con ibu ed equally o his wo k.
*Fo co espondence: Ma celo E. Gue in, [email p o ec ed]; Lena
Mäle , [email p o ec ed].
9868 J. Biol. Chem. (2020) 295(29) 9868–9878
© 2020 Liebau e al. Published by The Ame ican Socie y o Biochemis y and Molecula Biology, Inc.
ARTICLE
Au ho 's Choice
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
subs a e accele a es dynamics be ween ac i e and inac i e
con o ma ions.
Resul s
19
F labeling is a sensi i e app oach o s udy dynamics o
PimA by NMR
19
F nuclei a e NMR ac i e wi h a gy omagne ic a io ha is
0.94 imes ha o p o ons, ha e 100% na u al abundance, and
display high sensi i i y o hei magne ic en i onmen (29–32).
A he same ime, luo ine is nea ly absen om biological mole-
cules, educing he numbe o esonances obse ed in a spec um
o hose ha s em om a i icially in oduced
19
F labels. The e-
o e,
19
F nuclei a e sui able p obes ha allow employing NMR
expe imen s o dynamic s udies e en o la ge p o eins o which
mo e common labeling app oaches a e no success ul (31). Based
on analyses o he di e en c ys allog aphic s uc u es o PimA
(Fig. 1A), we in oduced a Cys mu a ion in posi ion A g144 (he e
e med PimA
R144C
)(Fig. 1A) and labeled i wi h 1,1,1- i luo oa-
ce one (TFA), gi ing ise o a
19
F labeled mu an PimA
R144C-TFA
(see Expe imen al p ocedu es). The in oduced
19
Flabelo he
mu an PimA
R144C-TFA
is loca ed in di e en chemical en i on-
men s in he di e en c ys allog aphic con o ma ions o PimA, as
highligh ed in Fig. 1A. In he ac i e s a e, he esidue is bu ied
inside he p o ein and, hus, no exposed o sol en , while i is sol-
en -exposed in he inac i e s a es.
19
F nuclei ha e been shown o
be sensi i e o such di e ences in he hyd ophobici y o hei im-
media e en i onmen (32,33). Thus, i can be expec ed ha he
19
F p obe loca ed a Cys144 can sense mo ions ha we
Figu e 1. C ys al s uc u es o PimA. (A) In i s inac i e s a e, PimA adop s wo con o ma ions (PDB en y 4NC9, apo s a e o PimA) wi h di e en e ia y s uc u es
in he eshu ling egion (displayed in ainbow colo ) bu iden ical opology. In he inac i e, ex ended o m (o ange), wo helices (
a
4and
a
5) poin away om he
N- e minal domain and old back along hinge loops on o he N- e minal domain in he inac i e, compac con o ma ion (yellow). No e ha he loop connec ing
a
4
and
a
5 is only pa ially esol ed in he c ys al s uc u e o he inac i e, ex ended s a e. T ansi ion o he ac i e s a e (PDB en y 2GEJ, PimA in he p esence o GDP-
Man) is accompanied by a old-swi ch in he eshu ling egion ( ed). In he c ys al s uc u es, R144 is shown/modeled as sphe es in g een. Selec ed seconda y s uc-
u e elemen s o PimA, including he eshu ling egion ( esidues 118–163), as obse ed in he unliganded (inac i e)compac andex ended con o ma ions and
he liganded (ac i e) PimA-GDP-Man complex, a e shown.
b
5,
b
6, and
b
7a eshowninblue,ligh blue,ando ange, espec i ely. (B)
19
Fspec ao PimA
T126C-V359C-
R144C-TFA
in 10% D
2
O( op), PimA
R144C-TFA
in 100% D
2
O(cen e ), and PimA
R144C-TFA
in 10% D
2
O(bo om). (C)
19
Fspec ao PimA
R144C-TFA
in he absence (blue)anda
inc easing concen a ions o GDP-Man. In ensi ies a e co ec ed o dilu ion e ec s.
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
J. Biol. Chem. (2020) 295(29) 9868–9878 9869
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
hypo hesize o occu in he eshu ling egion, which a e 1)
ex ended o compac domain mo ion and 2) old-swi ching dy-
namics be ween he inac i e and ac i e s a es.
Gi en ha he p obe expe iences su icien ly dis inc chemi-
cal en i onmen s, domain mo ions ha a e ypically on he
o de o milliseconds can be p obed by NMR elaxa ion dispe -
sion expe imen s (34). In addi ion, old-swi ching dynamics a e
expec ed o occu on a second imescale and can be p obed by
sa u a ion ans e expe imen s (35).
We i s de e mined ha he in oduced label does no al e
he h ee-dimensional s uc u e o PimA. Suppo ing his
no ion, wild- ype PimA, PimA
R144C
, and labeled PimA
R144C-TFA
mu an s displayed simila a -UV CD spec a (Fig. S1A)(25). In
addi ion, he use o he mal un olding expe imen s ollowed by
CD has been ex ensi ely used o measu e p o ein-ligand bind-
ing in e ac ion (36–39). The midpoin empe a u e o un olding
o he unliganded p o ein is e e ed o as T
m
.I isexpec ed
ha he addi ion o ligands o he p o ein sample will inc ease
he T
m
. The highe he a ini y o a ligand, he highe he
appa en T
m
alue, because i equi es mo e ene gy o disso-
cia e highe a ini y compounds (39). We ha e p e iously
desc ibed ha he addi ion o GDP o GDP-Man o wild- ype
PimA inc eases he T
m
o he enzyme, wi h he
b
-phospha e
o he nucleo ide g oup playing a p ominen ole (25). As
depic ed in Fig. S1B, he mal un olding ollowed by he a -
UV CD signal a 222 nm indica ed sligh di e ences in p o-
ein s abili y be ween unliganded WT PimA, PimA
R144C
,and
PimA
R144C-TFA
. In e es ingly, a e he addi ion o GDP, wild-
ype PimA, PimA
R144C
,andPimA
R144C-TFA
displayed a clea
inc ease in he T
m
alues (Table S1), suppo ing he no ion
ha he mu a ion and labeling do no signi ican ly pe u b
he enzyma ic s uc u e and unc ion (Fig. S1).
Ac i e and inac i e s a es o PimA show dis inc esonances
PimA con ains wo Cys esidues bu ied inside he p o ein co e
ha a e no expec ed o be labeled du ing he labeling p ocedu e
(see Expe imen al p ocedu es). To e i y his, he labeling scheme
was applied o WT PimA. Only esonances ha a e less han 5
Hz wide we e obse ed in a
19
Fspec um(Fig. S2), showing ha
hey s em om small molecules. The esonances could be
assigned o esidual 3-b omo-1,1,1- i luo oace one (BTFA) label
and 1,1,1- i luo oace one (TFA), he la e being he p oduc o
an incomple e labeling eac ion. The esonance a 2119.8 ppm
s ems om an impu i y in he BTFA solu ion. The e o e, no
labeling was obse ed o WTPimA.Weconclude ha
PimA
R144C-TFA
is only labeled a he in oduced Cys esidue.
Despi e being singly labeled, wo esonances a e obse ed a
284.1 ppm and 282.7 ppm (Fig. 1B,bo om). As discussed
below, hese wo peaks display sa u a ion ans e , u he co -
obo a ing ha hey s em om he same
19
F label. To assign
he esonances o con o ma ions o PimA, we in es iga ed he
19
F spec um o he iple mu an PimA
T126C-V359C-R144C-TFA
,
p e iously shown o be locked in he inac i e con o ma ion.
This a ian canno unde go eshu ling due o a disul ide bond
o med be ween Cys126 and Cys359; hese esidues a e in close
spa ial p oximi y in he inac i e s a e bu a apa in he ac i e
s a e (23). PimA
T126C-V359C-R144C-TFA
displays only a single es-
onance a 284.1 ppm (Fig. 1B, op), which he e o e can be
assigned o he inac i e s a e. Consequen ly, he esonance a
282.7 ppm is assigned o he ac i e s a e. This assignmen is
co obo a ed by he ac ha he esonance a 284.1 ppm shi s
0.1 ppm up ield in 100% D
2
O, whe eas he esonance a 282.7
ppm does no shi (Fig. 1B, cen e ). Fo ully sol en -exposed
19
F labels, up ield sol en iso ope shi s o 0.2–0.3 ppm ha e
been obse ed when compa ing shi s in a 90% H
2
O/10% D
2
O
o 100% D
2
O sol en (33,40), indica ing ha he esonance a
284.1 ppm s ems om pa ially exposed
19
F label. In he c ys al
s uc u e, A g144 is shielded om sol en in he ac i e s a e,
whe eas his is no he case o he inac i e con o ma ions (Fig.
1A), suppo ing he assignmen . The assignmen is u he sup-
po ed by ex ensi e molecula dynamics (MD) simula ions pe -
o med on he di e en s a es o PimA labeled wi h TFA. Based
on hese simula ions, we calcula ed he sol en -accessible su -
ace a ea (SASA) o he TFA label in he di e en s a es. A
highe SASA alue indica es ha he label is mo e sol en -
exposed. As shown in Fig. S3, he SASA o he TFA label in he
ac i e s a e is 19.3 Å
2
, which di e s ma kedly om he alues
ob ained o he p o ein in he inac i e o ms (96.7 Å
2
and 60.1
Å
2
o he ex ended and compac s uc u es, espec i ely).
Thus, MD simula ions e ealed ha he TFA label in he ac i e
s a e is less su ace exposed han ha in he inac i e s a e, in
ag eemen wi h he NMR expe imen s.
PimA unde goes slow con o ma ional exchange be ween he
inac i e and ac i e s a es
To es whe he he wo esonances in he
19
F spec um o
PimA
R144C-TFA
a e in slow exchange, we conduc ed sa u a ion
ans e expe imen s a inc easing sa u a ion imes. Exchange
a es we e ob ained om a i o he model gi en in Equa ions
4–7. Expe imen al da a o ob ain R
1
longi udinal elaxa ion
a es ha a e equi ed o he analysis a e shown in Fig. S4 and
Table S2.Figu e 2 shows esonance in ensi y a enua ion due
o slow con o ma ional exchange o he inac i e and ac i e
s a e esonances upon sa u a ion o he ac i e o inac i e s a e
esonance, espec i ely.
In he absence o GDP-Man, he exchange a e was de e -
mined o be 2.3 s
21
. These dynamics co espond o old-
swi ching be ween he ac i e and inac i e s a es o he p o ein,
as demons a ed by he assignmen s (Fig. 1B). The second
imescale o he mo ion is in acco dance wi h s udies o he
chemokine lympho ac in, o which old-swi ching dynamics
we e shown o be on he o de o seconds (14). In he absence
o GDP-Man, he ac i e and inac i e s a es o PimA a e nea ly
equally popula ed (Fig. 2Cand Table 1).
To es how GDP-Man a ec s hese dynamics, we i s con-
duc ed a i a ion expe imen o co obo a e ha he suga do-
no sa u a es he binding si e. As shown in Fig. 1C, i a ion o
GDP-Man o 100
m
MPimA
R144C-TFA
leads o a g adual dec ease
in in ensi ies o bo h esonances up o a nucleo ide suga con-
cen a ion o 200
m
M. A his concen a ion, he inac i e s a e
esonance shi s 0.2 ppm up ield, and u he addi ion o GDP-
Man does no al e he spec um anymo e. A he ield s eng h
employed in Fig. 1C(16.4 T), he ac i e s a e esonance is oo
b oad o clea ly obse e a esonance shi . Since he line
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
9870 J. Biol. Chem. (2020) 295(29) 9868–9878
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
b oadening due o he ield-dependen chemical shi aniso -
opy o
19
F is educed a lowe ield s eng h, we compa ed he
spec a o PimA
R144C-TFA
in he p esence and absence o GDP-
Man a 14.1 T. A his lowe ield, a down ield shi o he ac i e
s a e esonance o 0.1 ppm was obse able a 500
m
MGDP-
Man and 100
m
MPimA
R144C-TFA
(Fig. S5). Mo eo e , bo h
peaks b oaden upon addi ion o GDP-Man, indica ing ha he
p esence o GDP-Man al e s dynamics and/o popula ions o
PimA. This p ocess may be ela ed o GDP-Man binding i sel
o o he dynamic p ocesses obse ed in he eshu ling egion
Figu e 2. Slow con o ma ional exchange o PimA
R144C-TFA
.(A) In ensi y a enua ion o he ac i e s a e esonance in absence (black) and p esence (o ange)
o sa u a ing concen a ions o GDP-Man upon sa u a ion o he inac i e s a e esonance. (B) In ensi y a enua ion o he inac i e s a e esonance in he ab-
sence (black) and p esence (o ange) o sa u a ing concen a ions o GDP-Man upon he sa u a ion o he ac i e s a e esonance. Expe imen s we e conduc ed
wi h 100
m
MPimA
R144C-TFA
and 500
m
MGDP-Man, whe e applicable. (C) Seconda y s uc u e ep esen a ion o he eshu ling egion. The posi ion o he mu a-
ion (R144C) is indica ed as a g een do . Pa ame e s (p, popula ions; k
s
, slow exchange a es) a e ob ained om he i o da a in panels A and B.
Table 1
Pa ame e s ob ained om sa u a ion ans e expe imen s o (A) he ac i e s a e esonance and (B) o he inac i e s a e esonance o
PimA
R144C-TFA
Condi ion and esonance s a e Chemical shi (ppm) k
IAa
(s
21
)p
I
p
A
k
s
(s
21
)DG
IA
(kcal mol
21
)
Ac i e s a e
Apo 282.7 1.08 60.05 0.4760.04 2.3160.09 20.07060.090
GDP-Man 282.6 4.260.2 0.3460.04 13 6120.39060.180
Inac i e s a e
Apo 284.1 1.23 60.04 0.5360.04 2.3160.09 0.07060.090
GDP-Man 284.3 8.360.9 0.6660.12 13 61 0.39060.180
a
I, he inac i e s a e; A, he ac i e s a e, i.e. k
IA
is he exchange a e om he inac i e o he ac i e s a e, and DG
IA
is he ee ene gy di e ence be ween he inac i e and ac i e s a e
and ice e sa.
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
J. Biol. Chem. (2020) 295(29) 9868–9878 9871
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
o PimA, as discussed below. GDP-Man and GDP we e p e i-
ously seen o bind PimA in en halpy-d i en eac ions wi h disso-
cia ion cons an s (K
D
)o 0.23
m
M(DH=–15.6 kcal·mol
21
)and
0.03
m
M(DH= –14.0 kcal·mol
21
), espec i ely (41). Thus, he
submic omola K
D
indica es ha unde he expe imen al condi-
ions employed he e, 500
m
MGDP-Man and 100
m
MPimA
R144C-
TFA
binding o GDP-Man o PimA is ully sa u a ed. Thus, we can
exclude any con ibu ion o GDP-Man binding o he obse ed
dynamics discussedsubsequen ly.
Unde he condi ion ha GDP-Man is p esen a sa u a ing con-
cen a ions, he old-swi ching dynamics o PimA
R144C-TFA
a e
s ongly enhanced, and he exchange a e was obse ed o inc ease
o 13 s
21
(Fig. 2 and Table 1). A concu en shi in he popula ion
equilib ium owa d he inac i e s a es was obse ed (by a ound
10%), as e idenced by analyzing peak in eg als as well as he sa u a-
ion ans e expe imen s (Fig. 2,Fig. S5,andTable S3).
The inac i e s a e o PimA unde goes millisecond mo ions
be ween a compac and an ex ended s a e
To cap u e possible dynamics on he millisecond imescale,
we conduc ed Ca -Pu cell-Meiboom-Gill (CPMG) elaxa ion
dispe sion expe imen s on PimA
R144C-TFA
a wo magne ic ield
s eng hs in he absence and p esence o sa u a ing amoun s o
GDP-Man. Assuming a wo-s a e exchange, we i ed he da a
o Equa ion 3 (see Expe imen al p ocedu es) o ob ain an es i-
ma e o he size o he popula ions, he exchange a e be ween
hem, and he chemical shi di e ence due o he con o ma-
ional exchange. Da a ha we e acqui ed a one magne ic ield
s eng h only we e i ed o Equa ion 2 o ob ain an appa en
exchange a e.
The appa en exchange a e o he esonance ep esen ing
he inac i e s a e o PimA
R144C-TFA
and he iple mu an
PimA
T126C-V359C-R144C-TFA
locked in o he inac i e s a e a e
iden ical (Fig. S6). Al hough he iple mu an canno unde go
eshu ling o he ac i e s a e, he as e domain dynamics o
he inac i e s a e i sel a e no a ec ed. This is expec ed om
inspec ing he s uc u e o PimA. While he disul ide bond
be ween Cys126 and Cys359 p e en s seconda y s uc u e
eshu ling, i does no impede he ex ended- o-compac
mo ion o he inac i e s a es (Fig. 1A). Thus, he elaxa ion dis-
pe sion displayed by he inac i e s a e esonance moni o s he
ex ended- o-compac dynamics o he eshu ling egion.
As shown in Fig. 3A, he esonance in he
19
F spec um o
PimA
R144C-TFA
moni o ing he inac i e s a e displays elaxa ion
dispe sion be ween wo con o ma ions wi h di e en popula-
ions, and he p o ile is a ec ed by he addi ion o sa u a ing
amoun s o GDP-Man, e lec ing changes in dynamics and
popula ion equilib ia. In he absence o GDP-Man, he
exchange a e is 8000 s
21
, which dec eases o 5100 s
21
in he
p esence o GDP-Man (Table 2). Simila exchange a es ha e
p e iously been obse ed o mo ions o a helix o he
b
2-ad e-
ne gic ecep o (32). In he absence o GDP-Man, one o he
s a es is highly popula ed and he o he one is no (1% e sus
99%). Upon addi ion o GDP-Man, he less popula ed o he
inac i e s a es becomes sligh ly mo e popula ed (3% e sus
97%), which is accompanied by an up ield esonance shi o he
inac i e s a e esonance (Fig. 1C), indica ing ha he obse ed
change in popula ions and dynamics esul s in a mo e pola
en i onmen , on a e age, o he
19
F label. While he popula-
ion change upon addi ion o GDP-Man (1.7%) as de e mined
om he elaxa ion dispe sion da a is signi ican , gi en he i -
ed e o s in he analysis o he elaxa ion da a (Table 2), i
migh no be possible o decon olu e model pa ame e s
en i ely, gi en ha expe imen s we e acqui ed a wo magne ic
ield s eng hs only; howe e , changes in dynamics and changes
in popula ions owa d he less popula ed s a e a e also mani es
in he obse ed upshi o he inac i e s a e esonance in Fig.
1C,i.e. o he s a e in which he
19
F label is loca ed in a mo e po-
la en i onmen . Resonances o
19
F nuclei in a mo e pola en i-
onmen a e loca ed up ield o
19
F nuclei bu ied inside a p o ein
(32,33). This is in ag eemen wi h ou obse a ion ha he
inac i e s a e esonance, o which he
19
F label is mo e
exposed o he sol en han o he ac i e s a e esonance (Fig.
1B), is up ield o he ac i e s a e esonance. Based on he c ys al
s uc u es (Fig. 1A) and ou MD simula ions (Fig. S3), he
19
F
label is mos sol en -exposed in he inac i e, ex ended con o -
ma ion. Thus, we iden i y he spa sely popula ed s a e wi h he
inac i e, ex ended con o ma ion. This conclusion is u he
co obo a ed by s uc u al ea u es o PimA.
The con o ma ional ansi ion om he compac o ex ended
inac i e s a e o PimA, in he absence o GDP-Man, was ep o-
duced by s ee ed-MD simula ions, as shown in Fig. S7.This ypeo
calcula ion o ces he sys em o e ol e away om i s ini ial equilib-
ium condi ion (compac , inac i e) o a inal s a e (ex ended, inac-
i e), he eby accele a ing ansi ions be ween di e en ene gy
minima (42–44). This me hod is gene ally applied in s udying
many biophysical p ocesses, including he (un) olding mechanism
o p o eins (45,46). As e ealed by he s ee ed-MD simula ions,
he inac i e, ex ended s a e's la ge hyd ophobic pa ches a e
exposed, making his s a e po en ially agg ega ion-p one, whe eas
he hyd ophobic pa ches a e shielded om sol en in he inac i e,
compac s a e (23). The ac ha he addi ion o GDP-Man leads
o an inc ease in popula ion o he inac i e, ex ended s a e sugges s
ha he slowe eshu ling e en occu s in ha s a e.
The ac i e s a e o PimA unde goes millisecond mo ions ha
may e lec open- o-closed dynamics
As shown in Fig. 3C, he esonance in he
19
F spec um o
PimA
R144C-TFA
moni o ing he ac i e s a es also displays elaxa-
ion dispe sion, and he p o ile is again a ec ed by he addi ion o
sa u a ing amoun s o GDP-Man, e lec ing changes in dynamics
and popula ion equilib ia. In con as o he inac i e s a e eso-
nance, he ac i e s a e esonance is downshi ed upon addi ion o
GDP-Man (Fig. S5). This indica es ha a popula ion shi owa d
a mo e hyd ophobic en i onmen o he
19
F label occu s, which
migh be iden i ied wi h a popula ion shi om a s uc u ally
uncha ac e ized ac i e, ex ended s a e o he ac i e, compac
s a e o PimA. Based on his in e p e a ion, mos o PimA is
oundin heac i e,ex endeds a e(81%ex ended e sus 19%
compac ). This holds ue e en a e addi ion o GDP-Man (70%
ex ended e sus 30% compac ) (Fig. 3D). The ac i e, ex ended
s a e migh be ep esen ed by he open e sion o he al eady
s uc u ally cha ac e ized ac i e, closed s a e, sugges ing ha
hese dynamics co espond o he open- o-closed mo ion o
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
9872 J. Biol. Chem. (2020) 295(29) 9868–9878
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
PimA (Fig. 1A).Theexchange a ebe ween heex endedand
compac o ms o he ac i e s a esin hiscaseinc eases om
4300 s
21
o 8000 s
21
upon addi ion o sa u a ing concen a ions
o GDP-Man.
In e es ingly, accele a ed MD simula ions pe o med on
PimA labeled wi h TFA and s a ing om he compac ,
ac i e s a e indica ed ha he p o ein unde goes a con o -
ma ional change o each a s a e in which
b
7 dissocia es
om
b
6, which indica es he ini ia ion o he ansi ion o
a
-helix (Fig. S8). As s ee ed-MD simula ions, his compu a-
ional app oach is use ul o imp o ing con o ma ional
spacesampling.Theme hoddoesno equi eanyp ede-
ined coo dina es and educes he ene gy ba ie be ween
he di e en low-ene gy s a es (47).
The ac i e and inac i e s a es a e s uc u ally dis inc
The elaxa ion dispe sion da a indica e ha he inac i e,
compac and inac i e, ex ended s a es a e s uc u ally dis inc ,
as hei esonances a e sepa a ed by 1.2 ppm (Table 2). As dis-
cussed abo e, he esonance o he inac i e, ex ended s a e is
expec ed o be up ield o he esonance o he inac i e, compac
s a e, since he
19
F label is mo e sol en -exposed in he la e
(Fig. 1C). Simila ly, he elaxa ion dispe sion da a indica e ha
he ac i e, compac and ac i e, ex ended s a e a e s uc u ally
dis inc , as hei esonances a e sepa a ed by 2.9 ppm (Table 2).
The down ield shi o he esonance o he ac i e, compac
s a e compa ed wi h he ac i e, ex ended s a e (Fig. S5) indi-
ca es ha in he ac i e, compac s a e he
19
F label is loca ed
in a mo e hyd ophobic en i onmen . Impo an ly, om he
Figu e 3. In e media e con o ma ional exchange o he inac i e and he ac i e s a es o PimA
R144C-TFA
.(A) Relaxa ion dispe sion o he inac i e s a e es-
onance in he absence (black) and p esence (o ange) o sa u a ing concen a ions o GDP-Man. (B) Ex ended- o-compac dynamics in he eshu ling egion
(ex ended con o ma ion in o ange; compac con o ma ion in yellow). Pa ame e s (p, popula ions; D
d
, chemical shi di e ence) ob ained om he i o da a in
panel A a e indica ed (black, no ligand; o ange, wi h GDP-Man). (C) Relaxa ion dispe sion o he ac i e s a e esonance in he absence (black) and p esence (o -
ange) o sa u a ing concen a ions o GDP-Man. (D) Ex ended- o-compac dynamics ha occu in he eshu ling egion (ex ended con o ma ion in g ay; com-
pac con o ma ion in ed). Only he compac , ac i e con o ma ion is s uc u ally cha ac e ized. Pa ame e s (p, popula ions; D
d
, chemical shi di e ence)
ob ained om he i o he da a in panel C a e indica ed (black, no ligand; o ange, wi h GDP-Man). Expe imen s we e conduc ed wi h 100
m
MPimA
R144C-TFA
and 500
m
MGDP-Man whe e applicable. All obse ed exchange a es a e on he o de o 4000–8000 s
21
(see Table 1 o de ails).
Table 2
Pa ame e s ob ained om CPMG elaxa ion dispe sion expe imen s o PimA
R144C-TFA
i ed o Equa ion 3
Condi ion Con o ma ion Chemical shi (ppm) k
i
(10
3
s
21
)p
ex ended
p
compac
|D
d
|
a
(ppm)
Apo Ac i e 282.7 4.361.3 0.8160.03 0.19 60.03 1.2 60.1
GDP-Man Ac i e 282.6 7.361.5 0.7060.13 0.30 60.13 1.2 60.1
Apo Inac i e 284.1 8.068.0 0.009 60.002 0.991 60.002 2.9 60.3
GDP-Man Inac i e 284.3 5.162.7 0.026 60.008 0.974 60.008 2.9 60.3
a
|D
d
| is a global a iable o each peak.
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
J. Biol. Chem. (2020) 295(29) 9868–9878 9873
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
elaxa ion dispe sion da a (Table 2)and he1Dspec a(Fig. 1C,
Fig. S5), i can be concluded ha all con o ma ional s a es a e
s uc u ally dis inc , indica ing ha old-swi ching is accompa-
nied by a d ama ic s uc u al ea angemen nea he
19
Flabel.
This is consis en wi h he s uc u al da a ha show ha esidue
A g144 is eshu led om an
a
-helix (po en ially ia a sol en -
exposed andom coil s uc u e) o a
b
-shee en i onmen .
Discussion
The abili y o an enzyme o adop many di e en con o ma-
ions du ing he ca aly ic cycle is due o i s g ea lexibili y, play-
ing a c i ical ole in he ec ui men o subs a es, elease o
p oduc s, s abiliza ion o ansi ion s a es, and adap a ion o
he cell en i onmen . Employing
19
F NMR, we iden i ied ou
unc ionally ele an s a es o PimA ha coexis in dynamic
equilib ia in solu ion unde going con o ma ional exchange on
imescales om milliseconds o seconds. Speci ically, old-
swi ching, a hi he o poo ly cha ac e ized p ocess, occu s in
he eshu ling egion o he N- e minal domain o PimA on a
second imescale. Domain mo ions, on he o he hand, occu
on a millisecond imescale.
The obse a ions p esen ed he e allow us o p opose ha he
ac i a ion o PimA occu s along a sequence o s uc u al ea -
angemen s displayed in Fig. 4.InFig. S9, a ee-ene gy diag am
o he ac i a ion p ocess is depic ed ha is de i ed om he -
modynamic da a ob ained om he sa u a ion ans e and
elaxa ion dispe sion expe imen s (Tables 1 and 2). Ini ially,
PimA is inac i e and in a compac s a e ha shields hyd opho-
bic pa ches in he eshu ling egion. Almos all inac i e PimA
molecules a e in ha con o ma ion bu a e in millisecond
exchange wi h he inac i e, ex ended s a e. This s a e exposes
hyd ophobic pa ches and, hus, is spa sely popula ed. Howe e ,
i o e s su icien deg ees o eedom o a ea angemen o
seconda y s uc u e elemen s on a second imescale. A d as ic
change in chemical shi indica es ha he inac i e, ex ended
s a e is s uc u ally dis inc om he uncha ac e ized ac i e,
ex ended s a e, which migh co espond o he open s a e o
he s uc u ally cha ac e ized closed, GDP-Man-bound con o -
ma ion o PimA. Following he old-swi ch, which we specula e
o occu om he inac i e, ex ended s a e, PimA is in he ac i e,
ex ended s a e and in millisecond exchange wi h he ac i e, com-
pac s a e, which is he con o ma ion in which ca alysis occu s.
We p opose ha old-swi chingisanac i a ionmechanismo
PimA, and since i is a slow p ocess, i may ep esen he a e-lim-
i ing s ep o he glycosyla ion eac ion ca alyzed by PimA. The
subs a e, GDP-Man, ac s as an ac i a o and, hus, unc ional
egula o o PimA, as he binding o GDP-Man enhances he
old-swi ching dynamics. In iguingly, a signi ican popula ion o
he inac i e s a e is always p esen . This may poin o a mo e sig-
ni ican ole o he inac i e s a es in he ca aly ic cycle o PimA
han p e iously an icipa ed, o ins ance, in memb ane binding o
PimA o o acili a e e icien ec ui men o phospha idyl-myo-
inosi ol lipids, which a e accep o subs a es o PimA.
Conclusions
This s udy demons a es ha he
19
F-based NMR app oach p e-
sen ed he e allows o insigh s in o s uc u es and dynamics o
p o eins ha canno be s udied wi h mo e con en ional NMR
me hods. The me hods employed he e, a combina ion o sa u a-
ion ans e me hods and elaxa ion dispe sion expe imen s, allow
in es iga ions o dynamics occu ing simul aneously and on a wide
ange o imescales in solu ion. He e, his app oach e ealed a e-
ma kable si ua ion in which PimA exis s in a leas ou di e en
con o ma ions ha a e all p esen in solu ion. The dynamics
be ween and he ela i e popula ions o hese con o ma ions a e
a ec ed by he addi ion o subs a e, which demons a es ha he
equilib ia a e inely uned o op imize he ca aly ic cycle.
Ma e ials and Me hods
Ma e ials
1,1,1-T i luo oace one (TFA), 3-b omo-1,1,1- i luo oace one
(BTFA), guanosine 59-diphospho-D-mannose (GDP-Man), and
is(2-ca boxye hyl)phosphine(TCEP)we eacqui ed om
Me ck (Da ms ad , Ge many).
Cloning and pu i ica ion o WT PimA and PimA mu an s
PimA
R144C
and PimA
T126C-V359C-R144C
mu an s o PimA
om Mycobac e ium smegma is we e ob ained om GenSc ip
using pET29a-pimA as he DNA empla e (25). Recombinan
Figu e 4. Ac i a ion mechanism o PimA. S uc u al changes o PimA upon ac i a ion. Ini ially, almos all inac i e PimA molecules a e in he compac con-
o ma ion bu a e in millisecond exchange wi h he inac i e, ex ended s a e. Following he old-swi ch, PimA is in he s uc u ally uncha ac e ized ac i e,
ex ended s a e and in millisecond exchange wi h he ac i e, compac s a e, which is he con o ma ion in which ca alysis occu s. The loca ion o R144 is
shown/modeled o cla i y.
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
9874 J. Biol. Chem. (2020) 295(29) 9868–9878
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
WT PimA and he PimA
R144C
mu an we e exp essed in
Esche ichia coli BL21(DE3) pLysS and pu i ied as desc ibed
p e iously (25). PimA
T126C-V359C-R144C
mu an was exp essed
in E. coli Rose a Gami 2(DE3) pLysS and pu i ied as desc ibed
p e iously (23).
Labeling o PimA
An es ablished p o ocol was adap ed o label PimA wi h TFA
(48). 1 mMTCEP was added o 100–200
m
MPimA and incu-
ba ed o 1 h on ice, excep o PimA
T126C-V359C-R144C-TFA
, o
which no TCEP was added. Nex , BTFA was added in 10- old
mola excess (1–2mM), and he solu ion was incuba ed o e -
nigh a 4°C unde gen le agi a ion. Finally, he solu ion was
passed h ough a desal ing column equilib a ed wi h 50 mM
T is-HCl, pH 7.5, and 150 mMNaCl. The sample condi ions o
NMR expe imen s we e 100
m
MPimA, 50 mMT is-HCl, pH
7.5, 150 mMNaCl, and 10% D
2
O. 500
m
MGDP-Man was added
whe e app op ia e and i no s a ed o he wise.
To exchange o 100% D
2
O sol en , 100
m
MPimA was con-
cen a ed oughly i e imes and dilu ed back o a concen a-
ion o 100
m
Musing 50 mMT is-HCl, 150 mMNaCl in 100%
D
2
O. This p ocedu e was epea ed 5 imes.
NMR spec oscopy
Expe imen s we e pe o med a 25 °C on a B uke A ance 600
MHz III HD spec ome e (
19
F La mo equency o 564 MHz).
Relaxa ion dispe sion and i a ion da a o PimA
R144C-TFA
we e
addi ionally acqui ed on a B uke A ance 700 MHz spec ome e
(
19
F La mo equency o 659 MHz). Bo h spec ome e s we e
equipped wi h a
19
Fc yop obe.Typical
p
/2 pulse leng hs we e
15–19 ms, and he acquisi ion ime was 0.03 s. 10% D
2
Owas
added o he samples o equency locking. Fo 1D spec a, 400
ansien s we e acqui ed wi h a elaxa ion delay o 1 s. Da a we e
p ocessed in Topspin (B uke , Bille ica, MA) and analyzed using
in-house Ma lab ou ines (Ma hsWo ks, Na ick, MA) and Igo
P o (Wa eMe ics, Lake Oswego, OR), employing an in-house
sc ip o nonlinea i ing o elaxa ion dispe sion da a (49).
O e lapping esonances we e decon olu ed, employing iNMR
so wa e (Mes eLab Resea ch, Spain) by i ing spec a o dou-
ble-Lo en zian unc ions.
Fa -UV CD analysis
Spec a we e acqui ed in a J-810 CD spec opola ime e (Jasco
Co p., Tokyo, Japan) by using Hellma 105.200-QS qua z cu e es
wi h a 1 cm op ical pa h. Spec a we e eco ded in a con inuous
mode wi h 1 nm bandwid h, 1 s esponse, and a scan speed o 100
nm/min
21
.Sampleswe e2.5
m
MPimA, PimA
R144C
,o
PimA
R144C-TFA
in 10 mMT is-HCl, pH 7.5. GDP was added a a
1:15 a io, and 25 scans we e accumula ed o ob ain he inal
spec a, which we e u he co ec ed o he baseline signal.
Spec a we e eco ded in he 200–250 nm ange a 20°C.
Tempe a u e scans
Spec a we e acqui ed in a J-810 CD spec opola ime e (Jasco
Co p., Tokyo, Japan) by using Hellma 110-QS qua z cu e es wi h
a 1 mm op ical pa h by using a Pel ie he mal de ice, allowing
empe a u e con ol du ing he expe imen s. Spec a we e
eco ded in a con inuous mode wi h 1 nm bandwid h, 1 s
esponse, and a scan speed o 100 nm/min
21
.Sampleswe e10
m
M
PimA, PimA
R144C
,o PimA
R144C-TFA
in 10 mMT is-HCl, pH 7.5.
Subs a es we e added a a 1:15 a io. In his case, he mal depend-
encies o he ellip ici y we e moni o ed in he ange om 20 °C o
90°C a 222 nm. Tempe a u e was inc eased s epwise by 1°C/min.
Relaxa ion dispe sion expe imen s
1D
19
F CPMG elaxa ion dispe sion expe imen s (34,50)we e
pe o med analogously o p e ious s udies (32). 2400 ansien s
we e acqui ed du ing a cons an ime delay, T
CPMG
,o 3.84ms
o 14 e ocusing equencies,
n
CPMG
, angingbe ween;130 s
21
and ;6000 s
21
. E ec i e ans e se elaxa ion a es, R
2,e
,we e
compu ed om peak in ensi ies I(
n
CPMG
) ela i e o he peak in-
ensi y, I
0
, in he absence o a cons an ime delay as ollows
R2;e ¼T1
CPMGln I
CPMG
ðÞ
I0
(Eq. 1)
Fo da a measu ed a one magne ic ield s eng h, elaxa ion
dispe sion da a we e i ed o Luz-Meiboom’s wo-s a e model
o ob ain an appa en exchange a e, k
app
, o he dynamics (50):
R2;e
n
CPMG
ðÞ
¼c
kapp 14
n
CPMG
kapp anh kapp
4
n
CPMG
!
1R2;0
(Eq. 2)
whe e
n
CPMG ¼0:5
1
CPMG,
CPMG
is he sepa a ion be ween
wo 180°pulses and k
app
,R
2,0
, and c¼pApBD
2a e pa ame e s
o he i . The popula ions, p
A
and p
B
, o he wo s a es and he
chemical shi di e ence be ween he esonances, D
, canno
be decon olu ed wi h his model.
Relaxa ion dispe sion da a measu ed a wo magne ic ield
s eng hs we e i ed o Ca e -Richa ds model assuming wo
s a es o ob ain he exchange a e o he p ocess, k
i
(iindica es
in e media e exchange), popula ions, p
A
and p
B
, and he chemi-
cal shi di e ence, D
, be ween he s a es (34):
pA¼1pB
j
¼2D
pBpA
ðÞ
ki
c
¼pBpA
ðÞ
2ki2D
214pBpAki2
h
6¼1
ffiffiffi
8
p
n
CPMG ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
6
c
1ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
21
j
2
q
D6¼1
2611
c
12D
2
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
c
21
j
2
p
!
R2;e ¼R2;011
2ki2
n
CPMG a cosh D1cosh
h
1
ðÞ
Dcos
h
ðÞ
ÞÞ (Eq. 3)
To educe he numbe o pa ame e s in he i , i was assumed
ha he addi ion o GDP-Man did no al e he con o ma ions o
PimA p esen in solu ion, i.e. he chemical shi D
d
¼D
=2
p
ðÞ
,
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
J. Biol. Chem. (2020) 295(29) 9868–9878 9875
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
whe e is he spec ome e equency, was ea ed as a global con-
s an . Mo eo e , he magne ic ield s eng h does no a ec p o ein
kine ics, i.e. k
i
,p
A
,andp
B
we e iden ical o da a ob ained a di e -
en magne ic ield s eng hs bu could a y be ween samples ha
con ained o did no con ain GDP-Man.
T
1
expe imen s
In e sion eco e y expe imen s wi h a leas 10 delays,
,
anging be ween 25 ms and 800 ms we e pe o med o ob ain
he longi udinal elaxa ion ime, T
1
. 1200 ansien s we e
acqui ed, and he elaxa ion delay was 2 s. The elaxa ion a e,
R
1
=T
121
, was ob ained om a monoexponen ial i o he
in ensi ies in dependence on delay
.
Sa u a ion ans e expe imen s
Sa u a ion ans e expe imen s we e conduc ed o quan i y
slow exchange dynamics o PimA (32,51). Resonances we e sa u-
a ed wi h con inuous wa e i adia ion (B
1
ield o 28 Hz). The
esonance associa ed wi h he ac i e s a e was i adia ed a 282.3
ppm, and he in ensi y decay o he inac i e s a e esonance was
moni o ed a 284.1 ppm in he absence o subs a e and a
284.3 ppm in he p esence o GDP-Man. Con e sely, he eso-
nance associa ed wi h he inac i e s a e was i adia ed a 284.3
ppm, and he in ensi y decay o he ac i e s a e esonance was
moni o ed a 282.7 ppm in he absence o subs a e and a
282.6 ppm in he p esence o GDP-Man. Spec a we e acqui ed
o 8 sa u a ion imes, a ying be ween 25 ms and 1000 ms. 1200
ansien s we e eco ded, and he elaxa ion delay was 1.5 s.
Exchange a es we e ob ained by i ing esidual in ensi ies, I
ðÞ
,
a inc easing i adia ion imes
o a wo-s a e model (35):
Im
ðÞ¼Im0
ðÞ
kmn 1R1;m
kmnexp
kmn 1R1;m
ðÞðÞ
1R1;m
ðÞ
(Eq. 4)
Whe e mis inac i e and nis ac i e, ep esen ing he exchange
a e, k
mn
, om he inac i e o he ac i e s a e and ice e sa.
The exchange a e o slow dynamics, k
s
(sindica es slow
exchange), is he sum o k
mn
and k
nm
:
ks¼kmn 1knm (Eq. 5)
Popula ions, p, o he s a es and he ee-ene gy di e ence,
DG, a e gi en by
pn¼kmn
ks
(Eq. 6)
DGnm ¼RT ln pn
pm
(Eq. 7)
whe e Tis he absolu e empe a u e and Ris he uni e sal gas
cons an .
Con en ional MD simula ions
The MD simula ions we e pe o med wi h he AMBER 18
package (52). Th ee di e en simula ions, s a ing om he
ac i e-compac , inac i e-compac , and inac i e-ex ended con o -
ma ions, we e pe o med. Pa ame e s o he TFA label we e
gene a ed wi h he an echambe module o AMBER imple-
men ed wi h he gene al Ambe o ce ield (GAFF) (53), wi h pa -
ial cha ges se o i he elec os a ic po en ial gene a ed wi h
HF/6-31G(d) by RESP (54). The cha ges we e calcula ed acco d-
ing o he Me z-Singh-Kollman scheme using Gaussian 09 (55).
Each p o ein was imme sed in a wa e box wi h a 10 Å bu e o
TIP3P (56) wa e molecules. The sys em was neu alized by add-
ing explici coun e ions (Na
1
). A wo-s age geome y op imiza-
ion app oach was pe o med. The i s s age minimizes only he
posi ions o sol en molecules and ions, and he second s age is
an un es ained minimiza ion o all he a oms in he simula ion
cell. The sys ems hen we e gen ly hea ed by inc emen ally
inc easing he empe a u e om 0 o 300 K unde a cons an
p essu e o 1 a m and pe iodic bounda y condi ions. Ha monic
es ain s o 30 kcal·mol
21
we e applied o he solu e, and he
Ande sen empe a u e coupling scheme (57)wasused ocon ol
and equalize he empe a u e. The imes epwaskep a 1 sdu -
ing hehea ings ages,allowingpo en ial inhomogenei ies o sel -
adjus . Wa e molecules a e ea ed wi h he SHAKE algo i hm
such ha he angle be ween he hyd ogen a oms is kep ixed.
Long- ange elec os a ic e ec s a e modelled using he pa icle-
mesh-Ewald me hod (58). An 8 Å cu o was applied o Lenna d-
Jones and elec os a ic in e ac ions. Each sys em was equilib a ed
o 2 ns wi h a 2 s ime s ep a a cons an olume and empe a-
u e o 300 K. P oduc ion ajec o ies we e hen un o an addi-
ional 200 ns unde he same simula ion condi ions. SASA o lu-
o ine a oms o he TFA label was calcula ed using he keywo d
“SURF”in he cpp aj module (59)o AMBER.
Accele a ed MD simula ions
Wi h he accele a ed MD implemen ed in AMBER 18, 500
ns accele a ed MD simula ions we e pe o med on PimA
labeled wi h TFA and s a ing om he ac i e-compac con o -
ma ion. These simula ions we e es a ed om he inal s uc-
u e o he 200 ns con en ional MD simula ions wi h andom
a omic eloci y ini ializa ions a 300 K. Boos po en ial wi h an
ex a boos o he o sions was applied (iamd =3).
S ee ed-MD simula ions
The p o ocol employed abo e o con en ional MD simula ions
was also used in s ee ed-MD calcula ions (60). The compac -inac-
i e con o ma ion was used as he s a ing s uc u e. Fo he inal
p oduc ion ajec o y, he dis ance be ween C-alpha o Ala148 and
Ala366 was a ied om 13.5 (close, inac i e s a e, 0 ns) o 51.3 Å
(ex ended, inac i e s a e, 200 ns), using an k2 o 8 kcal·mol
21
·Å
21
.
Da a a ailabili y
All da a desc ibed in he manusc ip a e con ained wi hin
he manusc ip o he suppo ing in o ma ion.
Acknowledgmen s—The Swedish NMR Cen e a he Uni e si y o
Go henbu g is acknowledged o suppo . We hank Axel Abelein
o p o iding a sc ip o analyze elaxa ion dispe sion da a.
Au ho con ibu ions—J. L., M. E. G., and L. M. concep ualiza ion;
J. L., F. C., and T. S. da a cu a ion; J. L. and L. M. so wa e; J. L.,
Ac i a ion dynamics o a old-swi ch glycosyl ans e ase
9876 J. Biol. Chem. (2020) 295(29) 9868–9878
a UNIVERSIDAD DEL PAIS VASCO/EHU on Decembe 16, 2020h p://www.jbc.o g/Downloaded om
