Skip to main content

Sustainable Liquid Biofuels from Biomass Biorefining

Final Report Summary - SUNLIBB (Sustainable Liquid Biofuels from Biomass Biorefining)

Executive Summary:
4.1.1 Executive Summary

Compared with fossil fuels, biofuels (such as bioethanol) offer considerable potential benefits in terms of greenhouse gas emissions and sustainability. However, so-called ‘First Generation’ bioethanol is produced from sugar or starch, and using foodstuffs as feedstock in this way could negatively affect food security. Second Generation (2G) bioethanol is produced from inedible crop residues or waste plant materials (lignocellulosic biomass). The potential environmental and sustainability benefits of producing ethanol in this way are clear.

Despite its possible advantages, 2G bioethanol production is currently inefficient and economically uncompetitive. A major reason for this is the recalcitrant nature of lignocellulosic biomass, which is mainly composed of secondary plant cell walls and comprises approximately 75% polysaccharides and 25% lignin. The polysaccharides can be broken down to release sugars that can be fermented to bioethanol, but the complex structure of the walls and the presence of lignin (a phenolic polymer) make this saccharification process very difficult and energy intensive.

SUNLIBB addressed this problem using a holistic biorefinery approach, with the overall aim of increasing the economic viability of 2G bioethanol production in a sustainable and environmentally sensitive way. The SUNLIBB consortium comprised 13 European partners from academia and industry, chosen for their expertise in areas such as cell wall and lignin biochemistry, plant genetics, biomass deconstruction and process engineering. Separate, though closely-linked, work packages worked simultaneously on improving biomass quality to make it easier to break down, as well as improving the efficiency of feedstock processing methods. Other partners worked to identify and optimize the production of added-value products beyond bioethanol that could be obtained during biomass deconstruction. Further work monitored the sustainability and environmental impacts of the new processes, and set the project into context within the relevant regulatory frameworks.

The project focused on lignocellulosic residues from sugarcane, Miscanthus and maize. These 3 crops are closely related C4 grasses, with shared similarities in cell wall composition and structure. Discoveries in one species could therefore be readily exploited in the others.

SUNLIBB worked in close collaboration with a sister consortium in Brazil (CeProBIO). Thus, the project integrated European expertise in genomics, plant science and green chemistry with Brazilian expertise in sugarcane breeding and bioethanol process engineering. Both sides benefitted enormously from the bilateral sharing of knowledge, experience and data.

During the project, significant advances were made towards identifying the genes controlling biomass quality, which could pave the way for breeding plant biomass tailored for biorefinery applications. Our knowledge of the components of the cell wall matrix was increased substantially, leading to a new model suggesting how some of the cell wall polysaccharides interact. A new technique, CSPP (Candidate Substrate Product Pairs), was developed and used to uncover the complex metabolic pathways involved in lignin production and to identify a set of novel lignin biosynthesis genes. This work has greatly increased our understanding of the major barriers to biomass digestibility and how these may be overcome through plant breeding. To investigate adding value to the biorefinery processes, lignin degradation products (potentially valuable by-products of biorefining) were analysed. Waxes were extracted from biomass and shown to have potential value as antifoaming agents.
A High-Throughput Saccharification assay developed in a previous project was used on several thousand different plant samples to monitor biomass, and novel data were generated on the impacts of different biomass pre-treatments and deconstruction methods. Pilot scale fermentations were carried out and techno-economic analyses were performed. An exciting result showed that the efficiency of ethanol production could be increased by improving the genetics of the biomass, as mentioned above. Data integrated from laboratory and pilot experiments were used in process integration studies, quantitative environmental assessment, based on life-cycle assessment and qualitative assessment, including socio-economic considerations, to examine the potential cost and environmental benefits of the biorefinery process within the context of identified regulatory and other reporting frameworks.

Insights from this work will help to keep Europe at the forefront of the development of commercial 2G bioethanol production.

Project Context and Objectives:
4.1.2 Summary description of project context and main objectives
4.1.2.1 Context
Society in the 21st century faces huge problems because of uncertainty over future oil supplies and the detrimental effects of global climate change caused by releasing greenhouse gases from fossil fuels. There is a recognized need for renewable, carbon-neutral fuels and industrial feedstocks, and biofuels offer a potential solution. Liquid biofuels, such as bioethanol (produced by fermentation of plant-derived sugars) and biodiesel (produced from plant or animal oils) can be used as liquid transportation fuels. Statistics (Eurostat) show that transportation fuels derived from petroleum currently account for more than 30% of EU energy consumption. Replacing a proportion of this with renewable alternatives would lessen the dependence of EU member states on oil imports and increase the stability and security of fuel supply. In addition, preventing dangerous climate change is a strategic priority for the EU, and ambitious targets have been set for progressively reducing greenhouse gas emissions up to 2050.
Worldwide, bioethanol production has increased substantially over the last decade. In Brazil, a world-leading, economically competitive bioethanol industry has developed, using sugar from sugarcane as feedstock and producing demonstrable reductions in greenhouse gas emissions. Unfortunately, similar gains cannot easily be achieved with other crops where the energy balance is less favourable. For example, although the USA is the world’s top producer of bioethanol from corn, the high inputs required to grow maize for this purpose render its carbon footprint poor. The case is similar for wheat grown for bioethanol in Europe. There is also increasing concern that the production of bioethanol from edible feedstocks like cereal grains and sugarcane is putting pressure on global food markets, by increasing food prices, as well as intensifying competition for land use. Clearly, increased production of these so-called ‘First Generation’ biofuels is in itself unsustainable.
Second Generation (2G) bioethanol is produced from inedible crop residues, dedicated low-input biomass crops or waste plant materials that do not form part of the human food chain. Hence, the potential environmental and sustainability benefits of producing bioethanol from this lignocellulosic biomass are obvious.
Plant biomass (lignocellulose) is renewable and abundantly plentiful. It is mainly composed of secondary plant cell walls, which comprise roughly 75% polysaccharides and 25% lignin. In theory, the polysaccharides can be broken down to sugars (saccharification) and fermented to bioethanol or made into other useful products, such as bioplastics or fine chemicals. Integrated processing systems could add even more value, by allowing multiple different products to be produced from the same biomass; the ‘Biorefinery’ concept. However, the complex structure of cell walls, consisting of a network of cellulose microfibrils and matrix polysaccharides (such as hemicellulose and pectin) encrusted by lignin (a phenolic polymer) makes them very resistant to degradation. This means that the process of converting lignocellulosic biomass into fermentable sugars is currently too energy intensive and costly to be commercially viable.
SUNLIBB aimed to address this problem using a holistic, biorefinery approach. A substantial part of the project investigated ways to overcome the barriers to saccharification described above. For example, one of the main objectives was to find out if biomass raw material could be tailored, through plant breeding, to be broken down more easily. Ways to improve the saccharification process itself were also investigated, and the feasibility of producing valuable chemicals alongside bioethanol was determined.
The work of SUNLIBB focused on lignocellulosic biomass from sugarcane, maize and Miscanthus. All these 3 crops are C4 grasses, which share the most efficient form of photosynthesis (C4) and have similar cell wall composition and structure. Sugarcane was chosen because of its importance in the Brazilian bioethanol industry and the vast amount of research that has been done on this crop in Brazil. Sugarcane is a highly productive crop, yet actual sugar makes up only 10-15% of its biomass. Substantial quantities of ‘bagasse’ remain after conventional sugar extraction and this mainly comprises lignocellulose which could be used to make bioethanol, if this were profitable. As a tropical crop, sugarcane is not suitable for growth in Europe. Maize, however, is a major crop grown in many areas of the world, including Europe and Brazil. It is mainly grown for grain for food or feed, or as mentioned earlier, as feedstock for first generation bioethanol production in the USA. After the grain (corn) is used, large volumes of stover (stems, leaves and cobs) remain; this lignocellulosic material could be available for bioethanol production. Maize is a ‘model’ crop, with a sequenced genome and a wealth of breeders’ experience and genetic resources behind it, making it a promising candidate for biomass improvement through breeding. Miscanthus is a perennial grass, which can produce high yields of biomass in low-input agriculture and is viewed as a promising feedstock for sustainable biofuel production in Europe.
In order to begin to engineer improved biomass for biorefinery applications, it is first necessary to develop a clear understanding of the factors limiting biomass digestibility and of the genes and enzymes involved in constructing plant cell walls. To do this, the SUNLIBB programme brought together leading European cell wall scientists and experts in lignin research and plant genomics. In this context, SUNLIBB continued the work of RENEWALL, a previous EU-funded FP7 project that investigated lignocellulose digestibility in a range of plant species. In line with the biorefinery approach, the SUNLIBB consortium also included process chemists, to develop methods to extract added-value products from biomass, as well as process engineers, expert in the integration of bioprocesses in the production of biofuels. The team also included environmental scientists, with skills in assessing the impacts of biorefineries in economic, environmental and societal terms. The 13 SUNLIBB partners included several different European Universities, as well as industrial participants and access to two experimental biorefinery pilot plants. The work programme was divided into separate, though closely-linked Work Packages, each with a leader responsible for timely completion of research objectives and reporting of progress.

Throughout the SUNLIBB programme, the project collaborated closely with a sister consortium called CeProBIO in Brazil. This is an initiative aimed at developing sustainable second generation bioethanol from (sugarcane) biomass. The aim was to combine European and Brazilian research strengths to bring mutual benefits and research synergies to both projects. Thus, the project integrated European expertise in genomics, plant science and green chemistry with Brazilian expertise in sugarcane breeding and bioethanol process engineering and both sides benefitted enormously from the bilateral sharing of knowledge, experience and data.

4.1.2.2 Summary of Objectives
The aims of SUNLIBB were to:

• Improve the feedstock quality of lignocellulose in biofuels crops, to try to make 2G ethanol production cost-effective. To do this, modern crop breeding approaches were combined with cutting-edge plant cell wall research, to produce biomass crops with improved digestibility characteristics (WP1). Gene identification programmes were established, to discover genes for the following traits; improved polysaccharide content, improved digestibility and lignin that is easier to break down (WPs 2-4).
• Add value to the overall process of biomass conversion in biorefining, by upgrading residues and by-products and producing other value streams from the feedstock, besides bioethanol. To do this, technologies were developed to extract valuable materials such as waxes and to generate useful molecules from the breakdown of lignin. These processes were developed at lab scale and tested in pilot processing plants (WP 6).
• Improve the conversion process by which sugars are produced for fermentation. Experiments using the improved feedstocks from WPs1-4 were integrated with unique combinations of pre-treatments and new enzymes (from leading science programmes in Europe and Brazil) for conversion of lignocellulosic material (WP5).
• Develop integrated processes that capture maximum value from lignocellulosic biomass. To do this, the improved feedstock, added-value product extractions and improved conversion processes listed above were brought together in pilot biorefineries. Techno-economic feasibility studies were carried out using the data generated (WP7).
• Ensure that the new processes developed fulfil sustainability requirements, by reducing emissions of GHG and other pollutants, reducing impacts on biodiversity and not affecting food prices. To do this, Life-Cycle Assessment (LCA) models relevant to biorefineries were developed in WP8, to assess the full spectrum of impacts, using data from the other WPs, partner organisations and public sources.

It was anticipated that SUNLIBB, working in cooperation with CeProBIO, would deliver significant breakthroughs in the efficiency and economic viability of second generation bioethanol production through biorefining.

Project Results:
4.1.3 Description of Main Results
As indicated in the previous section, the overall strategy of the work plan was arranged so that WPs 1 to 4 were concerned with genetic improvement of lignocellulosic biomass as a feedstock. These WPs aimed to identify plant genes that control different components of biomass. In WPs 1 & 2, genetic populations were assembled and association studies were carried out to identify quantitative trait loci (QTL) for desirable biomass traits. In WPs 3 & 4, a reverse genetics approach was taken, to predict gene targets likely to affect lignocellulose quality. This information was sent to WP2, where transcriptome analysis was also used to discover target genes for which knock-out mutants could be generated (WP1). The mutant lines were then analysed in WPs 3 & 4. This research was also integrated with the cell wall deconstruction work from WP5; highlighting the close collaboration between SUNLIBB work packages. The intention was to start with maize (a model plant) and then test hypotheses in Miscanthus and sugarcane.

Progress in each Work Package was monitored by fulfilment of a set of projects called Deliverables, which were described in the original Description of Work. Deliverables Reports, which contain details of methods and techniques and experimental data have been submitted to the EC, and are available on the SUNLIBB website. They are referred to in the text of the results section, below. A list of completion dates and dissemination levels of all the Deliverables is presented in the Periodic Reports submitted to the EC.

1.1 Work Package 1
Genetic approaches to improving biomass quality for cellulosic biofuel production
1.1.1 Objectives
The objectives of WP1 were to develop robust methods for the improvement of biomass quality for biorefinery and biofuel production in maize in an agronomically relevant context. The transferability of this knowledge to other grasses was assessed for Miscanthus and, by our Brazilian partners, for sugarcane. WP1 also provided biomass from breeding populations of maize and Miscanthus for HT saccharification assays in WP5.

1.1.2 Task 1.1 Develop HT methods for cell wall analysis
The objective of this task was to develop High-Throughput (HT) methods to quantify cell wall composition and structure. Near-InfraRed Spectroscopy (NIRS) was a suitable such method. NIRS predictive equations for cell wall composition were developed previously (Yves Barrière) and were successfully transferred to the INRA-Versailles facility (Partner 6, P6) (Deliverable D1.5). In addition, novel NIRS equations were established for cell wall related traits and biomass degradability, using selected samples from SUNLIBB field trials carried out in France (2011-2013).
1.1.3 Task 1.2 Identification of QTL for biomass composition and saccharification potential in mapping populations and evaluation of the interactions between agronomic performance and biomass composition
The objectives of this task were to develop or to use mapping populations in maize and Miscanthus, in order to map QTL (Quantitative Trait Loci) involved in the variation of biomass composition and saccharification potential.
For maize, INRA (P6) grew a panel of 18 genotypes in two locations in France (Mauguio and Gif-sur-Yvette) during summer 2013. Stover samples were harvested at silage stage and sent to P1 (University of York) for quantification of saccharification potential. Plants grown in Mauguio were cultivated under two contrasted water regimes, allowing insights to be obtained into the interactions between yield, drought tolerance and saccharification (D1.7). Observed variation allowed us to select lines with contrasting lignocellulosic biomass characteristics, to send for pilot scale experiments.
70 selected lines derived from a cross between two recombinant inbred lines (RILs) were cultivated, selfed and harvested in a nursery at Gif-sur-Yvette in 2013. These lines were genotyped with molecular markers located within the confidence interval of a QTL for digestibility. Co-segregation analysis between allelic and digestibility variations present in this set of lines allowed us to validate and to reduce the size of the digestibility QTL under investigation (D1.13).
In order to validate the presence of the maize QTL for saccharification located on chromosome 1 (D1.10) selected recombinant inbred lines (9 RILs) were cultivated in a field trial at Gif-sur-Yvette. Crosses were performed according to alleles carried by the lines at the studied QTL. Samples of stover from all the cultivated lines were sent to P1 (University of York). F1 plants derived from lines showing the most contrasting saccharification values were cultivated and selfed in Chile in winter 2013-2014, (Chile being a standard location for contract maize breeding).
For Miscanthus, a Miscanthus sinensis mapping population (BIOMIS), consisting of 1 plant per accession was available at Wageningen University (P13) prior to the start of the SUNLIBB project. There was a pre-existing preliminary genetic map for this population (Atienza et al, 2002). Biomass from the 114 accessions was collected and biomass quality assessed. Biochemical analyses were carried out by P13, to determine the extent to which the cellulose, hemicellulose and lignin contents were segregating in this population, while saccharification data were generated by partner P1. These analyses showed that there was considerable variation in the population for cell wall traits, though no genetic markers (QTL) were found strongly associated with cell wall traits. This was partly due to the small number of markers that were only available for a sub-set of the lines. However, various sites on the map suggested the presence of QTL (as seen for different LG; D1.11). Given this result, P13 decided to proceed with a more dedicated population; the SUNLIBB mapping population.
The SUNLIBB dedicated mapping population originated from parents showing contrasting differences in cell wall composition. Approximately 200 individual plants were transplanted to soil, in three replicate randomized plots in May 2012. Harvesting of young leaf material was completed at the end of May 2013 for DNA isolation. Genotyping by sequence (GBS) was performed. The data generated were analysed by P13. A set of morphological and biomass quality traits was measured. These include heading date, plant height, dry matter yield, stem yield, biochemical cell wall composition and saccharification efficiency, and have been or will be collected for 2 consecutive years. Triplicate stem samples from the SUNLIBB population were ground up and sent to P1 for saccharification analysis. Preliminary analysis of these data showed that there was large variation for cell wall traits in the SUNLIBB population. QTL mapping was performed by the group in Wageningen (P13). In addition, 10kg of two Miscanthus genotypes with contrasting saccharification potential, selected on the basis of previous measurements of 20 contrasting genotypes analysed in 2013 by P1, were harvested and sent to P1 for the pilot scale experiments performed in WP5 & WP7.
1.1.4 Task 1.3 Assessing the potential of candidate genes as markers for improved biomass digestibility
Biogemma (P4) screened their maize population of 40 000 transposon insertion lines for knockout mutations in candidate genes identified in WPs 2-4. Three successive screenings (2009, 2010 and 2011) were conducted. 12 mutants were found for a total of 31 genes submitted. A backcross process was necessary to obtain BC2S1 plants for phenotyping. Hence, 3 years were necessary from the screening to the production of samples for analysis. The first screening was conducted prior to the start of the project in order to deliver all the mutants and samples within the time frame of SUNLIBB.
More than 650 samples were harvested in 2013 and 2014 (January in the greenhouse and August in the field). These samples were sent to SUNLIBB partners for analysis of polysaccharide composition (WP3 / P9) and saccharification assays (WP5 / P1).

1.1.5 References
Atienza, S.G. Z. Satovic, K.K. Petersen, O. Dolstra & A. Martín, (2002). Preliminary genetic linkage map of Miscanthus sinensis with RAPD markers. Theor Appl Genet. 105: 946–952

1.2 Work Package 2
Transcriptomic and genomic resources for biomass improvement
1.2.1 Objectives
To use transcriptomic data to help identify genes involved in cell wall biosynthesis, which likely have an impact on cell wall composition and saccharification potential in maize, Miscanthus and sugarcane. These genes were critical to research in WPs 1, 3 and 4 and will eventually provide markers for breeding programmes.

1.2.2 Task 2.1 Generating transcriptomic data from Miscanthus
Task 2.1 aimed to generate transcriptomic data relating to genes involved in the secondary cell wall biosynthesis of Miscanthus. This analysis was performed through the identification of ortholog genes from model species and other C4 plants, followed by expression analysis of key cell wall genes during different stages of Miscanthus development.
Given the lack of an assembled genome for Miscanthus, new sources of sequence data for Miscanthus had to be generated. Recently two sets of Miscanthus sinensis transcriptome data (Swaminathan et al., 2012 and Straub et al., 2013), have been released and these were used by P13 as a basis to develop an orthology database, to assist in identification of key cell wall biosynthetic genes in Miscanthus sinensis. In this database, besides the transcriptomes of Miscanthus, sequences of other C4 plants were used to search for orthologous genes. Currently, the database is being made into a user-friendly online platform. The main goal of the development of this tool is to enable biologists, without extensive bioinformatics knowledge, to efficiently study de novo transcriptome data from any orphan crop, (such as Miscanthus). The tool has been designed to facilitate the following common tasks: i) identification of candidate transcripts through phylogenetic relationships, and ii) design of specific and degenerate primers for the analyses of known and novel transcripts. Miscanthus orthologs and paralogs of known maize genes involved in major cell wall biosynthetic pathways were identified and transcripts will now be quantified on the selected contrasting internode samples.

1.2.3 Task 2.2 Mining transcriptomic data
The aim of this task was to uncover key genes involved in the synthesis and modification of the different components of the cell wall. In Miscanthus, the main focus has been to: i) compare Miscanthus genotypes with differences in cell wall composition, and ii) analyse selected genotypes at different developmental stages. For maize, data were extracted from 2 datasets; one produced in 2007 with the Arizona 46K array and one produced in 2009 with a 45K Biogemma Nimblegen array. For each case, the same two lines were used: F874 (low-degradable line) and F7019 (high-degradable line). At 14 days, three developmental stages were available for 2007: S2 (male flowering), S3 (female flowering) and S4 (female flowering) and 4 stages for 2009: S1 (9th visible leaf), S2 (male flowering), S3 (female flowering), S4 (female flowering). For each dataset, comparisons were made between the 2 lines for stage S2, S3 and S4 (ex: F874 S1 / F7019 S1) and between successive stages in each line (ex: F874 S1 vs F874 S2). The two datasets are complementary, with data on the same lines produced in 2 different trials, on the basis that the 2 array technologies were not the same (which gives more coverage, since some genes are covered by the Nimblegen array and not by the Arizona Array and vice-versa) and probes for the same gene in the 2 arrays are not located in the same part. In some cases, this can produce different profiles for the same gene.
1.2.3.1 Results for Miscanthus
Based on the determination of cell wall composition of mature plants, seven accessions were selected as the material to be used for Miscanthus transcriptome analysis. Shoots of each genotype were sorted into different groups according to their height, to provide a detailed profile of cell wall formation during development. Three internodes, starting from the second internode emerged from soil, were harvested and later divided into three sections. Samples were collected for both biochemical and transcriptomic analyses. All samples were biochemically analysed, and contrasting internode sections were selected.
1.2.3.2 Results for maize
In WP1, P6 identified a QTL for saccharification in a maize Recombinant Inbred Line (RIL) population derived from the cross F288xF271. The goal of this analysis was to try to identify clear candidate genes by cross-checking the information available in the SUNLIBB project. The confidence interval of this QTL was converted into physical positions relative to the reference maize genome B73 and the coordinates were sent to P4 for data mining.
The aims were to:
- identify the gene models present at this locus.
- provide functional annotation of these genes.
- add transcriptome information available in the project to help candidate selection.
- determine whether one of the genes was part of one of the strong candidate gene lists established by experts in WP3 and WP4 (deliverable D2.2 and D4.1).
If a strong candidate were found to be located under the detected QTL, it would be tested by P6 by sequencing the alleles of both parental lines of the mapping population, to identify the putative causal polymorphism(s). If no obvious candidate appeared, P6 would continue a classical fine mapping approach.
All the analyses were done on the maize B73 reference genome version1 (refgenv1, v4a53).
Gene models were taken from the Filter Gene Set available on for details regarding FilterGeneSet see http://www.maizesequence.org/info/website/help/index.html.
The confidence interval of the QTL covered a region on Chromosome 1 that includes 339 gene models. All these gene models have been compared to the list of genes of interest established in WP3 and WP4 (for details see D2.2 and D4.1). None of the genes of interest were identified under the QTL. Even if some genes showed annotation more-or-less related to the cell wall (cellulase, sucrose synthase, etc.), these annotations have mainly putative function and no gene with a robust and saccharification-related annotation (listed in WP3 and WP4) was found to co-localize with the detected QTL. At this stage, no genes appeared to be strong/obvious candidates to explain the QTL. Partners involved in this deliverable decided to pursue the size reduction of the QTL and not spend effort on sequencing unclear candidates.
In order to provide maximum information on the genes in the confidence interval, P4 enriched the data by adding its own functional annotation based on Blast against Swissprot and TrEMBL (from http://www.ebi.ac.uk/). Transcriptomic data provided by P4 in SUNLIBB have also been linked to the genes to reinforce possible candidates based on functional annotation. These data identify differentially expressed maize genes in lines of contrasted cell wall degradability (F874, DtLG, F2, F7019) and in mutants in the lignin pathway (CAD, CCR and C4H) (data produced in SUNLIBB for WP4).
1.2.3.3 Joint-collaborations with CeProBIO partners
Total RNA of a set of three contrasting Miscanthus genotypes was prepared by P13 and sent to Glaucia Souza’s group (CeProBIO) at the end of January 2014. This was tested against their available sugarcane chip array. The generated data were received by P13 in mid-August 2014 and are currently under analysis for the identification of key genes in the cell wall biosynthetic pathway.
Stem and leaf samples from the selection of seven Miscanthus accessions were harvested about 4 weeks after the onset of heading (which corresponds to full grown stems), and sent by P13 to Marcos Buckeridge’s group (CeProBIO). The results of this work are being described in a manuscript under preparation.

1.2.4 References
Straub, D., Yang H., Liu Y. & Ludewig U. (2013). Transcriptomic and proteomic comparison of two Miscanthus genotypes: high biomass correlates with investment in primary carbon assimilation and decreased secondary metabolism. Plant and Soil. 372:151-165.
Swaminathan K., Chae WB., Mitros T., Varala K., Xie L., Barling A., Glowacka K., Hall M., Jezowski S., Ming R., Hudson M., Juvik JA., Rokhsar DS. & Moose SP. (2012). A framework genetic map for Miscanthus sinensis from RNAseq-based markers shows recent tetraploidy. BMC Genomics. 13:142

1.3 Work Package 3
Understanding cell wall polysaccharides to underpin biomass improvement
1.3.1 Objectives
The aim of WP3 was to extend basic understanding of biomass polysaccharide composition and structure in maize and Miscanthus and investigate the function of genes potentially involved in their biosynthesis, using data from WP2. Information from WP3 will help identify key genes for breeding strategies, identify genes that link polysaccharides with lignin (WP4), generate vital information on polysaccharide composition to allow better biomass digestion (WP5) and identify strategies for added-value polysaccharide-based products for WP6 and WP7.

1.3.2 Task 3.1: Characterizing cell wall polysaccharide composition, quantity and structure
P9 described the oligosaccharides composing the xylan hemicellulose matrix in Miscanthus generated with the xylanases GH10 and GH11, and has optimised the protocol for feruloylated groups. Therefore, we now have an extended view of xylan polysaccharide structure regarding the glucuronic acid, arabinose, xylose and feruloyl side substitutions. Moreover, we know that this type of structure is present in most grasses, including maize, sugarcane and Brachypodium.
Furthermore, P13 measured cell wall polysaccharide composition in different sets of Miscanthus samples. These included two Miscanthus sinensis mapping populations, BIOMIS and SUNLIBB, and a large collection of Miscanthus genotypes, property of P13. The SUNLIBB population was further analysed in WP5 for saccharification efficiency and genotyped by sequencing (GBS) in WP1. Analysis of cell walls in the Miscanthus collection resulted in the selection of 7 genotypes contrasting for cellulose, hemicellulose and lignin content. These 7 lines were further used for transcriptomics analysis in WP2.

1.3.3 Task 3.2 Mining transcriptomic data for polysaccharide biosynthetic genes

Using the gene lists of D3.1 and using WP2 maize data, P4 generated a list of 62 target maize genes; moreover, P9 gained access to the SucEST database (http://sucest-fun.org) from our CeProBIO Brazilian partner and identified 24 sugarcane genes coding for proteins involved in hemicellulose biosynthesis. They are glycosyl transferases (GT) from the families 8, 43, 47, 61. GT8, GT43 and GT47 are involved in the reducing end formation and the β-1,4 xylosidic linkage of xylose residues forming the xylan backbone (table 3.1 excel file D3.4). Some GT8 are called GUX (glucuronic acid substitution of xylan) as they add the glucuronic acid (GlcU) residue onto the xylan backbone. GT61 adds the arabinose residue onto the xylan backbone. In addition some enzymes of the RWA (Reduce cell Wall Acetylation) family, that add acetyl groups onto the xylan residues, were added to the list.
The main focus for identification of key genes involved in the synthesis of polysaccharides in the Miscanthus cell wall has been on genes of the GT8-GUX and GT61 families. Degenerate primers have been designed by P13 for three GT8-GUX genes (GT8-GUX1, 2 and 3), and gene expression has been assessed in the 7 M. sinensis genotypes chosen for contrasting cell wall composition. Miscanthus RNA has been hybridized to the SUCEST array, to identify genes in sugar cane with orthologs most highly expressed in Miscanthus.

1.3.4 Task 3.3 Demonstrating gene function
Since the beginning of SUNLIBB, P9 has demonstrated the activities of eight new enzymes.
Among those are GT61, TaXAT1 and TaXAT2 and OsXAT2 and OsXAT3 (TaXAT2 homologs responsible for the addition of the arabinose substitution on the xylan backbone.) The DUF579 family has been shown to be involved in the addition of the 4-O-methyl group onto the glucuronic acid residues on xylan. Hence they are known as glucuronoxylan methyltransferases. The genes are: At1g33800, At4g09990, At1g09610, At1g71690 and form the Arabidopsis GXM family.
We have conducted experiments to decipher the activities of Arabidopsis IRX9, IRX14, IRX10 and IRX15, plus related genes, and have generated evidence that IRX14 requires sugar nucleotide binding to function in synthesis of the xylan backbone. Additionally, we have shown that DUF288/GT75R is involved in cellulose synthesis.
To identify maize mutants with altered cell wall polysaccharides, three successive screenings (in 2009, 2010 and 2011) were conducted using the Biogemma mutants collection. 12 mutants were found from a total of 31 genes submitted (Deliverable D1.8 and D1.12).
Mutants and samples for WP3 analysis by P1 and P9 were produced in the context of WP1. All 12 identified mutants were delivered (D1.8 and D1.12). In this reporting period, more than 650 samples were produced in 2013 (January in the greenhouse, August in the field) and in 2014 (field). These internodes and leaf samples were sent to P1 and P9 for analysis of polysaccharide composition and saccharification assay in WP3 and WP5.
Note that for 2 genes, GT61_3_12a and GT61_3_12, the evaluation of mutants was done in the heterozygous state. The homozygous mutation GT61_3_12a seemed to be lethal. For GT_61_3_12, the gene-transposon junction could not be sequenced and the homozygous mutant and heterozygote could not be distinguished. For these 2 genes, a homozygous wild-type was identified. For all the remaining genes, analysis by P1 and P9 was done on the homozygous wild-type vs homozygous mutants.
P9 also analysed the cell walls of the final set of maize mutants altered in cell wall polysaccharide biosynthesis. The results were summarised in D3.6.

1.4 Work Package 4
Understanding and obtaining value from lignin
1.4.1 Objectives
The aims of WP4 were to (1) mine transcriptomic data from maize and Arabidopsis for lignin biosynthesis genes, including transcription factors, in order to (2) make gene silencing or overexpression constructs for maize and Miscanthus transformation to improve biomass for saccharification potential. (3) WP4 also aimed to reveal novel aromatic pathways and obtain insight into altered fluxes through these pathways, by profiling the aromatic metabolome of a set of maize mutants with altered lignification and by integration of the data with transcript profiling data.
Clarification of Objectives (in response to request by Dr. J-L Dubois, EC Reviewer)
One aim of WP4 was to find out more about the genes involved in making lignin, by looking at data from maize and Arabidopsis that tell us which genes are being switched on at any particular time (e.g. when lignin is being made by the plant). Once such genes were discovered, WP4 aimed to switch them off or switch them permanently on, in transgenic maize and Miscanthus plants, and observe whether this had any effect on the digestibility of the lignocellulose from these plants. This information could then be used (in future plant breeding) to improve the digestibility of biomass. WP4 also aimed to investigate some mutant maize plants that were known to have differences in lignification from normal plants. WP4 would compare some of the chemical compounds made by these mutant plants with those from normal plants, and combine this knowledge with data about the genes involved. This would reveal valuable information about the pathways involved in lignin biosynthesis.
1.4.2 Task 4.1 Mining transcriptomic data for genes involved in lignin biosynthesis
P3 identified orthologs likely to be involved in lignification in maize. An additional co-expression experiment was set up by P4. Internode samples, at 4 developmental stages, from mutants and corresponding wild-types for maize CCR1, C4H and CAD2 genes were produced in the field by P4 in summer 2011.
144 samples were sent to P12 for metabolomics studies in November 2011. RNA extraction, microarray hybridization and data normalization for transcriptomic analysis were carried out. Normalised expression data were sent to P12 for systems biology approaches and P12 analysed the data according to their in-house protocol. Statistical analysis of the dataset showed that the expression profiles of lignin biosynthesis genes in the youngest vegetative stage were different from those of the reproductive stages, as expected. Principal Components Analysis (PCA) showed that the c4h mutant samples clustered separately from the control line, whereas for cad this was not the case.
Hence, despite the clear cell wall phenotype of the cad mutant, the overall transcriptome was not highly altered. The ccr samples could not be distinguished on the basis of expression profiles, which also shows that the phenotype was probably lost during the backcrossing process. A preliminary search for differential gene expression revealed that lignin biosynthetic genes that are situated high in the phenylpropanoid pathway are upregulated in the mutants. This could be explained by a feedback mechanism upon disturbance downstream of these genes.
To further characterize the maize plants, P4 also carried out NIRS analyses on mature-stage stem and leaf samples. NIRS data were sent to P12. An aliquot of the NIRS powder was sent to P12 for complementary saccharification and lignin content analysis. The parameters used in NIRS cover a wide range of maize cell wall characteristics and are thus very informative in describing possible phenotypes caused by the mutations. In the c4h mutant, there was an altered value for ADL (Acid Detergent Lignin) that was reflected in a generally affected abundance of extractable lignin monomers H, G and S. This altered abundance of lignin resulted in enhanced cell wall digestibility and improved energy values for milk and meat production. The direct product of the C4H enzyme, p-coumaric acid (pCA), was altered in its abundance; a good indicator for functional disruption of the C4H gene. Both 5-5- and 8-O-4-linked diferulic acid, which play a role in polysaccharide crosslinking showed altered abundances in the c4h mutants, compared with the controls. In the cad2 mutant, the cell wall characteristics were predicted with NIRS estimation. Differences in cell wall per dry weight can be explained by shifts in cellulose and ADL, accompanied by altered abundance of extractable lignin monomers (H, G and S). In the cad mutants, we observed differences in soluble sugars and their digestibility. We also found different abundances in pCA and 5-5- and 8-O-4-linked diferulic acid in the mutant, compared with the control. In the ccr mutant, only three of all predicted NIRS parameters were significantly altered, suggesting that the phenotype observed in the initial insertional mutant was lost during backcrossing.
Concerning Miscanthus, to circumvent problems due to the lack of a genome sequence, P13 focussed on quantitative expression analyses to identify orthologous genes. The amplification of 9 genes was tested (COMT, PAL, LAC, 4CL, CCoAOMT, CAD, PER, C4H and CCR) and partial sequences were obtained for 4 of them (CCoAOMT, CAD, PER, C4H and CCR), fulfilling the objectives of D4.1.

1.4.3 Task 4.2 Transformation of maize and Miscanthus with new genes involved in lignin biosynthesis
The beneficial enhanced saccharification efficiency of the naturally occurring ‘brown midrib’ (bm) mutants in maize and sorghum shows that disturbing lignin biosynthesis can be a good strategy for lignocellulosic crop improvement. In that perspective, the cultivation of these bms would form an improved supply of biomass for second generation biofuels. Unfortunately, these lines perform less well in the field and thus generate lower yields for both whole silage and grain. In order to compensate for the yield penalty of these mutants, it would be desirable to combine different beneficial traits to enhance saccharification in one and the same commercial line. This process of ‘gene or trait stacking’ is most easily studied by the use of transgenics made in a homozygous genetic background (a full inbred line). To explore the feasibility of improving lignocellulosic crops for second generation biofuels by means of genetic engineering, we initiated transformation experiments in a maize inbred line. This enabled us to easily compare and combine different traits. Traits can be combined by crossing or by ‘supertransformation’ of a particular transgenic line.
As an initial target, we chose the maize CINNAMYL ALCOHOL DEHYDROGENASE2 (ZmCAD2) gene for downregulation using a hairpin construct. The ZmCAD2 gene is defective in the maize bm1 mutant and is thus a good target. Also Arabidopsis and Brachypodium plants disrupted in the CAD gene display enhanced saccharification potential.
Several RNAi CAD lines with reduced CAD activity were isolated by P12 in screening experiments. However, no differences in acetyl bromide total lignin, lignin composition and saccharification potential were found in the stem material (see D4.2). We therefore conclude that the investigated approach, which is down-regulation (of the ZmCAD2 gene) using an RNAi approach with the UBIL promoter, was not successful in creating a cad mutant phenotype. The endogenous CAD transcript level was not down-regulated enough to cause a loss-of-function phenotype. Nevertheless, we can conclude that the transformation platform itself is very efficient and successful in producing transgenic lines that are resistant to the BASTA herbicide. When our experiments began, only the maize UBIL promoter was available, which is known to be active mainly in meristematic tissue. In our case, this was less suitable because we were trying to down-regulate a gene that is expressed in differentiated cells undergoing secondary thickening. Recently, a set of Brachypodium promoters were tested in maize tissue. These promoters were found to be 300 times stronger and possibly also active in mature tissues and may be more suitable for use in future experiments of this kind (Coussens et al. (2012) “Brachypodium distachyon promoters as efficient building blocks for transgenic research in maize.” J Exp Bot. 63(11):4263-73).

P13 (in WP4) tried to develop a protocol for efficient transformation of Miscanthus sinensis and to generate transgenic lines with up/down regulation in cell wall genes. Several protocols were tested but all resulted in very few transformants, with transformation efficiency below 0.1%. In another attempt, a total of 60 genotypes were tested for callus formation and regeneration efficiency. This step proved to be extremely time consuming, and together with the search for novel vectors for transformation, Agrobacterium strains and the testing out of suitable protocols, Miscanthus transformation proved not to be feasible within the time frame of SUNLIBB. P13 is still working on the improvement of the protocol, based on a new patent filed for Miscanthus transformation, and using the vectors developed at VIB, which are based on Brachypodium promoters and which have already been successfully tested in maize (Coussens et al., (2012). Once the protocol is established, our next step will be to silence genes identified in WP2.

1.4.4 Task 4.3 Phenolic pathway discovery by targeted metabolomics and integration into metabolic maps
Phenolic profiling can be used to identify the aromatic molecules that are present in bio-energy grasses. P12 used the newly developed CSPP method to increase the throughput of metabolite identification. The CSPP method makes use of high-resolution mass spectral data only, and is based on the mass differences between known substrates and products. These differ by a mass that is typical for enzymatic reaction(s), e.g. methylation, reduction, hexosylation, hydroxylation, etc. In addition, the expected shift in retention time during UHPLC-separation is taken into account. The combination of these two parameters (mass difference and retention time) is calculated for each peak pair in the chromatogram. In addition, information about correlation in abundances of the possible substrate and product can then be added, as two correlated molecules have a higher chance of belonging to the same molecular class. Furthermore, the similarity of the MS2 data of both the possible substrate and the possible product can be added, because similar molecules often lead to similar fragmentations. This whole procedure has been optimized for phenolics in Arabidopsis and published (Morreel et al., (2014) “Systematic structural characterization of metabolites in Arabidopsis via candidate substrate-product pair networks,” Plant Cell 26:929-945).
Subsequently, the CSPP technique was used to analyze maize and sugarcane. For maize, a set of 44 stem extracts (22 young and 22 mature) was analyzed via high-resolution UHPLC-MS. In doing so, P12 was able to annotate 94 compounds: 12 benzoxazinoids, 3 benzenoids, 15 phenylpropanoids, 11 flavonoids, 14 oligolignols, 22 flavonolignans, 15 aromatic esters and 2 other compounds (see D4.3). As such, a catalogue of aromatic molecules from maize was created; the majority of which had never been reported before.
In addition, transcriptomic and metabolomic studies were performed on a set of lignin mutants in genes ZmCCR1, ZmC4H1 and ZmCAD2 (described in D1.6). Biogemma (P3) produced the material in the field. Six or four replicates were used for the metabolomic and transcriptomic assays, respectively. The design was reported in D4.4.
During the project it became clear that Zmccr1 appeared to have lost its intended mutation, so the generated data were not further investigated. In addition, questions arose about the mutation in zmC4H1. The zmcad2 mutant appeared to be a very valuable line for further investigation. To investigate whether ZmCAD2 perturbation affected expression of genes within the same pathway, fold changes in gene expression in the zmcad2 mutant compared to control samples were determined (Table 3.4.4). In contrast to the zmc4h1 mutant, no major changes in expression levels were present in the V10 stage of zmcad2 mutants, except ZmCAD2 itself (GRMZM2G110175), which was highly down-regulated.
In order to investigate the general effects of ZmC4H1 perturbation on gene expression levels beyond the investigated phenylpropanoid pathway, a GO enrichment analysis was performed, as reported in D4.4.
Next, the metabolites that were previously identified via the CSPP algorithm (presented in D4.3) were mapped on a metabolic pathway. Subsequently, their abundance in the mutants as compared to the abundance in wild-type plants was calculated and appropriate statistics were applied. Then, their relative abundances were mapped on the metabolic pathways, to facilitate interpretation of the data.

1.5 Work Package 5
Biomass Deconstruction
1.5.1 Objectives
The deconstruction of lignocellulose is critical to provide sugars for fermentation for biofuel production. This process needs to be understood at the biochemical level in order to optimize for cost and energy efficiency, as well as to identify potential added-value products and integrate their extraction with bioethanol production. In addition to this detailed analysis, we required a High-Throughput assay that would help identify QTL and markers for improved digestibility in germplasm collections from WP1, and critically assess the digestibility of plants with modified cell walls that emerged from WPs 3 and 4. We took a combinatorial approach to identify combinations of pre-treatments and enzyme cocktails to release the valuable components from biomass, in order to produce the range of products developed in WP6 and to provide digestion and extraction steps for WP7 and WP8.

1.5.2 Task 5.1 Characterisation of sugars and oligosaccharides released during deconstruction
Tasks 5.1 and 5.2 led to a comprehensive peer-reviewed publication in Bioenergy Research entitled “Side by side comparison of chemical compounds generated by aqueous pretreatments of maize stover, Miscanthus and sugarcane bagasse” by Gómez LD, Vanholme R, Bird S, Goeminne G, Trindade LM, Polikarpov I, Simister R, Morreel K, Boerjan W & McQueen-Mason SJ (2014). In this publication, we examined the potential for co-product generation and characterised chemical compounds released by a range of alkaline and acidic aqueous pre-treatments, as well as the effect of these pre-treatments on the saccharification of the lignocellulosic material. Comparative experiments were performed using three biomass types chosen for their potential as second-generation biofuel feedstocks: maize stover, Miscanthus and sugarcane bagasse. The release of lignin from the feedstock correlated with the residual biomass saccharification potential, which was consistently higher after alkaline pre-treatment for all three feedstock types. Alkaline pre-treatment released more complex mixtures of pentose and hexose sugars into the pre-treatment liquor than did acid pre-treatment. In addition, complex mixtures of aromatic and aliphatic compounds were released into pre-treatment liquors under alkaline conditions, in a temperature dependent manner, but far less so under acidic conditions. We showed that the three feedstocks each interact with the pre-treatment conditions in a specific manner to generate different ranges of products, highlighting the need to tailor pre-treatments to both the starting feedstock and the desired outcomes.

1.5.3 Task 5.2 Determining the structure of lignin breakdown products
Lignin makes up a large fraction of lignocellulosic biomass, so creating valuable products from its breakdown would be extremely desirable.
The pre-treatment liquors of bagasse, maize and Miscanthus produced by P1 were analysed for their soluble phenolic composition by P12. 135 chromatograms were produced. These were aligned and integrated via Masslinx® software and PCA analyses were carried out. The results showed that acid pre-treatment required higher temperatures to be effective in degrading phenolic compounds. Alkaline pre-treatment was more effective at lower temperatures. Results were similar for Miscanthus and sugarcane bagasse, but different for maize, which is less closely related. P12 then worked on identifying the lignin degradation products (see Task 4.3).

1.5.4 Task 5.3 High-throughput saccharification screening
Several thousand samples from maize, sugarcane and Miscanthus populations were analysed, in order to obtain markers associated with saccharification. We have identified a QTL in the maize RIL F288xF271. For details see Task 1.2.
During the first half of the project we determined the saccharification potential in the BIOMIS population of Miscanthus. P13 produced, grew and harvested the SUNLIBB population at month 38 of the SUNLIBB project. We were particularly interested in screening this population because it is being genotyped by sequencing; offering the opportunity to detect genetic markers for saccharification. We analyzed around 500 samples corresponding to this population and returned the data to P13 in month 43. We expect this data to be used for QTL detection and published after the end of the project. This would be a novel achievement in a polyploid energy crop.
We also used our screening platform to study a sugarcane mapping population. This cooperation with the University of Sao Paulo in Pyracicaba combines a novel genotyping approach developed by Antonio Augusto Garcia’s team, the statistical support from INRA, and the biomass phenotyping platform at York. This collaboration has produced outstanding QTL for saccharification in sugarcane that will be published shortly.
We have also published a paper in collaboration with the Brazilian partners from USP and CTBE in Brazil entitled “Evaluating the composition and processing potential of novel sources of Brazilian biomass for sustainable biorenewables production.” Lima MA, Gomez LD, Steele-King CG, Simister R, Bernardinelli OD, Carvalho MA, Rezende CA, Labate CA, Deazevedo ER, McQueen-Mason SJ, Polikarpov I (2014). Biotechnol Biofuels. 7(1):10. This publication was achieved using the High Throughput Biomass Analytical Platform from Partner 1.

1.5.5 Task 5.4 Examining pretreatment efficiency for enzymatic saccharification
This task was closely associated with task 5.1 and has been included in the publication mentioned above.
A further outcome of this task was obtained in collaboration with our partners, by applying a similar methodology to eucalyptus bark, an important biomass source for Brazil. This collaboration between the University of York, University of Sao Paulo, CTBE and the University of Campinas led to the publication of a landmark paper in the area in Biotechnology for Biofuels entitled “Effects of pretreatment on morphology, chemical composition and enzymatic digestibility of eucalyptus bark: a potentially valuable source of fermentable sugars for biofuel production.” by Lima MA, Lavorente GB, da Silva HKP, Bragatto J, Rezende CA, Bernardinelli OD, de Azevedo ER, Gomez LD, McQueen-Mason SJ, Labate CA, Polikarpov I. (2013). Figure 3.5.5 summarises the optimal conditions to facilitate eucalyptus processing.

1.5.6 Task 5.5 Trialling new pretreatments
Microwave pre-treatments were evaluated as a novel alternative for processing. Microwave conditions of 180ºC for 2, 5 and 10 minutes were applied to Miscanthus and sugarcane. Analysis of the soluble sugars indicated good levels of hydrolysis of hemicelluloses under these conditions, but little hydrolysis of cellulose except under acidic conditions. In addition, increasingly acidic conditions (and prolonged reaction times) indicated lower levels of free glucose in solution, surprising from the expected kinetic dependence of hydrolysis on acid concentration. Analysis of the residual biomass by scanning electron microscope showed the expected gradual breakdown of the fibrous nature of the material, but for the more strongly acidic/ longer reaction time (acid) samples, there was evidence of particles of ca. micron size as well. Solid state nuclear magnetic resonance (NMR) studies on relevant samples indicate that these particles are ‘hydrothermal carbon,’ formed by the acid catalyzed conversion of glucopyranose to glucofuranose to hydroxymethylfurfural (HMF) to poly-HMF to hydrothermal carbon.
We have submitted two papers and a third one is in preparation.
-“Efficient removal of sugars from sugarcane bagasse by microwave assisted acid and alkali pretreatment” Zongyuan Zhu, Camila Alves Rezende, Rachael Simister, Simon J. McQueen-Mason, Duncan J. Macquarrie, Leonardo D. Gomez, Igor Polikarpov (submitted to Biomass and Biofuels)
-“Microwave assisted acid and alkaline pre-treatment for using Miscanthus biomass in biorefineries” Zongyuan Zhu, Duncan J. Macquarrie, Rachael Simister, Susannah Bird, Simon J. McQueen-Mason, Leonardo D. Gomez (submitted to Bioenergy Research)
-“Microwave assisted chemical pre-treatment of Miscanthus under different temperatures” Zongyuan Zhu, Duncan J. Macquarrie, Leonardo D. Gomez, Rachael Simister, Simon J. McQueen-Mason.

1.5.7 Task 5.6: Trialling new enzymes
This task involved testing new enzymes that originated during the course of the project from different partners in SUNLIBB, but above all from Project 3 in different Brazilian laboratories involved in CeProBIO. The development of new enzyme cocktails or the improvement of existing ones is a major focus of research in the biorefinery area. P1 has led an enzyme discovery program in the UK and several Brazilian collaborators lead similar programs in Brazil. The work was done in four core units.
a) Limnoria quatripunctata GH7
P1 has successfully led an enzyme discovery program funded by the BBSRC, where a marine wood borer organism, Limnoria quatripunctata, is used as a source of lignocellulose degrading mechanisms. A glycosyl hydrolase family 7 (GH7) with unique characteristics has recently been described in this organism (Kern et al., (2013) “Structural characterisation of a unique marine animal family 7 cellobiohydrolase suggests a mechanism of cellulase salt tolerance.” PNAS 110:10189-94). This novel enzyme has high halotolerance and represents a promising cellulolytic element to improve biomass hydrolysis. Within the present project, we tested the addition of LqGH7 into a standard commercial cocktail from Novozymes. The automated platform in York for the determination of enzyme activities allows a fast screen of the effects of the addition of enzymes into a cocktail at increasing concentrations and different incubation times (Whitehead et al, (2012) “Automated determination of cellulotic activities.” Methods in Enzymology 510:37-50). Figure 3.5.7 shows that when biomass (Miscanthus giganteus) is used as a substrate, the synergic effect remains, increasing by approximately 30% the amount of reducing sugars obtained from the biomass. These experiments show that LqGH7 can have a significant impact on the improvement of biomass hydrolysis.
b) Laccase/hemocyanin pretreatment system
P1’s work in revealing the mechanism present in Limnoria for the digestion of lignocellulosic biomass has shown that free radical chemistry takes part in the deconstruction of wood to extract sugars. In this free radical loosening of the cell wall, hemocyanins seem to mediate Fenton reactions. We therefore decided to test a similar mechanism in vitro to assess the effects of these free radical pretreatments on the saccharification of plant biomass. We tested the effect of hemocyanin pretreatment with and without the addition of methyl syringate (MeSy) as a phenolic mediator, on the subsequent saccharification. As a positive control of this mechanism we used laccase and MeSy as indicated in the literature. We did not see an enhancer effect of these free radical pretreatments on any of the biomass or substrates trialled.
c) Accessory proteins from Trichoderma hartzianum secretome
As part of the exchange between SUNLIBB and CeProBIO, cooperation was established between P1 and Prof. Nei Pereira Jr. from the Federal University of Rio de Janeiro. A Brazilian PhD student spent one year in York, exploring the function of accessory proteins present in the secretome of the fungus Trichoderma hartzianum. Previous work carried out in Brazil established the optimal conditions for the production of lignocellulolytic enzymes in Trichoderma, and the objective of the present work was to obtain the proteomic profile of the secreted enzymes, select components from the secretome, express the proteins and evaluate their effect on the saccharification of plant biomass. From these proteins we chose two for further work: a swollenin abundantly produced by Trichoderma and vanillin oxidase. These proteins were expressed in Aspergillus niger and the purified proteins were used in the automated platform to evaluate their effect on saccharification. None of the proteins had any effect on saccharification when used as a pre-treatment. When Trichoderma hartzianum swollenin was added into the enzyme cocktail, there was a small, but significant increase in the saccharification of both filter paper and biomass with increasing amounts of swollenin added.
d) Xylanase chimaeras
Collaboration with Professor Richard Ward, from Sao Paulo University, brought to SUNLIBB a set of chimeric enzymes designed to combine different functions to improve biomass hydrolysis. The original approach taken in Brazil to design enzymes involves the use of product-stimulated enzymes. In this particular case, a microbial XylA was subjected to mutagenesis and the clones were evaluated to determine the degree of xylose stimulation. These product-stimulated clones were linked to a xylose binding protein (XBP) to increase the enzyme-substrate affinity. These clones were sent to York and the automated platform was programmed to measure the effect of these chimaeras on the hydrolysis of plant biomass (Miscanthus), with or without the addition of commercial enzymatic cocktails, without pre-treatment, after water pre-treatment, and after 0.4M NaOH pre-treatment.
Although we could see clear differences between pre-treatments, the addition or not of the chimeric enzymes had no effect on the amount of reducing sugars released from the biomass. The chimeric enzymes had been tested against purified substrate (xylan) but not against biomass. The reason for the lack of effect under our test conditions remains unclear.

1.5.8 Task 5.7 Enzymatic lignin breakdown
SUNLIBB partner 11 (University of Sheffield) used GC-MS for the identification and quantification of the intermediate compounds from the breakdown of lignin by low temperature catalytic and biocatalytic routes. Kraft lignin was used in these experiments, because free lignin from maize or Miscanthus was not available and would not be, for the foreseeable future. The aim was to develop a generic methodology, in tandem with the other SUNLIBB work, which could be valuable in a prospective biorefinery process. These results cannot, therefore, be directly related to maize, Miscanthus or sugarcane, but the experiments were designed to stand alone and be complementary to the work of other SUNLIBB partners.
The main compounds identified were organic acids: acetic acid, malonic acid, succinic acid, butylated hydroxytoluene, vanillin, veratric acid, and palmitic acid. The photodegradation of phenolic compounds by TiO2 begins with 'OH radical attack on the phenyl rings, producing catechol, resorcinol, and hydroquinone. Subsequently, the phenyl rings in these compounds break up first to give malonic acid and then short-chain organic acids, such as maleic, oxalic, acetic, and formic acids, before finally yielding CO2. In the reactions observed here, TiO2/UV yielded some organic acids, such as succinic, palmitic, and acetic acids. This implies that the phenolic compounds in the lignin molecule were decomposed in this reaction. These organic acids were also found by others who used a hydrothermal oxidation process to degrade alkali lignin. However, in their case, the hydrothermal reaction was carried out at a higher temperature and the alkali lignin that they used had a lower molecular weight than the lignin tested in this work in SUNLIBB (thus potential energy saving here).
In contrast, the compounds that were identified from the degradation of lignin by laccase were substantially different from those observed from lignin and photocatalytic lignin degradation. For example, several polysaccharides were detected in the extracted samples sourced from the biocatalytic reaction, as well as from the single- and dual-step processes. The peaks obtained were D,L-arabinose, methyl α-D-glucopyranoside, hexose (D-galac¬tose), hexopyranose, D-glucose, and glucopyranose. These monosaccharides could be carbohydrate derivatives from the laccase molecule. The carbohydrate moiety of the majority of laccases consists of mannose, N-acetylglucoamine, and galactose at about 10-20% for fungal laccases. In comparison of the concentrations of intermediate compounds in lignin with laccase (standard), it was noted that the amount of palmitic acids significantly decreased for the biocatalytic and dual-step process but were the same concentrations as found in the single-step process. Furthermore, the amount of stearic acid that considerably decreased in the single- and dual-step processes was unchanged in the biocatalytic process. These compounds were also found in Kraft lignin degradation mediated by Bacillus sp. (accession number AY 952465), as reported elsewhere. The occurrence of these compounds indicates that lignin was degraded. These studies fulfilled the requirements of Deliverables 5.3 and 5.7 and led to the publication of a paper, where the detailed results are presented (Kamwilaisak & Wright (2012) “Investigating laccase and titanium dioxide for lignin degradation.” Energy Fuels 26:2400-2406).

1.6 Work Package 6
Generating added-value from biomass
1.6.1 Objectives
Whilst biofuels are likely to be the major product from biomass in terms of volume and value, cost-effective and sustainable bioprocessing will require the development of other value streams, particularly from components not converted into ethanol. We identified a range of added-value products for which potential markets already exist, and worked to identify the extraction and processing steps necessary to generate waxes, lignin derivatives and poly- and oligosaccharides from C4 grasses. Some aspects of this WP were informed by data on cell wall degradation products from WP5, including those from modified plants emerging from other WPs. The work in this WP directly fed into WP7, which integrated the production of added value products with ethanol production.

1.6.2 Task 6.1 Characterising and extracting high value molecules with CO2
Supercritical carbon dioxide was used successfully to extract potentially valuable lipids from maize, Miscanthus and sugarcane biomass. The extraction method was optimised and the extracted molecules were characterised in detail. This work was reported as fulfilment of Deliverables 6.1 and 6.2 (and Joint Deliverables 6.1 & 6.2). The optimisation of the extraction of epicuticular waxes from C4 biomasses was attempted using the factorial experimental design. The extraction yield is the response variable and the two factors which were chosen as the independent variables were temperature and pressure. The effect of temperature and pressure on the extraction yield was modelled using the dimensionless factor coordinate system. The initial set of experiments was carried out over four hour extraction periods. A variety of temperature and pressure ranges were incorporated in this study. The optimal conditions for Miscanthus giganteus were found to be 350 bar and 50 oC. For maize, the optimal conditions were found to be 400 bar and 65 oC while the optimal conditions for sugarcane bagasse were found to be 325 bar and 57 oC. Results indicate that for Miscanthus and sugarcane bagasse, density plays a more important role than temperature in the extraction of waxes. Furthermore, in both cases, the results suggest that a specific density is requited for optimal extraction. In the case of maize, results indicate that temperature is more important than density in the extraction of waxes. The main classes of lipophilic components that were identified in the wax samples from C4 biomass included long-chain n-alkanes, fatty acids, fatty alcohols, wax esters, sterols, triterpenoids, beta-diketones and other compounds.
The supercritical system extracts free lipids only and does not liberate those that are bound to the biomass. An advantage of using supercritical CO2 is the possibility of fractionating crude epicuticular wax into groups of different compounds, resulting in wax fractions containing different properties. The different wax fractions could therefore be used in different applications.
Extracted waxes were tested by Ecover (SUNLIBB P5) as an antifoam agent for laundry products. This is potentially a high value application, especially for ‘Eco’ washing powders, because existing antifoam agents have sustainability problems and are possible pollutants.
In total, 38 biomass samples (including standards) were obtained from the University of York (P1). The samples originated from sugarcane bagasse, wheat straw, Miscanthus and maize. The wax was extracted and fractionated using three fractional separators set at different pressures and temperatures. This resulted in wax fractions having different structures, textures and melting point ranges. Only fractions which melted below 60°C were retained. The highest yields of wax were obtained from Miscanthus and maize leaves. The wax fraction from maize obtained at 80 bar – ATM/ 50°C had a melting point range from 28 - 41°C, which meant that it was highly suitable for use in detergent products.
The defoaming abilities of the waxes were tested using washing machine tests, whereby the wax was added to the detergent formulation and washing machine runs were carried out with conditions mimicking real wash loads. As well as the washing machine tests, the foaming characteristics were analysed using a lab foam measuring set-up and a lab foam measuring device (Krüss foam analyser).
Both the lab and the washing machine experiments revealed that maize wax was the most effective compound as antifoam for laundry powder. The maize wax significantly reduced the quantity of foam generated, as well as decreasing the time taken for the foam to break up. Maize wax clearly has the potential to be used as a defoamer in washing machine detergent formulations.

The economics of wax extraction was also investigated. An economic assessment of the cost of supercritical fluid extraction (SFE) of waxes on an industrial scale was carried out, using methodology proposed by Turton et al (2009). This method has been employed in studies associated with SFE economics and has been found to be an effective and appropriate method for evaluating the costs of SFE processes.
When calculating the manufacturing costs of a chemical product, three main types of expenses are involved: (i) Direct costs; (ii) Fixed costs; (iii) General expenses. Therefore, the cost of manufacturing (COM) is the sum of direct costs, fixed costs and general expenses. These three components of the COM are estimated in terms of five main costs:
COM = 0.280FCI + 2.73COL + 1.23(CRM + CWT + CUT)
Where FCI is the fraction of capital investment (cost of the equipment), COL is the cost of labour, CRM is the cost of raw materials (i.e. the cost of growing, harvesting, storing and pelletising biomass), CWT is the cost of waste treatment and CUT is the utility costs (i.e. the costs of running the supercritical kit). Each one was determined in order to calculate the COM of wax extraction.
It was found that the cost of extraction was €568/tonne of Miscanthus biomass or €135/kg of Miscanthus wax. However, several assumptions were made. Firstly, the value for the amount of biomass that can be loaded into the supercritical extractor was based on milled biomass. In industry, biomass is normally pelletised and this increases the biomass loading three-fold. If pelletised Miscanthus were taken into account, the costs would be around €93/kg of wax. Secondly, it was assumed that the labour costs (estimated at €7.50/hour) are solely for the extraction of wax. This forms only part of the activity of a biorefinery. Therefore, in reality, this value would be much lower (reduced by one third), as the workers would have other duties within the biorefinery. Furthermore, it was assumed that the cost of raw materials (CRM) is solely for the supercritical extraction. Once again, the biomass would be passed on within the biorefinery and therefore the cost of raw materials is not solely for the extraction of wax, but for all the processes within the biorefinery. Therefore, the CRM value would be lower, leading to lower overall manufacturing costs. Finally, Miscanthus straw was used in the process. This has a yield of around 0.45%. The yield from Miscanthus leaves was almost 5 times greater (1.96%). This increased amount of wax would lower the COM 4-5 fold, to around €24/kg of wax. Furthermore, combustion of the Miscanthus biomass following the supercritical extraction would reduce the cost of the wax to €10.50/kg of wax.

1.6.3 Task 6.2 Added-value products from lignin
Borregaard (SUNLIBB partner 3) has existing technology for converting wood lignin into vanillin and other potentially valuable products, and this technology was adapted to work with lignin from C4 grasses.
Pyrolysis samples were sent to Borregaard from the University of York. Fractionated samples from microwave pyrolysis of Miscanthus were analysed by GC-MS. The analysis showed that the samples fractioned by extraction with organic solvents gave complex mixtures where no component was found to be of unique enough character and high enough concentration to justify any further separation. A new set of fractionated pyrolysis samples from Miscanthus was analysed using the same method. These samples were fractionated by supercritical CO2 extraction. The findings from these samples were similar to the previous samples; the pyrolysis and in-situ fractionation techniques used for these samples were not able to fine-tune the decomposition of the biomass into a product mixture which is commercially interesting.
It was concluded that both the pyrolysis and fractionation techniques needed to be altered to provide fractions which were more homogenous in composition. Heterogeneous samples, such as these, would not give any more information, other than that the sample was a complex mixture and thus not of commercial interest.

1.6.4 Task 6.3 Lignin fermentation
Fermentation experiments were carried out at the University of Sheffield (SUNLIBB partner 11) using commercially obtained plant materials and Miscanthus hydrolysate provided by the University of York. The recalcitrant nature of lignocellulosic biomass is a major roadblock to its exploitation as a potential energy source. The degradation step is the basic hurdle and needs to be optimised. The innate capacity of native microbes cannot be underestimated and throughout this SUNLIBB project and from the literature, it is clear that a single microbial system may not be best solution. For wider commercial use under difficult conditions in the field there are difficulties in recombinant strain development. As shown in SUNLIBB and elsewhere, both chemical and biological treatments for lignocellulose degradation are promising. It is becoming clearer that in addition to traditional strains such as Saccharomyces cerevisiae etc., anaerobes are raising increased interest among researchers. In addition to the yeast work conducted in this task and against this background, it was decided to examine the utility of anaerobes at degrading C4 biomass.
It was possible to grow Clostridium acetobutylicum successfully on a range of relevant feedstocks. It was also possible to produce H2, butanol and ethanol. Furthermore, we worked with co-cultures of Fibrobacter succinogenes S85 and C. acetobutylicum ATCC824. This required media optimisation as the growth media of the two strains are quite different and a compromise was sought. 40FS/60CA media was the best compromise. C. acetobutylicum can be a good candidate for utilization of sugar components from biomass hydrolysate to produce a range of biofuels. Microbial consortia for biofuel production using F. succinogenes and C. acetobutylicum could provide a major breakthrough in Consolidated Bioprocessing (CBP) development.

1.6.5 Task 6.4 Value from pentose sugars
Borregaard (P4) has developed valuable products derived from woody biomass. In SUNLIBB, they examined pentose sugars from the breakdown of C4 grasses to determine whether these were suitable for the formation of these products.
Pre-treated liquors from Miscanthus biomass were received from WP5. The sugars in the sample were easily fermented by Kluyveromyces marxianus. However, the fermentable sugar concentration was too low to be commercially interesting.
Ecover (P5) investigated whether sophorolipids could be produced from the sugars in hydrolysate samples produced by other SUNLIBB work packages. Sophorolipids are surfactants that can be synthesized by certain yeast strains and are use as ingredients in cleaning products.
16 samples of hydrolysate were obtained from Processum (P8). The sugar content was estimated using refractive index (Brix) and the pH was measured. A screening growth experiment was set up, to see whether the yeast Starmerella bombicola was able to grow on this C-source. As the sugar content in the samples was below 3wt%, the duration of the experiment was restricted to 48h. The medium consisted of 100ml of the hydrolysate sample, plus a nitrogen source (urea and yeast extract). Growth and sugar content were estimated during the run using, respectively, optical density and refractive index (Brix).
Starmerella bombicola was able to grow on all the hydrolysate samples. No growth inhibition was noticed. However, the hydrolysate samples had suffered contaminating microbial growth during transport. The sugar content of the samples was very low, so the samples were not considered to be suitable for industrial production of sophorolipids. It is also important to note that the strain of S. bombicola used was not able to ferment pentose sugars.
It had to be concluded that these experiments had not worked, and no further knowledge had been gained about obtaining valuable products from pentose sugars.

(Subsequently, in response to a reviewer’s request, it was ascertained that feeding S. bombicola with C5 sugars would result in production of the same sophorolipids as when feeding C6 sugars, due to the specific biochemical pathway involved. Changing the lipid feeding source would change the backbone chain length of the sophorolipid.)

1.6.6 References
Turton et al (2009) Analysis, synthesis and design of chemical processes. Prentice Hall, Boston.

1.7 Work Package 7
Integrating process engineering to obtain full value from biomass processing
1.7.1 Objectives
The major technical barriers to producing sustainable and cost-competitive bioethanol from lignocellulose lie in the high costs of saccharification. The objective of this WP was to trial a range of new processes that involve added-value product extraction as partial pre-treatments for enzymic saccharification in order to add value to the overall process and decrease saccharification costs. Data from WP5 and WP6 were used to develop integrated process systems to obtain maximal value from lignocellulose. This involved iterated combinations of data generated in piloting activities and in silico modeling to predict optimal ways forward. Pilot and demonstration facilities were available through partners in Europe and Brazil. We aimed to generate experimental data for integrating added-value products extraction with cellulosic ethanol production as well as pyrolysis of any residual material, and assisted LCA work in WP8.

1.7.2 Task 7.1 Integrating the extraction of added-value products with ethanol production
P11 have set up platform-type “base case” process simulations, focusing on design of simulation of ideal conversion efficiencies based on literature and real plant data, in order to utilize lignocellulosic feedstock and convert it to biofuel in the form of ethanol. This has integrated some tools and data from different parts of the SUNLIBB consortium.

1.7.3 Task 7.2 Fermentation of sugars from C4 grass biomass
P14 has developed thermophilic bacterial strains (Geobacilli) capable of fermenting C5 and C6 sugars from plant biomass. These strains were used to ferment samples of hydrolysate produced by P8. Sugars from maize and Miscanthus hydrolysate were utilized and more than 3.3 g/l ethanol was produced under semi-aerobic or aerobic conditions. There was a significant increase in sugar utilization and ethanol production in the presence of added nutrients, suggesting that optimization of the growth medium could further improve the results and enable all the sugars to be used.
P11 also received the same samples from P8 as P14. There was evidence of useful biofuel product generation (e.g. ethanol and butanol) from the samples using Clostridium sp. This work is beyond the deliverables of SUNLIBB and an update report was provided in month 50. A journal publication on this work is also planned.

1.7.4 Task 7.3 Lab and pilot scale sophorolipid fermentation trials using improved feedstock
P5 developed a production process to bioconvert sugars and rapeseed oil into sophorolipids (biosurfactants) using the yeast Starmerella bombicola.
Experiments showed that the yeast was able to grow on all samples of maize and Miscanthus hydrolysate provided by P8. No growth inhibition was noticed. However, due to the low sugar content, the samples were not considered to be suitable for industrial production of sophorolipids. The hydrolysate samples had also been subject to microbial growth during transport. These experiments are reported in D6.6 and D7.7. (Note that there is considerable overlap between the aims and conclusions of these deliverables).

1.7.5 Task 7.4 Pilot studies
Pre-treatments were carried out by P1 and samples were sent to P8 for hydrolysis and fermentation and further onward distribution as detailed in other tasks in WP7.

1.7.6 Task 7.5 Demonstration of added-value products from lignin breakdown
Sets of fractionated pyrolysis samples were sent to Borregaard (P3) from P1 and analysed by GC-MS (WP6). However, none of the components detected were of sufficient uniqueness or high enough concentration to justify further separation or repetition of the experiments at demonstration scale.

1.7.7 Task 7.6 Integrating improved feedstock
Techno-economic analysis by P11, using data from the pilot-scale experiments in task 7.4 showed that use of improved maize and Miscanthus feedstocks increased ethanol yield and resulted in lower production costs compared to generic feedstocks. Further optimization would be needed, but these studies illustrated the main areas for potential improvement. Data and standardization were made with the collaboration of P7.

1.8 Work Package 8
Sustainability Assessment
1.8.1 Objectives
The purpose of WP8 was to determine and compare the impacts of biofuel production in integrated biorefineries with other sources of biofuels and with conventional fuels (1) with reference to existing and emerging frameworks for calculating GHG emissions and the treatment of sustainability criteria, (2) by incorporating a model of biorefinery operation, based on the outcomes from other WPs, (especially WP7), to provide necessary data for the quantitative and qualitative assessment of sustainability, and (3) by adopting transparent MS Excel workbooks to evaluate total GHG emissions and a transparent approach to the evaluation of other sustainability criteria. The primary focus of WP8 was on the quantification of GHG emissions because of current emphasis on this specific environmental consideration in official policy on biofuels; especially the EC's Renewable Energy Directive and Fuel Quality Directive, which both specify the necessary GHG emissions calculation methodology and target savings. The most relevant other impacts were also addressed, in a qualitative manner appropriate to diversity and site-specific nature of biorefineries and their biomass feedstock supplies, including, for instance, direct and indirect land use change, water use, eutrophication and biodiversity. Developments in relevant methodologies, such as the emergence of the European Life Cycle Data System (ELCD), and progress in other related projects, such as the BIOCORE, SUPRABIO and EUROBIOREF projects, were taken into account.
Three partners were involved in WP8; North Energy Associates Ltd (P7 and WP8 Leader), Processum Biorefinery Initiative AB (P8) and the University of Leeds (P10). The work of P10 involved the “Review of Policy and Regulatory Context at European Union (EU) and Individual Country Level” and completion of Deliverable 8.1 consisting of the “Sustainability Policy and Regulatory Review Report.” The main work of P7 concentrated on the quantification of GHG emissions associated with biorefinery operation, using sugarcane, maize and Miscanthus feedstocks, along with the qualitative assessment of other sustainability criteria. This work was covered by Task 8.2 on “Model Development”, Task 8.3 on “Application to Maize Biorefineries”, Task 8.4 on “Application to Miscanthus Biorefineries” and Task 8.5 on “Sensitivity and Comparative Analysis. The input of P8 has been principally related to assisting the specification of the biorefinery which was an integral part of workbook development and finalisation as part of Tasks 8.3 and 8.4.
The finalisation of results from the workbooks generated in WP8 were completely dictated by the planned and eventually realised deadlines for key outcomes from other WPs on the SUNLIBB Project. Hence, it was not possible to provide feedback to these WPs during the course of the Project. As recommended in the Final Review meeting, such integration should be a key part of any further activities.
The workbooks generated in WP8 have the capability to investigate the impact of varying a very wide range of parameters, including Miscanthus yield. Indeed, it was confirmed, during the Interim Review Meeting, that users of the workbooks can, themselves, easily change the default value for Miscanthus yield and can explore the effects of supplying Miscanthus from different EU member states provided that they have access to comprehensive cultivation data. However, total greenhouse gas emissions associated with ethanol produced in a biorefinery are relatively insensitive to Miscanthus yield, as demonstrated in the attached file. This is because of the comparatively small contribution of Miscanthus cultivation, harvesting and baling (due to its low agricultural inputs) to total greenhouse gas emissions of ethanol production (typically, 20%). However, this point, along with any relevant information on Miscanthus yields from EU Member States, can be introduced into a revised version of Deliverable D8.5.

1.8.2 Task 8.1 Review of Policy and Regulatory Context at EU and Individual Country Level
As noted, this task primarily consisted of preparing the “Sustainability Policy and Regulatory Review Report.” This was based on a review of the relevant policy and regulatory frameworks at both the EU and individual Member State level, assessing in particular the effectiveness of sustainability criteria adopted at both levels of EU governance. P10 collaborated successfully with WP9 participants in producing SUNLIBB’s second and third Policy Briefs, which are available on the project webpage (www.sunlibb.eu). Their purpose was to introduce the nature of the work of the SUNLIBB Project, discuss its objectives and outline mid-term and final progress. The policy briefs were released in September 2013 and September 2014, respectively, and were disseminated to a plethora of contacts around the world. In addition to being e-mailed to CeProBIO Project partners in Brazil, the policy briefs were sent by P10 to the Climate-L and Energy-L mailing lists. These news and announcement list services are run by the International Institute for Sustainable Development and are considered a key resource for anyone interested in climate and energy policy.
P10 also participated in joint SUNLIBB/CeProBIO Projects workshops and has updated EU and Brazilian partners on all the major and minor policy developments with regard to biofuels that could be relevant to their work. In workshops in 2013 and 2014, P10’s attention focused on the EC’s draft legislation for addressing indirect land-use change (iLUC), defining highly biodiverse grasslands and limiting the percentage of biofuels from food crops. In 2014, P10 also presented work on the current status of EU-Brazil relations with respect to 2G biofuels and the current status of EU-Brazil trilateral efforts to promote biofuels in African countries.

1.8.3 Task 8.2 Model Development
Deliverable D8.2 the Biorefinery Model workbook, the final version of which was SUNLIBB Sugar Cane Ethanol NREL Biorefinery v04.xls was developed by P7 to estimate the primary energy emissions and prominent GHG emissions associated with the production of ethanol from sugar cane as a biomass feedstock with bagasse as a fuel, or from sugar cane and bagasse as biomass feedstocks. Initial sugar cane cultivation, harvesting, transportation and conventional processing data were provided via the CeProBIO project. In order to represent the lignocellulosic processing of bagasse, suitably-modified information was incorporated from "Process Design and Economics for Biochemical Conversion of Lignocellulosic Biomass to Ethanol: Dilute-acid pretreatment and enzymatic hydrolysis of corn stover" by D. Humbird, R. Davis, L. Tao, C. Kirchin, D. Hsu, A. Aden, P. Schoen, J. Lukas, B. Olthof, M. Worley, D. Sexton and D. Dudgeonm, Technical Report NREL/TP-5100-47764, National Technical Energy Laboratory, Golden, Colorado, United States of America, May 2011, (referred to as the NREL Report). This involved using initial experimental data from WP7. The final version of the workbook and those subsequently developed in Tasks 8.3 and 8.5 have a very high degree of functionality which enables methodological and technical options and variations in key parameters to be explored.

1.8.4 Task 8.3 Application to Maize Biorefineries
Using Deliverable D8.2 the Biorefinery Model workbook, as a template, the workbook to evaluate primary energy inputs and prominent GHG emissions associated with the production of ethanol from maize in the form of corn stover or whole maize was developed by P7. This involved adding pathways for corn stover collection, whole maize cultivation and harvesting, and biomass feedstock transport as well as changes to biorefinery performance. Additionally, the effects of supercritical CO2 extraction and the likely influence of improved maize feedstocks were accommodated using results from WP6 and WP7, respectively, with assistance from P8. The final version of this workbook was SUNLIBB Maize Ethanol Pilot Study Biorefinery v01.xls.

As well as the quantitative assessment of environmental impacts by this workbook, a “Supporting Sustainability Criteria Report: Maize” was also prepared, as part of D8.3 to examine other environmental impacts of corn stover and whole maize biorefineries in a qualitative manner. This addressed land use; soil erosion, fertility and carbon; water use; emissions to water and air; biodiversity; and local traffic levels.

1.8.5 Task 8.4 Application to Miscanthus Biorefineries
The Biorefinery Model workbook from D8.2 was also used by P7 as a template to develop the workbook for evaluating primary energy inputs and prominent GHG emissions associated with the production of ethanol from Miscanthus. This involved adding the pathway for Miscanthus cultivation, harvesting, and transportation as well as changes to biorefinery performance. Additionally, the effects of supercritical CO2 extraction and the likely influence of improved Miscanthus feedstocks were accommodated using results from WP6 and WP7, respectively, with assistance from P8. The final form of this workbook was SUNLIBB Miscanthus Ethanol Pilot Study Biorefinery v02.xls.
As previously, a “Supporting Sustainability Criteria Report: Miscanthus” was prepared, as part of D8.4 to examine other environmental impacts of Miscanthus biorefineries in a qualitative manner. This addressed land use; soil erosion, fertility and carbon; water use; emissions to water and air; biodiversity; invasiveness; and local traffic levels.

1.8.6 Task 8.5 Sensitivity and Comparative Analysis
Final results from sensitivity and comparative analysis generated using these workbooks were presented, along with summaries of assumptions for determining the biorefinery performance, the expected influence of biomass feedstock improvements and the effects of supercritical CO2 extraction. Additionally, consideration of LCA methodologies in the BIOCORE, SUPRABIO and EUROBIOREF projects, and summaries of qualitative assessment of other environmental impacts, and the possible socio-economic consequences of biorefineries were documented in D8.5 consisting of the “Sensitivity and Comparative Analysis Report.”


Potential Impact:
4.1.4 Description of potential impact
4.1.4.1 Socio-economic impact and wider societal implications
As described previously, the SUNLIBB project was designed to address a current socio-economic problem and ramifications from its results could have important societal implications. Uncertainty over future energy supplies is one of the major issues facing European society today, and the development of sustainable, renewable and secure sources of energy is crucial to the maintenance and improvement of living standards in the EU and beyond. In addition, efforts are being made internationally to mitigate the harmful effects of global climate change caused by the release of greenhouse gases from fossil fuels. This is a strategic priority for the EU, which has set itself ambitious targets for progressively reducing greenhouse gas emissions up to 2050.
Liberating the energy stored in lignocellulosic plant biomass offers a potentially sustainable, carbon-neutral way to meet future energy needs. The SUNLIBB programme aimed to advance knowledge in this field, to identify areas for possible improvement, suggest solutions to technical bottlenecks and pave the way for plant-fuelled biorefineries to become a realistic and commercially viable proposition. One of the key findings of SUNLIBB was that improving the quality of plant biomass through better genetics could lead to more efficient bioethanol production. An implication of this is that further improvements could be made, assisted by the greater understanding of factors affecting lignocellulose breakdown that was obtained through the work of SUNLIBB.
SUNLIBB addressed very carefully the issues of sustainability surrounding the production of second generation bioethanol. Overall sustainability takes into account many environmental and social criteria and SUNLIBB examined these, in the context of regulatory frameworks. Going forward, the results will be extremely valuable in assessing the true potential sustainability of biorefinery operation.
Close collaboration with partners in the CeProBIO consortium in Brazil was a key feature of the SUNLIBB project. This cooperation was highly beneficial, on many levels, to both consortia, and resulted in mutual sharing of expertise, knowledge and data, the publication of several joint papers and many other examples of Added Value. The impact of this synergistic interaction is on-going, with funding in place for several future EU-Brazil collaborations. Developing close links with Brazil has helped us to forge relationships with neighbouring countries, such as Argentina and Chile, in the area of bioenergy; resulting in the establishment in 2013 of a South American/European working group on biomass for biorefineries. Hence, the impact of EC funding into SUNLIBB is clearly having added value in deepening and widening cooperation in the globally important area of sustainable biofuels and biorefineries. In a wider context, sharing of resources and knowledge in this way will help to keep the EU at the forefront of global second generation bioethanol production.

4.1.4.2 Specific impacts of SUNLIBB
The main aims of SUNLIBB were to improve the feedstocks and processing of lignocellulosic biomass from maize, Miscanthus and sugarcane, to underpin the development of sustainable and economically viable biorefineries based on these feedstocks. The results of the SUNLIBB project have been summarised in section 3, where it can be seen that significant progress was made in advancing knowledge in all the above areas.
The genetic resources developed in WP1 and WP2 will have significant potential impacts on the improvement of biomass crops through breeding. The increased understanding of cell wall polysaccharide biosynthesis and structure (WP3) and lignin biosynthesis (WP4) will further impact on future biomass improvement. The progress made in understanding biomass deconstruction in WP5 and the demonstration that valuable compounds can be extracted during the process (WP6) will help to optimise the economic viability of the biorefinery. The scale-up experiments in WP7 have provided useful information about the areas of the biorefining process where improvements can be targeted, and have gratifyingly shown that improving the quality of the feedstock through genetics can improve the efficiency of bioethanol production. WP8 sets the project into context in terms of sustainability, and resilience to potential policy developments.
In some work packages, achievements extended beyond the original objectives of SUNLIBB. For example, WP2 has developed their transcriptomics data into an orthology web-interface database, designed for use by the wider scientific community. This will be of tremendous value in the study of gene expression in any orphan crop. Extra collaborations developed between WPs 2 & 7 resulted in the generation of useful additional proteomics data from Miscanthus. The CSPP technique developed by WP4 was so successful that further funding for similar work has already been obtained. WP5 was able to screen an additional Miscanthus population for saccharification potential, and the QTL identification work in sugarcane proceeded much further than anticipated.

4.1.4.3 Exploitation of Results and Dissemination Activities
Most of the activities of SUNLIBB were of an academic nature and most project deliverables were concerned with ‘Generation of Knowledge.’ However, it is probable (and intended) that the project’s results (foreground) will be exploited in the future.
Despite the fact that no patent applications were made, some of the foreground could be considered to have potential as Intellectual Property; (for example, the introgression lines produced by Biogemma and the information obtained about waxes as defoamers and about the sophorolipid production process at Ecover). In addition, the project’s results will serve as a platform for further experiments, involving future funding applications and further collaborations.
4.1.4.3.1 Exploitation plan
The following exploitation plan details the ways in which the SUNLIBB partners (and the wider community) could benefit from the foreground.
The relevant terms and conditions relating to foreground are set out in the General Conditions (Annex II, Articles II.26 – II.29) that form part of the Grant Agreement (no. 251132). These conditions relate to ownership, transfer, protection and use of foreground. In addition, the terms and conditions listed in Section 8 (‘Foreground’) of the Consortium Agreement apply.
Access rights, background rights and intellectual property rights are also defined in the Grant Agreement and Consortium Agreement.
As stated, foreground is the property of each beneficiary (partner). As such, it is each partner’s responsibility to use, protect and disseminate the foreground.
The consortium understands that exploitation of the outcomes of the project has 2 major dimensions; exploitation and dissemination & communication.
Regarding exploitation, this plan identifies the expected outcomes from the project and discusses initial views on how these could be further utilized and extended. Dissemination and communication are fundamental to any research process, and the plan describes details of dissemination activities undertaken during the lifetime of the project.
4.1.4.3.1.1 Exploitation of foreground
Possible exploitation of foreground generated by each of the individual partners is discussed below.
Partner 1 (University of York – CNAP)
The validation of the High-Throughput saccharification assay for maize, Miscanthus and sugarcane shows that the established assay can be adapted to different species, which could be exploited in future experiments. Trials of different biomass pre-treatments could ultimately lead to patents for new industrial processes.
Partner 1 (University of York – Green Chemistry)
The supercritical CO2 wax extraction process could in the future be patented as a new industrial process for specific applications and the waxes could also be marketable in their own right or fractionated for use in a wide variety of higher value applications (including use as defoaming agents and cholesterol-reducing nutraceuticals). The information gained on wax profiles, extraction economics and any enhanced ability to process biomass could be very important when integrating these processes into future biorefineries.
Partner 3 (Borregaard)
The phenolic compounds produced from biomass were potentially marketable. However, in these experiments, they were not produced in sufficient quantity or definition to be of sufficient value to pursue.
Partner 4 (Biogemma)
The generation of Knock-Out lines and the identification of target gene candidates could have commercial potential.
Partner 5 (Ecover)
The sophorolipids generated from biomass were potentially valuable, although the quantities produced in these experiments were too low to be exploitable. Increasing the fermentable sugar content is the bottleneck to overcome, to make this a commercially viable process. Also, other biotechnology processes would benefit from this, having access to an additional carbon source besides the sources currently available on the market. During this project, new knowledge and insights were obtained into waxes as defoamers. This will enable the parties to move forward at a more rapid pace.
Partner 6 (INRA)
The genetic data generated is a valuable resource and could be used by plant breeders to improve biomass for biorefining.
Partner 7 (North Energy)
The MS Excel workbooks, incorporating considerable functionality and modelling capability, produced for the evaluation of primary energy inputs and total GHG, as part of LCA, can assist researchers by directing their efforts to promising areas of inquiry, can inform process engineers involved in optimization of biorefinery design, and can inform policy-makers.
Partner 8 (Processum)
Hydrolysation experiments could inform future design of similar experiments.
Partner 9 (University of Cambridge)
The identification of xylan fine structure in grasses and the identification of oligosaccharides released by industrially relevant xylanases are examples of exploitable foreground in the form of ‘General Advancement of Knowledge.’ The identification of oligosaccharides from xylan allows determination of the products of enzymatic hydrolysis of crop grasses and determination of alterations in plants for breeding improved varieties, such as low ferulate for improved digestibility. This could be exploited by enzyme companies improving the conversion of straw or crop residues into fermentable sugars, or by crop breeding companies to identify reduced feruloylation in grasses. Further research would be needed, to confirm the qualitative nature of the ferulated oligosaccharide studies. The potential impact for industry could be substantial if next generation bioenergy becomes commercial. The impact on crop breeding would likely be small, since there is little breeding for forage grasses in Europe.
Partner 10 (University of Leeds)
Analysis and interpretation of policy and regulatory frameworks will enable any kind of influential developments that could affect the future potential to harness benefits from this work to be taken into account. However, policy changes regularly, so any exploitation in the future would require re-analysis according to the most recent developments.
Partner 11 (University of Sheffield)
Knowledge of improved catalysts and operational conditions for lignin degradation and potential product generation is an important step in optimizing future industrial processes. Catalyst and biocatalyst companies could benefit from the knowledge in developing new routes. Furthermore, insight has been gained in hybrid catalytic and biocatalyst based processes that should be of interest to companies in the biorefinery field. We have also tested a range of organisms for treatment of lignocellulosic feedstocks. This widens knowledge on biotechnology of these processes. These organisms might be suitable for direct industrial application or be the basis, either as host chassis or a provision of genetic material for subsequent metabolic engineering or synthetic biology constructs for industrial exploitation either by academic teams or industry.
Partner 12 (VIB UGent)
Identification of novel genes involved in lignin biosynthesis could have important implications for plant breeding. The use of CSPP in phenolic profiling generated a catalogue of aromatic molecules from maize; most of which had not been reported before. This information could be very significant when designing a potential biorefinery process.
Partner 13 (Wageningen University)
The genetics and transcriptomics data generated by P13 will lead to the discovery of QTL for biomass quality in Miscanthus, with consequent value for crop breeders. An orthology web-interface database has been developed and is designed to become a tool that can be used by the wider scientific community to study gene expression in any orphan crop.
Partner 14 (BTCL)
Information gleaned from fermentation experiments with BTCL’s Geobacilli strains will inform future, similar experiments and could be of value in designing future industrial processes to produce bioethanol efficiently and economically from biomass-derived feedstocks such as Miscanthus and maize, which were tested during this project.
4.1.4.3.1.2 Quantification of Exploitation of Foreground
Partners have estimated the value in monetary terms of potential exploitation of the SUNLIBB foreground. For example, based on the experience obtained in SUNLIBB, P1 (University of York) has applied for a joint BBSRC-FAPESP (Fundaçâo de Amparo à Pesquisa do Estado de Sâo Paulo) grant worth £573,000 to the UK and similar to Prof. Polikarpov’s group in Brazil (total value approx. €1,000,000).
P7 (North Energy Associates) are partners in a lignocellulosic processing proposal to the EC that is related to SUNLIBB. If successful, the value of this to North Energy would be €198,416.
Experience gained as a result of SUNLIBB contributed to the acquisition by P9 (University of Cambridge) of a £487,000 BBSRC grant on xylan arabinosyltransferases. Also, researchers visiting P9 were awarded €30,000 for subsistence (from Brazilian funders); this was spent in the UK.
P13 (Wageningen University) has calculated that if 1 million hectares of land in the EU were planted with Miscanthus, this would generate a turnover of €160,000,000 per year (and would sequester 18,000,000 tonnes CO2 per year).
4.1.4.3.2 Dissemination activities
4.1.4.3.2.1 Dissemination of project results to the scientific community
The main outlet for research results was publication in peer-reviewed journals (see list of publications in Template A1). As a result of research funded by the SUNLIBB project, 13 papers were published in high-impact journals by consortium members between 2010 and 2014, with at least a further 11 at the manuscript or submission stage by December 2014. In addition, presentations were made by representatives of the SUNLIBB consortium at European and International conferences and events, including among many others, the European Biomass Conference and Exhibition in Hamburg (2014) and the Improving Feedstocks for Biorefineries Conference in Buenos Aires, Argentina (2013). Further details are presented in the list of dissemination activities in Template A2. These events provided opportunities for consortium members to network with colleagues and contacts outside SUNLIBB/CeProBIO.
4.1.4.3.2.2 Dissemination of project results to the wider community
A dedicated SUNLIBB website (www.sunlibb.eu) was established and maintained by the SUNLIBB Project Manager at CNAP, University of York. The website has a public area, where documents and information accessible to the general public are displayed. The Members’ Area is password-protected and holds information for access only to members of the SUNLIBB consortium. Web analytics were installed in September 2014, so that web traffic could be monitored with respect to country of origin, number of pages viewed, length of visit etc. For example, in the month from 3rd October to 2nd November 2014, analysis showed there had been 218 sessions, by 178 users, 76% from new visitors. The highest proportion of visitors (42.66%) was from the USA, with the next highest (17.89%) from the UK and the third highest (14.22%) from Brazil. The average session duration was 2 minutes 27 seconds.
Newsletters were produced annually, circulated to interested parties and displayed on the SUNLIBB website. Posters were produced, describing the Aims and Achievements of SUNLIBB. These were displayed at events such as the Hamburg conference mentioned above, and are available for download from the SUNLIBB website (public area).
Three Policy Briefs were produced during the course of the project. These are available for download from the public area of the SUNLIBB website and were circulated among stakeholders and a distribution list of interested parties. This included, for example, the Energy Leeds network, the Energy-L mailing list (run by the International Institute for Sustainable Development (IISD)), the representative in charge of Energy and Environment affairs in the Delegation of Brazil to the EU, the representative in charge of Innovation and Science in the EU Delegation to Brasilia and the Office of UNICA (Brazil’s sugarcane lobby) in Brussels.
SUNLIBB members took part in a variety of outreach events, as described in Template A2. These involved communicating aspects of the project’s work to various audiences, including the general public and children.
4.1.4.3.2.3 Training and Education
The SUNLIBB consortium recognizes its responsibility for the employees associated with the project, especially the 4 PhD students and 11 Post-Doctoral researchers who were employed to carry out much of the research work (see Workforce Statistics). The consortium as a whole provided these scientists with excellent training and a strong basis for establishing networks (see Template A2, dissemination activities). Each Project Review meeting was attended by junior scientists, who were given ample opportunity to present their results and participate in discussions with other consortium members in a collegial atmosphere.
4.1.4.3.2.4 SUNLIBB/CeProBIO Collaborations
As indicated previously, collaborations between scientists in SUNLIBB and CeProBIO widened the scope of the research and greatly facilitated the progress made. The coordination of research activities led to the arrangement of exchange placements between SUNLIBB and CeProBIO partners. The exchanges took place at all levels of seniority and involved PhD students as well as more experienced researchers.
Examples of collaborative research and exchange visits are listed for each Work Package, below.
WP1
Early discussions between European and Brazilian counterparts were helpful in bringing insights from maize, Miscanthus and sugarcane together, initially to define the approach to be taken with each crop. This was particularly the case for sugarcane and Miscanthus, which are closely related and similarly challenging in terms of genetics.
WP2
At the start of the project, transcriptomics studies in sugarcane were much further advanced than in Miscanthus, because of the larger research community and the more extensive sequence and expression data available. This allowed Brazilian researchers to share expertise and experience with their European colleagues and helped to define more effective ways forward with the Miscanthus work. Having access to the sugarcane data facilitated annotation of the Miscanthus transcriptome data. For example, Prof. Glaucia Souza’s group from the University of São Paulo (USP) shared sequence and expression data with A/Prof. Luisa Trindade’s group at Wageningen University (P13). This greatly assisted the identification of key genes in the cell wall biosynthetic pathway in Miscanthus.
WP3
Negotiations early in the project between the USP and the University of Cambridge (P9) allowed access by P9 to the SUCEST database and facilitated the generation of candidate gene lists.
Augusto Crivellari from Prof. Marcos Buckeridge’s lab at the USP visited the University of Cambridge for one month in 2011. He received training in analysis of sugarcane polysaccharides.
A CTBE researcher, Dr. Marco Tine, visited the University of Cambridge in 2012, for a period of 6 months. Lucio Mendoça from USP Chemistry institute also spent six months working on feruloylated oligosaccharides in Cambridge.
Dr. Matthias Sorieul, a Post-Doctoral Researcher at the University of Cambridge (P9) spent 3 weeks at the USP in 2013. He used the specialized Biogel columns available at the USP to isolate large oligosaccharides released from Miscanthus stems and sugarcane bagasse after GH30 digest. The samples were isolated at the USP and were structurally characterized in Cambridge. Matthias also learned a protocol used routinely at the USP; the preparation of delignified hemicellulose. This technique was then used at the University of Cambridge.
Dr. Mathias Sorieul visited the USP again in 2014. He set up the USP’s first DASH experiment (DNA sequencer-Assisted Saccharide analysis in High throughput). This technique was developed at Cambridge and allows the rapid analysis of cell wall oligosaccharides. After a few optimisations, the experiment was running successfully.
Matheus Pinheiro from USP Ribeiro Preto began a year’s placement in October 2014 to work on xylanase characterisation using the DASH technique.
Prof. Igor Polikarpov (USP) has started to work on the characterisation of GT61 and other xylan synthesis enzyme structures.
Prof. Munir Skaf (University of Campinas, Institute of Chemistry, USP) worked with the University of Cambridge on the dynamic modelling of xylan on cellulose fibre. This led to a joint article in The Plant Journal (Busse-Wicher et al. Plant J. 2014 Aug; 79(3):492-506).
Oigres Daniel Bernardinelli from USP Sao Carlos will begin a six month placement in February 2015, to perform solid state NMR experiments on xylan structure. He will be based in Cambridge and at the University of Warwick.
WP4
VIB-U.Ghent (P12) collaborated with Prof. Marcelo Ehlers Loureiro (Department of Plant Biology from the Federal University of Viçosa; CeProBIO P4) to fulfil the requirements of Task 4.3 (phenolic pathway discovery by targeted metabolomics and integration into metabolic maps), by performing phenolic profiling on sugarcane stem material, followed by in-depth identification of the phenolic metabolites. The material was harvested and processed in the Loureiro lab and then sent to VIB-U.Ghent for UHPLC-MS analysis. Maria Esther Ricci Da Silva from Prof. Loureiro’s group visited Prof. Wout Boerjan’s lab at the VIB-Ghent (SUNLIBB P12) for a six month training period.
VIB-Ghent also collaborated with Wanderley dos Santos’ team (CTBE/ Universidade Estadual de Maringá) to investigate the biosynthetic route of monolignols (lignin precursors) and other phenylpropanoids in sugarcane. In the light of this collaboration, Nathalia Gomes Mattos from Wanderley dos Santos’ lab visited the VIB-Ghent for one year (February 2012 – 2013). She carried out work on sugarcane samples treated with a lignin biosynthesis inhibitor.
VIB-Ghent also collaborated with Brazil to determine the major phenolic degradation products released from bagasse that could represent added-value products.
WP5
Scientists from the University of York (CNAP) collaborated with colleagues at the USP to establish the High-Throughput saccharification assay for sugarcane in Brazil.
Dr. Leonardo Gomez (P1, CNAP) visited the CTBE (Brazilian Bioethanol Science and Technology Laboratory in São Paulo, the University of São Paulo (USP) and USP – São Carlos in March 2011, in order to coordinate various activities in the consortium. As part of these activities, Marisa Lima, a PhD student working in Prof. Igor Polikarpov’s laboratory at the USP, spent 6 months at CNAP in York, carrying out a comprehensive characterization of different biomass crops. As a result of that coordinated activity, two manuscripts were published in the journal Biotechnology for Biofuels.
Dr. Leonardo Gomez also visited the CTBE and Prof. Polikarpov’s laboratory for two weeks in June 2013.
Prof. Simon McQueen-Mason’s group at CNAP hosted for one year (2013 – 2014) a Brazilian PhD student from the laboratory of Prof. Nei Pereira Jr. from the Federal University of Rio de Janeiro. Her task was to explore the function of accessory proteins in the secretome of the fungus Trichoderma hartzianum. A further collaboration with Prof. Richard Ward from the USP provided novel xylanase chimeras for testing in biomass deconstruction.
A PhD student from the Polikarpov lab spent 6 months at Prof. McQueen-Mason’s lab, characterizing three potential new biomass species selected by EMBRAPA and developing effective deconstruction methodologies for these.
WP6
Strict Brazilian export laws on the movement of biomass meant that only limited amounts of bagasse could be sent to York for analysis. Consequently, Maggie Zhu and Thomas Attard, both PhD students from the University of York, visited Prof. Polikarpov’s laboratory in São Carlos for a period of 8 weeks. Maggie Zhu investigated microwave pre-treatments of bagasse (before enzymatic hydrolysis). Thomas Attard conducted extraction of high value chemicals from sugar cane and bagasse.
WP8
There were extensive discussions over the details of biofuels policies in Brazil and the European Union and an exchange of relevant information such as the EC’s Renewable Energy Directive and supporting documentation.
Dr. Nicola Favretto from the University of Leeds visited Brazil to meet CeProBIO partners at the USP in 2013 and conducted a series of interviews with Brazilian stakeholders on EU-Brazil bilateral cooperation, with regard to second generation biofuels. With the help of CeProBIO partner Prof. Marcos Buckeridge of the USP, a wide range of academics and policymakers was identified and interviewed, from institutions such as Petrobras, the Brazilian Development Bank (BNDES) and the University of Rio de Janeiro.
Dr. Stavros Afionis from the University of Leeds (P10) visited Prof. Marcos Buckeridge at the USP for 3 weeks in August 2014. The purpose of the visit was to conduct follow-up interviews with relevant stakeholders in São Paulo, Rio de Janeiro and Brasilia, with regard to EU-Brazil cooperation on second generation biofuels.
Examples of workbooks developed in WP8 were provided to Brazilian colleagues at the USP. Subsequently, a joint SUNLIBB/CeProBIO workshop on the development and use of such workbooks for biofuel GHG emission calculations was held in San Pedro, Brazil in 2011. Regular exchanges of information were maintained between North Energy Associates (P7) and Prof. Paulo Seleghim (CeProBIO) and colleagues, including the provision of “Sugar Cane Bioethanol: R&D for Productivity and Sustainability,” edited by Luis Augusto Barbosa Cortez.

4.1.4.3.2.5 SUNLIBB/CeProBIO Cooperation Plan – Joint Deliverables
It was recognized from the start that close collaboration with CeProBIO would be an integral part of the SUNLIBB project. To this end, a cooperation plan was prepared and agreed between both consortia, setting out detailed objectives that required input from members of SUNLIBB and CeProBIO, and providing a means of measuring progress, so that the work programme proceeded according to plan in all areas. These shared objectives are listed in the table of Joint Deliverables, below.
Joint Deliverable Reports are available on the SUNLIBB website Members’ area.

JD1.1 Plant material from Miscanthus breeding population and sugar cane delivered for saccharification analysis. Marcos Buckeridge (MB) & Igor Polikarpov (IP) (CeProBIO) delivered biomass from sugarcane varieties toWP5 (P1) for analysis. Luisa Trindade (LT) (P13) delivered Miscanthus samples to WP5 for analysis . Completed month 12. Samples were delivered on time for analysis.
JD1.2 Insights into the interactions between cell wall composition and saccharification in sugarcane and maize. MB, IP and Anete de Souza (AdS) (CeProBIO) shared and discussed data on cell wall composition and saccharification potential with SUNLIBB partners P6. P1, P9 and P12. This was completed in month 24 for maize and Miscanthus, and month 50 for sugarcane.
JD1.3 Plant material from mapping populations delivered for saccharification analysis. Plant material was delivered to P1 by Matthieu Reymond (MR) (P6) (maize), LT (P13) (Miscanthus) and AdS (CeProBIO) (sugarcane). This was completed in month 24 for maize and Miscanthus, and month 34 for sugarcane.
JD1.4 Markers for Miscanthus, maize and sugarcane biomass saccharification identified. Markers were identified by MR (P6) (maize), LT (P13) (Miscanthus) and AdS (CeProBIO) (sugarcane). QTL were reported in maize, & possible QTL in Miscanthus and sugarcane at the final project meeting in Ghent, 2014.
JD2.1 Initial 15 target genes identified for mutant analysis in maize and sugarcane. Paul Dupree (PD) (P9) and Wout Boerjan (WB) (P12) made gene list available to Jeremy Derory (JD) (P4), AdS and Glaucia de Souza (GdS) (CeProBIO). Completed for polysaccharides in Month 3.
JD2.2 A detailed characterization of patterns of secondary cell wall formation in stems of sugarcane and Miscanthus. MB & Marcelo Ehlers Loureiro (MEL) (CeProBIO) sectioned sugarcane stems. LT (P13) sectioned Miscanthus stems. PD (P9), WB (P12) and Herman Hofte (P6) analysed composition. Completed month 24.
JD2.3 High quality RNA extracts from appropriate stages of Miscanthus and sugarcane. LT (P13) provided RNA from Miscanthus. GdS (CeProBIO) provided RNA from sugarcane. Completed by month 30 for Miscanthus.
JD2.4 Transcript profiles for the different stages of secondary cell wall development in Miscanthus, maize and sugarcane genes. LT (P13) analysed transcript profiles for Miscanthus , JD (P4) did the same for maize & GdS (CeProBIO) analysed sugarcane. Maize transcriptomic profile was delivered in month 26. Miscanthus transcriptomic profile was delivered in month 40.

JD2.5 Target genes identified for WPs3&4 from Miscanthus, maize and sugarcane. Target genes were identified in Miscanthus by LT (P13), in maize by JD (P4) and by GdS (CeProBIO) in sugarcane. Information was sent to PD (P9) and WB (P12). Target genes were delivered for maize by month 24 and for Miscanthus by month 44.

JD3.1 Identification of gene candidates in matrix polysaccharide synthesis based on work in model species. PD (P9) and MB (CeProBIO) identified gene candidates in matrix polysaccharide synthesis based on work in model species and on data generated from LT (P13) (Miscanthus), JD (P4) (maize) and GdS (CeProBIO) (sugarcane). This was achieved by month 18 for maize and sugarcane, but delayed for Miscanthus because the available genome sequence was of poor quality.

JD3.2 Detailed characterization of maize, Miscanthus and sugarcane matrix polysaccharides. PD (P9) and MB (CeProBIO) characterized matrix polysaccharides from maize, Miscanthus and sugarcane, using material and data obtained from JD (P4) (maize), LT (P13) (Miscanthus) and GdS (CeProBIO) (sugarcane). Achieved and completed as reported in D3.2 and D3.3 for maize and Miscanthus (month 30).
Sugarcane was completed in month 36.
JD3.3 Identification of additional 30 gene candidates involved in matrix polysaccharide synthesis. JD (P4) and MB (CeProBIO) identified 30 gene candidates involved in matrix polysaccharide synthesis, using data from maize (JD), from Miscanthus (LT) and sugarcane (GdS), and compared with homologous genes in model species by PD (P9). Completed month 30. Achieved and completed as reported in D3.4
JD3.4 Identification of activities of 3 new enzymes. PD (P9) and IP (CeProBIO) identified 3 new enzymes. Completed month 30
JD3.5 Five maize or Miscanthus or sugarcane modified lines with altered cell wall polysaccharides characterised. Maize lines with altered cell wall polysaccharides were provided from JD and characterized by PD (P9), MB (CeProBIO) and Leonardo Gomez (P1). Completed month 50 See D3.6
JD4.1 Lignin biosynthesis gene orthologues from maize, Miscanthus and sugarcane. WB (P12) identified Arabidopsis genes involved in lignin biosynthesis. Orthologs of these genes were identified in Miscanthus (LT), maize (JD & MR) and sugarcane (MEL), using previously generated transcriptomic data combined with co-expression data. Completed in month 20 for maize.
The identification of orthologs was delayed in Miscanthus, due to the poor quality of the genome sequence data available.
JD4.2 A catalogue of aromatic molecules from C4 plants: continuous throughout the project. WB (P12) and MEL (CeProBIO) provided a catalogue of aromatic molecules from maize, Miscanthus and sugarcane. This was completed for maize and Miscanthus in month 30 and updated for sugarcane in month 50.
JD4.3 Phenolic metabolic maps for maize and sugarcane, integrating transcripts, metabolites and altered fluxes. WB (P12), GdS & MEL (CeProBIO) provided phenolic metabolic maps for maize and sugarcane, integrating transcripts, metabolites and altered fluxes. Completed month 48
JD4.4 Set of novel target lignin genes in C4 grasses identified through a systems biology approach WB (P12), GdS & MEL (CeProBIO) identified a set of novel target lignin genes in C4 grasses, through a systems biology approach. Complete month 40
JD5.1 High-Throughput (HT) saccharification assay established in Brazil. MB, IP (CeProBIO) and LG (P1) established a HT saccharification assay in Brazil. Completed month 24 A HT platform was set up in Sao Carlos by IP, and was optimized to deliver a range of functions for biomass characterization.
JD5.2 HT saccharification assay validated for maize, Miscanthus and sugarcane. MB (CeProBIO) and LG (P1) validated the HT saccharification assay for biomass samples from maize (MR; P6), Miscanthus (LT; P13) and sugarcane (AdS; CeProBIO). Completed month 24 A saccharification assay was established and validated on hundreds of samples of maize, Miscanthus and sugarcane.
JD5.3 Miscanthus and sugarcane lines with more digestible biomass identified. MB (CeProBIO) and LG (P1) identified Miscanthus and sugarcane lines with more digestible biomass using HT saccharification. Completed month 36
JD5.4 Detailed description of saccharide products released from biomass of 3 crops. MB (CeProBIO) and LG (P1) provided a detailed description of saccharide products released from maize, Miscanthus and sugarcane biomass Completed month 36
JD5.5 Four new cellulases trialled for biomass saccharification. IP (CeProBIO) and LG (P1) trialled 4 new cellulases or cellulose cocktails for biomass saccharification. Completed month 40
JD6.1 A fully-characterised range of hydrophobic molecules from supercritical CO2 extracts of C4 biomass. MR (P6), LT (P13) and MB (CeProBIO) provided samples of maize, Miscanthus and sugarcane, respectively, to Andrew Hunt (AH) (P1). AH delivered a fully characterized range of hydrophobic molecules from sCO2 extraction of the biomass. Completed month 20
JD6.2 Optimised sCO2 extraction protocols for marketable waxes and other compounds. AH (P1) optimized sCO2 extraction of marketable waxes and other compounds. Material was provided by MR (maize), LT (Miscanthus) and MB (sugarcane). Completed month 30
JD6.3 Production of high value phenolic compounds from C4 biomass. Biomass was pre-treated using different methods by LG (P1). WB (P12) analysed the products generated to identify valuable lignin monomers and oligomers. Material was provided by MR (maize), LT (Miscanthus) and IP (sugarcane). Paulo Seleghim (PS) (CeProBIO) optimized lab-scale fermentation of sugarcane and investigated the extraction of lignin and lignin breakdown products. Completed month 38
JD7.1 Lab-scale analysis of 10 integrated process scenarios using maize, Miscanthus and sugarcane. Completed month 30
JD7.3 In-silico optimization studies from lab-scale data completed. PS (CeProBIO) and Phillip Wright (PW) (P11) carried out optimization studies on lab-scale fermentation data Competed month 38
JD7.4 Pilot-scale analysis of 3 integrated process scenarios for each species. PS and MB (CeProBIO) and David Blomberg (DB) (P8) carried out pilot-scale analysis of 3 integrated process scenarios for maize, Miscanthus and sugarcane. Completed month 48
JD7.5 Pilot-scale fermentation of sugars from maize, Miscanthus and sugarcane biomass optimized. Namdar Baghaei-Yazdi (P14), PS and MB (CeProBIO) optimized pilot-scale fermentation of sugars from maize, Miscanthus and sugarcane. Completed month 48
JD8.1 Biorefinery model workbook. Nigel Mortimer (NM) (P7) and PS (CeProBIO) developed a biorefinery model workbook. Completed month 27 Collaboration took place to populate the Biorefinery Model workbook (D8.2) with Brazilian sugarcane data, to produce SUNLIBB Sugar Cane Biorefinery v012.xlxs for JD8.1
JD9.1 Agreement with Brazilian sister consortium signed by all parties. Simon-McQueen-Mason (SMM) (P1) and IP (CeProBIO) coordinated the signing of the agreement between SUNLIBB and CeProBIO. Completed month 17 This deliverable was delayed, because the negotiations took longer than expected.
JD9.2 Coordination Plan. SMM (P1) and IP (CeProBIO) wrote the coordination plan, detailing the Joint Deliverables. Completed month 10. This deliverable was delayed, because the negotiations took longer than expected.



4.4 Technology Readiness Levels (TRLs)
TRLs are measures used to assess the maturity of evolving technologies.
The following definitions (table 4.4.1) are based on those in the H2020 2014-2015 General Annex (G); ec.europa.eu.

Table 4.4.1 Technology Readiness Levels (TRLs) - Definitions
TRL Definition
1 Basic Principles observed (Basic Research)
2 Technology concept formulated
3 Experimental proof of concept (Applied Research)
4 Technology validated in lab (Small-scale prototype)
5 Technology validated in relevant environment (e.g industry) (Large-scale prototype)
6 Prototype system
7 System prototype demonstration in operational environment (Demonstration scale)
8 System complete and qualified (First of the kind commercial system)
9 Actual system proven in operational environment (Full commercial application)

The technologies developed by the Work Packages in SUNLIBB are estimated to have reached the following TRLs at the end of the project (November 2014) (table 4.4.2):

Table 4.4.2
Work Package Task Description TRL
WP1 1.1 Develop HT methods for cell wall analysis 3
WP1 1.2 Identification of QTL for biomass composition and saccharification potential in maize 2
WP1 1.3 Assessing the potential of candidate genes as markers for improved biomass digestibility 2
WP2 2.1 Generating transcriptomic data from Miscanthus 2
WP2 2.2 Mining transcriptomic data 2
WP3 3.1 Characterising cell wall polysaccharide composition, quantity and structure 2/3
WP3 3.2 Mining transcriptomic data for polysaccharide biosynthetic genes 2/3
WP3 3.3 Demonstrating gene function 2/3
WP4 4.1 Mining transcriptomic data for genes involved in lignin biosynthesis 2/3
WP4 4.2 Transformation of maize and Miscanthus with new genes involved in lignin biosynthesis 2/3
WP4 4.3 Phenolic pathway discovery by targeted metabolomics and integration into metabolic maps 2/3
WP5 5.1 Characterisation of sugars and oligosaccharides released during deconstruction 3/4
WP5 5.2 Determining the structure of lignin breakdown products 3/4
WP5 5.3 High-Throughput saccharification screening 3/4
WP5 5.4 Examining pre-treatment efficiency for enzymatic saccharification 3/4
WP5 5.5 Trialling new pre-treatments 3/4
WP5 5.6 Trialling the use of alternative enzymes for saccharification 3/4
WP5 5.7 Enzymatic lignin breakdown 3
WP6 6.1 Characterising and extracting high value molecules with CO2 3/4
WP6 6.2 Added-value products from lignin 4
WP6 6.3 Lignin fermentation 3/4
WP6 6.4 Value from pentose sugars 4
WP7 7.1 Integrating the extraction of added-value products with ethanol production 5
WP7 7.2 Fermentation of sugars from C4 grass biomass 4/5
WP7 7.3 Lab and pilot scale sophorolipid fermentation trials using improved feedstock from WP4 4
WP7 7.4 Pilot studies 5
WP7 7.5 Demonstration of added-value products from lignin breakdown 5
WP7 7.6 Integrating improved feedstock 4/5
WP8 8.1 Review of policy and regulatory context at EU and individual country level n/a
WP8 8.2 Model development n/a
WP8 8.3 Application to maize biorefineries n/a
WP8 8.4 Application to Miscanthus biorefineries n/a
WP8 8.5 Sensitivity and comparative analysis n/a

Table 4.4.2 shows that some SUNLIBB Work Packages reached higher TRLs than others by the end of the project. This was expected, because of the nature of the tasks within the different Work Packages. However, the Work Packages were closely linked, and mutually interdependent. To summarize, WPs 1 to 4 were concerned with gene discovery and the identification of components of cell walls and of biosynthetic pathways. The information generated was used to try to increase the digestibility of lignocellulosic biomass through improved genetics. Biomass from these Work Packages was then intended to feed into laboratory and pilot-scale experiments in WPs 5 & 7, so that conclusions could be drawn about whether these approaches were promising.
The designation of relatively low TRLs to tasks in WPs 1 & 2 reflects the nature of the work involved; namely gene mapping and transcriptomics studies, leading to the discovery of QTL for digestibility in maize, Miscanthus and sugarcane. Although this is relatively basic research, the results have important applications in terms of marker generation for plant breeding, besides the knock-on importance for the other Work Packages and the output of SUNLIBB. The TRL assigned does not in any way imply that WPs 1 & 2 lagged behind the other Work Packages, because all the work was closely integrated, as explained above.
WP6 was concerned with developing methods to generate added value products from biomass processing, at lab and pilot scale, hence the estimated TRLs of 3 to 4.

List of Websites:
SUNLIBB - Sustainable Liquid Biofuels from biomass Biorefining Grant Agreement Number 251132

Public website address: www.sunlibb.eu

Contact Details: Project Coordinator

Prof. Simon McQueen-Mason
CNAP, Department of Biology
University of York
Heslington
YORK
YO10 5DD
UK

Tel: +44 (0) 1904 328775

E-mail: sjmm1@york.ac.uk