ARTICLE
Two pa allel pa hways connec glu amine
me abolism and mTORC1 ac i i y o egula e
glu amop osis
Clémen Bodineau 1,2, Me cedes Tomé1, Sa ah Cou ois 3, Ana S. H. Cos a 4,5, Ma co Sciaco elli 4,
Benoi Rousseau6, Elodie Richa d7, Pie e Vache 7, Ca los Pa ejo-Pé ez8, Emilie Bessede3, Ch is ine Va on3,
Pie e Soubey an7, Ch is ian F ezza 4, Piedad del Soco o Mu doch1,9, Vic o H. Villa 10 &
Raúl V. Du án 1,2,7✉
Glu amop osis is he induc ion o apop o ic cell dea h as a consequence o he abe an
ac i a ion o glu aminolysis and mTORC1 signaling du ing nu i ional imbalance in p o-
li e a ing cells. The ole o he bioene ge ic senso AMPK du ing glu amop osis is no defined
ye . He e, we show ha AMPK eac i a ion blocks bo h he glu amine-dependen ac i a ion
o mTORC1 and glu amop osis in i o and in i o. We also show ha glu amine is used o
aspa agine syn hesis and he GABA shun o p oduce ATP and o inhibi AMPK, indepen-
den ly o glu aminolysis. O e all, ou esul s indica e ha glu amine me abolism is connec ed
wi h mTORC1 ac i a ion h ough wo pa allel pa hways: an acu e alpha-ke oglu a a e-
dependen pa hway; and a seconda y ATP/AMPK-dependen pa hway. This dual me abolic
connec ion be ween glu amine and mTORC1 mus be conside ed o he u u e design o
he apeu ic s a egies o p e en cell g ow h in diseases such as cance .
h ps://doi.o g/10.1038/s41467-021-25079-4 OPEN
1Cen o Andaluz de Biología Molecula y Medicina Regene a i a—CABIMER, Consejo Supe io de In es igaciones Cien íficas, Uni e sidad de Se illa,
Uni e sidad Pablo de Ola ide, Se ille, Spain. 2Ins i u Eu opéen de Chimie e Biologie, INSERM U1218, Uni e si é de Bo deaux, Pessac, F ance. 3Bo deaux
Resea ch in T ansla ional Oncology, INSERM U1053, Uni e si é de Bo deaux, Bo deaux cedex, F ance. 4Medical Resea ch Council Cance Uni , Hu chison/
MRC Resea ch Cen e, Box 197, Camb idge Biomedical Campus, Uni e si y o Camb idge, Camb idge, UK. 5Cold Sp ing Ha bo Labo a o y, Cold Sp ing
Ha bo , NY, USA. 6Se ice Commun des Animale ies, Animale ie A2, Uni e si y o Bo deaux, Bo deaux, F ance. 7INSERM U1218, Ins i u Be gonié,
Bo deaux, F ance. 8Ins i u o de Bioquímica Vege al y Fo osín esis, Consejo Supe io de In es igaciones Cien íficas, Uni e sidad de Se illa, Se ille, Spain.
9Depa amen o de Bioquímica Vege al y Biología Molecula , Uni e sidad de Se illa, Se ille, Spain. 10 CRUK Bea son Ins i u e, Glasgow, UK.
✉email: [email p o ec ed]
NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://doi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions 1
1234567890():,;
Among all amino acids, glu amine is he mos abundan in
he blood1–3. The impo ance o glu amine o cance cells
g ow h has been ex ensi ely desc ibed and linked o i s
ole as a p ecu so o α-ke oglu a a e (αKG) o sus ain he i-
ca boxylic acid (TCA) cycle. Indeed, glu amine is mainly me a-
bolized h ough glu aminolysis in a wo-s ep eac ion: fi s he
enzyme glu aminase (GLS) ca alyzes he hyd olysis o glu amine
o glu ama e; and hen glu ama e is con e ed o αKG h ough an
oxida i e deamina ion eac ion ca alyzed by he enzyme glu a-
ma e dehyd ogenase (GDH). Ano he amino acid, leucine, ac s as
an allos e ic ac i a o o he second enzyme, being consequen ly
necessa y o glu aminolysis induc ion4,5.
Al hough glu amine can pa icipa e in he syn hesis o di e en
amino acids h ough he p oduc ion o glu ama e, only aspa -
agine equi es glu amine o de no o syn hesis. Aspa agine syn-
he ase (ASNS) is he enzyme ha con e s glu amine and
aspa a e in o glu ama e and aspa agine in an ATP-dependen
manne 6. ASNS exp ession is up egula ed du ing amino acid
s a a ion h ough he ac i a ion o ac i a ing ansc ip ion ac o
4 (ATF4)7. Aspa agine has been iden ified as an exchange amino
acid ac o ha egula es he se ine/ h eonine kinase mammalian
a ge o apamycin complex 1 (mTORC1), nucleo ide bio-
syn hesis and p oli e a ion7.
The ole o GABA shun is a bypass o wo s eps o he TCA
cycle, whe ein GABA is syn hesized om glu ama e by a GAD1-
encoded deca boxylase, ansamina ed in o succinic semi-
aldehyde, and me abolized in o he TCA cycle in e media e
succina e8. In addi ion o i s ole as a neu o ansmi e , GABA
has been de ec ed in a wide ange o pe iphe al issues9and
ecen ly linked o cas a ion- esis an p os a e cance p og ession
h ough he egula ion o nuclea and ogen ecep o by GABA10.
Tha s udy epo ed ha GABA shun is up egula ed in esponse
o he ac i a ing phospho yla ion o GAD65 by he PI3K pa h-
way. In a di e en cance ype, he inc eased up ake and me a-
bolism o GABA in b eas o b ain me as a ic cells was linked o
an inc ease o NADH le els in he mic oen i onmen con e ing
a p oli e a i e ad an age o he umo 11.
P e iously, ou esul s demons a ed ha glu amine, and mos
pa icula ly i s ca abolic con e sion o αKG h ough glu amino-
lysis, ac i a es mTORC1 a he sho and long- e m5,12. mTORC1
is a majo signaling hub ha in eg a es di e en inpu s, such as
nu ien s, oxygen, ene gy, and g ow h ac o s, o egula e he
me abolic pa hways con olling cell g ow h and p oli e a ion13.
As mTORC1 is o e -ac i a ed in 80% o solid umo s14, di e en
analogs o i s inhibi o apamycin ha e been de eloped o inhibi
mTORC1 in pa ien s. A a clinical le el, he esul s emain
modes and a clea need o co- he apies has been desc ibed15–17.
Ou p e ious wo k de e mined ha he ac i a ion o mTORC1 by
glu aminolysis du ing nu i ional imbalance lead o he anom-
alous inhibi ion o au ophagy and a subsequen o m o apop osis
named glu amop osis12,18. Bu s ill, he impac o glu amine
me abolism on he bioene ge ic s a us o he cells du ing long-
e m mTORC1 ac i a ion and glu amop osis induc ion emains
o be defined.
Euka yo es ha e de eloped a sys em o sense low ATP le els
ia ano he se ine/ h eonine kinase, he AMP-ac i a ed p o ein
kinase (AMPK) complex. Unde low in acellula ATP le els,
AMP o ADP can di ec ly bind o he γ egula o y subuni o
AMPK, leading o a con o ma ional change ha allows he ac i-
a ing phospho yla ion o AMPK13. Once ac i a ed, AMPK
edi ec s he me abolism owa ds he inc ease o ca abolism and
he dec ease in anabolism h ough phospho yla ion o down-
s eam key p o ein pa hways, ul ima ely leading o mTORC1
inhibi ion13.
He ein, we epo ha he amino acid glu amine is su ficien o
sus ain he p oduc ion o ATP in absence o any o he amino
acid, ollowing a glu aminolysis-independen mechanism. Du ing
glu amine su ficiency, ASNS and GABA shun a e esponsible o
me abolizing glu amine o gene a e ATP and o inhibi AMPK.
Thus, ou esul s indica e ha glu amine ac i a es mTORC1
ollowing a wo b anches mechanism: a sho - e m mechanism
in ol ing αKG p oduc ion, and a long- e m mechanism in ol ing
ASNS, GABA, and AMPK.
Resul s
Glu aminolysis sus ains he p oduc ion o ATP o inhibi
AMPK and o ac i a e mTORC1. P e iously, we showed ha he
addi ion o leucine and glu amine (“LQ ea men ”) o amino
acid-s a ed cells is su ficien o ac i a e glu aminolysis and
subsequen ly mTORC15,19. To be e unde s and he ole o he
bioene ge ic s a us o he cell in his pa hway, we fi s in es iga ed
i LQ ea men impac ed he p oduc ion o ATP in he cell and
he ac i a ion o AMPK. Confi ming ou p e ious obse a ions,
sho - e m (2 h) amino acid s a a ion had no e ec in he le els
o ATP in cul u ed cells. Howe e , we obse ed ha , a longe
imes, amino acid wi hd awal significan ly dec eased ATP/ADP
a io in U2OS and HCT116 cellula models (Fig. 1A, B and
Supplemen a y 1A, B). LQ addi ion in amino acid-s a ed cells
sus ained he le els o ATP a 4, 8, 12, 24, 48, and 72 h (Fig. 1C
and Supplemen a y 1C–E). By con as , he addi ion o me hio-
nine o a ginine, wo amino acids which ha e been epo ed o
ac i a e mTORC1 a sho - e m, did no sus ain he p oduc ion
o ATP (Supplemen a y 1F, G). Measu emen s o glu amine and
leucine le els in he medium indica ed ha hese amino acids
emained ully a ailable e en a e 72 h o LQ ea men (Sup-
plemen a y 1H–I). In ag eemen o hese p e ious obse a ions,
we also obse ed ha LQ ea men inc eased he basal espi a-
ion o U2OS cells compa ed o amino acid s a a ion (Fig. 1D).
The maximal oxygen consump ion a e (OCR) was also inc eased
in he p esence o LQ, which esul ed in an inc ease o ATP
p oduc ion linked o espi a ion (Fig. 1E). Thus, we concluded
ha long- e m glu aminolysis, e en in he absence o any o he
amino acid, was su ficien o inc ease he espi a o y capaci y and
he p oduc ion o ATP in he cell.
We nex in es iga ed i he obse ed p oduc ion o ATP in LQ-
ea ed cells had any impac on he phospho yla ion o AMPK.
The emo al o amino acids om he cul u e medium led o he
inhibi ion o mTORC1 in U2OS cells, as b oadly epo ed
be o e20, bu concomi an ly i led also o he phospho yla ion o
AMPK a h eonine 172 (Fig. 1F), in ag eemen wi h he obse ed
dec ease in ATP le els. Acco dingly, he phospho yla ion o
AMPK was e e ed by he addi ion o LQ, which again
co ela ed wi h he inc ease in ATP le els in his condi ion
(Fig. 1F). The LQ-induced inhibi ion o AMPK was obse ed a
24, 48, and 72 h (Fig. 1G). Long- e m ac i a ion o mTORC1
downs eam o AMPK du ing LQ ea men was confi med by
he phospho yla ion o i s a ge p o eins S6K, S6, and 4EBP1
(Fig. 1F, G), and by i s ansloca ion a he su ace o he lysosome
using LAMP2 (Fig. 1H, I) o CD63 (Supplemen a y 1J, K) as
lysosomal ma ke s. To confi m a mechanis ic ole o AMPK
du ing LQ-media ed mTORC1 ac i a ion a long- e m, we
in es iga ed he s a us o mTORC1 by he addi ion o he AMPK
ac i a o AICAR in LQ- ea ed U2OS and HCT116 cells. As
obse ed in Fig. 1J and Supplemen a y 1L, eac i a ion o AMPK
using AICAR was su ficien o p e en he LQ-media ed
ac i a ion o mTORC1. Simila esul s we e ob ained by
ans ec ing a MYC- agged cons i u i ely ac i e o m o AMPK
(CA-AMPK, kindly p o ided by P o . Benoi Violle , Pa is,
F ance) (Fig. 1K) o using o he AMPK ac i a o s, such as
me o min o A769662 (Fig. 1L). Su p isingly, he abla ion o
AMPK using AMPK−/−MEFs (again, kindly p o ided by P o .
ARTICLE NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4
2NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://d oi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions
Benoi Violle ) also p e en ed he LQ-media ed ac i a ion o
mTORC1 (Supplemen a y 1M). This esul sugges ed ha , in
addi ion o ac ing as a nega i e egula o o mTORC1, he
p esence o AMPK is necessa y o he connec ion be ween
glu amine me abolism and mTORC1, unde sco ing he complex-
i y o he me abolic adap a ions in hese ci cums ances.
We also confi med ha his AMPK- egula ed ac i a ion o
mTORC1 du ing long- e m LQ ea men had a di ec physio-
logical impac in he cell in e ms o au ophagy ac i a ion, a key
p ocess du ing glu amop osis induc ion12. As obse ed in Fig. 1M,
N, LQ- ea ed U2OS cells s ably exp essing he au ophagic
epo e GFP-LC3 displayed an inc ease o GFP-LC3 agg ega es
upon AMPK eac i a ion using AICAR, me o min, o A769662.
Consis en ly, eac i a ion o AMPK using hese di e en
ac i a o s also dec eased he le els o endogenous p62 and
LC3I, inc easing LC3II o ma ion (Fig. 1L), well-known au op-
hagy ma ke s. We also e alua ed he au ophagic flux using
bafilomycin A1 (Ba A1) o ap he o ma ion o au ophago-
somes. The inhibi ion o au ophagy using Ba A1 led o a g ea e
accumula ion o p62 and LC3I in cells ea ed wi h LQ. Howe e ,
35
35
45
70
55
55
70
70
70
35
25
25
55
70
15
10
55
55
55
70
55
15
10
70
55
70
55
35
25
25
35
35
35
*
0
5
10
15
20
OCR
(pmol/min/ g p o ein)
-LQ+
0
20
40
60
80
100
% GFP-LC3
agg ega es
-LQ AICAR
Me o min
A769662
+ LQ
-
-
-
+
-
-
-
+
-
-
-
+
* *
0 2 4 8 12 24
0.5
1.0
1.5
0
ATP/ADP a io
*
hou s
-AA
N.S
0.5
1.0
1.5
0
ATP/ADP a io
+-+-AA
2h 72h
C
*
AB
FG
-LQ
AICAR Me o min A769662--
+LQ
-AA
K
H
+AA
-AA
+LQ
LAMP2 mTORC1 me ged
LQ
Ac in
S6-pS235/236
S6K
S6
24h
S6K-pT389
AMPK-pT172
+-+-+-
48h 72h
-AA
AMPK
E
-
-AICAR
LQ
-AA
+
-
-
+
+
+
AMPK-pT172
AMPK
S6K-pT389
S6K
S6-pS235/236
S6
Ac in
IJ
L
Ac in
LC3I/II
p62
AICAR
Me o min
A769662
LQ
-
-
-
+
-
-
-
+
-
-
-
+
-
-
-
S6-pS235/236
S6
AMPK-pT172
AMPK
S6K-pT389
S6K LQ
AICAR
Ba A1
+
-
-
-
-
-
p62
LC3I/II
Ac in
+
+
-
-
+
-
+
-
+
-
-
+
+
+
+
-
+
+
-0.2
0.0
0.2
0.4
0.6
Pea son co ela ion
coe icien
Colocalisa ion be ween
LAMP2 and mTORC1
+
-
-
-
-
+
AA
LQ
**
M
LQ
AA
+
-
-
-
-
+
Rap o -pS792
S6
S6-pS235/236
AMPK-pT172
S6K-pT389
S6K
4EBP1-pT37/46
4EBP1
AMPK
Ac in
Rap o
0 102030405060708090100
0
10
20
30
40
Time (minu es)
OCR
(pmol/min/ go p o ein)
-AA
+LQ
Oligo
FCCP
Ro +
An A
***
***
CD
O
AA
LQ
CA-AMPK
-
-
-
+
+
-
-
-
-
+
+
-
EV
Ac in
S6
CA-AMPK
S6-pS235/236
N
-++++
-AA
70
70
55
70
70
35
35
0.5
1.0
1.5
2.0
2.5
0
ATP/ADP a io
-+ LQ
24h
-+
48h
-+
72h
-AA
***
135
55
35
135
70
70
55
70
70
35
25
35
55
35
48
15
15
NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4 ARTICLE
NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://doi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions 3
eac i a ion o AMPK using AICAR concomi an ly o Ba A1 and
LQ did no accumula e p62 a he same le el (Fig. 1O), and
es o ed he le els o LC3II. Al oge he , hese esul s showed ha
long- e m LQ ea men was su ficien o inc ease he le els o
ATP, leading o he inhibi ion o AMPK. AMPK inhibi ion in
LQ- ea ed cells led o he subsequen ac i a ion o mTORC1 and
he inhibi ion o au ophagy, wo majo p ocesses con olling
glu amop osis.
AMPK inhibi ion is necessa y o glu amop osis bo h in i o
and in i o. Nex , we assessed he ole o AMPK in he induc ion
o apop osis ollowing he ac i a ion o mTORC1 by glu ami-
nolysis du ing nu i ional imbalance, i.e., glu amop osis. In
ag eemen wi h ou p e ious esul s12, long- e m LQ ea men in
absence o any o he amino acid dec eased cell iabili y (Fig. 2A,
B). The LQ-induced inc ease in cell dea h was p e en ed by
AMPK eac i a ion using AICAR, me o min, o A769662
(Fig. 2A, B). Fu he , a clonogenic assay confi med ha he
pha macological eac i a ion o AMPK p e en ed LQ-media ed
cell dea h induc ion (Fig. 2C, D). AMPK eac i a ion was also
su ficien o p e en apop osis induced by LQ, as de e mined by
he dec ease in he double annexin V/PI s aining obse ed by flow
cy ome y (Fig. 2E, F), and by he dec ease in he p o-apop o ic
ma ke s clea ed PARP and clea ed caspase 3 (Fig. 2G). On he
con a y, AMPK abla ion did no inc ease he induc ion o cell
dea h du ing amino acid wi hd awal, as mTORC1 inhibi ion and
subsequen au ophagy ac i a ion unde hese condi ions p e-
en ed glu amop osis (Supplemen a y 2A). This esul u he
posi ioned mTORC1/au ophagy downs eam o AMPK du ing
glu amop osis induc ion. These esul s we e confi med using
ATG5−/−MEFs in which he impai men o au ophagy com-
ple ely abolished he capaci y o AICAR o p omo e cell su i al
du ing amino acids s a a ion (Fig. 2H, I) as au ophagy is
necessa y as su i al mechanism. Thus, AMPK inhibi ion
ups eam o mTORC1/au ophagy was necessa y o glu a-
mop osis in i o.
The pa icipa ion o AMPK in glu amop osis was also
in es iga ed in i o using implan ed umo s in xenog a mouse
models. Two di e en coho s o mice we e implan ed in he igh
do sal flank wi h HCT116 cells. Mice we e ea ed wi h ei he
ehicle o wi h a cell-pe meable αKG de i a i e, dime hyl-αKG
(DMKG) o induce glu amop osis in i o. Mice we e hen co-
ea ed wi h he mTORC1 inhibi o emsi olimus (a apamycin
de i a i e) o wi h me o min. Fi s , he induc ion o mTORC1
by DMKG was assessed by he analysis o S6 phospho yla ion in
hese umo s. As expec ed, he injec ion o DMKG in mice
induced he ac i a ion o mTORC1. A dec ease o S6 phospho -
yla ion and hus mTORC1 ac i a ion was obse ed in mice co-
ea ed wi h emsi olimus o me o min (Fig. 2J, K and
Supplemen a y 2B). Second, he in i o induc ion o glu amop-
osis was assessed by immunohis ochemical analysis o clea ed
caspase 3. We confi med ha DMKG ea men induced an
inc ease in apop o ic cell dea h in xenog a umo s in i o, as
assessed by caspase 3 clea age (Fig. 2L–M and Supplemen-
a y 2C). This is a alida ion o he physiological ele ance o
glu amop osis in i o, confi ming ha glu amop osis can be used
as a po en ial ool o induce umo al cell dea h in i o.
Fu he mo e, as shown in Fig. 2L–M and Supplemen a y 2C,
we also confi med ha he inhibi ion o mTORC1 using
emsi olimus p e en ed glu amop osis in i o, simila o wha
we obse ed in cul u ed cells12. This esul alida ed he
physiological model by which he abe an ac i a ion o mTORC1
du ing nu i ional imbalance induces apop o ic cell dea h in
animal models. Finally, and o no e, AMPK eac i a ion using
me o min which was confi med by IHC on T172 AMPK
phospho yla ion (Supplemen a y 2D, E) was simila ly su ficien
o p e en glu amop osis in i o (Fig. 2L–M and Supplemen-
a y 2C). Al oge he , hese da a confi med he necessi y o AMPK
inhibi ion o glu amop osis induc ion bo h in i o and in i o.
Glu aminolysis is no necessa y o LQ-media ed ATP p o-
duc ion. A e confi ming he capaci y o LQ o sus ain ATP
le els, we in es iga ed he possible ole o glu aminolysis, he
main pa hway o ca abolise glu amine, in he p oduc ion o ATP
in esponse o LQ. Fo his pu pose, and ollowing a loss-o -
unc ion app oach, we inhibi ed GLS ei he pha macologically o
gene ically (using wo di e en inhibi o s, BPTES and DON, o
RNA-in e e ence media ed knockdown). We hen in es iga ed
he capaci y o LQ ea men o induce ATP le els in GLS-
inhibi ed cells. Su p isingly, he inhibi ion o GLS did no p e en
he p oduc ion o ATP in esponse o LQ in U2OS, HCT116 o
HEK293 cells (Fig. 3A, B and Supplemen a y 3A, B). Confi ming
Fig. 1 Glu aminolysis sus ains he p oduc ion o ATP o inhibi AMPK and o ac i a e mTORC1. A ATP/ADP a io o U2OS cells incuba ed in he
p esence o he absence o all amino acids o 2 o 72 h. Fed cells (C) a e used as con ol. BATP/ADP a io o U2OS cells incuba ed in absence o amino
acid o he indica ed ime. CATP/ADP a io o amino acid-s a ed U2OS cells incuba ed in he p esence o absence o LQ du ing he indica ed imes.
DOCR analysis by Seaho se® echnology o amino acid-s a ed U2OS cells incuba ed in he p esence (blue) o absence (pu ple) o LQ du ing 72 h. OCR
was measu ed ei he in basal condi ions o a e he injec ion o oligomycin, FCCP, and o enone/an imycin A. Da a a e mean ± SEM o h ee biologically
independen expe imen s pe o med wi h fi e eplica es. EBasal espi a ion used o d i e ATP p oduc ion as de e mined by OCR quan ifica ion o da a
ob ained in (D). FImmunoblo o mTORC1 ac i i y ma ke s (S6K, S6, and 4EBP1 phospho yla ion) and AMPK phospho yla ion o U2OS cells incuba ed
wi h o wi hou amino acids, in he p esence o absence o LQ du ing 72 h. GImmunoblo o mTORC1 ac i i y ma ke s (S6K and S6 phospho yla ion) and
AMPK phospho yla ion o amino acid-s a ed U2OS cells incuba ed in he p esence o absence o LQ du ing 24, 48, o 72 h. HImmunofluo escence
mic oscopy cap ions o U2OS cells incuba ed wi h o wi hou amino acids, in he p esence o absence o LQ du ing 72 h. Cells we e s ained agains LAMP2
(lysosomal ma ke , ed), mTORC1 (g een) and DAPI (blue). Scale ba ep esen s 10 µm. IQuan ifica ion o he colocaliza ion be ween LAMP2 and
mTORC1 as shown in (H). Pe son’s R alue was e alua ed using ImageJ coloc2 plugin on 25 ROI in h ee biologically independen expe imen s (75 ROI in
o al pe condi ion). JImmunoblo o mTORC1 ac i i y ma ke s (S6K and S6 phospho yla ion) and AMPK phospho yla ion o amino acid-s a ed U2OS
cells incuba ed in he p esence o absence o LQ, wi h o wi hou AICAR, du ing 72 h. KImmunoblo analysis o mTORC1 ac i i y ma ke (S6
phospho yla ion) o U2OS cells exp essing a myc- agged, cons i u i ely ac i e AMPK mu an in he p esence o absence o amino acids and LQ as
indica ed. LImmunoblo o au ophagy (p62 and LC3-I/II) and mTORC1 (S6K and S6 phospho yla ion) ma ke s o amino acid-s a ed U2OS cells incuba ed
in absence o p esence o LQ, AICAR, me o min o A769662 o 72 h, as indica ed. MFluo escence mic oscopy cap ions o GFP-LC3 exp essing amino
acid-s a ed U2OS cells incuba ed in he p esence o absence o LQ, wi h o wi hou AICAR, me o min o A769662 o 72 h. Au ophagosome o ma ion
upon GFP-LC3 agg ega ion was assayed using con ocal mic oscopy. The scale ba ep esen s 10 µm. NQuan ifica ion o he numbe o GFP-LC3 do s pe
cell o cap ions ob ained in (M). >100 cells we e coun ed pe expe imen . OImmunoblo analysis o au ophagy ma ke s (p62 and LC3I/II) o U2OS cells
ea ed wi h LQ, AICAR and/o Bafilomycin A1 as indica ed o 72 h. G aphs show mean alues ± SEM (n=3 biologically independen expe imen s). *p<
0.05 (ANOVA analysis ollowed by a pos hoc Bon e oni es ). Sou ce da a a e p o ided as a Sou ce Da a file.
ARTICLE NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4
4NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://d oi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions
his esul , silencing GDH did no impai he p oduc ion o ATP
in LQ- ea ed cells ei he (Fig. 3C). Thus, glu aminolysis did no
show a necessa y ole o he capaci y o LQ o sus ain ATP le els.
Fu he confi ming his conclusion, we also obse ed ha he
addi ion o DMKG o amino acid-s a ed cells did no sus ain he
p oduc ion o ATP o a simila ex en han LQ ea men (Fig. 3D
and Supplemen a y 3C, D). Hence, glu aminolysis was nei he
necessa y no su ficien o sus ain ATP p oduc ion. The lack o
dec ease on ATP le els upon glu aminolysis inhibi ion co ela ed
wi h an absence o AMPK eac i a ion (Fig. 3E, F), confi ming
ha glu aminolysis did no media e he connec ion be ween LQ
and he ATP/AMPK axis.
0
10
20
30
40
50
Cell dea h %
+
--
-AA
AICAR
-
++
--
--
+
MEF
ATG5-/-
*
MEF
ATG5+/+
*
0
20
40
60
80
100
Viabili y %
-
-+
--
+LQ
AICAR
+
+
*
MET
TEM
Vehicle
-DMKG +DMKG
S6-pS235/236
MET
TEM
Vehicle
-DMKG +DMKG
cCas3
+LQ-LQ
AICAR Me o min A769662--
A
-
-
+
-
-
+
+
+
LQ
AICAR
C
+LQ
-LQ
AICAR
Me o min A769662
-
F
H
I
-
+
+
+
-AA
-
-
LQ
AICAR
S6
cPARP
cCasp 3
S6K-pT389
S6K
S6-pS235/236
AMPK-pT172
AMPK
Ac in
G
B
DE
-LQ
AICAR
Me o min
A769662
+LQ
-
-
-
+
-
-
+
-
-
-
+
-
-
-
L
M
1.84%1.84% 12.04%12.04%
BAX
0
1
2
3
4
5
S6-pS235/236
IHC isual sco e
-
-
-
-
+
-
DMKG
Temsi olimus
Me o min
-
-
+
+
-
-
+
+
-
+
-
+
**
0
1
2
3
4
5
cCaspase 3
IHC isual sco e
-
-
-
-
+
-
DMKG
Temsi olimus
Me o min
-
-
+
+
-
-
+
+
-
+
-
+
**
cPARP
AA
AICAR
+
-
cCasp 3
ATG5-ATG12
Ac in
-
-
-
+
+
-
MEF ATG5+/+
-
-
-
+
MEF ATG5-/-
J
K
Rap o -pS792
Rap o
-AA
-AA
0
10
20
30
40
50
% La e apop o ic cells
-
* *
100
15
20
70
70
55
135
135
70
70
25
25
55
70
15
25
35
55
0
50
100
150
200
250
Numbe o colonies
-
-+
--
++
+LQ
AICAR
* *
82.34%82.34% 3.78%3.78%
4.81%4.81% 29.93%29.93%
64.29%64.29% 0.97%0.97%
0.90%0.90% 3.41%3.41%
95.04%95.04% 0.65%0.65%
1.44%1.44% 5.21%5.21%
1.00%1.00%92.35%92.35%
0.78%0.78% 4.90%4.90%
93.04%93.04% 1.28%1.28%
NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4 ARTICLE
NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://doi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions 5

Glu amine is su ficien o sus ain ATP le els, bu no o ac i-
a e mTORC1. Du ing LQ ea men , leucine addi ion plays an
allos e ic ac i a ion ole o GDH, necessa y o αKG p oduc ion5,21.
As we concluded ha glu aminolysis did no a ec he ATP/AMPK
pa hway, we hen assessed whe he leucine had any ole in ATP le els
du ing LQ ea men . Fo his pu pose, we ea ed amino acid-s a ed
U2OS, HCT116 and HEK293A cells wi h glu amine o leucine, and
subsequen ly measu ed ATP le els. Unlike leucine, glu amine alone
was su ficien o sus ain ATP le els in all h ee cell lines (Fig. 4Aand
Supplemen a y 4A, B). Confi ming his esul , we obse ed ha basal
espi a ion o he cells in amino acid s a a ion did no change in
esponse o leucine addi ion (Fig. 4B). Howe e , cells in he p esence
o glu amine showed a significan inc ease o he espi a o y ac i i y,
independen ly o he p esence o he absence o leucine (Fig. 4B).
Simila ly, glu amine su ficiency induced a clea inc ease o he OCR
linked o ATP p oduc ion by he mi ochond ia o mee he ene ge ic
needs o he cell (Fig. 4C). This inc ease o ATP le els by glu amine
alone also co ela ed wi h an inc ease o endoplasmic e iculum cal-
cium pool due o he ac i i y o he sa coendoplasmic e iculum Ca2+
ATPase (SERCA) (Supplemen a y 4C). In a simila manne , and
co ela ing wi h ATP le els, a s ong dec ease o AMPK phospho -
yla ion was obse ed upon glu amine addi ion o o he wise amino
acid-deple ed cells, while he addi ion o leucine did no educe he
phospho yla ion s a us o AMPK (Fig. 4D and Supplemen a y 4D).
Fig. 2 AMPK inhibi ion is necessa y o glu amop osis bo h in i o and in i o. A Rep esen a i e mic oscopy images o amino acid-s a ed U2OS cells
incuba ed in absence o p esence o LQ, AICAR, me o min, o A769662 o 72 h, as indica ed. Scale ba ep esen s 100 µm. BCell iabili y as es ima ed by
a ypan blue exclusion assay o amino acid-s a ed U2OS cells incuba ed in he p esence o absence o LQ and AICAR du ing 72 h. CRep esen a i e
images o clonogenic assay o U2OS cells ea ed as in (B). DColony quan ifica ion o images ob ained in (C). EQuan ifica ion o la e apop osis popula ion
o h ee biologically independen expe imen s (double posi i e, annexin V and PI) as ob ained in (F) o he indica ed condi ion. FFlow cy ome y analysis
o annexin V/PI s aining o amino acid-s a ed U2OS cells incuba ed wi h o wi hou LQ in combina ion wi h AICAR, me o min, o A769662 du ing 72 h.
GImmunoblo o he p o-apop o ic ma ke s (BAX, clea ed caspase 3, and clea ed PARP), mTORC1 ac i i y ma ke s (Rap o , S6K, and S6
phospho yla ion), and AMPK phospho yla ion o amino acid-s a ed U2OS cells incuba ed in he p esence o absence o LQ, wi h o wi hou AICAR, du ing
72 h. HCell iabili y as es ima ed by a ypan blue exclusion assay o ATG5+/+and ATG5−/−MEFs incuba ed in he p esence o absence o amino acids
(AA) and AICAR du ing 72 h. IImmunoblo analysis o p o-apop o ic ma ke s o ATG5+/+and ATG5−/−MEFs incuba ed as in (H). J–LRep esen a i e
immunohis ochemis y mic oscopy pic u es (×40 magnifica ion) o xenog a umo s o mice ea ed as indica ed (TEM: Temsi olimus; MET: Me o min).
Samples we e s ained agains S6-pS235/236 (J) and clea ed caspase 3 (L). Scale ba s ep esen 100 µm. K–MIHC isual sco e o S6-pS235/236 (K) and
caspase 3 (M) o images om (J) and (L), espec i ely. The uppe and lowe limi s o he boxes ep esen qua iles, wi h he line wi hin he boxes
indica ing he median and he whiske s showing he ex emes (n≥10 images pe ea men ). G aphs show mean alues ± SEM (n=3 biologically
independen expe imen s). *p< 0.05 (ANOVA analysis ollowed by a pos hoc Bon e oni es ). Sou ce da a a e p o ided as a Sou ce Da a file.
0.5
1.0
1.5
2.0
0
ATP/ADP a io
-
+
-
-
-
+
+
+
-
LQ
siC l
siGDH
+
-
+
N.S
0.5
1.0
1.5
2.0
0
ATP/ADP a io
-
+
-
-
-
+
+
+
-
LQ
siC l
siGLS1
+
-
+
N.S
0.5
1.0
1.5
2.0
0
ATP/ADP a io
-
-
-
+
-
-
+
+
-
LQ
BPTES
DON
+
-
+
*
-
-
LQ
BPTES
Ac in
S6
S6K
AMPK
AMPK-pT172
S6-pS235/236
S6K-pT389
-AA
+
-
+
+DON
LQ
-AA
-
-
Ac in
AMPK
AMPK-pT172
S6-pS235/236
S6
S6K-pT389
S6K
+
-
+
+
A
BC
DE F
Ac in
GLS1
siC l
siGLS1
Ac in
GDH
siC l
siGDH
55
55 35
35
70
70
55
70
70
25
25
55
70
70
55
70
70
35
25
25
45
0.5
1.0
1.5
2.0
0
ATP/ADP a io
-
-
+
-
-
+
LQ
DMKG
**
Fig. 3 Glu amine is su ficien o sus ain ATP le els, bu no o ac i a e mTORC1. A ATP/ADP a io o amino acid-s a ed U2OS cells incuba ed in he
p esence o absence o LQ wi h o wi hou BPTES o DON du ing 72 h. B,CGLS o GDH exp essions we e knocked down using small in e e ing RNA
(siRNA) in U2OS cells o 48 h. Cells we e hen ea ed wi h LQ o 72 h and he ATP/ADP a io was measu ed. Sc amble non- a ge ing siRNA was used as
a con ol. Immunoblo s o GLS o GDH le els a e p esen ed as a con ol o he knockdown. DATP/ADP a io o amino acid-s a ed U2OS cells ea ed wi h
LQ o DMKG o 72 h. E,FImmunoblo o amino acid-s a ed U2OS cells ea ed wi h o wi hou LQ in combina ion wi h BPTES (E) o DON (F) o 72 h.
Ac i i y ma ke s o AMPK and mTORC1 we e analysed. G aphs show mean alues ± SEM (n=3 biologically independen expe imen s). *p< 0.05 (ANOVA
analysis ollowed by a pos hoc Bon e oni es ). Sou ce da a a e p o ided as a Sou ce Da a file.
ARTICLE NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4
6NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://d oi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions
Simila o wha was obse ed o LQ ea men (Fig. 3B, C), glu a-
minolysis inhibi ion by GLS1 o GDH knockdown did no a ec he
capaci y o glu amine su ficiency o induce ATP le els in U2OS and
HCT116, confi ming ha glu aminolysis was no necessa y o
glu amine-dependen ATP p oduc ion (Supplemen a y 4E–H). I is
wo h no ing ha nei he glu amine alone no leucine alone (in
absence o any o he amino acid) we e su ficien o ac i a e mTORC1
a e 72 h in ei he U2OS o HCT116 cells (Fig. 4D and Supple-
men a y 4D). Consis en ly, mTORC1 localiza ion a he su ace o he
lysosome was only obse ed in he p esence o bo h leucine and
glu amine, bu i was no obse ed i only one o hese amino acids
we e added (Fig. 4E, F and Supplemen a y 4I, J). In a simila manne
0
5
10
15
20
OCR
(pmol/min/ gp o ein)
-
-
+
-
L
Q
-
+
+
+
*
RagB WT
Q
RagB GTP
Ac in
S6
S6K
S6-pS235/236
S6K-pT389
-AA
-
+
-
-
-
+
+
+
-
+
-
+
L
Q
Ac in
S6
S6K-pT389
S6K
-
-
+
-
-
+
+
+
AMPK-pT172
AMPK
-AA
S6-pS235/236
A
D
C
E
G
F
JH
0 102030405060708090100
0
10
20
30
40
Time (minu es)
OCR (pmol/min/μg o p o ein)
S
L
Q
LQ
Oligo
FCCP Ro +
An A
B
*-
-
Q
L
+
-
+
+
Ac in
LC3I/II
p62
I
-
-
+
-
+
+
Q
L
-0.2
0.0
0.2
0.4
0.6
Pea son co ela ion
coe icien
Colocalisa ion be ween
LAMP2 and mTORC1
-
-
+
-
-
+
L
Q
+
+
*
L
Q
-
LQ
LAMP2 mTORC1 me ged
*
0.5
1.0
1.5
2.0
2.5
3.0
0
ATP/ADP a io
-
-
+
-
-
+
L
Q
+
+
0
20
40
60
80
100
120
% GFP-LC3
agg ega es
-
-
+
-
+
+
Q
L
55
70
55
70
70
70
25
35
25
55
70
15
55
70
70
35
35
25
35
Fig. 4 Glu amine me abolism ac i a es mTORC1 ollowing wo pa allel, necessa y b anches. A ATP/ADP a io o amino acid-s a ed U2OS cells
incuba ed wi h leucine and/o glu amine du ing 72 h. BOCR analysis by Seaho se® echnology o amino acid-s a ed (pu ple) U2OS cells incuba ed wi h
leucine ( ed) and/o glu amine (Q g een, LQ blue) du ing 72 h. OCR was measu ed ei he in basal condi ions o a e he injec ion o oligomycin, FCCP, and
o enone/an imycin A. Da a a e mean ± SEM o h ee biologically independen expe imen s pe o med wi h fi e eplica es. CBasal espi a ion used o d i e
ATP p oduc ion as de e mined by OCR quan ifica ion o da a ob ained in (B). DImmunoblo o mTORC1 ac i i y ma ke s (S6K and S6 phospho yla ion)
and AMPK phospho yla ion o amino acid-s a ed U2OS cells incuba ed wi h leucine and/o glu amine du ing 72 h. EImmunofluo escence mic oscopy
cap ions o U2OS cells incuba ed wi h leucine and/o glu amine du ing 72 h. Cells we e s ained agains LAMP2 (lysosomal ma ke , ed), mTORC1 (g een)
and DAPI (blue). Scale ba ep esen s 10 µm. FQuan ifica ion o he colocaliza ion be ween LAMP2 and mTORC1 as shown in (E). Pe son’s R alue was
e alua ed using ImageJ coloc2 plugin on 25 ROI in h ee biologically independen expe imen s (75 ROI in o al pe condi ion). GFluo escence mic oscopy
cap ions o GFP-LC3 exp essing amino acid-s a ed U2OS cells incuba ed in he p esence o glu amine and/o leucine du ing 72 h. Au ophagosome
o ma ion upon GFP-LC3 agg ega ion was assayed using con ocal mic oscopy. The scale ba ep esen s 10 µm. HQuan ifica ion o he numbe o GFP-LC3
do s pe cell o cap ions ob ained in (G). >100 cells we e coun ed pe expe imen . IImmunoblo o au ophagy (p62 and LC3-I/II) ma ke s o amino acid-
s a ed U2OS cells incuba ed in he p esence o glu amine and/o leucine du ing 72 h. JU2OS cells we e ans ec ed wi h RagB WT plasmid o RagB 54 L
plasmid. Amino acid-s a ed cells we e hen incuba ed wi h o wi hou glu amine o 72 h. Downs eam a ge s o mTORC1 (S6K and S6 phospho yla ion)
we e assessed by immunoblo . G aphs show mean alues ± SEM (n=3 biologically independen expe imen s). *p< 0.05 (ANOVA analysis ollowed by a
pos hoc Bon e oni es ). Sou ce da a a e p o ided as a Sou ce Da a file.
NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4 ARTICLE
NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://doi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions 7
han o LQ, glu amine le els we e measu ed in he cell cul u e
medium du ing glu amine su ficiency condi ion o e i y he a ail-
abili y o his amino acid e en a e 3 days o ea men
(Supplemen a y 4K). Au ophagy analysis downs eam o mTORC1
also confi med he inabili y o glu amine su ficiency o ac i a e
mTORC1. Thus, long- e m glu amine addi ion did no inhibi
au ophagy ac i a ion du ing amino acid s a a ion, as assessed by he
au ophagy epo e GFP-LC3 by con ocal mic oscopy, and by
immunoblo analysis o he au ophagic endogenous ma ke s p62 and
LC3I/II. In con as , and as shown p e iously, he addi ion o bo h
leucine and glu amine s ongly dec eased he numbe o GFP-LC3
agg ega es and inc eased p62 le els in U2OS cells (Fig. 4G–I).
0.5
1.0
1.5
2.0
0
ATP/ADP a io
Q
BPTES
siC l
siASNS
-
-
-
-
+
-
+
-
+
+
+
+
N.S *
GABA
Aspa agine
Glu ama e
Aspa a e
Ci a e
NAA
Alanine
Glu amine
Mala e
Py u a e
Se ine
Isoleucine
O ni hine
Glucose
Palmi a e
S ea a e
Lac a e
So bi ol
F uc ose
αKG
Leucine
KMV/KIC
GSH
Glycine
Thymine
Lysine
Th eonine
Me hionine
Phenylalanine
Valine
A ginine
Aconi a e
2
1
0
-1
-2
Row Z-sco e
AMPK-pT172
Q
BPTES
-
-
+
-
+
+
-
-
+
-
+
+
AMPK
ASNS
Ac in
siC l siASNS
+
-
-
-
-
-
AA
L
Q
-
+
-
-
-
+
-
+
+
siC l
ASNS
Ac in
siASNS
AB
C
DE
F
G
HI
+ L + [U13C]Q+AA + [U
13C]Q + [U13C]Q
+AA + [15N2]Q
+
[15N2]Q
+L +
[15N2]Q
AA
BPTES
S6
Ac in
S6-pS235/236
ASNS
-
-
-
+
+
-
+
+
-
-
-
+
+
-
+
+
siC l siASNS
AMPK
AMPK-pT172
S6K
S6K-pT389
4EBP1
4EBP1-pT37/46
0
1106
2106
3106
Ion coun s
Glu
m+5
(17)
(30)
(22)
**
0
110
5
210
5
Ion coun s
GABA
m+4
(7)
(45)
(32)
**
0
2105
4105
6105
8105
Asn
m+4
(34)
(28)
(33)
* *
0
1104
2104
3104
4104
5104
Asp
m+4
(14)
(21)
(15)
**
0
2104
4104
6104
8104
Ci a e
m+4
(7)
(20)
(3)
**
1104
2104
0
Mala e
m+4
(8)
(21)
(8)
*
0
1106
2106
3106
4106
Ion coun s
(35)
(59)
(51)
Glu
15N
**
0
2104
4104
6104
3105
4105
5105
(33)
(59)
(35)
Asp
15N
**
0
2103
4103
1105
2105
3105
4105
(17)
(41)
(23)
Asn
15N2
**
0
1104
2104
3104
1105
2105
(39)
(21) (24)
Asn
15N
*
0.5
1.0
1.5
2.0
2.5
0
ATP/ADP a io
-
-
+
-
-
+
Q
DMKG
+
+
* *
55
35
70
55
55
55
55
35
70
100
70
35
25
25
25
55
55
70
55
55
0
5
10
15
20
ASNS mRNA le els (RU)
N.S
*
*
-
-
-
-
+
-
-
-
+
+
-
-
+
+
+
-
-
-
-
+
+
-
-
+
Q
L
BPTES
DMKG
ARTICLE NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4
8NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://d oi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions
I is well documen ed ha he e-addi ion o leucine alone o
10–30 m a e a sho - e m amino acid s a a ion (1–2h) is
su ficien o ac i a e mTORC122–25. Howe e , ou da a showed
ha his eac i a ion o mTORC1 disappea ed a e longe
s a a ion pe iods (>4 h o amino acid s a a ion) (Supplemen-
a y 4L), which co ela ed wi h a s ong dec ease o glu amine
in acellula le els ollowing 4 h o amino acid s a a ion
(Supplemen a y 4M). In con as , e-addi ion o bo h leucine
and glu amine ac i a ed mTORC1 e en a e longe pe iods o
amino acid s a a ion (Supplemen a y 4N). These esul s suppo
a model by which leucine alone ac i a es mTORC1 a e sho -
e m amino acid s a a ion hanks o he emaining pool o
in acellula glu amine. I has been p oposed elsewhe e ha
in acellula glu amine is equi ed o leucine up ake26. Howe e ,
ou da a showed ha , in ou se up, in acellula le els o leucine
we e no significan ly a ec ed by he co-addi ion o glu amine in
s a ed cells (Supplemen a y 4O), sugges ing ha in acellula
glu amine was no necessa y o leucine up ake. Ra he , ou da a
sus ained a model by which leucine only ac i a es mTORC1 by
coope a ing wi h glu amine o ac i a e glu aminolysis.
Glu amine me abolism ac i a es mTORC1 ollowing wo pa -
allel, necessa y b anches. In ligh o ou esul s, we in es iga ed
he possibili y ha wo pa allel pa hways connec glu amine and
mTORC1: a p e iously epo ed glu aminolysis-dependen
b anch; and an ATP-dependen bu GLS/GDH-independen
b anch. We p e iously showed ha glu aminolysis ac i a es
mTORC1 ia RagB. Thus, αKG gene a ion by glu aminolysis
inc eases GTP loading o RagB, leading o mTORC1 ansloca-
ion o he su ace o he lysosome5. I is documen ed ha he
o e exp ession o a cons i u i ely GTP-bound mu an a ian o
RagB (RagB-GTP mu an ) is su ficien o o ce he ansloca ion
o mTORC1 o he su ace o he lysosome e en in amino acid-
s a ed cells27. The exp ession o his RagB-GTP mu an is su -
ficien o ac i a e mTORC1 in amino acid-s a ed cells a sho
imes (1–2h)
5,27. Howe e , we obse ed ha RagB-GTP mu an
exp ession did no ac i a e mTORC1 in amino acid-s a ed cells
a longe imes (72 h, Fig. 4J and Supplemen a y 4P). S ikingly,
he combina ion o bo h RagB-GTP exp ession and glu amine
addi ion induced he ull ac i a ion o mTORC1 a 72 h, as
de e mined by he phospho yla ion o he downs eam a ge s
S6K and S6 (Fig. 4J). Simila esul s we e ob ained using a RagD-
GDP mu an , which also up egula es he lysosomal ansloca ion
o mTORC1 (Supplemen a y 4P). Opposi e o wha we obse ed
wi h glu amine, he combina ion o bo h RagB-GTP and leucine
did no induce he ac i a ion o mTORC1 signaling (Supple-
men a y 4Q). These esul s u he sus ained a model by which
glu aminolysis (ac i a ed by bo h glu amine and leucine) is
necessa y o p oduce αKG and o induce he Rag-media ed
ansloca ion o mTORC1 a he su ace o he lysosome. On he
o he hand, glu amine, independen ly o GLS and GDH, sus ains
he p oduc ion o ATP and he inhibi ion o AMPK, necessa y o
he ull ac i a ion o mTORC1 a he su ace o he lysosome,
media ed by he mTORC1 coac i a o Rheb28.
ASNS and GABA shun a e al e na i e pa hways o me abolize
glu amine. So a , ou esul s indica ed ha glu amine su ficiency
p oduces ATP h ough mi ochond ial ac i i y, bu glu aminolysis
was no necessa y o his. To unco e he pa hways in ol ed in
glu amine-induced ATP p oduc ion, we pe o med a me abo-
lomics analysis (Fig. 5A). We obse ed ha , du ing amino acid
s a a ion (72 h), he le els o a la ge numbe o me aboli es
d opped. Addi ion o leucine did no es o e he as majo i y o
hem, again confi ming ha leucine has mos ly an allos e ic ole,
no ha ing a significan impac on he me abolism o he cell by
i sel . In con as , he addi ion o glu amine changed comple ely
he pa e n, allowing us o iden i y a g oup o me aboli es which
le els aised e y high, pa icula ly GABA, aspa agine, glu ama e
and aspa a e (Fig. 5A and Supplemen a y 5A–F). This esul
indica ed an inc eased syn hesis o aspa agine, media ed by he
ac i i y o ASNS, esponsible o he p oduc ion o aspa agine
and glu ama e om glu amine and aspa a e (Supplemen-
a y 5G). To expe imen ally es his possibili y, we aced glu-
amine du ing glu amine su ficiency in U2OS cells using [U13C]-
glu amine. As expec ed, we obse ed an accumula ion o bo h
glu ama e m +5 and aspa agine m +4 in he in acellula ex ac
o U2OS cells a e 72 h o ea men . Fu he , glu amine su fi-
ciency also p oduced m +4 labeled aspa a e (Fig. 5B), indica ing
ha aspa a e used by ASNS was indeed o igina ed ( ecycled)
om he glu ama e h ough oxaloace a e gene a ion a he TCA
cycle. Glu ama e p oduced by ASNS was u he con e ed in o
aspa a e likely h ough he ac i i y o he enzyme aspa a e
ansaminase (GOT1) (Supplemen a y 5G). In pa allel, we also
incuba ed amino acid-s a ed cells in he p esence o [15N2]-
glu amine o 72 h. In ag eemen wi h ou p e ious obse a ions,
we obse ed a la ge accumula ion o 15N-aspa agine, confi ming
he implica ion o ASNS in he me aboliza ion o glu amine
du ing glu amine su ficiency (Fig. 5C). The p oduc ion o bo h
15N-aspa a e and 15N-aspa agine in his condi ion finally
Fig. 5 ASNS and GABA shun a e al e na i e pa hways o me abolize glu amine. A Hea map ep esen a ion o me aboli e le els, as de e mined by
LC–MS analysis, in amino acid-s a ed U2OS cells incuba ed wi h glu amine and/o leucine. The hea map was c ea ed wi h Me aboAnalys 3.0 wi h he
o al pools o he de ec ed me aboli es. B13C-labeled me aboli e le els, as de e mined by LC–MS analysis, in U2OS cells incuba ed wi h o wi hou all
amino acids o leucine alone as indica ed, in he p esence o (U)-13C-glu amine du ing 72 h. To al ion coun s o glu ama e m +5, aspa agine m +4,
aspa a e m +4, GABA m +4, ci a e m +4, and mala e m +4 a e g aphed. The pe cen age o labeling wi h espec o o al me aboli e le els is shown in
pa en hesis o each me aboli e. C15N-labeled me aboli e le els, as de e mined by LC–MS analysis, in U2OS cells incuba ed wi h o wi hou all amino acids
o leucine alone as indica ed, in he p esence o 15N2-glu amine du ing 72 h. Ion coun s o 15N -glu ama e, 15N -aspa agine, 15N aspa a e, and 15N2-
aspa agine a e plo ed. The pe cen age o labeling ela i e o he o al me aboli e is shown in pa en hesis o each me aboli e. DASNS exp ession was
knocked down using small in e e ing RNA (siRNA) in U2OS cells du ing 48 h. Cells we e hen ea ed wi h glu amine and BPTES as indica ed o 72 h and
he ATP/ADP a io was measu ed. Sc amble non- a ge ing siRNA was used as a con ol. Immunoblo o ASNS le els is p esen ed as a con ol o he
knockdown. EImmunoblo analysis o AMPK phospho yla ion in U2OS cells ea ed as in (D). FRela i e mRNA exp ession le els o ASNS as de e mined
by qPCR in amino acid-s a ed U2OS cells incuba ed as indica ed du ing 72 h. GATP/ADP a io o amino acid-s a ed U2OS cells incuba ed wi h
glu amine and/o DMKG du ing 72 h. HImmunoblo analysis o mTORC1 downs eam a ge (S6K, S6, and 4EBP1 phospho yla ion) and AMPK
phospho yla ion in U2OS cells incuba ed du ing 24 h in he p esence o all amino acids wi h dual inhibi ion o GLS and/o ASNS. ASNS exp ession was
knocked down using small in e e ing RNA (siRNA) in U2OS cells du ing 48 h. Sc amble non- a ge ing siRNA was used as a con ol. Cells we e hen
incuba ed in he p esence o absence o all amino acids and/o BPTES as indica ed du ing 24 h. ISchema ic ep esen a ion o he wo b anches model
connec ing glu amine me abolism and mTORC1 signaling. G aphs show mean alues ± SEM (n=3 biologically independen expe imen s). *p< 0.05
(ANOVA analysis ollowed by a pos hoc Bon e oni es ). Sou ce da a a e p o ided as a Sou ce Da a file.
NATURE COMMUNICATIONS | h ps://doi.o g/10.1038/s41467-021-25079-4 ARTICLE
NATURE COMMUNICATIONS | (2021) 12:4814 | h ps://doi.o g/10.1038/s41467-021-25079-4 | www.na u e.com/na u ecommunica ions 9