- Email: [email protected]
Identification of soluble microbial products (SMPs) from the fermentation and methanogenic phases of anaerobic digestion
Science of the Total Environment 698 (2020) 134177
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Identification of soluble microbial products (SMPs) from the fermentation and methanogenic phases of anaerobic digestion C. Kunacheva a, Y.N.A. Soh a, D.C. Stuckey a,b,⁎ a Advanced Environmental Biotechnology Centre, Nanyang Environment & Water Research Institute, Nanyang Technological University, 1 Cleantech Loop, Clean Tech One, Singapore 637141, Singapore b Department of Chemical Engineering, Imperial College London, SW7 2AZ, UK
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
• Composition of SMPs produced during fermentation and methanogenesis were compared. • Low MW solutes (b2 kDa) were about 70% similar in both fermentation and methanogenesis. • Fermentation produces higher high MW compounds (N70 kDa) than methanogenesis. • LC-ESI-Q-ToF analysis showed many SMPs are released from cell membranes and walls.
a r t i c l e
i n f o
Article history: Received 12 June 2019 Received in revised form 27 August 2019 Accepted 27 August 2019 Available online 28 August 2019 Editor: Frederic Coulon Keywords: Anaerobic digestion Disinfection byproducts (DBPs) Fermentation Identification Methanogenesis Soluble microbial products (SMP)
a b s t r a c t The production and transformation of Soluble Microbial Products (SMPs) in biological treatment systems is complex, and their genesis and reasons for production are still unclear. SMPs are important since they constitute the main fraction of effluent COD (both aerobic and anaerobic), and hence are the main precursors for disinfection by-products (DBPs). In addition, they are a key component of fouling in membrane bioreactors. Hence, it is important to identify the chemical composition of SMPs, determine their origin, and understand what system parameters influence their production so we can possibly develop strategies to control their production. This study focuses on the production and identification of SMPs in an anaerobic batch process being fed a synthetic feed. To further understand the origins of SMPs, and how they are produced, we analysed the processes of fermentation and methanogenesis independently which has never been done in detail before. SMP concentration, molecular weight distribution and carbohydrate analyses were used to estimate the amount of SMPs in the supernatants. Gas chromatography-mass spectrometry (GC–MS) and liquid chromatography-Time-of-Flight mass spectrometry (LC-ESI-Q-ToF) were used to identify many of the SMPs which have relative masses up to 2 kDa. Our results showed that fermentation released much higher SMP concentrations compared to methanogenesis, especially in the range of 70 k–1000 k Da and 106–1500 Da. Alkanes, alkenes, alcohols, acids, and nitrogen-compounds were the major group of compounds identified in the supernatant of both fermentation and methanogenesis, and 71% of the compounds identified were found in both phases of digestion. Results from LC-ESI-Q-ToF analysis identified components of the cell membrane, such as phosphatidylglycerol,
⁎ Corresponding author at: Advanced Environmental Biotechnology Centre, Nanyang Environment & Water Research Institute, Nanyang Technological University, 1 Cleantech Loop, Clean Tech One, Singapore 637141, Singapore. E-mail addresses: [email protected] (C. Kunacheva), [email protected] (Y.N.A. Soh), [email protected] (D.C. Stuckey).
https://doi.org/10.1016/j.scitotenv.2019.134177 0048-9697/© 2019 Elsevier B.V. All rights reserved.
2
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
phosphatidylethanolamine and phosphatidylserine, as well as other compounds such as flavonoids, acylglycerol, terpene and terpenoids, benzenoid, glyceride, steroid and steroid derivatives. © 2019 Elsevier B.V. All rights reserved.
1. Introduction
2. Materials and methods
Soluble microbial products (SMPs) are defined as the soluble organic products created from microbial metabolism in biological processes that are not products, intermediates such as volatile fatty acids (VFAs), or incoming feed (Barker and Stuckey, 1999). SMPs can be derived from substrate metabolism, biomass growth, and biomass decay; growth associated SMPs, which are called utilization associated products (UAP), are produced directly from biomass growth and substrate metabolism. In contrast, biomass associated products (BAP) are produced as a result of biomass decay and cell lysis during endogenous decay. These SMPs affect the performance of most biological treatment systems (aerobic and anaerobic) as they are the dominant fraction of the effluent chemical oxygen demand (COD) (Aquino, 2004), and also cause membrane fouling in membrane processes (membrane bioreactors and water reclamation plants) (Jarusutthirak and Amy, 2006). Many effluent SMPs are degradable over time in both aerobic and anaerobic processes, however, the hydraulic retention time (HRT) in these processes often limits their complete degradation. Hence, minimizing SMP production should increase treatment performance, minimize chlorination by-products (DBPs), and reduce fouling in membrane bioreactors (Zhang et al., 2015). SMP production and transformation/biodegradation in biological systems are complex processes, and the origin of SMPs is still unclear. Past work has shown that most of the soluble organics found in the effluent are not present in the feed but are produced during microbial metabolism and cell lysis (Kunacheva et al., 2017a; Kunacheva et al., 2017b), and hence are SMPs. Furthermore, SMPs concentrations in the effluent increase under transient conditions, including nutrient limitations, the presence of toxic compounds, and/or when the feed flow or composition is changed radically (Kunacheva et al., 2017a; Kunacheva et al., 2017b; Aquino and Stuckey, 2003). Hence, it is important to identify the composition of SMPs and their molecular mass (Da), determine their origin, and understand what system parameters influence their production, and their contribution to membrane fouling. Until recently this was not possible due to poor and non-specific analytical methods, but now we have the tools to analyse SMPs in far greater depth, at least up to around 2 kDa relative masses (Kunacheva et al., 2017c). Early workers in this field (Kuo et al., 1996; Noguera et al., 1994) found higher concentrations of SMPs in the fermentation phase (glucose fed) in contrast to methanogenesis (acetate fed). Kuo et al. (Kuo et al., 1996) postulated that this was due to the conversion of glucose to CH4 and CO2 being more energetically favourable than the degradation of acetate to the same end products; this gives higher bacterial yields, and should lead to higher concentrations of SMP. In addition, glucose will support different groups of microorganisms (i.e. fermenters, acetogens and methanogens), while acetate will support mainly acetoclastic methanogens, and hence a more diverse population would produce higher concentrations of SMPs (Kuo et al., 1996). However, no one has ever looked at whether the SMPs produced in each phase are similar, or very different. Hence, this study focuses on the identification and production of SMPs in the effluent from an anaerobic process; to further understand the specific origins of SMPs, and how they are produced, a comparison of separate batch experiments with fermentative and methanogenic organisms was carried out to see how much each phase contributes, what the composition of these compounds is, and whether they contribute similar compounds. Ultimately, this work may lead to an understanding about how we could control the different biological processes in anaerobic digestion and improve effluent quality and membrane fouling.
2.1. Reagents and chemicals Methanol (LC-MS grade) was purchased from Sigma-Aldrich Pte Ltd. (Singapore). Acetone, chloroform, dichloromethane, and n-hexane (GC–MS grade or equivalent) were purchased from Merck Pte Ltd. (Singapore). Other solvents such as diethyl ether, ethyl acetate, nheptane, methyl tert-butyl ether and toluene were of chromatographic grade and purchased from Fisher Scientific Pte Ltd. (Singapore). Formic, acetic, propionic, isobutyric, butyric, isovaleric and valeric acid (analytical grade) and the alkane standard mixture (C10 - C40, all 50 mg/L each) were purchased from Sigma-Aldrich Pte Ltd. (Singapore). Deionised water was obtained from a MilliQ system (Millipore Advantage A10). 2.2. Batch reactor 2.2.1. Anaerobic batch reactor A cylindrical batch reactor was made from polymethyl methacrylate (Plexiglas) and had a working volume of 6 L, with 2 L of headspace. This anaerobic sequencing batch reactor was initially seeded with an inoculum of screened anaerobic sludge (bacteria) from a conventional domestic wastewater treatment plant in Singapore (Ulu Pandan). To ensure that steady state conditions were achieved, the reactors were run for at least 60 days at an effective HRT of 7 days (feed and decant) and at a 200 day SRT before the experiment started. The anaerobic bioreactor was fed with a synthetic feed comprised of glucose, peptone, meat extract and essential nutrients at a concentration of 4 gCOD/L. After the final feed aliquot was added at time zero, the reactors were not fed again. An initial (0 h) sample was collected just after the feed was added and fully mixed with the sludge; samples were then collected every 2 h in the first day, and then every 12 h until 456 h, and were filtered through a 0.45 μm glass fibre filter before analysis. The mixed liquor volatile suspended solids (MLVSS) in the reactor were controlled at 6000 mg/L. The reactor temperature was controlled at 35 ± 1 °C, while excess biogas was collected using a polyethylene gasbag, and measured using a gas pump and meter. 2.2.2. Enrichment batch cultures (fermentation and methanogenesis) Fermentation and methanogenesis batch tests were carried out using duplicate 1 L glass bottles connected to gasbags. The fermentation bottles were fed with glucose (2734 mg/L), peptone (818 mg/L), meat extract (274 mg/L) and essential nutrients combined at a concentration of 4 gCOD/L, while the methanogenesis bottles were fed with acetate (859 mg/L), peptone (818 mg/L), meat extract (274 mg/L) and essential nutrients combined at a concentration of 2 gCOD/L. We selected 2 gCOD/L of acetate because the results from the first section showed that acetate was at the highest concentration at 2 gCOD/L (Fig. 1), and 2 g/L of sludge inoculum from the 6 L batch reactor were used in each bottle. To inhibit the activity of methanogens inside the fermentation batch, 50 mM of bromoethanesulfonic acid (BESA) was added. BESA is a structural analog and competitive inhibitor of coenzyme M (Taylor and Wolfe, 1974) in methanogens (Balch and Wolfe, 1979), and is a potent inhibitor of methanogens at low concentrations (Aguilar et al., 1995). It is possible that BESA inhibition could have influenced SMP production, but at present it is the best inhibitor of methanogenic activity. All the bottles were adjusted to pH 7 at the beginning of the experiment and then placed in a controlled temperature shaker at 35 °C, 150 rpm.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
Fig. 1. COD equivalent concentrations of glucose and VFAs in the supernatant of a batch reactor.
2.3. Analysis of various parameters 2.3.1. General parameters All samples were filtered through 0.45 μm glass fibre filters, and then analysed in duplicate for VFAs and COD. The concentration of SMPs is typically estimated by subtracting the COD due to intermediate VFAs and residual substrate, from the soluble effluent COD. Total suspended solids (TSS), MLVSS and COD were measured as described in Standard Methods APHA (Eaton and Franson, 2005). VFAs were measured using a Shimadzu high-performance liquid chromatography (HPLC, SPD20AD) with a UV diode array detector (DAD, SPD-M20A) at 210 nm using an Aminex HPX-87H (300 × 7.8 mm) column. Analysis time was 25 min for each sample operating under isocratic and isothermal conditions using 0.005 M H2SO4 as the mobile phase at a flowrate 0.8 mL/min at 55 °C (Guerrant et al., 1982). A total of seven VFAs were quantified with this method, and Coefficients of variation (COV = SD/average value) for all VFAs were below ±6%. The composition of biogas (methane, oxygen, nitrogen, and carbon dioxide) was determined using a Shimadzu GC-2010plus gas chromatograph with a thermal conductivity detector (TCD) (COV below ±3%), and a select permanent gases/CO2 (CP7429) column (Agilent) was used for gas separation. 2.3.2. Molecular weight distribution Samples were filtered through a 0.2 μm syringe filter, and their MW distribution analysed using size exclusion chromatography (SEC). SEC was carried out using two columns (PolySep GFC-P1000 and 4000, Phenomenex) connected in series, with UV-DAD and refractive index detectors (RID, Shimadzu) to analyse both proteins and carbohydrates, respectively. EasiVial polymer standards (Agilent, U.S.A.) were used for molecular weight (MW) calibration. 2.3.3. Monosaccharides analysis Polysaccharides were hydrolysed by trifluoroacetic acid (4 M) at 100 °C for 2 h, and the monosaccharides released were derivatized by the alditol acetate method (Fox et al., 1989); eight neutral sugars were analysed using inositol as the internal standard. The detection and quantification was done using gas chromatography (GC-2010 plus, Shimadzu) on a 30 m × 0.25 mm × 0.25 μm RTX-5MS column (Restek, Bellefonte, PA, USA) coupled with a mass spectrometry (GC–MSQP2010, Shimadzu).
3
2017c). In brief, SPE was carried out using two Waters OasisHLB cartridges in series, and the analytes eluted with 2 mL of selected solvents (methanol, acetone, dichloromethane, n-hexane) in sequence. The eluent from each SPE cartridge was collected and further pre-treated using LLE. With LLE, a mixed solvent (hexane:dichloromethane:chloroform, 1:1:1) was used for extraction, and after lyophilisation 1 mL of dichloromethane was used for reconstitution for chromatographic analysis. Eluted samples were then analysed using a GC–MS system (GCMSQP2010ULTRA, Shimadzu); the sample was injected onto an RTX-5MS (30 m × 0.25 mm ID, Restek) column. The total runtime per sample was 60 min, and the temperature program was: 50 °C, hold 7 min, rate 7 °C/min to 325 °C, and hold for 14 min. Mass spectra were acquired from m/z 30 to 580 after a 10 min solvent cut time. However, based on the separation techniques this method does not separate very polar solutes well, and hence they are not well represented in the SMPs. The chromatographic peaks were identified using the NIST11 library (National Institute of Standards and Technology, Gaithersburg, MD, USA, http://www.nist.gov/srd/mslist.htm). Similarity index, mass spectrum and retention index were all used as selection criteria for compound identification from the NIST library list of suggested compounds (Jonsson et al., 2005; Kumari et al., 2011). Method blanks (deionised water), including reactor parts and tubing soaked in DI water, were run through the same pretreatment and analysis, while feed samples were also analysed to identify compounds in the feed and the concentration of these compounds were subtracted from the effluent samples in order to “normalise” net SMP production. Supernatant samples from the methanogenic and fermentation cultures were also extracted and measured with the method that we previously developed (Tipthara et al., 2017) using LC-ESI-Q-ToF. LLE was performed by adding 3 mL of 2:1 (v/v) chloroform/methanol; this mixture was then centrifuged at 1610g, at 20 °C for a further 10 min. The upper aqueous (metabolite) fraction and lower organic (lipid) fraction were transferred to separate 1.5 mL tubes and dried in a vacuum concentrator. UPLC-MS analyses were conducted using an ACQUITY UPLC with a 2777C autosampler (Waters Corp, USA) coupled to a Xevo G2XS QTof mass spectrometer (Waters, Manchester, UK) with an electrospray ionization (ESI) source. A mass range of m/z 100 to 2000 was acquired for both metabolites and lipids. The metabolite extracts were separated using an ACQUITY UPLC HSS T3 column (2.1 × 100 mm, 1.8 μm; Waters) at 40 °C. Three replicate injections were made for each ESI positive and ESI negative mode. Mobile phase A consisted of deionised water with 0.1% formic acid, while mobile phase B consisted of acetonitrile with 0.1% formic acid. The elution gradient was as follows: 0–40% B (1.0–11.0 min), 40–80% B (11.0–11.1 min), 80% B (11.1–13.0 min), 80–0% B (13.0–13.1 min), 0% B (13.1–16 min) with a flow rate of 0.4 mL/min. The lipid extracts were analysed separately by reconstituting them in 2:1:1(v/v/v) isopropanol/acetonitrile/water, and separating them using an ACQUITY UPLC CSH C18 column (2.1 × 100 mm, 1.7 μm; Waters) at 55 °C. Three replicate injections were made for each ESI positive and ESI negative mode. Mobile phase A consisted of 60:40 (v/v) acetonitrile/deionised water with 10 mM ammonium formate and 0.1% formic acid. Mobile phase B consisted of 90:10 (v/v) isopropanol/acetonitrile with 10 mM ammonium formate and 0.1% formic acid. The elution gradient was set as follows: 40–43% B (0.0–2.0 min), 43–50% B (2.0–2.1 min), 50–54% B (2.1–12.0 min), 54–70% B (12.0–12.1 min), 70–99% B (12.1–18 min), 99–40% B (18.0–18.1 min), 40% B (18.1–20.0 min), with a flow rate of 0.4 mL/min. 3. Results and discussion 3.1. Anaerobic batch reactor
2.4. Identification of low MW compounds in SMPs Organic compounds were separated using a combination of solidphase (SPE) and liquid-liquid extraction (LLE) (Kunacheva et al.,
The 6 L fed batch reactor after running for 60 days at an HRT of 7 days, was fed one time with 4 gCOD/L of glucose. Fig. 1 shows the equivalent COD of effluent, glucose and VFAs in the supernatant over
4
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
time. The 0 h sample was collected 5 min after feeding, which showed the concentration of glucose had already reduced to 2.7 gCOD/L. Glucose (MW = 180 Da) was rapidly consumed by microorganisms in the reactor, and was gone within 6 h (Fig. 2, RID peak at 20 min). VFAs concentrations started to rise rapidly after the 0 h sample (5 min), and peaked at 4 h, showing that fermentation was the major mechanism for glucose removal during this period. Acetate and propionate were the major VFAs present, while formate was also found in the first 4 h; the presence of formate has also been noted after shock loads to anaerobic systems in past work (Kunacheva et al., 2017b; Nachaiyasit and Stuckey, 1995). The VFAs then started to decrease when the methanogens started to metabolise them, while the SMPs started to increase; the highest SMP concentration occurred after 6–8 h, and accounted for 1.2 gCOD/L (44% of total soluble COD). VFAs, SMPs and methane accounted for 36%, 32% and 21% of feed COD, respectively during 6–8 h. The other 11% can be accounted for as dissolved methane in the supernatant, errors in the multiple analyses, and cell growth. After 8 h both the SMPs and VFAs gradually decreased, and the VFAs were totally gone after 96 h. There are several hypotheses as to why the SMP concentrations decreased, such as; high SMP production by fermentative/acidogenic bacteria compared to methanogens, hydrolysis of SMPs, and/or consumption of SMPs by microorganisms in the anaerobic system. However, SMPs still remained in the supernatant at concentrations of around 200 mg/L (~5% of the incoming feed) even after 456 h, which
may represent the anaerobically non-biodegradable SMPs (Schiener et al., 1998). Fig. 2 shows the SECs of the supernatant from the batch reactor. The RID chromatogram shows that glucose was totally consumed in 6 h (peaks between 106 and 1500 Da), while a peak of high molecular carbohydrates occurred at 14.7 min (70 k–100 k Da) in the 4 and 6 h samples; it was clear that this peak was formed when the acidogens were active (Fig. 2a). The chromatogram at UV210nm shows that the high MW compounds (70 k–100 k Da) decreased which may have broken down to 50 k–70 k Da solutes during the 4–6 h period (Fig. 2b), and then further breakdown to smaller ones (500–5 k Da) during 8–24 h (Fig. 2c). The 500–5 k Da compounds were then not found in the 48 h sample, which shows only two peaks in the range of 100–300 Da, while smaller compounds were not found after 96 h, and there was only one small peak at 14 min (100 kDa) which may represent the anaerobically non-biodegradable SMPs. These results show that there might be a big difference in SMP production between fermentation and methanogenesis. Fig. 3 shows the monosaccharide concentrations in the supernatant; glucose, mannose, and galactose were the dominant sugars which are often found as major constituents of bacterial extracellular polymers (ECPs) (Dignac et al., 1998). Based on Fig. 2a, glucose in the feed was totally consumed within 6 h, however, there were also long chain carbohydrates that appear to have been broken down to glucose which
Fig. 2. Chromatograms from SEC of the supernatant of the batch reactor at different times using different detectors [RID (a) and UV210nm (b), (c)] (n = 3). Dotted lines are from MW standards.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
5
were detected using the monosaccharide analysis. ECPs released by the acetogens were possibly the cause of this increment in glucose (and mannose) after 4 h since it is possible that such ECPs were biomass associated products (BAPs) released into the medium as a result of increased starvation and cell death. Furthermore, the decrease in carbohydrates after 8 h could have been the result of SMP consumption by methanogens, which was also shown in a previous study where methanogens might have contributed more than fermentative/ acidogenic bacteria to SMP consumption (Wu and Zhou, 2010), and/or consumption by acidogens. 3.2. Enrichment cultures (fermentation and methanogen batch tests) From the results of the anaerobic batch reactor (Section 3.1), enrichment cultures (fermentation/acidogenesis and methanogenesis) were carried out to obtain a better understanding of SMP production from each phase of anaerobic digestion. The fermentation culture was fed with 4 gCOD/L glucose, and the COD did not change much over the 7 days since fermentation basically transforms glucose to VFAs (methanogenesis was inhibited by BESA) (Fig. 4a). Unlike the mixed anaerobic batch reactor, acetate was the only dominant VFA in this fermentation culture, although formate was also detected in the first 2 days similar to the mixed batch reactor in Section 3.1. There were also large peaks of other acids in the chromatograms during the experiment (oxalic and pyruvic) which were detected using HPLC; although this method was not optimized to quantify these acids (Guerrant et al., 1982). This would explain why there was a big difference between COD and VFAs (as COD) values. Fig. 4b shows the performance of the methanogenic culture; acetate was consumed from 2 to 1 gCOD/L in 7 days, while methane was produced continuously from day 1 until day 7. Fig. 5a shows the SEC chromatogram of the batch fermentation test. The biggest peak was around 60 kDa on day 1, but this changed gradually to a higher MW (80 kDa) on day 7; this peak was also seen in the mixed anaerobic batch reactor (Fig. 2), but not in the chromatogram from the methanogenic reactor (Fig. 5b). Hence it can be concluded that the fermentation process released the 60–80 kDa SMPs, while the lower MW peaks were similar in both fermentation and methanogenesis, especially initially. These peaks started with smaller compounds in the range of 100–300 Da on day 1, and the peak kept increasing in size (higher MWs) to about 1000 to 2000 Da (Fig. 5); these compounds were identified using GC–MS and a LC-ESI-Q-ToF in the next section. Overall, fermentation released much higher MW SMPs compared to methanogenesis, and this was also found in previous studies (Noguera et al., 1994; Wu and Zhou, 2010).
Fig. 4. COD concentrations in the fermentation and methanogenic batch tests (n = 4).
3.3. Identification of low MW SMPs (MW b 2000 Da) in enrichment cultures Identifying SMPs was done using SPE and LLE coupled with GC–MS and LC-ESI-Q-ToF analysis (n = 4). Supernatant samples were collected on day 7 from both the fermentation and methanogenesis reactors, and a total of 77 compounds were identified by GC–MS (MW b 580). Alkanes, alkenes, alcohols, acids, and nitrogen-compounds were the major groups of compounds identified in both fermentation and methanogenesis, and these compounds have been found in previous studies (Kunacheva et al., 2017b; Wu and Zhou, 2010; Citron et al., 2012; Trzcinski and Stuckey, 2009). Fig. 6 shows the chromatograms
Fig. 3. Monosaccharide concentrations in the supernatant of the anaerobic batch reactor after hydrolysis of polysaccharides (n = 3).
6
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
Fig. 5. SEC chromatogram of the fermentation (F) and methanogenic (M) batch tests.
(LLE) of fermentation and methanogenesis analysed by GC–MS; 71% of the compounds identified were actually the same in both fermentation and methanogenesis. Due to our specific analytical methodology, this is the first time this has ever been noted in the literature. From the literature it is clear that alkanes and alkenes can be synthesized by Sarcina lutea through substrate metabolism that utilizes fatty acids (Albro and Dittmer, 1970). Furthermore, in a recent study Shi et al. also found that a laccase enzyme from the endophyte Pantoea ananatis Sd-1 could degrade lignin (complex hydrocarbon) and produce metabolites such as alkanes and acids (Shi et al., 2015), similar to those found in our study. Also, a study on volatile compounds released by 50 bacterial strains revealed that bacteria can produce esters, alcohols, nitrogen-containing as well as sulfur compounds (Citron et al., 2012). Micromonospora aurantiaca, which are naturally present in both soil and water, can also produce a large diversity of VFAs and fatty acid methyl esters (Dickschat et al., 2011), while further studies found that N-compounds, acids and esters are precursors for disinfection by– products (Zhang et al., 2015; Liu and Zhang, 2014). Fig. 7 shows that compounds between 100 and 2000 Da were identified using LC-ESI-Q-ToF with methods for metabolites and lipids (Tipthara et al., 2017). Qualitative analysis was carried out for chromatographic peaks of signal intensity 1000 and above, and compounds identified from both positive and negative lipids and metabolites scans are summarised in Tables S2 and S3. In contrast to the data collected from SEC and SPE-LLE-GC–MS, results from the LC-ESI-Q-ToF analysis showed that there were more compounds identified in the methanogenesis experiment than the fermentation enrichment culture. Sphingolipids, glycerophospholipids and carbohydrate derivatives were found in higher abundance in the fermentation culture, whereas the abundance of glycerides, terpenes and terpenoids, and benzenoids were higher in the methanogenic culture. Steroids were found in similar amounts in both enrichment cultures (Fig. 7). A previous study of SMPs using the same LC-ESI-Q-ToF method had also identified components of cell membranes, such as phosphatidylglycerol (PG), phosphatidylethanolamine (PE) and phosphatidylserine (PS) (Tipthara et al., 2017). It is known that the outer membrane of E. coli is comprised of approximately 5% cardiolipin (CL), 20 to 25% PG, 70 to 80% PE, as well as a small percentage of PS (Romantsov et al., 2009), and PE and PG were reported to play significant roles in biofilm formation (Benamara et al., 2014; Ni et al., 2009). The composition of membrane biolipids has been of considerable interest,
Fig. 6. Chromatograms (LLE) of fermentation and methanogenesis reactor supernatants analysed by GC–MS (MW b 580 Da).
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
7
Fig. 7. Metabolites and lipids identified by LC-ESI-Q-ToF (MWb2,000) where signal intensity was higher in the methanogenic reactor than in fermentation reactor (A) and vice versa (B).
especially in research groups studying antimicrobial mechanisms and microbial ecology, and laboratory studies and computational microbiology have demonstrated the ability of bacteria to alter their membrane lipid composition to improve their adaptability to the physical environment (Lindblom et al., 2002; Matsumoto et al., 2006; Zhang and Rock, 2008). This current study of SMPs found during fermentation and methanogenesis, the samples for SMP extraction were only taken on the last day of the experiment, i.e. Day 7. Further work on intermittent sampling is required to monitor the changes of SMPs over the batch run. In this study, the compounds identified were more diverse than our earlier studies (Kunacheva et al., 2017a; Kunacheva et al., 2017b; Trzcinski and Stuckey, 2010), and included flavonoids, as well as acylglycerol, terpene and terpenoids, benzenoid, glyceride, steroid and steroid derivatives, while the Human Metabolome Database (HMDB) used to analyse our data suggested the presence of some plant-derived compounds. 58 out of 136 compounds identified in group (A) (Fig. 7), and 12 out of 44 compounds identified in group (B) (Fig. 7) were classified as plant-derived compounds, i.e. isolated from roots, bark, leaves or fruits of plants as identified on the HMDB. Interestingly, despite the fact that these enrichment cultures should contain only acidogenic or methanogenic bacterial communities, and hence the compounds detected in the cultures should only be produced by such microorganisms, many of these compounds are well-known plant-derived substances. One explanation for this is that they might have been produced by photosynthetic bacteria (Pfennig, 1967). Several classes of sphingolipids were detected using high performance chromatography, HPLC, GC–MS and LC-MS, and have been reported in various forms, such as sphingomyelin
(SM), sphingophospholipid, aminoglycosphingolipid, sphingoglycolipid, and ceramide. Further studies on bacterial lipid components revealed the presence of sphingolipids in acetic acid-producing bacteria under stressed conditions such as elevated temperatures and low pH, where they were biotransformed into ceramides to defend themselves against the harsh environment. This is consistent with the current finding which identified sphingolipids and glycerophospholipids in higher quantities in the fermentation culture. Other components of the membranes of acid-producing bacteria identified were phospholipids, terpenoids and aminolipids (Naka et al., 2000). In contrast, the current work found more terpenes and terpenoids in the compounds identified from the methanogenic culture. To date, there appears to be nothing in the literature on this subgroup of sphingolipids being produced by other bacteria. Bacterial membranes contain a majority of PE, and PG and PG derivatives such as diphosphatidylglycerol and CL, and the lipid composition depends on the class of bacteria; this group of compounds were found mainly in the fermentation culture. The inner membranes of gramnegative bacteria generally contain higher amounts of PE, while the cytoplasmic membranes of gram-positive bacteria are comprised predominantly of PG and its derivatives; the widely studied E. coli was found to have approximately 82% PE by weight in the inner cytoplasmic membrane, whereas the membrane of Staphylococcus aureus is mainly PGs and its derivatives. These different lipid compositions determine the membrane properties such as its general structure, packing density, rigidity, and overall surface charge density (Navas et al., 2005), and Zhao, Rog, Gurtovenko, Vattulainen and Karttunen (Zhao et al., 2008) proposed that PG contributed significantly to the stability of bacterial
8
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
membranes. Phosphatidylcholines (PC) and PE are neutral zwitterionic species under physiological conditions, whereas PG are negatively charged. The presence of PG as part of the membrane structure was shown to improve cell structural stability as it increases ionic electrostatic charges and inter-molecular hydrogen bonding between the PE molecules present in the lipid bilayer. This enhances membrane integrity during events of intercellular interactions, and external ‘disruptions’ in case of the presence of foreign molecules such as antimicrobial peptides (Zhao et al., 2008). The broad class of terpenes, terpenoids and benzenoids are found in higher amounts in the methanogenic culture in this experiment. The detection of triterpenoids and their derivatives in this study have been identified as compounds originating from plants by the HMDB, however, a review by Taylor and Davies (Taylor and Davies, 1974) introduced earlier research carried out on the identification, biosynthesis, and functionality of this group of compounds which showed that they can originate from non-photosynthetic bacterial species. By mapping of biosynthetic pathways through past reports on terpenoids and the precursor isoprene, non-photosynthetic bacteria were proven to be able to synthesize isoprenoids, and triterpenoids such as squalene and hopanes de novo (Taylor and Davies, 1974). Finally, triacylglycerol or triglycerides (TG) have been found to be produced by both gram-positive and gram-negative bacteria that belong to the actinomycetes group, such as Streptomyces, Nocardia, Rhodococcus, Mycobacterium, Dietzia and Gordonia, for carbon and energy storage purposes during stagnant growth, for instance in adverse environmental conditions. These triglycerides showed higher signal intensities in the methanogenic culture supernatant, are also the precursors for phospholipid biosynthesis, which is a major component of cell membranes (Alvarez and Steinbuchel, 2002; Athenstaedt and Daum, 2006). An earlier review by Norbert Pfennig reported that the photosynthetic route of bacteria proceeded under anaerobic conditions and in minimum light, and photosynthetic bacteria utilize external electron donors such as sulfur and other organic compounds. The carotenoids play an important role in the energy transfer chain during photosynthesis, and prevent the destruction of anaerobic cells by bacteriochlorophyll during light absorption in the presence of oxygen (Griffiths et al., 1955; Slouf et al., 2012), and the carotenoid, mutatochrome, was also detected in the enrichment cultures in this study. 4. Conclusion This study showed that both the fermentation (acidogens) and methanogenic phases released different types of SMPs. Acidogens released much higher high MW compounds (MW N 70 kDa) compared to methanogens, however, the lower MW compounds (MW b 2000 Da) were quite similar in both processes (about 71%). Analysis using LCESI-Q-ToF revealed that some SMPs are derived from cell wall fragments, and from compounds released by the cell to adjust to the surrounding environment, such as lipids and carbohydrates. Sphingolipids, glycerophospholipids and carbohydrate derivatives were found in higher abundance in the fermentation culture, whereas the abundance of glycerides, terpenes and terpenoids, and benzenoids were higher in the methanogenic culture. For future work, it would be important to try and link changes in microbial ecology to these differences in SMPs, and investigate how changes in the type of SMPs changes membrane fouling. Declaration of Competing Interest None declared under financial, general and institutional competing interests. Acknowledgement This research work was supported by the Singapore National Research Foundation under its Environmental & Water Technologies
Strategic Research Programme, and administered by the Environment & Water Industry Programme Office (EWI) of the PUB. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2019.134177.
References Aguilar, A., Casas, C., Lema, J.M., 1995. Degradation of volatile fatty-acids by differently enriched methanogenic cultures - kinetics and inhibition. Water Res. 29, 505–509. Albro, P.W., Dittmer, J.C., 1970. Bacterial hydrocarbons - occurrence, structure and metabolism. Lipids 5 (320-&). Alvarez, H.M., Steinbuchel, A., 2002. Triacylglycerols in prokaryotic microorganisms. Appl. Microbiol. Biotechnol. 60, 367–376. Aquino, S.F., 2004. Formation of Soluble Microbial Products (SMP) in Anaerobic Reactors during Stress Conditions. Department of Chemical Engineering and Chemical Technology, Imperial College of Science Technology and Medicine, Imperial College of Science Technology and Medicine, London SW7 2BY. Aquino, S.F., Stuckey, D.C., 2003. Production of soluble microbial products (SMP) in anaerobic chemostats under nutrient deficiency. J. Environ. Eng. ASCE 129, 1007–1014. Athenstaedt, K., Daum, G., 2006. The life cycle of neutral lipids: synthesis, storage and degradation. Cell. Mol. Life Sci. 63, 1355–1369. Balch, W.E., Wolfe, R.S., 1979. Specificity and biological distribution of coenzyme M (2mercaptoethanesulfonic acid). J. Bacteriol. 137, 256–263. Barker, D.J., Stuckey, D.C., 1999. A review of soluble microbial products (SMP) in wastewater treatment systems. Water Res. 33, 3063–3082. Benamara, H., Rihouey, C., Abbes, I., Ben Mlouka, M.A., Hardouin, J., Jouenne, T., Alexandre, S., 2014. Characterization of membrane lipidome changes in Pseudomonas aeruginosa during biofilm growth on glass wool. PLoS One 9. Citron, C.A., Rabe, P., Dickschat, J.S., 2012. The scent of bacteria: headspace analysis for the discovery of natural products. J. Nat. Prod. 75, 1765–1776. Dickschat, J.S., Bruns, H., Riclea, R., 2011. Novel fatty acid methyl esters from the actinomycete Micromonospora aurantiaca. Beilstein J. Org. Chem. 7, 1697–1712. Dignac, M.F., Urbain, V., Rybacki, D., Bruchet, A., Snidaro, D., Scribe, P., 1998. Chemical description of extracellular polymers: implication on activated sludge floc structure. Water Sci. Technol. 38, 45–53. Eaton, A.D., Franson, M.A.H., 2005. Standard Methods for the Examination of Water & Wastewater. American Public Health Association. Fox, A., Morgan, S.L., Gilbart, J., 1989. Preparation of alditol acetates and their analysis by gas chromatography (GC) and mass spectrometry (MS). Analysis of Carbohydrates by GLC and MS, pp. 87–117. Griffiths, M., Sistrom, W.R., Cohenbazire, G., Stanier, R.Y., 1955. Function of carotenoids in photosynthesis. Nature 176, 1211–1214. Guerrant, G.O., Lambert, M.A., Moss, C.W., 1982. Analysis of short-chain acids from anaerobic-bacteria by high-performance liquid-chromatography. J. Clin. Microbiol. 16, 355–360. Jarusutthirak, C., Amy, G., 2006. Role of soluble microbial products (SMP) in membrane fouling and flux decline. Environ. Sci. Technol. 40, 969–974. Jonsson, P., Johansson, A.I., Gullberg, J., Trygg, J., A, J., Grung, B., Marklund, S., Sjostrom, M., Antti, H., Moritz, T., 2005. High-throughput data analysis for detecting and identifying differences between samples in GC/MS-based metabolomic analyses. Anal. Chem. 77, 5635–5642. Kumari, S., Stevens, D., Kind, T., Denkert, C., Fiehn, O., 2011. Applying in-silico retention index and mass spectra matching for identification of unknown metabolites in accurate mass GC-TOF mass spectrometry. Anal. Chem. 83, 5895–5902. Kunacheva, C., Soh, Y.N.A., Stuckey, D.C., 2017a. Effect of feed pH on reactor performance and production of soluble microbial products (SMPs) in a submerged anaerobic membrane bioreactor. Chem. Eng. J. 320, 135–143. Kunacheva, C., Soh, Y.N.A., Trzcinski, A.P., Stuckey, D.C., 2017b. Soluble microbial products (SMPs) in the effluent from a submerged anaerobic membrane bioreactor (SAMBR) under different HRTs and transient loading conditions. Chem. Eng. J. 311, 72–81. Kunacheva, C., Le, C.C., Soh, Y.N.A., Stuckey, D.C., 2017c. Chemical characterization of low molecular weight soluble microbial products in an anaerobic membrane bioreactor. Environ. Sci. Technol. 51, 2254–2261. Kuo, W.C., Sneve, M.A., Parkin, G.F., 1996. Formation of soluble microbial products during anaerobic treatment. Water Environ. Res. 68, 279–285. Lindblom, G., Oradd, G., Rilfors, L., Morein, S., 2002. Regulation of lipid composition in Acholeplasma laidlawii and Escherichia coli membranes: NMR studies of lipid lateral diffusion at different growth temperatures. Biochemistry-Us 41, 11512–11515. Liu, J.Q., Zhang, X.R., 2014. Comparative toxicity of new halophenolic DBPs in chlorinated saline wastewater effluents against a marine alga: Halophenolic DBPs are generally more toxic than haloaliphatic ones. Water Res. 65, 64–72. Matsumoto, K., Kusaka, J., Nishibori, A., Hara, H., 2006. Lipid domains in bacterial membranes. Mol. Microbiol. 61, 1110–1117. Nachaiyasit, S., Stuckey, D.C., 1995. Microbial response to environmental-changes in an anaerobic baffled reactor (Abr). Anton Leeuw. Int. J. G. 67, 111–123. Naka, T., Fujiwara, N., Yabuuchi, E., Doe, M., Kobayashi, K., Kato, Y., Yano, I., 2000. A novel sphingoglycolipid containing galacturonic acid and 2-hydroxy fatty acid in cellular lipids of Sphingomonas yanoikuyae. J. Bacteriol. 182, 2660–2663.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177 Navas, B.P., Lohner, K., Deutsch, G., Sevcsik, S., Riske, K.A., Dimova, R., Garidel, P., Pabst, G., 2005. Composition dependence of vesicle morphology and mixing properties in a bacterial model membrane system. Bba-Biomembranes 1716, 40–48. Ni, B.-J., Fang, F., Xie, W.-M., Sun, M., Sheng, G.-P., Li, W.-H., Yu, H.-Q., 2009. Characterization of extracellular polymeric substances produced by mixed microorganisms in activated sludge with gel-permeating chromatography, excitation–emission matrix fluorescence spectroscopy measurement and kinetic modeling. Water Res. 43, 1350–1358. Noguera, D.R., Araki, N., Rittmann, B.E., 1994. Soluble microbial products (SMP) in anaerobic chemostats. Biotechnol. Bioeng. 44, 1040–1047. Pfennig, N., 1967. Photosynthetic bacteria. Annu. Rev. Microbiol. 21 (285-&). Romantsov, T., Guan, Z.Q., Wood, J.M., 2009. Cardiolipin and the osmotic stress responses of bacteria. Bba-Biomembranes 1788, 2092–2100. Schiener, P., Nachaiyasit, S., Stuckey, D., 1998. Production of soluble microbial products (SMP) in an anaerobic baffled reactor: composition, biodegradability, and the effect of process parameters. Environ. Technol. 19, 391–399. Shi, X.W., Liu, Q., Ma, J.S., Liao, H.D., Xiong, X.Q., Zhang, K.K., Wang, T.F., Liu, X.M., Xu, T., Yuan, S.S., Zhang, X., Zhu, Y.H., 2015. An acid-stable bacterial laccase identified from the endophyte Pantoea ananatis Sd-1 genome exhibiting lignin degradation and dye decolorization abilities. Biotechnol. Lett. 37, 2279–2288. Slouf, V., Chabera, P., Olsen, J.D., Martin, E.C., Qian, P., Hunter, C.N., Polivka, T., 2012. Photoprotection in a purple phototrophic bacterium mediated by oxygendependent alteration of carotenoid excited-state properties. Proc. Natl. Acad. Sci. U. S. A. 109, 8570–8575. Taylor, R.F., Davies, B.H., 1974. Triterpenoid carotenoids and related lipids - triterpenoid monohydroxy-carotenoids and monoglucosyloxy-carotenoids from streptococcusfaecium unh 564p. Biochem. J. 139, 761–769.
9
Taylor, C.D., Wolfe, R.S., 1974. Structure and methylation of coenzyme M (Hsch2ch2so3). J. Biol. Chem. 249, 4879–4885. Tipthara, P., Kunacheva, C., Soh, Y.N.A., Wong, S.C.C., Pin, N.S., Stuckey, D.C., Boehm, B.O., 2017. Global profiling of metabolite and lipid soluble microbial products in anaerobic wastewater reactor supernatant using UPLC-MSE. J. Proteome Res. 16, 559–570. Trzcinski, A.P., Stuckey, D.C., 2009. Anaerobic digestion of the organic fraction of municipal solid waste in a two-stage membrane process. Water Sci. Technol. 60, 1965–1978. Trzcinski, A.P., Stuckey, D.C., 2010. Treatment of municipal solid waste leachate using a submerged anaerobic membrane bioreactor at mesophilic and psychrophilic temperatures: analysis of recalcitrants in the permeate using GC-MS. Water Res. 44, 671–680. Wu, B., Zhou, W., 2010. Investigation of soluble microbial products in anaerobic wastewater treatment effluents. J. Chem. Technol. Biotechnol. 85, 1597–1603. Zhang, Y.M., Rock, C.O., 2008. Membrane lipid homeostasis in bacteria. Nat. Rev. Microbiol. 6, 222–233. Zhang, B.B., Xian, Q.M., Zhu, J.P., Li, A.M., Gong, T.T., 2015. Characterization, DBPs formation, and mutagenicity of soluble microbial products (SMPs) in wastewater under simulated stressful conditions. Chem. Eng. J. 279, 258–263. Zhao, W., Rog, T., Gurtovenko, A.A., Vattulainen, I., Karttunen, M., 2008. Role of phosphatidylglycerols in the stability of bacterial membranes. Biochimie 90, 930–938.
Contents lists available at ScienceDirect
Science of the Total Environment journal homepage: www.elsevier.com/locate/scitotenv
Identification of soluble microbial products (SMPs) from the fermentation and methanogenic phases of anaerobic digestion C. Kunacheva a, Y.N.A. Soh a, D.C. Stuckey a,b,⁎ a Advanced Environmental Biotechnology Centre, Nanyang Environment & Water Research Institute, Nanyang Technological University, 1 Cleantech Loop, Clean Tech One, Singapore 637141, Singapore b Department of Chemical Engineering, Imperial College London, SW7 2AZ, UK
H I G H L I G H T S
G R A P H I C A L
A B S T R A C T
• Composition of SMPs produced during fermentation and methanogenesis were compared. • Low MW solutes (b2 kDa) were about 70% similar in both fermentation and methanogenesis. • Fermentation produces higher high MW compounds (N70 kDa) than methanogenesis. • LC-ESI-Q-ToF analysis showed many SMPs are released from cell membranes and walls.
a r t i c l e
i n f o
Article history: Received 12 June 2019 Received in revised form 27 August 2019 Accepted 27 August 2019 Available online 28 August 2019 Editor: Frederic Coulon Keywords: Anaerobic digestion Disinfection byproducts (DBPs) Fermentation Identification Methanogenesis Soluble microbial products (SMP)
a b s t r a c t The production and transformation of Soluble Microbial Products (SMPs) in biological treatment systems is complex, and their genesis and reasons for production are still unclear. SMPs are important since they constitute the main fraction of effluent COD (both aerobic and anaerobic), and hence are the main precursors for disinfection by-products (DBPs). In addition, they are a key component of fouling in membrane bioreactors. Hence, it is important to identify the chemical composition of SMPs, determine their origin, and understand what system parameters influence their production so we can possibly develop strategies to control their production. This study focuses on the production and identification of SMPs in an anaerobic batch process being fed a synthetic feed. To further understand the origins of SMPs, and how they are produced, we analysed the processes of fermentation and methanogenesis independently which has never been done in detail before. SMP concentration, molecular weight distribution and carbohydrate analyses were used to estimate the amount of SMPs in the supernatants. Gas chromatography-mass spectrometry (GC–MS) and liquid chromatography-Time-of-Flight mass spectrometry (LC-ESI-Q-ToF) were used to identify many of the SMPs which have relative masses up to 2 kDa. Our results showed that fermentation released much higher SMP concentrations compared to methanogenesis, especially in the range of 70 k–1000 k Da and 106–1500 Da. Alkanes, alkenes, alcohols, acids, and nitrogen-compounds were the major group of compounds identified in the supernatant of both fermentation and methanogenesis, and 71% of the compounds identified were found in both phases of digestion. Results from LC-ESI-Q-ToF analysis identified components of the cell membrane, such as phosphatidylglycerol,
⁎ Corresponding author at: Advanced Environmental Biotechnology Centre, Nanyang Environment & Water Research Institute, Nanyang Technological University, 1 Cleantech Loop, Clean Tech One, Singapore 637141, Singapore. E-mail addresses: [email protected] (C. Kunacheva), [email protected] (Y.N.A. Soh), [email protected] (D.C. Stuckey).
https://doi.org/10.1016/j.scitotenv.2019.134177 0048-9697/© 2019 Elsevier B.V. All rights reserved.
2
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
phosphatidylethanolamine and phosphatidylserine, as well as other compounds such as flavonoids, acylglycerol, terpene and terpenoids, benzenoid, glyceride, steroid and steroid derivatives. © 2019 Elsevier B.V. All rights reserved.
1. Introduction
2. Materials and methods
Soluble microbial products (SMPs) are defined as the soluble organic products created from microbial metabolism in biological processes that are not products, intermediates such as volatile fatty acids (VFAs), or incoming feed (Barker and Stuckey, 1999). SMPs can be derived from substrate metabolism, biomass growth, and biomass decay; growth associated SMPs, which are called utilization associated products (UAP), are produced directly from biomass growth and substrate metabolism. In contrast, biomass associated products (BAP) are produced as a result of biomass decay and cell lysis during endogenous decay. These SMPs affect the performance of most biological treatment systems (aerobic and anaerobic) as they are the dominant fraction of the effluent chemical oxygen demand (COD) (Aquino, 2004), and also cause membrane fouling in membrane processes (membrane bioreactors and water reclamation plants) (Jarusutthirak and Amy, 2006). Many effluent SMPs are degradable over time in both aerobic and anaerobic processes, however, the hydraulic retention time (HRT) in these processes often limits their complete degradation. Hence, minimizing SMP production should increase treatment performance, minimize chlorination by-products (DBPs), and reduce fouling in membrane bioreactors (Zhang et al., 2015). SMP production and transformation/biodegradation in biological systems are complex processes, and the origin of SMPs is still unclear. Past work has shown that most of the soluble organics found in the effluent are not present in the feed but are produced during microbial metabolism and cell lysis (Kunacheva et al., 2017a; Kunacheva et al., 2017b), and hence are SMPs. Furthermore, SMPs concentrations in the effluent increase under transient conditions, including nutrient limitations, the presence of toxic compounds, and/or when the feed flow or composition is changed radically (Kunacheva et al., 2017a; Kunacheva et al., 2017b; Aquino and Stuckey, 2003). Hence, it is important to identify the composition of SMPs and their molecular mass (Da), determine their origin, and understand what system parameters influence their production, and their contribution to membrane fouling. Until recently this was not possible due to poor and non-specific analytical methods, but now we have the tools to analyse SMPs in far greater depth, at least up to around 2 kDa relative masses (Kunacheva et al., 2017c). Early workers in this field (Kuo et al., 1996; Noguera et al., 1994) found higher concentrations of SMPs in the fermentation phase (glucose fed) in contrast to methanogenesis (acetate fed). Kuo et al. (Kuo et al., 1996) postulated that this was due to the conversion of glucose to CH4 and CO2 being more energetically favourable than the degradation of acetate to the same end products; this gives higher bacterial yields, and should lead to higher concentrations of SMP. In addition, glucose will support different groups of microorganisms (i.e. fermenters, acetogens and methanogens), while acetate will support mainly acetoclastic methanogens, and hence a more diverse population would produce higher concentrations of SMPs (Kuo et al., 1996). However, no one has ever looked at whether the SMPs produced in each phase are similar, or very different. Hence, this study focuses on the identification and production of SMPs in the effluent from an anaerobic process; to further understand the specific origins of SMPs, and how they are produced, a comparison of separate batch experiments with fermentative and methanogenic organisms was carried out to see how much each phase contributes, what the composition of these compounds is, and whether they contribute similar compounds. Ultimately, this work may lead to an understanding about how we could control the different biological processes in anaerobic digestion and improve effluent quality and membrane fouling.
2.1. Reagents and chemicals Methanol (LC-MS grade) was purchased from Sigma-Aldrich Pte Ltd. (Singapore). Acetone, chloroform, dichloromethane, and n-hexane (GC–MS grade or equivalent) were purchased from Merck Pte Ltd. (Singapore). Other solvents such as diethyl ether, ethyl acetate, nheptane, methyl tert-butyl ether and toluene were of chromatographic grade and purchased from Fisher Scientific Pte Ltd. (Singapore). Formic, acetic, propionic, isobutyric, butyric, isovaleric and valeric acid (analytical grade) and the alkane standard mixture (C10 - C40, all 50 mg/L each) were purchased from Sigma-Aldrich Pte Ltd. (Singapore). Deionised water was obtained from a MilliQ system (Millipore Advantage A10). 2.2. Batch reactor 2.2.1. Anaerobic batch reactor A cylindrical batch reactor was made from polymethyl methacrylate (Plexiglas) and had a working volume of 6 L, with 2 L of headspace. This anaerobic sequencing batch reactor was initially seeded with an inoculum of screened anaerobic sludge (bacteria) from a conventional domestic wastewater treatment plant in Singapore (Ulu Pandan). To ensure that steady state conditions were achieved, the reactors were run for at least 60 days at an effective HRT of 7 days (feed and decant) and at a 200 day SRT before the experiment started. The anaerobic bioreactor was fed with a synthetic feed comprised of glucose, peptone, meat extract and essential nutrients at a concentration of 4 gCOD/L. After the final feed aliquot was added at time zero, the reactors were not fed again. An initial (0 h) sample was collected just after the feed was added and fully mixed with the sludge; samples were then collected every 2 h in the first day, and then every 12 h until 456 h, and were filtered through a 0.45 μm glass fibre filter before analysis. The mixed liquor volatile suspended solids (MLVSS) in the reactor were controlled at 6000 mg/L. The reactor temperature was controlled at 35 ± 1 °C, while excess biogas was collected using a polyethylene gasbag, and measured using a gas pump and meter. 2.2.2. Enrichment batch cultures (fermentation and methanogenesis) Fermentation and methanogenesis batch tests were carried out using duplicate 1 L glass bottles connected to gasbags. The fermentation bottles were fed with glucose (2734 mg/L), peptone (818 mg/L), meat extract (274 mg/L) and essential nutrients combined at a concentration of 4 gCOD/L, while the methanogenesis bottles were fed with acetate (859 mg/L), peptone (818 mg/L), meat extract (274 mg/L) and essential nutrients combined at a concentration of 2 gCOD/L. We selected 2 gCOD/L of acetate because the results from the first section showed that acetate was at the highest concentration at 2 gCOD/L (Fig. 1), and 2 g/L of sludge inoculum from the 6 L batch reactor were used in each bottle. To inhibit the activity of methanogens inside the fermentation batch, 50 mM of bromoethanesulfonic acid (BESA) was added. BESA is a structural analog and competitive inhibitor of coenzyme M (Taylor and Wolfe, 1974) in methanogens (Balch and Wolfe, 1979), and is a potent inhibitor of methanogens at low concentrations (Aguilar et al., 1995). It is possible that BESA inhibition could have influenced SMP production, but at present it is the best inhibitor of methanogenic activity. All the bottles were adjusted to pH 7 at the beginning of the experiment and then placed in a controlled temperature shaker at 35 °C, 150 rpm.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
Fig. 1. COD equivalent concentrations of glucose and VFAs in the supernatant of a batch reactor.
2.3. Analysis of various parameters 2.3.1. General parameters All samples were filtered through 0.45 μm glass fibre filters, and then analysed in duplicate for VFAs and COD. The concentration of SMPs is typically estimated by subtracting the COD due to intermediate VFAs and residual substrate, from the soluble effluent COD. Total suspended solids (TSS), MLVSS and COD were measured as described in Standard Methods APHA (Eaton and Franson, 2005). VFAs were measured using a Shimadzu high-performance liquid chromatography (HPLC, SPD20AD) with a UV diode array detector (DAD, SPD-M20A) at 210 nm using an Aminex HPX-87H (300 × 7.8 mm) column. Analysis time was 25 min for each sample operating under isocratic and isothermal conditions using 0.005 M H2SO4 as the mobile phase at a flowrate 0.8 mL/min at 55 °C (Guerrant et al., 1982). A total of seven VFAs were quantified with this method, and Coefficients of variation (COV = SD/average value) for all VFAs were below ±6%. The composition of biogas (methane, oxygen, nitrogen, and carbon dioxide) was determined using a Shimadzu GC-2010plus gas chromatograph with a thermal conductivity detector (TCD) (COV below ±3%), and a select permanent gases/CO2 (CP7429) column (Agilent) was used for gas separation. 2.3.2. Molecular weight distribution Samples were filtered through a 0.2 μm syringe filter, and their MW distribution analysed using size exclusion chromatography (SEC). SEC was carried out using two columns (PolySep GFC-P1000 and 4000, Phenomenex) connected in series, with UV-DAD and refractive index detectors (RID, Shimadzu) to analyse both proteins and carbohydrates, respectively. EasiVial polymer standards (Agilent, U.S.A.) were used for molecular weight (MW) calibration. 2.3.3. Monosaccharides analysis Polysaccharides were hydrolysed by trifluoroacetic acid (4 M) at 100 °C for 2 h, and the monosaccharides released were derivatized by the alditol acetate method (Fox et al., 1989); eight neutral sugars were analysed using inositol as the internal standard. The detection and quantification was done using gas chromatography (GC-2010 plus, Shimadzu) on a 30 m × 0.25 mm × 0.25 μm RTX-5MS column (Restek, Bellefonte, PA, USA) coupled with a mass spectrometry (GC–MSQP2010, Shimadzu).
3
2017c). In brief, SPE was carried out using two Waters OasisHLB cartridges in series, and the analytes eluted with 2 mL of selected solvents (methanol, acetone, dichloromethane, n-hexane) in sequence. The eluent from each SPE cartridge was collected and further pre-treated using LLE. With LLE, a mixed solvent (hexane:dichloromethane:chloroform, 1:1:1) was used for extraction, and after lyophilisation 1 mL of dichloromethane was used for reconstitution for chromatographic analysis. Eluted samples were then analysed using a GC–MS system (GCMSQP2010ULTRA, Shimadzu); the sample was injected onto an RTX-5MS (30 m × 0.25 mm ID, Restek) column. The total runtime per sample was 60 min, and the temperature program was: 50 °C, hold 7 min, rate 7 °C/min to 325 °C, and hold for 14 min. Mass spectra were acquired from m/z 30 to 580 after a 10 min solvent cut time. However, based on the separation techniques this method does not separate very polar solutes well, and hence they are not well represented in the SMPs. The chromatographic peaks were identified using the NIST11 library (National Institute of Standards and Technology, Gaithersburg, MD, USA, http://www.nist.gov/srd/mslist.htm). Similarity index, mass spectrum and retention index were all used as selection criteria for compound identification from the NIST library list of suggested compounds (Jonsson et al., 2005; Kumari et al., 2011). Method blanks (deionised water), including reactor parts and tubing soaked in DI water, were run through the same pretreatment and analysis, while feed samples were also analysed to identify compounds in the feed and the concentration of these compounds were subtracted from the effluent samples in order to “normalise” net SMP production. Supernatant samples from the methanogenic and fermentation cultures were also extracted and measured with the method that we previously developed (Tipthara et al., 2017) using LC-ESI-Q-ToF. LLE was performed by adding 3 mL of 2:1 (v/v) chloroform/methanol; this mixture was then centrifuged at 1610g, at 20 °C for a further 10 min. The upper aqueous (metabolite) fraction and lower organic (lipid) fraction were transferred to separate 1.5 mL tubes and dried in a vacuum concentrator. UPLC-MS analyses were conducted using an ACQUITY UPLC with a 2777C autosampler (Waters Corp, USA) coupled to a Xevo G2XS QTof mass spectrometer (Waters, Manchester, UK) with an electrospray ionization (ESI) source. A mass range of m/z 100 to 2000 was acquired for both metabolites and lipids. The metabolite extracts were separated using an ACQUITY UPLC HSS T3 column (2.1 × 100 mm, 1.8 μm; Waters) at 40 °C. Three replicate injections were made for each ESI positive and ESI negative mode. Mobile phase A consisted of deionised water with 0.1% formic acid, while mobile phase B consisted of acetonitrile with 0.1% formic acid. The elution gradient was as follows: 0–40% B (1.0–11.0 min), 40–80% B (11.0–11.1 min), 80% B (11.1–13.0 min), 80–0% B (13.0–13.1 min), 0% B (13.1–16 min) with a flow rate of 0.4 mL/min. The lipid extracts were analysed separately by reconstituting them in 2:1:1(v/v/v) isopropanol/acetonitrile/water, and separating them using an ACQUITY UPLC CSH C18 column (2.1 × 100 mm, 1.7 μm; Waters) at 55 °C. Three replicate injections were made for each ESI positive and ESI negative mode. Mobile phase A consisted of 60:40 (v/v) acetonitrile/deionised water with 10 mM ammonium formate and 0.1% formic acid. Mobile phase B consisted of 90:10 (v/v) isopropanol/acetonitrile with 10 mM ammonium formate and 0.1% formic acid. The elution gradient was set as follows: 40–43% B (0.0–2.0 min), 43–50% B (2.0–2.1 min), 50–54% B (2.1–12.0 min), 54–70% B (12.0–12.1 min), 70–99% B (12.1–18 min), 99–40% B (18.0–18.1 min), 40% B (18.1–20.0 min), with a flow rate of 0.4 mL/min. 3. Results and discussion 3.1. Anaerobic batch reactor
2.4. Identification of low MW compounds in SMPs Organic compounds were separated using a combination of solidphase (SPE) and liquid-liquid extraction (LLE) (Kunacheva et al.,
The 6 L fed batch reactor after running for 60 days at an HRT of 7 days, was fed one time with 4 gCOD/L of glucose. Fig. 1 shows the equivalent COD of effluent, glucose and VFAs in the supernatant over
4
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
time. The 0 h sample was collected 5 min after feeding, which showed the concentration of glucose had already reduced to 2.7 gCOD/L. Glucose (MW = 180 Da) was rapidly consumed by microorganisms in the reactor, and was gone within 6 h (Fig. 2, RID peak at 20 min). VFAs concentrations started to rise rapidly after the 0 h sample (5 min), and peaked at 4 h, showing that fermentation was the major mechanism for glucose removal during this period. Acetate and propionate were the major VFAs present, while formate was also found in the first 4 h; the presence of formate has also been noted after shock loads to anaerobic systems in past work (Kunacheva et al., 2017b; Nachaiyasit and Stuckey, 1995). The VFAs then started to decrease when the methanogens started to metabolise them, while the SMPs started to increase; the highest SMP concentration occurred after 6–8 h, and accounted for 1.2 gCOD/L (44% of total soluble COD). VFAs, SMPs and methane accounted for 36%, 32% and 21% of feed COD, respectively during 6–8 h. The other 11% can be accounted for as dissolved methane in the supernatant, errors in the multiple analyses, and cell growth. After 8 h both the SMPs and VFAs gradually decreased, and the VFAs were totally gone after 96 h. There are several hypotheses as to why the SMP concentrations decreased, such as; high SMP production by fermentative/acidogenic bacteria compared to methanogens, hydrolysis of SMPs, and/or consumption of SMPs by microorganisms in the anaerobic system. However, SMPs still remained in the supernatant at concentrations of around 200 mg/L (~5% of the incoming feed) even after 456 h, which
may represent the anaerobically non-biodegradable SMPs (Schiener et al., 1998). Fig. 2 shows the SECs of the supernatant from the batch reactor. The RID chromatogram shows that glucose was totally consumed in 6 h (peaks between 106 and 1500 Da), while a peak of high molecular carbohydrates occurred at 14.7 min (70 k–100 k Da) in the 4 and 6 h samples; it was clear that this peak was formed when the acidogens were active (Fig. 2a). The chromatogram at UV210nm shows that the high MW compounds (70 k–100 k Da) decreased which may have broken down to 50 k–70 k Da solutes during the 4–6 h period (Fig. 2b), and then further breakdown to smaller ones (500–5 k Da) during 8–24 h (Fig. 2c). The 500–5 k Da compounds were then not found in the 48 h sample, which shows only two peaks in the range of 100–300 Da, while smaller compounds were not found after 96 h, and there was only one small peak at 14 min (100 kDa) which may represent the anaerobically non-biodegradable SMPs. These results show that there might be a big difference in SMP production between fermentation and methanogenesis. Fig. 3 shows the monosaccharide concentrations in the supernatant; glucose, mannose, and galactose were the dominant sugars which are often found as major constituents of bacterial extracellular polymers (ECPs) (Dignac et al., 1998). Based on Fig. 2a, glucose in the feed was totally consumed within 6 h, however, there were also long chain carbohydrates that appear to have been broken down to glucose which
Fig. 2. Chromatograms from SEC of the supernatant of the batch reactor at different times using different detectors [RID (a) and UV210nm (b), (c)] (n = 3). Dotted lines are from MW standards.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
5
were detected using the monosaccharide analysis. ECPs released by the acetogens were possibly the cause of this increment in glucose (and mannose) after 4 h since it is possible that such ECPs were biomass associated products (BAPs) released into the medium as a result of increased starvation and cell death. Furthermore, the decrease in carbohydrates after 8 h could have been the result of SMP consumption by methanogens, which was also shown in a previous study where methanogens might have contributed more than fermentative/ acidogenic bacteria to SMP consumption (Wu and Zhou, 2010), and/or consumption by acidogens. 3.2. Enrichment cultures (fermentation and methanogen batch tests) From the results of the anaerobic batch reactor (Section 3.1), enrichment cultures (fermentation/acidogenesis and methanogenesis) were carried out to obtain a better understanding of SMP production from each phase of anaerobic digestion. The fermentation culture was fed with 4 gCOD/L glucose, and the COD did not change much over the 7 days since fermentation basically transforms glucose to VFAs (methanogenesis was inhibited by BESA) (Fig. 4a). Unlike the mixed anaerobic batch reactor, acetate was the only dominant VFA in this fermentation culture, although formate was also detected in the first 2 days similar to the mixed batch reactor in Section 3.1. There were also large peaks of other acids in the chromatograms during the experiment (oxalic and pyruvic) which were detected using HPLC; although this method was not optimized to quantify these acids (Guerrant et al., 1982). This would explain why there was a big difference between COD and VFAs (as COD) values. Fig. 4b shows the performance of the methanogenic culture; acetate was consumed from 2 to 1 gCOD/L in 7 days, while methane was produced continuously from day 1 until day 7. Fig. 5a shows the SEC chromatogram of the batch fermentation test. The biggest peak was around 60 kDa on day 1, but this changed gradually to a higher MW (80 kDa) on day 7; this peak was also seen in the mixed anaerobic batch reactor (Fig. 2), but not in the chromatogram from the methanogenic reactor (Fig. 5b). Hence it can be concluded that the fermentation process released the 60–80 kDa SMPs, while the lower MW peaks were similar in both fermentation and methanogenesis, especially initially. These peaks started with smaller compounds in the range of 100–300 Da on day 1, and the peak kept increasing in size (higher MWs) to about 1000 to 2000 Da (Fig. 5); these compounds were identified using GC–MS and a LC-ESI-Q-ToF in the next section. Overall, fermentation released much higher MW SMPs compared to methanogenesis, and this was also found in previous studies (Noguera et al., 1994; Wu and Zhou, 2010).
Fig. 4. COD concentrations in the fermentation and methanogenic batch tests (n = 4).
3.3. Identification of low MW SMPs (MW b 2000 Da) in enrichment cultures Identifying SMPs was done using SPE and LLE coupled with GC–MS and LC-ESI-Q-ToF analysis (n = 4). Supernatant samples were collected on day 7 from both the fermentation and methanogenesis reactors, and a total of 77 compounds were identified by GC–MS (MW b 580). Alkanes, alkenes, alcohols, acids, and nitrogen-compounds were the major groups of compounds identified in both fermentation and methanogenesis, and these compounds have been found in previous studies (Kunacheva et al., 2017b; Wu and Zhou, 2010; Citron et al., 2012; Trzcinski and Stuckey, 2009). Fig. 6 shows the chromatograms
Fig. 3. Monosaccharide concentrations in the supernatant of the anaerobic batch reactor after hydrolysis of polysaccharides (n = 3).
6
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
Fig. 5. SEC chromatogram of the fermentation (F) and methanogenic (M) batch tests.
(LLE) of fermentation and methanogenesis analysed by GC–MS; 71% of the compounds identified were actually the same in both fermentation and methanogenesis. Due to our specific analytical methodology, this is the first time this has ever been noted in the literature. From the literature it is clear that alkanes and alkenes can be synthesized by Sarcina lutea through substrate metabolism that utilizes fatty acids (Albro and Dittmer, 1970). Furthermore, in a recent study Shi et al. also found that a laccase enzyme from the endophyte Pantoea ananatis Sd-1 could degrade lignin (complex hydrocarbon) and produce metabolites such as alkanes and acids (Shi et al., 2015), similar to those found in our study. Also, a study on volatile compounds released by 50 bacterial strains revealed that bacteria can produce esters, alcohols, nitrogen-containing as well as sulfur compounds (Citron et al., 2012). Micromonospora aurantiaca, which are naturally present in both soil and water, can also produce a large diversity of VFAs and fatty acid methyl esters (Dickschat et al., 2011), while further studies found that N-compounds, acids and esters are precursors for disinfection by– products (Zhang et al., 2015; Liu and Zhang, 2014). Fig. 7 shows that compounds between 100 and 2000 Da were identified using LC-ESI-Q-ToF with methods for metabolites and lipids (Tipthara et al., 2017). Qualitative analysis was carried out for chromatographic peaks of signal intensity 1000 and above, and compounds identified from both positive and negative lipids and metabolites scans are summarised in Tables S2 and S3. In contrast to the data collected from SEC and SPE-LLE-GC–MS, results from the LC-ESI-Q-ToF analysis showed that there were more compounds identified in the methanogenesis experiment than the fermentation enrichment culture. Sphingolipids, glycerophospholipids and carbohydrate derivatives were found in higher abundance in the fermentation culture, whereas the abundance of glycerides, terpenes and terpenoids, and benzenoids were higher in the methanogenic culture. Steroids were found in similar amounts in both enrichment cultures (Fig. 7). A previous study of SMPs using the same LC-ESI-Q-ToF method had also identified components of cell membranes, such as phosphatidylglycerol (PG), phosphatidylethanolamine (PE) and phosphatidylserine (PS) (Tipthara et al., 2017). It is known that the outer membrane of E. coli is comprised of approximately 5% cardiolipin (CL), 20 to 25% PG, 70 to 80% PE, as well as a small percentage of PS (Romantsov et al., 2009), and PE and PG were reported to play significant roles in biofilm formation (Benamara et al., 2014; Ni et al., 2009). The composition of membrane biolipids has been of considerable interest,
Fig. 6. Chromatograms (LLE) of fermentation and methanogenesis reactor supernatants analysed by GC–MS (MW b 580 Da).
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
7
Fig. 7. Metabolites and lipids identified by LC-ESI-Q-ToF (MWb2,000) where signal intensity was higher in the methanogenic reactor than in fermentation reactor (A) and vice versa (B).
especially in research groups studying antimicrobial mechanisms and microbial ecology, and laboratory studies and computational microbiology have demonstrated the ability of bacteria to alter their membrane lipid composition to improve their adaptability to the physical environment (Lindblom et al., 2002; Matsumoto et al., 2006; Zhang and Rock, 2008). This current study of SMPs found during fermentation and methanogenesis, the samples for SMP extraction were only taken on the last day of the experiment, i.e. Day 7. Further work on intermittent sampling is required to monitor the changes of SMPs over the batch run. In this study, the compounds identified were more diverse than our earlier studies (Kunacheva et al., 2017a; Kunacheva et al., 2017b; Trzcinski and Stuckey, 2010), and included flavonoids, as well as acylglycerol, terpene and terpenoids, benzenoid, glyceride, steroid and steroid derivatives, while the Human Metabolome Database (HMDB) used to analyse our data suggested the presence of some plant-derived compounds. 58 out of 136 compounds identified in group (A) (Fig. 7), and 12 out of 44 compounds identified in group (B) (Fig. 7) were classified as plant-derived compounds, i.e. isolated from roots, bark, leaves or fruits of plants as identified on the HMDB. Interestingly, despite the fact that these enrichment cultures should contain only acidogenic or methanogenic bacterial communities, and hence the compounds detected in the cultures should only be produced by such microorganisms, many of these compounds are well-known plant-derived substances. One explanation for this is that they might have been produced by photosynthetic bacteria (Pfennig, 1967). Several classes of sphingolipids were detected using high performance chromatography, HPLC, GC–MS and LC-MS, and have been reported in various forms, such as sphingomyelin
(SM), sphingophospholipid, aminoglycosphingolipid, sphingoglycolipid, and ceramide. Further studies on bacterial lipid components revealed the presence of sphingolipids in acetic acid-producing bacteria under stressed conditions such as elevated temperatures and low pH, where they were biotransformed into ceramides to defend themselves against the harsh environment. This is consistent with the current finding which identified sphingolipids and glycerophospholipids in higher quantities in the fermentation culture. Other components of the membranes of acid-producing bacteria identified were phospholipids, terpenoids and aminolipids (Naka et al., 2000). In contrast, the current work found more terpenes and terpenoids in the compounds identified from the methanogenic culture. To date, there appears to be nothing in the literature on this subgroup of sphingolipids being produced by other bacteria. Bacterial membranes contain a majority of PE, and PG and PG derivatives such as diphosphatidylglycerol and CL, and the lipid composition depends on the class of bacteria; this group of compounds were found mainly in the fermentation culture. The inner membranes of gramnegative bacteria generally contain higher amounts of PE, while the cytoplasmic membranes of gram-positive bacteria are comprised predominantly of PG and its derivatives; the widely studied E. coli was found to have approximately 82% PE by weight in the inner cytoplasmic membrane, whereas the membrane of Staphylococcus aureus is mainly PGs and its derivatives. These different lipid compositions determine the membrane properties such as its general structure, packing density, rigidity, and overall surface charge density (Navas et al., 2005), and Zhao, Rog, Gurtovenko, Vattulainen and Karttunen (Zhao et al., 2008) proposed that PG contributed significantly to the stability of bacterial
8
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177
membranes. Phosphatidylcholines (PC) and PE are neutral zwitterionic species under physiological conditions, whereas PG are negatively charged. The presence of PG as part of the membrane structure was shown to improve cell structural stability as it increases ionic electrostatic charges and inter-molecular hydrogen bonding between the PE molecules present in the lipid bilayer. This enhances membrane integrity during events of intercellular interactions, and external ‘disruptions’ in case of the presence of foreign molecules such as antimicrobial peptides (Zhao et al., 2008). The broad class of terpenes, terpenoids and benzenoids are found in higher amounts in the methanogenic culture in this experiment. The detection of triterpenoids and their derivatives in this study have been identified as compounds originating from plants by the HMDB, however, a review by Taylor and Davies (Taylor and Davies, 1974) introduced earlier research carried out on the identification, biosynthesis, and functionality of this group of compounds which showed that they can originate from non-photosynthetic bacterial species. By mapping of biosynthetic pathways through past reports on terpenoids and the precursor isoprene, non-photosynthetic bacteria were proven to be able to synthesize isoprenoids, and triterpenoids such as squalene and hopanes de novo (Taylor and Davies, 1974). Finally, triacylglycerol or triglycerides (TG) have been found to be produced by both gram-positive and gram-negative bacteria that belong to the actinomycetes group, such as Streptomyces, Nocardia, Rhodococcus, Mycobacterium, Dietzia and Gordonia, for carbon and energy storage purposes during stagnant growth, for instance in adverse environmental conditions. These triglycerides showed higher signal intensities in the methanogenic culture supernatant, are also the precursors for phospholipid biosynthesis, which is a major component of cell membranes (Alvarez and Steinbuchel, 2002; Athenstaedt and Daum, 2006). An earlier review by Norbert Pfennig reported that the photosynthetic route of bacteria proceeded under anaerobic conditions and in minimum light, and photosynthetic bacteria utilize external electron donors such as sulfur and other organic compounds. The carotenoids play an important role in the energy transfer chain during photosynthesis, and prevent the destruction of anaerobic cells by bacteriochlorophyll during light absorption in the presence of oxygen (Griffiths et al., 1955; Slouf et al., 2012), and the carotenoid, mutatochrome, was also detected in the enrichment cultures in this study. 4. Conclusion This study showed that both the fermentation (acidogens) and methanogenic phases released different types of SMPs. Acidogens released much higher high MW compounds (MW N 70 kDa) compared to methanogens, however, the lower MW compounds (MW b 2000 Da) were quite similar in both processes (about 71%). Analysis using LCESI-Q-ToF revealed that some SMPs are derived from cell wall fragments, and from compounds released by the cell to adjust to the surrounding environment, such as lipids and carbohydrates. Sphingolipids, glycerophospholipids and carbohydrate derivatives were found in higher abundance in the fermentation culture, whereas the abundance of glycerides, terpenes and terpenoids, and benzenoids were higher in the methanogenic culture. For future work, it would be important to try and link changes in microbial ecology to these differences in SMPs, and investigate how changes in the type of SMPs changes membrane fouling. Declaration of Competing Interest None declared under financial, general and institutional competing interests. Acknowledgement This research work was supported by the Singapore National Research Foundation under its Environmental & Water Technologies
Strategic Research Programme, and administered by the Environment & Water Industry Programme Office (EWI) of the PUB. Appendix A. Supplementary data Supplementary data to this article can be found online at https://doi. org/10.1016/j.scitotenv.2019.134177.
References Aguilar, A., Casas, C., Lema, J.M., 1995. Degradation of volatile fatty-acids by differently enriched methanogenic cultures - kinetics and inhibition. Water Res. 29, 505–509. Albro, P.W., Dittmer, J.C., 1970. Bacterial hydrocarbons - occurrence, structure and metabolism. Lipids 5 (320-&). Alvarez, H.M., Steinbuchel, A., 2002. Triacylglycerols in prokaryotic microorganisms. Appl. Microbiol. Biotechnol. 60, 367–376. Aquino, S.F., 2004. Formation of Soluble Microbial Products (SMP) in Anaerobic Reactors during Stress Conditions. Department of Chemical Engineering and Chemical Technology, Imperial College of Science Technology and Medicine, Imperial College of Science Technology and Medicine, London SW7 2BY. Aquino, S.F., Stuckey, D.C., 2003. Production of soluble microbial products (SMP) in anaerobic chemostats under nutrient deficiency. J. Environ. Eng. ASCE 129, 1007–1014. Athenstaedt, K., Daum, G., 2006. The life cycle of neutral lipids: synthesis, storage and degradation. Cell. Mol. Life Sci. 63, 1355–1369. Balch, W.E., Wolfe, R.S., 1979. Specificity and biological distribution of coenzyme M (2mercaptoethanesulfonic acid). J. Bacteriol. 137, 256–263. Barker, D.J., Stuckey, D.C., 1999. A review of soluble microbial products (SMP) in wastewater treatment systems. Water Res. 33, 3063–3082. Benamara, H., Rihouey, C., Abbes, I., Ben Mlouka, M.A., Hardouin, J., Jouenne, T., Alexandre, S., 2014. Characterization of membrane lipidome changes in Pseudomonas aeruginosa during biofilm growth on glass wool. PLoS One 9. Citron, C.A., Rabe, P., Dickschat, J.S., 2012. The scent of bacteria: headspace analysis for the discovery of natural products. J. Nat. Prod. 75, 1765–1776. Dickschat, J.S., Bruns, H., Riclea, R., 2011. Novel fatty acid methyl esters from the actinomycete Micromonospora aurantiaca. Beilstein J. Org. Chem. 7, 1697–1712. Dignac, M.F., Urbain, V., Rybacki, D., Bruchet, A., Snidaro, D., Scribe, P., 1998. Chemical description of extracellular polymers: implication on activated sludge floc structure. Water Sci. Technol. 38, 45–53. Eaton, A.D., Franson, M.A.H., 2005. Standard Methods for the Examination of Water & Wastewater. American Public Health Association. Fox, A., Morgan, S.L., Gilbart, J., 1989. Preparation of alditol acetates and their analysis by gas chromatography (GC) and mass spectrometry (MS). Analysis of Carbohydrates by GLC and MS, pp. 87–117. Griffiths, M., Sistrom, W.R., Cohenbazire, G., Stanier, R.Y., 1955. Function of carotenoids in photosynthesis. Nature 176, 1211–1214. Guerrant, G.O., Lambert, M.A., Moss, C.W., 1982. Analysis of short-chain acids from anaerobic-bacteria by high-performance liquid-chromatography. J. Clin. Microbiol. 16, 355–360. Jarusutthirak, C., Amy, G., 2006. Role of soluble microbial products (SMP) in membrane fouling and flux decline. Environ. Sci. Technol. 40, 969–974. Jonsson, P., Johansson, A.I., Gullberg, J., Trygg, J., A, J., Grung, B., Marklund, S., Sjostrom, M., Antti, H., Moritz, T., 2005. High-throughput data analysis for detecting and identifying differences between samples in GC/MS-based metabolomic analyses. Anal. Chem. 77, 5635–5642. Kumari, S., Stevens, D., Kind, T., Denkert, C., Fiehn, O., 2011. Applying in-silico retention index and mass spectra matching for identification of unknown metabolites in accurate mass GC-TOF mass spectrometry. Anal. Chem. 83, 5895–5902. Kunacheva, C., Soh, Y.N.A., Stuckey, D.C., 2017a. Effect of feed pH on reactor performance and production of soluble microbial products (SMPs) in a submerged anaerobic membrane bioreactor. Chem. Eng. J. 320, 135–143. Kunacheva, C., Soh, Y.N.A., Trzcinski, A.P., Stuckey, D.C., 2017b. Soluble microbial products (SMPs) in the effluent from a submerged anaerobic membrane bioreactor (SAMBR) under different HRTs and transient loading conditions. Chem. Eng. J. 311, 72–81. Kunacheva, C., Le, C.C., Soh, Y.N.A., Stuckey, D.C., 2017c. Chemical characterization of low molecular weight soluble microbial products in an anaerobic membrane bioreactor. Environ. Sci. Technol. 51, 2254–2261. Kuo, W.C., Sneve, M.A., Parkin, G.F., 1996. Formation of soluble microbial products during anaerobic treatment. Water Environ. Res. 68, 279–285. Lindblom, G., Oradd, G., Rilfors, L., Morein, S., 2002. Regulation of lipid composition in Acholeplasma laidlawii and Escherichia coli membranes: NMR studies of lipid lateral diffusion at different growth temperatures. Biochemistry-Us 41, 11512–11515. Liu, J.Q., Zhang, X.R., 2014. Comparative toxicity of new halophenolic DBPs in chlorinated saline wastewater effluents against a marine alga: Halophenolic DBPs are generally more toxic than haloaliphatic ones. Water Res. 65, 64–72. Matsumoto, K., Kusaka, J., Nishibori, A., Hara, H., 2006. Lipid domains in bacterial membranes. Mol. Microbiol. 61, 1110–1117. Nachaiyasit, S., Stuckey, D.C., 1995. Microbial response to environmental-changes in an anaerobic baffled reactor (Abr). Anton Leeuw. Int. J. G. 67, 111–123. Naka, T., Fujiwara, N., Yabuuchi, E., Doe, M., Kobayashi, K., Kato, Y., Yano, I., 2000. A novel sphingoglycolipid containing galacturonic acid and 2-hydroxy fatty acid in cellular lipids of Sphingomonas yanoikuyae. J. Bacteriol. 182, 2660–2663.
C. Kunacheva et al. / Science of the Total Environment 698 (2020) 134177 Navas, B.P., Lohner, K., Deutsch, G., Sevcsik, S., Riske, K.A., Dimova, R., Garidel, P., Pabst, G., 2005. Composition dependence of vesicle morphology and mixing properties in a bacterial model membrane system. Bba-Biomembranes 1716, 40–48. Ni, B.-J., Fang, F., Xie, W.-M., Sun, M., Sheng, G.-P., Li, W.-H., Yu, H.-Q., 2009. Characterization of extracellular polymeric substances produced by mixed microorganisms in activated sludge with gel-permeating chromatography, excitation–emission matrix fluorescence spectroscopy measurement and kinetic modeling. Water Res. 43, 1350–1358. Noguera, D.R., Araki, N., Rittmann, B.E., 1994. Soluble microbial products (SMP) in anaerobic chemostats. Biotechnol. Bioeng. 44, 1040–1047. Pfennig, N., 1967. Photosynthetic bacteria. Annu. Rev. Microbiol. 21 (285-&). Romantsov, T., Guan, Z.Q., Wood, J.M., 2009. Cardiolipin and the osmotic stress responses of bacteria. Bba-Biomembranes 1788, 2092–2100. Schiener, P., Nachaiyasit, S., Stuckey, D., 1998. Production of soluble microbial products (SMP) in an anaerobic baffled reactor: composition, biodegradability, and the effect of process parameters. Environ. Technol. 19, 391–399. Shi, X.W., Liu, Q., Ma, J.S., Liao, H.D., Xiong, X.Q., Zhang, K.K., Wang, T.F., Liu, X.M., Xu, T., Yuan, S.S., Zhang, X., Zhu, Y.H., 2015. An acid-stable bacterial laccase identified from the endophyte Pantoea ananatis Sd-1 genome exhibiting lignin degradation and dye decolorization abilities. Biotechnol. Lett. 37, 2279–2288. Slouf, V., Chabera, P., Olsen, J.D., Martin, E.C., Qian, P., Hunter, C.N., Polivka, T., 2012. Photoprotection in a purple phototrophic bacterium mediated by oxygendependent alteration of carotenoid excited-state properties. Proc. Natl. Acad. Sci. U. S. A. 109, 8570–8575. Taylor, R.F., Davies, B.H., 1974. Triterpenoid carotenoids and related lipids - triterpenoid monohydroxy-carotenoids and monoglucosyloxy-carotenoids from streptococcusfaecium unh 564p. Biochem. J. 139, 761–769.
9
Taylor, C.D., Wolfe, R.S., 1974. Structure and methylation of coenzyme M (Hsch2ch2so3). J. Biol. Chem. 249, 4879–4885. Tipthara, P., Kunacheva, C., Soh, Y.N.A., Wong, S.C.C., Pin, N.S., Stuckey, D.C., Boehm, B.O., 2017. Global profiling of metabolite and lipid soluble microbial products in anaerobic wastewater reactor supernatant using UPLC-MSE. J. Proteome Res. 16, 559–570. Trzcinski, A.P., Stuckey, D.C., 2009. Anaerobic digestion of the organic fraction of municipal solid waste in a two-stage membrane process. Water Sci. Technol. 60, 1965–1978. Trzcinski, A.P., Stuckey, D.C., 2010. Treatment of municipal solid waste leachate using a submerged anaerobic membrane bioreactor at mesophilic and psychrophilic temperatures: analysis of recalcitrants in the permeate using GC-MS. Water Res. 44, 671–680. Wu, B., Zhou, W., 2010. Investigation of soluble microbial products in anaerobic wastewater treatment effluents. J. Chem. Technol. Biotechnol. 85, 1597–1603. Zhang, Y.M., Rock, C.O., 2008. Membrane lipid homeostasis in bacteria. Nat. Rev. Microbiol. 6, 222–233. Zhang, B.B., Xian, Q.M., Zhu, J.P., Li, A.M., Gong, T.T., 2015. Characterization, DBPs formation, and mutagenicity of soluble microbial products (SMPs) in wastewater under simulated stressful conditions. Chem. Eng. J. 279, 258–263. Zhao, W., Rog, T., Gurtovenko, A.A., Vattulainen, I., Karttunen, M., 2008. Role of phosphatidylglycerols in the stability of bacterial membranes. Biochimie 90, 930–938.