Application of L-lactate-cytochrome c-oxidoreductase for development of amperometric biosensor for L-lactate determination

Aim. Development of amperometric biosensor based on L-lactate-cytochrome c-oxidoreductase (flavocytochrome b2, FC b2) for lactate determination. Methods. All experiments were performed using the amperometric method of detection. The methods of electrochemical polymerization and immobilization in glutaraldehyde vapors were used for FC b2 immobilization on the surface of electrodes. Results. The FC b2 preparation, which demonstrated the best operational characteristics after immobilization in poly (3,4-ethylen dioxythiophene), was selected. The selectivity, operational and storage stability, and pH-optimum for operation of the created biosensor were determined. The analysis of L-lactate in the model solutions and wine samples was carried out using the developed biosensor. Conclusion. The FC b2-based biosensor due to its high stability can be effectively used for lactate determination in blood and other liquids containing no ethanol. After the selectivity optimization, the devise can be also applied for wine analysis.

In tro duc tion.To date, de vel op ment of an amperometric bio sen sor for lac tate anal y sis in winemak ing is an im por tant vi tal chal lenge.Nu mer ous meth ods, tra di tion ally used for the anal y sis of lac tate con cen tra tion in wine (liq uid chro ma tog ra phy, cap illary elec tro pho re sis, spectrophotometry and fermentative method), of ten do not guar an tee suf fi cient se lec tiv ity and sen si tiv ity.Be sides, these meth ods are hardly com pat i ble with the wine tech nol ogy be cause of com plex and bulky equip ment, ne ces sity of sam ple pretreatment, laborious and durable analytical procedure [1].
Mean while, lac tate is a sub stance to be thor oughly con trolled at all stages of wine mak ing since it in di cates bac te rial ac tiv ity dur ing must fer men ta tion and, concur rently, de ter mines qual ity and aroma of fi nal product and wine sta bil ity at stor age [1,2].
An ef fec tive tool for the dis crete and con tin u ous lac tate mon i tor ing dur ing wine pro duc tion can be an amperometric bio sen sor -a de vice of small di mensions, sim ple and low-cost in us age, which re quires nei ther sam ple pre treat ment, nor com plex equip ment nor high-qual i fied main te nance staff [3].An ap pli cation of biosensors in food in dus try is in dis put ably advan ta geous due to their high sen si tiv ity, a pos si bil ity of anal y sis of tur bid so lu tion, high com pat i bil ity with advanced mi cro elec tronic tech niques, and small en ergy con sump tion [4].There fore, no won der that biosensors are ever more widely ap plied in analysis of food quality, in particular, in wine production [5].
To de velop amperometric biosensors in tended for lac tate de ter mi na tion in wine, most fre quently NAD + -de pend ent lac tate dehydrogenase or lac tate oxidase (LOD) are used.An anal y sis of work ing char ac ter is tics of the biosensors based on these en zymes shows that the trans duc ers with im mo bi lized LOD have wider linear range and lower de tec tion limit as com pared with those based on lac tate dehydrogenase.An es sen tial advan tage of oxidase biosensors is that there is no need in an ex og e nous co fac tor for their func tion ing that consid er ably fa cil i tates the anal y sis pro ce dure [3].Therefore, lac tate amperometric biosensors based on im mo bilized LOD seem to be more prom is ing for wine anal y sis.
Re cently we have de vel oped the im mo bi lized LOD-based amperometric bio sen sor for quan ti ta tive lac tate de ter mi na tion in wine [2].It has per fect work ing char ac ter is tics: dy namic range of 0.008 -1.0 mÌ and high se lec tiv ity.Its ap pli ca tion in lac tate anal y sis of wine and must was a suc cess, the re sults cor re lated with those ob tained by the traditional HPLC method.
How ever, this bio sen sor was dis ad van ta geous due to low op er a tional and stor age sta bil ity: dur ing first three hours of con tin u ous work the sen sor sig nal lost up to 70% of its ini tial value, and this level of ac tiv ity was re tained for next three days.Note wor thy, that 30% activ ity was suf fi cient for lac tate anal y sis.Nev er the less, for op ti mi za tion of lac tate bio sen sor it was necessary to increase its stability.
An anal y sis of the pub lished data showed that low op er a tional sta bil ity is reg u larly in her ent to the biosensors based on im mo bi lized LOD.For in stance, the oxidase bio sen sor for lac tate de ter mi na tion in wine [6] dem on strated only 35% of ini tial ac tiv ity af ter 150 mea sure ments.In [7] the LOD sen sor is de scribed to lose half its ac tiv ity just af ter the first mea sure ment, and in 5-hour con tin u ous op er a tion it dem on strated 40% of the initial signal [8].
De crease in the im mo bi lized en zyme ac tiv ity can be caused by a num ber of fac tors, i.e. tem per a ture de natur ation, proteolytic deg ra da tion, non-spe cific ox i dation cat a lyzed by met als, changes of pH, so lu tion ion strength, etc. [9].Be sides, elec trode pol lu tion with foreign pro teins and other sub stances pres ent in a tested so lu tion can also be the rea son [8].De ter mi na tion of the pre cise mech a nism of en zyme in ac ti va tion and elab o ra tion of a pre vent ing pro ce dure is an ideal tac tic to solve the prob lem of in creas ing biosensor stability.
One of the meth ods of im prove ment of en zyme stabil ity is a di rect site-spe cific mu ta gen e sis which re sults in higher en zyme lipophility or re place ment of its cat alytic groups to the sites more im pen e tra ble to en vi ronmen tal ef fects [9].It is ev i dent that this ap proach is rather la bo ri ous and long-term and is ef fec tive only if the en zyme pro duc tion is self-de pend ent.
An other method to in crease bio sen sor sta bil ity is sta bi liz ing ad di tions, such as polyelectrolytes, osmolytes, al co hols, pro teins-shaperonines, sug ars, salts or nanoparticles of gold, zinc ox ide, fer ric ox ide Fe 3 O 4 [9 -12].These ad di tions were shown to in ter act elec tro chem i cally with the en zyme, some times even to cap su late it, thus chang ing the en zyme con for ma tion and in creas ing its sta bil ity [9].These sta bi liz ers can be ei ther in tro duced at the en zyme im mo bi li za tion as a com po nent of sen si tive mem brane, or added into the work ing buffer during measurements and storage of sensors.
Use of sta bi liz ing sub stances seems to be an ef fective ap proach, how ever, the se lec tion of op ti mal complex of sta bi liz ers for a cer tain en zyme im mo bi lized in a cer tain way is rather a time-con sum ing pro ce dure.Be sides, a com pli cated, multi-step im mo bi li za tion of an en zyme to gether with sta bi liz ers po ten tially decreases sen sor-to-sen sor re sponse reproducibility; some salts-ad di tions can cause an op po site ef fect, i.e. en zyme destabilization [9].In tro duc tion of sta bi liz ers into work ing so lu tion at anal y sis of real liq uids com plicates the an a lyt i cal pro ce dure and probably can have an interfering effect on biosensor function.
Ac cord ing to the pub lished data, the most es sen tial rea son of con sid er able drop of the ac tiv ity of LOD, immo bi lized into sen si tive mem brane, is par tial en zyme de na tur ation by hy dro gen per ox ide which, be ing gener ated in the re ac tion of lac tate dis in te gra tion, ox i dizes amino ac ids of the cat a lytic cen tre [8,11].There fore an al ter na tive way to im prove sta bil ity of lac tate transducer is us ing in stead of LOD an other en zyme, the func tion ing of which is not as so ci ated with gen er a tion of hydrogen peroxide.
The en zyme, able to be a biorecognition el e ment of the lac tate bio sen sor, is L-lac tate-cytochrome c-oxidoreductase (EC 1.1.2.3; flavocytochrome b 2 (FC b 2 ) cat a lyz ing elec tron trans fer from L-lac tate to cytochrome c in yeast mi to chon dria.FC b 2 can be extracted from Saccharomyces cerevisiae, Hansenula anomala or Hansenula polymorpha as a homotetramer, each sub unit of which con tains one mol e cule of flavine mononucleotide and one mol e cule of proheme IX [13].FC b 2 func tions in vi tro in the pres ence of a num ber of syn thetic ac cep tors of elec trons that con di tions the poten tial of this en zyme in an a lyt i cal bio tech nol ogy.The in ves ti ga tions [14] show better FC b 2 stor age sta bil ity (on the 5 th day af ter im mo bi li za tion the en zyme ac tiv ity is half its ini tial value) and op er a tional sta bil ity (50% of ini tial sig nal is reg is tered af ter 6 hours of continuous work) as compared to these parameters for LOD.
Study on se lec tiv ity of the FC b 2 -based biosensors showed the ab sence of non-spe cific re sponses to malate, pyruvate, ac e tate, and isocitrate [13,14].As to the main wine com po nents, eth a nol and glu cose, they, to gether with ascor bic acid and phe no lic sub stances, es sen tially in ter fere in quan ti ta tive de ter mi na tion of lac tate in wines and must [2].This is why se lec tiv ity towards these sub stances is the great est chal lenge, once the bio sen sor is de vel oped to analyze lactate in wine materials.
This work was aimed at the de vel op ment of lac tate amperometric bio sen sor on the ba sis of im mo bi lized flavocytochrome b 2 and eval u a tion of ad van tages and draw backs of its ap pli ca tion for wine anal y sis as com -pared to the sen sor based on im mo bi lized lactate oxidase.
Ma te ri als and meth ods.Ma te ri als.Three prep ara tions of the en zyme FC b 2 ex tracted from Hansenula polymorpha were used.Prep a ra tion FC b 2 ¹1 pre pared in 40% am mo nium sul phate (AS) had spe cific ac tiv ity of 3 U/mg, pro tein con cen tra tion of 4.8 mg/ml; prep ara tion ¹2 -in 80% AS; prep a ra tion ¹3 -spe cific activ ity 0.75 U/mg, pro tein concentration 13.9 mg/ml.
Mea sure ment.All elec tro chem i cal ex per i ments were per formed by the tra di tional three-elec trode system in which screen-printed elec trode SensLab («SensLab GmbH», Leip zig, Ger many) com bined all three elec trodes, i.e. plat i num work ing, aux il iary, and referent.
Plat i num screen-printed elec trodes SensLab were tested re gard ing reproducibility and op er a tional ca pacity within the range from 0 to +600 mV (po ten tial scan rate of 20 mV/s).Cy clic voltamperometry was car ried out by the potentiostate PalmSens (Palm In stru ments BV, Neth er lands).The amperometric de vice used in the work con sisted of potentiostate PalmSens, elec trochem i cal cell and com puter (gen eral view -in Fig. 1).
The cy clic voltammogram was ob tained us ing a plat i num elec trode SensLab with FC b 2 im mo bi lized in the work ing buffer at ad di tion of 2 mM po tas sium ferricyanide and 1 mM lac tate (Fig. 2).As seen, an in ser tion of the sub strate and the me di a tor into the work ing cell re sults in gen er a tion of ox i diz ing cur rent and in crease of trans ducer sig nal.We chose the po ten tial of +450 mV as a work ing pa ram e ter since it is the value of redox po ten tial of potassium ferricyanide [15].Amperometric mea sure ment was car ried out by the potentiostate PalmSens in a 5-ml elec tro chem i cal cell at a con stant po ten tial.
FC b 2 im mo bi li za tion by elec tro chem i cal poly meriza tion in poly mer EDT.Elec tro chem i cal poly mer ization as a method of en zyme im mo bi li za tion is thoroughly de scribed in [2].For elec tro chem i cal poly meriza tion, the com po nent mix ture, con sist ing of 10 -2 Ì 3,4-ethylenedioxythiophene (EDT), 10 -3 Ì poly eth ylene gly col, and 3 µl FC b 2 so lu tion, was prepared in 20 mM phosphate buffer, pH 6.2.
EDT was poly mer ized at the po ten tial from + 0.2 V to + 1.5 V, ap plied for 15 cy cles at the scan rate of 0.1 V/s.FC b 2 im mo bi li za tion in glutaraldehyde (GA) vapour.Bioselective mem branes were formed in the mix ture of 3 µl FC b 2 so lu tion and 5% BSA (1:3) in 10 mÌ phos phate buffer, ðÍ 7.2.The mix ture was de posited on the work ing elec trode sur face.For poly mer ization, the trans duc ers were placed for 10 min into sat urated vapour of glutaraldehyde, where upon they were dried in the air.
Lac tate de ter mi na tion in model so lu tions.Measure ment was car ried out at room tem per a ture in an open ves sel at in ten sive stir ring.20 mM so lu tion KH 2 PO 4 -Na 2 HPO 4 ×12H 2 O, ðÍ 7.6, was used as a work ing buffer, 2 mM po tas sium ferri cyanide -as a mediator.
Sub strates con cen tra tion was changed by ad di tion of de fined aliquots of con cen trated so lu tions.Af ter each sig nal, the sen sor was washed in a buffer so lu tion un til the base sig nal was sta bi lized.
Re sults and dis cus sion.Op er a tion of FC b 2 -based amperometric biosensors is based on the en zy matic reac tion with me di a tor Med [3]: The elec trons, which are gen er ated as a re sult of these re ac tions and den sity of which is pro por tional to the lac tate con cen tra tion in the work ing cell, are reg istered by the amperometric transducer.
At the first stage of the re search, FC b 2 was im mo bilized in GA vapour in the BSA mem brane.To se lect the best en zyme prep a ra tion the lab o ra tory pro to types were de vel oped on the ba sis of three FC b 2 prep a ra tions dif fer ent by the ac tiv ity and pres ence of ad di tives.The prep a ra tions 1 and 2 af ter im mo bi li za tion in GA vapours ap peared to be in ac tive and did not re spond to lac tate.The bio sen sor based on prep a ra tion 3 im mo bilized in GA va pours dem on strated low (12 nA) sig nals to lac tate, its work ing dy namic range was 0.002 -0.032 mÌ.Study on the bio sen sor se lec tiv ity showed that it gives prac ti cally no re sponse to glu cose, es sen tial non-spe cific re sponses to other in ter fer ing sub stances, re sponses to ascor bic acid be ing neg a tive.Con sid er ing that in the ex per i ments in ter fer ing sub strates were of the con cen tra tions equal to max i mal in wine [1], and lac tate con cen tra tion cor re sponded to sat u ra tion of the bio sen sor cal i bra tion curve, we stated that the sen sor based on prep a ra tion 3, im mo bi lized in GA vapours, has insufficient selectivity and, thus, it is unusable for the lactate measurement in wine and must.
Ex am i na tion of stor age sta bil ity showed that the sen sor ac tiv ity in one day af ter im mo bi li za tion was 18% of ini tial value; op er a tional sta bil ity was completely lost in one hour of con tin u ous work.
Since FC b 2 im mo bi li za tion in GA va pours gave inad e quate re sults, an al ter na tive im mo bi li za tion method was stud ied, i.e. elec tro chem i cal poly mer iza tion in PEDT.All the prep a ra tions, im mo bi lized in this way, ap peared to be ac tive.Cal i bra tion curves of the lab o ratory pro to types with prep a ra tions 1 and 2 are presented in Fig. 3.The sen sor based on prep a ra tion 3 dem on strated very small (no more than 11 nA) re -sponses to lac tate and work ing range within 0.001 -0.06 mÌ substrate concentration.
To se lect the most ef fec tive vari ant of FC b 2 prep ara tions, the de vel oped biosensors us ing three of them, were com pared re gard ing their op er a tional dy namic ranges, lim its of de tect able lac tate con cen tra tions, selec tiv ity, op er a tional and storage stability.
It was re vealed that the bio sen sor with prep a ra tion 3 im mo bi lized in PEDT dem on strated con sid er ably lower, as com pared to other prep a ra tions, de tect able limit (1 µÌ lac tate), though its dy namic range was short, and the sig nal did not ex ceed 11 nA, which restrict its ap pli ca tion for the lac tate anal y sis in real liquids.The sen sors based on prep a ra tions 1 and 2 are featured by wide dy namic range (0.2 -6.4 mÌ and 0.8 -25.6 mÌ lac tate, cor re spond ingly) the sig nal up to 70 -130 nÀ.At the same time, their limit of de tect able substrate con cen tra tion is rather high, which makes it dif ficult to an a lyze wines with low lac tate content.
The re sults of ex am i na tion of op er a tional sta bil ity (Fig. 4a) show that the sen sor with prep a ra tion 1 looses 40% of its ini tial ac tiv ity over the first hours of con tin uous op er a tion, there af ter the sig nal prac ti cally does not change, i.e. the re sponse is highly re pro duc ible.The rea son of ac tiv ity de crease can be a seg re ga tion of weakly ad her ent parts of the sen si tive mem brane from the elec trode sur face and/or wash-out of poorly im mobi lized en zyme mol e cules into the so lu tion [8].Op er ational sta bil ity at us ing prep a ra tion 2 is worse, sen sor ac tiv ity de creases through out the whole pe riod of contin u ous work which re sults in poor sig nal reproducibility.Op er a tional sta bil ity of the bio sen sor with prep a ra tion 3 which com pletely looses its ini tial activity during three hours is unsatisfactory.
Stor age sta bil ity (Fig. 4b) of the bio sen sor with prep a ra tion 3 is also very low; its ac tiv ity is com pletely lost over 5 days.Sta bil ity of sen sor with prep a ra tions 1 and 2 is better, they dem on strated about 30% ini tial sig nal in 90-day stor age.
Fur ther work was aimed at ex am i na tion of se lec tivity of amperometric biosensors based on three FC b 2 prep a ra tions.The re sponses were ob tained as lac tate (in con cen tra tions cor re spond ing to a higher limit of the dy namic range of each trans ducer) and main in terfer ing sub stances, i.e. eth a nol, glyc erol, glu cose, ascor -Fig.3. Calibration curves of laboratory prototypes of amperometric biosensor obtained by FC b 2 electrochemical polymerisation in polymer PEDT: 1 -preparation 1, 2 -preparation 2. Measurement in bic acid (in max i mum con cen tra tions ac tual for wines), were in serted into electrochemical cell (Table 1).
As seen, se lec tiv ity of prep a ra tion 3 is the worst; it gives, alike at im mo bi li za tion in GA vapour, es sen tial non-spe cific re sponses to main in ter fer ing sub stances (pos i tive sig nals to eth a nol, glyc erol, glu cose, and nega tive -to ascor bic acid) which even ex ceed the lac tate sig nal.This can be caused by poor pu ri fi ca tion of the prep a ra tion.As to the prep a ra tions 1 and 2, they have better se lec tiv ity, prac ti cally do not re spond to ascor bic acid, glu cose and glyc erol, though the eth a nol sig nal is close to that to lactate.
The char ac ter is tics of amperometric biosensors based on three FC b 2 prep a ra tions im mo bi lized in PEDT, are com pared with those of the bio sen sor based on LOD immo bi lized in PEDT which we de vel oped ear lier (Ta ble 2).
Tak ing into con sid er ation the re sults ob tained, the FC b 2 prep a ra tion 1 was cho sen from three tested prepa ra tions as an op ti mal vari ant by dy namic range and sta bil ity for the de vel op ment of amperometric lac tate bio sen sor.Op er a tional and stor age sta bil ity of the trans ducer based on this prep a ra tion is better than of LOD-based bio sen sor.Nev er the less, its ap pli ca tion in wine anal y sis was in ef fi cient: be cause of poor se lec tivity the mea sure ment re sults ex ceeded ac tual val ues of lactate concentration in wine by several times.
The es sen tial non-spe cific sig nals of biosensors with im mo bi lized FC b 2 can be ex plained by in suf ficient se lec tiv ity of FC b 2 or by rel a tively high work ing po ten tial (+450 mV) which con di tions the ox i da tion of in ter fer ing par ti cles and, cor re spond ingly, gen er a tion of the amperometric electrode response.Table 1.

Responses of laboratory prototypes of amperometric biosensors with different FC b 2 preparations to lactate (at saturation) and main interfering substances in wine
Note wor thy, the LOD-based amperometric biosen sor, de vel oped by us, had high se lec tiv ity: it dem onstrated small non-spe cific re sponse only to eth a nol in high con cen tra tions nonrelevant for wines [2].In contrast to the bio sen sor with im mo bi lized FC b 2 , the LOD-based bio sen sor works at lower po ten tial (+200 V), at which an ef fect of in ter fer ing sub stances on biosen sor functioning considerably less.
At the last stage of the re search pH-op ti mum for the bio sen sor with FC b 2 prep a ra tion 1 was de ter mined to be 7.6 (Fig. 5) which well cor re lates with the data ob -tained ear lier [14].Be sides, the value of re sponse was shown to be prac ti cally in de pend ent on the buffer capac ity and ion strength of tested so lu tion (Fig. 6a, 6b), which is char ac ter is tic for amperometric biosensors, in clud ing the LOD-based lactate sensor [16].
Con clu sion.The in ves ti ga tion showed that the FC b 2 prep a ra tion 1 (40% am mo nium sul phate, spe cific activ ity 3 U/mg, pro tein con cen tra tion 4.8 mg/ml), im mobi lized by elec tro chem i cal poly mer iza tion in PDT poly mer, is the most ef fec tive for the de vel op ment of amperometric lac tate bio sen sor based on flavocytochrome b 2 .This bio sen sor dem on strates better op er a tional and stor age sta bil ity as com pares with the sen sor based on lac tate oxidase im mo bi lized in PEDT.On the other hand, the lat ter is ad van ta geous by sen si tiv ity and se lec tiv ity, these pa ram e ters be ing de ci sive to choose a de vice for the anal y sis of wine and must -com plex mix tures, lac tate con cen tra tion in which is ex tremely low.The re sults of lac tate mea surement in wine sam ples by the LOD-based bio sen sor proved its ef fi ciency while those ob tained by the biosen sor with im mo bi lized FC b 2 ap peared to be un re liable.That is why the LOD-based biosensors are consid ered as pref er a ble for the ap pli ca tion in wine making.The FC b 2 based bio sen sor due to its high sta bil ity can be ben e fi cial for the anal y sis of other liq uids, e.g.blood, in which eth a nol is ab sent while lac tate con centra tion is rather high (0.5 -2.2 mÌ [8]) and is within the dy namic range of the sen sor un der con sid er ation.As it was shown in the pre vi ous in ves ti ga tion [13], an es sen tial im prove ment of the se lec tiv ity and sen sitiv ity of FC b 2 -based bio sen sor could be achieved by ap ply ing a lower work ing po ten tial when me di a tor with a low re dox po ten tial and FC b 2 prep a ra tions with a higher spe cific ac tiv ity were used.There fore, we are plan ning to con tinue our investigations in this direction.
This work was financially supported by National Academy of Sciences of Ukraine in the frame of Scientific and Technical Program "Sensors systems for medical-ecological and industrial-technological problems" and STCU project 4378.

Fig. 5 .
Fig. 5. Dependence of response of amperometric biosensor based on FC b 2 (preparation 1), immobilized in polymer PEDT, on pH of working solution.Concentration of lactate added into measurement cell -6.4 mÌ.Measurement in 20 mM phosphate buffer, potential of +450 mV vs reference electrode.

Fig. 6 .
Fig. 6.Dependence of response of amperometric biosensor based on FC b 2 (preparation 1), immobilized polymer PEDT, on concentration of background electrolyte in buffer (a) and concentration of buffer solution (b).Final concentration of lactate in a measurement cell was 6.4 mÌ.Measurement in phosphate buffer, pH 7.6, potential of +450 mV vs reference electrode

Table 2 .
Comparative analysis of laboratory prototypes of amperometric biosensors with different FC b 2 preparations and LOD-based biosensor