| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
,1
* Department of Clinical Pharmacology, Research Institute, International Medical Center of Japan, Tokyo, Japan;
PRESTO, Japan Science and Technology Agency, Saitama, Japan;
Department of Bacteriology, Nara Medical University, Kashihara, Japan;
Department of Medical Ecology and Informatics, Research Institute, International Medical Center of Japan, Tokyo, Japan; and
|| Center for Cancer Research, Massachusetts Institute of Technology, Cambridge, Massachusetts, USA
1Correspondence: Department of Clinical Pharmacology, Research Institute, International Medical Center of Japan, 1–21-1 Toyama, Shinjuku-ku, Tokyo 162-8655, Japan. E-mail: knishika{at}ri.imcj.go.jp
ABSTRACT
Infection with Shiga toxin (Stx)-producing Escherichia coli O157:H7 causes bloody diarrhea and hemorrhagic colitis in humans, sometimes resulting in fatal systemic complications. Among the known Stx family members, Stx2 is responsible for the most severe forms of disease. Stx2 binds to target cells via multivalent interactions between its B-subunit pentamer and globotriaosyl ceramide. After binding, it is first retrogradely transported to the Golgi and then to the endoplasmic reticulum (ER). Using a multivalent peptide library approach, we identified a tetravalent peptide that exhibits a high affinity for the Stx2 B-subunit pentamer (KD=0.13 µM) and markedly inhibits Stx2 cytotoxicity. The tetravalent peptide exerted its inhibitory effects by inducing aberrant cellular transport of Stx2. Although the tetravalent peptide/Stx2 complex was incorporated into cells and translocated to the Golgi, this process was followed by the effective degradation of Stx2 in an acidic compartment rather than by its transfer to the ER. This peptide thoroughly protected mice from a fatal dose of E. coli O157:H7 even when administered after an established infection. Thus, the multivalent peptide library approach enabled the identification of a peptide-based Stx2 inhibitor that has remarkable therapeutic potency and appears to function by inducing aberrant cellular transport and degradation of Stx2.—Nishikawa, K., Watanabe, M., Kita, E., Igai, K., Omata, K., Yaffe, M. B., Natori, Y. A multivalent peptide library approach identifies a novel Shiga toxin inhibitor that induces aberrant cellular transport of the toxin.
Key Words: multivalent interaction • retrograde transport • globotriaosylceramide • Shiga toxin-producing E. coli infection • therapeutic agent
MULTIVALENT INTERACTIONS BETWEEN carbohydrate-recognizing molecules and their intended carbohydrates are involved in a wide range of physiological and pathological events, such as the invasion of bacterial toxins and viruses into target cells, attachment of leukocytes to endothelial cells, and metastasis of tumor cells (1)
. Inhibitors that effectively inhibit such interactions would be useful not only for treatment of related human diseases, but also for studying the molecular mechanisms by which individual molecules function. We focused here on Shiga toxin 2 (Stx2), a major virulence factor of Stx-producing Escherichia coli (STEC) infections (2
3
4
5)
. We describe the identification of an inhibitor that directly blocks the multivalent interactions between Stx2 and its cell surface receptor. We have also taken advantage of this specific inhibitor of Stx2 to study the mechanisms underlying the intracellular transport of Stx2.
Stx2 consists of a catalytic A-subunit and a pentameric B-subunit. The former has 28S rRNA N-glycosidase activity and inhibits eukaryotic protein synthesis, whereas the latter is responsible for binding to the functional cell surface receptor, Gb3 (Gal
(1–4)-Galß (1–4)-glucoseß1-ceramide) (5
6
7)
. The recently resolved crystal structure of Stx2 reveals the presence of three distinctive binding sites (i.e., sites 1, 2, and 3) on each B-subunit monomer for the trisaccharide moiety of Gb3 (8)
. Accordingly, there are 15 potential trisaccharide binding sites per pentamer. The ability of Stx2 to selectively bind to target cells with such high affinity is attributable to multivalent interactions between the B-subunit pentamer and the trisaccharide of Gb3 on the plasma membrane. This is sometimes referred to as the "clustering effect."
On the basis of these facts, several Stx inhibitors containing the trisaccharide in multiple configurations have been developed (9
10
11
12
13
14)
. We recently developed a carbosilane dendrimer carrying six trisaccharides (referred to as SUPER TWIG (1)6) that is the first synthetic Stx2 inhibitor known to function in vivo (11
, 15)
. The clinical application of these trisaccharide-containing inhibitors, however, has been substantially hampered by the synthetic complexity of its trisaccharide moiety. Moreover, this strategy is not suitable for the identification of inhibitors of molecules that recognize carbohydrate ligands with highly complicated structures. Other widely used techniques, including screening of chemical compounds and phage display libraries, have also failed to identify such inhibitors because they are not based on the "clustering effect."
We earlier developed an affinity-based peptide library technique we used to identify specific inhibitors of ZAP-70 protein Tyr kinase that directly bind to the catalytic site of this protein with high affinity and selectivity (16)
. In this study we have improved this technique to make use of the "clustering effect" by synthesizing a multivalent peptide library. Using this library, we identified a tetravalent peptide that binds to the Stx2 B-subunit pentamer with high affinity and markedly inhibits Stx2, both in vitro and in vivo. Structural and mutational analyses indicate that the tetravalent peptide binds to the B-subunit exclusively in site 3-oriented configurations.
After binding to Gb3, Stx2 is first transported to the Golgi apparatus in a retrograde manner, then transported to the endoplasmic reticulum (ER); its catalytic A-subunit is released into the cytosol, where it inhibits protein synthesis (17
, 18)
. Although retrograde transport is essential for Stx2 to exert cytotoxic activity, the molecular mechanisms underlying this transport are still poorly understood. In particular, there is no information regarding the specific role of the individual trisaccharide binding sites in the retrograde transport of Stx2. Here we have found that the tetravalent peptide inhibitor prevents Stx2 cytotoxicity by inducing its aberrant transport from the Golgi to an acidic compartment rather than to the ER, which resulted in its efficient degradation. Furthermore, using the tetravalent peptide, we have found that the interaction between Stx2 and Gb3 through site 3, which is specifically blocked by the peptide, plays an essential role in the intracellular transport of Stx2 from the Golgi to the ER.
MATERIALS AND METHODS
Materials
Recombinant Stx2, recombinant histidine-tagged Stx2 B-subunit (2BH), 2BH with Ala substitution for Trp33 (2BH-W33A), 2BH with Glu substitution for Asp17 (2BH-D17E), and 2BH with Ala substitutions for both Trp29 and Gly61 (2BH-W29A/G61A) were prepared as described previously (19)
. Constructs for 2BH with Ala substitution for Asp16 (2BH-D16A), Ala substitution for Asp17 (2BH-D17A), and Ala substitution for Glu64 (2BH-E64A) were prepared using a QuikChange kit (Stratagene, San Diego, CA, USA) with the following mutagenic oligonucleotides: TTTCCAAGTATAATGAGGCTGACACATTTACAGTGAAGGT, TCCAAGTATAATGAGGATGCTACATTTACAGTGAAGGTTG, and AGAATACTGGACCAGTCGCGCGAATCTGCAACCGTTACTG, respectively, and these mutants were prepared as described previously (19)
. The free trisaccharide analog with a short alkyl chain (n-hexenyl trisaccharide) and SUPER TWIG (1)6 were synthesized as described elsewhere (15
, 20)
. Rabbit anti-Stx2 antiserum was provided by S. Yamasaki and T. Hamabata, International Medical Center of Japan, Tokyo. Stx2, 2BH, and 2BH-W29A/G61A containing Alexa Fluor 488 conjugated to their amino groups, as well as a tetravalent peptide containing Alexa Fluor 555 conjugated to the single carboxyl group were prepared by using the Alexa Fluor Protein Labeling Kit (Molecular Probes, Eugene, OR, USA).
Peptides and peptide libraries
Tetravalent peptide libraries, peptide monomers, and tetravalent peptides were synthesized using N--FMOC-protected amino acid and standard BOP/HOB coupling chemistry. The four peptide-chains present in a given tetravalent peptide (see Fig. 1
A) were elongated at the same time from the amino groups of the core polylysine (Fmoc MAP resin, Applied Biosystems, Foster City, CA, USA). The Met-Ala sequence at the amino terminus of the library peptides was included to verify that peptides from this mixture were being sequenced and to qualify the peptides. The Ala situated next to the last degenerate position provided an estimate of peptide loss during sequencing. Acetylation of the terminal amino groups of a tetravalent peptide (PPP-tet) that has four peptide motifs, MAPPPRRRRA, was performed using acetylimidazole.
|
Library screening
Recombinant 2BH or 2BH-W33A (1.1 mg of protein) bound to beads was packed in a small column and incubated with 600 µg of a given library peptide in PBS for 10 min at room temperature. After extensive washing with PBS containing 0.5% Nonidet P-40, bound peptides were eluted with 30% acetic acid, dried overnight, and sequenced on an Applied Biosystems model 477A protein sequencer. The abundance of each amino acid in the peptides recovered from the 2BH-W33A column in a given cycle was subtracted from data obtained from the 2BH column. To calculate the relative amino acid preference at each degenerate position, the corrected quantities of amino acids were compared with those in the starting mixture to calculate the ratios of abundance of amino acids (16
, 21)
. The relative abundance of individual amino acids at the degenerate positions reflects the relative abundance of high-affinity peptides containing these amino acids (21)
.
Kinetic analysis of the binding between inhibitory peptides and immobilized 2B-subunits
The binding of inhibitory peptide or other ligands to immobilized 2BH was quantified using a BIAcoreTM system instrument (BIAcore, Uppsala, Sweden) as described previously (13)
. The resonance unit is an arbitrary unit used by the BIAcore system. The binding kinetics were analyzed by Scatchard plotting using the software BIAEVALUATION 3.0 (BIAcore).
ELISA of the binding between 2BH and inhibitory peptides
The indicated concentration of a given inhibitor, dissolved in PBS, was coated onto each well of a 96-well ELISA plate (24 h at 4°C). After blocking, the plate was incubated with 2BH or 2BH mutant (0.1 µg/ml) for 1 h at room temperature. Bound 2BH or 2BH mutant was detected using rabbit anti-Stx2 antiserum.
Energy minimization analysis
One of the identified Stx binding motifs containing the sequence MA-PPPRRRR-A was docked to the Stx2 B-subunit by means of constant temperature, molecular dynamics simulations carried out with the program package CHARMm (22)
. The temperature was set at 300K in the course of the simulations, and the CHARMm version 22.0 force field was applied. The electrostatic interactions were calculated from Ewald’s summation. The energy gained by the docking (
E; kcal/mol) was calculated by CHARMm. Atomic coordinates of the Stx2 B-subunit for the initial conditions were provided by the Protein Data Bank (code: 1R4P and 1R4Q). The docking model was prepared by using RasMol (http://www.umass.edu/microbio/rasmol/).
Cytotoxicity assay
Subconfluent Vero cells cultured in a 96-well plate in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% FCS were treated with Stx2 (1 pg/ml) in the absence or presence of a given inhibitor for 72 h at 37°C. To examine the effect of delayed treatment with inhibitors, Vero cells were first treated with Stx2 (30 pg/ml) for 1 h at 4°C. After washing, the cells were subsequently cultured for 72 h at 37°C in the presence of a given inhibitor (200 µM of binding units). The inhibitor was added to the culture medium at times ranging from 10 min to 12 h after washing. The relative number of living cells was determined by a WST-1 Cell Counting Kit (Wako Pure Industries, Osaka, Japan). This assay is based on the cleavage of the tetrazolium salt, WST-1, by mitochondrial dehydrogenases to form formazan. The formazan dye produced by viable cells was quantified by measuring absorbance at 440 nm. Using this assay, Vero cell toxicity was detected at a minimal concentration of 0.1 pg/ml Stx2.
125I-Stx2 binding assay
Vero cells grown in a 24-well plastic microplate were treated with 125I-Stx2 (2x106 cpm/µg of protein, 1 µg/ml) in the absence or presence of a given inhibitor for 30 min at 4°C. After extensive washing, cells were dissolved in lysis solution (0.1 M NaOH and 0.5% SDS), and the recovered radioactivity was measured by a gamma counter.
Intracellular localization of Stx2, 2BH, and PPP-tet
Subconfluent Vero cells in a glass dish (35 mm) were treated with Alexa Fluor 488-labeled Stx2 (1 µg/ml) in the absence or presence of Alexa Fluor 555-labeled PPP-tet (7 µM of binding units) for 1 h at 4°C. After washing, cells were incubated at 37°C for the period indicated and analyzed by confocal laser scanning microscopy (Olympus, Melville, NY, USA). Vero cells treated with Alexa Fluor 488-labeled 2BH or 2BH-W29A/G61A (10 µg/ml) in the absence or presence of PPP-tet (200 µM of binding units) were also analyzed as described above. To examine colocalization of Stx2 with an ER marker, Hsp47, Vero cells were treated with Stx2 (1 µg/ml) in the absence or presence of PPP-tet (70 µM of binding units) for 1 h at 37°C and fixed with 3% paraformaldehyde. Stx2 immunostaining was carried out with rabbit anti-Stx2 polyclonal antibody (pAb) followed by Alexa Fluor 488-labeled goat anti-rabbit IgG (Molecular Probes). Immunostaining for Hsp47 was performed using mouse anti-Hsp47 monoclonal antibody (mAb) (StressGen, San Diego, CA, USA) and Alexa Fluor 546-labeled goat anti-mouse IgG (Molecular Probes). To examine colocalization of Stx2 with the Golgi, Vero cells were treated with BODIPY TRC5 ceramide complexed to BSA (5 µM, Molecular Probes), which localizes to the Golgi, for 30 min at 37°C. After washing, the cells were incubated with Alexa Fluor 488-labeled 2BH or 2BH-W29A/G61A (10 µg/ml) in the absence or presence of PPP-tet (200 µM of binding units) for 1 h at 4°C. After washing, cells were incubated an additional hour at 37°C and analyzed by confocal laser scanning microscopy. To examine the transport of Stx2 from the Golgi to the ER, Vero cells were treated with Stx2 (1 µg/ml) for 1 h at 37°C (1 h incubation). Cells were then washed and incubated an additional 2 h (3 h incubation in total) or 5 h (6 h incubation in total) at 37°C. PPP-tet (200 µM of binding units) was added after the 3 h incubation had elapsed. At 1, 3, and 6 h, cells were fixed with 3% paraformaldehyde and immunostaining for Stx2 and Hsp47 was carried out using specific antibodies. Immunostaining for GM130, a Golgi marker, was performed using mouse anti-GM130 IgG (BD Biosciences PharMingen, San Diego, CA, USA) and Alexa Fluor 546-labeled goat anti-mouse IgG (Molecular Probes).
Intracellular degradation of Stx2
Vero cells grown in a 24-well plastic microplate were treated with 125I-Stx2 (2x106 cpm/µg of protein, 1 µg/ml) in the absence or presence of PPP-tet (200 µM of binding units) for 2 h at 37°C. After extensive washing, the culture medium was changed to DMEM without FCS and cells were incubated for 1 h. After additional washing (time 0), cells were further incubated in DMEM without FCS for the periods indicated. The culture medium was recovered and 100% ice-cold trichloroacetic acid (TCA) was added to a final concentration of 10%. After centrifugation, radioactivity present in the TCA-soluble supernatants was measured by a gamma counter. Chloroquine (100 µM) was added 1 h before treatment with 125I-Stx2 and was present during the entire incubation.
Mouse infection protocol
Specific pathogen-free, 3-wk-old female C57BL/6 mice (Charles River Breeding Laboratories, Wilmington, MA, USA) were maintained on a low-protein diet for caloric malnutrition (23)
. Mice 5 wk of age were infected intragastrically with 2 x 106 colony-forming units of E. coli O157:H7 strain N-9 as described elsewhere (23)
. The indicated amount of tetravalent peptide or saline was administered intragastrically to the mice twice a day from day 2 to 5. To examine the effect of delayed onset of the treatment, Ac-PPP-tet (36 nmol of binding units/g of body wt) was administered twice a day for 4 days. Data were analyzed by Kaplan-Meier survival analysis or, when no mice had died by the end of the observation period, by Fisher’s exact test. All animal experiments were approved by the committee of Animal Ethics of Nara Medical University prior to their commencement.
RESULTS
Tetravalent peptide library screening and Stx2 inhibitor identification
We developed a tetravalent peptide library to identify a novel Stx binding motif. The library was comprised of compounds containing a polylysine core bifurcating at both ends with four randomized peptides (Fig. 1A
). The core structure of the compounds, including the length of the spacers, were optimized for their high-affinity binding to the Stx2 B-subunit pentamer based on structural requirements already established using a series of SUPER TWIGs carrying the trisaccharides as an Stx binding unit (19)
. The tetravalent peptide library was screened for compounds capable of binding to recombinant histidine-tagged Stx2 B-subunit (2BH). Potential compounds could not demonstrate binding to a mutant Stx2 B-subunit that had an Ala residue substituted for Trp33 (2BH-W33A) in trisaccharide binding site 3. Mutation of this site abolishes receptor binding activity (19)
. All of the SUPER TWIGs that bind to the B-subunit with high affinity bind exclusively to site 3 (19)
, making this site an excellent target for inhibitor design. As shown in Fig. 1B
, Arg and Asn were strongly selected at positions 1 and 3, respectively, and basic amino acids were preferred at all positions. Based on this result, a second set of tetravalent peptide libraries containing Arg and/or Asn fixed in these positions were constructed and screened to further refine peptide selection (Fig. 1B
). Because these "locked-in" amino acid function as anchors, the 7 amino acid span (
25 Å) is sufficient to fully embrace the receptor binding surface of a single B-subunit, whose diameter is 26 Å. In the case of all three "secondary" libraries, Arg was strongly selected at positions 6 and 7, whereas a Pro cluster was selected at amino-terminal positions.
Peptides with the amino acid sequences obtained from the second library screening as well as a consensus peptide, MAFRRNRRNA (FRR-mono), were synthesized for kinetic analysis. All of these peptides, but not the free trisaccharide analog, effectively bound to 2BH, clearly indicating that these peptides are superior Stx binding units compared with the natural binding unit (Fig. 1C
). Tetravalent forms of these peptides with the same polylysine core, referred to as PPR-tet, PPP-tet, KRR-tet, and FRR-tet, bound to the B-subunit with KD values similar to or lower than that of SUPER TWIG (1)6 (1.3 µM of trisaccharide) (Fig. 1D
). Furthermore, these KD values were much lower than that of Gb3 (KD=12 µM) present in phospholipid vesicles (molar ratio of phosphatidylcholine: Gb3=24:1) under the same conditions (19)
. In contrast, (MA-AU)4-3Lys (MA-tet), which has the same core structure but lacks any Stx binding motifs, did not bind to the B-subunit.
Tetravalent peptides bind to the B-subunit of Stx2 in site 3-oriented configurations
The binding site specificity of these tetravalent peptides was examined by ELISA. All of the tetravalent peptides effectively bound to 2BH whereas the maximum binding to 2BH-W33A was reduced to 1/4th-1/5th of the binding to wild-type (WT) (Fig. 2
A). In addition, it was shown that the site 3-specific binding by PPP-tet, the peptide exhibiting the highest binding affinity, was inhibited by 70% by the presence of an excess amount of SUPER TWIG (1)6 (Fig. 2B
), indicating that this tetravalent peptide is competitive with a cluster of trisaccharides at this site.
|
Structural analysis of the binding between the B-subunit and the PPP-mono peptide was performed using the free energy minimization method (22)
(Supplemental Fig. 1). In the consensus interaction model with the lowest free energy, both Arg4 and Arg5 of the peptide electrostatically interacted with Asp17, another crucial amino acid found in site 3 (8
, 19)
, whereas Arg6 and Arg7 interacted with Asp16 and Glu64, respectively (Fig. 2C
). Substitution of each Arg of PPP-tet with Ala resulted in a 3- to 5-fold increase in the KD value (Fig. 2D
). Moreover, substitution of Asp16, Asp17, or Glu64 with Ala reduced the binding of PPP-tet by 50, 70, and 71%, respectively, (Fig. 2E
), confirming the binding model. Enhancement of binding by substitution of Asp17 by Glu provided further support for the electrostatic interaction. Substitution of all of the Pro residues of PPP-tet with Ala was associated with a 3-fold increase in the KD value, whereas a single substitution of each Pro had no effect on the KD value (Fig. 2D
). Given the essential role of Trp33 in the binding of PPP-tet, it is likely that the Pro cluster interacts with Trp33 through hydrophobic interactions. In summary, our results indicate that Asp17 and Trp33, both of which are present in site 3, and an adjacent cluster of acidic amino acids (including Asp16 and Glu64) cooperatively contribute to the high-affinity binding of the peptide.
Tetravalent peptides inhibit Stx2 cytotoxicity by inducing aberrant cellular transport of Stx2
All the tetravalent peptides, but not MA-tet, efficiently protected Vero cells against the cytotoxic effects of Stx2 (Fig. 3
A). However, they did not inhibit binding of 125I-Stx2 to the cells, but SUPER TWIG (1)6 did (11)
(Fig. 3B
). These results indicate that their cytoprotective effects are not associated with binding inhibition. To elucidate the mechanism by which the tetravalent peptides neutralize Stx2, we examined the effects of PPP-tet on the intracellular transport of Stx2. Time-dependent localization of Alexa-Stx2 to the Golgi, confirmed by colocalization with Golgi markers (Supplemental Fig. 2A, B), merged well with that of Alexa-PPP-tet (Fig. 4
A, Supplemental Fig. 2B). In contrast, Alexa-PPP-tet was diffusely distributed in the absence of Alexa-Stx2. These results indicate that the PPP-tet forms a complex with Stx2, which is then incorporated into the cells and transferred to the Golgi. They also indicate that PPP-tet itself is cell permeable. Colocalization of Stx2 with Hsp47, an ER marker, was completely inhibited by the presence of PPP-tet, indicating that the transport of Stx2 from the Golgi to the ER was blocked by PPP-tet (Fig. 4B
, left panel). Degradation of 125I-Stx2 as measured by the production of small degradation products released into the culture medium was enhanced 3-fold in the presence of PPP-tet (Fig. 4B
, right panel). This degradation was blocked by the addition of chloroquine, a lysosome inhibitor. Together, these results indicate that PPP-tet inhibits the cytotoxic activity of Stx2 by inducing its aberrant cellular transport from the Golgi to an acidic compartment, rather than the ER, resulting in its degradation.
|
|
To further investigate the mechanism by which the tetravalent peptides neutralize Stx2, we examined the effect of PPP-tet on the intracellular transport of a double mutant of 2BH, 2BH-W29A/G61A. This mutant has amino acid substitutions in site 1 (Trp29) and site 2 (Gly61) and is unable to bind to target cells (19)
. PPP-tet formed a complex with 2BH-W29A/G61A via site 3 (Supplemental Fig. 3), the only residual trisaccharide binding site, and this complex was substantially incorporated into Vero cells (Fig. 4C
, left panel). However, this mutant did not bind to cells in the absence of PPP-tet. Of note, the PPP-tet/2BH-W29A/G61A complex that had been incorporated into cells was faintly detected and the complex was not transferred to the Golgi, whereas the PPP-tet/2BH complex was efficiently incorporated and transferred to the Golgi (Fig. 4C
, left and right panels). These results demonstrate that each Gb3 binding site on the B-subunit has a specific role. That is, binding of Gb3 through sites 1 and 2 is required for the efficient uptake of Stx2 and its subsequent retrograde transport to the Golgi, whereas binding of Gb3 through site 3, which is specifically blocked by PPP-tet, plays an essential role in its transfer from the Golgi to the ER.
PPP-tet inhibited the cytotoxicity of Stx2 even when added up to 3 h after the incorporation of Stx2 (Fig. 5
A). Such an inhibitory activity was not observed in SUPER TWIG (1)6, which functions by inhibiting the binding of Stx2 to target cells. Up to 3 h after the incorporation, Stx2 was localized mainly in the Golgi rather than in the ER (Fig. 5B
). After 6 h, however, obvious localization of Stx2 in the ER was observed. This localization to the ER was also effectively blocked by PPP-tet even when added 3 h after Stx2 incorporation (Fig. 5B
). These results indicate that due to its cell-permeable nature, PPP-tet is able to form a complex with Stx2 present in the Golgi, and it abrogates cytotoxicity by blocking Stx2 transport from the Golgi to the ER.
|
The tetravalent peptides protect against STEC-induced lethality
The inhibitory effects of the tetravalent peptides on the lethality of E. coli O157:H7 infections in protein calorie malnutrition mice, which are very susceptible to infection (23)
, were examined. Each peptide was intragastrically administered beginning on day 2 of infection, when infection can be diagnosed by detection of Stx in the stool. Both PPR-tet and PPP-tet (180 nmol/g) protected mice from lethality with remarkable potency (P<0.0001), and a lower dose of PPP-tet (53 nmol/g) significantly extended the average survival period by > twice that of control mice (17.8±0.44 d vs. 8.3±0.32 days, P<0.001, Fig. 6
A, left panel). To prevent proteolytic degradation in the gastrointestinal tract, an acetylated form of PPP-tet (Ac-PPP-tet) was synthesized. Ac-PPP-tet, which had a KD value equivalent to that of PPP-tet (0.53 µM of binding units) but inhibited the Stx2-cytotoxicity with >40-fold higher efficiency (IC50=1 µM of binding units) compared with PPP-tet, completely inhibited the lethality at amounts as low as 18 nmol/g (P<0.0001, Fig. 6A
, right panel). Treatment with Ac-PPP-tet (36 nmol/g) beginning on day 3 of infection, when the serum concentration of the inflammatory cytokine TNF reaches peak levels (24)
, still protected mice from lethality (P<0.005, Fig. 6B
).
|
DISCUSSION
Using a multivalent peptide library approach, we identified four tetravalent peptides that exhibit high affinities for the Stx2 B-subunit pentamer and markedly inhibit Stx2 cytotoxicity. Each tetravalent peptide has a novel Stx binding motif that is superior to the natural Stx binding unit, the trisaccharide. Since the B-subunit pentamer is known to bind to the trisaccharide assembled into multimers with a 106-fold greater efficacy compared with its free form (9)
, it is conceivable that the multivalent nature of the library enabled the efficient identification of such high-affinity motifs. Additional optimization of the structure of the library also contributed to the efficient screening. For example, the introduction of "locked-in" amino acids, identified using the first library screening, into the second library led to a 10- to 100-fold greater enhancement in binding (data not shown). Using the identified Stx binding motifs, we also synthesized octavalent peptides with a different core structure consisting of seven lysines. However, the octavalent peptides scarcely inhibited Stx2 cytotoxicity (K. Nishikaw et al., unpublished data), underscoring the importance of optimizing the structure to which Stx binding units are assembled.
Based on structural and mutational analysis of the binding between the Stx2 B-subunit and PPP-tet, the tetravalent peptide that exhibited the highest binding affinity, it was shown that Asp17 and Trp33 of the B-subunit, both present in site 3, and an adjacent cluster of acidic amino acids, including Asp16 and Glu64, cooperatively contribute to the high-affinity binding of the peptide. Although Glu64 is not involved in trisaccharide binding at any site, Asp16 is known to be an essential amino acid functioning in site 1 (8
, 19)
. However, after forming a complex with PPP-tet through site 3, Stx2 still binds to the cell surface receptor mainly through site1 (see below), suggesting that the interaction of Asp16 with PPP-tet does not play a significant role in the interaction with PPP-tet on target cells. The finding that Ala substitution of Asp16 had less effect on PPP-tet binding than substitution of other amino acids (i.e., Trp33, Asp17, or Glu64) provides further support for a minor contribution of Asp16 to Stx2 and PPP-tet interactions.
Previously developed Stx inhibitors function by inhibiting the binding of Stx to target cells (9
, 11
12
13)
. In stark contrast, we found that the Stx2/PPP-tet complex was effectively incorporated into cells. Also, PPP-tet itself was shown to be cell-permeable. PPP-tet contains an Arg cluster in its Stx2 binding motif, a common feature observed in other Stx binding units identified in this study. Because some peptide motifs with a cluster of basic amino acids have been shown to be cell permeable (25
, 26)
, this Arg cluster may impart the ability of PPP-tet not only to permeate cells by itself, but also to permit the complex with Stx2 to effectively associate with cells.
After association with cells, the Stx2/PPP-tet complex was transported to the Golgi and then to an acidic compartment, rather than to the ER, resulting in effective degradation of Stx2. Thus, PPP-tet inhibits Stx2 by inducing its aberrant cellular transport after binding through site 3. So far there has been no information regarding the specific role of the individual trisaccharide binding site in the intracellular transport of Stx2. This is due to the fact that mutation of each site results in marked reduction of Stx2 binding to target cells (19)
. We circumvented this problem by taking advantage of the ability of PPP-tet to permit mutated 2BH to be incorporated into cells. By investigating the intracellular transport of complexes containing 2BH or 2BH mutated at site 1 and 2 (2BH-W29A/G61A), we found that binding of Gb3 through site 1 and 2 is required for efficient uptake of Stx2 and its subsequent retrograde transport to the Golgi. We also found that binding of Gb3 through site 3, specifically blocked by PPP-tet, plays an essential role in its transfer from the Golgi to the ER.
We recently found that site 2 of Stx2 may be less functional as a trisaccharide binding site, as mutation of this site did not affect binding to Gb3, whereas mutation of either site 1 or site 3 resulted in a marked reduction in binding (19)
. The crystal structure of Stx2 indicates that the presence of Ser-54 in site 2 of the B-subunit would hamper the binding of the trisaccharide in the same way that it binds to site 2 of Stx1, another Stx family member (8
, 27)
, providing further support for the small contribution of site 2. Thus, it is likely that the Stx2/PPP-tet complex associates with the cell surface Gb3 mainly through site 1.
Based on these observations, we propose the following model by which PPP-tet exerts inhibitory activity against Stx2. PPP-tet complexes with Stx2 through site 3, and the complex associates with target cells due to the cell-permeable nature of PPP-tet. On the cell surface, Stx2 and Gb3 form a complex via interactions mainly at site 1, but also at site 2. The complex is then incorporated and transported to the Golgi in a retrograde manner. However, because of the lack of a functional interaction between Stx2 and Gb3 through site 3, the complex in the Golgi is transported to an acidic compartment rather than to the ER, resulting in degradation of Stx2. Although the precise molecular mechanism by which each trisaccharide binding site exerts its specific function remains to be elucidated, our present results indicate that transport of Stx2 from the plasma membrane to the Golgi and from the Golgi to the ER are independently regulated. Our finding that PPP-tet can inhibit the transport to the ER of Stx2 that has already localized in the Golgi lends further support for the existence of machinery that controls Stx2 transport through a specific interaction between Stx2 and Gb3 at site 3.
Finally, PPP-tet was found to protect mice from a fatal dose of Stx-producing E. coli even when administered after an established infection. In general, peptide-based compounds have a great advantage over carbohydrate-containing ones in that various chemical modifications can be easily introduced without affecting their biological activities. In the present case, an acetylated form of PPP-tet, Ac-PPP-tet, which is resistant to proteolysis, markedly enhanced inhibitory activity in the STEC infection experiment. The lowest effective amount of Ac-PPP-tet used (5.3 nmol of binding units/g) can be extrapolated to be 0.53 g/70 kg of body weight for human therapy, indicating that this compound holds promise as a practical therapeutic agent for STEC infections.
Thus, use of the multivalent peptide library approach enabled identification of a peptide-based Stx2 inhibitor that has remarkable therapeutic potency and appears to function by inducing aberrant cellular transport and degradation of Stx2. The multivalent peptide library technique is widely applicable to the development of selective inhibitors for a variety of disease-related molecules, such as influenza hemagglutinins and selectins, which recognize their intended carbohydrates based on the clustering effect.
ACKNOWLEDGMENTS
We thank Michael Berne (Tufts University) for peptide synthesis and sequencing and Drs. Lewis C. Cantley (Beth Israel Deaconess Medical Center and Harvard Medical School) and Takehiko Sasazuki (International Medical Center of Japan) for their continuous encouragement and fruitful discussion. This work was supported by the PRESTO program of JST.
Received for publication June 5, 2006. Accepted for publication July 11, 2006.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  L. Winkler, C. Gehring, A. Wenzel, S. L. Muller, C. Piehl, G. Krause, I. E. Blasig, and J. Piontek Molecular Determinants of the Interaction between Clostridium perfringens Enterotoxin Fragments and Claudin-3 J. Biol. Chem., July 10, 2009; 284(28): 18863 - 18872. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Kano, S. Yamauchi, Y. Yoshida, M. Watanabe-Takahashi, K. Nishikawa, N. Nakamura, and M. Murata Yip1A regulates the COPI-independent retrograde transport from the Golgi complex to the ER J. Cell Sci., July 1, 2009; 122(13): 2218 - 2227. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Vareille, T. de Sablet, T. Hindre, C. Martin, and A. P. Gobert Nitric oxide inhibits Shiga-toxin synthesis by enterohemorrhagic Escherichia coli PNAS, June 12, 2007; 104(24): 10199 - 10204. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
0.13 million). B) Screening of the library was performed to identify compounds that bound to 2BH but not to 2BH-W33A. Tetravalent peptide libraries with fixed Arg and/or Asn were used for the second round of selections. The predicted degeneracy of a randomized peptide with 6 degenerate positions is 196 (
