Protein Expression and PuriWcation 40 (2005) 142–151

www.elsevier.com/locate/yprep

Bacterial expression and enzymatic activity analysis of ME1,

a ribosome-inactivating protein from Mirabilis expansa
Ramarao Vepachedua, Sang-Wook Parkb, Neelam Sharmac, Jorge M. Vivancod,e,¤
a
Department of Medicine, National Jewish Medical and Research Center, 1400 Jackson Street, Denver, CO 80206, USA
b
Boyce Thompson Institute for Plant Research, Ithaca, NY 14853, USA
c
Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO 80523, USA
d
Department of Horticulture and Landscape Architecture, Colorado State University, Fort Collins, CO 80523-1173, USA
e
Cell and Molecular Biology Program, Colorado State University, Fort Collins, CO 80523-1173, USA

Received 15 September 2004, and in revised form 30 November 2004

Available online 5 January 2005

Abstract

Ribosome-inactivating proteins (RIPs) are toxic proteins synthesized by many plants and some bacteria, that speciWcally depuri-
nate the 28S RNA and thus interrupt protein translation. RIPs hold broad interest because of their potential use as plant defense fac-
tors against pathogens. However, study of the activity of type I RIPs has been hampered since their expression in Escherichia coli has
typically been toxic to the model system. Mirabilis expansa, an Andean root crop, produces a type I RIP called ME1 in large quanti-
ties in its storage roots. In this study, the cDNA sequence of ME1 was used to successfully express the recombinant ME1 protein in
E. coli. The production of recombinant ME1 in E. coli was conWrmed by Western blot analysis using anti-ME1 antibodies. The stud-
ies with Xuorescence-labeled ME1 showed that ME1 can enter bacteria and be distributed in the cytoplasm uniformly, indicating its
ability to access the protein synthesis machinery of the bacteria. The recombinant enzyme was active and depurinated yeast ribo-
somes. However, both native and recombinant ME1 proteins failed to depurinate the E. coli ribosomes, explaining the non-toxicity
of recombinant ME1 to E. coli. Structural modeling of ME1 showed that it has folding patterns similar to other RIPs, indicating that
ME1 and PAP, which share a similar folding pattern, can show diVerent substrate speciWcity towards E. coli ribosomes. The results
presented here are very signiWcant, as few reports are available in the area of bacterial interaction with type I RIPs.
 2004 Elsevier Inc. All rights reserved.

Keywords: Ribosome-inactivating protein; ME1; Depurination; E. coli; Yeast ribosomes; Naked rRNA

Ribosome-inactivating proteins (RIPs) are a family of
cytotoxic enzymes widely distributed in the plant King-
dom [1]. RIPs are polynucleotide adenosine glycosidases
that cleave the glycosidic bond of an adenosine base in
an evolutionarily conserved sequence (GAGA) located
in the
-sarcin/ricin (S/R) loop of eukaryotic ribosomes
[2,3]. RIPs show depurination activity against eukaryotic
and prokaryotic ribosomal RNA (rRNA) in the pres-
ence and absence of ribosomal proteins [4]. Depurina-
*
Corresponding author.
E-mail address: j.vivanco@colostate.edu (J.M. Vivanco).
tion of the S/R loop prevents binding of the elongation
factor 2 to the ribosome, and results in protein synthesis
inhibition [5]. Apart from the depurination of rRNA,
RIPs exhibit the ability to depurinate DNA, poly(A),
and viral RNA [6–8]. Studies on substrate speciWcity
indicate that some RIPs can also interact with and depu-
rinate mRNA [9], leading to a recent study demonstrat-
ing that depurination by ME1 does not require the cap
structure for recognition [10].
RIPs have been characterized from various plant
sources and are classiWed as type I, II, and III proteins.
Type I RIPs consist of a single polypeptide chain of

1046-5928/$ - see front matter  2004 Elsevier Inc. All rights reserved.
doi:10.1016/j.pep.2004.12.005
 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151
approximately 30 kDa. In contrast, type II RIPs contain
a catalytically active A-chain and a carbohydrate-bind-
ing B-chain. Both chains are derived from a single pre-
cursor by the excision of a linker sequence between the A
and B domains and are held together by a disulWde
bridge between the C-terminus of the A-domain and the
N-terminus of the B domain. Type III RIPs have a single
chain but are distinctly diVerent from both type I and II
RIPs in their sequence.
RIPs hold broad interest because of their potential
use as plant defense factors against viruses and fungi.
Transgenic plants expressing RIPs show enhanced resis-
tance to disease-causing pathogens: the barley RIP
expressed in tobacco (Nicotiana tabacum) protected
plants from the soil-borne pathogen Rhizoctonia solani
[11,12], and maize RIP expressed in tobacco plants con-
ferred resistance against the corn earworm, Helicoverpa
zea (Boddie) [13]. Additionally, many RIPs are used as
immunotoxins by linking them to antibodies speciWc to
target cells [14,15]. For instance, the anti-CD19 and anti-
CD22 immunotoxins made of murine IgG(1) monoclo-
nal antibodies (Mabs) conjugated to a deglycosylated
ricin A-chain (dgRTA) were eVective in killing the B-
lineage of non-Hodgkin’s lymphoma (NHL) [16].
Understanding the mechanism of biological activity of
RIPs greatly enhances our ability to use them as defense
proteins against pathogens. Production of RIPs as bio-
logically active recombinant proteins enhances our abil-
ity to use them as immunotoxins.
The expression of RIPs in bacteria as recombinant
proteins is complicated because of their cellular toxicity.
The expression of type I RIPs in Escherichia coli has typ-
ically been toxic to the model system because of their
ability to depurinate prokaryotic ribosomes. Studies
with dianthin and Phytolacca antiviral protein, PAP,
both type I RIPs, indicate that these RIPs cleave a bond
between the A-2660 and the ribose of the 23S rRNA of
E. coli, a position equivalent to the A-4324 of eukaryotic
28S rRNA [17]. PAP mutants, which do not inhibit
E. coli growth but inhibit eukaryotic protein synthesis,
have been constructed and expressed in E. coli [18]; how-
ever, PAP mutants, despite their non-toxicity to E. coli,
are apparently able to retain their ability to inactivate
prokaryotic ribosomes [19]. Mirabilis antiviral protein
(MAP) from Mirabilis jalapa has been expressed with its
own signal sequence in E. coli, but the synthesis resulted
in low yields and inhibition of bacterial growth [20]. Our
recent work with ME1, an RIP from Mirabilis expansa,
has shown that ME1 has close sequence similarity with
MAP [10]. Previous studies of RIP activity on bacterial
ribosomes have shown that type I RIPs show varying
degrees of toxicity against bacterial ribosomes [21,22].
Interestingly, ME1 has both anti-fungal and anti-bacte-
rial activities [23] that make this RIP an interesting tar-
get for expression studies aimed at understanding the
biological activities of the RIPs. The toxicity of ME1
143
depends on its localization and the ability of the ME1 to
depurinate the ribosomes of the host cells. Expression in
bacteria allows us to understand how prokaryotes
respond to a synthesis of these plant toxins. In the pres-
ent study, we describe how ME1 was successfully
expressed in E. coli, and discuss the enzymatic activity of
this recombinant protein and its structural similarities to
other RIPs.
Materials and methods
Plant material
Seeds of M. expansa (CIP Accession 208001, ARB
5395) were obtained from Dr. M.K.V. Zant, Southern
Illinois University, USA, and the native ME1 protein
was puriWed from the storage roots of M. expansa [23].
Construction of the recombinant expression system
ME1 was cloned into the pET28 vector at the NdeI
and BamHI sites for expression in E. coli pLysS. The
open reading frame of ME1 has 317 amino acids and was
obtained from a M. expansa storage root cDNA library
using polymerase chain reaction (PCR). The enzyme sites
NdeI and BamHI were generated using the following 5⬘
and 3⬘ primers: CATATGGAAACTATGAGGTTGC
TCTTCC and GGATCCTTAAGAAGATGCAACTA
CAACACTA. The PCR product containing the open
reading frame was cloned into a TA cloning vector,
pGEM-T Easy (Promega, Madison, WI). The TA clone
plasmid was isolated by mini-prep using a plasmid isola-
tion kit (Bio-Rad, Hercules, CA) and digested with NdeI
and BamHI. The fragment with the gene sequence was
separated on a 1% agarose gel and ligated into a pET28b
expression vector (Invitrogen, Carlsbad, CA) digested
with the same enzymes. The resulting vector, ME-pET
28b with an N-terminal His6-tag, was sequenced to con-
Wrm the sequence of ME1 and proper ligation to the vec-
tor. The ME-pET 28b was transformed into pLysS
(Invitrogen) bacterial cells for expression.
Expression of recombinant protein
The ME1 protein was expressed in E. coli pLysS
strain by induction with IPTG (isopropyl--D-thiogalac-
topyranoside). The E. coli BL21(DE3) pLysS strain con-
taining the ME1 expression plasmid was grown at 37 °C
with constant shaking in Luria–Bertani (LB) medium
supplemented with 50 g/ml kanamycin and chloram-
phenicol, until the culture reached an absorbance of 0.5
at 620 nm. ME1 protein synthesis was induced by the
addition of 1 mM IPTG to the medium and the culture
was grown for 3 h at 200 rpm under constant shaking. To
study the toxicity of the induced ME1 against the host
144 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151
bacteria, the cultures were allowed to grow for 15 h with
and without IPTG induction. The bacterial growth was
monitored by absorbance at 620 nm. For Western analy-
sis the samples were collected from induced and non-
induced cultures after 3 h of IPTG induction and the
bacterial cells were harvested by centrifugation (3000g at
4 °C for 15 min). The pellet was dissolved in B-Per
(Pierce, Rockford, IL). These protein samples were used
for analysis by sodium dodecyl sulfate–polyacrylamide
gel electrophoresis (SDS–PAGE) and Western blotting.
The data presented here represent three independent
experiments.
SDS–PAGE and Western blotting
The proteins from the bacterial extracts were resolved
on 12% SDS–PAGE and stained with Coomassie bril-
liant blue (CBB). To conWrm the identity of the resulting
band, the bacterial protein samples were transferred
onto a polyvinylidene Xuoride (PVDF) membrane and
probed with anti-ME1 antibodies. For Western blots,
20 g of the total proteins was separated on a 12% SDS–
PAGE gel and transferred to PVDF membranes (Milli-
pore, Bedford, MA) and the membranes were blocked
with 0.5% non-fat dry milk in TBST buVer (10 mM Tris–
HCl, pH 7.5, 150 mM NaCl, and 0.02% Tween-20) for 2 h
at room temperature. The membranes were then incu-
bated in anti-ME1 antibodies [24] at 1:1000 dilutions for
1 h. The membranes were washed three times with TBST
and incubated in horseradish peroxidase-conjugated
goat anti rabbit IgG (Immun-Blot Assay Kit, Bio-Rad,
Hercules, CA) at a 1:3000 dilution and the labeled bands
were detected according to the manufacturer’s protocol.
For identiWcation of His-tagged proteins the membranes
were treated with anti-His antibodies (Bio-Rad) fol-
lowed by anti-goat IgG (Bio-Rad) and the blots were
developed as per the manufacturer’s protocol.
PuriWcation of the ME1 recombinant protein
For puriWcation of recombinant proteins the bacteria
were grown in 2-liter cultures and induced with IPTG
for 3 h. The bacterial pellet from the culture was centri-
fuged and the pellet was lysed by sonication and the
insoluble cell debris was removed by centrifugation. For
puriWcation of bacterial expressed ME1, the bacterial
extract was initially passed through a Ni2+-column
(Pharmacia) as per the manufacturer’s protocol. How-
ever, no proteins were bound to the column. Subse-
quently, we prepared a resin for aYnity chromatography
in which polyclonal anti-ME1 antibodies (5 mg ml¡1)
were coupled to cyanogen bromide (CNBr)-activated
Sepharose. The resulting anti-ME1/Sepharose resin had
a binding capacity of 72 mg ME1/ml resin. The total
bacterial proteins were fractionated using diethylamino-
ethyl (DEAE) column chromatography to eliminate all
acidic proteins. The column of anti-ME1/Sepharose
resin was washed thoroughly with 75 mM Tris–HCl (pH
8.0) equilibration buVer. The protein Xow-through frac-
tion from the DEAE column was loaded onto the aYn-
ity column. The loaded column was washed with
equilibration buVer, and ME1 was eluted with 100 mM
glycine–HCl buVer (pH 2.7) using a NaCl gradient. Frac-
tions were collected and analyzed by SDS–PAGE using
silver staining. The protein concentration was deter-
mined by the Bradford [25] method using a protein assay
kit (Bio-Rad) and bovine serum albumin (BSA) as a
standard.
Isolation of yeast ribosomes
Yeast (Saccharomyces cerevisiae) strain YPH500 [26]
was grown in yeast extract/peptone/dextrose (YPD)
media. Yeast cells were then pelleted by centrifugation.
To isolate yeast ribosomes, 10 g of pelleted yeast cells
was ground in a mortar with liquid N2, and dissolved in
100 ml of extraction buVer (200 mM KCl, 25 mM MgCl2,
25 mM EGTA, 200 mM sucrose, and 25 mM -mercap-
toethanol in 200 mM Tris–HCl, pH 9.0). The superna-
tant, following centrifugation at 10,000g for 20 min at
4 °C, was pipetted onto a sucrose cushion (1 M sucrose,
20 mM KCl, and 5 mM MgCl2 in 25 mM Tris–HCl, pH
7.6) in 70 Ti tubes (Beckman), and centrifuged at
55,000 rpm for 4 h at 4 °C (L-70 Ultracentrifuge, Beck-
man). The pellets were resuspended in 25 mM Tris–HCl
buVer (pH 7.6) with 25 mM KCl and 5 mM MgCl2, and
stored at ¡80 °C until use.
ME1 localization studies
The interaction of ME1 with bacteria was studied
using labeled ME1 protein. The ME was Xuorescence-
labeled using NHS-Xuorescein (Pierce, Rockford, IL)
according to the manufacturer’s instructions. BrieXy,
1 mg of each sample was dissolved in 100 l of conjuga-
tion buVer (0.1 M phosphate, 0.15 M NaCl, pH 7.2). Sub-
sequently, 50 l of 2 M NHS-Xuorescein dissolved in
dimethyl sulfoxide (DMSO, Sigma, St. Louis, MO) was
added dropwise to the protein sample, and mixed by stir-
ring. Samples were placed on ice and incubated for 2 h.
The unreacted NHS-Xuorescein was then removed by
dialysis, and the protein concentration was determined
by the Bradford method [25] using a protein assay kit
purchased from Bio-Rad (Hercules, CA). Fluorescein-
labeled proteins were stored at 4 °C in 0.1% NaN3 until
use.
Bacteria grown overnight were used for the study of
ME1 interaction with bacteria. One ml of the culture was
centrifuged in a 1.5 ml micro-centrifuge tube and the pel-
let was washed with 1 ml PBS buVer, resuspended in
100 l of the Xuorescence-labeled ME, incubated at
room temperature covered with aluminum foil. After 4 h
 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151 145

of incubation, the bacteria were brieXy washed with PBS

followed by antifade reagent anti Slowfade Antifade
(Molecular Probes, Eugene, OR) as per manufacturer’s
instructions. The negative control, 100 l of 2 M NHS-
Xuorescein solution, was applied to the bacteria using
the same procedure used for the labeled protein, and
shows no background labeling. Fluorescent images were
taken using a Xuorescent microscope (JEOL 2000 EXII,
Japan).

Study of enzymatic activity

The enzymatic activity of the recombinant ME1 was

analyzed by the depurination of eukaryotic (yeast) and
prokaryotic (E. coli) ribosomes, and was compared to
the depurination activity of native ME1. The depurina-
tion assay was conducted according to Tumer et al. [27].
BrieXy, ribosomes were resuspended in RIP buVer
(150 mM KCl, 100 mM MgCl2, and 100 mM Tris–HCl,
pH 7.2) and incubated with RIPs at 30 °C for 30 min in a
total volume of 100 l. For naked rRNA depurination
the ribosomes were phenol:chloroform extracted and
ethanol precipitated before incubation with RIP. Fol-
lowing incubation, the RIP protein was removed from
the mixture by phenol:chloroform extraction and the Fig. 1. ME1 expression in bacteria. (A) EVect of ME1 on bacterial
growth. E. coli were grown in LB media and IPTG was added to the
RNA was divided into half. Half of the extracted RNA
medium to Wnal concentration of 1 mM when bacterial absorbance
was incubated on ice for 30 min with 1 M aniline acetate was 0.5 at 620. Control D E. coli (pLysS) without plasmid;
(pH 4.5) and precipitated with ethanol. Both aniline- Induced D E. coli (pLysS) with ME-pET28b vector for heterologous
treated and untreated rRNAs were subjected to electro- expression of ME1. The error bars indicate the variation observed
phoresis in a 7 M urea/4.5% polyacrylamide gel and between three experiments. (B) Protein proWle of bacteria expressing
ME1. Extracts from induced and non-induced bacteria were resolved
stained with ethidium bromide. The experiments were
on SDS–PAGE and stained with Coomassie brilliant blue. The arrow
repeated three times. PAP was used as a positive control indicates the new protein band, which appeared in the induced bacte-
and ricin as a negative control. rial extracts. (C) Western blot of the bacterial extracts. The bacterial
extracts were transferred to PVDF membrane and probed with anti-
Comparative modeling ME1 antibodies.

We used the structure of pokeweed antiviral protein
(PAP), an RIP isolated from Phytolacca americana, as a
template to model ME1 (GenBank Accession No.
AAN65450, amino acids 34–299) using the comparative
protein modeling program called SWISS-MODEL [28].
The procedure was conducted as follows: Wnding the tem-
plate structure (code: 1QCG) in the Protein Data Bank
(PDB), aligning it with the sequence of ME1 and calculat-
ing the 3D model. The secondary structure was obtained
with the program DeepView/Swiss-PdbViewer [29].
Results
Bacterial growth and expression of the ME1 recombinant
protein
ME1 was labeled with an N-terminal His6-tag and
expressed in the E. coli pLysS strain by induction with
IPTG. As seen in Fig. 1A, the bacterial expression of
ME1 was not toxic to E. coli. Furthermore, E. coli
showed higher growth during the Wrst 90 min of induc-
tion compared to non-induced cultures. Bacterial growth
slowed down after 150 min as analyzed by bacterial
absorbency, suggesting that prolonged exposure to ME1
recombinant protein may have negative eVects on bacte-
rial growth.
To study the expression of the recombinant protein,
the bacterial extracts were checked for ME1 after 3 h of
induction using 1 mM IPTG. At this point, equal
amounts of non-induced and induced bacterial extracts
were resolved on 12% SDS–PAGE. The Coomassie bril-
liant blue (CBB) staining indicated the appearance of a
30 kDa size new protein band (Fig. 1B), which was close
to the size of the native ME1 from M. expansa on SDS–
PAGE, thus indicating the expression of recombinant
ME1. To conWrm the identity of the new band, the bacte-
rial protein samples were transferred onto a PVDF mem-
brane and probed with anti-ME1 antibodies. The
expression of ME1 protein was conWrmed by the Western
146 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151

blot analysis (Fig. 1C). The results indicated that the
ME1 protein was produced in the bacteria only after
induction with IPTG, because the control sample, to
which IPTG had not been added, did not show this pro-
tein band. The Western blots of the recombinant protein
were also probed with anti-His antibodies. The results
indicate that the anti-His antibodies failed to recognize
the recombinant protein (data not shown). The fact that
anti-ME1 antibodies recognized the recombinant protein
while the anti-His antibodies failed to recognize it sug-
gests that there was an N-terminal deletion of the protein.
It is known that bacteria can process signal sequences in
recombinant proteins [30]. A study using SignalP V1.1
software has shown that both prokaryotes and eukary-
otes can recognize the signal sequence of ME1 [31]. Thus
it is probable that the E. coli was able to recognize the
signal peptide of ME1 and delete the N-terminal
sequence to synthesize mature recombinant ME1.
PuriWcation of the ME1 recombinant protein
The recombinant protein was designed with an N-ter-
minal polyhistidine-tag. However, the protein was not
found to bind to the Ni2+ aYnity column, which further
conWrms the deletion or masking of the N-terminal His6-
tag by E. coli. Hence, we have puriWed the protein by
conventional puriWcation methodologies for RIPs (see
Materials and methods). Since ME1 is a basic protein,
passing the total bacterial protein extracts through the
DEAE column removed most of the acidic proteins in
the bacterial lysates. The Xow-through basic fraction was
collected and passed through an ME1 aYnity column
with 20 mM Tris buVer (pH 8.0). The proteins bound to
the column were eluted with 0–1 M NaCl gradient (Fig.
2A) in Tris–glycine buVer (pH 2.5), and the fractions
were collected and checked for the presence of ME1 pro-
tein. A single protein peak eluted at 200 mM NaCl and
this peak fraction contained ME1 protein as shown by
SDS–PAGE and silver staining analysis (Fig. 2B).
Distribution of ME1
The toxicity of various RIPs is attributed to their abil-
ity to depurinate ribosomes and thus inhibit the protein
synthesis. However, most of the RIPs are localized to
compartments to avoid toxicity to host cells in plants.
Our bacterial growth experiments show that ME1
expression is non-toxic to bacteria and they also show
that the presence of ME1 does not harm the bacteria. So
to study how ME1 interacts with a bacterial cell when it
is added from outside, we labeled the ME1 with NHS-
Xuorescein and incubated it with bacteria for 4 h. The
distribution of labeled ME1 was observed using a confo-
cal microscope; these studies showed that ME1 enters
into the bacteria and is distributed all over the cell (Fig.
3). There was no speciWc localization of ME1 within the
bacteria. The presence of ME1 inside the bacteria does
not harm the bacterial growth, indicating that ME1 does
not inhibit the protein synthesis of bacteria.
Enzymatic activity of ME1
The enzymatic activity of the recombinant and native
ME1 was analyzed by the depurination of eukaryotic
(yeast) and prokaryotic (E. coli) ribosomes. The
recombinant ME1 was incubated with yeast ribosomes

Fig. 2. PuriWcation of recombinant ME1. (A) AYnity chromatography of bacterial recombinant ME1. Protein extract of the induced bacteria were
fractionated through diethylaminoethyl (DEAE) column chromatography to eliminate all acidic proteins. The Xow-through of basic proteins was
collected and passed through an anti-ME1/Sepharose resin column. The bound proteins were eluted with 100 mM glycine–HCl buVer (pH 2.7) with
NaCl gradient. (B) Silver staining of the bacterial recombinant protein. The protein fraction eluted from the aYnity column was resolved on SDS–
PAGE and silver stained.
 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151
Fig. 3. ME1 distribution in the bacteria. Distribution of labeled ME1 in the bacteria. Bacteria grown overnight were incubated for 4 h with Xuores-
cein-labeled ME1 and observed under a confocal microscope. The left side panel shows Xuorescein-labeled ME1 (green) distribution in the bacteria
and the right side panel shows the same under transmitted light. (For interpretation of the references to color in this Wgure legend, the reader is
referred to the web version of this paper.)
for 30 min at 30 °C, and the RNA depurination was con-
Wrmed by aniline treatment. The ribose backbone at the
depurinated site was fragmented by aniline treatment at
low pH (4.0) and a smaller fragment of RNA was
released from the 28S ribosomal RNA, indicating SR
loop depurination of yeast ribosomes (Fig. 4A). Both
native ME1 and recombinant ME1 showed enzymatic
activity on yeast ribosomes. Type I RIPs like PAP show
depurination of bacterial ribosomes, and PAP bacterial
expression indicates that it is toxic to the bacteria. Since
ME1 does not show any toxicity to bacteria, we studied
the eVect of ME1 on E. coli ribosomes. Ribosomes were
obtained from E. coli (DH5
) and used to test the depu-
rination ability of ME1. Treatment of E. coli ribosomes
with native ME1 and recombinant ME1 did not show
any depurination (Fig. 4B), indicating that bacterial
ribosomes were not substrates for ME1. A positive con-
trol showed that PAP depurinated E. coli ribosomes
under experimental conditions. Similarly, the depurina-
tion activity of native and recombinant ME1 was
checked against naked rRNA of bacteria to understand
the role of the ribosomal proteins in ME1 depurination
activity. The results show that ME1 could not depuri-
nate the naked rRNA of E. coli ribosomes (Fig. 4C).
However PAP can depurinate the naked rRNA as well
as being able to depurinate the intact ribosomes. These
results indicate that ME1 interacts with bacterial ribo-
somes in a manner more similar to ricin (type II RIP)
than to PAP (type I RIP).
Structural modeling of ME1
The amino acid sequences of ME1 contain close
identity and similarity to other type I RIPs. For
instance, ME1 has 33.3% identity and 83% similarity
147
with PAP (an RIP from pokeweed, PDB code 1QCG),
31.2 and 70% with
-trichosanthin (an RIP isolated
from Trichosanthes kirilowii, PDB code 1MRJ), and
31.3 and 67.4% with -momorcharin (RIP isolated from
Momordica charantia, PDB code 1CF5). The three-
dimensional structure of PAP has been resolved by X-
ray crystallography [31]. The high degree of sequence
similarity between ME1 and PAP indicates that both
polypeptides are closely related in structure. The struc-
ture of ME1 (amino acid sequence between 34 and 299
amino acids) was modeled using the coordinates of PAP
(Fig. 5A). The three-dimensional model of the ME1
sequence can be nicely superimposed on the structures
of the type I RIPs PAP and -momorcharin (Fig. 5B).
This superimposition holds especially true for the
-
helices and strands of -sheet, which constitute the
overall folding pattern of the proteins. As a result, the
network of amino acid residues forming the adenine-
binding site of all these protein models aligns in a simi-
lar pattern.
Enzymatically active PAP consists of two distinct
domains and contains a well-deWned secondary struc-
ture: eight
-helices and a -sheet composed of six
strands [32]. The active site residues include Glu176 and
Arg180, involved in catalysis, and Tyr72 and Tyr123,
involved in binding to the target adenine base; all of
these residues are conserved in the RIP family [33]. Mul-
tiple sequence alignment of ME1 with RIPs (PAP and
momorcharin) revealed that these active site residues are
fully conserved in ME1 (Glu135, Arg138, Tyr41, and
Tyr86). Least squares superposition of the active site
regions of PAP and ME1 shows that there is variation in
spatial positions of active residues in both enzymes
(Figs. 5C and D). However, the structural position of
Tyr41 in ME1 is similar to the corresponding residue in
148 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151

Fig. 4. Enzyme activity of native and recombinant ME1. (A) Depurination of yeast ribosomes. Yeast ribosomes were incubated with native and
recombinant ME1. The depurination of the ribosomes was indicated by the release of fragments from ribosomal RNA after aniline treatment. (B)
Depurination of bacterial ribosomes. The bacterial ribosomes were treated as above with native and recombinant ME1 followed by aniline treat-
ment. Bacterial ribosomes failed to release any fragment after aniline treatment. (C) Depurination of naked bacterial ribosomal RNA. The results
presented here are representative of three separate experiments.

region from Thr30 to Asp47 in ME1 closely resembles the

structure of PAP (Thr61 to Asp78).

Discussion

Recombinant ME1 was expressed in E. coli, and the

Fig. 5. Structural models of ME1 and PAP. (A) Ribbon model showing
the architecture of
-helix and  strands of ME1 (amino acids 34–299)
obtained using comparative protein structure homology modeling soft-
ware SWISS-MODEL. Deep View Swiss-Pdb viewer was used to generate
the Wgure. (B) Structure alignment of ME with PAP and momorcharin.
The three-dimensional model of the ME1 sequence was superimposed on
the structures of the type I RIPs PAP and -momorcharin. (C and D) Ste-
reodiagrams of the active site in PAP and ME1. The three-dimensional
structure of the amino acid residues involved in the active site of PAP and
ME1 are presented. (For color Wgure please refer to the web version).
PAP (Tyr72), although a slight positional rotation of the
aromatic ring has occurred. The structural positions of
several motifs exhibit close similarity (Fig. 5B), and the
eVect of ME1 on bacterial growth was studied to under-
stand its toxicity. Our results were quite surprising consid-
ering that the expression of RIPs in bacteria has been
complicated by the toxic nature of these proteins. In the
past, several attempts were made to express PAP full
length and N-terminal truncated, and the recombinant
proteins showed very low expression, probably due to RIP
toxicity in E. coli [18,19,34,35]. High-level expression of
active PAP was Wnally achieved by removing the N-termi-
nal amino acid region [36,37]. Bouganin (type I RIP) with
truncated N- and C-terminal regions has been expressed
with a pectate lyase (pel B) leader sequence and the active
enzyme has been collected from the periplasmic fraction
[38]. This bacterial toxicity of RIPs appears to be the
result of their ability to depurinate prokaryotic ribosomes.
This depurination may result in protein synthesis inhibi-
tion in E. coli, as it has been suggested that A-2660 inter-
acts with elongation factors EF-G and EF-Tu [39]. While
the variation in bacterial expression and toxicity to bacte-
ria is RIP-speciWc, successful expression of recombinant
RIPs without bacterial toxicity has usually been the result
of mis-folding, in which the expressed protein is inactive.
In the present study, ME1 expression did not show nega-
tive eVects on bacterial growth, even though the puriWed
recombinant protein was able to depurinate yeast ribo-
somes. This depurination activity indicated that the
ME1 expressed by the bacteria had proper folding and
was active. That ME1 does not cause bacterial growth
 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151
inhibition may provide clues about the behavior of ME1
towards prokaryotic ribosomes.
ME1 is known to localize the cell surface in Rhizocto-
nia solani and cause toxicity to this fungus [40]. How-
ever, in the case of E. coli no such surface interaction
and toxicity were observed. Since ME1 is non-toxic to
bacteria we have checked its interaction with bacterial
cells. We have checked if it is possible for the bacteria to
not uptake ME1 or to localize ME1 into a compartment
and thus render it non-toxic. Fluorescein-labeled ME1
was incubated with bacterial cultures for 4 h and the
ME1 distribution in the bacteria was studied. These
studies indicated that the labeled ME1 can enter into
bacteria and no localization of labeled ME1 was
observed. This behavior leads to a situation where ME1
is distributed throughout the cytoplasm and is thus
available to interact with the protein synthesis machin-
ery. Native ME1 from M. expansa roots has been shown
to be toxic to some bacteria, including Bacillus subtilis,
Pseudomonas syringae, and Agrobacterium tumifaciens,
but not to E. coli when supplied externally in inhibition-
halo plate assays [23]. The present expression studies
provide further conWrmation that ME1 is not toxic to E.
coli, probably due to the related result that ME1 does
not depurinate E. coli ribosomes. RIPs are known to be
selective in their substrate ribosomes. Type II RIPs, and
the A-chains of ricin and abrin, are highly active against
mammalian ribosomes, weakly active against plant ribo-
somes, and totally inactive against E. coli ribosomes [41].
However, type I RIPs like PAP have shown activity
against prokaryotic ribosomes and have depurinated
their prokaryotic RNA [17]. The reason for such selec-
tive depurination by RIPs is not known but is possibly
due to the protein–protein interaction of RIPs with some
ribosomal proteins [42]. RIPs are known to interact with
ribosomal proteins such as L3; however, it is not clear
what the signiWcance is of such interactions. It has been
shown that mutations in the L3 protein do not aVect the
direct binding of PAP to L3, but aVect the binding of
PAP to L3 in the ribosomes. These data indicate that the
quaternary structure of ribosomes is an essential factor
in the depurination of ribosomes by RIPs [42]. In the
present study the depurination of rRNA both as intact
ribosomes and as naked RNA was similar and ME1 was
found to be unable to depurinate rRNA in either condi-
tion, while the PAP could depurinate under both condi-
tions. This Wnding is very signiWcant, as few studies have
been done on the eVect of RIPs on bacterial ribosomes.
ME1 shows structural similarity with PAP. In the
present study our models show the close similarity
between PAP and ME1 in overall structure but indicate
that the orientation of the active site amino acids is
diVerent in both proteins. The active site amino acids are
conserved in ME1 (Glu135, Arg138, Tyr41, and Tyr86);
however, the spatial position of the active site residues
diVers from PAP (Glu176, Arg180, Tyr72, and Tyr123). This
149
diVerence could be the reason for their diVerent enzy-
matic activities towards bacterial ribosomes.
RIPs have signal sequences that speciWcally localize
them into cellular compartments to avoid cellular toxic-
ity. Due to their N-terminal signal sequence, many RIPs
are compartmentalized and secreted in such a way as to
avoid damage to the host cell ribosomes [1]. The N-termi-
nal sequence of native ME1 isolated from M. expansa
root tissue has been found to lack the Wrst 33 leading
amino acids found in the cDNA sequence, indicating that
this sequence is removed during the maturation of the
protein [10]. The signal sequence of ME1 has a hydro-
phobic region of 22 amino acids, which forms part of the
signal motif in the N-terminal signal peptides of both
prokaryotes and eukaryotes. The present study indicates
that the N-terminal region of ME1 which is processed
and expressed in bacteria does not have the signal pep-
tide. Similarly, Mirabilis antiviral protein (MAP) has
been expressed with its signal sequence in E. coli and the
recombinant protein was found to be 2.5 kDa smaller
than the expected size [17]. The authors of the MAP
study suggest that the signal peptide may have been
cleaved and the MAP transported to the ER lumen. The
signal peptide of ME1 shows sequence similarities with
the MAP signal sequence and it is processed in the bacte-
ria. Thus it is possible that bacteria have the ability to
process these proteins. However, it is diYcult to predict if
this signal processing leads to the same result as the pro-
cessing and transportation of MAP. Interestingly, labeled
native ME1 from a plant can enter the bacterial cell and
this labeled ME1 also distributes uniformly in the bacte-
rial cell without any speciWc localization. Taken together,
our observations indicate that the recombinant ME1 is
active, suggesting that the protein is folded correctly in
E. coli and this active protein does not target bacterial
protein synthesis machinery as bacterial ribosomes are
not substrates for this enzyme. This Wnding suggests that
ME1 and PAP show diVerent substrate speciWcities even
though they show close structural similarity. This sub-
strate speciWcity will facilitate ME1 expression in bacteria
and its use as an immunotoxin.
Acknowledgments
This work was supported by the Colorado State Uni-
versity Agricultural Experiment Station and by a
National Science Foundation Faculty Early Career
Development Award MCB-0093017 (to J.M.V.).
References
[1] K. Nielsen, R.S. Boston, Ribosome-inactivating proteins: a plant
perspective, Annu. Rev. Plant Physiol. Plant Mol. Biol. 52 (2001)
785–816.
150 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151
[2] Y. Endo, K. Mitsui, M. Motizuki, K. Tsurugi, The mechanism of
action of ricin and related toxic lectins on eukaryotic ribosomes:
the site and the characteristics of the modiWcation in 28S ribo-
somal RNA caused by the toxins, J. Biol. Chem. 262 (1987) 5908–
5912.
[3] Y. Endo, K. Tsurugi, RNA N-glycosidase activity of ricin A-
chain: mechanism of action of the toxic lectin ricin on eukaryotic
ribosomes, J. Biol. Chem. 263 (1988) 8735–8739.
[4] Y. Endo, A. Gluck, I.G. Wool, Ribosomal RNA identity elements
for ricin A-chain recognition and catalysis, J. Mol. Biol. 254 (1991)
848–855.
[5] F. Stirpe, L. Barbieri, M.G. Batelli, M. Soria, D.A. Lappi, Ribo-
some-inactivating proteins from plants: present status and future
prospects, BioTechnology 10 (1992) 405–412.
[6] L. Barbieri, P. Gorini, P. Valbonesi, P. Castiglioni, F. Stirpe, Unex-
pected activity of saporins, Nature 372 (1994) 624.
[7] L. Barbieri, P. Valbonesi, E. Bonora, P. Gorini, A. Bolognesi, F.
Stirpe, Polynucleotide:adenosine glycosidase activity of ribosome-
inactivating proteins: eVect on DNA, RNA and poly (A), Nucleic
Acids Res. 25 (1997) 518–522.
[8] A. Bolognesi, L. Polito, C. Lubelli, L. Barbieri, A. Parente, F.
Stirpe, Ribosome-inactivating and adenine polynucleotide glyco-
sylase activities in Mirabilis jalapa L. tissues, J. Biol. Chem. 277
(2002) 13709–13716.
[9] B.A. Parikh, C. Coetzer, T.E. Tumer, Pokeweed antiviral protein
regulates the stability of its own mRNA by a mechanism that
requires depurination, but can be separated from depurination of
the alpha-sarcin/ricin loop of rRNA, J. Biol. Chem. 277 (2002)
41428–41434.
[10] R. Vepachedu, H.P. Bais, J.M. Vivanco, Molecular characteriza-
tion and post-transcriptional regulation of ME1, a type I ribo-
some-inactivating protein from Mirabilis expansa, Planta 217
(2003) 498–506.
[11] J. Logemann, G. Jach, H. Tommerup, J. Mundy, J. Schell, Expres-
sion of a barley ribosome-inactivating protein leads to increased
fungal protection in transgenic tobacco plants, BioTechnology 10
(1992) 305–308.
[12] G. Jach, B. Gornhardt, J. Mundy, J. Logemann, E. Pinsdorf, R.
Leah, J. Schell, C. Maas, Enhanced quantitative resistance against
fungal disease by diVerent barley antifungal proteins in transgenic
tobacco, Plant J. 8 (1995) 97–109.
[13] P.F. Dowd, W.N. Zuo, J.W. Gillikin, E.T. Johnson, R.S. Boston,
Enhanced resistance to Helicoverpa zea in tobacco expressing an
activated form of maize ribosome-inactivating protein, J. Agric.
Food Chem. 51 (2003) 3568–3574.
[14] L. Barbieri, A. Bolognesi, P. Valbonesi, L. Polito, F. Olivieri, F.
Stirpe, Polynucleotide: adenosine glycosidase activity of immuno-
toxins containing ribosome-inactivating proteins, J. Drug Target 8
(2000) 281–288.
[15] J.E. Smallshaw, V. Ghetie, J. Rizo, J.R. Fulmer, L.L. Trahan, M.A.
Ghetie, E.S. Vitetta, Genetic engineering of an immunotoxin to
eliminate pulmonary vascular leak in mice, Nat. Biotechnol. 4
(2003) 387–391.
[16] L. Herrera, S. Yarbrough, V. Ghetie, D.B. Aquino, E.S. Vitetta,
Treatment of SCID/human B cell precursor ALL with anti-CD19
and anti-CD22 immunotoxins, Leukemia 17 (2003) 334–338.
[17] M.R. Hartley, G. Legname, R. Osborn, Z. Chen, J.M. Lord, Single-
chain ribosome-inactivating proteins from plants depurinate Esch-
erichia coli 23S ribosomal RNA, FEBS Lett. 290 (1991) 65–68.
[18] J.M. Dore, E. Gras, F. Depierre, J. Wijdenes, Mutations dissociat-
ing the inhibitory activity of the pokeweed antiviral protein on
eukaryote translation and Escherichia coli growth, Nucleic Acids
Res. 21 (1993) 4200–4205.
[19] J.A. Chaddock, J.M. Lord, M.R. Hartley, L.M. Roberts, Pokeweed
antiviral protein (PAP) mutations which permit E. coli growth do
not eliminate catalytic activity towards prokaryotic ribosomes,
Nucleic Acids Res. 22 (1994) 1536–1540.
[20] J. Kataoka, N. Habuka, M. Furuno, M. Miyano, Y. Takanami, A.
Koiwai, DNA sequence of Mirabilis antiviral protein (MAP), a
ribosome-inactivating protein with an antiviral property, from
Mirabilis jalapa L. and its expression in Escherichia coli, J. Biol.
Chem. 266 (1991) 8426–8430.
[21] L. Barbieri, M.G. Battelli, F. Stirpe, Ribosome-inactivating pro-
teins from plants, Biochim. Biophys. Acta 1154 (1993) 237–282.
[22] T. Girbes, L. Barbieri, M. Ferreras, F.J. Arias, M.A. Rojo, R.
Iglesias, C. Alegre, C. Escarmis, F. Stirpe, EVects of ribosome-
inactivating proteins on Escherichia coli and Agrobacterium
tumefaciens translation systems, J. Bacteriol. 175 (1993) 6721–
6724.
[23] J.M. Vivanco, B.J. Savary, H.E. Flores, Characterization of two
novel type I ribosome-inactivating proteins from the storage roots
of the Andean crop Mirabilis expansa, Plant Physiol. 119 (1999)
1447–1456.
[24] J.M. Vivanco, H.E. Flores, Biosynthesis of ribosome-inactivat-
ing proteins from callus and cell suspension cultures of Mirabi-
lis expansa (Ruiz & Pavon), Plant Cell Rep. 19 (2000) 1033–
1039.
[25] M.M. Bradford, A rapid and sensitive method for the quantiWca-
tion of microgram quantities of protein utilizing the principle of
protein-dye binding, Anal. Biochem. 72 (1976) 248–254.
[26] R.S. Sikorski, P. Hieter, A system of shuttle vectors and yeast host
strains designed for eYcient manipulation of DNA in Saccharo-
myces cerevisiae, Genetics 122 (1989) 19–27.
[27] N.E. Tumer, D.J. Hwang, M. Bonness, C-terminal deletion mutant
of pokeweed antiviral protein inhibits viral infection but does not
depurinate ribosomes, Proc. Natl. Acad. Sci. USA 94 (1997) 3866–
3871.
[28] T. Schwede, J. Kopp, N. Guex, M.C. Peitsch, SWISS-MODEL: an
automated protein homology-modeling server, Nucleic Acids Res.
31 (2003) 3381–3385.
[29] N. Guex, M.C. Peitsch, Swiss-model and the Swiss-PdbViewer: an
environment for comparative protein modeling, Electrophoresis
18 (1997) 2714–2723.
[30] T. Hattori, S. Ichihara, K. Nakamura, Processing of a plant vacu-
olar protein precursor in vitro, Eur. J. Biochem. 166 (1987) 533–
538.
[31] N. Henrik, E. Jacob, B. Søren, V.H. Gunnar, IdentiWcation of pro-
karyotic and eukaryotic signal peptides and prediction of their
cleavage sites, Protein Eng. 10 (1997) 1–6.
[32] I.V. Kurinov, D.E. Myers, J.D. Irvin, R.M. Uckun, X-ray crys-
tallographic analysis of the structural basis for the interactions
of pokeweed antiviral protein with its active site inhibitor and
ribosomal RNA substrate analogs, Protein Sci. 8 (1999) 1765–
1772.
[33] A.F. Monzingo, E.J. Collins, S.R. Ernst, J.D. Irvin, J.D. Robertus,
The 2.5 Å structure of pokeweed antiviral protein, J. Mol. Biol.
233 (1993) 705–715.
[34] J. Kataoka, N. Habuka, C. Masuta, M. Miyano, A. Koiwai, Isola-
tion and analysis of a genomic clone encoding a pokeweed antivi-
ral protein, Plant Mol. Biol. 20 (1992) 879–886.
[35] J. Kataoka, H. Ago, N. Habuka, M. Furuno, C. Masuta, M. Miy-
ano, A. Koiwai, Expression of a pokeweed antiviral protein in
Escherichia coli and its characterization, FEBS Lett. 320 (1993)
31–34.
[36] J. Random, J. Schlick, J. Poyet, A. Hoeveler, Over expression of
pokeweed antiviral protein in Escherichia coli: puriWcation, char-
acterization and comparison with native sources, Plant Sci. 120
(1996) 19–28.
[37] F. Rajamohan, C.R. Engstrom, T.J. Denton, L.A. Engen, I. Kouri-
nov, F.M. Uckun, High-level expression and puriWcation of bio-
logically active recombinant pokeweed antiviral protein, Protein
Expr. Purif. 16 (1999) 359–368.
[38] M.T. den Hartog, C. Lubelli, L. Boon, S. Heerkens, Ortiz,
A.P. Buijsse, M. de Boer, F. Stirpe, Cloning and expression
 R. Vepachedu et al. / Protein Expression and PuriWcation 40 (2005) 142–151 151

of cDNA coding for bouganin, Eur. J. Biochem. 269 (2002)
1772–1779.
[39] D. Moazed, J.M. Robertson, H.F. Noller, Interaction of elonga-
tion factors EF-G and EF-Tu with a conserved loop in 23S RNA,
Nature 334 (1988) 362–364.
[40] S.W. Park, N.M. Stevens, J.M. Vivanco, Enzymatic speciWcity of
three ribosome-inactivating proteins against fungal ribosomes,
and correlation with antifungal activity, Planta 216 (2002)
227–234.
[41] J.M. Lord, M.R. Hartley, L.M. Roberts, Ribosome-inactivating
proteins of plants, Semin. Cell Biol. 2 (1991) 15–22.
[42] K.A. Hudak, J.D. Dinman, N.E. Tumer, Pokeweed antiviral pro-
tein accesses ribosomes by binding to L3, J. Biol. Chem. 274 (1999)
3859–3864.