Article pubs.acs.org/JAFC
Inhibitory Activity of Carbonyl Compounds on Alcoholic Fermentation by Saccharomyces cerevisiae Dongxu Cao,† Maobing Tu,*,† Rui Xie,† Jing Li,† Yonnie Wu,‡ and Sushil Adhikari# †
Forest Products Laboratory and Center for Bioenergy and Bioproducts, Auburn University, 520 Devall Drive, Auburn, Alabama 36849, United States ‡ Department of Chemistry and Biochemistry, Auburn University, Auburn, Alabama 36849, United States # Department of Biosystems Engineering, Auburn University, Auburn, Alabama 36849, United States S Supporting Information *
ABSTRACT: Aldehydes and acids play important roles in the fermentation inhibition of biomass hydrolysates. A series of carbonyl compounds (vanillin, syringaldehyde, 4-hydroxybenzaldehyde, pyrogallol aldehyde, and o-phthalaldehyde) were used to examine the quantitative structure−inhibitory activity relationship of carbonyl compounds on alcoholic fermentation, based on the glucose consumption rate and the final ethanol yield. It was observed that pyrogallol aldehyde and o-phthalaldehyde (5.0 mM) reduced the initial glucose consumption rate by 60 and 89%, respectively, and also decreased the final ethanol yield by 60 and 99%, respectively. Correlating the molecular descriptors to inhibition efficiency in yeast fermentation revealed a strong relationship between the energy of the lowest unoccupied molecular orbital (ELUMO) of aldehydes and their inhibitory efficiency in fermentation. On the other hand, vanillin, syringaldehyde, and 4-hydroxybenzaldehyde (5.0 mM) increased the final ethanol yields by 11, 4, and 1%, respectively. Addition of vanillin appeared to favor ethanol formation over glycerol formation and decreased the glycerol yield in yeast fermentation. Furthermore, alcohol dehydrogenase (ADH) activity dropped significantly from 3.85 to 2.72, 1.83, 0.46, and 0.11 U/mg at 6 h of fermentation at vanillin concentrations of 0, 2.5, 5.0, 10.0, and 25.0 mM correspondingly. In addition, fermentation inhibition by acetic acid and benzoic acid was pH-dependent. Addition of acetate, benzoate, and potassium chloride increased the glucose consumption rate, likely because the salts enhanced membrane permeability, thus increasing glucose consumption. KEYWORDS: fermentation, carbonyl compounds, inhibition, Saccharomyces cerevisiae
■
INTRODUCTION Developing alternative biofuels from renewable biomass has great potential to reduce U.S. dependence on foreign oil while improving national energy security and addressing the environmental issues related to energy use.1 The Renewable Fuel Standard mandates 36 billion gallons of biofuels should be produced annually by 2022, with 16 billion gallons coming from lignocellulosic biomass.2 Pretreatment of biomass is needed to break down the recalcitrant structure of plant cell walls for subsequent enzymatic hydrolysis and microbial fermentation.3,4 The pretreatment processes undesirably can generate inhibitors from the degradation of cellulose, hemicellulose, lignin, and extractives, many of which significantly reduce the microbial growth and fermentation productivity during the fermentation process.5−7 Significant efforts have been made to identify inhibitors in biomass hydrolysates using analytical separation and identification tools including high-performance liquid chromatography (HPLC), gas chromatography−mass spectrometry (GC/ MS), liquid chromatography−mass spectrometry (LC/MS), and nuclear magnetic resonance (NMR).8−12 The most potent inhibitors that have pronounced effects on microbial fermentation are yet to be understood. Most currently identified inhibitors (such as aliphatic acids, aromatic acids, aldehydes, ketones, and some phenolic compounds) have a functional carbonyl group (CO). Among those, furfural and © 2014 American Chemical Society
hydroxymethyl furfural (HMF) are aldehydes; vanillin and syringaldehyde are aromatic aldehydes; acetic acid, ferulic acid, syringic acid, 4-hydroxybenzonic acid, and protocatechuic acid are carboxylic acids; Hibbert’s ketone is a ketone.13,14 Carbonyl compounds are reactive: the polar double bond of the carbonyl group creates a partial positive charge on the carbonyl carbon and a partial negative charge on the oxygen; the positive charge on the carbon can initiate a nucleophilic addition between cellular nucleophiles (proteins and/or DNA) and electrophiles (carbonyl compounds). Consequently, the inhibitory effects of carbonyl compounds are governed by their electrophilic reactivity toward biological nucleophiles. Molecular descriptors that have been widely used to characterize electrophilic reactivity include hydrophobicity (log P, octanol/ water partition coefficient), molar refractivity, dipole moment, energy of the lowest unoccupied molecular orbital (ELUMO), energy of the highest occupied molecular orbital (EHOMO), and electrophilicity index (ω).15,16 Quantitative structure−activity relationship (QSAR) modeling used in drug discovery and environmental toxicity assessment is another useful tool in assessing the reactivity of carbonyl compounds.17 Any Received: Revised: Accepted: Published: 918
August 29, 2013 January 7, 2014 January 8, 2014 January 8, 2014 dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
Figure 1. Carbonyl compounds tested for their inhibitory activity on alcoholic fermentation by Saccharomyces cerevisiae. where μ is the molecular chemical potential and η is the molecular hardness. Yeast Strain, Culture, and Fermentation. S. cerevisiae (baker’s yeast, Fleischmann’s) was precultured in a YPG liquid medium composed of 10 g/L yeast extract, 20 g/L peptone, and 20 g/L glucose at 30 °C. After 12−15 h, the yeast was harvested and concentrated by centrifugation. The harvested yeast was washed at least twice with sterile water before it was inoculated for batch fermentation. The initial yeast innoculation was 2.0 g/L in fermentation. Batch fermentation was performed in 125 mL serum bottles with 60 mL of 2% (w/w) glucose only at 30 °C, with an initial pH at 6.0 and a shaking speed at 165 rpm for 48 h. No additional nutrients were added in the fermentation medium. The glucose solution was sterilized by filtration through a 0.2 μm sterile filter (VWR, Suwanee, GA, USA). Carbonyl compounds or salts were added into the glucose solution individually from their stock solutions or powders before sterilization. Samples were taken from the fermentation at various time points (1, 3, 6, 9, 12, 24, and 48 h) and centrifuged to collect the supernantant for identification and quantitation by HPLC. Yeast biomass concentration of each data point was measured by UV−vis spectrophotometry at 600 nm. All fermentations were carried out in duplicate. To examine the pH effects of acetic acid and benzoic acid on fermentation, acids were added to the glucose solution from their stock soltuions, and the pH was adjusted with 10% NaOH before sterilization. The resulting pH of the fermentation medium for individual carbonyl compounds was 5.3 (5.0 mM o-phthalaldehyde), 4.8 (5.0 mM pyrogallol aldehyde), 5.0 (5.0 mM syringaldehyde), 5.2 (5.0 mM 4-hydroxybenzaldehyde), 5.0 (5.0 mM vanillin), and 5.9 (glucose control), respectively. To examine the effects of inorganic and organic salts on fermentation, we added 27.5 mM NaCl, KCl, MgCl2, and CaCl2 and 50 mM NaOAc individually to 2% glucose solution for batch fermentation. No additional nutrients were added to the fermentation medium. The fermentation conditions were kept the same as above. The initial consumption rate of glucose (RS) was calculated on the basis of the sugar consumed in the first 3 h of fermentation [RS = (C0 − Ct)/t)], where C0 and Ct were the sugar concentrations at 0 and t h, respectively. The inhibition efficiency E (%) of carbonyl compound on fermentation was calculated using the decreased gluocse consumption rate
correlation between the structural features of carbonyl compounds and their biological activities will help to understand inhibition of biofuel production. Therefore, we are using QSAR modeling to assess this inhibition on ethanolic fermentation in our research. In this study, the inhibitory effects of selected aldehydes and acids (Figure 1) on ethanolic fermentation by Saccharomyces cerevisiae were investigated. The objective of this study was to identify the specific molecular descriptors that correlate the molecular structure of carbonyl compounds to their inhibitory activity. The quantitative information on carbonyl inhibition was evaluated on the basis of the initial rate of glucose consumption and the final ethanol yield. The alcohol dehydrogenase (ADH) activity in response to vanillin inhibition was determined from the batch fermentation results. Effects of acetic acid and benzoic acid on alcoholic fermentation were also studied with and without pH control. Using those data, the correlation between the inhibition efficiency and carbonyl molecular descriptors was established.
■
MATERIALS AND METHODS
Chemicals. Glycerol, 4-hydroxybenzaldehyde, 4-hydroxy-3-methoxybenzaldehyde (vanillin), and 2,3,4-trihydroxybenzaldehyde (pyrogallol aldehyde) were purchased from Acros Organics (Morris Plains, NJ, USA). 3,5-Dimethoxy-4-hydroxybenzaldehyde (syringaldehyde) was purchased from Aldrich (Milwaukee, WI, USA). oPhthalaldehyde was purchased from Pickering Laboratories (Mountain View, CA, USA). Benzoic acid and glucose (anhydrous) were purchased from Alfa Aesar (Ward Hill, MA, USA). Vanillyl alcohol and vanillic acid were purchased from Fluka (Milwaukee, WI, USA). Acetic acid, NaOAc, NaCl, KCl, MgCl2, and CaCl2 were purchased from Fisher Scientific (Fair Lawn, NJ, USA). All chemical reagents were of chromatographic grade. Stock solutions of aldehydes (1.0 M) were prepared in ethanol (Sigma-Aldrich, St. Louis, MO, USA), and stock solutions of carboxylic acids (1.0 M) were prepared in deionized water. All stocks were kept at 4 °C before use. Calculation of Physicochemical Descriptor. The physicochemical properties of model inhibitory compounds were calculated using open software MarvinSketch for the hydrophobicity (log P), steric parameter, and molecular refractivity. The ELUMO, EHOMO and dipole moment (u), were calculated using PM6 semiempirical methods (GaussView 5.0). Together, the molecular electrophilicity index (ω) was calculated by using the equation15,18 ω=
E (%) =
R SC − R S × 100 R SC
where RSC is the initial glucose consumption rate without inhibitor. HPLC Analysis. Glucose, ethanol, and glycerol were quantitated with a Shimadzu (LC-20A) HPLC system consisting of an autosampler, LC-20AD pump, and RID-10A detector, with a 300 mm × 7.8 mm i.d., 9 μm, Aminex HPX-87P column and a 30 mm × 4.6 mm i.d. guard column of the same material (Bio-Rad, Hercules,
2 E + 2E HOMOE LUMO + E LUMO2 μ2 = HOMO 2η 4(E LUMO − E HOMO)
919
dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
CA, USA). The mobile phase was composed of 5 mM sulfuric acid running isocratic at 0.6 mL/min. The temperature of the column was maintained at 45 °C throughout the run. Vanillin, vanillyl alcohol, and vanillic acid were separated and quantitated on the same HPLC system mentioned above with a diode array detector at 254 nm (SPD-M20A). The column for separation was a 250 mm × 4.6 mm i.d., 5 μm, C30 Carotenoid column (Waters, Milford, MA, USA). The mobile phase was acetonitrile/water/formic acid (20:80:1, v/v/v) at 0.6 mL/min, and the temperature of the column was maintained at 30 °C. Yeast Alcohol Dehydrogenase Activity Assay. To examine the alcohol dehydrogenase (ADH) enzyme activity in response to vanillin inhibition during the batch fermentation, yeast cells were collected at 0, 6, and 9 h of fermentation time with the addition of 0, 2.5, 5.0, 10, and 25 mM vanillin. Samples in 5 mL aliquots were first centrifuged at 2100 rpm for 10 min and then washed with deionized water and centrifuged in an Eppendorf centrifuge 5415R at 13000 rpm for 5 min. The cell pellets were then resuspended in a Y-PER yeast protein extraction reagent (Thermo Fisher Scientific, Rockford, IL, USA) and lysed at room temperature for 20 min with a gentle vortex. The Halt protease inhibitor cocktail (Thermo Fisher Scientific) was added to the lysed solution to prevent protein degradation. The lysed cells were centrifuged in an Eppendorf centrifuge 5415R at 13,000 rpm for 10 min to remove the cell debris, and the supernatant was collected for ADH enzyme activity assays. The total protein concentration of the supernatant was determined using the Quick Start Bradford dye reagent (Bio-Rad), measuring at 595 nm using bovine serum albumin (BSA) as the reference standard. The ADH enzyme activity was measured on the basis of method described previously by Vallee and Hoch.19 Briefly, 1.3 mL of 50 mM, pH 8.8, sodium pyrophosphate buffer, 0.1 mL of 95% ethanol, and 1.5 mL of 15 mM cofactor nicotinamide adenine dinucleotide (NAD+) (Sigma-Aldrich) were mixed fresh. The reaction was started by adding 0.1 mL of cell extract (lysate) to the solution. The reaction rate was then measured for 6 min at 340 nm for the change of NAD+. One unit of ADH activity was defined as the amount of enzyme required to catalyze the conversion of 1.0 μmol of ethanol to acetaldehyde per minute, at pH 8.8 and 25 °C. Statistical Analysis and Structure−Inhibition Relationships. The data results of carbonyl compound inhibition on glucose consumption and ethanol production were analyzed using one-way analysis of variance (ANOVA) (Excel 2007, Microsoft Office). A p value of pyrogallol aldehyde > syringaldehyde > furfural, HMF, vanillin > 4-hydroxybenzaldehyde. We believe that the inhibitory activity of these aldehydes is related to the chemical environment of their reactive carbonyl groups (CO). Previously, various molecular descriptors have
■
RESULTS AND DISCUSSION Effects of Aldehydes on Alcoholic Fermentation. The inhibitory effects of different carbonyl compounds on alcoholic fermentation were observed when vanillin, syringaldehyde, 4hydroxybenzaldehyde, pyrogallol aldehyde, and o-phthalaldehyde (all 5.0 mM) were added to the yeast fermentation using glucose as the substrate. Syringaldehyde, pyrogallol aldehyde, and o-phthalaldehyde showed significant inhibition on both the glucose consumption rate and ethanol productivity, whereas furfural, HMF, and 4-hydroxybenzaldeyde showed no inhibition on the final ethanol yields. Specifically, pyrogallol aldehyde and o-phthalaldehyde reduced the initial glucose consumption rate by 60 and 89%, respectively, and also decreased the final ethanol yield by 60 and 99%, respectively (Figure 2). Although syringaldehyde did not reduce the final ethanol yield (0.40 g/g), it decreased the initial glucose consumption rate by 31%. Interestingly, it was observed that vanillin, syringaldehyde, and 4-hydroxybenzaldehyde could 920
dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
Table 1. Physicochemical Descriptors of Carbonyl Compounds and Their Inhibitory Activity inhibitor
log P
ELUMO (eV)
EHOMO (eV)
dipole (Debye)
MR
ω
inhibition efficiency E (%)
HMF furfural vanillin syringaldehyde 4-hydroxybenzaldehyde pyrogallol aldehyde o-phthalaldehyde
−0.10 0.75 1.22 1.07 1.38 1.43 1.44
−0.79 −0.82 −0.77 −0.82 −0.71 −0.83 −1.50
−9.74 −9.95 −8.90 −8.82 −9.58 −9.19 −10.01
4.50 3.50 4.05 5.50 2.77 5.66 5.21
31.73 2 5.03 41.09 47.55 34.62 38.58 39.23
3.10 3.17 2.91 2.88 2.98 2.98 3.73
17.2 16.1 16.4 31.1 1.30 60.0 88.9
Table 2. Regression Analysis between Inhibition Efficiency and Molecular Descriptors eq
regression
n
r2
s
F
p
1 2 3 4
E = −0.541 − 0.979 ELUMO E = −0.607 + 0.21dipole E= −1.723 + 0.655ω E = 0.096 + 0.227 log P
7 7 7 7
0.74 0.55 0.55 0.17
0.171 0.225 0.225 0.31
14.31 6.16 6.22 1.01
0.013 0.056 0.055 0.360
inhibition. Here, we quantitated the carbonyl reactivity by calculating their molecular descriptors and correlating the electrophilicity to the inhibitory activity of carbonyl compounds. Table 1 summarizes the inhibitory effects and physicochemical descriptors of carbonyl compounds; the ELUMO correlated best with the inhibition efficiency (E), which could be described as a QSAR model:
been widely used to characterize chemical reactivity of quinones, unsaturated esters, and phenols.25−27 These molecular descriptors typically include hydrophobicity (log P, octanol/water partition coefficient), molar refractivity, dipole moment, energy of the lowest unoccupied molecular orbital (ELUMO), energy of the highest occupied molecular orbital (EHOMO), and electrophilicity index (ω).15,16 Carbonyl Reactivity and Quantitative Structure− Inhibition Relationship (QSIR). The inhibition efficiency of carbonyl compounds on yeast fermentation was correlated with several molecular descriptors as summarized in Table 1. Significant correlation (r2 = 0.74, p = 0.013) has been found between ELUMO and the inhibition efficiency (E) of aldehydes (eq 1). ELUMO is a global parameter that indicates the electrophilic reactivity of a molecule.25,28 The ELUMO of ophthalaldehyde showed the most negative value of −1.50 eV; ophthalaldehyde also was the most inhibitory compound to yeast fermentation (Table 1). Linear regression analysis found the dipole moment and ω were weakly correlated to the inhibitory efficiency (eqs 2 and 3, Table 2). The dipole moment is a measure of the polarity of the molecule.29 Pyrogallol aldehyde showed the highest dipole moments (5.66 D), and it was the second most inhibitory compound among the tested carbonyl compounds. The electrophilicity index (ω) is defined by the square of its electronegativity divided by its chemical hardness.18 o-Phthalaldehyde showed the highest electrophilicity index (3.73), but its overall weak correlation with inhibitory efficiency (E) was probably due to the EHOMO value. No correlation was found between EHOMO and inhibitory efficiency. Another global parameter, log P, which measured the hydrophobicity of a molecule,28 showed a poor correlation to inhibitory efficiency in regression eq 4 (Table 2). Previously, Zaldivar et al.30,31 reported that the toxicity of aldehydes and organic acids on ethanolic fermentation by Escherichia coli LY01 was directly related to hydrophobicity. However, the difference may come from the use of a different microorganism and a difference in calculation methods; we used sugar consumption instead of bacterial growth in calculating inhibition. Studies in the literature have proposed that the chemical reactivity of aldehydes contributed to the inhibition.31 This hypothesis was supported by the observation that the addition of bisulfite reduced the toxicity of acid hydrolysates.32 However, few studies have quantitated the chemical reactivity of aldehydes or correlated their reactivity to fermentation
E = −0.541 − 0.979E LUMO
This indicates that carbonyl compounds with higher negative ELUMO values would have higher inhibitory effects. ELUMO typically serves as one of the reactivity parameters for molecular electrophilicity.26 Carbonyl compounds, as electrophiles, could react with biological macromolecules (protein and DNA) by Michael addition15 or Schiff base formation,33 and their inhibitory effects have been related to strong electrophilic reactivity.34 Our results support using ELUMO to indicate the inhibitory effects of carbonyl compounds on alcoholic fermentation. Consequently, we have identified ELUMO, one of the molecular descriptors, as a reliable, perhaps global parameter that correlates carbonyl compound electrophilicity to inhibition. Obviously, more work to build the QSAR model with a larger number of carbonyl compounds on ethanol and butanol fermentation in S. cerevisiae and Clostridium acetobutylicum will fine-tune the model and make it more accurate in predicting the structure−inhibitory activity of carbonyl compounds in biomass hydrolysate. The fact that pyrogallol aldehyde and o-phthalaldehyde exhibited inhibition on the final ethanol yield was probably attributed to their strong electrophilic reactivity toward the biological nucleophilic sites. However, the addition of weak inhibitors including furfural, HMF, 4-hydroxybenzaldehyde, vanillin, and syringaldehyde indeed enhanced the final ethanol yield; in particular, vanillin had the highest enhancement. Thus, we used vanillin as a model compound to further examine the effect of weak inhibitors on ethanol (as well as glycerol) production with S. cerevisiae. Effects of Vanillin Concentration on Alcoholic Fermentation. To examine the tolerance of yeast to vanillin, we added various amounts of vanillin (2.5, 5.0, 10, 25 mM) to yeast fermentation (Figure 3). The pH of the medium decreased to 5.3, 5.0, 4.8, and 4.7, respectively. The results showed that vanillin at 25 mM significantly inhibited the final ethanol yield by 63%, whereas at 2.5, 5, and 10 mM it increased 921
dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
The conversion yields were also dose-dependent; at low concentration (2.5 mM), the yield of vanillyl alcohol was 96.6% and that of vanillic acid, 3.4%. At high concentration (5.0 and 10 mM), the yields of vanillyl alcohol were 71.4 and 59.5%, respectively, and the yields of vanillin acid were 1.4 and 0.7%, respectively. The enzymes responsible for converting vanillin to vanillyl alcohol and vanillic acid probably were alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH).37 The higher conversion yield (95% at 2.5 mM vanillin) of vanillyl alcohol was related to higher ADH enzyme activity in the anaerobic conditions, because the NAD+dependent ADH activity was typically 10-fold higher than ALDH activity in aerobic condition.38 The increase of vanillin concentration would decrease the ADH and ALDH enzyme activities, which probably explains why both vanillyl alcohol and vanillic acid yields decreased with increasing vanillin concentration. Consequently, we examined the ADH activity in cell extract at 6 and 9 h of fermentation with various amounts of added vanillin (Figure 3B). The results showed ADH activity decreased significantly with the increase of vanillin concentration. The ADH activity dropped significantly at 6 h of fermentation from 3.85 U/mg to 2.72, 1.83, 0.46, and 0.11 U/g, respectively, with vanillin at 0, 2.5, 5.0, 10.0, and 25.0 mM. A positive correlation has been found between the vanillyl conversion yield and ADH activity (y = 0.0569x − 2.6473, r2 = 0.90). At 9 h of fermentation, the ADH activity dropped similarly with increasing vanillin concentration (Figure 3B). Addition of vanillin decreased the glycerol yield in yeast fermentation as compared to the control (Figure 3C). The glycerol yield decreased by 12.3, 29.6, 45.7, and 82.7%, respectively, with vanillin at 2.5, 5.0, 10.0, and 25 mM. Glycerol, another major byproduct in yeast fermentation, is produced by the reduction of dihydroxyacetone phosphate to glycerol 3-phosphate (G3P) and then dephosphorylation to glycerol.39,40 The formation of glycerol maintains the redox balance and removes the extra NADP and NAD.39 The decrease of glycerol yield was probably related to the reducing activities of the yeast cells, which acted as an alternative redox sink for reoxidizing NADH/NADPH. As a result, the ethanol yield increased in the presence of low concentrations of vanillin (2.5, 5.0, and 10 mM) (Figure 3C). The addition of vanillin also decreased the pH of the fermentation medium (5.3−4.7). This could also potentially contribute to the increase in ethanol yield and the decrease in glycerol yield.41 However, the control experiments did not show a pH effect on glycerol yield within the pH range examined. Effects of Acetic Acid and Benzoic Acid on Alcoholic Fermentation. Acetic acid and benzoic acid have been implicated as potential inhibitors from lignocellulosic biomass pretreatment.42,43 To examine whether acids or their salts play a role in the fermentation inhibition, we first added various amounts of acetic acid into yeast fermentation with the initial pH adjusted at 6 (Figure 4A). The results showed that acetate (25, 50, 100, and 150 mM) did not inhibit ethanol production and glucose consumption at pH 6; contrarily, the final ethanol yield increased by 3%, and the glucose consumption rate increased by 20%. On the other hand, acetate (25, 50, 100, and 150 mM) inhibited the yeast growth and reduced the final biomass by 10, 20, 30, and 38% (Figure 4B). Interestingly, the beginning of the stationary phase was shifted earlier, from 12 h (the control) to 6 h. Glycerol production was also inhibited by acetate; the final glycerol concentration was reduced from 0.53 to 0.3, 0.25, 0.22, and 0.22 g/L, respectively.
Figure 3. Effects of vanillin on (A) glucose consumption, (B) ADH enzyme specific activity, and (C) glycerol production in fermentation by S. cerevisiae.
the final ethanol yield by 3.2, 6.7, and 8.7%, respectively. This indicated a concentration-dependent activity: At low concentrations (2.5, 5, and 10 mM) there was nearly complete glucose consumption within 12 h. However, high concentration (25 mM) inhibited glucose consumption significantly and stopped glucose usage at 24 h. Previously, it was reported that the less inhibitory activities of furfural and HMF were from detoxification by microbes that converted them to alcohols.35,36 In this study, we observed that vanillin could be reduced to vanillyl alcohol (major product) and oxidized to vanillic acid (minor product) at the same time under the tested conditions. 922
dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
Figure 5. Effect of acetic acid and benzoic acid with and without pH control on fermentation by S. cerevisiae.
1.04 to 0.31 g/L/h. When the pH was adjusted to 6, the acetic acid and benzoic acid changed to their corresponding salts (acetate and benzoate). Both appeared to have no inhibitory activity on the final ethanol yield (Figure 5). Actually at pH 6.0, benzoate and acetate increased the final ethanol yields by 16 and 3%, respectively. These results suggest that the inhibitory effect of acetic acid and benzoic acid are pH-dependent. The inhibition by these weak organic acids at low pH probably came from their undissociated forms. It is believed that the addition of weak organic acids has a significant impact on the growth energetics of fermenting microorganisms.24,44 When a weak acid was present in the fermentation medium, a low medium pH value (especially lower than the pKa value of the acids added) increased the fraction of undissociated forms of the acids, which diffused into the cells and decreased the intracellular pH. To maintain a constant pH, the cell must pump out the proton using a plasma membrane ATPase. Consequently, if the diffusion rate exceeded the transport capacity of the plasma membrane, acidification of the cells occurred, which led to the inhibition of both cell growth and ethanol production. Benzoic acid and acetic acid have pKa values of 4.2 and 4.8, respectively. At a medium pH of 3.0, both acids were present in their undissociated forms; this could cause significant inhibition on fermentation. The pH dependence of weak acid inhibition was also reported on cofermentation of glucose and xylose by genetically engineered S. cerevisiae,43 where the inhibition by acetic acid could be removed by adjusting the pH. To examine if lowering the medium pH affected the fermentation, we also compared the effect of 5 and 50 mM acetic acid on fermentation without pH control (Figure 5). Addition of 5 mM acetic acid decreased the medium pH to 3.6, whereas it showed no inhibition on either glucose consumption or final ethanol yield. This indicated that it was the concentration of the undissociated form of acetic acid, rather than the medium pH, which led to the inhibition. A similar result was observed before, in which the specific growth rate of S. cerevisiae in batch culture did not decrease as the medium pH decreased from 5 to 3.24 To examine whether the beneficial effect of acetate on glucose consumption rate was specifically related to acetate or to commonly used inorganic or organic salts, we added 27.5 mM NaCl, KCl, MgCl2, and CaCl2 and 50 mM NaOAc to the
Figure 4. Effects of acetate on ethanol yields, yeast biomass, and glycerol production in fermentation by S. cerevisiae.
Acetic or benzoic acid inhibition was found to be pHdependent. Without pH adjustment, the addition of 5.0 mM benzoic acid resulted in pH 3.3 in the fermentation medium. The glucose consumption was significantly inhibited, and the final ethanol yield was reduced by 89% when compared to the glucose control (Figure 5). The addition of 50 mM acetic acid resulted in pH 3.0, the glucose consumption was significantly inhibited, and the ethanol final yield was reduced by 94% when compared to the control. At pH 4.0, the acetic acid inhibited the volumetric ethanol productivity, but not the final ethanol yield. The volumetric ethanol productivity was reduced from 923
dx.doi.org/10.1021/jf405711f | J. Agric. Food Chem. 2014, 62, 918−926
Journal of Agricultural and Food Chemistry
Article
38%. This suggested that acetate had a stimulatory effect on anaerobic fermentation and an inhibitory effect on respiration (biomass yield typically was related to sugar respiration). In summary, effects of aldehydes and carboxylic acids on ethanolic fermentation have been quantitated on the basis of glucose consumption rate and final ethanol yield. With an attempt to build a structure−inhibition relationship of carbonyl compounds on fermentation, we have identified that ELUMO is a very good global parameter to correlate the molecular structure of carbonyl compounds to their inhibitory effects. The study also revealed that the addition of vanillin (
