Original Article

Gene Therapy (2012) 19, 494–503; doi:10.1038/gt.2011.136; published online 6 October 2011

Novel Clostridium perfringens enterotoxin suicide gene therapy for selective treatment of claudin-3- and -4-overexpressing tumors

W Walther1, S Petkov1, O N Kuvardina1, J Aumann1, D Kobelt2, I Fichtner2, M Lemm2, J Piontek3, I E Blasig3, U Stein1 and P M Schlag1,2,4

  1. 1Experimental and Clinical Research Center, Charité University Medicine Berlin, Max-Delbrück-Center for Molecular Medicine, Berlin, Germany
  2. 2Max-Delbrück-Center for Molecular Medicine, Berlin, Germany
  3. 3Leibnitz-Institut für Molekulare Pharmakologie, Berlin, Germany
  4. 4Charité Comprehensive Cancer Center, Berlin, Germany

Correspondence: Dr W Walther, Experimental and Clinical Research Center, Charité University Medicine Berlin, Max-Delbrück-Center for Molecular Medicine, Robert-Rössle-Str. 10, 13125 Berlin, Germany. E-mail: wowalt@mdc-berlin.de

Received 25 November 2010; Revised 5 August 2011; Accepted 16 August 2011
Advance online publication 6 October 2011



Bacterial toxins are known to be effective for cancer therapy. Clostridium perfringens enterotoxin (CPE) is produced by the bacterial Clostridium type A strain. The transmembrane proteins claudin-3 and -4, often overexpressed in numerous human epithelial tumors (for example, colon, breast, pancreas, prostate and ovarian), are the targeted receptors for CPE. CPE binding to them triggers formation of membrane pore complexes leading to rapid cell death. In this study, we aimed at selective tumor cell killing by CPE gene transfer. We generated expression vectors bearing the bacterial wild-type CPE cDNA (wtCPE) or translation-optimized CPE (optCPE) cDNA for in vitro and in vivo gene therapy of claudin-3- and -4-overexpressing tumors. The CPE expression analysis at messenger RNA and protein level revealed more efficient expression of optCPE compared with wtCPE. Expression of optCPE showed rapid cytotoxic activity, hightened by CPE release as bystander effect. Cytotoxicity of up to 100% was observed 72h after gene transfer and is restricted to claudin-3-and -4-expressing tumor lines. MCF-7 and HCT116 cells with high claudin-4 expression showed dramatic sensitivity toward CPE toxicity. The claudin-negative melanoma line SKMel-5, however, was insensitive toward CPE gene transfer. The non-viral intratumoral in vivo gene transfer of optCPE led to reduced tumor growth in MCF-7 and HCT116 tumor-bearing mice compared with the vector-transfected control groups. This novel approach demonstrates that CPE gene transfer can be employed for a targeted suicide gene therapy of claudin-3- and -4-overexpressing tumors, leading to the rapid and efficient tumor cell killing in vitro and in vivo.


Clostridium perfringens enterotoxin; suicide gene; cancer; claudin



Cancer is still one major target for gene therapy due to the high incidence.1 Success for cancer gene therapy highly depends on the choice of the appropriate therapeutic gene.2, 3 Bacterial toxins are attractive candidates, which possess efficient cell killing capacity.4 Particularly for the treatment of refractory tumors, use of these toxins represents an applicable option and was tested in vitro and in vivo to treat various cancers, and have also been efficiently used in gene therapeutic applications.5, 6, 7, 8 The most extensively used toxins are diphteria toxin (Corynebacterium diphteriae) and Pseudomonas exotoxin A (Pseudomonas aeruginosa), which both inhibit protein synthesis by specific ribosylation of elongation factor-2.4, 9 The two toxins were used for the treatment of prostate, pancreatic, ovarian, lung carcinoma and glioblastoma with promising therapeutic efficacies, and exerted also therapeutic efficacy in gene therapy approaches.5, 6, 7, 8, 10, 11, 12, 13, 14 Alternatively, the pore-forming bacterial toxins streptolysin O (Streptococcus pyrogenes) and Clostridium perfringens enterotoxin (CPE) came into focus as promising cancer therapeutics.15, 16, 17, 18

CPE is produced by the anaerobic gram-positive Clostridium perfringens type A strain, known to cause food poisoning.19 CPE is a 35-kDa protein, which increases cell membrane permeability by formation of large prepore complexes. These insert into the plasma membrane, leading to the loss of osmotic equilibrium and cell death.20, 21, 22 CPE binds to a subset of the claudin family of tight junction proteins, such as claudin-3 and -4, initially defined as CPE receptors.23 The C-terminus of CPE permits this binding, whereas the N-terminus of CPE is rather associated with cytotoxicity.24, 25 CPE preferentially binds to claudin-4 and with lower affinity to claudin-3.23, 26 Claudin-3 and -4 regulate paracellular permeability, maintain epithelial cell polarity and are often overexpressed in epithelial tumors, such as colon, breast, pancreas, prostate, ovarian or endometrial cancer.27, 28, 29, 30, 31, 32 Thus, particularly these two claudins are attractive targets for the selective CPE treatment of solid tumors. Application of recombinant CPE protein leads to the dose-dependent rapid eradication of claudin-4- or claudin-3-overexpressing pancreas, breast or colon cancer cells in vitro and in vivo.17, 18, 33, 34, 35, 36, 37 The intratumoral in vivo application of recombinant CPE did not induce toxin-associated side effects, supporting its great therapeutic potential. However, these approaches require repeated, almost continuous regional or loco-regional application of recombinant CPE at doses ranging from 0.5 to 1.0μg CPE per application to achieve therapeutic effects.33, 36, 37 Furthermore, high interstitial pressure in the tumor might hinder proper distribution of externally applied CPE limiting its therapeutic action.38 Alternatively, intratumoral gene transfer of CPE-expressing vectors could significantly improve duration of toxin availability associated with optimized intratumoral distribution and accumulation. Surprisingly, to date no such gene therapy study has been performed, which employs CPE gene transfer to directly exploit the cytotoxic activity of the expressed toxin.

This is the first report describing the implementation of this approach. In this study, we cloned the CPE cDNA and optimized its sequence for human translation to express the toxin in human tumor cells. We demonstrate that CPE expression elicits rapid and selective tumor cell eradication in vitro and in vivo. Notably, CPE expression leads to the release of the toxin affecting non-transfected cells as bystander effect, which enhances toxin efficiency. The study supports the great potential of this novel suicide gene therapy, particularly for the local treatment of claudin-3- and -4-overexpressing epithelial tumors.



Claudin-3 and -4 expression in human tumor cell lines and sensitivity toward recombinant CPE

As CPE action is strictly dependent on the presence of CPE-binding claudins, like claudin-3 and -4, we tested the human cancer cell lines SKMel-5, HCT116, MCF-7 and Pac-1 for claudin expression. We found high claudin-3 and -4 expression in MCF-7 cells at messenger RNA and protein level (Figure 1a). HCT116 cells did express claudin-4 and -3 at lower level compared with MCF-7 cells, whereas Panc-1 cells did only express claudin-4 at detectable level (Figure 1a). Semiquantitative analysis of the claudin-3 protein expression revealed that in SKMel-5 cells claudin-3 expression is 0%, in HCT116 cells 84% and in Panc-1 cells 0% in relation to the highest claudin-3-expressing MCF-7 cells (set as 100%). For claudin-4 expression, the level in SKMel-5 cells was 0%, in HCT116 cells 40% and in Panc-1 cells 14% in relation to the expression in MCF-7 cells (set as 100%). The melanoma line SKMel-5 did neither express claudin-3 nor claudin-4 and was therefore used as negative control in all in vitro experiments. The testing of the four cell lines for sensitivity toward treatment with the recombinant CPE demonstrates tight correlation between claudin-3 and -4 expression and sensitivity for the toxin (Figure 1b). The claudin-3-and -4-expressing lines MCF-7 and HCT116 showed high sensitivity, starting at CPE concentrations of 0.05 to 0.1μgml−1, and developed highest, >80%, toxicity at 0.25μgml−1 CPE. The Panc-1 cells were less sensitive toward toxin action due to the much lower claudin-4 expression. The claudin-negative SKMel-5 cells were completely insensitive toward CPE.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Claudin-3 and -4 expression in the tumor cell lines SKMel-5, HCT116, MCF-7 and Panc-1 and sensitivity toward recombinant CPE. (a) Quantitative real-time RT-PCR, respective PCR products and western blot for claudin-3 (left) and claudin-4 expression (right), revealing high claudin-3 and 4 expression in MCF-7 and HCT116 cells, low claudin-4 expression in Panc-1 and no claudin-3 or -4 expression in SKMel-5 cells. Columns, mean of triplicates; bars, s.d. (b) Tumor cell sensitivities toward recombinant CPE determined by the Alamar-blue cytotoxicity assay. Recombinant CPE was added to the cells at indicated concentrations (0.05–0.25μg CPE per ml, which equals 1.4–7.0nM CPE) for 72h. Strong cytotoxicities are seen in HCT116 and MCF-7 cells, less toxicity in Panc-1 cells. Columns, mean of percentage of untreated control cells; bars, s.d. Measurements were performed in quadruplicates. Levels of significance were calculated by Students t-test; *P<0.001.

Full figure and legend (103K)

In vitro CPE expression and intracellular distribution

For CPE expression, we generated the pCpG-wtCPE vector carrying the bacterial wtCPE cDNA and the pCpG-optCPE vector carrying the translation-optimized CPE (optCPE) cDNA. The two vector constructs were transfected into the four tumor cell lines and CPE expression was analyzed. At 24h after transfection, we detected best, 4- to 40-fold improved messenger RNA and 3- to 36-fold improved protein expression of optCPE compared with wtCPE in all four lines (Figure 2a). We further determined changes in intracellular localization of the expressed optCPE protein by immunofluorescence in transfected MCF-7 and HCT116 cells with highest CPE expression (Figure 2b). We observed rapid cytoplasmic toxin accumulation 12h after transfection in both cell lines, which then changes to more membranous localization 24h after transfection. This membranous localization of CPE remained, even 48h after transfection, in dying or dead MCF-7 or HCT116 cells, in which cytoplasmic content is almost completely lost, reflected by loss of cytoplasmic Alexa 555 staining particularly in MCF-7 cells.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Analysis of CPE expression and intracellular localization in wtCPE- or optCPE-transfected tumor cells. (a) Quantitative real-time RT-PCR and western-blot of tumor cells 24h after transfection shows improved expression of the optCPE variant in all four cell lines. Semiquantitative western blot analysis (ChemiImager, Alpha Innotech, San Leonardo, CA, USA) revealed a 36-fold improved expression in SKMel-5, 3.4-fold improved expression in HCT116, 6.7-fold improved expression in MCF-7 and 12-fold improved optCPE expression in Pac-1 cells compared with wtCPE. Columns, mean of triplicates; bars, s.d. (b) Immunofluorescence of time-dependent intracellular distribution of optCPE (green) in HCT116 and MCF-7 cells. Counterstaining with Alexa 555 shows cytoplasm (red) and with DAPI the nuclei (blue). The images demonstrate initial cytoplasmic accumulation of the optCPE protein 12h after transfection, which turns into more membranous localization in the dying cells 24 to 48h after transfection. Scale bars, 20μm.

Full figure and legend (186K)

Selective cytotoxicity of CPE gene transfer and bystander effect by CPE release

The demonstration of CPE cytotoxicity after gene transfer is of primary interest. For this, we transfected the wtCPE- and optCPE-expressing vectors into the four cell lines and determined cell viability. These experiments show high cytotoxicity particularly of optCPE in MCF-7 and in HCT116 cells 72h after transfection with toxicity rates of 88% to almost 100% (Figure 3a). In Panc-1 cells, optCPE exerts up to 60% cytotoxicity. Evaluation by light microscope shows massive monolayer disruption, cell flattening and attached empty cell remnants caused by CPE gene transfer (Figure 3b). Due to the inefficient expression of wtCPE, levels of cytotoxicity are much lower in HCT116 and MCF-7 cells. The data again support correlation between claudin-4 and -3 expression and CPE toxicity. MCF-7 and HCT116 cells, which express claudin-4 and -3, are most sensitive for CPE gene transfer. Panc-1 cells with low-level claudin-4 and no claudin-3 expression, however, are less sensitive, in part also due to the lower transfection efficiency for this cell line (Figure 2a, Table 1). The claudin-3- and -4-negative SKMel-5 melanoma cells did not respond to wtCPE or optCPE gene transfer pointing to the strict selectivity of this approach.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Cytotoxicity of CPE gene transfer in tumor cells and verification of claudin specificity. (a) Cytotoxicity of CPE in wtCPE- and optCPE-transfected tumor cell lines. Alamar blue cytotoxicity assay was performed 72h after transfection. Controls are transfection reagent treated and vector-transfected cells. Columns, mean of percentage of non-treated control cells. Measurements were done in quadruplicates; bars, s.d. Level of significance was calculated by Students t-test; *Pless than or equal to0.004; **P<0.001. The assay demonstrates highest cytotoxicity for optCPE-expressing cells, which acts selectively on the claudin-3-or -4-positive cells, leaving the claudin-negative SKMel-5 cells unaffected. (b) Representative images of vector or optCPE-transfected cells 72h after transfection (scale bar, 100μm). In sensitive cells, large areas of dead cells appear. (c) Cell lines derived from SKMel-5 cells, transfected with wild-type (wtCldn3)- or mutated murine claudin-3 (mutCldn3)-expressing vectors. Claudin-3 expression was verified by western blot analysis. (d) Treatment of the wtCldn3-expressing SKMel-5 cells with recombinant CPE causes toxicity, whereas mutCld3-expressing cells remained unaffected. Controls (Co) are non-transfected SKMel-5 cells. (e) MTT cytotoxicity assay after transfection of the wtCldn3- or mutCldn3-expressing SKMel-5 cells with optCPE exerts toxicity on wtCldn3-expressing cells, mutCldn3 cells remained unaffected. Controls are non-transfected (Co) or vector-transfected (vec) SKMel-5 cells. Measurements for (d) and (e) were done in sextuplicates; bars, s.d. Level of significance for (d) and (e) was calculated by Students t-test; *P<0.05.

Full figure and legend (153K)

To analyze the claudin-specific action of CPE gene transfer in more detail, we established an isogenic pair of the claudin-3- and -4-negative SKMel-5 cells. We transfected the murine wild-type claudin-3 (wtCldn3) and a mutated claudin-3 (mutCldn3) variant into these cells and selected stably expressing clones (Figure 3c). Treatment of the wtCldn3-transfected SKMel-5 clones with 100ngml−1 recombinant CPE permitted 55% toxicity compared with untransfected control cells, whereas the mutCldn3-transfected SKMel-5 cells remained as insensitive, as the untransfected SKMel-5 cells (Figure 3d). We then transfected these clones with the optCPE-expressing vector and with the empty vector, and observed strong toxicity of 64% in the wtCldn3-expressing cells compared with the vector-transfected controls (Figure 3e). By contrast, the mutCldn3-expressing cells did not respond to the optCPE transfection. These data support that the CPE gene transfer-mediated toxicity specifically depends on claudin expression and on the CPE-binding motif (including N148 and L150) in the extracellular loop 2 of claudins.

In all these experiments, we observed a much higher CPE-mediated toxicity, as would be expected by the transfection rate of 40 to 70% (Table 1). This suggests that CPE might be released into the medium affecting also non-transfected tumor cells as bystander effect. To prove this, we determined the presence of CPE in a medium of transfected MCF-7 and HCT116 cells. We observed release of CPE into the medium 24 and 48h after transfection (Figure 4a). Quantitation of the released toxin by CPE-ELISA (enzyme-linked immunosorbent assay) confirmed its continued presence over time of up to 34ngml−1 liberated from MCF-7 cells and up to 468ngml−1 released from HCT116 cells (Table 2). To test if released CPE still has toxic activity, we added the medium collected 48h after transfection to the respective non-transfected cells (Figure 4b). The cytotoxicity assay clearly shows dramatic, up to 92%, cytotoxicity after 72h incubation in HCT116 and about 72% cytotoxicity in MCF-7 cells. Addition of the HCT116 supernatants, however, left SKMel-5 cells unaffected. This strongly supports that the bystander effect is associated with the CPE gene transfer and causes escalation in toxicity. The analysis of optCPE-transfected SKMel-5 melanoma cells revealed that these cells not only accumulate CPE with time in the cytoplasm, as shown for MCF-7 and HCT116 cells, but also release biologically active CPE into the medium (Figures 4c and d). Therefore, release of CPE into the medium by transfected cells seems to be independent of CPE cytotoxicity and of claudin-3 or -4 expression.

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Analysis of CPE release in optCPE-transfected MCF-7, HCT116 (a) and in SKMel-5 cells (c). (a) Western blot shows time-dependent CPE release from optCPE-transfected HCT116 and MCF-7 cells, which can act on non-transfected cells. (b) Analysis of optCPE-containing supernatants shown in (a) for cytotoxic activity. Incubation of non-transfected HCT116 and MCF-7 cells with supernatants taken from transfected HCT116 or MCF-7 cells 48h after transfection produced strong cytotoxicity, whereas SKMel-5 cells did not respond. The control cells were incubated with supernatants from vector-transfected HCT116 or MCF-7 cells, respectively. (c) The western blot analyses show also time-dependent release of CPE in transfected SKMel-5 cells, indicating that CPE release is independent from claudin-3 or -4 expression. (d) The MTT cytotoxicity tests of CPE-containing supernatants (marked with asterisks) on non-transfected HCT116 cells indicate toxicity of the released CPE. Columns, mean of percentage of controls, determined in triplicates in two independent experiments; bars, s.d.

Full figure and legend (63K)

Mechanism of cytotoxicity in CPE-transfected tumor cells

Due to the rapid cell killing of CPE gene transfer, analysis of the underlying death pathway was of interest. We determined the possible activation of apoptotic and necrotic pathways in optCPE-transfected MCF-7 and HCT116 cells with highest sensitivity toward CPE. Potential activation of capases 3 and 7 as apoptotic markers was analyzed 12, 24 and 48h after transfection (Figure 5a). The assay revealed no caspase 3 or 7 activation at any time in MCF-7 cells, possibly due to their rapid eradication. In contrast, late, up to 6-fold increase of caspase 3 and 7 activity was observed in HCT116 cells 24 to 48h after transfection. Analysis of caspase 1 did not reveal its activation by CPE gene transfer (Figure 5b). The FACS analyses of annexin-V, propidium iodide (PI)-stained cells however revealed 74% of PI positivity in HCT116 cells and 43% in PI positivity in MCF-7 cells 24h after transfection. This represents a 2.3- or 3-fold increase in PI labeling, respectively (Figure 5c). Annexin-V-positive cells were not detected in both cell lines. In addition, we analyzed release of lactate dehydrogenase (LDH) into the medium in optCPE-transfected MCF-7 and HCT116 cells (Figure 5d). In both cell lines, 24h after optCPE gene transfer a >2-fold increase in LDH release was detected, indicating CPE-mediated disruption of cell membranes. This coincides with the hightened release of CPE into the medium due to leakiness (Figure 4). In addition, close evaluation of CPE-transfected MCF-7 or HCT116 cells in a light microscope shows necrotic appearance of lyzed attached cell remnants with pyknotic nuclei (Figure 4e). These data suggest that CPE gene transfer preferentially initiates necrotic cell death.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Analysis of caspase 3, 7 and 1 activity, LDH release and PI labeling in optCPE-expressing MCF-7 and HCT116 cells. (a) Active caspase 3 and 7 were determined in cell culture medium at indicated time after transfection. For HCT116, increase in caspase 3 or 7 activity was detected 24 and 48h after optCPE gene transfer. In MCF-7 cells, no activation of caspase 3 or 7 was determined. (b) Active caspase 1 was determined at indicated times after transfection, given as optical density (OD). Neither for HCT116 nor MCF-7, activation of caspase 1 was determined after optCPE gene transfer. (c) Determination of PI labeling by FACScan analysis in transfected HCT116 and MCF-7 cells shows that optCPE expression leads to high proportion of PI-positive cells. (d) Analysis of LDH release in transfected MCF-7 and HCT116 cells, in which optCPE expression leads to increase in LDH release 24h after transfection. For all assays, columns are mean from quadruplicates; bars, s.d. Significance was calculated by Students t-test; *P< 0.001; **P=0.002. (e) Representative H&E-stained MCF-7 or HCT116 cells 24h after optCPE transfection and respective controls. The arrows indicate attached flattened remnants of lyzed cells with pyknotic nuclei indicating necrotic phenotypes (scale bar, 50μm).

Full figure and legend (119K)

In vivo expression of CPE and antitumoral efficacy after non-viral intratumoral CPE gene transfer

First, we analyzed intratumoral optCPE expression after combined intratumoral in vivo jet injection and electroporation gene transfer. We determined efficient and continuous intratumoral expression of CPE protein from day 1 to 5 (Figure 6a). The immunohistochemistry shows spotted areas of strong CPE expression within the tumor tissues, as shown in representative sections collected at day 1 and 2 after gene transfer (Figure 6b). These data indicate presence of CPE protein for at least 5 days after in vivo gene transfer. Therefore, the schedule for the therapeutic treatments was adjusted to repeated in vivo gene transfer applications with 7-day intervals. In CPE gene therapy experiments, optCPE-transfected MCF-7 tumors show significant (P=0.0317) reduction in tumor growth compared with the vector-transfected control group (Figure 6c). The average tumor size was reduced to 0.322±0.08cm3 of the treated group compared with an average size of 0.714±0.12cm3 of controls, representing a 55% reduction in tumor size. This is associated with appearance of great areas of tumor necrosis due to CPE action (Figure 6d). The extent of necrotic areas in comparison with the restricted areas of CPE expression (see Figure 6b) suggests similar bystander CPE effects as observed in the in vitro experiments. The in vivo experiments in HCT116 colon carcinoma-xenotransplanted mice also showed significant reduction (P=0.0317) in tumor volumes in the HCT116 tumors transfected with optCPE compared with the respective vector controls (Figure 6e). Here, the average tumor volume was reduced to 1.09±0.42cm3 of the treated group compared with an average size of 3.16±0.65cm3 of controls, representing a 64% reduction in tumor size. Evaluation of HE-stained sections of optCPE-transfected tumors revealed large necrotic areas in the tumors, as seen in the transfected MCF-7 tumors (Figure 6f). The gene transfer experiments in the HCT116 model support that CPE transfection exerts antitumoral in vivo effects in different tumor models, and point to the therapeutic potential of CPE gene therapy. In parallel, we included an additional animal group, which was treated with intratumoral injections of recombinant CPE (recCPE) at the dose of 1.0μg recCPE for six times in a 3-day interval, adapting the in vivo treatment schedules described earlier.17, 18 The application of the recCPE did also reduce the tumor volumes to an average size of 1.97±0.36cm3, representing a 37.7% reduction (P=0.1508) compared with the vector-transfected animal group (Figure 6e). Throughout the course of the CPE gene transfer experiments no systemic toxicities, such as body weight loss, diarrhea, increase in body temperature, were observed.

Figure 6.
Figure 6 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Non-viral in vivo gene transfer of optCPE in MCF-7 and HCT116 tumors by combined electroporation and jet injection for CPE expression analysis and tumor therapy. (a) Time-dependent analysis of CPE protein expression in MCF-7 tumors by western blot at indicated times. (b) Time-dependent intratumoral distribution of CPE protein expression (indicated by arrows) detected by CPE-specific immunohistochemistry in MCF-7 tumors. The images show representative sections collected at day 1 and 2 after gene transfer; tumors of day 1 and 2 correspond to those shown in (a). The right panel shows cells with typical morphology of necrosis (indicated by arrows) in the transfected tumor (day 2) at higher magnification. Vector-transfected tumor served as control. (c) Inhibition of MCF-7 xenotransplant tumor growth after optCPE in vivo gene transfer. The vector-transfected group served as control. The CPE gene transfer led to significant reduction in tumor growth (P=0.0317); level of significance was determined by using the non-parametric Mann–Whitney U-test; bars, s.e.m. (d) H&E stain shows appearance of massive necrosis in optCPE-transfected tumors (scale bar for B 100μm, for D 200μm). (e) Inhibition of HCT116 xenotransplant tumor growth after optCPE in vivo gene transfer or intratumoral injections of 1.0μg recombinant CPE (recCPE). The vector-transfected group served as control. The CPE gene transfer led to significant reduction in tumor growth (P=0.0317), whereas recCPE treatment is less effective compared with optCPE gene transfer (P=0.1508); level of significance was determined by using the non-parametric Mann–Whitney U-test; bars, s.e.m. (f) H&E stain shows again appearance of massive necrosis in the optCPE-transfected tumors (scale bar for (f) 200μm).

Full figure and legend (285K)



Toxins have shown their therapeutic efficacy for cancer treatment both as recombinant proteins or as therapeutic genes in cancer gene therapy. 5, 6, 7, 8, 10 Apart from diphteria toxin and Pseudomonas exotoxin, other toxins such as streptolysin O and CPE came into focus. The two latter toxins are both pore-forming proteins, of which streptolysin O is rather unspecific, whereas CPE possesses specificity in action.15, 16, 17 The targeted action of CPE is mediated by binding to a subset of claudins, mainly to claudin-3 and -4, which are frequently overexpressed in epithelial tumors.22, 24, 25 This binding leads to disintegration of plasma membrane in association with rapid cytolysis of treated cells.34 Several in vitro and in vivo studies demonstrated the antitumoral activity of recombinant CPE, limited by the need for frequent, repeated CPE application to achieve antitumoral effects.17, 18 As a novel alternative approach, we were aiming at CPE gene therapy for improved and prolonged toxin action at lower dose as required for external CPE application. We demonstrated for the first time that CPE can be efficiently expressed in vitro and in vivo, dramatically improved by the translation optimized optCPE. More importantly, we showed that intracellular CPE expression and accumulation leads to effective tumor cell lysis in claudin-3-and -4-positive cells, with strong preference to claudin-4-overexpressing cells, while claudin-negative cells remained unaffected.28

However, if claudin-3 was transfected into originally claudin-3-and -4-negative SKMel-5 cells, externally applied CPE and CPE transfection causes strong cytotoxicity. This supports the targeted action of CPE gene transfer. This approach is advantageous over other toxins used in cancer therapy, as they act on more ubiquitous, non-tumor-specific targets, such as E2F for diphteria toxin or membranous cholesterol for streptolysin O.4, 16 For proper tumor targeting, modification of these toxins is required, for example, by creating fusion proteins as immunotoxins.4 CPE has also been used to generate fusion proteins. Here, particularly the claudin-specific binding property of the CPE C-terminal domain was employed to target other toxins, such as diphteria toxin fragment A (DT-A), Pseudomonas aeroginosa exotoxin (PE) or TNF-α.39, 40, 41, 42, 43 These approaches exploit CPE-mediated targeting to introduce the toxic proteins into tumor cells, and were shown to be effective in vitro and in vivo. These strategies aimed rather at improvement of cytotoxicities of DT-A, PE or TNF-α, than the toxicity of CPE itself. However, in vivo frequent applications of CPE and of CPE-fusion proteins were required to achieve therapeutic effects.39, 40, 41, 43 In contrast, CPE as a sole full-length protein combines both strong target specificity and very efficient cytotoxicity, which is important for in vivo applications to achieve rapid tumor destruction.

This study demonstrated that although CPE is produced inside the transfected cells, its outside action for pore formation and cell lysis is not hampered. Similar observations were made for streptolysin O gene transfer, where intracellular toxin expression leads to perforation of cell membranes verified by LDH release.16 We also showed that intracellularly accumulated CPE causes membrane disruption associated with LDH relase in HCT116 and MCF-7 cells (Figure 4). More interestingly, we detected released biologically active CPE in the medium of transfected cells, although no signal peptide sequence was attached to CPE. Furthermore, in part, this release seems to be independent of CPE-mediated toxicity or of claudin-3 or -4 expression, as optCPE-transfected SKMel-5 cells also released CPE into the medium. This supports the concept of bystander toxicity by CPE gene transfer. This is strengthened by our observation that although our transfection affects 41 to 77% of the cells in the four tumor lines, in vitro CPE gene transfer eradicates up to almost 100% of the cells. Particularly for in vivo application of CPE gene therapy such bystander effect can be crucial for therapeutic efficacy. As we observed in vivo, our non-viral CPE gene transfer seems to affect greater than transfected areas, reflected by large proportions of necrosis in the transfected MCF-7 and HCT116 tumors (Figure 6).

The mode of transfected CPE action regarding induction of apoptotic or non-apoptotic pathways was of interest. In this context, we did not detect rapid activation of caspases 3 or 7, as would be expected for apoptosis. Comparable effects were described for streptolysin O-transfected cells.16 Only in HCT116 cells, delayed (48h post transfection) release of capases 3 and 7 was detected. This can be attributed to the late action of liberated CPE on non-transfected HCT116 cells in a bystander effect. The data rather suggest induction of necrosis after CPE gene transfer. This is supported by LDH release of CPE-mediated pore formation, lack of annexin-V but high percentage of PI-positive cells and appearance of necrotic cell morphologies (Figure 5), such as membrane rupture and attached empty cell remnants with pyknotic nuclei.44, 45 The absence of caspase 1 activation excludes pyroptosis as potential death pathway, which is often associated with membrane breakdown.46 In other reports for external application of low dose CPE induction of apoptosis was demonstrated, whereas for high dose CPE oncosis or necrosis has been described in colon carcinoma cells.47 Our in vitro and in vivo experiments support that the intracellular accumulation of CPE after transfection induces necrosis. Other in vivo studies of external therapeutic application of recombinant CPE also reported strong necrosis in the tumor tissues.33, 35, 36, 37

In conclusion, we report the successful tumor-targeted CPE gene therapy in vitro and in vivo. This approach efficiently eradicates tumor cells, and provides an attractive therapeutic option for the potential local treatment of refractory solid tumors (including unresectable tumor lesions, residual tumors or recurrences), originating from colon, mammary, pancreas or prostate cancer, which overexpress claudin-3 or -4. This might be of particular value if local control of the tumor disease by either non-viral or viral gene therapy is envisaged for possible clinical application, as numerous clinical studies have demonstrated applicability and efficacy of local cancer gene therapy.48, 49


Materials and methods

Cell cultivation

The human HCT116 colon, MCF-7 mammary carcinoma, Panc-1 pancreatic carcinoma and SKMel-5 melanoma cell line were grown in RPMI (PAA, Cölbe, Germany), 10% FCS (Biochrom, Berlin, Germany) at 37°C, 5% CO2. Identity of the cell lines was confirmed by STR DNA typing (DMSZ, Braunschweig, Germany).

Generation of CPE-expressing vectors

The 990-bp CPE cDNA was PCR amplified from Clostridium perfringens DNA (kindly provided by MR Popoff, Institut Pasteur, Paris) and cloned into the Stu I/Bgl II sites of the pCpG-mcsG2 expression vector (Invivogen, San Diego, CA, USA), resulting in the wild-type CPE (wtCPE)-expressing construct pCpG-wtCPE. For humanized translation, the CPE cDNA was codon optCPE (Entelechon GmbH, Regensburg, Germany) and cloned into the Stu I/Bgl II sites of the pCpG-mcsG2 vector, resulting in the pCpG-optCPE construct. In both constructs, a humanized Kozak consensus sequence was added to the CPE cDNA.

Transfection of human tumor cell lines

For transfection, 4 × 105 cells were seeded into 6-well plates and transfected with 2μg of the respective plasmid DNA using Fugene HD as recommended by the manufacturer (Roche Diagnostics, Mannheim, Germany). Transfection efficiency for each cell line was determined by transfection of the green fluorescent protein-expressing plasmid pEGFP-N1 (Clontech, Mountain View, CA, USA) and analysis using the FACScalibur (Becton Dickinson, San Jose, CA, USA) 48h after transfection. Number of green fluorescent protein-expressing cells was quantified in three independent experiments and given as % green fluorescent protein-positive cells (Table 1).

To generate the wtCldn3- and mutCldn3-expressing SKMel-5 cell clones as counterpart to the non-expressing wild-type SKMel-5 cells, the murine wtCldn3-expressing vector pECFP-N1cld3 and the mutCldn3-expressing vector pECFP-N1cld3mut were transfected.50 The wtCldn3 was previously shown to have strong homology to the human claudin-3 and to specifically bind CPE, whereas the mutCldn3 variant harbors a mutation in the second extracellular loop (for Asn148: N148D; Leu150: L150A), which prevents CPE binding. The mutCldn3-expressing clone served as additional control to further support specificity of toxicity after CPE gene transfer. Stably expressing cell clones were selected in G418 (Invitrogen, Carlsbad, CA, USA)-containing medium. Claudin-3 expression was verified by western blot and clones were used for further experiments.

Quantitative real-time RT-PCR

Total RNA from cells was isolated using Trizol (Invitrogen) and reverse transcribed. Each real-time PCR was done using the LightCycler480 (Roche Diagnostics). The following primers were used: for claudin-3: forward 5′-CTGCTCTGCTGCTCGTGTCC-3′; reverse 5′-TTAGACGTAGTCCTTGCGGTCGTAG-3′; for claudin-4: forward 5′-CCTCTGCCAGACCCATATAA-3′; reverse 5′-CACCGTGAGTCAGGAGATAA-3′; for wtCPE: forward 5′-GAAATCCTTGATTTAGCTGCTGC-3′; reverse 5′-AAGCTTTTGAGTCCAAGG GTATG-3′ and for optCPE: forward 5′-GCTAAGGAGGTGTTCCTCAT-3′; reverse 5′-GTGGCGTAGACCTTGTAGTA-3′. Normalization was done with the housekeeping gene glucose-6-phosphate dehydrogenase using the hG6PDH Roche Kit.

Western Blot

For western blot analysis, 25μg protein of lysates from cells or tissue cryosections were electrophorezed in 10% precast NuPAGE-gels (Invitrogen), 1h at 200V, transferred to nitrocellulose membranes (Hybond-C Extra, Amersham, Freiburg, Germany) by semidry blotting (BioRad, Munich, Germany) at 20V, 1h. Filters were blocked 1h at room temperature (RT) in TBS (50mM Tris, 150mM NaCl, pH 7.5, 5% fat-free dry milk) and washed in TBST (0.05% Tween20 in TBS), 10min at RT. Rabbit anti-CPE antibody (1:5000, Biogenesis, Poole, UK), rabbit anti-claudin-3 antibody (1:2000, Abcam, Cambridge, UK), rabbit anti-claudin-4 antibody (1:2000, Abcam) or mouse monoclonal anti-β-tubulin-IgM-antibody (1:500; BD Pharmingen, Heidelberg, Germany) was added over night at 4°C and washed in TBST. Secondary antibody HRP-labeled anti-rabbit-IgG antibody (1:10000, Promega, Madison, WI, USA) or goat anti-mouse IgM-antibody (1:5000, Sigma, Taufkirchen, Germany) was added for 1h, RT. Filters were washed in TBST and detection was done using ECL solution (Amersham) and exposed to Kodak X-Omat AR film (Kodak, Stuttgart, Germany).

Immunofluorescence and immunohistochemistry

For immunofluorescence, 5 × 104 cells were seeded into 4-well chamber slides (Thermo Fisher, Rochester, NY, USA). The following day optCPE was transfected as described. Cells were washed with phosphate-buffered saline (PBS), fixed 15min in 0.04% glutaraldehyde, permeabilized 10min with 0.5% Triton X-100 in PBS and blocked 1min with 5% bovine serum albumin in PBS at RT. Rabbit anti-CPE antibody (1:1000, Biogenesis) was added for 3h, RT, and cells were washed with TBST. Alexa 488-labeled anti-rabbit-IgG (1:1500, Invitrogen) was added as secondary antibody, 1h at RT. For staining of cytoplasm Alexa 555 (Invitrogen) and for staining of nuclei DAPI (Sigma) were used. Cells were evaluated in a Zeiss AxioObserver.Z1 fluorescence microscope (Zeiss, Jena, Germany).

For immunohistochemistry, tumor 7μm cryosections were fixed 15min in 0.04% glutaraldehyde, washed in PBS at RT, 20min blocked with 1% H2O2, washed with PBS, permeabilized by 5% Triton X-100 in PBS 10min at RT. Rabbit anti-CPE antibody (1:2000, Biogenesis) was added for 2h, RT, and sections were washed in PBS for 5min. HRP-labeled goat anti-rabbit antibody (1:500, Abcam) was added for 1h, RT, cryosections were washed in PBS, incubated with diamino-benzidine, 2min at RT, washed, counterstained for 1min with hemalum (Roth, Karlsruhe, Germany), rinsed in water, covered with glycergel (DAKO, Hamburg, Germany) and evaluated in a light microscope (Zeiss).

Alamar-blue cytotoxicity assay

For sensitivity testing of the four cell lines toward recombinant CPE (R-Biopharm, Darmstadt, Germany), 5 × 103 cells were seeded into 96-well plates, and 24h later the toxin was added at different concentrations (0.05, 0.1, 0.15, 0.2 and 0.25μg CPE per ml, which equals 1.4, 2.83, 4.24, 5.66 and 7.0nM CPE) for 72h. To determine cytotoxicity of CPE gene transfer, 1 × 105 cells of all four tumor cell lines were transfected in 24-well plates using 2μg DNA and Fugene HD as recommended by the manufacturer (Roche Diagnostics). As controls non-transfected and vector-transfected cells of the respective cell line were used.

In both settings, after 72h Alamar-blue (Serotech, Düsseldorf, Germany) was added and adsorption was measured in triplicates in a microplate reader (Tecan, Crailsheim, Germany) at 620/560nM. Values are expressed as percentage of untreated controls.

MTT cytotoxicity assay

The CPE-containing or control supernatants were harvested 48h after transfection. To test biological activity of released CPE from transfected HCT116 or MCF-7 cells, MTT cytotoxicity assay was performed. For this, 5 × 103 non-transfected HCT116 or MCF-7 cells were seeded into 96-well plates. After 24h, 50μl of supernatants from optCPE or vector (controls)-transfected HCT116 or MCF-7 cells was added to the respective non-transfected cells and incubated for 72h. MTT (3-(4,5-dimethylthiazyol-2yl)-2,5-diphenyltetrazolium bromide (Sigma, 5mgml−1) was added and absorbance was measured in triplicates at 560nm in a microplate reader (Tecan). Values are expressed as percentage of untreated controls.

CPE-specific ELISA

To quantify CPE in the supernatants 24h and 48h after optCPE transfection of HCT116 or MCF-7 cells, CPE-specific ELISA was performed. Supernatants (100μl) were added to the 96-well Ridascreen CPE-ELISA plates (R-Biopharm) and detection was performed as recommended by the manufacturer. Recombinant CPE was used as standard at serial dilutions (0.4 to 25ng CPE per ml). Measurements were done in duplicates at 450nm in a microplate reader (Tecan). Values are expressed as ng CPE per ml supernatant.

Caspase 3, 7 and caspase 1 assays

After transfection, supernatants of transfected cells were collected at indicated times (0 to 48h), and caspase 3 and 7 activity was determined using the Apo-One kit (Promega) as recommended by the manufacturer. The caspase 1 activity was determined using the Caspase-1 color assay kit (Bio Vision, Mountain View, CA, USA) as recommended, and optical densities were measured in a microplate reader (Tecan) at 405nm.

LDH release assay

After transfection, supernatants of transfected cells were collected at indicated times (0 to 48h), and LDH release was determined using the Cytotoxicity detection kit (Roche Diagnostics) as recommended by the manufacturer.

FACScan analysis of apoptosis

For analysis of possible optCPE-mediated apoptosis, cells were harvested 12, 24 and 48h after transfection and labeled using the Annexin-V fluor staining kit as recommended (Roche Diagnostics). Propidium- and Annexin-V-positive cells were quantified using the FACScalibur (Becton Dickinson).

In vivo CPE gene transfer

For establishment of subcutaneous tumors, 1 × 107 MCF-7 cells were inoculated into female NMRI:nu/nu mice (n=5 animals per group) or 1 × 107 HCT116 cells were inoculated into female NMRI:nu/nu mice (n=5 animals per group). All MCF-7 inoculated mice received an estradiol supplementation. At tumor size of 5 × 5mm, intratumoral non-viral in vivo gene transfer by combined jet injection and electroporation was performed in anesthetized animals. For this, 50μg plasmid DNA of the respective vector construct was applied by 5 jet injections (jet injector, EMS Medical Systems SA, Nyon, Switzerland) of 10μl per injection (1μg DNAμl−1 PBS) followed by electroporation of 5 pulses 200Vcm−120ms (CUY21 EDIT; Sonidel, Dublin, Ireland). The in vivo gene transfer was performed three times at an interval of 7 days. Tumor volumes, body temperature and body weight were measured at indicated time points. Animals were killed for tumor removal and further analysis. All experiments were performed in accordance with the UKCCCR guidelines and approved by the responsible local authorities (State Office of Health and Social Affairs, Berlin, Germany). For treatment of HCT116 tumors, animals received six times intratumoral injections of 1.0μg CPE (R-Biopharm) at 3-day intervals.

Statistical analysis

For statistical analyses of the in vitro experiments the Student's t-test was used, and for the analyses of the in vivo experiments the non-parametric Mann–Whitney test was used. Error values for the in vitro experiments are s.d. and for the in vivo experiments s.e.m.


Conflict of interest

The authors declare no conflict of interest.



  1. Edelstein ML, Abedi MR, Wixon J. Gene therapy clinical trials worldwide to 2007- an update. J Gene Med 2007; 9: 833–842. | Article | PubMed | ISI |
  2. Walther W, Stein U. Therapeutic genes for cancer gene therapy. Mol Biotechnol 1999; 13: 21–28. | Article | PubMed | ISI |
  3. Lo HW, Day CP, Hung MC. Cancer-specific gene therapy. Adv Genet 2005; 54: 235–255. | PubMed | ISI | CAS |
  4. Michl P, Gress TM. Bacteria and bacterial toxins as therapeutic agents for solid tumors. Current Cancer Drug Targets 2004; 4: 689–702. | Article | PubMed | ISI |
  5. Martin V, Cortes ML, deFelipe P, Farsetti A, Calcaterra NB, Izquierdo M. Cancer gene therapy by thyroid hormone-mediated expression of toxin genes. Cancer Res 2000; 60: 3218–3224. | PubMed | ISI | CAS |
  6. Lee EJ, Jameson JL. Cell-specific cre-mediated activation of the diphteria toxin gene in pituitary tumor cells: potential for cytotoxic gene therapy. Hum Gene Ther 2002; 13: 533–542. | Article | PubMed | ISI | CAS |
  7. Li Y, McCadden J, Ferrer F, Kruszewski M, Carducci M, Simons J et al. Prostate-specific expression of the diphteria toxin A chain (DT-A): studies of inducibility and specificity of expression of prostate-specific antigen promoter-driven DT-A adenoviral-mediated gene transfer. Cancer Res 2002; 62: 2576–2582. | PubMed | ISI | CAS |
  8. Zheng JY, Chen D, Chan J, Yu D, Ko E, Pang S. Regression of prostate cancer xenografts by a lentiviral vecor specifically expresing diphteria toxin A. Cancer Gene Ther 2003; 10: 764–770. | Article | PubMed | ISI | CAS |
  9. Kreitman RJ. Chimeric fusion proteins—pseudomonas exotoxin based. Curr Opin Investig Drugs 2001; 2: 1282–1293. | PubMed | CAS |
  10. Candolfi M, Xiong W, Yagiz K, Liu C, Muhammed AK, Puntel M et al. Gene therapy-mediated delivery of targeted cytotoxins for glioma therapeutics. Proc Natl Acad Sci USA 2010; 107: 20021–20026. | Article | PubMed |
  11. Laske DW, Youle RJ, Oldfield EH. Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors. Nat Med 1997; 3: 1362–1368. | Article | PubMed | ISI | CAS |
  12. Husain SR, Puri RK. Interleukin-13 receptor-directed cytotoxin for malignant glioma therapy: from bench to bedside. J Neurooncol 2003; 65: 37–48. | Article | PubMed |
  13. Zheng JY, Chen D, Chan J, Yu D, Ko E, Pang S. Regression of prostate cancer xenografts by a lentiviral vector specifically expressing diphtheria toxin A. Cancer Gene Ther 2003; 10: 764–770. | Article | PubMed | ISI | CAS |
  14. Ayesh B, Matouk I, Ohana P, Sughayer MA, Birman T, Ayesh S et al. Inhibition of tumor growth by DT-A expressed under the control of IGF2 P3 and P4 promoter sequences. Mol Ther 2003; 7: 535–541. | Article | PubMed | ISI |
  15. Bhakdi S, Bayley H, Valeva A, Walev I, Walker B, Kehoe M et al. Staphylococcal α-toxin, streptolysin O, and Escherichia coli hemolysin: prototypes of pore-forming bacterial cytolysins. Arch Microbiol 1996; 165: 73–79. | Article | PubMed | ISI | CAS |
  16. Yang WS, Park SO, Yoon AR, Yoo JY, Kim MK, Yun CO et al. Suicide cancer gene therapy using pore-forming toxin, streptolysin O. Mol Cancer Ther 2006; 5: 1610–1619. | Article | PubMed | ISI |
  17. Michl P, Buchholz M, Rolke M, Kunsch S, Löhr M, McClane B et al. Claudin-4: a new target for pancreatic cancer treatment using Clostridium perfringens enterotoxin. Gastroenterology 2001; 121: 678–684. | Article | PubMed | ISI | CAS |
  18. Kominsky SL, Vali M, Korz D, Gabig TG, Weitzman SA, Agani P et al. Clostridium perfringens enterotoxin elicits rapid and specific cytolysis of breast carcinoma cells mediated through tight junction proteins claudin 3 and 4. Am J Pathol 2004; 164: 1627–1633. | Article | PubMed | ISI | CAS |
  19. Johnson EA. Clostridial toxins as therapeutic agents: benefits of nature's most toxic proteins. Annu Rev Microbiol 1999; 53: 551–575. | Article | PubMed | ISI | CAS |
  20. Czeczulin JR, Hanna PC, McClane BA. Cloning, nucleotide sequencing and expression of the Clostridium perfringens enterotoxin gene in Escherichia coli. Infect Immun 1993; 61: 3429–3439. | PubMed | ISI |
  21. McClane BA. The complex interactions between Clostridium perfringens enterotoxin and epithelial tight junctions. Toxicon 2001; 39: 1781–1791. | Article | PubMed | ISI |
  22. Smedley III JG, Uzal FA, McClane BA. Indentification of a prepore large-complex in the mechanism of action of Clostridium perfringens enterotoxin. Infect Immun 2007; 75: 2381–2390. | Article | PubMed | ISI |
  23. Katahira J, Sugiyama H, Inoue N, Horiguchi Y, Matsuda M, Sugimoto N. Clostridium perfringens enterotoxin utilizes two structurally related membrane proteins as functional receptors in vivo. J Biol Chem 1997; 272: 26652–26658. | Article | PubMed | ISI | CAS |
  24. Kokai-Kun JF, McClane BA. Deletion analysis of the Clostridium perfringens enterotoxin. Clin Infect Dis 1997; 65: 1014–1022.
  25. Kokai-Kun JF, Benton K, Wieckowski EU, McClane BA. Identification of a Clostridium perfringens enterotoxin region required for large complex formation and cytotoxicity by random mutagenesis. Infect Immun 1999; 67: 5634–5641. | PubMed | ISI |
  26. Fujita K, Katahira J, Horiguchi Y, Sonoda N, Furuse M, Tsukita S. Clostridium perfringens enterotoxin binds to the second extracellular loop of claudin-3, a tight junction integral membrane protein. FEBS Lett 2000; 476: 258–2561. | Article | PubMed | ISI | CAS |
  27. Morita K, Furuse M, Fujimoto K, Tsukita S. Claudin multigene family encoding four-transmembrane domain protein components of tight junction strands. Proc Natl Acad Sci USA 1999; 96: 511–516. | Article | PubMed | CAS |
  28. Long H, Crean CD, Lee WH, Cummings OW, Gabig TG. Expression of Clostridium perfringens enterotoxin receptors claudin-3 and claudin-4 in prostate cancer epithelium. Cancer Res 2001; 61: 7878–7881. | PubMed | ISI | CAS |
  29. Rangel LB, Agarwal R, D′Souza T, Pizer ES, Alo PL, Lancaster WD et al. Tight junction proteins claudin-3 and claudin-4 are frequently overexpressed in ovarian cancer but not in ovarian cystadenomas. Clin Cancer Res 2003; 9: 2567–2575. | PubMed | ISI | CAS |
  30. Morin PJ. Claudin proteins in human cancer: promising new targets for diagnosis and therapy. Cancer Res 2005; 65: 9603–9606. | Article | PubMed | ISI | CAS |
  31. Soini Y. Expression of claudins 1, 2, 3, 4, 5 and 7 in various types of tumors. Histopathol 2005; 46: 551–560. | Article |
  32. Hewitt KJ, Agarwal R, Morin PJ. The claudin gene family: expression in normal and neoplastic tissues. BMC Cancer 2006; 6: 186. | Article | PubMed | CAS |
  33. Kominsky SL, Tyler B, Sosnowski J, Brady K, Doucet M, Nell D et al. Clostridium perfringens enterotoxin as a novel-targeted therapeutic for brain metastasis. Cancer Res 2007; 67: 7977–7982. | Article | PubMed | ISI | CAS |
  34. Smedley III JG, McLane BA. Fine mapping of the N-terminal cytotoxicity region of Clostridium perfringens enterotoxin by site-directed mutagenesis. Infect Immun 2004; 72: 6914–6923. | Article | PubMed | ISI |
  35. Santin AD, Cane S, Bellone S, Palmieri M, Siegel ER, Thomas M et al. Treatment of chemotherapy-resistant human ovarian cancer xenografts in C.B-17/SCID mice by intraperitoneal administration of Clostridium perfringens enterotoxin. Cancer Res 2005; 65: 4334–4342. | Article | PubMed | ISI | CAS |
  36. Santin AD, Bellone S, Siegel ER, McKenney JK, Thomas M, Roman JJ et al. Overexpression of Clostridium perfringens enterotoxin receptors claudin-3 and claudin-4 in uterine carcinomas. Clin Cancer Res 2007; 13: 3339–3346. | Article | PubMed | ISI |
  37. Santin AD, Bellone S, Marizzoni M, Palmieri M, Siegel ER, McKenney JK et al. Overexpression of claudin-3 and claudin-4 receptors in uterine serous papillary carcinoma. Cancer 2007; 109: 1312–1322. | Article | PubMed | ISI |
  38. Jain RK. Delivey of novel therapeutic agents in tumors: physiological barriers and strategies. J Natl Cancer Inst 1989; 81: 570–576. | Article | PubMed | ISI | CAS |
  39. Saeki R, Kondoh M, Kakutani H, Tsunoda S-I, Mochizuki Y, Hamakubo T et al. A novel tumor-targeted therapy using a claudin-4-targeting molecule. Mol Pharmacol 2009; 76: 918–926. | Article | PubMed | ISI |
  40. Saeki R, Kondoh M, Kakutani H, Matsuhisa K, Takahashi A, Suzuki H et al. A claudin-targeting molecule as an inhibitor of tumor metastasis. J Pharmacol Exp Ther 2010; 334: 576–582. | Article | PubMed | ISI |
  41. Kakutani H, Kondoh M, Saeki R, Fuji M, Watanabe Y, Mizoguchi H et al. Claudin-4-targeting of diphteria toxin fragment A using a C-terminal fragment of Clostridium perfringens entrotoxin. Eur J Pharm Biopharm 2010; 75: 213–217. | Article | PubMed | ISI |
  42. Yao Q, Cao S, Li C, Mengesha A, Low P, Kong B et al. Turn a diarrhoea toxin into a receptor-mediated therapy for plethora of CLDN-4-overexpressing cancers. Biochem Biophys Re Comm 2010; 398: 413–419. | Article |
  43. Yuan X, Lin X, Manorek G, Kanatani I, Cheung LH, Rosenblum MG et al. Recombinant CPE fused to tumor necrosis factor targets human ovarian cancer cell expressing the claudin-3 and claudin-4 receptors. Mol Cancer Ther 2009; 8: 1906–1915. | Article | PubMed | ISI |
  44. Rello S, Stockert JC, Moreno V, Gamez A, Pacheco M, Juarranz A et al. Morphological criteria to distinguish cell death induced by apoptotic and necrotic treatments. Apoptosis 2005; 10: 201–208. | Article | PubMed | ISI |
  45. McDonel JL. The molecular mode of action of Clostridium perfringens enterotoxin. Am J Clin Nutr 1997; 32: 210–218.
  46. Bergsbaken T, Fink SL, Cookson BT. Pyroptosis: host cell death and inflammation. Nature Rev Microbiol 2009; 7: 99–109. | Article | ISI | CAS |
  47. Chakrabarti G, Zhou X, McLane BA. Death pathways activated in Caco-2 cells by Clostridium perfringens enterotoxin. Infect Immun 2003; 71: 4260–4270. | Article | PubMed | ISI |
  48. Walther W, Stein U, Schlag PM. Local gene therapy for cancer. In: Stein U, Schlag PM (eds). Regional Therapy For Cancer. Humana Press: Totowa, 2007, pp 181–196.
  49. Walther W, Siegel R, Kobelt D, Knösel T, Dietel M, Bembenek A et al. Nonviral intratumoral jet-injection gene transfer in metastatic melanoma and breast cancer: results of phase I clinical trial. Clin Cancer Res 2008; 14: 7545–7553. | Article | PubMed | ISI |
  50. Winkler L, Gehring C, Wenzel A, Müller SL, Piehl C, Krause G et al. Molecular determinants of the interaction between Clostridium perfringens enterotoxin fragments and claudin-3. J Biol Chem 2009; 284: 18863–18872. | Article | PubMed | ISI |


We thank MR Popoff (Institut Pasteur, Paris) for kindly providing the Clostridium perfringens DNA; R Fischer and P Sander (R-Biopharm) for providing recombinant CPE and technical support. This work was supported by the Berliner Krebsgesellschaft, Grant WAFF200822 (to WW).