Biocatalytic Protein‐Based Materials for Integration into Energy Devices

There is a current need to fabricate new biobased functional materials. Bottom‐up approaches to assemble simple molecular units have shown promise for biomaterial fabrication due to their tunability and versatility for the incorporation of functionalities. Herein, the fabrication of catalytic protein thin films by the entrapment of catalase into protein films composed of a scaffolding protein is demonstrated. Extensive structural and functional characterization of the films provide evidence of the structural integrity, order, stability, catalytic activity, and reusability of the biocatalytic materials. Finally, these functional biomaterials are coupled with piezoelectric disks to fabricate a second generation of bio‐inorganic generators. These devices are capable of producing electricity from renewable fuels through catalase‐driven gas production that mechanically stimulates the piezoelectric material.


Introduction
Extensive efforts are currently being made in the fabrication of new functional biomaterialsa nd their integration into technological devices. [1][2][3][4] Av arietyo fb ioinspired strategies have recently emerged to create biomaterials with tailoredf unctionalities, improved properties, and potential applications. [1,[5][6][7] These strategiesi nclude bottom-up approaches for the generation of supramolecular biomaterials, [2] with particular efforts devotedt op rotein-based materials due to the amazing structural and functional diversity of these biomolecules. [5][6][7][8][9] Of particular interest to the field of active biomaterialsi st he fabrication of catalytic materials in whicha ctive enzymes are immobilized. [10] Severali mmobilization methods have been used, to date, including adsorption, [11] covalent linkage, [12] affinity immobilization, [13] entrapment, [14] encapsulation, [15] and enzyme crosslinking. [16] Among them,s imple entrapment offers easy fabrication and handling, and endows high stability to the immobilized enzymes. [17] Regarding the support, the use of natural materials as scaffolds provides advantages of biocompatibility,c hemical versatility,a nd biodegradability that organic or inorganic support materials do not offer.I nt his regard, proteins play ak ey role because they fulfill all of these characteristics and their use for the fabrication of protein-based materials has already been demonstrated. [9,18] However,t he robust fabrication of functional protein-based materials films that preserve both structural integrity and functionality of the entrapped enzymesr emains ac hallenge in biomaterials research. Few examples of enzymes immobilizedo np rotein-based materials have been successfully reported to date. [19] Several scaffolding proteins can be used for the fabrication of biomaterials. However,t he engineering of proteins provides clear advantages in terms of tunability,reproducibility,scalability,a nd rational design of the final properties of the material. [3] Modularb uildingb locks with simple interactions allow for betterc ontrol of the final assembliest hrough the implementation of bottom-up approaches. Engineered repeat proteins present these inherentf eatures, which makes them suitable candidates for use as scaffolding proteins. [20,21] In particular,tetratricopeptide repeat( TPR) proteins have proven to be exceptional scaffolds because consensus tetratricopeptide repeat (CTPR) proteins form ordered films, due to well-characterized side-to-side and head-to-tail interactions. [21] The basic CTPR unit consists of 34 amino acids with ah elix-turn-helix motif. CTPRs have am odulars tructure, in which the repeatsc an be combined in tandem to form highly stable CTPR proteinst hat display ar ight-handeds uperhelical structure, with eightr epeats per full turn of the superhelix. [22,23] These features make these repeat protein scaffolds ideal buildingb locks for numerous applications, [24,25] as previously demonstrated through the fabrication of functional protein-based materials. [26,27] Similarly, There is ac urrent need to fabricate new biobasedf unctional materials. Bottom-up approaches to assemble simple molecular units have shown promise for biomaterial fabrication due to their tunability and versatility for the incorporation of functionalities. Herein, the fabrication of catalytic protein thin films by the entrapmento fc atalase into protein films composed of as caffoldingp rotein is demonstrated. Extensive structural and functional characterization of the films providee vidence of the structurali ntegrity,o rder,s tability,c atalytic activity,a nd reusability of the biocatalytic materials. Finally,t hese functional biomaterialsa re coupled with piezoelectric disks to fabricatea second generation of bio-inorganicg enerators. These devices are capable of producing electricity from renewable fuels throughc atalase-driven gas productiont hat mechanically stimulates the piezoelectric material.
we hypothesize that materials based on CTPR proteins will be applicable for the ordered entrapmenta nd immobilization of enzymest owardt he generation of novel heterogeneous biocatalysts that can be readily integratedi nto devices.
Catalase (CAT) is an excellent candidate for entrapment into protein-based materials because it is extensively used for technological applications in which biocatalyst robustness is demanded. CATc atalyzes hydrogen peroxide disproportionation to releasew ater and molecular oxygen as innocuous products. [28] Such ac atalytic reaction makes this enzyme industrially relevant for the development of biosensors, [18] preventing food oxidation, [29] removing hydrogen peroxide from residual water, [30] and the intensificationo fo xidase-driven biotransformations. [31,32] In ap ioneering attemptt oe xpand the applications of CAT, our group has recently exploited CATi ns olution to produce bubbles upon H 2 O 2 disproportionation as as ource of mechanicale nergy,w hich is then harvested by coupling to piezoelectric materials to generate electricity. [11] Because the described system integrates as oluble CATt hat forms bubbles within ar eaction chamber,t he reaction mixture and, consequently,t he enzyme must be removed once the fuel (H 2 O 2 )i s consumed and maximal energyi sp roduced.M oreover,t he soluble enzyme normally becomes inactive after one operational cycle upon exposure to the high concentration of H 2 O 2 required fort he energy-generation process. Hence,t he current bio-inorganic generator is limitedt oo nly one cycle of energy generation,whichprecludes the reusability and continuous operation of the system.I nt his emerging filed, enzymesc ould play ak ey role because they are able to catalyze reactions, the byproduct of which may be used as as ource of mechanical energy.However,the coupling of enzymaticsystems and nanogenerators would require automatic,i ndependent,a nd reusable systems to minimize costs and enhancee fficiency.F or the aforementioned reasons,t he physicali mmobilization of CATon the surface of the piezoelectric materialw ould make the enzymem ore robust and recyclable; thus overcoming the major limitations of the current system for energy harvesting and ease of automation and integration of this bio-inorganic generatori nto more complex devices. Although enzyme immobilization is often considered to be one of the moste ffective methods to increaset he robustness of the biocatalyst, preservation of the enzymatic activity in the materials during the immobilizationp rocess is still ac hallenge. [33] This limitation arises mainly from loss of catalytic activity due to enzymes undergoing denaturing conformational changes, intermolecular aggregation, steric hindrance of the active sites, lack of dynamic freedom, and mass transport restriction of reactants imposed by the structure of the solid materials. [34] In addition, depending on the immobilization methodology,t he enzyme might be leached to the reaction media, resulting in activity loss of the immobilized biocatalyst. Hence, the immobilization of enzymes on solid materials should be optimized according to their final use, and thus, require protocols that maintain the activity and increase the stabilityr elative of the enzymei ns olution.
To advance the fabrication of novel bio-inorganic generators based on CAT, herein we describet he development of an ovel protein-based biomaterial that can be readily coupled with piezoelectric materials to increase the operational life span of the energy-harvestingp rocess. This increased life span is possible because the protein environmento ft he CTPR scaffolding protein forming the biomaterial stabilizes entrapped CAT, as demonstrated by thermal inactivation studies. The solids elfstanding biomaterial with CATactivity was casted into ap iezoelectric disk and provent op roduce electricity from H 2 O 2 .A lthoughe ntrapped CATh as al ower apparent catalytic efficiency than that of the enzyme in solution, the proximity of the bubbling enzyme to the piezoelectric surface improves the energy outcomei nl arge operational volumes comparedw ith the system involving use of the enzyme in solution.

Protein-based biomaterial fabrication and characterization
Herein, we report on the fabrication of ab ioactive material based on the self-assembly of CTPR proteins,w hich irreversibly entrapC AT as af unctional unit, and casting onto piezoelectric materials to give rise to an ew generation of energy-harvesting He currently leads the Heterogeneous Biocatalysis Lab at the University of Zaragoza as an ARAID tenured researcher,where he merges enzymes with advanced materials to fabricate the next generation of multifunctional heterogeneous biocatalysts. ChemBioChem 2019ChemBioChem , 20,1977ChemBioChem -1985 www.chembiochem.org 2019 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim devices based on those previously reported ( Figure 1A). [35] The formation of this active biomaterial relies on the self-assembling properties of CTPR andi ts ability to form self-standing orderedp rotein thin films. As previously described by our group, [21] fabrication of the protein thin film requires 3% (w/v) protein concentration of aC TPR10 variant (a CTPR protein with 10 identicalr epeats) to ensurea ppropriate packing and handling of the resulting film. The biocatalytic protein film was prepared under these conditions by homogeneously mixing CTPR10 and CATi ns elected ratios, andt hen casting the film through two different strategies:s pin-coating on ap iece of quartz,o rd rop casting on ah ydrophobic material (see the Experimental Sectionf or further details;F igure 1A,B). The resulting self-standing protein films quantitatively entrappedC AT and led to as table, flexible, easy to handle, and transparent material( Figure 1B). The homogeneity and surface uniformity of the material is evident from the SEM image ( Figure 1C), and the SEM cross-section image revealed at hickness of approximately 20 mmf or these fabrication conditions (Figure1Ca nd Figure S1 in the Supporting Information). As expected, the functionalized protein film showed ap eak at around 400 nm that corresponded to the maximum absorption of the heme group of CAT, which resulted in the light green color of the film ( Figure 1D). The CD spectrumo ft he protein film entrapping CATshows ac haracteristics ignal of the a-helicals tructure of the scaffolding protein that agreesw ith resultsp reviously observed by our group for which the secondary structure of CTPR10 is preserved. [21] CD datao nly reveal the secondary structure of the CTPR because the protein stoichiometry in the film is 1:100 CAT/CTPR; thus the contribution of CATt ot he CD spectrum is negligible ( Figure 1E).
It is knownt hat CTPR proteins present intrinsic properties to form nanostructured ordered films because they macroscopically align through head-to-tail and side-to-side interactions. [22,36] However,t he mobility and order of entrapped enzymesi nt he biomaterial is unknown. To shine light on these two parameters, fluorescencep olarization studies were performed by first labeling the amine groups of CATwith the fluorescent dye rhodamine (Rh). Using confocal fluorescence microscopy,w ed emonstrated that entrapped CATw as homogeneously distributed across the protein film ( Figure 2A ChemBioChem 2019ChemBioChem , 20,1977ChemBioChem -1985 www.chembiochem.org it was observed that the CAT-Rh conjugate in solution showed ah igher anisotropy than that of free Rh, as expected from its larger mass. Interestingly,R h-labeled CATe ntrappedi nt he solid protein film presented al arge anisotropy value, which indicatedt hat the confined enzymew as significantly less mobile than that of its counterpart in solution.N ext, the organization of CATw ithin the protein film was determined by measuring the fluorescenceo ft he Rh-labeled enzyme with plane polarized light. The fluorescencei ntensity changed as the excitation and emission polarizers werem ove from horizontal to vertical. These changes indicate that labeled CATe ntrapped into the protein film exhibits ad irectional macroscopic order within the material; this order is imposed by the anisotropic arrangement of the CTPR scaffolding protein ( Figure 2B).
The ordered self-assembled material, as described above, is soluble in aqueous media because it is formed mainly due to noncovalent interactions between CTPR proteins, which are too weak to maintain the integrity of the macrostructure. For operational purposes, the disassembly of the film andl eaching of the entrapped protein need to be minimized. Therefore, the biomaterial was crosslinked with 1% glutaraldehyde (GA) for 24 hb ym eans of vapor diffusiont or esult in am aterial that was stable in water.U pon crosslinking,t he biomaterial was immersedi nabuffer solution and showedn ol eachingo ft he protein into the medium; more than 95 %p rotein was retained 7days after immersion, whichi ndicated effective crosslinking ( Figure 3A). Moreover,t he mild crosslinking reactiond id not affect the structural properties of the protein, as shown in the CD spectra;t hus the material is expected to preserve its activity after the reaction ( Figure 3B).

Activityand stability of CATentrappedint he CTPR protein film
The functionalproperties (activity and stability) of CATimmobilized on the protein-based materiald escribed above wered etermined through an indirectc olorimetric assay by using peroxidaset od etermine the concentration of H 2 O 2 that remains after the action of CAT.
The enzymatic activity was assayeda tr oom temperature for the biomaterial and the enzymatic solutionu nder the same conditions. Swelling of the film was observed upon immersion into the aqueous solution:a0.5 mg film absorbed approximately 1 mLo fw ater.T he apparent specific activity of entrapped CATwas 0.08 Umg À1 ,c ompared with 6.7 Umg À1 observed for the enzyme in solution under the same experimental conditions. Moreover, the activity of films with different thicknesses was evaluated at ac onstant CTPR/CATr atio. These results showed that the total CATa ctivity was maximized at as tandard thickness of 20 mm; thicker films showedl ower activity and significantly lower specific activity ( Figure S4).A dditionally, the apparent kinetic parameters (K M and V max )o ft he immobilized enzyme were determined through aM ichaelis-Menten fit ( Figure S5) and compared with the values calculated fort he soluble enzyme under the same conditions. Ta ble 2s hows that the apparent K M and k cat fore ntrapped CATt owardsH 2 O 2 were 3.6-fold higher and 370 times lower than those for the soluble enzyme;t hus confinement andc rosslinking of CATw ithin the protein biofilm decreased the catalytic efficiencyb y   ChemBioChem 2019, 20,1977 -1985 www.chembiochem.org 2019 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim 1519 times. According to the specific activity and apparent kinetic parameters of entrapped CAT, the crowded environment of the protein film may limit the conformational flexibility of entrapped CATt op erform catalysis. Moreover,t he highera pparent K M values of entrapped CATs uggest mass transport issues that may hamper the diffusion of substrates towards the enzyme active sites. These issues can be relatedt o1 )enzyme entrapmenti nt he compact protein film, which makes it less accessible to the medium; [37] and/or 2) bubblesg enerated on the surface of the materialb yo xygen released during the reaction hamper substrate diffusion. [38] These bubbles grow over the surface and,d ue to the hydrophilic natureo ft he protein material, stay at the film-solution interphase;t hus hampering the diffusion of the substrate. In addition, this change in activity could be partially attributed to the loss of catalytic activity due to immobilization and crosslinking processes. Notably,t he time-course curves show al ag time of around 400-500 si fC AT is entrapped,w hereas that lag is not appreciable in the time course of the soluble enzyme ( Figures S6 and S7). Mass transfer issues are also supported by loading experiments, in which the specific activity of the entrapped enzymel inearly increases with the increase in the ratio of CTPR/CAT( Figure S8). To tal apparenta ctivities of films with different ratios are similar, but the specific activity of films with higher CATl oadings is dramatically reduced. It seems that, under high CATl oading conditions, substrate molecules cannot reacha ll entrapped enzyme molecules due to transport limitations, rather than conformational changes related to entrapmento rc rosslinking. Ap opulation of inaccessible enzyme molecules may explain the low catalytic efficiency observed for entrappedC AT.L ower CATl oadings ameliorate this effect because al arger enzyme population actively participates in H 2 O 2 disproportionation because of improved substrate accessibility.T hese diffusional restrictionsh ave also been observed if CATi sr andomly aggregated and irreversibly crosslinked in the presence of af eeder protein through crosslinked enzymea ggregate (CLEA) technology. [39] Unlike entrapmenti nto thin protein films, CLEAs of CAT have not provent he order of the enzymew ithin the solid material, as demonstrated in this work by using scaffolding proteins. Despite exhibiting low catalytic efficiency,e ntrapped CATi s still activee nough for furthera pplications; thus illustrating the potentiala nd biocompatibility of this simple immobilization strategy. To assess the robustness of the functional biomaterial, we first tested the thermals tabilityo fe ntrapped CATi nc omparisonw ith the soluble enzyme. Figure 4s hows that immobilized CATw as significantly more thermostablet han its soluble counterpart because it retained approximately 100 %o fi ts initial activity after 3h incubation at 50 8C, whereas the soluble enzymel ost 70 %o fi ts initial activity under the same inactivation conditions, which represents as tabilization factor of 3.5. According to other immobilization protocols, CATcould be stabilized by entrapment in CTPR protein thin films;t hus demonstratingi ts potential as ah eterogeneous biocatalyst that can be reused for severalo perational cycles. [40] Energyo utputmeasurements Recently,o ur group described the production of electrical energy from chemical energy by using bio-inorganicg enerators. [35] Such devices are able to convert chemical energy stored in renewable molecules (as carbohydrates, aminoa cids, alcohols, etc.) into mechanical energy (in the form of gas bubbles), which are furtherh arvested by ap iezoelectricm aterial based on al ead zirconate titanate (PZT) disk and transduced into open-circuitv oltages. That energy output wasm onitored in line by using an oscilloscope. This concept has been proven with soluble CAT, which disproportionates hydrogen peroxide into oxygen and water,f orming oxygen bubbles that trigger a mechanical stimulus for further harvesting by the piezoelectric material. The drawback of the reported system is its disposability,w hich prevents enzyme reutilization after one energy-generationc ycle. To make these systems reusable, protein films entrapping CATw erec ast on the piezoelectric surfacec oated with an anometer-thick inorganic layer,w ith the aim of fabricating bio-inorganic generators that could be repeatedly used for the continuous production of electrical energy to achievea more sustainable process.
Hence, we entrapped soluble CATi nto aC TPR film that was subsequently crosslinked with GA, as above described. Primarily,weassessed the effect of coating of the piezoelectric surface with aC TPR protein layer on the energy outcome. Using the soluble enzymea sacatalyst, we observed that the energy generated was reduced by only 25 %w ith regard to the naked piezoelectric surface; thus demonstrating that the biomaterial  ChemBioChem 2019, 20,1977 -1985 www.chembiochem.org 2019 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim layer slightly affects the performance of the piezoelectric surface, but still allows the conversion of mechanical into electrical energy.I fC AT was entrapped in the protein film, the energy outcome( at 750 mm H 2 O 2 )w as 70 nJ cm À2 ;avalue 3 and 1.6 times lower than that of the soluble enzymea cting over the naked piezo surface and the piezo surface casted with the protein film, respectively ( Figure 5A). The optimal fuel concentration for the system with the soluble enzyme was 50 mm H 2 O 2 ,w hereas the system with the immobilized enzyme required 750 mm H 2 O 2 .I ns pite of the lower output and higherd emand for fuel, the entrapment of CATenables its repeated use, maintaining more than 50 %o ft he initial energy production after 10 energy generation cycles,b yr eplacing the fuel (H 2 O 2 )a nd rinsing the reaction chamber after every use; this is not possible with soluble CAT( Figure 5B). These results demonstrate that CATr emains active in the protein film and can effectively convert chemical energy into an electrical output.
Although the bio-inorganicg enerator functionalized with entrapped CATg enerates three times less energy than that of the previously reporteds ystem with soluble CATi nar eaction volumeo f0 .2 mL, this effect is ameliorated if the reaction volumei nt he chamber is increased ( Figure 5C). Similar electrical powero utput responses are reached upon workingw ith 1mL( 5nJcm À2 ), whereas for the immobilized enzyme in ar eaction volume of 1.5 mL, twice as much energy was generated than that with the soluble enzyme workingu nder the same conditions. This effect suggests that immobilized CATc onsumes all fuel, and therefore, generates all of the bubbles (mechanical energy) at the interface with the piezoelectric surface; thus maximizing energy harvesting. This fact explains why more fuel (larger reactionv olume) resultsi nah igher electrical output.O nt he contrary,i fC AT is freely diffusing through the reactionv olume, part of the mechanical energy is noth arvested because the vast majority of the bubbles are produced far from the piezoelectric surface. This effect explains the ob-served plateau for energy generation ati ncreased reaction volumes for the soluble enzyme. Therefore, future bio-inorganic generatord esignsm ust contemplate major enzyme-piezoelectric component contact across the entire chamber to maximize harvesting of mechanical energy.T oo ptimizet his new generation of bio-inorganicg enerators, different CATc oncentrations were entrapped into CTPR protein films and casted on the piezoelectric surface ( Figure 6A). Bio-inorganicg enerators with immobilized CATg enerate higher electric powero utputsa s the immobilized CATl oading increases untilr eaching saturation conditions at 60 mgofimmobilized CAT. Finally,westudied the response of bio-inorganic generators (with 30 mgo fi mmobilized CAT) for different H 2 O 2 concentrations. Figure 6B shows that ah igherH 2 O 2 concentration resulted in ag reater energy  In all cases, aS iO 2 -coated piezoelectric surfacew as employed.7 10 mg of CTPR protein were used for immobilized CAT. The reaction mixture consists of 1.5 mL of 50 mm H 2 O 2 in 50 mm sodiump hosphate at pH 7. B) Effect of hydrogen peroxideconcentration on the productiono fe lectrical energy by bio-inorganic generators with immobilized enzyme.Ina ll cases, aS iO 2coated piezoelectric surface was employed. 710 mgo fC TPR protein were used for immobilized CAT(30 mg). The reaction mixture consisted of 1.5 mL of hydrogen peroxideatt he indicated concentrationi n50 mm sodium phosphate at pH 7. ChemBioChem 2019ChemBioChem , 20,1977ChemBioChem -1985 www.chembiochem.org output.T he maximume lectrical output was found at 750 mm H 2 O 2 ;t his meanst hat the performance of the system is optimal under those conditions which generate am aximum energy of 76 nJ cm À2 .

Conclusion
We have reported as imple methodology fort he immobilization of enzymes in solid protein-based biomaterials that could be cast into devices, such as bio-inorganic generators for electricity production. The developed methodology has been provenw ith CATa samodel enzyme. Herein, we described a simple protocol that involved only mixing and drop casting of the target enzymew ith as caffolding protein, followed by a mild crosslinking procedure, to yield ar obusta nd functional biomaterial. The biomaterial showed the same intrinsic properties as those of protein films generated only with the scaffolding protein. [27] Upon completiono ft he fabrication process,e ntrapped CATm aintained its functionality and displayedm acroscopic order within the biomaterial. This functional biomaterial was active and successfully integrated into bio-inorganicg enerators that converted chemical energy into electricity.T he device, as previously described for CATi ns olution, [11] was based on the conversion of chemical energy from the reaction catalyzed by CATi nto mechanical energy associated with the production of oxygen bubbles and the downstream harvesting of this mechanical energy by ap iezoelectric material to produce an electrical output as an open-circuit voltage. We have demonstrated the effective functionalization of the piezoelectric surface with ab iocatalytic material to advance the fabrication of bio-inorganicg enerators based on an ovel concepti n which chemical energy was transformed into electricity through mechanical energy harvesting. This biomaterial allowed the reusability of the device, whichw as otherwise impossible fort he soluble enzyme,a lthough energy harvesting was not as efficient as that of previously reporteds ystem with CATi ns olution.I na ddition, the amount of CATe ntrappedi n the protein film was ac rucial parameter to tune the electrical power output of the bio-inorganic generators developed herein. Remarkably,w ef ound that the electrical power output of bio-inorganic generators with immobilizedC AT was maximized upon increasing the reactionv olume, withoutr eaching as aturation point,u nlike the system with soluble CAT, which becames aturated with sub-milliliter reactionv olumes. We envision the potential of this technology to advance the fabrication of more robust bio-inorganic generatorsi nw hich the bioactive phase is in close contact with the piezoelectric transducer.F urthermore, the application of the describedb iocatalytic protein-based biomaterialsc an be expanded to the field of heterogeneous biocatalysis to improve chemicalm anufacturing, as well to the field of biosensing to develop more sensitive devices.
Protein expression and purification:T he gene encoding CTPR10 protein was previously generated based on aC TPR protein and cloned into the pPROEX-HTav ector for expression as aH is-tag fusion protein for affinity purification. [41,42] The protein was expressed and purified as described previously. [23] Briefly,t he plasmid was transformed into BL21(DE3) Escherichia coli and the cells were grown in lysogeny broth with ampicillin (0.1 mg mL À1 )u nder agitation to an O.D. between 0.6 and 0.8. Protein expression was induced with isopropyl-b-d-thiogalactopyranoside (IPTG;0 .6 mm)f or 5h at 30 8C, then the cells were centrifuged at 5000 g and resuspended in 300 mm NaCl, Tris lysis buffer (50 mm,1 5mL, pH 8.0) with 1mgmL À1 of lysozyme, 5mm b-mercaptoethanol, and DNase stock solution. The resulting lysate was sonicated for 5min with 30 si ntervals and 40 %a mplitude, and centrifuged at 10 000 g for 45 min. Protein purification was performed by means of affinity chromatography by using aN i 2 + His-Trap column. The eluted protein was dialyzed overnight in phosphate-buffered saline (PBS; 150 mm NaCl, 50 mm phosphate buffer pH 7.4 with 2.5 mm b-mercaptoethanol). The protein was concentrated and purified by means of FPLC gel filtration chromatography over aS uperdex 75 HiLoad column. Fractions containing the protein were analyzed in 15 %a crylamide gels to confirm the purity of the protein. Finally, the protein was concentrated to the desired concentration, from 50 to 400 mm,b yu sing the estimated molar extinction coefficient at l = 280 nm from the amino acid composition.
Protein thin-film fabrication:Aprotein solution (20 mL) of CTPR10 (400 mm;1 8.1 mg mL À1 )a nd 4 mm CAT( 1mgmL À1 )w as deposited over ah ydrophobic nonporous material by means of drop casting and left to dry for at least 4h at room temperature to ensure formation of the protein thin film. As econd approach to obtain thinner films was to use spin-coating to deposit the films. In this case, ad rop (15 mL) of as olution containing 400 mm CTPR10 and 4 mm CATw as deposited on aq uarz slide of 1 1m m 2 through as pincoating method by using aL aurell Te chnologies corporation model WS-400B-6NPP/LITE instrument, at 1000 rpm for 10 min with an acceleration of 3000 rpm s À1 .AL EICA S8APO stereomicroscope was used to image the biomaterial at the macroscopic level.
SEM imaging:AJEOL JSM-6490LVs canning electron microscope was used to image the surface of the protein thin film. The film was mounted on ac arbon tape and imaged under vacuum conditions by applying an electron high tension (EHT) of 5.00 kV,w orking distance (WD) of 2.5 mm, and an aperture size of 15 mm. Sputter coating was performed on all samples by using am etallization of Au/Pd alloy.
CD measurements:C Dw as used to determine the secondary structure of the CTPR10 units within the films by using aJ asco J-815 spectropolarimeter.S olid films were deposited on as andwich quartz cuvette (0.1 mm path length) through spin coating. CD spectra were acquired at 1nmi ncrements and 10 sa verage time over awavelength range of 190 to 260 nm.
Crosslinking of the protein thin films:G Aa taconcentration of 1% was used as crosslinking agent for gentle vapor diffusion crosslinking. [43,44] The reaction was performed in wells of 1mLf or 24 h at room temperature. At the bottom of each well, as olution of GA (500 mL) was added and the protein film was fixed on the coverslip used to seal the well. After the reaction, the films were recovered ChemBioChem 2019ChemBioChem , 20,1977ChemBioChem -1985 www.chembiochem.org 2019 Wiley-VCH Verlag GmbH &Co. KGaA, Weinheim and, to evaluate the crosslinking efficiency, dipped into an aqueous solution to monitor potential release into the solvent. The biomaterial was dipped into an aqueous solution for different times of 1h,2 4h,a nd one week, and the amount of protein in the supernatant was quantified for each time by using the Bradford assay on aV arioskan Flash spectrophotometer with as canning multimode reader (Thermo Scientific).
Distribution of CATi nt hin films determined by means of fluorescence imaging and fluorescence anisotropy:C AT was fluorescently labeled with Rh Bi sothiocyanate (Sigma-Aldrich). The labeling reaction was performed for 24 ha t3 78Cu nder constant shaking. The enzyme was purified from the free dye by using aN AP-5 column (GE Healthcare) and concentrated with an Amicon ultra-0.5 mL centrifugal filter MWCO 10 K( Merck Millipore). The enzyme concentration was calculated by measuring the absorbance at l = 280 nm (e 280 nm = 246 000 m À1 cm À1 )o naUV/Vis spectrophotometer (ThermoFisher). Fluorescence confocal microscopy (ZeissNLO 880) was performed to determine the homogeneity of the CAT-Rh-labeled biomaterial. The image was acquired at an excitation wavelength of 550 nm with am agnification of 20 .T he z stack for the 3D reconstruction was acquired with a40 oil objective at an excitation wavelength of 561 nm and emission wavelengths between 565 and 700 nm, 40 mmZ -slides, and an Airyscan detector at maximum intensity projection and 512 512 frames. Fluorescence anisotropy of the solid film was measured by using aP erkinElmer (LS55) fluorimeter with automated polarizers. Af ilm of CTPR10 and CATl abeled with Rh was used and polarization was determined at excitation and emission wavelengths of 550 and 572 nm, respectively.
Enzymatic activity measurements:T he CATactivity either in solution or in the protein thin film was indirectly analyzed by quantifying the amount of hydrogen peroxide produced over time. Samples of the CATr eaction were withdrawn at different times and incubated with peroxidase and ABTS (e = 36 000 m À1 cm À1 )f or af ixed time. HRP used the CAT-produced H 2 O 2 to oxidize ABTS, which increased the absorbance of the reaction mixture at l = 420 nm. The increase in absorbance was measured by using aV arioskan Flash spectral scanning multimode reader (Thermo Scientific) and quantified to determine the H 2 O 2 concentration upon the action of CAT. For the soluble enzyme, a0 .01 mg mL À1 solution of the enzyme (20 mL) was mixed with as olution of substrate (980 mL; 35 mm H 2 O 2 in 100 mm phosphate pH 7.4);a liquots (50 mL) were collected at different times and heated at 90 8Cf or 2min to inactivate the enzyme. Upon enzyme inactivation, each sample (20 mL) was mixed with ABTS (200 mL, 0.066 mg mL À1 )a nd HRP (200 mL, 0.013 mg mL À1 )i np hosphate buffer (100 mm,p H7.4) at room temperature and incubated for 5min. In the case of immobilized CAT, as ample of supernatant (20 mL) from the film incubated with the substrate solution was mixed with ABTS (200 mL, 0.066 mg mL À1 ) and HRP (0.013 mg mL À1 )i np hosphate buffer (100 mm,p H7.4) at room temperature and incubated for 5min. Hydrogen peroxide consumed in the reaction was calculated by using ac alibration curve. The specific activities of both CATins olution and entrapped CATw ere determined from the time-course curves obtained through monitoring the H 2 O 2 concentration over time ( Figure S3). The initial reaction rates were obtained from the slope of the curve, considering zero-order kinetics, and converted into specific activity by considering the total amount of enzyme tested. The kinetic parameters of immobilized CATand that in solution were determined by measuring the reaction rates at various H 2 O 2 substrate concentrations, ranging from 0.25 to 168 mm at pH 7.4, while keeping the amount of enzyme constant. The kinetic parameters K M and V max were calculated from Michaelis-Menten fitting. To determine the thermal stability of the enzymatic activity,i ncubations at 50 8Cf or different times were performed to measure the decay in activity.T he activity of each sample withdrawn at each inactivation time was measured through the HRP-coupled colorimetric assay described above, and fixing the CATreaction times to 15 and 3min for immobilized and soluble enzyme, respectively.
Electrical energy output measurements:S iO 2 thins films were deposited through nonreactive and reactive magnetron sputtering on PZT disk (1.25 cm 2 active piezoelectric surface) by using an AJA-ATC1 800 system with ab ase pressure of 10 À7 Pa. Deposition of the films was performed with three separate 2inch elemental targets, with ap urity of 99.999 %f or carbon (Demaco-Holland) and 99.95 %f or Nb (AJA International-USA) and SiO 2 ,i naconfocal configuration at ap ressure of 0.25 Pa of pure Ar.T he substrate bias voltage and substrate holder heating facility were turned off during depositions;t he distance between target and substrates was about 15 cm. Prior to deposition, the substrates were sputtercleaned with an egative bias of 180 V( 25 W) in a4Pa Ar atmosphere for 10 min. SiO 2 films were deposited in an Ar/O 2 atmosphere (10 seem Ar + 20 seem O 2 )attotal pressure of 0.4 Pa and applying ad irect current (DC) power of 230 Wt ot he Si target. CATembedded protein thin films were deposited on SiO 2 on the PZT surfaces and connected to an oscilloscope (Siglent model SHS806) to monitor the open-circuit voltage versus time, as described previously. [35] Different chamber volumes were obtained by varying the liquid fuel volume inside the open chamber (0.2, 0.4, 0.6, 1, 1.5, and 2mL). To assess the enzyme in solution, as olution that contained CATin5 0mm sodium phosphate buffer at pH 7w as placed inside the chamber.T he reaction was initiated by the addition of the fuel, H 2 O 2 ,a tt he indicated concentrations. Before triggering the reaction, the system was equilibrated until the voltage signal reached 0V .T he corresponding energy of the electrical output was calculated by using Equation (1): [45] E ¼ in which E is the generated electrical energy and V is the generated voltage from the start (t 1 )t ot he end (t 2 )o facycle at ac onstant resistance load (R). The R value was fixed to 60 MW (experimentally determined at the maximum produced voltage by the enzymes).