Increasing cellulose accessibility is more important than removing lignin: a comparison of cellulose solvent-based lignocellulose fractionation and soaking in aqueous ammonia.
ABSTRACT While many pretreatments attempt to improve the enzymatic digestibility of biomass by removing lignin, this study shows that improving the surface area accessible to cellulase is a more important factor for achieving a high sugar yield. Here we compared the pretreatment of switchgrass by two methods, cellulose solvent- and organic solvent-based lignocellulose fractionation (COSLIF) and soaking in aqueous ammonia (SAA). Following pretreatment, enzymatic hydrolysis was conducted at two cellulase loadings, 15 filter paper units (FPU)/g glucan and 3 FPU/g glucan, with and without BSA blocking of lignin absorption sites. The hydrolysis results showed that the lignin remaining after SAA had a significant negative effect on cellulase performance, despite the high level of delignification achieved with this pretreatment. No negative effect due to lignin was detected for COSLIF-treated substrate. SEM micrographs, XRD crystallinity measurements, and cellulose accessibility to cellulase (CAC) determinations confirmed that COSLIF fully disrupted the cell wall structure, resulting in a 16-fold increase in CAC, while SAA caused a 1.4-fold CAC increase. A surface plot relating the lignin removal, CAC, and digestibility of numerous samples (both pure cellulosic substrates and lignocellulosic materials pretreated by several methods) was also developed to better understand the relative impacts of delignification and CAC on glucan digestibility.
- SourceAvailable from: Yi-Heng Percival Zhang[show abstract] [hide abstract]
ABSTRACT: Engineering costly cellulases on natural cellulosic substrates is of importance for emerging biomass-based biorefineries. Directed enzyme evolution is becoming a popular tool, but identification of desired mutants from a large mutant library remains challenging sometimes. In this work, we demonstrated a novel combinatorial selection/screening strategy for finding thermostable beta-glucosidase on its natural substrate-cellobiose. First, selection was conducted through complementation of beta-glucosidase for non-cellobiose-utilizing Escherichia coli so that only the cells expressing active beta-glucosidase can grow on a M9 synthetic medium with cellobiose as the sole carbon source (selection plate). Second, the clones on the selection plates were duplicated by using nylon membranes. After heat treatment, the nylon membranes were overlaid on M9/cellobiose screening plates so that remaining activities of thermostable beta-glucosidase mutants hydrolyzed cellobiose on the screening plates to glucose. Third, the growth of an indicator E. coli strain that can utilize glucose but not cellobiose on the screening plates helped detect the thermostable beta-glucosidase mutants on the selection plates. Several thermostable mutants were identified from a random mutant library of the Paenibacillus polymyxa beta-glucosidase. The most thermostable mutant A17S had an 11-fold increase in the half-life of thermoinactivation at 50 degrees C.Biotechnology and Bioengineering 05/2009; 103(6):1087-94. · 3.65 Impact Factor
Increasing Cellulose Accessibility Is More
Important Than Removing Lignin: A Comparison
of Cellulose Solvent-Based Lignocellulose
Fractionation and Soaking in Aqueous Ammonia
Joseph A. Rollin,1Zhiguang Zhu,1Noppadon Sathitsuksanoh,1,2Y.-H. Percival Zhang1,2,3
1Biological Systems Engineering Department, Virginia Polytechnic Institute and State
University, 210-A Seitz Hall, Blacksburg, Virginia 24061; telephone: 1-540-231-7414; fax: þ1-
540-231-3199; e-mail: email@example.com
2Institute for Critical Technology and Applied Sciences (ICTAS), Virginia Polytechnic
Institute and State University, Blacksburg, Virginia
3DOE BioEnergy Science Center (BESC), Oak Ridge, Tennessee
Received 18 May 2010; revision received 11 August 2010; accepted 17 August 2010
Published online 1 September 2010 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/bit.22919
ABSTRACT: While many pretreatments attempt to improve
the enzymatic digestibility of biomass by removing lignin,
cellulase is a more important factor for achieving a high
sugar yield. Here we compared the pretreatment of switch-
grass by two methods, cellulose solvent- and organic sol-
vent-based lignocellulose fractionation (COSLIF) and
soaking in aqueous ammonia (SAA). Following pretreat-
ment, enzymatic hydrolysis was conducted at two cellulase
loadings, 15 filter paper units (FPU)/g glucan and 3 FPU/g
glucan, with and without BSA blocking of lignin absorption
after SAA had a significant negative effect on cellulase
performance, despite the high level of delignification
achieved with this pretreatment. No negative effect due to
lignin was detected for COSLIF-treated substrate. SEM
micrographs, XRD crystallinity measurements, and cellulose
accessibility to cellulase (CAC) determinations confirmed
that COSLIF fully disrupted the cell wall structure, resulting
in a 16-fold increase in CAC, while SAA caused a 1.4-fold
CAC increase. A surface plot relating the lignin removal,
CAC, and digestibility of numerous samples (both pure
cellulosic substrates and lignocellulosic materials pretreated
by several methods) was also developed to better understand
the relative impacts of delignification and CAC on glucan
Biotechnol. Bioeng. 2010;xxx: xxx–xxx.
? 2010 Wiley Periodicals, Inc.
KEYWORDS: biofuels; biomass pretreatment; cellulose sol-
vent- and organic solvent-based lignocellulose fractionation
(COSLIF); cellulose accessibility to cellulase; lignin re-
moval; soaking in aqueous ammonia (SAA)
Biomass-derived fuels offer the only renewable liquid
alternative to petroleum-based transportation fuels (Lynd
et al., 2009; Zhang, 2008). Biomass resources, such as crop
residues, dedicated bioenergy crops grown on marginal
land, and timber industry waste, are cheap and abundant
renewable feedstocks. A recent report on the technical
feasibility of a billion-ton annual biomass supply estimated
that enough biomass could be sustainably produced to
replace more than a third of the current US transportation
fuel demand, with only modest land-use change (Perlack
et al., 2005).
The largest obstacle to economical production of
cellulosic biofuels is cost-effectively releasing sugars from
recalcitrant lignocellulose (Lynd et al., 2008; Zhang, 2008).
Saccharification of biomass normally consists of pretreat-
ment or fractionation, which generates more reactive
biomass, followed by enzymatic hydrolysis, which produces
soluble fermentable sugars. Increasing overall sugar yields
and decreasing processing costs can be accomplished by
(1) improving pretreatment (Sathitsuksanoh et al., 2009;
Wyman, 2007), (2) enhancing cellulase performance
Abbreviations used: ARP, ammonia recycle percolation pretreatment; BSA, bovine
serum albumin; CAC, cellulose accessibility to cellulase; CAFI, consortium for applied
fundamentals and innovation; CBM, cellulose binding module; COSLIF, cellulose
DA, dilute acid pretreatment; EG, ethylene glycol; FPU, filter paper unit; GFP, green
fluorescent protein; NCAC, non-cellulose accessibility to cellulase; NREL, National
Renewable Energy Laboratory; QS, quantitative saccharification; RAC, regenerated
amorphous cellulose; SAA, soaking in aqueous ammonia pretreatment; TGC, thior-
edoxin–GFP–CBM fusion protein; TSAC, total substrate accessibility to cellulase; XRD,
Correspondence to: Y.-H. Percival Zhang
? 2010 Wiley Periodicals, Inc.
Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2010 1
(Heinzelman et al., 2009; Liu et al., 2009; Zhang et al., 2006),
(3) recycling cellulase (Lee et al., 1994; Tu et al., 2007; Zhu
et al., 2009b), (4) decreasing enzyme production costs
(Himmel et al., 2007), and (5) producing less recalcitrant
bioenergy plants (Chen and Dixon, 2007). In-depth
understanding of substrate characteristics after biomass
pretreatment and their relationship with enzymatic cellulose
hydrolysis is vital for decreasing costs associated with
biomass saccharification (Zhang and Lynd, 2006).
Substrate accessibility has long been recognized as an
important factor in the enzymatic hydrolysis of lignocellu-
lose (Arantes and Saddler, 2010; Grethlein, 1985; Lee et al.,
1994; Zhang and Lynd, 2004). Accessibility of pretreated
biomass substrate to cellulase is usually measured based on
the adsorption of active cellulase at decreased temperatures
(Kumar and Wyman, 2009; Lu et al., 2002), but this method
cannot distinguish cellulose and non-cellulose fractions
(Zhu et al., 2009a). Recently, a technology has been
developed todetermine the cellulose accessibility to cellulase
(CAC) and non-cellulose accessibility to cellulase (NCAC)
based on adsorption of a non-hydrolytic fusion protein
containing a green fluorescent protein and a family 3
cellulose-binding module (Hong et al., 2007; Zhu et al.,
2009a). This technique provides a useful metric for
pretreatment comparison. Due to the difficulty of accurately
measuring CAC previously, many pretreatments have been
compared on the basis of lignin or hemicellulose removal and
glucan digestibility (Converse, 1993; Mosier et al., 2005). The
importance of removing lignin has been frequently suggested,
as many studies have shown a correlation between
delignification and increased sugar release (Chang and
Holtzapple, 2000; Fan et al., 1981; Ishizawa et al., 2009; Liu
and Wyman, 2003; Yang et al., 2002).
Cellulose solvent- and organic solvent-based lignocellu-
lose fractionation (COSLIF) and soaking in aqueous
ammonia (SAA) represent two different biomass pretreat-
ment goals (Fig. 1). COSLIF is mainly focused on breaking
the linkages among lignocellulose components and disrupt-
cellulose accessibility greatly (Sathitsuksanoh et al., 2009;
Zhang et al., 2007a; Zhu et al., 2009a). The COSLIF
pretreatment accomplishes this disruption of biomass under
mild conditions by using the successive actions of cellulose,
organic, and aqueous solvents to fractionate lignocellulosic
biomass into lignin, hemicellulose oligomers, and highly
reactive cellulose (Moxley et al., 2008; Zhang et al., 2007a).
of lignin and dissolvesome hemicellulose(Kim and Lee, 2005;
Sousa et al., 2009). SAA is typically conducted at a moderate
temperature (?608C), but harsher conditions can be
In order to better understand the root causes of biomass
recalcitrance and the substrate characteristics impacting
enzymatic cellulose hydrolysis,
COSLIF- and SAA-pretreated switchgrass, using composi-
tional analysis, enzymatic hydrolysis, scanning electron
microscopy (SEM), X-ray diffraction (XRD), cellulose
this study compared
accessibility to cellulase (CAC), and non-cellulose accessi-
bility to cellulase (NCAC).
Materials and Methods
Chemicals and Materials
All chemicals were reagent grade, purchased from Sigma–
Aldrich (St. Louis, MO), unless otherwise noted. Aqueous
ammonia (28–30%), 85% phosphoric acid, and 95%
ethanol were purchased from Fisher Scientific (Houston,
TX). Trichoderma reesei cellulase(Novozyme 50013, 84 filter
paper units (FPU)/mL) and b-glucosidase (Novozyme
North American (Franklinton, NC). Alamo switchgrass
(Panicum virgatum) was gifted by the National Renewable
Energy Lab (Golden, CO). The biomass was harvested in
November 2007, baled and shipped a week after harvest,
hammer-milled to less than one inch, then processed with a
Wiley mill to 1mm. The resulting biomass was then sieved
to less than 20mesh (0.85mm particle size) and greater than
80mesh (0.18mm particle size).
COSLIF pretreatment of switchgrass was conducted as
described elsewhere (Sathitsuksanoh et al., 2009, 2010),
using 95% (v/v) ethanol as the organic solvent. In brief, one
8mL of 85% phosphoric acid at 608C for 45min in 50mL
plastic centrifuge tubes. Biomass dissolution was stopped by
Areas of the relative components correspond to their percentage of the microfibril or
pretreated material. Cellulose surfaces susceptible to enzymatic attack (110 face) are
highlighted in red. SAA lignin is greatly decreased, but remains widely distributed. The
quantity of COSLIF lignin is high, but it may form clusters as depicted here. Cellulose
accessibility increases greatly following COSLIF pretreatment.
Conceptual images of the effects of SAA and COSLIF pretreatments.
Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2010
was then conducted in a swinging bucket centrifuge at
4,500rpm at room temperature for 15min. The supernatant
was decanted, and an additional 40mL of ethanol was mixed
with the cellulose- and hemicellulose-containing slurry. Solid/
the supernatant was decanted, the pellets were re-suspended
and washed twice with 40mL of deionized water. The
remaining pellet (primarily amorphous cellulose) was
neutralized to pH?7 with 2M sodium carbonate.
Soaking in Aqueous Ammonia (SAA)
A range of SAA conditions were attempted, where the
ammonia concentration, pretreatment temperature, pre-
treatment time, and liquids to solids loading ratio were
adjusted. The pretreatment that was most effective for
removing lignin as well as a significant hemicellulose
200-mL serum bottle, followed by 21mL of 10% (w/w)
ammonia. The tube was capped with a rubber stopper and
placed into apreheated1408Cfurnace.After 14h,thebottles
were allowed to cool to room temperature. Liquid–solid
separation was conducted by filtration using glass filters
purchased from Fisher Scientific, and the remaining solids
were washed with deionized water until a pH of 7 was
achieved (?500mL). It should be noted that heating
ammonia to this temperature may cause potential explosive
hazards. Caution should be employed during any repetition
of this experiment.
Carbohydrate and Lignin Assays
determined with a modified quantitative saccharification
(QS) procedure (Moxley and Zhang, 2007). In modified QS,
secondary hydrolysis was conducted in the presence of 1%
(w/w) sulfuric acid at 1218C, for 1h to more accurately
determine the quantities of sugars susceptible to acid
degradation (e.g., xylan). The standard NREL biomass
protocol was used to measure lignin and ash (Sluiter et al.,
2008). In brief, solids remaining after two-stage acid
hydrolysis were held at 1058C overnight. The weight of
the dried solids corresponds to the amount of acid-insoluble
lignin and ash in the sample. The weight of the ash only
fraction was then determined by heating the solids to 5758C
for 24h. Percent acid-soluble lignin in the sample was
determined by measuring the UV absorption of the acid
hydrolysis supernatant at 320nm. Monomeric sugars were
measured with a Shimadzu HPLC, with a Bio-Rad Aminex
HPX-87H column (Richmond, CA), at 608C with 50mM
sulfuric acid as the mobile phase, operated at a rate of
0.6mL/min (Zhang et al., 2007a). All carbohydrate and
lignin assays were conducted in triplicate.
Enzymatic hydrolysis of substrate produced by each
pretreatment was conducted at two enzyme loadings (15
and 3FPU/g glucan, each supplemented with 10U of b-
experiments was conducted in triplicate and repeated in
duplicate. To prevent lignin from competitively binding
cellulase, blocking was achieved by adding BSA (10g/L) for
30min at room temperature prior to hydrolysis. BSA has
previously been shown to adsorb irreversibly to lignin
binding sites, without interfering with the sites important
Pretreated switchgrass samples were diluted to 10g glucan/L
in a 50mM sodium citrate buffer (pH 4.8) and 0.1% (w/v)
NaN3. Hydrolysis experiments were conducted in a shaking
water bath at 250rpm and 508C. Eight hundred microliters
of well-mixed hydrolysate were removed, followed by
immediate centrifugation at 13,000rpm for 5min. Then
exactly 500mL of the supernatant was transferred to another
microcentrifuge tube and incubated at room temperature
for 30min, to allow the conversion of any cellobiose present
to glucose, by the action of b-glucosidase remaining in the
supernatant. The supernatant was then acidified by adding
30mL of 10% (w/w) sulfuric acid, after which it was placed
in a ?48C freezer. After being frozen overnight, samples
were thawed, mixed well and then centrifuged at 13,000rpm
for 5min, to remove any precipitated solid sediment. The
measured by HPLC using a Bio-Rad HPX-87H column.
Galactose and mannose co-eluted with xylose. After
completion of 72h hydrolysis, the remaining hydrolyzate
was transferred to a 50mL centrifuge tube, centrifuged at
4,500rpm for 15min, and soluble sugar content was
determined using the same procedure as other hydrolyzate
samples, as described above. After all remaining hydrolyzate
was decanted, and the pellet was resuspended in 20mL of
water and centrifuged to remove residual soluble sugars from
the pellet. The sugar content of the washed pellet was
determined by QS. Enzymatic glucan digestibility after 72h
was calculated using the ratio of dissolved glucose in the
supernatant to the sum of this dissolved glucose and the
Scanning Electron Microscopy (SEM)
Micrographs were taken of untreated, COSLIF-treated, and
SAA-treated switchgrass samples using a Zeiss-DSM940
(Carl Zeiss, Okerkochen, Germany). All samples were
sputter-coated with gold and imaged by SEM, as described
elsewhere (Moxley et al., 2008; Sathitsuksanoh et al., 2009).
X-Ray Diffraction (XRD)
X-ray diffractograms of all samples were measured using a
Bruker D8 Discover X-ray diffractometer (Madison, WI)
Rollin et al.: COSLIF Versus SAA
Biotechnology and Bioengineering
with Cu Karadiation (l¼1.54178A˚)with the scanningrate
of 48/min, ranging from 108 to 608.
Substrate Accessibility Assays
on the basis of the maximum adsorption capacity of the TGC
protein (Zhu et al., 2009a). The TGC protein is a non-
hydrolytic fusion protein containing a green fluorescence
protein and cellulose-binding module (Hong et al., 2007).
Recombinant thioredoxin-green fluorescent protein-cellulose
binding module (TGC) fusion protein was produced in
Escherichia coli BL21 (pNT02) (Hong et al., 2007), purified by
adsorption onto regenerated amorphous cellulose, and
desorbed with ethylene glycol (EG) (Hong et al., 2008). EG
citrate buffer (pH 6.0), and the TGC solution was concentrated
using 10,000Da molecular weight cut-off centrifugal ultrafilter
columns (Millipore Co., Billerica, MA). CAC assays were
conducted in duplicate. Mass concentration of TGC protein in
the liquid phase (the fraction that had not adsorbed) was
measured by sample fluorescence using a BioTek multi-
detection microplate reader. Cellulose accessibility to cellulase
(CAC, m2/g biomass) was measured based on the maximum
with 5g/L BSA. Non-cellulose accessibility to cellulase
(NCAC, m2/g biomass) was calculated from the equation
NCAC¼TSAC?CAC (Zhu et al., 2009a).
Surface Graphing Software
Lignin removal percentages, normalized CAC values, and
72h glucan digestibility (15FPU/g glucan) data were
imported to TableCurve 3D v4.0.01 (Systat Software, Inc.,
San Jose, CA). A conservative smoothing method, Smooth
Non-Uniform Rational B-Splines (NURBS), was applied to
these data to obtain the surface presented.
COSLIF and SAA
Untreated switchgrass contained approximately 35wt%
cellulose, 22wt% hemicellulose, and 20wt% acid-insoluble
lignin. COSLIF conditions were 85% phosphoric acid at
508C for 45min, with 95% ethanol used as the organic
solvent (Sathitsuksanoh et al., 2009, 2010). The COSLIF
pretreatment resulted in decreased hemicellulose (67% removal)
and slightly reduced acid-insoluble lignin (34% removal). SAA
was conducted with a 1:20 solids to liquids loading in 10wt%
aqueous ammonia at 1408C for 14h. Under these conditions,
SAA resulted in a substrate with reduced hemicellulose (42%
removal) and greatly reduced acid-insoluble lignin (74%
removal). As shown in Table I, COSLIF and SAA pretreatments
resulted in similar cellulose contents (approximately 50wt%),
but there were large differences in hemicellulose and lignin
contents. Comparing the compositions of COSLIF- and SAA-
result of the different mechanisms of these pretreatments.
For the hydrolysis of COSLIF-treated switchgrass, there was
no statistically significant difference between the cases of
BSA blocking and unblocked substrate at either enzyme
loading (Fig. 2A). Over 80% digestibility was achieved after
12h at a typical cellulase loading (15FPUs/g glucan), and
the glucan digestibility increased to 90% final digestibility
hydrolysis rate was slower, requiring 24h to reach >80%
digestibility. The final glucan digestibility was 85%, slightly
less for the lower cellulase loading.
In contrast to the results obtained for COSLIF-pretreated
materials, BSA blocking caused a large increase in the rates
of hydrolysis for SAA-pretreated switchgrass (Fig. 2B). BSA
blocking of SAA-treated substrate resulted in final digest-
ibility increases of 30% at an enzyme loading of 15FPUs/g
glucan and of 40% for 3FPUs/g glucan. The final
digestibility for SAA-pretreated switchgrass with 15FPU/g
glucan was 64% without BSA blocking, and 82% with BSA
blocking. With 3FPU/g glucan, final digestibility was 42%
without BSA blocking and 58% with BSA blocking. These
large digestibility increases in the presence of BSA suggest
that a large negative effect is caused by an adsorptive lignin
fraction that remained following SAA pretreatment.
A qualitative assessment of the substrates was conducted
using SEM microscopy. Figure 3 shows clear cell wall
structures in the untreated switchgrass. Vascular bundles
and parenchyma are observed in cross-section (A), and
closer magnification reveals a complex ordered structure
Comparison of untreated, SAA-treated, and COSLIF-treated switchgrass compositions.
XMG, xylan, mannan, and galactan; AIL, acid-insoluble lignin.
Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2010
that has been cut open by the milling process (B). This
in a delignified structure with a different supramolecular
structure than the untreated material. Much different pictures
are observed after COSLIF. In Figure 3E and F, the fibril
structure is completely disrupted. The difference in ordered
structures apparent in these micrographs is in good agreement
with the expected results of these pretreatments (Fig. 1).
Quantitative Cellulose Accessibility Assay
A quantitative measure of the total substrate accessibility to
cellulase (TSAC) and the cellulose accessibility to cellulase
(CAC) were determined with the TGC fusion protein
adsorption assay (Hong et al., 2007; Zhu et al., 2009a).
Figure 4 shows schemes of TGC protein components (A),
the TSAC assay (B), and the CAC assay (C). For both TSAC
and CAC, COSLIF caused a much greater increase than SAA
(Table II). SAA increased CAC by 1.4-fold compared to
untreated switchgrass (0.49?0.049m2/g biomass), whereas
COSLIF increased CAC by 16-fold. Non-cellulose accessi-
bility to cellulase (NCAC) also increased to a much greater
extent for COSLIF-pretreated substrate (twofold increase)
than for SAA-pretreated material (8% increase). Since
NCAC primarily represented lignin binding sites, this
surprisingresult meansthat although lignin accessibility was
much higher after COSLIF, the relatively small amount of
lignin accessible after SAA was responsible for the large
negative effect seen in the hydrolysis results. Furthermore,
although SAA under the conditions tested removed a large
fraction of the lignin portion, the lignin accessibility
(NCAC) increased slightly compared to the untreated case,
suggesting that the lignin remaining after SAA was more
dard and low enzyme loadings, with and without BSA blocking. In these graphs circles
represent a standard enzyme loading (15FPU/g glucan), and triangles represent a low
enzyme loading (3FPU/g glucan). All hydrolysis runs were supplemented with 10U b-
glucosidase. Solid data points represent hydrolysis conducted without BSA, and open
data points represent hydrolysis conducted after BSA blocking. Error bars represent
one standard deviation.
COSLIF-pretreated (A) and SAA-pretreated (B) switchgrass at stan-
COSLIF-treated (E and F) switchgrass. Magnification of samples A, C, and E is
approximately 350?, while magnification was increased to approximately 3,000?
for B, D, and F. In the untreated material (A), vascular bundles and parenchyma are
highlighted with the upper and lower arrows, respectively.
SEM micrographs of untreated (A and B), SAA-treated (C and D) and
Rollin et al.: COSLIF Versus SAA
Biotechnology and Bioengineering
evenly distributed, with a much higher surface area to
volume ratio than the untreated case.
X-Ray Diffraction (XRD) Crystallinity Comparison
The XRD spectrum of untreated switchgrass reveals three
peaks, corresponding to (101), (002), and (040) lattice
planes (Fig. 5). SAA-pretreated materials retained the same
three peaks as untreated switchgrass, while COSLIF-
pretreated switchgrass effectively eliminated (101) and
(040) peaks, while broadening and greatly reducing the
(002) peak. Several procedures are available for calculating
the crystallinity index, CrI, from XRD spectra, and these
methods can give significantly differing, sometimes con-
flicting results (Thygesen et al., 2005). Two commonly used
methods were employed. The Segal method, which
compares peak height (Segal et al., 1959), is the most
than other methods (Park et al., 2009). In the peak
deconvolution method, the XRD spectrum is first compu-
tationally separated into component peaks. CrI is then
determined by taking the ratio of crystalline peak area to
total area (Park et al., 2010). Qualitatively from the XRD
diffractograms, SAA had a much less drastic effect on
whether biomass crystallinity increased or decreased, due to
conflicting results provided by the two methods employed.
COSLIF resulted in a clear and drastic decrease in CrI, as
expected due to the complete dissolution of biomass during
Despite retaining a large lignin fraction, COSLIF-treated
substrate was found to have greatly increased cellulose
accessibility, resulting in rapid hydrolysis rates. In contrast,
SAA removed large amounts of lignin, while causing a mild
increase in cellulose accessibility and slower hydrolysis rates.
Addition of BSA prior to hydrolysis caused a large increase
in thefinal digestibility of SAA-treatedsubstrate,butdidnot
impact COSLIF-treated material.
Based on the larger lignin content and increased NCAC
in digestibility following pre-hydrolysis BSA blocking might
be expected, but such an increase was not observed. On the
contrary, SAA-treated material exhibited strong negative
effects due to lignin, despite drastically reduced lignin content
accessibility and higher lignin inhibition. One explanation
may be that ammonia or the high level of delignification
delignification beyond about 50% might result in cellulose
pore collapse, causing a decrease in cellulose accessibility.
Another explanation may be different lignin properties after
different pretreatments. Kumar and Wyman (2009) showed
that lignin from acidic pretreatments had a significantly lower
represented in green and BSA proteins in blue. The TGC protein is similar in size to
T. reesei EG1 (A). To determine total substrate accessibility to cellulase (TSAC), TGC
equilibration is conducted without BSA (B). When BSA blocking is used prior to TGC
equilibration, cellulose accessibility to cellulase (CAC) may be determined (C).
Cellulose surfaces (110 face) susceptible to cellulase binding are highlighted in red.
Thioredoxin–GFP–CBM fusion protein schematic, with TGC proteins
TGC adsorption-based substrate accessibility to cellulase.
samples. The major peak seen for all three samples is the (002) peak, used for peak
and SAA samples.
XRD diffraction spectra for untreated, COSLIF, and SAA switchgrass
Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2010
inhibitory effect than lignin resulting from basic pretreat-
ments. The basic conditions used for higher lignin removal
caused a change of lignin chemistry, resulting in a lignin
surface that was more prone to adsorb protein. In addition,
biomass dissolution achieved by COSLIF may result in
lignin clustering, possibly in a manner similar to its
occurrence in dilute acid pretreatment (Donohoe et al.,
2008), which would further limit any inhibitory effect
observed, despite the relatively large lignin content. This
speculation is supported by the fact that while COSLIF
caused a CAC increase of 16-fold, the NCAC increase was
only 2-fold, much lower than expected based on composi-
tional ratio alone.
In addition to the COSLIF-SAA comparison presented
here, an empirical correlation between lignin removal and
cellulose accessibility for 72h enzymatic glucan digestibility
was developed by combining quantitative accessibility and
delignification data from this study, various other COSLIF
studies (Sathitsuksanoh et al., 2009; Zhu et al., 2009a), and
several publications of the Consortium for Applied
Fundamentals and Innovation (CAFI) (Kumar et al.,
2009; Kumar and Wyman, 2009; Wyman et al., 2009). In
order to normalize CAC values for different accessibility
assay conditions (temperature, protein detection method,
etc.), the CAC reading from Avicel was used as a reference
for this study and CAFI’s data. The CAFI-reported values
based on cellulase adsorption were adjusted to the TGC-
based CAC values by a factor of the cellulase adsorption on
the pretreated biomass samples relative to that on Avicel
(Kumar and Wyman, 2009). TableCurve 3D was used to
determine a best fit for these data points (R2¼0.99),
resulting in a surface representation of a variety of
pretreatments, presenting glucan digestibility as a function
of delignification and CAC (Fig. 6).
The substrates used to create this surface include
ammonia recycle percolation (ARP) hybrid poplar, dilute
acid (DA) hybrid poplar, lime hybrid poplar, untreated
Avicel, regenerated amorphous cellulose (RAC), COSLIF
corn stover, DA corn stover, untreated corn stover, COSLIF
COSLIF bamboo, untreated bamboo, COSLIF common
reed, and untreated common reed. While it was impossible
to logically relate hemicellulose removal, CAC, and glucan
digestibility with this data set (figure not shown), the effect
of lignin removal and CAC on digestibility displayed a clear
trend. Analyzing the effect of lignin removal at low CAC,
Figure 6 shows an increase from ?10% to 20% digestibility
(untreated substrates) to ?60% digestibility (SAA switch-
grass, Avicel). When CAC is increased, however, even when
much greater glucan digestibility is enabled, up to 97%
(COSLIF corn stover). This correlation suggests that
increasing the accessibility of the substrate is critical for
achieving high lignocellulose digestibility. Lignin removal is
important for increasing the digestibility of low-CAC
materials, but once high CAC is achieved, removing lignin
is a much less impactful pretreatment objective.
In conclusion, increasing cellulose accessibility was a
more important pretreatment consideration than deligni-
fication for effectively releasing sugars from recalcitrant
lignocellulose at high yield. High levels of delignification
without a significant increase in CAC did not result in a
correspondingly large increase in glucan digestibility. The
COSLIF pretreatment was capable of greatly increasing the
cellulose accessibility of switchgrass, resulting in rapid
digestibility and high final sugar yield, even at low enzyme
loadings. Further pretreatment development efforts are
recommended to focus on increasing substrate accessibility,
the most important factor in enzymatic biomass sacchar-
ification, using low-cost processes.
This work was supported mainly by the DOE BioEnergy Science
Center(BESC).TheBioEnergyScienceCenter is a U.S.Departmentof
Energy Bioenergy Research Center supported by the Office of Biolo-
gical and Environmental Research in the DOE Office of Science. This
work is also partially by the USDA Bioprocessing and Biodesign
Center and DuPont Young Professor Award. Noppadon Sathitsuksa-
noh was partially supported by the ICTAS scholar program.
Arantes V, Saddler JN. 2010. Access to cellulose limits the efficiency of
enzymatic hydrolysis: The role of amorphogenesis. Biotechnol Biofuels
Berlin A, Gilkes N, Kurabi A, Bura R, Tu M, Kilburn D, Saddler J. 2005.
Weak lignin-binding enzymes: A novel approach to improve activity of
cellulases for hydrolysis of lignocellulosics. Appl Biochem Biotechnol
obtained from this study andprevious studies in theliterature (Hong et al., 2007; Kumar
and Wyman, 2009; Kumar et al., 2009; Sathitsuksanoh et al., 2009, 2010; Wyman et al.,
2009; Zhu et al., 2009a). Here a correlation between delignification, CAC, and 72h
enzymatic digestibility is suggested, for a broad range of feedstocks and pretreat-
Digestibility as a function of delignification and CAC. Data points were
Rollin et al.: COSLIF Versus SAA
Biotechnology and Bioengineering
Chang VS, Holtzapple MT. 2000. Fundamental factors affecting biomass
enzymatic reactivity. Appl Biochem Biotechnol 84/86:5–37.
Chen F, Dixon RA. 2007. Lignin modification improves fermentable sugar
yields for biofuel production. Nat Biotechnol 25(7):759–761.
Converse AO. 1993. Substrate factors limiting enzymatic hydrolysis. In:
SaddlerJN, editor. Bioconversion of forest and agricultural plant
residues. Wallingford, Oxfordshire, UK: CAB International. pp. 93–
Donohoe BS, Decker SR, Tucker MP, Himmel ME, Vinzant TB. 2008.
Visualizing lignin coalescence and migration through maize cell walls
following thermochemical pretreatment. Biotechnol Bioeng 101:913–
Fan LT, Lee Y-H, Beardmore DR. 1981. The influence of major structural
features of cellulose on rate of enzymatic hydrolysis. Biotechnol Bioeng
Grethlein HE. 1985. The effect of pore size distribution on the rate of
enzymatic hydrolysis of cellulose substrates. Bio/Technol 3:155–160.
Gupta R, Kim TH, Lee YY. 2008. Substrate dependency and effect of
biomass. Appl Biochem Biotechnol 148:59–70.
Heinzelman P, Snow CD, Wu I, Nguyen C, Villalobos A, Govindarajan S,
Minshull J, Arnold FH. 2009. A family of thermostable fungal cellulases
created by structure-guided recombination. Proc Nat Acad Sci
Himmel ME, Ding S-Y, Johnson DK, Adney WS, Nimlos MR, Brady JW,
for biofuels production. Science 315(5813):804–807.
Hong J, Ye X, Zhang Y-HP. 2007. Quantitative determination of cellulose
accessibility to cellulase based on adsorption of a nonhydrolytic fusion
protein containing CBM and GFP with its applications. Langmuir
Hong J, Ye X, Wang Y, Zhang Y-HP. 2008. Bioseparation of recombinant
cellulose binding module-protein by affinity adsorption on an ultra-
high-capacity cellulosic adsorbent. Anal Chem Acta 621:193–199.
IshizawaCI, JeohT, Adney WS,HimmelME,Johnson DK,DavisMF.2009.
Can delignification decrease cellulose digestibility in acid pretreated
corn stover? Cellulose 16:677–686.
Kim TH, Lee YY. 2005. Pretreatment and fractionation of corn stover by
ammonia recycle percolation process. Bioresour Technol 96:2007–2013.
Kumar R, Wyman CE. 2009. Access of cellulase to cellulose and lignin for
poplar solids produced by leading pretreatment technologies. Biotech-
nol Prog 25(3):807–819.
Kumar R, Mago G, Balan V, Wyman CE. 2009. Physical and chemical
characterizations of corn stover and poplar solids resulting from
leading pretreatment technologies. Bioresour Technol 100(17):3948–
Lee D, Yu AHC, Wong KKY, Saddler JN. 1994. Evaluation of the enzymatic
susceptibility of cellulosic substrates using specific hydrolysis rates and
enzyme adsorption. Appl Biochem Biotechnol 45/46:407–415.
Liu C, Wyman CE. 2003. The effect of flow rate of compressed hot water on
xylan, lignin, and total mass removal from corn stover. Ind Eng Chem
Liu W, Hong J, Bevan DR, Zhang Y-HP. 2009. Fast identification of thermo-
stable beta-glucosidase mutants on cellobiose by a novel combinatorial
selection/screening approach. Biotechnol Bioeng 103:1087–1094.
Lu Y, Yang B, Gregg D, Saddler JN, Mansfield SD. 2002. Cellulase adsorp-
tion and an evaluation of enzyme recycle during hydrolysis of steam-
exploded softwood residues. Appl Biochem Biotechnol 98/100:641–
Lynd LR, Laser MS, Bransby D, Dale BE, Davison B, Hamilton R, Himmel
M, Keller M, McMillan JD, Sheehan J, Wyman CE. 2008. How biotech
can transform biofuels. Nat Biotechnol 26(2):169–172.
Lynd LR, Larson E, Greene N, Laser M, Sheehan J, Dale BE, McLaughlin S,
Wang M. 2009. The role of biomass in America’s energy future:
Framing the analysis. Biofuels Bioprod Biorefining 3(2):113–123.
MosierN, WymanCE,Dale BE,ElanderRT, Lee YY,HoltzappleM, Ladisch
M. 2005. Features of promising technologies for pretreatment of
lignocellulosic biomass. Bioresour Technol 96:673–686.
Moxley G, Zhang Y-HP. 2007. More accurate determination of acid-labile
carbohydrate composition in lignocellulose by modified quantitative
saccharification. Energy Fuels 21:3684–3688.
Moxley GM, Zhu Z, Zhang Y-HP. 2008. Efficient sugar release by
the cellulose solvent based lignocellulose fractionation technology
and enzymatic cellulose hydrolysis J. Agric Food Chem 56(17):7885–
crystallinity index of cellulose by solid state C-13 nuclear magnetic
resonance. Cellulose 16(4):641–647.
Park S, Baker JO, Himmel ME, Parilla PA, Johnson DK. 2010. Cellulose
crystallinity index: Measurement techniques and their impact on
interpreting cellulase performance. Biotechnol Biofuels 3:10.
feedstock for a bioenergy and bioproducts industries: The technical
feasibility of a billion-ton annual supply. Oak Ridge National Labora-
tory. Oak Ridge, TN.
Sathitsuksanoh N, Zhu Z, Templeton N, Rollin J, Harvey S, Zhang Y-HP.
2009. Saccharification of a potential bioenergy crop, Phragmites aus-
tralis (common reed), by lignocellulose fractionation followed by
enzymatic hydrolysis at decreased cellulase loadings. Ind Eng Chem
Sathitsuksanoh N, Zhu Z, Ho T-J, Bai M-D, Zhang Y-HP. 2010. Bamboo
saccharification through cellulose solvent-based biomass pretreatment
followed by enzymatic hydrolysis at ultra-low cellulase loadings Biores.
Segal L, Creely J, Martin A, Conrad C. 1959. An empirical method for
estimating the degree of crystallinity of native cellulose using the X-ray
diffractometer. Text Res J 29(10):786.
Sluiter A, Hames R, Ruiz R, Scarlata C, Sluiter J, Templeton D, Crocker D.
2008. Determination of structural carbohydrates and lignin in bio-
Sousa LD, Chundawat SPS, Balan V, Dale BE. 2009. ‘Cradle-to-grave’
assessment of existing lignocellulose pretreatment technologies. Curr
Opin Biotechnol 20:339–347.
Thygesen A, Oddershede J, Lilholt H, Thomsen AB, Stahl K. 2005. On the
determination of crystallinity and cellulose content in plant fibres.
Tu M, Chandra RP, Saddler JN. 2007. Evaluating the distribution of
cellulases and the recycling of free cellulases during the hydrolysis of
lignocellulosic substrates. Biotechnol Prog 23:398–406.
Wyman CE. 2007. What is (and is not) vital to advancing cellulosic ethanol.
Trends Biotechnol 25(4):153–157.
Wyman CE, Dale BE, Elander RT, Holtzapple M, Ladisch MR, Lee YY,
Mitchinson C, Saddler JN. 2009. Comparative sugar recovery and
fermentation data following pretreatment of poplar wood by leading
technologies. Biotechnol Prog 25:333–339.
Yang B, Boussaid A, Mansfield SD, Gregg DJ, Saddler JN. 2002. Fast and
efficient alkaline peroxide treatment to enhance the enzymatic digest-
ibility of steam-exploded softwood substrates. Biotechnol Bioeng
Zhang Y-HP. 2008. Reviving the carbohydrate economy via multi-product
biorefineries. J Ind Microbiol Biotechnol 35(5):367–375.
Zhang Y-HP, Lynd LR. 2004. Toward an aggregated understanding of
enzymatic hydrolysis of cellulose: Noncomplexed cellulase systems.
Biotechnol Bioeng 88:797–824.
Zhang Y-HP, Lynd LR. 2006. A functionally-based model for hydrolysis of
cellulose by fungal cellulase. Biotechnol Bioeng 94:888–898.
Zhang Y-HP, Himmel M, Mielenz JR. 2006. Outlook for cellulase improve-
ment: Screening and selection strategies. Biotechnol Adv 24(5):452–
Zhang Y-HP, Ding S-Y, Mielenz JR, Elander R, Laser M, Himmel M,
McMillan JD, Lynd LR. 2007a. Fractionating recalcitrant ligno-
cellulose atmodest reaction
Zhang Y-HP, Schell DJ, McMillan JD. 2007b. Methodological analysis for
determination of enzymatic digestibility of cellulosic materials. Bio-
technol Bioeng 96(1):188–194.
Biotechnology and Bioengineering, Vol. xxx, No. xxx, 2010
features affecting biomass enzymatic digestibility. Bioresour Technol
Zhu Z, Sathitsuksanoh N, Vinzant T, Schell DJ, McMillan JD, Zhang Y-HP.
2009a. Comparative study of corn stover pretreated by dilute acid and
cellulose solvent-based lignocellulose fractionation: Enzymatic hydro-
lysis, supramolecular structure, and substrate accessibility. Biotechnol
Zhu Z, Sathitsuksanoh N, Zhang Y-HP. 2009b. Direct quantitative deter-
mination of adsorbed cellulase on lignocellulosic biomass with its
Rollin et al.: COSLIF Versus SAA
Biotechnology and Bioengineering