molecules
Article
Ribosome Inactivating Proteins from Rosaceae
Chenjing Shang 1,† , Pierre Rougé 2 and Els J. M. Van Damme 1, *
1
2
*
†
Department of Molecular Biotechnology, Faculty of Bioscience Engineering, Ghent University,
9000 Ghent, Belgium; shangchenjing@scsio.ac.cn
Unité Mixte de Recherche 152 Pharma Développement, Institut de Recherche pour le Développement,
Université Paul Sabatier, 31062 Toulouse, France; pierre.rouge@free.fr
Correspondence: ElsJM.VanDamme@ugent.be; Tel.: +32-9264-6086; Fax: +32-9264-6219
Present address: State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology,
Chinese Academy Sciences, 510301 Guangzhou, China.
Academic Editor: Derek J. McPhee
Received: 11 July 2016; Accepted: 18 August 2016; Published: 22 August 2016
Abstract: Ribosome-inactivating proteins (RIPs) are widespread among higher plants of different
taxonomic orders. In this study, we report on the RIP sequences found in the genome/transcriptome
of several important Rosaceae species, including many economically important edible fruits
such as apple, pear, peach, apricot, and strawberry. All RIP domains from Rosaceae share high
sequence similarity with conserved residues in the catalytic site and the carbohydrate binding sites.
The genomes of Malus domestica and Pyrus communis contain both type 1 and type 2 RIP sequences,
whereas for Prunus mume, Prunus persica, Pyrus bretschneideri, and Pyrus communis a complex set of
type 1 RIP sequences was retrieved. Heterologous expression and purification of the type 1 as well as
the type 2 RIP from apple allowed to characterize the biological activity of the proteins. Both RIPs
from Malus domestica can inhibit protein synthesis. Furthermore, molecular modelling suggests that
RIPs from Rosaceae possess three-dimensional structures that are highly similar to the model proteins
and can bind to RIP substrates. Screening of the recombinant type 2 RIP from apple on a glycan array
revealed that this type 2 RIP interacts with terminal sialic acid residues. Our data suggest that the
RIPs from Rosaceae are biologically active proteins.
Keywords: carbohydrate binding activity; molecular modeling; protein synthesis inhibition;
ribosome-inactivating proteins
1. Introduction
Ribosome-inactivating proteins (RIPs) are a large family of enzymes (EC.3.2.2.22) comprising
an rRNA N-glycosylase domain that is capable of catalytically inactivating ribosomes through the
removal of a specific adenine residue from a highly conserved α-sarcin/ricin loop within the large
rRNA [1]. Though RIPs have first been detected and characterized from plants, RIPs have also been
isolated and characterized from bacteria and fungi, and more recently RIP sequences have also been
reported in insects [2]. However, apart from the shiga and shiga-like toxins from bacteria [3] and a
few fungal RIPs from mushrooms [4], virtually all research concentrated on RIPs are from flowering
plants. Plant RIPs are classically subdivided in two major groups: Type 1 RIPs consist of a single
protein domain with rRNA N-glycosylase activity (RIP domain), whereas type 2 RIPs are chimeric
proteins built up of an N-terminal rRNA N-glycosylase domain (RIP domain) fused to a C-terminal
carbohydrate binding domain (lectin domain).
RIPs are widely distributed in the plant kingdom and RIP sequences have been reported for
at least 71 monocotyledonous and dicotyledonous species within the Angiospermae [5]. However,
RIP sequences are not ubiquitous in plants. For example, no RIP sequence could be retrieved from
the complete genome of Arabidopsis thaliana [6]. At present, RIPs have been reported frequently
Molecules 2016, 21, 1105; doi:10.3390/molecules21081105
www.mdpi.com/journal/molecules
Molecules 2016, 21, 1105
2 of 17
in the plant families Cucurbitaceae, Euphorbiaceae, Sambucaceae, Phytolaccaceae, Poaceae, and
Caryophyllaceae [7]. RIPs are not associated with (a) particular tissue(s) but are found in virtually all
plant parts (e.g., seeds, roots, leaves, bulbs, fruits, and bark) [8]. Both the distribution over different
tissues and the abundance are highly variable depending on the species. Though multiple RIP
sequences have been reported within one species, usually most of these sequences belong to the same
class of RIP proteins. Several type 1 RIP isoforms have been reported in Phytolacca americana, referred to
as pokeweed antiviral protein or PAP [9]. Different isoforms can occur within the same tissue and with
differential expression during development. For instance, PAP-I, PAP-II, and PAP-III are isolated from
spring, early summer, and late summer leaves of Phytolacca. At least 31 members of the RIP family have
been identified in the rice genome, all of them can be grouped as type 1 RIPs [10]. At present, only a
few examples are known of plant species containing both type 1 and type 2 RIP sequences, such as e.g.,
Iris x hollandica [11], Sambucus sp. [12], Momordica charantia [13,14], and Trichosanthes kirilowii [15,16].
Most RIPs identified and studied today are expressed at high level which enabled the purification
of the protein and characterization of their activities. Since more genome/transcriptome sequences
have become available for plants in the last decades, these data also yielded more information with
respect to the distribution and evolution of RIP sequences [5,6,17,18]. At present, it is clear that the
distribution of RIPs is underestimated, since many RIPs are expressed at levels that are too low to
allow purification of the protein.
Recently, the genome sequences of four Rosaceae species (Malus domestica, Pyrus communis,
Prunus persica, and Fragaria vesca) have become public through the Genome Rosaceae Database.
In silico analyses of these plant genomes/transcriptomes not only allowed identifying novel types
of RIPs but yielded also a fairly detailed overview of the occurrence of the type 1 and type 2 RIPs in
Rosaceae and generated new insights in the molecular evolution of this protein family. In this study,
a detailed phylogenetic analysis of RIP sequences within Rosaceae has been performed. Malus domestica
was selected as a model since it contains both type 1 and type 2 RIP sequences, further referred to as
Md1RIP and Md2RIP, respectively. Detailed analyses of the RIP sequences combined with molecular
modeling of the active site pocket of the RIP domain and the carbohydrate binding activity of the lectin
domain provide the first structural information for Rosaceae RIPs.
2. Results
2.1. RIP Genes Are Present in Most Rosaceae Genomes
The presence and distribution of RIP sequences within the family Rosaceae was analyzed
using BLAST searches against the genomes of all Rosaceae species for which a complete genome
sequence is available. A total of 16 genes encoding putative type 1 RIP genes and two genes
encoding putative type 2 RIP genes were identified in Rosaceae species including Malus domestica
(three type 1 RIPs and one type 2 RIP), Prunus mume (four type 1 RIPs), Prunus persica (one type 1
RIP), Pyrus bretschneideri (four type 1 RIPs) and Pyrus communis (three type 1 RIPs and one type 2 RIP)
(Figure S1 and Table S1 in the Supplementary Materials). Overall, the RIP sequences from Rosaceae
show 55%–70% sequence similarity and 42%–60% sequence identity at amino acid level. Furthermore
BLAST searches in the NCBI sequence database revealed that RIP genes from Rosaceae species share
37%–50% sequence identity with some well-characterized RIP domains, such as trichosanthin (type 1
RIP from Trichosanthes kirilowii) and SNA-I, ebulin, and cinnamomin (type 2 RIPs from Sambucus nigra,
Sambucus ebulus, and Cinnamonum camphora, respectively).
All Rosaceae genomes studied contained one or more sequences with a RIP domain, except for
the genome of Fragaria vesca and Fragaria ananassa. Interestingly the genomes from Malus domestica
(apple) and Pyrus communis (pear) contain both type 1 RIP and type 2 RIP sequences. To investigate
the evolutionary relationships between the Rosaceae RIPs a phylogenetic analysis was performed for
all RIP domains identified within Rosaceae (Figure 1A, Figure S1 in the Supplementary Materials).
The dendrogram revealed two major clades (Figure 1A). The largest clade contains the RIP domain
sequences for all type 1 RIPs from Rosaceae and falls apart in three subgroups with apple sequences,
Molecules 2016, 21, 1105
3 of 17
pear sequences, and a group containing sequences from Malus, Pyrus, and Prunus. The smaller clade
clusters the RIP domains belonging to the type 2 RIPs from apple and pear. The RIP domain sequences
of some well-characterized type 2 RIPs (ricin, SNA-I, ebulin and cinnamomin) with high sequence
similarity to Rosaceae RIPs cluster in the same clade as the type 2 RIP domains from Malus and Pyrus.
The sequence for the type 1 RIP trichosanthin forms a separate branch of the dendrogram (Figure 1A).
A.
Type 1 RIP
Trichosanthin (AAB31048.1)
M. domestica (MDP0000711911)
P. communis (PCP031611)
SNA-I (U27122)
Type 2 RIP
Ebulin (AJ400822)
Ricin (XP_002534649.1)
Cinnamomin (AAK82460.1)
M. domestica (MDP0000223290)
M. domestica (MDP0000134012)
P. bretschneideri (XP_009374990.1)
P. bretschneideri (XP_009346753.1)
P. bretschneideri (XP_009375039.1)
P. bretschneideri (XP_009346751.1)
Type 1 RIP
P. communis (PCP026877.1)
P. communis (PCP011148.1)
P. mume (XP_016652175.1 )
P. persica (ppa009409mg)
P. mume (XP_008243881)
P. mume (XP_016652174.1 )
0.1
P. persica (ppa009637mg)
P. mume (XP_008243880.1)
M. domestica (MDP0000918923)
P. communis (PCP001408.1)
B.
215
214
211
212
213
177
178
179
180
119
120
121
122
123
81
82
83
80
Entropy
(a)
Entropy
(b)
Amino acid number
Figure 1. Sequence analyses of RIPs. (A) Phylogenetic tree of sequences encoding RIP domains
from Rosaceae, Ricinus communis, Sambucus nigra, Sambucus ebulus, Cinnamomum camphora, and
Trichosanthes kirilowii. The phylogenetic tree was made by using constraint-based alignment tool
(COBALT). The name and species of Rosaceae RIPs and accession numbers are indicated; (B) (a) Logo
representation of the amino acid sequence alignment of RIP domains presented in the phylogenetic tree.
The size of the letters is proportional to the frequency of the amino acid at that position of the sequence.
Residues reported to be important for the formation of active sites are indicated by black arrows [19].
Amino acid numbers refer to the sequence of ricin; (b) Amino acid conservation in the lectin domains of
type 2 RIPs from Malus domestica, Pyrus communis, Sambucus nigra, Sambucus ebulus, and Ricinus communis.
The conserved amino acids of the carbohydrate binding site are indicated by black arrows [20].
Sequence alignments for all RIP domains used in the phylogenetic tree were performed using
ClustalW and analyzed for conserved positions in the amino acid sequences. The sequence logo
representation is shown in Figure 1B (panel a) and highlights those amino acid residues that are highly
Molecules 2016, 21, 1105
4 of 17
conserved at several positions within the RIP domain sequence. 53 residues out of 277 (=average size
of the RIP domain sequence) are conserved in at least 80% of the sequences analyzed. Five amino
acids: Y80 (position numbering according to ricin sequence), Y123, E177, R180, and W211 are known
to be important for the catalytic activity of the model RIP ricin [19] and are highly conserved for all
the Rosaceae RIP domain sequences. Amino acid residues E177 and R180 are directly involved in the
activity of catalytic sites, whereas Y80, Y123, and W211 play an important role in binding of the target
adenine [5]. In type 2 RIP sequences from apple and pear residue Y80 is replaced by S71 (Figure 1B
and Figure S2 in the Supplementary Materials).
A sequence logo representation for the lectin domain sequences of ricin, ebulin, SNA-I, and the
type 2 RIPs from M. domestica and P. communis (Figure 1B panel b) showed that 149 residues are highly
conserved in the sequences analyzed. Among them, 10 amino acid residues (D15, Q28, W30, N39, Q40,
D230, I241, Y244, N251, Q252, position numbering according to ricin sequence) from Rosaceae type 2
RIPs are identical with the residues forming the carbohydrate binding sites of ricin.
2.2. Identification and Sequence Analysis of RIP Genes in the Malus Domestica Genome
The in silico analysis allowed identifying the RIP sequences in most Rosaceae genomes. However,
these sequence data do not allow the prediction of the biochemical activities of the proteins. Since
Malus domestica (apple) is one of the most important fruits, it was selected as a model to study the
recombinant RIPs and characterize their biological activities.
The apple genome contains three type 1 RIP sequences and one type 2 RIP sequence. In addition,
the screening of the apple genome also revealed the occurrence of several pseudogenes. Analysis of the
apple type 1 RIP (Md1RIP) genes indicated that they encode three closely related proteins (Genome
database for Rosaceae (GRD) accession no. MDP0000918923, MDP0000223290, MDP0000134012 further
referred to as Md1RIP genes A, B and C, respectively) (Figure 1A). All Md1RIP sequences are highly
similar with 48% sequence identity and 63%–64% sequence similarity. The Md1RIP sequences share the
highest sequence identity with the RIP domain from cinnamomin (isoform iii) (43% sequence identity
and 58% sequence similarity). None of these Md1RIP sequences are synthesized with a signal peptide
(Figure S2 in the Supplementary Materials). Residues that build up the core catalytic site of the rRNA
N-glycosylase domain of Md1RIP are highly conserved when compared to ricin (Figure 1B panel a).
The type 1 RIP corresponding to gene A (MDP0000918923) was selected for protein expression
and characterization of the protein. The transcript for the MdRIP1 gene A encodes a 301 amino
acid polypeptide.
The apple type 2 RIP (Md2RIP) gene (MDP0000711911) resembles the classical type 2 RIP genes
found in other plant species and shares the highest sequence identity to SNA-I (48% sequence identity
and 62% sequence similarity). The transcript for the MdRIP2 gene yielded a deduced amino acid
sequence corresponding to a 22 amino acid residue signal peptide (Figure S3 in the Supplementary
Materials) followed by a 526 amino acid polypeptide covering the RIP domain and the lectin domain.
The sequence contains seven putative N-glycosylation sites, four in the RIP domain, and three in
the lectin domain. Compared to ricin, most amino acid residues that compose the catalytic site are
conserved in the Md2RIP sequence (Figure S2 in the Supplementary Materials). Sequence alignment
with ricin revealed a single amino acid substitution in the active site (Y80 of ricin is replaced by S71).
Alignment with SNA-I further indicates that all four intra-chain disulphide-bridges that stabilize the
lectin domain (C24–C43, C65–C77, C147–C162, C198–C205) are conserved in the apple RIP (Figure S3 in
the Supplementary Materials). Furthermore, the amino acids that built the two carbohydrate binding
domains are highly conserved between Md2RIP and SNA-I and ricin (Figure S2 in the Supplementary
Materials) [20–22].
2.3. Purification and Characterization of Recombinant Md1RIP and Md2RIP
Heterologous expression of the MdRIP sequences allowed the purification of recombinant Md1RIP
and Md2RIP from Pichia pastoris and Nicotiana tabacum cv. Bright Yellow-2 (BY-2) cells, respectively.
Molecules 2016, 21, 1105
5 of 17
SDS-PAGE analysis revealed that the molecular mass of the recombinant Md1RIP polypeptides was
around 40 kDa (Figure 2A), which was approximately 4 kDa higher than the calculated molecular
mass of the Md1RIP sequence including the c-myc and (His)6 tags (36.1 kDa) (Figure S4C in the
Supplementary Materials). Western blot analysis with a monoclonal antibody directed towards the His
tag (Figure 2B) confirmed that the 40 kDa band corresponded to the recombinant MdRIP. Furthermore,
Edman degradation of the recombinant Md1RIP yielded the sequence EAEAEFALSFSI (Figure S4
in the Supplementary Materials). Since the EA repeats in this sequence correspond to
α part of the
α-mating sequence it can be concluded that the signal peptide sequence included in the construct to
achieve secretion of the recombinant protein was not completely cleaved by the Ste13 protease.
C
B
A
1
2
72 kDa
3
4
6
5
7
D
8
9
10
11
12
13
72 kDa
70 kDa
70 kDa
34kDa
34 kDa
35 kDa
35 kDa
E
Figure 2. SDS-PAGE (A,C) and Western blot analysis (B,D) of recombinant MdRIPs. Lanes 1 and
β β-mercaptoethanol, lane 7:
4: unreduced Md1RIP, lanes 3 and lane 6: reduced Md1RIP treated with
positive control (Orysata, [23]), lanes 8 and 12: reduced Md2RIP, lane 10: unreduced Md2RIP, lane 11:
positive control (Md1RIP). In panels A and B 5 µg protein was loaded in each well. In panels C and D
15 µg and 3 µg protein were loaded, respectively. Protein ladder (Fermentas) was run in lanes 2, 5, 9,
and 13; (E) Effect of RIPs in a cell-free translation assay. Dose response curves for luciferase synthesis
were measured after treatment with increasing concentrations of Md1RIP, Md2RIP, and saporin for
30 min. Luciferase activity is shown as a function of the concentration of Md1RIP and Md2RIP.
SDS-PAGE analysis (Figure 2C) of recombinant Md2RIP yielded a major polypeptide band of
65.5 kDa. The size of these polypeptides is approximately 4 kDa higher than the calculated molecular
mass of the Md2RIP coding sequence (61.5 kDa, RIP and lectin domain containing a (His)6 tag). Western
blot analysis using a polyclonal antibody against the Md1RIP confirmed that the 65.5 kDa polypeptide
reacts with the RIP antibody (Figure 2D). Edman degradation of the 65.5 kDa polypeptide yielded
the sequence of GATAXXDIXXL (Figure S4 in the Supplementary Materials), which suggested that,
besides the signal peptide, an extra propeptide of 27 amino acid residues is cleaved at the N-terminus
of the recombinant Md2RIP secreted to the BY-2 cell medium.
The slightly higher mass of the recombinant polypeptides for both Md1RIP and Md2RIP compared
to the calculated mass of the RIP sequences was due to N-glycosylation of the recombinant proteins,
Molecules 2016, 21, 1105
6 of 17
as shown by N-glycosidase treatment of the recombinant proteins and subsequent SDS-PAGE analyses
(Figure S5 in the Supplementary Materials).
2.4. Biological Activities of Recombinant MdRIPs
To assess the catalytic activity of the MdRIPs, a cell-free translation system was used to study
the inhibition of protein synthesis by recombinant MdRIPs. In a parallel experiment saporin, the
type 1 RIP from Saponaria officinalis was analyzed as a control. As shown in Figure 2E, all RIPs tested
showed a clear reduction of protein synthesis. Higher inhibition of protein translation was observed
with increasing concentrations of the RIPs, suggesting a concentration-dependent protein activity.
The 50% inhibitory concentration (IC50 ) for Md1RIP and Md2RIP corresponded to 175 nM and 263 nM,
compared to an IC50 value of 9.6 pM for saporin.
The carbohydrate binding activity of the recombinant Md2RIP was analyzed by agglutination
assays as well as glycan array screening. The recombinant protein behaved as a genuine lectin, the
minimal concentration of recombinant protein required for agglutination of trypsin treated rabbit
erythrocytes being 1.55 µg/mL. A detailed analysis of the carbohydrate binding specificity using
glycan microarray screening revealed that the Md2RIP reacts preferentially with Neu5Ac (glycan #11)
and glycans carrying at least one terminal Neu5Ac residue as well as with 2-keto-3-deoxy-Dglycero-D-galactononic acid (KDN)α2-6Galβ1-4GlcNAc (glycan #357) (Figure S6 and Table S2 in
the Supplementary Materials). Furthermore, the Md2RIP interacted also with Neu5Gc (glycan #286)
β
but less strongly than with Neu5Ac. α
2.5. Molecular Modeling of Enzymatically Active Sites and Carbohydrate-Binding Sites
Since all type 1 RIPs from Rosaceae share a high sequence similarity, one type 1 RIP sequence
for each fully sequenced Rosaceae genome was selected to perform the molecular modelling studies.
The modelled type 1 RIPs of apple (Malus domestica), peach (Prunus persica), and pear (Pyrus communis),
consist of a single RIP domain sharing the conserved organization of a left-twisted bundle of β-sheets
associated to α-helices (Figure 3A–C). They only differ from each other by the shape and the
β size of the
loops and turns connecting
the
β-sheets
and
the
α-helices.
Accordingly,
the
three
models
superpose
α
β for the loop and
α turn structures (Figure 3D). All RIPs contain the
nicely but some discrepancies exist
sequence stretch EAAR involved in the rRNA N-glycosylase activity of the RIP domain (Figure 3E–G),
and docking experiments performed with pteroic acid suggest they readily accommodate the substrate
analog through a typical network of hydrogen bonds and stacking interactions with aromatic residues
(Figure 3H–J).
Figure 3. Cont.
Molecules 2016, 21, 1105
7 of 17
α
Figure 3. Ribbon diagrams built for the type 1 RIPs from apple (A), peach (B), and pear (C). α-helices,
β
β-sheets and loops/turns, are colored red, blue, and cyan, respectively. (D) Superposition of the type 1
RIP models of apple (colored red), peach (colored green), and pear (colored yellow). Docking of pteroic
acid (magenta stick) into the active site of type 1 RIPs of apple (E), peach (F), and pear (G), showing
the EAAR sequence of the active site. Network of H-bonds (dashed yellow lines) anchoring pteroic
acid to the active site in type 1 RIP of apple (H), peach (I), and pear (J). A stacking interaction occurs
between pteroic acid and Tyr and Trp residues located in the vicinity of the active site.
β
α
The type 1 RIPs from the Rosaceae
species exhibit the canonical RIP-fold, characterized by the
α
occurrence
of
three
domains,
namely
the
N-terminal domain 1 made of
α
β β-sheets and α-helices, the
central domain 2 built up from α-helices, and the C-terminal
domain
3
which consists of one or two
α
β
α-helices followed by two shorts strands of antiparallel-sheet organized in a β-hairpin (Figure S7A–C
in the Supplementary Materials). The occurrence of this α-helix-β-hairpin domain 3 inαtype 1 βRIPs is
believed to be essential for some of their biological activities, including the binding to a lipid bilayer and
β
Rosaceae type 1 RIPs possess the α-helix-β-hairpin
the rRNA N-glycosylase activity [5,24,25]. Although
domain at the C-terminal end of the polypeptide chain (Figure S7D–F in the Supplementary Materials)
α
β
the degree of conservation of this β-hairpin domain is rather weak, compared to other structural
domains 1 and 2, which exhibit a higher degree of conservation (Figure S7G–I in the Supplementary
Materials). The α-helix-β-hairpin in the RIP domain of apple, peach, and pear shows an amphipathic
conformation with opposing nonpolar and βpolar faces similar to BE27 (Figure S8 in the Supplementary
α
Materials) [26].
λ αanβoverallγthree-dimensional
λ α fold
β similar
γ to that found in
The RIP domain of MdRIP2 exhibits
β
other type 2 RIPs, built up of six strands of β-sheet clustered in a left-handed twisted bundle, associated
to eight α-helices (Figure 4A). The lectin domain consists of two tandemly arrayed domains 1 and 2,
each consisting of four subdomains (1λ, 1α, 1β, and 1γ for domain 1; 2λ, 2α, 2β, and 2γ for domain 2),
in which strands of β-sheet predominate. The RIP domain contains the sequence stretch 160EAAR163,
which plays a key role in the rRNA N-glycosylase activity of the RIP domain (Figure 4B). The key
residues of the active site of the RIP domain (S71, V72, S109, Y111, R163), readily accommodate pteroic
acid as shown from docking experiments (Figure 4C). The hydrogen bond network anchoring pteroic
acid to the active site of the RIP domain within Md2RIP resembles that observed in the ebulin-A-pteroic
acid complex [27]. Additional stacking interactions occur with the aromatic residues Y111, F159, W194,
and F240.
Molecules 2016, 21, 1105
8 of 17
Figure 4. Molecular organization of Md2RIP. (A) Ribbon diagram showing the organization of Md2RIP
with a RIP (A) and a lectin (B) domain. Stars show the position of the carbohydrate-binding sites in
the lectin domain; (B) Docking of pteroic acid (magenta stick) into the active site of the RIP domain,
showing the EAAR sequence of the active site; (C) Network of H-bonds (dashed lines) anchoring pteroic
acid to the active site of the RIP domain. H-bond distances are expressed in angström (Å). Aromatic
residues developing a stacking interaction with pteroic acid are colored orange; (D,E) Network of
hydrogen bonds (dashed lines) anchoring MeGal (magenta stick) to the amino acid residues (atom-code
colored sticks) forming the carbohydrate-binding sites 1 (D) and 2 (E) of Md2RIP lectin domain.
Aromatic residues interacting through a stacking with the pyranose ring of MeGal in both sites are
colored orange.
α
γ
γ
Molecules 2016, 21, 1105
9 of 17
The lectin domain of MdRIP2 contains two carbohydrate-binding sites occurring at both ends
of subdomains 1α (D285, Q298, W300, Q305, N307) and 2γ (D498, N501, Y512, N519) (Figure 4D,E).
Similar to the ricin-B chain, the carbohydrate-binding site of subdomain 2γ of Md2RIP differs from
that observed in the ebulin-B chain by the replacement of a Phe residue (F) by Y250. However,
these aromatic residues create a stacking interaction with the pyranose ring of galactose that
reinforces the interaction with the sugar. Both sugar-binding residues readily interacted with
α
methyl-α-D-galactopyranoside (MeGal) in docking experiments (Figure 4E,F) through a network
of hydrogen bonds similar to that observed in the galactose-ebulin-B complex [28], but rather different
from that observed in ricin [29]. Similar to the ebulin-galactose complex, an additional residue
participates in the network of hydrogen bonds anchoring the sugar to both carbohydrate-binding sites.
Obviously, these additional interactions reinforce the binding of galactose to the lectin moiety.
Docking experiments performed with sialic acid Neu5Ac suggest that both lectin domains of
Md2RIP readily accommodate the sugar derivative (Figure 5A,B). Sialic acid anchors to both binding
sites through a network of seven and eight hydrogen bonds, respectively. An additional stacking
interaction with an aromatic residue, W300 in site 1 and Y512 in site 2, should reinforce the interaction
of Neu5Ac with the lectin. A three-dimensional model with similar fold and binding properties
towards pteroic acid (RIP domain) and MeGal/Neu5Ac (lectin domain), was built by homology
modeling of the type 2 RIP sequence from Pyrus communis (PCP031611) (results not shown).
Figure 5. Docking of Neu5Ac to the carbohydrate-binding sites of Md2RIP. (A) Network of hydrogen
bonds (dashed lines) anchoring Neu5Ac to the carbohydrate-binding site of subdomainα1α of Md2RIP.
H-bond distances are expressed in angström (Å). The W300 aromatic residue (colored orange)
participates in a stacking interaction with the pyranose ring of Neu5Ac; (B) Network of hydrogen
bonds (dashed lines) anchoring Neu5Ac to the carbohydrate-binding
γ site of subdomain 1γ of Md2RIP.
The Y512 aromatic residue (colored orange) participates in a stacking interaction with the pyranose
ring of Neu5Ac.
3. Discussion
Ribosome-inactivating proteins are widely distributed in flowering plants. In this study,
the genomes of Rosaceae species were screened for RIP sequences. Evidence for the occurrence of
multiple RIP sequences was obtained for the genomes of Malus domestica, Prunus persica, Prunus mume,
Pyrus communis, and Pyrus bretschneideri but RIP sequences were absent from the genomes of
Fragaria vesca and Fragaria ananassa. The latter result is in contradiction to a recent report by
Polito et al. [30] who reported increased RIP activity in strawberry (Fragaria ananassa) plants grown
under reproductive, biotic, and drought stress conditions. However, since only partially purified basic
protein fractions from strawberry tissue extracts were used to perform the RIP assays, the RIP activity
Molecules 2016, 21, 1105
10 of 17
reported in this study is questionable. Previously Di Maro et al. reported that the Malus domestica
genome encodes four type 1 RIPs [5]. However, for one of them the open reading frame is interrupted
and the sequence is incomplete. In this study we focused on the three fully sequenced Md1RIP genes.
Both type 1 and type 2 RIP sequences have been retrieved from Rosaceae species. The genomes
of apple (Malus domestica) and pear (Pyrus communis) express both type 1 and type 2 RIPs, whereas
peach (Prunus persica and Prunus mume) and pear (Pyrus bretschneideri) express a complex set of type
1 RIPs. At present, only a few species (Sambucus ebulus, Iris x hollandica, Momordica charantia, and
Trichosanthes kirilowii) have been reported to contain both type 1 and type 2 RIPs [11,13–16,31,32].
These species present interesting models to study the evolutionary relationships between type 1 and
type 2 RIPs.
Phylogenetic analysis suggested that most type 1 RIPs found in dicots are evolutionary related
to type 2 RIPs [6]. The type 1 RIPs from Iris x hollandica are derived from the deletion of the lectin
domain of the type 2 RIPs. Similarly, the type 1 RIP genes from Rosaceae lack the signal peptide as
well as the lectin domain of the type 2 RIP sequence. Therefore, type 1 RIPs from Rosaceae can be
considered as “deletion” products of type 2 RIPs [5]. The type 2 RIP gave rise to the multiple lines
of type 1 RIP genes (e.g., Md1RIP-A, B, or C) by lectin domain deletion/gene truncation events [18].
Another striking difference between type 1 and type 2 RIP sequences from apple/pear involves the
lack of a signal peptide in the type 1 RIP sequences. Consequently, the type 1 RIPs are likely to be
synthesized on free ribosomes in the cytoplasm whereas type 2 RIPs will follow the secretory route for
protein synthesis. The different location for type 1 and type 2 RIPs in the plant cell can be important
for their particular role in the plant.
The recombinant Md1RIP secreted by the P. pastoris cells is a functional protein, although the
processing of the α-mating sequence from S. cerevisiae was not correctly cleaved by the Ste13 protease.
Problems with the correct processing of this α-mating sequence have been reported before on several
occasions [23]. The recombinant Md1RIP was made as a secreted glycosylated protein. Since the native
Md1RIP produced by the apple cells is synthesized without a signal peptide, it does not enter the
secretory pathway and hence it is unlikely that apple Md1RIP occurs as a glycoprotein. Despite the
incorrect/incomplete processing of the N-terminal α-mating sequence, the presence of a His tag at the
C-terminus, and the presence of N-glycans the recombinant Md1RIP is catalytically active.
Tobacco BY-2 cells were selected for the production of the recombinant Md2RIP since transgenic
tobacco plants and cells are fully capable of carrying out the cleavage of the type 2 RIP precursors
and glycosylation of the protein [33]. Indeed, BY-2 cells successfully expressed and secreted Md2RIP
into the BY-2 medium, the final concentration exceeding 10 mg per liter. Characterization of Md2RIP
revealed a unique protein structure. The results from SDS-PAGE and western blot analysis indicate
that the RIP and lectin domain are located on a single polypeptide in the recombinant Md2RIP, which
implies that the processing step reported for type 2 RIPs whereby the linker between the RIP and
lectin chain is excised from the precursor protein does not take place for the Md2RIP. Furthermore,
preliminary data confirmed the lack of processing of the Md2RIP precursor protein in apple. Western
blot analysis of a crude extract from immature (10 days post-pollination) apple (M. domestica cv.
‘Golden Delicious’) fruits yielded a polypeptide with a molecular mass of approximately 66 kDa,
corresponding to a polypeptide that contains both the RIP and lectin domain.
The results of in silico analysis and molecular modelling revealed that the amino acids known to
be important for the activity of ricin are highly conserved in the active site of the MdRIP sequences,
suggesting that the apple RIPs are functional. Furthermore, translation inhibition experiments
confirmed that the recombinant type 1 and type 2 RIPs from apple inhibit protein synthesis. However,
the catalytic activity of Md1RIP (IC50 = 175 nM) and Md2RIP (IC50 = 263 nM) is much lower (at 104 level)
than that of saporin, 30–50 folder lower than SNA-I [34] and four orders of magnitude inferior to the
values reported for ricin (IC50 = 100 pM, [35]). According to previous work [36] the replacement of
Y80 of ricin by S71 accounts for a 160-fold reduction in the catalytic activity. Hence, this amino acid
substitution in the Md2RIP sequence can contribute to the higher IC50 value. However this reasoning
Molecules 2016, 21, 1105
11 of 17
does not hold true for the Md1RIP in which the Y80 of ricin is conserved. Therefore, it is likely that
residues other than those involved in the canonical active site also influence the catalytic properties of
the MdRIPs. It has been reported that glycosylation can affect RIP activity. Protein synthesis inhibition
activity of the recombinant PD-L1 (a type 1 RIP from P. dioica) was increased after removal of the glycan
chains [9]. In this study, the recombinant Md1RIP was synthesized following the secretory pathway
for protein synthesis, though the native protein is most probably synthesized in the cytoplasm due
to the lack of a signal peptide in the sequence. The deduced sequence of Md1RIP comprises five
putative N-glycosylation sites (Figure S4 in the Supplementary Materials) and our data show that the
recombinant protein indeed contains some N-glycan chains. It is possible that these glycans affect the
enzymatic activity of the recombinant Md1RIP.
The Md2RIP behaves as a genuine lectin with a well-defined carbohydrate binding activity and
specificity. Sequence alignment results indicated that S197 of SNA-I, which is critical for the binding to
sialic acid in the Neu5Ac(α2-6)Gal/GalNAc sequence of 2-6-sialyllactose, is conserved the type 2 RIP
sequences from M. domestica and P. communis (Figure S3 in the Supplementary Materials). Furthermore,
a detailed glycan microarray analysis revealed that the recombinant Md2RIP preferentially interacts
with Neu5Ac and glycans carrying at least one terminal Neu5Ac(α2-6)Gal/GalNAc residue, and also
strongly binds to KDN. In addition, the Md2RIP reacted with a non-sialylated complex glycan. Though
the preferential interaction with Neu5Ac and glycans carrying terminal Neu5Ac(α2-6)Gal/GalNAc
residue is reminiscent of the carbohydrate binding specificity of SNA-I the results summarized
in Table S2 in the Supplementary Materials leave no doubt that there are major differences
between both lectins. For example, the Md2RIP has a much higher affinity for Neu5Ac than for
Neu5Ac(α2-6)Galβ1-4GlcNAc whereas for SNA-I this is the reverse. Though still speculative, the
latter observation might indicate that the binding site of the Md2RIP is less extended compared to
that of SNA-I. The carbohydrate binding properties of the apple RIP are similar to those of SNA-I.
However, this carbohydrate binding specificity is very different from the majority of type 2 RIPs that
specifically react with galactose or galactose derivatives. Although it is commonly accepted that sialic
acid is absent in plants [37], sialic acid is widely spread from bacteria to animal tissues, and plays an
important role in cell communication, adhesion, and protein targeting [38]. Similar to SNA-I, the sialic
acid binding specificity of the apple type 2 RIP may also protect plants from plant diseases e.g., fungi,
insects, or viruses [39]. Recently Hamshou et al. [40] reported strong aphicidal activity of the RIPs
from apple when tested in an artificial diet and in planta using transgenic tobacco lines overexpressing
the RIPs.
Interestingly, the Md1RIP sequence possesses an α-helix-β-hairpin structure at the C-terminus.
This structural motif is also present in other Rosaceae type 1 RIPs (Pyrus communis and Prunus persica).
Citores et al. [25] reported the antifungal activity of BE27, a type 1 RIP from Beta vulgaris L., against the
green mould Penicillium digitatum. They hypothesize this C-terminal α-helix-β-hairpin motif can assist
BE27 insertion into the lipid bilayer of fungal membranes and inactivate the fungal ribosomes. Our data
show that the α-helix-β-hairpin motif is also present in the cytoplasmic type 1 RIPs from Rosaceae
species. It remains to be shown that this motif can help to protect the plant against fungal infection.
Our study indicates that most genomes from Rosaceae encode one or more RIP sequences.
Sequence comparisons and molecular modelling studies leave no doubt that all these Rosaceae RIPs
are highly similar with conservation of the amino acids important for the catalytic activity of the RIP
domain and carbohydrate binding activity of the lectin domain. However, our data do not allow
any conclusions with respect to the toxicity of the RIPs. According to EST data available, RIPs are
expressed in young tissues and unripe fruits in low amounts. At present there is no evidence for RIP
expression in the mature fruits. Since the fruits are widely consumed in large quantities it is unlikely
that the RIPs are toxic, but they could exert beneficial antiviral and insecticidal activity [40] or could
cause allergy in atopic individuals, as some RIPs do [41]. Our data can contribute to the understanding
of evolution of RIP genes in Rosaceae and will help to deduce their biological role. Future work will
Molecules 2016, 21, 1105
12 of 17
focus on the expression patterns of RIPs in different plant tissues and the physiological importance of
these proteins.
4. Materials and Methods
4.1. Sequence Alignment and Phylogenetic Analysis
Type 1 RIP (trichosanthin, accession number AAB31048.1) and type 2 RIP (cinnamomin,
AAK82460.1; Ebulin, AJ400822; ricin, XP_002534649.1; SNA-I, U27122) sequences were used as queries
to search for RIP sequences in NCBI database (Pyrus bretschneideri and Prunus mume), Genome database
for Rosaceae (Malus domestica, Pyrus communis, Prunus persica, Fragaria ananassa and Fragaria vesca),
and Phytozome database (Prunus persica). Sequence alignments were represented by sequence logos as
created by WebLogo 3 [42]. ClustalW was used for the alignment of the RIP sequences. A phylogenetic
tree of all RIP sequences was constructed using the constraint-based alignment tool-COBALT [43].
4.2. Purification of Recombinant Md1RIP
The sequence for Md1RIP was expressed in Pichia pastoris. Therefore, an expression vector
containing the coding sequence for Md1RIP was constructed according to EasySelect Pichia Expression
Kit from Invitrogen (Invitrogen, Carlsbad, CA, USA). To allow secretion of Md1RIP into the yeast
culture medium, the Md1RIP construct contained an α-mating sequence from Saccharomyces cerevisiae,
upstream of the RIP coding sequence. Furthermore, a polyhistidine tag was provided downstream
of the RIP coding sequence for easy purification of the fusion protein. The recombinant Md1RIP
was purified using a combination of ion exchange chromatography on S Fast Flow (GE Healthcare,
Uppsala, Sweden) and affinity chromatography on a nickel Sepharose column, as described by
Desmyter et al. [44] and Al Atalah et al. [23], respectively.
4.3. Purification of Recombinant Md2RIP
Bright yellow-2 tobacco cells (BY-2 cells) were used for the expression of Md2RIP. A binary
expression vector containing the Md2RIP sequence including the signal peptide of Md2RIP was
constructed according to the Gateway™ cloning technology of Invitrogen. Stable transformation of
BY-2 cells was performed as reported previously [45]. The recombinant Md2RIP was expressed in BY-2
cells under the control of the 35S Cauliflower Mosaic virus promoter and purified from the BY-2 cell
culture medium using hydrophobic interaction chromatography on phenyl Sepharose combined with
affinity chromatography on fetuin-Sepharose 4B [21,46].
4.4. Western Blot Analysis
Samples were separated by SDS-PAGE and proteins transferred onto a PVDF membrane
(Bio Trace™ PVDF, PALL, Gelman Laboratory, Ann Arbor, MI, USA). First, the blots were blocked
in blocking buffer, consisting of 5% milk powder dissolved in Tris buffered saline (TBS: 10 mM Tris,
150 mM NaCl and 0.1% (v/v) Triton X-100, pH 7.6). After blocking, blots were incubated for 1 h in
TBS supplemented with the following primary antibodies: (i) for recombinant Md1RIP purified from
Pichia pastoris medium: mouse monoclonal anti-His (C-terminal) antibody (1:5000, Invitrogen; (ii) for
recombinant Md2RIP purified from tobacco cells: rabbit polyclonal anti-type 1 RIP antiserum (1:1500,
produced by Thermo Scientific by injecting two rabbits with recombinant type 1 RIP from apple). The
secondary antibody was a 1:1000 diluted rabbit anti-mouse IgG (Dako Cytomation, Glostrup, Denmark)
or a 1:5000 diluted horseradish peroxidase-coupled goat anti-rabbit IgG (Sigma-Aldrich, St. Louis,
MO, USA), respectively. Since Md1RIP shares 43% sequence identity with the RIP domain of Md2RIP
the antiserum raised against Md1RIP also recognized the recombinant Md2RIP. Immunodetection was
achieved by a colorimetric assay using 3,3′ -diaminobenzidine tetrahydrochloride (Sigma-Aldrich) as a
substrate. All washes and incubations were conducted at room temperature with gentle shaking.
Molecules 2016, 21, 1105
13 of 17
4.5. N-terminal Sequence Analysis
Purified Md1RIP and Md2RIP were analyzed by SDS-PAGE, electroblotted onto a Problot™
polyvinylidene fluoride (PVDF) membrane (Applied Biosystems, Foster City, CA, USA) and the blot
stained using 1:1 mix of Coomassie Brilliant Blue and methanol to visualize the protein. The bands of
interest were cut and used for N-terminal sequencing by Edman degradation using a capillary Procise
491cLC protein sequencer without alkylation of cysteines (Applied Biosystems).
4.6. Biochemical Assays
A cell-free system, the TnT® T7 Quick Coupled Transcription/Translation System Kit (Promega,
Mannheim, Germany) was used to determine and quantify the protein synthesis inhibition activity
of recombinant Md1RIP and Md2RIP as described by Shang et al. [34]. The prepared mixture was
incubated at 30 ◦ C for 10 min and chilled on ice. Subsequently, 2 µL PBS or PBS containing different
concentrations of proteins were added to the reaction mixture and incubated for 30 min at 30 ◦ C.
After addition of 35 µL nuclease-free water at room temperature, the reaction samples were transferred
to a luminometer plate (Greiner Labortechnik, Frickenhausen, Germany) containing 5 µL luciferase
assay reagent at 25 ◦ C. The relative luciferase activities of the samples were determined at 562 nm for
10 s using a microtiter top plate reader (Infinite 200, Tecan, Mannedorf, Switzerland) with an initial
delay of 2 s.
Agglutination assays using rabbit erythrocytes (BioMérieux, Marcy I’Etoile, France) were
performed in small glass tubes by mixing 10 µL purified recombinant Md2RIP, 10 µL of 1 M ammonium
sulphate and 30 µL of a 10% suspension of trypsin-treated rabbit erythrocytes. After 30 min at room
temperature, the agglutination was assessed visually.
Glycan microarrays were printed as described previously [47]. The printed glycan array contains a
library of natural and synthetic glycan sequences representing major glycan structures of glycoproteins
and glycolipids. Array version 5.0 with 611 glycan targets, was used for the analyses with recombinant
Md2RIP [48].
4.7. Protein Deglycosylation
Recombinant MdRIPs were digested with N-glycosidase F (PNGase F) as described by
Al Atalah et al. [49]. Briefly, 2 µg of recombinant proteins were mixed in a volume of 10 µL denaturation
buffer (0.5% SDS and 0.04 M dithiothreitol). The samples were boiled at 100 ◦ C for 10 min and
cooled down to room temperature. To the denatured samples we added: 2 µL of 10× reaction buffer
(0.5 M sodium phosphate pH 7.5), 2 µL 10% NP-40, and 5.5 µL distilled water to reach a total volume
of 20 µL. The samples were incubated at 37 ◦ C for 4 h after adding 0.5 µL of PNGase F (1000 U·µL−1 )
to each sample. RNAse B (2 µg) was used as a positive control. Finally, protein samples were analyzed
by SDS-PAGE.
4.8. Molecular Modeling and Docking
Homology modeling of type 1 RIPs from apple (Malus domestica, type 1 RIP GDR accession
number MDP0000918923; type 2 RIP MDP0000711911), peach (Prunus persica, ppa009409mg), and
pear (Pyrus communis, type 1 RIP-PCP001408.1; type 2 RIP accession no. PCP031611), was performed
with the YASARA Structure program [50], running on a 2.53 GHz Intel core duo Macintosh computer.
Different models of type 1 RIPs were built from the X-ray coordinates of ricin A-chain (PDB codes
2PJO, 2VC3, 1IFT) [51–53] and mutant N122a (PDB code 1UQ5) and R213d (1UQ4) of recombinant ricin
A-chain [28], used as templates. Finally, a hybrid model for the three type 1 RIPs was built up from
previous models. Different models of type 2 RIP from apple and pear, were similarly built using the
X-ray coordinates of the snake gourd seed lectin (PSB code 4HR6) [54], ebulin (PDB code 1HWM) [27],
native mistletoe lectin ML-I (PDB codes 2RG9, 1ONK), and ML-1 complexed to GlcNAc (PDB code
4EB2), as templates. A hybrid model of the two type 2 RIPs was finally built from the previous models.
Molecules 2016, 21, 1105
14 of 17
PROCHECK [55], ANOLEA [56], and the calculated QMEAN scores [57,58], were used to assess the
geometric and thermodynamic qualities of the three-dimensional models (Table S3). Using ANOLEA
to evaluate the models, only a few residues of the type 1 RIP and type 2 RIP models exhibited an energy
over the threshold value. Both residues are mainly located in the loop regions connecting the β-sheets
to the α-helices in the models. The calculated QMEAN6 score of all of the models gave values > 0.5.
The ConSurf server was used to discriminate between conserved and variable regions in the RIP
models [59]. Molecular cartoons were drawn with YASARA [50] and Chimera [60].
Supplementary Materials: Supplementary materials can be accessed at: http://www.mdpi.com/1420-3049/21/
8/1105/s1.
Acknowledgments: This work was funded primarily by the Fund for Scientific Research-Flanders
(FWO grant G007916N), the Research Council of Ghent University (project 01G00515 to E.J.M.V.D.). C.S. is
the recipient of funds from China Scholarship Council and also received doctoral co-funding from the Special
Research Council of Ghent University.
Author Contributions: C.S. and E.J.M.V.D. conceived and designed the experiments; C.S. performed the
purification and characterization of the recombinant proteins; P.R. performed the molecular modeling; C.S.,
E.J.M.V.D., and P.R. analyzed the data and wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest. The founding sponsors had no role in the design
of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, and in the
decision to publish the results.
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
Stirpe, F.; Battelli, M.G. Ribosome-inactivating proteins: Progress and problem. Cell Mol. Life Sci. 2006, 63,
1850–1866. [CrossRef] [PubMed]
Hamilton, P.T.; Peng, F.; Boulanger, M.J.; Perlman, S.J. A ribosome inactivating protein in a Drosophila
defensive symbiont. Proc. Natl. Acad. Sci. USA 2015, 113, 350–355. [CrossRef] [PubMed]
Sandvig, K.; Lingelem, A.B.D.; Skotland, T.; Bergan, J. Shiga toxins: Properties and action on cells. In The
Comprehensive Sourcebook of Bacterial Protein Toxins, 4th ed.; Joseph, A., Landant, D., Popoff, M.R., Eds.;
Elsevier: Waltham, MA, USA, 2015; pp. 267–286.
Wang, H.; Ng, T.B. Isolation and characterization of velutin, a novel low-molecular-weight
ribosome-inactivating protein from winter mushroom (Flammulina velutipes) fruiting bodies. Life Sci. 2001,
68, 2151–2158. [CrossRef]
Di Maro, A.; Citores, L.; Russo, R.; Iglesias, R.; Ferreras, J.M. Sequence comparison and phylogenetic analysis
by the maximum likelihood method of ribosome-inactivating proteins from angiosperms. Plant Mol. Biol.
2014, 85, 575–588. [CrossRef] [PubMed]
Peumans, W.J.; van Damme, E.J.M. Evolution of plant ribosome-inactivating proteins. In Plant Cell
Monographs; Lord, J.M., Hartley, M.R., Eds.; Springer: Heidelberg, Germany, 2010; pp. 1–26.
Stirpe, F.; Lappi, D.A. Ribosome-Inactivating Proteins: Ricin and Related Proteins; John Wiley & Sons, Ltd.:
Oxford, UK, 2014.
Stirpe, F. Ribosome-inactivating proteins. Toxicon 2004, 44, 371–383. [CrossRef] [PubMed]
Parente, A.; Chambery, A.; di Maro, A.; Russo, R.; Severino, V. Ribosome-inactivating proteins from
Phytolaccaceae. In Ribosome-Inactivating Proteins: Ricin and Related Proteins; Stirpe, F., Lappi, D.A., Eds.;
John Wiley & Sons, Ltd.: Oxford, UK, 2014; pp. 28–43.
Jiang, S.Y.; Ramamoorthy, R.; Bhalla, R.; Luan, H.F.; Venkatesh, P.N.; Cai, M.; Ramachandran, S. Genome-wide
survey of the RIP domain family in Oryza sativa and their expression profiles under various abiotic and
biotic stresses. Plant Mol. Biol. 2008, 67, 603–614. [CrossRef] [PubMed]
Hao, Q.; van Damme, E.J.M.; Hause, B.; Barre, A.; Chen, Y.; Rougé, P.; Peumans, W.J. Iris bulbs express
type 1 and type 2 ribosome-inactivating proteins with unusual properties. Plant Physiol. 2001, 125, 866–876.
[CrossRef] [PubMed]
Tejero, J.; Jiménez, P.; Quinto, E.J.; Cordoba-Diaz, D.; Garrosa, M.; Cordoba-Diaz, M.; Gayoso, M.J.; Girbés, T.
Elderberries: A source of ribosome-inactivating proteins with lectin activity. Molecules 2015, 20, 2364–2387.
[CrossRef] [PubMed]
Molecules 2016, 21, 1105
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
15 of 17
Fang, E.F.; Zhang, C.Z.Y.; Ng, T.B.; Wong, J.H.; Pan, W.L.; Ye, X.J.; Chan, Y.S.; Fong, W.P. Momordica charantia
lectin, a type II ribosome inactivating protein, exhibits antitumor activity toward human nasopharyngeal
carcinoma cells in vitro and in vivo. Cancer Prev. Res. 2012, 5, 109–121. [CrossRef] [PubMed]
Husain, J.; Tickle, I.J.; Wood, S.P. Crystal structure of momordin, a type I ribosome inactivating protein from
the seeds of Momordica charantia. FEBS Lett. 1994, 342, 154–158. [CrossRef]
Li, M.; Wang, Y.P.; Chai, J.J.; Wang, K.Y.; Bi, R.C. Molecular-replacement studies of Trichosanthes kirilowii
lectin 1: A structure belonging to the family of type 2 ribosome-inactivating proteins. Acta Crystallogr. Sect.
D-Biol. Crystallogr. 2002, 56, 1073–1075. [CrossRef]
Ng, T.B.; Wong, H.J. Ribosome-inactivating proteins in Caryophyllaceae, Cucurbitaceae and Euphorbiaceae.
In Ribosome-Inactivating Proteins: Ricin and Related Proteins; Stirpe, F., Lappi, D.A., Eds.; John Wiley & Sons,
Ltd.: Oxford, UK, 2014; pp. 44–66.
Shang, C.; Peumans, W.J.; van Damme, E.J.M. Occurrence and taxonomical distribution of
ribosome-inactivating proteins belonging to the ricin/shiga toxin superfamily. In Ribosome-Inactivating
Proteins: Ricin and Related Proteins; Stirpe, F., Lappi, D.A., Eds.; John Wiley & Sons, Ltd.: Oxford, UK, 2014;
pp. 11–27.
Peumans, W.J.; Shang, C.; van Damme, E.J.M. Updated model of the molecular evolution of RIP genes.
In Ribosome-Inactivating Proteins: Ricin and Related Proteins; Stirpe, F., Lappi, D.A., Eds.; Wiley Blackwell Press:
New York, NJ, USA, 2014; pp. 134–150.
Lapadula, W.J.; Puerta, M.V.S.; Ayub, M.J. Revising the taxonomic distribution, origin and evolution of
ribosome inactivating protein genes. PLoS ONE 2013, 8, e72825. [CrossRef] [PubMed]
Kaku, H.; Kaneko, H.; Minamihara, N.; Iwata, K.; Jordan, E.T.; Rojo, M.A.; Minami-Ishii, N.; Minami, E.;
Hisajima, S.; Shibuya, N. Elderberry bark lectins evolved to recognize Neu5Acα2,6Gal/GalNAc sequence
from a Gal/GalNAc binding lectin through the substitution of amino-acid residues critical for the binding to
sialic acid. Biochem. J. 2007, 142, 393–401. [CrossRef] [PubMed]
Van Damme, E.J.M.; Barre, A.; Rougé, P.; van Leuven, F.; Peumans, W.J. The NeuAc(a-2,6) Gal/
GalNAc-binding lectin from elderberry (Sambucus nigra) bark, a type-2 ribosome-inactivating protein with
an unusual specificity and structure. Eur. J. Biochem. 1996, 235, 128–137. [CrossRef] [PubMed]
Hu, D.; Tateno, H.; Kuno, A.; Yabe, R.; Hirabayashi, J. Directed evolution of lectins with sugar-binding
specificity for 6-sulfogalactose. J. Biol. Chem. 2012, 287, 20313–20320. [CrossRef] [PubMed]
Al Atalah, B.; Fouquaert, E.; Vanderschaeghe, D.; Proost, P.; Balzarini, J.; Smith, D.F.; Rougé, P.; Lasanajak, Y.;
Callewaert, N.; van Damme, E.J.M. Expression analysis of the nucleocytoplasmic lectin ‘Orysata’ from rice in
Pichia pastoris. FEBS J. 2011, 278, 2064–2079. [CrossRef] [PubMed]
Mak, A.N.; Wong, Y.T.; An, Y.J.; Sze, K.H.; Wing-Ngor Au, S.; Wong, K.B.; Shaw, P.C. Structure-function
study of maize ribosome-inactivating protein: Implications for the internal inactivation region and the sole
glutamate in the active site. Nucleic Acids Res. 2007, 35, 6259–6267. [CrossRef] [PubMed]
Citores, L.; Iglesias, R.; Gay, C.; Ferreras, J.M. Antifungal activity of the ribosome-inactivating protein BE27
from sugar beet (Beta vulgaris L.) against the green mould Penicillium digitatum. Mol. Plant Pathol. 2016, 17,
261–271. [CrossRef] [PubMed]
Iglesias, R.; Citores, L.; Ragucci, S.; Russo, R.; di Maro, A.; Ferreras, J.M. Biological and antipathogenic
activities of ribosome-inactivating proteins from Phytolacca dioica L. BBA-Gen. Subjects 2016, 1860, 1256–1264.
[CrossRef] [PubMed]
Pascal, J.M.; Day, P.J.; Monzingo, A.F.; Ernst, S.R.; Robertus, J.D.; Iglesias, R.; Pérez, Y.; Férreras, J.M.;
Citores, L.; Girbés, T. 2.8-Å Crystal structure of a nontoxic type-II ribosome-inactivating protein, ebulin l.
Proteins Struct. Funct. Genet. 2001, 43, 319–326. [CrossRef] [PubMed]
Marsden, C.J.; Fülöp, V.; Day, P.J.; Lord, J.M. The effect of mutations surrounding and within the active site
on the catalytic activity of ricin A chain. Eur. J. Biochem. 2004, 271, 153–162. [CrossRef] [PubMed]
Rutenber, E.; Katzin, B.J.; Collins, E.J.; Mlsna, D.; Ernst, S.E.; Ready, M.P.; Robertus, J.D. Crystallographic
refinement of ricin to 2.5 Å. Proteins 1991, 10, 240–250. [CrossRef] [PubMed]
Polito, L.; Bortolotti, M.; Mercatelli, D.; Mancuso, R.; Baruzzi, G.; Faedi, W.; Bolognesi, A. Protein synthesis
inhibition activity by strawberry tissue protein extracts during plant life cycle and under biotic and abiotic
stresses. Int. J. Mol. Sci. 2013, 14, 15532–15545. [CrossRef] [PubMed]
Molecules 2016, 21, 1105
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
16 of 17
Vandenbussche, F.; Peumans, W.J.; Desmyter, S.; Proost, P.; Ciani, M.; van Damme, E.J.M. The type-1 and
type-2 ribosome-inactivating proteins from Iris confer transgenic tobacco plants local but not systemic
protection against virus. Planta 2004, 220, 211–221. [CrossRef] [PubMed]
Vandenbussche, F.; Desmyter, S.; Ciani, M.; Proost, P.; Peumans, W.J.; van Damme, E.J.M. Analysis of the in
planta antiviral activity of elderberry ribosome-inactivating proteins. Eur. J. Biochem. 2004, 271, 1508–1515.
[CrossRef] [PubMed]
Chen, Y.; Peumans, W.J.; van Damme, E.J.M. The Sambucus nigra type-2 ribosome-inactivating protein SNA-I’
exhibits in planta antiviral activity in transgenic tobacco. FEBS Lett. 2002, 516, 27–30. [CrossRef]
Shang, C.; Chen, Q.; Dell, A.; Haslam, S.M.; de Vos, W.H.; van Damme, E.J.M. The cytotoxicity of elderberry
lectins is not solely determined by their N-glycosidase activity. PLoS ONE 2015, 10, e0132389. [CrossRef]
[PubMed]
Ferreras, J.M.; Citores, L.; Iglesias, R.; Jiménez, P.; Girbés, T. Use of ribosome-inactivating proteins from
Sambucus for the construction of immunotoxins and conjugates for cancer therapy. Toxins 2011, 3, 420–441.
[CrossRef] [PubMed]
Kim, Y.; Robertus, J.D. Analysis of several key active site residues of ricin A chain by mutagenesis and X-ray
crystallography. Protein Eng. 1992, 5, 775–779. [CrossRef] [PubMed]
Zeleny, R.; Kolarich, D.; Strasser, R.; Altmann, F. Sialic acid concentrations in plants are in the range of
inadvertent contamination. Planta 2006, 224, 222–227. [CrossRef] [PubMed]
Varki, A.; Schauer, R. Sialic acids. In Essentials of Glycobiology, 2nd ed.; Varki, A., Cummings, R.D.,
Esko, J.D., Freeze, H.H., Stanley, P., Bertozzi, C.R., Hart, G.W., Etzler, M.E., Eds.; Cold Spring Harbor Press:
Cold Spring Harbor, Woodbury, NY, USA, 2009; Chapter 4.
Shahidi-Noghabi, S.; van Damme, E.J.M.; Smagghe, G. Carbohydrate-binding activity of the type-2
ribosomes-inactivating protein SNA-I from elderberry (Sambucus nigra) is a determining factor for its
insecticidal activity. Phytochemistry 2008, 69, 2972–2978. [CrossRef] [PubMed]
Hamshou, M.; Shang, C.; Smagghe, G.; van Damme, E.J.M. Ribosome-inactivating proteins from apple have
strong aphicidal activity in artificial diet and in planta. Crop Prot. 2016, 87, 19–24. [CrossRef]
Szalai, K.; Schöll, I.; Förster-Waldl, E.; Polito, L.; Bolognesi, A.; Untersmayr, E.; Riemer, A.B.;
Boltz-Nitulescu, G.; Stirpe, F.; Jensen-Jarolim, E. Occupational sensitization to ribosome-inactivating proteins
in researchers. Clin. Exp. Allergy 2005, 35, 1354–1360. [CrossRef] [PubMed]
Crooks, G.E.; Hon, G.; Chandonia, J.M.; Brenner, S.E. WebLogo: A sequence logo generator. Genome Res.
2004, 14, 1188–1190. [CrossRef] [PubMed]
Papadopoulos, J.S.; Agarwala, R. COBALT: Constraint-based alignment tool for multiple protein sequences.
Bioinformatics 2007, 23, 1073–1079. [CrossRef] [PubMed]
Desmyter, S.; Vandenbussche, F.; Hao, Q.; Proost, P.; Peumans, W.J.; van Damme, E.J.M. Type-1
ribosome-inactivating protein from iris bulbs: A useful agronomic tool to engineer virus resistance?
Plant Mol. Biol. 2003, 51, 567–576. [CrossRef] [PubMed]
Delporte, A.; de Vos, W.H.; van Damme, E.J.M. In vivo interaction between the tobacco lectin and the core
histone proteins. J. Plant Physiol. 2014, 171, 1149–1156. [CrossRef] [PubMed]
Stefanowicz, K.; Lannoo, N.; Proost, P.; van Damme, E.J.M. Arabidopsis F-box protein containing a
Nictaba-related lectin domain interacts with N-acetyllactosamine structures. FEBS Open Bio 2012, 2, 151–158.
[CrossRef] [PubMed]
Blixt, O.; Head, S.; Mondala, T.; Scanlan, C.; Huflejt, M.E.; Alvarez, R.; Bryan, M.C.; Fazio, F.; Calarese, D.;
Stevens, J.; et al. Printed covalent glycan array for ligand profiling of diverse glycan binding proteins.
Proc. Natl. Acad. Sci. USA 2004, 101, 17033–17038. [CrossRef] [PubMed]
Shang, C.; van Damme, E.J.M. Comparative analysis of carbohydrate binding properties of Sambucus nigra
lectins and ribosome-inactivating proteins. Glycoconj. J. 2014, 31, 345–354. [CrossRef] [PubMed]
Al Atalah, B.; Smagghe, G.; van Damme, E.J.M. Orysata, a jacalin-related lectin from rice, could protect plants
against biting-chewing and piercing-sucking insects. Plant Sci. 2014, 221–222, 21–28. [CrossRef] [PubMed]
Krieger, E.; Koraimann, G.; Vriend, G. Increasing the precision of comparative models with YASARA
NOVA—A self-parameterizing force field. Proteins 2002, 47, 393–402. [CrossRef] [PubMed]
Carra, J.H.; McHugh, C.A.; Mulligan, S.; Machiesky, L.M.; Soares, A.S.; Millard, C.B. Fragment-based
identification of determinants of conformational and spectroscopic change at the ricin active site.
BMC Struct. Biol. 2007, 7, 72–83. [CrossRef] [PubMed]
Molecules 2016, 21, 1105
52.
53.
54.
55.
56.
57.
58.
59.
60.
17 of 17
Allen, S.C.; Moore, K.A.; Marsden, C.J.; Fülöp, V.; Moffat, K.G.; Lord, J.M.; Ladds, G.;
Roberts, L.M. The isolation and characterization of temperature-dependent ricin A chain molecules in
Saccharomyces cerevisiae. FEBS J. 2007, 274, 5586–5599. [CrossRef] [PubMed]
Weston, S.A.; Tucker, A.D.; Thatcher, D.R.; Derbyshire, D.J.; Pauptit, R.A. X-ray structure of recombinant
ricin A-chain at 1.8 Å resolution. J. Mol. Biol. 1994, 244, 410–422. [CrossRef] [PubMed]
Sharma, A.; Pohlentz, G.; Bobbili, K.B.; Jeyaprakash, A.A.; Chandran, T.; Mormann, M.; Swamy, M.J.;
Vijayan, M. The sequence and structure of snake gourd (Trichosanthes anguina) seed lectin, a three-chain
nontoxic homologue of type II RIPs. Acta Crystallogr. Sect. D 2013, 69, 1493–1503. [CrossRef] [PubMed]
Laskowski, R.A.; MacArthur, M.W.; Moss, D.S.; Thornton, J.M. PROCHECK: A program to check the
stereochemistry of protein structures. J. Appl. Cryst. 1993, 26, 283–291. [CrossRef]
Melo, F.; Feytmans, E. Assessing protein structures with a non-local atomic interaction energy. J. Mol. Biol.
1998, 277, 1141–1152. [CrossRef] [PubMed]
Arnold, K.; Bordoli, L.; Kopp, J.; Schwede, T. The SWISS-MODEL workspace: A web-based environment for
protein structure homology modelling. Bioinformatics 2006, 22, 195–201. [CrossRef] [PubMed]
Benkert, P.; Biasini, M.; Schwede, T. Toward the estimation of the absolute quality of individual protein
structure models. Bioinformatics 2011, 27, 343–350. [CrossRef] [PubMed]
Glaser, F.; Pupko, T.; Bell, R.E.; Bechor, D.; Martz, E.; Ben-Tal, N. ConSurf: Identification of functional regions
in proteins by surface-mapping of phylogenetic informations. Bioinformatics 2003, 19, 163–164. [CrossRef]
[PubMed]
Pettersen, E.F.; Goddard, T.D.; Huang, C.C.; Couch, G.S.; Greenblatt, D.M.; Meng, E.C.; Ferrin, T.E. UCSF
Chimera—A visualization system for exploratory research and analysis. J. Comput. Chem. 2004, 25, 1605–1612.
[CrossRef] [PubMed]
Sample Availability: Samples of the compounds are not available from the authors.
© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).