The main cause of acute renal failure in children is HUS (haemolytic uraemic syndrome), a consequence of intestinal infections with Escherichia coli strains producing Stx (Shiga toxins). Stx released in the gut by the non-invasive bacteria reach the bloodstream and are targeted to cerebral and renal endothelium triggering HUS. PMN (polymorphonuclear leucocytes) seem to be involved in Stx delivery through an unidentified membrane receptor (Kd=10−8 M; 2×105 binding sites) which does not allow internalization. Some experts in the field have defined the Stx–PMN interaction as non-specific and of little biological significance. In the present study, we show that the A chain of ricin, the well-known plant RIP (ribosome-inactivating protein), interacts with PMN (Kd=10−9 M; 2×105 binding sites) competing for the same receptor that recognizes Stx, whereas diphtheria toxin and several agonists of TLRs (Toll-like receptors) or the mannose receptor were ineffective. No toxic effects of ricin A chain on PMN were observed, as assessed by measuring protein synthesis and the rate of spontaneous apoptosis of leucocytes. Moreover, two single-chain RIPs (gelonin and saporin S6) had the same competing effect. Thus RIPs and Stx1 share structural similarities, the same enzymatic activity and a common receptor on PMN. These observations reveal that the Stx–PMN interaction is specific, confirming that PMN recognize molecular patterns common to different foreign molecules.

INTRODUCTION

Stx (Shiga toxins) are produced by pathogenic STEC (Stx-producing Escherichia coli) strains which are responsible for severe human illnesses, such as haemorrhagic colitis and HUS (haemolytic uraemic syndrome) [1]. HUS is the main cause of acute renal failure in children [2,3] and is the consequence of intestinal infection by STEC.

STEC produce two main types of bipartite toxins, Stx1 and Stx2, that are capable of binding to glycolipid receptors, mainly Gb3 (globotriaosylceramide), present on the surface of target cells through their B subunits [4]. After endocytosis, the A subunit damages ribosomes, by removing a specific adenine from 28S rRNA [5], and DNA, by releasing multiple adenines [6,7]. The consequences are the arrest of protein synthesis and the formation of apurinic sites in the nucleus. Target cells showed a broad spectrum of responses, including the production of pro-inflammatory cytokines involved in HUS pathogenesis [8,9] and triggering of the apoptotic programme [10]. There is no doubt that the major part of the histopathological lesions observed in HUS are caused by the interaction of these toxins with the endothelial lining of the brain and kidney [4]. However, as STEC are non-invasive bacteria confined to the gut, a great amount of effort has been made to clarify the mechanism of transfer of Stx from the intestinal lumen to the endothelia of target organs. Stx have never been found in the plasma of patients [11,12], although they were detected on the surface of PMN (polymorphonuclear leucocytes) from patients with HUS [1315]. Purified Stx interact with human PMN through an unknown receptor [1618], which allows the binding but not internalization of the toxins [16]. Stx had lower affinity for the leucocyte receptor compared with the Gb3 receptor on target endothelial cells [16], and this might permit the transfer from circulating cells to endothelia. It has been demonstrated that Stx bound to PMN were transferred to Gb3-containing cells during co-incubation in vitro [16,18], impairing protein synthesis of target cells [16]. Although direct evidence in patients or animal models is lacking, this mechanism might represent what happens in the kidneys of patients with HUS. However, the interaction between Stx and PMN has been the object of controversy, with conflicting results being presented. Some authors have claimed a lack of specific binding of Stx to PMN [19,20] or reinterpreted their own previous results as artefacts [21].

In this light, a better characterization of the receptor responsible for the recognition of Stx by PMN would be a major goal in understanding a crucial step in the pathogenesis of HUS. The partially characterized molecule is a membrane receptor of human PMN [14,16,17], expressed on mature neutrophils or granulocyte-differentiated HL-60 cells, but not on immature PMN or undifferentiated HL-60 cells [17], and capable of binding isolated Stx in the absence of added cofactors [14,17], but not able to internalize the toxic ligand [16]. The binding of Stx1 to PMN is highly sensitive to trypsin treatment [16].

PMN, as well as macrophages, are capable of recognizing PAMPs (pathogen-associated molecular patterns) by means of membrane or cytoplasmic PRR (pattern-recognition receptors) that belong to the innate host immune system [2224], thus allowing a discrimination between pathogens and self. This prompted us to investigate whether the unknown Stx receptor of PMN is involved in the recognition of molecular signatures shared by other bacterial toxins (diphtheria toxin) or by toxins homologous with Stx.

The molecular mechanism of action of Stx described above is the same as that of the large family of RIPs (ribosome-inactivating proteins) from plants: the removal of a specific adenine residue from 28S rRNA in ribosomes and multiple adenine residues from DNA (for reviews, see [25,26]). Most plant RIPs, such as gelonin (from Gelonium multiflorum seeds) and saporin S6 (from Saponaria officinalis seeds), consist only of a single chain (A chain) containing the catalytic site. Ricin, the well-known two-chain RIP from Ricinus communis seeds, is highly toxic as it contains, in addition to the enzymatically active A chain, a second chain (B chain) that binds to galactose-containing surface molecules mediating endocytosis of the RIP. Stx and ricin differ largely in their B chain structure and binding specificity, whereas their A chains have a high degree of homology [27,28]. In addition to the canonical route of internalization through the lectinic B chain, ricin can enter cells through a secondary route: the mannose receptor present on macrophages or Kupffer cells [29,30]. This receptor binds to high-mannose oligosaccharide chains present on both the A and B chains of the toxin, and this leads to the internalization of the ligand. The mannose receptor is also present on PMN as a non-canonical PRR able to bind exogenous and endogenous glycosylated molecules [31].

In the present study, we provide evidence that the A chain of ricin also binds to PMN through the same receptor responsible for Stx recognition. The mannose receptor and TLR (Toll-like receptor) 1, TLR2, TLR5 and TLR6 are not involved in this unconventional role of neutrophils.

EXPERIMENTAL

Materials

Pam2CSK4 [dipalmitoylcysteinylseryl-(lysyl)4] and Pam3CSK4 [tripalmitoylcysteinylseryl-(lysyl)4] were purchased from InvivoGen; peptidoglycan from Staphylococcus aureus, lipoteichoic acid, ovalbumin and mannosylated BSA (23 mol of saccharide/mol of BSA) were obtained from Sigma–Aldrich; and flagellin from Salmonella enterica serotype Typhimurium was purchased from Alexis.

Toxin purification

The prototype Stx1 producer E. coli C600 (H19J) was kindly supplied by Dr Alison O'Brien (Department of Microbiology and Immunology, USU of the Health Sciences, Bethesda, MD, U.S.A.). Stx1 was purified by receptor analogue affinity chromatography [32] on globotriose–Fractogel (IsoSep AB). Ricin was generously provided by Professor Andrea Bolognesi (Dipartimento di Patologia Sperimentale, Università di Bologna, Bologna, Italy). Separation of the two ricin chains by two chromatographic runs on blue-Sepharose and blue dextran–Sepharose was performed as described previously [33]. The purity of the isolated chains was verified by SDS/PAGE. Isolated ricin A chain was stored at 4 °C in 10 mM Tris/HCl (pH 7.5), 100 mM NaCl and 0.1% 2-mercaptoethanol. Under these conditions, after 3 months, almost all of the ricin A chain was in the monomer form (96–99%) as assessed by non-reducing SDS/PAGE, followed by densitometric analysis of the Coomassie-Blue-labelled bands. The experiments described in the present study (see below) have been performed with different ricin A preparations within 3 months. The IC50 value of ricin A chain on the unfractionated rabbit reticulocyte lysate system (0.1–0.2 nM) was similar to values reported previously [25,34]. The IC50 value was calculated by the least-squares method applied to the linear regression between the fractional activity and log-transformed inhibitor concentration. Gelonin from G. multiflorum was from Sigma–Aldrich. Saporin S6 was generously provided by Professor Fiorenzo Stirpe (Dipartimento di Patologia Sperimentale, Università di Bologna, Bologna, Italy). Diphtheria toxin was prepared using a method described previously [35].

Binding of Stx1 to PMN

To obtain endotoxin-free PMN (98% lobulated nuclei) from healthy donors, all of the solutions used thoroughout the method [36] were sterile and prepared with medical-grade water. The binding experiments were performed in Eppendorf tubes pre-coated with PBS containing 1% (w/v) BSA to avoid a non-specific loss of toxins [37]. Freshly isolated endotoxin-free PMN (0.5×106) were immediately incubated with native or 125I-Stx1 (6–60 nM) in 250 μl of PBS containing 1% (w/v) BSA in the absence or presence of competing molecules for 90 min at 37 °C, with occasional stirring by gentle inversion of the tube. The cells were centrifuged at 200 g for 5 min and washed three times with 100 μl of the same buffer containing BSA at 37 °C. The extent of binding of native or radiolabelled Stx1 to PMN was measured as described below.

Indirect flow cytometric analysis of Stx1 bound to PMN

Stx1 bound to PMN was detected by flow cytometry as described previously [13,14]. Briefly, PMN were incubated with an appropriately diluted mouse monoclonal anti-Stx1 antibody in the presence of human serum to saturate Fc receptors. After incubation with FITC-conjugated goat anti-(mouse IgG), flow cytometric analysis was used to reveal the cell-bound fluorescence. The flow cytometer (FC500; Beckman Coulter) was set to acquire and gate events both by forward scatter compared with 90 ° side scatter and by green fluorescence compared with 90 ° side scatter. PMN were checked by staining with monoclonal antibodies to antigens associated to granulocytes (FITC–CD16 and FITC–CD65; Beckman Coulter). This set resulted in a prompt analysis of both morphology and fluorescence, allowing a clear evaluation of control and positive samples. The MCV (mean channel value of fluorescence) of the cells was chosen as an objective parameter to measure the extent of binding of Stx1 to cells [14]. The single values were calculated by subtracting the control MCV (range, 0.4–0.6), i.e. the MCV of cells incubated with primary and secondary antibodies in the absence of the toxin. The same values (MCV=0.4–0.6) were obtained if the mouse monoclonal anti-Stx1 antibodies were omitted from the assay in the presence of toxins and secondary antibodies. The assay has been validated previously by challenging Stx-positive PMN with a negative control antibody [13] and by comparing control subjects and patients with HUS in a double-blind fashion [14].

Radiolabelling of Stx1 and binding to PMN

Stx1 was radiolabelled with IODO-BEADS (Pierce Biotechnology) as described previously [17]. The efficiency of labelling was checked by precipitating an aliquot of 125I-Stx1 with 6% (v/v) TCA (trichloroacetic acid) in the presence of 0.01% sodium deoxycholate. The biological activity of the radiolabelled toxin (specific radioactivity, approx. 100000 c.p.m./pmol) was assessed by testing the binding to immobilized Gb3 (globotriose–Fractogel) and by measuring the IC50 of radiolabelled and native toxins on human endothelial cell protein synthesis [6]. The binding of 125I-Stx1 to PMN was quantified by counting the cell-associated radioactivity with a γ-counter after extensive washing (see above). The binding of radiolabelled Stx1 in the presence of a 50-fold excess of native toxin (non-specific binding) was subtracted in each experiment.

Fluorescent labelling of ricin A chain

The storage buffer of ricin A chain was removed by centrifugation on Microcon YM-3 (Millipore), followed by repeated washes with PBS. Then, ricin A chain (100 μg) was conjugated with Alexa Fluor® reactive fluorescent dye (Molecular Probes) and purified on a spin column, according to the manufacturer's instructions. The degree of labelling was approx. 1 mol of Alexa Fluor® 488 dye/mol of subunit.

Treatment of ricin A chain with NEM (N-ethylmaleimide)

Ricin A chain was adjusted to 30 μM in PBS (pH 7.2) by centrifugation on a Microcon YM-3 and repeated washes with the above buffer. Then, a 30-fold molar excess of freshly prepared NEM (Pierce Biotechnology) was added and the reaction mixture was incubated for 2 h at room temperature (24 °C) [38]. The reacted ricin A chain was purified using a spin column as described above. The degree of free sulfhydryl groups, assessed by Ellman's reagent according to the manufacturer's instructions (Pierce Biotechnology) after 5 min of incubation, was 0.86 or 0.16 mol/mol of untreated or NEM-treated ricin A chain respectively.

Direct flow cytometric analysis of the binding of fluorescent ricin A chain to PMN

PMN (0.5×106) were treated with increasing concentrations of Alexa Fluor® 488-labelled ricin A chain (0.5–20 nM) in 250 μl of PBS containing 1% (w/v) BSA and incubated and washed as described above. The extent of binding of fluorescent ricin A chain to PMN was measured by direct flow cytometric analysis and was expressed as the MCV. The binding of fluorescent ricin A chain in the presence of an excess of unlabelled ricin A (non-specific binding) was subtracted in each experiment.

Scatchard plot and non-linear regression analysis of the PMN–ricin A chain interaction

Different iodination procedures of ricin A chain were performed with IODO-BEADS using various protocols suggested by the manufacturer to limit the oxidation of the protein. In each case, we observed a loss of PMN-binding activity after iodination. Thus Alexa Fluor® 488-labelled ricin A chain was used in this set of experiments. Different concentrations of fluorescent ricin A chain (0.5–10 nM) were incubated for 90 min at 37 °C in 1.5 ml of PBS containing 1% (w/v) BSA in Eppendorf tubes coated with BSA (see above) in the absence and presence of PMN (4.75×106). After centrifugation of the samples for 5 min at 200 g, the fluorescence values of the supernatants were measured using a Kontron spectrophotofluorimeter. Excitation and emission wavelengths were set at 494 and 519 nm respectively. The single values of the two different series (with or without PMN) were calculated after subtracting the blank values obtained with the supernatant from untreated PMN or with the buffer alone respectively. The amounts of fluorescent toxin bound to PMN were calculated by subtracting the values of fluorescence obtained in the presence of PMN with those obtained in the absence of cells. The Kd and the number of binding sites on PMN were determined by Scatchard plot or by non-linear regression of the saturation binding data using the GraphPad Prism computer program.

Protein synthesis of PMN

PMN (3.5×106) were incubated for 90 min at 37 °C in 1.75 ml of PBS containing 1% (w/v) BSA in the absence and presence of ricin or ricin A chain, as indicated in the legend to Figure 2. At the end of the incubation, cells were washed three times with 1 ml of PBS containing 1% (w/v) BSA and protein synthesis was assessed immediately or 15 h after incubation in RPMI 1640 medium containing 10% (w/v) FBS (fetal bovine serum). PMN translation was measured as the rate of incorporation of radiolabelled leucine during a 1 h incubation of the cells at 37 °C in 1 ml of RPMI 1640 medium containing 10% (w/v) FBS and 5 μCi of [3H]leucine (62 Ci/mmol; Amersham Biosciences). PMN were centrifuged as described above and washed three times with 0.5 ml of ice-cold PBS containing 10 mM leucine. The cellular pellets were treated with 0.5 ml of ice-cold 10% (v/v) TCA, followed by 5 min of incubation on ice and centrifugation for 10 min at 13000 g. The procedure was repeated three times. Finally, the precipitated cellular proteins were resuspended in 200 μl of 0.2 M KOH, and the radioactivity was measured in a liquid-scintillation counter. In each experiment, blank values obtained by incubating PMN in the presence of radiolabelled leucine (see above) for 1 h at 0 °C were subtracted.

Determination of PMN apoptosis

Apoptosis was evaluated (i) by assessing the annexin V binding to phosphatidylserine exposed on the outer leaflet of the plasma membrane of apoptotic cells and by evaluating, simultaneously, the exclusion of PI (propidium iodide) (ApoScreen Annexin V apoptosis kit-FITC; Beckman Coulter) by means of flow cytometric analysis; and (ii) by the observation of changes in nuclear shape of PMN (from lobulated to shrunken, ovoid and pyknotic). Nuclear shape of PMN incubated as described in the legend to Figure 2(B) was determined from DAPI (4′,6-diamidino-2-phenylindole) staining. After washing in PBS and distilled water, the coverslips were air-dried and mounted in a solution containing 0.2 mg/ml DAPI (Sigma–Aldrich) in water, diluted 1:500 in DABCO (1,4-diazadicyclo[2.2.2]octane) (Sigma–Aldrich) and examined with a Nikon epifluorescence microscope, equipped with a 100 kV lamp and filters for FITC and DAPI.

Statistics

A Student's t test was used to compare two groups. A value of P<0.05 was considered statistically significant.

RESULTS

Effect of bacterial or plant toxins on the binding of Stx1 to PMN

According to the hypothesis that Stx receptors on PMN might recognize molecular signatures common to other bacterial toxins or to structurally related toxins of non-bacterial origin, we measured the binding of Stx1 to PMN by indirect flow cytometric analysis in the presence of diphtheria toxin and the plant RIP ricin. The experimental conditions were chosen to have 50% saturation of Stx receptors on PMN [14,17]. Ricin in different forms (whole toxin, reduced toxin, and isolated A or B chains) strongly interfered with the recognition of Stx1 by PMN (Table 1), whereas binding was unaffected in the presence of a large molar excess of diphtheria toxin (Table 1). NEM-treated ricin A chain was equally effective, indicating that the single sulfhydryl group exposed after reduction of whole ricin and neutralized by NEM [38] is not required for activity. The results obtained with the B chain of ricin, which recognizes many glycolipids and glycoproteins on the cell surface, were not surprising and suggest that the Stx receptors on PMN are glycosylated. More interesting is the competing effect displayed by the isolated ricin A chain, which showed structural homology with the A chains of Stx. To characterize better the competing effects and to exclude the presence of ricin A chain–Stx interactions preventing the binding of the bacterial toxin to PMN, the latter cells were pre-incubated with ricin A chain followed by extensive washing. This treatment completely inhibited the binding of subsequently added Stx1 (Figure 1), suggesting the presence of a common binding site for the two toxins on PMN.

Flow cytometric analysis illustrating the effect of ricin A chain on the binding of Stx1 to PMN

Figure 1
Flow cytometric analysis illustrating the effect of ricin A chain on the binding of Stx1 to PMN

(A) Control PMN, (B) PMN treated with 6 nM Stx1 (90 min at 37 °C), (C) PMN treated with 6 nM Stx1 in the presence of 20 nM ricin A chain (90 min at 37 °C), and (D) PMN pre-incubated with 2 nM ricin A chain (90 min at 37 °C), and post-incubated after extensive washing with 6 nM Stx1 (90 min at 37 °C).

Figure 1
Flow cytometric analysis illustrating the effect of ricin A chain on the binding of Stx1 to PMN

(A) Control PMN, (B) PMN treated with 6 nM Stx1 (90 min at 37 °C), (C) PMN treated with 6 nM Stx1 in the presence of 20 nM ricin A chain (90 min at 37 °C), and (D) PMN pre-incubated with 2 nM ricin A chain (90 min at 37 °C), and post-incubated after extensive washing with 6 nM Stx1 (90 min at 37 °C).

Table 1
Binding of Stx1 to PMN in the presence of putative competing molecules

PMN were treated with 6 nM Stx1 with or without the indicated additions, and the binding of the toxin to cells was measured by indirect flow cytometric analysis as described in the Experimental section. Under this condition, approx. 50% saturation of receptors was obtained (MCV=1.54±0.23, mean±S.D., n=3). *P<0.001 compared with control (no addition).

Treatment PMN-binding activity (%) 
No addition 100 
Toxins  
 Diphtheria toxin (0.5 μM) 81.1±20.4 
 Ricin (0.5 μM) 5.4±7.6* 
 Ricin (0.5 μM) reduced by 4% (v/v) 2-mercaptoethanol 8.7±7.2* 
 Ricin B chain (0.02 μM) 12.7±9.0* 
 Ricin A chain (0.02 μM) 3.9±3.4* 
 NEM-treated ricin A chain (0.02 μM) 3.7±1.0* 
Mannose receptor agonists  
 Mannosylated BSA (1.4 μM) 94.0±6.4 
 Ovalbumin (0.5 μM) 85.1±12.6 
TLR agonists  
 Pam2CSK4 (1 μM) (TLR2 and 6) 87.5±17.7 
 Pam3CSK4 (1 μM) (TLR2 and 1) 92.4±10.7 
 Peptidoglycan (10 μg/ml) (TLR2) 100.0±1.1 
 Lipoteichoic acid (10 μg/ml) (TLR2) 100.1±3.2 
 Flagellin (0.7 μM) (TLR5) 96.7±4.7 
Treatment PMN-binding activity (%) 
No addition 100 
Toxins  
 Diphtheria toxin (0.5 μM) 81.1±20.4 
 Ricin (0.5 μM) 5.4±7.6* 
 Ricin (0.5 μM) reduced by 4% (v/v) 2-mercaptoethanol 8.7±7.2* 
 Ricin B chain (0.02 μM) 12.7±9.0* 
 Ricin A chain (0.02 μM) 3.9±3.4* 
 NEM-treated ricin A chain (0.02 μM) 3.7±1.0* 
Mannose receptor agonists  
 Mannosylated BSA (1.4 μM) 94.0±6.4 
 Ovalbumin (0.5 μM) 85.1±12.6 
TLR agonists  
 Pam2CSK4 (1 μM) (TLR2 and 6) 87.5±17.7 
 Pam3CSK4 (1 μM) (TLR2 and 1) 92.4±10.7 
 Peptidoglycan (10 μg/ml) (TLR2) 100.0±1.1 
 Lipoteichoic acid (10 μg/ml) (TLR2) 100.1±3.2 
 Flagellin (0.7 μM) (TLR5) 96.7±4.7 

Lack of cytotoxic effects elicited by ricin A chain on PMN

The internalization of isolated ricin A chain through the mannose receptor has been described in different cells [29,30]. To exclude the presence of cytotoxic effects on PMN explaining the results obtained above, the translation rate of PMN was measured in the presence and absence of ricin A chain. Although PMN are terminally differentiated non-proliferating cells, a detectable level of [3H]leucine incorporation into proteins was recorded. As shown in Figure 2(A), PMN protein synthesis was strongly inhibited by whole ricin, whereas ricin A chain, tested at concentrations impairing the binding of Stx1 to PMN, did not significantly affect translation even 15 h after incubation.

Inhibition of protein synthesis and triggering of the apoptotic programme in PMN treated with whole ricin or ricin A chain

Figure 2
Inhibition of protein synthesis and triggering of the apoptotic programme in PMN treated with whole ricin or ricin A chain

(A) PMN were incubated for 90 min at the indicated concentrations of RIPs, extensively washed and processed to determine the rate of protein synthesis, as described in the Experimental section. Values are means±S.D. from three different experiments. The [3H]leucine incorporated by untreated cells was 483±23 and 160±34 d.p.m. after 90 min and 15 h of incubation of freshly isolated PMN respectively. *P<0.005 compared with control cells. (B) The nuclear shape of PMN under different conditions as assessed by DAPI staining. PMN (2% non-lobulated nuclei) resuspended in RPMI containing 10% (v/v) FBS were either untreated or treated with ricin A chain or whole ricin at the indicated concentrations for 10 and 24 h. After incubation, changes from lobulated to shrunken, ovoid and pyknotic nuclei (apoptotic nuclei) were determined. Values are means±S.D. from three different experiments. At zero time, control PMN had 98% lobulated nuclei. *P<0.001 compared with control cells.

Figure 2
Inhibition of protein synthesis and triggering of the apoptotic programme in PMN treated with whole ricin or ricin A chain

(A) PMN were incubated for 90 min at the indicated concentrations of RIPs, extensively washed and processed to determine the rate of protein synthesis, as described in the Experimental section. Values are means±S.D. from three different experiments. The [3H]leucine incorporated by untreated cells was 483±23 and 160±34 d.p.m. after 90 min and 15 h of incubation of freshly isolated PMN respectively. *P<0.005 compared with control cells. (B) The nuclear shape of PMN under different conditions as assessed by DAPI staining. PMN (2% non-lobulated nuclei) resuspended in RPMI containing 10% (v/v) FBS were either untreated or treated with ricin A chain or whole ricin at the indicated concentrations for 10 and 24 h. After incubation, changes from lobulated to shrunken, ovoid and pyknotic nuclei (apoptotic nuclei) were determined. Values are means±S.D. from three different experiments. At zero time, control PMN had 98% lobulated nuclei. *P<0.001 compared with control cells.

The absence of cytotoxic effects was also confirmed by comparing the rate of spontaneous apoptosis of control PMN and of PMN treated with ricin A chain. Mature PMN in the blood are highly specialized non-dividing cells with a short (1 or 2 days) lifespan [39] as they are constitutively engaged in the apoptotic programme [40]. PMN undergoing apoptosis lose the multi-lobulated shape of the nucleus which become ovoid, shrunken and pyknotic, and this is considered one of the morphological characteristics of apoptotic PMN [40]. Figure 2(B) shows that the appearance of apoptotic nuclei after 10 or 24 h of incubation of freshly isolated PMN was similar to that observed in ricin A chain-treated cells, whereas in the presence of ricin the whole PMN population underwent apoptosis. PMN were also analysed for binding to annexin V and simultaneously with a PI dye-exclusion test after 10 h of incubation in the absence and presence of ricin A chain. The annexin V assay is based on the binding of FITC-conjugated annexin V to phosphatidylserine exposed on the outer leaflet of the plasma-membrane lipid bilayer of apoptotic cells, followed by flow cytometric detection. Annexin V-positive cells were classified as early apoptotic cells (Figure 3, quadrant B4), and annexin V- and PI-double-positive cells were considered to be late-apoptotic cells (Figure 3, quadrant B2). The results depicted in the representative cytograms in Figure 3 show that the number of untreated live PMN [(0.25±0.02)×106] that are not yet initiated in the spontaneous apoptotic programme (Figure 3A, quadrant B3) did not change significantly in the presence of 2 nM [(0.23±0.02)×106] or 20 nM [(0.22±0.01)×106] ricin A chain (Figure 3D and C respectively). In contrast, whole ricin induced a strong apoptotic signal, as nearly all of the ricin-treated PMN were positive for annexin V (Figure 3B), whereas the number of live PMN [(0.01±0.01)×106; P<0.005] were reduced dramatically. Necrotic cells (cells positive for PI alone; quadrant B1) or late apoptotic cells were almost completely absent under these conditions.

Effect of whole ricin or ricin A chain on the survival of PMN after 10 h of culture

Figure 3
Effect of whole ricin or ricin A chain on the survival of PMN after 10 h of culture

PMN (1×106) from healthy donors were incubated for 10 h in the absence and presence of whole ricin and ricin A chain at the indicated concentrations (see below) and analysed by flow cytometry after staining with annexin V and PI. Quadrant B1, necrosis; quadrant B2, late apoptotic cells; quadrant B3, live cells; and quadrant B4, early apoptotic cells. (A) untreated PMN; (B) PMN treated with 50 nM ricin; (C) PMN treated with 20 nM ricin A chain; and (D) PMN treated with 2 nM ricin A chain.

Figure 3
Effect of whole ricin or ricin A chain on the survival of PMN after 10 h of culture

PMN (1×106) from healthy donors were incubated for 10 h in the absence and presence of whole ricin and ricin A chain at the indicated concentrations (see below) and analysed by flow cytometry after staining with annexin V and PI. Quadrant B1, necrosis; quadrant B2, late apoptotic cells; quadrant B3, live cells; and quadrant B4, early apoptotic cells. (A) untreated PMN; (B) PMN treated with 50 nM ricin; (C) PMN treated with 20 nM ricin A chain; and (D) PMN treated with 2 nM ricin A chain.

Taken together, these findings exclude the presence of cytotoxic effects induced by ricin A chain and, as shown previously for Stx1, suggest the involvement of a receptor incapable of inducing internalization of the ligand.

Single-chain RIPs inhibit the binding of Stx1 to PMN

Table 2 shows that the single-chain RIPs gelonin and saporin S6, as well as ricin A chain, also significantly inhibited the binding of Stx1 to PMN, as assessed by indirect flow cytometric analysis. Since this is an antibody-based indirect method that has been questioned by some authors [21], the RIP capacity to displace Stx1 from PMN was confirmed with iodinated Stx1 obtaining similar results (Table 2). The experimental conditions were chosen to have full saturation of the Stx receptors on PMN [14,17]. The results indicate that some common structural features of the enzymatically active chains of plant RIPs and Stx are recognized by PMN. Ricin A had a more powerful trend in displacing Stx1 from PMN when using either a direct (iodinated Stx1) or indirect (cytofluorimetric assay) method, even though differences in the inhibitory power of the RIPs tested were not significant in either assay. However, the strong competing effect showed with ricin A chain tested at a lower concentration with respect to Stx1 (Table 2) suggests a high-affinity binding of this RIP to PMN.

Table 2
Inhibition by RIPs of the binding of Stx1 to PMN

PMN were treated with 50 nM native Stx1 or 125I-Stx1 in the absence and presence of RIPs, and the binding of Stx1 to cells was measured as described in the Experimental section by indirect flow cytometric analysis or by detecting cell-associated radioactivity respectively. Under these conditions, full saturation of receptors was obtained either with fluorescent Stx1 (MCV=3.25±0.49, mean±S.D., n=3) or radiolabelled Stx1 (0.365±0.016 pmol of 125I-Stx1 bound/106 PMN, mean±S.D., n=3). *P<0.005 compared with control (no addition).

 PMN-binding activity (%) 
Treatment Native Stx1 125I-Stx1 
No addition 100 100 
RIP (2 nM)   
 Ricin A chain 13.5±3.3* 26.6±1.4* 
 Gelonin 38.7±18.0* 35.1±18.2* 
 Saporin S6 43.0±12.0* 37.9±9.5* 
 PMN-binding activity (%) 
Treatment Native Stx1 125I-Stx1 
No addition 100 100 
RIP (2 nM)   
 Ricin A chain 13.5±3.3* 26.6±1.4* 
 Gelonin 38.7±18.0* 35.1±18.2* 
 Saporin S6 43.0±12.0* 37.9±9.5* 

Mannose receptor and other known PRR are not involved in Stx1 recognition by PMN

It should be noted that the mature form of saporin S6 is not glycosylated [41,42] and, thus, the results obtained in Table 2 with this RIP cannot be explained by engagement of the mannose receptor. Moreover, the lack of inhibition of Stx1 binding to PMN in the presence of agonists of the mannose receptor (mannosylated BSA and ovalbumin) completely rules out the involvement of this PRR in Stx recognition by PMN (Table 1). The binding of Stx1 to PMN was also unaffected by the presence of a large excess of competitor molecules (Pam2CSK4, Pam3CSK4, peptidoglycan, lipoteichoic acid and flagellin) for TLR1, TLR2, TLR5 and TLR6, which sense protein or peptide derivatives produced by pathogens (Table 1). TLR3, TLR7, TLR8 and TLR9 were excluded from these experiments as they are not expressed on the cell surface [22].

High-affinity binding of ricin A chain to PMN

As shown in Figure 4, the binding of ricin A chain to PMN was demonstrated directly by incubating cells with increasing amounts of fluorescent Alexa Fluor 488®-labelled ricin A chain. The binding was saturable (Figure 4) and inhibited specifically by 500 nM unlabelled ricin A chain. The amount of fluorescence associated with PMN was identical in experiments performed at 4 or 37 °C, indicating that ricin A chain is not actively internalized by PMN. Scatchard plot analysis fitted to a linear curve suggested the presence of a single-binding-site class. As shown in Table 3, the Kd value and the number of binding sites in the ricin A chain–PMN interaction were calculated by both linear and non-linear regression employing PMN from two different healthy donors. The Kd value (10−9 M) showed that the affinity of ricin A chain for PMN was higher than that reported previously for Stx1 (10−8 M) [16], explaining the powerful inhibition of Stx1 binding obtained with lower concentrations of added ricin A chain (Table 2). The number of ricin A chain-binding sites per cell calculated in the present study (178000±5000) was of the same order of magnitude as those calculated previously for the Stx1–PMN interaction by te Loo et al. [16] (210000 binding sites) and in our laboratory (195000 binding sites) [17].

Binding of fluorescent ricin A chain to isolated PMN

Figure 4
Binding of fluorescent ricin A chain to isolated PMN

Saturation binding experiments with increasing concentrations of Alexa Fluor® 488-labelled ricin A chain in the absence (■) or presence (○) of 500 nM native ricin A chain. Values are means±S.D. from three different experiments performed as described in the Experimental section.

Figure 4
Binding of fluorescent ricin A chain to isolated PMN

Saturation binding experiments with increasing concentrations of Alexa Fluor® 488-labelled ricin A chain in the absence (■) or presence (○) of 500 nM native ricin A chain. Values are means±S.D. from three different experiments performed as described in the Experimental section.

Table 3
Dissociation constant and number of binding sites of the ricin A chain–PMN interaction
 Kd (M) Binding sites (n
Variable Scatchard plot Non-linear regression Scatchard plot Non-linear regression 
Donor 1 0.87×10−9 1.26×10−9 173000 178000 
Donor 2 1.00×10−9 1.40×10−9 184000 177000 
 Kd (M) Binding sites (n
Variable Scatchard plot Non-linear regression Scatchard plot Non-linear regression 
Donor 1 0.87×10−9 1.26×10−9 173000 178000 
Donor 2 1.00×10−9 1.40×10−9 184000 177000 

As unlabelled ricin A chain has been found to inhibit the binding of Stx1 to PMN (Figure 1), the reciprocal experiment was performed. Taking into account the 10-fold higher affinity of ricin A chain for PMN with respect to Stx1, PMN were pre-incubated with the bacterial toxin, then the unwashed leucocytes were treated with fluorescent ricin A chain at 4 °C to minimize the detachment of Stx1 from PMN [17]. Figure 5 shows that an excess of free Stx1 (60 nM) displaced fluorescent ricin A chain (2 nM) from PMN.

Flow cytometric analysis illustrating the effect of Stx1 on the binding of fluorescent ricin A chain to PMN

Figure 5
Flow cytometric analysis illustrating the effect of Stx1 on the binding of fluorescent ricin A chain to PMN

(A) Control PMN, (B) PMN treated with 2 nM Alexa Fluor® 488-labelled ricin A chain (90 min at 4 °C), and (C) PMN pre-incubated with 60 nM Stx1 (90 min at 37 °C) and post-incubated with 2 nM Alexa Fluor® 488-labelled ricin A chain (90 min at 4 °C).

Figure 5
Flow cytometric analysis illustrating the effect of Stx1 on the binding of fluorescent ricin A chain to PMN

(A) Control PMN, (B) PMN treated with 2 nM Alexa Fluor® 488-labelled ricin A chain (90 min at 4 °C), and (C) PMN pre-incubated with 60 nM Stx1 (90 min at 37 °C) and post-incubated with 2 nM Alexa Fluor® 488-labelled ricin A chain (90 min at 4 °C).

Taken together, the results obtained are in keeping with the notion that Stx1 and ricin A chain share the same receptor on PMN.

DISCUSSION

The interaction between Stx and PMN and their role in the pathogenesis of HUS have been the object of great scrutiny and have stimulated intense debate. Most of the criticisms have been centred on the weak affinity of the binding (Kd=10−8 M) that was defined by some authors as non-specific and of little biological significance. In fact, some experts in the field have failed to appreciate the relevance of the binding of a toxin to a low-affinity receptor, rather than to the presumably equally accessible higher-affinity receptors in renal endothelium. However, to the best of our knowledge, free Stx have never been found in the plasma of patients suffering from HUS. Moreover, the low affinity displayed by PMN for Stx1 could be considered a strength in a model involving the passage from circulating cells carrying the toxin to target cells. Obviously, low affinity does not necessary mean lack of specificity.

The present study provides strong evidence supporting the concept that PMN specifically recognize Stx1 and other closely related toxic proteins, namely RIPs. We found that the A chain of ricin and two single-chain RIPs, gelonin and the non-glycosylated saporin S6, inhibit the binding of Stx1 to PMN, whereas diphtheria toxin and several peptide/protein derivatives acting as agonists of TLRs or the mannose receptor are completely ineffective. This is in keeping with the notion that there are regions of sequence homology between the A chains of ricin, Stx1 and the close relatives Stx from Shigella dysenteriae and Stx2 variants [27,28]. The A subunits of these toxins are structurally and functionally equivalent to single-chain RIPs [25,43]. Our present results are in agreement with the convincing evidence presented by Griener et al. [18], indicating that the A subunit of Stx is mainly responsible for the binding of the whole toxin to human PMN with a Gb3-independent mechanism. The implication of Gb3 in Stx1 and ricin A binding was excluded further as human granulocytes do not have the enzyme repertoire for globo-series glycosphingolipids and do not express Gb3 and Gb4 (globotetraosylceramide) [18,44].

The binding of ricin A chain to human PMN was demonstrated directly by flow cytometric analysis, as iodination of the protein, even though performed under mild conditions, led to a loss of PMN-binding activity. It should be noted that Stx1 is fully active in binding PMN after iodination (Table 2) [17] and that its A chain possesses fewer tyrosine residues available for labelling compared with ricin A chain. Denaturation of the latter protein caused by steric hindrance, as a consequence of iodination, or caused by susceptibility to oxidation of some crucial protein domains might explain the result.

The evidence collected supports the involvement of a common receptor for Stx1 and ricin A chain on PMN as: (i) pre-incubation of PMN with ricin A chain, although non-toxic for leucocytes, prevented the binding of Stx1; (ii) pre-incubation of PMN with Stx1 displaced ricin A chain from leucocytes; (iii) co-incubation of Stx1 with single-chain RIPs homologous with ricin A chain inhibited the binding of the bacterial toxin to PMN; (iv) the number of binding sites for the Stx1–PMN and ricin A chain–PMN interactions is of the same order of magnitude; and (v) ricin A chain was not internalized by leucocytes in the same way as Stx1. The possibility that the data obtained may be explained by non-specific competition of Stx1 binding by the strong ionic charges of the A chains of some RIPs was ruled out as saporin S6 and gelonin, which have fairly high isoelectric points (≥ 9.5), are as effective as ricin A chain (Tables 1 and 2), which displays a nearly neutral isoelectric point (7.3–7.6) [25,45].

Although the importance of these findings in the mechanism of poisoning by ricin has to be established and is beyond the scope of the present study, our data strongly suggest the presence of a receptor on PMN interacting with a class of enzymes which have the same mechanism of action on polynucleotides and structural homology. These observations reveal that the Stx–PMN interaction is specific, confirming the important role of these circulating cells in the diagnosis and the pathogenesis of HUS.

Abbreviations

     
  • DABCO

    1,4-diazadicyclo[2.2.2]octane

  •  
  • DAPI

    4′,6-diamidino-2-phenylindole

  •  
  • FBS

    fetal bovine serum

  •  
  • Gb3

    globotriaosylceramide

  •  
  • HUS

    haemolytic uraemic syndrome

  •  
  • MCV

    mean channel value of fluorescence

  •  
  • NEM

    N-ethylmaleimide

  •  
  • Pam2CSK4

    dipalmitoylcysteinylseryl-(lysyl)4

  •  
  • Pam3CSK4

    tripalmitoylcysteinylseryl-(lysyl)4

  •  
  • PI

    propidium iodide

  •  
  • PMN

    polymorphonuclear leucocyte(s)

  •  
  • PRR

    pattern-recognition receptor(s)

  •  
  • RIP

    ribosome-inactivating protein

  •  
  • Stx

    Shiga toxin(s)

  •  
  • STEC

    Stx-producing Escherichia coil

  •  
  • TCA

    trichloroacetic acid

  •  
  • TLR

    Toll-like receptor

AUTHOR CONTRIBUTION

Valentina Arfilli, Domenica Carnicelli and Laura Rocchi performed most of the experiments, and contributed to the experimental design and data analysis. Francesca Ricci, Pasqualepaolo Pagliaro and Pier Luigi Tazzari performed the cytofluorimetric assays, and contributed to the experimental design and data analysis. Maurizio Brigotti planned the experiments, analysed the results and wrote the manuscript.

We thank Professor Andrea Bolognesi and Professor Fiorenzo Stirpe for generously providing ricin and saporin 6 respectively. We also thank Professor Emanuele Papini (Dipartimento di Scienze Biomediche Sperimentali dell'Università di Padova, Padova, Italy) and Dr Christine Betts (Dipartimento di Patologia Sperimentale, Università di Bologna, Bologna, Italy) for the critical reading of the manuscript prior to submission.

FUNDING

This work was supported by the University of Bologna (RFO funds to M.B).

References

References
1
Trompeter
R. S.
Schwartz
R.
Chantler
C.
Dillon
M. J.
Haycock
G. B.
Kay
R.
Barratt
T. M.
Haemolytic-uraemic syndrome: an analysis of prognostic features
Arch. Dis. Child.
1983
, vol. 
58
 (pg. 
101
-
105
)
2
Tarr
P. I.
Gordon
C. A.
Chandler
W. L.
Shiga-toxin-producing Escherichia coli and haemolytic uraemic syndrome
Lancet
2005
, vol. 
365
 (pg. 
1073
-
1086
)
3
Noris
M.
Remuzzi
G.
Hemolytic uremic syndrome
J. Am. Soc. Nephrol.
2005
, vol. 
16
 (pg. 
1035
-
1050
)
4
Paton
J. C.
Paton
A. W.
Pathogenesis and diagnosis of Shiga toxin-producing Escherichia coli infections
Clin. Microbiol. Rev.
1998
, vol. 
11
 (pg. 
450
-
479
)
5
Endo
Y.
Tsurugi
K.
Yutsudo
T.
Takeda
Y.
Ogasawara
T.
Igarashi
K.
Site of action of a Vero toxin (VT2) from Escherichia coli O157:H7 and of Shiga toxin on eukaryotic ribosomes. RNA N-glycosidase activity of the toxins
Eur. J. Biochem.
1988
, vol. 
171
 (pg. 
45
-
50
)
6
Brigotti
M.
Alfieri
R.
Sestili
P.
Bonelli
M.
Petronini
P. G.
Guidarelli
A.
Barbieri
L.
Stirpe
F.
Sperti
S.
Damage to nuclear DNA induced by Shiga toxin 1 and ricin in human endothelial cells
FASEB J.
2002
, vol. 
16
 (pg. 
365
-
372
)
7
Barbieri
L.
Valbonesi
P.
Brigotti
M.
Montanaro
L.
Stirpe
F.
Sperti
S.
Shiga-like toxin I is a polynucleotide:adenosine glycosidase
Mol. Microbiol.
1998
, vol. 
29
 (pg. 
661
-
662
)
8
Brigotti
M.
Carnicelli
D.
Ravanelli
E.
Vara
A. G.
Martinelli
C.
Alfieri
R. R.
Petronini
P. G.
Sestili
P.
Molecular damage and induction of proinflammatory cytokines in human endothelial cells exposed to Shiga toxin 1, Shiga toxin 2, and α-sarcin
Infect. Immun.
2007
, vol. 
75
 (pg. 
2201
-
2207
)
9
Matussek
A.
Lauber
J.
Bergau
A.
Hansen
W.
Rohde
M.
Dittmar
K. E.
Gunzer
M.
Mengel
M.
Gatzlaff
P.
Hartmann
M.
, et al. 
Molecular and functional analysis of Shiga toxin-induced response patterns in human vascular endothelial cells
Blood
2003
, vol. 
102
 (pg. 
1323
-
1332
)
10
Nakao
H.
Takeda
T.
Escherichia coli Shiga toxin
J. Nat. Toxins
2000
, vol. 
9
 (pg. 
299
-
313
)
11
Caprioli
A.
Luzzi
I.
Rosmini
F.
Pasquini
P.
Cirrincione
R.
Gianviti
A.
Matteucci
M. C.
Rizzoni
G.
Hemolytic-uremic syndrome and Vero cytotoxin-producing Escherichia coli infection in Italy
J. Infect. Dis.
1992
, vol. 
166
 (pg. 
154
-
158
)
12
Karmali
M. A.
Petric
M.
Lim
C.
Fleming
P. C.
Arbus
G. S.
Lior
H.
The association between idiopathic hemolytic uremic syndrome and infection by verotoxin-producing Escherichia coli
J. Infect. Dis.
1985
, vol. 
151
 (pg. 
775
-
782
)
13
Brigotti
M.
Caprioli
A.
Tozzi
A. E.
Tazzari
P. L.
Ricci
F.
Conte
R.
Carnicelli
D.
Procaccino
M. A.
Minelli
F.
Ferretti
A. V.
, et al. 
Shiga toxins present in the gut and in the polymorphonuclear leukocytes circulating in the blood of children with hemolytic-uremic syndrome
J. Clin. Microbiol.
2006
, vol. 
44
 (pg. 
313
-
317
)
14
Tazzari
P. L.
Ricci
F.
Carnicelli
D.
Caprioli
A.
Tozzi
A. E.
Rizzoni
G.
Conte
R.
Brigotti
M.
Flow cytometry detection of Shiga toxins in the blood from children with hemolytic uremic syndrome
Cytometry B Clin. Cytom.
2004
, vol. 
61
 (pg. 
40
-
44
)
15
Te Loo
D. M.
van Hinsbergh
V. W.
Van Den Heuvel
L. P.
Monnens
L.A.
Detection of verocytotoxin bound to circulating polymorphonuclear leukocytes of patients with hemolytic uremic syndrome
J. Am. Soc. Nephrol.
2001
, vol. 
12
 (pg. 
800
-
806
)
16
te Loo
D. M.
Monnens
L. A.
Van Der Velden
T. J.
Vermeer
M. A.
Preyers
F.
Demacker
P. N.
Van Den Heuvel
L. P.
van Hinsbergh
V. W.
Binding and transfer of verocytotoxin by polymorphonuclear leukocytes in hemolytic uremic syndrome
Blood
2000
, vol. 
95
 (pg. 
3396
-
3402
)
17
Brigotti
M.
Carnicelli
D.
Ravanelli
E.
Barbieri
S.
Ricci
F.
Bontadini
A.
Tozzi
A. E.
Scavia
G.
Caprioli
A.
Tazzari
P. L.
Interactions between Shiga toxins and human polymorphonuclear leukocytes
J. Leukocyte Biol.
2008
, vol. 
84
 (pg. 
1019
-
1027
)
18
Griener
T. P.
Mulvey
G. L.
Marcato
P.
Armstrong
G. D.
Differential binding of Shiga toxin 2 to human and murine neutrophils
J. Med. Microbiol.
2007
, vol. 
56
 (pg. 
1423
-
1430
)
19
Fernandez
G. C.
Gomez
S. A.
Rubel
C. J.
Bentancor
L. V.
Barrionuevo
P.
Alduncin
M.
Grimoldi
I.
Exeni
R.
Isturiz
M. A.
Palermo
M. S.
Impaired neutrophils in children with the typical form of hemolytic uremic syndrome
Pediatr. Nephrol.
2005
, vol. 
20
 (pg. 
1306
-
1314
)
20
Flagler
M. J.
Strasser
J. E.
Chalk
C. L.
Weiss
A. A.
Comparative analysis of the abilities of Shiga toxins 1 and 2 to bind to and influence neutrophil apoptosis
Infect. Immun.
2007
, vol. 
75
 (pg. 
760
-
765
)
21
Geelen
J. M.
Van Der Velden
T. J.
Te Loo
D. M.
Boerman
O. C.
Van Den Heuvel
L. P.
Monnens
L. A.
Lack of specific binding of Shiga-like toxin (verocytotoxin) and non-specific interaction of Shiga-like toxin 2 antibody with human polymorphonuclear leucocytes
Nephrol. Dial. Transplant.
2007
, vol. 
22
 (pg. 
749
-
755
)
22
Akira
S.
Takeda
K.
Toll-like receptor signalling
Nat. Rev. Immunol.
2004
, vol. 
4
 (pg. 
499
-
511
)
23
Creagh
E. M.
O'Neill
L. A.
TLRs, NLRs and RLRs: a trinity of pathogen sensors that co-operate in innate immunity
Trends Immunol.
2006
, vol. 
27
 (pg. 
352
-
357
)
24
Gazi
U.
Martinez-Pomares
L.
Influence of the mannose receptor in host immune responses
Immunobiology
2009
, vol. 
214
 (pg. 
554
-
561
)
25
Barbieri
L.
Battelli
M. G.
Stirpe
F.
Ribosome-inactivating proteins from plants
Biochim. Biophys. Acta
1993
, vol. 
1154
 (pg. 
237
-
282
)
26
Peumans
W. J.
Hao
Q.
Van Damme
E. J.
Ribosome-inactivating proteins from plants: more than RNA N-glycosidases?
FASEB J.
2001
, vol. 
15
 (pg. 
1493
-
1506
)
27
Calderwood
S. B.
Auclair
F.
Donohue-Rolfe
A.
Keusch
G. T.
Mekalanos
J. J.
Nucleotide sequence of the Shiga-like toxin genes of Escherichia coli
Proc. Natl. Acad. Sci. U.S.A.
1987
, vol. 
84
 (pg. 
4364
-
4368
)
28
O'Brien
A. D.
Tesh
V. L.
Donohue-Rolfe
A.
Jackson
M. P.
Olsnes
S.
Sandvig
K.
Lindberg
A. A.
Keusch
G. T.
Shiga toxin: biochemistry, genetics, mode of action, and role in pathogenesis
Curr. Top. Microbiol. Immunol.
1992
, vol. 
180
 (pg. 
65
-
94
)
29
Simmons
B. M.
Stahl
P. D.
Russell
J. H.
Mannose receptor-mediated uptake of ricin toxin and ricin A chain by macrophages. Multiple intracellular pathways for a chain translocation
J. Biol. Chem.
1986
, vol. 
261
 (pg. 
7912
-
7920
)
30
Skilleter
D. N.
Paine
A. J.
Stirpe
F.
A comparison of the accumulation of ricin by hepatic parenchymal and non-parenchymal cells and its inhibition of protein synthesis
Biochim. Biophys. Acta
1981
, vol. 
677
 (pg. 
495
-
500
)
31
Valera
I.
Vigo
A. G.
Alonso
S.
Barbolla
L.
Crespo
M. S.
Fernandez
N.
Peptidoglycan and mannose-based molecular patterns trigger the arachidonic acid cascade in human polymorphonuclear leukocytes
J. Leukocyte Biol.
2007
, vol. 
81
 (pg. 
925
-
933
)
32
Ryd
M.
Alfredsson
H.
Blomberg
L.
Andersson
A.
Lindberg
A. A.
Purification of Shiga toxin by α-D-galactose-(1–4)-β-D-galactose-(1–4)-β-Dglucose-(1–) receptor ligand-based chromatography
FEBS Lett.
1989
, vol. 
258
 (pg. 
320
-
322
)
33
Sperti
S.
Montanaro
L.
Rambelli
F.
Dye affinity chromatography of ricin subunits
Biosci. Rep.
1986
, vol. 
6
 (pg. 
1035
-
1040
)
34
Brigotti
M.
Rambelli
F.
Zamboni
M.
Montanaro
L.
Sperti
S.
Effect of α-sarcin and ribosome-inactivating proteins on the interaction of elongation factors with ribosomes
Biochem. J.
1989
, vol. 
257
 (pg. 
723
-
727
)
35
Pope
C. G.
Stevens
M. F.
The purification of diphtheria toxin and the isolation of crystalline toxin-protein
Br. J. Exp. Pathol.
1958
, vol. 
39
 (pg. 
139
-
149
)
36
Cassatella
M. A.
Cappelli
R.
Della Bianca
V.
Grzeskowiak
M.
Dusi
S.
Berton
G.
Interferon-γ activates human neutrophil oxygen metabolism and exocytosis
Immunology
1988
, vol. 
63
 (pg. 
499
-
506
)
37
van Setten
P. A.
Monnens
L. A.
Verstraten
R. G.
Van Den Heuvel
L. P.
van Hinsbergh
V. W.
Effects of verocytotoxin-1 on nonadherent human monocytes: binding characteristics, protein synthesis, and induction of cytokine release
Blood
1996
, vol. 
88
 (pg. 
174
-
183
)
38
Hedblom
M. L.
Cawley
D. B.
Boguslawski
S.
Houston
L. L.
Binding of ricin A chain to rat liver ribosomes: relationship to ribosome inactivation
J. Supramol. Struct.
1978
, vol. 
9
 (pg. 
253
-
268
)
39
Bainton
D. F.
Ullyot
J. L.
Farquhar
M. G.
The development of neutrophilic polymorphonuclear leukocytes in human bone marrow
J. Exp. Med.
1971
, vol. 
134
 (pg. 
907
-
934
)
40
Akgul
C.
Moulding
D. A.
Edwards
S. W.
Molecular control of neutrophil apoptosis
FEBS Lett.
2001
, vol. 
487
 (pg. 
318
-
322
)
41
Benatti
L.
Nitti
G.
Solinas
M.
Valsasina
B.
Vitale
A.
Ceriotti
A.
Soria
M. R.
A Saporin-6 cDNA containing a precursor sequence coding for a carboxyl-terminal extension
FEBS Lett.
1991
, vol. 
291
 (pg. 
285
-
288
)
42
Stirpe
F.
Gasperi-Campani
A.
Barbieri
L.
Falasca
A.
Abbondanza
A.
Stevens
W. A.
Ribosome-inactivating proteins from the seeds of Saponaria officinalis L. (soapwort), of Agrostemma githago L. (corn cockle) and of Asparagus officinalis L. (asparagus), and from the latex of Hura crepitans L. (sandbox tree)
Biochem. J.
1983
, vol. 
216
 (pg. 
617
-
625
)
43
Hartley
M. R.
Lord
J. M.
Cytotoxic ribosome-inactivating lectins from plants
Biochim. Biophys. Acta
2004
, vol. 
1701
 (pg. 
1
-
14
)
44
Macher
B. A.
Klock
J. C.
Isolation and chemical characterization of neutral glycosphingolipids of human neutrophils
J. Biol. Chem.
1980
, vol. 
255
 (pg. 
2092
-
2096
)
45
Woo
B. H.
Lee
J. T.
Na
D. H.
Lee
K. C.
Sepharose-unbinding ricin E as a source for ricin A chain immunotoxin
J. Immunol. Methods
2001
, vol. 
249
 (pg. 
91
-
98
)