Bioresource Technology 101 (2010) 9272–9278

Contents lists available at ScienceDirect

Bioresource Technology
journal homepage: www.elsevier.com/locate/biortech

Improving the thermostability of Geobacillus stearothermophilus xylanase

XT6 by directed evolution and site-directed mutagenesis
Zhi-Gang Zhang a,b, Zhuo-Lin Yi a,b, Xiao-Qiong Pei a, Zhong-Liu Wu a,*
a
Chengdu Institute of Biology, Chinese Academy of Sciences, P.O. Box 416, Chengdu 610041, China
b
Graduate University of the Chinese Academy of Sciences, Beijing 100049, China

a r t i c l e i n f o a b s t r a c t

Article history: Protein engineering of the thermostable xylanase XT6 from Geobacillus stearothermophilus was performed
Received 12 April 2010 to obtain enzymes with improved thermal tolerance. Mutants producing such enzymes were obtained
Received in revised form 13 July 2010 after several rounds of directed evolution using error-prone PCR and sequence family shuffling, in com-
Accepted 14 July 2010
bination with a consensus-based semi-rational approach. The most thermostable mutant enzyme con-
Available online 17 July 2010
tained 13 amino acid substitutions and its half-life of inactivation was 52-fold of that of the wild-type.
Its reaction temperature for maximum activity increased from 77 °C to 87 °C, and catalytic efficiency
Keywords:
(kcat/Km) increased by 90%. The mutant is of potential interest for industrial applications.
Xylanase
Protein engineering
Ó 2010 Elsevier Ltd. All rights reserved.
Directed evolution
Thermostability
Semi-rational

1. Introduction have to be used (Tracewell and Arnold, 2009; Wintrode et al.,

Xylanases (endo-1,4-b-xylanase, EC 3.2.1.8) are glycosidases
that catalyze the hydrolysis of xylan, a major constituent of
hemicellulose complex, which, together with cellulose and lig-
nin, make up the majority of plant cell walls. Xylanases have
been successfully used for animal food manufacturing and pulp
bleaching, and are considered to be a key player in the biodeg-
radation of renewable resources to useful end products (Beg
et al., 2001; Kulkarni et al., 1999). In most of these processes,
a thermostable xylanase would be beneficial since the enzyme
load could be decreased and the possibility of microbial contam-
ination could be reduced (Collins et al., 2005; Heinzelman et al.,
2009; Viikari et al., 2007). Since the availability of naturally
thermostable xylanases is limited, improvements in the thermal
tolerance of xylanases has been attempted by mutational meth-
ods using rational approaches based on three-dimensional struc-
tures (Belien et al., 2009; Fenel et al., 2006; Georis et al., 2000).
However, most of the functional changes are beyond current
predictive abilities, especially when insoluble substrates are in-
volved (Bornscheuer and Pohl, 2001; Leisola and Turunen,
2007) and therefore, approaches based on directed evolution
* Corresponding author. Address: Chengdu Institute of Biology, Chinese Academy
of Sciences, 9 South Renmin Road, 4th Section, Chengdu, Sichuan 610041, PR China.
Tel./fax: +86 28 85238385.
E-mail address: wuzhl@cib.ac.cn (Z.-L. Wu).
2001).
The majority of studies on the improvement of thermo-toler-
ance have involved wild-type xylanases with relatively low opti-
mal temperatures of 50–60 °C, and most of them belong to
family-11 (http://www.cazy.org/) (Andrews et al., 2004; Miya-
zaki et al., 2006; Palackal et al., 2004; Xie et al., 2006). So far,
the generation of thermostable xylanases has only been at-
tempted by mutation of XynA from Thermomyces lanuginosus
(Stephens et al., 2007, 2009), and the best of the mutants ob-
tained had a half-life of inactivation at 70 °C 2.4-fold of that of
the wild-type.
In the current study, we employed directed evolution in com-
bination with a consensus-based semi-rational approach to en-
hance the thermal tolerance of the thermostable xylanase XT6
from the thermophilic bacterium Geobacillus stearothermophilus.
This family-10 xylanases retains high activity over a broad range
of pH, and is most active in the neutral range. The native XT6 dis-
plays the highest initial reaction rate at 75 °C and has been suc-
cessfully used in a large-scale biobleaching mill trial (Khasin
et al., 1993; Lapidot et al., 1996). After several rounds of screen-
ings for both activity and stability, a series of mutants with en-
hanced thermal tolerance was obtained without compromising
activity. Combinations of amino acid substitutions resulted in
the most stable mutant enzyme (Fig. 1) with a 52-fold increase
in half-life of inactivation at 75 °C and a catalytic efficiency in-
crease of 90%.

0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.biortech.2010.07.060
 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278
2. Methods
2.1. Materials
Oligonucleotides were purchased from Invitrogen Life Technol-
ogies (Shanghai, China) in PAGE-purified grade. Birchwood xylan
was purchased from Sigma–Aldrich (St. Louis, USA). Taq polymer-
ase and xylose were purchased from the Takara Biotechnology (Da-
lian, China). Restriction enzymes were from New England Biolabs
(Beverly, MA). All other reagents were obtained from general com-
mercial suppliers and used without further purification.
2.2. Construction of error-prone PCR libraries
The Genemorph II Random Mutagenesis Kit (Stratagene, La Jolla,
CA) was employed to construct error-prone PCR (epPCR) libraries
of XT6 with the following primers: 50 -TAGGAGGTCATATGAA-
AAATGCGGACAGCTATGCG-30 (forward) and 50 -ATACGCGGATC-
CCTATTTGTGATCAATGATCGCCCAATACGCCGGCTT-30 (reverse).
For the first round of epPCR, plasmid pET-28a-XT6 encoding
wild-type XT6 was used as the template (Zhang et al., 2009). For
the second round of epPCR, equal amounts of three mutant plas-
mids harboring all the beneficial mutations from the first round
screening were pooled and used as templates. The amount of target
DNA was about 100 ng for a 50 ll reaction to achieve a low muta-
tion frequency. The reactions were performed in an MJ Mini™ 48-
Well Personal Thermal Cycler (Bio-Rad, Hercules, CA) following the
manufacturer’s protocols. The amplified PCR products were di-
gested with Nde I and BamH I and ligated into pET-28a(+) (Nova-
gen, Madison, WI). The products were introduced into E. coli
DH10B (Invitrogen, Carlsbad, CA) using Electroporator 2510 (Bio-
Rad Laboratories, Veenendaal, The Netherlands) and spread on Lur-
ia–Bertani (LB) plates containing 50 lg kanamycin/ml. After over-
night incubation at 37 °C, colonies were collected and transferred
into fresh LB-kanamycin liquid medium and allowed to grow for
4 h at 37 °C with shaking at 240 rpm.
Plasmids were extracted, purified using QIAprep Spin Miniprep
Kit (Qiagen, Germany), transformed into E. coli BL21 (DE3) (Nova-
gen, Madison, WI) and plated onto LB-kanamycin plates for
screening.
2.3. Construction of DNA shuffling library
DNA shuffling was carried out as previously described (Stem-
mer, 1994). Briefly, the XT6 xylanase coding sequences from ran-
dom libraries were amplified with primers 50 -GTGAGCGGATAA-
CAATTCCC-30 and 50 -CCTCAAGACCCGTTTAGAGG-30 using Pfu poly-
merase (Fermentas, Maryland). Equal concentrations of each PCR
product were pooled and purified. After Dnase I treatment, frag-
ments in the range of 50–250 bp were purified from 2% agarose
electrophoresis gel using QIAEX II Gel Extraction kit (Qiagen, Ger-
many) and reassembled as described by Abecassis et al. (2000).
Full-length reassembly products were amplified with sense primer
50 -TAGGAGGTCATATGAAAAAT GCGGACAGCTATGCG-30 and anti-
sense primer 50 -ATACGCGGATCCCTATTTGTGATCAATGATCGCCC-
AATACGCCGGCTT-30 , cloned into pET-28a(+), and electro-trans-
formed into E. coli DH10B. Subsequent procedures were the same
as those used for the construction of error-prone libraries.
2.4. Design and construction of site-directed mutants
A BLAST search was performed using the protein sequence of
XT6. Seventeen homologous sequences of family-10 xylanases
including ten from thermophilic bacteria were selected. Their
protein sizes varied significantly and were not considered as an
9273
important criteria for alignment purpose. The xylanase sequences
shared 41–75% sequence similarities with XT6 (see Electronic An-
nex, Table S1 and Fig. S1), and were thus suitable for generating the
consensus sequence using ClustalX program (Thompson et al.,
1997). Point mutations were proposed based on the comparison
of the XT6 sequence with the consensus sequence (see Electronic
Annex, Fig. S2).
Site-directed mutagenesis was performed following the instruc-
tions of QuickChangeÒ site-directed mutagenesis protocol (Strata-
gene, La Jolla, CA). The sequences of oligonucleotides containing
the appropriate base changes are listed in Table 1. A total of
125 ng of each pair of complementary primers with 10 ng of re-
combinant plasmid DNA as template were used for PCR amplifica-
tion. All mutants with single substitutions were constructed with
plasmids encoding wild-type XT6 as the template. Mutants FAM,
FAMG, FTAMG, FC06T were constructed with plasmids encoding
05D03, FAM, FAMG and 06H07 as the template, respectively. The
mutagenic PCRs were performed under the following conditions:
95 °C for 30 s, 12 cycles of 95 °C for 30 s, 55 °C for 60 s and 68 °C
for 7 min. After treatment with 20 U of Dpn I at 37 °C for 2 h, the
reaction mixtures were transformed into DH5a competent cells.
The successful introductions of the desired mutations were con-
firmed by sequencing at Shanghai Invitrogen Life Technologies.
2.5. Screening for thermal stable variants
Single colonies from the mutant library were picked with sterile
toothpicks and cultured in 200 ll of Terrific Broth (TB) containing
kanamycin (50 lg/ml) and 1 mM IPTG in sterile 96-well micro-
plates and incubated overnight at 37 °C with shaking at 240 rpm
for 24 h on an INFORS Multitron incubator shaker (INFORS HT,
Swissland). A replicate plate was prepared by transfer of half of
the content of each well into a fresh microplate. Both plates were
Table 1
Sequences of oligonucleotides used for site-directed mutagenesis.
Target sites Oligonucleotide sequencesa
N16K 50 - GCATATTAGCGCCCTGAAAGCGCCACAGCTGG-30
50 -CCAGCTGTGGCGCTTTCAGGGCGCTAATATGC-30
F52Y 50 -GATGCTGAAGCGTCATTATAACTCAATTGTGGCGG-30
50 -CCGCCACAATTGAGTTATAATGACGCTTCAGCATC-30
T120I 50 -CCGATGGTGAACGAGATCGATCCGGTGAAACGTG-30
50 -CACGTTTCACCGGATCGATCTCGTTCACCATCGG-30
G363D 50 -GCGCCGTTTGTGTTTGACCCGGATTACAAAGTGAAGC-30
50 -GCTTCACTTTGTAATCCGGGTCAAACACAAACGGCGC-30
K25I 50 -CAGCTGGACCAACGGTACATAAATGAATTTACTATTGGTGCG-30
50 -CGCACCAATAGTAAATTCATTTATGTACCGTTGGTCCAGCTG-30
E112D 50 -GGTTCTTCCTTGACAAAGATGGAAAACCGATGGTG-30
50 -CACCATCGGTTTTCCATCTTTGTCAAGGAAGAACC-30
E137D 50 -CTGCTGAAACGCTTGGACACCCACATCAAAACC-30
50 -GGTTTTGATGTGGGTGTCCAAGCGTTTCAGCAG-30
F94G 50 -GGCATGGATATTCGTGGTCACACCCTGGTGTGG-30
50 -CCACACCAGGGTGTGACCACGAATATCCATGCC-30
I151V 50 -GAACGTTATAAGGATGATGTAAAGTACTGGGATGTAG-30
50 -CTACATCCCAGTACTTTACATCATCCTTATAACGTTC-30
Y199F 50 -GGCGATAACATTAAGTTATTCATGAATGACTACAATACCG-30
50 -CGGTATTGTAGTCATTCATGAATAACTTAATGTTATCGCC-30
N261I 50 -GCACTGGGCCTGGATATCCAAATTACCGAACTGG-30
50 -CCAGTTCGGTAATTTGGATATCCAGGCCCAGTGC-30
F360L 50 - GCAAAGATGCGCCGCTTGTGTTTGGCCCGG-30
50 -CCGGGCCAAACACAAGCGGCGCATCTTTGC-30
I375V 50 -CCGGCGTATTGGGCGGTCATTGATCACAAATAG-30
50 -CTATTTGTGATCAATGACCGCCCAATACGCCGG-30
a
The nucleotide changes are underlined.
9274 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278
subjected to centrifugation at 1800g for 10 min. The supernatant of
each well was removed and the cell pellet was resuspended in
resuspended in 30 ll of 50 mM sodium phosphate buffer (pH
7.6). Heat challenge was applied to one microplate by pre-incuba-
tion at 85 °C for 2 h for the first round of epPCR library, at 88 °C for
2 h for the second round of epPCR library, and at 88 °C for 4 h for
the DNA shuffling library. Enzymatic assays in both plates were
then carried out at 37 °C for 2 h after the addition of 50 ll of 1%
(w/v) of xylan prepared in 50 mM sodium phosphate buffer. The
incubation temperature of 37 °C was chosen based on previous re-
ports on high-throughput screening of thermostable xylanases
(Miyazaki et al., 2006; Palackal et al., 2004). The incubation period
was optimized to two hours base on the detection sensitivity of the
miniature DNS method and the absorbance range of the microplate
reader.
The reactions were quenched by the addition of 120 ll of DNS
reagent (Bailey et al., 1992) and heated at 105 °C for 20 min to
determine the amount of reducing sugar. After centrifugation at
1800g for 10 min, the supernatant was transferred into a new
96-well microplate and absorbance was measured at 540 nm in a
Thermo Scientific Varioskan Flash microplate reader (Thermo
Scientific, USA).
2.6. Expression and purification of wild-type and mutant enzymes
Single colonies were grown in TB medium containing 50 lg
kanamycin/ml and induced by the addition of 1 mM IPTG when
the absorbance at 600 nm reached 0.7, and the incubation was con-
tinued for another 15 h. Cell pellets were collected by centrifuga-
tion at 2  104 g for 20 min, resuspended in 50 mM sodium
phosphate buffer (pH 7.6) and lysed by sonicating 10 times for
30 s with 30 s between each burst using a JY92-II Ultra Sonic Cell
Crusher (Scientz Biotechnology Co., China). Cell extracts were trea-
ted at 60 °C for 30 min and centrifuged at 2  104 g for 30 min be-
fore loading onto nickel-nitrilotriacetic acid columns (Bio-Rad).
Purified XT6 and variants were analyzed by SDS–PAGE and used
for enzymatic assay. Protein estimations were done with a com-
mercial BCA Protein Assay kit with bovine serum albumin as a
standard (Beyotime Beijing, China).
2.7. Measurement of xylanase activity
Purified enzymes were used throughout unless otherwise men-
tioned. All measurements were performed in triplicate. Enzymatic
activity was determined using birchwood xylan with the substrate
prepared as described previously (Dupont et al., 1998). Briefly, the
2-ml reaction mixture consisted of 0.2 ml of appropriately diluted
enzyme and 1.8 ml of 50 mM sodium phosphate buffer (pH 7.6)
containing 1% birchwood xylan. The mixture was incubated at
50 °C for 5 min, followed by immediate chilling on ice for 5 min.
The amount of reducing sugars released was determined by the
standard DNS method (Bailey et al., 1992). One unit of xylanase
activity was defined as the amount of enzyme producing 1 lmol
of reducing equivalents per min under the assay conditions.
Steady-state kinetic parameters were determined at a constant
enzymatic concentration of 0.4 lg/ml. Km and kcat values were
determined using substrate concentrations ranging from 0.2 to
10 mg xylan/ml in 50 mM sodium phosphate buffer (pH 7.6) at
50 °C, and estimated by fitting a hyperbolic Michaelis–Menten
equation using non-linear regression with Graphpad Prism soft-
ware (Graphpad, San Diego, CA).
The thermostabilities of XT6 and selected mutants were tested
by heating 0.1 mg/ml purified enzyme samples at 60–86 °C for
20 min in 50 mM sodium phosphate buffer (pH 7.6). The residual
activity was measured as described above. To measure irreversible
thermal inactivation, samples of crude enzyme were incubated at
75 °C for different time, placed on ice and assayed under standard
conditions. The thermal inactivation half-life (t1/2) of the enzyme
was determined by plotting the natural logarithms of remaining
activity values versus incubation time and calculated from linear
regression. A thermal cycler and thin-wall PCR tubes were used
for all thermostability measurements.
3. Results and discussion
3.1. Screening mutants with enhanced thermal tolerance from the first
round epPCR library
Randomly picked clones from the first round epPCR library
showed 1.7 base substitutions/kb and approximately 60% colonies
displayed xylanase activity. After heat treatment at 85 °C for 2 h,
wild-type XT6 was completely inactivated. A total of 12,000 mu-
tants were screened. The activities with or without heat challenge
were tested for each mutant to address the concern that improved
thermal tolerance might negatively impact catalytic activity. Four
mutants with improved thermostability without compromising
activity were selected (Table 2). All of them retained more than
40% activity after heat challenge on 96-well plate, while the
wild-type lost more than 90% activity under the same conditions.
Mutants 09D05 and 05D03 contained two amino acid substitu-
tions, while 10E11 and 08A07 contained a single substitution. All
of them displayed a 1.4–2.5-fold increase in their half-lives of ther-
mal inactivation at 75 °C. A total of five amino acid substitutions
from this round of screening, F52Y, T120I, A256V, M270I and
G363D, were found to be possibly beneficial for the enhanced ther-
mal tolerance of XT6 (Table 2).
3.2. Isolation of mutants from the second round epPCR library
The second round epPCR library was carried out with mutants
09D05, 05D03 and 10E11 as templates. Screening of 8000 colonies
yielded five colonies with significantly improved thermostability
compared with those from the first round of epPCR (Fig. 2). Several
amino acid substitutions from the first round of epPCR appeared to
be combined randomly, and the results revealed a synergistic or
addictive effect between these residues, which was consistent with
earlier reports (Palackal et al., 2004). In addition, four new amino
acid substitutions with possible beneficial effect were identified
(Fig. 2): E137D from mutant 03B11, N16K from mutant 02E04,
E112D from mutant 08G01 and K25I from mutant 04D03.
By constructing four single mutations N16K, K25I, E112D and
E137D, we found that every single mutation did not significantly
alter the enzymatic activity and thermostability (data not shown).
Their changes of t1/2 values were less than 10%. However, in the
case of multimutations, the addition of these substitutions resulted
in further increase in t1/2 values compared with the corresponding
mutants that only contained amino acid substitutions inherited
from the first round of epPCR such as 05F03 (A256V/M270I/
G363D, AMG), FAM (F52Y/A256V/M270I) and FAMG (F52Y/
A256V/M270I/G363D) (Fig. 2).
3.3. Identification of mutants via a consensus-based semi-rational
approach
To identify more beneficial amino acid substitutions, we applied
a consensus-based semi-rational design as a complementary ap-
proach. Based on the consensus concept that consensus amino
acids contribute more than average to the stability of the protein
than the non-consensus amino acids, the semi-rational approach
of sequence comparison of homologous proteins has been used
successful to identify beneficial amino acid substitutions
 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278 9275

Fig. 1. Lineage of thermostable mutants of Geobacillus stearothermophilus xylanase XT6. Amino acid substitutions accumulated through two rounds of epPCR, DNA shuffling
and recombination are shown. Newly introduced mutations in each generation are in bold and marked with an asterisk. Consensus-based site-directed mutagenesis approach
is boxed with dotted lines.

Table 2
Half-lives of thermal inactivation at 75 °C for mutants from the 1st round of epPCR
library.

XT6 Amino acid changes t1/2 (min) Fold

WT None 3.5 1
09D05 T120I/G363D 7.4 2.1
05D03 A256V/M270I 8.9 2.5
10E11 F52Y 8.5 2.4
08A07 M270I 5.0 1.4

(Lehmann and Wyss, 2001). In order to identify thermal stability-

related amino acids in xylanase XT6, a multiple sequence align-
ment with seventeen family-10 b-glycanases was carried out with
the ClustalX program (see Electronic Annex, Fig. S2).
After comparison of the XT6 sequence with the computer gen-
erated consensus sequence, five amino acid substitutions were se-
Fig. 2. Relative half-lives of thermal inactivation at 75 °C for WT and mutants from
lected for potential thermal stability effect. An additional the second round of epPCR. Amino acid substitutions identified from the 2nd round
substitution, I151V, was designed as well, based on previous con- of screening are listed below the names of the mutants. The letters F, A, M, G
sensus alignment of 18 family-10 xylanases (Gat et al., 1994). Sin- represent amino acid substitutions inherited from the first round of screening: F52Y
gle mutations were introduced by sited-directed mutagenesis and (F), A256V (A), M270I (M), and G363D (G). The relative t1/2s of mutants AMG, FAM
and FAMG are shown on top of the column in brackets for comparison.
their effects on enzymatic activity and thermostability were tested
with crude enzyme extracts. Compared with wild-type XT6, the
mutant enzymes showed no significant change of specific activity, tolerance. Its t1/2 value at 75 °C was 2.3 times of that of the wild-
and only one of them, mutant F360L, displayed enhanced thermal type (Table 3).
9276 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278

3.4. Combination of beneficial amino acid substitutions by DNA

shuffling and point mutations

DNA shuffling was first applied to combine amino acid substitu-

tions identified from the first and second round epPCR libraries.
After examining 4500 single colonies from the shuffling library,
one mutant, designated as 06H07, was identified with improved
thermostability. It retained 40% activity after heat treatment dur-
ing high-throughput screening in a microtiter plate, while the con-
trol mutant 04D03 had less than 20% activity on the same plate.
DNA sequencing revealed eight amino acid substitutions, five of
them inherited from the first round epPCR library (F52Y/T120I/
A256V/M270I/G363D), and the other three were newly generated
mutations (K72I/V218A/K379N). For comparison, the mutant with-
out three new mutations was constructed by site-directed muta-
genesis and designated as FTAMG. This mutant had a t1/2 value of
72.5 min (Fig. 3), demonstrated lower thermal stability when com-
pared with 06H07, which had a t1/2 value of 92.7 min at 75 °C, 26.5
times of that of the wild-type (Fig. 3).
Finally, five amino acid substitutions identified from the second
round of epPCR (N16K, E137D, K25I and E112D) and from the con- Fig. 3. Half-lives of thermal inactivation for wild-type XT6 (closed triangle) and
sensus-based approach (F360L) were introduced into mutant mutants FTAMG (closed square), 06H07 (open triangle) and FC06T (closed circle) at
75 °C, and FC06T at 80 °C (open circle with dotted line).
06H07 to achieve a possible addictive effect. The new XT6 mutant
enzyme was designated as FC06T. Its half-life of inactivation was
182.2 min at 75 °C (Fig. 3), 52 times of that of the wild-type. When
interactions as analyzed by both WHAT IF and PIC servers. Only
treated at 80 °C, the wild-type enzyme was inactivated immedi-
the E112D substitution could add putative hydrogen bonds with
ately within less than one minute, while mutant FC06T had a t1/2
OD1 of Asp110 within 1.31 Å and with N atom of Lys114 within
value of 39.6 min (Fig. 3).
2.84 Å. The analysis demonstrates the value of a random library
The FC06T gene sequence contains thirteen mutation sites
screening approach in identifying beneficial substitutions that
throughout the open reading frame as shown on a model generated
would be hard to predict with the currently limited knowledge
using the WHAT IF web interface (http://swift.cmbi.ru.nl/servers/
of thermostability-related structural requirements.
html/index.html) (Vriend, 1990) (Fig. 4). As it is well known that
hydrogen bonds, hydrophobic interactions, and electrostatic inter-
actions are the dominant structural factors responsible for protein 3.5. Characterization of wild-type XT6 and mutants
thermostability (Vieille et al., 1996; Vieira and Degreve, 2009; Vogt
et al., 1997), we analyzed these intramolecular interactions using Further enzymatic characterizations were performed for mu-
the same web interface in combination with the PIC server tants FC06T, FTAMG, and mutants identified from the first round
(http://crick.mbu.iisc.ernet.in/~PIC/) (Tina et al., 2007). The results of epPCR and the consensus-based approach, since they contained
showed that the F52Y substitution could add one hydrogen bond amino acid substitutions that had displayed more thermostabiliza-
with the OG1 atom of Thr314 within 3.08 Å, as well as an aro- tion effect than single mutations identified from subsequent
matic-aromatic interaction with Phe362 within 6.42 Å. The T120I screenings.
substitution could produce additional hydrophobic interactions The thermal stability of wild-type XT6 and mutants were deter-
with Pro122, and the M270I substitution could make the hydro- mined by measuring their residual activities after heat treatment
phobic interaction between Met270 and Phe289 be replaced by for 20 min at various temperatures. All the mutants with enhanced
that between Ile270 and Tyr278. The F360L substitution could half-lives of inactivation described above showed increased resid-
result in additional hydrophobic interaction with Phe362. The ual activities at the same temperature (Fig. 5). Mutant enzymes
G363D mutation could have additional hydrogen bonds with N FTAMG and FC06T retained more than 80% activity at 72 °C, while
atom of D365 within 2.90 Å and with N atom of K367 within wild-type XT6 lost 90% activity. The mutants 09D05, 05D03, 08A07
2.60 Å. In addition, this substitution could possibly generate new and 10E11, identified from the first round of epPCR, displayed inac-
salt-bridges with His11, Lys367 and Lys369, which would stabilize tivation patterns similar to those of mutants 08A07 and 09D05
the secondary structures and enhancing their rigidity (Vieille et al., presented in Fig. 5. They all lost about 50% activity under the same
1996). conditions. The half inactivation temperature (T1/2) as defined by
Substitutions identified from the 2nd round of epPCR and fam- the temperature at which the enzyme loses 50% activity was in-
ily shuffling libraries resulted in few changes of intramolecular creased from 70.1 °C for WT to 75.4 °C and 78.2 °C for FTAMG
and FC06T, respectively.
Table 3
The temperature dependence of enzymatic activities was mea-
Half-lives of thermal inactivation at 75 °C and specific activities
for mutants designed via consensus-based approach. sured in 5-min assays. The results revealed a marked change of
temperature/activity profile due to enhanced thermal stability
Enzyme Specific activity t1/2 (min)
(Fig. 6). The activity of wild-type XT6 reached a maximum at
(U/mg)
77 °C. Mutants 09D05 and 10E11, selected from the first round of
WT 420 4.5 epPCR, and FTAMG and FC06T, had maximum activities at 82 °C
F94G 384 3.2
I151V 512 3.8
and 87 °C, respectively (Fig. 6). The most thermostable mutant
Y199F 407 5.4 FC06T displayed a 44% activity increase at 87 °C compared with
N261I 448 3.3 that at 77 °C. When its pH dependence of activity was investigated,
F360L 498 10.5 this mutant showed a similar pH/activity profile as wild-type XT6
I375V 528 5.3
(data not shown).
 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278 9277

Fig. 4. Model of mutant FC06T of xylanase XT6. The elements of secondary structure and the position of thirteen mutations characterized in this study are displayed.
a-Helices are represented as cyan, b-strands as magenta, and loops as violet. The WHAT IF web interface was used to illustrate the thirteen amino acid substitutions of XT6
(PDB code: 1R85). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5. Thermostability of XT6 (closed triangle) and mutants at different temper-
atures. Shown are mutants 08A07 (open square), 09D05 (open triangle), FTAMG
(closed square) and FC06T (closed circle). Residual activities were measured after
heat treatment at various temperatures for 20 min.
tions over the range of 0.2–10.0 mg/ml (Table 4). All the enzymes
retained 100% activity at 50 °C, which would avoid the differences
in thermostability impacting kinetic parameters. Amino acid sub-
stitutions F52Y and F360L exerted little effect to the apparent Km
and catalytic turnover number (kcat). The M270I substitution, on
the contrary, resulted in a 62% decrease in Km value, indicating a
possible increase in the relative affinity of this mutant for the sub-
strate, which contributed to an over 200% increase in its catalytic
efficiency (kcat/Km). This residue is located within the active pocket,
close to the catalytic residue Glu265, and could possibly be in-
volved in substrate binding.
Mutant FTAMG and two double mutants T120I/G363D (09D05)
and A256 V/M270I (05D03) displayed 35–45% decrease in Km and
75–105% increase in kcat, leading to an approximately 200% in-
crease in catalytic efficiencies. The increased kcat values indicated
higher specific enzymatic activities than that of wild-type XT6 at
elevated substrate concentrations, which was also observed in
the measurements focusing on the thermostability of these en-
zymes. The mutant FC06T had a similar kcat value as the wild-type;
however, its catalytic efficiency was increased by 90% due to the
decrease of the Km. The results indicated that the mutants with
enhanced thermal tolerance, especially mutant FC06T, were
positively impacted in terms of their catalytic activities, which
was in accordance with the screening strategy.
The successful identification of potential thermo-stabilizing
amino acid substitutions without compromising activity demon-
strated the efficiency of direct screening for stable and highly
active mutants, which avoids the undesired activity loss that often
encountered in rational design (Wakarchuk et al., 1994).

Table 4
Steady-state kinetic parameters for wild-type XT6 and mutants.

Enzyme Km (mg/ml) kcat (s1) kcat/Km

Fig. 6. Temperature dependence of the activities of XT6 (closed triangle) and WT 0.95 ± 0.03 1238 ± 25 1303 ± 49
mutants. Shown are mutants 09D05 (open square), 10E11 (open triangle), FTAMG M270I 0.36 ± 0.04 1657 ± 30 4360 ± 466
(closed square) and FC06T (closed circle). Activities were normalized as percentages F52Y 0.99 ± 0.1 1585 ± 49 1601 ± 169
of the activities at 77 °C. F360L 0.97 ± 0.03 1272 ± 26 1311 ± 48
T120I/G363D 0.59 ± 0.05 2167 ± 42 3673 ± 319
A256V/M270I 0.62 ± 0.05 2557 ± 44 4124 ± 340
FTAMG 0.50 ± 0.05 2164 ± 43 4243 ± 441
The steady-state kinetic parameters for wild-type XT6 and mu- FC06T 0.55 ± 0.02 1356 ± 25 2465 ± 100
tants were determined at 50 °C with birchwood xylan concentra-
9278 Z.-G. Zhang et al. / Bioresource Technology 101 (2010) 9272–9278
4. Conclusions
Protein engineering of the thermostable xylanase XT6 from G.
stearothermophilus using directed evolution in combination with
a consensus-based approach revealed a series of mutants with en-
hanced thermal tolerance. The most thermostable mutant enzyme
had 13 amino acid substitutions and exhibited a t1/2 52-fold of that
of the wild-type without loss of activity. Mutants with improved
properties could be obtained by further manipulations at these
thirteen positions via saturation mutagenesis or site-directed
mutations.
Acknowledgements
This work was supported by the Program of 100 Distinguished
Young Scientists of the Chinese Academy of Sciences, the Province
Science Foundation of Sichuan, China (No. 08ZQ026-023 &
2010SZ0128) and the Knowledge Innovation Program of the Chi-
nese Academy of Sciences (No. KSCX1-YW-11B2).
Appendix A. Supplementary material
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.biortech.2010.07.060.
References
Abecassis, V., Pompon, D., Truan, G., 2000. High efficiency family shuffling based on
multi-step PCR and in vivo DNA recombination in yeast: statistical and
functional analysis of a combinatorial library between human cytochrome
P450 1A1 and 1A2. Nucl. Acids Res. 28, e88.
Andrews, S.R., Taylor, E.J., Pell, G., Vincent, F., Ducros, V.M.A., Davies, G.J., Lakey, J.H.,
Gilbert, H.J., 2004. The use of forced protein evolution to investigate and
improve stability of family 10 xylanases – the production of Ca2+-independent
stable xylanases. J. Biol. Chem. 279, 54369–54379.
Bailey, M.J., Biely, P., Poutanen, K., 1992. Interlaboratory testing of methods for
assay of xylanase activity. J. Biotechnol. 23, 257–270.
Beg, Q.K., Kapoor, M., Mahajan, L., Hoondal, G.S., 2001. Microbial xylanases and their
industrial applications: a review. Appl. Microbiol. Biotechnol. 56, 326–338.
Belien, T., Joye, I.J., Delcour, J.A., Courtin, C.M., 2009. Computational design-based
molecular engineering of the glycosyl hydrolase family 11 B. Subtilis XynA
endoxylanase improves its acid stability. Protein Eng. Des. Sel. 22, 587–596.
Bornscheuer, U.T., Pohl, M., 2001. Improved biocatalysts by directed evolution and
rational protein design. Curr. Opin. Chem. Biol. 5, 137–143.
Collins, T., Gerday, C., Feller, G., 2005. Xylanases, xylanase families and
extremophilic xylanases. FEMS Microbiol. Rev. 29, 3–23.
Dupont, C., Roberge, M., Shareck, F., Morosoli, R., Kluepfel, D., 1998. Substrate-
binding domains of glycanases from Streptomyces lividans: characterization of a
new family of xylan-binding domains. Biochem. J. 330, 41–45.
Fenel, F., Zitting, A.-J., Kantelinen, A., 2006. Increased alkali stability in Trichoderma
reesei endo-1, 4-[beta]-xylanase II by site directed mutagenesis. J. Biotechnol.
121, 102–107.
Gat, O., Lapidot, A., Alchanati, I., Regueros, C., Shoham, Y., 1994. Cloning and DNA
sequence of the gene coding for Bacillus stearothermophilus T-6 xylanase. Appl.
Environ. Microbiol. 60, 1889–1896.
Georis, J., de Lemos Esteves, F., Lamotte-Brasseur, J., Bougnet, V., Devreese, B.,
Giannotta, F., Granier, B., Frere, J.M., 2000. An additional aromatic interaction
improves the thermostability and thermophilicity of a mesophilic family 11
xylanase: structural basis and molecular study. Protein Sci. 9, 466–475.
Heinzelman, P., Snow, C.D., Wu, I., Nguyen, C., Villalobos, A., Govindarajan, S.,
Minshull, J., Arnold, F.H., 2009. A family of thermostable fungal cellulases
created by structure-guided recombination. Proc. Natl. Acad. Sci. USA 106,
5610–5615.
Khasin, A., Alchanati, I., Shoham, Y., 1993. Purification and characterization of a
thermostable xylanase from Bacillus stearothermophilus T-6. Appl. Environ.
Microbiol. 59, 1725–1730.
Kulkarni, N., Shendye, A., Rao, M., 1999. Molecular and biotechnological aspects of
xylanases. FEMS Microbiol. Rev. 23, 411–456.
Lapidot, A., Mechaly, A., Shoham, Y., 1996. Overexpression and single-step
purification of a thermostable xylanase from Bacillus stearothermophilus T-6. J.
Biotechnol. 51, 259–264.
Lehmann, M., Wyss, M., 2001. Engineering proteins for thermostability: the use of
sequence alignments versus rational design and directed evolution. Curr. Opin.
Biotechnol. 12, 371–375.
Leisola, M., Turunen, O., 2007. Protein engineering: opportunities and challenges.
Appl. Microbiol. Biotechnol. 75, 1225–1232.
Miyazaki, K., Takenouchi, M., Kondo, H., Noro, N., Suzuki, M., Tsuda, S., 2006.
Thermal stabilization of Bacillus subtilis family-11 xylanase by directed
evolution. J. Biol. Chem. 281, 10236–10242.
Palackal, N., Brennan, Y., Callen, W.N., Dupree, P., Frey, G., Goubet, F., Hazlewood,
G.P., Healey, S., Kang, Y.E., Kretz, K.A., Lee, E., Tan, X., Tomlinson, G.L., Verruto, J.,
Wong, V.W.K., Mathur, E.J., Short, J.M., Robertson, D.E., Steer, B.A., 2004. An
evolutionary route to xylanase process fitness. Protein Sci. 13, 494–503.
Stemmer, W.P.C., 1994. Rapid evolution of a protein in-vitro by DNA shuffling.
Nature 370, 389–391.
Stephens, D.E., Rumbold, K., Permaul, K., Prior, B.A., Singh, S., 2007. Directed
evolution of the thermostable xylanase from Thermomyces lanuginosus. J.
Biotechnol. 127, 348–354.
Stephens, D.E., Singh, S., Permaul, K., 2009. Error-prone PCR of a fungal xylanase for
improvement of its alkaline and thermal stability. FEMS Microbiol. Lett. 293,
42–47.
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F., Higgins, D.G., 1997. The
CLUSTAL_X windows interface. Flexible strategies for multiple sequence
alignment aided by quality analysis tools. Nucl. Acids Res. 25, 4876–4882.
Tina, K.G., Bhadra, R., Srinivasan, N., 2007. PIC: protein interactions calculator. Nucl.
Acids Res. 35, W473–W476.
Tracewell, C.A., Arnold, F.H., 2009. Directed enzyme evolution: climbing fitness
peaks one amino acid at a time. Curr. Opin. Chem. Biol. 13, 3–9.
Vieille, C., Burdette, D.S., Zeikus, J.G., 1996. Thermozymes. Biotechnol. Ann. Rev. 2,
1–83.
Vieira, D.S., Degreve, L., 2009. An insight into the thermostability of a pair of
xylanases: the role of hydrogen bonds. Mol. Phys. 107, 59–69.
Viikari, L., Alapuranen, M., Puranen, T., Vehmaanpera, J., Siika-Aho, M., 2007.
Thermostable enzymes in lignocellulose hydrolysis. Biofuels 108, 121–145.
Vogt, G., Woell, S., Argos, P., 1997. Protein thermal stability, hydrogen bonds, and
ion pairs. J. Mol. Biol. 269, 631–643.
Vriend, G., 1990. WHAT IF: a molecular modeling and drug design program. J. Mol.
Graphics 8, 52–56.
Wakarchuk, W.W., Campbell, R.L., Sung, W.L., Davoodi, J., Yaguchi, M., 1994.
Mutational and crystallographic analyses of the active-site residues of the
Bacillus circulans xylanase. Protein Sci. 3, 467–475.
Wintrode, P.L., Arnold, F.H., Frances, H.A., 2001. Temperature adaptation of
enzymes: lessons from laboratory evolution. Adv. Protein Chem. 55, 161–225.
Xie, H., Flint, J., Vardakou, M., Lakey, J.H., Lewis, R.J., Gilbert, H.J., Dumon, C., 2006.
Probing the structural basis for the difference in thermostability displayed by
family 10 xylanases. J. Mol. Biol. 360, 157–167.
Zhang, Z., Pei, X., Wu, Z., 2009. De novo synthesis and expression of a thermostable
xylanase from Geobacillus stearothermophilus in Escherichia coli. Chin. J. Appl.
Environ. Biol. 15, 271–275.