Ecofriendly lignocellulose pretreatment to enhance the carboxylate production of a rumen-derived microbial consortium

Innovative dry chemo- and chemo-mechanical pretreatments form an interesting approach for 24 modifying the native physico-chemical composition of lignocellulose facilitating its microbial 25 conversion to carboxylates. Here, the impact of four dry-pretreatment conditions on the 26 microbial transformation of wheat straw was assessed: milling to 2 mm and 100 µm, and 27 NaOH chemical impregnation at high substrate concentrations combined with milling at 2 28 mm and 100 µm. Pretreatment effect was assessed in the light of substrate structure and 29 composition, its impact on the acidogenic potential and the major enzyme activities of a 30 rumen-derived microbial consortium RWS. Chemo-mechanical pretreatment strongly 31 modified the substrate macroporosity. The highest carboxylate production rate was reached 32 after dry chemo-mechanical treatment with NaOH at 100 µm. A positive impact of the dry 33 chemo-mechanical treatment on xylanase activity was observed also. These results underline 34 that increasing substrate macroporosity by dry chemo-mechanical pretreatment had a positive 35 impact on the microbial acidogenic potential. effect the pretreatments on the subsequent dynamics of key enzymatic Our DC pretreatment this increase pronounced pretreatment This increase was by an increase in early phase activity, but CMCase activity was unchanged. Finally, results indicate a 106 correlation between lignocellulose macroporosity and its degradability by a hydrolytic 107 microbial consortium.

and acetogenic microorganisms. In this respect, the carboxylate platform is related to the 50 biogas platform, a two-stage anaerobic digestion system that produces carboxylates as 51 intermediates that are then broken down by methanogenic archaea into methane. However, in 52 the carboxylate platform, the products are considered to be building blocks for the production 53 of added value chemicals and materials (e.g polyhydroxyalcanoate bioplastics), and liquid 54 biofuels (Agler et al, 2011;Torella et al, 2013), rather than intermediates for low value 55 renewable energy (i.e. bio-H 2 , and bio-CH 4 ), which is inevitably in direct competition with 56 low-priced fossil resources, such as shale gas. 57 Despite its attractive features, the carboxylate platform is nonetheless limited by the ability to 58 extract fermentable components from LC biomass, which constitutes a highly recalcitrant raw 59 material. Indeed, LC biomass is a composite material composed mainly of cellulose,  The effect of different pretreatments on LC biomass has been extensively studied (Alvira et  the absolute need for pretreatment is cleverly illustrated by the truism ''the only process more 78 expensive than pretreatment is no pretreatment'' (Wyman, 2007). 79 Among the vast array of LC biomass pretreatments that have been tested, alkaline 80 pretreatment using sodium hydroxide is one of the most effective and attractive methods. It   Barakat et al., 2013). DC pretreatment consists of moderate chemical treatment using alkali 91 impregnation of LC biomass at high solids loadings, while dry-chemo-mechanical 92 pretreatment describes a process in which dry chemical pretreatment is performed 93 simultaneously with mechanical particle size reduction (Barakat et al., 2014). 94 Advantageously, these pretreatments reduce the use of chemicals and energy demand, 95 enhance polysaccharide saccharification when using enzyme cocktails (Barakat et al., 2014) 96 and, thanks to high solids loadings, permit process intensification and reactor downsizing.

97
To investigate whether DC and DCM pretreatments can be beneficial for the carboxylate 98 platform, we have investigated the use of these technologies in combination with anaerobic 99 conversion of wheat straw using a lignocellulolytic cow rumen-derived microbial consortium 100 (RWS). The effect of the pretreatments on the subsequent microbial activity has been 101 investigated, monitoring the kinetics of LC biomass conversion, carboxylate production and 102 the dynamics of key enzymatic activities. Our findings reveal that DC pretreatment increased 103 the initial VFA production rate and that this increase was most pronounced when DCM 104 pretreatment was employed. This increase was accompanied by an increase in early phase 105 xylanase activity, but CMCase activity was unchanged. Finally, our results indicate a 106 correlation between lignocellulose macroporosity and its degradability by a hydrolytic 107 microbial consortium.  Boissy-le-Repos, France) was harvested (in August 2011), milled to 2 mm using a knife mill Comment citer ce document : Lazuka, A., Roland, C., Barakat Sodium hydroxide (NaOH) was dissolved in distilled water (5g in 20 mL). Wheat straw at 2 116 mm (100g) was impregnated during 5h at ambient conditions (25 °C) with this alkaline 117 solution using a pulverizing system (5 g of NaOH per 100 g of wheat straw) according to the 118 procedure described previously (Barakat et al., 20014). The chemically treated wheat straw-119 2mm was dried at 105 °C (12h) resulting in a final moisture content of 8-10% (w/w) and 120 designated biomass B. Biomass A and B were comminuted using an impact mill operating at ambient temperature 123 and 18,000 rpm (Hosokawa-alpine, type UPZ, Augsburg, Germany). Fine particulate fractions 124 were collected using a 100 µm mesh (the material was milled until it passed through the grid) 125 and designated as biomass C and D, respectively.   Instruments, Orsay, France).

146
The energy consumed during milling was determined in triplicates using a watt meter 147 following a previously described procedure (Barakat el al., 2014). where D (mm) is the pore diameter, S the solute (water) surface tension, ∆P the suction 158 pressure (MPa) used to measure the ratio water held:pore volume.

159
Water absorption at different suction pressures (used to explore different pore diameters) was

197
The composition of the residual substrate in the reactor was characterized as described above 198 for substrate characterization (section 2.4.1).

199
VFA production was monitored using a Varian 3900 gas chromatograph as described by    Results and discussion 224 The impact of dry pretreatment on the kinetics of wheat straw degradation and carboxylate 225 production by a microbial consortium RWS was assessed using four different pretreated 226 wheat straws. Samples A and C were milled to 2 mm, B and D to 100 µm, and C and D were 227 submitted to 5% (w/w) NaOH impregnation at high LC solids. These conditions facilitated the 228 assessment of a possible synergy between dry milling and chemical pretreatment. Moreover, 229 it is important to underline that D was first chemically pretreated and then milled, because this 230 sequence is expected to reduce the energy demand associated with milling and increase  The energy consumption associated with the production of A was 223.3 kJ.kg -1 , while the 234 production of B required 2.8 times more energy (Table1). The energy demand associated with 235 the production of C was the same as that for A, which is logical since chemical impregnation 236 was carried out after milling. However, energy consumption associated with the production of

266
The comparison of the macroporosity of samples A to D (Fig. 1B) revealed a negative 267 correlation between macroporosity with fine milling (100 µm) in the absence of chemical 268 pretreatment. For sample B, macroporosity was 1.5 ± 0.3 g.g -1 , which is lower than that 269 measured for A, 2.9 ± 0.4 g.g -1 . On the other hand, soda pretreatment had a positive effect on 270 macroporosity, reaching 4.4 ± 0.9 g.g -1 and 4.6 ± 0.6 g.g -1 for samples C and D, respectively.

271
These results suggest that milling to 100µm (B) actually reduced accessibility to bacteria,

298
In conclusion, assuming that cellulose crystallinity, the aptitude towards enzymatic 299 saccharification and macroporosity are reliable indicators of pretreated substrate accessibility 300 for microbial bioconversion, it appears reasonable to suggest that samples C and D should be 301 readily amenable to bioconversion by suitable microbial consortia. In this regard, considering 302 samples A and B, which had not been submitted to soda treatment, the only clear difference 303 concerned their macroporosity (lower for B). This makes prediction more difficult, but 304 suggests that B (100 µm particle size) might be less amenable to microbial bioconversion.  Therefore, subsequently wheat straw degradation was expressed more simply as the 313 percentage of holocellulose-related carbon (in moles) removal (expressed as percentage of the 314 initial holocellulose-related carbon content, % iCmol Holo ). 315 In this respect, the results obtained for holocellulose removal (% iCmolHolo) on samples A, C was recorded for B (0.31 ± 0.02 Cmol VFA .iCmol -1 Holo ) (Fig. 2B). Accordingly, the 325 holocellulose degradation and VFA production rates ( Fig. 2C and D)  The composition of the carboxylates produced at the end of the experiment (Table 2) although VFA production levels were much higher, probably because of the strict anaerobic 385 conditions that were applied (static or anoxic conditions were used in previous studies). 386 Indeed, strict anaerobic conditions prevent microbial VFA consumption. Dry milling combined with NaOH pretreatment enhanced wheat straw enzymatic hydrolysis 438 and bioconversion using a microbial consortium RWS, leading to increased xylanase activity 439 and VFA production rate. Compared to raw wheat straw, the optimal pretreatment was dry 440 milling to 100 µm, combined to alkaline impregnation, which procured a greater than two-441 fold increase in VFA production rate. Acetic, propionic and butyric acids were the main VFA 442 produced by RWS, irrespective of the pretreatment method. An increase in butyric acid 443 production was observed with chemically-pretreated substrates. Macroporosity appeared as 444 the parameter that best predicts the biological acidogenic potential of RWS.         Table 1: Size and energy consumption of the four types of wheat straw pretreatment.