Role of Tryptophan Residues in a Class V Chitinase from Nicotiana tabacum

Naoyuki UMEMOTO,1;* Takayuki OHNUMA,1;* Henri URPILAINEN,1 Takanori YAMAMOTO,1

Tomoyuki NUMATA,2 and Tamo FUKAMIZO1;y

1Department of Advanced Bioscience, Kinki University, 3327-204 Nakamachi, Nara 631-8505, Japan2Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology (AIST),1-1-1 Higashi, Tsukuba 305-8566, Japan

Received November 29, 2011; Accepted January 4, 2012; Online Publication, April 7, 2012

[doi:10.1271/bbb.110914]

Tryptophan residues located in the substrate-bindingcleft of a class V chitinase from Nicotiana tabacum(NtChiV) were mutated to alanine and phenylalanine(W190F, W326F, W190F/W326F, W190A, W326A, andW190A/W326A), and the mutant enzymes were char-acterized to define the role of the tryptophans. Themutations of Trp326 lowered thermal stability by 5–7 �C, while the mutations of Trp190 lowered stabilityonly by 2–4 �C. The Trp326 mutations strongly im-paired enzymatic activity, while the effects of theTrp190 mutations were moderate. The experimentaldata were rationalized based on the crystal structure ofNtChiV in a complex with (GlcNAc)4, in which Trp190is exposed to the solvent and involved in face-to-facestacking interaction with the þ2 sugar, while Trp326 isburied inside but interacts with the �2 sugar throughhydrophobicity. HPLC analysis of anomers of theenzymatic products suggested that Trp190 specificallyrecognizes the �-anomer of the þ2 sugar. The strongeffects of the Trp326 mutations on activity and stabilitysuggest multiple roles of the residue in stabilizing theprotein structure, in sugar residue binding at subsite�2, and probably in maintaining catalytic efficiency byproviding a hydrophobic environment for proton donorGlu115.

Key words: Nicotiana tabacum; class V chitinase; tryp-tophan residues; chitin oligosaccharide;enzyme-substrate interaction

Chitinases (EC 3.2.1.14) hydrolyze �-1,4-glycosidiclinkages in chitin, an insoluble �-1,4-linked homo-polymer of N-acetylglucosamine (GlcNAc) residues.These enzymes belong to families GH-18 and GH-19according to the classification proposed in the CAZydatabase.1) Family GH-18 chitinases are widely dis-tributed in most living organisms, whereas family GH-19 enzymes are generally found only in plants and insome bacteria.2) Especially in plants, multiple forms ofGH-19 chitinases are constitutively or inducibly pro-duced and considered to act in defense against fungalpathogens.3) However, the physiological role of plantGH-18 chitinases remains unclear. Another classifica-tion system has been proposed for plant chitinases,

which are divided into at least five classes (I, II, III, IV,and V) based on sequence similarity.4,5) Classes I, II, andIV belong to family GH-19, while classes III and Vbelong to family GH-18. Among these enzymes, theexpression profile of class V chitinases is somewhatcomplicated. Their genes are responsive to both bioticand abiotic stress,6) whereas class III enzymes areexpressed in response only to abiotic stress.7) Thephysiological role of class V chitinases appears to besomewhat different from that of class III chitinases,although both classes belong to the same GH-family.In particular, class V chitinases appear to have multi-functional properties in plants. However, the structureand function of the class V enzymes are poorlyunderstood.Since the crystal structure of a class III enzyme from

the rubber tree, Hevea brasiliensis, was reported in1994,8) no crystal structure was reported for plant GH-18chitinases until 2010. Very recently, our group wassuccessful in determining the structure of plant class Vchitinases from Nicotiana tabacum (NtChiV)9) andArabidopsis thaliana.6) Class III and V chitinasespossess a similar fold ð�=�Þ8 barrel and the canonicalDxDxE catalytic motif. An insertion domain was foundin the class V enzymes, but not in the class III enzymes.The well-known family GH-18 chitinase B fromSerratia marcescens (SmChiB) possesses a chitin-bind-ing module in addition to an ð�=�Þ8 fold and an insertiondomain.10) The structural difference between these GH-18 chitinases from plants and the bacterium appears toreflect their functional difference. In fact, a tobacco classV chitinase hydrolyzes partially N-acetylated chitosannon-processively,9) whereas S. marcescens chitinase Bdoes so processively.11) The other striking differencebetween NtChiV and SmChiB is the location of thearomatic amino acid residues. Several aromatic residuesare aligned in the substrate-binding cleft of SmChiB,whereas only a few aromatic residues are found in thebinding cleft of NtChiV.9) In SmChiB, the individualaromatic amino acids act cooperatively toward the chitinpolysaccharide chain, resulting in processive hydrolysisfrom the non-reducing end, but the role of aromaticamino acids in NtChiV appears to be different from thatin SmChiB.

* These two authors contributed equally to this work.

y To whom correspondence should be addressed. Tel: +81-742-43-8237; Fax: +81-742-43-8976; E-mail: fukamizo@nara.kindai.ac.jp

Abbreviations: NtChiV, class V chitinase from Nicotiana tabacum; GlcNAc, 2-acetamido-2-deoxy-D-glucopyranose; (GlcNAc)n, �-1,4-linkedoligosaccharide of GlcNAc with a polymerization degree of n; PCR, polymerase chain reaction; SDS–PAGE, sodium dodecyl sulfate-polyacrylamidegel electrophoresis; HPLC, high performance liquid chromatography

Biosci. Biotechnol. Biochem., 76 (4), 778–784, 2012

The crystal structure of NtChiV in a complex with(GlcNAc)4 is shown in Fig. 1.9) (GlcNAc)4 binds tosubsites �2, þ1, þ2, and þ3, skipping subsite �1. Thebinding mode seems to be non-productive, and might beconverted to a productive binding mode by sliding of thethree sugar units (þ1, þ2, and þ3) to subsites �1, þ1,and þ2. NtChiV has two tryptophan residues (Trp190and Trp326) in the substrate binding cleft. Trp190 isrelatively exposed to the solvent, and appears to interactdirectly with the þ2 sugar. However, Trp326 is buriedinside, and appears to be involved in forming thehydrophobic core and in interaction with the �2 sugar.In the present study, the two tryptophan residues ofNtChiV were mutated to alanine and phenylalanine, andthe mutated enzymes (W190F, W326F, W190F/W326F,W190A, W326A, and W190A/W326A) were charac-terized to define the roles of the two tryptophan residues.

Materials and Methods

Materials. Chitin oligosaccharides, (GlcNAc)n (n ¼ 1{6), were

purchased from Seikagaku Kogyo (Tokyo). Glycol chitin was

synthesized by the method of Yamada and Imoto.12) E. coli Tuner

(DE3) pLacI cells and the expression vector pETBlue-1 were from

Novagen (Madison, WI). SP-Sepharose and Sephacryl S-100 HR were

from GE Healthcare (Tokyo). The TSK-GEL G2000PW and TSK-

GEL Amide 80 columns were from Tosoh (Tokyo). All other reagents

were of analytical grade.

Bacterial strain and plasmids. Escherichia coli JM109 was the host

strain used throughout the construction of plasmids carrying the

NtChiV genes with various mutations. E. coli Tuner (DE3) carrying

pLacI was used as the host strain in the production of the wild-type and

mutant chitinase proteins. Plasmid pETBlue-1 carrying wild-type

NtChiV was used for mutagenesis.

Site-directed mutagenesis. Site-directed mutagenesis was carried

out by polymerase chain reaction using a QuickChange site-directed

mutagenesis kit (Stratagene, La Jolla, CA) and pETBlue-1 carrying the

wild-type NtChiV as template. The primers used in production of the

NtChiV mutant genes were 50-CCTTATGGCCTATGATTTCTATGG-

ACCAAATTTCTCACCATCAC-30 (W190F), 50-GTAGAGGATTG-

CTTGGTTATTTTGCATTCCACGTTGCAGG-30 (W326F), 50-CTA-

TGGACCAAATGCCTCACCATCACAAACCAATTCACATGCAC-30

(W190A), and 50-GTTATTTTGCAGCACACGTTGCAGGGGATC-

AAAATTGGGGAC-30 (W326A) (mutation site underlined). The

double mutant genes (W190F/W326F and W190A/W326A) were

produced using the single mutant gene as template and a primer

containing an additional mutation.

Production and purification of NtChiV and its mutants. NtChiV

and its mutants (W190F, W326F, W190F/W326F, W190A, W326A,

and W190A/W326A) were produced in E. coli Tuner (DE3) carrying

pLacI cells carrying plasmid pETBlue-1 or derivatives, and purified

from the culture supernatants. After dialysis of the culture supernatant

against 10mM Tris–HCl buffer, pH 7.5, the dialysate was applied

to a cation-exchange column of SP Sepharose FF (1� 5 cm)

previously equilibrated with 10mM Tris–HCl buffer, pH 7.5. The

column was washed with 5-volumes of 10mM Tris–HCl buffer,

pH 7.5. The chitinase protein was then eluted with a linear gradient

from 0M to 0.2 M NaCl in 10mM Tris–HCl buffer, pH 7.5. The

chitinase fractions obtained were pooled and applied to a gel filtration

column of HiPrep 26/60 Sephacryl S-100 HR (2:6� 60 cm) previ-

ously equilibrated with 10mM Tris–HCl buffer. pH 7.5, containing

0.1M NaCl. SDS-polyacrylamide gel electrophoresis of the purified

chitinases in 15% slab gels was conducted by the method of

Laemmli.13) Protein bands were detected by staining with Coomassie

Brilliant Blue R-250. The protein concentration was estimated from the

absorbance at 280 nm using molar extinction coefficients calculated

from the amino acid composition of each protein by the equation

proposed by Pace et al.14)

Chitinase activity. Chitinase activity was assayed colorimetrically

using glycol chitin as substrate. Six mL of the enzyme solution was

added to 600 mL of 0.2% (w/v) glycol chitin in a 20mM sodium acetate

buffer, pH 5.0. After incubation at 37 �C for 15min, the reducing sugar

concentration of the reaction mixture was determined using ferri-

ferrocyanide reagent by the method of Imoto and Yagishita.15) One unit

of activity was defined as the amount of enzyme releasing 1 mmol of

GlcNAc per min at 37 �C.

Thermal unfolding experiments. To obtain the thermal unfolding

curve of the enzyme, the CD value at 222 nm was monitored using a

Jasco J-720 spectropolarimeter (cell length 0.1 cm), while the solution

temperature was raised at a rate of 1 �C/min by means of a temperature

controller (PTC-423L, Jasco). To facilitate comparison between

unfolding curves, the experimental data were normalized as follows:

The fraction of unfolded protein at each temperature was calculated

from the CD value by extrapolating the pre- and post-transition

baselines linearly into the transition zone, and plotted against the

temperature. The final concentrations of the enzyme and (GlcNAc)nwere 4mM and 4mM, respectively.

-2+1 +2 -2+3

+1 +2 +3

Trp190Trp326

Trp190Trp326

Fig. 1. Stereo View of the Surface of the NtChiV-(GlcNAc)4 Complex.The bound (GlcNAc)4 represented by sticks was found to bind to subsites �2, þ1, þ2, and þ3, skipping subsite �1. Trp190 and Trp326 were

the mutation targets in this study.

Role of Tryptophan Residues of a Plant Class V Chitinase 779

Anomer analysis of the enzymatic products. The anomeric form of

the enzymatic products was determined by HPLC, as reported

previously.16,17) Enzymatic hydrolysis of the chitin oligosaccharides,

(GlcNAc)n (n ¼ 4, 5, or 6), was carried out in 20mM sodium acetate

buffer, pH 5.0 at 25 �C. The concentrations of the wild-type NtChiV

and the substrate were 1.06 mM (32.1mM for W190A) and 4.6mM,

respectively. After incubation for a given period, a portion of the

reaction mixture was injected directly into a TSK-GEL Amide 80

column (Tosoh, Tokyo), and elution was performed with

acetonitrile:H2O (7:3) at a flow rate of 0.8mL/min. The substrate

and enzymatic products were detected by ultraviolet absorption at

220 nm. The splitting mode of the oligosaccharide substrates was

qualitatively estimated from the �=� ratio of each oligosaccharide

product in the HPLC profiles.18)

HPLC-based determination of the reaction time-course. The

reaction products from the chitinase-catalyzed hydrolysis of

(GlcNAc)n (n ¼ 4, 5, or 6) were determined quantitatively by gel

filtration HPLC by the method of Fukamizo et al.19) The enzymatic

reaction was performed in 20mM sodium acetate buffer, pH 5.0, at

40 �C. The enzyme and substrate concentrations were 0.32 mM and

4.6mM, respectively. To terminate the enzymatic reaction completely

at a given incubation time, a portion of the reaction mixture was mixed

with an equal volume of 0.1M NaOH solution, and immediately frozen

in liquid nitrogen. The resulting solution was applied to a gel filtration

column of TSK-GEL G2000PW (Tosoh, Tokyo), and eluted with

distilled water at a flow rate of 0.3mL/min. Oligosaccharides were

detected by ultraviolet absorption at 220 nm. Peak areas obtained for

individual oligosaccharides were converted to molar concentrations,

which were then plotted against reaction time to obtain the reaction

time-course.

Results

Enzyme productionAll the mutated enzymes (W190F, W326F, W190F/

W326F, W190A, W326A, and W190A/W326A) weresuccessfully expressed in the E. coli expression system,as in the case of the wild-type enzyme.9) The tryptophanmutations did not significantly affect the expressionlevels of the proteins. After purification of the enzymes,the individual proteins exhibited a single band on SDS–PAGE (Supplemental Fig. 1; see Biosci. Biotechnol.Biochem. Web site). The CD spectra of the mutantenzymes were almost identical to that of the wild type(Supplemental Fig. 2), indicating that the individualmutations did not affect the global conformation of theenzyme.

Thermal unfolding experimentsFigure 2 shows the thermal unfolding curves of the

wild-type and W326A proteins. The unfolding transitiontook place at 73.1 �C for the wild type, whereas forW326A the transition temperature (Tm) was 66.0 �C.The Trp326 mutation to alanine was found to impairstability of the enzyme strictly by 7.1 �C. The mutationof Trp326 to phenylalanine impaired stability by 5.2 �C(Table 1). As expected, the mutation to alanine affectedstability more strongly than that to phenylalanine.Similar findings were obtained for the Trp190 muta-tions, as listed in Table 1. The change in Tm induced bythe Trp190 mutation to alanine (4.8 �C) was moreintensive than that to phenylalanine (1.7 �C). However,the effects of the double mutations (W190F/W326F andW190A/W326A) were not additive with respect to theTm decreases, and much less intensive than expectedfrom the effects of single mutations. Similar effects havebeen reported for family GH-19 chitinases from barley

seeds, in which Trp72 and Trp82 were mutated toalanine.20) Mutations of the two tryptophan residues,which are closely located in the substrate-binding cleft,appear to affect the stability cooperatively.On the other hand, the addition of the nondegradable

N-acetylglucosamine trimer (GlcNAc)3 to the wild-typeenzyme did not significantly affect Tm, but affected thecooperativity of the unfolding transition to some extent(Fig. 2, top). When (GlcNAc)3 was added to W326A,however, the stability was almost restored, as shown inthe lower figure. Similar restoration effects due to theaddition of (GlcNAc)3 were observed for the othermutant enzymes, as listed in Table 1. It should be notedthat the restoration effects were greater in the Trp326-mutated enzymes than in the Trp190-mutated enzymes.

Enzymatic activity toward glycol chitinThe tryptophan mutations strictly impaired enzymatic

activity toward a soluble polysaccharide substrate,

0

0.2

0.4

0.6

0.8

1

1.2

50 55 60 65 70 75 80 85 90

0

0.2

0.4

0.6

0.8

1

1.2

50 55 60 65 70 75 80 85 90

Wild type+(GlcNAc)3

Wild type

W326A+(GlcNAc)3

W326A

Frac

tion

unfo

lded

Temperature (°C)

Fig. 2. Thermal Unfolding Curves of the Wild-Type and W326ANtChiV.

The unfolded fractions at the individual temperatures werecalculated from the observed CD values. CD measurements wereconducted in 20mM sodium acetate buffer pH 5.0. The concen-trations of the enzyme and (GlcNAc)3 were 4mM and 4mM,respectively.

Table 1. Thermal Unfolding Transition Temperatures (Tm) of theWild Type and Mutated NtChiV

Enzyme Tm �Tma

Wild type 73.1 0.6- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -W190F 71.4 (�1.7)b 0.2

W326F 67.9 (�5.2)b 5.5

W190F/W326F 69.7 (�3.4)b 1.6- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -W190A 68.3 (�4.8)b 1.7

W326A 66.0 (�7.1)b 5.4

W190A/W326A 67.7 (�5.4)b 2.3

a�Tm, increase in Tm induced by (GlcNAc)3 binding to NtChiVbNegative values in parentheses indicate the decrease in Tm from the value

for the wild type.

780 N. UMEMOTO et al.

glycol chitin, as listed in Table 2. The Trp326 mutationto alanine was found to abolish completely the enzy-matic activity. The mutation of Trp326 to phenylalaninereduced activity to 10%. The activity decrease inducedby the Trp190 mutation to alanine (90%) was moreintensive than that to phenylalanine (74%). For thedouble mutants (W190F/W326F and W190A/W326A),activity decreases were more intensive than those of thesingle mutations. Overall, effects of the Trp326 muta-tions were more intensive than those of the Trp190mutations.

Enzymatic activity toward N-acetylglucosamine hex-asaccharide

Chitin hexasaccharide was hydrolyzed by the wild-type and individual mutant enzymes, and the reactionproducts were analyzed by HPLC for a given reactiontime to obtain the time-course of hexasaccharidedegradation (Fig. 3). The wild-type enzyme produced(GlcNAc)2, (GlcNAc)3, and (GlcNAc)4 (Fig. 3A). Asimilar product distribution was obtained when W190Fwas used instead of the wild type (Fig. 3B). The specificactivities toward (GlcNAc)6 for the two enzymes werealmost identical. However, in W326F and W190F/W326F, considerable losses of activity (5% and 4% ofthat of the wild type, respectively) were observed, while

product distribution was not affected by these mutations.Enzymatic activity toward oligosaccharide substrateswas strongly affected by the Trp326 mutation, but not bythe Trp190 mutation. (GlcNAc)6 was not significantlyhydrolyzed by W190A, W326A, or W190A/W326A(data not shown).

Anomer analysis of the enzymatic products from theoligosaccharide substratesFigure 4 shows the time-dependent profile of

(GlcNAc)4 degradation by the wild-type and W190Aenzymes. Since NtChiV is a retaining enzyme, thenewly produced reducing end is exclusively a �-anomer.�-(GlcNAc)2 was almost exclusively produced by thewild type enzyme, and slowly converted to �-(GlcNAc)2by mutarotation (the upper three profiles in Fig. 4). Itappears that the �-(GlcNAc)2 produced is derived notonly from the newly produced reducing end but from thereducing end of the initial substrate (GlcNAc)4. Thus�-(GlcNAc)4 appears to be specifically hydrolyzed bythe wild-type enzyme. In fact, the rate of �-(GlcNAc)4degradation was higher than that of �-(GlcNAc)4degradation. Conversely, W190A produced a consider-able amount of �-(GlcNAc)2 from the (GlcNAc)4substrate, but �-(GlcNAc)2 was still dominant in theproducts (the lower three profiles in Fig. 4). The rate of�-(GlcNAc)4 degradation was similar to that of �-(GlcNAc)4 degradation. This suggests that �-(GlcNAc)4was not specifically hydrolyzed by W190A.An anomer analysis was also conducted for the

products from (GlcNAc)5, and the results are shown inFig. 5. In the wild-type enzyme (the upper threeprofiles), �-(GlcNAc)2 and �- and �-(GlcNAc)3 wereproduced from (GlcNAc)5. The �=� ratio of the(GlcNAc)3 product was slightly higher than that of theinitial (GlcNAc)5 substrate, indicating a low level of�-(GlcNAc)3 formation. It is very likely that (GlcNAc)5binds mostly to subsites �2, �1, þ1, þ2, and þ3 ofthe wild-type enzyme, but W190A exhibited a different

Table 2. Enzymatic Activities toward Glycol Chitin for the WildType and the Mutated NtChiV

Enzyme (NtChiV)Specific activity Relative activity

(mmol/min/mg) (%)

Wild type 7.53 100

W190F 1.97 26

W326F 0.79 10

W190F/W326F 0.28 4

W190A 0.76 10

W326A Not detected Not detected

W190A/W326A Not detected Not detected

0

1

2

3

4

5

6

7

8

0 20 40 60 80 100 120

0

1

2

3

4

5

6

7

8

0 100 200 300 400

Wild type

W326F

(GlcNAc)6 (GlcNAc)2 (GlcNAc)4 (GlcNAc)3

0

1

2

3

4

5

6

7

8

0 20 40 60 80 100 120

W190F

0

1

2

3

4

5

6

7

8

0 100 200 300 400

W190F/W326F

Reaction time (min)

Con

cent

ratio

n (m

M)

Fig. 3. Time-Courses of (GlcNAc)6 Degradation Catalyzed by the Wild-Type and Trp-Mutated NtChiVs.The enzymatic reactions were conducted in 20mM sodium acetate buffer pH 5.0 at 40 �C. The concentrations of the enzyme and (GlcNAc)6

were 0.32 mM and 4.6mM, respectively. The substrate and product concentrations at a given reaction time were determined by gel-filtration HPLCusing a column of TSK-GEL G2000PW (Tosoh). Symbols: squares, (GlcNAc)2; triangles, (GlcNAc)3; diamonds, (GlcNAc)4; circles,(GlcNAc)6. Lines were obtained by following the experimental data points roughly. The time-course for the wild type enzyme is cited fromOhnuma et al.9)

Role of Tryptophan Residues of a Plant Class V Chitinase 781

profile of anomer formation (the lower three profiles).The �=� ratio of the (GlcNAc)3 product was muchhigher than that of the initial (GlcNAc)5 substrate,indicating enhanced formation of �-(GlcNAc)3. InW190A, some fraction of (GlcNAc)5 appears to bindto subsites �3, �2, �1, þ1, and þ2, resulting inenhanced �-(GlcNAc)3 formation. The �=� ratio of(GlcNAc)5 did not change with progress of the enzy-matic reaction in either enzyme (Fig. 5), whereas in thewild-type enzyme the rate of decrease in �-(GlcNAc)4was higher than in �-(GlcNAc)4 (Fig. 4). In contrast tothe reaction toward (GlcNAc)4, the anomer selectivity ofthe substrate was not distinct in the reaction toward(GlcNAc)5. In the wild-type enzyme, the reducing endresidues of most (GlcNAc)5 molecules are in contactwith subsite þ3, and those of (GlcNAc)4 with subsiteþ2. Anomer selectivity is likely to take place at subsiteþ2, but not at subsite þ3.

Discussion

Tryptophan residues play an important role in thesubstrate-binding of glycoside hydrolases.21) Theirstacking interactions between the indole side chain andthe pyranose ring of the sugar residue are recognized asthe most important in substrate-binding, but the role of

tryptophan residues differs according to their location.For example, there are three tryptophan residues in thesubstrate-binding cleft of hen egg-white lysozyme,Trp62, Trp63, and Trp108.22–26) Trp62 is expose to thesolvent at subsite �3, but Trp63 is located relativelyinside at the same subsite. Trp108 is located at subsite�2 but very close to the carboxyl side chain of Glu35.Based on the crystal structure of the lysozyme in acomplex with (GlcNAc)3,

22) the side chains of Trp62and Trp63 interact with the OH groups of C6 and C3of the �2 sugar, respectively. The indole side chainof Trp62 also makes a stacking interaction with thepyranose ring of the �3 sugar. Both tryptophan residuesare significantly involved in sugar-residue binding atsubsites �2 and �3. Experimental evidence obtainedby site-directed mutagenesis23,24) appears to support thisidea basically, although a more complicated hydrogen-bonding network is involved in the Trp62 interaction.25)

On the other hand, Trp108 was found to have a specificfunction important for lysozyme catalysis: the hydro-phobicity of the tryptophan residue elevates the pKa

value of Glu35.26) The elevated pKa value of Glu35retains the protonated form of the side chain carboxylgroup, and hence the proton-donating ability of theglutamic acid. These findings led us to examine the roleof tryptophan residues in NtChiV.

Trp190Figure 6 shows a stereo view of the binding cleft

of NtChiV in a complex with (GlcNAc)4. As seen in

1 2 3 4

ααβ

α α αβ β β

0.5 min

4 min

8 min

8 min

0.5 min

16 min

5 10 15Retention time (min)

Abs

orba

nce

at 2

20 n

m

Wild-type

W190A

Fig. 4. Time-Dependent HPLC Profile Showing the Hydrolysis of(GlcNAc)4 by the Wild-Type and W190A NtChiV.The enzyme and substrate concentrations were 1.06 mM (32.1mM

for W190A) and 4.6mM, respectively. The enzymatic reaction wasconducted in 20mM sodium acetate buffer pH 5.0 at 25 �C. HPLCwas conducted with a column of TSK-GEL Amide-80 (Tosoh) using70% acetonitrile as the elution solvent. The numerals in the figurerepresent degrees of polymerization of the GlcNAc oligosaccha-rides. � and � are the anomeric forms. The upper three profiles arefor the wild type, and the lower three for W190A.

α

β

1 2 3 4 5

α α αα

β β β β

0.5 min

4 min

8 min

8 min

0.5 min

16 min

5 10 15 20Retention time (min)

Abs

orba

nce

at 2

20 n

m Wild-type

W190A

Fig. 5. Time-Dependent HPLC Profile Showing the Hydrolysis of(GlcNAc)5 by the Wild-Type and W190A NtChiV.

Reaction conditions were the same as in Fig. 4. Numerals in thefigure represent the degrees of polymerization of the GlcNAcoligosaccharides. � and � are the anomeric forms. The upper threeprofiles are for the wild-type, and the lower three for W190A.

782 N. UMEMOTO et al.

Figs. 1 and 6, the Trp190 side chain is exposed to thesolvent and appears to make a face-to-face stackinginteraction with the þ2 sugar. The mutation of Trp190resulted in a significant loss of stability (W190F,�1.8 �C; W190A, �4.8 �C), but (GlcNAc)3 bindingdid not fully restore the stability (W190F, 0.2 �C;W190A, 1.7 �C). (GlcNAc)3 is unlikely to bind to themutation site, the 190th position including subsite þ2.However, the two mutant enzymes (W190F andW190A) exhibited significant activity toward glycolchitin (W190F, 26%; W190A, 10%), indicating thatTrp190 contributes only partly to the enzymatic func-tion. The profile of the time-course of (GlcNAc)6degradation catalyzed by W190F was almost identicalto that obtained for the wild-type enzyme. All of theseresults confirm that Trp190 interacts only with the þ2sugar. Although the contribution of Trp190 to substrate-binding is restricted to subsite þ2, it is still significantfor substrate recognition. Figure 4 suggests that the�-anomer of (GlcNAc)4 is specifically hydrolyzed bythe wild type, but the specificity is missing in W190A.Thus, Trp190 recognizes the �-anomer at subsite þ2.The mode of (GlcNAc)5 binding was found to be shiftedby the W190A mutation (Fig. 5). This also suggestsa significant contribution of Trp190 to the substrate-binding in NtChiV.

Trp326As seen in Figs. 1 and 6, the Trp326 side chain is

buried inside the hydrophobic region, but rather close tothe catalytic center (Gln(Glu)115). The catalytic residueappears to be surrounded by aromatic side chainsof Phe31, Tyr183, Phe185, and Trp326. As expectedfrom the structure, the mutation of Trp326 greatlyimpaired stability (W326F, �5.3 �C; W326A, �7.1 �C).(GlcNAc)3 binding almost fully restored stability(W326F, 5.5 �C; W326A, 5.4 �C). (GlcNAc)3 is likelyto bind to the mutation site, the hydrophobic pocketincluding the catalytic center (Table 1). On the otherhand, the effects of the two mutations on enzymaticactivity (W326F, 10%; W326A, not detected) weremuch more intensive than those of the mutations ofTrp190, indicating that Trp326 makes a very importantcontribution to the enzyme function (Table 2). Theinitial velocity of (GlcNAc)6 degradation was alsostrictly reduced by the W326F mutation, while theproduct distribution obtained from the (GlcNAc)6hydrolysis catalyzed by W326F was similar to that

obtained by the wild-type enzyme (Fig. 3). Fukamizoet al.27,28) conducted a computer simulation of theenzymatic hydrolysis of (GlcNAc)n, under the assump-tion that the substrate binding cleft of the model enzymeis composed of several subsites (�m, . . .�2, �1, þ1,þ2, . . .þn). They found that reduction of the affinityof �2 sugar strongly affects the initial velocity of(GlcNAc)6 degradation, but not the product distribution.The effects of the reduction of �2 sugar affinity on thetheoretical time-course were similar to those of theW326F mutation on the experimental time-course of(GlcNAc)6 degradation (Fig. 3, two panels at left). Itappears that the affinity of the �2 sugar is reduced inW326F as compared to the wild type. Trp326 is likely tocontribute to �2 sugar affinity, probably through hydro-phobic interaction. This is consistent with predictionfrom the X-ray crystal structure of the NtChiV-(GlcNAc)4 complex.9) However, the strong effect ofthe Trp326 mutation on enzymatic activity suggests thatTrp326 is involved not only in the sugar residue bindingbut also in catalytic efficiency. The tryptophan residueprobably stabilizes the conformation of the catalyticpocket also through hydrophobic interactions with otherhydrophobic amino acid residues. The hydrophobicinteractions in the catalytic pocket might provide anappropriate environment for the catalytic action ofGlu115. In fact, such a hydrophobic environment hasbeen recognized to elevate the pKa value of the protondonor.26) The elevated pKa of Glu115 should result in anefficient proton-donating action in NtChiV-catalyzedhydrolysis.In conclusion, Trp190 interacts with the þ2 sugar,

recognizing �-anomer of the sugar residue, whileTrp326 plays multiple roles in stabilizing the proteinstructure, in sugar residue binding at subsite �2, andprobably in maintaining catalytic efficiency by provid-ing a hydrophobic environment for proton donorGlu115.

Acknowledgments

This work was supported by the Strategic Project toSupport the Formation of Research Bases at PrivateUniversities: Matching Fund Subsidy from MEXT(Ministry of Education, Culture, Sports, Science andTechnology), 2011–2015 (S1101035), and in part bya Grant-in-Aid for Scientific Research from MEXT toT.O. (23780349).

-2

+1

+3

W326

Q115

Y183

Y116

D184

+1

+2R75

G74

N189

F31

F185

W190

N162-2

+1

+3

W326

Q115

Y183

Y116

D184

+1

+2R75

G74

N189

F31

F185

W190

N162

Fig. 6. Stereo View of the (GlcNAc)4 Binding Site of E115Q NtChiV.The indole ring of Trp190 is involved in face-to-face stacking interactions with the GlcNAc residue at subsite þ2. The Trp326 side chain

appears to contribute to the hydrophobic environment of the catalytic pocket.

Role of Tryptophan Residues of a Plant Class V Chitinase 783

References

1) Henrissat B and Davies G, Curr. Opin. Struct. Biol., 7, 637–644

(1997).

2) Watanabe T, Kanai R, Kawase T, Tanabe T, Mitsutomi M,

Sakuda S, and Miyashita K, Microbiology, 145, 3353–3363

(1999).

3) Kasprzewska A, Cell. Mol. Biol. Lett., 8, 809–824 (2003).

4) Collinge DB, Kragh KM, Mikkelsen JD, Nielsen KK,

Rasmussen U, and Vad K, Plant J., 3, 31–40 (1993).

5) Neuhaus J-M, Fritig B, Linthorst HJM, and Meins FJr, Plant

Mol. Biol. Rep., 14, 102–104 (1996).

6) Ohnuma T, Numata T, Osawa T, Mizuhara M, Lampela O,

Juffer AH, Skriver K, and Fukamizo T, Planta, 234, 123–137

(2011).

7) Takenaka Y, Nakano S, Tamoi M, Sakuda S, and Fukamizo T,

Biosci. Biotechnol. Biochem., 73, 1066–1071 (2009).

8) Terwisscha van Scheltinga AC, Kalk KH, Beintema JJ, and

Dijkstra BW, Structure, 2, 1181–1189 (1994).

9) Ohnuma T, Numata T, Osawa T, Mizuhara M, Varum KM, and

Fukamizo T, Plant Mol. Biol., 75, 291–304 (2011).

10) van Aalten DM, Komander D, Synstad B, Gaseidne S, Peter

MG, and Eijsink VG, Proc. Natl. Acad. Sci. USA, 98, 8979–

8984 (2001).

11) Horn SJ, Sørbotten A, Synstad B, Sikorski P, Sørlie M, Varum

KM, and Eijsink VG, FEBS J., 273, 491–503 (2006).

12) Yamada H and Imoto T, Carbohydr. Res., 92, 160–162 (1981).

13) Laemmli UK, Nature, 227, 680–685 (1970).

14) Pace CN, Vajdos F, Fee L, Grimsley G, and Gray T, Protein

Sci., 4, 2411–2423 (1995).

15) Imoto T and Yagishita K, Agric. Biol. Chem., 35, 1154–1156

(1971).

16) Fukamizo T and Hayashi K, J. Biochem., 91, 619–626 (1982).

17) Koga D, Yoshioka T, and Arakane Y, Biosci. Biotechnol.

Biochem., 62, 1643–1646 (1998).

18) Aronson NNJr, Halloran BA, Alexyev MF, Amable L, Madura

JD, Pasupulati L, Worth C, and van Roey P, Biochem. J., 376,

87–95 (2003).

19) Fukamizo T, Sasaki C, Schelp E, Bortone K, and Robertus JD,

Biochemistry, 40, 2448–2454 (2001).

20) Fukamizo T, Miyake R, Tamura A, Ohnuma T, Skriver K,

Pursiainen NV, and Juffer AH, Biochim. Biophys. Acta, 1794,

1159–1167 (2009).

21) Vyas NK, Curr. Opin. Struct. Biol., 1, 732–740 (1991).

22) Imoto T, Johnson LN, North ACT, Phillips DC, and Rupley JA,

‘‘The Enzymes’’ Vol. 7, ed. Boyer PD, Academic Press, New

York, pp. 665–868 (1972).

23) Kumagai I, Maenaka K, Sunada F, Takeda S, and Miura K, Eur.

J. Biochem., 212, 151–156 (1993).

24) Maenaka K, Kawai G, Watanabe K, Sunada F, and Kumagai I,

J. Biol. Chem., 269, 7070–7075 (1994).

25) Maenaka K, Matsushima M, Song H, Sunada F, Watanabe K,

and Kumagai I, J. Mol. Biol., 247, 281–293 (1995).

26) Inoue M, Yamada H, Yasukochi T, Kuroki R, Miki T, Horiuchi

T, and Imoto T, Biochemistry, 31, 5545–5553 (1992).

27) Fukamizo T, Kuhara S, and Hayashi K, J. Biochem., 92, 717–

724 (1982).

28) Fukamizo T, Hayashi K, and Goto S, Eur. J. Biochem., 158,

463–467 (1986).

784 N. UMEMOTO et al.