Selection and characterization of biofuel-producing environmental bacteria isolated from vegetable oil-rich wastes

Selec ion and Cha ac e iza ion o Bio uel-P oducing En i onmen al Bac e ia Isola ed om Vege able Oil-Rich Was es Almudena Escoba -Nin ˜o 1,2 , Ca los Luna 3 , Diego Luna 3 , Ana T. Ma cos 1 , Da id Ca ´no as 1 *, Enca nacio ´n Mellado 2 * 1Depa men o Gene ics, Facul y o Biology, Uni e si y o Se ille, Se ille, Spain, 2Depa men o Mic obiology and Pa asi ology, Facul y o Pha macy, Uni e si y o Se ille, Se ille, Spain, 3Depa men o O ganic Chemis y, Uni e si y o Co ´ doba, Co ´ doba, Spain Abs ac Fossil uels a e consumed so apidly ha i is expec ed ha he plane esou ces will be soon exhaus ed. The e o e, i is impe a i e o de elop al e na i e and inexpensi e new echnologies o p oduce sus ainable uels, o example biodiesel. In addi ion o hyd oly ic and es e i ica ion eac ions, lipases a e capable o pe o ming anses e i ica ion eac ions use ul o he p oduc ion o biodiesel. Howe e selec ion o he lipases capable o pe o ming anses e i ica ion eac ions is no easy and consequen ly e y ew biodiesel p oducing lipases a e cu en ly a ailable. In his wo k we i s isola ed 1,016 lipoly ic mic oo ganisms by a quali a i e pla e assay. In a second s ep, lipoly ic bac e ia we e analyzed using a colo ime ic assay o de ec he anses e i ica ion ac i i y. Thi y o he ini ial lipoly ic s ains we e selec ed o u he cha ac e iza ion. Phylogene ic analysis e ealed ha 23 o he bac e ial isola es we e G am nega i e and 7 we e G am posi i e, belonging o di e en clades. Bio uel p oduc ion was analyzed and quan i ied by gas ch oma og aphy and e ealed ha 5 o he isola es p oduced bio uel wi h yields highe han 80% a bench op scale. Chemical and iscosi y analysis o he p oduced bio uel e ealed ha i di e ed om biodiesel. This bac e ial-de i ed bio uel does no equi e any u he downs eam p ocessing and i can be used di ec ly in engines. The eeze-d ied bac e ial cul u e supe na an s could be used a leas i e imes o bio uel p oduc ion wi hou diminishing hei ac i i y. The e o e, hese 5 isola es ep esen excellen candida es o es ing bio uel p oduc ion a indus ial scale. Ci a ion: Escoba -Nin ˜o A, Luna C, Luna D, Ma cos AT, Ca ´no as D, e al. (2014) Selec ion and Cha ac e iza ion o Bio uel-P oducing En i onmen al Bac e ia Isola ed om Vege able Oil-Rich Was es. PLoS ONE 9(8): e104063. doi:10.1371/jou nal.pone.0104063 Edi o : Ligia O Ma ins, Uni e sidade No a de Lisboa, Po ugal Recei ed Ma ch 21, 2014; Accep ed July 6, 2014; Published Augus 6, 2014 Copy igh : ß2014 Escoba -Nin ˜o e al. This is an open-access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal au ho and sou ce a e c edi ed. Da a A ailabili y: The au ho s con i m ha all da a unde lying he indings a e ully a ailable wi hou es ic ion. The nucleo ide sequences epo ed in his wo k ha e been deposi ed in he GenBank da abase unde accession numbe s KC880159 o KC880188. Funding: This wo k was suppo ed by g an s om Jun a de Andalucı ´a (P08-RNM-03515, P11-CVI-7427 MO and P11-TEP-7723), as well as Spanish Minis y o Economy and Compe i i eness (P ojec ENE 2011-27017) and FEDER unds. The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish, o p epa a ion o he manusc ip . Compe ing In e es s: The au ho s ha e decla ed ha no compe ing in e es s exis . * Email: [email p o ec ed] (DC); [email p o ec ed] (EM) In oduc ion Wi h a g owing wo ld popula ion, ossil uels a e cu en ly consumed oo apidly. Thus, i is expec ed ha we will deple e hese non- enewable esou ces om he plane in a ela i ely sho pe iod o ime. The inc easing demand o ossil uels has addi ional consequences, such as he concomi an inc eased p ices o c ude oil, he en i onmen al conce ns abou he pollu ion due o c ude oil de i a i es and he global g eenhouse e ec s, which al oge he a e igge ing he explo a ion o no el al e na i e sou ces o uels. The e o e, i is impe a i e o de elop al e na i e, sus ainable and inexpensi e uels. Biodiesel can be such a uel wi h he app op ia e echnologies. T iglyce ides (oils and a s) can no be di ec ly used in he a ailable diesel engines due o he high iscosi y and he acidic composi ion o hese lipids, he o ma ion o ee a y acids esul ing in gum o ma ion by oxida ion and polyme iza ion, he ca bon deposi ion and he hickening o he lub ican . Thus, he use o ege able oils as al e na i e sou ces o uels equi es hei p ocessing o each a iscosi y and a ola ili y simila o c ude oil de i ed uels, a ac ha will allow hem o be di ec ly used in he cu en con igu a ion o diesel engines. These easons a e d i ing he de elopmen o new ege able oil de i a i es displaying p ope ies simila o hose o diesel uels. Nowadays he mos p omising and accep ed ege able oil de i a i e is biodiesel [1,2]. Biodiesel is a enewable sou ce o ene gy conside ed o be ca bon neu al because he ca bon deli e ed du ing i s combus ion was ixed only a ew yea s be o e om he a mosphe e. In addi ion, i s use in in e nal combus ion engine does no p oduce sul u oxide and minimizes h ee old he o ma ion o soo pa icula es in compa ison wi h pe ochemical diesel [3]. Biodiesel is a mix u e o mono-alkyl es e s ha can be ob ained om ege able oils, animal a s, was e cooking oils, g eases and algae. Th ee p ocessing echniques a e mainly used o ca alyze he con e sion o ege able oils in o eady- o-use bio uels: py olysis, mic oemulsi ica ion and anses e i ica ion. The mos popula me hod is he anses e i ica ion (e hanolysis o me hanolysis) o ege able oils o p oduce biodiesel. The anses e i ica ion p ocess can be pe o med using an alkali ca alys , an acid ca alys , a PLOS ONE | www.plosone.o g 1 Augus 2014 | Volume 9 | Issue 8 | e104063 bioca alys o he e ogeneous ca alys s, howe e only he alkaline p ocess is cu en ly ca ied ou a indus ial scale, because i is cos e ec i e and highly e icien [4]. In he alkaline p ocess sodium hyd oxide (NaOH) o po assium hyd oxide (KOH) is used as a ca alys along wi h me hanol o e hanol. NaOH o KOH eac wi h he alcohol o gi e an alcoxy g oup. Then he alcoxy moie y eac s wi h any iglyce ide o o m he co esponding me hyl- o e hyl-es e s and glyce ol. A e he syn hesis, he p oduc ion o biodiesel equi es a se ies o downs eam p ocesses ha include epea ed washings o each su icien pu i y. These downs eam ope a ions a e usually hampe ed wi h p oblems, such as he sepa a ion o ca alys and un eac ed me hanol om biodiesel, he isk o ee a y acid o wa e con amina ion and soap o ma ion du ing he anses e i ica ion eac ion [1,2]. The p oduc ion o biodiesel using a bioca alys elimina es he disad an ages o he chemical alkaline p ocess by ob aining a p oduc o e y high pu i y wi h less o no downs eam ope a ions. In pa icula , o a oid he need o sepa a e glyce ol om biodiesel, an al e na i e s a egy based on inco po a ing some o he glyce ol de i a i es in he inal p oduc has been conside ed. In his way, a bio uel o med by a mix o glyce ol de i a i es and a y acid me hyl es e s (FAME) o a y acid e hyl es e s (FAEE) no only p e en s he need o sepa a e he glyce ol om he biodiesel, bu also inc eases he yield o he p ocess as he o al numbe o ca bons in ol ed in he eac ion is main ained in he inal p oduc [5,6]. Some examples o me hodologies conside ing his op ion a e based on he anses e i ica ion eac ion o iglyce ides (TG) wi h dime hyl ca bona e [7], me hyl ace a e [8] o e hyl ace a e [9] esul ing in a mix u e o h ee molecules o FAME o FAEE and one o glyce ol ca bona e o glyce ol iace a e. Ano he example is Ecodiesel-100, a pa en ed bio uel ob ained using 1,3-selec i e pa ial e hanolysis o iglyce ides wi h a po cine panc ea ic lipase. Ecodiesel-100 is a mix u e o FAEE and monoglyce ides (MG) [5,6]. Despi e o he ad an ages o Ecodiesel-100, he main d awback is he high cos o he pu i ied enzyme. The e o e, he e is a need o sol e his issue be o e he p ocess can be implemen ed a he indus ial scale. The enzyma ic p oduc ion o bio uel is possible using ex acellula o in acellula lipases. Lipases ( iacylglyce ol acylhyd olases EC 3.1.1.3) a e e y e sa ile bioca alys s: hey a e s able in o ganic sol en s; hey do no equi e any co ac o ; hey ha e b oad subs a e speci ici y; and hey show high enan ioselec i i y. The numbe o cloned lipases has inc eased since he 1980s, in pa because o his e sa ili y, bu also because o he inc eased numbe o cha ac e ized lipoly ic mic oo ganisms. Se e al enzyma ic ac i i ies ha e been desc ibed o lipases, such as hyd olysis, anses e i ica ion and es e i ica ion. Howe e , he hyd oly ic ac i i y o lipases has been s udied in de ail in con as o he anses e i ica ion ac i i y. I is hypo h- esized ha he anses e i ica ion ac i i y o lipases is simila o he se ine-p o ease ca aly ic ac i i y, because bo h enzymes sha e he same ca aly ic iad Se –His–Asp/Glu [10]. Ex ensi e and pe sis en sc eening o new mic oo ganisms and hei lipoly ic enzymes can open no el al e na i es o syn he ic p ocesses and consequen ly, no el possibili ies o con ibu e o sol e en i onmen al p oblems. The aim o his s udy is o ind and selec lipoly ic mic oo ganisms showing anses e i ica ion ac i i y, which can be used o p oduce a cheap enzyma ic ex ac , a oiding labo ious downs eam p ocessing. Employing his enzyma ic ex ac able o p oduce bio uel could educe he cos o enzyme p oduc ion. The e o e he selec ed mic oo ganisms ha e he po en ial o be used in whi e bio echnology o he indus ial p oduc ion o a ye - o-be comme cialized bio uel. Ma e ials and Me hods Si e desc ip ion and sample collec ion Fou samples we e collec ed om di e en loca ions in an oil mill in Ecija (Se illa, Spain) a he end o he ha es ing season (Janua y 2011). Pool 1 (AE1B) and pool 2 (AE2B) we e sampled om ponds con aining liquid was es om he oil mill. Alpechı ´n (AEA) was ob ained om he solid was es o he oil mill and he las sample (AEDH) was ob ained om he was es o he oli e ha es s con aining mainly oli e ee le o e s (i.e. dehyd a ed lea es and b anches). Samples we e aken in s e ile plas ic 50 ml ubes and s o ed in he da k a 4uC un il hey we e p ocessed. The pH, elec ic conduc i i y (C.E.) and humidi y o soil we e measu ed as p e iously epo ed [11]. The ex ac ion o bioa ailable Ca, Mg, Na, K, Fe, Cu, Mn and Zn was pe o med as p e iously epo ed [12,13]. De e mina ion o me als was pe o med using a omic abso p ion spec ome y in a spec opho ome e UNICAM (The mo). De e mina ion o o al C, N and S was pe o med in a LECO’s CNS-2000 elemen al au oanalyze . Ex ac ion and de e mina ion o bioa ailable P was pe o med as p e iously epo ed [14,15]. De e mina ion o o ganic ma e was done by calcina ion (UNE-EN 13039 s anda d) and he de e mina ion o o al ca bona es ollowing he calcime y p o ocol o Be na d [11]. Isola ion o mic oo ganisms showing lipoly ic ac i i y (hyd olysis) One g o soil sample was esuspended in 3 ml o s e ile saline solu ion (ClNa 0.85% w/ ). This suspension was dilu ed wi h saline solu ion (ClNa 0.85% w/ ) o ob ain single colonies and pla ed on a ba e y o solid media con aining 0.5% ibu y ine. The media used we e LB (Lu ia Be ani medium), Po a o Dex ose B o h (PDB) (Di co), 2% Aga (w/ ) in dis illed wa e , 9K A (1 L 9K solu ion, 1% (w/ ) glucose and 1% (w/ ) yeas ex ac ), 9K G (1 L 9K solu ion, 0.5% (w/ ) glucose, 0.5% (w/ ) yeas ex ac , and 1% (w/ ) mal ex ac ) o 9K Gamp (1 L 9K solu ion, 1% (w/ ) glucose, 0.5% (w/ ) yeas ex ac , 1% (w/ ) mal ex ac and 20 mg/ml ampicillin). The 9K solu ion was p epa ed as desc ibed in Sil e man [16] wi hou FeSO 4 7H 2 0. To ob ain solid media 2% aga was added o he media. The pla es we e incuba ed a 30uC o 4–7 days. Those mic oo ganisms displaying a clea ing zone a ound he colony we e g own in pla es con aining non-selec ing LB (bac e ia), PDB ( ungi) o YPD (yeas ) media, and hen he lipoly ic ac i i y was con i med by g owing he isola es again on he co esponding media con aining 0.5% ibu y ine. Sc eening o lipoly ic mic oo ganisms showing anses e i ica ion ac i i y Fo sc eening mic oo ganisms displaying anses e i ica ion ac i i y we de eloped a simple colo ime ic me hod based on a p e iously epo ed one [17] wi h some modi ica ions. The me hod consis s on he anses e i ica ion o pa a-ni ophenyl palmi a e (p-NPP) wi h e hanol in he absence o wa e o elease he yellow colo ed compound pa a-ni ophenol (p-NP), which can be subsequen ly de ec ed by using a spec opho ome e . A wa e - ee en i onmen p omo es he anses e i ica ion eac ion o e he wa e -dependen hyd olisis eac ion o p-NPP. A 5 ml o e nigh cul u e o each posi i e lipoly ic bac e ial isola e was employed o inocula e lasks con aining 25 ml o LB media plus 2% ibu y ine a a inal OD 600 o 0.1. These cul u es we e g own o 3 days a 30uC and 200 pm. A e his ime cells we e pelle ed by cen i uga ion and 1.8 ml o he supe na an s we e eeze-d ied o 24 hou s. The eeze-d ied supe na an was mixed wi h 1 ml o 10 mM p-NPP (in n-hexane) and 60 ml o absolu e e hanol. The Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 2 Augus 2014 | Volume 9 | Issue 8 | e104063 mix u e was made in 2 ml sa e-lock eppendo ubes and incuba ed wi h shaking a 200 pm in a o a o shake a 37uC o 16 hou s. A nega i e con ol was p epa ed by using a mix u e o absolu e e hanol and p-NPP. As a con ol o he hyd oly ic ac i i y o lipases we used a mix u e o p-NPP and eeze-d ied supe na an (wi hou e hanol). A e 16 hou s o eac ion he lipase was allowed o decan a he bo om o he ube o 10 minu es, hen 25 ml o supe na an was mixed wi h 1 ml o 0.05 M NaOH in a 1.5 ml eppendo ube. The p-NP p oduced du ing he anses e i ica ion eac ion was ex ac ed by he aqueous alkaline phase, ans e ed o a 1 ml cu e e. and he abso bance was quan i ied a 410 nm using a Beckman DU 640 spec opho om- e e . P oduc ion o biodiesel a bench op scale The p oduc ion o bio uel was analyzed by gas ch oma og aphy as p e iously epo ed [18]. A 5 ml o e nigh cul u e o each posi i e lipoly ic bac e ial isola e was employed o inocula e a lask con aining 25 ml o LB media plus 2% ibu y ine a a inal OD 600 o 0.1. Cul u es we e g own o 3 days a 30uC and 200 pm. The whole olume o he supe na an s was eeze-d ied and all he ex ac was used di ec ly in a anses e i ica ion eac ion (non- p ocessed supe na an ). To elimina e sal s and o he componen s o he media and ibu y ine om he supe na an s hey we e concen a ed in dialysis bags (12 KDa, Sigma) using polye hylene glycol (a e age Mw 8,000, Sigma) a 4uC o e nigh , hen dialyzed in 0.05 M po assium phospha e bu e (pH 7.6) ou imes. The esul ing concen a ed and dialyzed supe na an was inally eeze- d ied and used in a anses e i ica ion eac ion (p ocessed supe na an ). The anses e i ica ion eac ion was pe o med wi h con inous shaking a 37uC o 24 hou s and con ained 6 ml sun lowe oil, 1.75 ml absolu e e hanol, 0.05 ml NaOH 10N and he en i e p ocessed o non-p ocessed supe na an coming om 25 ml o cul u e. A nega i e con ol was p epa ed by using a mix u e o 6 ml sun lowe oil, 1.75 ml absolu e e hanol and 0.05 ml NaOH 10N. The e hyl es e s and glyce ides p oduced in he anses e - i ica ion eac ion we e analyzed using gas ch oma og aphy (GC) as desc ibed below. To s udy he kine ics o he eac ion, samples we e aken a he indica ed ime poin s, and analyzed by GC. To quan i y he bio uel p oduc ion a e he e-u iliza ion o he bac e ial ex ac s, he eac ion mix was cen i uged, he supe na an disca ded and he pelle was employed o epea he anses e i ica ion eac ion and he GC analysis. Bo h analyses we e pe o med using only p ocessed supe na an . Chemical analysis o he bio uel The me hod used in eg a es wo o icial me hods o he de ec ion o es e s (UNE EN ISO 14103) and glyce ides (UNE EN ISO 14105), using ce ane (n-hexadecane) as an in e nal s anda d o quan i y he con en s o glyce ol, e hyl es e s and glyce ides (mono-, di- and iglyce ides) as p e iously desc ibed [5]. B ie ly, a gas ch oma og aph Va ian 430 GG i ed wi h a capilla y column HT5, 0.1 mm (25 m60.32 mm, SGE, Supelco) wi h a lame ioniza ion de ec o (FID) and spli less injec ion was used. The anses e i ica ion eac ion p oduc (12.5 ml) was mixed wi h 4 ml o a 1:1 ( / ) e hanol/dichlo ome hane mix u e ha con ained he in e nal s anda d (ce ane), and 0.5 ml o he p epa ed sample was employed o he analysis. The esul s we e exp essed as ela i e quan i ies o he co esponding a y acid e hyl es e s (FAEE) and some monoglyc- e ides (MG) o lowe e en ion ime (RT,25 minu es) and he sum o he quan i ies o he o he MG (RT.25 minu es) and he diglyce ides (DG). The yield e e s o he ela i e amoun (%) o FAEE +MG (wi h lowe RT) p oduced. The con e sion includes he o al amoun (%) o iglyce ide ans o med in o FAEE, MG and DG. Blank eac ions con aining only he mix u e o oil, e hanol and NaOH we e pe o med. Con e sion o he s a ing oil ma e ial was below 15% unde hese expe imen al condi ions. Da a shown a e he a e age o a leas wo independen expe imen s. De e mina ion o he iscosi y The iscosi y was de e mined in a capilla y iscome e Oswald P o on Cannon-Fenske Rou ine Viscome e 33200, size 2150. The me hod used was based on de e mining he ime needed o a gi en olume o luid o pass be ween wo poin s ma ked in he ins umen . I co ela es wi h he speed educ ion su e ed by he liquid low as a esul o in e nal ic ion o i s molecules, depending on hei iscosi y. F om he low ime, , in seconds, he kinema ic iscosi y (u, cen is okes, cS ) can be ob ained om he equa ion: u6 =C, whe e Cis he cons an calib a ion o he measu ing sys em in cS ?s, which is gi en by he manu ac u e (0.10698 mm 2 s 2 1, a 40uC) and he low ime in seconds. The kinema ic iscosi y also ep esen s he a io be ween he dynamic iscosi y and he densi y ( ,u=g/ ). Analysis o hyd oly ic ac i i ies (amylase, p o ease, DNAse, pullulanase and xylanase) Amylase ac i i y was sc eened on LB solid medium supple- men ed wi h 0.2% (w/ ) soluble s a ch, 0.5% (w/ ) pep one and 0.3% (w/ ) mea ex ac . A e 7 days o incuba ion a 30uC he pla es we e looded wi h 0.3% (w/ ) I 2 –0.6% ( / ) KI solu ion. Hyd olysis o s a ch esul s in a clea ing zone a ound he colonies [19]. The p esence o p o ease ac i i y was de e mined in LB solid medium supplemen ed wi h 2% (w/ ) skim milk. The appea ance o zones o p ecipi a ion o pa acasein a ound he colonies a e 3 days o incuba ion a 30uC indica ed he p esence o p o eoly ic ac i i y [19]. DNAse ac i i y was analyzed by g owing bac e ia on DNAse es aga pla es con aining 4.2% (w/ ) Aga DNA. The pla es we e incuba ed o 7 days a 30uC and hen looded wi h 1N HCl solu ion. A clea ing zone a ound he colonies indica ed DNase ac i i y [20]. The pullulanase and xylanase ac i i ies we e de ec ed by sc eening o zones o blue halos p oduced due o he hyd olysis and solubiliza ion o he AZCL-pullulano and AZCL-xylano, espec i ely [21]. Isola ion o DNA and 16S RNA gene sequence analysis The bac e ial DNAs we e isola ed ollowing s anda d p o ocols [22]. The o al DNA isola ed was used as he empla e o he ampli ica ion o he 16S RNA by PCR using he uni e sal p ime s designed o Bac e ia 16F27 (59-AGAGTTT- GATCMTGGCTCAG-39) and 16R1488 (59- CGGTTACCTTGTTAGGACTTCACC-39) [23]. The p og am used o he ampli ica ion was: one cycle o 95uC o 5 minu es; 25 cycles o 94uC o 1 minu e, 50uC o 1 minu e and 72uC o 2 minu es; and a inal ex ension cycle o 10 minu es a 72uC. Pa ial 16S RNA gene sequences (c. 650 bp co esponding o posi ions 39 o 689 o he 16S RNA gene om Esche ichia coli) we e ob ained and aligned o he mos simila 16S RNA gene sequence in he GenBank da abase using he BLASTn algo i hm. Using he mos simila sequences ound in he GenBank da abase a mul iple sequence alignmen o he DNA sequences was Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 3 Augus 2014 | Volume 9 | Issue 8 | e104063 cons uc ed using he Clus alW so wa e [24]. The phylogene ic ees we e ob ained using he MEGA4 so wa e [25] wi h he neighbo -joining me hod [26]. The da a se was boo s apped 500 imes o ensu e eliabili y o each b anch. The e olu iona y dis ances we e compu ed using he maximum composi e likelihood me hod [27] and a e shown in numbe o base subs i u ions pe si e. Nucleo ide sequence accession numbe s The nucleo ide sequences epo ed in his wo k ha e been deposi ed in he GenBank da abase unde accession numbe s KC880159 o KC880188. Resul s and Discussion Chemical composi ion o he collec ed samples One o he aims o his wo k was o ind mic oo ganisms capable o p oducing bio uel om ege able oil. The e o e, loca ions ich in ege able oil we e selec ed o sc een o mic oo ganisms ha bo ing lipase ac i i y agains ege able oil. In pa icula , hese samples we e collec ed om an oli e oil mill loca ed in Ecija, Sou he n Spain. The chemical and physico-chemical cha ac e is ics o he collec ed samples a e shown in Table 1. The analysis o he samples e ealed ha he mos abundan me al elemen in all samples was Fe, al hough in he sample AEDH he e we e only mino di e ences be ween he Fe, Mn, Cu and Zn con en . The es o elemen s analyzed we e C.Ca.N.K.S.Mg.Na.P ( om mos o less abundan ). The excep ions we e sample AE2B whe e S was no de ec ed, sample AEA whe e P was mo e abundan han Na, and AEDH whe e Mg was mo e abundan han S. The pH o all samples was a ound 6.5, excep in AEA (alpechı ´n), which showed pH 4.7. In mos o he samples he o ganic ma e ial accoun ed o 30–40% excep o he AEA, which was he iches one wi h a 95.8%. Table 1. Chemical and physico-chemical analysis o he oil mill samples. Pa ame e Samples a AE1B AE2B AEA AEDH C o al (%) 16.24 22.490 66.410 22.970 N o al (%) 0.894 0.616 1.169 0.819 S To al (%) 0.05594 N.D 0.100 0.052 K(%) 0.144 0.110 0.456 0.269 Na (mg/Kg) 61.234 130.597 168.019 235.668 Mg (%) 0.018 0.030 0.022 0.059 Ca (%) 1.602 1.500 0.074 1.861 P (mg/Kg) 36.755 43.541 280.390 168.848 Mn (mg/Kg) 63.528 95.033 12.015 125.528 Fe (mg/Kg) 599.663 651.608 139.812 151.260 Cu (mg/Kg) 73.277 49.632 27.195 140.325 Zn (mg/Kg) 19.163 14.253 29.750 11.955 pH (1/5) 6.837 6.55 4.68 6.27 E.C. 1/5 (mS/cm) b 1.68 2.62 3.89 2.14 Humidi y (%) 5.460 4.690 5.819 5.179 O.M. (%) c 30.709 40.186 95.801 40.782 CaCO 3 o al (%) 9.499 8.808 1.262 6.314 a All he esul s a e e e ed o he d ied mass o he samples. b E.C.: Elec ic Conduc i i y. c O.M.: O ganic Ma e . doi:10.1371/jou nal.pone.0104063. 001 Table 2. Dis ibu ion o ypes o lipoly ic mic oo ganisms isola ed om he oil mill samples. Sample Bac e ia Yeas Fungi To al AE1B 61 21 174 256 AE2B 166 208 10 392 AEDH 57 190 1 248 AEA 7 70 51 128 To al 291 489 236 1016 doi:10.1371/jou nal.pone.0104063. 002 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 4 Augus 2014 | Volume 9 | Issue 8 | e104063 Fi s s ep o he sc eening: selec ion o mic oo ganisms showing lipoly ic ac i i y (hyd olysis) As a i s s ep o he sc eening p ocess, we assayed o he hyd oly ic ac i i y o he mic oo ganisms. This is an easy, quick and cheap es o pe o m on solid media based on he isual inspec ion o pla es con aining he lipid subs a e ibu y ine o mic oo ganisms showing a clea ing zone a ound he colony edges. To ob ain single colonies and selec lipoly ic mic oo ganisms, samples we e dilu ed in s e ile saline solu ion and pla ed on LB, PDB, wa e -aga , K9 A, K9 G o K9 Gamp solid media supplemen ed wi h 0.5% ibu y ine. Du ing he i s ound o selec ion, 1,016 colonies ( ungi, yeas s and bac e ia) we e iden i ied o p oduce a clea ing zone o hyd olysis (Table 2). Mos mic oo ganisms we e ob ained om media K9 A, K9 G and K9 Gamp (68% ungi, 65% bac e ia, 65% yeas s). We decided o s udy only he lipoly ic bac e ia and hus, we selec ed 291 bac e ia o u he s udies. These s ains we e i s inocula ed on non- selec i e media and e-checked o hyd oly ic ac i i y on LB supplemen ed wi h 0.5% ibu y ine. Sample AEA yielded he poo es numbe o lipoly ic bac e ia. The AEA sample was ob ained om alpechı ´n, which is he oxic was ewa e esul ing a e p essing he oli es du ing he p oduc ion o oil. The analysis o his sample showed ha i was acid and ich in o ganic ma e . This esul is in ag eemen wi h p e ious epo s desc ibing ha alpechı ´nis sligh ly acid and ich in soluble o ganic compounds [28,29]. Al hough AEA was he sample con aining he highes o ganic ma e ial con en , usually lignin accoun s o ca. 50% o he o al o ganic ma e ial in his ype o samples. Lignin is a ecalci an polyme ha may also a ec mic obial ac i i y and su i al nega i ely. In ac , he phy o oxic and an imic obial e ec s o phenols, o ganic acids and a y acids ha a e usually p esen in his was ewa e alpechı ´nha e been p e iously epo ed [30]. This may explain he low numbe o lipoly ic mic oo gan- isms ound in his sample. Second s ep o he sc eening: Analysis o he anses e i ica ion ac i i y o lipoly ic mic oo ganisms We ha e employed a me hod based on he anses e i ica ion o p-ni ophenolpalmi a e (p-NPP) wi h e hanol (o o he alcohol) o elease he yellow-colo ed p-ni ophenol (p-NP) as desc ibed in ma e ial and me hods. In o de o selec only ex acellula enzymes, his second s ep o he sc eening was pe o med using eeze-d ied supe na an o he bac e ia selec ed in he i s s ep o he sc eening. Du ing he sc eening p ocedu e, a di e se se o con ol assays was ca ied ou in he absence o e hanol o lipoly ic bac e ia. The p-NP eleased unde hese con ol condi ions was conside ed o esul om he hyd olysis o a non-enzyma ic eac ion o p-NPP and i was employed as he h eshold o de e mine he exis ence o anses e i ica ion ac i i y. The maximum abso bance alue Table 3. Classi ica ion and selec ion o bac e ia showing anses e i ica ion ac i i y acco ding o he pa a-ni ophenylpalmi a e es . Abso bance 410 nm Numbe o s ains Posi i e/Nega i e .1 30 Posi i e 1–0.8 28 Posi i e ,0.8 233 Nega i e doi:10.1371/jou nal.pone.0104063. 003 Figu e 1. Ch oma og am ob ained by GC o he eac ion mix a e he anses e i ica ion o sun lowe oil and e hanol pe o med by he p ocessed supe na an o s ain AE2B 122. Ce ane was used as an in e nal s anda d. FAEE: Fa y Acid E hyl Es e ; MG: MonoGlyce ide; DG: DiGlyce ide; TG: T iGlyce ide. The double-lined a ow indica es he e en ion ime expec ed o glyce ol as p e iously epo ed [5,6]. doi:10.1371/jou nal.pone.0104063.g001 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 5 Augus 2014 | Volume 9 | Issue 8 | e104063 ob ained in he con ol eac ions was 0.8. The anses e i ica ion eac ion wi h all he s ains selec ed du ing he i s s ep o he sc eening was pe o med in pa allel. All s ains showing abso - bance alues o e his h eshold o 0.8 we e conside ed posi i e and alues below he h eshold we e conside ed backg ound le els (Table 3). Acco ding o he abo e c i e ion, 58 s ains we e conside ed posi i e. Ou o he 58 posi i e s ains, he op 30 bac e ia, i.e. hose showing abso bance alues o e 1.0, we e selec ed o u he analysis and quan i ica ion o bio uel p oduc- ion. Du ing his s ep we selec ed only o bac e ia showing ex acellula anses e i ica ion ac i i y because i is desc ibed ha a indus ial le el he highes yields o biodiesel p oduc ion a e ob ained wi h ex acellula lipases a he han wi h whole cells [1,3]. In ac , whole cell anses e i ica ion p ocedu es p oduce low subs a e con e sion yields due o he oxici y o he sol en o he hos cells, and also due o he low mass ans e a e o high molecula weigh subs a es (oil) om he sol en phase o he whole cell bioca alys [31]. Ano he no iceable di e ence when he anses e i ica ion eac ion is pe o med wi h whole cells is he eac ion ime. No ozym 435 ga e yields o 87% a e 3.5 h o eac ion in a ed-ba ch ope a ion. Howe e , o ob ain yields o 80– 90% o biodiesel using whole cells, i was equi ed o pe o m he eac ion p ocess o 70 h [32]. Thi d s ep o he sc eening: Selec ion o he bac e ial supe na an s p oducing bio uel a bench op scale The aim o he hi d s ep o he sc eening was o iden i y he bes pe o ming bac e ia in bio uel p oduc ion a bench op scale. To op imize he sc eening p o ocol, we i s g ew he 30 selec ed bac e ial s ains as desc ibed in Ma e ials and Me hods. In a i s a emp he supe na an s (non-p ocessed) we e eeze-d ied o ob ain a ine d ied powde ha can be used di ec ly o he p oduc ion o bio uel. The anses e i ica ion eac ion o p oduce bio uel was pe o med a basic pH wi h sun lowe oil, e hanol and he powde ed supe na an . A e comple ion o he eac ion, he bio uel/lipid/ a y acid mix u e was analyzed by GC. The ch oma og ams show clea di e en e en ion imes o he inal (FAEE +MG), and he ini ial (TG) and in e media e (DG) p oduc s (Fig. 1). The yield ob ained wi h he bac e ia ini ially es ed was lowe han 22% in all cases (Fig. 2). In o de o y o inc ease he yield, he supe na an s o he cul u es we e p ocessed as desc ibed in ma e ials and me hods, and he samples we e hen eeze-d ied and assayed o bio uel p oduc ion. We obse ed ha p ocessing o he bac e ial supe na an s d as ically inc eased he pe o mance up o 17.8 old in compa ison o he non-p ocessed samples (Fig. 2). The e o e, we ound ou ha an easy and cheap p ocess o concen a ion and dialysis could inc ease he yield o bio uel p oduc ion. Once he p o ocol was op imized, we p oceeded o quan i y bio uel p oduc ion in he 30 selec ed bac e ial s ains. The comple e esul s o he eac ions using p ocessed supe na an s a e shown in Table 4. Since he compo- si ion o biodiesel is a mix o FAEEs and he composi ion o bio uel Figu e 2. Compa ison o bio uel p oduc ion using p ocessed o non-p ocessed supe na an s. The bac e ial supe na an s we e p ocessed as desc ibed in Ma e ials and Me hods o no p ocessed. Bo h p ocessed and non-p ocessed supe na an s we e eeze-d ied and employed o anses e i ica ion eac ion wi h sun lowe oil and e hanol. The da a show he pe cen age o FAE (FAEE+MG) p oduced a e 24 hou s o eac ion using 7 andomly selec ed bac e ial isola es. Da a a e he a e age o a leas 2 independen expe imen s. doi:10.1371/jou nal.pone.0104063.g002 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 6 Augus 2014 | Volume 9 | Issue 8 | e104063 is a mix o FAEs, he mos impo an alue o be conside ed is he pe cen age o FAE, which a e a mix o wo moles o FAEE and one mol o MG. The analysis o he da a e ealed ha he supe na an s o 5 o he bac e ial s ains we e capable o p oducing o e 80% FAE, 9 bac e ial supe na an s p oduced be ween 60– 80% FAE, and he p oduc ion o he o he 16 isola es was below 60% FAE. Eigh bac e ial s ains we e capable o p ocessing o e 80% o he TG in he eac ion mix u e (con e sion o TG in o DG plus FAE). Du ing he sc eening o he bes bio uel-p oducing s ains, we no iced ha glyce ol did no appea a he expec ed e en ion ime (5 min). This sugges s ha TGs we e con e ed in o e hyl es e s (FAEEs) and MG, which co-elu e unde ou expe imen al condi ions (Fig. 1). This was p e iously obse ed in anses e i i- ca ion eac ions pe o med using po cine panc ea ic lipase [6]. F om an indus ial pe spec i e, he bio uel p oduced by hese bac e ia would no equi e downs eam p ocessing o sepa a e he glyce ol con aminan because glyce ol was in eg a ed in he inal mix as MG. This could be due o a 1,3 selec i e e hanolysis o he sun lowe oil by he enzyme, ha means a pa ial e hanolysis o he oil ha p oduce 1 mol o MG o each 2 moles o FAEE [6]. The chemical analysis also sugges s ha his me hodology is no sui able o p oduce con en ional biodiesel acco ding o EN 14214, bu a he he new ype o bio uel, which inco po a es glyce ol as p e iously desc ibed [5,6]. Phylogene ic analyses o he selec ed bio uel p oducing bac e ia In o de o assign he bac e ial bio uel p oduce s o speci ic phylogene ic g oups, he sequence o he 16S RNA gene was de e mined. Phylogene ic econs uc ion pe o med wi h di e en me hods was consis en , and consequen ly only he ee ob ained wi h Neighbo -Joining o he e olu iona y his o y and Maximum Composi e Likelihood o e olu iona y dis ances is shown (Fig. 3). Table 4. T anses e i ica ion eac ion o sun lowe oil by he bac e ial ex ac s de e mined by GC analysis. Sample FAE % a DG % TG % Con e sion % Sun lowe oil 3.4 29.9 66.7 33.3 Nega i e Con ol 13.8 41.9 44.3 55.7 AE1B 20 63.1 6.7 30.2 69.8 AE1B 21 93.7 6.3 0.0 100 AE1B 22 65.9 2.5 31.7 68.4 AE1B 26 80.1 12.9 7.1 92.9 AE1B 27 57.9 4.3 37.7 62.3 AE1B 28 58.7 3.1 38.2 61.8 AE1B 35 67.4 2.5 30.2 69.8 AE1B 89 72.2 11.6 16.2 83.8 AE1B 90 58.6 2.7 38.7 61.3 AE1B 92 80.1 7.5 12.4 87.6 AE2B 29 76.0 7.8 16.2 83.8 AE2B 30 66.8 5.1 28.1 71.9 AE2B 85 41.2 13.8 45.1 54.9 AE2B 120 35.1 15.1 50.1 50.2 AE2B 122 91.16 1.75 6.8 93.2 AE2B 130 71.8 8.1 20.1 79.9 AE2B 131 61,27 9,55 14,59 85,41 AE2B 133 34.8 19.5 45.7 54.3 AE2B 134 31.7 2.1 66.3 33.7 AE2B 199 25.0 3.3 71.7 28.3 AE2B 222 77.5 3.2 12.9 87.1 AE2B 232 48.1 0.7 51.2 48.8 AE2B 250 46.4 10.7 42.9 57.1 AE2B 259 30.7 8.3 61.0 39.0 AE2B 261 82.6 7.8 9.6 90.4 AE2B 263 49.3 10.5 40.2 59.8 AE2B 264 24.1 2.6 73.3 26.7 AE2B 332 54.0 0.7 45.4 54.7 AE2B 340 51.4 1.0 47.6 52.4 AEDH 145 61.0 4.0 35.0 65.0 The pe cen age o FAE shows he yield o bio uel p oduc ion in he eac ion mix. Con e sion shows he pe cen age o TG me abolized du ing he eac ion. a FAE, Fa y Acid Es e s ( a y acid e hyl es e s +monoglyce ides) (Bio uel). DG, Diglyce ides. TG, T iglyce ides. Con e sion, pe cen age o TG con e ed in o FAE+DG. doi:10.1371/jou nal.pone.0104063. 004 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 7 Augus 2014 | Volume 9 | Issue 8 | e104063 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 8 Augus 2014 | Volume 9 | Issue 8 | e104063 The s ains isola ed we e e y di e se wi h ep esen a i es o bo h G am posi i e and G am nega i e bac e ia. This analysis oge he wi h he high numbe o euka yo ic mic oo ganisms (yeas s and ungi) also isola ed du ing his wo k sugges s ha he e is a high di e si y o li ing o ganisms in he samples o he oli e-mill was es. Table 5 shows he closes ela i e o each o he isola es acco ding o he esul s ob ained a e he phylogene ic analysis. Mos o he lipoly ic isola es we e ob ained om he sample AE2B (19 isola es): 16 o hem we e ela ed o he genus Pseudomonas, 2 o hem ela ed o he genus Te ibacillus, and one isola e ela ed o he genus Bacillus. Ele en lipoly ic isola es we e ob ained om he sample AE1B, which we e ela ed o he gene a En e obac e (3 isola es), Ae omonas (3 isola es), Bacillus (2 isola es) and Te ibacillus (one isola e). Only one isola e was ob ained om he sample AEDH. I was ela ed o he genus Acine obac e . In e es ingly, no lipoly ic isola es we e selec ed om he sample AEA (Table 5). The isola es ha showed he highes yield in he p oduc ion o bio uel (selec i i y o e 60%) we e 14 s ains belonging o he gene a Pseudomonas,Acine obac e ,En e obac e ,Bacillus, and Te ibacillus (Table 4 and 5). Thei closes ela i es we e Pseudomonas e onii [33] (AE2B 30, AE2B 29, AE2B 222), Pseudomonas ex emaus alis [34] (AE2B 130, AE2B 261), Acine obac e lwo ii [35] (AEDH 145), En e obac e ludwigii [36] (AE1B 89 and AE1B 92), Bacillus simplex [37] and B. mu alis [38] (AE1B 26 and AE1B 35), Ae omonas hyd ophila and A. media (AE1B 22 and AE1B 20), and Te ibacillus go iensis [39] and Te ibacillus saccha ophilus [40] (AE1B 21 and AE2B 122). Se e al o he selec ed isola es belong o gene a ha ha e been al eady epo ed o exp ess anses e i ica ion ac i i y, e.g. Pseudomonas, En e obac e and Bacillus. O he mic oo ganisms ha ha e been epo ed o ha e enzymes wi h anses e i ica ion ac i i y a e P. luo escens, P. cepacia, Rhizomuco miehei, Rhizopus o yzae, Candida ugosa, The momyces lanuginosus, Candida an a c ica, Ch omobac e ium iscosum, Bu kholde ia cepacia, E. ae ogenes, Muco miehei, Penicillium expansum and B. sub ilis [31]. I is in e es ing o no e ha he s ains AE2B 122 and AE1B 21 showing he bes pe o mance in he p oduc ion o bio uel unde ou expe imen al condi ions belong o he genus Te ibacillus. To he bes o ou knowledge he e a e no epo s on he employmen o membe s o he genus Te ibacillus o anses e i ica ion eac ions, and he e o e, hey cons i u e no el p oduce s o bio uel syn hesizing enzymes. Viscosi y measu emen s The high simila i y o he ch oma og aphy e en ion alues ob ained be ween di e en MG and FAEE compounds sugges s ha he heological p ope ies o hese compound mixes we e simila . This is an impo an ac o o u u e applica ions as bio uel in diesel engines. The e o e, we measu ed he iscosi y o he bac e ia-de i ed bio uel, as a c i ical pa ame e o employing he bio uel. Figu e 4 shows he iscosi y o he supe na an s o he op 5 bio uel-p oducing bac e ia. All bac e ial supe na an s selec ed o bio uel p oduc ion educed he iscosi y o he eac ion p oduc a leas h ee imes compa ed o he iscosi y o he sun lowe oil. P ocessing he bac e ial supe na an s as desc ibed in ma e ials and me hods u he educed he iscosi y o he eac ion p oduc (Fig. 4), p o iding ano he e idence o he ad an age o employing p ocessed supe na an s o e non- p ocessed supe na an s. Kine ics o biodiesel p oduc ion and e-u iliza ion o bac e ial supe na an s All p e ious expe imen s o bio uel p oduc ion we e pe o med o 24 h. Th ee s ains we e andomly selec ed o es whe he he eac ion imes could be sho en unde ou expe imen al condi- Figu e 3. E olu iona y ela ionships o he selec ed s ains. Phylogene ic ees we e in e ed om he 16S RNA sequences o he 23 G am nega i e (A) and he 7 G am posi i e (B) bac e ia wi h Neighbo -Joining clus e ing. The dis ances we e calcula ed using Maximum Composi e Likelihood. The bac e ial s ains isola ed in his wo k a e indica ed in bold. 16S RNA gene sequences om he isola es co espond o 650 bp. Ba ep esen s a 2% (A) o 1% (B) o sequence di e ence. doi:10.1371/jou nal.pone.0104063.g003 Table 5. The closes ela i e o each isola e is indica ed based on he phylogene ic econs uc ion shown in Figu e 3. Sample Closes ela i e Accession numbe % Simila i y AE2B 29 AE2B 30 AE2B 199 AE2B 222 AE2B 232 AE2B 263 AE2B 264 AE2B 332 AE2B 340 Pseudomonas e onii AF064460 99–100 AE2B 130 AE2B 133 AE2B 134 AE2B 261 Pseudomonas ex emaus alis AJ583501 99–100 AE2B 259 Pseudomonas g imon ii CFML 97-514T; AF268029 AF268029 99 AE2B 85 AE2B 120 Pseudomonas s u ze i AF094748 98–100 AEDH 145 Acine obac e lwo ii X81665 99 AE1B 89 AE1B 90 AE1B 92 En e obac e ludwigii AJ853891 99–100 AE1B 20 Ae omonas hyd ophila s ain: LMG 19562; AJ508765 AJ508765 99 AE1B 22 AE1B 27 Ae omonas media X60410 99–100 AE1B 26 AE1B 250 Bacillus simplex AJ439078 99–100 AE1B 35 Bacillus mu alis AJ316309 99 AE1B 28 S aphylococcus epide midis ATCC 14990; D83363 D83363 99 AE1B 21 AE2B 122 AE2B 131 Te ibacillus go iensis a T. saccha ophilus a DQ519571 AB243845 99–100 Samples AE1B and AE2B we e oba ined om ponds con aining liquid was es om he oil mill. Sample AEDH was ob ained om he was es o he oli e ha es s con aining mainly oli e ee le o e s (lea es and b anches). a S ains AE1B21, AE2B122 and AE2B131 showed he same % o simila i y o bo h T. go iensis and T. saccha ophilus. doi:10.1371/jou nal.pone.0104063. 005 Bio uel-P oducing Bac e ia Isola ed om Oil Was es PLOS ONE | www.plosone.o g 9 Augus 2014 | Volume 9 | Issue 8 | e104063