This article was published in an Elsevier journal. The attached copyis furnished to the author for non-commercial research and

education use, including for instruction at the author’s institution,sharing with colleagues and providing to institution administration.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

Author's personal copy

Hansenula polymorpha maltase gene promoter with sigma 70-like

elements is feasible for Escherichia coli-based biotechnological

applications: Expression of three genomic levansucrase

genes of Pseudomonas syringae pv. tomato

Triinu Visnapuu, Andres Mae, Tiina Alamae *

Institute of Molecular and Cell Biology, University of Tartu, Riia 23, 51010 Tartu, Estonia

Received 21 May 2007; received in revised form 14 November 2007; accepted 7 January 2008

Abstract

PMAL1, the sigma 70-like sequences possessing promoter of the maltase gene of Hansenula polymorpha, was evaluated for its application in

heterologous protein production in Escherichia coli. Levansucrases Lsc1, Lsc2 and Lsc3 of Pseudomonas syringae pv. tomato DC3000 were

expressed from PMAL1 in E. coli as biotechnologically relevant model proteins. Production of soluble levansucrases with high specific activity

confirmed appropriate strength of PMAL1 in a prokaryotic host. As about 90% of levansucrase activity was present in the cytoplasm of E. coli, no

levan-synthesis related sucrose intolerance of bacteria was observed. All three levansucrases hydrolyzed and polymerized both, sucrose and

raffinose. The raffinose-related activity of levansucrases has not been previously described in P. syringae. The PMAL1 expression system was used

to produce Lsc3 protein in E. coli for purification. The purified levansucrase showed much higher affinity for sucrose cleavage (Km = 21 mM) than

the levansucrase from P. syringae pv. phaseolicola described in the literature with Km of 160 mM.

# 2008 Elsevier Ltd. All rights reserved.

Keywords: Heterologous protein expression; Sigma 70; Levansucrase; Pseudomonas syringae; Hansenula polymorpha; Yeast promoter

1. Introduction

In our previous work, we have described the MAL gene cluster

that is responsible for the utilization of maltose and sucrose in

Hansenula polymorpha [1,2]. Cloning of the maltase gene

(HpMAL1) of the above mentioned cluster showed that the

promoter (PMAL1) of this gene was also functional in Escherichia

coli, as high maltase activity in HpMAL1-possessing transfor-

mants was observed [3]. Since E. coli is the most popular host in

recombinant protein production [4], PMAL1 may find application

in biotechnology. To evaluate feasibility of PMAL1 for E. coli-

based biotechnological expression systems, levansucrases were

chosen as test proteins expressed from the promoter. Levansu-

crases (EC 2.4.1.10) belong to glycoside hydrolase family GH68

(http://www.cazy.org/fam/GH68.html). They polymerize fruc-

tose residues of sucrose to b-(2,6)-linked fructan polymers

(levans) and are considered as proteins of biotechnological

interest, because (i) levan polymer has antitumor and anti-obesity

activities [5,6]; (ii) levan may be used as emulsifier or

encapsulating agent in industry [7]; (iii) besides levan, several

levansucrases also produce a considerable amount of prebiotic

fructo-oligosaccharides [8–12]. Levansucrases are present in

several bacteria with those of Gluconacetobacter diazotrophicus

[13], Zymomonas mobilis [14] and Bacillus subtilis [15] studied

most thoroughly. Bacterial genomes usually encode a single

levansucrase protein whereas a plant pathogen Pseudomonas

syringae pv. glycinea has three levansucrase genes [16]. Some of

these loci may encode proteins with specific functions and novel

properties. Up to now, levansucrases of P. syringae have been

mostly studied from the aspect of bacterial physiology and plant

pathogenesis [16,17], and only one levansucrase protein of this

species has been purified and thoroughly characterized [18].

Here, we tested feasibility of PMAL1 for biotechnological

applications—all three levansucrases encoded in the genome of

P. syringae pv. tomato DC3000 were expressed from it in E.

coli. The region in PMAL1 responsible for its expression in E.

coli was located using a reporter gene assay. Substrate

www.elsevier.com/locate/procbio

Process Biochemistry 43 (2008) 414–422

* Corresponding author. Tel.: +372 7 375013; fax: +372 7 420286.

E-mail address: talamae@ebc.ee (T. Alamae).

1359-5113/$ – see front matter # 2008 Elsevier Ltd. All rights reserved.

doi:10.1016/j.procbio.2008.01.002

Author's personal copy

specificity of Lsc proteins was studied and shown to be different

from that of P. syringae pv. phaseolicola levansucrase.

2. Materials and methods

2.1. Strains and plasmids

P. syringae pv. tomato DC3000 (P. syringae DC3000) was provided by Dr.

A. Tover (University of Tartu, Estonia). E. coli strains used in this study were

HB101 (lac+), a spontaneous LacY1 revertant of E. coli HB101 (supE44

hsdS20 (rB- mB-) recA13 ara-14 proA2 leuB6 rpsL20 xyl-5 mtl-1 galK2 lacY1)

[19], RA11r (DlacZY melA recA) [20,21] kindly donated by Dr. J.-C. Piard (Inra,

France), and DH5a (Stratagene, CA, USA). pBluescript SK (Stratagene) was

used as a source of b-lactamase gene in cell fractionation. p51 [3] contains the

maltase gene HpMAL1 and its full-length promoter in pYT3 [22]. Whereas in

p51MunI and p51HindIII-MunI the promoter was truncated till MunI and

HindIII site, respectively. pHIPX8-p51SpeI-SmaI has the maltase gene behind

its full-length promoter in pHIPX8 [1]. pX4MAL1 and pHIPX8MAL1 contain

the maltase reporter gene behind either the methanol oxidase gene promoter in

pX4-HNBESX [1] or the TEF2 gene promoter in pHIPX8. TEF2 (AY179869)

codes for translation elongation factor 1-alpha.

2.2. Construction of lsc expression vectors and transformation

The expression vector pHIPMalprom harboring the maltase gene promoter

(PMAL1) was designed as follows: PMAL1 was amplified from pRS425-p51SpeI-

SmaI [1] with primers MalpromBamHI (50 ctaggatccggaattaatattgtcaagagg 30)and Reverse (50 caggaaacagctatgac 30), and inserted after cleavage with BamHI

and NotI into pHIPX8, thus replacing the TEF2 promoter by PMAL1.

The coding regions of lsc1, lsc2 and lsc3 with preceding Shine-Dalgarno

sequences and �100 bp of 30-noncoding sequences were PCR-amplified from

lysed cells of P. syringae DC3000 using a Pfu DNA polymerase (Fermentas,

Lithuania) according to manufacturer’s recommendations. Lsc-specific forward

(Fw) and reverse (Rev) primers (BamHI and SalI restriction sites used for cloning

underlined) were as follows: Lsc1FwBam (50 tcgctgctggatcccaaagc 30); Lsc1Rev-

Sal (50 atagtcgaccatagccgagcagattatcg 30); Lsc2Fw (50 ttgtacaggatccaataaagg 30);Lsc2Rev (50 gctgatgcgtcgacaggcg 30); Lsc3FwBam (50 tcgctgctggatcccacaac 30);Lsc3RevSal (50 ttagtcgactgacggccagtcaagattc 30). The amplification products of

lsc1 and lsc3 were cleaved with BamHI and SalI and inserted between the same

sites of pHIPMalprom. The SmaI restriction site of pHIPMalprom was used for

lsc2 insertion. The resulting plasmids were designated pHIPMalprom-lsc1,

pHIPMalprom-lsc2 and pHIPMalprom-lsc3, respectively. Standard methods of

DNA manipulation were used [23]. DNA fragments further used for cloning were

gel-purified using Ultra CleanTM 15 kit (Mo Bio, CA, USA). Restriction endo-

nuclease digestions and DNA ligations were performed according to the man-

ufacturer’s (Fermentas) recommendations. Bacterial plasmid DNA was purified

using Perfectprep plasmid minikits (Eppendorf, Germany). E. coli was trans-

formed with plasmid DNA essentially as described in [3].

2.3. Growth media and cultivation of bacteria

Bacteria were routinely grown in LB medium supplied with antibiotics

(kanamycin, 50 mg/ml or ampicillin, 100 mg/ml) to maintain the plasmids. LB-

kanamycin agar containing 10% of either sucrose or raffinose was used to detect

levan formation by recombinant bacteria. Growth temperature was 37 8C, if not

indicated otherwise. Liquid cultures were aerated on a shaker. Growth of lsc3-

expressing E. coli on sucrose (0.5%) and glucose (0.5%) was inspected in liquid

minimal M9 medium [24] supplemented with 10 mg/ml thiamine, 100 mg/ml L-

leucine and 50 mg/ml kanamycin. The media were inoculated from overnight

cultures of LB-grown cells at initial optical density of culture at 600 nm (A600)

�0.1 and growth curves were obtained.

2.4. Fractionation of cells and preparation of extracts

Periplasmic and cytoplasmic fractions of recombinant E. coli were prepared

using osmotic shock treatment [25], but sucrose was replaced with sorbitol to

eliminate interference with levansucrase assay. A periplasmic marker (b-

lactamase) for the assay was provided by fractionation of a 1:1 mixture of

two different recombinant cultures of E. coli HB101 (lac+). The cultures carried

pHIPMalprom-lsc3 encoding the Lsc3 protein and pBluescript SK encoding the

E. coli b-lactamase, respectively. The cellular fractions were assayed for the

activities of levansucrase, b-lactamase and glucose-6-phosphate dehydrogen-

ase, the latter serving as a cytosolic marker.

For preparation of cell extract (it was also used as a fraction of soluble cell

proteins), the bacteria were harvested by centrifugation, washed twice in

appropriate buffer, sonicated, and the supernatant from centrifugation

(20 min, 13 700 g, 4 8C) was used as cell extract. The resulting pellet was

treated according to Sunitha et al. [26] to isolate the fraction of insoluble

proteins.

2.5. Enzyme assay

Levansucrase activity on sucrose and raffinose was determined at 37 8C by

monitoring the release of glucose or reducing sugar, respectively. Reaction

mixtures (1 ml) contained McIllvaine’s buffer (pH 6.0) with 0.015% Na-azide,

sucrose or raffinose, and appropriately diluted enzyme preparation. If glucose

release was measured, samples (50 ml) withdrawn at regular intervals were

mixed with 150 ml of 200 mM Tris-buffer (pH 8.3), and heated at 96 8C for

5 min. Glucose content of the samples was determined with Glucose Liquicolor

kit (Human GmbH, Germany) according to manufacturer’s instructions.

Release of reducing sugar was quantified using a 3,5-dinitrosalicylic acid

(DNS) method as follows. Samples (200 ml) withdrawn from the levansucrase

assay mixture were combined with 400 ml of DNS reactive [27], heated at

100 8C for 5 min, cooled on ice, 800 ml of distilled water was added and

absorption was measured at 540 nm.

Kinetic parameters of the purified Lsc3 protein for sucrose hydrolysis were

obtained by monitoring the release of glucose from 5 to 300 mM sucrose in the

absence and presence of 100 mM raffinose that was used as an inhibitor. The Km

of Lsc3 for raffinose was assayed similarly, but release of reducing sugar from

raffinose was measured in a DNS assay. The initial velocities of the reactions

were fitted with Enzyme Kinetics Module 1.1 of the Sigma Plot 2001 (SYSTAT)

according to Michaelis-Menten equation.

Levansucrase activity was expressed in mmoles of glucose (or reducing

sugar in case of DNS assay) produced per min per mg of protein (U/mg).

Maltase activity was measured with a chromogenic substrate p-nitrophenyl-a-

D-glucopyranoside [3]. b-Lactamase was assayed using benzylpenicillin [28],

glucose-6-phosphate dehydrogenase as in Clifton et al. [29], and protein

concentration according to Lowry.

2.6. Thin layer chromatography

Thin layer chromatography (TLC) was performed on silica gel 60 plates

(Merck, Germany). The levansucrase reaction with purified Lsc protein (�5 mg

protein in 1 ml reaction mixture) was conducted as described in Section 2.5 at

varied concentrations (10, 20, 50 and 100 mM) of sucrose or raffinose. At

selected time points, the reaction was stopped by heating and aliquotes (1 ml)

were spotted onto a TLC plate alongside with 0.5 ml of 0.1 M reference sugars.

The chromatogram was developed twice with a solvent system of chloroform–

methanol–water (90:65:15, v/v/v) [30]. The sugars were detected by dipping the

plate into a solution of 3% (w/v) urea, 1 M phosphoric acid in water-saturated

butanol and subsequent heating at 120 8C for 10 min as in [31].

2.7. Polyacrylamide gel electrophoresis

Polyacrylamide gel electrophoresis (PAGE) was conducted as described in

[23]. Relative amount of Lsc3 protein among soluble proteins of recombinant E.

coli was evaluated in a denaturating PAGE assay. For that, varied amounts of

cell extract protein from E. coli HB101 (lac+) carrying pHIPMalprom-lsc3

were electrophoresed alongside with different quantities of protein size marker

PageRulerTM SM0671 (Fermentas). The intensity of the levansucrase band was

compared with those of the PageRuler after staining of the gel with Coomassie

brilliant blue. Denaturating PAGE was also used for screening of soluble and

insoluble protein fractions of E. coli carrying pHIPMalprom-lsc3 for the

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422 415

Author's personal copy

presence of levansucrase band. Respective fractions of E. coli transformed with

the empty vector, were used as a control.

To visualize the activity of Lsc1, Lsc2 and Lsc3 proteins with sucrose and

raffinose, cell extracts of E. coli transformants producing these proteins were

electrophoresed in three parallel gels. One gel was stained for protein with

Coomassie brilliant blue, other two were incubated in McIllvaine’s buffer with

10% sucrose or 10% raffinose to identify levan-synthesising bands.

2.8. Protein purification

Lsc3 protein was purified from sonicated extracts of E. coli HB101 (lac+)

harboring pHIPMalprom-lsc3, using precipitation of proteins with (NH4)2SO4

and subsequent size-exclusion cromathography on a Sephacryl S-300 column

[32]. A detailed procedure of protein purification as well as a thorough

characterization of the protein will be published elsewhere.

2.9. Computer databases and programs used

CAZY (http://www.cazy.org/fam/GH68.html), NCBI (http://www.ncbi

.nlm.nih.gov/) and P. syringae (http://www.pseudomonas-syringae.org/) data-

bases were used for analysis of the genomes and retrieving the sequences.

Proteins were aligned using Clustal W program [33]. Molecular weights of

levansucrase proteins were calculated using Compute pI/Mw tool provided by

the ExPASy Proteomics Server (http://www.expasy.org/). Presence of signal

peptide in the proteins was predicted with the Signal P 3.0 program (http://

www.cbs.dtu.dk/services/SignalP/) [34].

3. Results and discussion

3.1. Evaluation of the performance of yeast promoters in E.

coli

We have previously shown that E. coli transformants

harboring the H. polymorpha maltase gene (HpMAL1;

AL432586) with its native promoter (PMAL1) exhibited a high

maltase activity. Thus, PMAL1 was functional in a prokaryotic

host [3]. According to the literature, the promoters of some

other genes of methylotrophic yeasts (LEU2 of Candida

boidinii, URA3 of C. boidinii and H. polymorpha) also function

in E. coli, since these genes have been cloned by complemen-

tion of respective E. coli mutants [35–37]. In LEU2 promoter,

sigma 70-like hexamers were assumed to serve as a promoter in

E. coli, albeit no experimental proof was provided [36].

Interestingly, sigma 70-like boxes were also found in PMAL1

[3]. Here, we inspected PMAL1 and two reference promoters

(PMOX1 and PTEF2) of H. polymorpha for the expression in E.

coli to verify the significance of these sigma 70-like boxes. H.

polymorpha maltase gene was used as a reporter since E. coli

lacks endogenous maltase activity and a simple chromogenic

maltase assay is available [3]. It should be noted that PMOX1 (the

promoter of the methanol oxidase gene), is one of the strongest

promoters in yeasts. It is very highly induced in methanol-

growing H. polymorpha, and has widely been used for

recombinant protein production [38]. The reporter plasmids

were introduced into E. coli, and performance of the promoters

was evaluated according to maltase activity of the transfor-

mants. Table 1 shows that PTEF2 and PMOX1 did not function in

E. coli: maltase expression from these promoters was only

marginal. Consistent with experimental data, sequence analysis

of PTEF2 and PMOX1 revealed no sigma 70-like hexamers and

Shine-Dalgarno sequences. At the same time, expression of

maltase from PMAL1 in E. coli was high in both plasmid

backgrounds studied. As previously reported by us, PMAL1

possesses sigma 70-like �10 and �35 hexamers at positions

�310 to �282 ( ), and �213 to �185

( ) relative to the ATG codon of the

maltase gene [1], positions agreeing with the sigma 70

consensus TTGACA-N17-TATAAT [4] are highlighted. A

reporter assay of truncated versions of PMAL1 supported

functional significance of these sigma 70-like boxes for gene

expression in E. coli: shortening of PMAL1 till MunI cleavage

site at�315 bp retaining sigma 70-like elements, did not affect

maltase activity of the transformants, whereas further trunca-

tion till the HindIII cleavage site (eliminating the sigma 70-like

hexamers), strongly reduced the activity (Table 1). It should be

noted that similarly to PTEF2 and PMOX1, PMAL1 has no Shine-

Dalgarno like sequence.

3.2. P. syringae pathovars have multiple levansucrase

genes

We chose levansucrases as biotechnologically important test

proteins to evaluate PMAL1-directed expression system in E.

coli. According to the literature, P. syringae pv. glycinea has

Table 1

Expression from Hansenula polymorpha promoters in Escherichia colia studied with maltase gene as a reporter

Plasmid used for transformation Promoterb Sigma 70-like boxes in

the promoterc

Specific maltase

activity (U/mg)d

pHIPX8-p51SpeI-SmaI PMAL1 (full-length) Present 1.410 � 0.130

p51 PMAL1 (full-length) Present 1.332 � 0.051

p51MunI PMAL1 (shortened till the MunI site at �315 bp) Present 1.308 � 0.228

p51HindIII-MunI PMAL1 (shortened till HindIII site at �68 bp) Absent 0.108 � 0.018

pX4MAL1 PMOX1 Absent 0.017 � 0.001

pHIPX8MAL1 PTEF2 Absent 0.015 � 0.003

Two to three distinct transformants of E. coli harboring a test plasmid were grown in liquid LB-medium supplemented with antibiotic to late-exponential phase and

specific maltase activity (average � standard deviation) was measured in cell extracts.a Strain DH5a was used.b The reporter plasmids are described in Section 2.1.c Sigma 70-like hexamers in the HpMAL1 promoter are located at positions �310 to �282 and �213 to �185 upstream of the translational start codon of the

maltase gene [1].d One unit equals to one mmol of p-nitrophenyl-a-D-glucopyranoside hydrolyzed per min.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422416

Author's personal copy

lscA, lscB and lscC genes, the lscB residing on a plasmid [16].

Analysis of genomic data of P. syringae pv. tomato DC3000, P.

syringae pv. syringae B728a and P. syringae pv. phaseolicola

1448A at http://www.pseudomonas-syringae.org/showed that a

typical P. syringae pathovar possesses three levansucrase

alleles, two chromosomal and one plasmid-borne. P. syringae

pv. syringae B728a has no plasmids and its chromosome

encodes two levansucrase (Lsc) proteins. The length of

levansucrase proteins deduced from respective ORFs of P.

syringae pathovars is either 415 or 431 amino acids (aa) (see

Fig. 1), being much smaller than that of levansucrases of Gram-

positive bacteria Lactobacillus sanfranciscensis (879 aa) and L.

reuteri (804 aa) [39,40]. The 424 aa levansucrase of a Gram-

positive bacterium Leuconostoc mesenteroides could be

considered exceptional because of its relatively small size

and a high similarity to levansucrases from Gram-negative

bacteria [12]. Interestingly, the 431 and 415 aa levansucrases of

P. syringae differ mostly in their N-termini: the 431 aa proteins

have N-terminal extensions of 16 aa lacking in 415 aa proteins

(Fig. 1), otherwise they share a high (89–99%) level of identity

according to the Clustal W analysis. The role of N-terminal

extensions of 431 aa levansucrases is not known, but most

probably they do not function as secretion leaders (see Section

3.5). Up to now, only one levansucrase protein of P. syringae

has been biochemically characterized—Hettwer et al. [18] has

purified an extracellular levansucrase protein from P. syringae

pv. phaseolicola. According to the sequence fragment

AGNINYEPTVWSRADALKVN shown for this protein [18],

it should belong to a group of 431-aa levansucrases. The P.

syringae pv. phaseolicola levansucrase exhibited several

biotechnologically valuable properties, e.g. perfect stability

and toleration of denaturants. It cannot use raffinose as

substrate and is exceptional among levansucrases because of

very low affinity for sucrose cleavage (Km = 160 mM) [18].

3.3. Cloning and expression of levansucrase genes of P.

syringae pv. tomato DC3000

E. coli is a perfect host for cloning and expression of lsc

genes: the laboratory strains of E. coli do not metabolize

sucrose [41] and therefore the lsc-expressing transformants will

have a dominant sucrose-positive phenotype. The fact that

levansucrases of other pathovars of P. syringae were not

expressed from their own promoters in E. coli [16,42], justifies

application of a heterologous promoter, PMAL1 in our case. lsc1,

lsc2 and lsc3 genes of P. syringae pv. tomato DC3000 (P.

syringae DC3000) were cloned into pHIPMalprom under

control of PMAL1. E. coli HB101 (lac+), transformed with

resulting plasmids, pHIPMalprom-lsc1, pHIPMalprom-lsc2

and pHIPMalprom-lsc3, formed mucoid colonies on LB-

kanamycin plates containing 10% sucrose, revealing synthesis

of functional levansucrases whereas control transformants with

empty vector pHIPMalprom, had a nonmucoid phenotype. The

exopolysaccharide produced by lsc-expressing clones in LB-

sucrose medium, was hydrolyzed and subjected to biochemical

analysis. Expectedly, the polymer had a 99:1 ratio of fructose to

glucose typical for levan (data not shown).

Table 2 presents levansucrase activities due to the expression

of lsc1, lsc2 and lsc3 in two different strains of E. coli. The

highest levansucrase activity was recorded for lsc3 in E. coli

HB101 (lac+), and the lowest for lsc2 in E. coli RA11r.

Due to levansucrase activity, recombinant E. coli gained the

ability to grow on sucrose minimal medium. Fig. 2 shows that

the lsc3-expressing E. coli grows on sucrose and glucose

equally well given that mostly glucose moiety of sucrose can be

used for growth as fructose is partly built into the levan

polymer.

3.4. Levansucrases are prominent proteins in cell extracts

of recombinant E. coli

Proteins in cell extracts of lsc-expressing E. coli were

separated by non-denaturating PAGE and stained in parallel for

protein and levansucrase activity (Fig. 3). In approximately

30 min of incubation, whitish levan bands were clearly visible

in sucrose-immersed gel whereas synthesis of levan from

raffinose was slower, it was evident after about 90 min of

incubation with raffinose. Raffinose utilization by levansu-

crases will be addressed in a more detail in Section 3.6.

Comparison of the gels shows that levan-synthesizing bands

correspond to most prominent protein bands of the reference gel

shown in Fig. 3A. Lsc2 protein that is smallest among of three

Fig. 1. Alignment of N-termini of eleven levansucrase proteins predicted from genomic sequences of Pseudomonas syringae pv. tomato DC3000, pv. phaseolicola

1448A, pv. glycinea PG4180 and pv. syringae B728a. The protein sequences were retrieved from CAZY (http://www.cazy.org/fam/GH68.html) and NCBI (http://

www.ncbi.nlm.nih.gov/) databases. GenBank accession numbers, names and sizes of the proteins in amino acids (aa) are shown left to the alignment. Identity and

similarity between the proteins is designated by differential background shading. Genomic localization of respective lsc genes is designated as C for chromosomal and

P for plasmid-borne.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422 417

Author's personal copy

levansucrase proteins (415 aa, calculated molecular weight

45.8 kDa), migrated in the gel most rapidly. The calculated

molecular weights of Lsc1 and Lsc3 proteins (both 431 aa long)

were 47.6 and 47.7 kDa, respectively. The PAGE shows that

despite Lsc1 and Lsc3 have quite similar calculated sizes, they

had slightly different mobility.

Denaturating PAGE was run to evaluate relative amount of

Lsc3 protein among total soluble proteins of host bacteria (see

Section 2.7). Lsc3 was calculated to comprise about 20% of

total soluble proteins of E. coli. That level of expression is

considered optimal for E. coli production systems [4]. On the

other hand, a high expression level may result in production of

insoluble aggregated protein. For example, levansucrase of Z.

mobilis (LevU) was produced in E. coli at a high level (20% of

total cellular proteins) if a constitutive promoter of Rahnella

aquatilis was used and bacteria were grown in fed-batch

culture, but the protein accumulated as inclusion bodies and

required renaturation. In shake flask culture, the yield of

levansucrase was lower, but it was produced as active and

soluble protein [26]. A widely used strong and inducible

expression system based on T7 RNA polymerase also resulted

in rapid inactivation of LevU protein and formation of inclusion

bodies in E. coli [43].

To rule out the possibility of levansucrase inactivation in

case of our expression system, a denaturating PAGE of soluble

Table 2

Levansucrase activity of Escherichia coli strains HB101 (lac+) and RA11ra harboring the expression plasmids with lsc genes of P. syringae pv. tomato DC3000

E. coli strain Plasmidb Lsc gene expressed Specific levansucrase activityc (U/mg)

HB 101 (lac+) pHIPMalprom-lsc1 lsc1 56 � 12

pHIPMalprom-lsc2 lsc2 21 � 4

pHIPMalprom-lsc3 lsc3 70 � 25

pHIPMalprom No (empty vector) Below detection

RA11r pHIPMalprom-lsc1 lsc1 44 � 8

pHIPMalprom-lsc2 lsc2 7 � 2

pHIPMalprom-lsc3 lsc3 46 � 7

pHIPMalprom No (empty vector) Below detection

a E. coli strain RA11r is melibiase-negative.b Described in Section 2.2.c Levansucrase activity was measured according to the release of glucose from 300 mM sucrose in cell extracts of E. coli transformants grown in LB-kanamycin

medium to late-exponential growth phase, average activity �standard deviation for two to six transformants is presented.

Fig. 2. Lsc3-producing Escherichia coli gains the ability to grow on sucrose. E.

coli HB101 (lac+) harboring pHIPMalprom-lsc3 was grown in liquid M9

minimal medium supplemented with kanamycin, auxotrophic supplements and

either glucose or sucrose. The reference clone (Ref) carried the empty vector

pHIPMalprom.

Fig. 3. Non-denaturating polyacrylamide gel electrophoresis of Pseudomonas

syringae pv. tomato DC3000 levansucrase proteins. Cell extracts (�15 mg

protein per lane) of recombinant E. coli HB101 (lac+) (A and B) or RA11r

(C) expressing individual levansucrases were separated in 7.5% polyacrylamide

gel and stained with Coomassie brilliant blue (A) or soaked in McIllvaine’s

buffer (pH 6.0) containing 0.015% Na-azide and 10% sucrose (B) or 10%

raffinose (C). Only region with levan-synthesizing bands of gels B and C is

shown. Clones harboring the empty vector were used as a reference (Ref).

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422418

Author's personal copy

and insoluble protein fractions of lsc3-expressing E. coli was

run. Expectedly, levansucrase band was present only in the

fraction of soluble proteins (data not shown). As discussed in

[4], one way to avoid inactivation of the produced protein is

using a promoter with appropriate strength. We conclude that

PMAL1 has appropriate strength to produce an adequate amount

of soluble catalytically active levansucrase protein in E. coli.

3.5. Cellular location of levansucrase in E. coli

Bacterial levansucrases are extracellular proteins in their

native hosts [16,44]. Among levansucrases of Gram-negative

bacteria, secretion mechanism is known only for LsdA of G.

diazotrophicus. A 30-aa N-terminal signal peptide is cleaved

off from LsdA during the transport to the periplasm, where the

enzyme adapts its conformation, and thereafter it is transferred

across the outer membrane without further proteolytic

processing [44].

We analyzed the sequences of Lsc proteins of P. syringae

DC3000 for the presence of a potential N-terminal signal

peptide using a Signal P program version 3.0 [34]. No secretion

leader was predicted for Lsc1, Lsc2 and Lsc3 proteins in

contrast to LsdA of G. diazotrophicus that was analyzed as a

reference. According to Li and Ullrich [16] computer analysis

of levansucrases of P. syringae pv. glycinea PG4180 also

predicted lack of N-terminal signal peptide [16]. In addition,

the authors proved that levansucrases of PG4180 are not

proteolytically processed during the secretion: the N-terminal

amino acid sequence of levansucrase isoenzyme mixture from

the supernatant of PG4180 was STSSSAVSQLKNSPLAG-

NINY, identical to the N-terminal sequence deduced from the

nucleotide sequences of lscB and lscC of PG4180 (shown also

in Fig. 1). The authors conclude that export of the levansucrases

might therefore occur via a sec-independent mechanism. So,

the function of N-terminal extension of 431-aa Lsc proteins

(Fig. 1) has yet to be elucidated.

As growth media of lsc-expressing E. coli exhibited no

levansucrase activity, distribution of one of the expressed

levansucrases, Lsc3, between periplasmic and cytosolic

fractions of E. coli was studied. It turned out that most

(88%) of levansucrase activity was located in the cytoplasm.

Compartment-specific reference proteins segregated correctly:

majority of b-lactamase activity was detected in the periplasm

and that of glucose-6-phosphate dehydrogenase in the

cytoplasm (Table 3), proving reliability of cell fractionation

data. We assume that a low amount of levansucrase in the

periplasmic fraction of recombinant E. coli protects bacteria

from toxic effect of sucrose. Indeed, according to our data, lsc-

expressing E. coli exhibited undisturbed growth on sucrose

(Fig. 2). As concluded from the analysis of the literature data,

levan synthesis in a sucrose-growing Lsc-expressing E. coli can

take place only in the periplasm. Namely, sucrose diffuses into

the periplasm of E. coli through aspecific porins in the outer

membrane, but cannot penetrate further being therefore absent

from the cytoplasm [4,45]. Thus, cytosolic population of

levansucrase will not get into contact with sucrose and thereby

will not contribute to levan synthesis by the cells. This

hypothesis agrees with severe toxicity of the expression of B.

subtilis levansucrase (SacB) to E. coli growing in sucrose

medium. SacB has mostly periplasmic location in E. coli [46]

that should cause cell bursting due to synthesis of a high amount

of levan in the periplasm if sucrose is present in the growth

medium. In accordance with that, expression of Z. mobilis

levansucrase caused no sucrose-toxicity to E. coli, because

majority (73%) of the levansucrase was detected in the

cytoplasm [47]. We hypothesize that in case of moderate levan

synthesis in the periplasm, the polymer will somehow be

excreted to growth medium, probably through pores in the outer

membrane. We also suggest that in case of levan-synthesizing

E. coli colonies, some levan on the colonies is probably

produced by levansucrase released from dead cells.

3.6. All three levansucrases of P. syringae pv. tomato

DC3000 use raffinose

In addition to sucrose, levansucrases of many bacteria, e.g.

Z. mobilis [14,47–49], L. sanfranciscensis [40], L. reuteri [39]

B. subtilis [50] and Microbacterium laevaniformans [11] can

use a trisaccharide, raffinose. Raffinose utilization by

levansucrases of P. syringae DC3000 was first suspected from

colony phenotype of lsc-expressing E. coli RA11r on LB-

raffinose plates. Regular strains of E. coli possess melibiase (a-

galactosidase), splitting raffinose to galactose and sucrose,

whereas the RA11r strain is melibiase-negative [20,21] having

suitable background for levansucrase-mediated raffinose

utilization assay. The lsc-expressing RA11r colonies turned

mucoid during incubation on LB-kanamycin plates containing

either sucrose or raffinose, the slime production being

somewhat less pronounced in case of raffinose. Cell extracts

of lsc-expressing RA11r produced reducing sugars from both,

sucrose and raffinose, albeit with different relative activity:

hydrolysis of 100 mM raffinose was 41–48% from the

hydrolysis of 100 mM sucrose. The levansucrase activity assay

on native polyacrylamide gels (Fig. 3) showed levan production

from both, sucrose and raffinose.

One of the levansucrases, the Lsc3, was purified from

recombinant E. coli HB101 (lac+) (see Section 2.8) for kinetic

study of sucrose and raffinose utilization. The purification

increased specific levansucrase activity on sucrose by more

Table 3

Cellular localization of Lcs3 protein of Pseudomonas syringae pv. tomato

DC3000 expressed in Escherichia coli HB101 (lac+)

Enzyme Distribution of enzyme

activity units between the

cytoplasm and periplasm (%)a

Periplasm Cytoplasm

Levansucrase 12 � 2 88 � 2

b-Lactamase (a marker for

periplasmic localization)

96 � 2 4 � 2

Glucose-6-phosphate dehydrogenase

(a marker for cytosolic localization)

2 � 1 98 � 1

a Two different samples of two bacterial cultures were processed separately,

and average values � standard deviation are presented.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422 419

Author's personal copy

than four-fold, up to the Vmax value of 286.2 � 7.4 U/mg. This

final activity value agrees with the predicted Lsc3 expression

level in E. coli of about 20% from total soluble proteins (see

Section 3.4). The Km for sucrose hydrolysis was

20.6 � 2.1 mM and raffinose inhibited sucrose hydrolysis

competitively (Fig. 4) with Ki of 47.6 � 5.0 mM. Thus, sucrose

and raffinose probably share a binding site on the levansucrase.

The Km for raffinose hydrolysis measured according to the

liberation of reducing sugars was 41.3 � 5.6 mM. Release of

reducing sugars from 100 mM raffinose was 48% of the value

obtained with 100 mM sucrose.

Next, a thin layer chromatography assay of levansucrase

reaction products was carried out. The purified Lsc3 protein

was incubated with four different concentrations (10, 20, 50 and

100 mM) of sucrose and raffinose for 6 min, and the reaction

products were resolved on a silica gel. A fructose-specific urea–

phosphoric acid visualization method was used, thus glucose

and melibiose formed from sucrose and raffinose, respectively,

did not stain. The chromatogram shown in Fig. 5 displayed

different affinity of Lsc3 protein for sucrose and raffinose

hydrolysis: in case of reaction with 10 mM sucrose, a clearly

detected fructose spot is visible in a TLC plate, whereas

reaction with 10 mM raffinose produces no detectable amount

of fructose. Fructose liberation from raffinose was detected

only at higher raffinose concentrations. TLC analysis also

detects levan formation by Lsc3 at higher concentrations of

sucrose, while levan production from raffinose was not

observed at any raffinose concentration used in the assay

(Fig. 5). Nevertheless, chromatography of samples from

prolonged incubation (15 and 30 min) of Lsc3 with raffinose,

indicated levan production at elevated raffinose concentrations,

50 mM and higher (data not shown). Literature data also show

that levan synthesis by levansucrases is stimulated at high

substrate concentrations [18,40]. Consistent with TLC analysis

data, levan-producing activity of Lsc3 with sucrose is higher

than with raffinose. The TLC method used in our assay can also

resolve fructo-oligosaccharides [30]. As we were not able to

detect a fructose-containing spot with mobility of fructo-

oligosaccharides, they were probably not produced in a

detectable amount from neither sucrose nor raffinose by

Lsc3 under applied reaction conditions.

Levansucrase of Z. mobilis has been studied from the aspect of

sucrose and raffinose utilization. The Z. mobilis enzyme has

similar affinities for sucrose and raffinose [48,49], with capacity

of raffinose hydrolysis exceeding that of sucrose hydrolysis

[14,48] or being equal to that [49]. Unlike the Z. mobilis enzyme,

levansucrases of P. syringae DC3000 prefer sucrose over

raffinose. Interestingly, the only levansucrase protein purified

from a P. syringae species, from pv. phaseolicola [18] is

considered exceptional among bacterial levansucrase for its poor

affinity (Km = 160 mM) for sucrose hydrolysis. In contrast to

that, the Lsc3 protein of P. syringae DC3000 described by us, had

much higher affinity (Km = 21 mM) for sucrose hydrolysis, being

similar to respective value of levansucrases from many other

bacteria, e.g. Bacillus sp. TH4-2 (Km = 16.7 mM) [51], L.

sanfranciscensis (Km = 13.1 mM) [40], G. diazotrophicus

(Km = 11.8 mM) [8], L. mesenteroides (Km = 26.6 mM) [12]

and Erwinia herbicola (Km = 28 mM) [52].

4. Conclusions

PMAL1, the promoter of the H. polymorpha maltase gene

possessing sigma 70-like sequence elements was demonstrated

as suitable for the expression of biotechnologically relevant

proteins in E. coli. As an example, three levansucrase proteins

of P. syringae DC3000 were successfully produced in E. coli,

and one of the recombinant proteins, Lsc3, was purified. Due to

appropriate strength of the promoter, the Lsc1, Lsc2 and Lsc3

were produced in E. coli as soluble proteins in a high amount,

Fig. 4. Kinetics of sucrose hydrolysis by purified Lsc3 protein of P. syringae pv

tomato DC3000. Release of glucose from 5 to 300 mM sucrose was assayed in

McIllvaine’s buffer (pH 6.0) in the presence (*) and absence (*) of 100 mM

raffinose. Data were analyzed with Enzyme Kinetics Module 1.1 of the Sigma

Plot program and plotted according to Michaelis-Menten and Eadie-Hofstee

(insert). Average values and standard deviation for three parallel measurements

are shown. The kinetic parameters, calculated from the graphs, were as follows:

the Km for sucrose hydrolysis 20.6 � 2.1 mM, the Vmax for sucrose hydrolysis

286.2 � 7.4 U/mg, and the Ki for raffinose inhibition 47.6 � 5.0 mM.

Fig. 5. Thin layer chromatography of reaction products derived from sucrose

and raffinose by Lsc3 protein of P. syringae pv. tomato DC3000. Purified Lsc3

protein was incubated in McIllvaine’s buffer (pH 6.0) with different concentra-

tions of sucrose and raffinose for 6 min, the reaction was stopped and 1 ml

aliquotes of reaction mixtures were chromatographed on a silica gel plate.

0.5 ml of 0.1 M solutions of standard sugars (fructose, sucrose, raffinose) were

analyzed alongside. Fructose-containing sugars were visualized using urea–

phosphoric acid treatment. F, fructose; S, sucrose; R, raffinose.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422420

Author's personal copy

and formation of inclusion bodies was not observed. Kinetic

assay of Lsc3 showed that a very low affinity for sucrose

hydrolysis described for a levansucrase of P. syringae pv.

phaseolicola [18] is not a common feature for levansucrases of

that species: the affinity of Lsc3 of P. syringae DC3000 for

sucrose hydrolysis was considerably high and similar to that of

levansucrases from many other bacteria. We also demonstrated

that unlike the P. syringae pv. phaseolicola levansucrase [18], P.

syringae DC3000 Lsc proteins could split and polymerize

raffinose. The expression system worked out by us can be used

to produce levansucrase proteins for structure–function

analysis of wild-type and engineered levansucrases for both,

basic research and practical applications. PMAL1, applied in this

study, has additional attractive feature—besides E. coli, it is

functional also in yeasts, e.g. H. polymorpha and Sacchar-

omyces cerevisiae [1]. Thus, it can be used for the design of a

common expression plasmids for more than just one host

species, if plasmid replication and maintenance will be

guaranteed. In this case, production of a protein of interest

can be tested in different hosts to select the most appropriate

one.

Acknowledgements

This work was supported by the grant 5676 from Estonian

Science Foundation to T.A. and FEMS grant for young

scientists from 2006 to T.V. The authors wish to thank Drs. A.

Tover, R. Bruckner and J.-C. Piard for providing bacterial

strains and Dr. A. Vigants for help in protein purification.

References

[1] Alamae T, Parn P, Viigand K, Karp H. Regulation of the Hansenula

polymorpha maltase gene promoter in H. polymorpha and Saccharomyces

cerevisiae. FEMS Yeast Res 2003;4:165–73.

[2] Viigand K, Tammus K, Alamae T. Clustering of MAL genes in Hansenula

polymorpha: cloning of the maltose permease gene and expression from

the divergent intergenic region between the maltose permease and maltase

genes. FEMS Yeast Res 2005;5:1019–28.

[3] Liiv L, Parn P, Alamae T. Cloning of maltase gene from a methylotrophic

yeast Hansenula polymorpha. Gene 2001;265:77–85.

[4] Schumann W, Ferreira LCS. Production of recombinant proteins in

Escherichia coli. Genet Mol Biol 2004;27:422–53.

[5] Yoo SH, Yoon EJ, Cha J, Lee HG. Antitumor activity of levan poly-

saccharides from selected microorganisms. Int J Biol Macromol

2004;34:37–41.

[6] Kang SA, Hong K, Jang KH, Kim YY, Choue R, Lim Y. Altered mRNA

expression of hepatic lipogenic enzyme and PPARa in rats fed dietary

levan from Zymomonas mobilis. J Nutr Biochem 2006;17:419–26.

[7] Banguela A, Hernandez L. Fructans: from natural sources to transgenic

plants. Biotecnologia Aplicada 2006;23:201–10.

[8] Hernandez L, Arrieta J, Menendez C, Vazquez R, Coego A, Suarez V, et al.

Isolation and enzymic properties of levansucrase secreted by Acetobacter

diazotrophicus SRT4, a bacterium associated with sugar cane. Biochem J

1995;309:113–8.

[9] Euzenat O, Guibert A, Combes D. Production of fructo-oligosaccharides by

levansucrase from Bacillus subtilis C4. Process Biochem 1997;32:237–43.

[10] Bekers M, Laukevics J, Upite D, Kaminska E, Vigants A, Viesturs U, et al.

Fructooligosaccharide and levan producing activity of Zymomonas mobi-

lis extracellular levansucrase. Process Biochem 2002;38:701–6.

[11] Park HE, Park NH, Kim MJ, Lee TH, Lee HG, Yang JY, et al. Enzymatic

synthesis of fructosyl oligosaccharides by levansucrase from Microbac-

terium laevaniformans ATCC 15953. Enzyme Microb Technol 2003;

32:820–7.

[12] Kang HK, Seo MY, Seo ES, Kim D, Chung SY, Kimura A, et al. Cloning

and expression of levansucrase from Leuconostoc mesenteroides B-512

FMC in Escherichia coli. Biochim Biophys Acta 2005;1727:5–15.

[13] Martınez-Fleites C, Ortız-Lombardıa M, Pons T, Tarbouriech N, Taylor

EJ, Arrieta JG, et al. Crystal structure of levansucrase from the Gram-

negative bacterium Gluconacetobacter diazotrophicus. Biochem J

2005;390:19–27.

[14] Yanase H, Maeda M, Hagiwara E, Yagi H, Taniguchi K, Okamato K.

Identification of functionally important amino acid residues in Zymomo-

nas mobilis levansucrase. J Biochem 2002;132:565–72.

[15] Meng G, Futterer K. Structural framework of fructosyl transfer in Bacillus

subtilis levansucrase. Nat Struct Biol 2003;10:935–41.

[16] Li H, Ullrich MS. Characterization and mutational analysis of three allelic

lsc genes encoding levansucrase in Pseudomonas syringae. J Bacteriol

2001;183:3282–92.

[17] Li H, Schenk A, Srivastava A, Zhurina D, Ullrich MS. Thermo-responsive

expression and differential secretion of the extracellular enzyme levan-

sucrase in the plant pathogenic bacterium Pseudomonas syringae pv.

glycinea. FEMS Microbiol Lett 2006;265:178–85.

[18] Hettwer U, Gross M, Rudolph K. Purification and properties of an

extracellular levansucrase from Pseudomonas syringae pv. phaseolicola.

J Bacteriol 1995;177:2834–9.

[19] Bruckner R, Wagner E, Gotz F. Characterization of a sucrase gene from

Staphylococcus xylosus. J Bacteriol 1993;175:851–7.

[20] Hanatani M, Yazyu H, Shiota-Niiya S, Moriyama Y, Kanazawa H, Futai

M, et al. Physical and genetic characterization of the melibiose operon and

identification of the gene products in Escherichia coli. J Biol Chem

1984;259:1807–12.

[21] Silvestroni A, Connes C, Sesma F, De Giori GS, Piard JC. Characteriza-

tion of the melA locus for alpha-galactosidase in Lactobacillus plantarum.

Appl Environ Microbiol 2002;68:5464–71.

[22] Tan X, Waterham HR, Veenhuis M, Cregg JM. The Hansenula poly-

morpha PER8 gene encodes a novel peroxisomal integral membrane

protein involved in proliferation. J Cell Biol 1995;128:307–19.

[23] Sambrook J, Fritsch ET, Maniatis T. Molecular cloning: a laboratory

manual. New York: Cold Spring Harbor Laboratory Press; 1989.

[24] Adams MH. Bacteriophages. New York: Interscience Publishers; 1959.

[25] Liu Y, Fu X, Shen J, Zhang H, Hong W, Chang Z. Periplasmic proteins of

Escherichia coli are highly resistant to aggregation: reappraisal for roles of

molecular chaperones in periplasm. Biochem Biophys Res Commun

2004;316:795–801.

[26] Sunitha K, Chung BH, Jang KH, Song KB, Kim CH, Rhee SK. Refolding

and purification of Zymomonas mobilis levansucrase produced as inclu-

sion bodies in fed-batch culture of recombinant Escherichia coli. Protein

Expr Purif 2000;18:388–93.

[27] Miller GL. Use of dinitrosalicylic acid reagent for determination of

reducing sugar. Anal Chem 1959;31:426–8.

[28] Waley SG. A spectrophotometric assay of b-lactamase action of peni-

cillins. Biochem J 1974;139:789–90.

[29] Clifton D, Weinstock SB, Fraenkel DG. Glycolysis mutants in Sacchar-

omyces cerevisiae. Genetics 1978;88:1–11.

[30] Tajima K, Tanio T, Kobayashi Y, Kohno H, Fujiwara M, Shiba T, et al.

Cloning and sequencing of the levansucrase gene from Acetobacter

xylinum NCI 1005. DNA Res 2000;7:237–42.

[31] Trujillo LE, Gomez R, Banguela A, Soto M, Arrieta JG, Hernandez L.

Catalytical properties of N-glycosylated Gluconoacetobacter diazotrophi-

cus levansucrase produced in yeast. Electron J Biotechnol 2004;7:116–23.

[32] Vigants A, Marx SP, Linde R, Ore S, Bekers M, Vina I, et al. A novel and

simple method for the purification of extracellular levansucrase from

Zymomonas mobilis. Curr Microbiol 2003;47:198–202.

[33] Thompson JD, Higgins DG, Gibson TJ, Clustal W. Improving the sensi-

tivity of progressive multiple sequence alignment through sequence

weighting, positions specific gap penalties and weight matrix choice.

Nucleic Acids Res 1994;22:4673–80.

[34] Bendtsen JD, Nielsen H, von Heijne G, Brunak S. Improved prediction of

signal peptides: SignalP 3.0. J Mol Biol 2004;340:783–95.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422 421

Author's personal copy

[35] Sakai Y, Kazarimoto T, Tani Y. Transformation system for an asporogen-

ous methylotrophic yeast, Candida boidinii: cloning of the orotidine-5’-

phosphate decarboxylase gene (URA3), isolation of uracil auxotrophic

mutants, and use of the mutants for integrative transformation. J Bacteriol

1991;173:7458–63.

[36] Sakai Y, Tani Y. Directed mutagenesis in an asporogenous methylotrophic

yeast: cloning, sequencing, and one-step gene disruption of the 3-iso-

propylmalate dehydrogenase gene (LEU2) of Candida boidinii to derive

doubly auxotrophic marker strains. J Bacteriol 1992;174:5988–93.

[37] Merckelbach A, Godecke S, Janowicz ZA, Hollenberg CP. Cloning and

sequencing of the ura3 locus of the methylotrophic yeast Hansenula

polymorpha and its use for the generation of a deletion by gene replace-

ment. Appl Microbiol Biotechnol 1993;40:361–4.

[38] van Dijk R, Faber KN, Kiel JAKW, Veenhuis M, van der Klei I. The

methylotrophic yeast Hansenula polymorpha: a versatile cell factory.

Enzyme Microb Technol 2000;26:793–800.

[39] van Hijum SAFT, Szalowska E, van der Maarel MJEC, Dijkhuizen L.

Biochemical and molecular characterization of a levansucrase from

Lactobacillus reuteri. Microbiology 2004;150:621–30.

[40] Tieking M, Ehrmann MA, Vogel RF, Ganzle MG. Molecular and func-

tional characterization of a levansucrase from the sourdough isolate

Lactobacillus sanfranciscensis TMW 1.392. Appl Microbiol Biotechnol

2005;66:655–63.

[41] Reid SJ, Abratt VR. Sucrose utilisation in bacteria: genetic organisation

and regulation. Appl Microbiol Biotechnol 2005;67:312–21.

[42] Hettwer U, Jaeckel FR, Boch J, Meyer M, Rudolph K, Ullrich MS.

Cloning, nucleotic sequence and expression in Escherichia coli of levan-

sucrase genes from the plant phatogens Pseudomonas syringae pv.

glycinea and P. syringae pv. phaseolicola. Appl Environ Microbiol

1998;64:3180–7.

[43] Kim CH, Lee JY, Kim MG, Song KB, Seo JW, Chung BH, et al.

Fermentation strategy to enhance plasmid stability during the cultivation

of Escherichia coli for the production of recombinant levansucrase. J

Ferment Bioeng 1998;86:391–4.

[44] Hernandez L, Arrieta J, Betancourt L, Falcon V, Madrazo J, Coego A, et al.

Levansucrase from Acetobacter diazotrophicus SRT4 is secreted via

periplasm by a signal-peptide-dependent pathway. Curr Microbiol

1999;39:146–52.

[45] Decad GM, Nikaido H. Outer membrane of Gram-negative bacteria XII.

Molecular-sieving function of cell wall. J Bacteriol 1976;128:325–36.

[46] Steinmetz M, LeCoq D, Djemia HB, Gay P. Genetic analysis of sacB, the

structural gene of a secreted enzyme, levansucrase of Bacillus subtilis

Marburg. Mol Gen Genet 1983;191:138–44.

[47] Yanase H, Fujimoto J, Maeda M, Okamoto K, Kita K, Tonomura K.

Expression of the extracellular levansucrase and invertase genes from

Zymomonas mobilis in Escherichia coli cells. Biosci Biotechnol Biochem

1998;62:1802–5.

[48] Andersone I, Auzina L, Vigants A, Mutere O, Zikmanis P. Formation of

levan from raffinose by levansucrase of Zymomonas mobilis. Eng Life Sci

2004;4:56–9.

[49] Sangiliyandi G, Ray KC, Gunasekaran P. Elevated temperature and

chemical modification selectively abolishes levan forming activity of

levansucrase of Zymomonas mobilis. Biotechnol Lett 1999;21:179–82.

[50] Seibel J, Moraru R, Gotze S, Buchholz K, Na’amnieh S, Pawlowski A,

et al. Synthesis of sucrose analogues and the mechanism of action of

Bacillus subtilis fructosyltransferase (levansucrase). Carbohydr Res

2006;41:2335–49.

[51] Ammar YB, Matsubara T, Ito K, Iizuka M, Limpaseni T, Pongsawasdi P,

et al. Characterization of a thermostable levansucrase from Bacillus sp.

TH4-2 capable of producing high molecular weight levan at high tem-

perature. J Biotechnol 2002;99:111–9.

[52] Cote GL, Imam SH. Purification and properties of an extracellular

levansucrase from Erwinia herbicola NRRL B-1678. Carbohydr Res

1989;190:299–307.

T. Visnapuu et al. / Process Biochemistry 43 (2008) 414–422422