Steering the conversion of protein residues to volatile fatty acids by adjusting pH

Steering the conversion of protein residues to volatile fatty acids by adjusting pH R. Bevilacqua, A. Regueira, M. Mauricio-Iglesias, J.M. Lema, M. Carballa Accepted Manuscript How to cite: Bioresource Technology, 320 (2021), 124315 https://doi.org/10.1016/j.biortech.2020.124315 Copyright information: © 2020 Elsevier Ltd. This manuscript version is made available under the CC-BY-NC-ND 4.0 license (http://creativecommons.org/licenses/by-nc-nd/4.0/) Highlights to: Steering the conversion of protein residues to volatile fatty acids by adjusting pH Protein acidification is favoured at neutral pH conditions The production of longer chain carboxylic acids is promoted at low pH Preferential amino acid uptake shapes protein conversion and product selectivity Amino acid consumption depends on pH and protein composition 1 Steering the conversion of protein residues to volatile fatty 1 acids by adjusting pH 2 Bevilacqua, R., Regueira A., Mauricio-Iglesias, M.*, Lema, J.M., Carballa, M. 3 CRETUS Institute, Department of Chemical Engineering, Universidade de Santiago de 4 Compostela, 15782 Santiago de Compostela, Spain – [email protected] 5 6 KEYWORDS: amino acid fermentation; mixed culture; biorefinery; selectivity; 7 volatile fatty acids 8 ABSTRACT 9 This study investigates the influence of pH on protein conversion into volatile fatty acids by 10 anaerobic mixed-culture fermentation, a topic that, in contrast to glucose fermentation, only had 11 scarce and contradictory information available. Several experiments were performed with two 12 model proteins (casein and gelatin) at three different pH values (5, 7 and 9) using chemostats and 13 batch tests. Highest conversion was reached at neutral pH although complete acidification was 14 never achieved. Longer chain carboxylates production was favoured at low pH, while acetic acid 15 was the main product at pH 7 and 9. Amino acids preferential consumption also varied with pH 16 and protein composition. In fact, protein conversion stoichiometry is mainly driven by energetic 17 yields and amino acid molecular configuration. Overall, this study identifies pH adjustment as a 18 way to steer volatile fatty acid production during mixed-culture fermentation of proteins. 19 20 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 2 1 INTRODUCTION 22 Mixed-culture fermentation (MCF) processes are an emerging biotechnology in the resource 23 recovery framework, being part of the so-called “carboxylate platform” (Agler et al., 2011). In fact, 24 many industrial wastes and organic sidestreams, such as cannery wastewaters or the organic 25 fraction of municipal solid wastes, are suitable candidates for the production of volatile fatty acids 26 (VFAs)(Garcia-Aguirre et al., 2017; Dahiya et al., 2018). To properly design and/or optimise this 27 kind of processes, it is paramount to know how the substrate composition and the operational 28 conditions affect their outcome in terms of both conversion efficiency and product selectivity. 29 MCF of glucose has been quite extensively studied and the related metabolic pathways are well 30 known (Zoetemeyer et al., 1982a; Zoetemeyer et al., 1982b; Ren et al., 1997; Horiuchi et al., 2002; 31 Temudo et al., 2007; González-Cabaleiro et al., 2015; Mohd-Zaki et al., 2016; Regueira et al., 32 2018). Glucose is either used in biomass growth as a carbon source (anabolism) or transformed 33 into pyruvate via glycolysis (catabolism). Then, pyruvate is processed to the different potential 34 fermentation products (e.g. butyric acid and ethanol) depending on the operational parameters. In 35 particular, pH has been identified as one of the most influential variables (Zoetemeyer et al., 36 1982b; Horiuchi et al., 2002; Temudo et al., 2007; Mohd-Zaki et al., 2016). Its effect is particularly 37 evident in terms of selectivity, with a shift from the production of butyrate, acetate and gaseous 38 hydrogen at low pH range (4-6.5) to acetate, ethanol and formate generation at higher values (6.5- 39 8.5), maintaining always a complete substrate consumption. Coherently with what was observed 40 with glucose, carbohydrate-rich streams and sugars (e.g. lactose of cheese whey) appear to be 41 generally fermented to a wider variety of VFAs when the MCF is performed in acid conditions 42 (Garcia-Aguirre et al., 2017; Asunis et al., 2019; Lu et al., 2020). 43 In contrast, information about protein MCF is more limited. The proteic substrate conversion to 44 VFAs appears to be often incomplete (Breure & van Andel, 1984; Ramsay & Pullammanappallil, 45 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 3 2001; Yu & Fang, 2003; Bevilacqua et al., 2020), with values ranging between 30 and 60% of the 46 total chemical oxygen demand (COD) of the substrate. A fixed stoichiometry to describe and 47 predict the conversion of amino acids (AAs) to VFAs was proposed by Ramsay and 48 Pullammanappalil (2001) and then revised and improved by Regueira et al. (2020), highlighting 49 Stickland reactions as the main degradation pathway. As these reactions involve the redox 50 coupling between two or more AAs (Nagase & Matsuo, 1982), protein composition plays an 51 important role in determining both the selectivity and the conversion efficiency of the MCF. In 52 fact, the preferential consumption of AA depends on their interactions (Bevilacqua et al., 2020), 53 which, in turns, could be potentially affected by pH. However, the information on the subject is 54 contradictory. While the degradation model of Ramsay and Pullammanappallil (2001) discards pH 55 influence on the process, experimental results demonstrated its relevance (Breure & van Andel, 56 1984; Breure et al., 1986; Yu & Fang, 2003; Duong et al., 2019). Acidification degree appears to be 57 favoured at neutral conditions, whereas either propionic (Breure & van Andel, 1984) or n-valeric 58 acid (Duong et al., 2019) productions are favoured over acetic acid generation at acid pH values. 59 No data on alkaline fermentation of proteins is available. 60 Therefore, the main aim of this study is to understand how protein degradation is affected by pH 61 changes during anaerobic MCF, focusing on its impact on acidification degree and VFA selectivity. 62 As a novel approach to the subject, AA analysis will help to understand how the conversion 63 mechanisms vary according to pH. Given the lack of information and the available inconclusive 64 results, determining the link between AA consumption and VFA production at different pH 65 conditions will broaden the available knowledge on protein MCF as well as denoting how the 66 process can be steered towards desired product compositions. 67 2 MATERIALS AND METHODS 68 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 4 2.1 Proteic substrates characteristics 69 The selected model proteins were synthetic hydrolysates compounds (peptone from casein 70 (A2208,0500 PanReac) or peptone from gelatin (70951-1KG-F Sigma-Aldrich)), used as sole carbon 71 source and supplemented with macro nutrients. The composition of the feedstock solution was as 72 follows (g/L): hydrolysed casein or gelatin 7.500-7.600; NaCl 0.292; KH2PO4 0.780; NH4Cl 0.530, 73 Na2SO4 0.057; MgCl2·6H2O 0.120. The feedstocks were maintained refrigerated throughout the 74 experiment (4°C). The amino acid composition of the chosen proteins was analysed in a previous 75 study (Bevilacqua et al., 2020). 76 2.2 Continuous reactors operation 77 Two CSTRs of 1 L of working volume were used. They were initially inoculated (in-reactor biomass 78 concentration of around 1.0 g VSS/L) with acidogenic-only biomass obtained from the operation 79 of a parent reactor (Bevilacqua et al., 2020), and were continuously operated at an hydraulic 80 retention time of 1.5 d and an organic loading rate of 5.3 g COD/L·d. 81 The tested pH values, 5, 7 (Bevilacqua et al., 2020) and 9, were selected to be representative of 82 each pH condition. The 2-points steps from pH 7 were chosen in order to ensure different 83 environmental conditions for the microbial populations while avoiding extreme pressure of more 84 marked acid and alkaline environments. Initially, the reactors were operated at neutral pH values 85 (7.2 - 7.4) without the need for chemical control for 140 days (casein reactor) and 170 days 86 (gelatin reactor). Afterwards, the pH was shifted to the alkaline range (i.e. 9), and the reactors 87 were operated at this pH for 160 days each. Before applying the acid conditions, both reactors 88 were acclimated again at pH 7 for at least 30 days, to avoid environmental shocks which could 89 negatively affect the microbiome. Finally, the reactors were operated at pH 5 for 320 and 110 days 90 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 5 respectively. Each pH condition was maintained for at least 100 days to ensure that steady-state 91 operation was achieved. 92 A multiparametric analyser (CHEMITEC, Italy) was used to monitor and continuously control pH, 93 allowing to maintain it at the chosen setpoint values by addition of either 3M HCl or 3M NaOH. 94 Chemical Oxygen Demand (COD, total and soluble) and VFA concentrations were determined 95 three times per week, while Total Ammonia Nitrogen (TAN) and solids concentrations were 96 measured once a week. AA analysis was performed on selected number of samples from steady 97 state periods of operation. 98 2.3 Batch experiments 99 When steady state was achieved at each pH value, batch fermentation tests were performed in 100 order to verify whether the absence of kinetic limitations might affect the incomplete substrate 101 acidification and/or the selectivity of the process. 102 In each assay, bottles of 0.5 L total volume (0.375 L of working volume) with rubber stoppers were 103 used. The chosen operational conditions were comparable to the ones applied to the continuous 104 reactors (25°C, N2 sparging, pH value). The vessels were inoculated with biomass from the 105 continuous reactor: effluent samples were centrifuged to separate the VFA-rich supernatants, and 106 the obtained biomass pellets were resuspended in the test bottles to achieve an initial biomass 107 concentration of 0.5 g VSS/L. The proteic substrates were added according to the selected 108 substrate-to-inoculum ratio, 10 g COD substrate/g VSS (Bevilacqua et al., 2020), while 109 macronutrients were supplemented with the same concentrations as in the reactors’ feedstock. 110 10 mL samples were taken at increasing time intervals (initially 2-3 hours). Half of the sample 111 volume was centrifuged and filtered for TAN and VFA determination, with the surplus being frozen 112 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 6 for AAs analysis. The remaining 5 mL were used for optical density determination (Bevilacqua et 113 al., 2020). 114 2.4 Analytical methods 115 Conventional physicochemical parameters were determined according to Standard Methods 116 (APHA, 2017). Samples of the reactor’s effluent were used for total (TS and VS), suspended (TSS 117 and VSS) solids (SM2540B, D and E) and total COD (modified SM5220C) content analysis. Filtered 118 (0.45 µm) effluent samples were used to measure the soluble chemical oxygen demand 119 (SM5220C) and TAN (SM4500-NH3.F). A Shimazdu UV-1800 was used for all the 120 spectrophotometric measurements. 121 VFAs from acetic to heptanoic acid were measured through gas chromatography (AGV-DB1 122 method), though n-caproic acid and heptanoic acid were never detected. The equipment used was 123 an Agilent 6850 with a flame ionisation detector. The column used was a DB-Wax (Agilent 124 Technologies, 30 m x 0.250 mm x 0.25 µm). The injector temperature was at 200°C while the 125 detector was set at 300°C. The carrier gas was nitrogen. The samples were filtered (0.45 µm) and 126 then acidified with 10 µM of concentrated H3PO4 (85%) before being analysed. 127 Secondary metabolites (formic acid, lactic acid and ethanol) were determined through high 128 performance liquid chromatography (HPLC) according to the GLEFG1 method with a HP 1100 (IR 129 HP1047A detector). The AMINEX HPX-87H (300 x 7.8 mm) column used H2SO4 (5 mM) as an 130 isocratic eluent. The column temperature was 30°C while for the detector was 35°C. The samples 131 were prepared in the same way as for VFA determination. 132 Gas composition was determined by gas chromatography with a HP 5890 Series II. Gas syringes of 133 1 mL were used to extract the gaseous samples through silicon rubber septa attached to the 134 outflow gas tubes of the reactors and through the bottle rubber stoppers. 135 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 7 For total AA determination, samples underwent acid hydrolysis for 24 hours at 110°C using HCl 6 136 N. Then, AccQ-Tag method was used to convert them to stable fluorescent derivatives (Cohen et 137 al., 1993) which were finally analysed through HPLC with a Waters 2695 equipped with a 138 fluorescence detector (Waters 2475). 139 Optical density determination was performed through spectrophotometric readings at 600 nm 140 and calibrated with actual VSS measurements. 141 2.5 Calculations 142 Acidification degree was the parameter chosen to describe substrate conversion (in COD basis), 143 based on the concentration of measured VFA (in this case aliphatic VFA) and expressed as: 144 𝐴𝑐𝑖𝑑𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑔𝑟𝑒𝑒 (%)= Ʃ C𝑉𝐹𝐴 𝐶𝑝𝑟 × 100 (1) where CVFA stands for the total concentration of the measured VFAs (in g COD-VFA/L) in the 145 reactor effluent and Cpr for the total protein concentration (in g COD/L) in the feeding of the 146 reactor. 147 Ammonification was also used as a proxy to monitor protein conversion to VFA, as amino acid 148 fermentation is always related to NH4+ release. It was expressed as follows: 149 𝐴𝑚𝑚𝑜𝑛𝑖𝑓𝑖𝑐𝑎𝑡𝑖𝑜𝑛 (%)= CTANeffluent − CTANfeeding CTANmaximum × 100 (2) Where CTANeffluent is the concentration of ammonium nitrogen (mg N-NH4+/L) measured in the 150 reactor effluent and CTANfeeding is the concentration of ammonium nitrogen (mg N-NH4+/L) in the 151 reactor feeding derived from the macronutrients supplementation and CTANmaximum is the maximum 152 concentration of ammonium nitrogen (mg N-NH4+/L) achieved if complete degradation of proteins 153 occurs. CTANmaximum was estimated based on Total Kjeldahl Nitrogen measurements (SM4500C) of 154 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 14 aromatic acid production, which show even higher toxicity than aliphatic VFA for anaerobic 293 digestion microorganisms (Sabra et al., 2015). 294 - Figure 4 goes here - 295 Figure 4. Amino acids consumption in casein (a) and gelatin (b) reactors at different pH values. (■ 296 pH 5; ■ pH 7 (Bevilacqua et al., 2020); ■ pH 9). Arg: Arginine; Ala: Alanine; Asp: Aspartic acid; Glu: 297 Glutamic acid; Gly: Glycine; His: Histidine; Ile: Isoleucine; Leu: Leucine; Lys: Lysine; Met: 298 Methionine; Phe: Phenylalanine; Pro: Proline; Ser: Serine; Thr: Threonine; Trp: Tryptophan; Tyr: 299 Tyrosine; Val: Valine. 300 3.4 Influence of pH on Stickland reaction stoichiometry 301 Despite Stickland reactions being widely accepted as the main mechanism of AA conversion into 302 VFA, it was already demonstrated that their fixed stoichiometry does not completely reproduce 303 the outcome of proteins fermentation at neutral pH (Bevilacqua et al., 2020). Moreover, there are 304 some evidences suggesting that pH might affect this mechanism (Breure & van Andel, 1984; Yu & 305 Fang, 2003). Therefore, this section analyses the influence of pH on the Stickland stoichiometry by 306 establishing molar balances between influent and effluent concentrations of amino acids and VFA. 307 Acetic acid was not considered as several AAs are responsible for its production. 308 Compatibly with what was observed for neutral conditions (Bevilacqua et al., 2020), iso-butyric 309 acid production followed the Stickland stoichiometry at pH 5 and 9, being its strict connection with 310 valine consumption confirmed. In contrast, n-butyric acid production did not correlate with its 311 parent amino acids (Glu, His, Thr, Lys) at pH 5 for both proteins and at pH 9 for casein (Fig. 5). This 312 imbalance between substrate and products at low pH might be either linked to other amino acids 313 being converted to n-butyric acid through unconsidered metabolic pathway and/or to chain 314 elongation between previously produced acetic acid and endogenous acetyl-CoA (Duncan et al., 315 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 15 2004). Conversely, alkaline conditions hindered n-butyric acid production during casein 316 fermentation without affecting the parent AA consumption, meaning that alternative carboxylic 317 acids might have been generated instead. Overall, n-butyric stoichiometry is not reliable outside 318 the neutral pH range, as it underestimates its production at acid conditions and overestimates it at 319 alkaline conditions. 320 - Figure 5 goes here - 321 Figure 5. n-Butyric acid balance in the continuous casein (a) and gelatin (b) reactors. ■ Glutamic 322 acid; ■ Histidine; ■ Threonine; ■ Lysine; ■ n-Butyric acid 323 Contrary to the results of pH 7, iso-valeric acid production fits Stickland stoichiometry at both pH 5 324 and 9 for both proteins, even when considering it together with iso-caproic acid production. 325 However, propionic and n-valeric acid balances (jointly evaluated since both are produced at equal 326 molar ratios from Arg and Pro, albeit propionic acid is also generated from methionine) differ 327 depending on protein composition (Fig. 6), although propionic acid was generally produced at 328 higher concentrations than n-valeric acid in both cases. For casein fermentation, the results 329 obtained at pH 5 and 9 are comparable with what was previously observed in the neutral range 330 (Bevilacqua et al., 2020), i.e. propionic acid production alone justifies amino acid consumption, 331 suggesting that the production of n-valeric acid was potentially related to other amino acids 332 degradation and/or to chain elongation processes (Fig. 6a). Moreover, propionic acid production 333 progressively increased with pH, while n-valeric acid production remained constant. Conversely, 334 gelatin fermentation at both acid and alkaline conditions (Fig. 6b) yielded less propionic acid than 335 at neutral pH, and n-valeric acid production was either strongly inhibited or null, suggesting that 336 the consumed AAs were either preferentially converted to propionic acid rather than to n-valeric 337 acid or to different VFAs. As previously mentioned (section 3.2), Arg can in fact be degraded 338 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 16 through an alternative pathway, which generates acetyl-CoA and Ala, and ultimately acetic acid 339 (Fonknechten et al., 2010), resulting in a greater ATP yield than the one assumed by Ramsay and 340 Pullammanappalil (2001), involving 5-aminovalerate and proline as intermediate compounds 341 instead (Barker et al., 1987). 342 - Figure 6 goes here - 343 Figure 6. Propionic and n-Valeric acid balance in the continuous casein (a) and gelatin (b) reactors. 344 ■Arginine; ■ Proline; ■ Methionine; ■ Propionic acid; ■ n-Valeric acid 345 All these results demonstrated how the metabolic pathways of AAs fermentation into VFAs vary 346 with protein composition and pH, with the observed deviations from the previously accepted 347 stoichiometry being probably driven by bioenergetics constraints. A change in pH has, in fact, the 348 potential of modifying the energetics (i.e. the ATP generated) of some amino acid pathways as 349 described by Regueira et al. (2020), in turn affecting the interaction between them. Additionally, 350 shifting pH might limit AAs consumption by either affecting the transportation mechanisms (Chen 351 & Chung, 2015) and/or directly altering AA bioavailability through changes in molecular charge 352 and subsequent precipitation (Tseng et al., 2009). 353 5. CONCLUSIONS 354 This work describes the effect of pH on amino acid conversion mechanisms during protein mixed- 355 culture fermentation. It is demonstrated that pH affects the stoichiometry of AAs conversion into 356 VFAs, by altering AA interactions and/or molecular configuration. Neutral pH maximises AA 357 degradation, while a more diverse product spectrum is obtained at acid conditions, regardless of 358 protein composition. Yet, those trends are more evident in casein fermentation. Therefore, the 359 outcome of protein-rich substrates fermentation depends mainly on protein composition and pH. 360 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 17 E-supplementary data for this work can be found in e-version of this paper online. 361 ACKNOWLEDGEMENTS 362 This research is funded by the Spanish Government through BIOCHEM project (PCIN-2016-102, 363 ERA-IB-2 7th call, ERA-IB-16-052) and by the Spanish Ministry of Education (FPU14/05457). 364 Authors belong to the CRETUS Strategic Partnership (ED431E 2018/01) and to the Galician 365 Competitive Research Group (ED431C2017/029). All these programs are co-funded by ERDF (EU). 366 REFERENCES 367 1) Agler, M.T., Wrenn, B.A., Zinder, S.H., Angenent, L.T., 2011. Waste to bioproduct 368 conversion with undefined mixed cultures: the carboxylate platform. Trends in 369 Biotechnology, 29:2, 70-78. https://doi.org/10.1016/j.tibtech.2010.11.006 370 2) APHA, 2017. Standard Methods for the Examination of Water and Wastewater, 371 America Public Health Association, Washington, DC, USA. 372 3) Asunis, F., De Gioannis, G., Isipato, M., Muntoni, A., Polettini, A., Pomi, R., Rossi, A., 373 Spiga, D., 2019. Control of fermentation duration and pH to orient biochemicals and 374 biofuels production from cheese whey. Bioresource Technology, 289, 121722. 375 https://doi.org/10.1016/j.biortech.2019.121722 376 4) Barker, H.A., D’Ari, L., Kahn, J., 1987. Enzymatic reactions in the degradation of 5- 377 aminovalerate by Clostridium aminovalericum, Journal of Biological Chemistry, 262, 378 8994-9003. 379 5) Bevilacqua, R., Regueira, A., Mauricio-Iglesias, M., Lema, J.M., Carballa, M., 2020. 380 Protein composition determines the preferential consumption of amino acids during 381 anaerobic mixed-culture fermentation, Water Research, 183, 115958. 382 https://doi.org/10.1016/j.watres.2020.115958 383 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 18 6) Breure, A.M., Mooijman, K.A., van Andel, J.G. 1986. Protein degradation in anaerobic 384 digestion: influence of volatile fatty acids and carbohydrates on hydrolysis and 385 acidogenic fermentation of gelatin. Appl. Microbiol. Biotechnol, 24, 426-431. 386 https://doi.org/10.1007/BF00294602 387 7) Breure, A.M., van Andel, J.G., 1984. Hydrolysis and acidogenic fermentation of a 388 protein, gelatin, an anaerobic continuous culture. Appl. Microbiol. Biotechnol, 20, 40 – 389 45. https://doi.org/10.1007/BF00254644 390 8) Britz, M.L., Wilkinson, R.G., 1981. Leucine dissimilation to isovaleric and isocaproic 391 acids by cell suspensions of amino acid fermenting anaerobes: the Stickland reaction 392 revisited. J. Microbiol., 28, 291-300. https://doi.org/10.1139/m82-043 393 9) Chen, R., Chung, S.H., 2015. Molecular dynamic simulations of Na (+) and Leucine 394 transport by LeuT. Biochem. Biophys. Res. Commun., 464, 281-285. 395 https://doi.org/10.1016/j.bbrc.2015.06.143 396 10) Cohen, S.A., De Antonis, K., Michaud, D.P., 1993. Compositional protein analysis using 397 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate, a novel derivatization reagent. 398 Techniques in Protein Chemistry IV, 289 – 298, (R.H. Angeletil, Ed.) Academic Press, San 399 Diego, CA. 400 11) Dahiya, S., Kumar, A.N., Chatterjee, J.S.S., Sarkar, O., Mohan, S.V., 2018. Food Waste 401 Biorefinery: Sustainable Strategy for Circular Bioeconomy, Bioresource Technology, 248, 402 2-12. http://dx.doi.org/10.1016/j.biortech.2017.07.176 403 12) Domingos, J.M.B., Martinez, G.A., Scoma, A., Fraraccio, S., Kerckhof, F.M., Boon, N., 404 Reis, M.A.M., Fava, F., Bertin, L., 2017. Effect of operational parameters in the 405 continuous anaerobic fermentation of cheese whey on titers, yields, productivities, and 406 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 19 microbial community structures. ACS Sustainable Chem. Eng., 5:2, 1400−1407. 407 https://doi.org/10.1021/acssuschemeng.6b01901 408 13) Duncan, S.H., Holtrop, G., Lobley, G.E., Calder, A.G., Stewart, C.S., Flint, H.J., 2004. 409 Contribution of acetate to butyrate formation by human faecal bacteria. British Journal 410 of Nutrition, 91:6, 915-923. https://doi.org/10.1079/BJN20041150 411 14) Duong, T.H., Grolle, K., Nga, T.T.V., Zeeman, G., Temmink, H., Van Eekert, M., 2019. 412 Protein hydrolysis and fermentation under methanogenic and acidifying conditions. 413 Biotechnol Biofuels, 12, 254-264. https://doi.org/10.1186/s13068-019-1592-7 414 15) Fonknechten, N., Chaussonnerie, S., Tricot, S., Lajus, A., Andreesen, J.R., Perchat, N., 415 Pelletier, E., Gouyvenoux, M., Barbe, V., Salanoubat, M., Le Paslier, D., Weissenbach, J., 416 Cohen, G.N., Kreimeyer, A., 2010. Clostridium sticklandii, a specialist in amino acid 417 degradation: revisiting its metabolism through its genome sequence. BMC Genomics, 418 11:555. https://doi.org/10.1186/1471-2164-11-555 419 16) Garcia-Aguirre, J., Aymerich, E., González-Mtnez de Goñi, J., Esteban-Gutiérrez, M., 420 2017. Selective VFA production potential from organic waste streams: Assessing 421 temperature and pH influence. Bioresource Technology, 244, 1081-1088. 422 https://doi.org/10.1016/j.biortech.2017.07.187 423 17) González-Cabaleiro, R., Lema, J. M., Rodríguez, J., 2015. Metabolic Energy-Based 424 Modelling Explains Product Yielding in Anaerobic Mixed Culture Fermentations. PLoS 425 ONE, 10:5, e0126739. https://doi.org/10.1371/journal.pone.0126739 426 18) Horiuchi, J.I., Shimizu, T., Tada, K., Kanno, T., Kobayashi, M. 2002. Selective production 427 of organic acids in anaerobic acid reactor by pH control. Bioresource Technology, 82:3, 428 209–213. https://doi.org/10.1016/S0960-8524(01)00195-X 429 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 20 19) Lu, Y., Zhang, Q., Wang, X., Zhou, X., Zhu, J., 2020. Effect of pH on volatile fatty acid 430 production from anaerobic digestion of potato peel waste. Bioresource Technology, 431 316, 123851. https://doi.org/10.1016/j.biortech.2020.123851 432 20) Mohd-Zaki, Z., Bastidas-Oyanedel, J.R., Lu, Y., Hoelzle, R., Pratt, S., Slater, F.R., 433 Batstone, D.J., 2016. Influence of pH Regulation Mode in Glucose Fermentation on 434 Product Selection and Process Stability. Microorganisms, 4, 2. 435 https://doi.org/10.3390/microorganisms4010002 436 21) Nagase, M., Matsuo, T., 1982. Interactions between amino‐acid‐degrading bacteria and 437 methanogenic bacteria in anaerobic digestion. Biotechnology and Bioengineering, 438 24:10, 2227-2239. https://doi.org/10.1002/bit.260241009 439 22) Ramsay, I.R., Pullammanappallil, P.C., 2001. Protein degradation during anaerobic 440 wastewater treatment: derivation of stoichiometry. Biodegradation, 12, 247 – 257. 441 https://doi.org/10.1023/a:1013116728817 442 23) Regueira, A., González-Cabaleiro, R., Ofiţeru, I.D., Rodríguez, J., Lema, J. M., 2018. 443 Electron bifurcation mechanism and homoacetogenesis explain products yields in 444 mixed culture anaerobic fermentations. Water Research, 141, 349-356. 445 https://doi.org/10.1016/j.watres.2018.05.013 446 24) Regueira, A., Lema, J.M., Carballa, M., Mauricio-Iglesias, M., 2020. Metabolic modeling 447 for predicting VFA production from protein-rich substrates by mixed-culture 448 fermentation. Biotechnology and Bioengineering, 117, 73-84. 449 https://doi.org/10.1002/bit.27177 450 25) Ren, N., Wang, B., Huang, J.C., 1997. Ethanol-type fermentation from carbohydrate in 451 high rate acidogenic reactor. Biotechnology and Bioengineering, 54:5, 428-433. 452 https://doi.org/10.1002/(SICI)1097-0290(19970605)54:5<428::AID-BIT3>3.0.CO;2-G 453 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 21 26) Sabra, W., Roske, I., Sahm, K., Antranikian, G., Zeng, A.P., 2015. High temperature 454 biogas reactors to treat stillage from an industrial bioethanol process: metabolic and 455 microbial characterization. Eng. Life Sci., 15:7, 743-750. 456 https://doi.org/10.1002/elsc.201500040 457 27) Shi, X., Lin, J., Zuo, J., Li, P., Li, X., Guo, X., 2016. Effects of free ammonia on volatile 458 fatty acid accumulation and process performance in the anaerobic digestion of two 459 typical bio-wastes. Journal of Environmental Sciences, 55, 49-57. 460 https://doi.org/10.1016/j.jes.2016.07.006 461 28) Smith, E.A., MacFarlane, G.T., 1998. Enumeration of amino acid fermenting bacteria in 462 the human large intestine: effects of pH and starch on peptide metabolism and 463 dissimilation of amino acids. FEMS Microbiology Ecology, 25:4, 355-368. 464 https://doi.org/10.1016/S0168-6496(98)00004-X 465 29) Temudo, M.F., Kleerebezem, R., van Loosdrecht, M., 2007. Influence of the pH on 466 (open) mixed culture fermentation of glucose: A chemostat study. Biotechnol. Bioeng., 467 98, 69–79. https://doi.org/10.1002/bit.21412 468 30) Tseng, H.C., Lee, C.Y., Weng, W.L., Shiah, I.M., 2009. Solubilities of amino acids in water 469 at various pH values under 298.15 K. Fluid Phase Equilibria, 285, 90-95. 470 https://doi.org/10.1016/j.fluid.2009.07.017 471 31) Yu, H.Q., Fang, H.H.P., 2003. Acidogenesis of gelatin-rich wastewater in an upflow 472 anaerobic reactor: influence of pH and temperature, Water Research, 37, 55-66. 473 https://doi.org/10.1016/s0043-1354(02)00256-7 474 32) Zoetemeyer, R.J., Arnoldy, P., Cohen, A., Boelhouwer, C., 1982a. Influence of 475 temperature on the anaerobic acidification of glucose in a mixed culture forming part 476 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 22 of a two-stage digestion process. Water Research, 16:3, 313-321. 477 https://doi.org/10.1016/0043-1354(82)90191-9 478 33) Zoetemeyer, R.J., van den Heuvel, J.C., Cohen, A., 1982b. pH influence on acidogenic 479 dissimilation of glucose in an anaerobic digestor. Water Research, 16:3, 303-311. 480 https://doi.org/10.1016/0043-1354(82)90190-7 481 482 483 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 Figure 1