Báo cáo khoa học: Cloning and characterization of the genes encoding toxic lectins in mistletoe (Viscum album L) pot

Cloning and characterization of the genes encoding toxic lectins
in mistletoe (
Viscum album
L)
Alma G. Kourmanova
1,2
, Olga J. Soudarkina
1
, Sjur Olsnes
3
and Jurij V. Kozlov
1,2
1
Engelhardt Institute of Molecular Biology, Russian Academy of Sciences and
2
The University of Oslo Centre for Medical Studies
in Moscow, Russia;
3
Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, Oslo, Norway
Leaves of mistletoe (Viscum album L) contain three toxic
lectins (type 2 ribosome-inactivating proteins) MLI, MLII,
and MLIII, differing in molecular mass and carbohydrate
specificity. Clones, containing sequences of three gene vari-
ants designated ml1p, ml2p,andml3p, were obtained using
PCR amplification from cDNA and from mistletoe genomic
DNA. The quantitative ratio of the ml1p, ml2p,andml3p
genes in genomic DNA was found to be 1.5 : 1 : 4,
respectively, whereas the ratio of their mRNA was
50 : 10 : 1. The quantitative prevalence of the ml1p tran-
script correlates well with the observation that MLI is
quantitatively dominant over MLII and MLIII in the

mistletoe extract. The sequences of the proteins encoded by
the ml1p, ml2p, and ml3p genes are identical to MLI by 98,
88, and 77%, respectively. The similarity to MLI of the
amino acid sequence encoded by the gene ml1p, the quan-
titative prevalent of its mRNA, as well as structural prop-
erties of the B-chain indicate that the gene, ml1p,
corresponds to MLI. Western blot analysis of recombinant
A-chains encoded by the three variants of mlp genes with the
monoclonal antibody MNA4 having differential affinity to
MLI, MLII and MLIII A-chains suggests that the ml2p and
ml3p genes correspond to MLII and MLIII, respectively.
Structural differences in the carbohydrate-binding sites
of the B-subunits of ML1p, ML2p, and ML3p probably
explain the difference in sugar specificity of MLI, MLII
and MLIII.
Keywords: mistletoe; ribosome-inactivating protein; toxic
lectin; Viscum album;viscumin.
Investigations over the last decades have shown that
extracts of several plants contain toxic proteins with lectin
properties. They bind by their one subunit, the B-chain,
to carbohydrate-containing structures at cell surfaces
[1–3]. The other subunit of the toxins, the A-chain, then
enters the cytosol and inactivates the ribosomes, leading
to cell death. Such toxins are also found in mistletoe
(Viscum album L). They are referred to as viscumins or
mistletoe lectins (ML). We shall here use the latter
designation.
As the toxins inactivate ribosomes, they are often referred
to as ribosome-inactivating proteins (RIP). Two major
groups of plant RIPs are distinguished classically according

to their molecular structure [2]. The type 1 RIPs are single
chain proteins (with few exceptions) resembling the toxin
A-chain in structure and function. They are essentially
nontoxic and they are abundantly present in a wide variety
of plants. The toxic type 2 RIPs are found only in few plants
and are heterodimers consisting of structurally and func-
tionally different A- and B-chains [2]. Type 3 was recently
introduced for RIPs with quite different structures: a single-
chain barley RIP, called JIP60, has an N-terminal domain
resembling type 1 RIPs and C-terminal domain with
unknown function [3].
The A-chain is a highly specific N-glycosidase that
irreversibly inactivates eukaryotic ribosomes [4]. The effi-
ciency of the A-chain is so high that penetration into the
cytosol of a single A-chain molecule is sufficient to induce
cell death [4].
The other polypeptide, the B-chain, is a lectin that is
responsible for binding to eukaryotic cell surface receptors
which induce endocytosis of the toxin and transfer of the
enzymatic subunit to the endoplasmic reticulum, where
penetration into the cytosol occurs [6,7].
The type 2 RIPs are synthesized as preprotoxins. The
mature form emerges after removal of first, a leader
sequence and then a linker that joins the A- and B- chains.
Another typical feature is the intronless gene structure [2].
The total sequence homology of these proteins, which
originate from taxonomically remote plant species, is not
high but all proteins have a similar fold [2,8,9].
Extract of mistletoe leaves has been used in folk medicine
since times immemorial. The toxic lectins MLI, MLII, and

MLIII were found to be present in the commercial mistletoe
preparation, Iscador, that is extensively used in paramedical
adjuvant therapy of cancer and for general immunostimu-
lation [10–12].
It was found previously that certain cancer cells are more
sensitive to the two RIP toxins, abrin and ricin, than cells
that were not able to generate tumors in animals [13]. As the
extensive use of mistletoe extracts in alternative medicine,
Correspondence to K. Jurij, Engelhardt Institute of Molecular Biology,
Vavilov street 32, 119991, Moscow, Russia. Fax: + 00 7 95 1352266,
Tel.: + 00 7 95 1359909, E-mail:
Abbreviations:IC
50
, 50% inhibitory concentration; MLI, MLII and
MLIII, mistletoe lectins 1, 2, 3; RIP, ribosome-inactivating protein.
Enzymes: N-riboside hydrolase (EC 3.2.2.22)
(Received 27 February 2004, revised 19 March 2004,
accepted 7 April 2004)
Eur. J. Biochem. 271, 2350–2360 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04153.x
it is important to obtain a clearer picture of the toxin
content in mistletoe.
The three toxic lectins MLI, MLII, and MLIII in
mistletoe differ in molecular mass and carbohydrate specifi-
city [14]. The most extensively studied one is MLI, whose
primary [15] and tertiary [16,17] structures are known. No
data are available on the primary structure of the two
remaining toxic lectins, MLII and MLIII.
It has been assumed that all three mistletoe toxins are
encoded by the same gene, and that post-translational
modifications cause the differences between the proteins [15].

The present work describes the full-size encoding
sequences of the mistletoe toxic lectin genes. Analysis of
the primary structure of the genes and the encoded proteins
enabled us to isolate three gene variants in good agreement
with the existence of three toxic lectins in mistletoe leaves.
We have shown that gene ml1p, encoding MLI, is
transcribed more efficiently than the other two genes, and
that this is probably the reason for the quantitative
prevalence of MLI in extract of mistletoe leaves [18]. The
A-chains encoded by the three variants of the genes were
expressed in Escherichia coli and their biological activity was
shown. Interaction of the recombinant A-chains with
monoclonal antibodies having differential affinity to MLI,
MLII and MLIII A-chains suggests that the other two
genes ml2p and ml3p correspond to MLII and MLIII,
respectively.
Materials and methods
Plant material
Young leaves of mistletoe (Viscum album L) were harvested
at the end of May from plants growing on a poplar tree
(Populus alba L) in the Poltavskaya region of South
Ukraine. The leaves were frozen in liquid nitrogen and
stored at )50 °C until use.
Plant RNA isolation and cDNA synthesis
Total RNA was prepared from young mistletoe leaves using
the guanidine hydrochloride extraction method [19]. Poly
(A)
+
-RNA was prepared from the total RNA using
paramagnetic oligo(dT) beads (Promega). A Universal

Riboclone cDNA Synthesis Kit (Promega) was used to
convert poly(A)
+
-RNA into double-stranded cDNA.
Oligo(dT) was used to prime the first strand. The primer
was annealed to 1–2 lgofpoly(A)
+
-RNA.
Amplification of genomic ML fragments using
degenerate primers
On the basis of the published amino acid sequences of the
A- and B-chains of mistletoe lectin I (MLI) [20,21], three
different degenerate primers were designed. Using primers
5¢-CAIACIACIGGNGARGARTA-3¢ and 5¢-ATIGGRT
TITTIAAIACNCCRTC-3¢ it was possible to amplify a
650 bp fragment encoding part of the A-chain. Then, using
primers 5¢-CTCGAGCTGGAGACGAGTTGGG-3¢ and
5¢-CKIATISWICCRTCICCRTA-3¢ an overlapping frag-
ment encoding a 14 amino acids linker, the larger part of the
B-chain was obtained. The sequences were different from
MLI. On the basis of the sequences, a single-site polymerase
chain reaction (PCR; via single oligonucleotide ligation) was
performed according to Lin et al.[22].ThePstI-digested
genomic mistletoe DNA was ligated to the PstI oligonucleo-
tide (5¢-AGCGTTGACAGCCAGCTGCA-3¢). The first
round of amplification was accomplished with the specific
primer 5¢-GGTGAGAACGCAGTCAGATGCTAGG-3¢
and PstI, and the second round with another specific
primer 5¢-GACCGGATCCCTCTGGGTAGAGAG-3¢
and PstI. The product of the second round of PCR was

cloned and the 1066 bp sequence encoding 125 amino acids
of the A-chain, the 19 amino acids linker, and 216 amino
acidsoftheB-chainoftheml gene was obtained. The
sequence was different from that obtained using degenerate
primers. The obtained sequence data were used to design
primers for RACE.
5¢ and 3¢ rapid amplification of cDNA ends (RACE)
RACE was performed according to Frohman et al.[23].
The sequences of the primers used for 5¢ RACE were:
5¢-GCTCCACCAACACAAATC-3¢ for reverse transcrip-
tion of poly(A)
+
-RNA with AMV-Reverse Transcriptase
(USB, Cleveland, OH, USA) and 5¢-GGATCGTAGACT
GACGCAAGAGTGG-3¢ with the Abridged Anchor Pri-
mer (AAP) (Gibco BRL) for the following amplification.
The sequences of the primers used for 3¢ RACE were:
5¢-TCTAGA(T)
20
-3¢ for the first cDNA strand priming,
and 5¢-GCCCCTCGCGAGGTAACC-3¢ with 5¢-TCTA
GA(T)
20
-3¢ or 5¢-CCGTAATCAATATTGTTAGCTGC
AG-3¢ with 5¢-TCTAGA(T)
20
-3¢ for the following amplifi-
cation. PCR reactions were set up in 20 lL using 5–15 ng
cDNA as template, 0.25 l
M

of each primer, 0.2 m
M
dNTP,
the buffer supplied with Taq DNA polymerase (Promega),
2.0 m
M
MgCl
2
and 1.0 unit of Taq DNA polymerase were
added to the reaction. The thermal profile was 94 °C, 1 min;
55 °C, 1 min; 72 °C, 1 min; 30 cycles. The PCR products
were analyzed on 2% agarose gels.
Amplification of the full-length coding sequence
of the
mlp
gene
To amplify the full-length coding sequences a set of primers
was derived from 3¢ and 5¢ untranslated regions (UTR)
obtained by RACE. The sequences of the primers were:
5¢AAAATCTAGAGAAGCAAGGAACAATGAATG-3¢
(5¢UTR) containing the XbaI recognition site for cloning,
5¢-AAAAATGCATGAAGTTGATTGCTTGCATTAAC
TCAT-3¢ (3¢UTR), 5¢-AAAAATGCATAGGGATGAA
GTTGATTGCTTGCC-3¢ (3¢UTR) containing the NsiI
recognition site for cloning, and 5¢-CACAAGGTGGC
TAAGGCTTCTTCCG-3¢ (3¢UTR), SphI recognition site
contained in all 3¢UTR clones was used for cloning in the
latter case. PCR reactions (GeneAmp PCR System 2400;
Perkin Elmer) were set up in 20 lL using 200 ng genomic
DNA or 5–15 ng cDNA as template, 0.25 l

M
of each
primer, 0.2 m
M
dNTP, the buffer supplied with Vent DNA
polymerase (BioLab) and 1.0 U of Vent DNA polymerase.
The thermal profile was: denaturation at 94 °C, 3 min; then
94 °C, 1 min; 54 °C, 1 min; 72 °C, 3 min for 5 cycles;
94 °C, 20 s; 58 °C, 40 s; 72 °C,3minplus10sforeach
following cycle for 25–28 cycles. The amplified DNA
Ó FEBS 2004 Structure of the mistletoe lectin genes (Eur. J. Biochem. 271) 2351
fragments were purified from agarose gel (GFX PCR DNA
and Gel Band Purification Kit, USB) and cloned into
pGEM7zf(+) or pGEM3zf(+) vector (Promega).
DNA isolation and Southern blot analysis
Whole genomic DNA was isolated from young leaves
according to Murray & Thompson [24]. The DNA prepar-
ation was treated with RNase to remove any contaminating
RNA. Approximately 10 lg of DNA was digested with
restriction endonucleases and subjected to electrophoresis in
a 0.7% agarose gel. Southern transfer and hybridization were
performed using Zeta-Probe Blotting membranes (Bio-
Rad) according to the manufacturer’s recommendations.
The blots were probed with the
32
P-radiolabelled 1 kb SmaI–
PstI fragment of the ml2p clone. The fragment was labelled
using the ÔNick translation kitÕ (Amersham Biosciences).
Assessment of the quantitative ratio of the three
variants of

mlp
genes in mistletoe genomic DNA and
their transcripts in mRNA
Two primers, universal for all obtained variants of ml gene
sequences were designed. The primer sequences were:
5¢-TGCTTGAGCTGGAGACGAGTTGG-3¢ (M1) and
5¢-CCATTGGATCGAATGGTTCCATC-3¢ (M2), the
PCR products being of 415 or 430 bp length. PCR reactions
were set up in 20 lL using 200 ng genomic DNA or
5–15 ng cDNA as template, 0.25 l
M
of each primer,
0.2 m
M
dNTP, the buffer supplied with Pfu DNA polym-
erase (Fermentas), 2.0 m
M
MgSO
4
, Pfu DNA polymerase
1.0 U added per 20 lL reaction mixture. The thermal
profile was: 94 °C, 5 min for initial template denaturation;
then 94 °C, 1 min; 50 °C, 1 min; 72 °C, 1 min for 3 cycles;
94 °C, 30 s; 57 °C, 40 s; 72 °C, 1 min; +2 s for each
following cycle for 26 cycles.
The amplified fragments (cDNA or genomic) were
purified from agarose gel and labelled with [
33
P]ATP[cP]
using T4 polynucleotide kinase. The labelled DNA was

purified from the unincorporated labelled nucleotides by gel
filtration through Sephadex G-50 and concentrated by
ethanol precipitation. Equal amounts (0.3 lg) of the
fragment were digested with either SalIorPstI, or with
both restriction endonucleases. The labelled digests were
separated on 2% agarose gels and the bands corresponding
to the three variants of mlp genes were excised. The slices
were dissolved in 0.5 mL 2
M
HCl, scintillation fluid
Aquasol-2 (PerkinElmer) was added and the radioactivity
was measured in a scintillation counter. The quantitative
ratio of the amplification products of the three gene variants
in the amplified fragment was assessed by calculation of the
ratio of radioactivity of the corresponding bands.
Quantitative PCR
The quantity of mlp genes in mistletoe genomic DNA was
assessed by quantitative PCR generally according to
Diviacco et al. [25]. The competitor DNA was constructed
by subcloning the HindIII-SalI fragment from the ml3.1p
clone to the ml2p clone applying HindIII and AgeIsites.
Obtained plasmid DNA was linearized before using as
competitor template. The universal M1 and M2 primers
were used for amplification of a fixed amount of mistletoe
genomic DNA (200 ng) mixed with increasing amounts of
competitor DNA (500–10 000 molecules). PCR reactions
were set up in 20 lLusing0.25l
M
of each primer (half of
the M2 primer being

33
P-end-labelled), 0.2 m
M
dNTP, the
buffer supplied with Taq DNA polymerase (Promega),
MgCl
2
added to a final concentration of 2.0 m
M
1.0 U of
Taq DNA polymerase was added per 20 lL reaction. The
thermalprofilewas:94°C, 5 min for initial template
denaturation; then 94 °C, 1 min; 50 °C, 1 min; 72 °C,
30 s; for 3 cycles then 94 °C, 30 s; 57 °C, 40 s; 72 °C, 30 s;
for 26 cycles. PCR products were separated on 2% agarose
gel and the bands corresponding to genomic fragment
(length of 415/430 bp) and competitor fragment (length of
388 bp) were excised. Radioactivity of the bands was then
measured as for the assessment of the quantitative ratio of
the three variants of mlp genes. The background measure-
ment was performed for each lane by taking a slice of gel
just below the competitor band. The equivalence of
competitor and genomic template was reached when the
competitor DNA input was 4000 molecules.
Sequence analysis
DNA sequencing was performed with commercial systems
T7 Sequenase version 2.0 DNA Sequencing Kit and
Thermo Sequenase Cycle Sequencing Kit (Amersham
Biosciences).
Construction of expression plasmids for the recombinant

A-chains of MLp
The mlp gene sequences encoding A-chains were subcloned
into the pET-28b(+) vector (Novagen). The sequences were
amplified with primers introducing restriction endonuclease
recognition sites (for the subcloning that followed), trans-
lation stop codons and some nucleotide changes replacing
codons of low usage in E. coli. Sequences of the primers and
pET28b(+) vector sites used for the subcloning are
presented in Table 1. The ligated DNA was transformed
by electroporation into competent E. coli DH10 cells and
the positive transformants were selected by restriction
analysis of plasmid DNA minipreparations of five to six
clones. The entire A-chain-coding sequences of selected
positive clones were checked by DNA-sequencing. For
subsequent affinity purification, all the expression plasmids
encode the MLp A-chains fused with vector-encoded
His-tag peptide on the N-terminus.
Expression of A-chains in
E. coli
Recombinant plasmids were introduced into E. coli strain
BL21(DE3)pLysS (Novagen) by calcium chloride-medi-
ated transformation. Optimal growth and expression
conditions for the recombinant proteins were established.
Overnight culture (2 mL) was grown at 37 °CinLuria–
Bertani medium (LB) containing kanamicin (30 lgÆmL
)1
)
and chloramphenicol (34 lgÆmL
)1
). The next day, cells

were harvested and resuspended in 2 mL fresh LB
medium and used to inoculate 200 mL medium containing
the antibiotics. Cultures were grown at 30 °CuntilA
600
reached 0.6–0.7 ( 3 h). Then, recombinant protein
2352 A. G. Kourmanova et al. (Eur. J. Biochem. 271) Ó FEBS 2004
expression was induced by addition of 1.0 m
M
isopropyl
thio-b-
D
-galactoside. Cells were harvested 2.5–3 h after
induction by centrifugation at 4000 g.
Isolation of recombinant A-chains
Purification of soluble His-tagged recombinant MLp
A-chains was performed using HisTrap Kit (Amersham
Biosciences). Elution was performed according to the
manufacturer’s instruction by adjusting the washing and
elution parameters. Fractions were analyzed by SDS/
PAGE, pooled and dialyzed against PN buffer (20 m
M
sodium phosphate buffer pH 7.2, 0.5
M
NaCl). Protein
concentration was determined by measuring the A
280
.
SDS/PAGE and Western blotting
Samples were boiled for 5 min in 1.0% SDS/50 m
M

dithiothreitol and run on 15% SDS/PAGE. Before electro-
phoresis, holotoxins MLI, MLII and MLIII [kindly provi-
ded by A. G. Tonevitsky (Institute of Transplantology and
Artificial Organs, Moscow, Russia)] were incubated with
dithiothreitol (50 m
M
)in20m
M
Tris, pH 8.0, 60 m
M
NaCl
at 37 °C for 30 min to reduce toxins. Proteins were
visualized with 0.1% Coomassie Blue R-350 in 10%
methanol (v/v)/10% acetic acid (v/v). For Western blots,
proteins were transferred to a Hybond-P membrane
(Amersham Biosciences) and the membranes were blocked
with 5.0% nonfat dry milk in TBS-T buffer (20 m
M
Tris/
HCl, 137 m
M
NaCl, pH 7.6, 0.1% Tween 20) overnight at
4 °C. The membranes were then incubated with MNA4
(2 lgÆmL
)1
)orTA7(2lgÆmL
)1
) monoclonal antibodies
(kindly provided by A. G. Tonevitsky, Institute of Trans-
plantology and Artificial Organs, Moscow, Russia) for 3 h

at room temperature. Then the antibody–antigen complexes
were probed with sheep anti(mouse IgG) Ig, horseradish
peroxidase linked whole antibody (Amersham Biosciences),
at 1 : 1000 dilution for 2 h at room temperature. Labelled
bands were detected using standard protocols of ECL
Western Blotting Detection Reagents kit (Amersham Bio-
sciences) and then exposed to film.
Activity of recombinant A-chains in reticulocyte lysate
The biological activity of the recombinant MLp
A-chains was determined by their ability to inhibit
[
3
H]Leu incorporation into protein in a cell-free system
(Rabbit Reticulocyte Lysate, Nuclease Treated from
Promega). Cell-free protein synthesis was performed
according to the manufacturer’s instructions. Reaction
samples not containing template RNA were incubated with
recombinant A-chains for 20 min at 37 °C and chilled to
0 °C. Then, template RNA was added and translation
reactions were carried out over 20 min at 30 °C. Ranges of
recombinant A-chain concentrations (0.1–60 ngÆmL
)1
)were
assayed with respect to controls. Each concentration was
assayed in triplicate.
Results
The nomenclature of the mistletoe toxic lectins
and their genes
Toxic lectins isolated from mistletoe were designated as
mistletoe lectins (ML) [14]. The abbreviation ML is followed

by the Roman numerals I, II, and III (MLI, MLII, MLIII)
and refer to different carbohydrate specificity and molecular
mass of each of the proteins. MLI is also called viscumin. The
full-size coding sequence of the mlI gene was cloned [15] and
designated rML. The same designation was used for the
protein encoded by this gene. The sequenced mistletoe toxic
lectin genes are designated as ml1p, ml2p and ml3p and the
respective encoded proteins as ML1p, ML2p and ML3p.
The letter ÔpÕ refers to the geographic origin of the plant
(Poltavskaya region of the Ukraine).
The geographic origin of the plant may be associated with
minor differences in the primary structure of the proteins.
The ml1p gene corresponds to MLI. As no structural data
for MLII and MLIII were available, ml2p and ml3p genes
were brought into correlation with MLII and MLIII by an
immunological approach (see below). The ml3.1p gene
encodes a protein that differs from ML3p by 14 amino acid
residues, but has common structural features with ML3p
that distinguish it from ML1p and ML2p. Therefore we
consider ml3.1p as a gene that encodes an isoform of the
protein encoded by ml3p.
Cloning of the sequences of
mlp
genes
Approximately 2 kbp products were obtained by PCR with
several pairs of specific primers complementary to 5¢-and
3¢-untranslated regions of 5¢-and3¢-RACE clones using
Table 1. Primers used for MLp A-chain coding sequence amplification and pET28b(+) vector sites used for the following subcloning. Restriction
endonuclease recognition sites (bold) and translation stop codons (underlined) introduced by primers are indicated.
MLp chain Primers

pET-28b(+) vector sites
used for cloning
ML1p
A-chain
5¢-
GATATA
CATATG
TACGAGCGTCTTCGTCTTCGTGTTACGCATC
-3¢
5¢-
CACAC
CTCGAG
TTATTAAGAAGAAGACGGACGCTCACCGCA
-3¢
NdeI and
XhoI
ML2p
A-chain
5¢-
GATATA
CATATG
TACGAGCGTCTTCGTCTTCGTGTTACGCATC
-3¢
5¢-
CACAC
CTCGAG
TTATTAAGAAGAAGACGGACGGTCCCGGCATAC
-3¢
NdeI and
XhoI

ML3p
A-chain
5¢-
GATATA
CATATG
TACCGTCGTATTAGCCTTCGTGTCACGGAT
-3¢
5¢-
CACAC
GAATTC
TTATTAAGAAGAAGAAGAACGGTCCCTGCATAC
-3¢
NdeI and
EcoRI
ML3.1p
A-chain
5¢-
GATATA
CATATG
TACGAGCGTCTTCGTCTTCGTGTTACGCATC
-3¢
5¢-
CACAC
GAATTC
TTATTAAGAAGAAGAAGAACGGTCCCTGCATAC
-3¢
NdeI and
EcoRI
Ó FEBS 2004 Structure of the mistletoe lectin genes (Eur. J. Biochem. 271) 2353
genomic DNA or cDNA as a template. The PCR products

were cloned and sequenced. Partial sequencing of more than
40 cDNA and genomic clones revealed three groups of
the sequences according to the extent of similarity. Thus,
sequencing of a region of about 200 bp gave either identical
sequences or the sequences differing by 1–2 bases, or the
sequences to differ markedly by 10–20 bases. In the first and
second cases, the clones were ascribed to one group and in
the latter case, to different groups. One genomic and cDNA
clone from each group was taken for complete sequencing.
Another genomic clone (ml3.1p) bearing a substitution in
the region of the A-chain active site was also taken. Here we
report the structure of four genomic clones ml1p, ml2p,
ml3p and ml3.1p containing full-size coding sequences of
mistletoe lectin genes. The proteins encoded by genomic
(ml1p, ml2p, ml3p) and cDNA clones (cml1p, cml2p and
cml3p) of the same group are identical by 96–97% and have
the same structural features (see below). By combining
overlapping RACE-clones and cDNA we obtained full-size
transcripts of the mistletoe toxic lectin genes.
Like other type 2 RIPs (such as ricin and abrin) [2], the
mlp genes do not contain introns and they encode the toxic
lectins in the form of a single chain precursor (Fig. 1). The
ML1p precursor, like the rML-precusor [15], is 564 amino
acids in length. The ML2p and ML3p precursors are 569
and 567 amino acids in length, respectively. The precursors
contain a 33 amino acid N-terminal leader peptide and a
small linker peptide joining the A- and B-chains – these are
removed during protein maturation.
Comparison of amino acid sequences of the MLp
precursors with that of rML revealed a marked difference

for the three variants. ML1p have the highest percentage of
identity (similarity) to rML: 98(99%), the similarity value
additionally includes similar amino acid positions so that it
is higher then that of identity. ML2p, ML3p and ML3.1p
are identical (similar) to rML by 88(91%), 78(87%) and
77(86%), respectively.
A-chains
The A-chain of the translation products of three mlp gene
variants, like the rML A-chain, consists of 254 amino acids.
The ML1p, ML2p and ML3p A-chains are identical
(similar) to the rML A-chain by 98(99%), 91(94%) and
83(90%), respectively.
Sequence analysis suggests that the A-chains are hetero-
geneous in the number of potential N-glycosylation sites
having either none (ML2p) or one site (ML1p, ML3p,
ML3.1p (Figs 1 and 2).
B-chains
The ML1p and ML2p B-chains, like the rML B-chain,
are 263 amino acids in length and the ML3p B-chain is 266
amino acids in length owing to the insertion of RGT(128-
130). The ML1p, ML2p and ML3p B-chains are identical
(similar) to the rML B-chain by 98(98%), 86(90%) and
71(82%), respectively.
The MLp B-chains have different patterns of potential
N-glycosylation sites (Figs 1 and 2). The ML1p B-chain has
thesamesitesasrML.TheML2pB-chainhasthree
potential N-glycosylation sites, two of which are homolog-
ous to that of ML1p, while the ML3p B-chain has three
sites which are homologous to that of ML2p. The ML3.1p
B-chain has one site.

The linker between the A- and B-chains
The linker peptide of ML1p, ML3p and ML3.1p corres-
ponds to 14 amino acid residues, whereas for ML2p it
corresponds to 19 residues.
Comparison of the amino acid sequences of the MLp
variants and ricin
The MLp variants have a high overall sequence similarity
to structurally and biologically related ricin D [26]. The
A-chains of ML1p, ML2p and ML3p have 40(54–57%) of
identity (similarity) to the ricin A-chain.
The invariant amino acid residues involved with the
structure or action of the A-chain catalytic site [27] are
conserved in the MLp variants: Tyr17, Arg25, Tyr76,
Tyr115, Glu165, Arg168, Trp199 of the MLp A-chains
correspond to Tyr21, Arg29, Tyr80, Tyr123, Glu177,
Fig. 1. Schematic structure of preproricin
[26,28] and mature lectins encoded by mlp
genes. Domain structure of the B-chain and
position of the carbohydrate-binding sites are
indicated. Each B-chain domain is composed
of three homologous subdomains a, b, c and
linking k subdomain which joins the B-chain
to the A-chain or the first B-chain domain to
the second one. Potential N-glycosylation sites
(NXS or NXT) and disulfide bonds are
marked.
2354 A. G. Kourmanova et al. (Eur. J. Biochem. 271) Ó FEBS 2004
Arg180, Trp211 of the ricin A-chain (Fig. 2). One other
highly conserved residue Arg166 involved in the catalytic
site structure (corresponding to Arg178 of the ricin A-chain)

is also conserved in all MLp variants except ML3.1p, which
has the substitution of Arg166fiVal.
The B-chains of ML1p, ML2p and ML3p have 63(75%),
65(77%)and55(70%)ofidentity(similarity)tothericin
B-chain, respectively.
The residues forming the sugar-binding sites in the ricin
B-chain are generally conserved in the structure of the
MLp B-chains: Asp23, Gln36, Trp38, Asn47, Gln48 of the
MLp B-chains (with one exception for ML2p) correspond
to Asp22, Gln35, Tyr37, Asn46 and Gln47 of the ricin
B-chain in the first sugar-binding site. Asp235 (238 for
ML3p), Ile247 (250 for ML3p), Tyr249, Asn256 (259 for
ML3p), Gln257 (260 for ML3p) of the MLp B-chains (with
one exception for ML3p) correspond to Asp234, Ile246,
Tyr248, Asn255 and Gln256 of the ricin B-chain in the
second sugar binding site [28]. The exceptions are the
replacement of Trp38 fi Ser in the ML2p B-chain and
Tyr252 fi Phe in the ML3p B-chain (Fig. 2).
The highly conserved amino acid residues forming the
hydrophobic core of the ricin B-chain domains [28] are fully
conserved in the MLp B-chains, reflecting the similarity of
the tertiary structure of type 2 RIPs (Fig. 2). Thus, positions
corresponding to the ricin B-chain Trp49, 90, 131, 173, 216,
258 and Ile57, 98, 181, 223 are occupied by Trp and Ile,
respectively, in the MLp B-chains. The ricin B-chain Val21,
Leu46, 117, 152, 191, 233 may be replaced in the MLp
B-chains by related Leu (Leu22 of the ML3p B-chain
corresponding to the ricin B-chain Val21), Ile (Ile118 of the
ML3p B-chain corresponding to the ricin B-chain Leu117)
or Met (Met153 of the ML1p and ML2p B-chains and

Met156 of the ML3p B-chain corresponding to the ricin
B-chain Leu152; and Met234 of the ML1p and ML2p
B-chains corresponding to the ricin B-chain Leu233).
Estimation of the quantitative ratio of the three
mlp
gene variants in genomic DNA and their transcripts
in mRNA
The ratio estimation was performed based on the principles
of competitive PCR [25]. If the primers M1 and M2 are
universal for all obtained variants of mlp genes, and the
sequences flanked by the primers are similar and do not
differ much in length, then equal amplification of the
variants should be expected, and the ratio between the
amplification products should correspond to that between
gene variants in the DNA template.
The difference between the restriction maps of the three
mlp gene variants by SalIandPstI in the region flanked by
the universal primers allowed separation of the amplifica-
tion products (Fig. 3A). Before cutting, a terminal radio-
active label was introduced into the amplified fragments.
Fig. 2. Comparison of amino acid sequences of ricin [26], mistletoe lectin I (rML) [15] and deduced amino acid sequences of mlp genes. The signal
peptide, mature A-chain, linker and B-chain sequences are marked. Sequences of A- and B-chains are numbered to the right. Conserved amino acid
residues forming the active site of the A-chain are marked with asterisks [27]. The residues participating in galactose binding in the 1a and 2c
subdomains of the B-chain [28] are marked with m. d Denotes highly conserved residues forming the hydrophobic core of the B-chain domains.
Residues forming the active site of the A-chain and those participating in galactose binding, but distinct from the conserved ones, are marked in
bold. Joined arrows mark Cys residues that form intrachain disulfide bonds. n Shows the bond between the ricin B-chain domains. s Mark Cys
residues forming an interchain disulfide bond. Potential N-glycosylation sites are enboxed. The numbers above the sequences refer to the positions
of some residues along the A- and B-chains of ricin, which are discussed in the text. Sequences were aligned using the
MULTALIGN
software in the

default setup [44].
Ó FEBS 2004 Structure of the mistletoe lectin genes (Eur. J. Biochem. 271) 2355
Thus, specific restriction fragments used for estimation of
the ratio were labelled at a single end. The quantitative ratio
of amplification products of three mlp gene variants was
calculated as the ratio of radioactivity of the corresponding
restriction fragments.
Such estimations will be correct if the genomic DNA and
mRNA contain no ml gene variants with sequences that are
not fully complementary to the primers used. Such variants
will amplify at a lower efficiency. To correct partly for such
a possibility, a lowered annealing temperature was used in
the first cycles of amplification. Presence of gene variants
with different restriction maps would produce additional
fragments, or the amplification product would not be cut.
As a whole, the pattern of fragments formed by cutting
the amplification product was similar in genomic DNA and
cDNA. Restriction fragments corresponding to all three mlp
gene variants can be seen in Fig. 3B. However, their
amounts were quite different in genomic DNA and cDNA.
In genomic DNA, three mlp gene variants are present in
the ratio 1.5 : 1 : 4 for ml1p, ml2p and ml3p, respectively.
Fig. 3. Quantitative ratio assessment for the three mlp gene variants in mistletoe genomic DNA and their transcripts in mRNA. (A) Restriction maps of
the mlp gene variants in the region amplified with the universal M1 and M2 primers. Positions of the primers are marked by arrows. The SalIand
PstI restriction endonuclease digests of the PCR product, amplified with M1 and M2 primers and mistletoe genomic DNA or cDNA as template,
were used for the ratio assessment. PstI cuts only the M1-M2 amplification product of ml1p variants giving 169 and 246 bp fragments. SalI cuts the
amplification product of ml2p, giving 66, 121 and 243 bp fragments, and the amlpification product of ml3p and ml3.1p, giving 50 and 365 bp
fragments. The 246 bp PstI restriction fragment, 243 bp and 365 bp SalI restriction fragments were used for the ratio assessment. (B) Equal
amounts of the
33

P-end-labelled M1-M2 amplification product were subjected to complete digestion with SalI, PstIorbothsimultaneously.The
digestion products were analyzed by 2% agarose gel electrophoresis in the presence of ethidium bromide. Lanes M, size markers; lanes 1, uncut
PCR products; lanes 2, PstI digests; lanes 3, SalIdigests;lanes4,SalI/PstI digests; lane 5, the triple amount of SalI digest for the 365 bp digestion
fragment (marked by a solid arrow to the right) may be seen clearly. An additional SalI restriction fragment corresponding to none of mlp genes is
marked by an open arrow to the right. The bands corresponding to the above-mentioned fragments were excised and their radioactivity was
measured. Background measurements were performed on a slice of agarose gel approximately between the 400 and 240 bp bands in the lane with
the SalIorPstI digest. (C) The quantitative ratio of amplification products of the three gene variants in the amplified fragment is presented.
2356 A. G. Kourmanova et al. (Eur. J. Biochem. 271) Ó FEBS 2004
In contrast, in cDNA the ratio of the three gene variants
was 50 : 10 : 1 for ml1p, ml2p and ml3p, respectively.
No additional bands appeared after cutting the fragment
amplified on cDNA. However, an additional restriction
fragment of  380 bp was observed after hydrolysis by SalI
of the product amplified from genomic DNA (Fig. 3B).
Restriction analysis of this fragment using NruI, AatII,
AgeI, NarI, and VneI did not reveal similarity with any of
the mlp gene variants. This fragment is probably the product
of unspecific amplification or it corresponds to unexpressed
pseudogene material. After combined treatment with SalI
and PstI of the product amplified from genomic DNA, only
a small portion of smear that was excised from gel with the
specific M1-M2 fragment was observed as uncut material
( 400 bp) (Fig. 3B, lane 4).
From the ratio of the three gene variants in the cDNA,
gene ml1p is transcribed more efficiently than genes ml2p
and ml3p. This result correlates well with the observation
that MLI is quantitatively prevalent over MLII and MLIII
in mistletoe extracts [18]. It is interesting to note that
mapping of the 5¢ ends of the mRNA revealed three bands
of different intensity (two minor ones and one major one)

corresponding to 39–41 bases upstream from the initiator
AUG codon (data not shown). This could mean that there is
a unique transcription start site characteristic for each gene
variant or mlp genes are able to use more than a single
transcription start site.
Sizing of the family of toxic lectin genes in mistletoe
Assuming that there is a single copy of the ml2p gene per
haploid genome, then it follows from the ratio of three gene
variants that the size of the gene family makes up 6–7 genes
(1,5 : 1 : 4 then 1/2 +1 + 4 ¼ 6/7).
Attempts to estimate the size of the mistletoe lectin gene
family using genome Southern blot analysis did not give
informative results. Only restriction enzymes that did not
cut any of the obtained mlp gene variants in the region
homologous to the probe were used for the analysis.
Therefore, it may be expected that the number of hybrid-
izing fragments in the digests represents approximately the
sizeofthegenefamily.TheBglII, EcoRI (and also XbaI,
NcoI, data not shown) digests gave two hybridizing
fragments, one of which (the smaller) was represented by
a more intensive band (Fig. 4). At the same time, the AflII,
NsiI(andBclI, data not shown) digests gave a very similar
pattern of five bands, one of which was also more intensive
than the others. Thus, the different intensity of the bands
may be the result of a different copy number of the genes in
the hybridizing fragments. If so, the number of the bands do
not represent the family size. Thus, the mlp gene family size
may be more than five genes, confirming the value obtained
based on the quantitative ratio of the three mlp gene
variants.

Large genome size frequently occurrs in plants [29]. We
made attempts to determine the size of the genome using
competitive quantitative PCR. This may give an accurate
value provided that all variants of mlp genes in the genome
are fully complementary to the universal M1 and M2
primers used. If otherwise, the efficiency of the amplification
will be lower than that of the considered variants, and
this will result in an overestimation of the amount of
genomic DNA containing one copy of the mlp gene.
Nevertheless, taking into account the possible size of the mlp
gene family and the number of gene variants considered in
the construction of the M1 and M2 primers, the error may
be not large. In our estimate, one copy of the mlp gene is
contained in 50 pg (5 · 10
10
bp) genomic DNA. Assuming
thesizeofthemlp gene family is six to seven genes, the
genome size is 5 · 10
10
· 6–7 ¼ 3–3.5 · 10
11
bp, that is, the
mistletoe genome may be one of the largest known
eukaryotic genomes. It is probable that the failure of
our attempts to obtain positive clones by screening the
genomic library is due to the huge size of the mistletoe
genome.
Expression and biological activity of the recombinant
MLp A-chains
The 762 bp DNA fragments encoding the MLp A-chains

were subcloned into the pET-28b(+) expression vector and
expressed in E. coli host strain BL21(DE3)pLysS. The
major part of the recombinant A-chains appeared in
insoluble cytoplasmic fraction but significant amounts were
found as soluble material. The soluble recombinant
A-chains were purified to homogeneity by a single round
of affinity chromatography applying the His-tag sequences.
On Coomassie Blue R-350-stained SDS/polyacrylamide
gels, the recombinant ML1p and ML2p A-chains could be
seen as a single band  30 kDa. The recombinant ML3p
and ML3.1p A-chain preparations always gave an addi-
tional band of  60 kDa (which may be seen on Western
immunoblots with TA7 monoclonal antibody, see below) as
if a dimerization of the chains takes place. The additional
band also appeared when the ML1p and ML2p recombin-
ant A-chain preparations have been stored for a while. The
identity of the bands was confirmed by Western blot
analysis. The yield of recombinant MLp A-chains was 6,
4.4, 0.35 and 0.3 mg per litre of culture for ML1p, ML2p,
Fig. 4. Estimation of the size of the mistletoe lectin gene family. Sam-
ples (10 lg) of high-molecular mass mistletoe genomic DNA were
digested with either BglII, NsiI, AflII or EcoRI and the resulting
fragments were separated by electrophoresis on a 0.7% agarose gel and
Southern-blotted onto a nylon membrane. The membrane was probed
with a radiolabelled SmaI–PstIfragmentofml2pcloneandwashed
under low stringency conditions. Positions of the DNA size markers
are shown by arrows.
Ó FEBS 2004 Structure of the mistletoe lectin genes (Eur. J. Biochem. 271) 2357
ML3p and ML3.1p A-chains, respectively. The activity of
the four recombinant A-chains was assessed by their

inhibitory effects on the protein synthesis in rabbit reticu-
locyte cell-free system (Fig. 5). All the A-chains appeared to
be fully active in translation inhibiting excepting the ML3.1p
A-chain which was noticeably less (> 10·) active than the
others. The IC
50
values for the four proteins were about
1ngÆmL
)1
(ML1p A-chain), 4 ngÆmL
)1
(ML2p A-chain),
0.35 ngÆmL
)1
(ML3p A-chain), 50 ngÆmL
)1
(ML3.1p A-
chain). The IC
50
valuefortheML1pA-chainintheassay,
thus, coincides with that for the ricin A-chain [30]. It is
interesting to note that according to our preliminary results
(data not shown) the native reduced ML toxins have
generally the same relative inhibitory activity: MLIII >
MLI > MLII (the IC
50
values are 0.5, 0.7 and 0.2 lgÆmL
)1
for MLI, MLII and MLIII, respectively).
Immunological identification of the recombinant

MLp A-chains
Western blot analysis was performed in order to correlate
the ml2p and ml3p genes with MLII and MLIII. Two
monoclonal antibodies (MNA4 and TA7) with differential
specificity against MLI, MLII and MLIII were used. The
monoclonal antibody TA7 interacts with A-chains of MLI,
MLII and MLIII [31]. MNA4 possesses the specificity to
MLI and MLII A-chains and reacts weakly with MLIII
A-chain [32]. Western blot analysis has shown that MNA4
efficiently binds to ML1p and ML2p A-chains and reacts
poorly with the ML3p A-chain (Fig. 6). Binding of the
recombinant MLp A-chains with TA7 was detected in all
samples. Interaction of the native A-chains with TA7 and
MNA4 in comparison with recombinant A-chains showed
the same binding.
Discussion
The present work reports the cloning and characterization
of three genes from mistletoe that encode toxic lectins that
are related to ricin. We were unable to obtain the clones by
screening genomic and cDNA libraries and we have
therefore used a PCR approach instead.
Three toxic lectins, MLI, MLII and MLIII, differing in
sugar specificity are isolated from mistletoe by affinity
chromatography [14]. Based on the amino acid sequence
of MLI, cloning of full-length coding sequences of
mistletoe lectin genes has revealed three variants. One of
the genes, ml1p, obviously corresponds to MLI whose
structure is known at the protein and nucleotide level.
Other two genes, ml2p and ml3p encode proteins mark-
edly differing from that of ml1p: the ML2p and ML3p

precursors are 87(91%) and 77(86%) identical (similar) to
the ML1p precursor. The ML2p and ML3p B-chains also
Fig. 5. Biological activity of the recombinant MLp A-chains. Recom-
binant ML1p (j), ML2p (m), ML3p (s) and ML3.1p (n)A-chains
were tested for their ability to inhibit cell-free protein synthesis in the
rabbit reticulocyte assay. The IC
50
values for the four proteins
were  1ngÆmL
)1
(ML1p A-chain), 4 ngÆmL
)1
(ML2p A-chain),
0.35 ngÆmL
)1
(ML3p A-chain), 50 ngÆmL
)1
(ML3.1p A-chain).
Fig. 6. Western immunoblot analysis of the recombinant MLp A-chains. From left to right: a dilution series of recombinant ML1p (lanes 1–3), ML2p
(lanes 4–6), ML3p (lanes 7–9) A-chains and reduced native MLI (lanes 10–11), MLII (lanes 12–13), MLIII (lanes 14–15) were electrophoresed on
15% SDS/polyacrylamide gel. Recombinant or native proteins (1.0 lg) were loaded on lanes 1, 4 and 7; 0.7 lg on lanes 2, 5, 8, 10, 12 and 14;
0.35 lg on lanes 3, 6, 9, 11, 13 and 15. Separated proteins were then transferred onto Hybond-P membrane (Amersham Biosciences). The blot was
probed with monoclonal antibodies TA7 or MNA4 followed by anti-mouse IgG–HRP (Amersham Biosciences). The blots were developed by the
enhanced chemiluminescence (ECL) method (Amersham Biosciences).
2358 A. G. Kourmanova et al. (Eur. J. Biochem. 271) Ó FEBS 2004
lack the structural feature of the MLI B-chain allowing its
reversible dimerization (not characteristic for MLII and
MLIII) – the loss of one disulfide bridge in the first
domain of B-chain after Ser40 fi Cys, corresponding to
Cys39 in ricin [33].

The protein structure of MLII and MLIII is not known
to date but they can be identified by sugar specificity and by
immunological methods. The last approach has been used
here to help reveal the relationship of ML2p and ML3p to
MLII and MLIII. The ML1p, ML2p, ML3p and ML3.1p
A-chains were expressed in E. coli in soluble and biologic-
ally active form. The recombinant ML3p A-chain was
recognized by the monoclonal antibody, TA7, which is
specific for all the MLI, MLII and MLIII A-chains [31] and
was not recognized by the specific for the MLI and MLII
A-chains monoclonal antibody MNA4 [32]. The data
suggest that ML3p corresponds to MLIII and ML2p to
MLII. Further evidence for the correspondence would be
the sugar specificity of ML2p and ML3p.
Like other type 2 RIPs, the mistletoe toxic lectins are
synthesized in the form of a single chain precursor. The
N-terminal leader sequence directs the toxic lectin precursor
to the endoplasmic reticulum, where it is split off [34] and
the linker peptide (reported to contain a signal for toxin
transport into vacuoles) is then excised [35,36]. The overall
sequence homology of the MLp linker peptide is low except
for the central region containing the sequence LVIRPV
which is of high homology to ricin and may reflect the
biological role of the sequence (Fig. 2).
The sequences of the catalytic A-chain of the MLp
toxins were found to differ. However, the amino acids
forming the catalytically active center of the A-chains of
the MLp toxins are fully conserved and retained in
corresponding positions, as in the A-chain of ricin [27] and
MLI [16,17,37], but with one exception, the Val166 fi Ala

substitution (corresponding to Ala178 of ricin) was found
in ML3.1p. It has been shown that Ala178 in ricin located
in the region of the catalytic site of the A-chain plays a
structural role and any larger residue would interfere with
one of the bonds of the invariant Tyr21 involved in
stabilization of the active center structure [27]. The
substitution possibly causes the low inhibitory activity of
the recombinant ML3.1p A-chain in cell-free translation
systems when compared with the other variants of the
recombinant MLp A-chains.
As for ricin, the C-terminal region of the MLp A-chains
contains a highly hydrophobic stretch of amino acids
(residues 236–246). This hydrophobic tail could function as
a signal peptide and initiate A-chain translocation across
the intracellular membrane [38]. An important role of the
Pro250 residue in the membrane translocation of ricin
A-chain was demonstrated using site-directed mutagenesis
[39]. A toxin containing the Pro250 fi Ala mutation
showed a dramatic effect, with 170-fold reduction in
cytotoxicity to Vero cells. In the hydrophobic region of
the MLp, A-chains only ML3p contains the corresponding
Pro residue, whereas the others contain Ala (Fig. 2).
The mistletoe toxic lectins are known to differ in their
sugar specificity. Thus, MLI is specific for
D
-galactose (Gal),
MLIII has higher affinity for N-acetyl-
D
-galactosamine
(GalNAc) than for Gal, and MLII binds to both sugars

with approximately equal affinity [14]. The comparison of
the sequences of ML2p, ML3p and ML3.1p with the
sequence of ricin also revealed a difference in the residues
forming the carbohydrate binding site. Thus, the conserved
residue Trp38 in the first site, corresponding to Trp37 in
ricin, is replaced by Ser. Such a substitution may result in a
marked lowering in affinity for Gal in ML2p, as it was
shown by site-directed mutagenesis for ricin [40]. The
introduced mutation of Trp37 fi Ser noticeably decreased
the efficiency of its binding to Gal. In the second binding site
of the ML3p and ML3.1p B-subunits, the replacement of
the conserved Tyr252 fi Phe, corresponding to Tyr248 in
ricin, is observed. A similar conversion is present in the
structure of the second carbohydrate-binding site of SNAV
[type 2 RIP from bark of elderberry (Sambucus nigra)],
having a 20-fold higher affinity for GalNAc when compared
with Gal [41]. The substitutions of amino acid residues
which are responsible for carbohydrate binding may change
the sugar-binding specificity of ML2p and ML3p compared
to MLI. The different lectin specificity of MLI, MLII and
MLIII is probably the result of different structures of the
carbohydrate-binding sites of these toxic lectins.
Formation of gene families is a characteristic feature of
many known RIPs [2]. Thus, although the extent of
homology in ricin and ricinus hemagglutinin is very high
[42], they are encoded by different genes. The ricin-like gene
family is probably comprises about eight members [43].
Different genes appear to encode the mistletoe toxic lectins.
As it was estimated using the quantitative ratio of the three
mlp gene variants in genomic DNA and Southern blot

analysis the size of the gene family is about six/seven genes.
Acknowledgement
This work was supported by the Russian Foundation for Basic
Research (project no. 04-04-49854).
References
1. Olsnes, S. & Pihl, A. (1982) Toxic lectins and related proteins. In
Molecular Action of Toxins and Viruses (Cohen, P & van Hey-
ningen, S., eds). pp. 51–105. Elsevier Biomedical Press, New York.
2. Barbieri, L., Batelli, M.G. & Stirpe, F. (1993) Ribosome-
inactivating proteins from plants. Biochim. Biophys. Acta. 1154,
237–282.
3. Peumans, W.J., Hao, Q. & Van Damme, E.J.M. (2001) Ribo-
some-inactivating proteins from plants: more then RNA N-gly-
cosidases? FASEB J. 15, 1493–1506.
4. Endo, Y. & Tsurugi, K.&.Lambert, J.M. (1988) The site of action
of six different ribosome- inactivating proteins from plants on
eucaryotic ribosomes. Biochem. Biophys. Res. Commun. 150,
1032–1036.
5. Eiklid,K.,Olsnes,S.&Pihl,A.(1980)Entryoflethaldosesof
abrin, ricin and modeccin into the cytosol of Hela cells. Exp. Cell.
Res. 120, 321–326.
6. Rapak, A., Falnes, P.O. & Olsnes, S. (1997) Retrograde trans-
port of mutant ricin to the endoplasmic reticulum with sub-
sequent translocation to cytosol. Proc. Natl Acad. Sci. USA 94,
3783–3788.
7. Wesche, J., Rapak, A. & Olsnes, S. (1999) Dependence of ricin
toxicity on translocation of the toxin A-chain from the endo-
plasmic reticulum to the cytosol. J. Biol. Chem. 274, 34443–34449.
8. Rutenber, E., Ready, M. & Robertus, J.D. (1987) Structure and
evolution of ricin B chain. Nature 326, 624–626.

Ó FEBS 2004 Structure of the mistletoe lectin genes (Eur. J. Biochem. 271) 2359
9. Tahirov,T.H.,Lu,T H.,Liaw,Y C.,Chen,Y L.&Lin,J Y.
(1995) Crystal structure of abrin-at 2.14 A
˚
. J. Mol. Biol. 250,
354–367.
10. Holtskog, R., Sandvig, K. & Olsnes, S. (1988) Characterization of
a toxic lectin in Iscador, a mistletoe preparation with alleged
cancerostatic properties. Oncology 45, 172–179.
11.Jung,M L.,Baudino,S.,Ribereau-Gayon,G.&Beck,J P.
(1990) Characterization of cytotoxic proteins from mistletoe.
(Viscum Album L.) Cancer Lett. 51, 103–108.
12. Kreitman, R.J. (2000) Immunotoxins. Expert Opin. Pharmaco-
ther. 1, 1117–1129.
13. Lin, J Y., Tsern, K Y., Chen, C C., Lin, L T. & Tung, T C.
(1970) Abrin and ricin: new anti-tumor substances. Nature 227,
292–293.
14. Franz, H., Ziska, P. & Kindt, A. (1981) Isolation and properties of
three lectins from mistletoe (Viscum album L.). Biochem. J. 195,
481–484.
15. Eck, J., Langer, M., Mo
¨
ckel,B.,Baur,A.,Rothe,M.,Zinke,H.&
Lentzen, H. (1999) Cloning of mistletoe lectin gene and char-
acterization of the recombinant A-chain. Eur. J. Biochem. 264,
775–784.
16. Sweeney, E.C. & Palmer, R.A. (1981) Crystallization of the
ribosome- inactivating protein MLI from Viscum album
(Mistletoe) complexed with b-
D

-Galactose. J. Mol. Biol. 234,
1279–1281.
17. Krauspenhaar, R., Echenburg, S., Perbandt, M., Kornilov, V.,
Konareva, N., Mikailova, I., Stoeva, S., Wacker, R., Maier, T.,
Singh,T.,Mikhailov,A.,Voelter,W.&Betzel,C.(1999)Crystal
structure of mistletoe lectin I from Viscum album. Biochem. Bio-
phys. Res. Commun. 257, 418–424.
18. Olsnes, S., Stirpe, F., Sandvig, K. & Pihl, A. (1982) Isolation and
characterization of viscumin, a toxic lectin from Viscum album L.
(Mistletoe). J. Biol. Chem. 257, 13263–13270.
19. Logemann, J., Schell, J. & Lothar, W. (1987) Improved method
for the isolation of RNA from plant tissues. Anal. Biochem. 163,
16–20.
20. Soler, M.H., Stoeva, S., Schwamborn, C., Wilhelm, S., Stiefel, Th
& Voelter, W. (1996) Complete amino acid sequence of the
A chain of mistletoe lectin I. FEBS Lett. 399, 153–157.
21. Soler, M.H., Stoeva, S. & Voelter, W. (1998) Complete amino acid
sequence of the B chain of mistletoe lectin I. Biochem. Biophys.
Res. Commun. 246, 596–601.
22. Lin, Z., Zhu, Y., Cui, X. & Li, H. (1995) Single-site polymerase
chain reaction through single oligonucleotide ligation. Anal. Bio-
chem. 231, 449–452.
23. Frohman, M.A., Dush, M.K. & Martin, G.R. (1988) Rapid
production of full-length cDNAs from rare transcripts: amplifi-
cation using a single gene-specific oligonucleotide primer. Proc.
NatlAcad.Sci.USA85, 8998–9002.
24. Murray, M.G. & Thompson, W.F. (1980) Rapid isolation of
high molecular weight plant DNA. Nucleic Acids Res. 8, 4121–
4325.
25. Diviacco, S., Norio, P., Zentilin, L., Menzo, S., Clementi, M.,

Biamonti, G., Riva, S. & Falaschi & Giacca, M. (1992) A novel
procedure for quantiative polymerase chain reaction by coampli-
fication of competitive templates. Gene 122, 313–320.
26. Halling, K.C., Halling, A.C., Murray, E.E., Ladin, B.F., Houston,
L.L. & Weaver, R. (1985) Genomic cloning characterization of
ricin gene from Ricinus communis. Nucleic Acids Res. 13, 8019–
8033.
27. Katzin, B.J., Collins, E. & Robertus, J.D. (1991) Structure of ricin
A-chain at 2.5 A
˚
. Proteins 10, 261–259.
28. Rutenber, E. & Robertus, J.D. (1991) Structure of ricin B-chain at
2.5 A
˚
Resolution. Proteins 10, 260–269.
29. Messing, J. (2001) Do plants have more genes than humans?
Trends Plant Sci. 6, 195–196.
30. O’Hare,M.,Roberts,L.M.,Thorpe,P.E.,Watson,G.J.,Prior,B.
& Lord, J.M. (1987) Expression of ricin A chain in Escherichia
coli. FEBS Lett. 216 (1), 73–78.
31. Temyakov, D.E., Agapov, I.I., Moisenovich, M.M., Prokof’ev,
S.K., Malyuchenko, N.V., Egorova, S.G., Pfueller, U., Zinke, H.
& Tonevitsky, A.G. (1997) Heterogeneity of mistletoe lectin
catalytic subunits assessed with monoclonal antibodies. Mol. Biol.
[Russian] 31, 536–541.
32. Malyuchenko, N.V., Moisenovich, M.M., Egorova, S.G.,
Kolesanova, E.F., Agapov, I.I., Solopova, O.S., Zaitsev, I.Z. &
Tonevitsky, A.G. (2000) Unfolding of subunit A during the
intracellular transport of Mistletoe Lectin I. Mol Biol. [Russian]
34, 152–159.

33. Sweeney, E.C., Tonevitsky, A.G., Palmer, R.A., Niwa, H., Pflu-
eller, U., Eck, J., Lentzen, H., Agapov, I.I. & Kirpichnikov, M.P.
(1998) Mistletoe lectin I forms a double trefoil structure. FEBS
Lett. 432, 367–370.
34. Lord, J.M. (1985) Precursors of ricin and Ricinus communis
agglutinin. Eur. J. Biochem. 146, 411–416.
35. Frigerio, L., Vitale, A., Lord, J.M., Ceriotti, A. & Roberts, L.M.
(1998) Free ricin A chain, proricin, and native toxin have different
cellular fates when expressed in tobacco protoplasts. J. Biol. Chem.
273, 14194–14199.
36. Frigerio, L., Jolliffe, N.A., Di Cola, A., Felipe, D.H., Paris, N.,
Neuhaus, J.M., Lord, J.M., Ceriotti, A. & Roberts, L.M. (2001)
The internal propeptide of the ricin precursor carries a sequence-
specific determinant for vacuolar sorting. Plant. Physiol. 126, 167–
175.
37. Eschenburg, S., Krauspenhaar, R., Mikhailov, A., Stoeva, S.,
Betzel, C. & Voelter, W. (1998) Primary structure and Molecular
modeling of Mistletoe lectin I from Viscum album. Biochem. Bio-
phys. Res. Com. 247, 367–372.
38. Chaddock,J.A.,Roberts,L.M.,Jungnickel,B.&Lord,J.M.(1995)
A hydrophobic region of ricin A chain which may have a role in
membrane translocation can function as an efficient noncleaved
signal peptide. Biochem. Biophys. Res. Commun. 217, 68–73.
39. Simpson, J.C., Lord, J.M. & Roberts, L.M. (1995) Point mutation
in hydrophobic C-terminal region of ricin a chain indicate that
P250 plays a key role in membrane translocation. Eur. J. Biochem.
232, 458–463.
40. Frankel, A.E., Burbage, C., Fu, T., Tagge, E., Chandler, J. &
Willingham, M.C. (1996) Ricin toxin contains at least three
galactose-binding sites located in B chain subdomane 1a,1b and

2c. Biochem. 35, 14749–14756.
41. Van Damme, E.J.M., Barre, A., Rouge, P., Van Leuven, F. &
Peumans, W.J. (1996) Characterization and molecular cloning of
Sambucus nigra agglutinin V (nigrin b), a GalNac- specific type-2
ribosome-inactivating protein from the bark of elderberry (Sam-
bucus nigra). Eur. J. Biochem. 237, 505–513.
42. Roberts, L.M. & Lord, J.M. (1981) The synthesis of Ricinus
communis Agglutinin. Eur. J. Biochem. 119, 31–41.
43. Tregear, J.W. & Roberts, L.M. (1992) The lectin gene family of
Ricinus communis: cloning of a functional ricin gene and three
lectin pseudogenes. Plant Mol. Biol. 18, 515–525.
44. Corpet, F. (1988) Multiple sequence alignment with hierarchical
clustering. Nucleic Acids Res. 16,10881–10890.
Supplementary material
The following material is available from http://www.
blackwellpublishing.com/products/journals/suppmat/EJB/
EJB4153/EJB4153sm.htm
Fig. S1. Mapping of the transcription start sites in the
mistletoe lectin genes by primer extension analysis.
Fig. S2. Sequencing strategy.
2360 A. G. Kourmanova et al. (Eur. J. Biochem. 271) Ó FEBS 2004