Toxicology in Vitro 24 (2010) 1176–1182

Contents lists available at ScienceDirect

Toxicology in Vitro

journal homepage: www.elsevier .com/locate / toxinvi t

Different in vitro toxicities of structurally similar type Iribosome-inactivating proteins (RIPs)

Ou Sha a,b,*, David Tai Wai Yew a, Tzi Bun Ng a, Lin Yuan b, Wing Hang Kwong a

a School of Biomedical Sciences, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, N.T., Hong Kong, Chinab Department of Basic Medical Sciences, School of Medicine, Shenzhen University, Shenzhen, Guangdong, China

a r t i c l e i n f o

Article history:Received 24 February 2009Accepted 15 February 2010Available online 17 February 2010

Keywords:RIPsRicin A chainTrichosanthinApoptosisNecrosisIn vitroIn vivo

0887-2333/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.tiv.2010.02.014

Abbreviations: RIP, ribosome-inactivating protein;A chain; RCA, Ricinus communis agglutinin.

* Corresponding author. Address: School of BiomMedicine, The Chinese University of Hong Kong, Shatin+852 2696 1226; fax: +852 2603 5031.

E-mail addresses: shaou@cuhk.edu.hk, shaou09@y

a b s t r a c t

This study was aimed at investigating and comparing the cytotoxicities of two structurally similar type IRIPs, namely trichosanthin (TCS) and free ricin A chain (RTA). A type II RIP, namely Ricinus communisagglutinin (RCA), was also included for comparison. The three RIPs were added separately to culturesof NIH 3T3 cells. The effective doses and time courses were analyzed using cell counts. Polyclonal anti-bodies against TCS and RTA were produced in rabbits and purified by a protein A-Sepharose CL-4B col-umn. The mechanisms of cell death were determined by TUNEL, immunohistochemical staining, flowcytometry, and Western blotting. The effective doses for TCS, RTA and RCA were found to be 800, 50,and 50 nM, respectively. All three RIPs induced apoptosis. In all cases, activation of caspase-3 and cas-pase-8, but not caspase-9, was detected. Additionally, RTA caused in vivo tissue necrosis in rabbits afterintradermal administration. Hence the mechanism of cell death due to RTA intoxication may varydepending on the experimental conditions, being necrosis in vivo and apoptosis in vitro. The present find-ings may shed light on the apoptotic pathway induced by RIPs. RTA may be useful for studying the shift incell death.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction adenine-4324 of the 28S ribosomal RNA. Another common exam-

Ribosome-inactivating proteins (RIPs) are toxic proteinsbecause they all possess a rRNA N-glycosidase domain that canirreversibly inactivate ribosomes in mammalian cells (Stirpe andBarbieri, 1986). RIPs are classified into two major types, type Iand type II (Peumans et al., 2001). Type I RIPs have a singlerRNA-cleaving domain and are typically composed of one single in-tact polypeptide chain. Type II RIPs consist of two kinds of polypep-tide chains, namely, A and B chains. The A chain is a rRNA-cleavingmoiety which resembles type I RIPs in both function and structure.The slightly larger B chain is the lectinic, galactosyl-specific cell-binding moiety, which can bind to galactose receptors on almostall eukaryotic cell surfaces.

Two different type I RIPs were examined in this study. Tricho-santhin (TCS), isolated from the root tuber of Trichosanthes kirilowiiMaximowicz, has a single peptide chain and is a typical example oftype I RIPs (Shaw et al., 2005). It has a molecular weight of 26 kDa,and it inactivates ribosomes by cleaving the N-glycosidic bond at

ll rights reserved.

TCS, trichosanthin; RTA, ricin

edical Sciences, Faculty of, N.T., Hong Kong, China. Tel.:

ahoo.com.cn (O. Sha).

ple of type I RIP is free ricin A chain (RTA), which is isolated fromthe type II RIP ricin. It has a molecular weight of 32 kDa, and it alsospecifically attacks the 28S rRNA (Endo and Tsurugi, 1987; Endoet al., 1987; Wales et al., 1993).

TCS is homologous to RTA in amino acid sequence and crystalstructure (Jin et al., 1981; Zhang and Wang, 1986; Maraganoreet al., 1987; Collins et al., 1990). However, they exert different tox-icities to different cells. TCS is selectively toxic to choriocarcinomacells, proximal tubule epithelial cells, antigen-specific T cells, andHIV-infected macrophages and lymphocytes (Byers et al., 1990;Tang et al., 1997; Li et al., 2005; Wang et al., 2005; Xia et al.,2006). In contrast, RTA is preferentially toxic to phagocytic cellsand endothelial cells (Simmons et al., 1986; Zenilman et al.,1988; Fiani et al., 1993; Johnston and Jasuja, 1994). Additionally,in our previous in vivo experiments, TCS and RTA also exhibited dif-ferent neuronal toxicities to rat retinal cells. The mechanisms ofcell death elicited by these two proteins are also distinct (Shaet al., 2008). TCS eliminates specific retinal cells by apoptosis,while RTA causes retinal necrosis.

The aim of this study was to investigate the in vitro toxicity ofTCS and RTA as well as the mechanisms of the subsequent celldeath. This would allow a better understanding of the action ofthese two type I RIPs. A type II RIP, Ricinus communis agglutinin(RCA), was included in this investigation for the purpose of

O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182 1177

comparison. Each RCA molecule consists of two ricin A (RTA)chains and two ricin B (RTB) chains; it manifests a much lowertoxicity than that of ricin, which has only one A chain and one Bchain.

2. Materials and methods

All experiments had been approved by the Animal ResearchEthics Committee of the Chinese University of Hong Kong and wereconducted in conformance with the Animal (Control of Experi-ments) Ordinance Law of Hong Kong.

The purity of the TCS and RTA used in this study (both from Sig-ma Co., U.S.A.) has previously been examined by SDS–PAGE andisoelectric focusing (Sha et al., 2008). The results showed that theTCS was pure, and that the RTA was not contaminated by RTB orricin.

2.1. Cell culture

NIH 3T3 mouse fibroblasts, commonly used in RIP cytotoxicitystudies, were used here. Briefly, the cell line (American Type Cul-ture Collection, U.S.A.) was maintained as a monolayer in DFH:Dulbecco’s modified Eagle medium and nutrient mixture F-12(Ham) at 1:1 (GIBCO BRL, U.S.A.), containing 10% fetal calf serum(GIBCO BRL, U.S.A.), 100 units/ml penicillin and 100 lg/ml strepto-mycin (Sigma Co., U.S.A.). The fast growing log phase was deter-mined from the cell growth curve for the administration of theRIPs.

2.2. Determination of effective dose

The RIPs were added separately into the cultures 2 days afterseeding the cells in 24-well culture plates, while the cells werein their log phase. The doses of TCS used were 0, 150, 200, 400,800 and 1600 nM, and the doses of RTA and RCA used were 0,50, 100, 150, 200 and 400 nM. Twenty-four hours after the RIPadministration, the culture medium was discarded. The cells weredetached from the culture plates by adding 0.25% trypsin and0.03% EDTA. The resuspended cells were then counted under thelight microscope by using a hemocytometer without any staining.All cells in the 9-mm2 square of the chamber were counted, and thecell concentration (cells/ml) was calculated as cells/mm � 103. Theeffective doses at which TCS, RTA and RCA caused a significantchange in cell counts were determined.

Fig. 1. Dose response of cultured cells to different RIP entities. The asterisks (*) indicatesignificantly from the control (p < 0.001). The lowest concentration of TCS used is 800 n

2.3. Time-course study

The three RIPs were added separately to the cultures at theireffective doses (800, 50 and 50 nM for TCS, RTA and RCA, respec-tively). The cell count from each culture well was determined at0, 12 and 24 h after the RIP administration.

2.4. Production and purification of polyclonal antibodies

Male New Zealand white rabbits weighing 2.5–3 kg were usedto raise antibodies (Abs) against TCS and RTA. The immunogen(0.7 mg of TCS or 0.3–0.5 mg of RTA) was mixed with completeFreud’s adjuvant for the initial injection, or with incompleteFreud’s adjuvant (both from Sigma Co., U.S.A.) for booster injection.The injections were spaced 4 weeks apart. The emulsion was in-jected intradermally into the back of rabbits. The titers of anti-serawere determined by double immunodiffusion method. The Abswere purified by protein A affinity chromatography (Zhao andGao, 2000). The titers of the purified Abs were determined by anenzyme-linked immunosorbent assay (ELISA) and the specificitieswere evaluated by immunohistochemical (IHC) staining.

2.5. IHC and TUNEL double staining

After treatment with the RIPs, the cells placed on 13 mm glasscoverslips were fixed in 4% paraformaldehyde for 30 min, and thentreated with 0.1% Triton X-100 for 10 min. After blocking with 0.5%normal horse serum, the cells were incubated with anti-TCS oranti-RTA overnight (both at 1:100 dilutions). Cy3-conjugatedanti-rabbit Ab (Jackson ImmunoResearch Laboratories, U.S.A.,1:400) was used as the secondary antibody. The cells were subse-quently stained by TUNEL method (terminal deoxynucleotidyltransferase biotin-dUTP nick end labeling). In brief, the cells wereincubated with a mixture of terminal deoxynucleotidyl transferase(TdT) and FITC (Fluorescence In Situ Cell Death Detection Kit,Roche Applied Science, Germany). Finally, the cells were observedunder a confocal microscope (MRC 1024, Bio-Rad). The images cap-tured were processed using LaserShop 2000 software. For the neg-ative control, the primary Abs and TdT were omitted from therespective incubation media. Retinas of 4-week-old Royal Collegeof Surgeon rats, known to contain a high density of apoptotic cells,were used as the positive control. All TCS-, RTA- and RCA-treatedcells were also double stained with rabbit anti-cleaved caspase-3Ab (Cell Signaling Technology, U.S.A., at 1:100 dilutions) and thesame TUNEL method.

the lowest concentrations at which the cell numbers of the treated cultures differedM (a), and that of RTA and RCA is 50 nM (b). Results represent mean + SEM (n = 6).

1178 O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182

2.6. Flow cytometry

The procedures followed those given in the Vybrant ApoptosisAssay Kit #4 (Molecular Probes, U.S.A.). Briefly, 0–24 h after theadministration of the RIPs, the cells were washed twice with coldDulbecco’s phosphate-buffered saline (GIBCO BRL, U.S.A.), andresuspended in 1 ml of the same buffer. For the positive control,normal cells were treated with 10 IU DNAse I for 1 h at 37 �C beforeharvesting. One microliter each of YO-PRO-1 and propidium iodide(PI) (included in the kit) was added to each tube of the cell suspen-sion. Necrotic cells would be double stained, apoptotic cells stainedgreen, and healthy cells would not be stained. For the negative con-

Fig. 2. Time course of the changes in cell number following administration of TCS,RTA and RCA to the culture (at the respective lowest effective doses shown inFig. 1). All three RIP entities caused a significant decrease in cell number at 24 h (*indicates p < 0.001) as compared to the control. Results represent mean + SEM(n = 6).

Fig. 3. Skin reaction of rabbits after injection of TCS and RTA. Following the initial and boinitial injection of RTA (c), necrosis was observed. The skin recovered about 2 months lateinjection.

trol, neither YO-PRO-1 nor propidium iodide was added. The cellsuspensions were then incubated on ice in the dark for 30 min,and immediately run through the flow cytometer (EPICS Altra,Beckman Coulter, U.S.A.). The percentages of apoptotic, necroticand living cells were recorded by the machine automatically withExpo32 Multicomp software.

2.7. Western blotting

The cells were treated with the different RIPs for 12 h. Thenthey were lysed in lysis buffer (50 mM Tris–HCl at pH 7.4 contain-ing 150 mM NaCl, 1% Triton X-100, 5 mM EDTA, 0.1% SDS, 1% phen-ylmethylsulfonyl fluoride, 0.05% aprotinin, 0.01% leupeptin, and0.01% pepstatin). Protein concentrations were determined by theLowry assay using a Protein DC kit (Bio-Rad, U.S.A.). The proteinsamples (100 lg) were run on SDS–PAGE and electrotransferredto a nitrocellulose membrane (Hybond ECL, GE Healthcare,U.S.A.). After blocking with 5% non-fat milk (Bio-Rad, U.S.A.), themembranes were separately incubated with rabbit anti-caspase-8(Santa Cruz Biotechnology Inc., U.S.A.), mouse anti-caspase-9 (CellSignaling Technology, U.S.A.) and goat anti-b actin (Santa Cruz Bio-technology Inc., U.S.A.). The membranes were then incubated withhorseradish peroxidase-conjugated anti-rabbit, anti-mouse (bothfrom GE Healthcare, U.S.A.), and anti-goat (Jackson ImmunoRe-search Laboratories, U.S.A.) secondary Abs, respectively. The mem-branes were developed with a Western blotting kit (GE Healthcare,U.S.A.) for specific protein detection with enhanced chemilumines-cence. Positive controls for active caspase-8 and active caspase-9were prepared using lysed apoptotic Jurkat cells (BD Pharmingen,U.S.A.).

2.8. Statistics

All quantitative data were presented as mean + SEM with n = 6.The changes in cell number after the toxin treatment, and the

oster TCS injections (a and b, respectively), the skin reaction was mild. Following ther (d). The photos were taken 1 week (a and c) and 2 months (b and d) after the initial

O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182 1179

difference in number of the apoptotic and necrotic cells, weretested with Student’s t-test. p < 0.05 was considered to indicatestatistical significance of the difference.

3. Results

3.1. Determination of effective concentration of RIP

The effective concentrations for the different RIP entities werefound to be 800 nM for TCS, and 50 nM for both RTA and RCA(Fig. 1a and b). The lowest dose (800 nM) of TCS effected a signif-icant decrease in cell count (p < 0.001), while at the lower dosesof 400, 200 and 150 nM, the cell counts did not decrease signifi-

Fig. 4. Immunohistochemical localization of TCS (a–d) and RTA (e–h). Positive signals weThe nuclei of the treated cells were stained by TUNEL (b and f, respectively). TCS- and RTATUNEL method (j and n), respectively. Images a and b, e and f, i and j, and m and n werimages to form d, h, l and p, respectively. Confocal images at �400 magnification; scale

cantly when compared to the controls (p > 0.05). At these effectivedoses, there was no statistically significant difference in cell countbetween the cultures treated with the different RIPs. These effec-tive doses were used in all subsequent experiments.

3.2. Time-course study

There was no significant difference at 12 h between the exper-imental and control groups (Fig. 2). Twenty-four hours after theadministration of the RIPs, the cell counts in all experimentalgroups were reduced as compared to their respective controlgroups (p < 0.001). However, no significant difference was foundamong the experimental groups at this time point.

re located in the cytoplasm of the cells treated with the respective proteins (a and e).-treated cells were also double stained with anti-cleaved caspase-3 Ab (i and m) ande separately merged to form c, g, k and o, which were merged with phase contrastbar: 50 lm.

1180 O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182

3.3. Reactions of rabbits during Ab immunization

The skin of the three rabbits used for TCS immunization ap-peared normal after both the initial and booster injections(Fig. 3a and b). In contrast, three of the rabbits died a few days afterreceiving an injection of 0.3–0.5 mg of RTA. The remaining tworabbits survived. About 1 week after the initial RTA injection, theinjection sites became necrotic (Fig. 3c). The skin appeared blackand became detached from the surrounding tissues. The damageappeared to be confined to the dermal layer and did not extendinto the muscular layer. The skin recovered about 2 months laterwhen anti-RTA Ab was detected by double immunodiffusion. After-wards, the reaction of the RTA-injected skin was mild (Fig. 3d).

3.4. IHC and TUNEL staining

In the cultures treated with TCS or RTA, the immunostainingwith anti-TCS and -RTA Abs showed positive signals in the cyto-plasm (Fig. 4a and e). When either the RIPs or antibodies wereomitted, no positive signals were detected. TUNEL-positive signals

Fig. 5. FACS histograms (left panel) and scattergrams (right panel) showing the relativeadministration of TCS (a), RTA (b), and RCA (c). The asterisks (*) indicate statistically sp < 0.001). Results represent mean + SEM (n = 6). N: necrotic; A: apoptotic; V: vital cells

were also found in the nuclei of the TCS- and RTA-treated cells(Fig. 4b and f), and were co-localized with the signals for the RIPsin the same cells (Fig. 4c, d, g and h). In all TCS-, RTA- and RCA-trea-ted cells, positive signals for cleaved caspase-3 signals were de-tected in co-localization with the TUNEL signals.

3.5. Flow cytometry

In the cultures treated with TCS, RTA and RCA, the percentagesof necrotic cells were not significantly different from those in thecontrol groups (Fig. 5). However, compared with the control, thepercentages of apoptotic cells increased significantly with time,reaching a maximum at 12 h (p < 0.001 in all cases). The percent-ages of apoptotic cells were also significantly higher than the per-centages of necrotic cells (p < 0.001).

3.6. Western blotting

Using anti-caspase-8 Ab, two bands of molecular weight 55 and30 kDa were detected in the extracts of cells treated with TCS, RTA

number of apoptotic and necrotic cells measured at 0, 1, 2, 4, 12 and 24 h after theignificant difference between the treated and the control cultures (*, p < 0.05; ***,.

Fig. 6. Western blotting of extracts of cells treated with TCS, RTA, RCA or controlbuffer (lanes 1–4, respectively). Anti-caspase-8 staining showed two bands at 55and 30 kD in all extracts (a). Anti-caspase-9 staining showed only one band at 49 kDin all extracts (b). Anti-b-actin staining (loading control) showed one band at 42 kDin all extracts (c).

O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182 1181

or RCA (Fig. 6a). In the extract of normal cells, only the 55-kDaband was observed. In contrast, immunostaining using anti-cas-pase-9 Ab revealed only a single 49-kDa band in the extracts ofthe RIP-treated and control cultures (Fig. 6b). As a loading control,anti-b actin antibody was used. The 42-kDa actin band was de-tected in all extracts (Fig. 6c).

4. Discussion

4.1. In vitro toxicity

The data on effective concentrations of the RIPs showed that thein vitro toxicity of TCS is much lower than that of RTA, while thetoxicity of RTA is similar to that of RCA. RTA, a single enzymaticchain isolated from the type II RIP ricin, possesses an amino acidsequence homologous to TCS, which is a type I RIP (Maraganoreet al., 1987; Collins et al., 1990). However its toxicity is similar to

another type II RIP, RCA. These findings suggest that the presenceof the cell-binding B chain (RTB) does not always increase the tox-icity of the RIP, and its presence may not enhance the toxicity ofthe RIP. Furthermore, all of these toxins induced apoptotic deathin the cultured cells.

RIPs are toxic to cells due to their ribosome-inactivation ability.The cell death induced by RIPs, including ricin, RTA and TCS, wasusually necrotic (Garcia et al., 1993; Tanaka et al., 2001; Fracassoet al., 2002; Lord et al., 2003). Recently, it has been reported thatcertain RIPs induced apoptosis. For example, a type I RIP,Momordica charantia protein-30, could induce apoptosis inprostatic intraepithelial neoplasia and prostate cancer cell linesby inhibiting histone deacetylase-1 activity (Xiong et al., 2009).Another type I RIP, saporin, can enter U937 cells and induce cas-pase-dependent apoptosis via the mitochondrial pathway, andthe apoptosis onset occurs before the inhibition of protein synthe-sis (Sikriwal et al., 2008; Weng et al., 2008). TCS has also been re-ported to induce apoptosis with an unknown mechanism (Shawet al., 2005; Sha et al., 2008). In addition, it was reported thatRTA, in the form of immunotoxins and other conjugates, could in-duce apoptosis (Thepen et al., 2000; van Roon et al., 2003). How-ever, no apoptotic cell death produced by RCA has been reported.

In the present study, RTA and TCS were found to have causedmainly apoptosis in the cultured cells. RTA and TCS signals werelocalized in the cytoplasm. RCA was found to induce cell deathwith a pattern analogous to that of RTA. Hence, the B chain ofthe ricin molecule is not an obligatory requirement for the entryof the ribosome-inactivating chain into the cell. The percentageof apoptotic cells induced by all RIPs peaked at 12 h after theiradministration, followed by a decline at 24 h. This could be dueto the proliferation of the surviving cells. The apoptotic changes in-duced by the three toxins appeared to follow the caspase-8-depen-dent extrinsic pathway. The involvement of the caspase cascadewas confirmed by the activation of caspase-3, which is well recog-nized to be the common down-stream pathway of apoptosis. Tobetter understand the process of cell death induced by RIPs, theinvolvement of mitochondria needs to be further investigated infuture studies. Permeabilization of mitochondrial membrane hasbeen regarded as a common event leading to both apoptosis andnecrosis (Malhi et al., 2006).

4.2. In vivo toxicity

All of the rabbits injected with 0.7 mg TCS for antibody produc-tion survived, and their skin at the injection sites appeared healthy.In contrast, some rabbits died after an injection of either 0.3 or0.5 mg of RTA, and those that survived developed skin necrosis.This observation was consistent with our previous results fromthe rat retina experiments (Sha et al., 2008). In that study, RTA,even at a much lower dose, caused necrosis and inflammation inthe rat retina, while TCS, even at a much higher dose, causedmainly apoptosis in specific types of retinal cells. Again, these sug-gest that the presence of the cell-binding B chain is not always nec-essary to trigger cell death.

Comparing the present in vitro results with the in vivo resultsfrom the retina and whole animals, RTA appears to trigger differenttypes of cell death. A possible explanation is that the effect of RTAon cultured fibroblasts is direct, whereas the effect in vivo is com-plicated by the host’s immune mechanism. In the in vivo condition,the toxic effect of RTA on blood vessels (Baluna and Vitetta, 1999)may also cause a shift from apoptosis to necrosis. Various otherfactors have been proposed to explain the shift between differenttypes of cell death (Wan et al., 2003; Cole and Perez-Polo, 2004;McKnight et al., 2005; Walter et al., 2008; He et al., 2009). RTAmay well be a useful tool to study these factors.

1182 O. Sha et al. / Toxicology in Vitro 24 (2010) 1176–1182

References

Baluna, R., Vitetta, E.S., 1999. An in vivo model to study immunotoxin-inducedvascular leak in human tissue. Journal of Immunotherapy 22, 41–47.

Byers, V.S., Levin, A.S., Waites, L.A., Starrett, B.A., Mayer, R.A., Clegg, J.A., Price, M.R.,Robins, R.A., Delaney, M., Baldwin, R.W., 1990. A phase I/II study oftrichosanthin treatment of HIV disease. AIDS 4, 1189–1196.

Cole, K., Perez-Polo, J.R., 2004. Neuronal trauma model: in search of Thanatos.International Journal of Developmental Neuroscience 22, 485–496.

Collins, E.J., Robertus, J.D., LoPresti, M., Stone, K.L., Williams, K.R., Wu, P., Hwang, K.,Piatak, M., 1990. Primary amino acid sequence of alpha-trichosanthin andmolecular models for abrin A-chain and alpha-trichosanthin. Journal ofBiological Chemistry 265, 8665–8669.

Endo, Y., Tsurugi, K., 1987. RNA N-glycosidase activity of ricin A-chain. Mechanismof action of the toxic lectin ricin on eukaryotic ribosomes. Journal of BiologicalChemistry 262, 8128–8130.

Endo, Y., Mitsui, K., Motizuki, M., Tsurugi, K., 1987. The mechanism of action of ricinand related toxic lectins on eukaryotic ribosomes. The site and thecharacteristics of the modification in 28S ribosomal RNA caused by thetoxins. Journal of Biological Chemistry 262, 5908–5912.

Fiani, M.L., Blum, J.S., Stahl, P.D., 1993. Endosomal proteolysis precedes ricin A-chaintoxicity in macrophages. Archives of Biochemistry and Biophysics 307, 225–230.

Fracasso, G., Bellisola, G., Cingarlini, S., Castelletti, D., Prayer-Galetti, T., Pagano, F.,Tridente, G., Colombatti, M., 2002. Anti-tumor effects of toxins targeted to theprostate specific membrane antigen. Prostate 53, 9–23.

Garcia, P.A., Bredesen, D.E., Vinters, H.V., Graefin von Einsiedel, R., Williams, R.L.,Kahn, J.O., Byers, V.S., Levin, A.S., Waites, L.A., Messing, R.O., 1993. Neurologicalreactions in HIV-infected patients treated with trichosanthin. Neuropathologyand Applied Neurobiology 19, 402–405.

He, S., Wang, L., Miao, L., Wang, T., Du, F., Zhao, L., Wang, X., 2009. Receptorinteracting protein kinase-3 determines cellular necrotic response to TNF-alpha. Cell 137, 1100–1111.

Jin, S.W., Sun, X.X., Wang, S.F., Tian, G.Y., Gu, Z.W., Qian, W.W., Liu, Y.Z., She, W.Y.,Qian, R.Q., Wang, Y., 1981. Chemistry of trichosanthin I. Physical and chemicalproperties of crystalline trichosanthin. Acta Chimica Sinica 39, 917–925.

Johnston, D.E., Jasuja, R., 1994. Purification of cultured primary rat hepatocytesusing selection with ricin A subunit. Hepatology 20, 436–444.

Li, F., Mei, Y., Wang, Y., Chen, C., Tu, J., Xiao, B., Xu, L., 2005. Trichosanthin inhibitsantigen-specific T cell expansion through nitric oxide-mediated apoptosispathway. Cell Immunology 234, 23–30.

Lord, J.M., Deeks, E., Marsden, C.J., Moore, K., Pateman, C., Smith, D.C., Spooner, R.A.,Watson, P., Roberts, L.M., 2003. Retrograde transport of toxins across theendoplasmic reticulum membrane. Biochemical Society Transactions 31, 1260–1262.

Malhi, H., Gores, G.J., Lemasters, J.J., 2006. Apoptosis and necrosis in the liver: a taleof two deaths? Hepatology 43, S31–44.

Maraganore, J.M., Joseph, M., Bailey, M.C., 1987. Purification and characterization oftrichosanthin. Homology to the ricin A chain and implications as to mechanismof abortifacient activity. Journal of Biological Chemistry 262, 11628–11633.

McKnight, J.J., Gray, S.B., O’Kane, H.F., Johnston, S.R., Williamson, K.E., 2005.Apoptosis and chemotherapy for bladder cancer. The Journal of Urology 173,683–690.

Peumans, W.J., Hao, Q., Van Damme, E.J., 2001. Ribosome-inactivating proteins fromplants: more than RNA N-glycosidases? The FASEB Journal 15, 1493–1506.

Sha, O., Kwong, W.H., Cho, Y.P., Yew, D.T.W., Ng, T.B., 2008. Different neuronaltoxicity of single-chain ribosome-inactivating protein on the rat retina. Toxicon51, 45–53.

Shaw, P.C., Lee, K.M., Wong, K.B., 2005. Recent advances in trichosanthin, aribosome-inactivating protein with multiple pharmacological properties.Toxicon 45, 683–689.

Sikriwal, D., Ghosh, P., Batra, J.K., 2008. Ribosome inactivating protein saporininduces apoptosis through mitochondrial cascade, independent of translationinhibition. International Journal of Biochemistry and Cell Biology 40, 2880–2888.

Simmons, B.M., Stahl, P.D., Russell, J.H., 1986. Mannose receptor-mediated uptake ofricin toxin and ricin A chain by macrophages. Multiple intracellular pathwaysfor a chain translocation. Journal of Biological Chemistry 261, 7912–7920.

Stirpe, F., Barbieri, L., 1986. Ribosome-inactivating proteins up to date. FEBS Letters195, 1–8.

Tanaka, K.S., Chen, X.Y., Ichikawa, Y., Tyler, P.C., Furneaux, R.H., Schramm, V.L., 2001.Ricin A-chain inhibitors resembling the oxacarbenium ion transition state.Biochemistry 40, 6845–6851.

Tang, N.L.S., Chan, W., Ke, Y., Mak, M.K.F., Lai, F.M., Tam, S., 1997. Acute renal failureand proximal tubule lesions after trichosanthin injection in rats. Experimentaland Molecular Pathology 64, 78–89.

Thepen, T., van Vuuren, A.J., Kiekens, R.C., Damen, C.A., Vooijs, W.C., van De Winkel,J.G., 2000. Resolution of cutaneous inflammation after local elimination ofmacrophages. Nature Biotechnology 18, 48–51.

van Roon, J.A., van Vuuren, A.J., Wijngaarden, S., Jacobs, K.M., Bijlsma, J.W., Lafeber,F.P., Thepen, T., van de Winkel, J.G., 2003. Selective elimination of synovialinflammatory macrophages in rheumatoid arthritis by an Fcgamma receptor I-directed immunotoxin. Arthritis and Rheumatism 48, 1229–1238.

Wales, R., Roberts, L.M., Lord, J.M., 1993. Addition of an endoplasmic reticulumretrieval sequence to ricin A chain significantly increases its cytotoxicity tomammalian cells. Journal of Biological Chemistry 268, 23986–23990.

Walter, D., Schmich, K., Vogel, S., Pick, R., Kaufmann, T., Hochmuth, F.C., Haber, A.,Neubert, K., McNelly, S., von Weizsäcker, F., Merfort, I., Maurer, U., Strasser, A.,Borner, C., 2008. Switch from type II to I Fas/CD95 death signaling on in vitroculturing of primary hepatocytes. Hepatology 48, 1942–1953.

Wan, L., Bellomo, R., Di Giantomasso, D., Ronco, C., 2003. The pathogenesis of septicacute renal failure. Current Opinion in Critical Care 9, 496–502.

Wang, Y., Mi, S.L., Lou, M.Y., Gao, Y., Chen, Z.L., An, C.C., 2005. Enhanced greenfluorescence protein tracks trichosanthin in human choriocarcinoma cells as afeasible and stable reporter. Frontiers in Bioscience 10, 2279–2284.

Weng, A., Melzig, M.F., Bachran, C., Fuchs, H., 2008. Enhancement of saporin toxicityagainst U937 cells by Gypsophila saponins. Journal of Immunotoxicology 5, 287–292.

Xia, X., Hou, F., Li, J., Ke, Y., Nie, H., 2006. Two novel proteins bind specifically totrichosanthin on choriocarcinoma cell membrane. The Journal of Biochemistry139, 725–731.

Xiong, S.D., Yu, K., Liu, X.H., Yin, L.H., Kirschenbaum, A., Yao, S., Narla, G., DiFeo, A.,Wu, J.B., Yuan, Y., Ho, S.M., Lam, Y.W., Levine, A.C., 2009. Ribosome-inactivatingproteins isolated from dietary bitter melon induce apoptosis and inhibit histonedeacetylase-1 selectively in premalignant and malignant prostate cancer cells.International Journal of Cancer 125, 774–782.

Zenilman, M.E., Fiani, M., Stahl, P., Brunt, E., Flye, M.W., 1988. Use of ricin A-chain toselectively deplete Kupffer cells. Journal of Surgical Research 45, 82–89.

Zhang, X.J., Wang, J.H., 1986. Homology of trichosanthin and ricin A chain (letter).Nature 321, 477–478.

Zhao, Y.H., Gao, X.Y., 2000. Course Manual for Biochemical Experimental Techniques(Shengwuhuaxue shiyan jishu jiaocheng). South China University of TechnologyPress (Huanan Ligong Daxue Press), Guangdong. pp. 54–60.