the potential of osmolytes and their precursors to...
TRANSCRIPT
lable at ScienceDirect
Water Research 147 (2018) 142e151
Contents lists avai
Water Research
journal homepage: www.elsevier .com/locate/watres
The potential of osmolytes and their precursors to alleviate osmoticstress of anaerobic granular sludge.
D. Sudmalis a, *, S.K. Millah a, M.C. Gagliano b, C.I. Butr�e c, C.M. Plugge b, H.H.M. Rijnaarts a,G. Zeeman a, H. Temmink a
a Sub-department of Environmental Technology, Wageningen University and Research, Bornse Weilanden 9, 6708 WG, Wageningen, the Netherlandsb Laboratory of Microbiology, Wageningen University and Research, Stippeneng 4, 6708WE, Wageningen, the Netherlandsc Laboratory of Food Chemistry, Wageningen University and Research, Wageningen, the Netherlands
a r t i c l e i n f o
Article history:Received 3 August 2018Received in revised form27 September 2018Accepted 29 September 2018Available online 2 October 2018
Keywords:OsmolytesAnaerobic granular sludgeMethanogenic activitySaline wastewater
* Corresponding author.E-mail addresses: [email protected],
(D. Sudmalis).
https://doi.org/10.1016/j.watres.2018.09.0590043-1354/© 2018 The Authors. Published by Elsevier
a b s t r a c t
Increasing amounts of saline (waste)water with high concentrations of organic pollutants are generatedglobally. In the anaerobic (waste)water treatment domain, high salt concentrations are repeatedly re-ported to inhibit methanogenic activity and strategies to overcome this toxicity are needed. Currentresearch focuses on the use of potential osmolyte precursor compounds for osmotic stress alleviation ingranular anaerobic sludges upon exposure to hypersalinity shocks. Glutamic acid, aspartic acid, lysine,potassium, gelatine, and tryptone were tested for their potential to alleviate osmotic stress in laboratorygrown and full e scale granular sludge. The laboratory grown granular sludge was adapted to 5 (R5) and20 (R20) g Naþ/L. Full-scale granular sludge was obtained from internal circulation reactors treatingtannery (waste)water with influent conductivity of 29.2 (Do) and 14.1 (Li) mS/cm. In batch experimentswhich focused on specific methanogenic activity (SMA), R5 granular sludge was exposed to a hypersa-linity shock of 20 g Naþ/L. The granular sludge of Do and Li was exposed to a hypersalinity shock of 10 gNaþ/L with sodium acetate as the sole carbon source. The effects on R20 granular sludge were studied atthe salinity level to which the sludge was already adapted, namely 20 g Naþ/L. Dosing of glutamic acid,aspartic acid, gelatine, and tryptone resulted in increased SMA compared to only acetate fed batches. Inbatches with added glutamic acid, the SMA increased by 115% (Li), 35% (Do) and 9% (R20). With addedaspartic acid, SMA increased by 72% (Li), 26% (Do), 12% (R5) and 7% (R20). The addition of tryptoneresulted in SMA increases of 36% (R5), 17% (R20), 179% (Li), and 48% (Do), whereas added gelatineincreased the SMA by 30% (R5), 14% (R20), 23% (Li), and 13% (Do). The addition of lysine, meanwhile, gavenegative effects on SMA of all tested granular sludges. Potassium at sea water Na/K ratio (27.8 w/w) had aslight positive effect on SMA of Do (7.3%) and Li (10.1%), whereas at double the sea water ratio (13.9% w/w) had no pronounced positive effect. R20 granular sludge was also exposed to hyposalinity shock from20 down to 5 g Naþ/L. Glutamate and N-acetyl-b-lysine were excreted by microbial consortium inanaerobic granular sludge adapted to 20 g Naþ/L upon this exposure to hyposalinity. A potentialconsequence when applying these results is that saline streams containing specific and hydrolysableproteins can be anaerobically treated without additional dosing of osmolytes.© 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Economic sectors such as leather tanning, chemical and agro-food industries generate large amounts of saline (waste)waterthat is both rich in salt (NaCl) and organic matter (Lefebvre and
Ltd. This is an open access article u
Moletta, 2006). Le Borgne et al. (2008) reported that around 5%of the globally produced industrial (waste)water is either saline orhypersaline. This amount of saline industrial (waste)water isincreasing due to growing industrial activities and recycling of(waste)water (Rozzi et al., 1999; Giustinianovich et al., 2018). Aviable option for removing bulk organic pollutants from this wateris anaerobic biological treatment. This approach also generatesenergy and reduces the amount of waste sludge. If sludge aggre-gation into microbial granules can be achieved, high rate anaerobic
nder the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
D. Sudmalis et al. / Water Research 147 (2018) 142e151 143
reactors such as upflow anaerobic sludge blanket (UASB) should beconsidered. The excellent settling properties of granular sludgeallows for a reduction in the bioreactor footprint (Vieira et al., 2005;Xiao and Roberts, 2010) and the high rate anaerobic reactors arealready widely applied for industrial (waste)water treatment (VanLier et al., 2015).
There are two factors of particular importance when treating(waste)water with high concentrations of soluble chemical oxygendemand (COD) with high rate anaerobic treatment: i) microbialactivity; and ii) microbial sludge granulation and stability(O'flaherty et al., 1997; McHugh et al., 2003; Liu et al., 2009; Luet al., 2013). However, high concentrations of monovalent saltshave been repeatedly reported to inhibit the methanogenic activity(Rinzema et al., 1988; Liu and Boone, 1991; De Vrieze et al., 2016b)and cause granular sludge disintegration with subsequent biomasswashout from the bioreactors (Rinzema et al., 1988; Ismail et al.,2008; Jeison et al., 2008; De Vrieze et al., 2016a).
Microbial cells have two different strategies for adapting to highosmotic pressure. The first is by increasing their intracellular ionconcentration (mainly Kþ and Cl�). The second is by up taking and/or synthesizing small organic molecules called compatible solutes(Sowers et al., 1990; Oren, 1999; Roberts, 2005). The first strategy isused by Halobacteria (Archaea) and Haloanaerobiales (Bacteria) andrequires energy for adaptation of the cells enzyme machinery(Oren, 1999; Roebler and Müller, 2001; Müller et al., 2005).
The second strategy, which is used by the majority of the livingcells (Pflüger et al., 2005), does not require such adaptation.Additionally, microorganisms applying this mechanism can survivein a broader range of salinity levels (Roebler and Müller, 2001;Müller et al., 2005; Pflüger et al., 2005). In compatible soluteaccumulating bacteria, the response to an osmotic shock can bedivided into two phases. During the first phase, a rapid influx ofpotassium ion into themicrobial cell is triggered. During the secondphase, cellular levels of potassium are decreased and compatiblesolutes are accumulated instead (Pflüger et al., 2005). Similar re-sponses to osmotic stress have been demonstrated in pure culturesof the methanogenic archaea Methanococcus thermolitotrophicus(Martin et al., 2000). Since salinity of (waste)water often fluctuatesin time (Vyrides, 2015), this mechanism is likely to be utilized bythe microbial consortium of bioreactors treating several types ofsaline (waste)water.
Several types of molecules can act as compatible solutes: sugars(trehalose, betaine), amino acids (glutamic acid, proline), de-rivatives of amino acids (N-ε-acetyl-b-lysine, ectoine) (Roberts,2005), and polyols (Roebler and Müller, 2001). Uptake from thebulk liquid of such molecules is bio-energetically more favourablethan synthesis de novo (Oren, 1999). This explains why severalstudies focused on the potential of dosing osmolytes in anaerobic(waste)water treatment systems. Molecules like glycine betaine, a-glutamate, b-glutamate, trehalose and, N-acetyl-b-lysine (Yerkeset al., 1997; Vyrides and Stuckey, 2009; Vyrides et al., 2010; Heet al., 2012) have all been demonstrated to have the potential ofincreasing methanogenic activity when flocculent biomass wasexposed to osmotic stress. In a recent review Vyrides and Stuckey(2016) reported that glycine betaine is by far the most studiedosmolyte to date. The effects of osmolytes onmethanogenic activityof granular sludge are not researched so far. Moreover, the bottleneck for application of osmolytes in practice is the relatively highcost of such molecules (Vyrides and Stuckey, 2016).
Recently, robust anaerobic sludge granulation at high salinitylevels (up to 20 g Naþ/L) was reported. The microbial activity wassufficient to achieve a high soluble COD removal efficiency (94%) ata loading rate of 16 g COD/L$d. The source of carbon in the synthetic(waste)water was a mixture of glucose, acetate, and tryptone(Sudmalis et al., 2018). Proteins contain a broad range of amino
acids as their natural building blocks (Smith and Friedman,1984; DiGioia and Guilbert, 1999). These amino acids are known to act asosmo-protectant precursors in anaerobic microorganisms (Roberts,2005; Zaprasis et al., 2014). It was hypothesised that, after hydro-lysis, tryptone (a mixture of peptides formed by digesting caseinwith trypsin) supplied the microorganisms with the “building-blocks” for the osmo-protectants uptake/synthesis. Thus hydro-lysable proteins are potentially a cheaper alternative to osmolytes ifindeed they can alleviate osmotic stress.
In this work, the osmolytes present in granular sludge adaptedto 20 g Naþ/L were first identified. Next it was investigated if theseosmolytes, their precursors, and proteins were able to alleviateosmotic shock stress in four types of granular sludge. Furthermore,the potential of potassium to alleviate osmotic shock stress wasstudied, since its rapid uptake is reported as the first response toincrease in salinity. The granular sludge was obtained from twofull-scale anaerobic reactors treating saline tannery (waste)waterand two from laboratory-scale reactors treating synthetic (waste)water at salinity levels of 5 and 20 g Naþ/L.
The results show that not only osmolytes, but also their pre-cursors and proteins can alleviate osmotic stress in all four types ofthe anaerobic granular sludges.
2. Materials and methods
2.1. Inoculum
Full scale granular sludge was obtained from two internal cir-culation (IC) anaerobic reactors (Waterstromen, The Netherlands)treating tannery (waste)water in Dongen (Do) and Lichtenvoorde(Li) (The Netherlands). The average influent (waste)water conduc-tivity was 14.1 and 29.2mS/cm for the IC reactors in Lichtenvoordeand Dongen, respectively. The average influent total COD concen-tration was 6.8 g/L and 5 g/L in Lichtenvoorde and Dongen,respectively. The average influent chloride concentration was 4 g/Lin Lichtenvoorde and 6.7 g/L in Dongen. The granular sludges fromthe full-scale reactors were pre-washed with 5 g Naþ/L nutrientmedium (composition as described in Table 1) and were stored inthe samemedium at 4 �C before use in batch experiments. This wasdone in order to have a defined liquid phase ionic compositionduring the batch experiments.
The laboratory-scale granular sludge adapted to 5 and 20 g Naþ/L was formed from dispersed biomass, as described in detail inSudmalis et al. (2018) and was kept at 4 �C before use in batchexperiments. From this point forward, the granular sludge formedat 5 and 20 g Naþ/L is referred to as R5 and R20, respectively.
2.2. Batch experiments
Three types of batch experiments were performed in this study.The first involved use of tryptone at high COD proportion to studyits potential for alleviating salt stress in R5 granular sludge. Thesecond involved an abrupt decrease in salinity and identification ofosmolytes in R20 granular sludge. The third type of experimentswas performed to study the potential of potassium and potentialprecursors of osmolytes to alleviate osmotic stress in R5, R20, Doand Li granular sludge. A detailed description of each type ofexperiment is given in sections 2.2.1., 2.2.2. and 2.2.3.
The batch experiments with laboratory grown granules (R5 andR20) were performed in 118mL serum bottles with a workingvolume of 50mL. The experiments were performed in an incubatorat 35 �C and with a 120 RPM mixing speed. The VSS concentrationwas set to 1 g/L and the COD:VSS ratio was 4:1 (w/w).
The experiments with full scale granules (Do and Li) were per-formed in Automatic Methane Potential Test System. (AMPTS) II
Table 1Nutrient medium with final salt concentrations used in all batch experiments. *NaCl was used to adjust sodium concentration depending on the type of experiment.**(NH4)2CO3 was used to compensate for additional nitrogen originating from osmo-protectants used in experiments with increasing salinity.
Macronutrients, alkalinity, Na, NH4eN Micronutrients
Salt Final Concentration, g/L Salt Final Concentration, mg/L
NH4Cl 1.02 FeCl2� 4H2O 1.20MgSO4� 7H2O 0.05 HBO3 0.03CaCl2� 2H2O 0.05 ZnCl2 0.03
KH2PO4 0.22 CuCl2� 2H2O 0.02NaHCO3 1.5 MnCl2� 4H2O 0.30NaCl Variable* (NH4)6Mo7O24� 4H2O 0.05
(NH4)2CO3 Variable** CoCl2� 6H2O 1.20NiCl2� 6H2O 0.03
EDTA 0.60HCl, mM 0.007Resazurin 0.30
Na2SeO3� 5H2O 0.06
D. Sudmalis et al. / Water Research 147 (2018) 142e151144
(Bioprocess Control, Sweden) at a stirring speed of 30 rpm (30 s onand off) and at 35 �C. Due to the overall lower activity of the full-scale as compared to the laboratory grown biomass, the workingvolume of batch experiments was set to 200mL. Additionally, thebiomass concentration for batch experiments was chosen accord-ingly andwas set to 2.5 and 5 g VSS/L for Do and Li, respectively. TheCOD concentration during experiments was the same as for R5 andR20, namely 4 g/L.
The level of sodium in each experiment was adjusted by varyingthe NaCl concentration (VWR, minimum purity 98%) of nutrientmedium as presented in Table 1. Sodium originating from sodiumacetate was taken into account for the final concentration calcula-tions. Parallel to all batch experiments, blank experiments wereperformed at corresponding salinity without the addition of COD tocorrect for biogas production originating from the inoculum. Theheadspace before each batch experimentwas flushedwith nitrogengas. An overview, containing all of the batch experiments in thisstudy and the main parameters, is presented in Table 2.
Previously it was demonstrated that the dominant activemethanogen was the acetoclastic Methanosaeta in both R5 and R20granular sludge (Gagliano et al., 2018; Sudmalis et al., 2018).Therefore, acetate was chosen as the main electron donor.Furthermore, data from one of the full-scale IC reactors operatorrevealed, that the reactor was operated at 7.6± 3.8 meq/L ofeffluent volatile fatty acids (VFAs), potentially requiring additionalacetoclastic activity for conversion into methane.
Table 2Summary of performed batch experiments. Preconditioning salinity e salinity at which bsalinity - salinity to which granular sludge was exposed for SMA determination with anduring the experiment.
Effectiveness of potential osmo-protectand their precursors
R5 R20 Do
Preconditioning salinity, g Naþ/L 5 20 5Experimental salinity, g Naþ/L 20 20 10Total COD, g/L 4 4 4COD/VSS 4 4 1.6% of COD added as osmo-protectant during the experiments and conc. of Kþ in g/LLysine,% 2.1 2.1 2.5Glutamic acid, % 2.8 2.7 3.4Aspartic acid, % 1.7 1.6 2.1Tryptone, % 9.7 9.7 12.1Gelatine, % 1.7 1.7 4.4Potassium, g/L 0.72 0.72 0.36Potassium, g/L 1.44 1.44 0.72
2.2.1. Tryptone for alleviation of hypersalinity shock at high CODproportion
A preliminary experiment was performed to study if peptidescan alleviate an osmotic shock in acetoclastic methanogens. Thiswas done by exposing R5 granular sludge to an abrupt increase insalinity from 5 g Naþ/L to 20 g Naþ/L with and without a dosing oftryptone. In the batches with added tryptone, the COD consisted of75% sodium acetate and 25% tryptone. In the batches withoutadded tryptone, the COD consisted solely of sodium acetate. Similarexperiments with R5 granular sludge were performed at a constantsalinity level of 5 g Naþ/L. These experiments were performed toinvestigate if specific methanogenic activity (SMA) of R5 granularsludge would increase at a stable 5 g Naþ/L salinity with addedtryptone compared to only sodium acetate fed batches.
2.2.2. Identification of osmolytes present in 20 g Naþ/L adaptedbiomass (R20)
To identify osmolytes present in 20 g Naþ/L (R20) adaptedgranular biomass, this sludge was exposed to a sudden decrease ofsalinity down to 5 g Naþ/L with sodium acetate as the only carbonsource. After exposure to the lower salinity, liquid aliquots weresampled with an interval of 2 h. This experiment was designed tostudy if amino acids or their derivatives would be released to thebulk liquid. This would indicate their use as osmo-protectants inthe biomass (Pflüger et al., 2005). As a control for osmolyte excre-tion, batch experiments at a stable salinity of 20 g Naþ/L with so-dium acetate as the carbon source were performed.
iomass was incubated with 0.4 g/L acetate COD prior the experiment. Experimentald without added potential osmo-protectant. Total COD e initial COD concentration
ants Experiment with tryptoneadded in excess in R5
Identification ofosmo-protectants in R20
Li R5 R20
5 5 2010 20 54 4 40.8 4 4
4.5 e e
6.7 e e
4.2 e e
23.8 25 e
8.5 e e
0.36 e e
0.72 e e
D. Sudmalis et al. / Water Research 147 (2018) 142e151 145
To study the specific methanogenic activity (SMA) upon a sud-den exposure to hyposalinity as well as at stable salinity of 20 gNaþ/L, batch experiments for biogas productionwere run in parallelto those for osmolyte analysis.
2.2.3. Potential osmolytes and their ability to alleviate hypersalinityshocks
The third set of experiments was used to study the SMA ofgranular sludge fed with acetate along with addition of a potentialosmo-protectant precursor upon exposure to hypersalinity of 10 or20 g Naþ/L. The selection of osmo-protectant precursors is detailedin the following steps. Firstly, osmo-protectants were detected inlaboratory scale granules adapted to 20 g Naþ/L and their biosyn-thesis precursors were chosen to study as the model compounds.Secondly, the addition of tryptone and gelatine was chosen toassess if peptides and proteins can also alleviate hypersalinityshock in granular sludge. Finally, potassiumwas chosen as a knowntransition element upon exposure of anaerobic microorganisms tohypersalinity shocks (Martin et al., 2000).
The selected compounds were: tryptone (Sigma Aldrich-NewZealand), gelatine (Type B from bovine bone, Merck-UnitedKingdom), glutamic acid (Merck-Germany), aspartic acid (TCIchemicals-Japan), lysine (Merck-France), and potassium (Merck-Germany).
To shorten the lag phase of these experiments, all batches werepreconditioned for 1e3 days at an acetate COD level of 0.4 g/L in75% of the final working volume. The salinity level during pre-conditioning was set to 5 g Naþ/L for R5, Do and Li granular sludgeand to 20 g Naþ/L for R20 granular sludge. After the preconditioningperiod, additional nutrient medium, sodium acetate and potentialosmo-protectants were spiked, resulting in a final COD concentra-tion of 4 g/L and salinity levels of 10 g Naþ/L (Do, Li) and 20 g Naþ/L(R5, R20).
A lower salinity for ‘full scale granules’ Do and Li was chosenafter preliminary experiments had shown almost no activity whenthese were exposed to 20 g Naþ/L. The nitrogen concentrationoriginating from addition of tryptone, gelatine, glutamic acid,aspartic acid, or lysine was compensated in batches fed with onlyacetate. This was done through the dosing of (NH4)2CO3 in thenutrient medium, shown in Table 1.
2.3. Estimation of applicable osmo-protectant concentration forhypersalinity shock experiments
Since Naþ and Cl� ions were by far the predominant ions in theexperiments, the amount of osmo-protectant to dose in the thirdset of batch tests was estimated based on the concentration of thesetwo ions. For example, the concentration of Naþ and Cl� at 20 gNaþ/L equals 0.87mol/L and 0.86mol/L, respectively. It wasassumed, that the density of the biomass is close to 1 g/cm3 and thefraction of bacteria accounts for about 40% of VSS in the granularsludge (m/m) (Chung and Neethling, 1990). The amount of osmo-protectant required to balance the pressure of Naþ and Cl� wouldbe 0.69mmol per gram VSS if it is assumed that the efficiency ofuptake for these molecules is 100%. This calculation was not cor-rected for the fact that osmo-protectants can carry more than onecharge at physiological pH (McNeil et al., 1999).
In the case of tryptone, it was assumed that after hydrolysis intoamino acids, mostly glutamic acid acts as the osmo-protectantprecursor (Roberts, 2005). In fact, tryptone is a pancreatic digestof casein, where glutamic acid accounts for around 20.2% of aminoresidues (Smith and Friedman, 1984). This was assumed to besimilar in tryptone.
In the case of gelatine, the amount to dose was estimated basedon the nitrogen content of gelatine, assuming that 1mol of nitrogen
corresponds to 1mol of osmo-protectant. Potassium was dosedboth at seawater Na/K ratio (w/w) of 27.8 (Zhu et al., 2004) anddouble this ratio resulting in a Na/K of 13.9 (w/w).
2.4. Identification of osmolytes as amino acids
The liquid samples were filtered through 0.45 mm acetate cel-lulose membrane filters (VWR® Syringe Filters). 0.5mL of thefiltrate was mixed with 0.4mL of methanol and 0.1mL of internalstandard (0.4mM Norleucine) and centrifuged to remove proteins.High Performance Liquid Chromatography (HPLC) was performedwith Thermo Fisher (Dionex) Ultrimate 3000 HPLC equipped withThermo Fischer TCC3000 column compartment, Variable wave-length detector 3400RS, Acquity UPLC BEH C18 1.7 mm,2.1� 150mm column (Waters 186002350) and VanGuard AcquityUPLC 1.7 mm, 2.1� 5 mm guard column (Waters 186003975). Thederivatisation procedure was based on (Hanczko et al., 2007). Thedata was analysed using Chromeleon 7.1 software.
2.5. Identification of amino acid derivatives as osmolytes
During HPLC analysis fractions containing unknown peaks, werecollected and concentrated five times for further hydrophilicinteraction liquid chromatography (HILIC) analysis. Acetonitrile(ACN) and trifluoroacetic acid (TFA) were added to the samples toreach final concentrations of 10% ACN and 0.1% TFA. Samples werecentrifuged for 5min (19000� g, 20 �C) prior to injection. Sampleswere analysed on an Acquity UPLC System (Waters, Milford, MA,USA) using an Acquity UPLC BEH amide column (2.1� 150mm,1.7 mm particle size) with an Acquity BEH amide Vanguard pre-column (2.1� 50mm, 1.7 mm particle size). Eluent A was 1% (v/v)ACN containing 0.1% (v/v) (TFA) in Milliporewater and eluent B was100% ACN containing 0.1% (v/v) TFA. Supernatants (8 mL) wereinjected into the column at 35 �C. The elution profile used was:0e2min isocratic on 90% B; 2e3min linear gradient from 90% to80% B; 3e6min linear gradient from 80 to 65% B; 6e8min lineargradient from 65% to 60 %B; 8e10min isocratic on 40% B,10e12min linear gradient from 40% to 90% B and 12e30min iso-cratic on 90% B. The flow rate was 350 mL/min. Detection was per-formed using a PDA, which scanned the absorbance from 200 to400 nm at a 1.2 nm resolution with 20 spectra per second.
The mass spectra of the samples separated by HILIC were ob-tained with an online Synapt high definition mass spectrometerG2-Si (Waters), coupled to the UPLC system, equipped with a z-spray electrospray ionization (ESI) source, a hybrid quadrupole andan orthogonal time-of-flight (Q-TOF). The system was calibratedusing sodium iodide. The capillary voltage was set to 3 kV with thesource operation in positive ion mode and the source temperatureat 150 �C. The sample conewas operated at 35 V. Nitrogenwas usedas desolvation gas (500 �C, 800 L/h) and cone gas (200 L/h). The trapgas was set at 2 mL/min. In the first scanning of the samples, MSandMS/MS (MSe continuummethod) were performed betweenm/z 100e2000 with a scan time of 0.3 s. The trap collision energy wasset at 4 V in single MS mode and ramped from 20 to 30 V in MSemode. The transfer collision energy was set at 2 V in MS andswitched between 2 and 10 V in the MSe mode. The ion of interesthad a m/z of 189.13. To tentatively identify the compound, MS/MSmethod was performed on the fixed mass m/z 189.13. The trapcollision energy ramped from 15 to 35 V and the MS and MS/MSwere recorded for m/z 50 to 500. Online lock mass data (angio-tensin II, [Mþ2H]2þ 523.7751) were collected in parallel of themeasurements and used to correct the measured m/z. UV and MSdata were acquired using MassLynx software v 4.1 (Waters).
D. Sudmalis et al. / Water Research 147 (2018) 142e151146
2.6. Biogas composition and amount of methane
Biogas composition at the end of the batch experiments wasmeasured as described in Steinbusch et al. (2008). To quantify theamount of methane produced in the serum bottles, a pressuremeter with an absolute pressure probe (GMH3151, GreisingerElectronic, Germany) was used to follow pressure build-up in time.The amount of methane produced was calculated using the idealgas law. The amount of methane produced in AMPTS II systemwasautomatically recorded through the AMPTS web-based softwareapplication (Bioprocess Control, Sweden) and corrected for theactual biogas composition measured.
3. Results and discussion
3.1. Tryptone at high COD proportion as an effective compound toalleviate osmotic stress
Tryptone in a COD ratio tryptone/acetate of 1:4 was added to thebatches, upon exposure to an abrupt increase in salinity from 5 to20 g Naþ/L. This resulted in a clear alleviation of osmotic stress forlaboratory scale R5 granular sludge (Fig. 1-A). Methane productioncurves of only acetate fed batches compared with acetate andtryptone fed batches clearly show a strong reduction of lag-phasefrom approximately 350 h to approximately 150 h before a signifi-cant amount of methane was produced (Fig. 1-A). The maximumslope of themethane production curve increased by 86% from 0.051NmL-CH4/h in the acetate fed batch to 0.095 NmL-CH4/h in thebatch with added tryptone (Fig. 1-A). These results demonstrate apositive effect of tryptone addition to alleviate salt stress inanaerobic granules.
It is assumed that the increase of methanogenic activity ismostly related to the conversion of acetate rather than the con-version of the hydrogen produced from tryptone fermentation.Indeed, based on stoichiometry, the theoretical amount of methaneproduced from hydrogen during casein fermentation only accountsfor approximately 10% of the total produced methane (Ramsay andPullammanappallil, 2001).
The results of the experiment with stable salinity i.e., granularsludge adapted to 5 g Naþ/L (R5) and exposed to 5 g Naþ/L areshown in Fig. 1-B. The slope of the methane production curveincreased by 37% from 0.16 to 0.22 NmL-CH4/h due to the additionof tryptone. The improvement of methanogenic activity is muchsmaller compared to exposing the same granular sludge to 20 gNaþ/L (86%). However this still shows that tryptone can improvethe methanogenic activity of granular sludge that is alreadyadapted to a certain salinity level.
Fig. 1. Methane production curves during batch experiments with dosing tryptone for allevgranular sludge exposed to an abrupt increase in salinity B - R5 granular sludge exposed to ameasured values from duplicate samples.
3.2. Amino acids and their derivatives as osmo-protectants in 20 gNaþ/L adapted granular sludge
When exposed to a sudden decrease in osmotic pressure, bac-teria activate the so called mechanosensitive channels located incytoplasmic membranes. These channels allow to rapidly restorecell turgor by either selectively releasing small osmolytes or - ifmembrane tension increases - by non-selectively releasing osmo-lytes together with small proteins (Berrier et al., 1992; Pflüger et al.,2005). Archaea have also shown to possess such mechanosensitivechannels (Kloda and Martinac, 2002). Since the granular sludgeused in these experiments with decreasing salinity was previouslyadapted to 20 g Naþ/L, a rapid release of accumulated osmolytesupon exposure to a hyposalinity shock was expected.
HPLC chromatogram of amino acids in Figs. S1eA shows thatduring an abrupt decrease in salinity from 20 g Naþ/L to 5 g Naþ/Lglutamic acid and another compound containing amine bondswere excreted by the granular biomass. These compounds were notdetected at the stable 20 g Naþ/L salinity as shown in Figs. S1eB.The unknown compound was further analysed with HILIC coupledto ESI-Q-TOF-MS and tentatively identified as N-acetyl-b-lysine.Both molecules have been previously reported as compatible sol-utes in anaerobic microorganisms. This is true for moderate salinityenvironments, particularly in methanogenic archaea, which accu-mulate a-glutamate, b-glutamate, b-glutamine, and Nε-acetyl-b-lysine (Müller et al., 2005; Schlegel and Müller, 2011).
Uptake of osmolytes is more bio-energetically favourable thansynthesis de novo (Oren, 1999). Therefore, glutamic acid - anabundant amino residue in tryptone - can potentially be taken upby microbial cells after hydrolysis of tryptone, where under phys-iological pH it would take its anionic form of glutamate (Sowers andGunsalus, 1995). Glutamic acid in its anionic form (glutamate) cancounter balance the charge of accumulated potassium in microor-ganisms as well as balance the osmotic pressure (Roberts, 2005).One must keep in mind that glutamate can be also synthesized denovowithout exogenous addition of casamino acids (Csonka, 1989;Sowers and Gunsalus, 1995). However, given the positive effect oftryptone on methanogenic activity (Fig. 1-A), it is likely that in R20granular sludge glutamate is taken up after hydrolysis of tryptone.
Nε-acetyl-b-lysine is a compatible solute unique tomethanogenicarchaea (Pflüger et al., 2003) and some halophilic bacteria (Jogheeand Jayaraman, 2014). This osmolyte has shown to accumulate inmany methanogenic archaea such as different Methanosarcina sub-species, Methanogenium cariaci, and Methanohalophilus strain FDF1.(Sowers et al.,1990; Roberts et al.,1992; Sowers andGunsalus,1995).
In pure cultures of methanogenic archaeaMethanosarcina mazeiGӧ1 and Methanococcus maripaludis it was demonstrated that the
iation of osmotic stress. NmL e methane volume normalized to 0 �C and 1 atm. A e R5stable 5 g Naþ/L salinity. Error bars in the graphs indicate the maximum and minimum
D. Sudmalis et al. / Water Research 147 (2018) 142e151 147
synthesis of Nε-acetyl-b-lysine starts with a-lysine production frompyruvate and aspartic acid semi-aldehyde (Pflüger et al., 2003;Schlegel and Müller, 2011). Tryptone contains both lysine (6.96mol%) and aspartic acid (6.87mol %) (Smith and Friedman, 1984). It isexpected that bothmolecules may have acted as potential substratefor Nε-acetyl-b-lysine synthesis. In fact, it appears that most of thenitrogen containing osmolytes are synthesized either via asparticacid pathway or via the glutamic acid pathway (Galinski andTrüper, 1994). This suggests that proteins or peptides containingthese amino acids have the potential to provide microorganismswith the “building blocks” of osmolytes.
In our recent work (Gagliano et al., 2018) we showed that thedominant microorganism in granules sampled from the R20 labo-ratory - scale reactor was the acetoclastic Methanosaeta har-undinacea. Parallel with our findings, the in silico analysis of thecomplete genome of M. harundinacea strain 6Ac showed that thismicroorganism possesses a series of geneswhich products aremostlikely involved in the pathway for the synthesis of Nε-acetyl-b-lysine, both from aspartate or lysine (Fig. S2) (Gagliano et al., 2018).In particular, two locus encode for the L-lysine 2,3-aminomutaseenzyme (Fig. S2), which activity is essential for the biosynthesisof this osmolyte (Saum et al., 2009).
Fig. 2 illustrates that the sudden decrease in salinity down to 5 gNaþ/L gave a steeper slope of the methane production curve ascompared to R20 granular sludge exposed to the adapted salinity of20 g Naþ/L (Fig. 2). This result is logical because response tohyposalinity shock both in bacteria and archaea has been shown tobe a mechanistic reaction (Berrier et al., 1992; Kloda and Martinac,2002) and does not interfere with energy metabolism. The increasein activity can be explained with lower energy requirements tomaintain cell turgor at decreased salinity. For practical applications,this means that problems with granular sludge activity are ex-pected with a rapid increase (not decrease) in salinity.
3.3. Effect of amino acids on SMA of laboratory-scale and full-scalederived biomass
Three amino acids were tested for their potential to alleviateosmotic stress of full-scale and laboratory-scale granular sludge.The first two are potential precursors for production of Nε-acetyl-b-lysine, namely aspartic acid and lysine (Pflüger et al., 2003). Thethird was glutamic acid e a precursor for glutamate found in R20biomass. N-acetyl-b-lysine was not considered in this study due toits high price compared to the other tested molecules.
Fig. 3 shows that glutamic acid and aspartic acid were bothefficient in alleviating osmotic stress with all tested granularsludges. This is an important observation because it shows thatboth osmo-protectants themselves and also their precursors can
Fig. 2. Methane production curves during the decreasing salinity experiments with20 g Naþ/L adapted (R20) granular sludge. NmL e volume normalized to 0 �C and1 atm. Error bars indicate maximum and minimum measured values in duplicatesamples.
indirectly alleviate osmotic stress. Similarly, evidence of a positiveeffect of osmolyte precursors was recently shown in Bacillus subtilis,which imports proteogenic amino acids and converts them intoproline to enhance its growth under otherwise osmotically unfav-ourable conditions (Zaprasis et al., 2014).
The improvement of SMA by dosing glutamic acid and asparticacid (Fig. 3-A,B) also demonstrates the efficiency of osmolytes otherthan commonly used glycine betaine (Vyrides and Stuckey, 2016).
The addition of glutamic acid was more efficient in alleviatingsalt stress than aspartic acid (Fig. 3-A,B). This is in line with Pflügeret al. (2003), who found that in Methanococcus maripaludis JJglutamate is themain osmo-protectant up to 376mMNaCl (approx.10 g Naþ/L) and within microbial cells its concentration is unaf-fected up to salinity levels of 800mM NaCl (approx. 20 g Naþ/L).The dominant osmolyte at 800mM was Nε-acetyl-b-lysine (Pflügeret al., 2003). Upon increasing salinity, a similar pattern of osmolyteswas found in eight strains representing five species of Meth-anosarcina spp. (Sowers and Gunsalus, 1995). Aspartate is known asa precursor for biosynthesis of Nε-acetyl-b-lysine. However, gluta-mate can act as an osmolyte directly within the entire range ofinvestigated salinities in this study and, therefore, the higherobserved effect of dosing of glutamic acid. Moreover, full-scalegranular sludge was exposed to 10 g Naþ/L which is a salinitylevel at which glutamate has been shown as predominant osmolyte(Pflüger et al., 2003).
Addition of glutamic acid and aspartic acid had a more pro-nounced effect on Li and Do granules from full scale ICs than onlaboratory scale R5 and R20 granular sludge. Compared to acetatefed granular sludge, glutamic acid increased the SMA by 115%, 35%and 9% in Li, Do and R20, respectively (Fig. 3-A). Similarly, asparticacid increased the SMA by 72%, 26%, 12% and 7% in Li, Do, R5 andR20, respectively (Fig. 3-B). Because the amount of added aminoacids was calculated based on the VSS amount used in the exper-iment, more glutamic and aspartic acid was added to full-scalegranular sludge Do and Li compared to R5 and R20 (Table 2). Thisdifference may partially explain the general trend of a strongerpositive effect of glutamic acid and aspartic acid on Do and Li sludgeas compared to R5 and R20. The larger difference of osmo-protectant concentration between granules and bulk liquid canfavour faster transport in the granules by diffusion, thereby alsoallowing for a faster and more pronounced positive effect. Addi-tionally, R5 granular sludge was exposed to a salinity increase from5 g Naþ/L to 20 g Naþ/L compared to the case of Do and Li where theincrease was from 5 to 10 g Naþ/L. The combination of R5 granulesexposure to higher salinity (as compared to Do and Li) withpotentially slower diffusion of amino acids due to the concentrationgradient may have contributed to the less pronounced effect ofglutamic acid and aspartic acid on the SMA. The lowest effect onR20 granular sludge can be explained by the fact that this sludgehad already been adapted to 20 g Naþ/L for almost two years.
Fig. 3eC shows that the addition of lysine gave a negative effecton the SMA of all tested granules, as already observed in anaerobicbatch experiments by Wagner et al. (2012). The decrease in SMAwas 6%, 24% and 14.6% for R20, Li and Do, respectively. No expla-nation for this result could be found.
3.4. Effect of tryptone and gelatine on SMA of laboratory-scale andfull-scale biomass
Tryptone and gelatine both had a positive effect on the SMA forall granular sludges tested, with tryptone giving a higher improve-ment of SMA compared to gelatine (Fig. 4). The average increase ofSMA by addition of tryptonewas 36%,17%, 179% and 48% for R5, R20,Li and Do, respectively. The addition of gelatine resulted in an SMAincrease of 30%,14%, 23% and 13% for R5, R20, Li andDo, respectively.
Fig. 3. Effect of added amino acids on SMA of full-scale and laboratory-scale granular sludge, when exposing the biomass to hypersalinity shocks. The error bars indicate theminimum and maximum measured effect with respect to control samples without additional dosing of amino acids.
Fig. 4. Effect of added tryptone and gelatine on SMA of full-scale and laboratory-scale biomass, when exposing the biomass to hypersalinity shocks. The error bars indicate theminimum and maximum measured effect with respect to control samples without additional dosing of these compounds.
D. Sudmalis et al. / Water Research 147 (2018) 142e151148
The more pronounced improvement of SMA with tryptonecompared to gelatine (Fig. 4) can be explained in several ways.Firstly, gelatine needs to undergo more hydrolysis steps comparedto tryptone before its amino residues can be taken up by microbialcells. This may be a rate limiting step. This also implies gelatinewould be less efficient than tryptone to alleviate osmotic stress
upon abrupt increase in osmotic pressure.Secondly, tryptone was added in a higher concentration
compared to gelatine (See Table 2). This was done because the doseof tryptone was based only on its glutamic acid content whereasthe dose of gelatine was based on its nitrogen content. As a result,potentially more osmolyte precursors in a form of other amino
D. Sudmalis et al. / Water Research 147 (2018) 142e151 149
acids were supplied by tryptone. Thirdly, tryptone also differs fromgelatine in its amino acid residue composition. The most abundantamino acid sequences in gelatine are triplets of glycine, proline andhydroxyproline (Karim and Bhat, 2009), whereas the most abun-dant amino residue in tryptone is glutamic acid (18.5%) (Smith andFriedman, 1984). Tryptone also contains more (6.9%) aspartic acidcompared to gelatine (5%) (Smith and Friedman, 1984; Karim andBhat, 2009). As showed previously (Fig. 3-A,B), both glutamic acidas well as aspartic acid are efficient in alleviating osmotic stress.Therefore, tryptone can potentially provide more of these aminoacids as osmo-protectants or as osmo-protectant precursorscompared to gelatine.
The positive effect of a peptide such as tryptone and a proteinsuch as gelatine on the activity of anaerobic biomass, when exposedto high salinity is a very important outcome in this study. It isimportant because addition of a cheap (waste) protein can beeconomically more viable than addition of osmolytes.
Additionally, a number of saline (waste)waters, (e.g., cheese,olive oil milling, and fish industry) contain proteins (Paredes et al.,1999; Ramsay and Pullammanappallil, 2001). In fact, these saline(waste)water streams have been shown to be anaerobically treat-able without dosing of osmolytes (Guti�errez et al., 1991; Khoufiet al., 2006; Chowdhury et al., 2010). In literature, no adequateexplanations were yet reported for sufficient microbial activity toachieve high efficiencies of organics conversion when treatingthese saline streams. Our results show that the presence of certainproteins, which after hydrolyses can be taken up as osmo-activeorganic chemicals in the granules, probably play an importantrole in salt stress alleviation when treating these types of (waste)waters.
3.5. Effect of potassium on SMA of laboratory-scale and full-scalebiomass
Fig. 5 indicates that the addition of potassium at sea water Na/Kratio (27.8 w/w) resulted in a slight positive effect on SMA in Do andLi granules. With the “laboratory-scale granules” no clear positiveeffect of potassiumwas detected. Potassium at double the seawaterNa/K ratio (13.9 w/w) gave no positive effect on any of the sludges.At the salinity levels investigated in our study (10e20 g Naþ/L)organic osmolytes are predominant within non-halophilic meth-anogenic archaea. Since potassium only alleviates a part of theosmotic pressure (Sowers and Gunsalus, 1995), the low effect ofpotassium can be explained.
Gagliano et al. (2017) reported that potassium can alleviate the
Fig. 5. Effect of added potassium on SMA of full-scale and laboratory-scale biomass,when exposing the granular sludge to hypersalinity shocks. The error bars indicate theminimum and maximum measured effect with respect to control samples withoutadditional dosing of potassium. Potassium 1� e Na/K ratio of 27.8 (w/w); Potassium2� e Na/K ratio of 13.9 (w/w).
osmotic stress in a continuously operated anaerobic system at 20 gNaþ/L at a seawater Na/K ratio. In batch experiments with a Na/Kratio higher than sea water (359 w/w), a positive effect was onlyobserved after approximately 3 weeks of incubation at 14 g Naþ/L(Vyrides et al., 2010).
In the current study the maximum applied incubation time was10 days. Potassium had no positive effect during the first 10 days ofbatch incubation.
3.6. The potential of precursor compounds as osmo-protectants foranaerobic treatment in practice
Vyrides and Stuckey (2016) recently reported the bottleneck forapplication of osmolytes for (waste)water treatment is the costs ofthese molecules. In this contribution we demonstrated the poten-tial of precursor compounds, such as proteins and amino acids, toalleviate osmotic stress in methanogenic archaea. However, dosingof proteins or amino acids for anaerobic treatment of saline (waste)waters still may not be an economically viable option. However, theexisting saline (waste)waters containing proteins, for example, canbe identified and potentially treated anaerobically. Moreover, whenpossible, mixing (waste)water streams can be considered to avoidexternal dosing of osmolytes. Vlyssides et al. (2004) suggested re-covery of all marketable compounds from olive oil milling (waste)water to make the olive oil production more economically viable.Our findings suggest that not all proteins from existing (waste)water streams should be recovered if anaerobic treatment is to beapplied.
There is a large potential for anaerobic biological treatmentapplication in the saline (waste)water domain. In the EU, the annualfish catch in 2015 was reported to be 5.14 million tonnes (Eurostat,2018c). The annual production of olive oil in 2017 for the EU was10.2 million tonnes. 91% of this was produced in Greece, Spain, andItaly (Eurostat, 2018b). Similarly, within the EU approximately 56million tonnes of milk were used to produce cheese in 2016(Eurostat, 2018a). The large amounts of goods produced by theseindustries are expected to generate even larger quantities of saline(waste)water. As an example, production of 1 kg of olive oil gen-erates around 7.5 kg of saline (waste)water (Vlyssides et al., 2004).This means that, during harvest time, around 70 million tonnes ofheavily polluted saline (waste)water needs to be treated in Greece,Spain, and Italy. Anaerobic treatment technology could become anattractive alternative for treatment of these (waste)waters.
Fish processing and olive oil industry wastewaters are known tocontain high amounts of suspended solids (Paredes et al., 1999;Palenzuela-Rollon et al., 2002). Such (waste)waters are not suitablefor direct treatment with granular sludge reactors. However, floc-culent UASB reactors can be applied instead. If granular sludgeUASB is the desired treatment technology, suspended solids, suchas lipids, need to be removed prior the anaerobic treatment(Palenzuela-Rollon et al., 2002). Most of the suspended solids infishery wastewater are proteins and lipids (Palenzuela-Rollon et al.,2002). This means that only dissolved proteins would remain in the(waste)water after removing these solids. Further anaerobic treat-ment with granular sludge reactors would then only be possible if asufficient amount of dissolved proteins remain in the (waste)water.Similar considerations as for fish processing (waste)water shouldalso be made for other (waste)waters with high suspended solidscontent.
Within the microbial consortia of anaerobic reactors, aceto-clastic methanogens are often mentioned as the most sensitivemicroorganisms to high salinity (Feijoo et al., 1995; Chen et al.,2008; Gagliano et al., 2017). As such, this study focuses only onmethanogenic activity. However, effective anaerobic treatment ofcomplex wastewater is achieved by a diverse microbial community
D. Sudmalis et al. / Water Research 147 (2018) 142e151150
(O'Flaherty et al., 2006) and all microorganisms involved in thebiodegradation process should be able to cope with the osmoticstress for successful treatment. This means that both archaea andbacteria need to synthesize/accumulate osmo-protectants. A po-tential advantage of precursor compounds like tryptone and gela-tine for osmo-protection lies in their chemical composition. Afterhydrolysis, potentially diverse groups of microorganisms can besuppliedwith substrates for osmolyte synthesis. In our recentwork,successful sludge granulation in UASB reactors, treatingwastewaterwith a mixture of glucose, acetate, and tryptone at salinity levels ashigh as 20 g Naþ/L was shown (Sudmalis et al., 2018). The highbacterial and methanogenic activity within these UASB reactors isexplained by the presence of tryptone in the influent, which mayhave indirectly supplied the microbial community with precursorsof osmolytes. Further research should be targeted to study the ef-fects of compounds like gelatine and tryptone on specific microbialgroups driving granules formation, upon exposure to osmoticstress. Furthermore, it is important to understand shifts in micro-bial communities when different osmoprotectant precursors/sub-strates are used during long-term continuous flow studies, becausemicrobial communities can explain effectiveness of different pre-cursor compounds of osmolytes as well as potentially reveal the keymicroorganism driving the granulation process.
4. Conclusions
The results obtained in these batch experiments allow us toconclude the following:
� Glutamic acid and N-acetyl-b-lysine are the sole osmo-protectants in anaerobic granular biomass, grown in a UASBon a mixture of glucose, acetate, and tryptone (3:2:1 on CODbasis) and adapted to 20 g Naþ/L;
� Glutamic acid and aspartic acid, as well as the non-hydrolysedtryptone and gelatine, can alleviate salt toxicity of metha-nogens in anaerobic granular sludge;
� Lysine cannot alleviate salt toxicity of methanogens in anaerobicgranular sludge;
� Potassium added at seawater ratio can slightly alleviate osmoticstress of methanogens in anaerobic granular sludge during in-cubation period of 10 days;
� Potassium added at double the sea water ratio cannot alleviatesalt stress of methanogens in anaerobic granular sludge duringincubation period of 10 days;
� When adequate amounts and types of proteins are present inthe (waste)water, streams with high salt content could beanaerobically treated without external dosing of osmolytes;
� Hyposalinity shock from 20 g Naþ/L down to 5 g Naþ/L does notnegatively affect methanogenic activity in 20 g Naþ/L adaptedgranular sludge
Acknowledgements
This research is financed by the Netherlands Organisation forScientific Research (NWO), which is partly funded by the Ministryof Economic Affairs and Climate Policy, by the Ministry of Infra-structure and Water Management and partners of the Dutch WaterNexus consortium (project nr. STW 14300 Water Nexus 2.1).Additionally we would like to thank Susan Witte for amino acidanalysis and the technical and analytical support team of the Sub-department of Environmental Technology (ETE) Vinnie de Wilde,Bert Willemsen, Jean Slangen, Livio Carlucci, Katja Grolle, PieterGremmen and Hans Beijleveld. A special thank you to Peter Wier-enga for the fruitful discussions and cooperation in identifying theosmolytes and to Jaap Vogelaar (Paques B.V.) for setting up an
internship project of Millah within ETE. Finally, we would like tothank Jessica Wreyford for proof reading the manuscript andhelping with improving our English.
Appendix A. Supplementary data
Supplementary data to this article can be found online athttps://doi.org/10.1016/j.watres.2018.09.059.
References
Berrier, C., Coulombe, A., Szabo, I., Zoratti, M., Ghazi, A., 1992. Gadolinium ion in-hibits loss of metabolites induced by osmotic shock and large stretch-activatedchannels in bacteria. FEBS J. 206 (2), 559e565. https://doi.org/10.1111/j.1432-1033.1992.tb16960.x.
Chen, Y., Cheng, J.J., Creamer, K.S., 2008. Inhibition of anaerobic digestion process: areview. Bioresour. Technol. 99 (10), 4044e4064. https://doi.org/10.1016/j.biortech.2007.01.057.
Chowdhury, P., Viraraghavan, T., Srinivasan, A., 2010. Biological treatment processesfor fish processing wastewater e a review. Bioresour. Technol. 101 (2), 439e449.https://doi.org/10.1016/j.biortech.2009.08.065.
Chung, Y.-C., Neethling, J., 1990. Viability of anaerobic digester sludge. J. Environ.Eng. 116 (2), 330e343. https://doi.org/10.1061/(ASCE)0733-9372(1990)116:2(330).
Csonka, L.N., 1989. Physiological and genetic responses of bacteria to osmotic stress.Microbiol. Rev. 53 (1), 121e147.
De Vrieze, J., Coma, M., Debeuckelaere, M., Van der Meeren, P., Rabaey, K., 2016a.High salinity in molasses wastewaters shifts anaerobic digestion to carboxylateproduction. Water Res. 98, 293e301. https://doi.org/10.1016/j.watres.2016.04.035.
De Vrieze, J., Regueiro, L., Props, R., Vilchez-Vargas, R., J�auregui, R., Pieper, D.H.,Lema, J.M., Carballa, M., 2016b. Presence does not imply activity: DNA and RNApatterns differ in response to salt perturbation in anaerobic digestion. Bio-technol. Biofuels 9 (1), 244. https://doi.org/10.1186/s13068-016-0652-5.
Di Gioia, L., Guilbert, S., 1999. Corn protein-based thermoplastic resins: effect ofsome polar and amphiphilic plasticizers. J. Agric. Food Chem. 47 (3), 1254e1261.https://doi.org/10.1021/jf980976j.
Eurostat, 2018a. What is the milk in the EU used for? [updated 2018.01.29; cited2018.05.31]. Available from: http://ec.europa.eu/obic biological treatmentapplication in the saline (waste)water domaineurostat/en/web/products-eurostat-news/-/DDN-20180129-1.
Eurostat, 2018b. Olives by production [updated 2018.05.30.; cited 2018.05.31].Available from: http://ec.europa.eu/eurostat/tgm/table.do?tab¼table&init¼1&plugin¼1&language¼en&pcode¼tag00122.
Eurostat, 2018c. Catches in all fishing regions [updated 2018.05.03; cited2018.05.31]. Available from: http://ec.europa.eu/eurostat/tgm/table.do?tab¼table&init¼1&plugin¼1&language¼en&pcode¼tag00076.
Feijoo, G., Soto, M., M�endez, R., Lema, J.M., 1995. Sodium inhibition in the anaerobicdigestion process: antagonism and adaptation phenomena. Enzym. Microb.Technol. 17 (2), 180e188. https://doi.org/10.1016/0141-0229(94)00011-F.
Gagliano, M., Ismail, S., Stams, A., Plugge, C., Temmink, H., Van Lier, J., 2017. Biofilmformation and granule properties in anaerobic digestion at high salinity. WaterRes. 121, 61e71. https://doi.org/10.1016/j.watres.2017.05.016.
Gagliano, M.C., Neu, T.R., Kuhlicke, U., Sudmalis, D., Temmink, H., Plugge, C.M., 2018.EPS glycoconjugate profiles shift as adaptive response in anaerobic microbialgranulation at high salinity. Front. Microbiol. 9, 1423. https://doi.org/10.3389/fmicb.2018.01423.
Galinski, E.A., Trüper, H.G., 1994. Microbial behaviour in salt-stressed ecosystems.FEMS (Fed. Eur. Microbiol. Soc.) Microbiol. Rev. 15 (2-3), 95e108. https://doi.org/10.1111/j.1574-6976.1994.tb00128.x.
Giustinianovich, E.A., Campos, J.-L., Roeckel, M.D., Estrada, A.J., Mosquera-Corral, A.,Val del Río, �A., 2018. Influence of biomass acclimation on the performance of apartial nitritation-anammox reactor treating industrial saline effluents. Che-mosphere 194, 131e138. https://doi.org/10.1016/j.chemosphere.2017.11.146.
Guti�errez, J.L.R., Encina, P.A.G., Fdz-Polanco, F., 1991. Anaerobic treatment of cheese-production wastewater using a UASB reactor. Bioresour. Technol. 37 (3),271e276. https://doi.org/10.1016/0960-8524(91)90194-O.
Hanczko, R., J�ambor, A., Perl, A., Molnar-Perl, I., 2007. Advances in the o-phtha-laldehyde derivatizations: comeback to the o-phthalaldehydeeethanethiol re-agent. J. Chromatogr. A 1163 (1e2), 25e42. https://doi.org/10.1016/j.chroma.2007.06.013.
He, Q., Kong, X.-J., Chai, H.-X., Fan, M.-Y., Du, J., 2012. Efficiency influence of exog-enous betaine on anaerobic sequencing batch biofilm reactor treating highsalinity mustard tuber wastewater. Environ. Technol. 33 (15), 1695e1699.https://doi.org/10.1080/09593330.2011.643317.
Ismail, S., Gonzalez, P., Jeison, D., Van Lier, J., 2008. Effects of high salinity waste-water on methanogenic sludge bed systems. Water Sci. Technol. 58 (10),1963e1970. https://doi.org/10.2166/wst.2008.528.
Jeison, D., Rio, A.D., Lier, J.V., 2008. Impact of high saline wastewaters on anaerobicgranular sludge functionalities. Water Sci. Technol. 57 (6), 815e820. https://doi.org/10.2166/wst.2008.098.
D. Sudmalis et al. / Water Research 147 (2018) 142e151 151
Joghee, N.N., Jayaraman, G., 2014. Metabolomic characterization of halophilic bac-terial isolates reveals strains synthesizing rare diaminoacids under salt stress.Biochimie 102, 102e111. https://doi.org/10.1016/j.biochi.2014.02.015.
Karim, A.A., Bhat, R., 2009. Fish gelatin: properties, challenges, and prospects as analternative to mammalian gelatins. Food Hydrocolloids 23 (3), 563e576.https://doi.org/10.1016/j.foodhyd.2008.07.002.
Khoufi, S., Aloui, F., Sayadi, S., 2006. Treatment of olive oil mill wastewater bycombined process electro-Fenton reaction and anaerobic digestion. Water Res.40 (10), 2007e2016. https://doi.org/10.1016/j.watres.2006.03.023.
Kloda, A., Martinac, B., 2002. Common evolutionary origins of mechanosensitive ionchannels in Archaea, Bacteria and cell-walled Eukarya. Archaea 1 (1), 35e44.https://doi.org/10.1155/2002/419261.
Le Borgne, S., Paniagua, D., Vazquez-Duhalt, R., 2008. Biodegradation of organicpollutants by halophilic bacteria and archaea. J. Mol. Microbiol. Biotechnol. 15(2e3), 74e92. https://doi.org/10.1159/000121323.
Lefebvre, O., Moletta, R., 2006. Treatment of organic pollution in industrial salinewastewater: a literature review. Water Res. 40 (20), 3671e3682. https://doi.org/10.1016/j.watres.2006.08.027.
Liu, X.-W., Sheng, G.-P., Yu, H.-Q., 2009. Physicochemical characteristics of microbialgranules. Biotechnol. Adv. 27 (6), 1061e1070. https://doi.org/10.1016/j.biotechadv.2009.05.020.
Liu, Y., Boone, D.R., 1991. Effects of salinity on methanogenic decomposition. Bio-resour. Technol. 35 (3), 271e273. https://doi.org/10.1016/0960-8524(91)90124-3.
Lu, Y., Slater, F., Bello-Mendoza, R., Batstone, D.J., 2013. Shearing of biofilms enablesselective layer based microbial sampling and analysis. Biotechnol. Bioeng. 110(10), 2600e2605. https://doi.org/10.1002/bit.24947.
Martin, D.D., Ciulla, R.A., Robinson, P.M., Roberts, M.F., 2000. Switching osmolytestrategies: response of Methanococcus thermolithotrophicus to changes inexternal NaCl. Biochim. Biophys. Acta Gen. Subj. 1524 (1), 1e10. https://doi.org/10.1016/S0304-4165(00)00131-8.
McHugh, S., O'reilly, C., Mahony, T., Colleran, E., O'flaherty, V., 2003. Anaerobicgranular sludge bioreactor technology. Rev. Environ. Sci. Biotechnol. 2 (2e4),225e245. https://doi.org/10.1023/B:RESB.0000040465.45300.97.
McNeil, S.D., Nuccio, M.L., Hanson, A.D., 1999. Betaines and related osmoprotectants.Targets for metabolic engineering of stress resistance. Plant Physiol. 120 (4),945e949. https://doi.org/10.1104/pp.120.4.945.
Müller, V., Spanheimer, R., Santos, H., 2005. Stress response by solute accumulationin archaea. Curr. Opin. Microbiol. 8 (6), 729e736. https://doi.org/10.1016/j.mib.2005.10.011.
O'flaherty, V., Lens, P., De Beer, D., Colleran, E., 1997. Effect of feed composition andupflow velocity on aggregate characteristics in anaerobic upflow reactors. Appl.Microbiol. Biotechnol. 47 (2), 102e107. https://doi.org/10.1007/s002530050896.
O'Flaherty, V., Collins, G., Mahony, T., 2006. The microbiology and biochemistry ofanaerobic bioreactors with relevance to domestic sewage treatment. Rev. En-viron. Sci. Biotechnol. 5 (1), 39e55. https://doi.org/10.1007/s11157-005-5478-8.
Oren, A., 1999. Bioenergetic aspects of halophilism. Microbiol. Mol. Biol. Rev. 63 (2),334e348. http://www.ncbi.nlm.nih.gov/pubmed/10357854.
Palenzuela-Rollon, A., Zeeman, G., Lubberding, H., Lettinga, G., Alaerts, G., 2002.Treatment of fish processing wastewater in a one-or two-step upflow anaerobicsludge blanket (UASB) reactor. Water Sci. Technol. 45 (10), 207e212. https://doi.org/10.2166/wst.2002.0332.
Paredes, C., Cegarra, J., Roig, A., S�anchez-Monedero, M.A., Bernal, M.P., 1999. Char-acterization of olive mill wastewater (alpechin) and its sludge for agriculturalpurposes. Bioresour. Technol. 67 (2), 111e115. https://doi.org/10.1016/S0960-8524(98)00106-0.
Pflüger, K., Baumann, S., Gottschalk, G., Lin, W., Santos, H., Müller, V., 2003. Lysine-2,3-aminomutase and b-lysine acetyltransferase genes of methanogenic archaeaare salt induced and are essential for the biosynthesis of Nε-acetyl-b-lysine andgrowth at high salinity. Appl. Environ. Microbiol. 69 (10), 6047e6055. https://doi.org/10.1128/AEM.69.10.6047-6055.2003.
Pflüger, K., Wieland, H., Müller, V., 2005. Osmoadaptation in methanogenic archaea:recent insights from a genomic perspective. In: Adaptation to Life at High SaltConcentrations in Archaea, Bacteria, and Eukarya. Springer, Dordrecht,pp. 239e251.
Ramsay, I.R., Pullammanappallil, P.C., 2001. Protein degradation during anaerobicwastewater treatment: derivation of stoichiometry. Biodegradation 12 (4),247e256. https://doi.org/10.1023/A:1013116728817.
Rinzema, A., van Lier, J., Lettinga, G., 1988. Sodium inhibition of acetoclasticmethanogens in granular sludge from a UASB reactor. Enzym. Microb. Technol.10 (1), 24e32. https://doi.org/10.1016/0141-0229(88)90094-4.
Roberts, M., Lai, M., Gunsalus, R., 1992. Biosynthetic pathways of the osmolytes Nepsilon-acetyl-beta-lysine, beta-glutamine, and betaine in Methanohalophilusstrain FDF1 suggested by nuclear magnetic resonance analyses. J. Bacteriol. 174
(20), 6688e6693. https://doi.org/10.1128/jb.174.20.6688-6693.1992.Roberts, M.F., 2005. Organic compatible solutes of halotolerant and halophilic mi-
croorganisms. Saline Syst. 1 (1), 5. https://doi.org/10.1186/1746-1448-1-5.Roeßler, M., Müller, V., 2001. Osmoadaptation in bacteria and archaea: common
principles and differences. Environ. Microbiol. 3 (12), 743e754. https://doi.org/10.1046/j.1462-2920.2001.00252.x.
Rozzi, A., Malpei, F., Bonomo, L., Bianchi, R., 1999. Textile wastewater reuse innorthern Italy (Como). Water Sci. Technol. 39 (5), 121e128. https://doi.org/10.1016/S0273-1223(99)00093-1.
Saum, R., Mingote, A., Santos, H., Müller, V., 2009. A novel limb in the osmoregu-latory network of Methanosarcina mazei G€o1: Nε-acetyl-b-lysine can besubstituted by glutamate and alanine. Environ. Microbiol. 11 (5), 1056e1065.https://doi.org/10.1111/j.1462-2920.2008.01826.x.
Schlegel, K., Müller, V., 2011. Osmoadaptation in methanogenic archaea: physiology,genetics, and regulation in Methanosarcina mazei G€o1. In: ExtremophilesHandbook. Springer, Dordrecht, pp. 327e342.
Smith, G., Friedman, M., 1984. Effect of carbohydrates and heat on the amino acidcomposition and chemically available lysine content of casein. J. Food Sci. 49 (3),817e820. https://doi.org/10.1111/j.1365-2621.1984.tb13219.x.
Sowers, K.R., Robertson, D.E., Noll, D., Gunsalus, R.P., Roberts, M.F., 1990. N epsilon-acetyl-beta-lysine: an osmolyte synthesized by methanogenic archaebacteria.Proc. Natl. Acad. Sci. Unit. States Am. 87 (23), 9083e9087. https://doi.org/10.1073/pnas.87.23.9083.
Sowers, K.R., Gunsalus, R.P., 1995. Halotolerance in Methanosarcina spp.: role of N(sup (epsilon))-Acetyl-(beta)-Lysine,(alpha)-Glutamate, Glycine betaine, and K(supþ) as compatible solutes for osmotic adaptation. Appl. Environ. Microbiol.61 (12), 4382e4388. https://www.ncbi.nlm.nih.gov/pubmed/16535193.
Steinbusch, K.J.J., Hamelers, H.V.M., Buisman, C.J.N., 2008. Alcohol productionthrough volatile fatty acids reduction with hydrogen as electron donor bymixed cultures. Water Res. 42 (15), 4059e4066. https://doi.org/10.1016/j.watres.2008.05.032.
Sudmalis, D., Gagliano, M.C., Pei, R., Grolle, K., Plugge, C.M., Rijnaarts, H.H.M.,Zeeman, G., Temmink, H., 2018. Fast anaerobic sludge granulation at elevatedsalinity. Water Res. 128 (Suppl. C), 293e303. https://doi.org/10.1016/j.watres.2017.10.038.
Van Lier, J., Van der Zee, F., Frijters, C., Ersahin, M., 2015. Celebrating 40 yearsanaerobic sludge bed reactors for industrial wastewater treatment. Rev. Envi-ron. Sci. Biotechnol. 14 (4), 681e702. https://doi.org/10.1007/s11157-015-9375-5.
Vieira, D., S�ervulo, E., Cammarota, M., 2005. Degradation potential and growth ofanaerobic bacteria in produced water. Environ. Technol. 26 (8), 915e922.
Vlyssides, A.G., Loizides, M., Karlis, P.K., 2004. Integrated strategic approach forreusing olive oil extraction by-products. J. Clean. Prod. 12 (6), 603e611. https://doi.org/10.1016/S0959-6526(03)00078-7.
Vyrides, I., Stuckey, D., 2009. Adaptation of anaerobic biomass to saline conditions:role of compatible solutes and extracellular polysaccharides. Enzym. Microb.Technol. 44 (1), 46e51. https://doi.org/10.1016/j.enzmictec.2008.09.008.
Vyrides, I., Santos, H., Mingote, A., Ray, M., Stuckey, D., 2010. Are compatible solutescompatible with biological treatment of saline wastewater? Batch and contin-uous studies using submerged anaerobic membrane bioreactors (SAMBRs).Environ. Sci. Technol. 44 (19), 7437e7442. https://doi.org/10.1021/es903981k.
Vyrides, I., 2015. Anaerobic treatment of organic saline waste/wastewater: over-come salinity inhibition by addition of compatible solutes. In: EnvironmentalMicrobial Biotechnology. Springer, Dordrecht, pp. 105e117.
Vyrides, I., Stuckey, D.C., 2016. Compatible solute addition to biological systemstreating waste/wastewater to counteract osmotic and other environmentalstresses: a review. Crit. Rev. Biotechnol. 1e15. https://doi.org/10.1080/07388551.2016.1266460.
Wagner, A.O., Hohlbrugger, P., Lins, P., Illmer, P., 2012. Effects of different nitrogensources on the biogas productionea lab-scale investigation. Microbiol. Res. 167(10), 630e636. https://doi.org/10.1016/j.micres.2011.11.007.
Xiao, Y., Roberts, D.J., 2010. A review of anaerobic treatment of saline wastewater.Environ. Technol. 31 (8e9), 1025e1043.
Yerkes, D., Boonyakitsombut, S., Speece, R., 1997. Antagonism of sodium toxicity bythe compatible solute betaine in anaerobic methanogenic systems. Water Sci.Technol. 36 (6e7), 15e24. https://doi.org/10.1016/S0273-1223(97)00502-7.
Zaprasis, A., Bleisteiner, M., Kerres, A., Hoffmann, T., Bremer, E., 2014. Uptake ofamino acids and their metabolic conversion into the compatible solute prolineconfers osmoprotection to Bacillus subtilis. Appl. Environ. Microbiol. 81 (1),250e259. https://doi.org/10.1128/AEM.02797-14.
Zhu, C., Dong, S., Wang, F., Huang, G., 2004. Effects of Na/K ratio in seawater ongrowth and energy budget of juvenile Litopenaeus vannamei. Aquaculture 234(1e4), 485e496. https://doi.org/10.1016/j.aquaculture.2003.11.027.