N1,N1-Diethyl-N 2-(2,3,4,6-tetra-O-acetyl-β-D-glucopyranosyl)acetamidine

Ac a C ys . (2003). D59, 2183±2190 Xu e al. PII p o eins 2183 esea ch pape s Ac a C ys allog aphica Sec ion D Biological C ys allog aphy ISSN 0907-4449 The s uc u es o he PII p o eins om he cyanobac e ia Synechococcus sp. PCC 7942 and Synechocys is sp. PCC 6803 Yibin Xu, a ² Paul D. Ca , b ² Paula Clancy, a Ma io Ga cia- Dominguez, b,c Ka l Fo chhamme , d F ancisco Flo encio, c Nicole Tandeau de Ma sac, e Subhash G. Vasude an a, and Da id L. Ollis b * a Depa men o Biochemis y and Molecula Biology, James Cook Uni e si y, Towns ille, Queensland 4811, Aus alia, b Resea ch School o Chemis y, Aus alian Na ional Uni e si y, GPO Box 414, Canbe a, ACT 2601, Aus alia, c Ins i u o de Bioquõ Âmica Vege al y Fo osõ Ân esis, Uni e sidad de Se illa-Consejo Supe io de In es igaciones Cien õ  icas, Ame  ico Vespucio s/n, E-41092 Se illa, Spain, d Ins i u u È Mik obiologie und Molekula biologie, Jus us-Liebig-Uni e si a È Giessen, Ge many, e Uni e Âdes Cyanobac e  ies, CNRS-URA 2172, De Âpa emen de Mic obiologie Fondamen ale e Me Âdicale, Ins i u Pas eu , 28 Rue du Doc eu Roux, 75724 Pa is CEDEX 15, F ance, and No a is Ins i u e o T opical Diseases P e L d, 1 Science Pa k Road, #04-14 The Cap ico n, Singapo e Science Pa k II, Singapo e 117528, Singapo e ² These wo au ho s con ibu ed equally o his pape . Co espondence e-mail: [email p o ec ed]u #2003 In e na ional Union o C ys allog aphy P in ed in Denma k ± all igh s ese ed The PII p o eins om he cyanobac e ia Synechococcus sp. PCC 7942 and Synechocys is sp. PCC 6803 ha e been c ys allized and high- esolu ion s uc u es ha e been ob ained using X- ay c ys allog aphy. The co e o hese new s uc u es is simila o ha o he PII p o eins om Esche ichia coli, al hough he s uc u es o he T- and C-loops di e . The T-loop o he Synechococcus p o ein is o de ed, bu appea s o be s abilized by c ys al con ac s. The same loop in he Synechocys is p o ein is diso de ed. The C- e minus o he Synechocys is p o ein is s abilized by hyd ogen bonding o he same egion o a c ys allog aphically ela ed molecule. The same e minus in he Synechococcus p o ein is s abilized by coo dina ion wi h a me al ion. These obse a ions a e consis en wi h he idea ha bo h he T-loop and he C- e minus o PII p o eins a e ¯exible in solu ion and ha his ¯exibili y may be impo an o ecep o ecogni ion. Sequence compa isons a e used o iden i y egions o he sequence unique o he cyanobac e ia. Recei ed 26 May 2003 Accep ed 5 Sep embe 2003 1. In oduc ion PII is a signal- ansduc ion p o ein ha is bes known o i s ole in he egula ion o ni ogen assimila ion by en e ic bac e ia. Howe e , he p o ein is ound and highly conse ed in a a ie y o o ganisms co e ing all he h ee kingdoms o li e (Nin a & A kinson, 2000; A condeguy e al., 2001). In se e al o hese o ganisms he e a e genes ha code o wo o e en h ee PII-like p o eins. In Esche ichia coli he e a e wo such p o eins, known as GlnB and GlnK. Whils hese wo p o eins sha e 67% simila i y, hey do no appea o be edundan in hei unc ion ( an Heeswijk e al., 1995; A kinson & Nin a, 1999). The ole o he PII-like p o eins in o he o ganisms is only now being elucida ed, bu i is al eady clea ha hey ha e di e en unc ions. In E. coli, he PII pa alogues a e u idylyla ed and modula e he ac i i y o glu amine syn he ase (GS) and he ansc ip ion o he co esponding gene glnA (Me ick & Edwa ds, 1995). In cyanobac e ia, he PII p o ein is e e sibly phospho yla ed on a se yl esidue (Fo chhamme & de Ma sac, 1995a). The speci®c phospho-PII phospha ase, a membe o he PP2C amily, was iden i®ed ecen ly (I mle & Fo chhamme , 2001). The cyanobac e ial PII p o ein is in ol ed in he egula ion o ni a e and ni i e up ake (Fo chhamme & de Ma sac, 1995b; Lee e al., 1998, 2000), bica bona e anspo (Hisbe gues e al., 1999) and is equi ed o he ac i a ion o N cA-dependen gene exp ession unde condi ions o ni ogen s a a ion (Aldehni e al., 2003). The s uc u es o E. coli GlnB (EcPII) and GlnK ha e bo h been sol ed (Cheah e al., 1994; Xu e al., 1998). Mo e ecen ly, esea ch pape s 2184 Xu e al. PII p o eins Ac a C ys . (2003). D59, 2183±2190 he s uc u e o he PII p o ein om He baspi illum se opedicae (HsPII), a membe o he same o de (p o eo- bac e ia) bu a di e en subg oup o E. coli, has been sol ed (Benelli e al., 2002). In all h ee cases, signal ansduc ion occu s h ough modi®ca ion o he Ty 51 esidue ound on he T-loop which ex ends ou om he co e o he molecule. This loop has been ound o ha e di e en con o ma ions in he E. coli p o eins and is diso de ed in he complexes o med wi h ATP (Xu e al., 1998, 2001). The T-loop is also la gely dis- o de ed in HsPII. I is hough ha his loop is ¯exible in solu ion and ha his ¯exibili y allows ecogni ion o occu wi h se e al ecep o s. Apa om he T-loop, he C- e minus o he PII p o ein appea s o be capable o aking on mul iple con o ma ions and may be in ol ed in ecogni ion. The C- e minal loop adop s one o wo opologies: ei he a pai o sho an ipa allel -s ands o a 3 10 -helix. This loop is obse ed o exhibi igid-body mo emen s ela i e o he co e o he molecule when known s uc u es a e o e laid. The E. coli p o eins equi e ATP in o de o ca y ou hei unc ions. C ys allog aphic s udies ha e shown ha ATP binds in a highly conse ed cle on he side o he molecule. Like hei E. coli coun e pa s, he PII p o eins om cyanobac e ia o m ime s o 12 kDa subuni s. Fo Synechococcus sp. PCC 7942, i has been shown ha he PII p o ein (SnPII) can o m wi h ze o, one, wo o h ee phos- pho yla ed se ine esidues. The ex en o phospho yla ion depends upon he a ailabili y o ca bon and ni ogen sou ces (Fo chhamme & de Ma sac, 1995b; Lee e al., 1998). A he molecula le el, he phospho yla ion s a e o PII esponds o cen al ca bon me aboli es, in pa icula 2-oxoglu a a e, e¯ec ing he me abolic s a e o he cells (Ruppe e al., 2002). The se ine esidue in ques ion is loca ed in he T-loop a posi ion 49 (Fo chhamme & de Ma sac, 1995a). When exp essed in E. coli, he y osine a posi ion 51 can be u idylyla ed, bu i does no modula e ei he GS ac i i y o he ansc ip ion o i s gene (Fo chhamme , 2003). This sugges s ha he T-loop in SnPII di e s om ha ound in he E. coli pa alogues. Simila ly, s udies on PII om Synechocys is sp. PCC 6803 (SsPII) ha e shown ha he same se yl esidue a posi ion 49 is phospho yla ed (Ga cia-Dominguez & Flo encio, unpublished wo k). The sequences o he T-loops in he SnPII and SsPII p o eins a e iden ical and di e om ha o EcPII a 5 o he 19 esidues (Fig. 1). The sequence iden- i ies be ween he SnPII and SsPII p o eins and he EcPII p o ein a e 65 and 55%, espec i ely. This compa es o 86% iden i y be ween he wo cyanobac e ial sequences. Al hough he sequences o he cyanobac e ial PII p o eins a e e y simila o hose o he E. coli PII-like p o eins, hey a e di e en ly modi®ed and ha e di e en a ge ecep o s. The cyanobac e ial p o eins a e mo e closely ela ed o he PII p o eins ound in plan s. The sequence o he PII om A abidopsis haliana is mo e simila o ha o he cyano- bac e ia han i is o E. coli (Smi h e al., 2003). In addi ion, he plan p o eins do no ha e a y osine in he T-loop bu , like he cyanobac e ial p o eins, ha e a se ine ha appea s o be phospho yla ed. We ha e de e mined he s uc u es o he SsPII p o ein and a mu an o m o he SnPII p o ein. The mu an had an alanine in place o he unc ionally impo an Se 49 and mimics he unphospho yla ed o m o he p o ein (Lee e al., 2000). We commen on he unc ional signi®cance o s uc u al compa isons made be ween he PII p o eins om E. coli and cyanobac e ia. 2. Ma e ials and me hods 2.1. Bac e ial s ains and g ow h condi ions E. coli DH5was used o plasmid cons uc ion and epli- ca ion. E. coli BL21 was used o he exp ession o SsPII p o ein om Synechocys is sp. PCC 6803 (he ea e e e ed o as Synechocys is). The ec o exp essing he mu an o m o he PII p o ein (SnPIIA) om Synechococcus sp. PCC 7942 (he ea e e e ed o as Synechococcus) was ans o med in o RB9065 cells. E. coli s ains we e g own in Lu ia B o h (LB) medium as desc ibed by Samb ook e al. (1989) supplemen ed wi h 100 mgml ÿ1 ampicillin when equi ed. 2.2. Exp ession and pu i ica ion Fo exp ession o he SsPII p o ein in E. coli, a plasmid, pMAB11, was cons uc ed as a de i a i e o he exp ession ec o pET-3a (No agen). A DNA agmen o abou 400 bp encompassing he Synechocys is glnB gene was ob ained by PCR ampli®ca ion wi h oligonucleo ides gbp1 (50-GTACCA- CATATGAAAAAAG-30) and gbp2 (50-CTTGTCTGGATC- CGCCCAAC-30). Oligonucleo ides gbp1 and gbp2 ha bou NdeIandBamHI si es, espec i ely. The PCR p oduc was checked by sequencing, diges ed wi h NdeI and BamHI and cloned in o he NdeI±BamHI si es o pET-3a. Oligonucleo ide gbp1 subs i u es he na u al ansla ion s a ing codon o he Synechocys is glnB gene (TTGl; Ga cia-Dominguez & Flo encio, 1997) by a s anda d ATG codon, included in he NdeI si e. pMAB11- ans o med E. coli BL21 s ain was g own in 3 l o M9 minimal medium wi h ampicillin (100 mgml ÿ1 ) o an op ical densi y a 600 nm o 0.5; o p oduc ion o he SsPII p o ein, isop opyl--d- hiogalac opy anoside (IPTG) was hen added o a ®nal concen a ion o 1 mMand he cul u e Figu e 1 Alignmen o wo cyanobac e ial PII sequences wi h EcPII. was u he incuba ed o 4 h be o e ha es ing cells by cen i uga ion. Cells we e esuspended in 15 ml T bu e (50 mMT is±HCl pH 8.0, 50 mMNaCl, 5 mMMgCl 2 ,1mMEDTA), dis up ed by sonica ion (20 kHz, 75 W) o 2 min and cen i uged a 40 000 g o 30 min. Cell- ee ex ac was ac iona ed be ween 35 and 50% ammonium sul a e. A e dialysis wi h T bu e , he p o ein was hea ed a 338 K o 5 min and cen i uged; he supe na an was hen applied o a DEAE-cellulose column. The ¯ow h ough was collec ed, concen a ed wi h 50% ammonium sul a e and desal ed in a Sephadex G-50 (Pha - macia) column equilib a ed in T bu e . A 150 mg ac ion o SsPII p o ein was yielded and adjus ed o 10 mg ml ÿ1 wi h T bu e . The gene o he SnPIIA p o ein was ob ained om plasmid pPM308 (Lee e al., 2000). This plasmid was used o p oduce a PCR agmen wi h an NdeI and KpnI si e a ei he end. The PCR p ime s used o his pu pose we e SynPII-FP (50- TATATTCATATGAAGAAGATTGAGGCGATTATTC-30) and SynPII-RP (50-TATATTGGTACCTTAGATCGTGTCG- GCGTTTTTC-30). The PCR p oduc was diges ed wi h NdeI and KpnI es ic ion endonucleases and liga ed wi h pND707 (Lo e e al., 1996) ha had been simila ly cu . The esul ing clone was checked by nucleo ide-sequence analysis. The SnPIIA p o ein was o e exp essed in RB9065 cells (UTase ÿ mu an ). The cells we e g own in a 5 l e men e o LB medium wi h ampicillin (100 mgml ÿ1 )a 303K oanOD 595 o 0.6; he empe a u e was hen apidly shi ed o 315 K and he cells we e g own o a u he 3 h. The cells we e ha es ed by cen i uga ion, esuspended in HEPES bu e A(20 mM HEPES, 1 mMEDTA, 1 mM-me cap oe hanol) and dis up ed using a F ench p ess. The cells we e p essed wice and PMSF (0.5 mM) was added be o e and a e p essing o p e en p o eolysis. Nucleic acid con aminan s we e emo ed om he sample using s ep omycin sul a e [1.5%(w/ )] p ecipi a ion, ollowed by ac iona ion using ammonium sul a e (be ween 25 and 35%). The sample was dialysed agains 3 2 l changes o HEPES bu e Aand applied o a Cibac on Blue ma ix column, washed wi h he same bu e con aining 100 mMNaCl and ®nally elu ed wi h a sal g adien in HEPES bu e (1±0 MNaCl, 5 mMATP). The esul ing p o ein solu ion was >95% in pu i y based on a 15% SDS± PAGE gel and was concen a ed o 20 mg ml ÿ1 o c ys al- liza ion ials. 2.3. C ys alliza ion, da a collec ion and s uc u e de e mina ion C ys als o he SsPII p o ein we e g own using he hanging- d op me hod a 293 K. Equal olumes o he p o ein solu ion (10 mg ml ÿ1 ) and a ese oi solu ion con aining 20% 2-p opanol, 100 mMCaCl 2 , 100 mMsodium ace a e pH 4.7 we e used. Small c ys als appea ed a e 3 d and con inued o g ow o abou wo weeks. C ys als we e p epa ed o da a collec ion by ans e o a solu ion o he ese oi bu e con aining 25% glyce ol. The SnPIIA p o ein was c ys allized using he hanging-d op me hod a 293 K by mixing equal olumes o he p o ein solu ion (20 mg ml ÿ1 ) wi h a ese oi solu ion con aining 1.0 Mli hium sul a e, 0.1 MT is bu e pH 8.5 and 10 mM nickel chlo ide hexahyd a e. C ys als we e p epa ed o da a collec ion by ans e o a solu ion o he ese oi bu e . All da a we e collec ed using a Rigaku R-AXIS IIc image pla e moun ed on a o a ing-anode gene a o unning a 50 kV, 100 mA. Da a we e p ocessed wi h he HKL p og am (O winowski & Mino , 1997). Fo SsPII, da a we e collec ed a 100 K. The c ys als we e ound o belong o space g oup R3, wi h uni -cell pa ame e s a=b= 129.79, c= 74.14 A Ê, and con ained ou subuni s in he asymme ic uni . Fo he SnPII S49A mu an , da a we e collec ed a 277 K. The c ys als we e ound o belong o space g oup I4 1 22, wi h uni -cell pa a- me e s a=b= 109.70, c= 109.68 A Ê, and con ained a ime in he asymme ic uni . The s uc u e o he SsPII p o ein was de e mined by molecula eplacemen using a model o a ime o EcPII (Ca e al., 1996) om which he T-loop ( esidues 37±55) and he ou C- e minal esidues had been dele ed. Calcula ions we e unde aken using he p og am AMoRe (Na aza, 1994) as implemen ed in he CCP4 package (Collabo a i e Compu a- ional P ojec , Numbe 4, 1994). The bes o a ion- unc ion solu ion had a co ela ion coe ®cien o 0.424 compa ed wi h he highes inco ec peak o 0.170. The bes ansla ion- unc ion solu ion had a co ela ion coe ®cien o 0.509 and an R ac o o 0.465 compa ed wi h 0.237 and 0.566, espec i ely, o he highes inco ec solu ion. A e igid-body e®nemen , hese alues imp o ed o 0.587 and 0.428, espec i ely. The model gene a ed om his solu ion ga e good c ys al packing when inspec ed on a g aphics e minal. Fu he e®nemen s we e unde aken using CNS (B u Ènge e al., 1998) and phases we e imp o ed using DM (Cow an, 1994). Du ing subsequen model building, i became appa en ha he e was densi y o ano he molecule o PII in he asymme ic uni . A u he monome o PII was added o he model. A well packed molecule was ob ained om ou subuni s o PII. The o iginal h ee molecules o med a ime a ound a non- c ys allog aphic iad, whils he ou h p oduced ime s om c ys allog aphically ela ed molecules. A Ma hews coe ®cien (Ma hews, 1968) V M o 2.8 A Ê 3 Da ÿ1 is ob ained o ou molecules pe asymme ic uni , wi h a co esponding sol en con en o 56%. This compa es wi h a V M o 3.6 A Ê 3 Da ÿ1 o a single ime pe asymme ic uni o 1.8 A Ê 3 Da ÿ1 o wo ime s pe asymme ic uni . Fu he ounds o e®nemen using simula ed-annealing, posi ional and indi idual B- ac o e®nemen using s anda d CNS p o ocols (B u Ènge e al., 1998) we e in e spe sed wi h manual ebuilding o he model. The ®nal model had an R wo k and R ee o 0.219 and 0.251, espec i ely. Re®nemen s a is ics a e lis ed in Table 1. The s uc u e o he SnPIIA p o ein was sol ed by mole- cula eplacemen wi h MOLREP (Vagin & Teplyako , 2000), inco po a ed in he CCP4 sui e. The high- esolu ion s uc u e o he E. coli PII ime wi hou he T-loop was used as a sea ch model (Ca e al., 1996). MOLREP ga e ise o a Ac a C ys . (2003). D59, 2183±2190 Xu e al. PII p o eins 2185 esea ch pape s esea ch pape s 2186 Xu e al. PII p o eins Ac a C ys . (2003). D59, 2183±2190 dominan solu ion wi h R= 0.481 and a co ela ion coe ®cien o 0.476 o 3.0 A Ê esolu ion. Rigid-body e®nemen and u he s uc u e e®nemen we e implemen ed using CNS .1.1 and model building was ca ied ou wi h O(Jones e al., 1991). Non-c ys allog aphic symme y es ain s we e applied h oughou he e®nemen . The occupancies o diso de ed side chains we e se o ze o du ing he e®nemen . A numbe o non-pep ide ea u es appea ed in he elec on-densi y map du ing he cou se o e®nemen . Wi hin each ATP-binding si e, he e we e wo ea u es ha appea ed o be sul a es om he c ys alliza ion bu e . In addi ion, he e was a sphe ical ea u e in he maps ha appea ed o be Ni 2+ ions, also om he c ys alliza ion bu e . S a is ics o da a collec ion and s uc u e e®nemen a e gi en in Table 1. 3. Resul s and discussion 3.1. The models The models con® m ha he cyanobac e ial PII p o eins ha e he same o e all opology as o he known PII s uc u es. The molecules o m homo ime s a ound h ee old axes, wi h -shee in e ac ions in ol ing bo h o he neighbou ing p o ome s. These hyd ogen bonds hold he cen al co e s uc u e o he ime s oge he . Each monome consis s o a cen al ou -s anded an ipa allel -shee wi h a pai o an i- pa allel helices on one side. These o m an in e locking double  mo i . The e a e wo majo loops emana ing om he cen al shee . The ® s and la ges , be ween s ands 2 and 3, is known as he T-loop and con ains he si e o phospho yla ion. In s uc u es sol ed o da e, his loop has only been isible whe e c ys al-packing o ces ha e s abilized one pa icula con o ma ion. This has led o he conclusion ha i is ¯exible in solu ion. The T-loop is no isible in he SsPII s uc u e epo ed he e, bu is isible in he elec on densi y o SnPII (Fig. 5). La ice con ac s appea o s abilize his con o ma ion. The second loop is ound be ween s and 4 and he ca boxyl- e minus. In bo h o he s uc u es epo ed he e and in se e al o he PII s uc u es his loop con ains wo sho an ipa allel -s ands. In o he PII-like s uc u es, o example, EcGlnK and HsPII, his loop con ains a 3 10 -helix a he han -s ands. The model o SsPII consis s o ou p o ein chains. Chains A,Band C o m a ime a ound a non-c ys allog aphic iad. Chain D o ms a ime by symme y ope a ions on i sel . Like se e al o he o ms o PII (Xu e al., 2001), he T-loop ( esi- dues 37±53) is no isible in he elec on densi y, p esumably owing o mobili y. The model consis s o esidues A1±36, A54± 112, B1±36, B55±112, C1±36, C55±112, D1±38 and D54±110. In addi ion, wo Ca 2+ ions, wo glyce ol molecules and 217 wa e molecules we e included. The ollowing esidues did no exhibi good side-chain densi y and we e unca ed o alanines in he e®nemen : A76, A101, A106, B76, C76, D38 and D54. The ollowing glu ama e esidues exhibi ed al e na e con o - ma ions o hei side chains: A15, A62, B5, B15, B62, C5, C15 and D15. Al e na e con o me s we e modeled a hal -occu- pancy. The ®nal model exhibi ed good s e eochemis y when checked by WHAT_CHECK (Hoo e al., 1996) and PROCHECK (Laskowski e al., 1993). The Ramachand an plo showed 98.1% o esidues alling in o he mos a ou ed egion, whils he emaining 1.9% we e in addi ional allowed egions. No esidues ell in o he disallowed o gene ously allowed egions. The o e all old o he SsPII p o ein is essen ially he same as ha o EcPII p o ein. The indi idual p o ein chains we e o e laid using LSQKAB (Kabsch, 1976). The .m.s. alues o main-chain a om displacemen we e less han o equal o 0.6 A Ê. Fo compa ison, he Dchain was also supe imposed on Table 1 Summa y o X- ay di ac ion da a and e®nemen s a is ics. Values in pa en heses a e o he ou e da a shell. SnPII SsPII X- ay di ac ion da a Resolu ion (A Ê) 25±2.0 (2.1±2.0) 25±2.0 (2.1±2.0) Obse ed e¯ec ions 68888 156138 Unique e¯ec ions 19247 31543 Comple eness (%) 84.0 (80.2) 99.5 (98.2) R me ge (%) 5.9 (24.1) 5.3 (16.7) A e age I/(I) 8.5 (3.1) 16.7 (4.93) Re®nemen s a is ics Resolu ion (A Ê) 25±2.0 25±2.0 No. o e¯ec ions (wo king) 17030 29381 No. o e¯ec ions ( es ) 1869 (8.2%) 1577 (5%) R c ys /R ee ² (%) 20.0/22.7 21.9/25.3 No. p o ein a oms 2610 3499 No. sol en a oms 132 217 No. glyce ol molecules Ð 2 No. ions 3 Ni 2+ ,6SO 2ÿ 42Ca 2+ R.m.s. de ia ion om ideal geome y Bond leng h (A Ê) 0.005 0.006 Bond angle () 1.28 1.26 Dihed als () 23.64 26.64 Imp ope () 0.67 1.17 A e age B ac o s (A Ê 2 ) P o ein 18.4 29.8 Wa e 30.0 42.3 Sul a e 43.4 Ð Glyce ol Ð 51.2 Ions 23.7 34.2 Ramachand an plo (%) Mos a ou ed egions 96.9 98.1 Addi ional allowed egions 3.1 1.9 Gene ously allowed egions 0.0 0.0 Disallowed egions 0.0 0.0 ²R c ys =(|F obs |ÿ|F calc |)/|F obs |; R ee was calcula ed in he same manne as R c ys using a es se o da a no used in he e®nemen p ocess. Table 2 R.m.s. displacemen alues o supe posi ion o main-chain a oms o a ious PII s uc u es. Species Space g oup Chain R.m.s. (A Ê) Synechocys is R3A0.60 B0.59 C0.58 D0.00 Synechococcus I4 1 22 1.24 E. coli P6 3 0.95 I2 1 3 1.19 E. coli GlnK P2 1 3 1.08 H. se oedicae P2 1 2 1 2 1 A0.92 a numbe o o he PII s uc u es (Table 2). In all o he supe posi ions he basic s uc u e o he p o ein chains align well, wi h jus h ee egions showing a ia ions. Fi s ly, in s uc u es whe e c ys al packing s abilizes he T-loop (EcPII, EcGlnK and SnPIIA) i adop s di e en con o ma ions. Secondly, he C-loop, which exhibi s -shee opology, unde goes small igid-body mo emen s wi h espec o he cen al s ands. Thi dly, he las h ee C- e minal esidues di e be ween s uc u es. Fig. 2 shows an o e lay o he a ious PII p o ein chains. Inspec ion o he empe a u e ac o s e eals ha he cen al igid co e o in e - locking -s ands has he lowes B ac o s (18 A Ê 2 ). The B-loop and helices ha e alues ha a e ypically double hose o he co e. The C-loop, howe e , is mo e s able (25 A Ê 2 ) owing o in e - ime s and o ma ion. The e is a ia ion be ween he indi idual p o ome s ha clus e a ound he non-c ys allog aphic h ee old axis, al hough i di ®cul o see why his should be as hey ha e simila packing en i onmen s. The c ys al packing appea s o ha e in¯uenced he cu en s uc u e by in e - ac ions a wo si es. The ® s is a he Ca 2+ ion-binding si es, which occu be ween symme y- ela ed molecules. The me al ion is liganded o he acidic side chains o esi- dues Asp14 and Glu15 and he ca bonyl O a om o Phe11. I is also bound o he symme y ma es o hese h ee a oms in a neighbou ing molecule, gi ing he Ca 2+ ion ull oc ahed al coo dina ion (Fig. 3). The Ac a C ys . (2003). D59, 2183±2190 Xu e al. PII p o eins 2187 esea ch pape s Figu e 2 Supe posi ion o a ious PII monome s. Colou scheme: SnPIIA, magen a; SsPIIA, blue; SsPIID, o ange; EcPII, black; GlnKB, g een; HsPII, ed. No e ha T-loops (uppe le ), whe e p esen , show a iable con o ma ions. The C-loop s uc u e (lowe igh ) unde goes igid-body mo emen s ela i e o he s able double  co e be ween s uc u es. Figu e 3 Close-up o c ys al con ac in SsPII a he Ca 2+ -binding si e. Colou coded by (symme y- ela ed) p o ein chains. Figu e 4 Hyd ogen bonds s abilize he C-loops o symme y- ela ed molecules o SsPII. esea ch pape s 2188 Xu e al. PII p o eins Ac a C ys . (2003). D59, 2183±2190 second si e o packing-in¯uenced in e ac ions occu s a he ®nal -s and, which o ms pa o he C-loop. He e, he backbone amide N and ca bonyl O a oms o esidues 107 and 109 o m hyd ogen bonds in a classic an ipa allel -shee a angemen wi h he same esidues in a neighbou ing symme y- ela ed molecule (Fig. 4). This has he e ec o p oducing a la ge ex ended -shee o 12 s ands spanning wo ime s. Simila packing in e ac ions we e obse ed in he EcPII I2 1 3 s uc u e (Xu e al., 2001). The model o he SnPIIA p o ein consis s o a ime o almos iden ical subuni s. The T-loops a e clea ly o de ed as shown in Fig. 5. The main in e ac ion s abilizing he T-loops appea s o be main-chain hyd ogen bonds o ming -shee s in ol ing esidues 49±53 wi h hei symme y-equi alen esidues om ano he ime . Al hough Se 49 has been mu a ed o alanine in his s uc u e, i does no appea ha side-chain in e ac ions in ol ing he se ine would a ec he local con o ma ion o he loop. The side chains o esidues A g38, Gln39, Lys40, Gln42 A g47, Glu50, Gln57, Gln69 and Glu85 in all h ee subuni s ha e poo ly de®ned elec on densi y. The C- e minal loop has -shee opology. The e a e h ee Ni 2+ ions ha link he N- and C- e mini o each subuni , as shown in Fig. 6. In addi ion, he e a e 132 sol en molecules and six sul a es ha bind in he h ee ATP-binding cle s. PROCHECK (Laskowski e al., 1993) indica es ha 96.7% o he ', angles o all non-glycyl esidues a e in he mos a ou ed egions o he Ramachand an plo . The addi ional 3.3% we e in he addi ionally allowed egions. Inspec ion o he B ac o s again shows he cen al co e o in e locking -s ands o ha e he lowes mobili y (12 A Ê 2 ). All loops, including he T-, B- and C-loops, ha e alues ha a e app oxima ely doubled. The ex e nal helices a e in e - media e be ween hese wo alues. The nickel ions ha e B ac o s simila o he liganding esidues. The SO2ÿ 4ions ha e highe B ac o s han he nickel ions, close o ypical sol en alues. 3.2. Molecula s uc u es The e ia y s uc u es o he cyano- bac e ial SsPII and SnPIIA p o eins a e essen ially he same as he EcPII and GlnK s uc u es. Howe e , he e a e signi®can di e ences be ween he T- and C-loop egions o SsPII and SnPII and he same loops in EcPII and GlnK. The T-loop o he SsPII p o ein is dis- o de ed as has been obse ed in some EcPII and GlnK s uc u es. The T-loop in he SnPIIA p o ein ex ends om 2 and 3 wi hou a dis up ion o seconda y s uc u e, so ha hese s ands a e longe han hose ound in EcPII and GlnK. 2 o SnPIIA consis s o esidues 29±41, his being ® e mo e han he numbe o esidues in 2 (37± 41) o he EcPII molecule. Simila ly, 3o SnPIIA consis s o esidues 50±65, while 3 o he o he PII molecules consis s o esi- dues 56±65. The e is li le con ac be ween he ex ension o 2 and 3 and he co e o he Synechococcus PII molecule. The T-loop egion o SnPIIA akes pa in c ys al packing. Residues 49, 51 and 53 in a gi en monome o he ime o m hyd ogen bonds wi h hei symme y coun e pa s o one monome wi hin a neighbou ing ime . The speci®c in e ac ions in ol e backbone hyd ogen bonds be ween Ty 51 and i s c ys allog aphic equi alen , while Ala49 and Val53 a e linked o Val53 and Ala49, espec i ely, in a c ys allog aphically ela ed molecule. These in e ac ions appea o s abilize he T-loop. Figu e 5 T ime s o PII wi h isible T-loops. All a e s abilized by c ys al con ac s. Colou scheme: SnPII, magen a; GlnK, g een; EcPII, black. (a) Le , iew looking down he h ee old axis. Righ , pe pendicula o he h ee old axis looking in o he ATP-binding cle . (b) An omi map o he T-loop egion in SnPII. The C- e minus o he SsPII p o ein is e y simila o ha obse ed in a low-pH s uc u e o EcPII (Xu e al., 2001). In bo h, he C- e minus is in ol ed in c ys al con ac s. The con o ma ion o he C- e minus o SsPII di e s om ha obse ed in EcPII and GlnK s uc u es ob ained a highe pH alues. The con o ma ion o he C- e minus o SnPIIA di e s om hose o o he PII p o eins. I poin s owa ds he N- e minus o he same chain and is held in place by coo dina- ion wi h a me al ion, p esumably a Ni 2+ ha was added o c ys alliza ion. The coo dina ion si e in ol es Asp110, a esidue ha is no conse ed in ei he he SsPII o he EcPII p o eins. The me al-ion coo dina ion ies down he C- e minus and p obably acili a es c ys alliza ion, bu is no likely o ha e a physiological unc ion. SnPIIA binds wo anions in he ATP-binding cle . In he absence o ATP, GlnK was also obse ed o bind an anion in his cle . The GlnK anion-binding si e also appea s o be occupied in SnPIIA. This anion o ms in e ac ions wi h he backbone N a om o Gly87 and he side-chain N a oms o A g101 and A g103 om a neighbou ing monome wi hin he ime . The second anion o ms links o he side-chain N a oms o Lys90 and he neighbou ing A g101 as well as he symme y- ela ed A g103. The la e anion is exposed o sol en and has a highe B ac o . 4. Concluding ema ks The s uc u es obse ed o he cyanobac e ial PII p o eins a e e y simila o hose obse ed o he E. coli p o eins. The co e o he molecule is ema kably well p ese ed; his simi- la i y is pa icula ly ema kable when i is ecalled ha he cyanobac e ial PII p o eins in e ac wi h a se o ecep o s ha a e qui e di e en o hose o E. coli. This is consis en wi h he idea ha ecogni ion occu s h ough ¯exible loops and ha he ¯exibili y allows he ecogni ion o mul iple ecep o s (Xu e al., 1998). The T-loop in he p esen s uc u es is diso de ed in one case and in he o he is dependen upon c ys al packing. Clea ly, his piece o pep ide is highly ¯exible. The C- e minal pep ide is also capable o assuming mul iple con o ma ions al hough i ends o a ou a 3 10 -helical o a s and s uc u e. As no ed ea lie (Xu e al., 2001), i may a ou a helical s uc u e in solu ion, bu his s uc u e canno be pa icula ly s able. Indeed, he ¯exibili y o his pep ide may also be impo an o ecep o ecogni ion. The unc ion o ATP in he PII-like p o eins is no exac ly clea , ye i s binding pocke is highly conse ed in e ms o sequence (Xu e al., 1998). The cyanobac e ial PII s uc u es again show a highly conse ed binding cle . The e a e ea u es o he cyanobac e ia p o eins ha appea unique. A BLAST sea ch o ®nd o he p o eins ha we e simila o Synechococcus PII ga e a o al o 167 hi s. Amongs hese sequences, cyanobac e ia (Synechocys is, Nos oc punc i o me,F emyella diplosiphon,The mo- synechococcus elonga us BP-1 and P ochlo ococcus ma inus) ha e a conse ed Gln a he posi ion 57. All p o eobac e ia and ac inobac e ia ha e a p oline a posi ion 57 in hei GlnB and GlnK p o eins. O he o ganisms ha e a a iable non- p oline a he co esponding posi ion. These o ganisms include he a chaebac e ia, aqui®caceae, he ma ogales, deinococci, ed algae and dico yledenous plan s. Like he cyanobac e ia p o eins, hese p o eins a e no u idylyla ed as e idenced by he ac ha glnD (coding o UTase) does no exis in hei genomes. Residue 57 is loca ed in a sec ion o he molecule ha is likely o e ec ecep o ecogni ion. We sugges ha esidue 57 may be in ol ed in ecogni ion e en s ha a e speci®c o cyanobac e ia. We hank he Aus alian Resea ch Council o suppo ing his wo k. We hank Paul Pa ch o echnical assis ance. We also hank he High Pe o mance Compu ing Sec ion o James Cook Uni e si y o suppo and he Supe compu ing Facili y a he Aus alian Na ional Uni e si y o a g an o ime on hei machines. Re e ences Aldehni, M. F., Saue , J., Spielhaup e , C., Schmid, R. & Fo chhamme , K. (2003). J. Bac e iol. 185, 2582±2591. Ac a C ys . (2003). D59, 2183±2190 Xu e al. PII p o eins 2189 esea ch pape s Figu e 6 Ni 2+ ion linking he N- and C- e mini in SnPII. (a) Loca ion o he me al- binding si e. (b) Close-up showing he me al coo dina ion. Unbonded ed sphe es ep esen sol en molecules. esea ch pape s 2190 Xu e al. PII p o eins Ac a C ys . (2003). D59, 2183±2190 A condeguy, T., Jack, R. & Me ick, M. (2001). Mic obiol. Mol. Biol. Re . 65, 80±105. A kinson, M. R. & Nin a, A. J. (1999). Mol. Mic obiol. 32, 301±313. Benelli, E. M., Buck, M., Polika po , I., de Souza, E. M., C uz, L. M. & Ped osa, F. O. (2002). Eu . J. Biochem. 269, 3296±3303. B u Ènge , A. T., Adams, P. D., Clo e, G. M., DeLano, W. L., G os, P., G osse-Kuns le e, R. W., Jiang, J.-S., Kuszewski, J., Nilges, N., Pannu, N. S., Read, R. J., Rice, L. M., Simonson, T. & Wa en, G. L. (1998). Ac a C ys . D54, 905±921. Ca , P. D., Cheah, E., Su olk, P. M., Vasude an, S. G., Dixon, N. E. & Ollis, D. L. (1996). Ac a C ys . D52, 93±104. Cheah, E, Ca , P. D., Su olk, P. M., Vasude an, S. G., Dixon, N. E. & Ollis, D. L (1994). S uc u e,2, 981±990. Collabo a i e Compu a ional P ojec , Numbe 4 (1994). Ac a C ys . D50, 760±763. Cow an, K (1994). Jn CCP4/ESF±EACBM Newsl. P o ein C ys al- log . 31, 34±38. Fo chhamme , K. (2003). Symbiosis,35, 101±115. Fo chhamme , K. & de Ma sac, N. T. (1995a). J. Bac e iol. 177, 5812± 5817. Fo chhamme , K. & de Ma sac, N. T. (1995b). J. Bac e iol. 177, 2033± 2040. Ga cia-Dominguez, M. & Flo encio, F. J. (1997). Plan Mol. Biol. 35, 723±734. Heeswijk, W. C. an, S egeman, B., Ho ing, S., Molenaa , D., Kahn, D. & Wes e ho , H. V. (1995). FEMS Mic obiol. Le . 132, 153±157. Hisbe gues, M., Jeanjean, R., Jose , F., de Ma sac, N. T. & Be Âdu, S. (1999). FEBS Le . 463, 216±220. Hoo , R. W. W., V iend, G., Sande , C. & Abola, E. E. (1996). Na u e (London),381, 272. I mle , A. & Fo chhamme , K. (2001). P oc. Na l Acad. Sci. USA,98, 12978±12983. Jones, T. A., Zou, J. Y., Cowan, S. W. & Kjeldgaa d, M. (1991). Ac a C ys . A47, 110±119. Kabsch, W. (1976). Ac a C ys . A32, 922±923. Laskowski, R. A., MacA hu , M. W., Moss, D. W. & Tho n on, J. M. (1993). J. Appl. C ys . 26, 283±291. Lee, H. M., Flo es, E., Fo chhamme , K., He e o, A. & de Ma sac, N. T. (2000). Eu . J. Biochem. 267, 591±600. Lee, H. M., Flo es, E., He e o, A., Houma d, J. & de Ma sac, N. T. (1998). FEBS Le . 427, 291±295. Lo e, C. A., Lilley, P. E. & Dixon, N. E. (1996). Gene,176, 49±53. Ma hews, B. W. (1968). J. Mol. Biol. 33, 491±497. Me ick, M. J. & Edwa ds, R. A. (1995). Mic obiol. Re . 59, 604±622. Na aza, J. (1994). Ac a C ys . A50, 157±163. Nin a, A. J. & A kinson, M. R. (2000). T ends Mic obiol. 8, 172±179. O winowski, Z. & Mino , W. (1997). Me hods Enzymol. 276, 307±326. Ruppe , U., I mle , A., Klo , N. & Fo chhamme , K. (2002). Mol. Mic obiol. 44, 855±864. Samb ook, J., F i sch, E. F. & Mania is, T. (1989). Molecula Cloning: A Labo a o y Manual, 2nd ed. Cold Sp ing Ha bo : Cold Sp ing Ha bo Labo a o y P ess. Smi h, C. S., Aalim, M. W. & Moo head, G. B. G. (2003). Plan J. 33, 353±360. Vagin, A. & Teplyako , A. (2000). Ac a C ys . D56, 1622±1624. Xu, Y., Ca , P. D., Hube , T., Vasude an, S. G. & Ollis, D. L. (2001). Eu . J. Biochem. 268, 2028±2037. Xu, Y., Cheah, E., Ca , P. D., an Heeswijk, W. C., Wes e ho , H. V., Vasude an, S. G. & Ollis, D. L. (1998). J. Mol. Biol. 282, 149± 165.