Deactivation of cellulases by phenols - USDA

10112

Enzyme and Microbial Technology 48 (2011) 54–60

Contents lists available at ScienceDirect

Enzyme and Microbial Technology

journal homepage: www.elsevier.com/locate/emt

Deactivation of cellulases by phenols

Eduardo Ximenes a,b, Youngmi Kim a,b, Nathan Mosier a,b, Bruce Dien d, Michael Ladisch a,b,c,∗

a Laboratory of Renewable Resources Engineering, Purdue University, West Lafayette, IN 47907-2022, United States
b Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, IN 47907-2022, United States
c Weldon School of Biomedical Engineering, Purdue University, West Lafayette, IN 47907-2022, United States
d Fermentation Biotechnology Research Unit, National Center for Agricultural Utilization Research, USDA, Agricultural Research Service, 1815 N. University Street, Peoria, IL 61604,
United States

article info abstract

Article history: Pretreatment of lignocellulosic materials may result in the release of inhibitors and deactivators of cellu-
Received 29 June 2010 lose enzyme hydrolysis. We report the identification of phenols with major inhibition and/or deactivation
Received in revised form 27 August 2010 effect on enzymes used for conversion of cellulose to ethanol. The inhibition effects were measured by
Accepted 3 September 2010 combining the inhibitors (phenols) with enzyme and substrate immediately at the beginning of the assay.
The deactivation effects were determined by pre-incubating phenols with cellulases or ␤-glucosidases
Keywords: for specified periods of time, prior to the respective enzyme assays. Tannic, gallic, hydroxy-cinnamic,
Cellulose and 4-hydroxybenzoic acids, together with vanillin caused 20–80% deactivation of cellulases and/or ␤-
Cellobiose glucosidases after 24 h of pre-incubation while enzymes pre-incubated in buffer alone retained all of
Cellulases their activity. The strength of the inhibition or deactivation effect depended on the type of enzyme, the
␤-Glucosidase microorganism from which the enzyme was derived, and the type of phenolic compounds present. ␤-
Enzyme inhibition Glucosidase from Aspergillus niger was the most resistant to inhibition and deactivation, requiring about
Enzyme deactivation 5 and 10-fold higher concentrations, respectively, for the same levels of inhibition or deactivation as
Cellulose hydrolysis observed for enzymes from Trichoderma reesei. Of the phenol molecules tested, tannic acid was the sin-
T. Reesei gle, most damaging aromatic compound that caused both deactivation and reversible loss (inhibition) of
A. niger all of enzyme activities tested.
Tannic acid
Aromatic acids © 2010 Elsevier Inc. All rights reserved.

1. Introduction lignin concentration of 0.025 mg/mL, corresponding to 1.85 mg
lignin/mg protein. ␤-Glucosidase pre-incubated for 1 h with
Phenols in higher plants range from simple low-molecular- crude and purified lignin at a ratio of 200 mg lignin/mg pro-
weight phenolic glycosides to polymeric compounds. Several tein, gave 75% less glucose compared to cellobiose hydrolysis
of them and their metabolites are believed to play roles in by ␤-glucosidase that was not pre-incubated with these lignin
plant defense mechanisms [1,2]. Conversely, digestibility of grass preparations.
cell walls increases when hydroxycinnamates (e.g., ferulic and
␳-coumaric acids) are removed by chemical or biological pretreat- Phenols including lignin degradation products, hydroxynamic
ments [3–5]. Phenolic hydroxyl groups associated with tannins acid derivatives, tannins and gallic acid are released during pre-
and lignin particles are known to adsorb proteins and deacti- treatment of lignocellulosic biomass [8–10]. Typical concentrations
vate cellulolytic enzymes and ␤-glucosidases during hydrolysis of phenolics are a function of biomass type, pretreatment con-
of microcrystalline cellulose [3,6,7]. These prior works, based on ditions, and ratio of biomass to water. Concentrations of soluble
hydrolysis of 7.5 mg/mL of a microcrystalline cellulose, mixed phenols increase with increasing biomass solids content due to a
with 4 different lignin preparations (crude neutral and acidic reduction in liquid volume. For water-based pretreatments, a solids
lignin and purified neutral and acidic lignin) at concentrations loading of 20% (w/v) results in a total phenolics concentration of
varying from 0.025 to 2.5 mg/mL, gave 20% inactivation of the 10–20 mg/mL in the resulting aqueous fraction. At 5 mg/mL, phe-
cellulase complex after 24 h. 15% inactivation still occurred at nols obtained by an ethyl acetate extraction of the aqueous fraction
derived from washed, steam pretreated poplar, had little effect
∗ Corresponding author at: Laboratory of Renewable Resources Engineering, Pur- on cellobiohydrolase activity compared to decreased glucose pro-
due University, West Lafayette, IN 47907-2022, United States. Tel.: +1 765 494 7022; duction by 80% [7] at a ratio of 150 mg phenol/mg protein. This
fax: +1 765 494 7023. indicated that the phenols reduced ␤-glucosidase activity. Con-
sidering that enzyme loadings of 1 mg enzyme (protein) per gram
E-mail address: [email protected] (M. Ladisch). solids correspond to a phenol to protein ratio of 30 mg phenols/mg
protein when the lignocellulose slurry is at 20% (w/v) solids, deac-
0141-0229/$ – see front matter © 2010 Elsevier Inc. All rights reserved.
doi:10.1016/j.enzmictec.2010.09.006

E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60 55

Table 1 Phenol concentration, mg/mL
Summary of protein and phenol concentrations. (deactivation studies)
5
Type of enzyme activity Protein concentration, mg/mL Phenol concentration, mg/mL Protein concentration, mg/mL 5
(inhibition studies) (inhibition studies) (deactivation studies) 5
Filter paper (FPase) 5
CMCase 0.060 2 15
␤-Glucosidase 0.034 2 15

Novozyme 188 0.065 2 3.4
Spezyme CP 0.32 2 15

tivation or inhibition caused by lignin-derived compounds is a experiments using these enzymes were done in triplicate and results reported as an
practical concern for enzyme hydrolysis of pretreated lignocellu- average with an indicated standard deviation.
lose. Due to advances in commercial cellulase preparations, protein
loadings for effective cellulose hydrolysis are currently in the 2.2. Enzyme activities
range of 2–10 mg per gram lignocellulose solids, which repre-
sent up to a 5 times decrease in enzyme from only 5 years ago. Cellulase and cellobiase activities were measured by procedures of the Interna-
Further decreases in required enzymes are needed for econom- tional Union of Pure and Applied Chemistry (IUPAC) [13]. Action against a specified
ically attractive processes [11,12]. As protein loadings decrease, substrate defines enzyme activity regardless of whether the enzyme is purified or
the relative impact of phenolic molecules on enzyme deactiva- exists in a background of other enzymes [10,14]. Release of reducing sugars from
tion becomes more pronounced as the ratios of phenols to protein filter paper (for measuring total cellulase activity) was determined per Miller [15],
increases. while glucose generated from cellobiose hydrolysis (for measuring ␤-glucosidase
activity) was determined by a d-glucose (GOPOD Format) assay kit (Megazyme,
In this study we tested inhibitory and/or deactivator effects Wicklow, Ireland) and confirmed by HPLC analysis as previously described [10].
of polymeric (tannic acid) and monomeric phenols (gallic, trans- Enzyme activity in the presence of ␳-nitrophenyl conjugated substrate (␳-NPGase
cinnamic, ferulic, ␳-coumaric, sinapic and 4-hydroxybenzoic acids, activity, also for measuring ␤-glucosidase activity) was determined at 50 ◦C and pH
vanillin and syringaldehyde) on cellulases and ␤-glucosidases from 4.8 following published methods [14,16]. One unit of cellulase (CMCase) is defined as
different microorganisms commonly used to produce commercial the release of one micromole of glucose equivalent per minute from carboxymethyl
enzymes for conversion of cellulose to ethanol. Inhibition effects cellulose and is a measure of endoglucanase activity. For ␳-nitrophenyl conjugated
were measured by combining phenols with enzyme and substrate substrates, one unit of activity is defined as one micromole of ␳-nitrophenyl (␳-NP)
immediately at the beginning of the assay. Deactivation effects released per minute and is a measure of ␤-glucosidase activity. The protein content
were measured by pre-incubating phenols with cellulases or ␤- of the commercial preparations was determined using a Pierce BCA Protein Assay
glucosidases, prior to carrying out the respective enzyme assays. kit (Thermo Scientific, Rockford, IL, Product No. 23225).
Concentrations of phenols tested here fall in the expected range of
typical values of phenols under industrial lignocellulose processing 2.3. Testing possible inhibitors and/or deactivators for cellulases (FPase and
conditions. CMCase activities) and ˇ-glucosidase (cellobiase and ␳-NPGase activities)

2. Materials and methods Inhibition effects were measured by combining the inhibitors (phenols) with
enzyme and substrate immediately at the beginning of the assay. The hydrolysis
Materials and methods was reported previously by Ximenes et al. [10], and are was carried out for 15–60 min at 200 rpm in an orbital shaker at 50 ◦C. For cel-
summarized below. lulases (FPase and CMCase activities) and ␤-glucosidase (cellobiase and ␳-NPGase
activities), the possible inhibitory effects of various phenolic compounds (tannic,
2.1. Materials gallic, trans-cinnamic, ferulic, ␳-coumaric and sinapic acids, vanillin and syringalde-
hyde) were tested at 0.2% (w/v) with phenol/enzyme ratios as summarized in
Spezyme CP (batch No. 3016295230, cellulase preparation from Trichoderma Tables 1 and 2. This concentration was lower than 0.5% (w/v) used in our previ-
reesei containing exo-, endo- and ␤-glucosidase activities) was provided by Genen- ous work on the inhibitory effects of phenols [10], and enabled us to extend the
cor, Danisco Division (Palo Alto, CA). Novozyme 188 (␤-glucosidase from Aspergillus range of measurements of inhibitor effects as well as minimize background spectral
niger) was purchased from Sigma (St. Louis, MO, Cat. No. C6150). The substrates absorbance due to the presence of the phenols in colorimetric assays.
used for enzyme assays were filter paper number 1 from Whatman (Whatman
International Ltd, England, Cat. No. 1001 125), carboxymethylcellulose sodium salt Deactivation was measured for ␤-glucosidases and cellulases, respectively, by
(low viscosity, Cat. No. C5678), cellobiose (Cat. No. C7252) and ␳-nitrophenyl ␤-d- pre-incubating enzyme with phenols at 0.3–1.5 mg phenol/mg protein at pH 4.8
glucopyranoside (Cat. No. N7006) purchased from Sigma (St. Louis, MO). Hydrolysis and 50 ◦C under agitation (200 rpm) for 1–24 h (Tables 1 and 2, column 2). After this
pre-incubation step the enzymes were diluted 100-fold and added to Filter paper
(3.3%, w/v), carboxymethylcellulose (1%, w/v), cellobiose (0.5% w/v) or ␳-NPG (0.1%
w/v) (Table 2, column 3) and incubated per the procedures given in Ximenes et al.
[10]. The amount of protein used was the same as for inhibition studies [10], but at
1/100th the concentration of phenols as used in the inhibition studies.

Table 2
Ratios of phenols to enzyme protein or activities for study of inhibition and deactivation effects. Phenols added based on enzyme activity (expressed as mg phenols/unit) or
protein (mg phenols/mg protein).

Inhibition assays (enzyme not pre-incubated Deactivation assays
with phenols)

mg phenol mg phenol mg phenol mg phenol Pre-incubation of enzyme with phenolsa Activity assays of pre-incubated enzyme
mg protein FPU CBU IU

mg phenol mg phenol mg phenol mg phenol mg phenol mg phenol mg phenol mg phenol
mg protein FPU CBU IU mg protein FPU CBU IU

Type of enzyme activity

Filter paper (FPase) 34.5 60 – – 0.30 0.50 – – 0.003 0.005 – –

CMCase 59.0 – – 3.2 0.30 – – 0.20 0.003 – – 0.002

␤-Glucosidase

Novozyme188 29 – 10 13 1.50 – 0.50 0.70 0.001 – 0.005 0.007

Spezyme CP 6 – 14 13.5 0.30 – 0.70 0.70 0.003 – 0.007 0.007

a The enzymes were first incubated at 50 ◦C, pH 4.8, for 1–24 h. Samples were withdrawn and diluted 100×, and added to substrate to measure hydrolysis. Corresponding
protein concentrations are 0.034 mg protein/mL (at 59 mg phenol/mg protein), 0.060 mg protein/mL (at 34.5 mg phenol/mg protein), 0.065 mg protein/mL (at 29.0 mg phe-
nol/mg protein), 0.32 mg protein/mL (at 6 mg phenol/mg protein), 3.4 mg protein/mL (at 1.5 mg phenol/mg protein) and 15.0 mg protein/mL (at 0.3 mg phenol/mg protein).
Concentrations of phenols ranged from 2 to 5 mg/mL as described in Section 2.

56 E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60

Fig. 1. Inhibitory effect of phenols on filter paper (FPase activity) and carboxymethylcellulose (CMCase activity) hydrolysis measured by DNS assay. Phenols were added at
the start of the assay. 100% FPase and CMCase activities corresponded to 50 FPU/mL (833 ␮mol glucose per min per mg protein) and 1560 IU/mL (45,882 ␮mol glucose per
min per mg protein), respectively. These values, measured in the absence of inhibitors were used as reference for calculation of loss of enzyme activities due to the presence
of phenols tested.

3. Results conditions, these phenol concentrations at protein concentrations
of 0.1–15 mg/mL enabled us to gauge their importance in enzyme
Measurement of cellulase activities after incubating 1 h deactivation and inhibition at industrial process conditions.
(Figs. 1–3) and 24 h (Figs. 4–6) with phenols enabled us to investi-
gate factors that contribute to the often observed decrease in the 3.1. Inhibition
rate of glucose formation during enzyme hydrolysis of pretreated
lignocellulosic biomass. Phenols do inhibit cellulases, and partic- Tannic acid strongly inhibited the combined cellulase activities
ularly ␤-glucosidase from T. reesei [10]. While inhibition occurs derived from T. reesei as measured by the filter paper and car-
almost immediately, continuing loss of enzyme activity is caused boxymethyl cellulose (i.e., endo-cellulase) activity assays (Fig. 1).
by time-dependent deactivation of endo-, exo- and ␤-glucosidase The other phenols had little or no effect. Individual HPLC mea-
enzymes by phenols generated by pretreatment, as described in the surement of cellobiose and glucose showed that accumulation of
discussion section of this paper. Recognition of the impacts of aro- cellobiose was lowered by tannic acid (60% inhibition, Fig. 2), while
matic molecules on these hydrolytic enzymes, coupled with known glucose accumulation was completely shut down (Fig. 2). The other
inhibitory effects of cellobiose and glucose [17–21] and the recal- phenolic molecules had little or no measurable effect on inhibiting
citrance of the remaining cellulose as hydrolysis proceeds, helps to either glucose or cellobiose formation.
explain the resistance of cellulose to enzyme hydrolysis [22,23].
The effects of phenols on ␤-glucosidases from T. reesei and
The phenols selected for this work to measure inhibition and A. niger were compared given the dramatic effect of tannic acid
deactivation were studied within the range of expected concen- on glucose forming (␤-glucosidase) activity. In addition to cel-
trations resulting from aqueous pretreatment of lignocellulose at lobiose, ␳-NPG was used as a substrate to measure ␤-glucosidase
solids loadings of 150–250 g/L, where the lignin content ranges activities. Since ␳-NPG gives a much stronger response to the
between 20 and 40% of the biomass (lignin concentration between hydrolytic activity of ␤-glucosidase than cellobiose, it is also a more
30 and 100 g/L). Concentrations of 0.2–0.5% or 2–5 mg/mL phenols sensitive indicator of inhibition. Tannic acid had little effect on ␤-
used in this work corresponded to release of 2–16% of the mass glucosidase in a commercial preparation from A. niger (Fig. 3a),
of lignin as phenols. While approximate, and covering a range of while almost completely inhibiting ␳-NPG hydrolysis, i.e., activity

100 Enzyme Activity mg inhibitor mg inhibitor
mg protein unit activity

80 FPase activity 34.5 60.0

Inhibition (%) 60 Inhibition of FPase activity (celllulases + β-glucosidase)
followed by reduced formation of:

Cellobiose
40

Glucose

20

0

Tannic acid Cinnamic acid ρ-Coumaric acid Vanillin 4-Hydroxybenzoic acid

Gallic acid Ferulic acid Sinapic acid Syringaldehyde

Fig. 2. Inhibitory effect of phenols on filter paper (FPase activity) hydrolysis based on glucose and cellobiose formation measured by HPLC. Phenols were added at the start
of the assay. 100% FPase activity corresponded to 50 FPU/mL (833 ␮mol glucose per min per mg protein), measured in the absence of inhibitors and used as reference for
calculation of loss of enzyme activities due to the presence of phenols tested.

E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60 57

(a) β-glucosidase in A. niger (Novozyme 188)

100 mg inhibitor mg inhibitor
mg protein unit activity

80

Inhibition (%) Novozyme 188 29.0 13.0 ρ-NGase

60 10.0 cellobiase

p-NPGase activity

40

Cellobiase activity

20

0

Tannic acid Ferulic acid Sinapic acid

Gallic acid ρ-Coumaric acid

(b) β-glucosidase in T. reesei (Spezyme CP)

100 mg inhibitor mg inhibitor
mg protein unit activity

80

Inhibition (%) Spezyme CP 6.0 13.5 ρ-NGase

60 14.0 cellobiase

p-NPGase activity

40

Cellobiase activity

20

0 Ferulic acid Sinapic acid

Tannic acid

Gallic acid ρ-Coumaric acid

Fig. 3. Inhibitory effect of phenols on ␳-NPG (␳-NPGase activity) and cellobiose (cellobiase activity) hydrolysis. Phenols were added at the start of the assay. 100% ␳-NPGase
activity for Spezyme CP and Novozyme 188 corresponded to 37 IU/mL (115 ␮mol ␳-NP per min per mg protein) and 380 IU/mL (5,846 ␮mol ␳-NP per min per mg protein),
respectively. 100% cellobiase activity for Spezyme CP and Novozyme 188 corresponds to 34 CBU/mL (106 ␮mol cellobiose converted per min per mg protein) and 515 CBU/mL
(7923 ␮mol cellobiose converted per min per mg protein), respectively. Activities were measured in the absence of inhibitors, and used as reference for calculation of loss of
enzyme activities due to the presence of phenols tested.

100 Enzyme mg deactivator mg deactivator

Activity mg protein unit activity

(Spezyme CP)

80

Deactivation (%) FPase activity 0.30 0.50

60 Deactivation of FPase activity (cellulases + beta-glucosidase)
followed by reduced formation of :

Cellobiose
40

Glucose

20

0

Tannic acid Cinnamic acid ρ-Coumaric acid Vanillin 4-Hydroxybenzoic acid

Gallic acid Ferulic acid Sinapic acid Syringaldehyde

Fig. 4. Deactivation of filter paper activity (cellulase and ␤-glucosidase activities) by phenols based on reduction of cellobiose and glucose formation measured by HPLC. The
enzyme (Spezyme CP) was incubated with the phenols for 24 h, at 50 ◦C and 200 rpm prior enzyme hydrolysis. 100% FPase activity corresponded to 50 FPU/mL (833 ␮mol
glucose per min per mg protein), as described in legend for Fig. 2, measured in the absence of deactivators, and used as reference for calculation of loss of enzyme activities

due to the presence of phenols tested.

58 E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60

of ␤-glucosidase from T. reesei (Fig. 3b). For the other phenols tested, 0 100
only gallic acid inhibited ␤-glucosidase (from T. reesei) for activity
measured with ␳-NPG as substrate (Fig. 3b). Since inhibitors and 80
enzyme were mixed immediately at the beginning of the inhibition -0.5
assay, these experiments also suggested that enzyme from T. reesei
will also be inhibited immediately upon addition of the enzyme to ln A/Ao -1 40 % of Initial Activity
a pretreated lignocellulose substrate that contains tannic or gallic
acid. When hydrolysis of filter paper with 0.06 mg/mL enzyme was -1.5 Filter Paper Activity (FPase) 20
carried out in buffer, glucose formation was double that of hydrol- -2 Endoglucanase (CMCase) activity
ysis carried out in the presence of 5 mg/mL tannic acid added to the
buffer. -2.5 5 10 15 20 25 0
0 30
3.2. Deactivation Time of exposure to tannic acid (h)

A different pattern emerged when combined cellulase activities Fig. 5. Deactivation effect as a function of time of tannic acid on filter paper (FPase
(cellulases and ␤-glucosidases) were incubated with phenols for activity) and carboxymethylcellulose (CMCase activity) hydrolysis was measured by
24 h, followed by measurement of enzyme activity. Tannic, gallic, DNS assay. The enzymes were incubated with the tannic acid for 24 h, at 50 ◦C and
hydroxycinnamic and 4-hydroxybenzoic acids as well as vanillin 200 rpm prior to enzyme hydrolysis. As defined for Fig. 1, 100% FPase and CMCase
and syringaldehyde all reduced activities of either cellulases and/or activities corresponded to 50 FPU (833 ␮mol glucose per min per mg protein) and
␤-glucosidases (Figs. 4, 5, 6a and b, 7a and b), with tannic acid 1560 IU/mL (45,882 glucose per min per mg protein), respectively, measured in the
having the most severe effect. Deactivation of an enzyme should absence of deactivators, and used as reference for calculation of loss of enzyme
increase with increasing time of exposure to the deactivating con- activities due to the presence of phenols tested.
ditions. This was observed for combined cellulase activities (filter
paper assay, Fig. 5), endoglucanase (Fig. 5) and ␤-glucosidase in the T. reesei cellulase preparation after 24 h (Fig. 6b). In the absence of
presence of tannic acid (Fig. 7). While ␤-glucosidase from A. niger inhibitors or deactivators, the enzyme activities (exo-, endo-, and
was more resistant to gallic and ␳-coumaric acids (Fig. 6a), gallic ␤-glucosidases) showed no loss of activity after incubation for 24 h
and ␳-coumaric acid also deactivated 50–70% of ␤-glucosidase in a at the same conditions.

(a) β-glucosidase in A. niger (Novozyme 188) mg deactivator mg deactivator
mg protein unit activity
100

80
Deactivation (%)
60 Novozyme 188 1.50 0.70 and 0.50

40

p-NPGase activity

20 Cellobiase activity

0

Tannic acid Cinnamic acid ρ-Coumaric acid Vanillin 4-Hydroxybenzoic acid

Gallic acid Ferulic acid Sinapic acid Syringaldehyde

(b) β-glucosidase in T. reesei (Spezyme CP) mg deactivator mg deactivator

100 mg protein unit activity

Deactivation (%)80 Spezyme CP 0.30 0.70 and 0.70

60

40

p-NPGase activity
Cellobiase activity

20

0

Tannic acid Cinnamic acid ρ-Coumaric acid Vanillin 4-Hydroxybenzoic acid

Gallic acid Ferulic acid Sinapic acid Syringaldehyde

Fig. 6. Deactivation effect of phenols on ␳-NPG (␳-NPGase activity) and cellobiose (cellobiase activity) hydrolysis. The enzymes were incubated with the phenols for 24 h, at
50 ◦C and 200 rpm prior enzyme hydrolysis. 100% ␳-NPGase activity for Spezyme CP and Novozyme 188 corresponds to 37 IU/mL (115 ␮mol ␳-NP per min per mg protein)
and 380 IU/mL (5846 ␮mol ␳-NP per min per mg protein), respectively. As defined for Fig. 3, 100% cellobiase activity for Spezyme CP and Novozyme 188 corresponded to
34 CBU/mL (106 ␮mol cellobiose converted per min per mg protein) and 515 CBU/mL (7923 ␮mol cellobiose converted per min per mg protein), respectively, measured in
the absence of deactivators, and used as reference for calculation of loss of enzyme activities due to the presence of phenols tested.

E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60 59

Table 3 (a) β-glucosidase in A. niger (Novozyme 188)
Summary of deactivation rate constants of selected phenols generated with pre- 0.5
treated biomass.

Phenol k (h−1) for indicated enzymea 0 100
80
Combined (FPAse) Exo CBH Endo ␤-Glucosidase -0.5 60 % of Initial Activity

From T. From A. n A/A0 -1 20
reesei niger
Tannic acid 0
Tannic acid 0.032 – 0.031 0.068 0.045 -1.5 Gallic acid 30
Gallic acid –
Ferulic acid – – – 0.046 0.006 Ferulic acid
␳-Coumaric – Coumaric acid
– – 0.005 0.004 -2

– – 0.060 0.004

a Constant determined from −k = ln(A/A0)/t from fit of data plotted in Figs. 5 and 7. -2.5 5 10 15 20 25
0

Time to exposure to phenolic molecules (h)

4. Discussion (b) β-glucosidase in T. reesei (Spezyme CP)
0.5
The overall decrease in measured enzyme activity with increas-
ing concentration of pretreated lignocellulose biomass is caused by ln A/A0 0 Tannic acid 100 % of Initial Activity
both inhibition and deactivation. Inhibition occurred immediately, -0.5 Gallic acid 80
and is represented by the intercepts at time = 0 in Figs. 5 and 7, Ferulic acid 60
as well as direct measurements presented in Figs. 1–3. Subse- -1 Coumaric acid
quent exposure of the enzyme to phenols caused additional loss -1.5 20
of enzyme activity, and follows a logarithmic decline which may
be fitted by an equation of catalyst deactivation [24–26] (Table 3): -2 0
-2.5 30
− d[A] = k[A] (1)
dt 0

Integration of Eq. (1) which represents first order kinetics gave 5 10 15 20 25
the following equation:
Time to exposure to phenolic molecules (h)

ln [A] = −kt (2) Fig. 7. Deactivation effect as a function of time of phenols on ␳-NPG (␳-NPGase activ-
[A]o ity) hydrolysis. The enzymes were incubated with the phenols prior the enzyme
hydrolysis. As defined for Fig. 3, 100% ␳-NPGase activity for Spezyme CP and
in which A = enzyme activity at time t; A0 = initial enzyme activity; Novozyme 188 corresponded to 37 IU/mL (115 ␮mol ␳-NP per min per mg protein)
k = rate constant; and t = time (h). and 380 IU/mL (5846 ␮mol ␳-NP per min per mg protein), respectively, measured in
the absence of deactivators, and used as reference for calculation of loss of enzyme
This equation, when fitted to the data, showed that a charac- activities due to the presence of phenols tested.
teristic rate constant, k, was a function of the type of enzyme,
microbial source from which the enzyme was derived, and the phe- While glucose and xylose may be fermented, thus relieving inhi-
nolic molecule. The interpretation of the causes of overall reduction bition of ␤-glucosidase and cellobiohydrolase, phenols are not so
of enzyme activity is complicated by the dual role of some of metabolized.
the phenols as inhibitors and deactivating molecules. This is seen
from inhibition data recently reported by Ximenes et al. [10], by Another major factor in slowing hydrolysis rates is the recal-
the inhibition noted here through direct measurement (Figs. 1–3), citrance of the cellulose itself. While pre-treatments have been
and extrapolation of the deactivation lines in Figs. 5 and 7. The defined and compared that address the recalcitrance barrier [29],
Y-intercepts for the molecules tested: tannic, gallic, ferulic and ␳- there is always a resistant fraction that will remain, and which will
coumaric acids, show reduced enzyme activity due to inhibition. In require more time and/or enzyme to be completely hydrolyzed.
the case of tannic, ␳-coumaric and ferulic acids this effect depends
on the type of enzyme: exo- or endo- cellulases (Figs. 1 and 2), ␤- The inhibitory and deactivating effects of polyphenols, repre-
glucosidase (Fig. 3a and b), and whether the source of ␤-glucosidase sented in this work (Figs. 1–7) are a clear target for enhancing rates
is from T. reesei or A. niger (Fig. 3a and b). Akin and Rigsby [27] found and extents of enzyme hydrolysis of pretreated cellulose. Substan-
that esterases release hydroxycynnamic acids derivatives (ferulic tial accumulation of tannins may occur in almost any part of plant
and ␳-coumaric acids) from lignocellulose. If these phenolic acids biomass, including seeds, fruit, leaves, wood, bark, and root [30].
are washed away before cellulases are added, increased release of Hence the results presented here have broad applicability.
sugars should result.
Most water soluble polyphenols (like tannins) in the molecu-
Product inhibition of cellobiohydrolase (CBH) by cellobiose lar weight range of 500–3000 form multiple hydrogen bonds with
and ␤-glucosidase by glucose [18–21] is a factor if glucose proteins [31]. The data in the literature for lignin [3,7] and the work
(inhibits ␤-glucosidase) and/or cellobiose (inhibits exo-cellulase, presented here with tannic acid (Figs. 1–7) clearly show that inhibi-
i.e., cellobiohydrolase) accumulates during saccharifaction. Product tion and deactivation by polyphenols is a primary cause of reduced
inhibition effects were not considered here, and while important, enzyme activity over long hydrolysis times for slurries of pretreated
are less critical than the impact of phenols since phenols are not lignocellulosic feedstocks. Removal of polyphenols would preserve
consumed (at least at same level as sugars) by microbial reaction. enzyme activity and enhance overall yields of monosaccharides.
Similarly, Kumar and Wyman [28] and Ximenes et al. [10] showed
xylo-oligosaccharides inhibit cellulases, while the substrate from Kawamoto et al. [32] showed that tannins interact with proteins
which they are derived (xylans) and the monosaccharide (xylose) and decrease ␤-glucosidase activity from almonds due to irre-
is less inhibitory. In the same fashion, the addition of enzymatic versible protein denaturation. A threshold concentration of tannin
activities to hydrolyze xylo-oligosaccharides and microorganisms below which this enzyme is not significantly precipitated depends
capable of fermenting these sugars can also relieve the inhibition. on source of the phenols [2]. Betula pendula and B. nana tannins
exhibit a threshold effect at 0.3 mg tannin/mg protein, while for
other species (B. pubescens, Salix caprea, S. pentandra, and Pinus
sylvestris) showed a threshold effect at 0.125 mg tannin/mg protein.

60 E. Ximenes et al. / Enzyme and Microbial Technology 48 (2011) 54–60

Tannins from bark of the woody species studied differed from each [2] Juntheikki MR, Julkunen-Tiito R. Inhibition of (-glucosidase and esterase by
other in their inhibition capacity: S. pentandra ≥ P. sylvestris > B. tannins from Betula, Salix and Pinus species. J Chem Ecol 2000;26(5):1151–65.
pubescens > S. caprea > B. nana > B. pendula. Tannins from wood were
shown to inhibit almond ␤-glucosidase at a ratio of 1 mg tannin/mg [3] Pan X. Role of functional groups in lignin inhibition of enzymatic hydrolysis of
protein, resulting in 80% loss of activity. We observed the strong cellulose to glucose. J Biobased Mater Bioenergy 2008;2(1):25–32.
inhibition of T. reesei ␤-glucosidase at a ratio of 6 mg tannic acid/mg
protein, while A. niger ␤-glucosidase was inhibited at a ratio of [4] Anderson WF, Akin DE. Structural and chemical properties of grass lig-
29 mg tannic acid/mg protein. The tannin-enzyme interaction will nocelluloses related to conversion for biofuels. J Ind Microbiol Biotechnol
not only depend on the source of tannin, but also on the nature and 2008;35(5):355–66.
source of the enzyme.
[5] Ximenes EA, Brandon SK, Doran-Peterson J. Evaluation of a Hypocrea jecorina
The time-dependent inactivation process of cellulases and enzyme preparation for hydrolysis of Tifton 85 bermudagrass. Appl Biochem
␤-glucosidases identified in our work resembled the slow loss Biotechnol 2008;146:89–100.
of catalytic activity of other enzymes, such as ribulose 1,5-
bisphosphate carboxylase/oxygenase (RuBisCO; EC 4.1.1.39), in a [6] Kawamoto H, Nakatsubo F, Murakami K. Synthesis of condensed tannin
process named “fallover,” which results from the formation of derivatives and their protein-precipitating capacity. J Wood Chem Technol
tightly binding inhibitors at the active site. The fallover effect con- 1990;10(1):59–74.
tinues until eventually a steady rate of catalysis is reached, which
is substantially less than the initial rate. Dilution of the mixture by [7] Excoffier G, Toussaint B, Vignon MR. Saccharification of steam-exploded poplar
50-fold did not relieve inhibition. A similar effect was observed in wood. Biotechnol Bioeng 1991;38:1308–17.
our work. The extent to which activity is reduced during “fallover,”
as well as the characteristic time over which the process takes place [8] Chen S-F, Mowery RA, Scarlata CJ, Chambliss CK. Compositional analysis of
is dependent on the external conditions and plant species [33,34]. water-soluble materials in corn stover. J Agric Food Chem 2007;55:5912–8.
The phenomenon is not a result of substrate depletion or product
accumulation. [9] Yang S-J, Kataeva I, Hamilton-Brehm SD, Engle NL, Tschaplinski TJ, Doeppke C,
et al. Appl Environ Microbiol 2009;75(14):4762–9.
5. Conclusion
[10] Ximenes E, Kim Y, Mosier N, Dien B, Ladisch M. Inhibition of cellulases by
Development of enzyme processes for hydrolysis of cellulose to phenols. Enzyme Microb Technol 2010;46:170–6.
glucose must reduce inhibition and deactivation effects in order
to enhance hydrolysis and reduce enzyme usage. Phenols are [11] Mosier N, Wyman C, Dale B, Elander R, Lee YY, Holtzapple M, et al. Features of
inhibitors and deactivators of cellulolytic enzymes and to a greater promising technologies for pretreatment of lignocellulosic biomass. Bioresour
extent, ␤-glucosidases. The polymeric phenol tannic acid was a Technol 2005;96(6):673–86.
major inhibitor and deactivator for all of the enzyme activities
tested, with monomeric phenolic compounds tested in this study [12] Sathitsuksanoh N, Zhu Z, Ho T-J, Bai M-D, Zhang Y-HP. Bamboo saccha-
having a less pronounced effect. Tannic, ferulic and ␳-coumaric rification through cellulose solvent-based biomass pretreatment followed
acids inactivated both T. reesei and A. niger ␤-glucosidases. Hence, by enzymatic hydrolysis at ultra-low cellulase loadings. Bioresour Technol
the identification and development of ␤-glucosidases that resist 2010;101(13):4926–9.
inhibition from both sugars and phenols will enhance cellulose
hydrolysis. Alternative strategies include carrying out enzyme [13] Ghose TK. Measurement of cellulase activities. Pure Appl Chem
hydrolysis over shorter periods of time to decrease time-dependent 1987;59(2):257–68.
deactivation, removing phenolics prior to enzyme hydrolysis by
separation methods, including washing of the solids, or using [14] Dien BS, Ximenes EA, O’Bryan PJ, Moniruzzaman M, Li X-L, Balan V, et al.
microbial, enzymatic, or chemical methods of converting the phe- Enzyme characterization for hydrolysis of AFEX and liquid hot-water pre-
nolics to an inactive form. treated distillers’ grains and their conversion to ethanol. Bioresour Technol
2008;99(12):5216–25.
Competing interest
[15] Miller GL. Use of dinitrosalicylic acid reagent for determination of reducing
Michael Ladisch is Chief Technology Officer at Mascoma Corpo- sugar. Anal Chem 1959;31:426–8.
ration.
[16] Li X-L, Dien BS, Cotta MA, Wu YV, Saha BC. Profile of enzyme production by
Acknowledgments Trichoderma reesei grown on corn fiber fractions. Appl Biochem Biotechnol
2005;121–124:321–34.
The authors wish to thank Xingya (Linda) Liu, Rick Hendrick-
son and Thomas Kreke, for their excellent technical assistance, [17] Gong C-S, Ladisch MR, Tsao GT. Cellobiase from Trichoderma viride: purification,
and David Hogsett, Mascoma, Mira Sedlak and Xuan Li, Purdue for properties, kinetics, and mechanism. Biotechnol Bioeng 1977;19:959–81.
their internal review of this paper. We thank Genencor for their
gift of enzymes. The material in this work was supported by DOE [18] Ladisch MR, Gong C-S, Tsao GT. Cellobiase hydrolysis by endoglucanase
grant DE-AC36-99GO10337, DE-FG02-06ER06-03, GO12O26-174, (glucan-glucanohydrolase) from Trichoderma reesei: kinetics and mechanism.
DE-FG02-06ER64301; DOE BES Project 0012846, USDA IFAFS con- Biotechnol Bioeng 1980;22:1017–126.
tract #00-52104-9663, and Mascoma Sponsored Research and Test
Agreements. [19] Ladisch MR, Hong J, Voloch M, Tsao GT. Cellulase Kinetics. In: Hollaender A,
Rabson R, Rodgers P, Pietro AS, Valentine R, Wolfe R, editors. Basic life science
References (Trends Biol Ferment Fuels and Chem), vol. 18. New York: Plenun Press; 1981.
p. 55–83.
[1] Rank NE, Köph A, Julkunen-Tiitto R, Tahvanainen J. Host performance and lar-
val performance of the salicylate-using leaf beetle Phratora vitellinae. Ecology [20] Hong J, Ladisch MR, Gong C-S, Wankat PC, Tsao GT. Combined product and sub-
1998;79:618–23. strate inhibition equation for cellobiase. Biotechnol Bioeng 1981;23:2779–88.

[21] Ladisch MR, Tsao GT. Engineering and economics of cellulose saccharification
systems. Enzyme Microb Technol 1986;8:66–9.

[22] Fan LT, Lee Y, Gharpuray MM. The nature of lignocellulosics and their pretreat-
ment for enzymatic hydrolysis. Adv Biochem Eng 1982;23:157–87.

[23] Himmel ME, Ding SY, Johnam DK, Adney WS, Nimlos MR, Brady JW, et al.
Biomass recalcitrance: engineering plants and enzymes for biofuels production.
Science 2007;315:804–7.

[24] Ladisch MR, Emery A, Rodwell VW. Economic implications of purification of
glucose isomerase prior to immobilization. Ind Eng Chem Process Des Dev
1977;16:309–13.

[25] Toscano G, Pirozzi D, Maremonti, Gianfreda L, Greco Jr G. Kinetics of enzyme
deactivation: a case study. Catal Today 1994;22:489–510.

[26] Kim Y, Hendrickson R, Mosier N, Ladisch M. Plug-flow reactor for continu-
ous hydrolysis of glucans and xylans from pretreated corn fiber. Energy Fuels
2005;19:2189–200.

[27] Akin DE, Rigsby LL. Corn fiber: structure, composition, and response to
enzymes for fermentable sugars and co-products. Appl Biochem Biotechnol
2008;144:59–68.

[28] Kumar R, Wyman CE. Effect of enzyme supplementation at moderate cellulase
loadings on initial glucose and xylose release from corn stover solids pretreated
by leading technologies. Biotechnol Bioeng 2009;102(2):457–67.

[29] Kumar R, Wyman CE. Does change in accessibility with conversion depend
on both the substrate and pretreatment technology. Bioresour Technol
2009;18:4193–202.

[30] Haslam E. Vegetable tannins-lessons of a phytochemical lifetime. Phytochem-
istry 2007;68:2713–21.

[31] Rowe JW, Conner AH. Extractives in eastern hardwoods—a review. Gen Tech
Rep FLP 18. Madison, WI: USDA Forest Products Laboratory; 1979.

[32] Kawamoto H, Mizutani K, Nakatsubo F. Binding nature and denaturation of pro-
tein during interaction with galloylglucose. Phytochemistry 1997;46(3):473–8.

[33] Edmondson D, Badger RB, Andrews TJ. Slow inactivation of ribulosebiphos-
phatecarboxylase during catalysis is caused by accumulation of a slow,
tight-binding inhibitor at the catalytic site. Plant Physiol 1990;93:1390–7.

[34] Witzel F, Götze J, Ebenhöh O. Slow inactivation of ribulose 1,5-biphosphate
carboxylase/oxygenase elucidated by mathematical models. FEBS J
2010;277:931–50.

Supplied by the U.S. Department of Agriculture, National Center for
Agricultural Utilization Research, Peoria, Illinois