.elsevier.com/locate/bba

Biochimica et Biophysica Acta

cDNA cloning and functional expression of KM+, the mannose-binding

lectin from Artocarpus integrifolia seeds

Luis L.P. daSilva a,1, Jeanne Blanco de Molfetta-Machado a, Ademilson Panunto-Castelo b,

Jurgen Denecke c, Gustavo Henrique Goldman d,

Maria-Cristina Roque-Barreira b, Maria Helena S. Goldman a,*

a Depto. Biologia, FFCLRP/Universidade de Sao Paulo, Av. Bandeirantes, 3900 Ribeirao Preto, SP 14040-901, Brazilb Depto. Biologia Celular e Molecular e Bioagentes Patogenicos, FMRP/Universidade de Sao Paulo,

Av. Bandeirantes, 3900 Ribeirao Preto, SP 14049-900, Brazilc Centre for Plant Sciences, Faculty of Biological Sciences, University of Leeds, Leeds LS2 9JT, UK

d Depto. C. Farmaceuticas, FCFRP/Universidade de Sao Paulo, Av. do Cafe s/no, Ribeirao Preto, SP 14040-903, Brazil

Received 19 May 2005; received in revised form 7 September 2005; accepted 11 September 2005

Available online 29 September 2005

Abstract

KM+, a mannose-binding lectin present in the seeds of Artocarpus integrifolia, has interesting biological properties and potential

pharmaceutical use [A. Panunto-Castelo, M.A. Souza, M.C. Roque-Barreira, J.S. Silva, KM(+), a lectin from Artocarpus integrifolia, induces

IL-12 p40 production by macrophages and switches from type 2 to type 1 cell-mediated immunity against Leishmania major antigens,

resulting in BALB/c mice resistance to infection, Glycobiology 11 (2001) 1035–1042. [1]; L.L.P. daSilva, A. Panunto-Castelo, M.H.S.

Goldman, M.C. Roque-Barreira, R.S. Oliveira, M.D. Baruffi, J.B. Molfetta-Machado, Composition for preventing or treating appearance of

epithelia wounds such as skin and corneal wounds or for immunomodulating, comprises lectin, Patent number WO20041008. [2]]. Here, we

have isolated clones encoding the full-length KM+ primary sequence from a cDNA library, through matrix PCR-based screening methodology.

Analysis of KM+ nucleotide and deduced amino acid sequences provided strong evidence that it neither enters the secretory pathway nor

undergoes post-translational modifications, which is in sharp contrast with jacalin, the more abundant lectin from A. integrifolia seeds. Current

investigations into the KM+ properties are often impaired by the difficulty in obtaining sufficient quantities of jacalin-free KM+ through direct

seed extraction. To obtain active recombinant protein (rKM+) in larger amounts, we tested three different expression systems. Expression

vectors were constructed to produce: (a) rKM+ in E. coli in its native form, (b) rKM+ with GST as an N-terminal tag and (c) native rKM+ in

Saccharomyces cerevisiae. The presence of the GST-tag significantly improved the overall rKM+ yield; however, most of the obtained rGST-

KM+ was insoluble. Production of rKM+ in the yeast host yielded the highest quantities of soluble lectin that retained the typical high-

mannose oligosaccharide-binding properties of the natural protein. The possible biotechnological applications of recombinant KM+ are

discussed.

D 2005 Elsevier B.V. All rights reserved.

Keywords: KM+ or artocarpin; Jackfruit lectin; cDNA cloning; Heterologous expression; Mannose-binding

0304-4165/$ - see front matter D 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.bbagen.2005.09.006

Abbreviations: rKM+, recombinant KM+ produced in a heterologous

system; jfKM+, KM+ extracted from jackfruit seeds; HRP, horseradish

peroxidase glycoprotein; JRLs, jacalin-related lectins; CRD, carbohydrate

recognition domain

* Corresponding author. Tel.: +55 16 3602 3702; fax: +55 16 3633 1758.

E-mail address: mgoldman@ffclrp.usp.br (M.H.S. Goldman).1 Present address: Centre for Plant Sciences, Faculty of Biological Sciences,

University of Leeds, Leeds LS2 9JT, UK.

1. Introduction

Lectins are proteins displaying at least one non-catalytic

domain, which reversibly binds to specific mono or oligosac-

charides [3]. Lectins are known as being an extremely useful

tool for carbohydrate investigation on cell surfaces, for

glycoproteins isolation and characterization and for lympho-

cytes polyclonal activation. Numerous lectins have been

isolated from many organisms ranging from viruses and

bacteria to plants and animals, and they are known to play a

1726 (2005) 251 – 260

http://www

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260252

key role in a variety of biological processes (reviewed in [4]).

Plant lectins have been used in different biotechnological

applications (reviewed in [5]), including recombinant proteins

production for therapeutic purposes [6] or for targeted drug

delivery (reviewed in [7]).

Artocarpus integrifolia (jackfruit) seeds are known to

contain two lectins – jacalin and KM+ – that exhibit

significantly distinct carbohydrate binding specificities in spite

of a high overall structural similarity [8]. Jacalin, the first of

these lectins to be isolated [9], binds to d-galactose and its

derivative Galh1-3GalNAc [10], a specificity that has given it abroad application in the isolation and detection of mammals

O-glycosylated proteins and to selectively interact with human

IgA1 from human serum and secretions [11–13]. Jacalin is

synthesized on the rough endoplasmic reticulum, and after co-

translational cleavage of the signal peptide, the N-terminal

propeptide as well as an internal linker peptide are processed,

resulting in a a- and h-chain (20 and 133 amino acids,

respectively) [14]. As revealed by the immunocytological

analysis of jackfruit cotyledons, jacalin accumulates in small

punctuate structures distributed throughout the cytoplasm [15],

which could be storage vacuoles. However, the exact intracel-

lular sites in which jacalin molecular modifications take place

and its transport route within the secretory pathway still have to

be elucidated.

In the beginning of the last decade, a second lectin was

identified in jackfruit seeds as being responsible for the effects

that the crude extract from jackfruit seeds has on both T cell

proliferation and B cell polyclonal activation [16]. The novel

lectin, named KM+ or artocarpin, was isolated and character-

ized as a neutrophil migration inducer [17], a property

attributed to its binding specificity towards d-mannose and

d-glucose, but not d-galactose [17,18]. The KM+ primary

structure was determined by Rosa et al. [8] as being a

polypeptide chain of 149 amino acids sharing 52% identity

with the jacalin sequence. The differences between jacalin and

KM+ are mainly attributed to the fact that KM+ does not

undergo internal post-translational cleavage, preserving a short

glycine-rich linker sequence that holds the regions analogous to

the jacalin a- and h-chains together [8]. According to

molecular modeling and crystal studies, the consequent

structural differences account for the distinct carbohydrate-

binding specificities exhibited by the two lectins [8,19,20],

especially when the recognition of d-mannose, but not d-

galactose, by KM+ is concerned. Indeed, the best ligands for

KM+ are those with N-linked glycans containing the trimanno-

side core Mana1–3[Mana1–6]Man, such as the horseradish

peroxidase glycoprotein (HRP) [18]. Recently, a reinvestiga-

tion of the KM+ carbohydrate-binding properties revealed its

unexpected behaviour as a polyspecific lectin that reacts with a

wider range of monosaccharides, although the preferential

affinity is for mannose [21].

KM+ has been reported as a tool for multiple biomedical

applications, including the induction of neutrophil migration

[22,23], degranulation of mast cells [24], induction of IL-12

production by macrophages [1,25] and acceleration of wound

healing [2]. However, advances in the studies are often limited

by fruit harvesting and the hard task of purifying large

quantities of jacalin-free KM+. Jacalin, which is present in

jackfruit seed extracts in approximately 60 times higher

concentration than KM+, is a major contaminant in natural

KM+ preparations [17], making difficult the interpretation of

experiments using this material. The availability of appropriate

amounts of homogeneous KM+ is a pre-requisite for repro-

ducible mechanistic studies for a better insight into the lectin

mode of action on cells and the evaluation of its pharmaceutical

application in extensive pre-clinical models.

Considering the KM+ applications and the interest in further

exploring the basis of its biological actions at a molecular level,

we have aimed at cloning its cDNA and produce the

recombinant KM+ to avoid the problems with seed extracts.

In this work, the recombinant product has been successfully

expressed and purified, its sugar binding characteristics have

been compared to those of the plant-derived product, and

equivalence has been determined. Differences between KM+

and jacalin at the DNA level and their implications have also

been discussed.

2. Materials and methods

2.1. RNA isolation and northern blot analysis

Total RNA was isolated from mid-maturation seeds (about 29�18 mm in

size) of a jackfruit tree from Brazil. About 6 g of this material was ground in

liquid nitrogen and extracted as described in the literature [26]. RNA was

separated by electrophoresis on a 1.5% agarose gel containing 2.2 M

formaldehyde and transferred to Hybond N+ for 2.5 h, in a vacuum blotter

apparatus (Bio-Rad, Hercules, CA, USA), using 10� SSC. For northern

analysis, hybridization was performed with a DNA probe in 6� SSC, 5�Denhardt’s solution, 0.5% SDS, and 100 Ag/ml denatured carrier DNA at

50 -C, overnight. Filters were washed at 50 -C, in 6� SSC, 0.5% SDS for 15

min and subsequently in decreasing salt concentrations (2� SSC, 1� SSC,

0.5� SSC and 0.1� SSC) in 0.1% SDS for 30 min. Hybridized filters were

exposed to Kodak X-Omat films for the appropriate time period, at �70 -C, in

between intensifying screens.

2.2. Construction of a jackfruit seed cDNA library

RNA integrity was examined by testing the presence of the jacalin

transcript by northern blot analysis before it was used for cDNA synthesis (not

shown). The poly(A)+ RNA was isolated by the PolyATract mRNA Isolation

System (Promega, Madison, WI, USA), according to the manufacturer’s

instructions. A cDNA library was constructed from the mRNA, using the

SuperScripti Plasmid System for cDNA Synthesis and Plasmid Cloning

(Invitrogen, Carlsbad, CA, USA), which allowed the directional insertion of the

cDNA fragments into the SalI and NotI restriction sites of pSPORT-P

(Invitrogen). The library was propagated in Escherichia coli DH10B cells

(Invitrogen). The cDNA clones were individually transferred to plates of 96

wells containing 2� YT liquid medium, supplemented with 100 Ag/ml

ampicillin and 20% glycerol, and stored at �80 -C. The whole cDNA library

was organized in a total of 136 plates.

2.3. Screening of the cDNA library by matrix PCR and sequencing of

the positive clones

A matrix containing ordered pools of the library clones was created in 96-

well microtiter plates. It was achieved by inoculating a replica from each of the

136 microtiter plates in one master plate, resulting in 96 pools of 136 clones

each (Fig. 1, panel B). The matrix plate was then incubated for 4 h at 37 -Cwith slow shaking to allow further growth of pooled bacteria.

Fig. 1. Screening of the cDNA library via matrix PCR. (A) Schematic illustration of the expected topology of a library clone containing the KM+ encoding cDNA

insert. The three positions correspond to the degenerated oligonucleotides 1S, 2S and 3AS within the KM+ cDNA (white bar) and the relative locations of the oligos

SP6 and T7 within the plasmid vector (black bar) are illustrated. (B) Screening of positive clones from a library with 1.2�104 clones after only two PCR rounds. The

left panel explains how the matrix plate, representing all clones of the library, was generated. The 96 clone pools present in this matrix plate were used as templates in

the first PCR round using the pair of oligos SP6 and 3AS. The right panel shows the electrophoretic analysis of the products of the second PCR round, using the

oligos T7 and 1S. In the second round, only 28 clone pools, which previously resulted in amplification products of the expected size, were used as template. The

position of the clone pools on the matrix plate is indicated above each lane. Only eight of those reactions resulted in amplification, and the two clone pools were

chosen to continue the screening (white arrowhead). (C) Elaboration of a second matrix and identification of two positive clones. The left panel illustrates the

elaboration of a second matrix, using aliquots of the original 272 clones that formed the two mixtures selected as described in panel B. The right panel shows the

results of 36 PCR in which the clone pools of the second matrix were used as template. The result indicates the presence of positive clones in column 1 and lines A*

and E* (white arrowheads).

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260 253

Initial library screening was performed by successive PCR amplifications

using 1 Al of each matrix pool as templates and the enzyme Taq DNA

polymerase (Invitrogen). Three degenerate oligonucleotides were synthesized

from KM+ specific regions of the amino acid sequence [8], as follows: oligo-1

sense (1S) 5V[GCGAATTCGARCCNTTYWSNGGNCCNAA]3V, oligo-2 sense

(2S) 5V[GCGAATTCAARYTNCCNTAYAARAA]3V, oligo-3 antisense (3AS)

5V[GCGGATCCGCCATATGNACNCCDATNGC]3V, corresponding to the ami-

no acid sequences EPFSGPK, KLPYKN and AIGVHMA, respectively. The

above oligos were used on PCRs in different combinations, or with the SP6

promoter primer and the T7 promoter primer (as described in Results). The SP6

primer hybridizes upstream of the SalI site in the pSPORT-P polylinker (5V endof the cDNA inserts), while the T7 primer hybridizes downstream of the NotI

site (3V end of the cDNA inserts). The PCRs were performed under the

following conditions: initial denaturation at 94 -C for 4 min, followed by 40

cycles of 94 -C for 1 min, 48 -C for 1 min, and 72 -C for 2 min, and then final

extension at 72 -C for 10 min. The reaction products were resolved by agarose

gel electrophoresis. Clone pools, which resulted in amplification products with

the predicted size, were used as templates in a second round of nested PCRs

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260254

under the same conditions described above, but using a different primer

combination. Further elimination of negative clone pools was possible and only

two of these samples were selected to continue the screening. The 136

individual clones represented in each of the two selected clone pools were

rearranged into a new matrix for the final screening step (Fig. 1, panel C).

The final positive clones were sequenced according to the dideoxynucleo-

tide chain termination method, using Big-Dye chemistry (Applied Biosystems,

Foster City, CA, USA) and automated sequencer ABI3100 (Applied

Biosystems). DNA and deduced amino acid sequences were analyzed using

freely available computer software like Phred/Phrap/Consed and tools

accessible from the NCBI (http://www.ncbi.nlm.nih.gov/) and Expasy (http://

us.expasy.org/tools/) sites. The sequence data reported here is available in the

GenBank under accession numbers AY957581 and AY957582.

2.4. Southern blot analysis

Genomic DNA isolated from A. integrifolia seeds was digested with EcoRI,

HindIII and PstI, separated by electrophoresis on a 1% agarose gel [27], and

transferred to Hybond N+, according to the alkaline method from the Amersham

protocol. Hybridization, subsequent washes and film exposition were essentially

the same as described for the northern blot analysis (see above).

2.5. Expression vectors construction

For KM+ expression, the cDNA encoding KM+ from the pLL29 clone was

inserted into the pDONR201 entry vector of the Gateway system (Invitrogen),

according to the manufacturer’s instructions. For this purpose, the following

oligos were used: OL 17 sense 5V [GGGGACAAGTTTGTACAAAAAAGCA-GGCTTCGAAGGAGATAGAACCATGGCGAGCCAGACGATAACAGTC-

GGG]3V, OL 18 antisense 5V[GGGGACCACTTTGT ACAAGAAAGCTGGG-

TCTAAAGTGCCGTGAACGCCAATAGC]. This allowed the introduction of

sites for sequence specific recombination, as well as the addition of optimal

translational initiation sites for both eukaryotic and prokaryotic cells (sequences

from KM+ are underlined). The PCR reactions were performed under the

following conditions: 25 cycles of 94 -C for 30 s, 45 -C for 1 min and 72 -C for

3 min, and then final extension at 72 -C for 10 min. The reaction product was

introduced into pDONR201 (Invitrogen) by site specific recombination,

yielding pEntryKM+, from which the KM+ coding sequence was transferred

to the pDEST14, pDEST15 and pYESDEST52 (Invitrogen) expression vectors,

using the LR recombinase. This resulted in the generation of vectors for KM+

expression in E. coli, either in its native form (pExpKM+) or with an N-

terminal GST tag (pExpGST-KM+), in addition to a vector for KM+ expression

in S. cerevisiae (pYESKM+). All the constructs were confirmed by sequence

analysis.

2.6. KM+ expression in E. coli

A single E. coli BL21 (SI) colony transformed with either pExpKM+ or

pExpGST-KM+ was selected and grown overnight at 37 -C, in LB medium

devoid of NaCl and supplemented with ampicillin (100 Ag/ml). The culture was

diluted 100 fold into LB medium without NaCl and incubated at 30 -C for

2 h (OD600�0.5). NaCl was added to a 0.3-M final concentration, and

induction was allowed to proceed for increasing time periods, to optimized

expression level. Aliquots of 1 ml culture were harvested, and cells expressing

rKM+ or rGST-KM+ were pelleted by brief centrifugation. The cell pellet was

suspended in 200 Al of 10 mM Tris–HCl (pH 7.5), supplemented with Tosyl

Lysyl Chloromethylketone (TLCK) as a protease inhibitor. Cellular proteins

were analyzed separately as soluble and insoluble proteins. To extract soluble

proteins, the cell suspensions were sonicated and spun at 25,000�g for 15 min.

The supernatant containing soluble proteins was recovered (S). The resulting

pellet, enriched in insoluble proteins and protein aggregates, was resuspended

in 200 Al of the same buffer, sonicated, and left in suspension (I).

2.7. KM+ expression in Saccharomyces cerevisiae

The pYESKM+ expression vector was used to transform S. cerevisiae

cells—strain INVSc1 (MATa, his3-Dl, leu2, trp1-289, ura3-52; Invitrogen) by

the lithium acetate method, as described in the literature [28]. Transformants

were selected onto SC/agar minimal medium devoid of uracil (SC-U; 0.67%

yeast nitrogen base, 0.072% Dropout mix without URA—Sigma, St. Louis,

MO, USA, 2% glucose). Selected transformants were cultured in SC-U medium

at 30 -C, overnight, to reach an OD600 of approximately 3. The culture was

spun for 5 min at 1500�g, and the supernatant discarded. Cells were

resuspended to an approximate OD600 of 0.4 in induction medium (SC-U in

which glucose was replaced by 2% galactose and 1% raffinose), and incubated

at 30 -C with shaking (250 rpm). After induction, cells were washed in distilled

water, and cell pellets were stored at �80 -C. Cell samples harvested at

different time-points (0, 4, 8, 16 and 24 h) were analyzed to optimize induction

time (not shown). To prepare cell lysates, frozen cell pellets were resuspended

in breaking buffer (50 mM sodium phosphate buffer, pH 7.4, 1 mM EDTA, 5%

glycerol, 1 mM PMSF), and an equal volume of acid-washed glass beads was

added. The mixtures were submitted to 4 sections of 30 s vortex, followed by

30 s ice incubation and, after 10 min centrifugation at 25,000�g, the

supernatants were isolated and used for subsequent analysis. The higher protein

yield necessary to perform the functional assays was obtained by scaling up the

procedure described above to 3 l culture. Preparation of cell lysates from high

culture volumes was achieved by using a French Press.

2.8. Gel blot analysis

For the comparison of the relative rKM+ yield in the different expression

systems used, each of the protein extracts obtained as described above was split

into two identical aliquots, one of which was immediately frozen after mixing

with an equal volume of 2� SDS loading buffer (125 mM Tris–HCl pH 6.8,

4% SDS, 10% glycerol, 0.2% bromophenol blue and 4% beta-mercaptoetha-

nol). The remaining aliquots were used for protein quantification through the

use of the Protein Assay kit (Bio-Rad), following the manufacturer’s

instructions. To allow the loading of equal protein levels per lane (1 Ag/Al),samples were adjusted to the appropriate dilution by considering their initial

protein concentration, using a 50/50 Tris-buffer/SDS loading buffer. After such

adjustment, the samples were briefly boiled (5 min at 95 -C). Proteins in SDS-

PAGE were transferred onto a nitrocellulose membrane and blocked for 1 h with

1% gelatin in TBS (20 mM Tris–HCl, 150 mM NaCl, pH 7.5) containing 0.1%

Tween 20. Immunodetections were performed using a 1/1000 dilution rabbit

polyclonal anti-KM+ serum, 1/2000 [17,22] alkaline phosphatase-conjugated

anti-rabbit IgG antibody (Promega), and nitroblue tetrazolium/5-bromo-

4-chloro-3-indolyl-phosphate (Gibco-BRL, Rockville, MD).

2.9. Lectin binding assay

Each well of a 96-microtiter plate (MaxiSorp FluoroNunc, Roskilde,

Denmark) was coated with 100 Al of one of the following proteins (10 Ag/ml in

carbonate buffer pH 9.6): (1) KM+ from jackfruit (jfKM+), (2) recombinant

KM+ (rKM+), (3) jacalin, or (4) bovine serum albumin (BSA), overnight, at

4 -C. After washing with 0.05% Tween-20 in PBS (PBS-T), the non-specific

interactions were blocked with 3% gelatin in PBS-T, at room temperature (RT).

After 1 h incubation, the plate was washed and incubated with a 100-Al serialdilution of horseradish peroxidase (HRP, Sigma Chemical Co.) in 1% gelatin in

PBS-T, for 2 h at RT. In inhibition assays, the coated jfKM+ or rKM+ reacted

with 1 or 10 Ag HRP, respectively, in the presence of different concentrations

(from 0.1 to 1000 mM) of d-mannose or d-galactose. The wells were washed,

and the bound HRP was detected using H2O2/OPD (ortho-phenylenediamine)

as substrate (Abbott Laboratories, Abbott Park, IL, USA). The reactions were

stopped by adding 1 N sulfuric acid and the samples were read at 490 nm in a

microwell plate reader (Elisa Power Wave X; Bio-Tek Instruments, Inc.,

Winosky, VT, USA). The binding assays were repeated at least three times.

3. Results

3.1. KM+ cDNA clones isolation by matrix PCR

A cDNA library from A. integrifolia seeds with approxi-

mately 13�103 clones was produced through the directional

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260 255

cDNA insertion into the plasmid vector. Thus, we could predict

the orientation of the primer sequences situated in the vector in

relation to the three degenerate oligonucleotides designed

based on KM+ amino acid sequence (Fig. 1A). This feature

allowed us to develop a screening strategy based on successive

PCR amplifications in a matrix format using different primer

combinations. Three regions were rationally selected to design

degenerate oligonucleotides for KM+ amplification to avoid

similarities with jacalin. The primer selection was particularly

important, considering that there is a 52% primary sequence

identity between KM+ and jacalin [8], along with the likely

higher incidence of jacalin clones in the library.

An ordered mixture of aliquots from each individual library

clone was performed, resulting in a template matrix containing

96 pools of 136 clones each (see Materials and methods). For

the first screening step, an aliquot of each of the 96 clone pools

was used as template for PCR, employing SP6 and 3AS

oligonucleotides as primers. This would lead to the amplifica-

tion of the 5V end of the cDNAs and permit selection of full

length clones by size. Only 28 templates resulted in amplifi-

cation products, allowing the elimination of at least 9�103

negative clones (not shown). The selected templates with

amplification products were used in a second PCR round, using

the 1S and T7 oligonucleotides as primers. In the latter

reactions, only eight of the templates resulted in amplification

(Fig. 1B), leading to a further elimination of negative clones.

Based on the sizes obtained in the first and the second PCR

screen, two pools were chosen for individual analysis.

The original 136 clones represented in each of the two

selected templates were then rearranged in a new matrix. The

ordered mixture of the clones located in the twelve columns and

twenty-four lines of the new matrix resulted in 36 clone pools

(Fig. 1C), which were templates for PCRs, using the 2S and T7

oligonucleotides as primers. This third nested reaction would

exclude any false positives and three pools generated amplifi-

cation products with the expected size. The latter result allowed

us to identify two individual clones by recalling their positions in

the matrix and repeating the PCR. These clones were sequenced

and confirmed as containing a cDNA sequence encoding KM+

due to their match with the protein sequence previously

determined by Rosa et al. [8]. In the last screening of the matrix

comprising 272 clones, four jacalin clones were identified,

which corresponds to a frequency of approximately 1.5% of the

total library (data not shown). This shows that jacalin transcripts

are much more abundant compared to KM+ transcripts.

The entire cDNA and the deduced amino acid sequence

present in one of the isolated clones (pLL29) are shown in Fig.

2A. In Fig. 2B, the primary sequences deduced from both

positive clones are compared through alignment with the KM+

amino acid sequence identified in reference [8]. The sequence

corresponding to pLL29 displays two amino acid substitutions,

both from charged to non-charged residues and the one from

pLL30 shows a single amino acid substitution, in which it is

identical to the pLL29 sequence. Both clones contain an open

reading frame of 507 bp encoding 169 amino acids residues

with a likely start codon at position 20 of the deduced amino

acid sequence. This hypothesis is strengthened by the presence

of sequence 5V[ACCATGG]3V surrounding the codon region for

the first amino acid of the mature KM+ [8], which exactly

corresponds to the sequence described in references [29,30] as

being the optimal context for translational initiation in

eukaryotes. Moreover, no signal sequence could be recognized,

supporting the notion that KM+ does not enter the secretory

pathway. In fact, the absence of signal peptide, no post-

translational processing and cytosolic residence are now

believed to be main features of the mannose-specific jacalin-

related lectins, JRLs [31,32].

To gain insight into the KM+ transcript, we performed

northern blot analysis, in which mRNA extracted from seeds

was probed with the insert from pLL29. A single mRNA band

with a size of approximately 750 nucleotides was detected (Fig.

3A), which corresponds to the total length of the cDNA

sequences of the identified clones.

3.2. Contrary to jacalin, KM+ is encoded by a small multigene

family

It has previously been reported that jacalin, the prototype of

the galactose-specific JRLs, is encoded by a large family of

genes at multiple loci in the A. integrifolia genome [14]. To

investigate whether this also occurs with KM+, Southern blot

analysis of the genomic DNA extracted from jackfruit seeds

was performed. KM+ genes were probed using the cDNA

sequence from pLL29. Since a total of four clones encoding

full-length jacalin cDNA sequences were also isolated from our

library (data not shown), we could use one of these clone

inserts to probe jacalin genes.

The hybridization pattern obtained for each of the employed

probes is clearly distinct. While the jacalin probe produced a

great number of bands with different hybridization identities,

the KM+ probe revealed a much lower number of bands (Fig.

3B; compare the two panels). Considering the presence of an

EcoRI site in the KM+ cDNA sequence, it is likely that two of

these bands represent a single gene. In addition, due to the high

sequence similarity, there are bands recognized by both probes,

however, with different hybridization intensities (Fig. 3B, black

arrow-heads). Considering the probable heterozygous condi-

tion of jackfruit plants in nature and in the light of our results,

the most likely conclusion is that the KM+ gene is present as a

small gene family in the A. integrifolia genome, which is in

sharp contrast to jacalin.

3.3. Expression of rKM+ in heterologous systems

A number of attractive biological effects and the potential

for biotechnological and pharmaceutical uses have been

reported for KM+ [1,2,22]. However, its isolation from

jackfruit seeds and purification from the highly predominant

jacalin is time consuming and generally result in low yield. To

overcome this limitation, we aimed at establishing a recombi-

nant system capable of producing biologically active KM+ in

large quantities.

Our first approach was the expression of recombinant KM+

(rKM+) in its native form using E. coli cells—strain BL21 (SI).

Fig. 2. The deduced amino acid sequences of the two isolated cDNA clones present high degree of similarity with the KM+ protein sequence. (A) Nucleotide

sequence and deduced amino acid sequence of the cDNA present in the clone pLL29. The methionine at position 20 was assumed as the first amino acid; therefore,

the preceding deduced residues are presented in lower case. The sequences corresponding to the degenerated oligonucleotides are shown in bold. (B) Alignment of

the deduced amino acid sequences of the two identified KM+ cDNA clones and the KM+ protein sequence [8]. Amino acids that are not identical in the three

sequences are shown in bold and the ones not present in the mature protein in italics.

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260256

To exploit this system, the entire open reading frame of the

cDNA sequence present in the pLL29 clone, starting from the

first methionine, was inserted into the pDEST14 expression

vector by site-specific recombination (see Materials and

methods). After induction with NaCl and incubation for

increasing time periods (up to 5 h), the protein expression

pattern was monitored via Coomassie blue stained SDS-PAGE

analysis. No difference in the expression pattern could be

noticed under these conditions when compared with the

negative controls of both soluble and insoluble cell fractions

(not shown).

To test KM+ presence, immunoblots of the different

samples were performed using anti-KM+ serum. An extremely

low expression level was detected in the samples prepared from

induced cells (Fig. 4). In addition, the rKM+ yield was on

average three times higher in the insoluble cell fractions in

terms of the total loaded protein, as estimated by the SDS-

PAGE gels scanning densitometry (not shown).

To try to increase the rKM+ yield and to expand our tools to

optimize rKM+ production, we have tested and compared two

other expression systems: (a) E. coli expression of rKM+ with

a GST fusion at the N-terminus (rGST-KM+), and (b) S.

cerevisiae (INVSc1 strain) expression of rKM+ in its native

form. In these two systems, expression was achieved by

introducing the KM+ encoding sequence into the pDEST15

and pYESDEST52 expression vectors, respectively.

The expression of KM+ with an N-terminal GST tag

resulted in a dramatic increase in the recombinant protein

yield. Under the same growth conditions, the rGST-KM+ levels

in terms of mass per liter of culture were approximately 40

times higher than the levels of rKM+ after a 4 h induction

period, as determined by the SDS-PAGE gels scanning

densitometry. However, the 50-kDa fusion protein almost

completely accumulated in the insoluble cell fraction (Fig. 4),

most likely as inclusion bodies. Although structure and

therefore functionality of the protein as an insoluble aggregate

is greatly compromised, successful restoration of protein

activity through refolding procedures is achievable in several

cases [33]. Therefore, this expression system could offer an

alternative means to obtain functional rKM+ after GST

cleavage.

Instead, we decided to test whether rKM+ expression in S.

cerevisiae would directly result in higher concentrations of

soluble protein. A culture of transformed INVSc1 cells was

Fig. 4. The S. cerevisiae based expression system produces a better yield of

soluble rKM+. E. coli BL21 cells wild type (control) or transformed with either

pExpKM+ or pExpGST-KM+ were induced with 0.3 M NaCl for 4 h, harvested

and fractionated to obtain extracts enriched in soluble (S) and insoluble (I)

proteins. Similarly, cell extracts enriched in soluble proteins were prepared

from untransformed (S�) or pYESKM+ transformed (S+) S. cerevisiae cells,

which were cultured in induction medium for 24 h. Before loading, the various

cell extracts were equalized with respect to total protein content, allowing the

comparison of the relative yield of recombinant protein from each system. An

aliquot of KM+ purified from A. integrifolia seeds was also loaded. Following

SDS-PAGE, protein gel blot was probed with anti-KM+ serum. The faint rKM+

bands resulting from E. coli cell extracts are indicated by white arrowheads.

Fig. 3. The KM+ gene is present in a low number of copies on the genome. (A)

RNA blot analysis of KM+ transcript. Poly(A)+ RNA isolated from A.

integrifolia seeds was hybridized with the labeled KM+ cDNA insert of pLL29.

A single transcript with size corresponding to approximately 750 nucleotides

was detected. (B) Comparative Southern blot analysis of KM+ and jacalin

genes. A. integrifolia genomic DNAwas digested with the restriction enzymes

as indicated on each lane, and hybridized with probes prepared with either

KM+ (left panel) or jacalin (right panel) cDNAs clones. Bands highlighted by

both probes with different intensities are indicated (black arrowheads).

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260 257

induced for a 24-h period, throughout which aliquots were

collected and analyzed by SDS-PAGE. After Coomassie

staining, a faint band corresponding to a protein of the

expected size could be visualized in the soluble cell fraction

after an 18-h induction (data not shown). Fig. 4 shows a

Western blot analysis comparing the level of recombinant

protein production in terms of total cellular protein, yielded by

each of the three tested systems. Because the latter system

comparatively provided the best soluble rKM+ yield, the S.

cerevisiae based system was chosen to proceed with protein

purification and functional assays.

3.4. Recombinant KM+ produced in S. cerevisiae is functional

The total soluble protein fraction obtained from S.

cerevisiae cells expressing rKM+ was submitted to affinity

chromatography using a mannose-agarose column. The bound

material eluted with D-Man corresponded to 0.15% of the

total protein. When analyzed by SDS-PAGE, the D-Man

bound fraction showed one broad band only, which corre-

sponded to the 60- to 80-kDa molecular mass range. Such

band was shifted to a single 16-kDa sharp band when the

sample was heated at 100 -C for 3 min (data not shown).

This electrophoretical pattern is similar to the one provided

by jackfruit KM+ (jfKM+).

To evaluate the rKM+ lectin activity, this protein was coated

to the wells of a microplate and incubated with different

concentrations of HRP, a glycoprotein containing the trimanno-

side Mana1–3[Mana1–6]Man, which is a known ligand for

jfKM+. rKM+ was able to bind HRP in a dose dependent

manner, leading to a curve parallel to the one provided by

jfKM+. Jacalin and BSA coated to the microplate wells were

both unable to interact with HRP (Fig. 5A). Although rKM+

was less reactive with HRP than jfKM+, the glycan recognition

was very specific, once it was inhibited by D-Man, and not by

D-Gal, in a dose dependent manner (Fig. 5B).

4. Discussion

4.1. A fast and sensitive methodology for isolating cDNA

clones

Two full-length KM+ encoding clones were isolated from a

cDNA library through an effective and straightforward

screening methodology, in which ordered mixtures of the

library clones were directly used as template in three PCR

rounds. The successive exclusion of negative clones was

achieved by using different combinations of three KM+

degenerated oligonucleotides as primers on matrix PCR

amplifications (Fig. 1).

There are several advantages to this screening approach.

Firstly, it does not require the availability of a probe sequence

and/or the manipulation of radioactive material. Secondly,

results can be obtained much more quickly than with traditional

hybridization-based approaches. In addition, this methodology

Fig. 5. (A) Functionality of rKM+ is revealed by a glycoprotein binding assay.

The wells of a microtiter plate were coated with either 1 Ag jfKM+ or 1 Ag yeastrKM+. Coating with jacalin or BSAwas used as negative control. Peroxidase, a

known specific ligand for KM+, was added in increasing concentrations to the

wells. The KM+ binding to peroxidase was revealed with the addition of

hydrogen peroxide and o-phenylenediamine. Color development was measured

at 490 nm. Data are representative of three independent assays. (B) rKM+ binds

to peroxidase through specific carbohydrate recognition. jfKM+ or rKM+ coated

the wells of a microtiter plate reacted with optimal concentrations of peroxidase,

as determined in A (1 and 10 Ag, respectively), in the presence of increasing

concentrations of monosaccharides (d-mannose ord-galactose).d-mannose was

able to inhibit the binding of both jackfruit KM+ and yeast rKM+ to peroxidase in

a dose specific manner, while d-galactose had no effect on the binding ability.

Data are representative of three independent assays.

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260258

sensitivity is particularly useful for the screening of cDNA

clones derived from mRNAs sharing high homology with more

abundantly transcribed ones, which is the case of KM+.

Furthermore, the selection of abundant clones using a single

pair of specific primers can be achieved after one mere PCR

set. This was illustrated by the isolation of four full-length

jacalin clones by screening only �1/50 of the same jackfruit

seed library (data not shown).

It is important to mention that the possibility of direct

amplification of the cDNA using the degenerate oligos was

discarded to avoid mutagenesis of possible isoforms containing

sequence dissimilarities within the regions encoded by the

oligos. In addition, this method alone would not provide

information about the complete sequence of KM+ and its

possible post-translational processing. Analysis of nucleotide

and deduced amino acid sequences of the KM+ cDNA clones

strongly suggests that, in contrast to jacalin, KM+ does not enter

the secretory pathway and most likely remains in the cytosol.

Mature KM+ appears to be the direct product of primary

translation in which the initial methionine was removed and the

resulting first amino acid alanine was acetylated.

Another interesting difference between these two closely

related lectins emerged from comparative Southern blot

analysis of the A. integrifolia genomic DNA. Yang and Czapla

[14] have provided evidence that jacalin is encoded by a large

family of genes, and that the 10–13 variant amino acids found

in the different jacalin identified sequences ([14] and our own

data) are most likely derived from independent transcripts. In

contrast, the data from Fig. 3B clearly shows a much lower

number of hybridization bands for KM+ than for jacalin.

Therefore, it is tempting to speculate that KM+ is less

represented in the A. integrifolia genome, and that this could

possibly explain why a lower number of isoforms of this lectin

seems to be present in jackfruit seeds.

4.2. Recombinant KM+ expression in S. cerevisiae is efficient

to produce large quantities of functional lectin

The growing interest in obtaining jacalin-free KM+ in

larger amounts than those typically obtained by direct

extraction from jackfruit seeds was an additional motivation

for the present work. E. coli remains as the simplest and most

cost-effective host for foreign protein expression. Since our

findings and data from a previous study [8] demonstrated that

KM+ does not undergo post-translational modifications, an E.

coli based system was our first choice for rKM+ production.

Expression of rKM+ in its native form in the bacterial host

was accomplished; however, this system failed to produce

satisfactory quantities of recombinant protein within the

different tested conditions, reaching a maximum of approxi-

mately 1.5 mg per liter of culture, nearly 75% of which as

insoluble aggregates.

There are several possible explanations for low yield of

recombinant protein production in this host system, such as:

instability at the mRNA, inefficient translation and proteolytic

degradation [34,35]. KM+ has a high amount of glycine

residues, 57% of which are encoded by either GGA or GGG.

These codons are amongst the least used by E. coli, and can be

detrimental to foreign protein synthesis by this host [36–38].

Interestingly, glutathione-S-transferase (GST) fusion as an

N-terminal tag successfully increased the overall KM+

production level in the same host system and under similar

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260 259

conditions used for native expression (Fig. 4). Besides

facilitating protein purification, fusion partners, such as GST,

are thought to increase protein stability within the host cell by

accelerating proper folding and protecting the recombinant

protein against protease degradation [39–42]. Additionally, the

inhibitory effect on translation resulting from rare-codons is

thought to be much more pronounced when these are located

near the 5V end of the transcript [38,43], which provides an

alternative explanation for the positive effect of GST.

Nevertheless, despite the substantial increase in recombinant

protein yield resulting from this alternative strategy, almost all

the produced rGST-KM+ accumulated as insoluble protein.

The use of the rGST-KM+ produced in this system as a source

of active lectin would implicate on protein solubilization and

possibly refolding, as well as GST-cleavage. Therefore, we

decided to test an alternative means to directly obtain soluble

KM+.

Native rKM+ production in the alternative host S. cerevisiae

improved in about 8 fold the yield of soluble rKM+ per liter of

culture when compared to E. coli, reaching the level of

approximately 4 mg/l of soluble rKM+ [2]. Moreover, this

system produced not only a good protein yield, but also a

functional lectin, since rKM+ was able to bind to HRP, in a dose

dependent manner (Fig. 5A). The binding of both lectin forms,

jfKM+ and rKM+, to HRP was inhibited by d-mannose, but not

by d-galactose (Fig. 5B). Although rKM+ and jfKM+ have the

same sugar specificity, the affinity of the former for carbohydrate

is about 10 fold lower. Indeed, a 100-fold lower affinity for the

specific carbohydrate methyl-a-galactose was demonstrated for

the recombinant form of jacalin (another JRL member), in

comparison to the native form [44]. The authors suggested that it

is due to the lack of proteolytic processing of jacalin molecule in

E. coli environment. In the case of rKM+, here reported, we have

no noticeable explanation for the 10 times decreased affinity for

d-mannose containing glycans, a point that is under investiga-

tion by our group. The aims achieved in this study, cDNA

cloning and functional expression of KM+, are powerful tools

for the study of the previously reported recognition by JRL of

glycans containing mannose residues in a1–3 and a1–6

linkages [18–20], like those present in HRP and laminin. The

dissection of the surprising KM+ polyspecificity for mono-

saccharides such as galactose, N-acetyl galactosamine, glucose,

sialic acid and N-acetylmuramic acid, and its preferential

binding to mannose, as revealed by surface plasmon resonance

hapten inhibition experiments [21], will also be aided by the

KM+ cDNA cloning and functional expression achieved in this

work. The sugar discrimination described byMisquith et al. [18]

and Pratap et al. [19] is consistent with molecular modeling

studies that provided the basis for the recognition of D-

mannose, but not D-galactose, by KM+ [8], and is also in

agreement with recent detailed studies on the KM+ (or

artocarpin) carbohydrate specificity carried out by comparing

crystal structures of KM+ complexed to mannotriose or

mannopentose [20]. This demonstrates that the lectin possesses

a deep-seated binding site composed of two subsites of

interaction with sugar residues. This CRD feature is relevant

for the enhancement of KM+ affinity binding to longer

oligosaccharides, such as Mana1–6[Mana1,3][Xylh1–2]Manh1–4GlcNAch1–4[Fuca1,3]GlcNAc, found in HRP.

While the trimannoside core interacts with the primary site

through numerous hydrogen bonds, the GlcNAc, Xyl and Fuc

residues bind to the secondary site, essentially through van

der Waals interactions. These features could explain the high

KM+ specificity in the recognition of glycans present in HRP,

laminin and cell surfaces glycoconjugates, a question that can

be more precisely clarified by taking up selected KM+ point

mutations and examining their crystal structures. The detailed

understanding of the KM+ sugar binding properties will allow a

better application of its relevant pharmaceutical abilities [2],

such as the induction of tissue regeneration after burning and the

modulation of immune response to intracellular pathogens.

Acknowledgements

We thank Dr. Els J. M. Van Damme for stimulating

discussion. We also thank Ms. Andrea C. Quiapim and Mrs.

Patrıcia M. Vitorelli for their technical assistance in DNA

sequencing, and Mrs. Sandra M.O. Thomaz for technical

assistance on lectin purification and binding assays. L.L.P.

daSilva, J.B. Molfetta-Machado and A. Panunto-Castelo were

supported by fellowships from FAPESP. G.H. Goldman, M.-C.

Roque-Barreira and M.H.S. Goldman are indebted to CNPq for

their research fellowships. This work was supported by a grant

from FAPESP (00/09333-2).

References

[1] A. Panunto-Castelo, M.A. Souza, M.C. Roque-Barreira, J.S. Silva, KM(+),

a lectin from Artocarpus integrifolia, induces IL-12 p40 production by

macrophages and switches from type 2 to type 1 cell-mediated immunity

against Leishmania major antigens, resulting in BALB/c mice resistance to

infection, Glycobiology 11 (2001) 1035–1042.

[2] L.L.P. daSilva, A. Panunto-Castelo, M.H.S. Goldman, M.C. Roque-

Barreira, R.S. Oliveira, M.D. Baruffi, J.B Molfetta-Machado, Composi-

tion for preventing or treating appearance of epithelia wounds such as skin

and corneal wounds or for immunomodulating, comprises lectin, Patent

number WO20041008.

[3] W.J. Peumans, E.J.M. Van Damme, Lectins as plant defense proteins,

Plant Physiol. 109 (1995) 347–352.

[4] N. Sharon, H. Lis, History of lectins: from hemagglutinins to biological

recognition molecules, Glycobiology 14 (2004) 53–62.

[5] E.J. Van Damme, W.J. Peumans, A. Barre, P. Rouge, Plant lectins: a

composite of several distinct families of structurally and evolutionary

related proteins with diverse biological roles, Crit. Rev. Plant Sci. 17

(1998) 575–692.

[6] T. Hajto, K. Hostanska, T. Berki, L. Palinskas, F. Boldizsar, P. Nemeth,

Oncopharmacological perspectives of a plant lectin (Viscum album

Agglutinin-I): overview of recent results from in vitro experiments and

in vivo animal models, and their possible relevance for clinical

applications, eCAM 2 (2005) 59–67.

[7] C. Bies, C.M. Lehr, J.F. Woodley, Lectin-mediated drug targeting: history

and applications, Adv. Drug Deliv. Rev. 56 (2004) 425–435.

[8] J.C. Rosa, R. Garratt, L. Beltramini, K. Resing, M.C. Roque-Barreira,

L.J. Greene, KM+, a mannose-binding lectin from Artocarpus integri-

folia: amino acid sequence, predicted tertiary structure, carbohydrate

recognition, and analysis of the beta-prism fold, Protein Sci. 8 (1999)

13–24.

[9] M.C. Roque-Barreira, A. Campos-Neto, Jacalin: an IgA-binding lectin,

J. Immunol. 134 (1985) 1740–1743.

L.L.P. daSilva et al. / Biochimica et Biophysica Acta 1726 (2005) 251–260260

[10] M.V. Sastry, P. Banarjee, S.R. Patanjali, M.J. Swamy, G.V. Swarnalatha,

A. Surolia, Analysis of saccharide binding to Artocarpus integrifolia lectin

reveals specific recognition of T-antigen (beta-D-Gal(1–3)D-GalNAc),

J. Biol. Chem. 261 (1986) 11726–11733.

[11] G.L. Hortin, B.L. Trimpe, Lectin affinity chromatography of proteins

bearing O-linked oligosaccharides: application of jacalin-agarose, Anal.

Biochem. 188 (1990) 271–277.

[12] S. Kabir, Jacalin: a jackfruit (Artocarpus heterophyllus) seed-derived

lectin of versatile applications in immunobiological research, J. Immunol.

Methods 212 (1998) 193–211.

[13] H. Kondoh, K. Kobayashi, K. Hagiwara, T. Kajii, Jacalin, a jackfruit

lectin, precipitates IgA1 but not IgA2 subclass on gel diffusion reaction,

J. Immunol. Methods 88 (1986) 171–173.

[14] H. Yang, T.H. Czapla, Isolation and characterization of cDNA clones

encoding jacalin isolectins, J. Biol. Chem. 268 (1993) 5905–5910.

[15] W.J. Peumans, B. Hause, E.J. Van Damme, The galactose-binding and

mannose-binding jacalin-related lectins are located in different sub-

cellular compartments, FEBS Lett. 477 (2000) 186–192.

[16] I.K. De Miranda-Santos, J.O.J. Mengel, M.M. Bunn-Moreno, A. Campos-

Neto, Activation of T and B cells by a crude extract of Artocarpus

integrifolia is mediated by a lectin distinct from jacalin, J. Immunol.

Methods 140 (1991) 197–203.

[17] R. Santos-de-Oliveira, M. Dias-Baruffi, S.M. Thomaz, L.M. Beltramini,

M.C. Roque-Barreira, A neutrophil migration-inducing lectin from

Artocarpus integrifolia, J. Immunol. 153 (1994) 1798–1807.

[18] S. Misquith, P.G. Rani, A. Surolia, Carbohydrate binding specificity of the

B-cell maturation mitogen from Artocarpus integrifolia seeds, J. Biol.

Chem. 269 (1994) 30393–30401.

[19] J.V. Pratap, A.A. Jeyaprakash, P.G. Rani, K. Sekar, A. Surolia, M. Vijayan,

Crystal structures of artocarpin, aMoraceae lectin withmannose specificity,

and its complex with methyl-alpha-d-mannose: implications to the

generation of carbohydrate specificity, J. Mol. Biol. 317 (2002) 237–247.

[20] A.A. Jeyaprakash, A. Srivastav, A. Surolia, M. Vijayan, Structural basis

for the carbohydrate specificities of artocarpin: variation in the length of a

loop as a strategy for generating ligand specificity, J. Mol. Biol. 338

(2004) 757–770.

[21] A. Barre, W.J. Peumans, M. Rossignol, G. Borderies, R. Culerrier, E.J.M.

Van Damme, P. Rouge, Artocarpin is a polyspecific jacalin-related lectin

with a monosaccharide preference for mannose, Biochimie 86 (2004)

685–691.

[22] L. Ganiko, A.R. Martins, E.M. Espreafico, M.C. Roque-Barreira,

Neutrophil haptotaxis induced by the lectin KM+, Glycoconj. J. 15

(1998) 527–530.

[23] L. Ganiko, A.R. Martins, E. Freymuller, R.A. Mortara, M.C. Roque-

Barreira, Lectin KM(+)-induced neutrophil haptotaxis involves binding to

laminin, Biochim. Biophys. Acta 1721 (2005) 152–163.

[24] A.N. Moreno, M.C. Jamur, C. Oliver, M.C. Roque-Barreira, Mast cell

degranulation induced by lectins: effect on neutrophil recruitment, Int.

Arch. Allergy Immunol. 132 (2003) 221–230.

[25] C.R. Teixeira, K.A. Cavassani, R.B. Gomes, M.J. Teixeira, M.-C. Roque-

Barreira, B.S. Cavada, J.S. Silva, A. Barral, M. Barral-Netto, Lectins

potential as adjuvants in immunization against Leishmania amazonensis

correlates to induction of inflammation in addition to the capacity to

induce IFN-g production, Vaccine (submitted for publication).

[26] C. Dean, P. Van Den Elzen, S. Tamaki, P. Dunsmuir, J. Bedbrook,

Differential expression of the eight genes of the petunia ribulose

biphosphate carboxilase small subunit multi-gene family, EMBO J. 4

(1985) 3055–3061.

[27] J. Sambrook, E.F. Fritsch, T. Maniatis, Molecular Cloning: A Laboratory

Manual, 2nd edR, Cold Spring Harbor Laboratory Press, Cold Spring

Harbor, NY, 1989.

[28] F.M. Ausubel, R. Brent, R.E. Kingston, D.D. Moore, J.G. Seidman, J.A.

Smith, K. Struhl, Short Protocols in Molecular Biology, Wiley, New York,

1995.

[29] M. Kozak, Structural features in eukaryotic mRNAs that modulate the

initiation of translation, J. Biol. Chem. 266 (1991) 19867–19870.

[30] M. Kozak, Initiation of translation in prokaryotes and eukaryotes, Gene

234 (1999) 187–208.

[31] E.J. Van Damme, B. Hause, J. Hu, A. Barre, P. Rouge, P. Proost, W.J.

Peumans, Two distinct jacalin-related lectins with a different specificity

and subcellular location are major vegetative storage proteins in the bark

of the black mulberry tree, Plant Physiol. 130 (2002) 757–769.

[32] H. Tateno, H.C. Winter, J. Petryniak, I.J. Goldstein, Purification,

characterization, molecular cloning, and expression of novel members of

jacalin-related lectins from rhizomes of the true fern Phlebodium aureum

(L) J. Smith (Polypodiaceae), J. Biol. Chem. 278 (2003) 10891–10899.

[33] H. Lilie, E. Schwarz, R. Rudolph, Advances in refolding of proteins

produced in E. coli, Curr. Opin. Biotechnol. 9 (1998) 497–501.

[34] F. Baneyx, Recombinant protein expression in Escherichia coli, Curr.

Opin. Biotechnol. 10 (1999) 411–421.

[35] F. Baneyx, M. Mujacic, Recombinant protein folding and misfolding in

Escherichia coli, Nat. Biotechnol. 22 (2004) 1399–1408.

[36] J.F. Kane, Effects of rare codon clusters on high-level expression of

heterologous proteins in Escherichia coli, Curr. Opin. Biotechnol. 6

(1995) 494–500.

[37] S.L. Martin, B. Vrhovski, A.S. Weiss, Total synthesis and expression in

Escherichia coli of a gene encoding human tropoelastin, Gene 154 (1995)

159–166.

[38] C. Gustafsson, S. Govindarajan, J. Minshull, Codon bias and heterologous

protein expression, Trends Biotechnol. 22 (2004) 346–353.

[39] D.B. Smith, K.S. Johnson, Single-step purification of polypeptides

expressed in Escherichia coli as fusions with glutathione S-transferase,

Gene 67 (1988) 31–40.

[40] A.K. Chopra, A.R. Brasier, M. Das, X.J. Xu, J.W. Peterson, Improved

synthesis of Salmonella typhimurium enterotoxin using gene fusion

expression systems, Gene 144 (1994) 81–85.

[41] R.B. Kapust, D.S. Waugh, Escherichia coli maltose-binding protein is

uncommonly effective at promoting the solubility of polypeptides to

which it is fused, Protein Sci. 8 (1999) 1668–1674.

[42] D.S. Waugh, Making the most of affinity tags, Trends Biotechnol. 23

(2005) 316–320.

[43] E. Goldman, A.H. Rosenberg, G. Zubay, F.W. Studier, Consecutive low-

usage leucine codons block translation only when near the 5V end of a

message in Escherichia coli, J. Mol. Biol. 245 (1995) 467–473.

[44] A.A. Sahasrabuddhe, S.M. Gaikwad, M.V. Krishnasastry, M.I. Khan,

Studies on recombinant single chain jacalin lectin reveal reduced affinity

for saccharides despite normal folding like native jacalin, Protein Sci. 13

(2004) 3264–3273.