Molecular tools for selective recovery and detection of lignin-derived molecules

G een Chemis y PAPER Ci e his: G een Chem., 2018, 20, 2829 Recei ed 12 h Feb ua y 2018, Accep ed 8 h May 2018 DOI: 10.1039/c8gc00490k sc.li/g eenchem Molecula ools o selec i e eco e y and de ec ion o lignin-de i ed molecules† Milla Salmela, * a Hanna Sanma k, b Elena Efimo a, a Alexande Efimo , a Vesa P. Hy önen, c,d U po Lamminmäki, b Su i San ala a and Ville San ala a The pulp and pape indus y oge he wi h lignocellulosic bio uel p oduc ion p o ides plen i ul s eams o lignin and lignin-de i ed molecules (LDMs) ha cu en ly emain unde u ilized. The he e ogenei y and complexi y o lignin along wi h he lack o con enien ools significan ly hampe i s u iliza ion. Selec i e sepa a ion o hese LDMs om s eams using specific ools would allow he eco e y o a oma ic com- pounds, as well as acili a e biological p ocesses aiming a lignin alo iza ion. To his end, he e we epo he isola ion and cha ac e iza ion o single-chain a iable agmen (scF ) an ibodies agains e ula e, cou- ma a e, and caffea e, which a e he molecula ep esen a i es o LDMs. Binde s o he a ge LDMs we e en iched by in e oga ing a syn he ic scF lib a y wi h he phage display echnique. As a esul , scF binde s specific agains each o he a ge molecules we e ob ained wi h affini ies in he mic omola ange. The selec i i y o scF s owa ds specific LDMs was p o ed by eco e ing caffea e om simula ed LDM solu ion, K a lignin, and ice s aw hyd olysa e samples. Fu he p oo o concep s udies wi h model compounds demons a ed he applicabili y o an ibody-based binde s as a de ec ion ool o moni o ing mic obial LDM con e sion. O e all, his s udy demons a es he po en ial o scF binde s as a specific oolse o lignin compound eco e y and analysis. In oduc ion Cu en ly, he chemical indus y depends s ongly on c ude oil e inemen p ocesses designed o anspo a ion uels and ene gy p oduc ion. 1,2 The e a e, howe e , conce ns o e he economic and en i onmen al sus ainabili y o hese indus ies 3–5 as well as on he global sufficiency o oil depos- i s. 6 These conce ns mo i a e us o dec ease ou eliance on ossil uels, and he e o e mode n e ine ies equi e new aw ma e ials independen o he oil indus y. 7 Fo hese pu poses, lignin is conside ed a p omising eeds ock o a ple ho a o chemicals and ma e ials. De i ed om sou ces such as he pulp and pape indus y, as well as om he eme ging indus y o lignocellulosic bio uel p oduc ion, lignin can p o ide an en i onmen ally sus ainable and abundan al e na i e o he oil-based chemical indus y. 8 Fu he mo e, lignin alo iza ion enables he p oduc ion o bo h ine and bulk chemicals in a bio e ine y concep . 9–12 Howe e , comme cial applica ions and indus ial scale p oduc ion ha e been hinde ed due o he ecalci an and he e ogeneous na u e o lignin 13 and he p esen indus ial p ocesses disca d he lignin s eams mainly as a was e o combus hem o hea . Lignin is an a oma ic he e opolyme ound embedded wi h cellulose and hemicellulose in lignocellulosic biomass. Plan s syn hesize lignin p ima ily om 4-hyd oxyphenylp opanoids ia oxida i e coupling, which esul s in diffe en phenyl- p opane s uc u es in he polyme . 14 Consequen ly, complex mix u es o high and low molecula weigh a oma ic s uc- u es, among o he compounds, can be ound in he lignin s eams ob ained om biomass p e- ea men p ocesses. In addi ion, he go e ning a oma ic s uc u es o hese echnical lignins depend no ably on he o igin o he biomass and on he chosen ea men me hod. 15 Fo ins ance, e ula e and couma a e a e ela i ely dominan a oma ic monome s in lignin ac ions acqui ed om pa icula ag icul u al and g assy o igins 15–19 and in some cases ace amoun s o caffea e ha e also been de ec ed. 20 These a oma ic compounds a e examples o indus ially in e es ing molecules. They possess an ioxidan p ope ies, which make hem aluable molecules o example o he ood indus y. 21 In addi ion, in he chemical indus y, he use o a oma ic monome s om sus ainable sou ces is desi able, †Elec onic supplemen a y in o ma ion (ESI) a ailable. See DOI: 10.1039/ c8gc00490k a Labo a o y o Chemis y and Bioenginee ing, Tampe e Uni e si y o Technology, Ko keakoulunka u 8, 33720 Tampe e, Finland. E-mail: [email p o ec ed] b Depa men o Biochemis y/Bio echnology, Uni e si y o Tu ku, Tykis öka u 6 A, 6 h l ., 20520 Tu ku, Finland c Facul y o Medicine and Li e Sciences and BioMediTech, Uni e si y o Tampe e, Lääkä inka u 1, 33520 Tampe e, Finland d Fimlab Labo a o ies, Bioka u 4, 33520 Tampe e, Finland This jou nal is © The Royal Socie y o Chemis y 2018 G een Chem.,2018,20,2829–2839 | 2829 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online View Jou nal | View Issue hough selec i e ca aly ic modi ica ion o hese compounds is hinde ed due o he difficul lignin ma ix hese compounds eside in. 22 Biological sys ems, howe e , a e known o be highly speci ic and wo k in dilu e solu ions. Selec i e binding o hese molecules, o example, would enable he isola ion o speci ic a oma ic compounds om a mix u e o diffe en LDMs, which migh p o e use ul o he eco e y o small molecules om complex liquo s, 23 dilu e s eams o om biocon e sion p ocesses accumula ing speci ic a oma ic compounds. 24 In na u e, ce ain bac e ia ha e he abili y o pe o m selec- i e con e sion o a oma ic compounds om he e ogeneous ma e ials such as lignin. 25,26 Bac e ia, such as Acine obac e baylyi ADP1, 27 Rhodococcus jos ii, 28 and Pseudomonas pu ida 19,20,29 can, o example, u ilize e ula e, couma a e, and caffea e as he sole ca bon and ene gy sou ces. Mo eo e , me abolic enginee ing and syn he ic biology app oaches p o ide ools o con e hese subs a es in o aluable p oduc s such as anillin o polyhyd oxy ca bona es (PHA). 26,27,30 An ex ended applica ion o LDM selec i e binde s could be he speci ic and apid de ec ion o LDMs om cul u e b o hs o he op imiza ion o biocon e sion p ocesses in ol ing lignin. An ibodies a e p o eins ha can bind speci ically o a ge analy es e en in complex ma ices. They a e commonly used in diagnos ic and he apeu ic applica ions and o some deg ee in en i onmen al analyses. 31–33 Single-chain a iable agmen (scF ) molecules a e mode n an ibodies enginee ed o con ain only he smalles immunoglobulin uni s equi ed o an igen ecogni ion. 34 The scF s can be p oduced in simple bac e ial exp ession sys ems such as Esche ichia coli, and hey can easily be u he gene ically modi ied, o example, o endow hem wi h common labelling p o eins such as g een luo escen p o ein (GFP) o alkaline phospha- ase (AP). The ools o gene ic enginee ing can be used o gene a e highly di e se syn he ic epe oi es o scF s, om whe e binde s agains a ge s o in e es can be isola ed by echniques such as he phage display echnique. An igen- binding si e a chi ec u es o he an ibodies in hese lib a ies can be designed o a o he ecogni ion o ce ain ypes o an igens diffe ing o ins ance in e ms o size. 35 Wi h a sui - able an ibody lib a y, speci ic binde s can be ob ained e en o e y small hap ens, such as ska ole (MW = 131.2 Da), 36 which ypically a e e y difficul a ge s o an ibody de elop- men . Thus, scF binde s o small molecules ha a e ele- an o he lignin indus y, such as hyd oxycinnama es, migh also be ound om syn he ic an ibody lib a ies, e en hough hey diffe only in he deg ee o me hoxyla ion and he numbe o hyd oxyl g oups. In he p esen s udy, an ibodies capable o speci ic ecog- ni ion o lignin compounds e ula e, couma a e, and caffea e we e isola ed and cha ac e ized. As a p oo o concep expe i- men , caffea e was eco e ed om ea ed lignin con aining liquo s by a as and simple column pu i ica ion me hod u iliz- ing a caffea e speci ic scF . The addi ional applica ion o he scF binde s was demons a ed wi h a highly speci ic caffea e- binding an ibody by moni o ing he caffea e u iliza ion o A. baylyi ADP1 by a simple compe i i e enzyme-linked immunoso ben assay (ELISA). Ma e ials and me hods De ailed desc ip ions o he me hods o he chemical syn- hesis o bio inyla ed a ge molecules, biopanning, cloning and ecombinan p o ein exp ession a e desc ibed in he ESI.† Phages, s ains, plasmids, enzymes and bio inyla ed hap ens The binde s o e ula e, couma a e and caffea e we e isola ed om a syn he ic an ibody phage lib a y (ScF M) desc ibed by Huo inen e al. 35 E. coli XL1-Blue (S a agene, USA) cells we e used o phage in ec ion and p oduc ion, cloning, and p o ein exp ession. The VCS-M13 helpe phage (S a agene) was used o escue he phagemid ca ying phage in biopanning. S iI es ic ion si es we e used o clone he scF genes om he pEB32x phagemid o exp ession ec o s pLK06H 35 and pLK04. 37 The pLK06H ec o con ains an ampicillin esis ance ma ke , as well as an addi ional his idine ag and wo S iI si es ups eam o an AP gene, whe eas he pLK04H ec o lacks he AP gene. All o he enzymes and buffe s o cloning we e pu - chased om The mo Scien i ic, USA. Caffea e consump ion s udies we e conduc ed wi h A. baylyi ADP1 (DSM 24194, DSMZ, Ge many). The bio inyla ed a ge molecules ( a ge hap ens) we e con- s uc ed by chemical syn hesis. Th ee diffe en conjuga es we e manu ac u ed ca ying e ula e, couma a e, o caffea e. Each o he molecules con ained a linke wi h 8 PEG uni s and a bio in moie y. Cons uc ion o he bio inyla ed a ge mole- cules (bio in-PEG-[caffea e/ e ula e/couma a e]) is desc ibed in de ail in he ESI.† Media componen s E. coli XL-1 Blue cells we e g own in Supe B o h (SB) medium ( yp one 30 g L −1 , yeas ex ac 20 g L −1 , MOPS 10 g L −1 , and pH 7) and supplemen ed wi h 1% w/ glucose and an ibio ics ( e acycline 10 µg mL −1 , ampicillin 100 µg ml −1 , chlo am- phenicol 25 µg mL −1 o kanamycin 30 µg mL −1 )whenapp op i- a e. A. baylyi ADP1 cells we e g own in minimal sal medium MA/9 (Na 2 HPO·2H 2 O5.518gL −1 ,KH 2 OPO 4 3.402 g L −1 ,NH 4 Cl 0.963 g L −1 , ni ilo iace ic acid 0.008 g L −1 , NaCl 1 g L −1 , FeCl 3 0.001 mg L −1 , MgSO 4 240 mg L −1 , and CaCl 2 11 mg L −1 ) supplemen ed wi h 10 mM ace a e and 10 mM caffea e. Fo media p epa a ion, caffeic acid (Sigma) was i s dissol ed in MA/9 media wi hou supplemen a ion (pH is adjus ed o 8.5 wi h NaOH) and il e ed h ough 0.2 µm il e s. The pH o he s ock solu ion dec eased a e he caffeic acid addi ion close o neu al (pH 6.5–7). Phage en ichmen immunoassay The scF -phage lib a y scF M was independen ly en iched agains he diffe en bio inyla ed a ge molecules (bio in-PEG- [caffea e/ e ula e/couma a e]) in h ee consecu i e ounds o affini y-based biopanning using supe pa amagne ic beads. Pape G een Chemis y 2830 |G een Chem.,2018,20,2829–2839 This jou nal is © The Royal Socie y o Chemis y 2018 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online A mo e de ailed desc ip ion o he biopanning p ocess is offe ed in he ESI.†Subsequen ly, a ime- esol ed luo ome y-based phage immunoassay was used o e i y he en ichmen o he a ge speci ic scF -phage. All o he assay s eps we e con- duc ed in a wo king olume o 200 µL o assay buffe (50 mM T is-HCl, 150 mM NaCl, 0.1% BSA, 0.0.1% Tween20, pH 8.0) a oom empe a u e wi h slow shaking on a p ewashed 96-well s ep a idin (SA) pla e (Kai ogen, Finland). All washing s eps we e un ou imes wi h a Del ia pla e washe (Pe kinElme , Finland) using wash buffe (5 mM T is-HCl, 150 mM NaCl, 0.01% Tween20, pH 8.0). Fo a ge molecule immobiliza ion, a sa u a ing concen a ion o bio inyla ed a ge molecules (bio in-PEG-[caffea e/ e ula e/couma a e]) was added in he s ep a idin-well ollowed by 30 minu es o incuba ion and washing. Aliquo s o 1 × 10 9 c u mL −1 phage om all he panning eac ions we e added independen ly on he pla e, which was incuba ed o 1 hou and hen washed. Then, 25 ng o eu opium-labelled an i-M13 phage an ibody 9E7 (Depa men o Biochemis y/Bio echnology, Uni e si y o Tu ku, Finland) was added and he eac ions we e incuba ed o 30 min be o e washing. The ea e , DELFIA enhancemen solu ion (Pe kinElme ) was added, incuba ed o 10 minu es and hen he luo escence signals we e measu ed wi h ime- esol ed luo escence measu emen (1420 Vic o Mul ilabel Coun e (Pe kinElme ) p og am Eu opium). All samples we e un as duplica e. The wells wi hou he a ge molecules we e used as con ols. P ima y, seconda y and compe i i e AP-ELISA sc eening assays The en iched scF s genes we e cloned om he phagemid ec o as a pool in o he ec o pLK06H. Depending on he a ge , 100–200 andomly chosen colonies we e hen picked om he ans o ma ion pla e o exp ess he chosen clones as scF -AP- usions. La e on, ce ain scF s we e also exp essed in he single chain o m using he ec o pLK04H. De ailed desc ip ions o cloning, AP-scF and scF lysa e p epa a ion as well as p o ein exp ession and pu i ica ion a e offe ed in he ESI.†Assessmen s o he ini ial binding o he exp essed 100–200 scF s owa ds hei a ge molecules we e conduc ed by AP-ELISA. The assay and wash condi ions we e simila o he phage immunoassay. AP-scF lysa es we e added on he immobilized molecules in a inal dilu ion o 1 : 10, incuba ed o 1 hou and washed. Then, 100 µL o 1 mg ml −1 o 4-ni o- phenyl phospha e disodium sal hexahyd a e (pNPP) (Sigma- Ald ich) in pNNP buffe (500 mM T is, 200 mM NaCl, 10 mM MgCl 2 , pH 9.0) was added in o he wells and incuba ed o 40 min. The abso bance was measu ed wi h a Mul iskan Ascen pla e eade (The moLabsys ems, Finland) a a wa e- leng h o 405 nm. The binde s en iched agains couma a e and e ula e we e sc eened agains bo h couma a e and e u- la e, whe eas caffea e binde s we e sc eened agains caffea e and e ula e. Based on he ini ial sc eening, 18 o 21 clones we e chosen o a seconda y sc eening. The assay was he same as he ini ial sc eening, excep ha he eshly p epa ed lysa es we e dilu ed o a inal concen a ion o 1 : 2, un as duplica es and pNPP solu ion (Sigma-Ald ich, #P7998) was applied as he eac an . Using he highes signal- o-backg ound a io and minimal c oss- eac i i y as he c i e ia, se en scF -APs pe a ge molecule we e chosen o sequencing. A compe i i e AP-ELISA was conduc ed o scF -APs wi h unique sequences in o de o con i m he ecogni ion o he non-bio inyla ed o ms o he a ge molecules. In his assay, he lysa es we e incuba ed oge he wi h he bio inyla ed a ge molecule and inc easing concen a ions o he co esponding ee acid (0.125–64 µM) o 1–2 h. Then, he mix u e was ans- e ed o SA-coa ed pla es and incuba ed o 10 minu es. A e washing he wells, pNPP solu ion was added and he pla es we e incuba ed o 1 h. The abso bance was measu ed a 405 nm. The samples we e un as iplica e. La e on, simila assays we e conduc ed wi h pu i ied scF -APs. Fo a caffeic acid binde a compe i i e assay was also pe o med agains ca echol and 4-e hylca echol. E alua ion o binding speci ici y and kine ics by biolaye in e e ome y A Fo ebio Oc e RED384 ins umen equipped wi h 16 pa allel SA-coa ed op ical biosenso s (Fo ebio, Pall Li e Sciences, Menlo Pa k, USA) was used o u he cha ac e ize he binding p ope ies o he selec ed binde s. The da a collec ed we e p o- cessed and analyzed using he Oc e Da a Analysis So wa e ( e sion 7.1). Fo he pu i ied binde s, buffe exchange o Oc e kine ic buffe (1 mg ml −1 bo ine se um albumin (BSA), 0.02% Tween 20, 0.05% NaN 3 in PBS) was ca ied ou using Illus a NAP-10 columns (GE Heal hca e Li e Sciences, USA) acco ding o he manu ac u e ’s ins uc ions. Fo scF samples, kine ic buffe was supplemen ed wi h 8% glyce ol. The SA biosenso s we e soaked p io o un o 15–60 minu es in ITC-buffe (50 mM NaPO 4 , 100 mM NaCl, pH 7.4). The s i ing speed was se a 500 pm and he em- pe a u e was a 27 °C. Black, il ed-bo om 384-well pla es (Fo ebio, Pall Li e Sciences) we e used wi h 80 µL buffe / sample olumes. The baseline o he senso s was eco ded in ITC-buffe o 60 seconds. The ea e , he syn hesized ligands (∼100 µM bio in-PEG-[caffea e/ e ula e/couma a e] in ITC- buffe con aining 10% e hanol) we e a ached o he senso s in a 300 s ac i a ion s ep, ollowed by b ie washing and quenching (sa u a ion o he ee s ep a idins wi h 17 µg ml −1 bio in in ITC-buffe con aining 10% e hanol) s eps o 60 seconds. The baseline o he senso s in he kine ic buffe was eco ded o 300 s. The associa ion and dissocia ion o he scF samples we e measu ed o 600 s. Senso egene a ion wi h 1 M ace ic acid (pH 2) ollowed by a wash s ep wi h kine ic buffe was ca ied ou when app op ia e. Senso s wi hou immobilized ligands and unc ionalized senso s soaked in an ibody- ee kine ic buffe we e used as con ols. The effec o ee acids on he dissocia ion o he scF -APs bound o he unc ionalized biosenso su ace (CAF_3, CAF_5, FER_9, FER_78, COU_12, and COU_74) was s udied by expos- ing he scF -AP-soaked senso s o ee caffea e, e ula e, and couma a e (∼500 µM in kine ic buffe con aining 10% e hanol). The concen a ions o scF -APs we e 1.2, 1.5, 1.6, 1.6, G een Chemis y Pape This jou nal is © The Royal Socie y o Chemis y 2018 G een Chem.,2018,20,2829–2839 | 2831 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online 1.5, and 1.6 mg mL −1 o CAF_3, CAF_5, FER_9, FER_78, COU_12, and COU_74, espec i ely. The dissocia ion o he complex in he p esence o each acid o blank kine ic buffe was hen eco ded. The speci ici y owa ds he acids was de e - mined by calcula ing he dec ease o he su ace dep h in nanome e s du ing he dissocia ion s ep. The dec ease was p opo ioned o ha wi h he blank buffe and no malized. Fo de e mining he affini ies o he scF s CAF_3s, FER_9s, and COU_12s owa ds he hap ens, he samples we e dilu ed o concen a ions o 0.004, 0.016, 0.06, and 0.25 mg mL −1 . The biosenso analysis was ca ied ou o each scF sample as desc ibed abo e, bu in he absence o he compe ing ee ligand. Local pa ial i (20 s om he beginning o he associ- a ion and 60 s om he beginning o he dissocia ion phase) was employed o de e mine he binding affini ies. The affini ies we e calcula ed as a e ages o he alues de e mined o each concen a ion. P epa a ion o scF -based LDM-sepa a ion column Fo selec i e LDM eco e y, a column coupled wi h he caffea e binding scF (CAF_3) was assembled. Buffe exchange o pu i- ied scF CAF_3 was ca ied ou wi h NAP buffe -exchange columns o coupling buffe (0.2 M NaHCO3, 0.5 M NaCl, pH 8.3) and glyce ol was added o a inal concen a ion o 8%. The buffe -exchanged scF s we e coupled o a sepha ose ma ix, in which he coupling occu s ia co alen binding be ween he lysine esidues o he an ibodies and he NHS- g oups o he sepha ose ma ix making he scF -coupled bed eusable. The coupling was conduc ed acco ding o he manu- ac u e ’s ins uc ions using 0.4 mg o an ibodies pe ml bed olume, le o e nigh a 4 °C and blocked wi h 0.1 M T is- HCl, pH 8.5. The eac ion mix u e was washed i e imes wi h 0.1 M T is-HCl (pH8.5) and 0.1 M ace a e buffe , 0.5 M NaCl (pH 4.5). The coupled ma ix was packed in o a column unde g a i a ion and used o caffea e eco e y s udies. A o al olume o 1 mL o sepha ose ma ix was used o he coupling esul ing in a packed column bed o 1 ml. Caffea e eco e y wi h scF -based column Fo he quali a i e selec i i y s udies o he scF CAF_3s, ou diffe en samples (caffea e, simula ed LDM mix u e, K a lignin and ice s aw hyd olysa e) spiked wi h caffea e, couma- a e and e ula e we e pou ed h ough he p epa ed column and 1 mL o he ac ion samples was analyzed. Each sample was dilu ed in assay buffe in o a o al olume o 5 mL. The column was equilib a ed wi h assay buffe (50 mM T is-HCl, 150 mM NaCl, pH 8.0) p io o each expe imen . The samples we e washed wi h 10 ml o wash buffe (5 mM T is-HCl, 150 mM NaCl, pH 8.0) and elu ed wi h 3 ml o glycine buffe (pH 2.2). The collec ed ac ions we e analyzed wi h HPLC (Agilen 1100 se ies, Hewle Packa d, Ge many) equipped wi h a as acid H + column (Phenomex, USA), a degasse (G1322A) and an UV-de ec o (G1315A) using 0.005 N H 2 SO 4 as he eluen . The pump (G1211A) low was adjus ed o 1 ml min −1 , he column empe a u e o 80 °C, and peaks we e iden i ied a a wa eleng h o 310 nm by compa ing he e en ion ime and spec al p o ile wi h s anda ds p epa ed o couma a e, caffea e and e ula e. Fo he spiked K a lignin and ice s aw hyd olysa e, he samples we e analyzed h oughou he spec- um o 210–900 nm. To examine he column’s abili y o cap u e caffea e, 50 µM o caffea e solu ion was passed h ough he column. The ini ial caffea e eco e y es was e un wi h he same column a e e-equilib a ion wi h assay buffe . The same column was used o s udy caffea e eco e y o a simula ed mix u e o LDMs con aining 1–2mMo caffea e, e ula e and couma a e (Sigma). To u he es he column’s capabili y o pu i y LDMs, 0.7 mg o wa e soluble low sul ona e con aining K a lignin (Sigma) supplemen ed wi h caffea e, couma a e and e - ula e (1–2 mM each) was used. Due o he high molecula weigh o he K a lignin (a e age o 10 000 g mol −1 )–and he appa en lack o soluble acids ha could be iden i ied wi h he chosen HPLC equipmen –a es wi h ice s aw hyd olysa e was also conduc ed. In his expe imen , 50 µM o caffea e, cou- ma a e and e ula e we e mixed wi h 300 mg o eeze-d ied ice s aw hyd olysa e (desc ibed by Kannis o e al., 2015), 38 and passed h ough he column. Addi ionally, quan i a i e caffea e eco e y was s udied as abo e wi h a educed washing olume (3 ml) using 2.5 µg o caffea e o 2.6 µg o caffea e dissol ed wi h 3–5 µg o couma a e and e ula e. Caffea e eco e y pe - cen ages we e calcula ed om he elu ion ac ions. Caffea e consump ion s udies The applicabili y o he scF -AP-based immunoassay as a moni o ing ool o subs a e con e sion s udies was es ed wi h he caffea e binde designa ed as CAF_3. Fo hese pu - poses, A. baylyi ADP1 was p e-g own o e nigh on MA/9 minimal sal media supplemen ed wi h 10 mM Na-ace a e and 10 mM caffea e. F om he p e-cul u e, 50 ml cul i a ion (same as p e-cul u e) was inocula ed o he ini ial OD 600 o 0.02, incuba ed a 30 °C and shaken a 300 pm. The samples we e collec ed e e y 2 h o a du a ion o 12 hou s. The g ow h was moni o ed wi h op ical densi y measu emen s and 1 ml o cul u e supe na an was s o ed a −20 °C o subs a e analysis. Caffea e deple ion was analyzed wi h compe i i e AP-ELISA. 0.1 µM o bio inyla ed caffea e and 14 µg mL −1 o he caffea e binde CAF_3 we e mixed oge he wi h a 1 : 625 dilu ion o he collec ed samples in a inal olume o 150 µL. F ee caffeic acid concen a ions anging om 1 µM o 32 µM we e used o he calib a ion cu e. To con e he ELISA abso bance alues o mM, a 4 PL cu e i 39 was applied on he calib a ion cu e. Eqn (1) was used o sol e he sample concen a ion by eplacing λwi h he abso bance alue and sol ing x om he equa ion. The alues c(in lec ion poin ) and b(coefficien o slope s eepness) we e app oxima ed wi h he Ma lab 2016 so wa e da a i ing ool. Fo da a i ing, he leas squa es me hod was used. λ¼λmin þðλmax λminÞ 1þx c  b  ð1Þ Pape G een Chemis y 2832 |G een Chem.,2018,20,2829–2839 This jou nal is © The Royal Socie y o Chemis y 2018 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online λ, he abso bance signal (y-axis), 405 nm; λ min , he minimum eco ded abso bance signal; λ max , he maximum eco ded abso bance signal; x, [ ee caffeic acid] (x-axis); c, he in lec ion poin ; b, he coefficien o slope s eepness. Addi ional caffea e concen a ion measu emen s we e conduc ed wi h UV-Vis a 280 nm (NanoD op 2000 Spec opho ome e , The mo Fische Scien i ic). The ace a e concen a ions we e measu ed wi h a HPLC (Shimadzu, USA) equipped wi h an RID-10A de ec o , an SIL-20AC HT au o- sample and a DGU-20A3 degasse . The empe a u e o he Resex™RHM-Monosaccha ide H + column (Phenomex, USA) was adjus ed o 25 °C and he low a e o 0.6 mL min −1 . 0.01 N H 2 SO 4 was used as he mobile phase. Resul s and discussion In p inciple, phage displayed scF -lib a ies p o ide a po en ial sou ce o binde s owa ds any molecule o in e es . LDMs such as hyd oxycinnama es a e he molecules o in e es due o hei po en ial use as building blocks o he chemical indus- y o as subs a es in bio echnological p ocesses. Thei selec- i e con e sion o aluable end p oduc s is possible by ce ain mic oo ganisms, e en hough o many o ganisms used in cu en bio e ine y p ocesses e en low concen a ions o hese molecules can be oxic owing o he an ioxidan na u e o he molecules. 23 In addi ion o hei po en ial as subs a es o biocon e sion p ocesses, a oma ic LDMs a e also alued as la o s, ag ances and an ioxidan s. 21,40 As hese compounds a e ound in he e ogeneous mix u es in ea ed lignin ac- ions, a sys em a ge ing hese highly simila molecules speci ically and di ec ly om hyd olysa es could p o e use ul o he eco e y o pa icula LDMs. Such eco e y migh be use ul also in cases, whe e he he e ogeneous LDMs a e con- e ed in o speci ied a oma ics h ough mic obial p ocesses. In his s udy, we used a syn he ic scF M phage lib a y, specially designed o a o he ecogni ion o low molecula weigh compounds o selec binde s agains h ee s uc u ally ela ed bio inyla ed a ge s (LDMs) wi h a e y low molecula weigh . The bio inyla ed a ge -molecule conjuga es o couma- a e, e ula e and caffea e we e chemically syn hesized o ou LDM-binde de elopmen . Success ul an ibody de elopmen equi es conjuga es wi h unambiguously de ined s uc u es, oge he wi h an app op ia ely spaced linke be ween he bio in moie y and he a ge molecule. Wi hou hese conju- ga es, i would be necessa y o use immobilized ca ie p o- eins, such as BSA, du ing an ibody selec ion. Typically, he a ge hap ens a ach a iably a ound he ca ie p o ein esul - ing in diffe en epi ope s uc u es. Undesi ably, his phenom- enon p oduces a mo e miscellaneous g oup o binde s, whe eas bio inyla ed conjuga es p oduce binde s o mo e uni o m quali y. En ichmen o LDM speci ic scF pools Highly speci ic scF s owa ds h ee common LDMs we e iso- la ed om a o al numbe o 6 × 10 9 independen clones s o ed in a scF M phage-lib a y, and used in a no el app oach o eco e , de ec , and quan i y LDMs. These scF s ac as binde s owa ds phenolic acids, couma a e, e ula e and caffea e. F om an immunoassay pe spec i e, he phenolic acid concen a ions o lignin hyd olysa es 19,20 a e a he high. Consequen ly, scF s wi h a mode a e affini y can be success- ully used as binde s, howe e high speci ici y is necessa y due o he a iabili y o compounds p esen in echnical lignins. The phages wi h he desi ed p ope ies we e en iched wi h a ge -molecule coa ed magne ic beads. To minimize he unspeci ic binding o he pa amagne ic bead ma e ial, he s ep a idine-coa ed beads we e changed o neu a idin (NA) in he second ound and back o SA in he hi d ound. Acco ding o he ob ained da a om he phage immunoassay, e ula e and couma a e binde s we e en iched in he 2 nd and 3 d panning ounds wi hou signi ican unspeci ic binding o SA (Fig. 1). Fo he caffea e binding phage also unspeci ic binding o SA occu ed. Wi hin all o he samples, no u he en ichmen was obse ed be ween he 2 nd and 3 d ounds, hough he o e all yield o he phage inc eased owa ds he 3 d ound. Fig. 1 The esul s o phage immunoassay. The ba s ep esen he ela i e fluo escence signals (RFU) ob ained om he assay. Whi e ba s = phages incuba ed wi h he a ge molecule immobilized on an SA-coa ed mic o i e well; g ey ba s = backg ound signal om he SA-well wi hou he a ge molecule. G een Chemis y Pape This jou nal is © The Royal Socie y o Chemis y 2018 G een Chem.,2018,20,2829–2839 | 2833 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online LDM posi i e clones A e phage en ichmen , as he phage pool con ains scF s wi h a ying binding cha ac e is ics, he binde s wi h he desi ed cha ac e is ics we e sc eened. An immunoassay (AP-ELISA) was used o sc een 96–192 andomly chosen scF - AP clones pe a ge molecule o ecogni ion and diffe en- ia ion be ween simila molecules. The ini ial sc eening e ealed 86 posi i e clones ou o 96 clones o e ula e and 84posi i eclonesou o 96clones o couma a e(ESI Fig. 7–9†). In e es ingly, ac i e binde s agains caffea e we e also ound wi h a ela i ely high equency (56 posi i e clones ou o 192 clones), despi e he appa en en ichmen o he s ep a idin in his panning. A clone was de e mined posi i e when he abso bance signal o he well con aining he a ge molecule was highe han ha o he bio in sa u a ed back- g ound con ol wells (= S/B > 1.5). F om hese posi i e clones, o each a ge molecule, app oxima ely 20 clones we e chosen o u he alida ion based on he S/B and low c oss- binding signal (ESI Fig. 7–9†). Acco ding o he seconda y sc eening, o e ula e a ge ed binde s some deg ee o c oss-binding occu ed wi h he majo i y o he clones (16/20 clones), hough e ula e speci ic clones we e also ob ained (4/20 clones). The speci ic ecog- ni ion abili y o mos o he binde s a ge ed agains caffea e and couma a e s uc u es was e i ied by he AP-ELISA. Nea ly all o he clones selec ed agains couma a e showed a low backg ound as well as low c oss-binding (19/21 clones), hough 1/21 clones wi h c oss-binding o e ula e and 1/21 clones wi h no signal a all we e obse ed. Mos o he clones selec ed agains caffea e showed selec i e binding wi hou c oss-binding (12/18 clones), hough 3/18 clones indica ed a high backg ound and c oss-binding, whe eas 3/18 clones showed no signal a all (ESI Fig. 10–12†). Cha ac e iza ion o he LDM speci ic ScF -APs and scF s Fu he scF cha ac e iza ion included sequencing o unique clones as well as binding speci ici y and affini y s udies. Se en o eigh clones pe a ge molecule wi h he highes speci ic signal –compa ed o he backg ound and c oss-signals acco d- ing o he seconda y sc eening –we e chosen o be sequenced. The sequencing da a (ESI Table 1†) e ealed 2/8 unique sequences among binde s a ge ed agains caffea e and 3/7 o binde s a ge ed agains e ula e, whe eas o binde s a ge ed agains couma a e 6/7 sequences we e unique. Speci ici y. Six unique clones we e chosen o u he s udies based on hei pe o mance in he compe i i e assay, i.e. wo clones pe a ge molecule ha showed he mos signi ican esponse (dec ease in signal) upon he addi ion o he co es- ponding ee acid (ESI Fig. 13†). These clones included wo binde s agains each hap en, designa ed as CAF_3 and CAF_5 o caffea e, FER_9 and FER_78 o e ula e, and COU_12 and COU_74 o couma a e. These six AP- usion an ibodies we e exp essed and pu i ied. The pu i y and co ec size (∼75 kDa) o he usion an ibodies we e con i med by SDS-PAGE. Speci ic binding cha ac e is ics o he a ge hyd oxycinnama es we e u he s udied by biolaye in e e ome y. The binding and dissocia ion kine ics o he an ibodies we e de e mined using SA senso s unc ionalized wi h bio inyla ed a ge LDMs. Fo each binde , he p esence o he cogna e ee acid in he dis- socia ion phase was ound o accele a e he dissocia ion o he an ibodies om he senso -bound bio inyla ed ligands com- pa ed o he ligand- ee buffe (Table 1). The p esence o he o he ligands had a less p onounced effec on he dissocia ion kine ics. All he binde s showed p e e ence owa ds hei cogna e a ge hap ens, al hough clea diffe ences in speci ici ies we e obse ed. The caffea e binde s CAF_3 and CAF_5 showed he highes speci ici y, whe eas o he e ula e and couma a e binde s also sligh c oss-binding occu ed. In he AP-ELISA, he couma a e binde s showed e y li le indica ion o c oss- binding o he bio inyla ed e ula e, whe eas he binding p o- iles o caffea e and e ula e binde s indica ed simila beha io agains he bio inyla ed and ee o m o he a ge ed mole- cule. In s udies whe e single model compounds a e used, he c oss- eac i i y wi h o he simila compounds is no an issue and his ype o de ec ion is applicable e en o quan i a i e con e sion s udies. Howe e , he sligh c oss-binding migh p o e p oblema ic when mo e complex lignin s uc u es a e used as a subs a e o i mul iple hyd oxycinnama es a e moni- o ed a he same ime. To o e come his obs acle, he ools o gene ic enginee ing can be used o u he op imize he binding cha ac e is ics o he ob ained an ibodies. In addi ion, a gene ic binde , which could iden i y all o he hyd oxycinna- Table 1 Specifici y o he binde s agains he a ge hyd oxycinnama es, caffea e, e ula e, and couma a e. The specifici y was de e mined by com- pa ing he an ibody dissocia ion om he bio inyla ed hap en in he p esence o he compe i i e ee acid obse ed a 600 seconds o he dis- socia ion obse ed in he absence o he compe ing ligand Binde Ta ge an igen Rela i e dissocia ion No compe i o F ee caffea e F ee e ula e F ee couma a e CAF_3 Caffea e 1.0 2.0 1.0 1.0 CAF_5 1.0 1.2 0.8 0.9 FER_9 Fe ula e 1.0 1.2 1.5 1.1 FER_78 1.0 1.3 1.3 1.1 COU_12 Couma a e 1.0 1.6 1.4 1.7 COU_74 1.0 1.6 1.5 1.7 Pape G een Chemis y 2834 |G een Chem.,2018,20,2829–2839 This jou nal is © The Royal Socie y o Chemis y 2018 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online ma es equally and simul aneously, migh also be use ul. Such gene ic binde s ha e p e iously been de eloped o molecules such as sulphonamides and luo oquinolones. 41,42 None heless, he speci ic beha io o he caffea e binde CAF_3 was clea ly p o ed in hese expe imen s indica ing he applicabili y o scF s as speci ic ools in lignin esea ch. Based on he speci ici y p o ile, one AP-scF binde agains each a ge LDM (CAF_3, FER_9, and COU_12) was selec ed o addi ional compe i i e AP ELISA expe imen s. All o he selec ed binde s ecognized also he non-bio inyla ed ( ee acid) o m o he molecule (Fig. 2), simila o he biolaye in e e ome y expe imen s, e en hough bio inyla ed a ge molecules we e used in he panning and sc eening s eps. This implies ha he pa a ope–epi ope pai occu s a he a oma ic ing s uc u e o he molecule. This claim is u he suppo ed by he ac ha he diffe en hyd oxycinnama es –diffe ing only in he deg ee o me hoxyla ion o in he numbe o hyd oxyl g oups ound in he a oma ic ing s uc u es –we e ecognized o a lesse deg ee when compa ed o he a ge molecule in he c oss-binding expe imen s. Addi ionally, he CAF_3 binde is speci ic o he caffeic acid s uc u e, includ- ing he alkyl chain o he molecule. Ca echol and 4-e hylca e- col ha e simila a oma ic subs i u ion o caffeic acid diffe ing only in he alkyl subs i uen s. In he p esence o ca echol o 4-e hylca echol, he scF CAF_3 showed no compe i i e binding o hese compounds, in con as o he ee caffeic acid (ESI 14†). Acco ding o he da a collec ed om he com- pe i i e immunoassay, he IC50 alues we e app oxima ely 3 µM o he caffea e binde , 7 µM o he couma a e binde and 17 µM o he e ula e binde . The eliable ecogni ion o hese scF -AP binde s unde he chosen condi ions spans in he egion o 1–20 µM. No ably, wi h mul iple sc eening and selec ion me hods, binde s wi h speci ic binding p o iles owa ds he a ge mole- cules we e success ully ob ained. Rega dless o he high back- g ound in he phage immunoassay, i was also possible o isola e a highly speci ic an ibody o caffea e, which could dis- inguish be ween he hyd oxyl and me hoxy g oups o e ula e and couma a e. Affini y. In o de o de e mine he affini ies o he binde s, hey we e p oduced as scF s wi hou he AP usion; due o he dime ic na u e o AP, he affini ies canno be eliably de e - mined o an ibody-AP usions. The binde s CAF_3, FER_9, and COU_12 we e cloned o he ec o pLK04H, exp essed, and pu i ied. The pu i y and size (∼27 kDa) we e con i med by SDS-PAGE. The affini ies o he esul ing scF s, designa ed as CAF_3s, FER_9s, and COU_12s, espec i ely, we e de e mined by biolaye in e e ome y s udies. Ra he na ow concen- a ion anges had o be used in he de e mina ion due o he spon aneous and apid agglome a ion o he single chains a highe concen a ions. The affini ies o all binde s we e ound o be in he mic omola ange, CAF_3s possessing he highes affini y o 2.0 × 10 −6 M (Fig. 3). The mode a e affini y ange is also a o ed as i enables he elu ion o he a ge ed molecules o eco e y pu poses. The mic omola affini y ange is con enien o hei use in bo h a ge compound moni o ing and eco e y, conside ing he ele an concen a ions o hyd oxycinnama es in he lignin-con aining liquo s as well as o allow efficien elu ion o he a ge compound in hyd oxycinnama e eco e y. LDM sepa a ion om he e ogeneous mix u es One po en ial app oach o he alo iza ion o lignin-based s eams is he selec i e and high eco e y o monome ic com- Fig. 2 The binding o scF -APs FER_9, COU_12 and CAF_3 owa ds a ge molecules. In his compe i i e assay, he selec ed binde s we e mixed oge he wi h he bio inyla ed a ge molecule and inc easing concen a ions o he co esponding ee acid (0.125–64 µM). The binde s bound o he co esponding ee acid we e washed away and he abso bance was measu ed. The abso bance signals o he compe i i e binding esul s a e scaled o ela i e abso bance B/B0 (B = measu ed signal, B0 = max. signal). Whi e = e ula e binde FER_9, blue = couma a e binde COU_12, and ed = caffea e binde CAF_3. The s anda d de ia ion o iplica es is also shown in he figu e as e o ba s. G een Chemis y Pape This jou nal is © The Royal Socie y o Chemis y 2018 G een Chem.,2018,20,2829–2839 | 2835 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online pounds such as phenolic acids. The commonly used me hods o phenolic compound eco e y include poly inyl esins o ac i a ed ca bon ma ices. 43–45 The phenolic compounds bind o he ma ix ei he ia hyd ophobic in e ac ions o by hyd o- gen bonding om he a oma ic o ca boxyl g oup o he phe- nolic compound allowing non-selec i e adso p ion on o he ma ix. Fo example, da Cos a Lopes e al. (2016) de eloped a me hod o pu i ying phenols ex ac ed om biomass wi h ionic liquids using poly inyl esins. 46 The eco e ed ac ion was u he pu i ied wi h supe c i ical CO 2 . Howe e , he e we e no a emp s o selec i ely eco e a gi en a ge molecule om he ac ion esul ing in a mix u e o phenolic acids. On he o he hand, he adso p ion and deso p ion con- di ions in a ma ix can be op imized o a o he eco e y o ce ain molecules. Fo example, he selec i e eco e y o a e ulic acid om enzyma ically ea ed suga -bee pulp was s udied by Cou eau & Ma haly (1998) by op imizing a ixed-bed sys em u ilizing ac i a ed ca bon. 47 In hei s udy, 50% e ulic acid pu i y was achie ed. As ano he example, e ulic acid eco e y s udies conduc ed by Tilay e al. (2008) e ealed 50.89% pu i y and 57.97% eco e y efficiency o polyme ic adso ben s. 48 Howe e , by using mo e complex me hods, such as subsequen p epa a i e HPTLC, he pu i y o he eco e ed e ulic acid ac ion was inc eased o 95.35%. 48 Addi ionally, he molecula imp in ing echnique can be used o inc ease he selec i eness o he esin. As an example, Michailo e al. (2008) used molecula ly imp in ed polyme s o eco e caffeic acid, 4-HBA and o he phenolic compounds wi h eco e y efficiencies be ween 9–51%. Fo he caffeic acid imp in ed polyme s ong compe i ion o binding si es occu ed also wi h p-couma ic acid. 49 To examine he po en ial o he an ibody use in speci ic LDM eco e y in he con ex o lignin alo iza ion, he selec i- i y o he chosen scF s was examined in a p oo o p inciple expe imen by sepa a ing caffea e om a he e ogeneous mix u e o soluble compounds. Caffea e eco e y om diffe en mix u es spiked wi h caffea e was demons a ed by using an affini y column p epa ed wi h immobilized CAF_3 scF . Elu ion o he caffea e molecules was achie ed wi h pH change, ollowed by immedia e equilib a ion o he column o e old he co alen ly bound an ibodies, hus making he column ma ix eusable. A e e i ying ha he p epa ed column cap u es caffea e (ESI 14†), he column was used o eco e caffea e om simu- la ed LDM solu ion, k a lignin and ice s aw hyd olysa e spiked wi h caffea e, couma a e and e ula e. Caffea e was eco e ed wi h high pu i y om all o he samples wi hou aces o e ula e, couma a e o o he de ec able compounds p oposing he sys em o be highly selec i e owa ds caffea e (ESI Fig. 15–17†). In o de o quan i a i ely de e mine he caffea e eco e y a e o he column, he maximum heo e ical binding capaci y o he scF -coupled bed was de e mined. Assuming ha all o he scF s we e coupled unc ionally o he ma ix and each Fig. 3 De e mina ion o he binding affini y and dissocia ion kine ics o he caffea e (CAF_3s), e ula e (FER_9s), and couma a e (COU_12s) binde s. Local pa ial fi s we e employed o es ima e he affini ies. The K D and K diss alues a e p esen ed as an a e age o he affini ies es ima ed o ou diffe en binde concen a ions ( h ee o FER9s). The de e mina ion was based on he associa ion and dissocia ion o he binde s o/ om he bio inyla ed LDMs ( he selec ed hyd oxycinnama es) a ached o s ep a idin-coa ed op ical biosenso s. Pape G een Chemis y 2836 |G een Chem.,2018,20,2829–2839 This jou nal is © The Royal Socie y o Chemis y 2018 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online scF binds he a ge ed molecule in a mola a io o 1 : 1, he binding capaci y o he column was es ima ed o be 2.6 µg o caffea e pe 1 ml o so ben in he column. Thus, samples con- aining 2.5 µg pe 1 ml column o caffea e we e used o quan- i a i e eco e y expe imen s. Ve y high eco e y a es we e ob ained o he simula ed LDM mix u e (94%) (Table 2) and also o he caffea e solu ion (77%). Couma a e and e ula e we e ound om he low and wash ac ions o he LDM mix u e e i ying ha hey a e no e ained in he column. Caffea e binde as a moni o ing ool Hyd oxycinnama es, such as p-couma a e, e ula e, and caffea e, a e ound in plan cells and echnical lignin s uc- u es. These molecules a e po en ial subs a es o bac e ial lignin upg ading schemes and simple acking me hods o hese subs a es a e o g ea in e es . Fo example, he speci ic de ec ion o hese molecules di ec ly om he cul u e medium and o he complex ma ices is an appealing app oach in bac- e ial con e sion s udies. To his end, expe imen s wi h he highly speci ic caffea e binde CAF_3 we e conduc ed o e i y i s applicabili y as a moni o ing ool o subs a e con e sion. A. baylyi ADP1 was cul i a ed wi h ace a e and caffea e as he ca bon sou ces. Acco ding o he cul u e sample analysis pe o med by he immunoassay, he samples collec ed om A. baylyi ADP1 cul u es showed he subsequen u iliza ion o ace a e and caffea e (Fig. 4). This is well in co ela ion wi h he known subs a e p e e ences o A. baylyi; i s a oma ic ca abo- lism is ep essed in he p esence o ace a e which is conse- quen ly u ilized be o e he a oma ics. 50 The de eloped immunoassay quan i ied he concen a ion o caffea e in he cul i a ion samples up o he ime poin o 10 hou s. The ea e , he concen a ions we e below he de ec ion limi . The calib a ion cu e o he immunoassay ga e a linea esponse be ween 4–10 µM o he caffea e binde CAF_3 and he caffea e concen a ions om he samples we e u he con- i med wi h UV-Vis 208 nm, suppo ing he immunoassay esul s (Fig. 4). Conclusions The s uc u al complexi y and he e ogenei y o lignin make i s u he u iliza ion and analysis e y challenging. Due o he high binding speci ici y, an ibody-based ools could p o ide a unique means o he eco e y, de ec ion and moni o ing o a a ie y o lignin compounds e en om complex ma ices. In his s udy, binde s agains h ee lignin-de i ed hyd oxycinna- ma es e ula e, couma a e and caffea e we e success ully iso- la ed om a syn he ic an ibody phage lib a y. These scF s can be used o isola e and collec indi idual chemically pu e sub- s ances om a complex mix u e o lignin-de i ed compounds as was demons a ed wi h he speci ic caffea e-binding an i- body. Fu he mo e, he binde agains caffea e was success ully employed o ack caffea e consump ion in a bac e ial biop o- cess. This p oo o concep s udy shows he po en ial o an i- bodies o he eco e y o small molecules om dilu e and challenging s eams. Table 2 Selec i i y and eco e y o caffea e om he CAF_3 scF - column. LDM model compound solu ions we e un h ough he affini y column and he yields o he diffe en ac ions we e calcula ed om he analyzed samples. Reco e y % om he column is calcula ed om he elu ed samples, whe eas eco e y % o al includes he calcula ed yields di ided by he ini ial amoun LDM model compound solu ion Caffea e (μg) Couma a e (μg) Fe ula e (μg) Sample 2.45 3.84 5.62 Flow h ough 0.00 2.24 3.16 Wash h ough 0.00 1.54 2.14 Elu ion 1 0.24 0.00 0.00 Elu ion 2 1.57 0.00 0.00 Elu ion 3 0.50 0.00 0.00 Reco e y % om column 94 0.00 0.00 Reco e y % o al 94 98 94 Fig. 4 Uppe figu e: Subs a e consump ion and g ow h o A. baylyi ADP1 ba ch cul u e o e a ime cou se o 12 hou s. Closed diamonds ep esen ace a e concen a ion in mM and blue ci cles caffea e con- cen a ion in mM (immunoassay). The black as e isks ep esen he bac- e ial g ow h measu ed as he op ical densi y a 600 nm. The open ci cles indica e he caffea e concen a ion as mM when measu ed wi h UV-Vis a 280 nm. Bo om figu e: he calib a ion cu e used o calcula - ing he caffea e concen a ion. Open ci cles ep esen he abso bance measu ed a 405 nm in diffe en caffea e concen a ions and he do ed line he coefficien o a ia ion. The black dashed line ep esen s he da a fi ing o he ob ained abso bance alues o e a ying ee acid concen a ions. G een Chemis y Pape This jou nal is © The Royal Socie y o Chemis y 2018 G een Chem.,2018,20,2829–2839 | 2837 Open Access A icle. Published on 08 May 2018. Downloaded on 5/4/2020 6:55:20 AM. This a icle is licensed unde a C ea i e Commons A ibu ion 3.0 Unpo ed Licence. View A icle Online