University of Groningen

How lactobacilli synthesize inulinAnwar, Munir Ahmad

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2010

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Anwar, M. A. (2010). How lactobacilli synthesize inulin: characterization of fructosyltransferase enzymes.Groningen: s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 18-10-2020

Chapter 5

Probing the Lactobacillus reuteri inulosucrase reaction and

product specificity by site-directed mutagenesis of conserved

amino acid residues

Munir A. Anwar, Hans Leemhuis, Tjaard Pijning, Slavko Kralj, Bauke W. Dijkstra and

Lubbert Dijkhuizen

Submitted for publication

Chapter 5

132

Abstract

The probiotic bacterium Lactobacillus reuteri 121 produces two fructosyltransferase

(FTF) enzymes; a levansucrase (Lev) and an inulosucrase (Inu). Although these two FTF

enzymes share high sequence and 3D structural similarity, they differ significantly in the

type and size distribution of FOS products synthesized from sucrose, and in their activity

levels. In order to examine the contribution of specific amino acids in such differences, 15

single inulosucrase mutants and 4 multiple mutants were designed in residues that are

conserved in Inu, but not in Lev proteins. The effects of the mutations were interpreted using

the 3D structures of Bacillus subtilis levansucrase (SacB)(15) and Lactobacillus johnsonii

NCC 533 inulosucrase (InuJ)(Chapter 4).

Mutating residue G416, present at the rim of the active site pocket in loop 415-423,

into glutamate increased the hydrolytic activity 2-fold, without significantly changing

transglycosylation activity. A multiple mutant GM4 (T413K, K415R, G416E, A425P,

S442N, W486L, P516L), which also contains three residues from the above mentioned loop,

synthesized 1-kestose only, though at low efficiency. Mutation A538S, located behind the

general acid/base, increased enzyme activity 2-3 fold. Mutation N543S, located adjacent to

+1/+2 subsite residue R544, synthesized less FOS as compared to the wild type enzyme. The

current study demonstrates that the product specificity of Inu is easily altered by protein

engineering, obtaining Inu variants with higher transglycosylation specificity, higher

catalytic rates and different FOS oligosaccharide size distributions.

1. Introduction

Inulosucrases (E.C. 2.4.1.9) and levansucrases (E.C. 2.4.1.10) form a group of

enzymes, called fructosyltransferases (FTFs), that polymerize the fructose moiety of their

substrate sucrose into fructans which possess either inulin or levan structures with β(2-1) and

β(2-6) linkages, respectively. Most of the known microbial FTF enzymes are levansucrases,

characterized from a wide variety of microorganisms, whereas so far only a few

inulosucrases have been characterized from a few species of lactic acid bacteria

(www.cazy.org) (4).

Site-directed mutagenesis of Inu

133

These FTF enzymes belong to the glycoside hydrolase family 68 (GH68)

(www.cazy.org)(4) and share a high (>60%) amino acid sequence similarity. Depending on

the type of acceptor substrate used, they convert the donor substrate sucrose into glucose and

fructose (hydrolysis), fructooligosaccharides (FOS) and fructan polymer

(transfructosylation). Family GH68 enzymes are retaining enzymes {Koshland, 1954 915

/id}. The levansucrases from Bacillus subtilis, Gluconacetobacter diazotrophicus, and

Streptococcus salivarius catalyze transfructosylation via a ping-pong mechanism involving

the formation of a transient fructosyl-enzyme intermediate (5,7,10,22,24).

Previous studies have revealed that, in spite of their high sequence and secondary

structure similarity, FTF enzymes of Lactobacillus reuteri 121 differ strongly in product

spectrum from sucrose. The levansucrase (Lev) produced mainly levan polymer and FOS

with a degree of polymerization (DP) < 5, whereas the inulosucrase (Inu) synthesized mostly

FOS up to DP15 and relatively low amounts of inulin polymer (20). In addition, Lev

exhibited a lower total activity and a lower transglycosylation/hydrolysis ratio as compared

to Inu (19). Similar observations were made when comparing the Lactobacillus gasseri

levansucrase (LevG) with Lb. gasseri (InuGA and InuGB) and Lactobacillus johnsonii

(InuJ) inulosucrases (1,2). Knowledge about the structural basis of these functional

differences in FTFs is desirable, providing fundamental insights in reaction and product

specificity of GH68 enzymes, and would be beneficial for protein engineering aimed at

construction of enzymes tailored to synthesize desirable products.

In the past few years, several attempts, including mutagenesis and crystal structure

elucidation, have been made to determine structure-function relationships of FTF enzymes.

For instance, high resolution 3D structures for the levansucrase enzymes of the Gram-

positive bacterium B. subtilis (SacB), also with bound sucrose and raffinose donor substrates

(15,16) and the Gram-negative bacterium G. diazotrophicus (LsdA)(14) have been

elucidated. These levansucrase enzymes both show a similar five-bladed β-propeller

topology with a negatively charged central pocket. Proteins of family GH68 and GH32

(invertases, inulinases) maintain a similar structural fold and on this basis they are grouped

in Clan-GH-J of carbohydrate active enzymes (www.cazy.org) (4). The structural similarities

and differences among 3D structures of families GH32 and GH68 enzymes, along with

functional implications, have been reviewed recently (13).

Chapter 5

134

Based on the 3D structures of the SacB and LsdA proteins, and site-directed

mutagenesis studies on FTFs from diverse microorganisms, the FTF catalytic residues and

their functions have been identified. These include three residues (SacB numbering): D86,

D272 and E342, which serve as catalytic nucleophile, transition state stabilizer and general

acid base catalyst, respectively (3,15,21,24,30). The roles of other amino acid residues in the

catalytic domain in the FTF reaction mechanism and product specificity also have been

investigated. For example, it was demonstrated that the mutation R360K/S in SacB

drastically reduces polymerization activity (6). Amino acids forming contacts with sucrose

and constituting the -1 and +1 sugar binding subsites (nomenclature adopted from (9)) have

been identified. Subsite -1 consists of W85, W163, R246 and D247, whereas subsite +1 is

formed by R246, E340, E342 and R360 (15,19). Site-directed mutagenesis of the Lb. reuteri

121 inulosucrase -1 sugar binding subsite residues strongly influenced the size of the

products synthesized (19). Similarly, it was shown that N252 (a constituent of the +2

subsite) and R370 in Bacillus megaterium levansucrase are critical for the polymer versus

oligosaccharide product ratio of the enzyme (11). Analysis of the raffinose bound E342A

mutant of SacB has revealed that the raffinose galactosyl unit protruded out of the active site

while specificity-determining contacts were restricted to the sucrosyl moiety (16).

Crystallographic and biochemical data of a B. subtilis levansucrase mutant S164A

demonstrated that S164 is important in maintaining the position of the nucleophile in the

active site (17). In the same study, the mutants R360S, Y429N and R433A were found to

produce exclusively oligosaccharides but not levan polymer.

Recently, we have elucidated the 3D structure of Lb. johnsonii NCC 533

inulosucrase (InuJ) (Chapter 4). The native InuJ subsite -1 is virtually identical to that of

SacB and LsdA, indicating that the binding mode of the -1 fructosyl unit of the substrate is

very similar in inulo- and levansucrases. As the orientation of an incoming acceptor

determines the type of glycosidic bond of the product, functional differences could be

expected at the acceptor binding subsites (+1, +2, +3 etc.), which remain to be characterized.

As families GH68 and GH32 share similar protein folds, it would be worthy to mention here

a recent report on the 3D structure of Aspergillus japonicus fructosyltransferase (a GH32

enzyme) with bound substrate molecules, where the residues Y404 and E405 have been

Site-directed mutagenesis of Inu

135

shown to contribute to the +2/+3 subsites and speculated to be responsible for the formation

of the inulin-type product (8).

It is evident from the preceding discussion that most of the structure-function related

studies have focused on levansucrases and on the role of amino acid residues conserved

between levan- and inulosucrases. No attempts have been made yet to determine the

functional importance of the amino acid residues that are well conserved in inulosucrases but

are different in levansucrases. Only a single study seeks to explain the difference in sizes of

the products synthesized by levan- and inulosucrases of Lb. reuteri 121 with a model

showing that the levansucrase and inulosucrase catalyze processive and non-processive type

of reactions, respectively, due to difference in affinities of +2 and +3 subsites for acceptors

(20).

Here we report a characterization of Lb. reuteri 121 Inu variants in which the

conserved inulosucrase amino acids, identified on the basis of amino acid sequence

alignments (Table 1), were substituted for Lb. reuteri 121 Lev amino acids. A comparison

with the X-ray structures of the levansucrases from B. subtilis (38% identity at the amino

acid level) (15) and the recently elucidated Lb. johnsonii inulosucrase (74 % identity in 554

amino acids) (Chapter 4) indicated an overall similar protein structure, carrying identical

essential amino acid motifs but also some important differences. A detailed biochemical

investigation of the 19 Inu variants identified catalytic specificity and product profile

modifying determinants.

Chapter 5

136

Table 1. Alignments of FTFs from lactobacilli and other Gram-positive bacteria showing the 15 single mutants and triple mutant IAM/YCL(GM1)

generated in Lb. reuteri 121 Inu at the top. Based on the single mutants, multiple mutants GM2 (N365K, T366L, A425P) GM3 (A425P, A538S,

N543S, D548R, W551T) and GM4 (T413K, K415R, G416E, A425P, S442N, W486L, P516L) were constructed. The GenBank accession numbers

of FTF proteins are shown in parenthesis following the corresponding strain name.

§ Amino acid residues 420-422

Inulosucrases Lb. reuteri 121 (AAN05575) Lb. gasseri DSM 20243 (BK006921) Lb. gasseri DSM 20604 (ACZ67286) Lb. johnsonii NCC533 (AAS08734) Lb. reuteri TMW 1.106 (CAL25302) Leuc. citreum CW28 (AAO25086)

N N N N N -

T T T T T -

T T T T T Q

K K K K K -

G G G G G -

A A A A A A

IAM IAM IAM IAM IAM FTL

A A A A A P

S S S S S N

W W W W W T

A A A A A T

P P P P P V

A A A A A G

N N N N N N

D D D D D D

W W W W W P

Levansucrases Lb. reuteri 121 (AAO14618) Lb. gasseri DSM 2077 (ACZ67287) Lb. panis (29) Strep. salivarius (AAA71925) Lb. sanfranciscensis (CAD48195) Leuc. mesenteroides LevS (AAY19523) Leuc. mesenteroides B-512FMC (AAT81165) Leuc. mesenteroides LEUM_1409 (ABJ62502) Leuc. mesenteroides LEUM_1410 (ABJ62503) Leuc. mesenteroides LEUM_1411 (ABJ62504) Leuc. mesenteroides MIFT (AAT81165) B. subtilis SacB (AAA22724)

K K K K K P P P P P P H

L L L L L N N R M D - Y

K K G T K A A A A N - N

R R R D R T T D D N - S

E N Q D K T T T S D - S

N R D T N G G G G - - G

YCL YTL YCL YCL YCL FTM FTM FTL FTL FAL - HTL

P P P P P P P P P P P P

N N N N N N N N N N N N

L V E W L K K K K K F L

G G G G G G G G G G A G

L L L H L N N N N N V N

S S S T T T T T T T T T

S S S N S S S S S D - S

R G T A R P P P P K - D

T T Y T N D D D D E V T

Inu Mutant

N 365 K

T 366 L

T 413 K

K 415 R

G 416 E

A 417 N

IAM §/ YCL

A 425 P

S 442 N

W 486 L

A 489 G

P 516 L

A 538 S

N 543 S

D 548 R

W 551 T

Site-directed mutagenesis of Inu

137

2. Materials and methods

2.1. Bacterial strains, plasmids and culturing conditions

The construct Inu∆699His (28), consisting of a C-terminally truncated (aa 699 to aa

798) Lb. reuteri 121 inulosucrase (Inu) in plasmid pBAD/myc/his/C, was used as parent and

wild type (WT) enzyme in the current mutagenesis study. Escherichia coli TOP10

(Invitrogen) cells were used as host for cloning and expression purposes. E. coli strains were

grown at 37 °C at 210 rpm shaking speed in Luria–Bertani (LB) medium supplemented with

100 µg ml-1 of ampicillin. LB-agar plates were made by adding 1.5% agar to the LB

medium.

2.2. Molecular techniques

Multiple amino acid sequence alignments were made with ClustalW 1.74 (25) to

identify the mutation targets (Table 1). Plasmid DNA was isolated from E. coli using a

plasmid miniprep kit (Sigma-Aldrich). General procedures for cloning, DNA manipulations,

and agarose gel electrophoresis were as described (23). Point mutations were introduced by

PCR (PTC-200 Thermal Cycler, MJ Research Inc. USA) using pfu DNA polymerase

(Fermentas) and primers listed in Table 2, following the Quik Change protocol (Stratagene).

The resulting constructs with mutant inu genes were transformed into E. coli TOP10 cells by

the heat shock method using chemically competent cells (23) for protein expression. The

multiple mutations were introduced sequentially using Inu∆699His as parent in the Quik

Change protocol. Mutations were confirmed by nucleotide sequence analysis (GATC,

Germany).

2.3. Expression and purification of FTF proteins

Cultures (400 ml each) of E. coli TOP10 harboring mutant constructs were grown

overnight with 0.02 % arabinose for protein expression at 200 rpm shaking speed. The cells

were harvested by centrifugation at 3500 × g for 15 min and the recombinant FTF protein

was purified to homogeneity by Ni-NTA affinity chromatography as described (28). Protein

concentrations were determined using the Bradford reagent (Bio-Rad, Germany) with bovine

serum albumin as standard.

Chapter 5

138

Table 2. Primers used to introduce mutations in L. reuteri 121 inulosucrase construct Inu∆699His.

Changed codons and nucleotides are shown underlined and in bold, respectively. Reverse

complements of these oligonucleotides/primers were used as reverse primers in the PCR reactions

2.4. Biochemical characterization of the mutant enzymes

Initial total, hydrolytic and transglycosylation activities of wild-type and recombinant

enzymes were determined by measuring the amount of glucose and fructose released from

sucrose. All the enzyme activity assays were performed at 37 ºC in 25 mM sodium acetate

buffer (pH 5.4) containing 1 mM CaCl2 and 500 mM sucrose. The assay mixture was pre-

incubated at 37 ºC for 10 min and reactions were started by enzyme addition. Samples were

A. Primer Name/ Mutation

Primer Sequence (5’ to 3’)

N365K

CTGATAACAAGACCAATCATC

T366L TGATAACAATCTCAATCATCAA T413K CAATGGAAAGCTAAGAACAAAGGTGCCG K415R GCTACTAACAGAGGTGCCGAT G416E GCTACTAACAAAGAGGCCGATAATATTG A417N CTAACAAAGGTAACGATAATATTG A425P AATGCGTGATCCTCATGTAAT S442N TTTGAAGCAAATACTGGTTTA W486L GTCGGGCAACTTTGGCTAATGCAGC A489G TGGGCTAATGGAGCTATCGGC P516L ATTTCTGCACTAATGGTAAGC A538S TTACTTATTTTCCGCTACCCG N543S ACCCGTTTAAGTCGAGGAAGT D548R AGGAAGTAATCGTGATGCTTGG W551T TGATGATGCTACGATGAATGCT B. Primers used for multiple mutants:

GM1 (I420Y:A421C:M422L)

TGCCGATAATTATTGCCTGCGTGATGCTC

GM2 Primer used sequentially: T366L, A425P and modified N365K: CTGATAACAAGCTCAATCATC

GM3 Primers used sequentially: A425P, A538S, N543S, D548R and modified W551T primer: TCGTGATGCTACGATGAATGCT

GM4 Primers used sequentially: T413K, K415R, G416E, A425P, S442N, W486L, P516L, modiefied K415R primer: GCTAAGAACAGAGGTGCCGAT ; and finally, modified G416E primer: GCTAAGAACAGAGAGGCCGATAATATTG

Site-directed mutagenesis of Inu

139

taken every 3 min, for a total of 15 min, and used to determine the amount of glucose and

fructose released from sucrose (27). Total activity (VG) is the amount of glucose released

from sucrose during the reaction. The amount of fructose (VF) formed is a measure for the

hydrolytic activity. The transglycosylation activity was calculated by subtracting the amount

of free fructose from glucose (VG-VF). One unit of enzyme activity is defined as the release

of 1 µmol of monosaccharide per min from sucrose. For stability measurements, the wild

type (WT) and mutant enzymes were incubated at 37 ºC in 25 mM sodium acetate buffer

(pH 5.4) containing 1 mM CaCl2 for 1 h and 24 h and their total activity was measured as

described above.

2.5. Poly/oligosaccharide production and characterization

To produce fructan oligosaccharides and polymers, the (mutant) enzymes (2 U ml-1)

were incubated with sucrose (500 mM) at 37 °C in 25 mM sodium acetate buffer (pH 5.4)

containing 1 mM CaCl2 (final volume 1 ml). Samples taken from the reaction mixture after

24 h of incubation were diluted 1000 times with dimethlsulfoxide (DMSO) and subjected to

high performance anion-exchange (HPAE) chromatography (Dionex, Sunnyvale, USA) for

separation and analysis of oligosaccharides products. Separation of the oligosaccharides was

achieved using the following gradient: eluent A (0 min, 100%); (10 min, 78%); (25 min,

60%); (80 min, 10%); (83 min, 0%); (91 min, 100%). Eluent A was 0.1 M sodium hydroxide

and eluent B was 0.1 M sodium hydroxide in 0.6 M sodium acetate. 1-Kestose and 1,1-

nystose (Fluka Chemie AG, Switzerland) were used as standards. Larger reaction mixture

volumes (100 ml) and lower enzyme concentrations (1 U ml-1) with longer incubation times

(72 h) were used to produce polymer for NMR analysis. In case of the multiple mutants

GM2, GM3 and GM4, reaction mixtures were incubated for 4, 2 and 6 weeks respectively,

while 0.5 U ml-1 enzyme activity was used for mutants GM2 and GM4. Subsequently the

mixture was treated with two volumes of cold 96% ethanol for polymer precipitation. The

polymer precipitated from the reaction mixture was separated by centrifugation at 2500 × g

for 15 min. The pellet was dissolved in MilliQ water and the precipitation process was

repeated twice (26). The polymer was finally freeze dried and subjected to NMR

spectroscopy after dissolving in 99.9 atom % D2O (Aldrich). One-dimensional 13C-NMR

spectra were recorded at 125 MHz on a 500 MHz Varian Inova NMR spectrometer at a

Chapter 5

140

probe temperature of 80°C. Chemical shifts are expressed in ppm relative to the methyl

group of internal acetone (δ=31.07). Carbon spectra were recorded in 38K data sets, with a

spectral width of 30.166 kHz. Prior to Fourier transformation the time-domain data were

apodized with an exponential function, corresponding to a 0.5 Hz line broadening.

3. Results and discussion

The present study was undertaken to further investigate the structural basis for

functional differences between levansucrase and inulosucrases enzymes. Inu and Lev were

selected for comparison as they both are produced by Lb. reuteri 121 and are closely related

(86% similarity) at the amino acid sequence level. Mutation targets in Inu were selected on

the basis of alignments of levansucrase and inulosucrase sequences of lactic acid bacteria,

and point mutations were introduced at positions highly conserved in inulosucrases but

different/variable in levansucrases (Table 1). Multiple mutants (GM1-GM4) were

constructed based on the location of the targeted amino acids in the active site funnel.

Characteristics of the mutants that behaved differently from the WT are discussed below in

the light of the 3D structures of the InuJ inulosucrase which became available after these

mutagenesis studies (Chapter 4) and the SacB levansucrase.

3.1. Expression and stability of mutant Inu proteins

All single mutants expressed well yielding about 4.5 to 6.0 mg of purified protein l-1

of E. coli culture. However, the multiple mutants GM2 and GM4 expressed poorly, yielding

about 0.1 to 0.2 mg of purified protein l-1 of culture. All single and multiple mutants were

active and retained over 85% of their activity even after incubation at 37 ºC for 24 h.

3.2. Characterization of mutant Inu enzymes harboring single mutation

Although a few single mutants behaved similar to WT Inu, possibly due to their

remote locations, several mutations affected the catalytic activity (Table 3) and some even

altered the oligosaccharide patterns compared to the WT enzyme (Fig. 2); in the following

paragraphs their effects will be discussed “per region”. Residue numbers are for Lb. reuteri

121 inulosucrase, with the corresponding residue numbers in InuJ given between brackets.

Site-directed mutagenesis of Inu

141

Mutations T413K, G416E and A417N (414, 417 and 418) are located in loop 415-

423 (Fig. 1a) that in InuJ connects helix α4 to the A-strand of blade 3 (β12) (Chapter 4). This

loop lines the active site funnel from its rim to the deep pocket; the side chains of most of its

residues are in the interior of the enzyme or point away from the catalytic cavity, the only

exception being the N419 (420) side chain. The residue N419 was not selected for mutation

in this study, as it is highly conserved in both, the levan- and inulosucrases. In the 3D

structure of InuJ this loop displayed ambiguous electron density, possibly indicative of

flexibility (Chapter 4). Of the mutated residues, T413 (414) is at the back of the rim. Its side

chain is protruding into the solvent and likely does not interact with substrate or acceptor

molecules. The slightly increased total and hydrolytic activity of mutant T413K (Table 3)

thus must result from other, indirect effects. The same is probably true for A417N (418) for

which the side chain is directed towards the protein interior. The third mutation in this

region, G416E (417), is located at the top of the rim. The glutamate side chain would point

into the solvent and one could envisage it to be involved in acceptor binding at higher

subsites. However, the transglycosylation rate of this mutant is hardly changed; instead its

hydrolytic rate is increased more than twofold. Therefore, indirect effects may result in a

more favored water attack on the fructosyl-enzyme intermediate. The polar character of the

glutamate might be important in this light.

Mutation W486L (487) increased total activity; both hydrolysis and

transglycosylation rates were increased. This residue is adjacent to the 415-423 loop

discussed above; a mutation at this position might influence the conformation of this loop

which lines the active site cavity. Moreover, the main chain carbonyl oxygen of W486 (487)

is one of the Ca2+ ligands. An effect of the mutation on Ca2+ binding, which is known to be

important for activity (18), might explain the changed activity but the exact mechanism for

this is not clear from the structure.

Another amino acid found to influence the catalytic properties of Inu is A538 (539).

The levansucrases of Lb. reuteri 121 and B. subtilis contain a serine or threonine at this

position, respectively. Swapping A538 of Inu with serine increased total activity 2.5 fold

with a concomitant 1.5 fold increase in transglycosylation/hydrolysis ratio. In the crystal

structures of InuJ and SacB this residue is located in the interior of the enzyme, just behind

the acid/base catalyst E342 (Fig. 1a & b). In fact, the hydroxyl group of the threonine in

Chapter 5

142

SacB (T357) makes a strong hydrogen bond (2.77 Å) to the main chain carbonyl oxygen of

the acid/base catalyst E342 (Fig. 1b). In InuJ, the alanine behind the acid-base catalyst

cannot make such a hydrogen bond (Fig. 1a), but in the A538S mutant this appears possible.

The electron-withdrawing effect of this hydrogen bond would lower the pKa of the acid/base

catalyst E523, making it easier for the proton to leave the carboxylate. This would explain

the increased activity of this mutant, but it does not explain how the

transglycosylation/hydrolysis ratio is increased, and why the wild type Lb. reuteri 121

levansucrase (bearing a serine) has a lower total activity than the inulosucrase (with an

alanine) (19).

Fig. 1. Positions of selected amino acid residues (light blue) in 3D structures of InuJ of Lb. johnsonii

NCC 533 (a) and SacB of B. subtilis (PDB code: 10YG) (b), targeted for mutagenesis in the Lb.

reuteri 121 Inu protein. The three catalytic residues are shown in magenta, two of the +1 and/or +2

residues green and hydrogen bonds yellow doted lines. These models were generated PyMOL

(DeLano Scientific LLC, California, U.S.A.)

a b

Site-directed mutagenesis of Inu

143

Table 3. FTF enzyme activities of Inu wild type (WT) and mutant enzymes measured at 37 ºC using

500 mM sucrose as substrate. All activity assays were done with 1.0 µg ml-1 enzyme, except for

GM4, where 3.2 µg ml-1 enzyme was used. Values shown are means (with standard deviations) of the

results obtained in two independent experiments.

GM1: I420Y, A421C, M422L GM2: N365K, T366L, A425P GM3: A425P, A538S, N543S, D548R, W551T GM4: T413K, K415R, G416E, A425P, S442N, W486L, P516L

Mutant Activity (U mg-1 of protein) Transglycosylation/

Total Hydrolytic Transglycosylation Hydrolysis

Inu (WT)

97 + 8.0

52 + 4.0

44 + 3.0

0.84

N365K 111 + 4.0 65 + 3.0 46 + 2.0 0.71

T366L 91 + 4.8 52 + 7.0 39 + 2.0 0.75

T413K 111 + 6.0 65 + 1.5 45 + 2.0 0.70

K415R 102 + 6.0 56 + 1.8 46 + 4.0 0.82

G416E 170 + 0.7 119 + 0.3 51 + 0.4 0.43

A417N 100 + 1.5 62 + 1.0 38 + 0.4 0.61

A425P 91 + 3.0 49 + 1.0 42 + 2.5 0.87

S442N 87 + 5.0 46 + 3.0 41 + 5.0 0.89

W486L 136 + 7.0 75 + 3.5 61 + 6.0 0.82

P516L 97 + 5.0 50 + 2.0 47 + 3.0 0.89

A538S 233 + 0.7 102 + 2.0 131 + 3.0 1.3

A489G 89 + 1.8 50 + 1.8 39 + 3.0 0.78

N543S 119 + 6.0 77 + 4.0 42 + 6.0 0.55

D548R 121 + 5.0 68 + 4.5 49 + 5.0 0.72

W551T 108 + 7.0 52 + 6.0 57 + 1.5 1.1

GM1 91 + 6.0 51 + 3.0 40 + 3.0 0.78

GM2 53 + 3.0 33 + 2.0 20 +1.0 0.67

GM3 35 + 0.8 27 + 1.7 8 + 0.8 0.30

GM4 6 + 0.01 3.5 + 0.3 2.5 + 0.3 0.70

Chapter 5

144

Fig. 2. HPAEC analysis of the FOS products synthesized by the Lb. reuteri 121 Inu (WT) and its

mutants N543S, GM3 and GM4 (a). The X- and Y-axis scales of the same chromatograms have been

expanded/compressed to better visualize their initial parts (b). Gluc: glucose; Fru: Fructose, Suc:

sucrose; GF2: 1-kestose.

a b

Site-directed mutagenesis of Inu

145

Three other Lb. reuteri 121 Inu mutations, N543S, D548R and W551T (544, 549 and

552), are located in the 310-helix (η8) and the unique helix α7, following β18 (strand B of

blade 4) in InuJ (Chapter 4). This region also contains the two arginine residues R541 and

R544 (542 and 545), both of which point into the active site cavity. In SacB the

corresponding R360 and K363 also line the active site cavity and are located near the +1

and/or +2 subsites (15,16).

Mutant N543S (544) exhibited a 1.5-fold higher hydrolytic activity (Table 3). It

synthesized only short FOS upto DP6, whereas the wild type enzyme makes FOS upto DP18

(Fig. 2). In the inulosucrases, arginine R544 (545) would be stabilized by hydrogen bonds

with N543 (544) and E521 (522) (Fig. 1a). Replacing N543 (544) with a serine would at

least remove one of these hydrogen bonds, and could affect the orientation of the arginine,

indirectly affecting the affinity for acceptor molecules and consequently the product

spectrum.

Together with N543 (544), the other two mutated residues in this region, D548 (549)

and W551 (552), located on the unique α7 helix, are involved in a hydrogen bond network

that also involves the side chains of residues R483 (484) and D479 (480), and G545 (546)

(Chapter 4). This network might contribute to the stability of the residues connecting β18

and α7 and to the orientation of R544 (545). Mutant W551T (552) displayed a somewhat

higher transglycosylation activity, while for D548R (549) the hydrolytic activity is slightly

increased. Since the region containing these mutation positions is not conserved between

inulosucrases and levansucrases, it is difficult to explain the effects of these mutations.

All recombinant enzymes with single point mutations produced inulin from sucrose,

as identified by comparing their 13C NMR spectra with that of the polymer product produced

by Inu WT enzyme (Fig. 3a & b), indicating that the targeted amino acids have no role, on

their own, in determining the linkage type in the product.

3.3. Characterization of the mutant proteins carrying multiple mutations

The motif IAM (amino acids 420-422 in Lb. reuteri 121 Inu) is highly conserved in

inulosucrases, while levansucrases of lactic acid bacteria contain variable residues, mostly

YCL/HTL/FTL (Table 1). This motif is located in the same loop as the single mutants T413

(414), G416 (417) and A417 (418) and precedes the conserved RDP or RDA motif harboring

Chapter 5

146

the transition state stabilizing residue D424 (425) and the arginine R423 (434) that makes a

salt bridge to the general acid/base residue (Fig. 1a & b). None of the side chains of the

mutated residues (GM1) would point into the active site cavity; occluded by R423 (424),

they cannot interact with donor or acceptor molecules. However, for two of the three

residues (I�Y and A�C) a larger side chain has to be accommodated in the mutant, and a

structural rearrangement of this region would be necessary to avoid clashes. Therefore it is

somewhat unexpected that replacing the amino acids had virtually no effect on the activity

and product profile.

Fig. 3. 13C NMR spectrum of inulin produced by (a) Inu WT (b) Inu mutant N365K recorded at 125

MHz. Peak position (ppm) for each carbon is given in parenthesis. All the remaining single and

multiple mutants (except GM4, which did not produce polymer) yielded similar (inulin type) spectra.

a

b

Site-directed mutagenesis of Inu

147

The multiple mutant GM1 behaved similar to the WT enzyme with respect to product

specificity and activity (Table 3). For mutants GM2, GM3 and GM4, polymer products

could only be obtained for GM2 and GM3 mutants and they produced inulin as revealed by 13C NMR spectroscopy. No GM4 mutant polymer products could be obtained, not even after

6 weeks of incubation, possibly due to its very low activity (Table 3); only 1-kestose could

be detected as product synthesized by this mutant (Fig. 2). This kind of effect was also

observed when W340 in Inu was replaced with an asparagine (19).

The GM2 mutant involves two residues, N365 (366) and T366 (367), located in the

loop connecting β-strands B and C of blade 2 at the rim of the active site funnel in InuJ

(Chapter 4), and one residue after the transition state stabilizing residue D424,which is an

alanine (A425 (426)) in the inulosucrases but a proline in levansucrases (Table 1). The GM2

mutations lowered total activity to 55% of wild type activity, and slightly lowered the

transglycosylation/hydrolysis ratio (Table 3). The lowered activity is likely to be an effect of

the alanine to proline mutation, which will affect the main chain conformation of this region,

including that of the neighbouring transition state stabilizer. However, since the single

A425P mutation only mildly lowered the activity, this effect must result from the

combination with the two other mutations in this group.

Also the GM3 mutant has low total activity (36% of wild type) (Table 3), and

includes A425P (426). Three other mutations in this group, N543S (544), D548R (549) and

W551T (552) involve the region after β18 (Chapter 4) which had also been mutated

individually, with different effects. Finally, this multiple mutant includes A538S (539),

which on its own increased total activity. Apparently the combination with the other

mutations in this group cancels out this effect, resulting in a decreased total activity. The

product spectrum of the GM3 mutant resembles mostly that of the single N543S (544)

mutation (Fig. 2), designating the importance of this residue for the relative

transglycosylation activity and product spectrum.

The multiple mutant GM4 has the lowest total activity: 6% of that of the wild type,

with a slightly decreased transglycosylation/hydrolysis ratio (Table 3), even though the

individual mutations, except G416E, have little effect on activity and the product spectrum.

Some of the mutations in this group mutant (T413K, K415R, G416E) are located in a loop at

the rim of the active site funnel or adjacent to it (W486L), while the others (S442N, P516L,

Chapter 5

148

A425P) are buried in the interior of the enzyme (Fig. 1a). A combination of these mutations

apparently affects the activity drastically, but the individual contribution of mutated residues

remains to be elucidated. No polymer product was obtained for this mutant, therefore we

could not analyze whether the polymer linkage type had been affected by the mutations.

4. Conclusions

Levansucrase enzymes have been characterized biochemically and structurally from

a range of Gram-positive and Gram-negative bacteria. Only recently, amino acid sequence

information and 3D structural information has become available for well-characterized

inulosucrases enzymes from various probiotic lactobacilli (Anwar 2008, Anwar 2010,

Chapter 4). This allowed identification of residues fully conserved in inulosucrase proteins,

but differing in levansucrase proteins. Site-directed mutagenesis of conserved inulosucrase

residues in the Lb. reuteri 121 Inu protein, followed by purification and biochemical

characterization of the 19 mutant proteins, revealed that several of these residues have clear

functional roles in inulosucrase reaction and product specificity. The most pronounced

effects were exhibited by mutants G416E and GM4, with single and multiple mutations,

respectively, in residues of loop 415-423, present at the rim of the active site pocket. Of

major interest is also residue A538, located behind the general acid/base. Inu mutation

A538S resulted in a 2-3 fold increase in enzyme activity, and an 1.5 fold increase in

transglycosylation/hydrolysis ratio. Recently, we observed that InuJ contains a unique helix

α7, which is not present in levansucrases SacB and LsdA (Chapter 4). Here we demonstrate

that the amino acid residues located in blade 4, including the unique helix α7, have a clear

effect on transglycosylation activity and on the FOS product spectrum. In particular,

mutation of N543, a residue hydrogen bonding to the adjacent arginine R544, which is near

subsites +1 and/or +2, influences the length of the oligosaccharides produced. The results

provide a better fundamental understanding of differences between

inulosucrase/levansucrase enzymes, and are important for the construction of inulosucrase

mutants with higher transglycosylation specificity, higher catalytic rates, and different

FOS/polymer size distribution.

Site-directed mutagenesis of Inu

149

Acknowledgements

We thank Peter Sanders for HPAEC analysis, Wieger Eeuwema for his technical assistance

in constructing the IAM mutant, and Pieter van der Meulen (Department of Biophysical

Chemistry) for his technical help in NMR spectroscopy.

Chapter 5

150

References

1. Anwar, M. A., S. Kralj, A. V. Pique, H. Leemhuis, M. J. Van der Maarel, and L. Dijkhuizen. 2010. Inulin and levan synthesis by probiotic Lactobacillus gasseri strains: characterization of three novel fructansucrase enzymes and their fructan products. Microbiology (UK) 156:1264-1274.

2. Anwar, M. A., S. Kralj, M. J. Van der Maarel, and L . Dijkhuizen. 2008. The probiotic Lactobacillus johnsonii NCC 533 produces high-molecular-mass inulin from sucrose by using an inulosucrase enzyme. Appl. Environ. Microbiol. 74:3426-3433.

3. Batista, F. R., L. Hernández, J. R. Fernandez, J. Arrieta, C. Menéndez, R. Gomez, Y. Tambara, and T. Pons. 1999. Substitution of Asp-309 by Asn in the Arg-Asp-Pro (RDP) motif of Acetobacter diazotrophicus levansucrase affects sucrose hydrolysis, but not enzyme specificity. Biochem. J. 337:503-506.

4. Cantarel, B. L., P. M. Coutinho, C. Rancurel, T. Bernard, V. Lombard, and B. Henrissat. 2009. The Carbohydrate-Active EnZymes database (CAZy): an expert resource for Glycogenomics. Nucleic Acids Res. 37:D233-D238.

5. Chambert, R. and G. Gonzy-Treboul. 1976. Levansucrase of Bacillus subtilis. Characterization of a stabilized fructosyl-enzyme complex and identification of an aspartyl residue as the binding site of the fructosyl group. Eur. J. Biochem. 71:493-508.

6. Chambert, R. and M. F. Petit-Glatron. 1991. Polymerase and hydrolase activities of Bacillus subtilis levansucrase can be separately modulated by site-directed mutagenesis. Biochem. J. 279:35-41.

7. Chambert, R., G. Treboul, and R. Dedonder. 1974. Kinetic studies of levansucrase of Bacillus subtilis. Eur. J. Biochem. 41:285-300.

8. Chuankhayan, P., C. Y. Hsieh, Y. C. Huang, Y. Y. Hsieh, H. H. Guan, Y. C. Hsieh, Y. C. Tien, C. D. Chen, C. M. Chiang, and C. J. Chen. 2010. Crystal structures of Aspergillus japonicus fructosyltransferase complex with donor/acceptor substrates reveal complete subsites in the active site for catalysis. J. Biol. Chem.

9. Davies, G. J., K. S. Wilson, and B. Henrissat. 1997. Nomenclature for sugar-binding subsites in glycosyl hydrolases. Biochem. J. 321:557-559.

10. Hernández, L., J. Arrieta, C. Menéndez, R. Vazquez, A. Coego, V. Suarez, G. Selman, M. F. Petit-Glatron, and R. Chambert. 1995. Isolation and enzymic properties of levansucrase secreted by Acetobacter diazotrophicus SRT4, a bacterium associated with sugar cane. Biochem. J. 309:113-118.

Site-directed mutagenesis of Inu

151

11. Homann, A., R. Biedendieck, S. Gotze, D. Jahn, and J. Seibel. 2007. Insights into polymer versus oligosaccharide synthesis: mutagenesis and mechanistic studies of a novel levansucrase from Bacillus megaterium. Biochem. J. 407:189-198.

12. KOSHLAND, D. E., Jr. and S. S. STEIN. 1954. Correlation of bond breaking with enzyme specificity; cleavage point of invertase. J. Biol. Chem. 208:139-148.

13. Lammens, W., R. K. Le, L. Schroeven, L. A. Van, A. Rabijns, and E. W. Van den. 2009. Structural insights into glycoside hydrolase family 32 and 68 enzymes: functional implications. J. Exp. Bot. 60:727-740.

14. Martinez-Fleites, C., M. Ortiz-Lombardia, T. Pons, N. Tarbouriech, E. J. Taylor, J. G. Arrieta, L. Hernandez, and G. J. Davies. 2005. Crystal structure of levansucrase from the Gram-negative bacterium Gluconacetobacter diazotrophicus. Biochem. J. 390:19-27.

15. Meng, G. and K. Futterer. 2003. Structural framework of fructosyl transfer in Bacillus subtilis levansucrase. Nat. Struct. Biol. 10:935-941.

16. Meng, G. and K. Futterer. 2008. Donor substrate recognition in the raffinose-bound E342A mutant of fructosyltransferase Bacillus subtilis levansucrase. BMC. Struct. Biol. 8:16.

17. Ortiz-Soto, M. E., M. Rivera, E. Rudino-Pinera, C. Olvera, and A. Lopez-Munguia. 2008. Selected mutations in Bacillus subtilis levansucrase semi-conserved regions affecting its biochemical properties. Protein Eng Des Sel 21:589-595.

18. Ozimek, L. K., G. J. Euverink, M. J. Van der Maarel, and L. Dijkhuizen. 2005. Mutational analysis of the role of calcium ions in the Lactobacillus reuteri strain 121 fructosyltransferase (levansucrase and inulosucrase) enzymes. FEBS Lett. 579:1124-1128.

19. Ozimek, L. K., S. Kralj, T. Kaper, M. J. Van der Maarel, and L. Dijkhuizen. 2006. Single amino acid residue changes in subsite -1 of inulosucrase from Lactobacillus reuteri 121 strongly influence the size of products synthesized. FEBS J. 273:4104-4113.

20. Ozimek, L. K., S. Kralj, M. J. Van der Maarel, and L. Dijkhuizen. 2006. The levansucrase and inulosucrase enzymes of Lactobacillus reuteri 121 catalyse processive and non-processive transglycosylation reactions. Microbiology 152:1187-1196.

21. Ozimek, L. K., S. A. van Hijum, G. A. van Koningsveld, M. J. Van der Maarel, G. H. Van Geel-Schutten, and L. Dijkhuizen. 2004. Site-directed mutagenesis study of the three catalytic residues of the fructosyltransferases of Lactobacillus reuteri 121. FEBS Lett. 560:131-133.

Chapter 5

152

22. Pons, T. and W. Van den Ende. 2010. "Glycoside Hydrolase Family 68" in CAZypedia, available at URL http://www.cazypedia.org/, accessed 10 August 2010.

23. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual. Cold Spring Harbour Laboratory Press, New York.

24. Song, D. D. and N. A. Jacques. 1999. Mutation of aspartic acid residues in the fructosyltransferase of Streptococcus salivarius ATCC 25975. Biochem. J. 344:259-264.

25. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res. 22:4673-4680.

26. Van Geel-Schutten, G. H., E. J. Faber, E. Smit, K. Bonting, M. R. Smith, B. Ten Brink, J. P. Kamerling, J. F. Vliegenthart, and L. Dijkhuizen. 1999. Biochemical and structural characterization of the glucan and fructan exopolysaccharides synthesized by the Lactobacillus reuteri wild-type strain and by mutant strains. Appl. Environ. Microbiol. 65:3008-3014.

27. Van Hijum, S. A. F. T., K. Bonting, M. J. E. C. Van der Maarel, and L. Dijkhuizen. 2001. Purification of a novel fructosyltransferase from Lactobacillus reuteri strain 121 and characterization of the levan produced. FEMS Microbiol. Lett. 205:323-328.

28. Van Hijum, S. A. F. T., G. H. Van Geel-Schutten, H. Rahaoui, M. J. Van der Maarel, and L. Dijkhuizen. 2002. Characterization of a novel fructosyltransferase from Lactobacillus reuteri that synthesizes high-molecular-weight inulin and inulin oligosaccharides. Appl. Environ. Microbiol. 68:4390-4398.

29. Waldherr, F. W., D. Meissner, and R. F. Vogel. 2008. Genetic and functional characterization of Lactobacillus panis levansucrase. Arch. Microbiol. 190:497-505.

30. Yanase, H., M. Maeda, E. Hagiwara, H. Yagi, K. Taniguchi, and K. Okamoto. 2002. Identification of functionally important amino acid residues in Zymomonas mobilis levansucrase. J. Biochem. (Tokyo) 132:565-572.