Skip to main content

Optimized esterase biocatalysts for cost-effective industrial production

Final Report Summary - OPTIBIOCAT (Optimized esterase biocatalysts for cost-effective industrial production)

Executive Summary:
OPTIBIOCAT developed bioconversions characterized by reduced environmental impact for replacing chemical processes currently adopted for production of ingredients for the cosmetic industry . The OPTIBIOCAT bioconversions represent low carbon energy technologies with improved energy and resource efficiency. The OPTIBIOCAT biocatalysts are based on the enzymes feruloyl esterases (FAEs) and glucuronoyl esterases (GEs). A summary of the main results of each work-package (WP) is reported as following:
➢ WP2: Identification of 1,636 putative fungal FAE and 166 putative fungal GE protein sequences was carried out by fungal genome mining. The putative enzymes were used for construction of phylogenetic trees and gene model correction was performed on 52 putative FAEs and 20 putative GEs. In addition, the identification of 500 bacterial FAE gene sequences was also performed by genome mining. Over 500 putative FAE and GE encoding genes were codon optimized for the gene synthesis and the recombinant production in either prokaryotic or eukaryotic hosts. Furthermore, 4 FAE encoding genes in DNL’s C1 were annotated and 3 other proprietary fungal genomes were mined and 3 esterase coding sequences for modification were selected for synthesis and cloning. Homology modeling, active site determination and modeling of surface charge were performed on 2 FAEs and 1 GE. The 5 most promising mutants for each enzyme were selected for further characterization in WP3. An alternative approach based on a rational selection was also followed. To identify species with different properties from the ones already available, 6 fungi were selected for genome and trascriptome sequencing: de novo assembly, gene finding and annotation of putative FAE and GE coding sequences. A database was created to host the data generated in OPTIBIOCAT (sequences, biochemical data, application data) with detailed information ranging from biological aspects like gene name, organism, DNA/protein sequences to chemical properties like pH stability and solvent types.
➢ WP3: Improvement of microbial FAEs and GEs was obtained by using several parallel strategies. Previously characterized and new putative fungal and bacterial FAE and GE encoding genes were phylogenetically grouped and selected for heterologous expression in several production hosts. Over 50 fungal and 500 bacterial FAEs and GEs were produced and biochemically characterized. Rational mutagenesis was used to improve thermal stability of novel fungal esterases and directed evolution was applied to improve the biochemical properties of characterized fungal FAEs. By using these approaches several active, thermo- and solvent tolerant FAEs and GEs were developed.
➢ WP4: Following the preliminary characterization, the enzymes which showed most interesting activities were overproduced for further investigation. Three microbial hosts were used for overproduction of selected enzymes namely Myceliophthora thermophila C1, Pichia pastoris and Escherichia coli using proprietary expression vectors. Elevated expression levels were achieved by strain improvement and/or process optimization for 25 enzymes in Pichia and 7 in C1, all fungal, of which 12 were further overproduced in bioreactors. In all cases, production levels were not a limitation for the activities of the project.
➢ WP5: The synthesis of substrates for directed evolution and enzymatic transesterification was achieved. 44 fungal and bacterial enzymes were evaluated for their synthetic ability. The 8 most promising enzymes were used as catalysts for optimization of reaction conditions in a variety of solvents. A library of 310 compounds was built via enzymatic transesterification with high achieved purity of products. 60 library compounds were tested for antioxidant activity and cytotoxicity indicating a variety of potent antioxidants. A whole integrated industrial process of novel antioxidant production was assessed considering a full industrial scale scenario.
➢ WP6: enzymes and bioactive compounds identified in other WP were taken forward and their potential as industrial biocatalysts and new ingredients for cosmetic applications was evaluated. In particular scale up and optimization of the fermentation process for production of enzymes and the bioconversion process for production of compounds were performed to demonstrate the ability of the new biocatalysts to work in conditions mimicking the industrial ones. Also the safety profile of new compounds was studied and a technical and economic model to analyze large scale production processes was provide.
➢ WP7: Communication, dissemination and exploitation are key to a successful research project. Within OPTIBIOCAT, a plan for communication and dissemination activities was developed and implemented. Three successful workshops and one final conference were organized to raise awareness and interest for the subjected-matter of the project. Several newsletters, continued social media activity and a project film complemented the dissemination activities. The research conducted is exploited in different fields: companies are exploiting the expertise developed and the proof of their technology to support the sale or production of their commercial products and academia is exploiting the expertise developed by publishing in high quality peer reviewed journals.
Project Context and Objectives:
“The European cosmetics industry is a world leader and dominant cosmetics exporter, a highly innovative sector and a significant employer in Europe....Today's cosmetic market is driven by innovation including new colour pallets, treatments targeted to specific skin types and unique formulas concentrating on different needs. Most cosmetics products have a lifespan of less than five years and manufacturers reformulate 25% of their products every year. They need to improve products constantly in order to stay ahead in a highly competitive market where more choice and ever greater efficacy are expected by the consumer”
As established by the EU Strategy for Key Enabling Technologies (EC COM(2012) 341), Key Enabling Technologies (KETs) are a key source of innovation. They provide indispensable technology bricks that enable a wide range of product applications, including those required for developing low carbon energy technologies, improving energy and resource efficiency, boosting the fight against climate change or allowing for healthy ageing. Industrial biotechnology has been recognized as the driving KET for the bioeconomy. In fact, significant growth of the European bioeconomy is expected to be arisen from industrial biotechnology due to its capacity to lead to new bio-based industries, transform existing ones, and open new markets for bio-based products (EC COM(2012) 60: “Innovating for Sustainable Growth: A Bioeconomy for Europe”). It is estimated that every euro invested into research and innovation in the area of industrial biotechnology will result in a tenfold return. The enzymes represent one of the market segments chosen by EC in the Lead Market Initiative on Bio-based products (EC COM(2007) 860).
OPTIBIOCAT was aimed at replacing chemical processes currently adopted for production of ingredients for the cosmetic industry with bioconversions characterized by reduced environmental impact. The OPTIBIOCAT bioconversions represent low carbon energy technologies with improved energy and resource efficiency. The OPTIBIOCAT biocatalysts are based on the enzymes feruloyl esterases (FAEs) and glucuronoyl esterases (GEs).
The goal of OPTIBIOCAT was to develop competitive bioconversions based on esterification reactions catalyzed by FAEs and GEs , for production of molecules with antioxidant activity belonging to the classes of:
• phenolic fatty esters obtained by esterification of hydroxycinnamic (ferulic, p-coumaric, caffeic,sinapinic) acids having antioxidant properties with aliphatic alcohols, to make them more lipophilic and improve their application in oil-based processes

• sugar esters, resulting from the esterification of sugars with antioxidants such as cinnamyl and benzyl alcohols or hydroxyphenyl alcohols in non-conventional reaction media giving these molecules more hydrophilic compared to the starting antioxidant compounds.
The specific scientific and technical objectives and indicators of OPTIBIOCAT are:
o Development of an impressive portfolio of novel enzymes and methods for biocatalysts production optimization and an inventory of well characterized new enzymes:
➢ 50 novel esterases from fungi
➢ 500 novel esterases from bacteria
➢ 25 rationally designed mutants
➢ 20 best directed evolved mutants
More than 550 novel esterases (30 novel fungal candidate FAEs and 20 novel fungal candidate GEs and 500 putative bacterial FAEs) whose sequences were identified from fungal and bacteria genomes by bioinformatic analyses and alignment with sequences of known enzymes and recombinantly expressed in Pichia pastoris, for fungal, and Escherichia coli for bacterial enzymes; more than 20 rationally designed mutants of FAEs and GEs prepared both from new enzymes discovered in the project and previously availbale enzymes.
The portfolio of enzymes developed within OPTIBIOCAT also includes libraries of more than 60,000 variants of directed evolution of three known FAEs recombinantly expressed in the yeasts Yarrowia lipolytica and Saccharomyces cerevisiae that were screened for their hydrolytic activity towards two different substates developed within the project and more than 10 best directed evolved mutants that were selected from these libraries for their highest activity and highest thermo and solvent-resistance . These libraries of thousands of enzyme variants represent a source of the Optiobiocat partners since they can be further screened for other properties related to applications behind the scope of the project.
The sequencing of genomes of six fungal strains, selected based on their physiologic characteristics and/or functional properties, was also performed and transcripts of five of these were also sequenced. The annotation of the genome sequences and their analyses allowed to identify new FAEs and GEs and enzymes involved in the biomasses degradation.
The consortium also worked to improve the efficiency of fermentation/production and stabilize both the enzyme formulations and the life cycle of the biocatalysts. In order to select optimised biocatalysts, the new enzymes were subjected to analyses of thermo-resistance and resistance to solvents.
In order to demonstrate the applicability of the new biocatalysts on an industrial scale, the best enzyme candidates were subjected to optimisation of production by expression in appropriate yeast (Pichia pastoris) and fungal (the proprietary fungal strain Myceliophthora thermophila C1) strains with improved molecular systems. Moreover, production of some selected new enzymes was performed in fermenters and the best operative coditions were investigated to improve yield and productivity of their production. Scale-up of production beyond 1L to 20 L was performed for the best enzyme candidates.
o Optimization of conditions for synthetic reactions
Analyses of synthetic capabilities of around 40 selected biocatalysts for production of more than 20 vinyl esters were performed using both free and immobilized forms. The optimization of reaction conditions i.e water content, donor concentration, acceptor concentration, enzyme concentration, pH, temperature, time and agitation was perfomed to test some selected enzymes.

o Library generation: an inventory of new characterized compounds synthesised by the OPTIBIOCAT enzymatically catalyzed reactions
The two enzymes, C1FaeA1 and C1FaeB2 (expressed in the proprietary fungal strain Myceliophthora thermophila C1), were applied to compound library generation and total of 309 library compounds were generated in the last 2 ½ years.

o Biological active compounds: antioxidant activity for cosmetics
The compounds chosen to be tested are Prenylferulate, Prenylcaffeate, L-Arabinoseferulate, 3 -(3,4,5-Trihydroxyphenyl)-acrylic acid 3-methyl-but-2-enyl ester, 3,4,5-Trihydroxy-benzoic acid 3-methyl-but-2-enyl-ester and 2-[2-(2-Hydroxyethoxy)-ethoxy]ethyl caffeate were tested with in vitro Epiderm Skin Irritation Test (EPI-200-SIT) at the non-cytotoxic and antioxidant- competent concentration of 20μM following the manufacturer’s instructions. According to the EU and GHS classification, none of the aforementioned compounds exhibited skin irritancy potential as the mean tissue viability exposed to each test compound was greater than 50% (in vitro result).. Based on the in vitro results the in vivo prediction for each test compound is non-irritant (NI).

o Techno-economic viability of the developed OPTIBIOCAT processes with demonstration of a significant improvement of the economic efficiency and environmental performance of existing and future biorefineries

Project Results:
➢ WP2:
DNL (Partner n.12)
DNL investigated the genome of its proprietary fungus Myceliophthora thermophila C1 for identification of putative FAEs and GEs. These C1 putative enzymes were inserted in a wider phylogenetic analysis and based on the distribution of the C1 FAE’s along the phylogenetic tree, four C1 enzymes were selected for further study, namely FaeA1, FaeB3, FaeB4, and Fae1.
Furthermore, DNL sequenced 3 additional genomes and mined them for putative GE and FAE genes. Gene functions were inferred from high sequence similarity with known, characterized FAEs.
All DNL esterase sequences were included in a new alignment along with other fungal sequences used in OPTIBIOCAT. The alignment was used to construct a phylogenetic three using the Neighbor-Joining method. Based on the phylogenetic analyses, three new non-C1 FAEs were selected for cloning and further study, namely FAE64, FAE125 and FAE7262. The coding sequences were modified in silico for C1 codon usage and restriction sites were incorporated for cloning. Resulting genes were synthetized.
KNAW (Partner n.8) UNINA (Partner n.1) and UH (Partner n.14)
KNAW, performed fungal genome mining for FAEs and GEs by using characterized fungal FAEs (26) and GEs (16) as queries in a blast search against an internal fungal genome database at KNAW, containing appr. 300 fungal genomes. The fungal identified sequences,shared among UH , UNINA and KNAW were used for automated phylogeny analysis by KNAW followed by manual curation with collaboration of UNINA and UH. This was performed by removal of
- gene models that were incomplete, contained unusually large introns or had large deletions
- gene models with high homology to putative acetyl xylan esterases, lipases, tannases, or hydrolases
- duplicate gene models
This analysis resulted in the identification of 1,636 putative FAE and 166 putative GE protein sequences. Final phylogenetic trees were constructed using the amino acid sequences of both FAE and GE gene sets, which separated the FAEs and GEs into 12 and 8 subgroups, respectively. Among the obtained putative protein sequences, 20 putative GEs and 30 putative FAEs were corrected and sent to NZYtech for gene synthesis.
In addition KNAW also performed the identification of 500 bacterial FAE gene sequences by genome mining and sent to NZYtech for gene synthesis.

According to the outputs of enzyme characterization performed in task 3.1 two fungal FAEs (FAE799: SF1, Aspergillus sydowii; FAE801: SF2.3 Aspergillus wentii) and one fungal GE (GE717: Gr1, Dichomitus squalens) showing the highest specific activity and productivity of the enzymes, were selected for site-directed mutagenesis.
KNAW (partner n. 8)
Homology models for each of the selected enzymes were generated and based on amino acid analysis/alignment and algorithms for predicting structural changes, target sites suitable for mutagenesis have been chosen. This choice focused on thermal stability. The five most promising mutants for each enzyme were selected for further characterization in Task 3.2
CTH (partner n.9)
The selected enzymes were also investigated for their suitability for immobilization (on mesoporous silica particles, mostly) by active site determination and modelling of surface charge as function of pH. Once the analysis of protein surface properties was done, to be able to work on immobilization, CTH performed large scale production of the enzymes and analysis of their stability properties (WP4) for enzyme immobilization.

KNAW (Partner n.8) and UH (Partner n.14) and UNINA (Partner n.1)
Since the available substrates did not allow a suitable screening for a high number of fungi, and recent data from partners labs demonstrated that FAE and GE genes are usually lowly expressed and often in a time and substrate dependent manner, direct screening for FAE or GE producing fungi under a range of conditions was not always possible. Therefore, an alternative approach based on a rational selection was also followed. The criterion was to select species with different properties from the ones already available, so to identify FAE or GE with strong adaptive modifications in their coding sequences.
This resulted in the selection of two of cold-tolerant basidiomycetes, Trametes pubescens and Phlebia centrifuga (UH), and Penicillium subrubescens (KNAW) for its ability to grow on substrates with high phenolic content. Furthermore, partner UNINA screened 3 fungal strains (AT71X, AT84X and PT63X) isolated from lignocellulosic biomasses for FAEs production and selected the AT71X for genome analysis of new FAE- and GE-coding genes.
CTH (Partner n.9)
In an earlier project, 49 fungal strains (25 mesophilic, 24 thermophilic) were isolated from soil, wood or agricultural compost/silages in collaboration with the Food Industries Research Institute in HanoiCTH performed an initial screening of the mesophilic and thermophilic strains
The mesophilic strains tested showed in general a much higher FAE and GE activity than the thermophilic strains, which can be either due to non-ideal cultivation or assay conditions or a lack/low activity of the corresponding enzymes.
For genome and subsequent RNA sequencing, the thermophilic strains FCH 10.5 (Malbranchea cinnamomea) and FCH 5.7 (Rhizomucor pusillus)
SXS (Partner n.11)
Partners UNINA, KNAW, UH and CTH extracted DNAs from the selected fungi by using CTAB-based extraction buffer. RNAs were isolated at various timing upon culture on various substrates using TRIzol reagent.
SXS received DNA and RNA samples and performed the sequencing using the Illumina HiSeq2500 platform. Libraries for sequencing were prepared using NEBNext® Ultra DNA Library Prep kit and NEBNext Ultra Directional RNA Library Prep Kit respectively for DNA and RNA.
_ Data Analysis: DNA
Prior to assembly, the reads were trimmed for adapter sequences and filtered for sequence quality.
Assembly was performed using a short-read genome assembler based on De Bruijn graphs.
For establishing the parameters for eukaryote gene finding, reference genomes for each of the newly sequenced fungi were used as training reference. The parameters resulting from the training were used for the identification of sequences that could be genes, introns, and exons within the newly assembled genomes. Furthermore, an evidence-based method that uses mapped mRNA-Seq reads to determine genes and exon boundaries was used. The output of both methods was combined into a single gene model.
After the gene finding approach, the discovered putative genes were translated into amino acid sequences for annotation and gene homology matching.
_ Data Analysis: RNA
The assembled genomes were used for alignment of the reads for each sample of the selected fungal strains. Based on the mapped locations in the alignment file, the frequency of how often a read was mapped on a transcript was determined, creating the mRNA-Seq read counting. These counts served then as input for the downstream mRNA-Seq differential expression analysis.
During the analysis of the samples supplied by UNINA, some discrepancies between the DNA and RNA sequencing data emerged. After careful evaluation, SXS and UNINA, with KNAW support, realized that 2 different fungi have been sequenced: AT71X (Talaromyces borboniensis) and AT84X (Talaromyces adpressus). Both strains have been sent to KNAW for the deposit to the CBS culture collection. Furthermore, the fungal genome sequences have been submitted to specialized databases
In order to identify putative genes and enzymes involved in the breakdown, biosynthesis or modification of carbohydrates, the total predicted ORFs in the newly assembled genomes were compared to the entries of the Carbohydrate-Active Enzymes (CAZymes) database ( resulting in identification of new putative CAZymes.

KNAW (Partner n.8)
A database to host the data generated in OPTIBIOCAT (sequences, biochemical data, application data) with detailed information ranging from biological aspects like gene name, organism, DNA/protein sequences to chemical properties like pH stability and solvent types has been created.

➢ WP3:
Task 3.1: Enzyme improvement
Improvement of microbial FAEs and GEs was obtained by using several parallel strategies. Previously characterized and new putative fungal and bacterial FAE and GE encoding genes were phylogenetically grouped and selected for heterologous expression in several production hosts. Rational mutagenesis was used to improve thermal stability of novel fungal esterases and directed evolution was applied to improve the biochemical properties of previously characterized fungal FAEs. By using these approaches several active, thermo- and solvent tolerant FAEs and GEs were developed.

Cloning, expression and characterization of new enzymes
Novel putative fungal FAEs and GEs
cDNAs encoding previously characterized and new putative fungal FAEs and GEs were synthetized (by NZYtech) for screening, site-directed mutagenesis and directed evolution. Synthetic genes were cloned into the pPNic706 expression vector and recombinant proteins were produced in Pichia pastoris GS115 host strain in microscale (by RVE) and in flasks (by UNINA, KNAW, UH). After exhaustive development of the expression system by RVE, pPNic962 and pPNicPN-Seamless plasmids were created resulting in improved production yields by introducing genetic elements to the expression vectors. Even 1000 mg/L protein production yields were obtained in flasks by UNINA, KNAW, UH.
DNL utilized an improved fungal platform based on its proprietary fungus Myceliophthora thermophila C1 for the production of nine putative FAE’s and two GE’s and other proteins. The fully sequenced genome of C1 serves as database for potentially useful enzymes in various applications.Over 50 fungal FAEs and GEs were expressed in Pichia pastoris or Myceliopthora thermophila production hosts. Previously characterized Aspergillus niger FaeA (AnFaeA), AnFaeB, M. thermophila (Sporotrichum thermophile) C1 FAEs and GEs (StGE2) were chosen as reference enzymes. The production of the recombinant or overexpressed proteins was analyzed by SDS-PAGE and catalytic activity was determined towards a set of methyl or ethyl hydroxycinnamic acids and pNP-ferulate (pNP-F) for FAEs and benzyl-D-glucuronate for GEs (Sunner et al. 2015) as substrates. Over 20 fungal FAEs covering 11 subfamilies showed FAE activity and most of the new ones obtained activities comparable to those of previously characterized fungal FAEs (Dilokpimol et al. 2018a). From the new GEs, 13 were active against benzyl D-glucuronate and the highest specific activities (>1000 nkat/mg) were detected for PcGCE (Phanerochaete carnosa), HsGE1 (Hypholoma sublaterium) and DsGE1 (Dichomitus squalens)(Dilokpimol et al. 2018b).
Substrate specificities of recombinant FAEs were determined by using four methyl substrates (methyl ferulate, MF; methyl caffeate, MC; methyl sinapate, MS and methyl p-coumarate, MpC). From subfamily (SF) 1 AnFaeB (Aspergillus niger) and AsFaeF (Aspergillus sydowii) hydrolyzed MF, MC and MpC (Dilokpimol et al. 2018a). SF2 family members showed no obvious substrate specificity pattern, whereas the majority of the active FAEs from SF5 and SF6 hydrolyzed all four substrates. Previously characterized AnFaeA (A. niger) was the only representative of SF7 showing activity towards MF and MS. Enzymes from SF8, SF9, and SF10 showed limited activity towards synthetic FAE substrates. Five of the putative FAEs belonging to SF9, SF10 and SF13 were active towards methyl gallate that indicates tannase activity. Two putative FAEs were active towards pNP-palmitate suggesting lipase activity.
Optimal temperatures of FAEs and GEs were 37-50°C and 30-45°C, respectively. 18 FAEs and 8 GEs showed >70% residual activity after incubation at 55°C for 1 hour indicating good thermostability. Solvent tolerance was tested by incubation of the enzymes in 25% acetone for 1h. 7 FAEs and 13 GEs showed similar or better residual activity after incubation with acetone than previously characterized reference enzymes AnFaeA, AnFaeB and StGE2.
Novel putative bacterial FAEs and GEs
500 putative bacterial FAE and GE encoding genes were synthesized, cloned into pHTP1 NZYTech’s proprietary prokaryotic expression vector and produced in Escherichia coli by NZYtech. Twenty percentage of them were expressed at levels over 300 mg/L. From 480 prokaryotic FAEs, 45% were active towards pNP-F. More than 10 % were active on MF, MC and MpC, and 9% hydrolyzed MS. The most promising 78 FAEs based on the protein production level, thermostability at 55°C and solvent (acetone) tolerance were characterized in more detail. Of them the best six bacterial FAEs were purified via Immobilized metal ion affinity chromatography (IMAC) for immobilization and transesterification studies.
Task3.2 :Site-directed mutagenesis of bacterial and fungal FAEs and GEs
Based on the previously solved structural features of Fusarium oxysporum (FoFAEC), M. thermophila (MtFae1a), Sorangium cellulosum FAE (MNEE) and StGE2, as well as protein production level combined with biochemical properties of the novel fungal esterases, Aspergillus wentii (FAE801) and A. sydowii (FAE799), and GE from D. squalens (GE717) were selected for site-directed mutagenesis. Recombinant mutated enzymes were produced in P. pastoris or in E. coli.
Construction, expression and characterization of modified fungal FAEs and GE
Based on amino acid analysis/alignment and algorithms for predicting structural changes, homology models were generated for three new fungal enzymes and suitable sites for mutagenesis were chosen in WP2. Five mutants for each enzyme were codon optimized and synthesized by NZYtech and cloned into P. pastoris by RVE. Enzyme production of the mutants was determined by SDS-PAGE, Western blotting and enzyme activity assays by UNINA, KNAW and UH.
The production levels of A. sydowii (FAE799) N387V and D367A mutants and enzyme activities were comparable to the wild-type enzyme. They also showed slightly higher optima temperature and 10°C higher temperature stability, whereas the activities of the other mutants were negligible. Only N391V mutant of A. wentii (FAE801) showed higher activity than wild-type enzyme. However, other four mutants (N391L, G470F, G195C, S89V) showed improved thermotolerance. Four to 38 –fold higher residual activity was detected after incubation at 55°C for 1 h when pNP-F was used a substrate. Four D. squalens (GE717) mutants (P91V, P91T, P17T, G188I) were active at the temperature range 15 – 55°C when benzyl D-glucuronate was used as a substrate. However, only G188I showed notably higher residual activity than wt GE717 after incubation at 40°C for 1h. Only wild-type GE717 retained more than 40% of its activity after incubation at 60°C and 70 °C.
Modification of biochemically and structurally determined thermophilic FAEs and GE
In order to increase the MS activity of FoFaeC, the Small Molecule Docking (SMD) experiment was performed to the PDB entries 2WTM and 3WMT, as well as the crystal structure of FoFaeC (unpublished data) (LTU). FoFaeC is a type C FAE and has been shown to have activity towards MpC, MC, MF and only some activity against MS (Moukouli et al. 2008). Based on SMD, the double mutant F230H/T202V was designed and cloned for recombinant protein production in P. pastoris X-33 strain. The produced mutants showed activity of 0.3 – 0.8 mU/mL towards MS.
StGE2 crystal structure (PDB:4G4G) was recently characterized and it was shown to have strong activity against 4-O-methyl-D-glucopyranuronate (Charavgi et al., 2013). The methoxy group located at the 4 position is particularly important for substrate recognition of the sugar with much lower activity against esters of D-glucuronic acid. In silico mutational techniques were used to identify mutations that may increase the activity of StGE2 on glucuronic acid based esters such as benzyl D-glucoronate (LTU). To identify residues that determine the binding of the methoxy side group into the binding pocket, the solved crystal structure of StGE2 in complex with 4-O-methyl D-glucopyranuronate (PDB: 4G4J) was used as a starting point. Possible residues were analyzed in terms of mean binding energy and based on these results the most promising double mutant might be G238N/L311H.
Prokaryotic FAE from S. cellulosum (MNEE) contains five nucleophilic elbows (E) each of which was found to form a local active site with varied substrate specificities and affinities (Udatha et al. 2015). The most interesting active sites were studied in detail by inactivating different combinations of nucleophilic elbows (CTH). When only E3 was active, the enzyme was found to be more active towards a broad spectrum of methylated hydroxycinnamates than MNEEwt. With only active E5, the enzyme was active on methyl 4-hydroxy-4-methoxy phenyl propionate, whereas MNEEwt was completely inactive on this substrate. The study will continue by characterization of the double mutant with active elbows E3 and E5 to study their additive effects.

Task3.3: Directed evolution

Directed evolution was applied to improve the biochemical properties of the previously characterized fungal FAEs A. niger FAE A (AnFaeA), F. oxysporum FaeC (FoFaeC) and M. thermophila FAE 1a (MtFAE1a). Several expression vectors and production hosts were used to optimize construction, production and high through-put (HTP) screening of the mutant libraries.

Construction of libraries of randomly mutated forms of FAEs and GEs

The development of the evolved mutant library was based on using a new strain of Y. lipolytica JMY1212 especially dedicated to HTP screening procedures. This strain was constructed to enable the integration of the expression cassette at a defined locus aiming up to 8000 transformants per µg DNA (Bordes et al. 2007). The shuttle vector JMP62-TEF-ppLIP2-LIP2, a derivative of the expression vector described in Nicaud et al. (2002), was used for cloning in Y. lipolytica. Several cloning strategies were used to develop a recombinant expression system in Y. lipolytica. For AnFaeA and FoFAEc Y. lipolytica was a correct platform for expressing the genes and therefore performing directed evolution.

The mutations were introduced to FoFAEc (UNINA) and AnFaeA (INRA) cDNAs by using error-prone polymerase chain reaction (ep-PCR) using GeneMorph II Random mutagenesis kit (Agilent Technologies). For AnFaeA, DNA shuffling (Staggered Extension Process) and rational mutagenesis strategies were also used by INRA. Several rounds of PCR were carried out to obtain the optimal conditions for a good amount of mutated cDNA needed for the construction of a library of 30,000 mutants.

Unfortunately, recombinant expression of active AnfaeB, MtFaeA1 and StGE2 was not successful in Y. lipolytica and therefore recombinant expression in S. cerevisiae was tested by using several expression systems. In the case of MtFaeA1, the combination of S. cerevisiae Y294 strain with pSAL4 was demonstrated by UNINA suitable for the expression and the production of the recombinant enzyme. For AnFaeB, production of an active enzyme was obtained using S. cerevisiae INVSc1 strain with pYES2 (Life Technologies).

Screening of random mutant libraries

HTP screening in solid media with chromogenic substrate and in multiwall plates in liquid media were developed and optimized. Plate screening was performed using either the chromogenic substrate, 5-bromo-4-chloro-3-indolyl ferulate (X-Fe) or a new substrate 4-nitrocatechol-ferulate, which was synthesized to allow activity measurements in both solid and liquid media (Gherbove et al. 2016). The screening steps included selection of FAE expressing clones from plate cultivations followed by cultivation on multi-well plates in liquid medium. Overall, more than 40 000 clones were screened to search for optimized esterases.

Characterization of optimized proteins

UNINA: For FoFAEc more than 30,000 mutants were prepared to obtain optimized variants in terms of activity, thermostability and/or solvent resistance (Cerullo et al, submitted). Five mutants with improved properties were selected when 4pNC-Fe was used as a substrate. 96 enzyme variants showed improved characteristics. The best five mutants were selected substrate specificity analyses. L432I variant showed higher activity towards MpCA, MFA and MCA than wt FoFaeC. E455D variant maintained completely its hydrolytic activity after two hour incubation at 55°C, whilst L284Q/V405I variant showed both higher thermo- and solvent tolerance than wt FoFaeC.
UNINA: Thirty improved enzyme variants of MtFae1a were selected (Varriale et al. submitted).The best four novel selected variants and wt MtFae1a were investigated in experiments of docking with hydroxycinnamic acid esters using a hybrid model of 3D structure of MtFae1a, allowing to confirm the preference of the enzymes towards bulkier substitutions and the broad substrate specificity of MtFae1a. H105Y and a double mutant F251L/H105Y exhibited two-fold increased hydrolytic activity towards MF, MpC and pNP-F and three-fold increased activity towards MC. Moreover, H105Y variant showed enhanced abilities in glyceryl ferulate, prenyl ferulate, prenyl caffeate and butyl ferulate syntheses. G49A variant showed fourfold higher activity that wt MtFae1a towards MC.
INRA: The first round of evolution on AnFaeA resulted in 13 mutants with higher resistance to temperature and/or presence of solvent. After two rounds of evolution and isolation of several optimized mutants also rational mutagenesis was performed to explore precise association of mutations, and further improve performances. The quadruple-mutant N6T-A119V-S163T-V243M displayed the most improved temperature resistance properties, with a Tm increased by 9 °C and a residual activity after heat shock multiplied by four, compared to the template enzyme. In parallel, mutant N6T-R66Q-Q117E was obtained after rational recombination of some highlighted substitution, and constitutes the esterase with the highest resistance to both tested solvents, DMSO and tert-butanol. Interestingly, after the StEP recombination, mutant N6T-Q22L-F187Y showed both a better resistance in presence of tert-butanol and a better thermostability than the template enzyme. A slight increase in catalytic efficiency was detected for D93G variant when 4NTC-Fe was used as a substrate. Though most of the substitutions only moderately affect the kinetic parameters of the mutant enzymes, kcat was significantly decreased for proteins containing the R66Q substitutions, and a ca. three times higher Km was observed for the V243M mutant.

➢ WP4:
Task 4.1: Optimization of enzyme production
Enzyme production was optimized for each chosen expression system differently. The common strategy involved using optimized expression vectors for the gene(s) of interest. When larger quantities of protein were needed for more extensive characterization, the production strains were grown in fermenters with larger working volumes and the fermentation process was further developed.
Production of bacterial esterases was tested on two different culture media. Production optimization was not needed for all target enzymes as levels obtained in standard cultivation conditions were sufficient for planned experiments.

Optimization of fermenter cultivations for production of selected feruloyl and glucuronoyl esterases (UNINA, WWUM, UH, DNL/GIBV)
Expression of enzymes in Pichia was done under the control of a strong inducible promoter in a three-phase fermentation process. The transition from one carbon source to another was initially a bottleneck for enzyme production. Fine-tuning of the fermentation protocol led to overcoming the limitations of the process and enabled the production of all selected enzymes expressed in Pichia.
The yields of the Pichia strains were evaluated at different temperatures and in different fermentation media for the batch phase in shake-flasks. The results were used to optimize the bioreactor process.For Myceliophthora thermophila C1, the general fermentation process was developed following a preliminary design of experimentsThe few parameters identified as most significant for specific protein production were tested for the fermentation of selected strains. Changes in the process temperature, carbon source and feed rate led to improvement of production by twofold to fourfold, depending on the production strain. In best cases the yields were double digit numbers of g/L of protein of interest
Optimization of enzyme production by strain improvements (RVE, DNL/GIBV)

Protein production in Pichia was done in two methanol-induced systems consisting of different expression vector and strains. Upon production of mutants a tag was added for better quantification of secreted enzyme, regardless of the activity levels.
A third expression vector was used for protein production in methanol-free conditions. The production levels in this case were not significantly higher than those of the standardly used system.
New vectors were designed that included proprietary genetic elements proven to have a positive role in enhanced protein production. These new vectors were synthesized and used to construct the expression cassette of three selected target proteins. The use of the new genetic elements improved target protein production by a factor of three when comparing single-copy strains. The first generation strains were improved by additional integration of the expression cassettes, and multi-copy strains were generated. The final production levels were improved compared to the standard production strains by a factor of six.
Other strain improvement tools used for production optimization in Pichia comprised of vector backbone and expression cassette minimization and signal sequence variations.
Minimization of the expression cassette was done by removing two specific features from the well-expressing proprietary plasmid. Transformants containing either the original or the minimized plasmid exhibited similar production levels of the protein of interest.
Efficient secretion of recombinant protein is obtained by fusing the genes of the desired proteins to a α-mating factor pre-pro leader sequence (MFA). Composition and length of the C-terminal part of MFA signal sequence (“spacer peptide”) were varied. These modifications improved protein production by twofold.
Production optimization in Myceliophthora thermophila C1 was achieved for all reference proteins by generation of multi-copy strains which improved target protein levels by an average of two-fold. For one selected FAE, production was increased by using an alternative expression cassette where the gene of interest was under the control of a novel, more potent promoter. The production levels of the new strain was improved by a factor of 4.6 target protein compared to the prior first generation strain and by 1.5-fold compared to the best producing multi-copy strain that used the standard promoter.

Task 4.2 : Enzyme formulation for industrial applications (UNINA, WWUM, UH, DNL/GIBV)

For immobilization and for using directly in ester synthesis as free enzyme, the esterases are required to be delivered in solution. The limitations of liquid formulations are given by the reduced shelf life caused by processes like proteolysis, protein aggregation and oxidation. In presence of preservatives, the liquid enzyme concentrate showed good stability over a period of six months with negligible activity loss.
For longer storage time the enzymes were freeze-dried and the resulting preparations were stable for over two years.

Task 4.3: Enzyme immobilization

Immobilization of selected esterases by cross-linked enzyme aggregates (CLEA).
CLEA Technologies has developed various protocols for the immobilization of enzymes as cross-linked enzyme aggregates (CLEAs). The general procedure is firstly the formation of insoluble protein aggregates by adding a precipitant followed by cross-linking the insoluble protein aggregates using a bifunctional cross-linking agent to produce the CLEAs.
Three FAEs have been immobilized as CLEAs and further studied in the hydrolysis of ethyl ferulate in order to calculate the activity and activity recovery with respect to the native enzyme. The CLEA immobilization of two selected FAEs resulted in an activity which is similar to that of the native enzyme.
For optimal CLEA immobilization, various concentrations of precipitation agent and cross-linker were trialed.
All enzymes seemed to tolerate better the non-polar solvents tested; in this case, the residual activities obtained from most enzymes exceeded 70 % of the initial. Alkanes were even found to enhance enzyme activity in some cases, with residual activities exceeding 100 % of the initial. This could be partly explained by the fact that these solvents are not miscible with water, and as a result, when the enzymes are introduced in such a solvent, they tend to form hydrophilic micelle-like structures, where the enzyme is protected on the enclosed space. Most enzymes seemed to perform well in the reaction solvent, with residual activities ranging from 103 % to 20 %.
The most robust enzymes in terms of operational stability retained 100 % or more of their initial activity over five consecutive cycles, for 8 hours at 45 °C. One immobilized enzyme in particular showed an enhanced activity in the first rounds of synthesis reactions, which can be attributed to the extensive hours of agitation, leading to shearing of bigger particles into smaller ones, and thus minimizing the diffusion limitations observed in large CLEA particles. The least stable FAE retained 15 % of their initial activity respectively, after 5 cycles of synthetic reactions.
The immobilization of FAEs on acrylic polymer beads was also very successful. Good enzyme loadings of 5 -10 % were achieved for the immobilized enzymes by means of hydrophobic interactions or covalent attachment.
In the synthesis of target ester compounds these immobilized enzymes gave overall higher yields and better selectivity compared to those of the native enzymes.

Immobilization of selected esterases on mesoporous material, determination of immobilization efficiency and stability (CTH).
Selected esterases were immobilised on mesoporous material with different pore sizes with the goal to improve performance of the enzymes in selected synthesis reactions. The immobilisation material was synthesised by researchers of the Division of Applied Surface Chemistry at the Department of Chemistry and Chemical Engineering at Chalmers University of Technology.
Immobilisation efficiency and stability of immobilised enzymes parameters were investigated and evaluated, by using a model reaction.
All selected fungal feruloyl esterases could be immobilized on mesoporous material. The process was most efficient at lower pH values and within a few hours maximum enzyme loading could be achieved. Enzymes with a low molecular weight could be selectively immobilised and thus careful selection of immobilisation support could be used as a way to avoid costly enzyme purification.
Generally, with increasing pore size, more protein (and implicitly more FAE) could be immobilised, although there was an exception to this rule possibly due to different geometry of the material’s pores.
Immobilisation pH strongly influenced the activity of the enzymes in subsequent transesterification reactions.
The best results in terms of product formation rate were achieved for two B-type FAEs. The product yield was lower for the immobilised enzymes compared to the free enzymes but the synthesis to hydrolysis ratio was higher for the former.
Re-usability was tested for one immobilized FAE in the ester synthesis system and the results showed that the enzyme was stable and active for over nine production cycles, each lasting 24 h. The productivity did not drop below 90 % of the initial value and the synthesis to hydrolysis ratio did not change during the tested time period. The immobilisation seems to have stabilized the enzyme very well, and even though the enzymes are only immobilised by adsorption, it was concluded that the degree of leaching was very low since the enzyme activity did not drop significantly.
Six bacterial FAEs were immobilized in mesoporous silica (MPS) particles. The immobilized enzymes were assayed for trans-esterification activity. Of the six bacterial FAEs, four had yielded the expected product over the course of 138 hours. The FAEs worked better at neutral pH and in most cases the protein loading was almost reaching 100 % of initial protein concentration. Three FAEs were chosen to further optimize the conditions for enzyme immobilization and trans-esterification. However, the FAEs were mostly inactive when they were immobilized. There was minor enzymatic activity detected in some cases but hydrolytic activity was more dominant than trans-esterification.

Task 4.4: Techno-economic evaluation (SUPREN)
The techno-economic viability and environmental friendliness of the whole production process of technical enzymes was assessed considering a full industrial scale scenario in close collaboration with the WP4 partners. Data collected during the fermentation process for both fungal protein production systems were used to create the preliminary mass- and heat balances. These balances were the basis of computed LCA models.
The overall environmental impact value was calculated through the LCA, the energy consumption of the overall process was computed and the cost calculation and its economic analysis were performed for each available data set, all based on the production scenarios.
A cost analysis for the whole OPTIBIOCAT process at the previously given production rate of 500 t/a (at large scale) was conducted. The whole OPTIBIOCAT process is represented by the combined processes of “technical enzyme production” and “novel antioxidant production” (work package 5). The cost breakdown of the whole process of novel OPTIBIOCAT antioxidant production considered in this deliverable with regard to these cost categories making up the total cost of manufacturing.
The Myceliopthora themophila C1 expression system was more efficient for esterase production and comparing the production of feruloyl esterase and of glucuronoyl esterase, the former was superior. The results only reflect the analysis of a limited set of data and are not definite as there is still significant room for improvement following further molecular and process developments.

➢ WP5:
Task 5.1: Synthesis of substrates (TAROS; INRA)
Synthesis of novel chromogenic and fluorescent substrates was achieved for high-throughput screening of esterase activities such as indolyl- and 4-nitrocatechol ferulates in good overall yields (46-55% in 4 steps). Synthesis of substrates was performed for screening of feruloyl esterase and glucuronoyl esterase activities for directed evolution in WP3. Reference samples of targeted compounds: prenyl ferulate, prenyl caffeate, L-arabinose ferulate and glyceryl ferulate were produced for use in WP5. A total of 30 activated donors were produced in small and large scale for transesterification reactions in WP5 and scale up of bioconversions in WP6. Synthesis of a glucuronate substrate for enzymatic transesterification with GEs was achieved for the first time.

Task 5.2: Evaluation of synthetic abilities of FAEs and GEs

Analysis of FAEs/GEs for their ability to catalyze esterification of ferulic acid/D-glucuronic acid with various fatty alcohols or carbohydrates/aromatic alcohols (LTU)

For analysis of the synthetic abilities of esterases, a protocol was developed based on detergentless microemulsions allowing the detection of targeted products in high yields. The protocol was shared and was adopted by collaborating partners in WP5 for analysis of enzymes for bioconversions. In total, 352 conditions were tested successfully including testing 44 FAEs (27 fungal, 6 bacterial and 11 mutants) for the synthesis of 4 targeted compounds: prenyl ferulate, prenyl caffeate, glyceryl ferulate, L-arabinose ferulate and butyl ferulate, as reference. A complete assessment of the synthetic potential of FAEs was performed correlating enzyme-based with system-based characteristics and providing insights into the selection of robust biocatalytic tools for efficient bioconversions. 8 promising enzymes were selected for optimization of reaction conditions. 4 fungal GEs were tested for their synthetic ability targeting to the synthesis of 2 targeted compounds: prenyl glucuronate and benzyl glucuronate.
Analysis of enzymes for conversion of the library of 300 substrates (TAROS).
Numerous fungal esterases in free and CLEA form were tested as adequate candidates for utilization in the construction of the compound library of 300 substrates. The most promising candidates were identified while reaction conditions were adjusted in order to serve an efficient conversion of library generation.

Task 5.3: Optimization of reaction conditions

Optimization of reaction parameters affecting the reaction rate and conversion yield (LTU)

8 selected fungal FAEs for optimization of reaction conditions in a ternary system forming detergentless microemulsions aiming to the synthesis of 2 targeted products: prenyl ferulate and L-arabinose ferulate. Parameters such as the water content, the substrate concentration, the enzyme load, the pH, the temperature and agitation were investigated. Kinetic studies indicated enzymes with higher catalytic efficiency and substrate specificity while competitive bioconversions were developed yielding >70% of targeted compound. Optimization of reaction conditions offered a 5-fold increase in yield and a 4-fold increase in selectivity. The 8 selected FAEs were also used for synthesis in 12 single solvent systems (apolar, polar organic solvents and ionic liquids). The 3 most promising FAEs were used as catalysts in further optimization of reaction conditions applying response surface methodology in the best solvent for each enzyme. Efficient bioconversions were developed with improved yields (>90%) and selectivity allowing easy solvent and product recovery and possibility of enzyme reuse. Overall, 4 competitive bioconversions based on microemulsions and 4 based on single solvents catalyzed by free enzymes were developed in this task.

Optimization of reaction conditions with immobilized esterases (CTH; CLEA)

Optimization of reaction conditions for synthesis of antioxidants was applied utilizing esterases immobilized and produced in WP4. 4 FAEs from M. thermophila C1 immobilized in mesoporous silica (MPS) particles were used for the synthesis of 2 targeted compounds and their transesterification efficiency was compared with the respective free form. Screening of synthetic abilities and stability in 12 solvent systems including organic solvents and ionic liquids was performed at fixed conditions. The catalytic efficiency was correlated with system-based and enzyme-based characteristics at varying water content and revealed that MPS immobilization results in higher selectivity. 6 bacterial FAEs immobilized in MPS were used for transesterification studying the effect of immobilization efficiency, pore size, modified MPS surface on the bioconversion yield and selectivity. The operation stability of fungal and bacterial MPS immobilized FAEs was tested in > 5 reaction cycles. Seven CLEA-immobilized fungal FAEs were used for screening in 12 different solvents targeted to the synthesis of 2 targeted compounds. The most promising FAE was used as biocatalyst for optimization of reaction conditions in the best solvent applying response surface methodology while its operational stability was tested for >5 reaction cycles. Overall, 2 competitive bioconversions based on single solvents catalysed by CLEA-enzymes were developed in this task.

Task 5.4 : Library generation and screening

Library generation and products recovery (TAROS)

The library program was conducted in three steps: validation and identification, optimisation, and scale-up and library production including validation and identification of products. Reaction conditions such as solvent, concentrations of substrates, pH, temperature and reaction time were optimised to maximise product yield and selectivity. Enzymatic transesterification of hydroxycinnamic acids and glucuronic acid was performed with a structurally highly diversified collection of alcohols and carbohydrates catalysed by FAEs and GEs from M. thermophila C1. Library production was conducted in parallel arrays. A total of 430 enzymatic transesterification reactions were performed on a 30 mg scale. 310 reactions were successful and afforded 5-30 mg of the final compounds in a purity >90% after purification by preparative HPLC. The product recovery of products was optimised and determined by LC-MS and NMR analysis. The synthesised compound collection allowed a detailed assessment of the potential of the FAEs and GEs for the scaffold derivatisation and the structure specificities.

Screening of the library of the new compounds for their antioxidant activity (KORRES)

60 library compounds were tested for antioxidant activity in a cell-free and cell-based system and for cytotoxicity. The findings indicated that human skin fibroblasts remained viable to a large extent with the majority of targeted compounds. Furthermore, the majority of tested compounds have been indicated as potent antioxidants, as they exhibited a great free radical scavenging activity in the cell-free DPPHA assay and additionally the capacity to reduce intracellular ROS levels in non-cytotoxic concentrations. In most cases, their radical scavenging activity lied between 40-100 μM while cytotoxicity was observed between 100-500 μM.

Task 5.5: Development of the whole integrated industrial process concept considering techno-economic evaluation (SUPREN in collaboration with all WP5 partners)

SUPREN assessed The techno-economic viability and environmental friendliness of the whole industrial bio-catalysed OPTIBIOCAT process concept was assessed considering a full industrial scale scenario in close collaboration with the WP4 and WP5 partners and experimental support from WP6. The evaluations were conducted exploiting the numerical models developed. To innervate these models with real-life data the experimental findings of various partners were implemented. These findings and the process chemistry along with the thermal stability of product were exploited to generate a robust and feasible process concept taking into account different modes of operation (continuous/batch) and solvent recovery. Two different concepts were developed based on the previously defined production scales (small scale: 50 t/a and large scale: 5000 t/a). Heat and energy balancing was performed to feed both techno-economic analysis and environmental impact assessment. From the summary of the results obtained it was concluded that the process route of the large scale antioxidant production is more economic and more environmental friendly in comparison to small scale. This results from a more energy efficient process which operates mostly in a continuous manner whereas the small scale production solely runs in batch operation mode.

➢ WP6:
Task 6.1: Scale up of enzyme production (RVE, WWUM, DNL/GIBV)

In Task 6.1 fermentation processes were scaled up to show the potential of these developed processes to reduce production costs of the enzymes essential for the green production of cosmetics.
The production of one C1-derived FAE in M. thermophila was up-scaled to a fermenter of 150 L capacity .
WWUM performed in total 19 small scale fermentations (1.25 L) and 15 medium scale fermentations (16-20 L) using standard P. pastoris strains or P. pastoris strains optimized by the introduction of protein expression enhancing genetic element (done by RVE) expressing either FAE710, FAE711, FAE713, FAE714, FAE799 or FAE801 or a reference protein of human origin if induced with methanol (AOX1 promotor). Productions of FAE710, FAE713, FAE799 and the reference protein could be scaled up successfully resulting in a >6-fold higher protein contents in supernatants of the optimized strains compared to the standard strain supernatants, impressively demonstrating the potential of the introduced genetic modifications. Productions of FAE711, FAE714 and FAE801 could not be scaled up successfully most likely due to individual problems with the provided strains. Furthermore, one strain expressing a GE (StGE1; GE712) was fermented in a small scale culture (1.25 L) but could not be up-scaled or produced due to similar problems. Additionally 4 small scale fermentations using standard or optimized Pichia strains constitutively expressing FAE710 (GAP promoter) were successfully performed resulting in >6-fold higher enzyme activities in the supernatants of the optimized strain compared to the standard strain also demonstrating the high potential of the genetic modifications and that the functionality of these changes is independent of the kind of expression induction.

Task 6.2: Scale up of bioconversions (WWUM, LTU, TAROS, CTH, CLEA)
The up-scalability of a ternary solvent system forming micro-emulsions (hexane/t-butanol/water) and of a single solvent system were tested at preparative scale (50 mL) according to optimization results on bioconversions developed by LTU in WP5 for the synthesis of prenyl ferulate (Task 5.3.1 Optimization of reaction conditions). The biocatalyst used was C1FaeB2 in both cases. Scale up results on micro-emulsions at 50 mL performed by WWUM gave lower results (50%) instead of 70% found by LTU in Task 5.3.1. Scale up of single solvent system bioconversions resulted in 90% yield and significantly higher velocity, while a comparison between different batches of C1FaeB2 used in Task 5.3.1 and Task 6.2 revealed that the enzyme used in scale up was >2 times more active justifying the difference. Prenol was identified as the main contaminant of the reaction crude by TAROS. For further scale-up two adaptations were investigated by WWUM: the batch process was substituted by a fed-batch process, and reduction of prenol was investigated. Fed batch experiments revealed that after 3 donor additions the bioconversion yield was significantly higher as in a single step batch process. The effect of byproduct acetaldehyde on the yield was tested showing that enzyme activity was unaffected up to 15% v/v of acetaldehyde. Studies on the effect of prenol concentration in the batch of C1FaeB2 used by WWUM showed that the prenol/donor ratio could be reduced without affecting the yield. A final experiment at 1 L was performed at adapted conditions resulting in competitive yields (77%). The downstream processing of reaction mixture included centrifugation andstripping of solvent. Further work-up including isolation and purification of the product was done by TAROS. Purity of all prenyl ferulate batches were determined to be up to >95%.

Task 6.3. Characterization of allergenic and safety properties (KORRES)
The purpose of task 6.3 was to explore the potential use of compounds prepared and tested in previous WPs for the use in cosmetics formulations. In order to be able to use chemicals in cosmetics manufacturers need to ensure that the chemicals are safe to use as per the Directives of the European Commission. In WP6 Korres has followed the EU “Notes of Guidance for Testing of Cosmetic Ingredients and Their Safety” where an extensive reference is given on the the different aspects of testing and safety evaluation of cosmetic substances in Europe with particular emphasis on cosmetic ingredients.
In the context of WP6 of Optibiocat Korres was responsible for testing via suitable in vitro assays the safety profile of four compounds derived from the work of partner Taros and Korres’ profiling on their antioxidant activity. Instead of four compounds, Korres was able to test seven compounds as follows: Prenylferulate, Prenylcaffeate, L-Arabinoseferulate, 3-(3,4,5-Trihydroxyphenyl)-acrylic acid 3-methyl-but-2-enyl ester, 3,4,5-Trihydroxy-benzoic acid 3-methyl-but-2-enyl-ester, 2-[2-(2-hydroxy-ethoxy)-ethoxy] ethyl caffeate, 3-(3,4-Dihydroxyphenyl)-acrylic acid 3,4 dihydroxybenzyl ester.

The above compounds were tested for their potential to induce skin irritation by the use of the reconstructed human epidermal model EpiDermTM (EPI-200, MatTek, Ashland, USA) that consists of normal human-derived epidermal keratinocytes, which have been cultured to form a multilayered highly differentiated model of human epidermis. The tests based on the in vitro results the in vivo prediction for each test compound is non-irritant (NI), at the tested concentrations. Subsequently, the seven aforementioned compounds were also tested for their acute eye irritation potential by means of the EpiOcularTM Eye Irritation Test. The EpiOcularTM tissue consists of normal, human- derived epidermal keratinocytes which have been cultured to form stratified, squamous epithelium similar to that found in the cornea. Regarding eye imitation testing and according to GHS classification, none of the aforementioned compounds exhibited eye irritation potential and can be classified as non-irritants (NI), at the tested concentrations. In order to determine the potential genotoxicity of the tested compounds we evaluated the activation of p53 tumor suppressor which plays a major role in cellular response to DNA damage and other genomic aberrations. The exposure of cells to the seven targeted compounds, at the tested concentrations, does not induce the phosphorylation of p53 on Ser15 at tests times of incubation (1 hour of incubation and after a longer period of 6 hours of incubation). Furthermore, in order to identify the contact sensitization potential (allergenic potential) of the seven targeted compounds we used the EpiDermTM in vitro human skin model and measured the pro-inflammatory IL-18 secretion into the tissue culture medium after a 24-hour topical exposure of the tested compounds and all of the seven targeted compounds exhibited no contact sensitization potential at the tested concentrations.

Task 6.4 Techno-economic evaluation (SUPREN in collaboration with all WP5 partners).
Finally in task 6.4 the focus was on the scale-up and the economic and ecologic evaluation of whole production processes of OPTIBIOCAT antioxidants. It combines large scale production processes of “technical enzymes” and “novel antioxidants” in order to arrive at the overall process: The target antioxidants are obtained via industrially viable bioconversions applying those developed enzymes.
The evaluations are conducted by implementing numerical models which are developed by SUPREN GmbH. To innervate these models with real-life data, the findings of laboratorial work provided by the other technology developing partners contributing to this task are exploited. Furthermore a process concept is developed for the production of antioxidants and heat and energy balancing is performed by the use of the aforementioned models. The resulting data is used to feed evaluation procedures, the techno-economic analysis as well as the environmental impact assessment.
In parallel to the OPTIBIOCAT antioxidants, a state-of-the-art antioxidant technology is investigated and production of Vitamin C is selected as a base case scenario for benchmarking in order to demonstrate the benefits of the OPTIBIOCAT antioxidants for the global market. General assumptions on the production scenario from an economic as well as from an environmental point of view are applied like-wise to both OPTIBIOCAT antioxidants and Vitamin C production in the same detail to facilitate a fair, transparent and significant comparison of the forthcoming results.
The benchmarking results reveal that the production of prenyl ferulate is three times more expensive than the production of Vitamin C. This is mainly due to the high costs of the raw materials of the prenyl ferulate production process which are not available in bulk and at a low price in the market as opposed to the Vitamin C case. However it is observed that the environmental footprint resulting from both the midpoint and endpoint indicators is considerably high as for Vitamin C case in comparison to prenyl ferulate production. When the quantitative footprints of both production processes are compared against each other there is approximately a factor of 20 between the total impacts of each category. This immense difference between two cases is the result of the variations between two processes such as the amount of processing efforts e.g. various utilities and exploitation of different material types having different environmental hazards.

Potential Impact:
Approach to Dissemination and Communication

An important aspect for effective communication is a strong core branding. BIOCOM developed a distinct logo and branding style to guarantee a consistent design. Effective communication is also about developing and sending key messages in order to raise awareness for the topic and create interest for the more complex results of the project. The key messages for OPTIBIOCAT were:

• The OPTIBIOCAT project will develop eco-friendly, resource-efficient and cost-effective bioconversions for the production of antioxidants used in the cosmetic industry
• OPTIBIOCAT is an international consortium bringing together excellent experts from academic and research institutes and the chemical industry as well as biotech companies
• OPTIBIOCAT advances international cooperation while connecting researchers from different European countries such as Italy, Germany, Greece, Portugal, Sweden, the Netherlands, Finland and France.

The different target groups included research communities, industrial communities, policy makers and the general public. The main communication instruments were the website of the project, press releases, social media channels and events. On the website, news on the project were disseminated via the news and results section and in the Press & Multimedia folder. Via the social media accounts (Facebook and Twitter) articles and publications on sustainable cosmetics were posted to assemble a group of interested individuals, policy makers, companies or institutions. Once the project moved forward, the focus of posts moved to news from the website and upcoming events. In general, the dissemination activities can be clustered as following:

• Creation and distribution of dissemination material: website, leaflets, brochures, posters, PowerPoint template for presentations, newsletters
• Publications in technical journals and press, presentations at conferences and workshop
• Co-organisation of scientific and practice-oriented workshops
• Presentations translated in the partners langauges to support OPTIBIOCAT diffusion at local level
• Liaison activities with other related scientific communities/societies as well as relevant research projects, to support exchange of technological knowhow and knowledge among different partner organisations and experts

Dissemination highlights

This overview will give a glimpse into the OPTIBIOCAT dissemination activities.

Creation of dissemination material

The following communication material was designed: leaflet, powerpoint presentation template, letter, a poster for each workshop and the final conference, a poster displaying some of the results of the project to be shown at conferences and three newsletters.

Workshop poster
Results poster

Publication of scientific papers

In the course of the project, 13 scientific papers have been published, about the same number is still in preparation:

1. Antonopoulou, I., Leonov, L., Jütten, P., Cerullo, G., Faraco, V., Papadopoulou, A., Kletsas, D., Ralli, M., Rova, U., Christakopoulos, P. (2017): Optimized synthesis of novel prenyl ferulate performed by feruloyl esterases 1 from Myceliophthora thermophila in microemulsions. Appl Microbiol Biotechnol 101: 3212-3226.
2. Antonopoulou, I., Papadopoulou, A., Iancu, L., Cerullo, G., Ralli, M., Jütten, P., Piechot, A., Faraco, V., Kletsas, D., Rova, U., Christakopoulos, P. (2017): Optimization of chemoenzymatic synthesis of L-arabinose ferulate catalyzed by feruloyl esterases from Myceliophthora thermophila in detergentless microemulsions and assessment of its antioxidant and cytotoxicity activities. Process Biochem
3. Antonopoulou, I., Hunt, C., Cerullo, G., Varriale, S., Gerogianni, A., Faraco, V., Rova, U., Christakopoulos, P. (2017): Tailoring the specificity of the type C feruloyl esterase FoFaeC from Fusarium oxysporum towards methyl sinapateby rational redesign based on small molecule docking simulations
4. Cerullo G, Houbraken J, Granchi Z, Pepe O, Varriale S, Ventorino V, Chin-A-Woeng T, Meijer M, de Vries RP, Faraco V. Draft Genome Sequence of Talaromyces adpressus. Genome Announc, in press. DOI:10.1128/genomeA.01430-17
5. Dilokpimol, A. Mäkelä, M.R. Mansouri, S., Belova, O., Waterstraat, M., Bunzel, M., de Vries, R., Hilden, K. (2017): Expanding the feruloyl esterase gene family of Aspergillus niger by characterization of a feruloyl esterase, FaeC. New Biotechnol 37: 200–209.
6. Dilokpimol A., Mäkelä M.R. Varriale S., Zhou, M., Cerullo G., Gidijala L., Brás, J.L.A. Jütten, P., Piechot, A., Verhaert, R., Hilden, K.S. Faraco, V., de Vries, R.P. (2018): Fungal feruloyl esterases: functional validation of genome mining based enzyme discovery including uncharacterized subfamilies. New Biotechnol 41: 9-14.
7. Dilokpimol A., Mäkelä M.R. Cerullo G., Zhou, M., Varriale S., Gidijala L., Brás, J.L.A. Jütten, P., Piechot, A., Verhaert, R., Faraco, V.,Hilden, K.S. de Vries, R.P. (2018): Fungal glucuronoyl esterases: Genome mining based enzyme discovery and biochemical characterization. New Biotechnol 40: 282-287.
8. Granchi Z, Peng M, Chi-A-Woeng T, de Vries RP, Hildén K, Mäkelä MR (2017): Genome sequence of the basidiomycete white-rot fungus Trametes pubescens. Genome Announc 5: e01643-16.
9. Granchi Z, van Pelt S, Thanh VN, Olsson L, Hüttner S. (2017): Genome Sequence of the Thermophilic Biomass-Degrading Fungus Malbranchea cinnamomea FCH 10.5. Genome Announc. 17: 5(33) e00779-17.
10. Hüttner, S., Klaubauf, S., de Vries, R.P. Olsson, L. (2017): Characterisation of three fungal glucuronoyl esterases on glucuronic acid ester model compounds. Appl Microbiol Biotechnol, 101: 5301-5311.
11. Hüttner, S., Thuy Nguyen, T., Granchi, Z., Larsbrink, J., Thanh, V.N. Olsson, L. (2017): Combined genome and transcriptome sequencing to investigate the plant cell wall degrading enzyme system in the thermophilic fungus Malbranchea cinnamomea. Biotechnol Biofuels 10: 265.
12. Hüttner, S., Zezzi Do Valle Gomes, M., Iancu, L., Palmqvist, A., Olsson, L. (2017): Immobilisation on mesoporous silica and solvent rinsing improve the transesterification abilities of feruloyl esterases from Myceliophthora thermophile, Bioresource Technol 239: 57-65.
13. Peng, M., Dilokpimol, A., Mäkelä, Hildén, K., Bervoets, S., Riley, R., Grigoriev, I.V. Hainaut, M., Henrissat, B., de Vries, R.P. Granchi, Z. (2017): The draft genome sequence of the ascomycete fungus Penicillium subrubescens reveals a highly enriched content of plant biomass related CAZymes compared to related fungi. J Biotechnol 246: 1-3.

OPTIBIOCAT in the press
The OPTIBIOCAT project was covered by several magazines and news websites. Selected examples are listed below, an overview of all press coverage of the project can be found on the website.

Presentation of results at conferences

Apart from the three workshops and the final conference that were organized as part of the project, OPTIBIOCAT attended numerous events and presented own results. A selection of events the consortium joined:

• 28th Fungal Genetics Conference, 16-22 March 2015, "Heterologous expression of a feruloyl esterase from Aspergillus terreus" by Mäkelä, M.R.
• 11th Carbohydrate Bioengineering Meeting, 10-13 May 2015, Espoo, Finland, “Efficient chemoenzymatic synthesis of antioxidants using feruloyl esterases in detergentless microemulsions”, Io Antonopoulou, Evangelos Topakas, Laura Leonov, Ulrika Rova, Paul Christakopoulos
• Biotrans 2015, 26-30 July 2015, Vienna: “Chemoezymatic synthesis of Antioxidants using the feruloyl esterases from Myceliophthora thermophile in non conventional media”, Antonopoulou, I., Topakas, E., Leonov, L., V. Faraco, U. Rova and P. Christakopoulos
• 19th European Symposium of Organic Chemistry, 12-16 July 2015, Lisbon: “Design of versatile synthetic probes for efficient screening and evaluation of feruloyl esterase activities”, O. Gherbovet, F. Ferreira, J. Durand, M. Ragon, G. Hostyn, S. Bozonnet, R.Fauré, M.J. O'Donohue
• 5th International Conference on Novel Enzymes, 11 October 2016, Groningen, Netherlands, Adiphol Dilokpimol
• European Society of Biochemical Engineering Sciences (ESBES), 11-14 September 2016 in Dublin, Silvia Hüttner
• Poster presentation at 29th Fungal Genetics Conference, London, UK (2017): „Expanding feruloyl esterase gene family of Aspergillus niger: characterization of a new feruloyl esterase” by Dilokpimol et al
• Poster presentation at 2nd Symposium on Plant Biomass Conversion by Fungi, Utrecht, Netherlands (2017): “A potential new fungal cell factory: Penicillium subrubescens reveals enriched plant biomass related CAZymes” by Dilokpimol et al
• Presentation of the project at the Globelics in Athens, Greece by Marianna Ralli
• Presentation at European Biomass Conference & Exhibition Stockholm/Sweden (2017): “Technical Production Process for Innovative Antioxidants using Novel Enzymes as Biocatalyst” by Adam, S., Tlatlik, S., Gottschalk, A.
• Presentation at CBM in Vienna (2017): “Feruloyl esterases, effects of glycosylation on activity, stability and immobilization” by Bonzom, C., Chong, S.L. Hüttner, S., Iancu, L., Olsson, L.
• Poster Presentation at 13thBiotrans in Budapest (2017): “Development of improved variants of a fungal feruloyl esterase to replace conventional chemical reactions with eco-friendly bioconversions for cosmetic industries”, by Cerullo,G., Varriale, S., Brás, J. L.A. Fontes, C.M.G.A. Faurè, R., Piechot, A, Jütten, P., Faraco, V.
• Poster presentation at the 2nd Symposium on Plant Biomass Conversion by Fungi, Utrecht, The Netherlands (2017): “Heterologous expression of feruloyl esterase of Aspergillus terreus” by Mäkelä MR, Dilokpimol A, Koskela SM, de Vries RP, Hildén K.
• Poster presentation at the British Mycological Society and British Society for Plant Pathology Joint Presidential Meeting, Nottingham, England (2017): “Heterologous expression of feruloyl esterase of Aspergillus terreus” by Mäkelä MR, Dilokpimol A, Koskela SM, de Vries RP, Hildén K.
• Final presentation of the project at the ECOMONDO conference in Rimini, Italy (2017): “Mining enzymes for green products” by Vincenza Faraco


The final OPTIBIOCAT video was released on YouTube, integrated into the project website and disseminated by all consortium members to their network.

Exploitation of results

OPTIBIOCAT’s research is projected to be exploited in different fields. Companies and academic partners both are involved in the exploitation. The exploitation of data as well as commercialisation is ongoing. Companies are exploiting the expertise developed (e.g. GeneScan, NZYtech) and/or the proof of their technology to support the sale (e.g. ProteoNic) or production (e.g. Dupont), of its commercial product. Academia is exploiting its expertise developed and “advertised” in high quality peer reviewed publications mainly to gain research funding. Also, the materials and expertise is used to continue studies and start novel research lines. Especially academia is also active in sharing materials and knowledge to facilitate the scientific research of the partners.

Each partner has an individual exploitation focus:

• Protein Production Technologies (PROTEONIC, DNL, WWUM)
• Enzyme immobilization and biocatalytic process development (CTH, LTU)
• Bioconversions and conceptiual process design (LTU SUPREN)
• Green Cosmetics (KORRES)
• Communication Methods (BIOCOM)
• Library Generation, HTP Gene Synthesis and Cloning (NZYTECH)
• Genotyping and expression analysis (Genomescan, KNAW, UH, CTH, UNINA)
• Organic synthesis and library generation (TAROS)
• Methods for directed evolution (INRA, UNINA)

The potential impact

The main objective of the directed evolution part of OPTIBIOCAT was the generation of a set of optimized mutant proteins issued from 5 targeted esterases. Novel versatile substrates were designed and synthesized by INRA (X-Fe, 4NTCFe, 4NTC-linker-Fe) and used to develop interesting methodologies that were proven successful as many optimized mutants have been obtained.

These libraries thus represent significant value and will also be used as start point for new projects and new research lines to identify novel enzymes. Several publications in peer-reviewed journals can testify for these achievements – and should be followed by others in coming months- and will act as a testimony for the groups involved.

The main objects, mutant proteins (or mutation information), for the creation of cosmetic and pharmaceutical products containing natural ingredients will be exploited as well as further developed and shared with OPTIBIOCAT partners, to expand the number and type of chemical transformations carried out by enzymes at industrial scale, thereby meeting the need to increase the sustainability of the industrial production.

List of Websites: