A novel strategy for triacylglycerides and polyhydroxyalkanoates production using waste lipids

A novel strategy for triacylglycerides and polyhydroxyalkanoates production using waste lipids Lucía Argiz, Rebeca González-Cabaleiro, Ángeles Val del Río, Jesús González-López, Anuska Mosquera-Corrala Accepted Manuscript How to cite: Science of The Total Environment, 763 (2021), 142944 https://doi.org/10.1016/j.scitotenv.2020.142944 Copyright information: © 2020 Elsevier B.V. This manuscript version is made available under the CC-BY-NC-ND 4.0 license (http://creativecommons.org/licenses/by-nc-nd/4.0/) A novel strategy for triacylglycerides and polyhydroxyalkanoates production using waste lipids Lucía Argiz a *, Rebeca González-Cabaleiro b, Ángeles Val del Río a, Jesús González-López c, Anuska Mosquera-Corral a a CRETUS Institute, Department of Chemical Engineering, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Galicia, Spain b Department of Infrastructure and Environment, University of Glasgow, Rankine Building, Glasgow, G12 8LT, UK c Department of Microbioloy and Institute of Water Research, Universidad de Granada, Granada, Spain * Corresponding author: luciaargiz.mon[email protected] Highlights - Preferent triacylglycerides or polyhydroxyalkanoates storage from waste lipids - No waste oil pretreatment was needed before the enrichment and accumulation stages. - Dominant metabolic pathways depended on the imposed selective pressures. - Carbon and nitrogen (limited) uncoupling promoted polyhydroxyalkanoates storage. - Low pH in the famine, an additional advantage for triacylglycerideaccumulators. Enrichment SBR Accumulation FBR Substrate hydrolysis Culture selection +Substrate hydrolysis TAG / PHA storage + Uncoupled C and N feedings Uncoupled C and N feedings Low pH in the famine phase 43 wt % TAGs (TAG:PHA 13:87) 0.67 Cmmol TAG / Cmmol oil Effect of the selection strategy 82 wt % PHAs (TAG:PHA 4:96) 0.80 Cmmol PHA / Cmmol oil Fish-canning industry effluent Canned tuna cooking wastewater Enriched biomass Waste fish oil Primary treatment VALORIZATION PROCESS Fed batch reactor (FBR); polyhydroxyalkanoates (PHAs); sequential batch reactor (SBR); triacylglycerides (TAGs) Clarified effluent 1 A novel strategy for triacylglycerides and polyhydroxyalkanoates 1 production using waste lipids 2 Lucía Argiz a *, Rebeca González-Cabaleiro b, Ángeles Val del Río a, Jesús González-López c, 3 Anuska Mosquera-Corral a 4 a CRETUS Institute, Department of Chemical Engineering, Universidade de Santiago de 5 Compostela, 15782 Santiago de Compostela, Galicia, Spain 6 b Department of Infrastructure and Environment, University of Glasgow, Rankine Building, 7 Glasgow, G12 8LT, UK 8 c Department of Microbioloy and Institute of Water Research, Universidad de Granada, 9 Granada, Spain 10 11 * Corresponding author: luciaargiz.mont[email protected] 12 13 ABSTRACT 14 Lipids are one of the main components of the organic matter present in the effluents of the food15 processing industry. These waste streams can be biotransformed into valuable triacylglycerides (TAGs) 16 and polyhydroxyalkanoates (PHAs), precursors of biofuels and biomaterials alternative to petroleum17 based products. These compounds are yielded by mixed microbial cultures, and considering that both 18 TAG and PHA accumulators may coexist within the community, it seems crucial to define those 19 operational strategies that might control the selection of the dominant metabolic pathways (TAG or PHA 20 accumulation). In this work, residual fish-canning oil wasused as a carbon source in a two-stage process 21 (culture selection and intracellular compounds accumulation) in which the substrate was simultaneously 22 hydrolyzed in these two stages without the need for a previous fermentation unit. It was pretended to 23 maximise preferential TAG or PHA storage in the accumulation reactor by the imposition of certain 24 selective pressures in the enrichment one. Uncoupling C and N feedings and limiting nitrogen availability 25 in the medium, allowed to maximise PHA production (82.3 wt % of PHAs, 0.80 CmmolPHA/CmmolS). 26 Besides, when low pH in the famine phase was considered as additional selective pressure, it was possible 27 to shift the ratio TAG:PHA from 4:96 obtaining 43.0 wt % of TAGs (0.67 CmmolTAG/CmmolS). 28 Therefore, this novel and simplified process demonstrated versatility and efficiency in the storage of 29 TAGs and PHAs from a unique residual feedstock and using an open culture proving that product 30 selection can be harnessed if choosing the right operational conditions in the enrichment stage. 31 32 Keywords: biotransformation; lipidic waste; mixed microbial culture; pH; uncoupled feeding. 33 34 2 ADF: aerobic dynamic feeding 35 C: carbon 36 COD: chemical oxygen demand 37 DO: dissolved oxygen (mg/L) 38 FBR: fed-batch reactor 39 F/F: feast/famine 40 FFA: free fatty acids 41 MMC: mixed microbial culture 42 N: nitrogen 43 PHA: polyhydroxyalkanoate 44 qN: maximum specific nitrogen consumption rate (CmmolN/CmmolX·h) 45 qPHA: maximum specific polyhydroxyalkanoates production rate 46 (CmmolPHA/CmmolX·h) 47 qS: maximum specific carbon consumption rate (CmmolS/CmmolX·h) 48 qTAG: maximum specific triacylglycerides production rate (CmmolTAG/CmmolX·h) 49 qX: maximum specific active biomass production rate (CmmolX/CmmolS·h) 50 S: substrate 51 SBR: sequencing batch reactor 52 TAG: triacylglyceride 53 TN: total nitrogen (mg/L) 54 TS: total solids (g/g) 55 3 TSS: total suspended solids (g/L) 56 VS: volatile solids (g/g) 57 VSS: volatile suspended solids (mg/L) 58 WWTP: wastewater treatment plant 59 X: active biomass (Cmmol/L) 60 YPHA: maximum polyhydroxyalkanoates production yield (CmmolPHA/CmmolS) 61 YTAG: maximum triacylglycerides production yield (CmmolTAG/CmmolS) 62 YX: maximum active biomass production yield(CmmolX/CmmolS) 63 64 4 1. INTRODUCTION 65 The food industry generates large volumes of solid and liquid waste types that suppose 66 a significant environmental and economic concern. Among the industries causing a 67 global major impact are seafood and fish-processing activities (Kosseva, 2013). 68 Specifically, the fish-canning industry is characterized by the generation of large 69 amounts of high strength liquid effluents (Cristóvaõ et al., 2015), which are rich in 70 lipids, one of the main components of natural fish (Chipasa and Mȩdrzycka, 2006). 71 These effluents are usually pre-treated before biological treatment, to separate solids 72 and lipids using different physicochemical processes (Ahmad et al., 2020). Then, lipids 73 are commonly composted or anaerobically digested (Cristóvaõ et al., 2015; Tamis et al., 74 2015). However, with the advancement of science and technology, wastewater 75 treatment methods are constantly improving (Wu et al., 2020). In the last years, 76 industrial waste fish oils have been identified as potential feedstocks for the obtention of 77 value-added compounds such as renewable 2nd generation biofuels (Sawangkeaw and 78 Ngamprasertsith, 2013) or bioplastics (Surendran et al., 2020), which would contribute 79 to cope with fossil fuel depletion and environmental problems (Leng et al., 2020). 80 Using single-cell microorganisms, the valorisation of waste oils is feasible since they 81 are capable of storing important percentages of high-quality compounds that serve as 82 raw materials or intermediates for a variety of applications (Garay et al., 2014; Gujjala 83 et al., 2019). Special emphasis is given to triacylglycerides (TAGs) and 84 polyhydroxyalkanoates (PHAs), excellent reserve materials that constitute the main 85 compounds accumulated in eukaryotes and prokaryotes, respectively (Revellame et al., 86 2012). Their storage fulfils several roles in microorganisms such as cell growth and 87 division (Garay et al., 2014), stress response against the pressure exerted by certain 88 environments (Kumar et al., 2018), and energy reserve (Kumar et al., 2020). 89 5 Fungi and yeast are oleaginous microbes in their majority as they can accumulate over 90 20 wt % of intracellular TAGs (Alvarez and Steinbüchel, 2003). However, TAGs are 91 not a common storage compound in bacteria and their accumulation was only described 92 in a few prokaryotic species belonging to the actinobacterial group (Gujjala et al., 93 2019). Bacteria typically biotransform lipids into PHAs (Garay et al., 2014) although 94 certain species are able to synthesize PHA and TAG together, as it is the case of 95 Rhodococcus Ruber (Alvarez and Steinbüchel, 2003; Garay et al., 2014). 96 Although research on TAG and PHA production from lipidic feedstocks was 97 traditionally focused on the use of edible plant oils as substrates and pure cultures with 98 very high accumulation capacity (Basnett et al., 2018; Donot et al., 2014; Pérez-Arauz 99 et al., 2019), the high production costs, mainly associated with substrate acquisition and 100 culture conditions pushed for the development of cost-effective alternatives. These 101 involved the use of waste feedstocks and mixed microbial cultures (MMCs) (Sabapathy 102 et al., 2020) and not only reduced the costs of the process, but also increased resource 103 efficiency and improved waste management performance contributing to the 104 establishment of a circular economic model (Surendran et al., 2020; Yadav et al., 2020). 105 Recently, it was developed studies regarding the use of animal fats for TAG (Lopes et 106 al., 2018) and PHA (Sangkharak et al., 2020; Van Thuoc et al., 2019) production using 107 pure cultures. Regarding MMCs, to the best of the author’s knowledge, only the use of 108 vegetable oils was explored so far. TAG accumulation has been demonstrated feasible 109 through a two-stage process: (1) selection and enrichment of the MMC; and (2) 110 maximization of the TAG storage before its extraction and purification (Tamis et al., 111 2015). In the case of PHA production, it has been proposed an analogous configuration 112 that included a previous fermentation of the lipids contained in the residual streams 113 12 hydroxybutyrate and 12 % hydroxyvalerate (Sigma Aldrich, USA). The monomer 262 propyl esters were analysed by gas chromatography (HP innovax column equipped with 263 a FID, Agilent, USA). 264 2.4 Calculations and statistical analysis 265 The percentage of intracellular compounds accumulated (TAGs or PHAs) expressed in 266 dry weight (wt %), was calculated by dividing the measured mass of TAGs or PHAs by 267 the mass of VSS present in the medium (in grams). Active biomass (X) was estimated 268 as the difference between the measurements of VSS and storage compounds 269 accumulated (TAG + PHA). The elemental composition of the biomass was assumed to 270 be CH1.8O0.5N0.2. 271 Maximum specific conversion rates of carbon (– qS, CmmolS/(CmmolX·h)) and nitrogen 272 (– qN, CmmolN/(CmmolX·h)) consumption, and TAG (qTAG, CmmolTAG/(CmmolX·h)) 273 and PHA production (qPHA, CmmolPHA/(CmmolX·h)), were determined using the 274 maximum slopes of the experimental data divided by the active biomass. Similarly, the 275 specific biomass production rate was determined from the maximum slope of the 276 produced biomass divided by the amount of substrate fed (qX, CmmolX/(CmmolS·h)). 277 The production yield of the storage compounds (YTAG CmmolTAG/CmmolS; and YPHA, 278 CmmolPHA/CmmolS) and the biomass (YX, CmmolX/CmmolS) were determined by 279 dividing the related production rate (CmmolTAG/h, CmmolPHA/h, and CmmolX/h) by the 280 carbon source consumption rate (CmmolS/h). 281 Statistical analysis was performed by SPSS software, IBM Corp Released 2017 (IBM 282 SPSS Statistics for Windows, Version 25.0 Armonk, NY). One-way variance analysis 283 (ANOVA) was used to test the existence of statistically significant differences between 284 the composition of the intracellular compounds accumulated in the SBR during the 285 different operational periods (five independent groups). To determine which of these 286 13 groups differed from each other, the Turkey post hoc test was applied. On the other 287 hand, to compare the type of intracellular compounds accumulated in SBR and FBR 288 reactors during the same operational periods (comparison of two unrelated groups on 289 the same dependent variable), the independent-samples t-test was considered. A 5 % 290 significance level was assumed in both cases. 291 292 3. RESULTS AND DISCUSSION 293 3.1 Feasibility of using fish-canning oil to enrich a MMC in TAG and PHA 294 producers 295 During the first 203 days (Period I), the SBR was operated under a coupled feeding 296 strategy (simultaneous C and N addition) to enrich conventional activated sludge in 297 microorganisms with high storage ability. It is known that lipids, especially long-chain 298 fatty acids with saturated carbon chains, are less responsive to biodegradation than other 299 organic substrates (Chipasa and Mȩdrzycka, 2006). However, the typical F/F profile 300 was observed after 3 enrichment cycles (Figure S3.a), which indicated that the culture 301 was able to consume the substrate despite its high hydrophobicity and organic matter 302 composition (Table 1). 303 The operation of the reactor became stable after 50 days. During this period, the feast 304 phase length presented wide variability (from 1.3 to 5.5 hours), and the VSS 305 concentration sharply decreased due to culture selection (from 2.13 ± 0.10 g/L to 0.32 ± 306 0.04 g/L between days 1 and 44). From day 45 onwards, the length of the feast phase 307 followed a downward trend and reduced its variability. Nevertheless, even after more 308 than 100 days it still ranged between 2 to 4 hours (feast/cycle ratio of 0.21 ± 0.07). 309 According to Dionisi et al. (2005), ratios lower than approximately 0.25 allow for the 310 selection of microorganisms with a good storage ability. These variations in the feast 311 14 phase length could be due to the low solubility of the substrate and grease accumulation 312 in the walls of the reactor, which difficulted mass transfer and limited the microbial 313 uptake. 314 The pH was maintained close to neutrality (Figure S3.b) by NaHCO3 addition in the 315 dilution water. With the progress of the operation, NaHCO3 addition was reduced (from 316 1 to 0.64 g/L) until reaching stationary state. Occasionally, excess of NaHCO3 increased 317 the pH in the liquid bulk above neutral (e.g. on day 50 of operation, 0.7 g/L NaHCO3 318 were added in the feeding, and a pH of 7.9 was observed at the end of the cycle) (Figure 319 S3.c). In these conditions, the FFA produced in a first hydrolysation of the substrate by 320 lipase-catalysed reactions, saponified and became unavailable for microbial 321 metabolization. Once the pH was optimised, the possible saponification did not occur 322 anymore. Although after feeding, the pH slightly increased during the feast (e.g. from 323 7.0 to 7.4 on day 72 and from 7.0 to 7.3 on day 171), it decreased to the initial values 324 remaining almost constant during the whole cycle in period I (Figure 2.a, 2.b and 2.c). 325 In period I, the ratio C/N was maintained at 16.7 ± 2.4 g COD/g TN, and about 274 ± 28 326 g of TN were fed in each cycle. Neither the COD nor the TN were fully consumed 327 during the SBR cycle and average concentrations of 162 ± 40 mg COD/L (measured in 328 the centrifuged and non-filtered sample) and 81 ± 17 mg TN/L were detected in the 329 effluent (end of the cycle). All the consumed nitrogen was assumed to be used for 330 growth since nitrification was inhibited by allylthiourea addition. The presence of 331 remnant COD at the end of each cycle is explained by the nature of the substrate and the 332 presence of slowly biodegradable or recalcitrant compounds. The composition of the 333 substrate and in special its hydrophobicity (slow diffusion) could have limited microbial 334 growth (as no N limiting conditions were imposed) and explain the low VSS 335 concentrations obtained at steady-state, 0.58 ± 0.06 g/L (Figure S3.d). 336 15 During the feast phase of period I, microbial growth due to extracellular substrate 337 consumption, along with intracellular accumulation was observed (Figure 2.a, 2.b and 338 2.c), but the progressive enrichment of the culture enhanced the use of intracellular 339 substrate for growth during the famine phase. For example, on day 50, microorganisms 340 grew during the feast while 13.7 wt % TAGs and 0.9 wt % PHAs was accumulated and 341 no significant growth was observed during the famine phase (Table 2). However, once 342 these percentages increased on day 78 up to 20.6 wt % and 1.6 wt %, respectively, 343 growth took place during both phases (Table 2). 344 The percentage of intracellular compounds accumulated (as a sum of PHAs and TAGs) 345 did not significantly vary after day 78 onwards (period I), although changes in the 346 composition were observed, which suggested alterations in the cellular metabolism due 347 to a slight enrichment of the culture in PHA-accumulators (Silva et al., 2017). For 348 example, on day 171 the highest PHA production (5.3 wt %) was reached in period I but 349 concomitantly, TAGs accumulated decreased (10.8 wt %) in comparison with days 78, 350 111, and 118 (Figure 3.a). 351 Results obtained in period I showed the feasibility of the culture to metabolize the fish352 canning waste oil. However, the carbon source was mainly used for growth as a 353 consequence of the high N availability present in the medium. Once the substrate was 354 hydrolyzed, FFA weretransported into the cell and mainly degraded for biomass 355 production (respiration via the Krebs Cycle) (Figure 1). At the same time, part of the 356 FFA were directly accumulated as lipid reserve (TAGs with a composition proportional 357 to that of the substrate (Table 1, Table 2)) via the ex novo fatty acids pathway (Figure 358 1), a primary anabolic process occurring simultaneously with the production of lipid359 free biomass regardless of the C/N ratio in the medium (Carsanba et al., 2018; Lopes et 360 al., 2018; Pérez-Arauz et al., 2019;). Besides, results obtained in period I (Table 2) 361 16 evidenced that the presence of too high N concentrations blocked PHA biosynthesis, 3362 Hydroxyacyl-CoA instead of being polymerized, was further transformed into acetyl363 CoA and channelled into the Krebs cycle (Figure 1) (Tan et al., 2014). 364 Consequently, the enrichment stage of the process needed to be optimized to improve 365 culture selection and maximize TAG or PHA storage. 366 3.2 Optimization of the SBR enrichment strategy 367 In this section, shifts in the SBR enrichment strategy were explored to improve 368 selection and establish the operational conditions for the preferential development of 369 TAG or PHA-storing populations. Thus, although both storage compounds are 370 considered valuable potential sources (Garay et al., 2014), their mixture seems not and 371 might hinder downstream extraction and purification processes. 372 3.2.1 Channelling lipids metabolism towards PHA production 373 3.2.1.1 Nitrogen availability limitation 374 In periods II and III, N availability was limited during the feast phase (unbalanced 375 nutrient conditions) to favour the PHA accumulation pathway (Figure 1) and maximise 376 PHA storage (Tan et al., 2014). For that purpose, two different strategies were tested in 377 the SBR and its cycle configuration was changed on day 204 to set uncoupling type A, 378 (period II) and on day 225 to set uncoupling type B (period III). In both cases, C and N 379 feedings were uncoupled to supply the N source at the end of the feast ensuring 380 unbalanced nutrient conditions (C excess in the presence of N limitation). The amount 381 of N added was limited to the minimum needed for microbial growth. Therefore, N was 382 controlled maintaining a minimum concentration in the effluent of the SBR. The 383 quantity of TN added was reduced from 274 ± 28 to 141 ± 10 g of TN per cycle, and 384 consequently, about 12 ± 4 mg TN/L were measured in the effluent, which supposed a 385 reduction of more than 85 % respect to the values observed in the coupled 386 17 configuration. After modifying the N feeding strategy, specific nitrogen consumption 387 and biomass production rates were substantially lower during the feast phase when 388 compared with period I, especially in the case of the type B configuration (period III) 389 (Table 2). In periods II and III, the specific N consumption rate significantly decreased 390 during the feast phase and biomass growth mainly occurred in the famine phase, 391 favouring culture selection (Table 2). Consequently, a positive effect over PHA 392 accumulation was observed (Figure 3.a), which confirmed that N limitation and C and 393 N uncoupling allowed for lipids biotransformation into PHAs and promoted the 394 enrichment of the culture in PHA producers. 395 After C and N uncoupling, the ratio between the feast phase length and the length of the 396 whole cycle decreased (Figure S3.a). This result evidenced the enrichment of the 397 microbial culture in species with high storage ability (Dionisi et al., 2005). The ratio 398 was reduced from 0.21 ± 0.07 in period I (coupled system, accumulation + extracellular 399 growth), to 0.10 ± 0.05 and 0.11 ± 0.04 in periods II and III (single accumulation). 400 The F/F profile also changed when C and N were uncoupled. Initially, microorganisms 401 used organic matter as an electron donor and oxygen as an acceptor (DO concentration 402 decreased). But after this, a lag phase was observed where DO concentration 403 progressively increased. Then, the addition of nutrients reduced the DO again because 404 of respiration for microbial growth. Therefore, the DO profiles presented two slopes, 405 although when the feast length was too short they overlapped (similar DO profiles were 406 observed by Lorini et al. (2020)). The sum of both slopes showed that accumulation 407 plus intracellular growth (0.35 ± 0.07 and 0.35 ± 0.06 in periods II and III, respectively) 408 took longer than simultaneous accumulation and extracellular growth (0.21 ± 0.07, 409 period I) (which is in agreement with the observed by Silva et al. (2017) and Oliveira et 410 al. (2017)) (Figure S4). 411 18 3.2.1.2 Initial carbon availability 412 The higher C availability at the beginning of the cycle in the case of the uncoupling type 413 A cycle (equal amount of COD added in half of the volume) lead to differences in 414 composition and amount of storage compounds accumulated between the two 415 uncoupling strategies tested (Figure 3.a; Table 2). 416 During the feast phase, uncoupling type B cycle caused the enhancement of PHA 417 accumulation at the expense of TAGs (32.1 wt % PHAs and 5.8 wt % TAGs on day 418 238) in comparison to the uncoupling type A cycle (18.4 wt % PHAs and 25.8 wt % 419 TAGs on day 218). The total quantity of storage compounds accumulated at the end of 420 the feast was lower in type B for similar food to microorganism ratios (values of 4.4 and 421 4.6 g COD/g VSS for A and B, respectively). In the uncoupling strategy A, 22.6 wt % 422 TAGs and 2.12 wt % PHAs remained at the end of the cycle whereas in type B only 8.0 423 wt % TAGs and 0.48 wt % PHAs were measured. The higher carbon availability might 424 be selecting for microorganisms that consume it faster, using shorter pathways, which 425 could explain the higher TAG than PHA accumulation at the end of the feast phase in 426 period II. Moreover, due to the high initial concentration in the feast phase, extracellular 427 carbon was not fully consumed and therefore, it was also available in the famine phase. 428 This explains the high percentage of TAGs accumulated measured at the end of the 429 cycle (Vasiliadou et al., 2018). 430 The strategy of uncoupling type B seems to be the most efficient to maximize PHA 431 production. It favours a higher PHA storage during the feast and avoids the presence of 432 extracellular substrate during the famine limiting the growth of non-storing populations. 433 434 435 436 19 3.2.2 Channelling lipids metabolism towards TAG production 437 The pH was maintained stable during the SBR cycles at almost neutral values during 438 period I (Figure 2.a, 2.b and 2.c) showing an average of 7.1 ± 0.3 at the end of the cycle. 439 However, after uncoupling, this parameter became unstable. In period II (uncoupling A) 440 the pH decreased but it was recovered by NaHCO3 addition (Figure 2.d) reaching values 441 of of 7.5 ± 0.2 at the end of the cycle. In period III (uncoupling B) the pH increased at 442 the beginning of the cycle due to NaHCO3 supply, but it decreased during the famine 443 phase ( Figure 2.e) (average of 6.4 ± 0.5 at the end of the cycle). This reduction was a 444 consequence of the microbial growth sustained by the intracellular carbon stored. In the 445 respiration of the intracellular carbon, CO2 is released (Figure 1) acidifying the medium. 446 In the coupled system, it was not observed a decrease of the pH since NaHCO3 was 447 added immediately before the organic matter oxidation, which was mainly extracellular 448 and took place during the feast phase. 449 Considering that respiration acidifies the medium and low pH environments constitute a 450 competitive advantage for oleaginous microorganisms, it was explored the feasibility of 451 selecting a culture with high TAG-storage capacity by the imposition of a double 452 selective pressure in the SBR (period IV, days 254 – 307): N limitation during the feast 453 phase and low pH during the famine. For nitrogen limitation, the same feeding strategy 454 of period III was imposed (uncoupling type B). The pH was reduced by lowering the IC 455 in the dilution water (Figure S3.b) to reduce the buffering capacity of the system and 456 allow for greater acidification in the famine phase. The pH at the beginning of the cycle 457 increased over 7.0 as a consequence of dilution water addition. However, during the 458 feast phase, this parameter sharply decreased since the system buffering capacity was 459 very low. For example, on day 287 of operation (Figure 2.f), the pH at the beginning of 460 the cycle increased up to 7.6 but then decreased reaching a value of 3.8 at the end of the 461 20 cycle. On average, the pH at the end of the cycle decreased from 6.5 ± 0.4 to 4.0 ± 0.3 462 between periods III and IV. 463 In period IV, the highest percentages of TAGs and the lowest of PHAs were observed at 464 the end of the feast phase (Figure 3.a). For example, on day 287 of the enrichment 465 cycle, a maximum of 30.6 wt % TAGs and 0.7 wt % PHAs were achieved, whereas at 466 day 238 of the enrichment cycle (period III) only a 5.8 wt % TAG was accumulated. 467 These results showed that nitrogen limitation during the feast phase favoured the 468 accumulation of storage compounds (advantage for accumulators), but low pH 469 conditions during the famine only allowed for the survival of those species able to grow 470 on the intracellular carbon source at low pH environments and resist these extreme 471 conditions for a long time (TAG accumulators). The ratio between the length of the 472 feast and the length of the cycle (feast/cycle ratio) was in period IV (0.12 ± 0.03) similar 473 to the one measured in periods II and III (0.11 ± 0.04 and 0.12 ± 0.03, respectively) and 474 also indicated an enrichment of the culture in microorganisms with high storage ability 475 (Dionisi et al., 2005). Accumulation together with intracellular growth (0.24 ± 0.06), 476 took longer than simultaneous accumulation and extracellular growth (0.21 ± 0.07, 477 period I). The sum of accumulation and intracellular growth in periods II and III (0.35 ± 478 0.07 and 0.35 ± 0.06, respectively) was higher than in period IV. This result suggested 479 that the use of internal TAGs was easier than the use of internal PHAs for growth. In 480 fact, VSS significantly increased, from 0.50 ± 0.16 g/L and 0.48 ± 0.08 g/L in periods II 481 and III to 0.94 ± 0.13 g/L in period IV. 482 3.2.3 Evaluation of the system recovery capacity 483 During period V (days 308 – 331) the selective pressure imposed by low pH during the 484 famine phase was removed to test if it was possible to return to similar conditions to 485 those of period III and recover the PHA-accumulation capacity of the system. For that 486 21 purpose, the IC of the dilution water was increased obtaining a pH of 6.2 ± 0.4 at the 487 end of the feast phase (similar to that of period III, 6.5 ± 0.4). With it, the VSS 488 concentration decreased, the feast/cycle ratio increased and the evolution of the different 489 parameters monitored during the whole cycle (Figure 2.g) presented a similar trend to 490 the observed in period III. From day 317 onwards an increasing PHA proportion in the 491 storage compounds was observed (Figure 3.a), reaching a 21.6 wt % at the end of the 492 feast phase on day 322 (Figure 2.g). 493 3.2.4 Statistic comparison of the SBR enrichment strategy 494 The effect of imposing certain selective pressures in the enrichment SBR was 495 statistically analysed to determine their impact over the preferent development of TAG 496 or PHA producers, which correlated with the type of intracellular compounds 497 accumulated. For that purpose, it was considered the composition of the intracellular 498 compounds stored at the end of the feast phase in the SBR (operating at stationary state) 499 during periods I – V. 500 Applying one-way ANOVA and Turkey post hoc tests (Table S1), it was observed 501 statistically significant differences in the type of intracellular compounds accumulated 502 when C and N feedings were uncoupled and N availability was limited to promote 503 preferential PHA biosynthesis (same amount of C per unit of time) (periods I vs III). 504 Besides, low pH imposition (period IV) maintaining the same conditions of those of 505 period III, was also demonstrated to have a statistically significant impact over the 506 preferent storage of TAGs. 507 508 509 510 511 28 https://doi.org/10.1007/s10856-018-6183-9 660 Becker, P., 2010. Understanding and Optimizing the Microbial Degradation of Olive 661 Oil: A Case Study with the Thermophilic Bacterium Geobacillus thermoleovorans 662 IHI-91, Olives and Olive Oil in Health and Disease Prevention. Elsevier Inc. 663 https://doi.org/10.1016/B978-0-12-374420-3.00042-5 664 Bušić, A., Kundas, S., Morzak, G., Belskaya, H., Mardetko, N., Šantek, M.I., Komes, 665 D., Novak, S., Šantek, B., 2018. Recent trends in biodiesel and biogas production. 666 Food Technol. Biotechnol. 56, 152–173. 667 https://doi.org/10.17113/ftb.56.02.18.5547 668 Campanari, S., E Silva, F.A., Bertin, L., Villano, M., Majone, M., 2014. Effect of the 669 organic loading rate on the production of polyhydroxyalkanoates in a multi-stage 670 process aimed at the valorization of olive oil mill wastewater. Int. J. Biol. 671 Macromol. 71, 34–41. https://doi.org/10.1016/j.ijbiomac.2014.06.006 672 Carsanba, E., Papanikolaou, S., Erten, H., 2018. Production of oils and fats by 673 oleaginous microorganisms with an emphasis given to the potential of the 674 nonconventional yeast Yarrowia lipolytica. Crit. Rev. Biotechnol. 38, 1230–1243. 675 https://doi.org/10.1080/07388551.2018.1472065 676 Castro, A.R., Silva, P.T.S., Castro, P.J.G., Alves, E., Domingues, M.R.M., Pereira, 677 M.A., 2018. Tuning culturing conditions towards the production of neutral lipids 678 from lubricant-based wastewater in open mixed bacterial communities. Water Res. 679 144, 532–542. https://doi.org/10.1016/j.watres.2018.07.068 680 Chan, L.G., Cohen, J.L., Ozturk, G., Hennebelle, M., Taha, A.Y., De Moura Bell, 681 J.M.L.N., 2018. Bioconversion of cheese whey permeate into fungal oil by Mucor 682 circinelloides. J. Biol. Eng. 12, 1–14. https://doi.org/10.1186/s13036-018-0116-5 683 29 Chipasa, K.B., Mȩdrzycka, K., 2006. Behavior of lipids in biological wastewater 684 treatment processes. J. Ind. Microbiol. Biotechnol. 33, 635–645. 685 https://doi.org/10.1007/s10295-006-0099-y 686 Cristóvaõ, R.O., Botelho, C.M., Martins, R.J.E., Loureiro, J.M., Boaventura, R.A.R., 687 2015. Fish canning industry wastewater treatment for water reuse - A case Study. 688 J. Clean. Prod. 87, 603–612. https://doi.org/10.1016/j.jclepro.2014.10.076 689 da Cruz Pradella, J.G., Ienczak, J.L., Delgado, C.R., Taciro, M.K., 2012. Carbon source 690 pulsed feeding to attain high yield and high productivity in poly(3691 hydroxybutyrate) (PHB) production from soybean oil using Cupriavidus necator. 692 Biotechnol. Lett. 34, 1003–1007. https://doi.org/10.1007/s10529-012-0863-1 693 Dionisi, D., Majone, M., Vallini, G., Gregorio, S. Di, Beccari, M., 2005. Effect of the 694 Applied Organic Load Rate on Biodegradable Polymer Production by Mixed 695 Microbial Cultures in a Sequencing Batch Reactor. 696 https://doi.org/10.1002/bit.20683 697 Donot, F., Fontana, A., Baccou, J.C., Strub, C., Schorr-Galindo, S., 2014. Single cell 698 oils (SCOs) from oleaginous yeasts and moulds: Production and genetics. Biomass 699 and Bioenergy 68, 135–150. https://doi.org/10.1016/j.biombioe.2014.06.016 700 Fra-Vázquez, A., Pedrouso, A., Palmeiro-Sánchez, T., Moralejo-Gárate, H., Mosquera701 Corral, A., 2018. Feasible microbial accumulation of triacylglycerides from crude 702 glycerol. J. Chem. Technol. Biotechnol. 93, 2644–2651. 703 https://doi.org/10.1002/jctb.5618 704 Garay, L.A., Boundy-Mills, K.L., German, J.B., 2014. Accumulation of high-value 705 lipids in single-cell microorganisms: A mechanistic approach and future 706 perspectives. J. Agric. Food Chem. 62, 2709–2727. 707 30 https://doi.org/10.1021/jf4042134 708 Gujjala, L.K.S., Kumar, S.P.J., Talukdar, B., Dash, A., Kumar, S., Sherpa, K.C., 709 Banerjee, R., 2019. Biodiesel from oleaginous microbes: opportunities and 710 challenges. Biofuels 10, 45–59. https://doi.org/10.1080/17597269.2017.1402587 711 Herrero, O.M., Villalba, M.S., Lanfranconi, M.P., Alvarez, H.M., 2018. Rhodococcus 712 bacteria as a promising source of oils from olive mill wastes. World J. Microbiol. 713 Biotechnol. 34, 0. https://doi.org/10.1007/s11274-018-2499-3 714 Hori, K., Abe, M., Unno, H., 2009. Production of triacylglycerol and poly(3715 hydroxybutyrate-co-3-hydroxyvalerate) by the toluene-degrading bacterium 716 Rhodococcus aetherivorans IAR1. J. Biosci. Bioeng. 108, 319–324. 717 https://doi.org/10.1016/j.jbiosc.2009.04.020 718 Huang, C., Luo, M.T., Chen, X.F., Qi, G.X., Xiong, L., Lin, X.Q., Wang, C., Li, H.L., 719 Chen, X. De, 2017. Combined “de novo” and “ex novo” lipid fermentation in a 720 mix-medium of corncob acid hydrolysate and soybean oil by Trichosporon 721 dermatis. Biotechnol. Biofuels 10, 1–11. https://doi.org/10.1186/s13068-017-0835722 8 723 Jimenez-diaz, L., Caballero, A., Segura, A., 2019. Aerobic Utilization of Hydrocarbons, 724 Oils, and Lipids. Aerob. Util. Hydrocarb. Oils, Lipids 291–313. 725 https://doi.org/10.1007/978-3-319-50418-6 726 Kourmentza, C., Plácido, J., Venetsaneas, N., Burniol-Figols, A., Varrone, C., Gavala, 727 H.N., Reis, M.A.M., 2017. Recent Advances and Challenges towards Sustainable 728 Polyhydroxyalkanoate (PHA) Production. Bioengineering 4, 55. 729 https://doi.org/10.3390/bioengineering4020055 730 Kosseva M., Webb C., 2013. Food industry wastes. Assessment and Recuperation of 731 31 Commodities, 1st ed, Academic press. 732 Kumar, M., Ghosh, P., Khosla, K., Thakur, I.S., 2018. Recovery of 733 polyhydroxyalkanoates from municipal secondary wastewater sludge. Bioresour. 734 Technol. 255, 111–115. https://doi.org/10.1016/j.biortech.2018.01.031 735 Kumar, M., Rathour, R., Singh, R., Sun, Y., Pandey, A., Gnansounou, E., Andrew Lin, 736 K.Y., Tsang, D.C.W., Thakur, I.S., 2020. Bacterial polyhydroxyalkanoates: 737 Opportunities, challenges, and prospects. J. Clean. Prod. 263, 121500. 738 https://doi.org/10.1016/j.jclepro.2020.121500 739 Leng, L., Zhang, W., Peng, H., Li, H., Jiang, S., Huang, H., 2020. Nitrogen in bio-oil 740 produced from hydrothermal liquefaction of biomass: A review. Chem. Eng. J. 741 401, 126030. https://doi.org/10.1016/j.cej.2020.126030 742 Lopes, M., Gomes, A.S., Silva, C.M., Belo, I., 2018. Microbial lipids and added value 743 metabolites production by Yarrowia lipolytica from pork lard. J. Biotechnol. 265, 744 76–85. https://doi.org/10.1016/j.jbiotec.2017.11.007 745 Lorini, L., di Re, F., Majone, M., Valentino, F., 2020. High rate selection of PHA 746 accumulating mixed cultures in sequencing batch reactors with uncoupled carbon 747 and nitrogen feeding. N. Biotechnol. 56, 140–148. 748 https://doi.org/10.1016/j.nbt.2020.01.006 749 Maqbool, A., Strandvik, B., Stallings, V.A., 2011. The skinny on tuna fat: Health 750 implications. Public Health Nutr. 14, 2049–2054. 751 https://doi.org/10.1017/S1368980010003757 752 Md Din, M.F., Ujang, Z., van Loosdrecht, M.C.M., Ahmad, A., Sairan, M.F., 753 2006.Optimization of nitrogen and phosphorus limitation for better biodegradable 754 plasticproduction and organic removal using single fed-batch mixed cultures 755 32 andrenewable resources. Water Sci. Technol. 53, 15– 756 20.https://doi.org/10.2166/wst.2006.164Oliveira, C.S.S., Silva, C.E., Carvalho, 757 G.,Biotechnol. 37, 69–79 758 Oliveira, C.S.S., Silva, C.E., Carvalho, G., Reis, M.A., 2017. Strategies for efficiently 759 selecting PHA producing mixed microbial cultures using complex feedstocks: 760 Feast and famine regime and uncoupled carbon and nitrogen availabilities. N. 761 Biotechnol. 37, 69–79. https://doi.org/10.1016/j.nbt.2016.10.008 762 Papanikolaou, S., Aggelis, G., 2010. Yarrowia lipolytica: A model microorganism used 763 for the production of tailor-made lipids. Eur. J. Lipid Sci. Technol. 112, 639–654. 764 https://doi.org/10.1002/ejlt.200900197 765 Patel, A., Matsakas, L., 2019. A comparative study on de novo and ex novo lipid 766 fermentation by oleaginous yeast using glucose and sonicated waste cooking oil. 767 Ultrason. Sonochem. 52, 364–374. https://doi.org/10.1016/j.ultsonch.2018.12.010 768 Pérez-Arauz, A.O., Aguilar-Rabiela, A.E., Vargas-Torres, A., Rodríguez-Hernández, 769 A.I., Chavarría-Hernández, N., Vergara-Porras, B., López-Cuellar, M.R., 2019. 770 Production and characterization of biodegradable films of a novel 771 polyhydroxyalkanoate (PHA) synthesized from peanut oil. Food Packag. Shelf Life 772 20, 100297. https://doi.org/10.1016/j.fpsl.2019.01.001 773 Revellame, E.D., Hernandez, R., French, W., Holmes, W.E., Benson, T.J., Pham, P.J., 774 Forks, A., Callahan, R., 2012. Lipid storage compounds in raw activated sludge 775 microorganisms for biofuels and oleochemicals production. RSC Adv. 2, 2015– 776 2031. https://doi.org/10.1039/c2ra01078j 777 Riedel, S.L., Jahns, S., Koenig, S., Bock, M.C.E., Brigham, C.J., Bader, J., Stahl, U., 778 2015. Polyhydroxyalkanoates production with Ralstonia eutropha from low quality 779 33 waste animal fats. J. Biotechnol. 214, 119–127. 780 https://doi.org/10.1016/j.jbiotec.2015.09.002 781 Sabapathy, P.C., Devaraj, S., Meixner, K., Anburajan, P., Kathirvel, P., Ravikumar, Y., 782 Zabed, H.M., Qi, X., 2020. Recent developments in Polyhydroxyalkanoates 783 (PHAs) production – A review. Bioresour. Technol. 123132. 784 https://doi.org/10.1016/j.biortech.2020.123132 785 Sangkharak, K., Paichid, N., Yunu, T., Klomklao, S., Prasertsan, P., 2020. Utilisation of 786 tuna condensate waste from the canning industry as a novel substrate for 787 polyhydroxyalkanoate production. Biomass Convers. Biorefinery. 788 https://doi.org/10.1007/s13399-019-00581-4 789 Santamauro, F., Whiffin, F.M., Scott, R.J., Chuck, C.J., 2014. Low-cost lipid production 790 by an oleaginous yeast cultured in non-sterile conditions using model waste 791 resources. Biotechnol. Biofuels 7, 1–11. https://doi.org/10.1186/1754-6834-7-34 792 Sawangkeaw, R., Ngamprasertsith, S., 2013. A review of lipid-based biomasses as 793 feedstocks for biofuels production. Renew. Sustain. Energy Rev. 25, 97–108. 794 https://doi.org/10.1016/j.rser.2013.04.007 795 Silva, F., Campanari, S., Matteo, S., Valentino, F., Majone, M., Villano, M., 2017. 796 Impact of nitrogen feeding regulation on polyhydroxyalkanoates production by 797 mixed microbial cultures. N. Biotechnol. 37, 90–98. 798 https://doi.org/10.1016/j.nbt.2016.07.013 799 Song, J.H., Jeon, C.O., Choi, M.H., Yoon, S.C., Park, W., 2008. Polyhydroxyalkanoate 800 (PHA) production using waste vegetable oil by Pseudomonas sp. strain DR2. 801 J.Microbiol. Biotechnol. 18, 1408–1415. 802 Sudesh, K., Abe, H., Doi, Y., 2000. Synthesis, structure and properties of 803 34 polyhydroxyalkanoates: Biological polyesters. Prog. Polym. Sci. 25, 1503–1555. 804 https://doi.org/10.1016/S0079-6700(00)00035-6 805 Surendran, A., Lakshmanan, M., Chee, J.Y., Sulaiman, A.M., Thuoc, D. Van, Sudesh, 806 K., 2020. Can Polyhydroxyalkanoates Be Produced Efficiently From Waste Plant 807 and Animal Oils? Front. Bioeng. Biotechnol. 8. 808 https://doi.org/10.3389/fbioe.2020.00169 809 Tamis, J., Sorokin, D.Y., Jiang, Y., Van Loosdrecht, M.C.M., Kleerebezem, R., 2015. 810 Lipid recovery from a vegetable oil emulsion using microbial enrichment cultures. 811 Biotechnol. Biofuels 8, 1–11. https://doi.org/10.1186/s13068-015-0228-9 812 Tan, G.Y.A., Chen, C.L., Li, L., Ge, L., Wang, L., Razaad, I.M.N., Li, Y., Zhao, L., 813 Mo, Y., Wang, J.Y., 2014. Start a research on biopolymer polyhydroxyalkanoate 814 (PHA): A review. Polymers (Basel). 6, 706–754. 815 https://doi.org/10.3390/polym6030706 816 Van Thuoc, D., My, D.N., Loan, T.T., Sudesh, K., 2019. Utilization of waste fish oil 817 and glycerol as carbon sources for polyhydroxyalkanoate production by 818 Salinivibrio sp. M318. Int. J. Biol. Macromol. 141, 885–892. 819 https://doi.org/10.1016/j.ijbiomac.2019.09.063 820 Vasiliadou, I.A., Bellou, S., Daskalaki, A., Tomaszewska-Hetman, L., Chatzikotoula, 821 C., Kompoti, B., Papanikolaou, S., Vayenas, D., Pavlou, S., Aggelis, G., 2018. 822 Biomodification of fats and oils and scenarios of adding value on renewable fatty 823 materials through microbial fermentations: Modelling and trials with Yarrowia 824 lipolytica. J. Clean. Prod. 200, 1111–1129. 825 https://doi.org/10.1016/j.jclepro.2018.07.187 826 Vishniac, W., Santer, M., 1957. The thiobacilli. Bacteriol. Rev. 21, 195–213. 827 35 Wakelin, N.G., Forster, C.F., 1997. An investigation into microbial removal of fats, oils 828 and greases. Bioresour. Technol. 59, 37–43. https://doi.org/10.1016/S0960829 8524(96)00134-4 830 Wallace, T., Gibbons, D., O’Dwyer, M., Curran, T.P., 2017. International evolution of 831 fat, oil and grease (FOG) waste management – A review. J. Environ. Manage. 187, 832 424–435. https://doi.org/10.1016/j.jenvman.2016.11.003 833 Waller, J.L., Green, P.G., Loge, F.J., 2012. Mixed-culture polyhydroxyalkanoate 834 production from olive oil mill pomace. Bioresour. Technol. 120, 285–289. 835 https://doi.org/10.1016/j.biortech.2012.06.024 836 Wu, H., Gao, X., Wu, M., Zhu, Y., Xiong, R., Ye, S., 2020. The efficiency and risk to 837 groundwater of constructed wetland system for domestic sewage treatment - a case 838 study in Xiantao, China. J. Clean. Prod. 123384. 839 https://doi.org/10.1016/j.jclepro.2020.123384 840 Yadav, B., Pandey, A., Kumar, L.R., Tyagi, R.D., 2020. Bioconversion of waste 841 (water)/residues to bioplasticsA circular bioeconomy approach. Bioresour. 842 Technol. 298, 122584. https://doi.org/10.1016/j.biortech.2019.122584 843 844 1 Figure 1. Simplified metabolic pathways describing TAG and PHA production from lipids in microbial cultures. Free fatty acids (FFA) yielded after the substrate hydrolysis can be catabolized (direct respiration) or accumulated as reserve materials (TAGs or PHAs produced by the ex novo or PHA biosynthesis pathways, respectively). 2 Figure 2. Characterization of the SBR enrichment cycles. Concentrations of DO (⸺), X (---), pH (●), TN (♦), TAG + PHA (○). 0 2 4 6 8 10 0 20 40 60 80 100 0246810 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA(%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0 2 4 6 8 10 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0246810 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0246810 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0 2 4 6 8 10 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0 2 4 6 8 10 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) 0 2 4 6 8 10 0 20 40 60 80 100 0246810 12 DO (mg/L), pH, NT (Nmmol/L) TAG+PHA (%wt), X (Cmmol/L) Time (h) Day 50 (period I) Day 78 (period I) Day 171 (period I) I) Day 218 (period II) Day 238 (period III) Day 287 (period IV) Day 322 (period V) a) b) c) d) e) f) g) 5 Table 4 (Continuation) Inoculum Feedstock Production process wt % max. YTAG/S - YPHA/S Reference TAGs MMC Activated sludge Soybean oil SBR + FBR 54 - Tamis et al. (2015) PHAs MMC Activated sludge Olive oil pomace Pretreatment + SBR 39 0.36 g / L Waller et al. (2012) Activated sludge Olive oil mill wastewater Pretreatment + Acidification + SBR + FBR 30 0.56 g COD / g COD Campanari et al. (2014) Activated sludge (10 %) and palm oil mill effluent (90 %) Palm oil mill wastewater Single SBR for growth and accumulation 45 - Md Din et al. (2006) Activated sludge for aerobic granules formation Palm oil mill wastewater Pretreatment + SBR 68 0.66 g /g COD Gobi and Vadivelu (2014) PHAs + TAGs MMC Activated sludge The oily fraction of a fishcanning industry effluent SBR + FBR (1) TAG: 43 (2) PHA: 82 (1) 0.67 CmmolTAG / CmmolS (2) 0.80 CmmolPHA / CmmolS This research work (1) Data concerning Period IV, highest TAGs accumulation and production yield. (2) Data regarding Period III, highest PHAs accumulation and production yield. 1 A novel strategy for triacylglycerides and polyhydroxyalkanoates production using waste lipids Lucía Argiz a *, Rebeca González-Cabaleiro b, Ángeles Val del Río a, Jesús González-López c, Anuska Mosquera-Corral a a CRETUS Institute, Department of Chemical Engineering, Universidade de Santiago de Compostela, 15782 Santiago de Compostela, Galicia, Spain b Department of Infrastructure and Environment, University of Glasgow, Rankine Building, Glasgow, G12 8LT, UK c Department of Microbioloy and Institute of Water Research, Universidad de Granada, Granada, Spain * Corresponding author: luciaargiz.m[email protected] 2 Figure S1. Configuration of the SBR cycles during the different operational periods (I – V). Coupled system configuration (period I) Feeding1 Aeration Withdrawal Time (min) 5 708 7 Feeding1: 2 mL C & 2 L of a nutrients solution containing N. Uncoupled system configuration type A (period II) Feeding2 Aeration Feeding3 Withdrawal Time (min) 5 180 5 523 7 Feeding2: 2 mL C; Feeding3: 2 L of a nutrients solution containing N. Uncoupled system configuration type B (periods III, IV and V) Feeding4 Aeration Feeding5 Withdrawal Time (min) 5 120 – 180 5 523 – 583 7 Feeding4: 2 mL C & 2 L of a nutrients solution without N; Feeding5: 20 mL of a N solution. 3 Figure S2. Distribution of the different operational periods in the SBR (I – V) and accumulation assays performed in the FBR. 4 Figure S3. Evolution of different parameters in the SBR during periods I – V. a) Feast phase length considering accumulation + growth ( ), single accumulation (●); b) Inorganic carbon (×) and pH (●) of the nutrients solution, pH at the end of the cycle ( ); c) TN of the nutrients solution (●), TN at the end of the cycle ( ); d) VSS at the end of the cycle ( ). 0 2 4 6 8 10 12 050 100 150 200 250 300 350 Feast length (h) Day of operation 0 20 40 60 80 100 0 2 4 6 8 10 050 100 150 200 250 300 350 IC feeding (g/L) pH Day of operation 0 50 100 150 200 050 100 150 200 250 300 350 TN (mg/L) Day of operation 0.0 0.5 1.0 1.5 2.0 2.5 050 100 150 200 250 300 350 VSS (g/L) Day of operation I II III IV V I II III IV V I II III IV V I II III IV V a a) a b) a c) a d) 5 Figure S4. Comparison of the evolution of the dissolved oxygen (DO) profiles in the coupled and uncoupled systems and feast phase length quantification. 0 2 4 6 8 10 0 4 8 12 16 20 24 DO (g/L) Time (h) Accumulation + extracellular growth Period I (day 118) – Coupled configuration FEAST PHASE 0 2 4 6 8 04812 16 20 24 DO (g/L) Time (h) Period III (day 237) – Uncoupled configuration FEAST PHASE Intracellular growth Accumulation 6 Figure S5. Simplified flowchart of the research methodology and operational periods. 7 Table S1. Results of the statistical analysis. a) One-Way ANOVA, Turkey post hoc test considering TAG and PHA as independent variables: comparison of the storage compounds composition at the end of the feast phase in the enrichment SBR during different operational periods. b) t-test for independent samples: comparison of the composition of the intracellular compounds accumulated during the same operational periods in SBR and FBR reactors. a) One-Way ANOVA Sum of Squares df Mean Square F Sig. TAG Between Groups 11273.204 4 2818.301 8.731 .001 Within Groups 4519.156 14 322.797 Total 15792.360 18 PHA Between Groups 11273.204 4 2818.301 8.731 .001 Within Groups 4519.156 14 322.797 Total 15792.360 18 Turkey post hoc tests (TAG as an independent variable) Period (i) Period (j) Mean Difference (i – j) Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound I II 28.29000 14.66963 .348 -17.4197 73.9997 III (1) 71.51000 14.66963 (2) .002 25.8003 117.2197 IV -10.11000 11.59736 .903 -46.2467 26.0267 V 26.52000 10.87929 .162 -7.3792 60.4192 II I -28.29000 14.66963 .348 -73.9997 17.4197 III 43.22000 17.96655 .171 -12.7628 99.2028 IV -38.40000 15.55949 .154 -86.8825 10.0825 V -1.77000 15.03189 1.000 -48.6085 45.0685 III I (1) -71.51000 14.66963 (2) .002 -117.2197 -25.8003 II -43.22000 17.96655 .171 -99.2028 12.7628 IV (1) -81.62000 15.55949 (2).001 -130.1025 -33.1375 V -44.99000 15.03189 .062 -91.8285 1.8485 IV I 10.11000 11.59736 .903 -26.0267 46.2467 II 38.40000 15.55949 .154 -10.0825 86.8825 III (1) 81.62000 15.55949 (2).001 33.1375 130.1025 V 36.3000 12.05233 .057 -.9244 74.1844 V I -26.2000 10.87929 .162 -60.4192 7.3792 II 1.7000 15.03189 1.000 -45.0685 48.6085 III 44.99000 15.03189 .062 -1.8485 91.8285 IV -36.63000 12.05233 .057 -74.1844 .9244 8 Table S1 (continuation) a) Turkey post hoc tests (TAG as an independent variable) Period (i) Period (j) Mean Difference (i – j) Std. Error Sig. 95% Confidence Interval Lower Bound Upper Bound I II -28.29000 14.66963 .348 -73.9997 17.4197 III (1) -71.51000 14.66963 (2).002 -117.2197 -25.8003 IV 10.11000 11.59736 .903 -26.0267 46.2467 V -26.52000 10.87929 .162 -60.4192 7.3792 II I 28.29000 14.66963 .348 -17.4197 73.9997 III -43.2000 17.96655 .171 -99.2028 12.7628 IV 38.40000 15.55949 .154 -10.0825 86.8825 V 1.77000 15.03189 1.000 -45.0685 48.6085 III I (1) 71.51000 14.66963 (2) .002 25.8003 117.2197 II 43.22000 17.96655 .171 -12.7628 99.2028 IV (1) 81.62000 15.55949 (2) .001 33.1375 130.1025 V 44.99000 15.03189 .062 -1.8485 91.8285 IV I -10.11000 11.59736 .903 -46.2467 26.0267 II -38.40000 15.55949 .154 -86.8825 10.0825 III (1) -81.62000 15.55949 (2) .001 -130.1025 -33.1375 V -36.63000 12.05233 .057 -74.1844 .9244 V I 26.52000 10.87929 .162 -7.3792 60.4192 II -1.77000 15.03189 1.000 -48.6085 45.0685 III -44.99000 15.03189 .062 -91.8285 1.8485 IV 36.63000 12.05233 .057 -.9244 74.1844 (1) The mean difference is significant at the 0.05 level. (2) Significatively different (sig. < 0.05). b) Group statistics Reactor Nº of samples Mean Std. deviation Std. error deviation SBR 5 53.0204 32.98924 14.75324 FBR 5 45.8840 35.20856 15.74575 Levene’s test for equality of variances F Sig. 0.134 0.724 9 Table S1 (continuation) b) t-test for Equality of Means t Degrees of freedom Sig (2-tailed) Mean difference Std. Error Difference 95 % Confidence Interval of the Difference Lower Upper Equal variances .331 8 .749 7.13640 21.57745 -42.62130 56.89410 Equal variances not assumed .331 7.966 .749 7.13640 21.57745 -42.65793 56.93073