P oc. Na l. Acad. Sci. USA
Vol. 96, pp. 7161–7166, June 1999
Biochemis y
Glu amine syn he ase inac i a ion by p o ein–p o ein in e ac ion
(ni ogen me abolism
y
cyanobac e ia
y
Synechocys is 6803
y
enzyme egula ion)
MARIO GARCI
´A-DOMI
´NGUEZ,JOSE
´C. REYES,AND FRANCISCO J. FLORENCIO*
Ins i u o de Bioquı´mica Vege al y Fo osı´n esis, Uni e sidad de Se illa-Consejo Supe io de In es igaciones Cien ı´ icas, Ame´ ico Vespucio syn, E-41092
Se illa, Spain
Edi ed by Robe Haselko n, The Uni e si y o Chicago, Chicago, IL, and app o ed Ap il 26, 1999 ( ecei ed o e iew Feb ua y 22, 1999)
ABSTRACT Glu amine syn he ase (GS; EC 6.3.1.2) is he
pi o al enzyme o ni ogen me abolism in p oka yo es. Con-
ol o bac e ial GS ac i i y by e e sible adenylyla ion has
p o ided one o he classical pa adigms o signal ansduc ion
by cyclic cascades. By con as , in he p esen wo k we show
ha cyanobac e ial GS is con olled by a di e en mechanism
ha in ol es he in e ac ion o wo inhibi o y polypep ides
wi h he enzyme. Bo h inac i a ing ac o s (IFs), named IF7
and IF17, a e equi ed in i o o comple e GS inac i a ion.
Inac i e GS-IF7 and GS-IF17 complexes we e econs i u ed in
i o by using Esche ichia coli-exp essed pu i ied p o eins. Ou
da a sugges ha con ol o GS ac i i y is exe ed by egu-
la ing he le els o IF7 and IF17.
Glu amine syn he ase (GS; EC 6.3.1.2) is he key enzyme o
ni ogen me abolism in p oka yo es, and i is subjec ed o a
sophis ica ed egula o y con ol ha has been s udied ex en-
si ely ( e iewed in e s. 1–3). Depending on he o ganism, he
GS ac i i y is egula ed by a leas one, and o en by all, o he
ollowing h ee le els: eedback inhibi ion o he ac i i y,
e e sible co alen modi ica ion o he enzyme, and ansc ip-
ional egula ion o he s uc u al gene. In mos o he sys ems
s udied, con ol o GS ac i i y esponds o ca bon and ni ogen
signals. In he p esence o abundan ca bon sou ces, ni ogen
de iciency esul s in a high le el o GS ac i i y. On he
con a y, when ni ogen sou ce is abundan , GS ac i i y is
down- egula ed.
GS ype I ( e e ed o as GS), he mos common ype o GS
in p oka yo es, is a dodecame ic enzyme composed o 12
iden ical subuni s (M
, abou 55,000) a anged in wo supe -
imposed hexagonal ings (4, 5). Regula ion o en e obac e ial
GS by e e sible adenylyla ion has p o ided one o he classical
pa adigms o signal ansduc ion by cyclic cascades. S ad man
and colleagues disco e ed ha GS om Esche ichia coli exis s
in wo in e con e ible o ms, an adenylyla ed o m ha is
highly sensi i e o eedback inhibi ion and a deadenylyla ed
o m ha is ela i ely insensi i e o eedback inhibi ion ( e-
iewed in e s. 1, 6, and 7). Adenylyla ion in ol es he ans e
o an adenylyl g oup om ATP o a Ty esidue on each o he
12 subuni s o he enzyme. The adenylyla ion s a e o GS is
con olled by a bicyclic cascade in ol ing wo bi unc ional
p o eins, he adenylyl ans e ase and he u idylyl ans e ase,
and he signal- ansducing p o ein PII. F om an e olu iona y
poin o iew, con ol o GS ac i i y by adenylyla ion has been
a e y success ul sys em because i is p esen in nume ous
eubac e ial g oups (2). In con as , GSs om Bacillus,Clos-
idium, and cyanobac e ia a e no modi ied by adenylyla ion
(8, 9). Al hough egula ion o GS ac i i y in Bacillus and
Clos idium seems o be mos ly a he le el o eedback
inhibi ion, an in i o e e sible inac i a ion o GS has been
epo ed in he cyanobac e ium Synechocys is sp. PCC 6803
(10–12). As o he adenylyla ion o en e obac e ial GS,
con ol o Synechocys is GS depends on he ni ogen–ca bon
balance o he cell. Thus, addi ion o ammonium o ni a e-
g owing Synechocys is cells esul s in a apid and d as ic decay
o GS ac i i y. The cons uc ion o a Synechocys is s ain
ha bo ing a his idine- agged modi ied e sion o GS has
allowed us o pu i y he inac i e enzyme. We desc ibe in he
p esen wo k he inac i a ing mechanism o he cyanobac e ial
GS ha in ol es he di ec in e ac ion o wo di e en inhib-
i o y polypep ides wi h he enzyme.
MATERIALS AND METHODS
Synechocys is G ow h Condi ions. Synechocys is sp. s ain
PCC 6803 and i s de i a i es we e g own pho oau o ophically
a 30°C on BG11 medium (13) (18 mM ni a e as ni ogen
sou ce) unde con inuous illumina ion (50 Wzm
22
; whi e
ligh ). The cul u es we e bubbled wi h 1% ( oly ol) CO
2
in ai .
BG11
0
medium was BG11 medium lacking ni ogen sou ce.
When ammonium was used as ni ogen sou ce, BG11
0
medium
was supplemen ed wi h 10 mM NH
4
Cl and he medium was
bu e ed wi h 20 mM N- is(hyd oxyme hyl)-me hyl-2-
aminoe hanesul onic acid (Tes) bu e a pH 7.0. Kanamycin
(50 o 200
m
gyml) and chlo amphenicol (20
m
gyml) we e
added when equi ed.
Pu i ica ion o Inac i e His idine-Tagged GS. ASynecho-
cys is s ain ha bo ing a his idine ag-modi ied GS was gene -
a ed as ollows. Fi e his idine codons we e inse ed a e he
glnA ATG s a codon by s anda d PCR echniques. Then, a
1.3-kb kanamycin esis ance (Km
R
) casse e (C.K1) (14) was
inse ed in o he ScaI si e, 189 bp ups eam o he glnA
ansla ion s a codon. This plasmid (pHIT4) was used o
ans o m Synechocys is as desc ibed p e iously (15). To al
eplacemen o he wild- ype (w ) glnA gene by he agged
e sion was e i ied by Sou he n blo (16) and PCR o pu i ied
genomic DNA om he Km
R
Synechocys is clones. This s ain
was named Synechocys is HTGS1. Ni a e-g own Synechocys is
HTGS1 cells o ni a e-g own cells ea ed o 2hwi h5mM
ammonium we e ha es ed by cen i uga ion and esuspended
in 50 mM HepeszNaOH bu e , pH 7.0. Cells we e dis up ed
by sonica ion (20 kHz, 75 W) o 2 min and cen i uged a
40,000 3g o 15 min. The esul ing supe na an cons i u ed
he cell- ee ex ac . His-GS was pu i ied om he cell- ee
ex ac s by Ni-a ini y ch oma og aphy by using His-Bind
Resin ma ix (No agen) and ollowing he manu ac u e ’s
ins uc ions.
Amino-Te minal De e mina ion o Inac i a ing Fac o s IF7
and IF17. P o ein samples we e subjec ed o SDSyPAGE (17)
and ans e ed o a poly( inylidene di luo ide) memb ane
The publica ion cos s o his a icle we e de ayed in pa by page cha ge
paymen . This a icle mus he e o e be he eby ma ked ‘‘ad e isemen ’’ in
acco dance wi h 18 U.S.C. §1734 solely o indica e his ac .
PNAS is a ailable online a www.pnas.o g.
This pape was submi ed di ec ly (T ack II) o he P oceedings o ice.
Abb e ia ions: GS, glu amine syn he ase; w , wild ype; IF, inac i a -
ing ac o .
*To whom ep in eques s should be add essed a : Ins i u o de
Bioquı´mica Vege al y Fo osı´n esis, Cen o de In es igaciones Cien-
ı´ icas Isla de la Ca uja, Cy. Ame ico Vespucio syn, 41092 Se illa,
Spain. e-mail: [email p o ec ed]s.
7161
(Immobilon-P; Millipo e). Memb ane-bound p o ein was sub-
jec ed o au oma ic Edman deg ada ion by using an Applied
Biosys ems P ocise Sequence . De e mined amino- e minal
sequences we e STQQQAR o IF7 and MQLSYR o IF17.
Inse ional Mu agenesis o gi A and gi B Synechocys is
Genes. Loci ssl1911 and sll1515 we e ampli ied by PCR, using
pu i ied Synechocys is sp. PCC 6803 genomic DNA, and cloned
in o pGEM-T (P omega), gene a ing he plasmids pSIF1 and
pLIF1, espec i ely. Ta ge ing ec o s pSIF3 and pSIF4 we e
gene a ed by eplacing a 334-bp AccI agmen , which con ains
he en i e gi A gene, by C.K1 and C.C1 (14) (chlo amphenicol
esis ance, Cm
R
) casse es, espec i ely. pLIF2 was gene a ed
by eplacing a 286-bp, NheI-Bs EII agmen con aining almos
he en i e gi B gene by a C.C1 (Cm
R
) casse e. To gene a e
Dgi A o Dgi B s ains, w Synechocys is cells we e ans o med
wi h pSIF3 (Km
R
) o pLIF2 (Cm
R
), espec i ely. To gene a e
Dgi Agi B double mu an s, Synechocys is Dgi A cells we e ans-
o med wi h he gi B gene- a ge ing ec o pLIF2, and Km
R
y
Cm
R
colonies we e selec ed. To gene a e Dgi A and Dgi B
mu an s in he Synechocys is s ain ha bo ing he modi ied
His-GS, pSIF4 and pLIF2 we e used o ans o m Synecho-
cys is HTGS1 (Km
R
) s ain, and Km
R
yCm
R
colonies we e
selec ed. Mu an s we e es ed by Sou he n blo analysis.
GS Assay. GS biosyn he ic ac i i y and GS ans e ase
ac i i y we e de e mined as desc ibed p e iously (11, 18). One
uni o GS ac i i y co esponds o he amoun o enzyme ha
ca alyzes he syn hesis o 1
m
mol o glu amine o
g
-glu amyl-
hyd oxama e pe min.
RNA Isola ion and No he n Blo Analysis. To al RNA
om Synechocys is was isola ed as desc ibed p e iously (19).
Fo No he n blo s, 15
m
g o o al RNA was loaded pe lane
and elec opho esed in 1% aga ose dena u ing o maldehyde
gels. T ans e o nylon memb anes (Hybond N
1
; Ame sham),
p ehyb idiza ion, hyb idiza ion, and washes we e in acco -
dance wi h Ame sham ins uc ion manuals.
GS, IF7, and IF17 Exp ession and Pu i ica ion. To exp ess
Synechocys is GS ype I in E. coli,aSalI agmen om pJCR3
(20) con aining he en i e Synechocys is glnA gene was cloned
in o pBluesc ip SK(1) in he same o ien a ion as he plac
p omo e . GS was pu i ied om E. coli by ammonium sul a e
p ecipi a ion and a ini y ch oma og aphy in ADP-Sepha ose.
Speci ic ac i i y o he pu e enzyme was 175 milliuni symg. To
exp ess IF7 and IF17, PCR-syn hesized agmen s encompass-
ing gi A and gi B genes we e inse ed in o pET-3a o gene a e
pSET2 and pLET2, espec i ely. Exponen ially g owing E. coli
BL21 cells ans o med wi h he indica ed plasmids we e
ea ed wi h 0.5 mM o isop opyl
b
-D- hiogalac oside o 5 h.
IF7 was pu i ied om he soluble ac ion by ca ion-exchange
ch oma og aphy on CM-52 cellulose (Wha man). IF17 was
ound o accumula e as insoluble inclusion bodies. The insol-
uble inclusion bodies we e isola ed, washed ex ensi ely wi h
1% T i on X-100, and hen solubilized in U bu e (7 M
u eay50 mM Hepes, pH 7.0y50 mM KCl). Solubilized inclu-
sion bodies we e subjec ed o ion-exchange ch oma og aphy
on DEAE-cellulose. The low- h ough ac ion con ained 95%
pu i ied IF17. Dena u ed IF17 was ena u ed by a 10- old
dilu ion in 50 mM Hepes, pH 7.0y50 mM KCl bu e . U ea was
elimina ed u he by dialysis.
P o ein–P o ein Band Shi Expe imen s. The binding e-
ac ions we e ca ied ou in a inal olume o 20
m
l con aining
2.25
m
g (0.18
m
M) o pu i ied GS and inc easing quan i ies o
IF7 o IF17 in HepeszNaOH bu e , pH 7.0y50 mM KCl. GS-IF
complexes we e allowed o o m du ing 5 min a oom
empe a u e and hen sepa a ed in 6% nondena u ing poly-
ac ylamide gels un a 4°C. Complexes we e isualized by gel
s aining wi h Coomassie blue.
C oss-Linking Expe imen s. C oss-linking eac ions we e
pe o med a 25°C in 50 mM HepeszNaOH bu e , pH 7.0, by
addi ion o 1-e hyl-3-(3-dime hylaminop opyl) ca bodiimide
o a inal concen a ion o 4 mM. Reac ions we e s opped a e
15 min by addi ion o SDS-con aining Laemmli loading bu e
(17). P o eins we e sepa a ed by SDSyPAGE, and gels we e
s ained wi h Coomassie blue.
RESULTS
Pu i ica ion o Inac i e Synechocys is 6803 GS. E o s o
pu i y inac i e GS om Synechocys is by classical ch oma o-
g aphic p ocedu es we e unsuccess ul, because he GS became
ac i e du ing he pu i ica ion p ocedu e. The e o e, we de-
cided o use Ni-a ini y ch oma og aphy as a as and e icien
me hod o pu i y inac i e GS om Synechocys is. Fo his
pu pose a Synechocys is s ain (HTGS1) was cons uc ed by
eplacing he w glnA gene (s uc u al gene o GS) by a
his idine- agged modi ied e sion. Ammonium-media ed in-
ac i a ion o he modi ied His-GS in i o was indis inguishable
om he inac i a ion o w GS (da a no shown). The ac i e
His-GS was pu i ied om ni a e-g own Synechocys is HTGS1
cells by Ni-a ini y ch oma og aphy. The inac i e enzyme was
pu i ied by he same p ocedu e om ni a e-g own Synecho-
cys is HTGS1 cells ha we e incuba ed wi h ammonium o 2 h.
Whe eas pu i ied His-GS om ni a e-g own cells showed a
speci ic ac i i y o 175 uni symg, His-GS pu i ied om am-
monium- ea ed cells showed a speci ic ac i i y o 63.5 uni sy
mg, indica ing ha abou 65% o he enzyme emained inac i e
a e elu ion om Ni-a ini y ch oma og aphy. These da a
we e con i med by in i o eac i a ion o he inac i e enzyme
by using eac i a ion ea men s desc ibed p e iously (pH o
ionic s eng h inc ease) (10). Bo h ac i e and inac i e pu i ied
His-GS we e subjec ed o SDSyPAGE. Su p isingly, wo
polypep ides o abou 7 and 17 kDa copu i ied wi h he inac i e
His-GS bu no wi h he ac i e enzyme (Fig. 1A). These
polypep ides we e named IF7 (inac i a ing ac o o 7 kDa)
and IF17 (inac i a ing ac o o 17 kDa).
IF7 and IF17 A e Encoded by Two Di e en Genes. Se-
quencing o IF7 and IF17 amino e mini (see Ma e ials and
Me hods) and compa ison wi h he Synechocys is 6803 genome
da abase (21) e ealed ha hese polypep ides co espond o
he ORFs ssl1911 and sll1515, espec i ely. IF7 and IF17
encoding genes we e named gi A and gi B, espec i ely ( o
glu amine syn he ase inac i a ing ac o ). Compa a i e anal-
ysis o he amino acid sequences demons a ed signi ican
sequence simila i y be ween IF7 and he ca boxyl e minus o
IF17 (Fig. 1B). Compa ison o gi A- and gi B-deduced amino
acid sequences wi h he da abase e ealed he exis ence o a
p e iously uniden i ied ORF om Anabaena sp. PCC 7120, a
ilamen ous cyanobac e ium, ha sha es homology wi h bo h
ac o s (Fig. 1B). In e es ingly, his ORF is placed downs eam
and in he opposi e o ien a ion wi h espec o he Anabaena
glnA gene (22).
gi A and gi B Mu an s A e Impai ed in GS Inac i a ion. To
es whe he IF7 and IF17 we e in ol ed in he inac i a ion o
Synechocys is GS, we cons uc ed single (Dgi A and Dgi B) and
double (Dgi Agi B) dele ion mu an s. Dgi A,Dgi B, and Dgi A-
gi B mu an s g ew no mally by using ni a e as ni ogen sou ce.
Le els o GS biosyn he ic ac i i y we e de e mined a di e en
imes a e ammonium addi ion o ni a e-g own w and
mu an cells. As desc ibed p e iously (11), w GS ac i i y
dec eased d ama ically, eaching abou 20% o he ini ial le el
40 min a e ammonium addi ion. Howe e , GS inac i a ion
was impai ed se e ely in bo h he Dgi A and he Dgi B mu an s
and comple ely absen in he Dgi Agi B double-mu an (Fig.
2A). These esul s clea ly demons a e ha IF7 and IF17 a e
in ol ed in he Synechocys is GS-inac i a ing mechanism. Be-
cause IF7 and IF17 copu i ied wi h GS, he mechanism o
inac i a ion could in ol e he physical in e ac ion o bo h
inac i a ing ac o s wi h he enzyme. To e i y whe he each
ac o is able o bind independen ly o GS in i o,wecon-
s uc ed Dgi A and Dgi B mu an s in he Synechocys is HTGS1
s ain ha bo ing he modi ied His-GS. Small quan i ies o IF7
7162 Biochemis y: Ga cı´a-Domı´nguez e al. P oc. Na l. Acad. Sci. USA 96 (1999)
o IF17 copu i ied wi h His-GS om ammonium- ea ed
HTGSI-Dgi B o HTGSI-Dgi A cells, espec i ely, co ela ing
wi h he low le el o GS inac i a ion obse ed in hese s ains.
These esul s demons a ed ha bo h inac i a ing ac o s we e
able o bind independen ly o he GS in i o (Fig. 2B).
Exp ession o gi A and gi B.No he n blo expe imen s
demons a ed ha gi A and gi B mRNA le els we e ansien ly
induced by ammonium. Thus, gi A and gi B mRNA le els in
ni a e-g own cells we e low, inc easing d ama ically 5 min
a e ammonium addi ion and eaching he highes le els a 5
20 min. One hou la e , mRNA quan i y e u ned o le els only
2- old highe ha hose p esen in ni a e (Fig. 3). De e mi-
na ion o GS ac i i y le els in he same expe imen showed a
empo al delay be ween he maximal gi A and gi B mRNA
accumula ion and he maximal GS inac i a ion. Thus, gi A and
gi B mRNA le els s a ed o decay be ween 20 and 40 min a e
ammonium addi ion, when GS has no eached comple e
inac i a ion (Fig. 3). These expe imen s sugges ha le els o
gi A and gi B mRNA a e inely con olled by a eedback
mechanism and ha GS inac i a ion may be he consequence
o inc easing he in acellula le els o IF7 and IF17.
In Vi o Recons i u ion o he GS Inac i a ion. To cha ac-
e ize u he he GS-IF in e ac ion in i o, we pu i ied
Synechocys is GS, IF7, and IF17 exp essed in E. coli (Fig. 4A).
FIG.1. (A) IF7 and IF17 copu i y wi h inac i e GS. His-GS was
pu i ied by Ni-a ini y ch oma og aphy om ni a e-g own HTGS1
Synechocys is cells (NO
3
2
) o om ni a e-g own HTGS1 Synechocys is
cells ea ed wi h 5 mM ammonium chlo ide o 2h(NH
4
1
). App ox-
ima ely 20
m
g o p o ein was sepa a ed by SDSyPAGE and s ained
wi h Coomassie blue. (B) Alignmen o he deduced amino acid
sequence o Synechocys is IF7 (ORF sll1911), IF17 (ORF ssl1515), and
Anabaena IF7 (p e iously uniden i ied ORF om GenBank accession
no. X00147). Alignmen was ca ied ou by using he CLUSTALX
p og am (29). Conse ed esidues a e indica ed by as e isks.
FIG.2. (A)In i o ammonium-dependen GS inac i a ion in Synechocys is w and Dgi A,Dgi B, and Dgi Agi B mu an s. w , Dgi A,Dgi B, and
Dgi Agi B Synechocys is cells we e g own in BG11 medium by using ni a e as ni ogen sou ce. Ammonium chlo ide (5 mM) was added a 50,
and GS biosyn he ic ac i i y was de e mined, in si u, a he indica ed imes. (B)Dgi A and Dgi B Synechocys is HTGS1 cells we e g own in BG11
medium and ea ed wi h 5 mM o ammonium chlo ide o 2 h. His-GS was pu i ied by Ni-a ini y ch oma og aphy and subjec ed o SDSyPAGE
and Coomassie blue s aining.
FIG. 3. T ansien induc ion o gi A and gi B mRNA le els. Am-
monium chlo ide (5 mM) was added o mid-log Synechocys is w cells
a 50. Samples we e aken a he indica ed imes, and o al RNA was
isola ed and analyzed by No he n blo ing. PCR-syn hesized ag-
men s, encompassing he en i e gi A o gi B genes, we e used as p obes.
The il e was ehyb idized wi h a p obe o he cons i u i ely ex-
p essed RNase P RNA gene as con ol (31). GS ans e ase ac i i y o
he same cul u es, a he indica ed imes, is shown in he g aph a he
bo om. One hund ed pe cen ac i i y co esponds o 1.5 uni symg o
p o ein.
Biochemis y: Ga cı´a-Domı´nguez e al. P oc. Na l. Acad. Sci. USA 96 (1999) 7163
Pu i ied IF7 and IF17 inhibi ed GS ac i i y in i o (Fig. 4Band
C). Addi ion o bo h ac o s oge he in equimolecula con-
cen a ions did no inc ease he inhibi o y e ec (da a no
shown). These esul s demons a e ha ei he IF7 o IF17 is
su icien pe se, wi hou addi ional modi ica ions, o GS
inac i a ion. GS inac i a ion exhibi ed a sigmoidal dependence
on IF17 concen a ion, sugges ing ha his ac o binds coop-
e a i ely o he GS. Howe e , a linea esponse was obse ed
o IF7-dependen inac i a ion. GS ac i i y could be eco e ed
by inc easing he pH o he sample (up o pH 9), a ea men
ha has been shown p e iously o eac i a e inac i e GS in
ex ac s om ammonium- ea ed cells (10). GS-IF in e ac ion
was demons a ed and isualized by p o ein–p o ein band shi
expe imen s. Inc easing amoun s o IF7 and IF17 e a ded he
GS p o ein band in nondena u ing gels (Fig. 5 Aand B). The
p esence o IF7 o IF17 in e a ded complexes was demon-
s a ed by excision o he slowe -mig a ing band om he
nondena u ing gels and mig a ion on SDSyPAGE gels. GS and
IF7 o IF17 we e sepa a ed (da a no shown). Pu i ied GS om
he cyanobac e ium Anabaena azollae was no e a ded by IF7
o IF17, indica ing ha bo h ac o s in e ac speci ically wi h
he Synechocys is GS. The mobili y shi caused by IF7 was
mino compa ed wi h he one caused by IF17. Whe eas one
majo band was obse ed in GS-IF7 in e ac ion expe imen s,
up o six di e en GS-IF17 o ms we e isible in GS-IF17 band
shi assays (Fig. 5B). Di ec in e ac ion was demons a ed
u he by c oss-linking expe imen s. Thus, ea men s o
mixes con aining GS and IF7 o IF17 wi h he wa e -soluble
ca bodiimide 1-e hyl-3-(3-dime hylaminop opyl) ca bodiim-
ide esul ed in he o ma ion o co alen complexes be ween
FIG. 4. Recons i u ion o Synechocys is GS inac i a ion in i o.(A) SDSyPAGE (15%) o pu i ied GS, IF17, and IF7 p o eins. Lanes: 1, 2
m
g
o pu i ied GS; 2, 1
m
g o pu i ied IF17; 3, 1
m
g o pu i ied IF7. Synechocys is GS (2.25
m
g) was incuba ed wi h inc easing quan i ies o IF7 (B)
and IF17 (C) in a inal olume o 20
m
l. Inac i e GS-IF complexes we e allowed o o m du ing 5 min, and GS ans e ase ac i i y was de e mined.
One hund ed pe cen ac i i y co esponds o 0.4 uni o GS.
FIG. 5. IF7 and IF17 in e ac in i o wi h GS. Synechocys is GS (0.18
m
M) was incuba ed wi h inc easing quan i ies o IF7 (A) and IF17 (B)
in a inal olume o 20
m
l. Inac i e GS-IF complexes we e allowed o o m du ing 5 min and hen sepa a ed in a 6% nondena u ing polyac ylamide
gel and s ained wi h Coomassie blue. Pu i ied Anabaena azollae GS (GSA) was incuba ed wi h IF7 (A) o IF17 (B). (C) Pu i ied GS (0.18
m
M)
was incuba ed in he absence o IF (lane 1), wi h 7
m
M IF7 (lane 2), 2
m
M IF17 (lane 3), o 7
m
M IF7 and 2
m
M IF17 (lane 4) in a inal olume
o 20
m
l. C oss-linking eac ions wi h 1-e hyl-3-(3-dime hylaminop opyl) ca bodiimide we e ca ied ou o 15 min. C oss-linking p oduc s we e
isualized by SDSyPAGE and Coomassie blue s aining. (D) Pu i ied GS (0.18
m
M) was incuba ed wi h 7
m
M o IF7 (lanes 2–4) o 2
m
M IF17
(lanes 5–7) o 10 min. Then, inc easing amoun s o IF17 (1 and 2
m
M) we e added o samples 3 and 4, and inc easing amoun s o IF7 (3.5
m
M
and 7
m
M) we e added o samples 6 and 7. Compe i ion binding was allowed o an addi ional 10-min pe iod, and GS-IF complexes we e isualized
by 6% nondena u ing PAGE.
7164 Biochemis y: Ga cı´a-Domı´nguez e al. P oc. Na l. Acad. Sci. USA 96 (1999)
GS subuni s and IF7 o IF17, which we e esol ed by SDSy
PAGE (Fig. 5C). Howe e , only IF17 was c oss-linked o GS
subuni s when bo h IF7 and IF17 we e added simul aneously
o he enzyme (Fig. 5C). In addi ion, mobili y-shi compe i-
ion assays showed ha GS-IF7 complexes we e dis up ed in
he p esence o IF17 bu GS-IF17 complexes we e no dis-
up ed by IF7 (Fig. 5D). These esul s sugges ha IF17 has a
highe a ini y o he GS han IF7. This conclusion also was
suppo ed by he inac i a ion kine ic da a (Fig. 4 Band C).
DISCUSSION
Phylogene ic analysis has e ealed ha he s uc u al gene o
GS is one o he oldes unc ioning genes (23), and, he e o e,
GS could be one o he oldes unc ioning enzymes. P obably,
his long e olu iona y his o y oge he wi h i s cen al ole in
me abolism ha e de e mined he exis ence o e y sophis i-
ca ed mechanisms o con ol GS ac i i y. We desc ibe a
egula ion mechanism o he cyanobac e ial GS ac i i y ha is
di e en om he classical adenylyla ion mechanism p esen
in many p oka yo es. The Synechocys is sys em in ol es he
di ec in e ac ion o wo inhibi o y pep ides (IF7 and IF17)
wi h he GS. The esul s p esen ed he e sugges a model in
which he p esence o ammonium, a ni ogen- ich sou ce,
igge s he ansc ip ional induc ion o gi A and gi B genes,
inc easing he syn hesis o IF7 and IF17. These ac o s hen a e
able o bind o he GS, p o oking enzyme inac i a ion.
Two di e en aspec s o he amino acid composi ion o IF7
and IF17 a e ema kable. Fi s , glu amine and a ginine a e he
mos abundan amino acids in bo h IF7 (13.7% Gln and 12.5%
A g) and IF17 (9.2% Gln and 12.2% A g). Gln is he enzyma ic
p oduc o GS. Glu amine in acellula pool inc eases abou
60- old a e ammonium addi ion o ni a e-g own cyanobac-
e ia (11), coinciding wi h he ime when IF7 and IF17 a e
syn hesized. A ginine is he amino acid wi h he highes
ni ogen con en , and, he e o e, i s syn hesis is a o ed unde
ich ni ogen condi ions. In addi ion, a ginine and ci ulline
(an in e media e in a ginine biosyn hesis) a e labeled sho ly
a e assimila ion o
13
NH
4
in some cyanobac e ia (24). Tha
Gln and A g, which cons i u e abou 25% o he amino acids
o IF7 and IF17, a e abundan du ing he ime ha bo h ac o s
a e syn hesized may enhance he a e o gi A and gi B mRNA
ansla ion. In addi ion, his pa icula amino acid composi ion
o bo h ac o s may be conside ed a kind o p oduc - eedback
mechanism egula ing he ammonium assimila ion pa hway.
The second in e es ing poin in he amino acid composi ion o
IF7 and IF17 is he abundance o posi i ely cha ged esidues.
The isoelec ic poin s o IF7 and IF17 a e 11.2 and 10.9,
espec i ely. GS can be eac i a ed in c ude ex ac s by in-
c easing he pH o he ionic s eng h (10). These esul s also
ha e been con i med wi h pu i ied GS and IFs (da a no
shown). These expe imen s indica e ha GS-IF complex o -
ma ion may be de e mined by elec os a ic in e ac ions. The
posi i e cha ge o bo h IFs sugges s ha he GS-in e ac ion si e
could be a nega i ely cha ged egion o he enzyme.
P edic ed seconda y s uc u e (using PHDSEC p og am om
he P edic P o ein Se e , Eu opean Molecula Biology Lab-
o a o y) (25) o bo h IFs e ealed he p obable p esence o an
a
-helix comp ising amino acids 85–115 o he IF17 sequence
and amino acids 3–45 o he IF7 sequence. The h ee IF
sequences a ailable show a high amino acid iden i y in his
egion (Fig. 1B), sugges ing ha i may be in ol ed in he
in e ac ion wi h he GS. Tha bo h Synechocys is IFs show
s uc u al and sequence simila i ies also sugges s ha bo h
ac o s in e ac wi h he same egion o he GS. Which egion
his is and how binding o IF7 and IF17 leads o he inhibi ion
o GS ac i i y a e poin s ha emain o be elucida ed. Mobil-
i y-shi expe imen s e ealed one majo band con aining
GS-IF7 complexes. In con as , up o six di e en bands we e
isible in GS-IF17 in e ac ion expe imen s, sugges ing ha
binding o a leas six IF17 polypep ides is equi ed o
comple e GS inac i a ion. Fu he expe imen s a e equi ed o
de e mine he exac s oichiome y o he inac i e GS-IF
complexes in i o.
IF7 and IF17 exp essed in E. coli a e able, wi hou u he
modi ica ion, o bind and inac i a e he GS. This esul o-
ge he wi h he apid inc ease in gi A and gi B exp ession a e
ammonium addi ion sugges ha binding o IFs o GS is
de e mined only by he in acellula concen a ion o bo h
ac o s. The e ec s o IF7 and IF17 in i o seem o be
cumula i e, aising he ques ion o why wo inac i a ing ac o s
a e equi ed. One possibili y is ha exp ession o IF7 and IF17
esponds o di e en en i onmen al cues in addi ion o he
ni ogen sou ce. Mo eo e , he di e en inac i a ion kine ics
displayed by each ac o poin s o a dis inc ole o each IF
unde speci ic condi ions.
Tha a p o ein homologous o IF7 and IF17 is p esen in
Anabaena sp. s ain PCC 7120, a cyanobac e ium phylogene i-
cally dis an om Synechocys is, sugges s ha a sys em o GS
ac i i y con ol simila o he one ha we desc ibe he e is
ex ended b oadly in cyanobac e ia. Ammonium-p omo ed
down- egula ion o GS om o he cyanobac e ial s ains also
has been obse ed ( e . 26 and ou unpublished obse a ions).
Howe e , i is unknown whe he a simila sys em o GS
inac i a ion ope a es in o he p oka yo ic g oups.
One majo di e ence be ween he sys em o GS modi ica-
ion by adenylyla ion and he cyanobac e ial sys em is ha
whe eas adenylyla ion p o okes a high sensi i i y o eedback
inhibi ion, GS-IF complex o ma ion seems o yield a com-
ple ely inac i e enzyme. Adenylyla ion cons i u es an example
o he egula o y cascades in which enzymes a one le el
modula e he ac i i ies o enzymes a he subsequen le el.
Such cascades p o ide dis inc con ol po en ial, including
signal ampli ica ion and inc eased con ollabili y (27, 28). In
addi ion, co alen modi ica ions a e as e esponse sys ems
han hose in ol ing al e a ions in gene exp ession. F om his
poin o iew, he cyanobac e ial mechanism could be consid-
e ed mo e udimen a y han he adenylyla ion sys em. How-
e e , egula ion by binding o inhibi o y pep ides is a solu ion
used ex ensi ely o modula e c i ical enzyma ic ac i i ies. Fo
example, con ol o cell cycle p og ession in euka yo es is
egula ed by a numbe o cyclin-dependen kinase inhibi o s
such as p15, p19, p21, p27, e c. (25). Fu he mo e, p o ease
enzyme ac i i ies om euka yo ic and p oka yo ic o igins also
a e o en con olled by inhibi o y pep ides (29, 30). S udies on
he enzymology and he me abolic con ol o GS ha e con-
ibu ed eno mously o biochemical knowledge in he pas . The
molecula in e ac ions in ol ed in he binding o IF7 and IF17
o GS and he molecula bases o he inhibi ion o GS
cons i u e an in e es ing model o unde s and biochemical
egula o y p ocesses based on p o ein–p o ein in e ac ion.
We hank J. d’Alaye o p o ein mic osequencing. We hank J. L.
C espo o p o iding pu i ied Anabaena azollae glu amine syn he ase
and J. Wei zman and A. Vioque o c i ical eading o he manusc ip .
M.G.D. was he ecipien o a ellowship om he Spanish Minis e io
de Educacio´n y Cul u a. This wo k was suppo ed by G an PB94–1444
and PB97–0732 om he Di eccio´n Gene al de Ensen˜anza Supe io
e In es igacion Cien ı´ ica and by Jun a de Andalucı´a (g oup CV1–
0112).
1. Rei ze , L. J. (1996) in Esche ichia coli and Salmonella: Cellula
and Molecula Biology, eds. Neidha d , F. C., Cu iss, R., III,
Ing aham, J. L., Lin, E. C. C., Low, K. B., Magasanik, B.,
Rezniko , W. S., Riley, M., Schaech e , M. & Umba ge , H. E.
(Am. Soc. Mic obiol., Washing on, DC), 2nd Ed., pp. 391–407.
2. Me ick, M. J. & Edwa ds, R. A. (1995) Mic obiol. Re . 59,
604–622.
3. Magasanik, B. (1996) in Esche ichia coli and Salmonella: Cellula
and Molecula Biology, eds. Neidha d , F. C., Cu iss, R., III,
Ing aham, J. L., Lin, E. C. C., Low, K. B., Magasanik, B.,
Biochemis y: Ga cı´a-Domı´nguez e al. P oc. Na l. Acad. Sci. USA 96 (1999) 7165

Rezniko , W. S., Riley, M., Schaech e , M. & Umba ge , H. E.
(Am. Soc. Mic obiol., Washing on, DC), 2nd Ed., pp. 1344–1356.
4. Yamashi a, M. M., Almassy, R. J., Janson, C. A., Cascio, D. &
Eisenbe g, D. (1989) J. Biol. Chem. 264, 17681–17690.
5. Almassy, R. J., Janson, C. A., Hamlin, R., Xuong, N.-H. &
Eisenbe g, D. (1986) Na u e (London) 323, 304–309.
6. Rhee, S. G., Chock, P. B. & S ad man, E. R. (1989) Ad ances in
Enzymology, ed. Meis e , A. (Wiley In e science, New Yo k),
Vol. 62, pp. 37–92.
7. S ad man, E. R. (1990) Me hods Enzymol. 182, 793–809.
8. Deuel, T. F. & P usine , S. (1974) J. Biol. Chem. 249, 257–264.
9. Fishe , R., Tuli, R. & Haselko n, R. (1981) P oc. Na l. Acad. Sci.
USA 78, 3393–3397.
10. Me´ ida, A., Candau, P. & Flo encio, F. J. (1991) Biochem.
Biophys. Res. Commun. 181, 780–786.
11. Me´ ida, A., Candau, P. & Flo encio, F. J. (1991) J. Bac e iol. 173,
4095–4100.
12. Reyes, J. C. & Flo encio, F. J. (1995) FEBS Le . 367, 45–48.
13. Rippka, R., De uelles, J., Wa e bu y, J. B., He dman, M. &
S anie , R. Y. (1979) J. Gen. Mic obiol. 111, 1–61.
14. Elhai, J. & Wolk, C. P. (1988) Gene 68, 119–138.
15. Chau a , F., De V ies, L., Van de Ende, A. & Van A kel, G. A.
(1988) Mol. Gen. Gene . 204, 165–191.
16. Samb ook, J., F i sch, E. F. & Mania is, T. (1989) Molecula
Cloning: A Labo a o y Manual (Cold Sp ing Ha bo Lab. P ess,
Plain iew, NY), 2nd Ed.
17. Laemmli, U. K. (1970) Na u e (London) 227, 680–685.
18. Ma que´s, S., Flo encio, F. J. & Candau, P. (1989) Anal. Biochem.
180, 152–157.
19. Ga cı´a-Domı´nguez, M. & Flo encio, F. J. (1997) Plan . Mol. Biol.
35, 723–734.
20. Reyes, J. C. & Flo encio, F. J. (1994) J. Bac e iol. 176, 1260–1267.
21. Kaneko, T., Sa o, S., Ko ani, H., Tanaka, A., Asamizu, E.,
Nakamu a, Y., Miyajima, N., Hi osawa, M., Sugiu a, M.,
Sasamo o, S., e al. (1996) DNA Res. 3, 109–136.
22. Tume , N. E., Robinson, S. J. & Haselko n, R. (1983) Na u e
(London) 306, 337–342.
23. Kumada, Y., Benson, D. R., Hillemann, D., Hos ed, T. J.,
Roche o , D. A., Thompson, C. J., Wohlleben, W. & Ta eno, Y.
(1993) P oc. Na l. Acad. Sci. USA 90, 3009–3013.
24. Meeks, J. C., Wolk, C. P., Thomas, J., Lockau, W., Sha e , P. W.,
Aus in, S. M., Chien, W. S. & Galonsky, A. (1977) J. Biol. Chem.
252, 7894–7900.
25. Ros , B. & Sande , C. (1993) P oc. Na l. Acad. Sci. USA 90,
7558–7562.
26. Rowell, P., Sampaio, M. J. A. M., Ladha, J. K. & S ewa , W. D. P.
(1979) A ch. Mic obiol. 120, 195–200.
27. Mu a, U., Chock, P. B. & S ad man, E. R. (1981) J. Biol. Chem.
256, 13022–13029.
28. Mu a, U. & S ad man, E. R. (1981) J. Biol. Chem. 256, 13014–13021.
29. Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994) Nucleic
Acids Res. 22, 4673–4680.
30. Laskowski, M., J ., & Ka o, I. (1980) Annu. Re . Biochem. 49,
593–626.
31. Vioque, A. (1992) Nucleic Acids Res. 20, 6331–6337.
7166 Biochemis y: Ga cı´a-Domı´nguez e al. P oc. Na l. Acad. Sci. USA 96 (1999)