Structural and functional insights into the C-terminal signal domain of the Bacteroidetes type-IX secretion system

S uc u al and unc ional insigh s in o he C- e minal signal domain o he Bac e oide es ype-IX sec e ion sys em Danu a Mizgalska1,†, A u o Rod íguez-Banque i2,†, Flo ian Veilla d1, Mi osław Książęk1, Theodo os Goulas4, Tibisay Gue a a2, Ul ich Eckha d3, Jan Po empa1,5 and F. Xa ie Gomis-Rü h2 1Depa men o Mic obiology, Facul y o Biochemis y, Biophysics and Bio echnology, Jagiellonian Uni e si y, K aków, Poland 2P o eolysis Labo a o y, Depa men o S uc u al Biology, and 3Syn he ic S uc u al Biology G oup, Depa men o S uc u al Biology, Molecula Biology Ins i u e o Ba celona (CSIC), Ba celona, Ca alonia 08028, Spain 4Depa men o Food Science and Nu i ion, School o Ag icul u al Sciences, Uni e si y o Thessaly, Ka di sa 43100, G eece 5Depa men o O al Immunology and In ec ious Diseases, Uni e si y o Louis ille School o Den is y, Louis ille, KY 40202, USA UE,0000-0001-5863-4514; FXG-R,0000-0002-6848-6874 G am-nega i e bac e ia om he Bac e oido a phylum possess a ype-IX sec e ion sys em (T9SS) o p o ein sec e ion, which equi es ca goes o ha e a C- e minal domain (CTD). S uc u ally analysed CTDs a e om Po phy omonas gingi alis p o eins RgpB, HBP35, Po U and Po Z, which sha e a compac immunoglobulin-like an ipa allel 3+4 β-sandwich (β1– β7). This a chi ec u e is essen ial as a P. gingi alis s ain wi h a single- poin mu an o RgpB dis up ing he in e ac ion o he CTD wi h i s p eceding domain p e en ed sec e ion o he p o ein. Nex , we iden i ied he C- e minus (‘mo i C- .’) and he loop connec ing s ands β3 and β4 (‘mo i Lβ3β4’) as conse ed. We gene a ed wo s ains wi h inse ion and eplacemen mu an s o Po U, as well as h ee s ains wi h abla ion and poin mu an s o RgpB, which e ealed bo h mo i s o be ele an o T9SS unc ion. Fu he mo e, we de e mined he c ys al s uc u e o he CTD o mi olase, a ca go o he Tanne ella o sy hia T9SS, which sha es he same gene al opology as in Po phy omonas CTDs. Howe e , mo i Lβ3β4 was no conse ed. Consis en ly, P. gingi alis could no p ope ly sec e e a chimae ic p o ein wi h he CTD o pep idyla ginine deiminase eplaced wi h his o eign CTD. Thus, he incompa ibili y o he CTDs be ween hese species p e en s po en ial in e e ence be ween hei T9SSs. 1. In oduc ion The in e ace be ween bac e ia and hei ex acellula en i onmen is pe mea ed by sec e ion machine ies, which a e mul ip o ein ga eways o he anspo o ca goes om he cy osol ac oss he cy oplasmic inne memb ane (IM; ‘p o ein expo ’) and, when p esen , ou e memb ane (OM; ‘p o ein sec e ion’) [1–4]. These sys ems a e c i ical o bac e ial iabili y a a si e o in ec ion o coloniza ion as hey se e nu ien acquisi ion and communi- ca ion wi h o he bac e ia, o en ac oss species ba ie s [2,3]. They u he enable bac e ial adhesion o bio ilms, S-laye o ma ion and gliding mo ili y, as well as he sec e ion o i ulence ac o s o disa m hos de ences and compe ing bac e ia a he si e o coloniza ion/in ec ion [3–5]. One o he 11 sec e ion sys ems [3] ha a e cu en ly known o exis is he ype-IX sec e- ion sys em (T9SS) [6], which was ea lie known as he ‘Pe ioGa e’ and ‘Po © 2024 The Au ho s. Published by he Royal Socie y unde he e ms o he C ea i e Commons A ibu ion License h p://c ea i ecommons.o g/licenses/by/4.0/, which pe mi s un es ic ed use, p o ided he o iginal au ho and sou ce a e c edi ed. Resea ch Ci e his a icle: Mizgalska D, Rod íguez- Banque i A, Veilla d F, Książęk M, Goulas T, Gue a a T, Eckha d U, Po empa J, Gomis-Rü h FX. 2024 S uc u al and unc ional insigh s in o he C- e minal signal domain o he Bac e oide es ype- IX sec e ion sys em. Open Biol. 14: 230448. h ps://doi.o g/10.1098/ sob.230448 Recei ed: 12 Decembe 2023 Accep ed: 20 Feb ua y 2024 Subjec A eas: biochemis y, s uc u al biology, mic obiology Keywo ds: pe iodon al disease, bac e ial i ulence ac o , in ec ious disease, p o ein sec e ion, X- ay c ys al s uc u e, T9SS Au ho s o co espondence: Ul ich Eckha d e-mail: uecc [email protected] Jan Po empa e-mail: jan.po [email protected] F. Xa ie Gomis-Rü h e-mail: xg c [email protected] †Join i s au ho s. Elec onic supplemen a y ma e ial is a ailable online a h ps://doi.o g/10.6084/ m9. igsha e.c.7272136. Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 Sec e ion Sys em’ [7–9]. I is widely ound in he G am-nega i e Bac e oide es (synonymous wi h Bac e oido a) phylum o he ‘FCB supe phylum’ [10], and sequences o T9SS co e p o eins ha e been iden i ied in he genomes o 90 species o his phylum, wi h he no able excep ion o Bac e oides spp. [11]. T9SSs may also be p esen in o he FCB phyla [12], including Chlo obio a, Igna ibac e io a, Rhodo he mo a, Gemma imonade es, Cloacimonado a and Fib obac e o a [6,11,13–15]. Rema ka- bly, T9SS genes a e absen om o he bac e ial phyla, a chaea and euka yo es [11]. T9SSs o hei componen p o eins ha e been expe imen ally s udied om Candida us pa apo phy omonas polyenzymogenes; Capnocy ophaga och acea; Cellulophaga algicola and Cellulophaga omni esco ia; Cy ophaga hu chinsonii; Dyadobac e sp.; Fib obac e succinogenes; Fla obac e ium columna e, Fla obac e ium johnsoniae, Fla obac e ium collinsii, Fla obac e ium spa ansii and Fla obac- e ium psych ophilum; Pa abac e oides dis asonis; P e o ella in e media and P e o ella melaninogenica; Po phy omonas gingi alis; Rieme ella ana ipes i e ; Rosei he mus saccha iphilus; Spo ocy ophaga myxococcoides; Tanne ella o sy hia; and Tenacibaculum ma i imum [4,7,9,11,14,16–35]. Many o hese species in ec animals, among which P. gingi alis and T. o sy hia a e human pe iodon opa ho- gens ha p oli e a e in he gingi a unde dysbio ic condi ions o he o al mic obiome. Toge he wi h T eponema den icola, hey o m he ‘ ed complex’ [36] ha causes g ie ous gingi i is and pe iodon al disease, which a ec s an es ima ed 750 million people wo ldwide in i s se e e o m and is he six h mos p e alen disabling heal h condi ion [37–40]. The T9SS is a specialized OM shu le o p o ein ca goes ha includes a o a y mo o , a anslocon, an OM-associa ed sca old and an a achmen complex [4,41]. I ope a es in conjunc ion wi h he gene al sec e o y (Sec) pa hway o he p e ious expo o ca goes om he cy osol ac oss he IM [4,41]. Once in he pe iplasm, he N- e minal Sec signal pep ide is emo ed, and ca goes old and a e esco ed o he T9SS anslocon o sec e ion. The bes -cha ac e ized T9SSs a e om F. johnsoniae and P. gingi alis [4,7,9,28]. In he la e , i is he majo p o ein sec e ion pa hway and is con o med o egula ed by a leas 24 ‘Po ’ p o eins [4,9,41]. I sec e es >35 ca go p o eins, which include essen ial i ulence ac o s o in ec ion such as he gingipain cys eine pep idases Kgp, RgpA and RgpB [42]; Po phy omonas pep idyla ginine deiminase (PPAD) [8,43]; a 70 kDa me alloca boxypep idase (CPG70) [44]; and a 35 kDa haemin-binding p o ein (HBP35) [27]. T9SS anspo equi es a C- e minal domain (CTD) o 70–100 esidues, also known as he ‘T9SS signal’ [16,17,41,45–50], which con as s wi h he sho uns uc u ed signal pep ides o o he sec e ion sys ems [2,51]. In ac , all Bac e oide es species po en ially encompassing a T9SS also encode pu a i e ca go p o eins ea u ing a CTD and, ice e sa, species lacking T9SS genes also miss-p edic ed p o eins wi h a CTD [11]. Thus, T9SS ca goes a e also dubbed ‘CTD p o eins’ [46,52]. In P. gingi alis, he CTD is gene ally chopped o du ing o a e OM ansloca ion by he C- e minal T9SS signal pep idase Po U [53], which ope a es wi hin an ‘a achmen complex’ wi h Po Q, Po V and Po Z [4,41,54], and ca goes a e eleased o he en i onmen . In a subse o hem, howe e , anionic lipopolysaccha ide p o ided by Po Z is a ached o he new C- e minus by Po U, which hus also ac s as a so ase- ype anspep idase [53,55,56]. These ca goes a e ancho ed o he ex acellula side o he OM [17,57– 60]. No ably, some T9SS cons i uen s ha a e hemsel es sec e ed, such as Po A, Po U and Po Z, also possess a CTD, which, howe e , is no emo ed upon sec e ion [5,41,61]. Thus, he e a e Po U-p ocessed and Po U-unp ocessed CTDs. The CTD is appa en ly necessa y and su icien o T9SS sec e ion [4,11,52,62]. Mo eo e , CTDs a e unc ionally exchangea- ble be ween ca goes wi hin a gi en species, as shown o P. gingi alis Kgp wi h he CTD o RgpB [45]. Wha is mo e, he CTDs o P. gingi alis ca go p o eins RgpB, PPAD, CPG70 o P27 used o a g een luo escen p o ein caused he esul ing chimae as o be sec e ed and ancho ed o he OM by his bac e ium [45,52]. In F. johnsoniae, he same luo escen p o ein and mChe y, u nished wi h he CTD o he RemA adhesin [11] o he ChiA chi inase [47], we e sec e ed by he T9SS. CTDs may also unc ion ac oss species wi hin Bac e oide es, as shown o he Fla obac e iia class. Indeed, C. algicola AmyA α-amylase was sec e ed by he T9SS o F. johnsoniae [11]. Ac oss classes, he la e anslocon ecognized he CTD o C. hu chinsonii (class Cy ophagia) Cel9B cellulase bu no ha o P. gingi alis (class Bac e oidia) RgpB [11]. Mo eo e , he CTD o C. hu chinsonii Cel9A cellulase is equi ed o be N-glycosyla ed in he pe iplasm p io o sec e ion and cell-su ace exposu e in wha esembles an addi ional s ep o egula ion in his species [63]. O e all, hese indings pinpoin species-speci ic di e ences ac oss he T9SS signals and dispa a e in e speci ic compa ibili y. Al hough CTDs ha e been classi ied in o ypes A, B and C, sequence iden i ies a e gene ally e y low and he e a e no ob ious mo i s a ibu able o ca go ansloca ion and/o clea age h ough he C- e minal T9SS signal pep idase [4,11]. This has uelled he hypo hesis ha conse ed s uc u al elemen s a he han sequence s e ches may be esponsible o unc ion [11]. In his sense, he s uc u es o he CTDs o RgpB [49,64] and HBP35 [27], which a e emo ed upon sec e ion, as well as hose o he non-clea ed CTDs o Po Z [5] and Po U [41], ha e been de e mined. All co espond o P. gingi alis p o eins and no CTDs om o he species ha e been epo ed o da e. He e, we sough o iden i y molecula de e minan s o T9SS sec e ion and Po U-clea age in P. gingi alis by compa ing he ou a o emen ioned s uc u es wi h wo ad hoc high-con idence homology models and e i ied hem wi h a coho o mu an s ains ha we e subjec ed o cellula and unc ional assays. Fu he mo e, we de e mined he c ys al s uc u e o he CTD o he T. o sy hia T9SS ca go mi olase [65] (dubbed ‘LKK’; om L706 o K–K791, see UniP o en y (UP) A0A0A7KVG3 o esidue numbe ing), he i s ou side P. gingi alis, and assessed i he abo e s uc u al de e minan s we e conse ed ac oss species. Finally, we checked he CTD compa ibili y o T. o sy hia and P. gingi alis. 2. Resul s and discussion 2.1. Molecula de e minan s o P. gingi alis C- e minal domains o T9SS unc ion We compa ed he CTD s uc u es o Po Z [5], Po U [41], HBP35 [27] and RgpB [49] wi h he calcula ed high-con idence compa a i e models (see §3.10) o he CPG70 and PPAD CTDs ( igu e 1a). We ound ha he gene al opology and connec i i y, 2 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 as i s desc ibed o RgpB [49], we e main ained. All molecules consis ed o a cen al immunoglobulin-like moie y [66–68], which ea u es an an ipa allel se en-s anded β-sandwich wi h a ou -s anded (β4–β3–β6–β7) and a h ee-s anded (β5–β2–β1) β-shee , wi h in e shee angles o ~30–40°C and G eek-key s and connec ions ( igu e 1a). Howe e , wi h he excep ion o he loop connec ing s ands β3 and β4 (Lβ3β4), each s uc u e displayed la ge a ia ions in he leng h o he s ands and he connec ing loops, which led o a he high msd and low sequence-iden i y alues in pai wise compa isons ( able 1). We hen pe o med a s uc u e-assis ed sequence alignmen and included a o al o 12 CTDs om epo ed P. gingi alis ca goes, which con i med he lack o signi ican sequence simila i y ( igu e 1b). We ound ha Po U-p ocessed ca goes a e gene ally sho e and span 67–88 esidues om he p edic ed/de e mined Po U clea age si e [50,53,55] o he C- e minus, while Po Z and Po U encompass 96 and 108 esidues, espec i ely. Thus, CTDs signi ican ly a y in leng h. P e ious s udies wi h a coho o 16 poin mu a ions ac oss he CTD o RgpB had e ealed no pa icula esidue side chain indispensable o expo [69,70]. Thus, we hypo hesized ha an in ac well- olded CTD and, possibly, pa icula h ee-dimen- sional s uc u al mo i s would be equi ed o sec e ion and Po U-media ed clea age/ anspep ida ion, as p e iously sugges ed [11]. To alida e he impo ance o o e all CTD in eg i y, we iden i ied RgpB esidue A719 (numbe ing acco ding o UP P95493), which is he cen al esidue o he penul ima e β-s and β6 o he CTD and is engaged in in e ac ions wi h he ups eam p o ein domain [49]. We gene a ed a P. gingi alis s ain wi h he alanine eplaced wi h a ginine (mu an A719R). To pe o m such RgpB mu a ions, we ou inely employ he RgpA-de icien mu an s ain RgpA-C o p e en he highly homologous gpA gene om in e e ing wi h he gene ic manipula ion o gpB and RgpB p o ein de ec ion [16,71]. We compa ed he occu ence o he mu an p o ein in dis inc cell-cul u e ac ions and ex acellula gingipain ac i i y wi h hose o he wild- ype and RgpA-C as posi i e con ols. We ound ha he mu an p o ein was no sec e ed o he OM bu accumula ed in he pe iplasm ( igu e 2a), and he s ain lacked ex acellula RgpB p o eoly ic ac i i y ( igu e 2b). This is indica i e o T9SS mal unc ion o his ca go [16,41,56]. In con as , Kgp ac i i y was una ec ed, which documen s he co ec unc ion o he T9SS o o he ca goes ( igu e 2b). As o pa icula CTD elemen s ha may be ele an o T9SS unc ion, we iden i ied wo s uc u al mo i s, ‘mo i C- .’ and ‘mo i Lβ3β4’ ( igu e 1a,b). They a e conse ed and loca ed on he su ace, nea he op edge o he domain ( igu e 1a,c) and, hus, po en ially accessible o in e ac ion wi h he T9SS anslocon and subsequen Po U clea age [41]. 2.2. Mo i C- . P e ious s udies ha e unde lined he impo ance o he CTD C- e minal end o unc ion [11]. In pa icula , he unca ion o wo esidues—bu no one—a he C- e minus o RgpB c ea ed an inac i e a ian ha is e ained in he pe iplasm [16]. Mo eo e , he T9SS a ge ing signal o HBP35 was localized o he las 22 esidues o he CTD [52]. Inspec ion o he supe posed CTD s uc u es e ealed ha he main chains o he C- e minal β-s and coincide ( igu e 1a), which sugges s hey may be impo an o s uc u al in eg i y and/o unc ion o he whole domain. Mo eo e , a conse ed lysine is ound in equi alen posi ions o s and β7 ( igu e 1b). To alida e his hypo hesis, we cons uc ed and assayed P. gingi alis s ains ea u ing wo Po U mu an s, in which he las six esidues we e eplaced wi h hexahis idine ( eplacemen mu an 1152sHis) o eigh his idines we e added o he C- e minus (ex ension mu an Po UHis). By using he wild- ype and a Po U dele ion mu an (ΔPo U) as posi i e and nega i e con ols, espec i ely, we ound ha in he Po UHis mu an , Po U and RgpB we e co ec ly ansloca ed o he OM su ace and he CTD o RgpB was p ope ly emo ed by Po U, as ound in he wild- ype ( igu e 2c). Mo eo e , he mu an e inced ex acellula RgpB and Kgp ac i i y indis inguishable om he wild ype ( igu e 2d). In con as , mu an 1152sHis mimicked he ΔPo U dele ion mu an in ha RgpB was no expo ed o p ocessed co ec ly, so i s CTD was s ill a ached ( igu e 2c). Mo eo e , mu an Po U p o ein accumula ed in he pe iplasm. Finally, again, as in ΔPo U, no ele an ex acellula RgpB o Kgp ac i i y was de ec ed ( igu e 2d). O e all, hese esul s indica e gene al mal unc ioning o he T9SS due o abe an Po U. Al oge he , he indings o he wo mu an s con i m he impo ance o he esidues cons i u ing he C- e minal mo i o ansloca ion, bu he C- e minus may be ex ended wi hou consequences. This is consis en wi h epo s o Po Z, o which elonga ion o he C- e minus by his idine ags likewise did no cause unc ional T9SS impai men , bu eplacemen o he las esidues wi h his idines did [5]. Fu he mo e, ex ension o he C- e minus by polyhis idine ags, as well as dele ion o he las wo esidues—bu no ou esidues—did no a ec he sec e ion and OM a achmen o PPAD. In ac , he CTDs o Po Z and Table 1. Pai wise compa ison o expe imen al and p edic ed P. gingi alis CTD s uc u es. Po U Po Z RgpB HBP35 PPAD Po Z 80; 105/89; 3.0; 6 RgpB 64; 105/65; 3.0; 6 61; 89/65; 1.8; 20 HBP35 29; 105/59; 2.8; 0 54; 89/59; 2.3; 22 57; 65/59; 2.0; 30 PPAD 77; 105/81; 1.9; 13 72; 89/81; 2.0; 10 59; 65/81; 2.7; 8 54; 59/81; 2.6; 7 CPG70 72; 105/76; 2.4; 15 70; 89/76; 2.2; 13 60; 65/76; 2.1; 27 56; 59/76; 2.7; 9 72; 81/76; 1.9; 18 Fo each pai o s uc u es, he numbe o aligned esidues ou o he o al esidues o each p o ein, he o e all msd (in Å) and he sequence iden i y (in %) a e indica ed. 3 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 Po U e ince one and h ee ex a C- e minal esidues, espec i ely, when supe posed on o hose o he o he ca goes ( igu e 1a,b). 2.3. Mo i Lβ3β4 All CTD models showed an excellen i o he segmen encompassing Lβ3β4 ( igu e 1a,c), which sugges s a possible consensus ole in T9SS unc ion. This inding is consis en wi h a p e ious dele ion mu an o RgpB lacking L692–V702, which supp essed he sec e ion o his ca go [69]. Howe e , as his mu an had been designed be o e he s uc u e o he RgpB CTD was a ailable [49], we hypo hesized ha an in e nal dele ion o 11 esidues abla ing no only he connec ing loop bu also mos o s ands β3 and β4 migh ha e caused he collapse o he en i e domain ins ead o selec i ely abla ing a speci ic s uc u al elemen in an o he wise in ac sca old. Thus, we ob ained a mo e conse a i e s uc u e-based dele ion s ain ha should keep he o e all s uc u e well olded, in which F695–D–M–N–G–R–R701 was eplaced wi h wo glycines (mu an 695–701:GG). We ound ha in his mu an , RgpB accumula ed in he pe iplasm and was no sec e ed and ha ex acellula RgpB ac i i y—bu no Kgp ac i i y—was signi ican ly diminished ( igu e 2a,b). Mo eo e , he a o emen ioned s udy by Slakeski e al. also epo ed ha eplacemen in RgpB o he only conse ed esidue o he Lβ3β4 mo i , a glycine (G699) in he i s posi ion o he ou h s and ( igu e 1b), wi h p oline had no signi ican dele e ious e ec [69]. We hypo hesized ha his eplacemen may no ha e had a s uc u al impac la ge enough o cause mo i dis up ion. To e i y his, we eplaced his glycine wi h a ginine (mu an G699R) and ound ha his mu a ion ac ually hinde ed RgpB sec e ion and p ocessing o i s ac i e ma u e o m ( igu e 2a). Mo eo e , ex acellula RgpB ac i i y was abla ed ( igu e 2b). Thus, his glycine is unc ionally ele an and pinpoin s mo i Lβ3β4, na owed down o he enci cled egion in igu e 1c, as essen ial o CTD ecogni ion and ca go sec e ion by he T9SS anslocon. mo i Lβ3β4 mo i C- . SS-elemen s (PDB 5AG8) SS-elemen s (PDB 5AG8) RgpB (UP P95493) 654 VIKDVKVEGTSIADVANDKPYTVAVSG--------KTITVESP-----AAGLTIFDMNGRRVATAKN-------RMVFEAQ---NGVYAVRIATE-GKTYTEKVIVK PKVCVDYIPDGVADVTAQKPYTLTVVG--------KTITVTCQ------GEAMIYDMNGRRLAAGRN-------TVVYTAQ---GGYYAVMVVVD-GKSYVEKLAVK AIETATLNITSLADVTAQKPYTLTVVG--------KTITVTCQ------GEAMIYDMNGRRLAAGRN-------TVVYTAQ---GGHYAVMVVVD-GKSYVEKLAVK YYRIRLITDLSSSDIAGKDVSTIVLYPNPAHDYVHVAIPPTYAG-----STLRLFDIQGRMQLSTKIESA-DMRLDVERLP---KGTYIVVVED-----MVGKLFIR VAVTVTVTNSSLSNVDGQAPYTLRVEG--------KKIIAEAH------GMITLYDINGRTVAVAPN-------RLEYMAQ---TGFYAVRFDVG-NKHHVSKIQVR AIETATLNITSLADVTAQKPYTLTVVG--------KTITVTCQ------GEAMIYDMNGRRLAAGRN-------TVVYTAQ---GGHYAVMVVVD-GKSYVEKLAVK NYFYTDMQGNALTDVAVNESIKIYPRPAT-----DFLRIEGSQL-----LRLSLFDMNGKLIRATELTGD-LAIIGVASLP---RGTYIAEITAANSKTIRAKVSLR WSDTQWTHANGVEDIVMQEGSMKLYPNPAQE--YAVISLPTAAN-----CKAVVYDMQGRVVAEASFSGN-EYRLNVQHLA---KGTYMLKVVSD-TERFVEKLIVE KVCKNIVVANSANIYGADKPFALTVVG--------KTIVASAFK-----GEITLYDIRGRLIASGCD-------TLRYKAE---NGFYLIKIQVN-GTVYTEKIQIQ EDYIEVILDDSVEDIVAQTGIVIRPQNG------TKQILIEANA---PIKAIVLYDINGRVVLKTTPNQL----RSTVDLSILPEGIYTINIKTE-KSARTEKIHIG 736 991 1732 840 293 2164 390 540 352 821 556 344 910 1651 748 212 298 446 270 729 459 268 680 1048 GTGSGSASELDGVYVYPNPLRPEYPDGVTIAGLQA----GCSVKITDTTGRLLYQTESV--------TTEVKWNARGADGNRVASGVYAVAVYDP--VSKKSKLIRFAVIR-- 776 1158 RVVDGIAPDVADVILFPNPVRE--SATFRIFHNRPGSDLNVVVEIYDFTGRLVNSLPVKTYSSSYGEPIEIKWDLTSKYGVKIGNGFYLYRCVVNSPGGQTASMAKKMIVVAQ 2083 RgpA (UP P28784) Kgp (UP Q51817) P T (UP A0A1R4AEF2) P H-like (UP T2N9GB) HagA1 PG1837(UP P59915) PG0654 (UP Q7MWG3) PG1326 (UP Q7MUZ4) CPG70 (UP A0A1R4AHF7) PPAD (UP Q9RQJ2) HBP35 (UP Q8G962) Po Z (UP Q9S3Q8) Po U (UP B2RGP6) TapA P59 (UP A0A1R4DTQ3) ssssss ssssss ssssss sssssss sssssssss ssss ssss sssssss ssssssssss sssssssss ssssss sssssss ssssssss sssss sssssssss sssssssss SS-elemen s (PDB 5M11) SS-elemen s (PDB 6ZA2) KFTCMNETNTCTVTGAAKALRAWFNAG-------RSELAVSVSLNIAGTYRIKLYNTAGEEVAAMTKELVAGTSVFSMDVYSQAPGTYVLVVEGN-GIRETMKILK- QESLDKAEPTATEQIVATPSVKAYVQN--------GKIVVEEE-----YSKMEVFNATGQLVK-------------NESLV---PGVYVVRITAN-GVMHFLKVLVP ssss ssss sss sss sssssssssss sssssssssss sssssssssssssssssssssssssssssssssssssssssssssss mo i Lβ3β4 mo i C- . ssssssssssssssssssssssssss β4 β5 β6 β7 β3 β2 β1 β4 β1 β8 β9 β6 β7 β3 β2 β1 (a) (b) (c) Figu e 1. S uc u al compa ison o P. gingi alis CTDs. (a) O e all supe posi ion in c oss-eye s e eo o he Cα- aces o he non-clea ed CTDs o Po U (P o ein Da a Bank (PDB) access code 6ZA2; A1054–Q1158, o UP codes, see (b); cyan) and Po Z (PDB 5M11; E688–R776; plum), as well as o he clea ed CTDs o RgpB (PDB 5AG8; K672–K736; ligh g een), HBP35 (PDB 5Y1A; P287–P344; salmon), CPG70 (p edic ed model; Q746–G821; gold) and PPAD (p edic ed model; K476–K556; da k g ey). The se en consensus s ands a e labelled β1–β7, as well as he wo essen ial s uc u al mo i s o T9SS unc ion (①, mo i C- . and ②, mo i Lβ3β4) and he non- unc ional β-hai pin wi hin Lβ5β6 (③). (b) S uc u e-based sequence alignmen s o he clea ed CTDs o selec ed ca goes ( op alignmen ) and he non-clea ed CTDs o Po Z and Po U (bo om alignmen ). The espec i e UP access codes and lanking esidue numbe s a e p o ided. Residues in β-s and con o ma ion (‘SS-elemen s’) acco ding o he espec i e PDB en ies o RgpB and Po Z (abo e he espec i e alignmen block) and HBP35 and Po U (below he espec i e alignmen block) a e displayed in o ange. They a e ea ma ked wi h a magen a ‘s’ and labelled β1–β7. Po U and Po Z ha e an ex a β- ibbon (s ands β8+β9) inse ed a e he i h s and. P edic ed s and esidues o he wo homology models (CPG70 and PPAD) a e in blue. The las esidues o he espec i e ups eam domains o s uc u ally analysed p o eins a e shown in magen a. De e mined o pu a i e clea age si es by Po U a e lanked by esidues in ed [50,53,55]. Mo i s Lβ3β4 and C- . a e boxed and a e simila o egions B and E o [46]. (c) Supe posi ion in s e eo o mo i Lβ3β4 ( ed ellipse, end-on iew) om Po U ( esidues I1090–V1098, ca bons in cyan), Po Z (I720–L728, plum), RgpB (I694–V702, ligh g een), HBP35 (V287–V295, an), CPG70 (L773–V781, gold) and PPAD (L505–V513, g ey). The esidue numbe s co espond o he espec i e UP en ies in (b). The o ien a ion esul s om ha o (a) a e successi e e ical and ho izon al o a ions o 120°C and 40°C, espec i ely. 4 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 2.4. The β-hai pin wi hin Lβ5β6 o Po U and Po Z is unc ionally no ele an Upon examining Po U-p ocessed and Po U-unp ocessed CTDs o any no iceable a ia ions, we ound ha Po U and Po Z sha e an addi ional β-hai pin a he end o he i h s and, which is no p esen in he o he ou models ( igu e 1a,b). This p o usion, which accoun s o mos o he ex a esidues in he wo Po p o eins when compa ed wi h Po U-p ocessed CTDs, should hampe any unc ional in e ac ion Lβ5β6 migh pe o m, so we hypo hesized ha his elemen migh be engaged in Po U binding o ca go p ocessing wi hin he a achmen complex. To alida e his hypo hesis, we cons uc ed a mu an s ain in which E712–A–Q714 o RgpB was eplaced wi h N–A–R–G–G–D–G–N–S–V–A (mu an 712–714:NARGGDGNSVA). This mimics he inse ion o Po Z excep o a se ine eplacing an a ginine ( igu e 1b). Unexpec edly, we ound ha in his mu an RgpB was co ec ly expo ed, clea ed by Po U and ancho ed on he su ace in a manne indis inguishable om he RgpA-C posi i e con ol ( igu e 2a), and ex acellula gingipain ac i i y was no a ec ed ( igu e 2b). Thus, mo i Lβ5β6 is appa en ly no ele an o dis inguish be ween ansloca ion and Po U p ocessing, so he disce ning ea u es be ween hese wo unc ions emain o be de e mined. 2.5. C ys al s uc u e o a C- e minal domain o hologue om T. o sy hia The LKK p o ein domain om T. o sy hia mi olase was ecombinan ly p oduced in Esche ichia coli wi h selenome hionine eplacing me hionine as an N- e minal usion wi h glu a hione-S- ans e ase, and pu i ied by a ini y and size-exclusion An i-RgpB (a) (b) (c) (d) WT 695-701:GG G699R 712-714 A719R RgpB ac i i y RgpB ac i i y Kgp ac i i y Kgp ac i i y WC WT RgpA-C 695-701:GG G699R A719R 712-714:NARGDGNSVA WT WT ∆Po U Po UHis 1152sHis WT ∆Po U Po UHis 1152sHis RgpA-C 695-701:GG G699R A719R 712-714:NARGDGNSVA Medium WC Medium WC Medium WC Medium Rela i e ac i i y 100 80 60 40 20 0 Rela i e ac i i y Rela i e Kgp ac i i y Rela i e Rgp ac i i y 100 80 60 40 20 0 0 50 100 0 50 100 NARGDGNSVA RgpA-C S C OM PP WT ∆Po U 1152sHis Po UHis C OM PP C OM PP C OM PP C OM PP CS OM PP CS OM PP CS OM PP CS OM PP S C OM PP S C OM PP S C OM PP S C OM PP S C OM PP 250 130 100 70 55 35 25 An i-CTD 250 130 100 70 55 35 25 An i-Po U 250 130 100 70 55 35 25 An i-RgpB 250 130 100 70 55 35 25 An i-CTD 250 130 100 70 55 35 25 Figu e 2. Cell- ac ion analysis and p o eoly ic ac i i y o RgpB and Po U mu an s ains. (a) Wes e n blo ing analysis o RgpB-mu an s ains in he supe na an (S), whole-cell ex ac (C), ou e -memb ane (OM) ac ion and pe iplasmic/cy oplasmic ac ion (PP) employing polyclonal an ibodies agains RgpA and RgpB ( op ow, pAb GP1) o he CTD o RgpB (bo om ow). Red squa e b acke s pinpoin memb ane- ype RgpB and as e isks deno e he isola ed RgpA ca aly ic domain. The la e is missing in he RgpA-C s ain and he mu an s gene a ed wi h a RgpA-C empla e. The absence o gingipain ac i i y esul s om de icien p o eoly ic p ocessing o he espec i e zymogens, which equi e he en i e p odomain o be emo ed and deg aded o elease ac i i y [56]. (b) RgpB ( op panel) and Kgp (bo om panel) ac i i y ela i e o he wild- ype (100%) o he RgpB-mu an s ains o (a) in whole-cell (WC) cul u es and he cul u e medium a e cell cen i uga ion. Expe imen s we e pe o med in iplica e. (c) Wes e n blo ing analysis o Po U-mu an s ains employing he monoclonal an ibody 7G9 agains Po U ( op ow) and he polyclonal an ibodies o (a) agains RgpB and i s CTD (middle and bo om ows). Red squa e b acke s pinpoin memb ane- ype RgpB; as e isks deno e he isola ed RgpA ca aly ic domain; open iangles hallma k he unp ocessed, ull-leng h 80 kDa RgpB zymogen including he in ac N- e minal p odomain and he CTD; and ull iangles designa e he N- e minally unca ed RgpB zymogen. (d) RgpB (le panel) and Kgp ( igh panel) ac i i y o he Po U-mu an s ains o (c) in whole cul u es (WC) and he cul u e medium a e cell cen i uga ion. 5 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 ch oma og aphy s eps (see §3.7). A e ag emo al, he p o ein was c ys allized in he o m o well-di ac ing igonal c ys als (see igu e 3a and able 2). Howe e , s uc u e solu ion and e inemen p o ed excep ionally di icul owing o he simul aneous occu ence o ansla ional non-c ys allog aphic symme y ( NCS; igu e 3b) and hemihed al winning, which led o space-g oup ambigui y (see §3.9). E en ually, he s uc u e was sol ed in he p imi i e space g oup P3121 by single-wa e- leng h anomalous di ac ion (SAD) wi h da a collec ed a he selenium K-edge, bu i could only be e ined in P31 a e he di ac ion da a had been de winned ( able 2). All his no wi hs anding, he inal s uc u e ( igu e 3c), wi h six p o ome s in he c ys allog aphic asymme ic uni (a.u.), was backed by solid Fou ie map densi y ( igu e 3d) and exhibi ed good alida ion pa ame e s ( able 2). (a) (c) (d) ( ) (g) (e) (b) Figu e 3. C ys allog aphic s udies o T. o sy hia LKK. (a) T igonal c ys al o space g oup P31 o selenome hionine-de i a ized LKK di ac ing o 1.6 Å esolu ion. (b) Na i e Pa e son map sec ion (axes u, , w) iewed down he w axis o ac ion 0.65–0.71. A peak o 72% he heigh o he o igin peak a ac ional coo dina es 0.331, 0.663 and 0.678 accoun s o s ong ansla ional non-c ys allog aphic symme y, which obscu es hemihed al winning ollowing win law h,–h–k,–l (es ima ed win ac ion α = 0.423). (c) Ribbon- ype plo in c oss-eye s e eo o LKK, which encompasses eigh β-s ands as salmon a ows (β1–β7 plus β2’) connec ed by coil egions in cyan. The lanking esidues o each s and a e numbe ed in ed, and he N- and he C- e minus a e labelled in black. In his iew, he molecule has been o a ed e ically ~180° wi h espec o igu e 1a o cla i y. (d) De ail o he inal σA-weig hed (2mFobs–DFcalc)- ype Fou ie map, con ou ed a 1σ abo e he h eshold, cen ed on he h ee β-s ands (β1, β2 and β5) o he on shee o molecule A. The inal e ined model is shown o segmen s L706–S709, T718–L721 and P759–Q762, app oxima ely in he iew o (c). (e) Same as (c) a e a e ical 90° o a ion. ( ) Va ian o (e) depic ing he side chains ha o m he hyd ophobic co e o he LKK moie y, wi h ca bons in ligh g een. The pa icipa ing esidues a e L706, L708, A713, V717, L719, L721, Y740, I742, I744, F754, T756, F761, I763, P764, M765, L768, Y773, V775, V777, K779, Y784, L788 and K790. (g) Supe posi ion o LKK as Cα-plo in sandy b own in he iew o (c) on o he s uc u es o he Po U (cyan; PDB 6ZA2) and Po Z (plum; PDB 5M11) CTDs. 6 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 The LKK moie y has an elonga ed ellipsoidal shape o ~42 Å heigh , ~26 Å wid h and ~22 Å dep h in he iew o igu e 3c. As he a o emen ioned P. gingi alis CTDs, i is an an ipa allel β-sandwich wi h a h ee-s anded shee (β1, β2 and β5, om le o igh in igu e 3c) and a ou -s anded shee (β7, β6, β3 and β4). Two ex a N- e minal esidues om he pu i ica ion ag (glycine and se ine) eplace domain esidues L704 and Y705 and o m he i s esidues o s and β1, so ha he domain ac ually spans 88 esidues (L704–K791). This lies wi hin he ange o clea able P. gingi alis CTDs (see §2.1). The wo LKK shee s a e a anged wi h an in e shee angle o ~35° and ha e G eek-key connec i i y ( igu e 3c,e). The s ands a e connec ed by sho loops spanning ei he wo esidues (Lβ3β4 and Lβ6β7), h ee esidues (Lβ4β5), i e esidues (Lβ5β6) o six esidues (Lβ1β2). The only excep ion is Lβ2β3. He e, a u he s and ex e nal o he cen al sandwich (β2′) in e ac s wi h he elonga ed ip o s and β2, which gi es ise o a 13- esidue connec o ha gene a es a hai pin-like appendage a he bo om o he sandwich ( igu e 3c,e). O e all, he moie y is cohe ed by an in e nal hyd ophobic co e made up by 23 esidues ( igu e 3 ). Finally, he C- e minal segmen o s and β7 encompasses he sequence K787–L–I–K–K791, which is ound a he C- e minus o i e unique se ine- and me al-dependen pep idases sec e ed by he T9SS o T. o sy hia in addi ion o mi olase, collec i ely named ‘KLIKK-pep idases’ [65,74]. This conse a ion is in line wi h he abo e sequence equi emen s o mo i C- . in P. gingi alis, which also includes a conse ed lysine in he middle o C- e minal s and β7 (see §2.2). Table 2. C ys allog aphic da a. da ase LKK (T. o sy hia mi olase CTD, SeMe ) beam line (synch o on) ID23-1 (ESRF) de ec o ADSC Quan um Q315 space g oup / p o ome s pe a.u.aP31 / 6 (A–F) cell cons an s (a and c, in Å) 81.45, 66.55 wa eleng h (Å) 0.91917 no. o measu emen s / unique e lec ions 731 662 / 65 076 esolu ion ange (Å) (ou e mos shell)b48.4–1.60 (1.70–1.60) comple eness (%) 99.8 (99.6) Rme gec0.121 (0.954) Rmeasc0.127 (0.999) CC(1/2)c0.999 (0.943) A e age in ensi yd9.0 (2.7) B- ac o (Wilson) (Å2) 23.5 a e . mul iplici y 11.2 (11.1) winning law h, -h-k, -l winning ac ion α0.423 no. o e lec ions used in e inemen [in es se ]e64 231 [656] c ys allog aphic R ac o / ee R ac o 0.275 / 0.302 Fobs , Fcalc co ela ion [ es se ] 0.933 [0.930] no. o p o ein esidues / a oms / sol en molecules / non-co alen ligands 518 / 4048 / 476 / 2 K+, 6 Cl−, 6 PEG Rmsd om a ge alues bonds (Å) / angles (°) 0.008 / 0.97 a e age B- ac o s (Å2): o e all // mol. A/B/C/D/E/F 21.7 // 17.9/22.9/22.2/17.8/22.8/22.0 all-a om con ac s and geome y analysis p o ein esidues in a ou ed Ramachand an egions / ou lie s / all esidues 503 (98.4%) / 0 / 511 wi h ou lying o ame s / bonds / angles / chi ali y / o sion 5 (1.1%) / 0 / 0 / 0 / 0 all-a om clashes / clashsco e 20 / 2.2 PDB access code 8QB1 aAbb e ia ions: a.u., c ys allog aphic asymme ic uni ; PEG, die hylene glycol. bValues in pa en heses e e o he ou e mos esolu ion shell. cFo de ini ions, see [72]. dA e age in ensi y is < I/σ(I) > o unique e lec ions a e me ging acco ding o Xscale [73]. eRe inemen was pe o med agains p ocessed da a ha we e subsequen ly de winned acco ding o win-law (h,-h-k,-l) and a winning ac o α = 0.423 wi h p og am De win wi hin Ccp4. Acco ding o he wwPDB Valida ion Se ice: h ps://wwpdb- alida ion.wwpdb.o g/ alidse ice. 7 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 S uc u e simila i y sea ches wi h LKK iden i ied he CTDs o Po Z (Z-sco e 10.0, msd 1.7 Å, 72 aligned esidues, 18% iden i y; see [75]) and Po U (Z-sco e 9.9, msd 1.9 Å, 78 aligned esidues, 19% iden i y) as he closes s uc u al ela i es. Con a y o he Tanne ella moie y, hese Po phy omonas CTDs co espond o non-clea ed o ms. No ably, he simila i y was signi ican ly lowe wi h he clea able CTDs o RgpB (Z-sco e 8.3, msd 2.2 Å, 72 aligned esidues, 7% iden i y) and HBP35 (Z-sco e 6.0, msd 2.2 Å, 54 aligned esidues, 17% iden i y). Supe posi ion o LKK on o he Po U and Po Z CTDs ( igu e 3g) e ealed ha he opology and connec i i y o he sandwich β-s ands a e main ained bu loops la gely de ia e and pai wise sequence iden i ies span me e 18–19%. In pa icula , he ex ended loop Lβ2β3 o LKK is much longe and adop s comple ely di e en ajec o ies in he Po p o eins. In con as , Lβ6β7 and Lβ5β6 a e sho e in LKK. The la e is uniquely longe in bo h Po U and Po Z bu is appa en ly no ele an o unc ion (see §2.4). 2.6. C- e minal domain in e change be ween P. gingi alis and T. o sy hia We ound ha , in addi ion o he much longe Lβ2β3, LKK lacked one esidue in mo i Lβ3β4, which is ele an o T9SS unc ion in P. gingi alis (see §2.3), and he conse ed glycine was eplaced wi h me hionine. Thus, we an icipa ed speci ic di e ences be ween he T9SSs o he wo species, as p e iously epo ed o he P. gingi alis and F. johnsoniae pai [11]. We cons uc ed a s ain wi h a eplacemen mu an o P. gingi alis PPAD, in which he CTD (A474–K556; numbe ing acco ding o UP Q9RQJ2) was eplaced wi h LKK (D702–K791) o yield mu an PPAD-LKK. We ound ha his chimae a was no p ope ly sec e ed and mimicked a s ain ha canno ancho ca goes o he OM ( igu e 4a). Mo eo e , he mu an s ain e inced subs an ially less ex acellula PPAD ac i i y ( igu e 4b). This con i ms ha he CTDs a e no unc ionally in e changeable be ween P. gingi alis and T. o sy hia due o he dis up ion o mo i Lβ3β4, despi e bo h species belonging o he Bac e oidia class. 2.7. Conclusion In an e o o de e mine molecula de e minan s o he ecogni ion o he speci ic T9SS sec e ion signal, a olded CTD, by he anslocon in P. gingi alis, as well as o i s clea age by he C- e minal signal pep idase, we compa ed six s uc u es. We ound he domain’s C- e minus and loop Lβ3β4 o be essen ial o unc ion. In con as , we could no de e mine he dis inguishing ea u es be ween Po -clea ed and Po U-non-clea ed CTDs. Gi en ha all he epo ed CTD s uc u es a e om P. gingi alis, we de e mined he s uc u e o a CTD om a T. o sy hia ca go, LKK. This module sha es he g oss a chi ec u e wi h i s P. gingi alis coun e pa s bu lacks he speci ic ea u es o Lβ3β4, and i e inces a subs an ially longe Lβ2β3. Consis en ly, he eplacemen o he CTD o a P. gingi alis ca go wi h LKK hampe ed i s sec e ion, which poin s o speci ic di e ences in CTD ecogni ion be ween he anslocons o hese wo species. This may p e en po en ial unc ional in e e ence due o he acciden al ans e o ca go genes be ween wo phylogene ically close pe iodon opa hogens, which sha e he human gingi al c e ice as an ecological niche, a e ound in he same bio ilm, and compe e o esou ces and h i e a he si e o in ec ion [76]. 3. Ma e ial and me hods 3.1. Bac e ial cul u ing All P. gingi alis mu an s we e de i ed om s ain W83 (see able 3 o a lis o s ains and cells used in his s udy) and we e cul u ed in yp icase soy b o h (30 g l−1) en iched wi h haemin (5 mg l−1), menadione (2 mg ml−1), ʟ-cys eine (0.5 g l−1) and yeas ex ac (5 g l−1). Solid cul u es we e addi ionally supplemen ed wi h 5% de ib ina ed sheep blood, 1.5% aga , and, when an ibio ic selec ion was needed, e y h omycin (5 μg ml−1 o e acycline (1 μg ml−1). Cells we e cul i a ed in an incuba o om Don Whi ley Scien i ic a 37°C unde anae obic condi ions (10% ca bon dioxide, 5% hyd ogen and 85% ni ogen). Fo molecula cloning p ocedu es, E. coli s ain DH5α was g own in Lu ia–Be ani medium (Lennox). Solid cul u es in he same medium we e p epa ed wi h 1.5% aga and supplemen ed wi h ampicillin (100 μg ml−1) o an ibio ic selec ion. 3.2. Subcellula ac iona ion o P. gingi alis cells P. gingi alis cul u es (con ols and mu an s) we e g own o a s a iona y phase, and hei cell densi y was adjus ed o OD600 ≈ 1.5 wi h yp icase soy b o h. Nex , WC cul u es we e p eincuba ed wi h 2 mM 2,2′-di hiodipy idine o 15 min and cen i uged (5000×g; 20 min) o sepa a e he cell- ee cul u e medium, which was designa ed ‘supe na an ac ion’ (S). The esul ing cell pelle s we e washed and esuspended in ice-cold phospha e-bu e ed saline (PBS) supplemen ed wi h a comple e EDTA- ee p o ease inhibi o mix (Roche) and 0.2 mM osyl-ʟ-lysyl-chlo ome hane hyd ochlo ide (Sigma) and subjec ed o cell dis up ion a 30 kPa p essu e wi h a BT40/TS2/AA cell dis up o om Cons an Sys ems. This was ollowed by 30 min diges ion wi h 0.02 mg ml−1 DNase I (Roche), which esul ed in he ‘cell ex ac ’ (C) ac ion. The emaining cell lysa e was ul acen i uged a 150 000×g o 1 h and he soluble ac ion (‘pe iplasmic/cy oplasmic’ ac ion; PP) was sepa a ed om he pelle , which con ained he insoluble memb ane ac ion. IM p o eins we e solubilized by incuba ion wi h 200 mM magnesium chlo ide and 10% T i on X-100 a 4°C o 30 min and sepa a ed by u he cen i uga ion a 150 000×g o 1 h. The emaining insoluble pelle con ained OM p o eins (‘OM ac ion’) and was esuspended by sonica ion in PBS supplemen ed wi h he abo e p o ease inhibi o s. 8 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 P o ein concen a ions we e de e mined wi h he BCA p o ein assay ki om The mo Fishe Scien i ic, and ac ions we e s o ed a −20°C. 3.3. Gene a ion o C- e minal domain mu an s Genomic modi ica ion o P. gingi alis was pe o med by homologous ecombina ion wi h suicide plasmids ca ying an an ibio ic esis ance casse e, as desc ibed be o e [41]. P ime s used o PCR eac ions a e lis ed in able 4. Fo RgpB and Po U W83 0 20 40 17 28 36 55 72 95 130 250 60 80 100 120 140 160 WC Rela i e ac i i y W83 M LKK WC LKK M WT LKK ∆WbpB WC S WC S WC S (a) (b) Figu e 4. C oss-species CTD exchange be ween P. gingi alis and T. o sy hia. (a) Wes e n blo ing analysis o PPAD expo in WC cul u es and he cul u e medium (S) o he wild- ype p o ein (labelled WT) and a PPAD chimae a wi h LKK as CTD (labelled LKK), as well as o a P. gingi alis mu an s ain (ΔWbpB), which does no a ach lipopolysaccha ide and hus does no ancho subs a es o he OM. The chimae a mimics he pheno ype o he la e s ain, hough wi h a subs an ially lowe yield, which indica es ha he p o ein is no p ope ly sec e ed. (b) PPAD ac i i y in whole-cell ex ac and medium o he wild- ype p o ein and he chimae a. Table 3. Lis o s ains and plasmids. s ains desc ip ion (geno ype; esis ance) sou ce/ e e ence E. coli DH5α gene al cloning hos The mo Fishe E. coli BL21 (DE3) exp ession o Po U p o ein Millipo e P. gingi alis W83 WT wild ype ∆Po U po U::e mF; Em [5] 1152sHis po U1153-1158::6His, e mF; Em his s udy Po UHis (1158iHis) po U1158ins8His, e mF; Em [41] RgpA-C RgpA::ca ; Cm [16] 695-701:GG RgpA::ca , RgpB 695-701::GG, e R, Cm ,Te his s udy G699R RgpA::ca , RgpB G699R, e R; Cm , Te his s udy A712-714:NARGDGNSVA RgpA::ca , RgpB 712-714::NARGDGNSVA, e R; Cm ,Te his s udy A719R RgpA::ca , RgpB A719R, e R; Cm , Te his s udy PPAD-LKK PPAD 474-556::Mi olase 702:791, e R; Te his s udy ∆WbpB wbpB:: e Q; Te his s udy plasmids p22-E-1158iHis suicide plasmid o Po UHis (1158iHis) s ain gene a ion [41] pPo U 1152sHis suicide plasmid o 1152sHis s ain gene a ion his s udy pRgpB- suicide plasmid o RgpB-coding gene modi ica ion [16] pRgpB 695-701:GG suicide plasmid o 695-701:GG s ain gene a ion his s udy pRgpB G699R suicide plasmid o G699R s ain gene a ion his s udy pRgpB 712-714:NARGDGNSVA suicide plasmid o 712-714:NARGDGNSVA s ain gene a ion his s udy pRgpB A719R suicide plasmid o A719R s ain gene a ion his s udy pPPAD- e suicide plasmid o PPAD-coding gene modi ica ion his s udy pPPAD-LKK suicide plasmid o PPAD-LKK s ain gene a ion his s udy LKK-pGEX6-P-1 exp ession plasmid o LKK domain his s udy pWbpB- suicide plasmid o wbpB gene dele ion his s udy 9 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025 91. D ozde skiy A, Cole C, P oc e J, Ba on GJ. 2015 JP ed4: a p o ein seconda y s uc u e p edic ion se e . Nucl. Acids Res. 43, W389–W394. (doi:10.1093/na /gk 332) 92. Pei J, Kim BH, Tang M, G ishin NV. 2007 PROMALS web se e o accu a e mul iple p o ein sequence alignmen s. Nucl. Acids Res. 35, W649–W652. (doi:10.1093/na /gkm227) 93. K issinel E, Hen ick K. 2004 Seconda y-s uc u e ma ching (SSM), a new ool o as p o ein s uc u e alignmen in h ee dimensions. Ac a C ys allog . D 60, 2256–2268. (doi:10. 1107/S0907444904026460) 94. K issinel E, Hen ick K. 2003 P o ein s uc u e compa ison in 3D based on seconda y s uc u e ma ching (Pdbe old) ollowed by Cα alignmen , sco ed by a new s uc u al simila i y unc ion. In 5 h In . Con . on Molecula S uc u al Biology, Vienna, Aus ia, 3–7 Sep embe 2003. 95. Jumpe J e al. 2021 Highly accu a e p o ein s uc u e p edic ion wi h AlphaFold. Na u e 596, 583–589. (doi:10.1038/s41586-021-03819-2) 96. B yan P, Pozza i G, Elo sson A. 2022 Imp o ed p edic ion o p o ein-p o ein in e ac ions using AlphaFold2. Na . Commun. 13, 1265. (doi:10.1038/s41467-022-28865-w) 97. Ma iani V, Biasini M, Ba ba o A, Schwede T. 2013 lDDT: a local supe posi ion- ee sco e o compa ing p o ein s uc u es and models using dis ance di e ence es s. Bioin o ma ics 29, 2722–2728. (doi:10.1093/bioin o ma ics/b 473) 98. Tunyasu unakool K e al. 2021 Highly accu a e p o ein s uc u e p edic ion o he human p o eome. Na u e 596, 590–596. (doi:10.1038/s41586-021-03828-1) 99. Mizgalska D, Rod íguez-Banque i A, Veilla d F, Książęk M, Goulas T, Gue a a T e al. 2024 Supplemen a y ma e ial om: S uc u al and unc ional insigh s in o he C- e minal signal domain o he Bac e oide es ype-IX sec e ion sys em. Figsha e (doi:10.6084/m9. igsha e.c.7272136) 16 oyalsocie ypublishing.o g/jou nal/ sob Open Biol. 14: 230448 Downloaded om h ps:// oyalsocie ypublishing.o g/ on 24 Ma ch 2025