Development of microbial community simulation methods to characterize and analyze the effects of metal concentrations on pathogen silencing/promotion

Uni e si y o Minho School o Enginee ing Alexand e A eias Cas o De elopmen o mic obial communi y simula ion me hods o cha ac e ize and analyze he e ec s o me al concen a ions on pa hogen silencing/p omo ion janua y 2024 Uni e si y o Minho School o Enginee ing Alexand e A eias Cas o De elopmen o mic obial communi y simula ion me hods o cha ac e ize and analyze he e ec s o me al concen a ions on pa hogen silencing/p omo ion Mas e s Disse a ion Mas e s in Bioin o ma ics Disse a ion supe ised by Vi o Manuel Sá Pe ei a Sa ela Ga cia San ama ina janua y 2024 Copy igh and Te ms o Use o Thi d Pa y Wo k This disse a ion epo s on academic wo k ha can be used by hi d pa ies as long as he in e na ionally accep ed s anda ds and good p ac ices a e espec ed conce ning copy igh and ela ed igh s. This wo k can he ea e be used unde he e ms es ablished in he license below. Reade s needing au ho iza ion condi ions no p o ided o in he indica ed licensing should con ac he au ho h ough he Reposi ó iUM o he Uni e si y o Minho. License g an ed o use s o his wo k: CC BY-SA h ps://c ea i ecommons.o g/licenses/by-sa/4.0/ i Acknowledgemen s To my supe iso s, Sa ela and Ví o , o p o iding all he condi ions and suppo o make his in o a inished esea ch p ojec . Th ough all he challenges, di icul ies and ime cons ain s, you belie in he conclusion o his p ojec , e en when i seemed impossible, only made i mo e enjoyable. I eel ex emely hono ed o ha e comple ed his ask wi h he suppo and counseling o such g ea esea che s and p o essionals. To my amily, o hei sac i ices and suppo allowed me o pu sue his educa ion and ollow my d eams. Thanks o you and you suppo , I was able o no only inish bu o ake e e y s ep ha has led me he e. May li e allow me o gi e back all you ha e gi en me. To my iends, o hei company and eassu ance in e e y one o he imes we we e oge he helped me ga he he s eng h o mo e on. E e y laugh, joke o e e y ime you ha e lis ened o me an abou how di icul li e was was uly a he apy, and makes me ealize ha iends a e uly one o li e’s bigges gi s. Las ly, o my gi l iend, Ma ilde, o you lo e and uncondi ionall suppo . No one belie ed in me mo e han you, and mos imes, ha belie o e came my own. Wo ds canno exp ess how p oud I am o ha e had on my side e e y s ep o he way, and hope o con inue o ha e so. Li e is uly special by you side. ii S a emen o In eg i y I he eby decla e ha ing conduc ed his academic wo k wi h in eg i y. I con i m ha I ha e no used plagia ism o any o m o undue use o in o ma ion o alsi ica ion o esul s along he p ocess leading o i s elabo a ion. I u he decla e ha I ha e ully acknowledged he Code o E hical Conduc o he Uni e si y o Minho. Uni e si y o Minho, B aga, janua y 2024 Alexand e A eias Cas o iii Abs ac Genome-scale me abolic modeling has become a widesp ead me hodology o analyzing mic obial me abolism. The ma hema ical modeling o hese ne wo ks has p o en i sel as an indispensable ool o unde s and- ing mic obes by p edic ing hei beha io and sys ema ically allowing he es ing o di e en hypo heses. On mic obial communi ies, i u he allows he p edic ion o in e ac ions a he le el o me aboli es and me abolic eac ions, he e o e p o iding aluable insigh s on he inne wo king o he as ne wo ks hey cons i u e. Cu en ex ensions o his me hodology ha e conside ed he addi ion o enzyma ic da a o single models in o de o u he cons ain eac ion luxes and p oduce a mo e accu a e esul s. The goal o he esea ch p esen ed in his disse a ion is o combine di e en me hodologies cu en ly in exis ence in o de o c ea e a me hodology ha allows he e ec i e and accu a e simula ion o he beha iou o gu mic obial species inside a communal space. These simula ions will ul ima ely aim o p edic g ow h unde low le els o ansi ion me als in si ua ions ha esemble hose o in ec ion, whe e he hos limi s he a ailabili y o i on and o he elemen s essen ial o all li e o ms in o de o igh pa hogenesis. Th ough he combina ion o he enhancemen o models wi h enzyma ic da a wi h he c ea ion o compa men alized models unde di e en simula ion me hods, he de ised wo k aimed o ex end bo h exis en me hodologies and p oduce esul s ha we e imp o ed h oughou he en i e y o he wo k by basing and aiming o eplica e p o ided esul s ob ained om in i o condi ions. The ob ained esul s e eal no only ha he enhancemen o models wi h enzyme cons ain s p o ide communi y s abili y bu also he ex ension o he capabili ies o communi y simula ion me hods p e iously es ablished. The esul s o his wo k cons i u e a no el app oach o he simula ion o gu mic obial communi ies ha e ealed he impac he enhancemen p ocess has on bo h he way models beha e in singula se ings and also in p o iding communi y s abili y. Mo eo e , he de ised app oach and subsequen esul s p o ide he opo uni y o ex ension and imp o emen wi h e e mo e da a in he applica ion o simila o o he s udy cases. Keywo ds Gu mic obial communi ies, Me abolism, Communi y modeling, Genome-scale me abolic models, T ansi ion me als i Resumo A simulação me abólica de modelos em escala genômica o nou-se uma me odologia p oeminen e na análise do me abolismo mic obiano. A modelação ma emá ica des as edes em-se p o ado con inua- men e como uma e amen a indispensá el na comp eensão do compo amen o de mic oo ganismos ao p e e o seu compo amen o, pe mi indo assim a sis ema ização da es agem de di e en es hipó eses. Em comunidades mic obianas, es a pe mi e ainda a p e isão das in e ações ao ní el dos me aboli os e eações, con e indo assim uma isão mais ala gada do uncionamen o in e no des as as as edes po eles cons i uídas. As ex ensões pa a es a me odologia a ualmen e u ilizadas como a adição de es ições enzimá icas êm como obje i o o aumen o das es ições aplicadas a cada modelo, de modo a que es es p oduzam esul ados cada ez mais p ecisos. O obje i o des e p oje o de in es igação ap esen ado nes a disse ação passa pela combinação de di e en es me odologias a ualmen e exis en es de modo a p ocede à c iação de uma no a me odologia que pe mi a a simulação do compo amen o de espécies mic obianas do a o in es inal inse idas num espaço comuni á io. Es as simulações e ão como obje i o inal a capacidade de p e e o seu c escimen o sob e baixos ní eis de me ais de ansição em si uações que se assemelham ás de in eção, onde o hospedei o limi a a disponibilidade de e o e ou os elemen os essenciais à ida de modo a comba e a pa ogénese. A a és da combinação do melho amen o de modelos pela adição de es ições enzimá icas com a c i- ação de modelos compa imen alizados sob e di e en es mé odos de simulação, o p esen abalho isou a ex ensão de ambas as me odologias e p oduzi esul ados que o am sis ema icamen e ap imo ados po se basea am e emula em esul ados ob idos em condições in i o . Os esul ados ob idos e elam que não só o melho amen o dos modelos com es ições enzimá icas p omo em a es abilidade comuni á ia, mas ambém que ex endem as capacidades de mé odos p e iamen e es abelecidos. Os esul ados des e abalho consi u em uma no a abo dagem pa a a simulação de comunidades mic obianas in es inais que e elam o impac o que o p ocesso de melho amen o e e na manei a como os modelos se compo am, assim como na p omoção da es abilidade das comunidades. Adicionalmen e, a abo dagem desen ol idade e subsequen emen e, os seus esul ados, são passí eis de ex ensão e mel- ho amen o pa a ainda mais dados de modo a se em aplicados a es udos simila es ou ex endidos a ou os casos de es udo. Pala as-cha e Comunidades mic obianas in es inais, Me abolismo, Modelação de comunidades, Modelos me abólicos em escala genómica, Me ais de ansição i Glossa y GECKO Me hod o enhancemen o GEMs wi h Enzyma ic Cons ain s using Kine ic and Omics da a. xiii Pa I In oduc o y ma e ial 1 Chap e 1 In oduc ion 1.1 Mo i a ion Mic obial communi ies a e he mos na u al o m o occu ence o mic obes. The in e ac ions ha occu among hem as well as hose ha occu be ween hem and hei hos can ha e a ious epe cussions ha can la gely con ibu e o he o e all heal h o humans and many o he o ms o li e. The powe o unde s and hei e ec s and mo e impo an ly he unde lying cons i u ion, unc ion and in e ac ions ha happen amids hese communi ies, as well as how hey a e a ec ed by ac o s such as hei di e si y and o he ou e ac o s can unlock a new way o unde s anding mic obial ecology. The s udy o mic obial communi ies and hei unc ioning ine i ably comp ehends he unde s anding o he eme gen p ope ies ha su ge wi hin he communi ies. When he shee numbe o in e ac ions is pu in o pe spec i e, one can unde s and how such mic oscopic in e ac ions can p oduce mac oscopic p ope ies. These p ope ies can be de ined as any pa e n o unc ion ha canno be deduced as he sum o he p ope ies o he cons i uen pa s [ an den Be g e al.,2022] and a e ypically a byp oduc o he size o he communi y, as well as he connec ions es ablished be ween he membe s o hese complex mic obial ne wo ks. Finding he h eshold ha igge s hese p ope ies can p o e o be an ambiguous and mos ly si ua ional ask due o he speci ici y and uniqueness ha cha ac e ize e e y single communi y as well as hei in insic p ope ies. Due o ha , mos cha ac e iza ions o he h esholds and he eme gen p ope ies om which hey a e inhe en a e mos ly done in quan i a i e o ms ha o en lack p ecision and eliabili y. 2 T ansi ion me als a e essen ial nu ien s o all o ms o li e. Thei impo ance and e ec s a e di ec ly in- luenced by hei p esence and concen a ion. Howe e , i is c ucial o unde s and ha bo h he de iciency and excess o hese me als a e he basis o he de elopmen o nume ous pa hologies and ep esen a majo public heal h bu den [Van Gossum and Ne e,1998]. The in e ac ion wi h hese elemen s al e s he physio-chemical p ope ies o p o eins, he eby p omo ing ca alysis o enzyma ic eac ions, s abilizing p o ein s uc u e, and/o acili a ing elec on anspo [Mu doch and Skaa ,2022], and can a ec many c ucial cellula p ocesses such as espi a ion, ansc ip ion, signal ansduc ion, and p oli e a ion [And eini e al.,2006]. O ganisms main ain app op ia e le els o me als by con olling up ake, use, s o age, and exc e ion a he cellula and sys emic le els. Tissue me al le els in e eb a e hos s a e mainly con olled by abso p ion om die a y sou ces in he in es inal ac , whe eas bac e ial pa hogens ob ain me al om he ex acellula and in acellula en i onmen s o hos issues du ing in ec ion [Lopez and Skaa ,2018]. Unlocking new mechanisms ha enable a be e unde s anding o how eme gen p ope ies a e ig- ge ed and hei unc ioning and he e ec o di e en me al concen a ions on he in es inal ac mic obial communi ies can esul in new o ms o comp ehending he appea ance o a g ea numbe o pa hologies while p o iding new insigh s on new app oaches o ha ma e . 1.2 Goals Th ough he c ea ion o new mechanisms o me abolic modelling, he esea ch he e p esen ed aims o cap u e he e ec s o luc ua ing le els o ansi ion me als in gu mic obial communi ies and he se e al mechanisms hese igge , including bo h pa hogen colonisa ion esis ance and commensalis in e ac ions. Examples o hese a e seen in he u iliza ion o side opho es p oduced by pa hogens unde Fe limi a ion ca ied ou by commensal species such as Bac e oides he aio aomic on as epo ed in Zhu e al. [2020]. By combining he use o Genome-Scale Me abolic Models (GEM), he enzyma ic enhancemen o hese models and he use o a ious simula ion app oaches o communi y models like Flux Balance Analysis and S eadyCom [Chan e al.,2017], he esul s o he his wo k will cons i u e a no el app oach o no only mic obial communi y modelling in gene al bu also o a mo e accu a e comp ehension o he mechanisms ha a e in insic o he in e ac ions ha occu wi hin he gu mic obial space and a he hos –pa hogen in e ace. 3 1.3 S uc u e The p esen documen is s uc u ed in di e en chap e s, each ocusing on a speci ic subjec ela ed o he ask a hand. •Chap e 2:Me abolic Modelling - This chap e se es as in oduc o y ma e ial o me abolic model- ing by explo ing he mos impo an aspec s o his app oach. The chap e con inues by p esen ing in oduc o y concep s o communi y models as well as some o he mos widely used ypes o mod- els. A e wa d, he opic o enzyma ic cons ain s enhancemen o models in which his app oach is in oduced, as well as some o i s co e ad an ages. •Chap e 3:Mic obial Communi ies - He e, a b ie in oduc ion o mic obial communi ies is made while add essing some o i s co e cha ac e is ics and concep s like hei s uc u e and unc ion. A e wa d, u he no ions a e p esen ed on he se e al in e ac ions ha occu bo h on he in e nal and ex e nal in e aces o hese biological sys ems. •Chap e 4:T ansi ion Me als - This chap e add esses he gene al in oduc o y concep s o an- si ion me als, ocusing on hei s uc u e and unc ion, while also app oaching he a ious con ol mechanisms exe ed by hos and mic obial o ganisms, as well as he in e ac ions p omo ed by he p esence o hese elemen s. •Chap e 5:Me hods - In his chap e he es ablished wo k low is desc ibed, along wi h he pe - o med asks, he wo k’s imeline, main me hods, and ools used. •Chap e 6:Resul s and Discussion - This chap e p esen s he ob ained esul s, hei discussion and limi a ions. •Chap e 7:Conclusions - He e he main conclusions, con ibu ions and p oposi ions o u u e wo k, as well as he main di icul ies encoun e ed a e p esen ed and desc ibed. 4 Chap e 2 Me abolic Modelling 2.1 Me abolic Modelling o Communi ies Unde s anding he ela ionship be ween geno ype and pheno ype is a c ucial pa o unde s anding how biological sys ems wo k. To ha end, he comp ehensi e knowledge o he complex me abolic pa hways, gene ic in o ma ion, and how o ma hema ically model hem has su ged as a majo equi emen o biol- ogis s. The use o hese ma hema ical models has become an es ablished ool o sys ema ic analyses o me abolism o a wide a ie y o o ganisms, ha ing p o ed hemsel es as being c ucial no only o he unde s anding o mechanisms unde lying complex human diseases bu also o o he applica ions such as model-d i en de elopmen o e icien cell ac o ies [Domenzain e al.,2022]. The in e ac i e biological g oups ha exis in na u e hold he sec e o unde s anding how indi iduals o g oups o indi iduals a ec popula ion dynamics, esou ce a ailabili y, and e en di e si y. Communi ies’ bio ic composi ion and unc ions a e ex emely a iable and being able o associa e hese ac o s wi h he ole o singula o ms o li e c ea es he possibili y o no only unde s and hei unc ioning bu also c ea e ways o implemen dynamics ha allow communi ies o a ain ce ain s uc u es and unc ions. Fo biologis s, i has become a majo goal o c ea e models and simula ion me hods ha allow he ob aining o esul s ha can encompass in e ac ions and co- ac o s close o hose ha occu in na u al en i onmen s. The e is cu en ly a wide ange o ypes o models ha o he c ea ion o simula ion me hods unde many app oaches and desi ed capabili ies. Globally, he o mula ion me hods o models can be di ided in o wo main g oups: s eady-s ake and kine ic models. 5 The c ea ion o communi y models has p esen ed i sel as an e ec i e mechanism o he esolu ion o ha p oblem. These models can be de ined as agglome a es o singula models ha ake in o accoun e- ac ions ha occu wi hin each indi idual whils also accoun ing o in e ac ions ha ake place amids and be ween o he cons i uen pa s o hese communi ies. This ma hema ical app oach, allied wi h he e e - g owing compu a ional powe , has enabled he c ea ion o simula ion me hods ha can emula e g ow h a es and o he impo an ac o s such as enzyma ic usage, acco ding o desi ed pa ame e s, while also allowing he c ea ion o desi ed mu an s ha can be e display desi ed ai s o each desi ed e ec . In he nex sub-chap e , a lis o he mos common ypes o communi y models is p esen ed, alongside hei unc ioning, ma hema ical coun e pa s, as well as hei iabili y in e ms o hei applica ion o he c ea ion o mic obial communi y models. 2.1.1 Types o Models Lo ka–Vol e a models - Kine ic The Lo ka-Vol e a (LV) model was in oduced in he 20 h cen u y and is used o desc ibe p eda o -p ey sys ems and hei dynamics. I is one o he i s ma hema ical app oaches o desc ibe complex ecological dynamics [Nedo ezo ,2016], consis ing o a se o non-linea , coupled, i s -o de di e en ial equa ions. The classic model does no cap u e in e ac ions in ol ing mo e han wo species, so i s ele ance o he s udy o mic obial communi ies lies in he Gene alized Lo ka-Vol e a model, which can accommoda e any numbe o species. The gene ic LV sys em o npopula ions o species xi akes he o m: dBi d =Bi(αi+ n ∑ j=1 βijxj)(2.1) whe e he non-nega i e pa ame e αiis he g ow h a e o species iand each eal- alued in e ac ion pa ame e βij quan i ies he ype and s eng h o he e ec o species jon species i, i i=j. I i=j, βij e lec s in aspecies in e ac ions. In he case o he p esence o only a single species (n= 1), he LV model simpli ies o he logis ic equa ion: dBi d =αi+βx2(2.2) 6 The wo- a iable case (n= 2) o he Lo ka-Vol e a model includes h ee e ms o each species: one g ow h e m (αi), one in aspecies e m (βii) and one in e species in e ac ion e m (βij): dx1 d =x1(α1+β11x1+β12x2) dx2 d =x2(α1+β21x1+β22x2) The sign o each in e species in e ac ion e m ep esen s he ype o ela ionship be ween he wo species, such as desc ibed in Table 1: Table 1: Types o in e species in e ac ions and in e ac ion coe icien s signals. In e ac ion Type βij βji Neu alism 0 0 Mu ualism + + Commensalism + 0 P eda ion + - Amensalism 0 - Compe i ion - - When con empla ing he applica ion o hese models o mic obial communi ies, howe e , some di icul- ies associa ed wi h he assump ions c ea ed when using he model a ise. These assump ions p esuppose ha i s ly, he in e ac ions be ween popula ions occu while being s a ic in ime and space, meaning ha e en s such as en i onmen al changes, physiological adap a ion, egula o y shi s o e olu iona y changes a e no aken in o accoun [Voi e al.]. Secondly, posi i e eedback loops in he sys em can lead o un e- s ained popula ion g ow h. Consequen ly, i ing da a o a Lo ka-Vol e a model may unde es ima e he p e alence o mu ualism (Hoek e al. [2017], an den Be g e al. [2022]). Also, a hi d p oblem su ges as he model pos ula es ha he e a e no highe -o de in e ac ions meaning ha he al e a ion o in e ac ions be ween a pai as a consequence o he p esence o a hi d species does no occu . Las ly, small in e - ac ions such as me aboli e exchange o quo um sensing a e no included in hese models, which highly comp omises he s udy o he eme gen p ope ies o a mic obial communi y, and he e o e also p oduces esul s and conclusions a he om p ac ical scena ios. 7 MacA hu consume – esou ce models - Kine ic This phenomenological model p esen s i sel as an al e na i e app oach o ha p oposed by he Lo ka- Vol e a g oup o models. The esou ce-explici model encompasses he g ow h a e’s dependence on chemical ac o s such as nu ien s, signals o oxins, while also coupling hei concen a ions o popula ion dynamics. The model allows he quan i ica ion o he way di e en species use esou ces: he bigge he esou ce o e lap, he smalle he chance o coexis ence. The g ow h a e o each species is di ec ly p opo ional o i s esou ce consump ion, accoun ing o a non-pa ame e iza ion o species in e ac ions bu a he hei media ion being done by he de iciency in sha ed esou ces among he en i e communi y o popula ion. In MacA hu ’s models, each consume species is egula ed by i s esou ce ha es , which is linea in he esou ce densi ies, ela i e o i s main enance equi emen : dBi d = iBi(∑ α ∆wiαCiαRα)−miBi(2.3) ∆wiα =wα−∑ β Di βαwβ(2.4) whe e Biis he densi y o consume species i,1≤i≤n; pe uni ime, a uni o consume species iha es s (u ilizes) CiαRαuni s o esou ce; pe uni ime, a uni o consume species imus ha es ene gy i o main ain i sel . Each esou ce species is subjec o a g ow h unc ion (Equa ion 2.3), which may in ol e no only sel - egula ion bu also in e ac ions wi h o he esou ce species, and mo ali y by consump ion (mi). I is assumed ha he esou ces don’ in e ac wi h one ano he , bu hey can be u ilized and consumed by in ica e dynamics. The MacA hu consume - esou ce equa ions also allow o esou ces and consume s o ha e di e en imescales. I he esou ce dynamics a e as , he sys em simpli ies o dynamics simila o hose p esen ed by Lo ka-Vol e a models. I s applicabili y o mic obial communi ies, howe e , depends on i s ex ension o a beyond mo e han wo species and wo esou ces. I also in ol es a se ies o adap a ions o he model, like he abili y o include c oss- eeding, he assump ion o ixed me abolism o communi y membe s, he absence o axonomic and hie a chical me abolic s uc u es, and he missing he modynamic and biochemical de ail, being he la e a common limi a ion amongs o he ecological models. This is a classic ecological model ha is de e minis ic in na u e, in con as wi h he s ochas ic bio ic and abio ic in luences on communi y dynamics [Zelezniak e al.,2015]. 8 T ai -based models - Kine ic T ai -based (TB) models, unlike o he ypes o models p e iously men ioned, de ine he ocal a iables as he pheno ypic ai s o indi iduals a he han phylogene ic g oups. Membe s a e de ined by hei ai s and he models desc ibe how ai combina ions espond o en i onmen al a iables, as well as hei in luence o e hem. Also, unlike classic ecological models ha gene ally assume a cons an en i onmen , ai -based models ocus on communi y dynamics along majo en i onmen al g adien s, which leads o a unc ional ai dis ibu ion ha op imizes ce ain communi y-le el unc ional p ope ies. This allows he TB model’s capabili y o be e simula e he eme gence o selec ion p essu es and di e si y pa e ns along modeled en i onmen al g adien s, which a e impo an aspec s in unlocking he unde s anding o e olu iona y p ocesses [Li chman e al.,2013]. The g ow h pe popula ion (Equa ion 2.5) and impac en i onmen al ac o (Equa ion 2.6) a e calcula ed h ough he ollowing equa ions dBi d = ( i−mi)Bi(2.5) i= max iα Rα Rα+Kiα eSi(T−T e )(2.6) whe e Biis he popula ion size o i, iis he impac en i onmen al ac o o iand mi ep esen s he dea h o indi iduals belonging om popula ion i.T ai -based models a e, he e o e ex emely well sui ed o be applied o la ge communi ies, being howe e dependen on hei low local a ia ions in en i onmen al a iables. This has p o en i sel as a p oblem o i s applica ion o gu mic obio a sys ems, no only because hese communi ies p esen high local a iabili y, bu also because e en hough TB models do no equi e he p io imposi ion o ai combina ions, hey do equi e he de ini ion o which ai s a e included in he sys em. 9 3.3 In e nal In e ac ions Mic obe-mic obe in e ac ions accoun o an impo an pa o he in e ac ions ha happen wi hin commu- ni ies. The mechanisms ha a e unlocked h ough he in e ac ing be ween se e al popula ions can shape communi y dynamics and he e o e gi e ise o eme gen p ope ies ha can con e esilience, esis ance o coloniza ion and e en p omo e pa hogenesis. P a ical examples shown in s udies ha e p o en ha mic obio a can p omo e esis ance o coloniza ion by pa hogenic species, including s udies in which mice ha we e ea ed wi h an ibio ics o b ead in s e ile en i onmen s ha e p o en o be mo e p one o he in ec ion by en e ic pa hogenic bac e ia such as Shigella lexne i , Ci obac e oden ium , Lis e ia monocy- ogenes and Salmonella en e ica se o a Typhimu ium (Sp inz e al. [1961], Kamada e al. [2012], Zacha and Sa age [1979]). O he s udies ha e also shown ha some mic obio as can also lead o he expansion o enhance he i ulence o o he pa hogenic popula ions [Came on and Spe andio,2015]. C ea ing he link be ween he composi ion o he mic obial communi ies, hei in e ac ions and unc ion has he e o e been pinpoin ed as a majo ocal poin in he comp ehension o mic obial communi ies. O he cases o heal hy indi iduals being ansplan ed wi h mic obio a om in ec ed indi iduals has also esul ed in he i ulence o hose p e iously heal hy indi iduals [Ghosh e al.,2011], he e o e suppo ing he idea ha he in e nal in e ac ions play a majo ole no only in composi ion bu also in he su ging o eme gen p ope ies. 3.3.1 Auxo ophs Mic obial communi ies a e composed o cells wi h a wide ange o me abolic capaci ies, and egula ly include auxo ophs. This g oup can be de ined as mic oo ganisms ha p esen auxo ophy, a s a e in which an o ganism is unable o syn e hise compounds ha a e essencial o hei g ow h. The sea ch o he ole o auxo ophs in mic obial communi ies has lead o he disco e y o no only he p e alence o auxo ophs in mic obial communi ies, bu also ha hei me abolic coope a ion unlocks a se ies o mechanisms ha con ibu e o he eme gence o p ope ies ha con ey esis ence o d ugs no only o hem, bu also o he en i e y o he communi y, by p omo ing a iche me abolic en i onmen wi h high me aboli e e lux [Yu e al.,2022]. 16 3.4 Mic obio a-pa hogen-hos in e ace The whole ange o in e ac ions ha occu a he hos -pa hogen in e ace media e a a ms ace ha is cha ac e ized by a ange o mechanisms. In one on , he need o ob ain compounds ha a e essen ial o he g ow h o mic obial cells p omo es a se ies o in e ac ions ha seek no only o ob ain me aboli es om he ex acellula space bu also p omo es he he o any o hose elemen s as he hos s ind ways o seques e many compounds in o de o s op g ow h ha migh enable pa hogenesis. Ano he o m o in e ac ions encompass he cons an need o communi ies o e ol e ways o de elop esis ence o d ugs and o he o ms o an ibio ics ha migh be in oduced by he hos ’s immune sys em. This p omo es a se ies o popula ional dynamics ha shapes adap abili y o he communi ies ei he h ough selec i e p essu e ha esul s in mu an s ha p omo e he su ging o eme gen p ope ies o h ough he change o me aboli e up ake and sec e ion, in which a ange o mic obe-mic obe in e ac ions media e a change in he communi y’s unc ion. Ha ing ha said, o he o ms o mu ualis ic ela ions be ween hos s and mic obes gi e shape o i ulence esis ance, whe e non-pa hogenic mic obio a p omo e esis ance o coloniza ion by pa hogenic species, as p e iously men ioned. Fu he mo e, he unde s anding o hese dynamics c ea es a connec ion be ween many o he aspec s p e iously men ioned. The p esence o pa hogens in mic obial communi ies cons i u es he basis o many pa hologies ha a ec hos s in many ways. While hei p esence in communi ies can be bene i ial o no o hem, he p essu e exe ed by he hos s, and in some cases, by o he mic obio a, leads o a numbe o in e ac ions, he e o e ep esen ing ano he complex o m o in e ac ion ha con ibu e o he e e g ow h o he ne wo k in which mic obial communi ies and hei in e ac ions can be ep esen ed. In u he chap e s, a e iew o some o hose mechanisms, specially hose in ol ing he use o ace me als will be p esen ed. 3.5 Chap e Summa y In his chap e , a b ie in oduc ion o mic obial communi ies was made, add essing some o i s co e concep s and cha ac e is ics, as well as he p esen a ion o some o he in e nal and ex e nal in e ac ions ha occu be ween mic obio a and hos s alike. 17 Chap e 4 T ansi ion Me als 4.1 In oduc ion Figu e 4: T ansi ion me als in he Pe iodic Table o elemen s. T ansi ion me als a e comp ised o he chemical elemen s loca ed in he cen al block (d-block) o he Pe iodic Table (Figu e 4). All he elemen s a e solids wi h he excep ion o me cu y, Hg, which is a liquid a no mal p essu es and empe a u es. Na u ally occu ing me als in biological sys ems include elemen s like i on (Fe), zinc (Zn), cobal (Co) and coppe (Cu). All hese elemen s, in con a y o some o hose p esen in he -block o he pe iodic able (inne ansi ion me als) a e syn hesized na u ally and a e c ucial o se e al p ocesses in biological sys ems. The abso p ion o hese elemen s occu s mainly om die a y sou ces when conside ing animals, while mic obial o ms o li e use many o he o ms o up aking ace me als. Fu he discussion on he e ec s, homeos asis con ol and in e ac ions ha occu be ween hos s and pa hogens will be made. 18 4.2 E ec s The p esence o ace me als in o ganic sys ems ansla es in o he abili y o in e ac and cause a panoply o causes and e ec s. The mos common lies in he in e ac ion wi h a ce ain and widely common g oup o p o eins: me allop o eins. This ype o p o ein is cha ac e ized by i s me al ion co- ac o and me al-binding p o ein domains and unc ions in a di e se se o cellula p ocesses, including espi a ion, ansc ip ion, signal ansduc ion and p oli e a ion [And eini e al.,2006]. The in e ac ions ha happen be ween me al- lop o eins and ansi ion me als p oduce a se ies o e ec s on p ocesses like he p omo ion o ca alysis o enzyma ic eac ions, s abiliza ion o p o ein s uc u e and/o e en he media ion o elec on anspo . As i has been p o en, he panoply o e ec s ha he p esence o ace me al p oduces a e mos ly essen ial o li e o ms o many kinds. Excessi e le els o hese elemen s, howe e , ha e also been p o en o be oxic o mos li e o ms. One po en ial mechanism o oxici y esides in he p oduc s o Fen on chemis y by he edox-ac i e ansi ion me als i on, coppe , and manganese. Fe3++O2 •− ←−→ Fe2++O2(4.1) Fe2++H2O2−−→ Fe3++HO +OH−(4.2) O2 •− +H2O2←−→ O2+OH•+OH−(4.3) In he Habe -Weiss eac ion o Fen on chemis y, Fe2+ eac s wi h hyd ogen pe oxide o o m Fe3+(Equa- ion 4.2), a hyd oxyl adical, and hyd oxide; Fe3+can hen eac wi h hyd ogen pe oxide o egene a e Fe2+in addi ion o a hyd ope oxyl adical and a p o on (Equa ion 4.1). These oxida i e species inhibi cell g ow h by damaging p o eins, DNA, and lipids. Ano he po en ial mechanism o oxici y is media ed by a p ocess known as p o ein misme alla ion, media ed by he p e alence o me allop o eins in he bac e ial cell and hei ela i ely compa ible binding ligands ha equi e igh con ol o me al homeos asis o con inued me abolism. In he uncon olled le els o me als esides he o igin o a sho age o p o ein binding si es and a subsequen excess o supply unde a sho age o demand. 19 4.3 Con ol Mechanisms In spi e o all he necessi y o ace me als exe ed by cells o p omo e ac i i y, he excess o hese elemen s p o okes oxici y. This phenomenon can be explained ei he by he excessi e in e ac ion wi h me allop o- eins and consequen misme alla ion, which nulli ies hei enzyma ic capabili ies o by he o ma ion o edox-ac i e molecules. To ha e ec and p e en oxici y, i is c ucial o o ganisms o ha e mechanisms ha allow hem o main ain app op ia e le els o hese componen s. 4.3.1 Homeos asis Con ol The main enance o p ope ace me als homeos asis is c ucial o he su i al o all o ganisms. To ha e ec , p o eins dedica ed o he me al homeos asis con ol impo and expo (me allo anspo e s), a ge he deli e y o me als o speci ic enzymes (me allochape ones), a e esponsible o he me allocen e assembly, and egula e he exp ession o he o he p o eins in esponse o ace me al ions a ailabili y (me allo egula o s). When he me al- a icking p ocess mal unc ions, suscep ibili y o in ec ion and many pa hologies a i e. A mo e in-dep h e iew is done o he main me al e lux sys em in Chap e 4.4.2, as well as hei s uc u e and unc ioning. 4.4 Hos -Pa hogen In e ac ions Hos me al concen a ions and p esence can d as ically a ec suscep ibili y o bac e ial in ec ions [Palme and Skaa ,2016]. O he han he homeos asis con ol mechanisms desc ibed abo e, hos s ha e e ol ed o de elop many o he ways o con olling me al concen a ions and ha nessing hei capabili ies o igh he p esence o bac e ial pa hogens. Some o hose mechanisms include ways o c ea ing me al wi hd aws om pa hogens and causing a s oppage in me abolic pa hways ha enable cell g ow h, as well as o he s ha can unleash oxic le els o me als o p o oke pa hogen cell dea h. In he con ex o in ec ion, bac e ial and hos me al me abolism de e mine disease pa hogenesis un- de h ee eme ging p inciples. The i s one s a es ha al hough he hos ep esen s a ich nu ien me al sou ce o bac e ia, he hos immune sys em can block bac e ial me al acquisi ion in a p ocess denom- ina ed as e med nu i ional immuni y. Secondly, hos s ha e he abili y o o e load bac e ia wi h oxic concen a ions o me als. Finally, he de egula ion o hos me al homeos asis h ough gene ic mu a ion o nu i ion has an impac on he hos ’s suscep ibili y o in ec ion. 20 In o de o ob ain hei equi ed nu ien me als, bac e ia ha e e ol ed h ee main classes o me al acquisi ion sys ems: elemen al me al impo , ex acellula me al cap u e media ed by side opho es (low molecula weigh , high-a ini y Fe3+binding compounds sec e ed and impo ed by bac e ia o i on acqui- si ion and e en he om hos ), as well as me al acquisi ion om hos p o eins, being he la e a p ocess in which me al he om hos nu i ional immuni y p o eins unde o he mechanisms is also included. In esponse, mammalian hos s ha e e ol ed s a egies o es ic bac e ia om being success ul in hei me al acquisi ion p ocess. 4.4.1 Me al-Seques e ing P o eins Me al bioa ailabili y exe s a s ong selec i e p essu e a he in ec ious in e ace gi ing ise o an e olu- iona y a ms ace be ween hos and pa hogen ha shapes me al seques a ion s a egies. As a esponse o mic obial in ec ion, many hos s e ol ed he deploymen o me al-seques e ing hos -de ence p o eins o be e con ol he a ailabili y o essen ial me al nu ien s in he ex acellula space. In human hos s, he case o Calp o ec in (CP) se es as an example o hese mechanisms. CP is an abundan an imic obial he e oligome o wo S100 p o eins (S100A8 and S100A9) (Fig. 5) eleased om neu ophils and epi he- lial cells a si es o in ec ion ha seques e s di alen i s - ow ansi ion me al ions like Fe2+ and Zn2+ [Zygiel and Nolan,2018], ha , as i name highligh s, has p ope ies ha include he binding wi h elemen s like hose p e ioulsy men ioned and i s subsequen an imic obial ac i i y. This is jus one o he many examples o s a egies en o ced by hos s o inhibi he a ailabili y o essen ial ace me als o pa hogens, in ano he a emp o igh hei ne a ious e ec s o i s o ganism. Figu e 5: C ys al s uc u e o human calp o ec in (S100A8/S100A9). 21 4.4.2 Hos -Imposed Me al Toxici y Bac e ial Me al E lux Sys ems Hos s ha e e ol ed mechanisms o deploy oxic le els o me al ions o bac e ial con ol, and bac e ia use a la ge span o me al expo e s o esis hos in oxica ion [Guilhen e al.,2013]. These bac e ial expo e s can be di ided in i e main g oups: •Resis ance-Nodula ion-Cell di ision (RND) ype anspo e s; •P- ype ATPase amily; •Ca ion Di usion Facili a o (CDF) amily; •T anspo e Media ing Manganese Expo (Mn X) amily; •Cobal Resis ance p o ein A (Co A) amily. The Resis ance-Nodula ion-Cell di ision (RND) ype anspo e s a e in eg al memb ane p o- eins media ing he e lux o a b oad a ie y o subs a es wi h a subse expo ing me als. I is composed by a pump (anno a ed A in Fig. 6), an in eg al memb ane p o ein wi h a hyd ophilic pe iplasmic compo- nen , which is connec ed o a ime ic ou e memb ane ac o (C in Fig. 6), ha se es he pu pose o allowing he e lux o me al in o he ex acellula space [Paulsen e al.,1997]. The hi d and inal pa o he RND anspo e complex is composed o se e al uni s o a pe iplasmic memb ane p o ein (B in Fig. 6)l, se ing as a binde ha c ea es a ing in ol ing he ou e memb ane p o eins and he pump, p omo ing he s abiliza ion o he con ac be ween he wo o he componen s (Mu akami e al. [2002], Akama e al. [2004]). Se e al RND expo e sys ems ha e been iden i ied o media e he e lux o Co2+, Zn2+, Cd2+, Ni2+, Cu+and Ag+(Nies and Sil e [1989], Liesegang e al. [1993], S ähle e al. [2006], Long e al. [2012]). Membe s o he P- ype ATPase amily (Fig. 6) a e p esen in euka yo es and p oka yo es and couple me al anspo o he hyd olysis o ATP, in a p ocess whe e ca aly ic phospho yla ion o he anspo e oc- cu s a e binding o cy oplasmic me al o he ans-memb ane me al-binding si es [Fagan and Saie ,1994]. Me als a ec ed by he ac ion o hese anspo e s can be di ided in wo main g oups: Zn2+/Cd2+/Pb2+ o Cu+/Ag+(Bo ella e al. [2011], Fu e al. [2013]). 22 Ca ion Di usion Facili a o (CDF) amily membe s a e he media o s o he anspo o a a- ie y o me als in bo h p oka yo es and euka yo es [Haney e al.,2005]. These anspo e s a e usually cons i u ed by six ansmemb ane domains ollowed by a me allochape one-like cy oplasmic domain (Fig. 6) ha egula es me al anspo ac i i y [Lu e al.,2009]. These anspo e s media e he e lux o Zn2+ Cd2+and Fe2+, ha ing howe e he anspo o Fe2+ha ing been p o en less e ec i e han he o he s [Hoch e al.,2012]. A subclass o CDF expo e s, Mn E, a Mn2+ anspo e has also been desc ibed. Se e al ecen s ud- ies, howe e , ha e shown ha Mn E is no he only ype o Mn2+expo e , ha ing lead o he disco e y o he exis ence o he T anspo e Media ing Manganese Expo (Mn X) amily [Li e al.,2011]. Al hough much is ye unknown abou he s uc u e o his amily o expo e s, s udies ega ding he sec- onda y s uc u e and opological p edic ions ha e sugges ed an in e ed epea o h ee ansmemb ane segmen s [Vey ie e al.,2011](Fig. 6). These anspo e s ha e all been desc ibed o expo p incipally Mn2+wi h some lesse a ini y o o he di alen me als. The i h g oup o expo e s (Cobal Resis ance p o ein A (Co A) amily) we e i s iden i ied as Mg2+ anspo e s, in which some membe s a e dedica ed o he expo o o he di alen ca ions, p inci- pally Zn2+, wi h he excep ion o some eco ded cases o media ed anspo o Zn2+and Cd2+[Wo lock and Smi h,2002]. This p o ein ha bou s wo ansmemb ane domains and a long cy oplasmic egion (Fig. 6) ha media es he acquisi ion and subsequen deli e y o ca ions o he anspo channel. Figu e 6: Di e en amilies o bac e ial me al expo e s. 23 Pa hogen Homeos a ic Con ol As p e iously s a ed, hos s es ic bac e ial g ow h by dep i ing pa hogens o essen ial le els o ace me als, and by de ini ion, expo ac i i y has a a he unhelp ul unc ion o ha ma e . The main enance o iable concen a ions o me als unde luc ua ing le els imposed by hos s composes an in e ac ion ha u he ansla es he igh happening in he on line o he hos -pa hogen in e ace. Acco ding o he abo e desc ip ion, mos o he me al expo e s om bac e ial pa hogens a e dedi- ca ed o media ing he expo o h ee ansi ion me als (Zn, Cu, and Mn) ha a e p esen in subs an ial amoun s in he human body. In obse ed cases o he in luence o ace me als in i ulence, howe e , i has been shown an inc ease o Zn2+le els du ing in ec ion [Gaballa and Helmann,2003], pos ula ing he use o Zn2+expo sys ems o ace he ise o oxic le els o his me al. O he cases ha e also demons a ed ha animals use coppe as an an i-mic obial weapon by inducing oxida i e s ess [Haney e al.,2005]. I has also been es ablished ha he lack o Cu+expo e s on pa hogens is di ec ly ela ed o hei dec ease in i ulence and abili y o su i e on he hos o ganism. This al oge he e eals ha al hough hos s ha ness he powe o hese me als o igh in ec ion [Bo ella e al.,2012], pa hogens explo e he capabili ies o hei expo sys ems o emedia e he e ec s o his ise, enabling he conclusion ha pa hogens ha lack me al expo e s p esen impai ed i ulence. 4.5 Chap e Summa y In his chap e an o e iew o he cha ac e is ics, e ec s, homeos a ic con ol mechanisms as well as he in e ac ions ha a e shaped by he necessi ies exe ed by o ganisms o ob ain ace me als. A special emphasis on i s - ow ace elemen s was made, jus i ied by hei mos common p esence in biological sys ems, as well as he numbe o in e ac ions media ed by hei p esence. 24 Pa II Co e o he Disse a ion 25 5.3.4 Model calib a ion The calib a ion o enzyme-cons ained models is c ucial o he ob en ion o alid models and ele an esul s. The e o e, when conside ing his impo an s ep, many app oaches can be employed. In o he simila asks, s a egies like he use o in i o esul s o he use o p o eomic measu emen s we e used bu , conside ing bo h he ammoun o models being c ea ed and calib a ed a he same ime and he a ailable omics da a and in i o s udies, he bes s a egy in ol ed he calib a ion o models by i ing hem bo h acco ding o he g ow h esul s ob ained expe imen ally and making hem iable o g ow h wi h one ano he . To ha e ec , he i s pa o his ask in ol ed he calib a ion o each model’s p o ein pool exchange P alue. This was pe o med by manually i ing he he p edic ed maximal g ow h a es o each model acco ding bo h o he low i on condi ions p e iously desc ibed and whose esul s we e based o o ex- pe imen ally oba ined esul s and alues i ed in o he expe imen s and published enzyme-cons ained models [Adadi e al.,2012,Ca asco Mu iel e al.,2023]. Table 4: Maximum p edic ed g ow h unde di e en P alues P o ein pool PB Bu Ec Fn Ri Sp Ss 0.1 0.000037 0.000082 0.000767 0.000195 0.000341 0.000144 0.000126 0.2 0.000074 0.000164 0.001535 0.000390 0.000683 0.000288 0.000252 0.3 0.000111 0.000245 0.002302 0.000584 0.001024 0.000431 0.000379 0.4 0.000148 0.000327 0.003070 0.000779 0.001366 0.000575 0.000505 0.5 0.000185 0.000409 0.003837 0.000974 0.001707 0.000719 0.000631 0.6 0.000222 0.000491 0.004604 0.001169 0.002048 0.000863 0.000757 0.7 0.000259 0.000573 0.005372 0.001364 0.002390 0.001006 0.000883 0.8 0.000296 0.000654 0.006139 0.001559 0.002731 0.001150 0.001010 0.9 0.000333 0.000736 0.006906 0.001753 0.003073 0.001294 0.001136 1.0 0.000370 0.000818 0.007674 0.001948 0.003414 0.001438 0.001262 32 Basing he alues o o hose p e ioulsy es ablished in o he models o Esche ichia coli was he i s s ep o de e mine he p o ein pool alues o he es o he models. Following he P alue de e mined in Adadi e al. [2012] o 0.095 gdw−1h−1, simula ions we e un in o de o de e mine he easibili y o he model when pu in a communi y se ing wi h o he models. A e unning he simula ions, a alue o 0.13 was deemed ideal as i p o ided mos compa ibili y wi h he emaining models. A e wa ds, in o de o de e mine he alues o he o he models, adjus men s we e made based on bo h he ammoun o enzyma ic da a was added in he models and how hei p o ein pool should be in p opo ion o he al eady es ablished alue o E. coli . Fo ha , a ious simula ions we e un on he samples in o de o ind a consensus ha would make he simula ions easible and allowed g ow h, all while main aning ealis ic alues o all models. Mo eo e , his ype o app oach ollowed a mo e empy ical way o calib a ing he models. A e delibe a ing on he op imal alues (Table 5), he models we e hen eady o be sampled. Table 5: Chosen p o ein pool P alues o each model Model P o ein pool alue (gdw−1h−1) B 0.075 Bu 0.075 Ec 0.13 Fn 0.046 Ri 0.078 Sp 0.082 Ss 0.048 This calib a ion p ocess did no in ol e any kind o kca adjus men as he lack o a ailable enzyma ic in o ma ion ac oss all models would c ea e an imbalance in he models ha would u he cause oubles when he sampling and subsequen simula ions we e o occu . None heless, he models we e deemed p epa ed and he samples eady o be c ea ed, as will be u he explained in Sec ion 5.4.3. 33 5.4 Communi y simula ions 5.4.1 G ow h media G ow h media o he selec ed samples was chosen o po ai di e en le els o a ailabili y o bo h o ganic and ino ganic i on in gu mic obial communi ies, wi h he objec i e o ul ima ely obse e how communi ies dynamics a e a ec ed du ing a ious s ages o coli is when he hos exe s me abolic p essu e by educing he a ailabili y o i on in he ex acellula space, en isioning he obse a ion o how his a ec s communal composi ion and p omo es he adap ion o he me abolic p o iles o bo h indi iuduals and he communi y as a whole. The chosen g ow h media is ep esen ed in Table 6. Table 6: G ow h media o he simula ions (a) I on a ailabili y Media Cons ain s M1 33 µM Fe M2 0µM Fe/H M3 0.7µM H; 8.2µM Fe M4 2.7µM H; 33 µM Fe M5 2.7µM H M6 125 µM Fe M7 125 µM H M8 125 µM H; 2mM Fe M9 125 µM H; 125 µM Fe M10 No cons ain s (b) Composi ion Type Compounds Suga D-glucose, F uc ose, Cellobiose, Mal ose, Lac ose, Inulin Amino acids L-His idine, L-Isoleucine, L-Leucine, L-Me hionine, L-Valine, L-A ginine, L-Cys eine, L-Glu ama e, L-Phenylalanine, L-P oline, L-Aspa agine, L-Aspa a e, L-Glu amine, L-Se ine, L-Th eonine, L-Alanine, Glycine, L-Lysine, L-T yp ophan, L-Ty osine Nucleo ides Adenine, Guanine, U acil, Xan hine Sal s and mine als Magnesium Chlo ide Vi amins and Bio in, Ribo la in, Fola e, Niacin, an ioxidan s Inosi ol, L-Glu a hione educed O he s NAD The choosing o he abo e media e lec s he need o emula e he condi ions h ough which he in i o expe imen s we e conduc ed. This was achie ed by emula ing he GMM+LAB plus Mucin media [T amon ano e al.,2018] wi hou mucin in which he in i o expe imen s we e conduc ed and de ining a g ow h media wi h he a ailable exchange eac ions in he models. Th ough his, a compa ison and 34 e alua ion o he ob ained esul s agains hose ob ained expe imen ally can be made, allowing o he ine uning o models in eaching he end goal o cap u ing obse ed mechanisms and in e ac ions in hose same s udies. 5.4.2 Simula ions To p edic he s eady-s a e me abolic lux dis ibu ions in he selec ed mic obial communi ies using genome- scale me abolic models, wo app oaches we e employed: FBA and S eadyCom. In his sec ion is p esen an o e iew o each one and hei unc ioning, while explaining how hey we e applied o he cu en s udy as well as hei ad an ages and disad an ages. FBA This app oach is a di ec ex ension o Flux Balance Analysis (FBA) (o join FBA), which in eg a es me abolic econs uc ions o indi idual mic obial species in o a mul i-compa men model wi h a commu- ni y compa men allowing o he exchange o me aboli es be ween species. The op imiza ion objec i e unc ion is usually he sum o he biomass eac ions o indi idual species (called he communi y biomass). Gene ally, his app oach o en equi es addi ional cons ain s o cap u ing obse ed co-g ow h beha io . The ollowing equa ion p esen s he linea p oblem h ough which his me hod ope a es. Le Kbe he se o all o ganisms in he communi y. Fo an o ganism kin K, he adi ional FBA o p edic ing maximum g ow h can be s a ed as ollows: max k biomass subjec o ∑ j∈Jk Sk ij k j= 0,∀i∈Ik LBk j≤ k j≤UBk j,∀j∈Jk whe e k jis he lux o eac ion j(in mmol gdw−1h−1 o gene al me abolic eac ions, in gdw−1h−1 o he biosyn hesis o mac omolecules, and in h−1 o he biomass eac ion), Sk ij is he s oichiome y o me aboli e iin eac ion j,LBk jand UBk ja e he lowe bound and uppe bounds o luxes k j espec- i ely, Ikand Jka e espec i ely he se o me aboli es and eac ions o o ganism k. All his allow he cap u ing o impo an ea u es o mic obial communi ies such as c oss- eeding and compe i ion. F amewo ks based on Flux Balance Analysis (FBA) exhibi , howe e , a undamen al omission. This s ems om he ac ha he biomass eac ion lux no only se es as a sink o biomass cons i uen s bu 35 also quan i ies he speci ic g ow h a e (in h−1), he eby indica ing he size o he sys em. In he case o a mono-cul u e, a single biomass lux is p esen , no malized by speci ic a es o consump ion o p oduc ion (in mmol gdw−1h−1). Howe e , in he con ex o mul iple o ganisms g owing oge he , a uni ied FBA amewo k does no necessa ily impose cons ain s on he g ow h a e o all pa icipa ing membe s in he communi y. This absence o a consis en g ow h a e ac oss all mic obes wi hin he communi y may esul in me abolic luxes ha a e inconsis en wi h a s able a e age communi y composi ion as he as es g owing o ganism will ake o e he popula ion. S eadyCom This app oach su ged as an answe o many o he p oblems s a ed be o e and is e med an ex ension o Flux Balance Analysis ha akes in o accoun he ac ha o each a s able composi ion he o gan- isms need o g ow a he same speci ic g ow h a e (h−1), which means ha he absolu e g ow h a e (gDW −1/h−1) o each o ganism is p opo ional o i s abundance a s eady-s a e (gDW −1). Using he ollowing equa ion, he maximum communi y g ow h a e µmax o a communi y sa is ying he communi y s eady-s a e can be ound by sol ing he ollowing non-linea op imiza ion p oblem: maxµ subjec o         ∑j∈JkSk ijVk j= 0,∀i∈Ik LBk jXk≤Vk j≤UBk jXk,∀j∈Jk Vk biomass =Xkµ Xk≥0         ∀k∈K uc i−ec i+∑k∈KVk ex(i)= 0,∀i∈Icom ∑k∈KXk=X0 , ec i≥0,∀i∈Icom Fo con enience he communi y expo a es ec iand up ake a es uc ia e no malized o one uni o o al communi y biomass, he e o e X0is se a 1 gDW and Xkis hus equal o he ela i e abundance o o ganism k. By se ing X1= 1 and Xk= 0 o k > 1, S eadyCom is educed o he s anda d single- o ganism FBA model and he agg ega e biomass lux coincides wi h he speci ic g ow h a e. Simila ly o wha happens in single-o ganism FBA, cons ain s on he sys em up ake a es uc ia e su icien o gua an ee a ini e solu ion (i.e. ini e µmax). Fu he mo e, p edic ions by S eadyCom a e in gene al di e en om he p edic ions by join FBA because o he cons ain s ela ing biomass, he bounds o speci ic a es, he agg ega e luxes and he communi y g ow h a e. 36 5.4.3 Sampling The sampling s ep o he wo k he e p esen ed was o mos impo ance, as i no only was a de e minan in he whole wo k low bu also allowed o a be e unde s anding o each o ganism’s beha iou in he p es- ence o di e en mic obial o ganisms. The ini ial i e a ion o hese samples in ol ed he c ea ion o g oups o 3 models whe e 2 models p e ailed in he en i e y o he samples: he Bac e oides he aio aomic on VPI 5482 and Bac e oides uni o mis ATCC 8492, wi h he hi d model being each one o he emnan 5 a ge models. This p o ided 5 ini ial samples om which simula ions unde many di e en condi ions we e conduc ed and esul s gene a ed. F om ha i was obse ed ha hose communal composi ions we e no ideal, mainly due o he mo e esou ce demanding up ake p o ile o B. uni o mis ATCC 8492 ha hinde ed he g ow h o B. he aio aomic on VPI 5482 ac oss all samples. Conside ing he p e iously s a ed ac s, 6 new samples we e designed wi h he goal o s udy he g ow h o B. he aio aomic on VPI 5482 du ing coli is, wi h a ocus o unde s anding he g ow h o his s ain unde he p esence o o he mic obes wi h di e en consump ion p o iles obse ed in p o ided esul s om in i o expe imen s and con i med by means o unning simula ions in e e y single model unde same abi- o ic condi ions as hose conduc ed expe imen ally. These new samples cons i u ed he inal a ge o his expe imen and possibili a ed bo h he co obo a ion o in i o esul s in o he s udies and he obse ing o new esul s ha u he p o e he p ecision and accu acy o hese ep esen a ions and simula ions. Table 7: Chosen samples (S1 o S6) and p esen models O ganism S1 S2 S3 S4 S5 S6 B. he aio aomic on VPI 5482 • • • • • • B. uni o mis ATCC 8492 • E. coli ED1a • F. nuclea um subsp. nuclea um ATCC 25586 • R. in es inalis L1 82 • S. pa asanguinis ATCC 15912 • S. sali a ius DSM 20560 • 37 As displayed in Table 7, he chosen samples consis o communi y models ha combine B. he aio- aomic on VPI 5482 wi h each o he emaining models. This choice e lec ed he need o mee he end goal o in e ing on he g ow h o his s ain unde he p esence o bo h commensal and pa hogenic o gan- isms unde di e en concen a ions o he chosen ansi ion me als o his s udy case. Mo eo e , since all models p esen such di e en me abolic p o iles when compa ed wi h B. he aio aomic on , wi h he excep ion o B. uni o mis as i will be u he demons a ed, i will aso p o ide aluable insigh s on he in e ac ions ha occu be ween 2 di e en mic obes in condi ions ha induce me abolic s ess. 5.5 Wo k low The c ea ion and applica ion o a wo k low (Figu e 8) ha be e sui s he de ised endea ou is i al o he o e all success o he wo k. He e, an explana ion o he en i e y o he s eps o be aken unde he expe imen s made, wi h an emphasis on he me hods and ools used along he en i e y o he p ocess. Figu e 8: Visual ep esen a ion o he es ablished wo k low. 38 When obse ing he ep esen a ion o he es ablished wo k low in Figu e 8one can unde s and ha he na u e o he de eloped wo k ollowed a ial-e o app oach, he e o e enabling he cons an uning o ce ain aspec s in he sea ch o imp o emen o he end esul . Mo eo e , by implemen ing such plan, each i e a ion p oduced in e es ing esul s ha , e en i no op imal o conside ed wo hy o men ioning, se ed as bo h a guideline and a bluep in o he esul s ha shall be p esen ed and discussed in Chap e 6. The p elimina y s age o sampling in ol ed he c ea ion o samples ha cons i u ed he s udy cases o his wo k. Fi s ly, desi ed species o mic oo ganisms we e selec ed acco ding o in i o s udies pe o med pa allel o his s udy. Secondly, he selec ed g ow h media we e selec ed o be e emula e die a y sou ces o many kinds and selec ed acco ding o in i o app oaches. Finally, he samples we e c ea ed om he combina ion o all hese ac o s. The c ea ion o models encompassed he acqui emen o GEMs o he selec ed o ganisms. The models we e ob ained as p e iously men ioned om he AGORA eposi o y. The ob ained models hen unde wen a cu a ion p ocess ollowed by an implemen a ion o he GECKO me hod, in which hey we e be enhanced wi h enzyma ic cons ain s using kine ic da a supplied by in i o esul s p esen in a ious da abases, and also by making use o a Deep-Lea ning model ha p edic s u no e alues based on subs a es and enzyma ic da a. The en i e p ocess as well as he c ea ion o communi y models om each singula Genome-Scale Me abolic Models was accomplished using bo h a ious me hods and clasees om he MEWpy [Pe ei a e al.,2021] package and o iginal code. Simula ions o singula and communi y models unde he selec ed abio ic condi ions was done using bo h Flux Balance Analysis and S eadycom [Chan e al.,2017], a simula ion me hod ha cons i u es a gene aliza ion and an ex ension o FBA wi h he abili y o p edic he change in species abundance in esponse o changes in die s wi h minimal addi ional imposed cons ain s on he model. This me hod, by aking in o accoun he ac ha o each a s able composi ion he o ganisms need o g ow a he same speci ic g ow h a e (1/h), c ea es a di ec p opo ionali y be ween he absolu e g ow h a e (gDW/h) o each o ganism and i s abundance a s eady-s a e (gDW ). The ou h s ep o he wo k plan consis s on he ob en ion o esul s and analysis. This s ep was a successo o he simula ions, as he e he analysis o i s esul s, as well as he c ea ion o he i s hypo hesis su ged. Finally, he ending s ep o each i e a ion o his cyclic wo k low cons i u ed a i al pa o he wo k, as he e compa isons we e made be ween he ob ained esul s and o he esul s ob ained bo h h ough p e ious simula ions and hose ob ained using in i o app oaches unde simila a ge condi ions. He e, he p ecision o he encompassed me hod was also cons an ly e alua ed, wi h he in en ion o assessing he 39 ad an ages and disad an ages o each me hod being employed, he e o e seeking a balance ha c ea es he possibili y o de eloping a be e me hod o he p oposed wo k and leading o he de elopmen o be e and imp o ed ones. Finally, new samples we e c ea ed and new changes we e made o be e enhance he me hod each ime a cycle ended. 5.6 Code a ailabili y The en i e y o he code and ob ained esul s is publicly a ailable and can be used in i s in eg i y o ex ended o conduc new s udies. The sou ce code o he de eloped p ojec can be ound a : h ps://gi hub.com/a eias03/ mCOM.gi . 5.7 Chap e Summa y In his chap e , a summa y o he s eps de ised o he execu ion o he in ended expe imen s is p esen ed, wi h a isual display o he planned wo k schedule also being highligh ed. 40 Chap e 6 Resul s and Discussion 6.1 Model Analysis 6.1.1 Simila i y The enhancemen p ocess o which he models we e subjec ed was he o igin o many changes in each model’s me abolic p o ile. These changes, e en i mainly a ec ing eac ion’s s oichiome y and p o ein molecula weigh s, caused also many changes in each model’s naming o eac ions and me aboli es, as well as hei o ganiza ion by e e sibili y. In Figu e 9i is possible o obse e some o hose changes. Rega ding me aboli e changes when conside ing he enhancemen o models wi h enzyma ic cons ain s h ough he p ocess he e employed, one mus conside he addi ion o genes as pseudo- eac ions a majo ac o ha impac s he models. Specially, conside ing he shee numbe o me aboli es added his way causes, as obse ed, a dec ease in simila i y among he models. In he case o he o e lap o eac ions in Figu e 10, by aking in o accoun he ac ha mapped eac ions’s s oichiome ies a e changed and spli ed by e e sibili y, i is expec ed ha such changes in simila i y occu a e he p ocess is concluded. Finally, in he case o he obse ed changes in esou ce o e lap, no changes we e seen as obse ed in Figu e 11 as no changes made in he models as hey we e enhanced impac ed he models’ up ake eac ions’ bounds and he e o e i did no cause a dec ease in simila i y in e ms o up ake o e lap. I is o mos impo ance o no e howe e , ha he esul s he e displayed may no be indica i e o he eal changes ha occu ed in he models, as esul s displayed u he in his sec ion show many di e ences specially ega ding he changes in up ake p o iles o he a ious models. 41 simula ion me hod i sel . An in e es ing esul su ged, howe e , when unning he simula ions h ough S eadyCom. He e, due o he ac ha he cons ain s a e applied o all membe s o a communi y and he consis ency in e ms o g ow h a es amongs all mic obes, i was obse ed ha only enzyme-cons ained we e able o g ow in low i on condi ions. This obse a ion u he leads o he conclusion ha he enzyma ic enhancemen o models u he ex ended he capabili ies o Genome-Scale Me abolic Models (GEM) by capping he o e aking o esou ces by species ha possess a highe a e o consump ion o me aboli es. Mo eo e , i is also possible o conclude ha his p ocess con e ed s abili y o he communi ies as he hinde ing o g ow h o all models applied by bo h cons ain s employed by S eadyCom and he enzyma ic enhancemen o models u he con ibu ed o a mo e con olled g ow h o all membe s o each communi y in spi e o hei me abolic p o ile ha p omo ed he a o ing o o ganisms wi h mo e ca aly ically e icien enzymes wi h a lowe syn hesis cos . 6.3 Bac e oides he aio aomic on g ow h du ing coli is In he esul s published in Zhu e al. [2020] i was epo ed ha he e was obse ed g ow h o Bac e oides he aio aomic on in communi y wi h species om he En e obac e iciae amily du ing coli is, whe e he acquisi ion o i on was media ed h ough he u iliza ion o xenoside opho es o En e obac e iciae like Es- che ichia coli and Salmonella Typhimu ium . Mo eo e , hese esul s ep esen a mechanism o commen- salism be ween species in gu mic obial communi ies ha u he e lec he complex in e ac ions ha occu in hese se ings and whose cha ac e is ics a e he goal o his esea ch. To ha end, i was decided ha pa o hese esul s would in e on he g ow h o he samples unde di e en le els o i on and y o obse e and cap u e, o some ex en , simila i ies be ween he in i o and in silico expe imen s. Mo eo e , o achie e his, simula ions using bo h FBA (Figu e 17) and S eadyCom (Tables 9and 10) we e pe o med using he di e en samples unde he di e en g ow h media, he e o e allowing he compa ison be ween bo h he esul s ob ained using he di e en simula ion me hods and he obse ed di e ences in enzyme-enhanced and non-enhanced samples. The esul s ob ained om he Flux Balance Analysis simula ions we e a he su p ising. On one side, e en i he ull ex en o he mechanism was no mean o be cap u ed using his me hodology, he hope would be o i o ac ually poin owa ds ha poin . On he o he , he expec ancy ha he enhancemen o models would be he solu ion o he en i e y o he p oblem would be somewha a e ched. Mo eo e , since FBA does no apply he cons ain s equally ac oss all membe s o a commmuni y, he ac ha o ganisms like E. coli ED1a g ow mo e han B. he aio aomic on VPI 5482 in a communi y is wi hin he 48 Enhanced Non-enhanced Figu e 17: G ow h o B. he aio aomic on in di e en communi ies h ough Flux Balance Analysis expec ed ange o esul s. One con adic o y esul o he expec ed howe e , happened in sample 6, whe e he eally di e en up ake p o iles obse ed in Figu e 7be ween B. he aio aomic on and S. sali a ius would ha e c ea ed a sample ha would p omo e a lo o c oss- eeding in e ac ions and he e o e, g ow h in si ua ions o low le els o ace me als. I is wo hy o no e howe e ha , as obse ed in Figu e 13, he simila i ies in g ow h sugges ha bo h models do no possess g ea i ness in e ms o adap abili y o low le els o i on. Mo e in e es ing esul s su ged howe e , when unning he simula ions wi h S eadyCom (Tables 9and 10). He e, a endency was obse ed ha was cons an ac oss enhanced and non-enhanced samples. When obse ing he esul s p esen ed in Table 9, i is possible o obse e he occu ence o a pa e n in he g ow h a es. While sample 2 (B + Ec) was able o main ain a nea s able g ow h a e ac oss all media, he emaining samples we e only able o g ow in condi ions whe e he e we e no es ic ions in e ms o i on a ailabili y. Mo eo e , wi h he excep ion o samples 1 (B + Bu) and 5 (B + Sp) all samples p esen ed exo bi an g ow h a es, ha a e indica i e o hei un es ained abili y o g ow unde media ha is no well de ined. I is howe e , a posi i e indica ion o he p oo ha hese esul s hope o achie e, ha despi e he nea cons an g ow h o sample 2 i was he only one ha g ew unde luc ua ing le els o bo h o ganic and ino ganic i on. Conce ning he abundances o each esul howe e , as seen in Table 11 ha in sample 2 he abundance o Bac e oides he aio aomic on was 0 a s eady-s a e ac oss all media, meaning ha e en hough he e was communi y g ow h, he lack o cons ain s ac oss all models o each samples made i impossible o a balance in e ms o abundances o occu . 49 Table 9: G ow h o non-enhanced samples h ough S eadyCom Cons ain s B + Bu B + Ec B + Fn B + Ri B + Sp B + Ss 33 uM Fe 0 153.504883 0 0 0 0 0 uM Fe/H 0 153.504883 0 0 0 0 0,7 uM H; 8,2 uM Fe 0 153.504883 0 0 0 0 2,7 uM H; 33 uM Fe 0 153.504883 0 0 0 0 2,7 uM H 0 153.504883 0 0 0 0 125 uM Fe 0 153.504883 0 0 0 0 125 uM H 0 153.504883 0 0 0 0 125 uM H; 2 mM Fe 0 153.504883 0 0 0 0 125 uM H; 125 uM Fe 0 153.504883 0 0 0 0 No cons ain s 0.954102 153.635742 30.661133 54.433594 1.657227 41.800781 Table 10: G ow h o enzyme-enhanced samples h ough S eadyCom Cons ain s B + Bu B + Ec B + Fn B + Ri B + Sp B + Ss 33 uM Fe 0 5.179688 0 0 0 0 0 uM Fe/H 0 5.179688 0 0 0 0 0,7 uM H; 8,2 uM Fe 0 5.179688 0 0 0 0 2,7 uM H; 33 uM Fe 0 5.179688 0 0 0 0 2,7 uM H 0 5.179688 0 0 0 0 125 uM Fe 0 5.179688 0 0 0 0 125 uM H 0 5.179688 0 0 0 0 125 uM H; 2 mM Fe 0 5.179688 0 0 0 0 125 uM H; 125 uM Fe 0 5.179688 0 0 0 0 No cons ain s 0.477539 5.185547 0.369141 2.761719 1.378906 0.597656 50 E en hough he p e iously obse ed pa e n in he he non-enhanced samples pe sis ed, he e we e some changes. In sample 2, a nea cons an g ow h a e o app oxima ely 5.2 gdw−1h−1was main ained ac oss all media and simila ly posi i e g ow h a es we e main ained in he emaining samples in condi ons whe e he e was no es ainmen in e ms o i on le els. I was howe e in he abundances (Table 12) ha a shi in he pa e n occu ed. He e, no only has B. he aio aomic on sus ained a high abundace o app oxima ely 31% in sample 3 unde no cons ain s bu also, e en i p ominen ly low, a posi i e abun- dance in sample 2 in 3 di e en media (M3, M4 and M7) u he p o ing ha , e en i unde ex emely low abundances, Bac e oides he aio aomic on can g ow unde low ansi ion me al le els in communi y se ings wi h o he mic obes belonging o he En e obac e iaceae amily like Esche ichia coli . These esul s a e a ex emely posi i e demons a ion ha he ex ension o he models wi h enzyme cons ain s is a s ep owa ds he abili y o cap u e mechanisms ha occu wi hin gu mic obial communi ies. Fu he mo e, i is possible o conclude ha i he a be e adjus men o he kca alues was o be made, hese esul s could be u he imp o ed. 6.4 Limi a ions The ob ained esul s we e hinde ed o some ex en mainly by bo h he quali y o he models and he inabili y o be e une hese same models, mainly due o he lack o a ailable enzyma ic in o ma ion. Fu he mo e, e en wi h he con inuous su ging o simula ion me hods applied o he simula ion o mic obial communi ies, he e is s ill no de ini i e solu ion o he challenges encoun e ed when using all o he ones he e p esen ed. I is o upmos impo ance o no e, howe e , ha none o hese ac o s we e a main cause o impedance o he ob en ion o ele an esul s in any s age o his wo k, and a he a challenge o o e come. 51 6.5 Chap e Summa y In his chap e he esul s p oduced by he es ablished amewo k we e p esen ed and analyzed, in acco - dance wi h he designed expe imen s. This p ocess eso ed mainly in he compa ison o he ob ained g ow hs and abundancdes o each se o samples (enzyme-enhanced and no ) ela i ely o he case s udies pe o med, de e mined by he selec ed g ow h media. The i s componen o he esul s was ela ed o he obse ed changes in he models in e ms o simila i y be o e and a e he p ocess o enhancemen wi h enzyma ic cons ain s. The esul s he e p esen ed p o ided s ong insigh s on he impac he enhancemen had on he models and se he g ound o he esul s o be p esen ed nex . Nex , he di e ences in up ake p o iles o he models shedded ligh s on he impac ha he applied cons ain s in he models applied on each model’s up ake p o ile and how hey co ela e wi h each o he . This e ealed c ucial as he p o iling o each model u he allows he comp ehension on how cons ainnig a model impac s he en i e me abolic ne wo k. The nex se o esul s comp ised o a s udy on communi y composi ion pe o med ha , while no being impac ul on he o e all wo k, p o ided in e es ing iews on how he calib a ion o each model impac ed he inal esul . Finally, one o he mos impo an end goals o his wo k was me as he an a emp o cap u e an obse ed mechanism in i o was made. In his sec ion o esul s, he g ow h esul s o he a ious samples in he selec ed media was p esen ed as hey we e simula ed h ough S eadyCom. These esul s a e o mos impo ance o his esea ch and e en i lacking some imp o emen , se ed as an impo an basis ha u he p o es he ele ance o hese ypes o app oaches. 52 Chap e 7 Conclusions 7.1 Summa y The objec i e o his endea ou was o de elop a me hodology capable o accu a ely simula e he be- ha iou o gu mic obial communi ies unde luc ua ing le els o ansi ion me als. The i s s ep was o e iew he s a e-o - he-a , unde s and he app oaches cu en ly being used in his ield and y o le e - age he combina ion o some o hese me hodologies wi h o he and newe ones. This e iew e ealed a lo o in e es ing app oaches and me hodologies being employed wi h he capaci y o ex end ou cu en knowledge o he in e ac ions ha occu wi hin bo h he mic obe-mic obe and mic obe-hos axises. A e his e iew, he desi ed se o me hodologies we e selec ed wi h he in en ion o hem being applied and ex ended owa ds his s udy case. The selec ed me hodologies included he enhancemen o models wi h enzyma ic cons ain s and he use o a ious simula ion me hods like join Flux Balance Analysis (FBA) and S eadyCom, and a e he wo low was designed, i was hen applied. The es ablished wo k low included he use o in i o esul s om wo main sou ces: op ical densi y measu emen s o se e al mic obial s ains ha e lec ed hei g ow h unde di e en le els o ace me als p o ided by he seconda y counselo o his disse a ion and hose published in Zhu e al. [2020] and T amon ano e al. [2018] o base a ious aspec s o he wo k om, including a ge mechanisms o cap u e and abundances o he samples. Nex , he applica ion o he de ised app oach lead o he ob en ion o in e es ing esul s ha we e bo h in e es ing and encou aging in e ms o po en ial. 53 Fi s ly, he p oduced esul s demons a ed he impo ance o he enhancemen o models wi h enzy- ma ic cons ain s in he s abili y o a communi y and u he p o ed hei impo ance in he sea ch o accu a e ep esen a ions o he communi y space. In ano he , he esul s we e o some ex en se back by he lack o a ailable in o ma ion ha could be c ucial in he calib a ion o each model. Mo eo e , his hinde ance u he lea es he possibili y o u he wo ks o ex end and upg ade he wo k he e conduc ed. Gi en he ob ained esul s, i is possible o conclude ha he majo goal o c ea ing a communi y modelling me hod o cap u ing mic obial gu communi y dynamics ega ding luc ua ions o ansi ion me als we e a ained. 7.2 P oblems and challenges As p e iously men ioned in Sec ions 5.2 and 6.4, e en i he comple e de elopmen o he amewo k was success ul, he e we e some obs uc ions bo h in he de elopmen and he ob en ion o esul s. The i s se o challenges ega d he di icul ies encoun e ed when dealing wi h he models and hei cu a ion p ocess. This p o ed i sel as a he di icul o o e come as he semi-au oma ed o o gebe a ion om which hey we e c ea ed o en leads o p oblems in he models ha in se e al occasions de icul ed o e en impeded he p og essing o bo h he enhancemen o he models o he simula ions. The la e e e s mainly o he lack o in eg a ion o enzyma ic da a om ei he li e a u e o in i o esul s ha p omo ed a mo e ial-and-e o app oach o he calib a ion p ocess o he models. E en i , o some ex en , lacking accu acy, he calib a ed models pe o med wi hin hei expec ed manne and p oduced esul s ha a e wo h o no ice and upg ade. 7.3 Con ibu ions The p esen ed endea ou in oduced se e al ou comes ha ha e he po en ial o con ibu e o he u he de elopmen o mic obial communi y simula ion me hods. Those con ibu ions a e: • De elopmen o an au oma ed amewo k o gene a e enzyme-enhaced models; • mCOM, a amewo k ha possesses many ools and sc ip s used in his s udy and can be u he ex ended o o he use cases o mic obial communi y simula ions; • Analysis and alida ion o in i o esul s ega ding he e ec o ansi ion me als in gu mic obial communi ies. 54 7.4 Fu u e Wo k The de eloped wo k p esen s a ounda ion o u u e wo ks h ough which his can se e as bo h an ex ension and an imp o emen o ha employed he e, in he sea ch o be e and mo e accu a e esul s. Fo ha e ec , u u e asks migh include: • C ea ion o me abolic models om a non-au oma ed app oach; • Be e cu a ion o models; • In eg a ion o p o eomics and o he enzyma ic da a in he calib a ion p ocess. 55 Bibliog aphy Roi Adadi, Benjamin Volkme , Ron Milo, Ma hias Heinemann, and Tome Shlomi. P edic ion o Mic obial G ow h Ra e e sus Biomass Yield by a Me abolic Ne wo k wi h Kine ic Pa ame e s. PLOS Compu a ional Biology , 8(7):e1002575, May 2012. ISSN 1553-7358. doi: 10.1371/jou nal.pcbi.1002575. Hi oyuki Akama, Misa Kanemaki, Masa o Yoshimu a, Tomi ake Tsukiha a, Tomoe Kashiwagi, Hi oshi Yoneyama, Shin-ichi o Na i a, A sushi Nakagawa, and Taiji Nakae. C ys al s uc u e o he d ug discha ge ou e memb ane p o ein, op m, o pseudomonas ae uginosa: dual modes o memb ane ancho ing and occluded ca i y end. Jou nal o Biological Chemis y , 279(51):52816–52819, 2004. Claudia And eini, Lucia Banci, I ano Be ini, and An onio Rosa o. Coun ing he zinc-p o eins encoded in he human genome. Jou nal o p o eome esea ch , 5(1):196–201, 2006. Hélène Bo ella, Pascale Pey on, Flo ence Le illain, Renaud Poincloux, Yannick Poque , I ène B andli, Chuan Wang, Ludo ic Tailleux, Syl ain Tilleul, Guillaume M Cha iè e, e al. Mycobac e ial p1- ype a pases media e esis ance o zinc poisoning in human mac ophages. Cell hos & mic obe , 10(3): 248–259, 2011. Hélène Bo ella, Gus a o S ad hagen, Geannca lo Lugo-Villa ino, Chan al de Chas ellie , and Oli ie Ney- olles. Me allobiology o hos –pa hogen in e ac ions: an in oxica ing new insigh . T ends in mic obiology , 20(3):106–112, 2012. Elizabe h A Came on and Vanessa Spe andio. F enemies: signaling and nu i ional in eg a ion in pa hogen- mic obio a-hos in e ac ions. Cell hos & mic obe , 18(3):275–284, 2015. Jo ge Ca asco Mu iel, Ch is ophe Long, and Nikolaus Sonnenschein. Simul aneous applica ion o en- zyme and he modynamic cons ain s o me abolic models using an upda ed Py hon implemen a ion o GECKO. Mic obiology Spec um , 0(0):e01705–23, Oc obe 2023. doi: 10.1128/spec um.01705-23. Siu Hung Joshua Chan, Ma ga e N Simons, and Cos as D Ma anas. S eadycom: p edic ing mic obial abundances while ensu ing communi y s abili y. PLoS compu a ional biology , 13(5):e1005539, 2017. 56 James R Cla k, S ua J Daines, Timo hy M Len on, And ew J Wa son, and Hywel TP Williams. Indi idual- based modelling o adap a ion in ma ine mic obial popula ions using gene ically de ined physiological pa ame e s. Ecological modelling , 222(23-24):3823–3837, 2011. Ka ha ine Z Coy e, Jonas Schlu e , and Ke in R Fos e . The ecology o he mic obiome: ne wo ks, com- pe i ion, and s abili y. Science , 350(6261):663–666, 2015. I án Domenzain, Benjamín Sánchez, Mihail An on, Edua d J Ke kho en, Aa ón Millán-O opeza, Céline Hen y, Ve ena Siewe s, John P Mo issey, Nikolaus Sonnenschein, and Jens Nielsen. Recons uc ion o a ca alogue o genome-scale me abolic models wi h enzyma ic cons ain s using gecko 2.0. Na u e Communica ions , 13(1):3766, 2022. Ma hew J Fagan and Mil on H Saie . P- ype a pases o euka yo es and bac e ia: sequence analyses and cons uc ion o phylogene ic ees. Jou nal o molecula e olu ion , 38:57–99, 1994. Yue Fu, Ho-Ching Ti any Tsui, Ke in E B uce, Lok-To Sham, Khadine A Higgins, John P Lishe , K ys yna M Kazmie czak, Michael J Ma oney, Cha les E Dann III, Malcolm E Winkle , e al. A new s uc u al pa adigm in coppe esis ance in s ep ococcus pneumoniae. Na u e chemical biology , 9(3):177, 2013. Ahmed Gaballa and John D Helmann. Bacillus sub ilis cpx- ype a pases: cha ac e iza ion o cd, zn, co and cu e lux sys ems. Biome als , 16:497–505, 2003. Sanjoy Ghosh, Chuanbin Dai, Ki s y B own, E hendha Rajendi an, Saman ha Maka enko, Jessica Bake , Caixia Ma, Swaga a Halde , Ma inie e Mon e o, Vlad-And ei Ionescu, e al. Colonic mic obio a al e s hos suscep ibili y o in ec ious coli is by modula ing in lamma ion, edox s a us, and ion anspo e gene exp ession. Ame ican Jou nal o Physiology-Gas oin es inal and Li e Physiology , 301(1):G39–G49, 2011. Manuel Glöckle , And eas D äge , and Reihaneh Mos olizadeh. Ncmw: a py hon package o analyze me abolic in e ac ions in he nasal mic obiome. F on ie s in Bioin o ma ics , 2:11, 2022. Valen ina Gogulancea, Rebeca González-Cabalei o, Bowen Li, Denis Taniguchi, Pahala Geda a Jaya hilake, Jinju Chen, Da en Wilkinson, Da id Swailes, And ew S ephen McGough, Paolo Zuliani, e al. Indi idual based model links he modynamics, chemical specia ion and en i onmen al condi ions o mic obial g ow h. F on ie s in mic obiology , 10:1871, 2019. Cy il Guilhen, Muhamed-Khei Taha, and F édé ic J Vey ie . Role o ansi ion me al expo e s in i ulence: he example o neisse ia meningi idis. F on ie s in Cellula and In ec ion Mic obiology , 3:102, 2013. 57 Appendix 8 De ails o esul s Table 11: Abundance o B. he aio aomic on unde S eadyCom in non-enhanced samples Cons ain s Pa ne Abundance 33 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 No cons ain s M_Bac e oides_uni o mis_ATCC_8492 1.000000 125 uM H; 2 mM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H; 125 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 2,7 uM H; 33 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 0,7 uM H; 8,2 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 0 uM Fe/H M_Bac e oides_uni o mis_ATCC_8492 0.000000 2,7 uM H M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H; 125 uM Fe M_Esche ichia_coli_ED1a 0.000000 0,7 uM H; 8,2 uM Fe M_Esche ichia_coli_ED1a 0.000000 2,7 uM H; 33 uM Fe M_Esche ichia_coli_ED1a 0.000000 2,7 uM H M_Esche ichia_coli_ED1a 0.000000 125 uM Fe M_Esche ichia_coli_ED1a 0.000000 125 uM H M_Esche ichia_coli_ED1a 0.000000 125 uM H; 2 mM Fe M_Esche ichia_coli_ED1a 0.000000 No cons ain s M_Esche ichia_coli_ED1a 0.000000 33 uM Fe M_Esche ichia_coli_ED1a 0.000000 0 uM Fe/H M_Esche ichia_coli_ED1a 0.000000 No cons ain s M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.991742 64 125 uM H; 2 mM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 2,7 uM H; 33 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 0,7 uM H; 8,2 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 0 uM Fe/H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 33 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM H; 125 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 2,7 uM H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 No cons ain s M_Rosebu ia_in es inalis_L1_82 0.017450 2,7 uM H M_Rosebu ia_in es inalis_L1_82 0.000000 33 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 0 uM Fe/H M_Rosebu ia_in es inalis_L1_82 0.000000 0,7 uM H; 8,2 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 2,7 uM H; 33 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H; 2 mM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H; 125 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 33 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 0 uM Fe/H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 No cons ain s M_S ep ococcus_pa asanguinis_ATCC_15912 0.576236 125 uM H; 125 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H; 2 mM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 2,7 uM H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 2,7 uM H; 33 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 0,7 uM H; 8,2 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H; 2 mM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 125 uM H M_S ep ococcus_sali a ius_DSM_20560 0.000000 65 125 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 2,7 uM H M_S ep ococcus_sali a ius_DSM_20560 0.000000 33 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 0,7 uM H; 8,2 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 0 uM Fe/H M_S ep ococcus_sali a ius_DSM_20560 0.000000 125 uM H; 125 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 2,7 uM H; 33 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 No cons ain s M_S ep ococcus_sali a ius_DSM_20560 0.022845 Table 12: Abundance o B. he aio aomic on unde S eadyCom in enzyme-enhanced samples Cons ain s Pa ne Abundance 33 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 No cons ain s M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H; 2 mM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H; 125 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 2,7 uM H; 33 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 0,7 uM H; 8,2 uM Fe M_Bac e oides_uni o mis_ATCC_8492 0.000000 0 uM Fe/H M_Bac e oides_uni o mis_ATCC_8492 0.000000 2,7 uM H M_Bac e oides_uni o mis_ATCC_8492 0.000000 125 uM H; 125 uM Fe M_Esche ichia_coli_ED1a 0.000000 0,7 uM H; 8,2 uM Fe M_Esche ichia_coli_ED1a 0.000231 2,7 uM H; 33 uM Fe M_Esche ichia_coli_ED1a 0.000893 2,7 uM H M_Esche ichia_coli_ED1a 0.000000 125 uM Fe M_Esche ichia_coli_ED1a 0.000000 125 uM H M_Esche ichia_coli_ED1a 0.000300 125 uM H; 2 mM Fe M_Esche ichia_coli_ED1a 0.000000 No cons ain s M_Esche ichia_coli_ED1a 0.000000 33 uM Fe M_Esche ichia_coli_ED1a 0.000000 66 0 uM Fe/H M_Esche ichia_coli_ED1a 0.000000 No cons ain s M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.314043 125 uM H; 2 mM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 2,7 uM H; 33 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 0,7 uM H; 8,2 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 0 uM Fe/H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 33 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 125 uM H; 125 uM Fe M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 2,7 uM H M_Fusobac e ium_nuclea um_subsp_nuclea um_ATCC... 0.000000 No cons ain s M_Rosebu ia_in es inalis_L1_82 0.000000 2,7 uM H M_Rosebu ia_in es inalis_L1_82 0.000000 33 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 0 uM Fe/H M_Rosebu ia_in es inalis_L1_82 0.000000 0,7 uM H; 8,2 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 2,7 uM H; 33 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H; 2 mM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 125 uM H; 125 uM Fe M_Rosebu ia_in es inalis_L1_82 0.000000 33 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 0 uM Fe/H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 No cons ain s M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H; 125 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H; 2 mM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 125 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 2,7 uM H M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 2,7 uM H; 33 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 0,7 uM H; 8,2 uM Fe M_S ep ococcus_pa asanguinis_ATCC_15912 0.000000 67 125 uM H; 2 mM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 125 uM H M_S ep ococcus_sali a ius_DSM_20560 0.000000 125 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 2,7 uM H M_S ep ococcus_sali a ius_DSM_20560 0.000000 33 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 0,7 uM H; 8,2 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 0 uM Fe/H M_S ep ococcus_sali a ius_DSM_20560 0.000000 125 uM H; 125 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 2,7 uM H; 33 uM Fe M_S ep ococcus_sali a ius_DSM_20560 0.000000 No cons ain s M_S ep ococcus_sali a ius_DSM_20560 0.000000