Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/springer-journals/adaptation-to-low-ph-and-lignocellulosic-inhibitors-resulting-in-OeAwFU7AE0
- Publisher
- Springer Journals
- Copyright
- Copyright © 2016 by The Author(s)
- Subject
- Life Sciences; Microbiology; Microbial Genetics and Genomics; Biotechnology
- eISSN
- 2191-0855
- DOI
- 10.1186/s13568-016-0234-8
- pmid
- 27566648
- Publisher site
- See Article on Publisher Site
Abstract
Lignocellulosic bioethanol from renewable feedstocks using Saccharomyces cerevisiae is a promising alternative to fossil fuels owing to environmental challenges. S. cerevisiae is frequently challenged by bacterial contamination and a combination of lignocellulosic inhibitors formed during the pre‑ treatment, in terms of growth, ethanol yield and productivity. We investigated the phenotypic robustness of a brewing yeast strain TMB3500 and its ability to adapt to low pH thereby preventing bacterial contamination along with lignocellulosic inhibitors by short‑ term adaptation and adaptive lab evolution (ALE). The short‑ term adaptation strategy was used to investigate the inherent ability of strain TMB3500 to activate a robust phenotype involving pre‑ culturing yeast cells in defined medium with lignocellulosic inhibitors at pH 5.0 until late exponential phase prior to inoculating them in defined media with the same inhibitor cocktail at pH 3.7. Adapted cells were able to grow aerobically, ferment anaerobically (glucose exhaustion by 19 ± 5 h −1 to yield 0.45 ± 0.01 g ethanol g glucose ) and portray significant detoxification of inhibitors at pH 3.7, when com‑ pared to non‑ adapted cells. ALE was performed to investigate whether a stable strain could be developed to grow and ferment at low pH with lignocellulosic inhibitors in a continuous suspension culture. Though a robust population was obtained after 3600 h with an ability to grow and ferment at pH 3.7 with inhibitors, inhibitor robustness was not stable as indicated by the characterisation of the evolved culture possibly due to phenotypic plasticity. With further research, this short‑ term adaptation and low pH strategy could be successfully applied in lignocellulosic ethanol plants to prevent bacterial contamination. Keywords: Saccharomyces cerevisiae, Low pH, Lignocellulosic inhibitors, Phenotypic robustness, Adaptation, Ethanol yield low production rates accompanied by low yields (Papout- Introduction sakis and Pronk 2015). Increasing concerns over the need for sustainable and A significant challenge in the fuel ethanol production scalable fuels as a means to curb global warming has is acute and chronic bacterial contamination, since the led to focus on bioethanol production from renewable incoming substrate might contain microorganisms and biomass, such as agricultural and industrial residues the fermentation is carried out in non-aseptic conditions (Limayem and Ricke 2012). Due to the decrease in costs (Skinner and Leathers 2004). Bacterial contamination is of petroleum as a response to recent discoveries of fos- predominantly due to lactic and acetic acid bacteria lead- sil fuel reserves, there is an intense emphasis on lowering ing to loss of fermentable sugars and micronutrients, the costs of renewable bioethanol by overcoming chal- increased by-product formation (lactic acid and ace- lenges connected to high substrate costs, low titers and tic acid), reduced ethanol yields and productivities and stuck fermentations (Beckner et al. 2011; Bischoff et al. 2009). Bacterial contamination has been studied exten- *Correspondence: ed.van_niel@tmb.lth.se Division of Applied Microbiology, Department of Chemistry, Lund sively (Bischoff et al. 2009; Skinner and Leathers 2004) University, P.O. Box 124, 221 00 Lund, Sweden and several antimicrobial strategies, including usage of Full list of author information is available at the end of the article © 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Narayanan et al. AMB Expr (2016) 6:59 Page 2 of 13 antibiotics, have been adopted in the first generation furfural and hydroxymethylfurfural (HMF), and (3) phe- bioethanol production (Muthaiyan et al. 2011). These nolic compounds such as vanillin, coniferyl aldehyde and methods are expensive and some are environmentally 4-hydroxybenzoic acid (Palmqvist and Hahn-Hägerdal invasive when used in large-scale fermentations (Muth- 2000; Taherzadeh and Karimi 2011). S. cerevisiae endures aiyan et al. 2011). Contamination might pose a bigger inhibitors through different mechanisms, including threat to lignocellulosic ethanol owing to the versatility detoxification by enzymatic reduction, efflux and mem - in sugar substrates. As Saccharomyces cerevisiae displays brane repair (Piotrowski et al. 2014). Advancements in glucose repression, it takes longer time to assimilate pre-treatment processes resulted in reduction of furans, other sugars including xylose and arabinose. Thus any phenolics, formic acid and levulinic acid concentra- contamination might be able to utilise the other sugars tions in the hydrolysate (Jönsson et al. 2013). However, swiftly and more efficiently than S. cerevisiae leading to acetic acid is naturally bound to lignocellulose in the reduction in ethanol production. Several attempts have form of acetyl sugars in the hemicellulose fraction and been made to study and control bacterial contamina- becomes de-acetylated during the hydrolysis treatment tion in lignocellulosic ethanol production including: (1) (Almeida et al. 2007). As a weak organic acid, its effect is adding NaCl and ethanol to wood hydrolysate (Albers more pronounced at low pH and may facilitate synergy et al. 2011), (2) high solid loading in simultaneous sac- between furans and phenolics (Ding et al. 2011). Hence, charification and fermentation (SSF) (Ishola et al. 2013), it is crucial to focus on yeast tolerance in acidic environ- (3) usage of an antibiotic like gentamicin and biomass ments inflicted by the combination of inorganic and weak autoclaving (Serate et al. 2015), and (4) usage of bacte- organic acids in the presence of lignocellulosic inhibitors riophages (Worley-Morse et al. 2015). These strategies for cost competitive ethanol production. encounter challenges including: (1) additional cost and Different rational engineering strategies have been need for extensive fine tuning and testing of concentra - pursued with S. cerevisiae to understand the molecu- tions of NaCl and ethanol (Albers et al. 2011), (2) loss of lar mechanisms involved in coping with one or several cell viability due to mechanical stress caused by solid par- inhibitors, thereby creating inhibitor tolerant S. cerevisiae ticles in high cell loading (Ishola et al. 2013), (3) cost and strains (Alriksson et al. 2010; Caspeta et al. 2015; Koichi environmental challenges posed by gentamicin, energy et al. 2012; Lei et al. 2011; Liu 2011; Parawira and Tekere expenditure and formation of inhibitors due to autoclav- 2011; Taherzadeh and Karimi 2011; Takuya et al. 2013). ing (Serate et al. 2015), and (4) rise of bacteriophage- As Meijnen et al. (2016) found that tolerance towards insensitive mutants and possibilities of gene transfer acetic acid is a result of a polygenic response from yeast, from bacteriophages to yeast (Worley-Morse et al. 2015). evolutionary adaptation might be a suitable strategy to One of the potentially scalable and economically feasi- improve tolerance towards low pH and acetic acid with ble solutions to control bacterial contamination is to run other lignocellulosic inhibitors since yeast might accrue the lignocellulosic fermentation at low pH, around pH beneficial properties under stress conditions over the 4 where the growth and viability of bacteria are drasti- time of evolution. Evolutionary engineering strategies cally reduced (Kádár et al. 2007). Additionally, yeast cells have been successfully pursued to obtain yeast strains are recycled in several commercial ethanol production with enhanced tolerance against individual or combina- processes up to 6 months to reduce fermentation time tions of several inhibitors in defined media (Dominik and and cost of yeast propagation, increasing the chances of Uwe 2008; Wright et al. 2011) or in hydrolysates (Almario contamination (Basso et al. 2011). To prevent contami- et al. 2013; Hanqi et al. 2014) with beneficial properties nation, yeast cells are treated with dilute sulphuric acid including better growth, improved viability, higher yield (H SO ) at pH between pH 1.8 and 2.5 for 1–2 h (Basso and ethanol productivity in comparison to the control 2 4 et al. 2011), which results in reduction in intracellular pH strains. (Beales 2004), yeast viability and low ethanol yield (De Yeast tolerance towards inhibitors could also be Melo et al. 2010). Hence, it might be efficient to develop induced by pre-cultivation with lower concentration of S. cerevisiae strains tolerant to lower pH induced by inor- inhibitors in defined media or diluted hydrolysates. This ganic acids. induces the general stress response leading to improved Apart from bacterial contamination, inhibitors pose growth and fermentation performance in the inhibitory another obstacle to yeast in ethanol production, formed medium or hydrolysate (Nielsen et al. 2015; Tomás-Pejó from the components of lignocellulose including cel- and Olsson 2015). Kádár et al. (2007) had improved the lulose, hemicellulose and lignin due to the harsh condi- tolerance of S. cerevisiae towards lignocellulosic inhibi- tions of biomass pre-treatment (Almeida et al. 2007). tors at pH 4, but with lower concentration of inhibitors to They include (1) weak organic acids such as acetic acid, potentially threaten yeast growth and fermentation. Low- formic acid and levulinic acid, (2) furans, including ering the culture pH to pH 4 in the presence of different Narayanan et al. AMB Expr (2016) 6:59 Page 3 of 13 −1 −1 concentrations of acetic acid in corn-mash has resulted in 0.5 g L HMF and 1 g L vanillin were used as inhibi- complete inhibition of ethanol production at acetic acid tors in this study, hereafter mentioned as inhibitor concentrations greater than 0.8 % weight/volume (w/v) cocktail (IC), unless mentioned otherwise. The con - (Graves et al. 2006). Yet, other industrial strains have centrations of different inhibitors chosen in this study been analysed for their low pH tolerance, such as JP1 and were in the ranges found in different pre-treated ligno - PE-2 from commercial ethanol production in Brazil (De cellulose hydrolysates obtained from barley straw, dilute Melo et al. 2010; Della-Bianca et al. 2014). We have previ- spruce and wheat straw (Almeida et al. 2009a). The used −1 ously demonstrated a strain-independent pre-cultivation vanillin concentration of 1 g L was 5–10 times higher strategy where S. cerevisiae cells can grow and ferment at than in the lignocellulosic hydrolysate to account for −1 pH 3.7 with lethal concentrations of 6 g L acetic acid different kinds of phenolic compounds and lignosul - −1 after a short-term adaptation with 6 g L acetic acid at fonates. Defined media were chosen over hydrolysates pH 5.0 (Sànchez i Nogué et al. 2013). to have better control over the experiments performed Though numerous endeavours have been pursued to since hydrolysate might contain many unknown inhibi- develop a robust S. cerevisiae strain in the presence of tors (Palmqvist and Hahn-Hägerdal 2000) and they inhibitors by targeted, evolutionary and pre-cultivation might change in composition over time. All media com- approaches, to our knowledge, there has been no inves- ponents were sterile filtered to avoid changes in compo - tigation on improving phenotypic robustness of S. cerevi- sition due to evaporation. siae to low pH with acetic acid and other lignocellulosic inhibitors. Herein, we aimed at developing a short-term Culture conditions adaptation strategy and an ALE of yeast in a chemostat Aerobic cultures were performed at 30 °C in a rotary to investigate the nature of adaptability of S. cerevisiae to shake-incubator (New Brunswick, Enfield, CT, USA) at the harsh conditions of low pH and lignocellulosic hydro- 180 rpm with cell concentrations determined as optical lysates and the stability of the acquired robustness. Fur- density (OD) at 620 nm (Spectrophotometer U-1800, thermore, understanding the interactive effects among Hitachi, Berkshire, UK). Seed cultures were grown from the inhibitors at low pH could pave the way for develop- single colonies of TMB3500 (YPD agar plate) in 5 mL ing new strains and strategies to lignocellulosic ethanol defined medium in a 50 mL conical tube to reach late production. exponential phase. Pre-cultures were started from the seed culture in defined medium with an initial OD of 0.5, Materials and methods grown till late exponential phase, unless mentioned oth- Yeast strain and media erwise. All the aerobic batch cultures were cultivated in The commercial S. cerevisiae brewer’s strain, Coo - baffled shake flasks with a medium volume equivalent to bra 6 Magnum (CBF Drinkit AB, Mölndal, Sweden) 10 % of the volume of the baffled shake flask to maintain was renamed to TMB3500 (Almeida et al. 2009a) and adequate aeration. Cells for inoculation were obtained used in this study. It was stored at −80 °C in yeast pep- after centrifuging the pre-culture at 4000 rpm for 5 min −1 tone dextrose (YPD) medium containing 10 g L yeast at 4 °C, washing the cells with saline and repeating the −1 −1 extract, 20 g L peptone and 20 g L glucose supple- centrifugation process. Gas proof neoprene tubes (Mas- mented with 30 % (v/v) glycerol and maintained on YPD terflex , Cole-Parmer, Sweden) were used for connec- −1 medium with 20 g L agar. All the chemicals were pur- tions in the anaerobic experiments to avoid oxygen chased from Sigma Aldrich, Sweden, unless mentioned diffusion. All the growth and fermentation experiments otherwise. except the ALE were carried out at least in biological rep- A chemically defined medium (Verduyn et al. 1992) licates and measurements were carried out in technical −1 with 20 g L glucose, buffered with 50 mM potassium triplicates. Data represented in figures include standard hydrogen phthalate and 20 mM potassium hydroxide deviations from the replicates. (KOH) (Hahn-Hägerdal et al. 2005) was used in all aero- bic growth experiments. The pH of the defined medium Short‑term adaptation was adjusted using 3 M H SO and 3 M KOH. Ergos- Pre-cultures were grown aerobically until late exponential 2 4 terol and Tween80 were added in anaerobic experi- phase in 25 mL of defined medium with the IC at pH 5.0 −1 ments, in the final concentrations of 0.01 and 0.42 g L , from the seed culture, termed as short-term adaptation −1 respectively. A silicone based antifoam (0.5 mL L ) was step being used for subsequent cultivations. Cells grown added in the experiments performed in fermentors to until late exponential phase in defined medium without avoid excessive foaming (Dow Corning Antifoam RD the IC at pH 5.0 was used as negative control, termed as emulsion, VWR International Ltd., Poole, UK). Com- non-adapted cells. Aerobic batch growth in short-term −1 −1 pounds including 6 g L acetic acid, 1.5 g L furfural, adaptation experiments were followed for 5 days. Narayanan et al. AMB Expr (2016) 6:59 Page 4 of 13 Aerobic batch growth Characterisation of evolved population Short-term adapted TMB3500 cells were inoculated The evolved population (CC156) at pH 3.7 in the ALE into 25 mL of defined media at pH 3.7 with three differ - and the parental strain TMB3500 were compared for −1 ent inhibitor combinations: (1) 6 g L acetic acid and their robustness towards low pH and acetic acid with −1 −1 −1 0.5 g L HMF; (2) 6 g L acetic acid and 1.5 g L fur- other inhibitors. In addition, a biofilm was formed when −1 −1 fural; and (3) 6 g L acetic acid and 1 g L vanillin. the pH of the chemostat was reduced to 4.1. The biofilm Defined media (25 mL) with the IC at different pH and CC156 population were collected at the end of the values (pH 5.0, 4.5, 4.0 and 3.7) were inoculated with chemostat culture and stored as glycerol stocks. They short-term adapted cells of strain TMB3500 on defined were streaked on YPD plates to obtain single colonies medium. for DNA fingerprinting and used as a population from Short-term adapted TMB3500 cells were inoculated at glycerol stock in the pre-cultures for liquid media growth −1 different cell dry weights (gdw L ) namely 0.5, 1 and 3 experiments. −1 gdw L into 25 mL of defined medium with the IC at pH Agar plates were prepared by mixing autoclaved −1 3.7. 20 g L agar with filter sterilised chemically defined media of different conditions including pH 5.0 and 3.7 Anaerobic fermentation (with or without the IC) and pH 4.5 (50 % of the IC). Cell Short-term adapted and non-adapted TMB3500 cells suspensions of strain CC156 and strain TMB3500 (100 −1 were inoculated at 3 gdw L cells into 500 mL of defined μL) were streaked evenly on the agar plates and incu- medium with the IC at pH 3.7 in a 1 L Infors fermen- bated at 30 °C aerobically and anaerobically for 3–7 days. tor (InforsHT, Switzerland). Anaerobic conditions were Experiments were carried out in triplicate to enumerate −1 obtained by sparging with nitrogen gas (200 mL min ), the colony forming units (CFUs). the stirring was set at 200 rpm and the temperature was Aerobic and anaerobic liquid batch growth experi- maintained at 30 °C. The fermentation profile was fol - ments were carried out with pre-cultures of strain lowed by sampling for metabolites, residual inhibitors TMB3500 (from a single colony) and the evolved popu- and OD. lation CC156. Different conditions were tested in 25 mL defined media, including pH 5.0 and 3.7 (with and with - Adaptive lab evolution of TMB3500 out the IC) and pH 4.5 (50 % of the IC). Cells were grown A pre-culture of strain TMB3500 was used to inocu- at 30 °C in baffled shake flasks and sealed glass vials with late an aerobic batch of 1 L defined medium without magnetic stirrers and rubber stoppers to follow aerobic inhibitors at pH 5.0 in 1.4 L Infors fermentors (Infor- and anaerobic growth, respectively. sHT, Switzerland). The culture was operated at 30 °C, Cells of CC156 population, biofilm and strain with a stirring rate of 200 rpm and air was sparged at TMB3500 from the short-term adaptation step were −1 a flow rate of 200 mL min . At the end of the expo- inoculated into 25 mL of defined medium with inhibi - nential phase, the feed containing defined medium with tors at the concentrations similar to the residual inhibitor the IC was connected to the fermentor at a dilution rate concentrations present in the CC156 population of the −1 −1 −1 of 0.1 h and the fermentor was rendered anaerobic by ALE chemostat at pH 3.7 i.e. 0.32 g L HMF, 0.46 g L −1 −1 −1 sparging nitrogen gas at a flow rate of 200 mL min . furfural, 0.53 g L vanillin, 6.0 g L acetic acid, and this After the culture reached steady state, the pH was aerobic batch growth was followed for 2 days. reduced by 0.2 units using 3 M H SO . A 1.4 L Infors Cells of strain CC156, biofilm and strain TMB3500 2 4 fermentor stirring at 200 rpm and nitrogen gas sparged were propagated in 5 mL of YPD liquid medium contain- −1 −1 −1 −1 at 200 mL min was used between the fermentor with ing 10 g L yeast extract, 20 g L peptone and 20 g L yeast culture and feed bottle to facilitate a slow tran- glucose. Genomic DNA was extracted (Harju et al. 2004) sition to the new pH. The culture pH was maintained from three colonies in each strain and amplified by PCR using 3 M KOH. Once the pH was reduced to 4.5, the (C1000 Touch thermocycler, Bio-rad, USA) using −1 dilution rate was increased to 0.15 h . Gradual reduc- Dream taq polymerase (Life technologies, Sweden) and tion in pH was continued until it reached a value of selected primers targeting (1) TY1, TY3 elements (Trans- 3.7. The experiment was carried out for 3600 h thereby posable elements, individually and combined) (i Nogué obtaining 709 generations. Samples were taken for OD et al. 2012; Schofield et al. 1995). TY elements were cho- and metabolite analysis. Cell dry weight analyses were sen due to their presence in a wide variation in the yeast performed when the culture was in steady state. At the genome distribution making them ideal for intraspe- end of the evolution experiment, the cell suspension cies discrimination (Schofield et al. 1995), (2) randomly was transferred to a new chemostat running under the amplified polymorphic DNA (S1254 random primer) same conditions. (Akopyanz et al. 1992) and (3) (GACA) (Andrade et al. 4 Narayanan et al. AMB Expr (2016) 6:59 Page 5 of 13 2006) and (GTG) repeats (da Silva-Filho et al. 2005). Cells grown at pH 3.7 with acetic acid and furfural had −1 They were chosen owing to their usefulness in differen - a μ of 0.10 ± 0.01 h with a lag phase of 13 ± 0.5 h max tiating a strain during the evolution monitoring process whereas cells grown at pH 3.7 with acetic acid alone −1 in wine fermentation (da Silva-Filho et al. 2005). PCR was had a similar μ (0.11 ± 0.01 h ), but without any lag max performed adopting to conditions from the respective lit- phase (Sànchez i Nogué et al. 2013). This indicates that erature. The PCR products were separated in an agarose furfural had a combined inhibitory effect along with ace - gel (0.8 %) at 100 V for 60 min with a gene ruler DNA tic acid. Short-term adapted cells inoculated at pH 3.7 ladder (100 bp–10 kb) and gene ruler 100 bp plus ladder with acetic acid and vanillin had just marginal growth −1 (100 bp–3 kb) (Thermo Scientific, Sweden) as standards. of 0.01 ± 0.00 h after 160 h (Fig. 1). Similarly, minimal −1 growth of 0.01 ± 0.00 h was observed in the culture Metabolite analysis at pH 3.7 with all the inhibitors. Minimal to no growth Cell dry weight was determined in triplicate by filter - was observed with non-adapted cells in any of the above ing 5 mL of the culture on a pre-weighed 0.45 μm pore conditions, hence subsequent aerobic batch experiments size Supor membrane disc filter (Pall Corporation, Port were performed only with short-term adaptation. Washington, NY, USA). Filters were washed with distilled water and dried for 8 min at 350 W in a microwave oven. Combined effect of different pH and multiple inhibitors To analyze the metabolites, cells were separated by cen- on growth trifugation at 13,200 rpm for 2 min; the supernatant was To further map the inhibitory effect of low pH and acetic filtered through 0.20 μm membrane filters (Toyo Roshi acid in the presence of a cocktail of lignocellulosic inhibi- Kaish, Tokyo, Japan) and stored at −20 °C until analy- tors, a short-term adapted TMB3500 culture was grown sis. Concentrations of glucose, glycerol, acetate, ethanol, at different pH between 3.7 and 5.0 in media with the HMF, furfural and vanillin were determined by high per- IC. Whereas the cells at pH 5.0 started growing almost −1 formance liquid chromatography (Waters, Milford, MA, immediately with a μ of 0.20 ± 0.00 h , a decrease in max USA) using a HPX-87H resin-based column (Bio-Rad, pH by 0.5 units resulted in a long lag phase (20 ± 1 h) and −1 Hercules, CA, USA) preceded by a Micro-Guard Cation- a reduction in μ to 0.08 ± 0.00 h . At pH 4.0 there max H guard column (Bio-Rad). Separation was performed at was just a marginal growth and no growth was observed −1 45 °C with 5 mM H SO at a flow rate of 0.6 mL min . at pH 3.7 even after 150 h (Fig. 2). This illustrates the crit - 2 4 All compounds were quantified by refractive index ical role played by pH in the presence of acetic acid and detection (Shimadzu, Kyoto, Japan). For each HPLC inhibitors on yeast growth. run, a seven-point calibration curve was made for each compound. 0.12 Results In this study, the significance of pre-cultivation and ALE towards low pH and inhibitor tolerance were explored. 0.09 The TMB3500 yeast strain employed in this study was previously shown to possess high tolerance to lignocel- lulosic inhibitors by proliferating in the presence of high 0.06 amounts of non-detoxified hydrolysates, including bar - ley straw (40 % w/v), dilute spruce (60 % w/v) and wheat straw hydrolysates (60 % w/v) (Almeida et al. 2009a). 0.03 Robust phenotype through short‑term adaptation and role of individual inhibitors at low pH To test the influence of short-term adaptation on toler - ance to low pH and individual inhibitors, TMB3500 cells short-term adapted with the IC at pH 5.0 and non- adapted cells were inoculated in defined media at pH 3.7 −1 with acetic acid and individual inhibitors. Short-term Fig. 1 Eec ff t of low pH with 6 g L acetic acid and individual −1 −1 −1 adapted TMB3500 was able to grow at pH 3.7 with ace- inhibitors (1.5 g L furfural or 0.5 g L HMF or 1 g L vanillin) on short‑term adapted (black) and non‑adapted (white) cells. pH 3.7 with tic acid and HMF at a maximum specific growth rate −1 inhibitors includes the IC (μ ) of 0.09 ± 0.01 h without any lag phase (Fig. 1). max -1 Maximum specific growth rate (h ) Narayanan et al. AMB Expr (2016) 6:59 Page 6 of 13 ferment at pH 3.7 in the presence of IC, an anaero- 0.3 −1 bic batch culture was inoculated with 3 gdw L of a TMB3500 culture with (Fig. 4a) and without pre-adap- 0.2 tation (Fig. 4b). In the pre-adapted system, glucose was consumed completely within 19 ± 5 h at a specific con - −1 −1 sumption rate of 0.47 ± 0.01 g glucose g cells h to 0.1 −1 produce 0.45 ± 0.01 g ethanol g glucose (Table 1). There was a significant detoxification of inhibitors 0.0 including 65 ± 10 % of HMF, 85 ± 12 % of furfural and 3.54.0 4.55.0 5.5 60 ± 1 % of vanillin to their respective alcohols (Fig. 5). In pH the non-adapted system, the cell concentration started to Fig. 2 Eec ff t of the pH on growth of strain TMB3500 after short ‑term decrease and there was no significant sugar consumption adaptation at pH 5 with the IC or ethanol production even after 110 h and no inhibitor detoxification, except for 75 ± 12 % of furfural (Fig. 5; Table 1). Influence of initial cell density in growth performance at low pH with inhibitors ALE of TMB3500 for tolerance towards low pH with acetic To investigate whether the observed growth inhibition acid and other inhibitors could be relieved by increasing the initial cell density Though short-term adaptation is successful in obtain - and if there was a critical cell concentration that allowed ing a stable growth and fermentation profile at low pH for growth to occur, short-term adapted TMB3500 cul- under inhibitory conditions, a stable robust yeast strain tures were inoculated at different cell concentrations that does not require short-term adaptation towards low −1 (0.5, 1 and 3 gdw L ) in a medium at pH 3.7 with the IC. pH would be more ideal. Cells of strain TMB3500 were −1 When inoculated with 3 gdw L , cells grew at a μ of max evolved to improve their tolerance towards low pH with −1 0.04 ± 0.00 h (Fig. 3) in comparison to an inoculum size −1 inhibitory concentrations of acetic acid (6 g L ) and −1 of 0.5 or 1 gdw L , where no growth was observed even other inhibitors typically present in lignocellulose hydro- after 130 h of incubation. The short-term adapted bio - −1 −1 lysates including furfural (1.5 g L ), HMF (0.5 g L ) and mass at higher concentration might possess the required −1 vanillin (1 g L ). They were grown in an anaerobic con - volumetric reductase activity to efficiently detoxify HMF −1 tinuous culture at a dilution rate of 0.1 h in a defined and furfural to their corresponding alcohols (Modig et al. medium with the IC at pH 5.0. Once the culture attained 2008) reaching inhibitor threshold levels for allowing steady state, the pH was reduced stepwise from pH 5.0 to growth at the low pH in presence of acetic acid. 3.7. At pH 4.5, cells were further stressed to grow faster −1 by increasing the dilution rate to 0.15 h . A biofilm Eec ff t of initial cell density in fermentation performance which was formed at pH 4.1 in the chemostat covering at low pH with inhibitors the glass walls and baffles increased proportionally dur - To test the influence of the short-term adaptation strat - ing further reduction in pH. At pH 3.7, the cell culture egy to maintain metabolic activity and the ability to CC156 growing in the presence of lignocellulosic inhibi- tors was obtained after 709 generations (Fig. 6a, b). The cell suspension of the chemostat was transferred into a new chemostat operating under similar conditions to test its capacity for inhibitor tolerance at pH 3.7 and to iden- tify the role of the biofilm in ALE. The chemostat culture washed out (data not shown) and the CC156 cells stored at −80 °C were chosen as the evolved population. The increase in un-dissociated acetic acid concentra - 1 tion in the chemostat due to stepwise lowering the pH led to a series of effects over the course of cultivation: (1) From pH 5.0 to 4.5 an initial decrease in opti- 050100 150 cal density was followed by recovery of the culture, but Time (h) at lower pH, the optical density gradually decreased Fig. 3 Eec ff t of initial cell density on growth and inhibitor tolerance (Fig. 6a); (2) this was accompanied by a temporary of strain TMB3500 at pH 3.7 with the IC after short‑term adaptation. −1 −1 increase in the glucose concentration due to a lower Cell dry weight‑3 gdw L (triangles), 1 gdw L (diamonds) and 0.5 −1 gdw L (squares) glucose rate of consumption. After each pH change, the ln OD (620nm) Maximum specific growth rate -1 (h ) Narayanan et al. AMB Expr (2016) 6:59 Page 7 of 13 Time (h) HMFFurfuralVanillin -20 Fig. 5 Inhibitor conversion profile in anaerobic batch fermentation with (black) and without (white) short‑term adaptation Characterisation of long‑term evolution Characterisation of the CC156 population from the che- mostat was performed in an attempt to isolate a stable strain displaying robustness towards low pH with inhibi- 0306090120 tors in comparison to the parental strain, TMB3500. To Time (h) isolate a strain displaying an inhibitor tolerant phenotype Fig. 4 Anaerobic batch fermentation of strain TMB3500 with 3 gdw at low pH, cells of CC156 and TMB3500 strains were −1 L cells in pH 3.7 with the IC with (a) and without (b) short‑term plated in defined solid media with different concentra - adaptation. Cell dry weight (squares), glucose (diamonds), ethanol tions of inhibitors and pH without short-term adaptation. (circles), acetate (triangles) and glycerol (crosses) Surprisingly, both the strains displayed similar colony growth characteristics: (1) colonies formed in less inhibi- tory conditions including defined media at pH 5.0, 3.7 and 4.5 with 50 % of the IC. (2) No colonies were formed glucose concentration increased sequentially in the new in severe inhibitory conditions including pH 5.0 or 3.7 steady states (Fig. 6b); (3) HMF, furfural and vanillin most with the IC, indicating that the evolved population could probably have been detoxified by NAD(P)H-dependent not retain the phenotypic robustness against inhibitors. reductases in strain TMB3500 as observed by Modig To validate the strain robustness, cells from the CC156 et al. (2008) and furfural was detoxified with a conversion population and TMB3500 were grown in chemically efficiency of 75 % on average throughout the chemostat defined liquid media with different pH and concentra - cultivation and HMF and vanillin were being detoxified tions of inhibitors, aerobically and anaerobically. How- in varying efficiencies based on the cell concentrations ever, growth characteristics for both the parental and at the given time point (Fig. 7). On the other hand, the evolved strain were similar in liquid media as well. Cells ethanol yield was maintained at 0.42 ± 0.04 g ethanol g −1 were considered to have positive growth characteristics if glucose throughout the evolution process in spite of their OD value had doubled at least threefold. Based on the increase in dilution rate and decrease in pH. Moreo- this, there was a difference between the growth patterns ver, the concentration of acetate remained constant at −1 in solid and liquid media where growth was displayed in 6 g L , as it could not be consumed as a source of carbon low to medium inhibitory conditions, including chemi- by S. cerevisiae under anaerobic conditions (Henningsen cally defined media at pH 5.0, 3.7 and 4.5 with 50 % of et al. 2015) (Fig. 6b). Table 1 Anaerobic batch fermentation at pH 3.7 with the IC with and without short-term adaptation at pH 5 −1 Condition Ethanol titre Yield (g g glucose ) Specific consumption/pro ‑ −1 −1 −1 (g L ) duction rate (g g h ) Biomass Glycerol Acetate Ethanol Glucose Ethanol Short‑term adapted 8.93 ± 0.38 0.01 ± 0.00 0.01 ± 0.01 0.00 ± 0.01 0.45 ± 0.01 −0.47 ± 0.01 0.18 ± 0.02 a a Non‑adapted 0.73 ± 0.84 −0.58 ± 0.56 0.03 ± 0.04 0.30 ± 0.15 0.15 ± 0.13 0.01 ± 0.00 0.00 ± 0.00 Yield values based on glucose consumption of 1.80 ± 2.11 g in total -1 -1 Concentration (g L ) Concentration (g L ) Conversion % Narayanan et al. AMB Expr (2016) 6:59 Page 8 of 13 -1 -1 0.10 h 0.15 h 10 2.5 pH pH pH pH pH pH pH pH 4.9 4.7 4.5 4.5 4.3 4.1 3.9 3.7 8 2.0 1.5 4 1.0 2 0.5 0 0.0 0600 1200 1800 2400 3000 3600 Time (h) 15 pH ~ 2 0600 1200 1800 2400 3000 3600 Time (h) Fig. 6 a ALE of strain TMB3500. Optical density (solid squares) and cell dry weight (spotted triangles). Dotted vertical lines indicate a change in pH. b Metabolite profile of ALE of strain TMB3500. Glucose (diamonds), ethanol (circles), acetate (triangles) and glycerol (crosses) 122145 289300 1872 1922 2497 2521 2856 2890 3264 3291 3582 4.94.7 4.54.3 4.13.9 3.7 Time (h) & pH Fig. 7 Inhibitor conversion profile throughout the ALE at the start and end of every pH shift. (% conversion, calculated based on initial and meas‑ ured time point). HMF (white), furfural (grey) and vanillin (black) the IC and pH 5.0 with IC. Nevertheless, both parental Since the evolved strain did not display inhibitor and adapted strain did not grow in the medium with pH robustness in conditions of the ALE chemostat, it was 3.7 and IC like in the severest conditions applied in the hypothesised that the thick biofilm generated over the chemostat. course of the long-term adaptation could have a role in -1 % Conversion Concentration (g L ) OD (620nm) -1 CDW (g L ) Narayanan et al. AMB Expr (2016) 6:59 Page 9 of 13 detoxifying the medium and cell proliferation, whereas ATPases and multiple drug resistance transporters pump the free cells present in the suspension also might be the protons and acetate anions, respectively, out of the growing and fermenting, tolerating the residual inhibi- cell by utilising ATP, leading to an energy drain (Caspeta tors. To verify this hypothesis and to test whether short- et al. 2015; Piotrowski et al. 2014; Taherzadeh and Karimi term adaptation helps to retain tolerance in the evolved 2011), (2) deactivation of key cellular and glycolytic strain, cells from the biofilm, CC156 and TMB3500 were enzymes, membrane and DNA damage caused by fur- inoculated in liquid media with inhibitor concentra- fural and HMF (Piotrowski et al. 2014); (3) vanillin may tions similar to the residual inhibitor concentrations in cause damage to cell membrane integrity (Piotrowski the liquid suspension of the ALE chemostat. Unexpect- et al. 2014; Trinh Thi My et al. 2014); and (4) reduction of edly, strain TMB3500 could grow at a maximum spe- the furans demands excessive reducing power of NAD(P) −1 cific growth rate of 0.07 ± 0.00 h with a lag phase of H that is directed away from ethanol production and 12 ± 0.5 h, biofilm cells and the CC156 population did anabolism (Almeida et al. 2007, 2009b). All four mecha- not grow even after 150 h. nisms might have contributed to reduced growth capac- To investigate if large rearrangements in the genome ity, elongated lag phases, lower growth rates and ethanol had occurred, DNA fingerprinting was carried out with yields (Almeida et al. 2011; Piotrowski et al. 2014). As different sets of primers in the cells from the biofilm, the mechanism of detoxification, tolerance, energy and population CC156 and strain TMB3500. Interestingly, co-factor requirements are different among weak acids, the gel bands were similar for all analysed strains (Addi- furans and phenolics, the presence of more than one tional file 1: Figure S1) indicating that there were no class of inhibitor in the substrate results in an additional major genetic re-arrangements in the evolved strain in burden to the yeast cell. comparison to the parental strain. Short-term adaptation to the inhibitors at pH 5 might ALE of S. cerevisiae TMB3500 strain resulted in a have led to a reduction in the cytosolic pH that subse- population accompanied by a biofilm displaying robust - quently has increased the tolerance to acetic acid at pH ness at pH 3.7 with the IC coupled with efficient etha - 4.5, as was earlier observed in another S. cerevisiae strain nol production. However, when the selection pressure (Fernández-Niño et al. 2015). Hence during incubation was removed through storage, the parental strain and at pH 3.7 with the IC, the effect of low pH and diffusion the evolved population had similar growth characteris- of undissociated acetic acid inside the cell must have tics under the various inhibitory conditions. This led to been less pronounced. Increasing the cell density com- the conclusion that the robustness towards low pH and bined with short-term adaptation improved the poten- lignocellulosic inhibitors displayed by the evolved strain tial of growth and ethanol fermentation capacity at pH from ALE might be a result of population heterogeneity. 3.7 with the IC in 12 h compared to the non-adapted culture (Figs. 4, 5; Table 1). Higher inoculum concentra- Discussion tions could have several positive effects towards inhibitor Developing inhibitor tolerant S. cerevisiae strains ferment- detoxification including: (1) ‘Safety in numbers’, i.e. low - ing at low pH is very attractive to the cellulosic bioethanol ering the ratio of inhibitor concentration over cell con- industry owing to challenges including bacterial contami- centration, which leads to lower detoxification demand nation, reduction in cell viability, longer lag phase due to per cell; and (2) population heterogeneity, where more inhibitor detoxification, formation of undesirable prod - representatives of several sub populations could be dedi- ucts, lower ethanol yield, productivity and titre (Almeida catedly involved in detoxification of inhibitors, thereby et al. 2009b; Palmqvist and Hahn-Hägerdal 2000; Skinner facilitating growth and ethanol production by other sub and Leathers 2004). In this study, two different adaptation populations. The heterogeneity could be due to (a) vari - strategies were successfully applied to attain tolerance in ations in cell growth phase, cycle and cell ageing, and (b) S. cerevisiae towards more severe conditions than investi- stochasticity in gene expression impacting enzymatic gated before, i.e. low pH at 3.7 and inhibitory concentra- activities leading to variations in metabolic reactions tions of acetic acid, furfural, HMF and vanillin. (Avery 2006; Delvigne et al. 2014). Of all the inhibi- When a short-term adaptation strategy was applied tors, vanillin had a major impact on the growth perfor- with the IC, cells were able to grow at pH 5.0 and 4.5, mance of strain TMB3500 at pH 3.7 in the presence of but not at pH 4.0 and 3.7 (Fig. 2). This could be due to acetic acid as previously observed by Klinke et al. (2004) synergistic effects of low pH and the different inhibi - affecting the growth of S. cerevisiae at a concentra - −1 tors, i.e. (1) increased passive diffusion of undissociated tion of 0.5 g L , followed by furfural with an increased acetic acid into the cells leading to acidification of the lag phase, whereas HMF had no effect (Fig. 1) when cytoplasm. Once inside the cell, acetate, being a weak compared with cells grown at pH 3.7 with acetic acid acid, will reduce the intracellular pH. Plasma membrane (Sànchez i Nogué et al. 2013). Since vanillin is involved Narayanan et al. AMB Expr (2016) 6:59 Page 10 of 13 in membrane damage, it might aid the cellular entry of characterisation could have been the low inoculum size −1 acetic acid, thereby affecting the fitness and metabolism of OD 0.5 (OD of 2.5 corresponds to 1 gdw L of strain rapidly. Synergistic effects of acetic acid and furfural have TMB3500) used to start the aerobic and anaerobic liquid been reported to negatively affect specific growth rate batches, which may very well have been below the criti- and ethanol yield (Palmqvist et al. 1999). In addition, cal cell mass to initially detoxify the inhibitors at low pH effects of the individual and combination of inhibitors when compared with the size of inoculum in the chemo- including furfural, phenol and acetic acid have been ana- stat (Fig. 6a). lysed using metabolic profiling. The synergistic negative All in all, this study indicated that the phenotype of effect on amino acids and central carbon metabolism was both the short-term and long-term adaptation, i.e. growth more pronounced than the sum of individual inhibitors at pH 3.7 with inhibitors, turned out to be similar with no with acetic acid playing a key role in the combined inhi- rigorous genetic changes. Copy number variations of spe- bition (Ding et al. 2011). cific genes and single nucleotide polymorphisms obtained Evolutionary engineering is a useful tool to obtain a from 5 months of evolution might be very specific to be desired phenotype by acquiring a stable genotype in an visualised in the broad genomic DNA fingerprinting tech - organism by applying constant or increasing selection niques used for the current analysis. Indeed, short-term pressure (Almario et al. 2013; Sauer 2001). The natural adaptation at acidic pH might have a positive influence in evolution process towards desirable properties like adapt- epigenetic expression of various stress response genes and ability to low pH and inhibitor tolerance is prominent transcription factors/activators including YAP1, HAA1 among industrial yeast strains where cells are repeatedly (Anneli et al. 2006; Modig et al. 2008; Wright et al. 2011). washed with dilute H SO and recycled in the fermenta- Moreover, De Melo et al. (2010) used an industrial strain 2 4 tion process along with pre-cultures as observed with the named JP1 to show that an acidic environment affected PE-2 strain used in Brazilian ethanol production plants cell growth and induced general stress response. The cell (Della-Bianca et al. 2014). Exposure of strain TMB3500 tolerance to acidic environment may involve down regu- to step-wise reduction in pH from 5 to 3.7 over 3600 h lation of transcription and protein synthesis due to PKA in a chemostat led to successful growth and ethanol pro- based glucose signalling (De Melo et al. 2010). Interest- duction in the presence of inhibitory concentrations of ingly, a cross-tolerance phenomenon has been observed acetic acid, furfural, HMF and vanillin over the whole pH in S. cerevisiae, meaning that tolerance acquired to one range. Interestingly, at pH 4.1, biofilm was formed in the stress enhances resistance to other forms of stress (Gib- fermentor, possibly to protect the cells against harsher son et al. 2007). For instance, low pH stress due to inor- conditions as was observed for Zymomonas mobilis on ganic and weak organic acids induces the expression of rice bran hydrolysate forming a protective layer around genes involved in tolerance to heat shock, cell wall assem- cells (Todhanakasem et al. 2014). Yeast cells form biofilm bly, trehalose biosynthesis, tolerance to osmotic stress and through cell–cell adhesion in response to stress through glycerol production (Kapteyn et al. 2001; Kawahata et al. triggering the Ras/cAMP/protein kinase A (PKA) and 2006). Also, the resulting enhancement in population per- mitogen activated protein kinase pathways and express- formance that we observed in ALE could be due to poly- ing the FLO genes (Verstrepen and Klis 2006). Hence, genic response as observed by Meijnen et al. (2016) in the the biofilm might have contributed to the majority of case of tolerance towards acetic acid, preserving beneficial the detoxification and cell proliferation. In addition, cells mutations and by avoiding any undesirable pleiotropic released from the biofilm into the liquid suspension may response as in case of any targeted genetic manipulation have contributed to ethanol production, but without pro- (Sauer 2001). liferation. Therefore, when the cell suspension was trans - The adaptation patterns observed in our study, includ - ferred into another chemostat under the same condition, ing the long-term adaptation, could thus be due to a the culture simply washed out immediately, adding cred- combination of a variety of genetic changes and sto- ibility to this interpretation. chastic switching that are either triggered by the harsh Surprisingly, characterisation of the adapted strain environment (Acar et al. 2008) or are already present in acquired from the suspension after 709 generations a subpopulation (Delvigne and Goffin 2014; Levy et al. showed that it possessed a similar phenotype as the 2012). In the latter case, phenotypic variability or plas- parental strain TMB3500 in response to the expo- ticity exhibited in the subpopulation might enhance the sure of low pH and the IC. The apparent robust pheno - survival of the species when confronted with diverse type had disappeared as soon as the adapted strain was hostile environments (Delvigne and Goffin 2014; Levy exposed again to the harsh conditions as seen by Wright et al. 2012). Most likely, in TMB3500 cultures only the et al. (2011) when adapting for acetic acid tolerance. subpopulation readily adapted to the inhibitors at low However, one drawback of the experimental setup of pH were selected for, which is underlined by the need Narayanan et al. AMB Expr (2016) 6:59 Page 11 of 13 Funding to apply thick inocula to provide a critical mass of this This research was funded by Swedish National Energy Agency (www.ener‑ subpopulation. gimyndigheten.se, Project No. P35350‑1). The funders had no role in study We have performed a short-term adaptation and an design, data collection and analysis, decision to publish, or preparation of the manuscript. ALE of an industrial S. cerevisiae strain to grow and produce ethanol at low pH in the presence of lignocel- Received: 6 July 2016 Accepted: 18 August 2016 lulosic inhibitors. The next step would be to perform fer - mentations with yeast and simulated contaminations of bacteria in a medium stressed with inorganic and weak organic acids in the presence of lignocellulosic inhibitors References to analyze the effect of low pH fermentations. This might Acar M, Mettetal JT, van Oudenaarden A. Stochastic switching as a survival be a key step towards reduction of bacterial contamina- strategy in fluctuating environments. Nat Genet. 2008;40(4):471–5. tion in large-scale lignocellulosic ethanol production. doi:10.1038/ng.110. Akopyanz N, Bukanov NO, Westblom TU, Kresovich S, Berg DE. DNA diversity Though the evolved strain obtained from long-term evo - among clinical isolates of Helicobacter pylori detected by PCR‑based lution did not maintain stable inhibitor robustness, the RAPD fingerprinting. Nucl Acids Res. 1992;20(19):5137–42. doi:10.1093/ short-term adaptation strategy to pH 5 and all inhibitors nar/20.19.5137. Albers E, Johansson E, Franzén CJ, Larsson C. Selective suppression of made it possible for the first time to successfully ferment bacterial contaminants by process conditions during lignocel‑ glucose to ethanol at pH 3.7 in the presence of lignocel- lulose based yeast fermentations. Biotechnol Biofuels. 2011;4:59. lulosic inhibitors. Inhibitor tolerance and cell growth in doi:10.1186/1754‑6834‑4‑59. Almario MPP, Reyes LH, Kao KC. Evolutionary engineering of Saccharomyces yeast is thus achieved through phenotypic plasticity, i.e. cerevisiae for enhanced tolerance to hydrolysates of lignocellulosic bio‑ a delicate phenotypic balance underlined by expression mass. Biotechnol Bioeng. 2013;110(10):2616–23. doi:10.1002/bit.24938. of genes relating to stress tolerance, growth and fermen- Almeida JR, Modig T, Petersson A, Hähn‑Hägerdal B, Lidén G, Gorwa‑ Grauslund MF. Increased tolerance and conversion of inhibitors in lignocellulosic tation. Understanding the complex genetic regulation hydrolysates by Saccharomyces cerevisiae. J Chem Technol Biotechnol. behind phenotypic plasticity and stochastic expression in 2007;82(4):340–9. doi:10.1002/jctb.1676. adaptive evolution might enable us to steer the yeast cells Almeida JRM, Karhumaa K, Bengtsson O, Gorwa‑ Grauslund M‑F. Screening of Saccharomyces cerevisiae strains with respect to anaerobic growth to desirable metabolic responses. in non‑ detoxified lignocellulose hydrolysate. Bioresour Technol. 2009a;100(14):3674–7. doi:10.1016/j.biortech.2009.02.057. Additional file Almeida JRR, Bertilsson M, Gorwa‑ Grauslund MF, Gorsich S, Lidén G. Metabolic effects of furaldehydes and impacts on biotechnological processes. Appl Microbiol Biotechnol. 2009b;82(4):625–38. doi:10.1007/s00253009 ‑ 1875 ‑ 1 ‑ . Additional file 1: Figure S1. DNA fingerprinting with genomic DNA Almeida JRM, Runquist D, Sànchez Nogué V, Lidén G, Gorwa‑ Grauslund MF. extracted from parental strain TMB3500 (first column of gel), evolved Stress‑related challenges in pentose fermentation to ethanol by the yeast strain CC156 (middle column of gel) and biofilm (third column of gel). Saccharomyces cerevisiae. Biotechnol J. 2011;6(3):286–99. doi:10.1002/ Primers used include a) (GACA) b) (GTG) c) S1254 random primer d) TY1 4 5 biot.201000301. primer e) TY3 primer f ) TY1 + TY3 primer. One gel is displayed from each Alriksson B, Horváth IS, Jönsson LJ. Overexpression of Saccharomyces cerevisiae strain as a representative from triplicate analysis. transcription factor and multidrug resistance genes conveys enhanced resistance to lignocellulose‑ derived fermentation inhibitors. Process Biochem. 2010;45(2):264–71. doi:10.1016/j.procbio.2009.09.016. Abbreviations Andrade M, Rodriguez M, Sánchez B, Aranda E, Córdoba J. DNA typing meth‑ ALE: adaptive lab evolution; SSF: simultaneous saccharification and fer ‑ ods for differentiation of yeasts related to dry‑ cured meat products. Int J mentation; HMF: hydroxymethylfurfural; YPD: yeast peptone dextrose; KOH: Food Microbiol. 2006;107(1):48–58. doi:10.1016/j.ijfoodmicro.2005.08.011. potassium hydroxide; H SO : sulphuric acid; IC: inhibitor cocktail; OD: optical 2 4 Anneli P, João RMA, Tobias M, Kaisa K, Bärbel HH, Marie FGG, Gunnar L. A density; TMB3500: Brewer’s yeast strain; CC156: evolved population; TY: trans‑ 5‑hydroxymethyl furfural reducing enzyme encoded by the Saccharomy - posable elements; μ : maximum specific growth rate; PKA: protein kinase A. max ces cerevisiae ADH6 gene conveys HMF tolerance. Yeast. 2006;23(6):455– 64. doi:10.1002/yea.1370. Authors’ contributions Avery SV. Microbial cell individuality and the underlying sources of heteroge‑ All the authors participated in the design of the study, VN and VSN performed neity. Nat Rev Micro. 2006;4(8):577–87. doi:10.1038/nrmicro1460. the experimental work, VN and EvN wrote the manuscript and all the authors Basso LC, Basso TO, Rocha SN. Ethanol production in Brazil: the industrial pro‑ critically read and approved the manuscript. All authors read and approved cess and its impact on yeast fermentation. In: dos Santos Bernardes MA, the final manuscript. editor. Biofuel production‑recent developments and prospects. Rijeka: InTech; 2011. p. 85–100. doi:10.5772/17047. Author details Beales N. Adaptation of microorganisms to cold temperatures, weak acid Division of Applied Microbiology, Department of Chemistry, Lund University, preservatives, low ph, and osmotic stress: a review. Compr Rev Food Sci F. P.O. Box 124, 221 00 Lund, Sweden. Present Address: National Bioenergy 2004;3(1):1–20. doi:10.1111/j.1541‑4337.2004.tb00057.x. Center, National Renewable Energy Laboratory, 15013 Denver West Parkway, Beckner M, Ivey M, Phister T. Microbial contamination of fuel etha‑ Golden, CO 80401, USA. nol fermentations. Lett Appl Microbiol. 2011;53(4):387–94. doi:10.1111/j.1472‑765X.2011.03124.x. Acknowledgements Bischoff KM, Liu S, Leathers TD, Worthington RE, Rich JO. Modeling bacte ‑ We would like to thank Dr. Magnus Carlquist for the discussions and critically rial contamination of fuel ethanol fermentation. Biotechnol Bioeng. reading the manuscript. 2009;103(1):117–22. doi:10.1002/bit.22244. Caspeta L, Castillo T, Nielsen J. Modifying yeast tolerance to inhibitory Competing interests conditions of ethanol production processes. Front Bioeng Biotechnol. The authors declare that they have no competing interests. 2015;3:184. doi:10.3389/fbioe.2015.00184. Narayanan et al. AMB Expr (2016) 6:59 Page 12 of 13 da Silva‑Filho EAA, Brito dos Santos SK, Resende AdM, de Morais JOO, Kapteyn J, Ter Riet B, Vink E, Blad S, De Nobel H, Van Den Ende H, Klis F. Low de Morais MA, Ardaillon Simões D. Yeast population dynamics of external pH induces HOG1‑ dependent changes in the organization of industrial fuel‑ ethanol fermentation process assessed by PCR‑finger ‑ the Saccharomyces cerevisiae cell wall. Mol Microbiol. 2001;39(2):469–80. printing. Antonie Van Leeuwenhoek. 2005;88(1):13–23. doi:10.1007/ doi:10.1046/j.1365‑2958.2001.02242.x. s10482‑004‑7283‑8. Kawahata M, Masaki K, Fujii T, Iefuji H. Yeast genes involved in response to lac‑ De Melo HF, Bonini BM, Thevelein J, Simões DA, Morais MA. Physiologi‑ tic acid and acetic acid: acidic conditions caused by the organic acids in cal and molecular analysis of the stress response of Saccharomyces Saccharomyces cerevisiae cultures induce expression of intracellular metal cerevisiae imposed by strong inorganic acid with implication to metabolism genes regulated by Aft1p. FEMS Yeast Res. 2006;6(6):924–36. industrial fermentations. J Appl Microbiol. 2010;109(1):116–27. doi:10.1111/j.1567‑1364.2006.00089.x. doi:10.1111/j.1365‑2672.2009.04633.x. Klinke HB, Thomsen AB, Ahring BK. Inhibition of ethanol‑producing yeast Della‑Bianca BE, de Hulster E, Pronk JT, van Maris AJA, Gombert AK. Physiology and bacteria by degradation products produced during pre‑treatment of the fuel ethanol strain Saccharomyces cerevisiae PE‑2 at low pH indi‑ of biomass. Appl Microbiol Biotechnol. 2004;66(1):10–26. doi:10.1007/ cates a context‑ dependent performance relevant for industrial applica‑ s00253‑004‑1642‑2. tions. FEMS Yeast Res. 2014;14(8):1196–205. doi:10.1111/1567‑1364.12217. Koichi T, Yukari I, Jun O, Jun S. Enhancement of acetic acid tolerance in Sac- Delvigne F, Goffin P. Microbial heterogeneity affects bioprocess robustness: charomyces cerevisiae by overexpression of the HAA1 gene, encoding dynamic single‑ cell analysis contributes to understanding of microbial a transcriptional activator. Appl Environ Microbiol. 2012;78(22):8161–3. populations. Biotechnol J. 2014;9(1):61–72. doi:10.1002/biot.201300119. doi:10.1128/AEM.02356‑12. Delvigne F, Zune Q, Lara AR, Al‑Soud W, Sørensen SJ. Metabolic variability Lei J, Yu S, Lili X, Bingyin P, Yazhong X, Xiaoming B. Enhanced resistance of in bioprocessing: implications of microbial phenotypic heterogeneity. Saccharomyces cerevisiae to vanillin by expression of lacA from Trametes Trends Biotechnol. 2014;32(12):608–16. doi:10.1016/j.tibtech.2014.10.002. sp. AH28‑2. Bioresour Technol. 2011;102(17):8105–9. doi:10.1016/j. Ding MZ, Wang X, Yang Y, Yuan YJ. Metabolomic study of interactive effects biortech.2011.06.057. of phenol, furfural, and acetic acid on Saccharomyces cerevisiae. OMICS. Levy SF, Ziv N, Siegal ML. Bet hedging in yeast by heterogeneous, age‑ corre‑ 2011;15(10):647–53. doi:10.1089/omi.2011.0003. lated expression of a stress protectant. PLoS Biol. 2012;10(5):e1001325. Dominik H, Uwe S. Identification of furfural as a key toxin in lignocellulosic doi:10.1371/journal.pbio.1001325. hydrolysates and evolution of a tolerant yeast strain. Microb Biotechnol. Limayem A, Ricke SC. Lignocellulosic biomass for bioethanol production: 2008;1(6):497–506. doi:10.1111/j.1751‑7915.2008.00050.x. current perspectives, potential issues and future prospects. Prog Energy Fernández‑Niño M, Marquina M, Swinnen S, Rodríguez‑Porrata B, Nevoigt Combust Sci. 2012;38(4):449–67. doi:10.1016/j.pecs.2012.03.002. E, Ariño J. The cytosolic pH of individual Saccharomyces cerevisiae Liu ZL. Molecular mechanisms of yeast tolerance and in situ detoxification of cells is a key factor in acetic acid tolerance. Appl Environ Microbiol. lignocellulose hydrolysates. Appl Microbiol Biotechnol. 2011;90(3):809– 2015;81(22):7813–21. doi:10.1128/AEM.02313‑15. 25. doi:10.1007/s00253‑011‑3167‑9. Gibson BR, Lawrence SJ, Leclaire JPR, Powell CD, Smart KA. Yeast responses to Meijnen J‑P, Randazzo P, Foulquié ‑Moreno MR, van den Brink J, Vandecruys stresses associated with industrial brewery handling. FEMS Microbiol Rev. P, Stojiljkovic M, Dumortier F, Zalar P, Boekhout T, Gunde‑ Cimerman N, 2007;31(5):535–69. doi:10.1111/j.1574‑6976.2007.00076.x. Kokošar J, Štajdohar M, Curk T, Petrovič U, Thevelein JM. Polygenic analysis Graves T, Narendranath NV, Dawson K, Power R. Eec ff t of pH and lactic or and targeted improvement of the complex trait of high acetic acid toler‑ acetic acid on ethanol productivity by Saccharomyces cerevisiae in ance in the yeast Saccharomyces cerevisiae. Biotechnol Biofuels. 2016;9:5. corn mash. J Ind Microbiol Biotechnol. 2006;33(6):469–74. doi:10.1007/ doi:10.1186/s13068‑015‑0421‑x. s10295‑006‑0091‑6. Modig T, Almeida JR, Gorwa‑ Grauslund MF, Lidén G. Variability of the response Hahn‑Hägerdal B, Karhumaa K, Larsson CU, Gorwa‑ Grauslund M, Gör‑ of Saccharomyces cerevisiae strains to lignocellulose hydrolysate. Biotech‑ gens J, van Zyl WH. Role of cultivation media in the development of nol Bioeng. 2008;100(3):423–9. doi:10.1002/bit.21789. yeast strains for large scale industrial use. Microb Cell Fact. 2005;4:31. Muthaiyan A, Limayem A, Ricke SC. Antimicrobial strategies for limiting bacte‑ doi:10.1186/1475‑2859‑4‑31. rial contaminants in fuel bioethanol fermentations. Prog Energy Combust Hanqi G, Jian Z, Jie B. Inhibitor analysis and adaptive evolution of Saccharomy- Sci. 2011;37(3):351–70. doi:10.1016/j.pecs.2010.06.005. ces cerevisiae for simultaneous saccharification and ethanol fermentation Nielsen F, Tomás‑Pejó E, Olsson L, Wallberg O. Short ‑term adaptation from industrial waste corncob residues. Bioresour Technol. 2014;157:6–13. during propagation improves the performance of xylose‑fermenting doi:10.1016/j.biortech.2014.01.060. Saccharomyces cerevisiae in simultaneous saccharification and Harju S, Fedosyuk H, Peterson KR. Rapid isolation of yeast genomic DNA: bust co‑fermentation. Biotechnol Biofuels. 2015;8:219. doi:10.1186/ n’Grab. BMC Biotechnol. 2004;4(1):8. doi:10.1186/1472‑6750‑4‑8. s13068‑015‑0399‑4. Henningsen BM, Hon S, Covalla SF, Sonu C, Argyros AD, Barrett TF, Wiswall Palmqvist E, Grage H, Meinander NQ, Hahn‑Hagerdal B. Main and interaction E, Froehlich AC, Zelle RM. Increasing anaerobic acetate consumption effects of acetic acid, furfural, and p ‑hydroxybenzoic acid on growth and ethanol yields in Saccharomyces cerevisiae with NADPH‑specific and ethanol productivity of yeasts. Biotechnol Bioeng. 1999;63(1):46–55. alcohol dehydrogenase. Appl Environ Microbiol. 2015;81(23):8108–17. doi:10.1002/(SICI)1097‑0290(19990405)63:1<46:AID ‑BIT5>3.0.CO;2‑ J. doi:10.1128/AEM.01689‑15. Palmqvist E, Hahn‑Hägerdal B. Fermentation of lignocellulosic hydro ‑ i Nogué V, Bettiga M, Gorwa‑ Grauslund MF. Isolation and characteriza‑ lysates. II: inhibitors and mechanisms of inhibition. Bioresour Technol. tion of a resident tolerant Saccharomyces cerevisiae strain from a 2000;74(1):25–33. doi:10.1016/S0960‑8524(99)00161‑3. spent sulfite liquor fermentation plant. AMB Express. 2012;2(1):68. Papoutsakis ET, Pronk JT. Editorial overview: energy biotechnology. Curr Opin doi:10.1186/2191‑0855‑2‑68. Biotechnol. 2015;33:8–11. doi:10.1016/j.copbio.2015.04.001. Ishola MM, Babapour AB, Gavitar MN, Brandberg T, Taherzadeh MJ. Eec ff t Parawira W, Tekere M. Biotechnological strategies to overcome inhibitors of high solids loading on bacterial contamination in lignocellulosic in lignocellulose hydrolysates for ethanol production: review. Crit Rev ethanol production. Bioresources. 2013;8(3):4429–39. doi:10.15376/ Biotechnol. 2011;31(1):20–31. doi:10.3109/07388551003757816. biores.8.3.4429‑4439. Piotrowski JS, Zhang Y, Sato T, Ong I, Keating D, Bates D, Landick R. Death Jönsson LJ, Alriksson B, Nilvebrant N‑ O. Bioconversion of lignocellulose: by a thousand cuts: the challenges and diverse landscape of lignocel‑ inhibitors and detoxification. Biotechnol Biofuels. 2013;6(1):16. lulosic hydrolysate inhibitors. Front Microbiol. 2014;5:90. doi:10.3389/ doi:10.1186/1754‑6834‑6‑16. fmicb.2014.00090. Kádár Z, Maltha SF, Szengyel Z, Réczey K, De Laat W. Ethanol fermenta‑ Sànchez i Nogué V, Narayanan V, Gorwa‑ Grauslund M, . Short‑term adaptation tion of various pretreated and hydrolyzed substrates at low initial improves the fermentation performance of Saccharomyces cerevisiae pH. Appl Biochem Biotechnol. 2007;137(1–12):847–58. doi:10.1007/ in the presence of acetic acid at low pH. Appl Microbiol Biotechnol. s12010‑007‑9102‑ y. 2013;97(16):7517–25. doi:10.1007/s00253‑013‑5093‑5. Narayanan et al. AMB Expr (2016) 6:59 Page 13 of 13 Sauer U. Evolutionary engineering of industrially important microbial phe‑ Todhanakasem T, Sangsutthiseree A, Areerat K, Young GM, Thanonkeo P. notypes. Advances in biochemical engineering/biotechnology. Berlin: Biofilm production by Zymomonas mobilis enhances ethanol production Springer; 2001. p. 129–69. and tolerance to toxic inhibitors from rice bran hydrolysate. New Biotech‑ Schofield MA, Rowe SM, Hammond JR, Molzahn SW, Quain DE. Differentia‑ nol. 2014;31(5):451–9. doi:10.1016/j.nbt.2014.06.002. tion of brewery yeast strains by DNA fingerprinting. J Inst Brewing. Tomás‑Pejó E, Olsson L. Influence of the propagation strategy for obtain‑ 1995;101(2):75–8. doi:10.1002/j.2050‑0416.1995.tb00850.x. ing robust Saccharomyces cerevisiae cells that efficiently co ‑ferment Serate J, Xie D, Pohlmann E, Donald C, Shabani M, Hinchman L, Higbee A, xylose and glucose in lignocellulosic hydrolysates. Microb Biotechnol. Mcgee M, La Reau A, Klinger GE, Li S, Myers CL, Boone C, Bates DM, Cava‑ 2015;8(6):999–1005. doi:10.1111/1751‑7915.12280. lier D, Eilert D, Oates LG, Sanford G, Sato TK, Dale B, Landick R, Piotrowski J, Trinh Thi My N, Sakihito K, Shingo I. Importance of glucose‑6‑phosphate dehy‑ Ong RG, Zhang Y. Controlling microbial contamination during hydrolysis drogenase (G6PDH) for vanillin tolerance in Saccharomyces cerevisiae. J of AFEX‑pretreated corn stover and switchgrass: effects on hydrolysate Biosci Bioeng. 2014;118(3):263–9. doi:10.1016/j.jbiosc.2014.02.025. composition, microbial response and fermentation. Biotechnol Biofuels. Verduyn C, Postma E, Scheffers WA, Van Dijken JP. Eec ff t of benzoic acid on 2015;8(1):1–17. doi:10.1186/s13068‑015‑0356‑2. metabolic fluxes in yeasts: a continuous‑ culture study on the regula‑ Skinner KA, Leathers TD. Bacterial contaminants of fuel ethanol produc‑ tion of respiration and alcoholic fermentation. Yeast. 1992;8(7):501–17. tion. J Ind Microbiol Biotechnol. 2004;31(9):401–8. doi:10.1007/ doi:10.1002/yea.320080703. s10295‑004‑0159‑0. Verstrepen KJ, Klis FM. Flocculation, adhesion and biofilm formation in yeasts. Taherzadeh MJ, Karimi K. Fermentation inhibitors in ethanol processes and Mol Microbiol. 2006;60(1):5–15. doi:10.1111/j.1365‑2958.2006.05072.x. different strategies to reduce their effects. Biofuels, alternative feed‑ Worley‑Morse TO, Deshusses MA, Gunsch CK. Reduction of invasive bacteria stocks and conversion processes. Cambridge: Academic Press; 2011. p. in ethanol fermentations using bacteriophages. Biotechnol Bioeng. 287–311. 2015;112(8):1544–53. doi:10.1002/bit.25586. Takuya I, Daisuke W, Yoko Y, Koichi T, Jun O, Hiroshi T, Hitoshi S, Jun S. An Wright J, Bellissimi E, de Hulster E, Wagner A, Pronk JT, van Maris AJ. Batch organic acid‑tolerant HAA1‑ overexpression mutant of an industrial and continuous culture‑based selection strategies for acetic acid toler ‑ bioethanol strain of Saccharomyces cerevisiae and its application to ance in xylose‑fermenting Saccharomyces cerevisiae. FEMS Yeast Res. the production of bioethanol from sugarcane molasses. AMB Express. 2011;11(3):299–306. doi:10.1111/j.1567‑1364.2011.00719.x. 2013;3(1):74. doi:10.1186/2191‑0855‑3‑74.
Journal
AMB Express – Springer Journals
Published: Aug 26, 2016