Biorefinery of brewery wastes to produce bioactive molecules with food interest (bacteriocins and biosurfactants)

Résumé

The demographic and economic expansion, with the consequent changes in the habits of consumption, has caused an increment in the generation of wastes. Traditional waste management methods, such as landfilling or incineration, become ineffective and unsustainable, adding pressure on the already damaged environment. Due to this, it is necessary to abandon the current linear economic system and adopt a circular model, in which resources are maximized and waste is reduced until reaching the Zero Waste concept. The objective is not only to mitigate existing environmental problems, but also to respond to the need to produce food and energy in a safe and sustainable way. In recent years, the concept of biorefinery has emerged, based on obtaining energy and other products through the processing of biomass. Lignocellulosic biomass, like forest and agricultural residues, is suitable for this purpose. Among this lignocellulosic biomass it can be highlighted the agro-industrial wastes generated during the processing of food industries, since these industries produce huge amounts of wastes annually, much of which can cause environmental pollution and pose a health risk if not treated properly. The use of these wastes in the biorefinery implies proper management of the waste while it is being valorized. In this context, brewery spent grain (BSG) is presented as a very attractive material. The BSG is the main waste derived from the beer production. A worldwide annual generation of BSG is estimated at 39 million tons, which makes it an abundant, low cost and high availability material. It is a lignocellulosic material formed, mainly by the husks of barley grain, which presents an interesting composition for multiple purposes in biorefinery. An interesting aim in biorefinery processes is to obtain products with high added value such as lactic acid, bacteriocins or biosurfactants, all of them with multiple applications, such as in the food industry. Lactic acid bacteria (LAB) are a very large and heterogeneous group of microorganisms capable of producing lactic acid as a result of carbohydrate metabolism. They are Gram-positive bacteria considered GRAS (Generally Recognized as Safe); therefore, in general, both, LAB and their products can be applied in food products. On the other side, bacteriocins are a group of peptides or proteins, synthesized by bacteria, that show antimicrobial activity against bacteria of the same species or genus. Bacteriocins are presented as an alternative to antibiotics in food, with the advantage of being able to be classified as of natural origin. In addition, biosurfactants are active compounds with the ability to reduce surface and interfacial tension and promote the emulsion of immiscible liquids, which can be produced, among others, by LAB. Biosurfactants are natural compounds with multiple applications in the pharmaceutical, cosmetic, food industry and in environmental bioremediation. In the present thesis, the use of BSG has been proposed from various points of view. First of all, BSG was used as a substrate for the production of cellulolytic enzymes by solid state fermentation (SSF) with filamentous fungi. These enzymes were further applied to fractionate BSG in order to obtain fermentable sugars solutions, from which high added value compounds (bacteriocins and biosurfactants) were produced by lactic acid bacteria. To achieve these goals one of the main drawbacks to be avoided is the recalcitrance of lignocellulosic materials such as BSG. Therefore, this material was pretreated to improve the efficiency of the enzymatic hydrolysis. In this way, pretreatment stages have been considered one of the key steps in biomass processing, where the main objective is to address the causes of resistance to degradation presented by these materials such as lignin content, crystallinity or contact area that can limit the access of enzymes. This problem has been tackled from different approaches including physical, chemical, physical-chemical and biological methods. However, current efforts are focused on developing methods that manage to modify or fractionate the material in such a way that it is possible to obtain sugars efficiently, without generating toxic or inhibitory compounds for the subsequent stages. This is how pretreatments with ionic liquids (IL) are presented. ILs are compounds formed exclusively by ions that have particular characteristics, which allow the pretreatment of biomass in milder conditions compared to other methods. Although these compounds were considered "green" at first due to their recyclability, this qualification was later reconsidered for certain ILs as they are not biodegradable and are toxic and persistent in the environment. In recent years, attention has focused on bio-derivative ILs, such as those with choline as a cation. Choline-based ILs that include an amino acid as an anion are of special interest, since they have an ability to dissolve biomass, overcoming the environmental problems caused by traditional ILs, as they are completely biodegradable. In this thesis, IL choline glycinate [N1112OH] [Gly] was assayed as pretreatment for BSG fractionation. The main results obtained in the development of the BSG biorefinery within the framework of this thesis are summarized below. First, in the article "Biorefining brewery spent grain polysaccharides through biotuning of ionic liquids", the fractionation of BSG in its constituent polymers was studied using a pretreatment based on IL [N1112OH][Gly]. Lignin is one of the main limiting factors for enzymatic hydrolysis, which is why the aim of this work was to obtain a material with a reduced lignin content. In this way, temperature is a crucial factor in these processes since, at high temperatures, the polysaccharides in the biomass can be degraded. Under all the temperatures studied (60, 90, 120 and 150 ºC) for 16 h, it was possible to separate the BSG into two fractions: a carbohydrate rich material (CRM) and a lignin rich material (LRM). The most drastic reduction in lignin content (75.89%) occurred in the treatment carried out at 90 ºC. With comparative purposes, the same pretreatment was also performed with the traditional IL 1-ethyl-3-methylimidazolium acetate [C2C1im][C1COO]. In this case, the fractionation of the BSG was also achieved but only with a 40.18% of delignification. BSG and the fractions obtained after the pretreatments (CRM and LRM) were analyzed by various techniques such as ATR-FTIR and scanning electron microscopy (SEM), to verify the structural modifications suffered by the material. The FTIR spectra show that the CRM obtained after treatment with [N1112OH][Gly] were mainly made up of carbohydrates and presented residual lignin, while the LRM only presented characteristic lignin peaks. On the contrary, the CRM obtained after treatment with [C2C1im][C1COO] contained higher amounts of lignin, so the treatment was less effective. SEM images showed changes in the cell walls of the BSG after treatment with [N1112OH][Gly], presenting a less compacted structure with fibrils and visible pores. The CRM from the commercial IL treatment showed a similar structure to BSG. Afterwards, an enzymatic hydrolysis was carried out with a commercial cellulase preparation on the BSG and the CRM obtained with the two IL treatments. The saccharification of cellulose for BSG and CRM from the [C2C1im][C1COO] pretreatment was 59.52% and 62.96% respectively, meanwhile the [N1112OH][Gly] pretreatment CRM reached up to 94.25%, which it is about 1.5-fold improvement over control (untreated BSG) or after pretreatment with imidazolium-based IL. Finally, in order to make this process sustainable, it is necessary to study the recycling of IL. Therefore, the IL [N1112OH][Gly] was analyzed by ATR-FTIR, finding impurities (small amounts of residues from the process) in the samples of IL used in the treatment. Finally, the IL [N1112OH][Gly] was analyzed before and after pretreatment by ATR-FTIR, finding impurities (small amounts of residues from the process) in the samples of IL used in the treatment, which could limit the sustainability of this process, being necessary to study the recycling of IL in the next works. Therefore, once the efficacy in terms of IL delignification and improvement of enzymatic hydrolysis in BSG was demonstrated, the following work was focused on finding a suitable method to fractionate the BSG and recover and reuse the IL efficiently. In the article “Recovery and reuse of ionic liquid cholinium glycinate in the treatment of brewery spent grain” three strategies based on the use of antisolvents were proposed to study the reuse of IL [N1112OH][Gly] after BSG fractionation. Acetone (TA), ethanol (TB) and NaOH (TC) were selected as key compounds. With all treatments a significant delignification of BSG was achieved. The best results in terms of delignification and glucan recovery were achieved with TA. The recovery of the other fractions (xylan and arabinan), as well as the mass of the CRM recovered was similar in the three treatments, meanwhile TA presented a greater recovery of the IL. Hence, TA was selected to continue the study deeply. The purity and recovery of the IL were analyzed, as well as the composition and structure of the CRMs obtained after five treatment cycles. In all cases, IL recovery was greater than 80%, although purity decreased from 72% in the first cycle to 33% from the third. This reduction in purity was due to the accumulation of compounds derived from the treatment of biomass and residual water. Despite of this, the efficacy of the treatment was not affected in terms of delignification and carbohydrate recovery. The CRMs obtained were also analyzed by ATR-FTIR. All CRMs had a predominantly carbohydrate composition, appearing peaks associated with lignin with low intensity corresponding to residual lignin. Crystallinity, a crucial characteristic in the enzymatic hydrolysis of biomass, was also studied by ATR-FTIR and X-ray diffraction. CRMs presented higher crystallinity than BSG given their higher cellulose content, however, this cellulose was mostly amorphous, which favored its hydrolysis. In addition, SEM images showed a less compact structure and the presence of pores in CRMs, regardless of the treatment cycle from which they come. Acetone was the most suitable antisolvent. The TA method allowed a reuse of the IL in five treatment cycles, without the delignification being affected. The CRM obtained present a composition and structure more suitable for enzymatic hydrolysis. The suitability of the proposed process for the competitive sequential delignification of the BSG was evidenced. Thereafter, with the purpose of approaching a circular system, in the following article “A novel approach to biorefinery of brewery spent grain” a biorefinery in three steps was proposed for the complete fractionation of BSG. This approach involved a) the production of hydrolytic enzymes by SSF, b) the delignification of the resulting residue with IL and c) a hydrolysis of the delignified material, catalyzed with the enzymes obtained in the first point. Accordingly, an enzymatic cocktail, mainly of xylanases, was produced by SSF with the filamentous fungus Aspergillus brasiliensis CECT 2700. The SSF was carried out in a horizontal bioreactor of 7 L capacity, using BSG moistened with a salt solution as substrate. Kinetics of cellulolytic enzyme production were followed for 7 days. The maximum occurred after 4 days of fermentation, where an enzyme complex was obtained containing 3152 U of xylanase /g BSG, 7.26 FPU of cellulase/g BSG, 19.02 U of β-glucosidase /g BSG and 1.05 U of ferulic acid esterase/g BSG, using an initial inoculum of 1 x 106 spores/g BSG. After the production of enzymes by SSF, the pretreatment with the IL [N1112OH][Gly], previously described, was applied on the residue resulting from the SSF (denoted as SSR). This resulted in a material with a reduced lignin content (CRM) and a strong increase in glucan and, to a lesser extent, in xylan. Furthermore, the pretreatment caused structural changes in the cell walls of the SSR, resulting in a less compacted material with pores, observable in the SEM images. Finally, an enzymatic hydrolysis was carried out on the BSG, the SSR and the CRM, with the enzymatic extract obtained after the SSF with A. brasiliensis, to obtain fermentable sugars solutions, under non-optimized conditions. For comparison, the commercial enzyme preparations Ultraflo L and Sherzyme 500L were also used. First, the trial was carried out on raw BSG. The extract of A. brasiliensis and Ultraflo L offered similar performances in the saccharification of glucans and xylan. Sherzyme 500L showed poor results and was discarded for further studies. A completely different behavior was observed when the enzymatic hydrolysis was tested in SSR. No xylose, and only low amounts of glucose were released. Only by using Ultraflo L® it was possible to release arabinose. These poor results were attributed to the increment of lignin content after SSF that prevents or limits the access of the enzymes to the substrate. Therefore, the need for a pretreatment to increase the hydrolysis yield became evident. The compositional and structural changes that CRM presented with respect to SSR, resulted in a considerable increase in the release of sugars. It is evident that enzymatic hydrolysis was improved in CRM with respect to SSR with either of the two enzyme complexes used. The main difference observed was that xylose was the main sugar released when catalyzing the hydrolysis with the A. brasiliensis extract, while the use of the Ultraflo L® preparation led to a greater release of glucose and arabinose. Since the enzyme cocktails were equal in function of xylanase activity, the hydrolysis catalyzed by the Ultraflo L preparation, which contained a greater amount of cellulases and β-glucosidases, released a greater amount of glucose. Although enzymatic hydrolysis needs to be improved at this point, these results illustrate the benefits of the proposed model. Finally, in the handwritten work "Brewery spent grain biorefinery for the production of bacteriocins and biosurfactants" the complete process of using the BSG for the production of biomolecules was developed taking as a base the previous works. Additionally, in order to improve the amounts of cellulases the strain Trichoderma reesei was evaluated for SSF of BSG. Therefore, T. reesei and A. brasiliensis were assayed independently to produce enzymatic cocktails by SSF of BSG under static or stirring conditions. The higher activities of enzymes were achieved after 5 days of fermentation with stirring every 24 h for the two microorganisms. T. reesei showed a remarkable cellulase production meanwhile A. brasiliensis secreted an enzymatic cocktail with higher xylanase, CMCase and β-glucosidase activities. Afterwards, all the hydrolysis catalyzed by the T. reesei extract achieved good saccharifications of cellulose, corresponding to a large extent to the release of cellobiose, which indicates an incomplete hydrolysis, as it was not cleaved into glucose. This behavior was due to the deficiency of this enzymatic complex of β-glucosidases. Contrary, when using the extract of A. brasiliensis, the saccharification of cellulose was complete, resulting in a medium containing glucose. In general, the A. brasiliensis extract led to a greater release of sugars but only hydrolyzing approximately half of the glucan and xylan. This led to the idea of using a combination of both extracts to solve the deficiencies that they may have and favor synergistic actions. Better yields were achieved when using mixtures of the enzymatic cocktails, showing a tendency to improve with a higher percentage of the A. brasiliensis extract. The best results were achieved when using a mixture of enzymes in a 2.5:0.5 ratio (Aspergillus/Trichoderma) (v/v). In this case, the enzymatic hydrolysis of CRM was improved by 51.59% for glucan and 66.62% for xylan compared to crude BSG. This improvement supports the efficacy of the IL pretreatment. Subsequently, the sugar-rich solution obtained, was finally fermented with Lactobacillus plantarum CECT 221 and Lactobacillus pentosus CECT 4023 strains for 72 h, following the consumption of sugars and the production of organic acids, bacteriocins and biosurfactants. After 24 h, Lact. plantarum synthesized 1.33 g/L of acetic acid and 7.98 g/L of lactic acid, depleting glucose and arabinose. From this point on, the microorganism re-consumed lactic acid and continued with the production of acetic acid up to 8.81 g/L at 72 h. Lact. pentosus showed a different behavior where it did not deplete glucose completely, leaving 1.63 g/L after 24 h accompanied by the production of lactic acid (8.04 g/L) and acetic acid (0.55 g/L). By not depleting the main carbon source, this microorganism does not re-consumed the lactic acid produced. The concentrations of lactic acid and acetic acid increased slightly, remaining almost constant until the end of the fermentation. The antimicrobial activity of the obtained extracts was determined, calculating the inhibitory dose 50 (ID50) against Listeria monocytogenes CECT 934. In both cases, the best results were obtained after 24 h of fermentation. An ID50 of 6.69 BU/mL for Lact. plantarum and 6.25 BU/mL for Lact. pentosus were achieved by fermenting the CRM hydrolysate. These activities were higher when compared with those of the extracts obtained by fermenting the MRS commercial medium under the same conditions (4.49 BU/mL for Lact. plantarum and 5.31 BU/mL for Lact. pentosus). The limiting conditions in nutrients of the CRM hydrolysate can provoke a response of the cells to stress, which would explain the higher production of bacteriocin when using this medium. Regarding the production of biosurfactants, it is considered that a reduction of 8 mN/m in the surface tension of a medium is indicative of the presence of surfactants. Lact. plantarum and Lact. pentosus were able to produce biosurfactants, showing a reduction of 27.00 mN/m and 19.50 mN/m respectively, after 24 h of fermentation in the CRM hydrolysate. These values were higher than those obtained with the commercial MRS medium (19.50 mN/m and 18.00 mN/m respectively). Altogether, BSG is a suitable material for a biorefinery process. Pretreatment with IL [N1112OH] [Gly] turned out to be effective in delignifying and improving the enzymatic hydrolysis of BSG, while its recyclability makes it a sustainable process. It was possible to produce cellulolytic enzymes using BSG as a substrate, and then use them to obtain fermentable sugar solutions. The hydrolysates obtained were successfully fermented with lactic acid bacteria to produce bacteriocins and biosurfactants using an innovative and sustainable process.