university of pittsburgh - cinvestav eugenio... · 2013-11-11 · general de mis estudios es...

University of Pittsburgh School of Medicine Department of Microbiology and Molecular Genetics

23 de Enero de 2009 Dra. Laura Silva Rosales Laboratorio de Interacción Planta-Virus Cinvestav, Unidad Irapuato Irapuato México.

Estimada Dra. Silva, Tuve la oportunidad de encontrar en la página de internet de la unidad Irapuato del

Cinvestav, dos plazas de trabajo vacantes para el puesto de investigador, en el Departamento de Ingeniería Genética. Apreciaré si por favor considera mi solicitud para esta plaza, ya que coincide perfectamente con mi formación científica, los intereses de mi carrera, pero sobre todo, con las expectaciones hacia el futuro de la línea de investigación que pretendo consolidar.

La investigación que realizo se enfoca en patogénesis bacteriana. El objetivo

general de mis estudios es demostrar el mecanismo de virulencia de cepas patógenas, para animales y humanos, de Clostridium perfringens. Dada la relativa rapidez con la que se lleva a cabo la infección, mi campo de investigación tiene especial énfasis en la regulación genética de la secreción de toxinas en modelos in vitro e in vivo. A mediano plazo, esta información servirá para diseñar estrategias alternativas para el tratamiento y control.

Estoy especialmente motivado por las plazas que ustedes tienen, ya que la unidad

Irapuato del Cinvestav es uno de los centros de investigación más importantes en el país y a nivel mundial. Su departamento cuenta con investigadores que trabajan con líneas de investigación afines como, interacciones hospedero-virus, interacciones planta-bacteria, mecanismos de regulación, producción de compuestos de interés farmacéuticos entre otros. Además de su planta de científicos, sus modernas instalaciones crean un excelente ambiente para realizar investigación de alto nivel.

Sin más por el momento, quedo respetuosamente a sus órdenes.

Jorge E. Vidal, PhD Post Doctoral Scientist Department of Microbiology and Molecular Genetics University of Pittsburgh, School of Medicine 200 Lothrop St., W1114 BSTWR Pittsburgh, PA 15261 Phone: 412-648-9021

Jorge Eugenio Vidal

Curriculum Vitae Present Position: Post Doctoral Scientist Department of Microbiology and Molecular Genetics University of Pittsburgh School of Medicine 200 Lothrop St., W1114 BSTWR Pittsburgh, PA 15261 Phone: 412-648-9021 FAX: 412-624-14011236 E-mail: [email protected] Citizenship: Mexico Education 2006 Ph.D. Department of Cell Biology, Cinvestav-IPN Mexico city, Mexico 2001 M. Sc. Biomedicine (Honors)

Department of Microbiology ENCB-IPN, Mexico City, Mexico. 1999 B.S. Chemistry, Pharmacology and Biology (Honors) Autonomous University of Puebla BUAP, Puebla, Mexico. Peer Reviewed Publications • Vidal, J. E. and F. Navarro-García. 2006. Efficient translocation of EspC into epithelial cells depends on enteropathogenic Escherichia coli and host cell contact. Infection and Immunity, 74:2293-2303. • Navarro-García F., Canizalez-Roman, A. Burlingame KE, Teter K and J.E. Vidal. 2007. Pet, a Non-AB Toxin, is Retrograde Transported and Translocated into Epithelial Cells. Infection and Immunity. 75:2101-2109. • F. Navarro-Garcia, A. Canizalez-Roman, J. E. Vidal and Ma. I. Salazar. 2007. Intoxication of epithelial cells by plasmid-encoded toxin requires clathrin-mediated endocytosis. Microbiology, 153: 2828-2839. • Vidal, J. E., Giono-Cerezo, S., Ribas-Aparicio, R. M., Enríquez-Rincón, F. and P. Figueroa-Arredondo. 2007. Vibrio cholerae O1 strains of different ribotypes have similar hlyA RFLP patterns but different vacuolating ability. American Journal of Infectious Diseases, 3(2):98-109.

• Vidal, J. E., A. Canizalez-Roman, Gutierrez-Jiménez J. and F. Navarro-García. 2007. Molecular pathogenesis, epidemiolgy and diagnosis of enteropatogenic Escherichia coli (EPEC). Salud Pública de México, (Public health in Mexico). 49(5):376-386. • Sayeed, S., F. A. Uzal, D. J. Fisher, J. Saputo, J. E. Vidal, M. E. Fernandez-Miyakawa, Y. Chen, P. Gupta, J. I. Rood and B. A. McClane. 2008. Beta toxin is Essential for the Virulence of Clostridium perfringens Type C Isolate CN3685 in a Rabbit Ileal Loop Model. Molecular Microbiology, 67(1):15-30. • Vidal, J. E. and F. Navarro-García. 2008. EspC translocation into epithelial cells by enteropathogenic Escherichia coli requires a concerted participation of type V and III secretion systems. Cellular Microbiology. 10:1975-1986. • Vidal, J. E., Bruce A. McClane, Juliann Saputo, Jaqueline Parker and Francisco A. Uzal. 2008. Effects of Clostridium perfringens Beta Toxin (CPB) on the Rabbit Small Intestine and Colon. Infection and Immunity. 76:4396-4404. • Vidal, J. E., Enríquez-Rincón, F., Giono-Cerezo, S., Ribas-Aparicio, R. M., and P. Figueroa-Arredondo. 2009. Culture supernatant from V. cholerae O1 ElTor isolates from different geographic origins induces cell vacuolation and cytotoxicity. Salud Pública de México, (Public health in Mexico). 51:39-47. • Vidal, J. E., Ohtani, K., Shimizu, T., and B. A. McClane. 2009. Contact with Enterocyte-like Caco-2 cells Induces Rapid Upregulation of Toxin Production by Clostridium perfringens Type C Isolates. Submitted. • Saputo, J., Vidal, J.E., Fernandez-Miyakawa, M., Sayeed, S., McClane, B.A., and Francisco A. Uzal. 2009. A mouse model for studying Clostridium perfringens type C infection. Submitted. • Vidal, J. E., Li, J., Chen, J. And Bruce A. McClane. 2009. Clostridium perfringens toxin production is regulated by the quorum sensing agr system. Manuscript under preparation. Peer Reviewed Abstracts • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. “Efecto vacuolizante del gen hlyA en cepas de diferente ribotipo de V. cholerae O1” Latinamerican Association of Microbiology, 15th Latinamerican Meeting of Microbiology. Mérida Yucatán, Mexico. Apr 9-13, 2000. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. “Vacuolating activity of hlyA genes from different ribotypes of Vibrio cholerae” American

Society for Microbiology, 102nd General Meeting May 19-23, 2002, Salt Lake City, Utah, USA. • Vidal, JE and F. Navarro-García. “Efficient internalization of EspC into epithelial cells depends on EPEC-host cell contact" American Society for Microbiology, 105th General Meeting Jun 5-9, 2005, Atlanta Georgia, USA. • Vidal, JE and F. Navarro-García. “Type III Secretion System (TTSS) Helps to Translocate EspC Autotransporter Protein from Enteropathogenic Escherichia coli (EPEC) to the Eukaryotic Cell” American Society for Microbiology, 45th ICAAC Dec 16-19, 2005, Washington DC, USA. • Navarro-García, F. and J. E. Vidal. Type III and Type V Secretion Systems Cooperation for Cytosolic Translocation of EspC from Enteropathogenic Escherichia coli, 6th international symposium on Shiga toxin (Verotoxin) producing E. coli infection, Melbourne Victoria, Australia, October 29 to November 1, 2006. • Jorge E. Vidal and Bruce A. McClane. 2007. Up-regulation of C. perfringens Beta Toxin (CPB) Transcription and Secretion in the Presence of CaCo-2 cells. Pittsburgh Bacterial Meeting 2007, Duquesne University, Pittsburgh Pennsylvania, USA. • S. Sayeed, D. Fisher, M. Miyakawa, F. A. Uzal, J. E. Vidal and B. A. McClane.

Construction of an Intron-Based Toxin Genes Knockout Mutants in Clostridium perfringens Type C Isolates. 2007 ASM General Meeting, May 21-25, Toronto, Canada. • Jorge E. Vidal and Bruce A. McClane. 2007. Up-regulation of Transcription and Secretion of C. perfringens Type C Toxins in the Presence of CaCo-2 cells. 2007 ASM General Meeting, May 21-25, Toronto, Canada. • Jorge E. Vidal, Juliann Saputo, Francisco A. Uzal and Bruce A. McClane. 2008. Pathologic effects of Clostridium perfringens beta toxin (CPB) in rabbit intestinal loops. ASM General Meeting, Jun 1-5, Boston MA, USA. • Uzal, F. A., S. Sayeed, D. J. Fisher, J. Saputo, J. E. Vidal, J. I. Rood and B. A. McClane. 2008. Beta toxin is Essential for the Virulence of Clostridium perfringens Type C Isolate in experimental caprine enterotoxemia. Anaerobe Meeting 2008, Long Beach, California, June 24-27, 2008. • Jorge E. Vidal, Kaori Ohtani, Tohru Shimizu and Bruce A. McClane. 2009. Contact with Enterocyte-like Caco-2 cells Induces Rapid Upregulation of Toxin Production by Clostridium perfringens Type C Isolates. Abstract submitted to be presented during the ASM General Meeting, May 2009, Philadelphia, USA. • Adrián Canizalez-Roman, E. R. Gonzalez-Nuñez, V. J. Picos-Cardenas, J. Zazueta-Beltran, Jorge E. Vidal, and N. Leon-Sicairos. 2009. Enteropathogenic (EPEC) and

Enteroaggregative (EAEC) Escherichia coli are the most prevalent pathotype among E. coli strains in food items at northwest of Mexico. Abstract submitted to be presented during the ASM General Meeting, May 2009, Philadelphia, USA. Awards and honors • Honorific mention in bacterial pathogenesis research. 34th Mexican Meeting of Microbiology, Cancún, QR, Mexico, Aug 27-29, 2004. “La proteína EspC de EPEC se internaliza a la célula eucariótica mediante un mecanismo de endocitosis no específica y no parece requerir de tráfico vesicular” by Jorge E. Vidal Graniel and Fernando Navarro-García. • Prize for excellence in bacterial pathogenesis research. Meeting of the Biological Science, CINVESTAV. Mexico City, Mexico. Jan, 21, 2005. “EspC una proteína secretada por EPEC se internaliza a la célula mediante un proceso inusual de pinocitosis” by Jorge E. Vidal Graniel and Fernando Navarro-García. • Awarded with a Corporate Activities Student Travel Grant to assist to the American Society for Microbiology 105th General Meeting. Jun 5-9, 2005, Atlanta Georgia, USA. • Member of the Tabasco State System of Researcher (SEI), Council of Science and Technology (CCYTET), Tabasco, Mexico. December 2005. • Awarded with a Corporate Activities Student Travel Grant to assist to the American Society for Microbiology 45th ICAAC. Dec 16-19, 2005, Washington DC, USA. • Member of the National System of Researchers (SNI) of the National Council of Science (CONACyT), Mexico. January 2007. • Awarded with an ASM General Meeting Post Doctoral Minority Travel Grant to assist to 107th ASM General Meeting, May 21-25, 2007 Toronto, Canada. • Journal cover. From our research published in Molecular Microbiology 67(1):15-30, the editor chose one figure for the cover of the January 2008 issue. • Awarded with an ASM General Meeting Post Doctoral Minority Travel Grant to assist to 108th ASM General Meeting, Jun 1-5, 2008 Boston, MA, USA. Employment history • Post Doctoral Scientist, Department of Microbiology and Molecular Genetics, University of Pittsburg School of Medicine. Dr. Bruce A. McClane. 2006-Present.

• PhD student. Cinvestav-IPN, Mexico city. Advisor Dr. Fernando Navarro-García. 2002-2006. • MSc student. ENCB-IPN, Mexico city. Advisor Dr. Silvia Giono. 1999-2001. • Laboratory Head, Laboratory of Clinical Diagnostic and Microbiological Analysis. Tabasco, Mexico 2001-2002. • Associate professor. Biochemistry. University of Tabasco, Mexico. 2001-2002. • Laboratory Assistant in the Department of Clinical Microbiology, Specialized Advisors for Clinical Laboratories. Puebla, Mexico 1996-1998.

1

Jorge E. Vidal, PhD. Current research and future plans: The Role of CPB in Type C Disease and Regulatory Molecules of Pathogenic Clostridium perfringens strains.

C. perfringens Type C isolates cause in humans enteritis necroticans, an endemic disease

in much of Southeast Asia (5). Enteritis necroticans also occurs in diabetic patients from

developed countries, whom often survive less than 48 h after the first appearance of symptoms (6,

10, 15, 20). Type C isolates also produce severe animal diseases ranging from enteritis to

enterotoxaemia (17, 18). So far, the overall mechanism of pathogenesis had remained unclear (17,

18).

Historically, the 35 kDa, trypsin-sensitive CPB has been considered the major cause of

the clinical symptoms associated with type C disease (12, 13). During the last few years we

experimentally confirmed the pathologic role of CPB, and found that the mechanism of

pathogenesis involves early cross-talk between C. perfringens and host cells, to rapidly stimulate

toxin production. We first demonstrated that CPB is the main lethal toxin present in the

supernatant of several type C isolates, when injected in the mouse tail vein (3). Then, we

constructed a series of C. perfringens toxin mutants in the type C isolate CN3685 (cpb+, cpa+ and

pfoA+), using our modified Targetron® technology (2). With those mutants, we showed that

CPB, but not CPA or PFO, is necessary and sufficient for type C isolate CN3685 to induce

intestinal damage in rabbit ileal loops (14).

To additionally confirm the pathogenic role of CPB, we purified CPB (∼95%

homogeneity) and it was injected in different segments of the rabbit intestine (21). In this

research we described that CPB is a very potent and fast-acting toxin in the small intestine. One

µg of purified CPB caused severe intestinal damage in the ileum and jejunum, in less than 1 h

and moderate damage in the duodenum. We next developed two mouse challenge models to test,

during an experimental infection, C. perfringens type C lethality (Saputo, J., Vidal, J.E., et al.

2009, submitted to Infection and Immunity). Clinical signs and lethality of type C infection were

reproduced by inoculating the bacteria, directly in the duodenum or by intragastric gavage. The

molecular basis of the CPB-induced intestinal damage and lethality are under study.

Toxin production is the hallmark of C. perfringens virulence (7, 11). However, no

research had been conducted to explore toxin production in vivo. We recently demonstrated that

the presence of enterocytes induces upregulation of CPB, CPA, PFO and CPB2 secretion. This

2

phenomenon occurs in different pathogenic C. perfringens strains and it is triggered by different

host cells (Vidal, J.E., et al. 2009, submitted to Cellular Microbiology).

The upregulated toxin production was shown to be mediated at the transcriptional level.

The presence of host cells induced the early transcription, as measured by RT-PCR, of type C

toxin genes (cpb, pfoA, cpa, cpb2). We experimentally demonstrated that C. perfringens host-

cell contact is necessary to induce upregulated toxin secretion. Once C. perfringens sense host-

cells, the VirS/VirR two-component regulatory system (TCRS) regulates the in vivo production

of CPB and PFO in the type C isolate CN3685. A virR mutant or the wild type strain

transformed with a virR antisense plasmid, both were unable to early secrete CPB or PFO in the

presence of Caco-2 cells (Vidal, J.E., et al. 2009, submitted to Cellular Microbiology). We are

now assessing, with the collaboration of Dr. Francisco A. Uzal in UC-Davis, the contribution of

the VirS/VirR system in animal models of type C infection.

The signaling molecule that activates the VirS/VirR TCRS when C. perfringens

type C strains infect cell culture is under research. Since in strain 13 (C. perfringens type A) the

luxS-controlled quorum sensing mechanism regulates in vitro virulence factors (9), we generated

a luxS mutant in the type C isolate CN3685. The secretion of toxins and transcription of toxin

genes in our mutant was similar to the WT strain, indicating that the luxS-controlled quorum

sensing mechanism does not regulate type C toxin production in vitro or in vivo. More recently

we developed a Tn5-based mutagenesis technique for C. perfringens strains. The technique has

proved to generate randomly mutations, allowing the Tn-5 transposon inserts in different sites of

the chromosome. Using this mutagenesis approach, and phenotypic assays that screen for clones

carrying mutations in regulatory molecules, we have successfully characterized the quorum

sensing signaling molecule that activates the VirS/VirR TCRS in C. perfringens type A strain 13

(Vidal, J.E., et al. 2009. manuscript under preparation). Current efforts are underway to

demonstrate if the same quorum sensing molecule activates the VirS/VirR system during type C

infection in vivo.

3

Future plans

1. Only limited information is currently available regarding the in vivo expression of C.

perfringens type C genes. To improve our understanding of type C disease, microarray

analyses will be conducted with RNA extracted from C. perfringens infecting either cell

cultures or rabbits (ileal loops), with the collaboration of Dr. Tohru Shimizu’s research

group in Japan. Dr. Shimizu´s group released the first genome sequence of C. perfringens

and they are the world’s experts in C. perfringens microarray technology.

2. Define which host cell membrane protein(s) and C. perfringens molecule(s) participate in

the early cross-talk inducing lethal toxins upregulation. This information can be used as

potential therapeutic targets in the future, given that infected patients often do not

respond to antibiotic therapy and the prognostic of type C infection is unfavorable.

3. We generated, using our recently developed Tn5-based mutagenesis technique, a library

with more than 50 type C-derived mutants (randomly mutations were confirmed by

southern blot and sequencing). These mutants were unable to secreted CPB. Given the

importance of CPB during type C disease and the lack of information about its secretion

mechanism, this library will be used to explore proteins involved in the secretion

mechanism and other regulatory molecules not yet characterized.

4. Investigate the mechanism by which C. perfringens infection targets CPB directly to the

host-cell membrane. Characterize the CPB receptor on Caco-2 cell cultures. Define the

distribution of the CPB receptor in diverse organs of its natural host and the contribution

of this receptor in type C disease.

5. We found that some C. perfringens type C and type A isolates (gas gangrene producer

strains) attach the Caco-2 cells in a specific pattern around the cell edge (Vidal, J.E., and

B.A. McClane, unpublished observations). Since there is little information in the

literature and adherence might be important for C. perfringens pathogenesis, it will be

interesting to find out what molecule(s) mediates this striking adhesion phenotype.

4

References

1. Amimoto, K., T. Noro, E. Oishi, and M. Shimizu. 2007. A novel toxin homologous to large clostridial cytotoxins found in culture supernatant of Clostridium perfringens type C. Microbiology 153:1198-1206.

2. Chen, Y., B. A. McClane, D. J. Fisher, J. I. Rood, and P. Gupta. 2005. Construction of an alpha toxin gene knockout mutant of Clostridium perfringens type A by use of a mobile group II intron. Appl Environ Microbiol 71:7542-7547.

3. Fisher, D. J., M. E. Fernandez-Miyakawa, S. Sayeed, R. Poon, V. Adams, J. I. Rood, F. A. Uzal, and B. A. McClane. 2006. Dissecting the contributions of Clostridium perfringens type C toxins to lethality in the mouse intravenous injection model. Infect Immun 74:5200-5210.

4. Gui, L., C. Subramony, J. Fratkin, and M. D. Hughson. 2002. Fatal enteritis necroticans (pigbel) in a diabetic adult. Mod Pathol 15:66-70.

5. Johnson, S., and D. N. Gerding. 1997. Enterotoxemic infections. In The Clostridia. Molecular Biology and Pathogenesis. pp-117-140. edited by Rood, J.I., McClane, B.A., Songer, J.G., R.W. Titball, Academic press, London, UK.

6. Matsuda, T., Y. Okada, E. Inagi, Y. Tanabe, Y. Shimizu, K. Nagashima, J. Sakurai, M. Nagahama, and S. Tanaka. 2007. Enteritis necroticans 'pigbel' in a Japanese diabetic adult. Pathol Int 57:622-626.

7. McClane, B. A., F.A. Uzal, M. Fernandez-Miyakawa, D. Lyerly and T.D. Wilkins. 2004. The Enterotoxigenic Clostridia., p. 698-752. In S. F. M. Dworkin, E. Rosenburg, K.F. Schleifer, and E. Stackebrandt. (ed.), The Prokaryotes., vol. 4. Springer-Verlag., New York, NY. .

8. Nagahama, M., S. Hayashi, S. Morimitsu, and J. Sakurai. 2003. Biological activities and pore formation of Clostridium perfringens beta toxin in HL 60 cells. J Biol Chem 278:36934-36941.

9. Ohtani, K., H. Hayashi, and T. Shimizu. 2002. The luxS gene is involved in cell-cell signalling for toxin production in Clostridium perfringens. Mol Microbiol 44:171-179.

10. Petrillo, T. M., C. M. Beck-Sague, J. G. Songer, C. Abramowsky, J. D. Fortenberry, L. Meacham, A. G. Dean, H. Lee, D. M. Bueschel, and S. R. Nesheim. 2000. Enteritis necroticans (pigbel) in a diabetic child. N Engl J Med 342:1250-1253.

11. Rood, J. I. 1998. Virulence genes of Clostridium perfringens. Annu Rev Microbiol 52:333-360.

12. Sakurai, J., and C. L. Duncan. 1977. Purification of beta-toxin from Clostridium perfringens type C. Infect Immun 18:741-745.

13. Sakurai, J., and C. L. Duncan. 1978. Some properties of beta-toxin produced by Clostridium perfringens type C. Infect Immun 21:678-680.

14. Sayeed, S., F. A. Uzal, D. J. Fisher, J. Saputo, J. E. Vidal, Y. Chen, P. Gupta, J. I. Rood, and B. A. McClane. 2008. Beta toxin is essential for the intestinal virulence of Clostridium perfringens type C disease isolate CN3685 in a rabbit ileal loop model. Mol Microbiol 67:15-30.

15. Severin, W. P., A. A. de la Fuente, and M. F. Stringer. 1984. Clostridium perfringens type C causing necrotising enteritis. J Clin Pathol 37:942-944.

16. Smedley, J. G., 3rd, D. J. Fisher, S. Sayeed, G. Chakrabarti, and B. A. McClane. 2004. The enteric toxins of Clostridium perfringens. Rev Physiol Biochem Pharmacol 152:183-204.

5

17. Songer, J. G. 1996. Clostridial enteric diseases of domestic animals. Clin Microbiol Rev 9:216-234.

18. Songer, J. G., and F. A. Uzal. 2005. Clostridial enteric infections in pigs. J Vet Diagn Invest 17:528-536.

19. Steinthorsdottir, V., H. Halldorsson, and O. S. Andresson. 2000. Clostridium perfringens beta-toxin forms multimeric transmembrane pores in human endothelial cells. Microb Pathog 28:45-50.

20. Tonnellier, M., E. Maury, J. Guglielminotti, and G. Offenstadt. 2001. A fatal sandwich. Lancet Infect Dis 1:202.

21. Vidal, J. E., B. A. McClane, J. Saputo, J. Parker, and F. A. Uzal. 2008. Effects of Clostridium perfringens beta-toxin on the rabbit small intestine and colon. Infect Immun 76:4396-4404.

22. Vidal, J. E., and F. Navarro-Garcia. 2006. Efficient translocation of EspC into epithelial cells depends on enteropathogenic Escherichia coli and host cell contact. Infect Immun 74:2293-2303.

23. Vidal, J. E., and F. Navarro-Garcia. 2008. EspC translocation into epithelial cells by enteropathogenic Escherichia coli requires a concerted participation of type V and III secretion systems. Cell Microbiol 10:1975-1986.

1

Curriculum vitae

Jorge Eugenio Vidal Graniel Doctor en Ciencias

Edad: 33 años Fecha de Nacimiento: 31 de Mayo de 1975 Lugar de Nacimiento: Comalcalco, Tabasco, México. Estado civil: Casado Dirección actual: 1820 Pioneer ave. Pittsburgh PA, USA. 15226. Dirección permanente en México: Jacarandas No.152, Fraccionamiento Real del Angel, CP. 86153, Villahermosa Tabasco, México. Teléfono: (01) 9933-517243 email: [email protected] (trabajo); [email protected] (personal) Estudios profesionales • Licenciatura en Químico Farmacobiólogo (QFB): 1993-1998. Facultad de Ciencias Químicas de la Benemérita Universidad Autónoma de Puebla (BUAP), Puebla, Puebla. Titulación profesional automática por excelencia académica (promedio general de estudios). • Servicio Social Profesional: 1998-1999. Centro de Investigaciones Microbiológicas de la BUAP. Tutora Dra. Elsa Castañeda Roldan y Dra. Silvia Giono. Estancia especial en el Departamento de Bacteriología Médica de la ENCB el IPN y en el Hospital del IMSS No. 27 de Tlatelolco en el D.F de Julio a Diciembre de 1998. Posgrado • Maestro en Ciencias Químico Biológicas con Mención Honorífica (1999-2001). Posgrado con la especialidad en Biomedicina, área de Microbiología en la Escuela Nacional de Ciencias Biológicas del Instituto Politécnico Nacional (IPN). Tesis: Expresión del gen de la hemolisina vacuolizante en cepas de Vibrio cholerae O1 de diferente ribotipo. • Doctor en Ciencias con la especialidad en Biología Celular (2002-2006). Departamento de Biología Celular del Centro de Investigación y Estudios Avanzados del IPN (Cinvestav). Tesis: Caracterización del mecanismo de internalización de la proteína autotransportadora EspC de Escherichia coli enteropatógena (EPEC) en células epiteliales. • Post Doctoral Scientist (2006-A la fecha), Departamento de Microbiología y Genética Molecular, Escuela de Medicina de la Universidad de Pittsburgh, Pittsburgh Pensilvania, Estados Unidos de América. 200 Lothrop St., Biomedical Science Tower (BSTWR) W1114, Pittsburgh, PA 15261. Teléfono: +412-648-9021. FAX: +412-624-1401.

2

Reconocimientos y premios: • Primer lugar en conocimientos de Microbiología Clínica. Facultad de Ciencias Químicas de la Benemérita Universidad Autónoma de Puebla (BUAP). Diciembre de 1996. • Mención honorífica en la obtención del grado de Maestro en Ciencias, por la excelencia en el desarrollo académico y trabajo experimental de tesis, ENCB-IPN, Septiembre de 2001. • Reconocimiento por parte del Programa de Evaluación de la Calidad entre Laboratorios (PACAL). Participación activa en el programa de calidad en el área de bacteriología, parasitología, química clínica y hematología. 2001, periodo en el cuál ocupé el cargo de jefe de laboratorio. • Acreditación de la calidad y distinción por parte del Programa de Evaluación de la Calidad entre Laboratorios (PACAL) del IPN a laboratorios Graniel en el año 2001. Por haber obtenido excelencia en la calidad analítica del área de bacteriología. Periodo en el cuál ocupé el cargo de jefe de laboratorio. • Mención honorífica por el trabajo titulado “EspC, una proteína secretada por EPEC se internaliza a la célula mediante un proceso inusual de pinocitosis" presentado durante los trabajos del XXXIV Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología A.C, Agosto de 2004. • Miembro del Sistema Estatal de Investigadores de la comunidad científica de Tabasco. Nombramiento otorgado por el Consejo de Ciencia y Tecnología del Estado de Tabasco (CCYTET), distinción obtenida a partir del año 2005. • Premio a la investigación en el 2º Congreso del área biológica del CINVESTAV-IPN al trabajo titulado “EspC, una proteína secretada por EPEC se internaliza a la célula mediante un proceso inusual de pinocitosis" autores Jorge E. Vidal y Fernando Navarro-García. 21 de Enero de 2005. • Galardonado con un Corporate Activities Student Travel Grant, otorgado por la American Society for Microbiology (ASM) para asistir a presentar el trabajo de investigación “Efficient Internalization of EspC into epithelial cells depends on bacteria-host cell contact” en el 105th General Meeting celebrado en Atlanta Georgia, USA, del 5 al 9 de Junio de 2005. • Galardonado con un Corporate Activities Student Travel Grant otorgado por la American Society for Microbiology (ASM) para asistir a presentar el trabajo de investigación “Type III Secretion System (TTSS) Helps to Translocate EspC Autotransporter Protein from Enteropathogenic Escherichia coli (EPEC) to the Eukaryotic Cell” en el 45th Interscience Conference on Antimicrobial Agents and Chemoterapy (ICAAC),Washington D.C. USA, del 16 al 19 de Diciembre de 2005.

3

• Galardonado con un ASM General Meeting Minority Post Doctoral Travel Grant otorgado por la American Society for Microbiology (ASM) para asistir al 107th General Meeting celebrado en Toronto, Canadá del 21 al 25 de Junio de 2007. • Miembro del Sistema Nacional de Investigadores (SNI) de México. Nombramiento otorgado por el Consejo Nacional de Ciencia y Tecnología (CONACyT), Agosto, 2006. • Galardonado con un ASM General Meeting Minority Post Doctoral Travel Grant otorgado por la American Society for Microbiology (ASM) para asistir al 108th General Meeting celebrado en Boston MA, USA del 01 al 05 de Junio de 2008. Artículos científicos y de divulgación: • Vidal-Graniel, J. E. 2003. Escherichia coli enteropatógena (EPEC): una causa frecuente de diarrea infantil. Revista Salud en Tabasco. 9(1):188-193. • Vidal-Graniel, J. E.. 2003. Bacterias patógenas y ser humano: la importancia de la Virulencia Bacteriana. Revista Diálogos del Consejo de Ciencia y Tecnología de Tabasco. Vol. 13:8-14. • Vidal-Graniel, J.E. y J. Gutiérrez Jiménez. 2004. La clonación terapéutica: una nueva herramienta para la medicina. Revista Diálogos del Consejo de Ciencia y Tecnología de Tabasco. Vol.14:22-28. • Vidal, J. E. and F. Navarro-García. 2006. Efficient translocation of EspC into epithelial cells depends on enteropathogenic Escherichia coli and host cell contact. Infection and Immunity, 74:2293-2303. • Navarro-García F., Canizalez-Roman, A. Burlingame KE, Teter K and J.E. Vidal. 2007. Pet, a Non-AB Toxin, is Retrograde Transported and Translocated into Epithelial Cells. Infection and Immunity. 75:2101-2109. • F. Navarro-Garcia, A. Canizalez-Roman, J. E. Vidal and Ma. I. Salazar. 2007. Intoxication of epithelial cells by plasmid-encoded toxin requires clathrin-mediated endocytosis. Microbiology, 153: 2828-2839. • Vidal, J. E., Giono-Cerezo, S., Ribas-Aparicio, R. M., Enríquez-Rincón, F. and P. Figueroa-Arredondo. 2007. Vibrio cholerae O1 strains of different ribotypes have similar hlyA RFLP patterns but different vacuolating ability. American Journal of Infectious Diseases, 3(2):98-109. • Vidal, J. E., A. Canizalez-Roman, Gutierrez-Jiménez J. and F. Navarro-García. 2007. Molecular pathogenesis, epidemiolgy and diagnosis of enteropatogenic Escherichia coli (EPEC). Salud Pública de México, 49(5):376-386.

4

• Sayeed, S., F. A. Uzal, D. J. Fisher, J. Saputo, J. E. Vidal, M. E. Fernandez-Miyakawa, Y. Chen, P. Gupta, J. I. Rood and B. A. McClane. 2008. Beta toxin is Essential for the Virulence of Clostridium perfringens Type C Isolate CN3685 in a Rabbit Ileal Loop Model. Molecular Microbiology, 67(1):15-30. • Vidal, J. E. and F. Navarro-García. 2008. EspC translocation into epithelial cells by enteropathogenic Escherichia coli requires a concerted participation of type V and III secretion systems. Cellular Microbiology. 10:1975-1986. • Vidal, J. E., Bruce A. McClane, Juliann Saputo, Jaqueline Parker and Francisco A. Uzal. 2008. Effects of Clostridium perfringens Beta Toxin (CPB) on the Rabbit Small Intestine and Colon. Infection and Immunity. 76:4396-4404. • Vidal, J. E., Enríquez-Rincón, F., Giono-Cerezo, S., Ribas-Aparicio, R. M., and P. Figueroa-Arredondo. 2009. Culture supernatant from V. cholerae O1 ElTor isolates from different geographic origins induces cell vacuolation and cytotoxicity. Salud Pública de México, (Public health in Mexico). 51:39-47. • Vidal, J. E., Ohtani, K., Shimizu, T., and B. A. McClane. 2009. Contact with Enterocyte-like Caco-2 cells Induces Rapid Upregulation of Toxin Production by Clostridium perfringens Type C Isolates. Enviado a Cellular Microbiology. • Saputo, J., Vidal, J.E., Fernandez-Miyakawa, M., Sayeed, S., McClane, B.A., and Francisco A. Uzal. 2009. A mouse model for studying Clostridium perfringens type C infection. Enviado a Infection and Immunity. • Vidal, J. E., Li, J., Chen, J. And Bruce A. McClane. 2009. Clostridium perfringens toxin production is regulated by the quorum sensing agr system. Manuscript under preparation. Asistencia a congresos nacionales • Congreso Nacional de Ciencias farmacéuticas, organizado por la Asociación Farmacéutica Mexicana en la cd. de Cancún, Quintana Roo los días del 30 de Noviembre al 4 de Diciembre de 1997. • Congreso Nacional de Biología Molecular en Medicina, organizado por la Asociación Mexicana de Biología Molecular en Medicina, en la cd. de Guanajuato, Guanajuato los días 14 al 17 de Noviembre de 1998. • Participación como congresista durante los trabajos del XXII Congreso Nacional de Química Clínica de la Asociación Mexicana de Bioquímica Clínica los días 1, 2 y 3 de Marzo de 1999, en México D.F.

5

• Congresista dentro de las actividades del XXX Congreso Nacional de Microbiología de la Asociación Mexicana de Microbiología, realizado en Oaxtepec, Morelos los días de 20 al 23 de Junio de 1999. • Asistencia a los trabajos del 1er Congreso Nacional de Medicina tropical organizado por la Sociedad Mexicana de Medicina Tropical A.C. en la cd. De Cholula , Puebla los días del 7 al 9 de Octubre de 1999. • Asistencia a la IV reunión Nacional sobre Conservación de Cepas y Colecciones Microbianas organizado por el Colegio de Químicos Bacteriólogos Parasitólogos en la cd. de México D.F. los días del 8 al 11 de Noviembre de 1999. • Participación durante los trabajos del XXIII Congreso Nacional de Química Clínica organizado por la Asociación Mexicana de Bioquímica Clínica en la cd. De México D.F. los días 8 al 10 de Marzo del 2000. • Participación como congresista en el XXXI Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología y el centro de Investigaciones Regionales Hideyo Noguchi, celebrado en la cd. de Mérida Yucatán los días 9 al 13 de Abril del 2000. • Asistente del XXXII Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología y la Universidad de Guanajuato, evento realizado del 3 al 5 de Abril del 2001 en la cd. De Guanajuato, Guanajuato. • Participación durante los trabajos del XXXIV Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología A.C. en la cd. de CanCun, QuintanaRoo, llevado a cabo del 27 al 29 de Agosto del 2004. Congresos internacionales • XV congreso Latinoamericano de Microbiología, organizado por la Asociación Latinoamericana y Mexicana de Microbiología y el centro de Investigaciones Regionales Hideyo Noguchi. Mérida Yucatán del 9 al 13 de Abril del 2000. • Congresista en el 105th General Meeting organizado por la American Society for Microbiology (ASM) celebrado en Atlanta Georgia, USA, del 5 al 9 de Junio de 2005. • Asistencia al 45th ICAAC organizado por la American Society for Microbiology (ASM), celebrado en la cd de Washington DC, USA los días del 16 al 19 de Diciembre de 2005. • Congresista en el 107th General Meeting organizado por la American Society for Microbiology (ASM) celebrado en Toronto Canadá del 21 al 25 de Mayo, 2007.

6

• 108th General Meeting organizado por la American Society for Microbiology (ASM) celebrado en Boston MA, USA del 01 al 05 de Junio, 2008. Trabajos de investigación presentados en congresos • Jorge E. Vidal y Silvia Giono Cerezo. "Evaluación externa de bacteriología del hospital IMSSS No. 27". Jornadas estudiantiles de la carrera de químico bacteriólogo y parasicólogo (QBP) en la Escuela Nacional de Ciencias Biológicas del IPN, 1° de Diciembre de 1998. • Jorge E. Vidal y Silvia Giono Cerezo. Control externo de bacterias Gram negativas en un laboratorio clínico. XXX congreso nacional de Microbiología de la Asociación Mexicana de Microbiología, realizado en Oaxtepec, Morelos los días del 20 al 23 de Junio de 1999. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. Fenotipo vacuolizante en cepas de diferente origen de Vibrio cholerae O1. 1er Congreso Nacional de Medicina tropical organizado por la Sociedad Mexicana de Medicina Tropical en Cholula , Puebla los días del 7 al 9 de Octubre de 1999. • Jorge E. Vidal, Rodolfo Doval y Silvia Giono. Conservación de Células Vero para Ensayos de Adherencia y Citotoxicidad Bacteriana. IV reunión Nacional sobre conservación de cepas y colecciones microbianas organizado por el Colegio de Químicos Bacteriólogos Parasitólogos en México D.F. los días del 8 al 11 de Noviembre de 1999. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. Patrones de Restricción del gen hlyA en cepas de V. cholerae O1. XXIII Congreso Nacional de Química Clínica organizado por la Asociación Mexicana de Bioquímica Clínica en México D.F. los días del 8 al 10 de Marzo del 2000. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. Efecto vacuolizante del gen hlyA en cepas de diferente ribotipo de V. cholerae O1. En el marco del XV Congreso Latinoamericano de Microbiología y XXXI Congreso Nacional de Microbiología, celebrado en Mérida Yucatán los días del 9 al 13 de Abril del 2000. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. Variabilidad genética de la hemolisina vacuolizante de Vibrio cholerae O1 en cepas de diferente ribotipo. En el marco del XXXII Congreso Nacional de Microbiología realizado en Guanajuato, Guanajuato, Abril de 2001. • Vidal, JE, Giono-Cerezo, S., Enríquez-Rincón, F. and P. Figueroa-Arredondo. Vacuolating activity of hlyA genes from different ribotypes of Vibrio cholerae. Presentado

7

en el 102nd General Meeting de la American Society for Microbiology, Salt Lake City, UTA, USA del 29 al 23 de mayo del 2002. • Jorge E. Vidal y Fernando Navarro-García. EspC, una proteína secretada por EPEC se internaliza a la célula mediante un proceso inusual de pinocitosis" presentado en la modalidad de cartel durante el XXXIV Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología A.C. en CanCun, QuintanaRoo, del 27 al 29 de Agosto del 2004. • Javier Gutiérrez-Jiménez, Jorge E. Vidal María E. Mejía Albarrán, Jesús Reséndiz Sánchez, Adolfo Pérez Miravete y Fernando Navarro-García. Prevalencia de patotipos de Escherichia coli diarreogénica en niños del Hospital Infantil de México. XXXIV Congreso Nacional de Microbiología, organizado por la Asociación Mexicana de Microbiología A.C. en CanCun, QuintanaRoo, llevado a cabo del 27 al 29 de Agosto del 2004. • Jorge E. Vidal y Fernando Navarro-García. Efficient internalization of EspC into epithelial cells depends on EPEC-host cell contact. Presentado durante el “105th general meeting of the American Society for Microbiology, by June 5-9, 2005”. • Jorge E. Vidal y Fernando Navarro-García. Type III Secretion System (TTSS) Helps to Translocate EspC Autotransporter Protein from Enteropathogenic Escherichia coli (EPEC) to the Eukaryotic Cell. Presentado en el 45th Interscience Conference on Antimicrobial Agents and Chemoterapy (ICAAC) 2005 celebrado en Washington D.C. USA, del 16 al 19 de Diciembre de 2005. • Navarro-García, F. and J. E. Vidal. Type III and Type V Secretion Systems Cooperation for Cytosolic Translocation of EspC from Enteropathogenic Escherichia coli, 6th international symposium on Shiga toxin (Verotoxin) producing E. coli infection, Melbourne Victoria, Australia, del 29 de Octubre al 1ro de Noviembre de 2006. • Jorge E. Vidal and Bruce A. McClane. 2007. Up-regulation of C. perfringens Beta Toxin (CPB) Transcription and Secretion in the Presence of CaCo-2 cells. Pittsburgh Bacterial Meeting 2007, Duquesne University, Pittsburgh Pennsylvania, USA. • S. Sayeed, D. Fisher, M. Miyakawa, F. Uzal, J. E. Vidal and M. Bruce. Construction of an Intron-Based Toxin Genes Knockout Mutants in Clostridium perfringens Type C Isolates. 2007 ASM General Meeting, May 21-25, Toronto, Canada. • Jorge E. Vidal and Bruce A. McClane. 2007. Up-regulation of Transcription and Secretion of C. perfringens Type C Toxins in the Presence of CaCo-2 cells. 2007 ASM General Meeting, May 21-25, Toronto, Canada. • Uzal, F. A., S. Sayeed, D. J. Fisher, J. Saputo, J. E. Vidal, J. I. Rood and B. A. McClane. 2008. Beta toxin is Essential for the Virulence of Clostridium perfringens Type C Isolate in experimental caprine enterotoxemia. Anaerobe Meeting 2008, Long Beach, California, June 24-27, 2008..

8

• Jorge E. Vidal, Juliann Saputo, Francisco A. Uzal and Bruce A. McClane. 2008. Pathologic effects of Clostridium perfringens beta toxin (CPB) in rabbit intestinal loops. 3rd Annual meeting of the University of Pittsburgh Post Doctoral Association. Pittsburgh PA, USA, May 7, 2008. • Jorge E. Vidal, Juliann Saputo, Francisco A. Uzal and Bruce A. McClane. 2008. Pathologic effects of Clostridium perfringens beta toxin (CPB) in rabbit intestinal loops. ASM General Meeting, Jun 1-5, Boston MA, USA. • Jorge E. Vidal, Kaori Ohtani, Tohru Shimizu and Bruce A. McClane. 2009. Contact with Enterocyte-like Caco-2 cells Induces Rapid Upregulation of Toxin Production by Clostridium perfringens Type C Isolates. Abstract submitted to be presented during the ASM General Meeting, May 2009, Philadelphia, USA. • Adrián Canizalez-Roman, E. R. Gonzalez-Nuñez, V. J. Picos-Cardenas, J. Zazueta-Beltran, Jorge E. Vidal, and N. Leon-Sicairos. 2009. Enteropathogenic (EPEC) and Enteroaggregative (EAEC) Escherichia coli are the most prevalent pathotype among E. coli strains in food items at northwest of Mexico. Abstract submitted to be presented during the ASM General Meeting, May 2009, Philadelphia, USA. Docencia y actividades de difusión: • Profesor del curso “Bacteriología Médica y el Diagnóstico de Laboratorio” desarrollado dentro de las instalaciones de la clínica del IMSS No. 27 de Tlatelolco, en México D.F. de Junio a Diciembre de 1998. • Profesor instructor en el taller de “Metodología de Vanguardia en Biología Celular” que se realizó en el departamento de Biología Celular del Centro de Investigación y Estudios Avanzados del IPN, en México D. F. los días del 22 al 26 de Noviembre de 1999. • Profesor instructor en el taller de “Metodología de Vanguardia en Biología Celular” que se realizó en el departamento de Biología Celular del Centro de Investigación y Estudios Avanzados del IPN, en México D. F. los días del 24 al 28 de Enero del 2000. • Ponente con el tema “La toxina vacuolizante de “Vibrio cholerae” en la asamblea mensual del colegio de Químicos de Tabasco. Febrero 23 del 2001 Villahermosa Tabasco, México. • Participación como ponente del Tema “Patogénesis molecular del cólera” en las Jornadas de salud efectuadas por la Secretaría de Salud del estado de Tabasco, los días del 1 al 3 de Marzo del 2001. • Profesor de la cátedra de Bioquímica en la División Académica de Ciencias de la Salud (DACS) de la Universidad Juárez Autónoma de Tabasco (UJAT), durante el periodo comprendido de Agosto de 2001 a julio de 2002.

9

• Profesor del curso Fisicoquímica y Bioquímica elemental, en la Maestría en Ciencias Básicas ofrecida por la División Académica de Ciencias de la Salud (DACS) de la Universidad Juárez Autónoma de Tabasco UJAT, Noviembre de 2001. • Clase por invitación en el curso de patogénesis microbiana con el tópico “Factores de patogenicidad de Vibrio cholerae” dentro del curso Biología del Parasitismo I y II en la maestría en Patología Experimental del CINVESTAV-IPN. Marzo, 19 de 2003. • Ponente del tema “Participación de EspC en la virulencia de EPEC” dentro del curso “Tópicos de Virulencia Bacteriana” organizado por la División de Ciencias Químico Biológicas de la Facultad de Estudios Superiores Cuautitlán de la UNAM. Cuautitlán Izcalli Edo. de México, los días 27 y 28 de Noviembre de 2003. • Cátedra titulada “Mecanismo de invasión y daño de Bacterias: la toxina colérica” dentro del curso Biología del Parasitismo I y II en la maestría en Patología Experimental del CINVESTAV-IPN. Marzo 17 de 2004. • Conferencia “Las bacterias: su descubrimiento, importancia y como se relacionan con el ser humano” ofrecida dentro del programa La Ciencia en tu Escuela, organizado por la Academia Mexicana de Ciencias (AMM), el día 26 de Marzo de 2004. • Cátedra titulada “Factores de patogenicidad de Vibrio cholerae” dentro del curso Biología del Parasitismo I y II en la maestría en Patología Experimental del CINVESTAV-IPN. Marzo, 16 de 2005. • Conferencia titulada “Como causan diarrea las bacterias” ofrecida dentro del programa Domingos de la Ciencia, organizado por la Academia Mexicana de Ciencias (AMM) y el Consejo Nacional de Ciencia y Tecnología (CONACyT), en la Universidad Autónoma de Queretaro los días 14 y 15 de Octubre de 2005. • Conferencia internacional. “Translocation of the EspC autotransporter toxin from EPEC requires the type three secretion system” llevada a cabo en el Department of Cellular Microbiology, Max Planck Institute for Infection Biology en Berlín Alemania Abril 27 de 2006. • Conferencia internacional. “EspC an autotransporter toxin secreted by EPEC requires the TTSS to get entry into infected cells” llevada a cabo en el Departamento de Bioquímica y Genética Molecular de la Escuela de Medicina de la Universidad de Pittsburgh en Pittsburgh Pennsylvania Estados Unidos, 2 de Mayo de 2006. • Conferencia. Up-regulation of Transcription and Secretion of C. perfringens Type C Toxins in the Presence of Caco-2 cells. Pittsburgh Bacterial Meeting, Pittsburgh PA, USA, 15 de Marzo de 2008.

10

Cursos en México • Control de calidad en bacteriología médica, curso pre-congreso durante los trabajos del XX Congreso Nacional de Químicos Clínicos, celebrado en Torreón Coahuila del 9 al 11 de Septiembre de 1996. • Asistencia al curso pre-congreso micología médica, celebrado durante los trabajos del XXI congreso Nacional de Química Clínica, organizado por la Asociación Mexicana de Bioquímica Clínica en México D.F. los días 1 y 2 de Marzo de 1998. • Participación como alumno del curso "biología molecular aplicada al diagnóstico clínico" durante los trabajos del XXII Congreso Nacional de Química Clínica del Asociación Mexicana de Bioquímica Clínica los días 27 y 28 de Febrero de 1999. • Alumno del curso "Streptococcus y resistencia a antibióticos" celebrado en la ENCB del IPN los días 19 y 20 de Marzo de 1999 durante el marco de actividades del 50 Aniversario de la Asociación Mexicana de Microbiología. • Participación en el curso-taller denominado Bases para implementar la documentación del laboratorio clínico de acuerdo a la norma NOM-SSA-166-1997 realizado por el Instituto de Desarrollo Profesional y Educación Continua, S.C., en Villahermosa Tabasco los días del 17 al 19 de Noviembre del 2000. • Asistencia al curso de Microbiología Celular, Sistemas de Secreción Bacterianos y el Citoesqueleto de Células Eucarióticas, impartido por diversos investigadores científicos de México. Sección de Ciencias Morfológicas Agropecuarias de la Facultad de Estudios Superiores Cuatitlán, UNAM, Septiembre 11 al 14, 2002. Cursos internacionales • Patogenia de Microorganismos Entéricos, llevado a cabo en la unidad de Investigación de la UNAM en el Hospital General de México, organizado por el departamento de Medicina Experimental de la Facultad de Medicina de la UNAM en colaboración con la Universidad de Washington. México D.F. del 2 al 7 de Noviembre de 1998. • Research Integrity, Center for Continuing Education in the Health Sciences, University of Pittsburgh, Pittsburgh PA, USA October 2, 2006. • Blood Borne Pathogens Training (CertificateID:67462). Center for Continuing Education in the Health Sciences, University of Pittsburgh, Pittsburgh PA, USA. Abril, 2007. • Chemical Hygiene Training (CertificateID:67480). Center for Continuing Education in the Health Sciences, University of Pittsburgh, Pittsburgh PA, USA. Abril, 2007.

11

• Training on Bioinformatics for Bioterrorism: Pathema-Clostridium. The Jon Craig Venter Institute (formerly known The Institute for Genome Research TIGR). Rockville Maryland, USA. Sep 6-7, 2007. • Dangerous Goods Shipping Training. Department of Environmental Health and Safety, University of Pittsburgh, Pittsburgh PA, USA. September, 12, 2007. Eventos profesionales y científicos: • Primer Simposium de Medicina Transfucional, servicios coordinados de salud pública en el estado de Puebla (SSA) y la BUAP, llevado a cabo los días del 8 al 10 de Junio de 1994 en Puebla, Puebla. • Asistencia al ciclo de conferencias “Actualización del QFB” celebrado en el auditorio Albert Einstein de la BUAP, Puebla, Puebla los días del 3 al 7 de Junio de 1996. • Asistencia a las “LV Jornadas Farmacéuticas”, organizado por la Asociación Farmacéutica Mexicana en Puebla, Puebla los días del 23 al 25 de Abril de 1997. • Simposium “Diagnóstico y Tratamiento de Algunas Enfermedades Congénitas”, organizado por la Asociación Poblana de Médicos Generales, el 5 de Julio de 1997 en la cd. de Puebla, Puebla . • Asistencia al “Primer Encuentro de Posgraduados” organizado por la facultad de ciencias Químicas de BUAP, del 20 al 22 de Agosto de 1997. • Asistencia a los “Foros Farmacéuticos” organizados por la Asociación Farmacéutica Mexicana en la facultad de Ciencias Químicas de la BUAP, del 2 al 4 de Septiembre de 1997. • Asistencia a la “Primera Jornada de Reporte Técnico en Análisis Clínicos” en Noviembre de 1997 en Puebla, Puebla. • Simposium “Automatización y Avances en el Área Clínica” llevado a cabo en el auditorio de la facultad de derecho de la BUAP los días 10 al 12 de Noviembre de 1997. • Ciclo de Conferencias “Perspectivas del QFB al siglo XXI” llevadas a cabo en el auditorio de la facultad de Derecho de la BUAP. Puebla, Puebla los días del 11 al 13 de Marzo de 1998. • Asistencia al 3rd Pittsburgh bacterial Meeting, Pittsburgh PA, USA. April 2007. • Asistencia al 4th Pittsburgh bacterial Meeting, Pittsburgh PA, USA. April 2008.

12

• Fourth annual retreat. Molecular virology and Microbiology Graduate Program. University of Pittsburgh School of Medicine. Pittsburgh PA, USA. Mayo 2, 2008. • 3rd Annual meeting of the University of Pittsburgh Post Doctoral Association. Pittsburgh PA, USA, May 7, 2008. Experiencia Profesional • Químico analista en el departamento de microbiología en Asesores Especializados en Laboratorios A. C. de la cd. de Puebla, Puebla en el periodo comprendido entre Junio de 1996 a Septiembre de 1997. • Estancia especial en el laboratorio de Producción y Control de Biológicos de la ENCB del IPN bajo la coordinación de la Dra. en C. Rosa María Ribas Jaimes durante Julio a Octubre del 2000. • Estancia especial en el laboratorio de Parasitología y Micología del Hospital de Infectología del Centro Médico Nacional la Raza del IMSS en Noviembre del 2000. • Jefe de laboratorio en Laboratorios Graniel (análisis clínicos y microbiológicos) de Comalcalco Tabasco, México. Diciembre del 2000 a Julio de 2002. • Profesor de Bioquímica Humana, División Académica de Ciencias de la Salud (DACS), Escuela de Medicina Universidad Juárez Autónoma de Tabasco (UJAT), Agosto 2001-Julio 2002. • Post Doctoral Scientist, Department of Microbiology and Molecular Genetics, University of Pittsburg School of Medicine, 2006-a la fecha. Membresías • Miembro de la Confederación Nacional de Químicos Clínicos el año de 1997. • Miembro de la Asociación Farmacéutica Mexicana los años 1997 y 1998. • Miembro de la Asociación Mexicana de Bioquímica Clínica los años 1997 y 1998. • Miembro de la Sociedad Mexicana de Medicina Tropical los años 1999 al 2001. • Miembro del Colegio de Químicos de Tabasco (COLQUITAB), 2001. • Miembro de la American Society for Microbiology de 2000 a la fecha. Becas obtenidas • Becario del Consejo Nacional de Ciencia y Tecnología (CONACYT). Número de registro 138072. En apoyo a los Posgrados Maestría en Ciencias (Enero 1999 a diciembre del 2000) y Doctorado en Ciencias (Agosto de 2002 a Marzo de 2006).

13

• Becario del Programa Institucional de Formación de Investigadores (PIFI) por parte de la Sección de Estudios de Posgrado e Investigación de la Escuela Nacional de Ciencias Biológicas del IPN. Julio de 1999 a Junio del 2000. • Beca para realizar una estancia Post Doctoral en el extranjero, Consejo Nacional de Ciencia y Tecnología de México (CONACyT). Enero-Diciembre 2008.

Beta toxin is essential for the intestinal virulence ofClostridium perfringens type C disease isolate CN3685 in arabbit ileal loop model

Sameera Sayeed,1 Francisco A. Uzal,2

Derek J. Fisher,1† Juliann Saputo,2 Jorge E. Vidal,1

Yue Chen,3 Phalguni Gupta,3 Julian I. Rood4 andBruce A. McClane1,4*1Department of Molecular Genetics and Biochemistry,University of Pittsburgh School of Medicine, Pittsburgh,PA, USA.2California Animal Health and Food Safety LaboratorySystem, San Bernardino Branch, School of VeterinaryMedicine, University of California, Davis, CA, USA.3Department of Infectious Diseases and Microbiology,Graduate School of Public Health, University ofPittsburgh, PA, USA.4Australian Research Council Centre of Excellence inStructural and Functional Microbial Genetics,Department of Microbiology, Monash University, ClaytonCampus, Vic., Australia.

Summary

Clostridium perfringens type C isolates, whichcause enteritis necroticans in humans and enteritisand enterotoxaemias of domestic animals, typicallyproduce (at minimum) beta toxin (CPB), alpha toxin(CPA) and perfringolysin O (PFO) during log-phasegrowth. To assist development of improved vaccinesand therapeutics, we evaluated the contribution ofthese three toxins to the intestinal virulence of type Cdisease isolate CN3685. Similar to natural type Cinfection, log-phase vegetative cultures of wild-typeCN3685 caused haemorrhagic necrotizing enteritis inrabbit ileal loops. When isogenic toxin null mutantswere prepared using TargeTron® technology, even adouble cpa/pfoA null mutant of CN3685 remainedvirulent in ileal loops. However, two independent cpbnull mutants were completely attenuated for virulencein this animal model. Complementation of a cpbmutant restored its CPB production and intestinalvirulence. Additionally, pre-incubation of wild-typeCN3685 with a CPB-neutralizing monoclonal antibody

rendered the strain avirulent for causing intestinalpathology. Finally, highly purified CPB reproducedthe intestinal damage of wild-type CN3685 and thatdamage was prevented by pre-incubating purifiedCPB with a CPB monoclonal antibody. These resultsindicate that CPB is both required and sufficient forCN3685-induced enteric pathology, supporting a keyrole for this toxin in type C intestinal pathogenesis.

Introduction

The Gram-positive spore-former Clostridium perfringensis one of the most important anaerobic pathogens ofhumans and domestic animals (McClane et al., 2004).The virulence of this bacterium is largely attributable to itsprodigious toxin-producing abilities. However, no singleC. perfringens isolate produces all 15 of the toxinsreported in the literature. This provides the basis for acommonly used classification scheme (McClane et al.,2004) that assigns C. perfringens isolates to one of fivetypes (A–E), based on their production of four (a, b, e andi) typing toxins. By definition, C. perfringens type C iso-lates must produce (i) alpha toxin (CPA), a potent lethaltoxin with phospholipase C, sphinomyelinase and hostcell signalling properties (Songer, 1997; Titball, 1998;Lyras and Rood, 2006), and (ii) beta toxin (CPB), an evenmore lethal pore-forming toxin (Hunter et al., 1993;Steinthorsdottir et al., 1998; Tweten, 2001). Fisher et al.(2006) have shown that most type C isolates also produceperfringolysin O (PFO), another lethal pore-forming toxin,and that some type C isolates additionally make theenterotoxin (CPE) or beta2 toxin (CPB2). Very recently,some type C isolates were found (Amimoto et al., 2007) toproduce TpeL, a truncated homologue of Clostridium dif-ficile toxin A.

In humans, C. perfringens type C isolates cause enteri-tis necroticans (also known as Darmbrand or Pigbel),which is currently endemic in much of Southeast Asia(Johnson and Gerding, 1997). Until the 1980s, typeC-associated enteritis necroticans was the most commoncause of death in children older than 1 year of age in thePapua New Guinea Highlands (Johnson and Gerding,1997; Lawrence, 1997; McClane et al., 2004). Enteritisnecroticans due to type C isolates also has been reported

Accepted 8 October, 2007. *For correspondence. E-mail [email protected]; Tel. (+1) 412 648 9022; Fax (+1) 412 624 1401. †Presentaddress: Department of Microbiology and Immunology, UniformedServices University of the Health Sciences, Bethesda, MD, USA.

Molecular Microbiology (2008) 67(1), 15–30 � doi:10.1111/j.1365-2958.2007.06007.x

© 2008 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd

in developed countries (Lee et al., 2000; Gui et al., 2002;Sobel et al., 2005; Matsuda et al., 2007). The risk factorsfor developing this disease include low trypsin productiondue to a protein-poor diet or pancreatic disease andconsumption of foods containing a high concentration oftrypsin inhibitor. Each of these risk factors contributes totoxin persistence in the gastrointestinal tract during type Cinfection. In Pigbel, type C isolates are usually introducedinto the gastrointestinal tract by consumption of a con-taminated meat (typically pork). Severe cases of typeC-associated enteritis necroticans, where patients exhibitsigns of toxaemia and segmental necrotizing enteritis isvisible in the small intestines, can be rapidly fatal withoutsurgical resection of the small bowel.

Besides their importance for human medicine, type Cveterinary infections are also economically important interms of both livestock loss and vaccination costs(Songer, 1996; McClane et al., 2004). Type C veterinaryinfections include ‘struck’, an enterotoxaemia where anadult sheep dies so quickly that it appears to have beenstruck by lightning, and haemorrhagic enteritis in lambs.Type C isolates also cause enterotoxaemias in piglets,calves and foals, as well as some cases of necrotic enteri-tis in poultry. Animals suffering from type C infection typi-cally show haemorrhaging and necrosis of their intestines,which can result in death due to direct intestinal damageor toxaemia following absorption of toxins from the intes-tines into the circulation. Newborn animals, particularlypiglets, are especially susceptible to type C outbreaks,which can claim herd mortality rates exceeding 30%.

It has been generally assumed, but not directly proven,that the pathogenesis of type C infections largely involvesCPB. However, several observations suggest that CPBmay not be the only, or even perhaps the major, toxininvolved in type C disease. For example, we recentlydemonstrated that type C isolates typically produce atleast three, and sometimes up to five, different lethaltoxins (Fisher et al., 2006). This opens the possibility oftoxins such as CPA or PFO contributing to type C intesti-nal necrosis or promoting CPB entry into the circulation tocause systemic lethality. The possible involvement ofother toxins in type C pathogenesis is further supportedby reported difficulties in reproducing typical type Cdisease in experimental animals using CPB alone(Songer, 1996). For example, typical disease lesions havebeen described in gut loops inoculated with broth culturesof type C isolates, but necrosis was absent when similarloops were inoculated with CPB alone (Bergeland, 1972).In addition, current type C human or veterinary ‘betatoxoid’ vaccines are typically prepared using crude type Cculture supernatants, so they contain several inactivatedtoxins (Fisher et al., 2006), i.e. these protective ‘betatoxoid’ vaccines may be evoking protective immunityagainst several different type C toxins, not only CPB.

Resolving the uncertainties regarding type C patho-genesis is important for the design and development ofimproved, new generation human and veterinaryvaccines. Therefore, the current study prepared isogenictoxin null mutants in the type C sheep disease isolateCN3685. Those mutants have allowed us to evaluate theenteric virulence contributions of three toxins (CPA, CPBand PFO) typically produced by nearly all type C isolatesduring late log-phase vegetative growth.

Results

Construction and genotypic characterization of CN3685cpb(antisense), pfoA and plc null mutants

To evaluate the role of CPB in CN3685 virulence, the cpbgene of this type C isolate was inactivated by inserting,in the antisense orientation, a Group II intron (~900 bp)between nucleotides 154|155 of the CN3685 cpb openreading frame (ORF) (Fig. 1A). The presence of an introninsertion in the cpb gene of this mutant (named BMC100)was first shown by PCR using two cpb-specific primersthat supported PCR amplification of an ~1 kb product fromthe wild-type cpb gene but amplified a larger ~2 kbproduct from BMC100 (Fig. 1A). Sequencing of the cpbgene of BMC100 then confirmed that the Group II intronhad inserted into the expected site of the BMC100 cpbORF. Southern blotting demonstrated that only a single-intron insertion was present in this cpb mutant (Fig. 2).

To assess the contributions of CPA and PFO to CN3685virulence, an intron was inserted (in the antisense orien-tation) into, respectively, this strain’s plc or pfoA gene tocreate an isogenic plc mutant named BMC101 or anisogenic pfoA mutant named BMC102. PCR analysis withplc- or pfoA-specific primers amplified the expected largeproducts of an intron-disrupted plc or pfoA gene fromBMC101 or BMC102 respectively (Figs 1B and C). Thedisrupted plc gene of BMC101 and disrupted pfoA gene ofBMC102 were also sequenced, which confirmed an intronhad been inserted into the expected plc or pfoA ORFsites. Southern hybridization with an intron-specific probedemonstrated that BMC101 and BMC102 carried only thesingle-intron insertion present in their plc or pfoA genesrespectively (Fig. 2).

Phenotypic characterization of CN3685 cpb(antisense), pfoAand plc null mutants

When grown on egg yolk agar plates (Fig. 3A), colonies ofthe wild-type CN3685 parent, BMC100 and BMC102 weresurrounded by the characteristic halo zone indicative oflecithin breakdown due to the phospholipase C activity ofCPA (Awad et al., 2001). In contrast, BMC101 failed toproduce this zone around its colonies, indicating a lack ofCPA production (Fig. 3A).

16 S. Sayeed et al. �

© 2008 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd, Molecular Microbiology, 67, 15–30

The wild-type parent and mutants were also grown onsheep blood agar plates to assess their PFO production.Colonies of wild-type CN3685 and BMC100 were sur-rounded (Fig. 3A) by the double zone of haemolysis char-acteristic of C. perfringens [the faint outer haemolyticzone is due to CPA production while the inner zone ofb-haemolysis is due to PFO production (Allen and Baron,1991)]. In contrast, colonies of BMC101 were not sur-rounded by the outer haemolytic zone, indicating a loss ofCPA activity, while colonies of BMC102 were not sur-rounded by the inner b-haemolytic zone, indicating a lossof PFO activity (Fig. 3A).

To compare CPB production between the wild-typeparent versus cpb, pfoA and plc single-toxin mutants, aCPB Western blot was performed. Cultures of wild-type

Fig. 1. Intron-based mutagenesis to createCN3685 single cpb, plc or pfoA mutants anda double pfoA/plc mutant.A–C. PCR confirmation of cpb, plc and pfoAmutants respectively. In each panel (i)graphically depicts: (top) CN3685 wild-typecpb, plc or pfoA genes with their mapped orapparent promoter region (P), showing theexpected 1, 1.2 or 0.65 kb PCR productamplification from, respectively, the cpb, plc orpfoA genes using primer pairs specific foreach wild-type toxin gene and (bottom) eachCN3685 mutant with an intron insertion in itscpb, plc or pfoA ORF, as specified, showingthe expected 2, 2.1 or 1.6 kb PCR productsthat should be amplified from theintron-disrupted genes using the same toxingene primers. (ii) Actual PCR results forreactions described above in (i). Lane 1: 1 kbmolecular weight marker; lane 2: PCR productfrom CN3685; and lane 3: PCR product fromeither (A) BMC100 (CN3685::cpb(antisense)), (B)BMC101 (CN3685::plc) or (C) BMC102(CN3685::pfoA).D. PCR analysis of BMC103 double mutantusing DNA from wild-type CN3685 or thedouble pfoA/plc mutant using pfoA- orplc-specific primers. pfoA gene amplification isshown for wild type CN3685 (lane 1) andBMC103 double mutant (lane 2). Amplificationof the plc gene is shown for wild-type CN3685(lane 3) and BMC103 (lane 4). A 1 kbmolecular marker is in lane 5. In all Fig. 1gels, DNA from wild-type CN3685, single-toxinmutants or double-toxin mutants wassubjected to PCR and products wereelectrophoresed on a 1% agarose gel andthen stained with ethidium bromide forvisualization.

Fig. 2. Southern blot hybridization of wild-type CN3685 and toxinnull mutants. DNA from each strain was digested with EcoRI priorto electrophoresis on a 1% agarose gel. The separated DNA on theagarose gel was then transferred onto a nylon membrane andhybridized with an intron-specific DIG-labelled probe. Lane 1:Wild-type CN3685; lane 2: BMC100; lane 3: BMC101; lane 4:BMC102; and lane 5: BMC103.

Clostridium perfringens type C isolate CN3685 intestinal virulence 17


CN3685, BMC101 and BMC102 each contained a 35 kDaprotein that was recognized by an anti-CPB monoclonalantibody (mAb) and co-migrated with purified CPB(Fig. 3B). In contrast, no anti-CPB mAb immunoreactivitywas observed when BMC100 cultures were similarlyWestern immunoblotted (Fig. 3B).

Virulence phenotypes of wild-type CN3685 and thethree single-toxin mutants in rabbit ileal loops

Late log-phase cultures of wild-type CN3685 reproducedthe intestinal pathology of type C infection after a 6 htreatment of ileal loops. Those CN3685-infected loopsshowed grossly red mucosa with thick, mucoid and redcontent, all of which could be observed from the serosalsurface (Fig. 4). The intestinal wall appeared thin and had

lost its natural tone. A modest accumulation of bloody fluidwas observed in the lumen [0.6 g per cm of bloody fluidversus 0.4 g per cm of clear fluid in control loops; thosequantitative differences in fluid volume were consistentbut not statistically significant (P > 0.05)].

Histologically, loops treated with wild-type CN3685showed severe diffuse necrotizing enteritis, character-ized by complete loss of absorptive cells along the villiand coagulation necrosis of the lamina propria wherelarge amounts of karyorrhectic debris were present(Fig. 5 and Table 1). The villi in these loops had almostcompletely disappeared and were replaced by a seriesof low-denudated bumps that were lined by apseudomembrane composed of haemorrhage, neutro-phils, necrotic epithelial cells, cell debris and largenumbers of Gram-positive bacteria. Occasionally a few

Fig. 3. Characterization of toxin productionby CN3685 and toxin null mutants.A. Phenotypic expression of CPA and PFO bywild-type CN3685 and each toxin null mutant.PLC activity of CPA was detected on egg yolkagar plate and PFO activity on blood agarplates. Wild-type CN3685 and each toxinmutant (except for null mutants with anintron-disrupted plc gene) showed acharacteristic CPA-induced halo zone on eggyolk agar plates. Strains producing PFOshowed an inner clear zone of b-haemolysisaround colonies grown on blood agar platesbut mutants with an intron-disrupted pfoAgene failed to show this b-haemolytic zone.Arrows point to non-b-haemolytic colonies forBMC102 and BMC103.B. Western blot analysis of CN3685 andtoxin null mutants for their CPB production.Overnight culture supernatants from eachstrain were mixed with protein loadingdye and electrophoresed on a 10%polyacrylamide gel containing SDS. Theseparated proteins were then transferred ontonitrocellulose membrane and Western blottedfor detection of CPB using a mousemonoclonal anti-CPB antibody. Left. CPBproduction by: lane 1: CN3685; lane 2:BMC100; lane 3: purified CPB. Right. Lane 1:BMC100; lane 2: Wild-type CN3685; lane 3:BMC101; lane 4: BMC102; and lane 5:BMC103.

A

B



preserved crypts were observed, although most of thecrypt epithelium was also necrotic and sloughed off.Occasionally thrombi were observed in the mucosalblood vessels. The submucosa showed severe oedemaand hyperaemia and the lymphatic vessels wereengorged with proteinaceous material and neutrophils.

Fibrin thrombi were frequently observed in submucosalveins. In contrast, rabbit ileal loops treated for 6 h withsterile, non-toxic culture medium (negative controls)showed no significant luminal fluid accumulation andgross or histological abnormalities, except for some mildsubmucosal oedema and dilation of submucosal lym-phatic vessels (Fig. 5 and Table 1).

Rabbit ileal loops treated with the BMC100 cpb mutantexhibited no significant intestinal pathology, appearinggrossly (Fig. 4) and histologically (Fig. 5 and Table 1)indistinguishable from negative control loops receivingonly sterile medium. In contrast, ileal loops treated witheither BMC101 or BMC102, which both still express CPB,showed bloody luminal fluid accumulation and histologicalchanges similar to the intestinal effects caused by wild-type CN3685 (Figs 4 and 5 and Table 1).

For wild-type CN3685, BMC100, BMC101 andBMC102, ~108 bacterial cells were injected into each loop.After 6 h, ~108C. perfringens cells were recovered fromeach loop. No C. perfringens cells were recovered fromloops inoculated with sterile media, confirming that theC. perfringens recovered from loops containing CN3685,BMC100, BMC101 or BMC102 were the inoculatedstrains. Intact CPB was detectable (not shown) in ilealloop fluids of wild-type CN3685, BMC101 and BMC102(but not BMC100) after the 6 h treatment.

Construction and characterization of a CN3685 plc/pfoAdouble mutant

Results using the single-toxin mutants indicated thatsingle plc and pfoA null mutants remained fully virulent inthe ileal loop model. To assess whether those twomembrane-active toxins might be redundant, such thatthe production of either CPA or PFO is sufficient for

Fig. 4. Gross pathology of rabbit-ligated intestinal loops inoculated with CN3685, isogenic single- and double-toxin mutants, purified CPB ormedium (uninoculated TGY medium) and then incubated for 6 h. Note that loops inoculated with the wild-type strain, BMC101, BMC102,BMC103 or purified CPB are severely haemorrhagic and distended with fluid. No significant gross abnormalities are observed in the loopsinoculated with BMC100, BMC104 or sterile medium. Not shown are loops inoculated with BMC105, which were also haemorrhagic anddistended with fluid or loops inoculated with BMC106 which did not show gross abnormalities. See Table 1 for the number of times thisexperiment was performed.

Fig. 5. Histological damage in rabbit ileum treated for 6 h with an8 h culture of wild-type CN3685, BMC101, BMC102 or BMC103.Control loops, which were inoculated with sterile TGY, and loopsinoculated with the cpb null mutant BMC100 showed normal, full-length intestinal villi with a well-preserved epithelium and laminapropria. Loops inoculated with wild-type CN3685, BMC101 (plc nullmutant), BMC102 (pfoA null mutant) or BMC103 (double plc/pfoAnull mutant) all showed histological damage consisting of necrosisand loss of epithelium, necrosis of lamina propria, blunting of thevilli, haemorrhage of the mucosa and diffuse neutrophilic infiltrationof mucosa and sub-mucosa. Sections were stained with haematoxy-lin and eosin and photographed at 200¥ magnification. See Table 1for the number of times this experiment was performed.



CN3685 virulence, we constructed a plc/pfoA double nullmutant (BMC103) using clostridia-modified TargeTron®

technology. The presence of both intron-disrupted plcand pfoA genes in BMC103 was demonstrated by PCR(Fig. 1D) and sequencing. The presence of only twointrons in BMC103 was then confirmed by Southernhybridization (Fig. 2). The inability of BMC103 to produceCPA or PFO was shown by its lack of phosopholipase Cactivity on egg yolk agar and failure to produce anyhaemolysis on blood agar plates (Fig. 3A). However, thisdouble mutant retained normal CPB production (Fig. 3B).Finally, when late log-phase cultures of BMC103 weretested in the ileal loop model, they produced similarpathological effects as wild-type CN3685 (Figs 4 and 5and Table 1).

Approximately 108 cells of BMC103 were injected intoeach loop. After 6 h, the same number of C. perfringenscells were recovered from the loop. Intact beta toxin wasdetected (not shown) in luminal fluids from BMC103-inoculated ileal loops after the 6 h treatment.

Construction and characterization of a CN3685 cpb(sense)

null mutant and complement

The results presented above suggested that CPB isrequired for the virulence of late log-phase CN3685 cul-tures in the ileal loop model. This suggestion was thenverified by several independent approaches. First, to ruleout the possibility that polar effects had caused the viru-lence attenuation of the cpb(antisense) null mutant, a secondCN3685 cpb mutant named BMC104 was constructedthat carries an intron insertion, in the sense orientation, in

its cpb ORF. PCR analysis using cpb-specific primersproduced an ~1 kb product for wild-type CN3685 but theexpected ~2 kb product for BMC104, reflecting the pres-ence of an intron disrupting its cpb ORF (Fig. 6A). Introndisruption of the cpb ORF in BMC104 was then confirmedby sequencing. Western immunoblotting showed loss ofCPB production by BMC104 (Fig. 6B), although thismutant retained the ability to produce CPA and PFO(Fig. 3A).

The purpose of constructing BMC104 was twofold. First,it provided a second, independent mutant to confirm thatloss of CPB production leads to the intestinal virulenceattenuation of CN3685. Second, the availability of BMC104permitted use of a trans-splicing complementationapproach (Yao et al., 2006) to specifically restore CPBproduction and rule out polar effects as the explanation forthe attenuated phenotype of cpb null mutants. To removethe intron from the cpb mRNA of BMC104, the plasmidpJIR750cpbi-s (which encodes a functional LtrA protein)was re-introduced into this mutant, producing BMC105.The complemented mutant was then grown at 30°C in thepresence of antibiotics to maintain selective pressure andproduce maximal expression of the LtrA protein requiredfor splicing-induced intron removal (Yao et al., 2006) andconsequent restoration of CPB production. Western immu-noblotting confirmed production of CPB by BMC105(Fig. 6B), although at one-third the CPB level of wild-typeCN3685 as determined by densitometric scanning(not shown). Similar Western blot analysis (not shown)detected no CPB production by BMC106, which was trans-formed with the empty vector pJIR750 shuttle plasmid thatlacks an LtrA coding sequence.

Table 1. Rabbit ileal loop histopathology.

Sample DesquamationNecrosis ofEpithelium

Necrosis ofLamina Propria Inflammation

VillusBlunting Overall

CN3685 3.9 � 0.1 3.9 � 0.1 3.5 � 0.7 2.0 � 0.8 3.9 � 0.4 3.9 � 0.4BMC100 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a

BMC101 3.8 � 0.4 3.9 � 0.2 3.5 � 0.6 2.0 � 0.4 3.8 � 0.4 3.9 � 0.2BMC102 3.7 � 0.3 3.6 � 0.5 3.1 � 0.3 2.0 � 0.4 3.5 � 0.6 3.5 � 0.6BMC104 1.1 � 0.2a 1.2 � 0.2a 1.1 � 0.4a 1.1 � 0.2a 1.2 � 0.3a 1.2 � 0.2a

BMC105 3.0 � 0.1a 3.1 � 0.3a,b 2.9 � 0.3b 1.9 � 0.3b 3.5 � 0.6b 2.9 � 0.4a,b

BMC106 1.2 � 0.2a 1.1 � 0.2a 1.1 � 0.3a 1.2 � 0.2 1.2 � 0.3a 1.2 � 0.3a

BMC103 3.8 � 0.4 3.8 � 0.4 3.3 � 0.3 2.1 � 0.3 3.4 � 1.0 3.9 � 0.4CN3685 + CPBmAb 1.2 � 0.3a 1.2 � 0.3a 1.1 � 0.3a 1.2 � 0.3 3.9 � 0.3 1.3 � 0.3a

CN3685 + CPAmAb 3.9 � 0.2 3.9 � 0.2a 3.2 � 0.4 1.8 � 0.3 1.1 � 0.3a 3.8 � 0.3CPB Toxin 3.9 � 0.3 3.5 � 0.7 3.4 � 0.3 2.1 � 0.3 3.9 � 0.2 3.8 � 0.3CPB + CPBmAb 1.2 � 0.3a 1.2 � 0.3a 1.2 � 0.4a 1.9 � 0.3 1.2 � 0.3a 1.5 � 0.3a

CPA + CPAmAb 3.9 � 0.2 3.8 � 0.3 3.1 � 0.3 1.1 � 0.3a 3.9 � 0.3 3.7 � 0.3CPA Toxin 1.1 � 0.3a 1.1 � 0.3a 1.1 � 0.3a 1.9 � 0.3 1.1 � 0.3a 1.2 � 0.3a

Medium 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.2a 1.1 � 0.3a 1.1 � 0.2a

a. Indicates a statistically significant difference (P < 0.05) relative to CN3685.b. Indicates a statistically significant difference (P < 0.05) between BMC104 and BMC105.Pathology was scored by a blinded pathologist, on a 1- to 4-point scale (with 0.5 point increments); a 4 score represents maximal effect and a 1score represents no effect. Results shown are for either four rabbits (BMC100, BMC101, BMC102, BMC103, BMC104, BMC105, BMC106,CN3685 + CPB mAb, CN3685 + CPA mAb, CPB or CPA) or eight rabbits (CN3685 and medium).



Northern blot (Fig. 6C) and reverse transcription poly-merase chain reaction (RT-PCR) (Fig. S1) analyses sup-ported splicing-induced intron removal as restoringproduction of functional cpb message to BMC105.RT-PCR confirmed that mRNA with cpb sequences waspresent in wild-type CN3685, BMC104 and BMC105 butintron-containing mRNA was only present in BMC104 and

BMC105. RT-PCR, using one primer located within theintron and a second present in cpb sequences down-stream of the intron insertion site, showed the presence ofintron-disrupted cpb mRNA in both BMC104 and BMC105(due to incomplete splicing removal of the intron). North-ern blot analyses then indicated that: (i) the cpb gene istranscribed as a monocistronic message by CN3685, (ii)native, ~1.4 kb cpb message is detectable in CN3685 andBMC105, but not in BMC104, and (iii) larger (> 2 kb)intron-disrupted cpb messages are present in BMC104and BMC105. The presence of more than one intron-disrupted cpb message in BMC104 and BMC105 likelyreflects message instability, as noted previously withintron-containing mRNA (Matsuura et al., 1997).

When BMC104, BMC105 and BMC106 were tested forvirulence in the rabbit ileal loop model, BMC104 failed tocause intestinal pathology (Figs 4 and 7). Unlike BMC104or BMC106, the BM105 complementing strain wasvirulent in this animal model (not shown and Fig. 7 andTable 1). However, BMC105 was not quite as virulent aswild-type CN3685, likely because of its lower CPB expres-sion levels. Consistent with this possibility, CPB levelsin loop fluids had dropped below detection after 6 h(not shown).

For wild-type CN3685, BMC104, BMC105 andBMC106, ~108 bacteria were injected into each loop andthe same number of C. perfringens were recovered fromeach loop after the 6 h experiment.

Monoclonal antibody neutralization of CN3685-inducedileal loop pathology

A non-genetic approach then independently verified aCPB requirement for CN3685 virulence. When late log-phase cultures of the CN3685 parent were pre-incubatedwith a neutralizing mAb against CPB, those cultures werecompletely attenuated in the rabbit ileal loop model(Fig. 8). However, pre-incubating the same CN3685 cul-

Fig. 6. Analysis of cpb(sense) null mutant BMC104 and complement-ing strain BMC105.A. Using cpb-specific primers, DNA from wild-type CN3685 orBMC104 were subjected to PCR. Products were then electrophore-sed on a 1% agarose gel and stained with ethidium bromide for visu-alization. cpb gene amplification is shown for wild-type CN3685(lane 1), BMC104 (lane 2) and BMC105 (lane 3).B. Western blot analysis to detect the CPB production in the wild-type CN3685 and BMC104 (lane 2) and BMC105 (lane 3). Similaranalyses (not shown) demonstrated that BMC106 also carries thelonger, disrupted cpb gene of BMC104 and produces no CPB.C. Northern blot analysis of cpb mRNA from wild-type CN3685, cpbmutant and complementing strain. RNA was extracted from wild-typeCN3685, BMC104 or BMC105. Total RNA (15–20 mg) from eachstrain was mixed with 60 ml of loading dye and electrophoresed on a1% Northern gel. The separated DNA on the agarose gel was thentransferred onto a nylon membrane and hybridized with a cpb-specific DIG-labelled probe. The separated RNA bands were trans-ferred onto a membrane for Northern blot analysis to detect cpbmRNA in wild-type CN3685 (lane 1), BMC104 (lane 2) and BMC105(lane 3). Note the presence of larger cpb mRNA transcripts inBMC104 due to the inserted intron and the presence in BMC105 ofboth wild-type and larger cpb transcripts (the larger transcriptsreflect removal of introns from only some cpb message in thiscomplementing strain, as evident from its producing only ~30% ofwild-type CPB levels).

Fig. 7. Histological damage in rabbit ileum after a 6 h treatments with 8 h cultures of wild-type CN3685, BMC104, BMC105 or BMC106.Control loops, which were inoculated with sterile TGY medium, as well as loops inoculated with the BMC104 cpb null mutant or BMC106(BMC104 transformed with an empty shuttle plasmid) showed normal intestinal villi with a well-preserved epithelium and lamina propria. Loopsinoculated with wild-type CN3685 or BMC105 (the complementing strain of BMC104) show similar histological changes consisting of necrosisand loss of epithelium, necrosis of lamina propria, blunting of the villi, haemorrhage of the mucosa and diffuse neutrophilic infiltration ofmucosa and sub-mucosa. Sections were stained with haematoxylin and eosin and photographed at 200¥ magnification. See Table 1 for thenumber of times this experiment was performed.



tures with a CPA-neutralizing mAb did not inhibit ileal looppathology (Table 1). For these experiments, ~108 cells ofCN3685 were injected into each loop and the samenumber of C. perfringens cells were recovered from allloops after the 6 h experiment.

Effects of purified beta toxin on ileal loops andmouse lethality

The results presented above implicated beta toxin as anessential contributor to the intestinal virulence of CN3685infection under the experimental conditions tested. Toassess whether the presence of CPB alone is sufficient tocause intestinal pathology similar to infection with latelog-phase cultures of wild-type CN3685, we purified CPBto near homogeneity. When a dose–response experimentwas performed to test the activity of this purified beta toxinin ileal loops after 6 h, treatment of ileal loops with as littleas 1 mg ml-1 purified CPB produced (see Fig. 8 andTable 1 for representative effects of a 4.5 mg ml-1 CPBdose) a haemorrhagic necrotizing enteritis identical toinfection with late log-phase cultures of wild-type CN3685(Fig. 8 and Table 1), except for the absence of largenumbers of bacteria seen in the lumen of loops treatedwith live cultures of the wild-type strain. In contrast, treat-ment of loops with 18 mg ml-1 purified alpha toxin had littleor no pathological effect (Table 1). Pre-incubation of thepurified beta toxin with an anti-CPB mAb, but not with aCPA-neutralizing mAb, eliminated the ability of this toxinpreparation to cause ileal loop pathology (Fig. 8 andTable 1).

Discussion

There has been uncertainty about the pathogenesis ofC. perfringens type C human and veterinary infections,particularly as recent studies established that type C iso-lates typically express several potent toxins (Fisher et al.,2006). One often-used approach for dissecting bacterialvirulence is reverse genetics, where genes are mutatedand the virulence effect of that mutation is then evaluated.For C. perfringens, reverse genetic approaches havebeen, until recently, applied only to study the pathogen-esis of the relatively simple type A isolates, where theyclearly identified the enterotoxin as a key contributor tothe GI pathogenesis of enterotoxin-positive type A isolates(Sarker et al., 1999) and established roles for both alphatoxin and PFO in the pathogenesis of gas gangrenecaused by enterotoxin-negative type A isolates (Awadet al., 1995; 2001). However, we recently used classicalallelic exchange approaches to inactivate the epsilontoxin gene (etx) in two type D disease isolates (Hugheset al., 2007). Trypsin-activated culture supernatants pre-pared from either type D isolate were highly cytotoxic forMDCK cells. In contrast, similarly activated supernatantsprepared from etx null mutants of either type D isolate didnot damage MDCK cells, strongly suggesting that epsilontoxin is important for the cytotoxic properties of those twotype D strains.

Despite that recent progress, there had not yet been asystematic dissection of toxin virulence contributions for atype B-E isolate prior to the current study. This was largely

Fig. 8. Histological damage in rabbit ileumafter a 6 h treatment with CPB (4.5 mg) or an8 h culture of wild-type CN3685 that had orhad not been pre-incubated with neutralizingCPB or CPA mAbs. Control loops, which wereinoculated with sterile TGY, loops inoculatedwith wild-type CN3685 that had beenpre-incubated with CPB mAb and loopsinoculated with purified CPB that had beenpre-incubated with CPB mAb all showedintact intestinal villi with a well-preservedepithelium and lamina propria. In contrast,loops inoculated with wild-type CN3685, loopsinoculated with CN3685 that had beenpre-incubated with a CPA mAb or loopstreated with purified CPB all showed similarhistological changes consisting of necrosisand loss of epithelium, necrosis of laminapropria, blunting of the villi, haemorrhage ofthe mucosa and diffuse neutrophilic infiltrationof mucosa and sub-mucosa. Sections werestained with haematoxylin and eosin andphotographed at 200¥ magnification. SeeTable 1 for the number of times thisexperiment was performed.



due to the complicated toxin genotype of these isolatesnot being readily amenable to classical allelic exchangemutagenesis approaches, which often require 1–2 yearsto produce a single C. perfringens toxin mutant. Thesetechnical limitations have limited our understanding of themolecular pathogenesis of type B-E infections, despitetheir medical, veterinary and biodefence importance(McClane et al., 2004).

Recently, we modified the TargeTron® (Sigma ChemicalCompany) vector for use in C. perfringens (Chen et al.,2005; 2007). Our pilot studies with type A isolates demon-strated that this approach produces stable, insertionalinactivation of target C. perfringens genes with great tar-geting specificity and high efficiency (Chen et al., 2005;2007). We hypothesized that this powerful new tool shouldallow dissection of the intestinal virulence contributions ofthe three toxins (CPA, PFO and CPB) produced by virtuallyall type C isolates during late log-phase vegetative growth(Fisher et al., 2006). Confirming that reverse geneticsapproaches are now feasible for studying the virulence ofall C. perfringens types, we have now successfully appliedthis approach to the type C disease isolate CN3685.

These analyses revealed that CPB is required for theintestinal virulence of CN3685 under the experimentalconditions employed for this study. In contrast, neither PFOnor CPA was found to make significant contributions toCN3685 intestinal virulence. These conclusions are basedon genetic evidence, where inactivation of the CN3685 cpbgene (but not the plc and pfoA genes) in two differentCN3685 mutants caused loss of all intestinal virulence inileal loops. In addition, the natural intestinal virulencephenotype of CN3685 could be substantially restored bycomplementing a cpb mutant to again produce CPB; i.e.,by fulfilling molecular Koch’s postulates, these data dem-onstrate a key role for CPB in CN3685 intestinal virulence.An independent, non-genetic line of evidence also impli-cated CPB in CN3685 intestinal virulence when it wasshown that pre-incubation with a CPB mAb completelyneutralized the intestinal pathology caused by CN3685.This neutralization was specific as an identical pre-incubation with a CPA-neutralizing mAb had no inhibitoryeffect on CN3685-induced intestinal pathology.

As CN3685 was isolated from a diseased sheep and is astrong producer of CPB, PFO and CPA (see Experimentalprocedures), our findings should have general relevancefor understanding type C enteric pathogenesis. However, itremains possible that other toxins besides CPB can alsocontribute to the intestinal virulence of some type C iso-lates, particularly when grown under different conditionsfrom those used for our study. For example, although> 50% of type C isolates are, like CN3685, cpb2-negativeand produce only CPB, PFO and PLC during late log-phase growth (Fisher et al., 2006), about one-third of typeC isolates also produce beta2 toxin under these growth

conditions. As purified beta2 toxin reportedly can inducehaemorrhagic necrosis in guinea pig ileal loops (Gibertet al., 1997), that toxin might contribute to the intestinalvirulence of late-log phase cultures of CPB2-producingtype C isolates. In addition, about 10–15% of type C strainsproduce CPE during sporulation (Fisher et al., 2006), sothe enterotoxin could contribute to enteric virulence ofCPE-positive type C strains if they were to sporulate in theintestines. Finally, some type C isolates were recentlyshown to produce the newly identified toxin named TpeL,which is a truncated homologue of C. difficile toxin A(Amimoto et al., 2007). Although CN3685 is tpeL +, it isunlikely that TpeL contributes to the CN3685 intestinalvirulence as tested in the current study as we found thatCN3685 expresses TpeL transcript in stationary phase, butnot during log-phase growth (J.E. Vidal and B.A. McClane,unpubl. obs.). However, TpeL could contribute to thepathogenesis of more chronic intestinal infections involv-ing tpeL + type C isolates like CN3685. Work to dissect theintestinal virulence contributions of additional type Ctoxins, particularly under other specific growth conditions,is currently underway in our laboratories.

The current results indicate that PFO and CPA makelimited, if any, contribution to the intestinal virulenceof type C isolate CN3685, at least under the tested condi-tions. The limited role, if any, of CPA in CN3685 intestinalvirulence is particularly interesting as (i) the plc gene of thistype C isolate encodes (relative to the classical plc gene ofmost type A isolates) an Ala→Asp change at CPA residue174, a Thr→Ala change at CPA residue 177 and aSer→Pro change at CPA residue 335, and (ii) those samesubstitutions are also present in a chymotrypsin-resistantCPA variant produced by C. perfringens type A bovinedisease isolates (Ginter et al., 1996) and encoded by C.perfringens type D isolates (V. Adams, R. Poon, M.Hughes, and J. Rood, unpublished results). An apparentlack of CPA involvement in CN3685 intestinal virulence issimilar to recent results from one of our laboratories,where a plc null mutant of a type A chicken necroticenteritis isolate was still able to cause GI disease in achicken GI disease model (Keyburn et al., 2006).However, despite this lack of genetic evidence supportingCPA or PFO as major contributors to C. perfringens GIdisease, vaccination studies have suggested that CPAdoes play at least a partial role in the virulence of type Anecrotic enteritis strains in a chicken disease model(Songer, 1996; Kulkarni et al., 2007). Given these contra-dictory findings and the importance of fully understandingC. perfringens virulence for vaccine development, CPAinvolvement in C. perfringens intestinal disease will clearlyrequire further studies using additional strains, animalmodels and growth conditions.

As mentioned earlier, there has been confusion in theliterature regarding whether, by itself, CPB can cause the



intestinal effects of type C infections. Therefore, a notablefinding of the current study was that highly purified CPBcan produce the same intestinal pathology as wild-typeCN3685. Furthermore, the intestinal pathology caused bypurified CPB could be completely blocked by a CPB-neutralizing, but not by a CPA-neutralizing, mAb. Onepossible explanation for the detection of CPB-induceddamage in our current studies, but not previous studies,was our addition of trypsin inhibitor to the purified CPBprior to its injection into ileal loops. Supporting this possi-bility, we failed to detect any ileal loop pathology withoutaddition of this inhibitor (F.A. Uzal et al., unpublished),which was added as CPB is highly sensitive to trypsin(McClane et al., 2004; Smedley et al., 2004).

Coupled with the cpb mutant observations, our purifiedCPB findings confirm that CPB is both required and suf-ficient for CN3685 intestinal virulence under our experi-mental conditions. When combined with our previousresults indicating that CPB is also the major contributor totype C supernatant lethality in the mouse intravenousinjection model (Fisher et al., 2006), these findingssupport the importance of CPB-neutralizing immunity forefficacious vaccination against type C intestinal disease.In particular, the current data suggest that, when inacti-vated crude culture supernatants are used to prepare atype C toxoid (as is typically the case for current type Cvaccines), they should be prepared from a strong CPB-producing strain in order to obtain a maximally protectivehuman or veterinary vaccine.

The two hallmarks of natural type C disease includenecrotizing enteritis (which was modelled in the currentstudy) or toxaemia, which generally follows the develop-ment of intestinal damage and results from absorption ofintestinal toxins into the circulation. Unfortunately, rabbitileal loop experiments terminate before type C isolate-induced toxaemia develops, so our current studies did notmodel this aspect of type C infections. We have shownpreviously that CPB is the major contributor to lethalityinduced by intravenous injection of sterile type C isolatesupernatants into mice (Fisher et al., 2006). However,another model involving longer-term animal infection willneed to be developed to definitively address the relativeimportance of CPB versus PFO or CPA for type C isolate-induced toxaemias.

Finally, returning to the use of TargeTron® technologyfor studying clostridial pathogenesis, our study has dem-onstrated several new developments. In pilot studiesdescribing targetron mutagenesis of a C. perfringens typeA isolate, we had used simple phenotypic screeningassays, i.e. loss of b-haemolysis on blood agar plates orloss of colony halos on egg yolk agar plates, to helpidentify (respectively) pfoA or plc mutants of type A iso-lates (Chen et al., 2005; 2007). We now show that simpleTargeTron® technology can be used to obtain mutants

even where no easy phenotypic screening assay isavailable. Specifically, after introduction of our clostridia-modified targetron-carrying plasmid to disrupt the cpbgene, 8 of the first 10 colonies screened by PCR hadtheir cpb gene disrupted by an intron insertion. While thisinsertional frequency likely varies among different genetargeting sequences, our results demonstrate thattargetron-derived mutants can be rapidly obtained, at highfrequency, in the absence of simple phenotypic screeningassays or marking of introns for selection.

The second technical advance for clostridial mutagen-esis demonstrated in this study was the use of a clostridial-modified TargeTron® to construct mutants carrying multipleinactivated genes. Specifically, the successful constructionof a double pfoA/plc mutant is reported here, but we haveused this same approach to quickly construct a triple toxinmutant of CN3685 (D.J. Fisher, S. Sayeed and B.A.McClane, unpublished). A recent study has shown that aTargeTron® derivative with a selectable resistance markercan also be used for insertional mutagenesis in severalclostridial species (Heap et al., 2007), but the absence ofselectable markers in our TargeTrons® facilitated rapidretargeting of the same intron vector to multiple toxingenes, i.e. had we relied on a selection approach, it wouldhave been necessary to construct multiple TargeTron®

vectors with different selectable markers.The final technical advance in clostridial TargeTron®

technology reported in this study concerns exploitation ofGroup II intron biology for C. perfringens complementa-tion purposes. After obtaining a cpb antisense mutant, werepeatedly tried to clone the cpb gene into a shuttle vectorfor classical complementation (Awad et al., 1995; Sarkeret al., 1999) of that mutant. However, the cpb gene provedexceptionally unstable in Escherichia coli. Therefore, weapplied a strategy previously reported for Staphylococcusaureus (Yao et al., 2006) that involved reintroduction of anLtrA-encoding plasmid into our BMC104 mutant to spliceout the sense-oriented intron insertion present in itsmRNA, which substantially restored CPB production. Thiscomplementation approach proved simple, rapid and par-ticularly useful for an unstable or toxic gene. However, assplicing of the intron is not 100% efficient (Yao et al.,2006), this approach yielded less CPB production thanoccurs with the wild-type parent CN3685. This lower levelof protein production could represent a problem for someapplications but the success of our CPB complementationstudies indicates that this can be a viable approach forC. perfringens virulence studies, particularly those involv-ing potent toxins. In summary, the usefulness of a Targe-Tron® for producing single or multiple C. perfringensmutants, and for complementation in some cases,makes this technology increasingly useful for studiesof C. perfringens biology and virulence. TargeTron®

mutagenesis is now being employed in additional studies



to dissect the virulence of C. perfringens type C isolateswith more complicated toxin genotypes than CN3685.

Experimental procedures

Bacterial strains, media and chemicals

Clostridium perfringens type C isolate CN3685 was providedby Russell Wilkinson (University of Melbourne), Australia.This isolate, originally from the Burroughs–Wellcome (BW)C. perfringens collection, had been isolated in 1954 fromperitoneal fluid of a sheep with struck. It has been previouslycharacterized (Fisher et al., 2006) for its genotype and toxinproduction, which showed that (under the growth conditionsused in our current study) this isolate produces a CPA activitytitre of 6.0, a PFO activity titre of 2.9 and ~13 mg ml-1 CPB.Those toxin production levels place CN3685 among thestrongest 25% of PLC producers, strongest 10% of PFOproducers and strongest 20% of CPB producers of the 45type C disease isolates surveyed in that previous study(Fisher et al., 2006).

To ensure culture purity, type C isolate CN3685 wasstreaked onto SFP agar [(Difco Laboratories), 0.04%D-cycloserine (Sigma-Aldrich)] and then grown overnight at37°C under anaerobic conditions. FTG (Fluid ThioglycolateMedium; Difco Laboratories) and TGY [3% Tryptic soy broth,(Becton-Dickinson), 2% glucose (Sigma-Aldrich), 1% yeastextract (Becton-Dickinson), 0.1% thioglycolate (Sigma-Aldrich)] medium were routinely used for growing brothcultures of CN3685 or derivatives of this strain. E. coli DH5awas grown in LB broth (1% Tryptone, 0.5% yeast extract and1% NaCl, pH 7.0) with shaking or LB agar at 37°C. All anti-biotics used in this study were purchased from the Sigma-Aldrich Chemical Company. All mutants and plasmids used inthis study are listed in Table 2.

The purified CPA used in ileal loop experiments was pur-chased from Sigma.

Purification of CPB protein

An isolated colony of CN3685 on SFP agar was inoculatedinto FTG and grown overnight at 37°C. A 0.1 ml aliquot of that

overnight culture was then transferred to 30 ml of TGY andgrown at 37°C for ~8 h. This 30 ml of culture was transferredto 3 l of fresh TGY and grown at 37°C for another ~8 h. Theculture was then chilled immediately on ice and centrifuged at10 000 g for 20 min. Proteins in the culture supernatant wereprecipitated using 60% ammonium sulphate (Fisher Scien-tific), with constant stirring, at 4°C for ~1 h. The precipitatewas then collected by centrifugation at 10 000 g for 30 min.The pellet was re-suspended in 30 mM Tris-HCl buffer(pH 7.5) and dialysed overnight against the same buffer, withseveral changes, at 4°C. The dialysate was filtered througha 0.45 mm filter (Millipore) and loaded onto a DEAE-CL6BSepharose column (Sigma). This column was pre-equilibrated with 30 mM Tris-HCl buffer (pH 7.5) in an ÄKTAprime system (Amersham Bioscience). After loading thesample, the DEAE-CL6B column was washed with 45 ml of30 mM Tris-HCl buffer (pH 7.5) and bound CPB was theneluted from the column using 0.1 M NaCl in 30 mM Tris-HClbuffer (pH 7.5). Fractions were assessed for the presence ofCPB by Western blotting using a mouse monoclonal anti-CPB antibody obtained from Dr P. Hauer. CPB purity wasanalysed by SDS-PAGE, which showed the preparation to be~95% homogeneous. Fractions containing the purified CPBwere pooled and dialysed with ice-cold phosphate-bufferedsaline (PBS, pH 7.4) at 4°C. The CPB concentration in thefinal preparation was estimated by Lowry assay using bovineserum albumin as the standard (Lowry et al., 1951).

Construction of a CN3685 cpb(antisense) null mutant

To construct an initial CN3685 mutant with an inactivated cpbgene, the wild-type cpb gene sequence was first entered intothe Intron prediction program (http://www.sigma-genosys.com/targetron/). This program predicted seven insertion sitesacross the ~1 kb cpb gene. For maximal cpb disruption, theantisense orientation intron insertion site between nucle-otides 154|155 of the cpb ORF was selected for initial introntargeting. The primers listed in Table S1 (IBS, EBS1-d andEBS2) were used to generate a 350 bp intron targetingsequence to this cpb ORF site. The amplified 350 bp frag-ment was then digested with HindIII and BsrGI and ligatedinto pJIR750ai [a TargeTron®-carrying vector we created pre-viously to construct a plc null mutant (Chen et al., 2005) in

Table 2. Strains, mutants and plasmids used in this study.

Description Origin

Isolate nameCN3685 Wild type (cpb +, plc +, pfoA +, tpeL +) Diseased sheep/from Russell WilkinsonBMC100 CN3685::cpb(antisense) This studyBMC101 CN3685::plc This studyBMC102 CN3685::pfoA This studyBMC103 CN3685::plc/pfoA This studyBMC104 CN3685::cpb(sense) This studyBMC105 CN3685::cpb(sense) + pJIR750cpbi This studyBMC106 CN3685::cpb(sense) + pJIR750 This study

Plasmid namepJIR750 E. coli–C. perfringens shuttle vector Bannum and Rood (1993)pJIR750ai pJIR750 with plc targeted intron in antisense orientation Chen et al. (2005)pJIR750bi pJIR750 with cpb targeted intron in antisense orientation This studypJIR750qi pJIR750 with pfoA targeted intron in antisense orientation Chen et al. (2007)pJIR750bi-s pJIR750 with cpb targeted intron in sense orientation This study



http://www.sigma-genosys.com/targetron

http://www.sigma-genosys.com/targetron

type A isolate ATCC3624], which had been similarly digestedwith the same two restriction enzymes. The resultantdonor plasmid (named pJIR750cpbi), which now carried acpb-targeted intron, was electroporated into wild-typeC. perfringens type C isolate CN3685, as described previ-ously (Chen et al., 2005). Transformants were plated ontoSFP agar plates containing 15 mg ml-1 chloramphenicol.Colonies were then PCR screened, as described in a lattersection, using cpb-specific primers (Table S2) and their cpbgene was sequenced to confirm the identity of putative trans-formants carrying an intron-disrupted cpb gene. A mutant(BMC100) shown to carry a cpb intron insertion was subcul-tured in TGY medium without antibiotics for ~10 days, withdaily subculturing, to cure the intron-carrying donor plasmidpJIR750cpbi. Curing was initially shown by lack of growth onchloramphenicol-containing SFP plates and then confirmedby Southern blotting, which demonstrated the presence of asingle intron in the mutant (Fig. 2).

Construction of a CN3685 cpb(sense) null mutant

The intron prediction program also identified a second site forGroup II intron insertion, in the sense orientation, betweennucleotides 182|183 of the cpb ORF. This site was used totarget a sense-oriented intron insertion into the cpb gene,creating a second cpb null mutant (BMC104) that could laterbe complemented at the mRNA level (see next section). Theprimers listed in Table S1 (IBS-s, EBS1-ds and EBS2-s) wereused to generate a 350 bp cpb targeting sequence for thisintron insertion between nucleotides 182|183 of cpb ORF.The amplified 350 bp fragment was then digested with HindIIIand BsrGI and ligated into pJIR750ai, which had been simi-larly digested with these two restriction enzymes, to createplasmid pJIR750cpbi-s. This intron donor plasmid was elec-troporated into wild-type CN3685, as described previously(Chen et al., 2005). Transformants were plated onto SFPagar plates containing 15 mg ml-1 chloramphenicol. Colonieswere then PCR screened, as described in a later section,using cpb-specific primers (Table S2) and their cpb gene wassequenced to identify putative transformants carrying anintron-disrupted cpb gene. A transformant (BMC104) carryingthis intron insertion in its cpb gene was subcultured in TGYmedium without antibiotics for ~10 days, with daily subcultur-ing, to cure the intron-carrying donor plasmid pJIR750cpbi-s.Curing was initially shown by lack of growth onchloramphenicol-containing SFP plates and then confirmedby Southern blotting, which demonstrated the presence of asingle intron in the mutant (not shown).

Complementation of the CN3685 cpb(sense) null mutant

CPB complementation of the cpb(sense) null mutant BMC104was achieved at the mRNA level by removing the intron fromdisrupted cpb mRNA, thus restoring a functional cpb tran-script (Yao et al., 2006). Removal of this intron was achievedby reintroducing pJIR750cpbi-s, which encodes the LtrAprotein required to splice out the intron insertion from thedisrupted cpb mRNA (Yao et al., 2006), into BMC104. Trans-formants were selected on SFP agar containing 15 mg ml-1

chloramphenicol.

Construction of a CN3685 plc null mutant

The previously constructed plc intron donor plasmidpJIR750ai (Chen et al., 2005), which specifies insertion (in anantisense orientation) of a Group II intron between nucle-otides 50|51 of the plc ORF, was used to inactivate thechromosomal alpha toxin (plc) gene of CN3685. After elec-troporation of pJIR750ai, transformants were grown on BHIagar plates containing 15 mg ml-1 chloramphenicol and eggyolk [4% (v/v) egg yolk]. This medium helped to identifyputative plc null mutants by loss of the white halo zone thatnormally surrounds C. perfringens colonies due to the phos-pholipase C activity of alpha toxin (Awad et al., 2001). Puta-tive plc null mutants were screened by PCR (described in thesection below) using the primers plcF and plcR (sequenceslisted in Table S2) and the plc gene of one putative mutantBMC101 was sequenced to demonstrate specific intron dis-ruption of its plc gene. The confirmed mutant carrying anintron insertion in its plc gene was then grown in TGY mediumwithout antibiotic and subcultured daily for ~10 days to curethe intron-carrying donor plasmid pJIR750ai. Curing was ini-tially shown by lack of growth on chloramphenicol-containingSFP plates and then confirmed by Southern blotting, whichshowed the presence of a single intron in the mutant (Fig. 2).

BMC102, a pfoA null mutant of CN3685

BMC102, pfoA null mutant of CN3685, was prepared usingthe previously constructed plasmid pJIR750qi, which targetsan intron insertion (in the antisense orientation) into the per-fringolysin O gene (pfoA) at nucleotide site 32|33 (Chen et al.,2007). This plasmid was electroporated into CN3685 andtransformants were then plated onto blood agar plates (TSAII with 5% sheep blood, BD BBL™). Colonies lacking theinner zone of b-haemolysis indicative of PFO activity werePCR screened with specific primers pfoAF and pfoA1222R(Table S2) and their pfoA gene was sequenced to confirm anintron disruption of this gene. One such mutant BMC102 wasthen grown in TGY medium without antibiotics and subcul-tured daily for ~10 days to cure the intron-carrying donorplasmid pJIR750qi. Curing was initially shown by lack ofgrowth on chloramphenicol-containing SFP plates and thenconfirmed by Southern blotting, which demonstrated thepresence of a single intron in the mutant (Fig. 2).

Construction of a CN3685 plc and pfoA doublenull mutant

To prepare a double plc/pfoA null mutant of CN3685, a plcnull mutation was introduced into BMC102 (created above)using the plc intron donor plasmid pJIR750ai (Chen et al.,2005). Transformants were then grown on BHI agar platescontaining 15 mg ml-1 chloramphenicol and egg yolk [4% (v/v)egg yolk] to identify putative plc null mutants by monitoringloss of the white halo zone around transformant colonies(Awad et al., 2001). The presence of introns in both the plcand pfoA genes of the BMC103 double mutant was confirmedby PCR (described in the section below using the sameprimer sets as used before) and by DNA sequencing its plcand pfoA genes. The confirmed plc/pfoA double null mutant



was grown in TGY medium without antibiotics and subcul-tured daily for ~10 days to cure the intron-carrying donorplasmid pJIR750ai. Curing was initially shown by lack ofgrowth on chloramphenicol-containing SFP plates and thenconfirmed by Southern blotting, which demonstrated thepresence of only two introns in the mutant (Fig. 2).

PCR screening of intron insertion into cpb, plc orpfoA genes

Template DNA for all PCR reactions was obtained fromcolony lysates, which had been prepared as described pre-viously (Wen and McClane, 2004). Each PCR mixture con-tained 5 ml of template DNA, 40 ml of TAQ Complete 1.1Master Mix (Gene Choice) and 2.5 ml of each primer pair(1 mM final concentration). The primers listed in Table S2were used to screen for intron insertions into the cpb, plc orpfoA genes. PCR reactions were performed in a Techne(Burkhardtsdorf, Germany) thermocycler using the followingPCR amplification conditions: 94°C for 3 min, 35 cycles of94°C for 1 min, 55°C for 1 min and 72°C for 2 min, followedby a 5 min extension at 72°C. PCR products were run on a1% gel and stained with ethidium bromide for visualization.

Sequencing of the null mutants

DNA was isolated using the MasterPureTm Gram-PositiveDNA Purification Kit (Epicentre Biotechnologies, Wisconsin).Sequencing was performed at the University of Pittsburghcore sequencing facility (http://www.genetics.pitt.edu/services.html) and sequences were then analysed usingBioEdit (Hall, 1999) and Vector NTI (Invitrogen).

Southern hybridization

For Southern blotting, C. perfringens genomic DNA from wild-type CN3685 or the toxin null mutants was isolated using theMasterPure™ Gram-Positive DNA Purification Kit (EpicentreBiotechnologies, Wisconsin). This DNA was then digestedwith EcoRI and run on a 0.8% agarose gel. After transfer to anylon membrane (Roche), the blot was probed with adigoxigenin-labelled probe specific for the intron sequence.As described previously (Miyamoto et al., 2006), this probewas prepared using the primer pair KO-IBS and KO-EBS1d(Table S3) and a DIG labelling kit obtained from RocheApplied Science. CSPD substrate (Roche Applied Science)was used for detection of DIG-labelled hybridized probes,according to the manufacturer’s instruction.

CPB Western immunobloting

Wild-type CN3685 and each single or double-toxin nullmutant carrying one or more antisense-oriented introns werefirst grown overnight at 37°C in FTG medium. A 0.1 ml aliquotof each overnight culture was transferred into 10 ml of TGYmedium, which was then grown overnight at 37°C. Sampleswere collected and supernatants were mixed with an equalvolume of loading buffer and electrophoresed on a 10% SDS-polyacrylamide gel. The separated proteins were transferred

onto nitrocellulose membrane and Western blotted for detec-tion of CPB using mouse monoclonal anti-CPB antibody, asdescribed previously (Fisher et al., 2006).

Wild-type CN3685, BMC104 and BMC105 were eachgrown overnight at 37°C in FTG. The complement was grownin the presence of 15 mg ml-1 chroramphenicol to retain theLtrA-encoding plasmid. A 1 ml aliquot of each culture wasthen transferred into 9 ml of TGY and grown at 37°C for 8 hbefore transfer to 30°C for an overnight growth. Culturesupernatants were assayed for CPB production by Westernimmunobloting as described above.

The presence of CPB in ileal loop fluids was evaluated bycollecting luminal fluids and subjecting it to CPB Westernblotting as described above.

RNA extraction, Northern hybridization and RT-PCR

Wild-type CN3685, BMC104 and the BMC105 were eachgrown for 8 h at 37°C and then overnight at 30°C asdescribed above. About 3 ml of each culture was then usedfor RNA extraction with the RiboPure™-Bacteria kit fromAmbion according to the manufacturer’s instructions. RNAsamples were gel electrophoresed (1% agarose gel) andtransferred to a nylon membrane (Roche) according to themanufacturer’s instructions (NorthenMax®, Ambion). The blotwas probed with a dioxigenin-labelled cpb-specific probeprepared, as described previously (Miyamoto et al., 2006). ADIG-labelled probe-specific for internal cpb sequences wasprepared using primers cpbF and cpbORF-R (Table S4).DIG-labelling and detection reagents were obtained fromRoche Applied Science. CSPD substrate (Roche AppliedScience) was used for detection of DIG-labelled hybridizationprobes, according to the manufacturer’s instructions. TheRNA samples were checked for the presence of contaminat-ing DNA by performing PCR with the same primers and Taqpolymerase but no product was amplified.

For RT-PCR analyses, RNA samples extracted from wild-type CN3685, BMC104 or BMC105 were treated with DNaseI at 37°C for 30 min. RT-PCR reactions were then carried outon those DNase-treated RNA samples using the AccesQuickRT-PCR system (Promega). Briefly, 100 ng of each RNAsample was reverse transcribed to cDNA at 45°C for 1 h andthen used as template for PCR (denaturing at 94°C for 1 min,annealing at 55°C for 1 min and extension at 72°C for 1 min)with primers targeting cpb sequences, intron sequences orintron and downstream cpb sequences (Table S5). ControlRT-PCR reactions were similarly performed, except for theomission of reverse transcriptase. As an additional control, aPCR amplifying each gene was performed by adding, into theRT-PCR reaction cocktail, DNA extracted from each strainusing MasterPure Gram-Positive DNA Purification Kit(Epicentre Biotechnologies).

Preparation of type C cultures for virulence testing inthe rabbit ileal loop assay

Wild-type CN3685 and all mutants (except the cpb(sense)

mutant and its complementing strain) were grown for 8 h inTGY at 37°C. A 1 ml aliquot of those overnight cultures wasused to inoculate a rabbit ileal loop. To test for complemen-



http://www.genetics.pitt.edu/services.html

http://www.genetics.pitt.edu/services.html

tation, 1 ml of wild-type CN3685, BMC104, BMC105 andBMC106 were inoculated into 9 ml of TGY and then grown at37°C for ~8 h. Those tubes were then transferred to 30°C forovernight incubation to induce the expression of the LtrAprotein required for trans-splicing removal of the sense-oriented intron from the cpb mRNA. A 1 ml aliquot of each ofthose cultures was inoculated into the rabbit ileal loop. Insome experiments, cultures (0.8 ml) were pre-incubated atroom temperature for 45 min with 0.2 ml of either anti-CPBmAb or anti-CPA mAb prior to inoculation into a rabbit ilealloop.

Fluid accumulation in rabbit ileal loops by wild-typeCN3685 and toxin mutants

Young adult, male or female, New Zealand White rabbits(Charles River, CA) were deprived of food and waterovernight before surgery. Pre-medications includedacepromazine, xylazine and burprenorphine. Induction ofanaesthesia was performed with ketamine and maintainedwith inhalatory isofluorane. A laparotomy was performed viathe mid-line and the small intestine was exposed. Lengths(~2 cm) of ileum were isolated by ligation, avoiding interfer-ence with the blood supply and leaving an empty segment ofgut between the loops. A 1 ml culture aliquot of either wild-type CN3685 or a toxin mutant, prepared as describedabove, was injected into an ileal loop of each experimentalanimals. Additional loops (controls) received a similar volumeof sterile TGY medium, pure alpha toxin (18 mg ml-1) orpure beta toxin (usually 4.5 mg ml-1) plus trypsin inhibitor(150 mg ml-1). In some studies, the same amount of purifiedbeta toxin (in 0.8 ml containing trypsin inhibitor) was pre-incubated for 45 min with 0.2 ml of neutralizing anti-CPB mAbor anti-CPA mAb (provided by Dr P. Hauer). Care was takento avoid over-distension of the loops. During surgery, theserosal surface of the loops was kept wet by frequent soakingwith normal saline solution. The abdominal incision wasclosed by separate muscle and skin sutures and the animalswere kept deeply anaesthetized throughout the experiment.After 6 h, the rabbits were euthanized with an overdose ofsodium barbiturate (Beuthanasia, Schering-Plough AnimalHealth, Kenilworth, NJ), the abdominal cavity was re-opened,and the small intestinal loops were excised in the same orderthat they had been inoculated. Fluid accumulation was mea-sured by weight in grams and loop length was recorded incentimetre. For each loop, a fluid weight/length ratio wascalculated. All procedures were reviewed and approved bythe University of California, Davis Committee for Animal Careand Use (permit 04-11593).

Preparation of ileal loop tissue for histological analysis

After gross analysis, each treated loop was immersed in 10%buffered, pH 7.4 formalin. Following 24–48 h of formalin fixa-tion, the tissues were dehydrated through graded alcohols toxylene and finally embedded in paraffin wax. Sections (4 mmthick) were cut using a microtome and stained conventionallyby haematoxylin and eosin. Sections were then photomicro-graphed using an Olympus microscope (Tokyo, Japan) at100¥ or 200¥ final magnification.

Determination of C. perfringens cell numbers inoculatedinto loops and present in loops after a 6 h incubation

To determine the colony-forming units (cfu) present per ml ofinoculum, a standard dilution-plating technique was used.Serial dilutions of the inocula were prepared in PBS and 100 mlof each dilution was plated onto blood agar plates.After 24 h ofanaerobic incubation at 37°C, the total number of colonies oneach plate was determined and a cfu ml-1 calculation wasmade based on that number and the dilution factor used.

After a 6 h incubation in rabbits, the loops were openedand the fluid content was collected using aseptic technique.The number of C. perfringens cells present in the collectedintestinal fluids was then determined as described above.

Statistical analyses

Statistical significance of pathology results was evaluatedusing the Student’s t-test, performed with the InStat program,version 2.03 (GraphPad).

Acknowledgements

This work was generously supported by Grants R01AI056177-04 (McClane) and R03 AI067515-01 (Chen) fromNational Institute of Allergy and Infectious Diseases and byT32 AI060525-01A1 (Ruth L. Kirschstein National ServiceAward to Sayeed). The authors thank Dr P. Hauer for supply-ing monoclonal antibodies against CPB and CPA andJackelin Parker and Jim Cravotta for help with anaesthesiaand animal experiments.

References

Allen, S.D., and Baron, E.J. (1991) Clostridium. In Manual ofClinical Microbiology. Balows, A., Hausler, W.J., Jr,Herrmann, K.L., Isenberg, H.D., and Shadomy, H.J. (eds).Washington, DC: American Society of Microbiology,pp. 505–521.

Amimoto, K., Noro, T., Oishi, E., and Shimuzu, M. (2007) Anovel toxin homologous to large clostridial cytotoxins foundin culture supernatant of Clostridium perfringens type C.Microbiology 153: 1198–1206.

Awad, M.M., Bryant, A.E., Stevens, D.L., and Rood, J.I.(1995) Virulence studies on chromosomal alpha-toxin andtheta toxin mutants constructed by allelic exchange providegenetic evidence for the essential role of alpha-toxin inClostridium perfringens-mediated gas gangrene. MolMicrobiol 15: 191–202.

Awad, M.M., Ellemor, D.M., Boyd, R.L., Emmins, J.J., andRood, J.I. (2001) Synergistic effects of alpha-toxin andperfringolysin O in Clostridium perfringens-mediated gasgangrene. Infect Immun 69: 7904–7910.

Bannum, T.L., and Rood, J.I. (1993) Clostridium perfringens–Escherichia coli shuttle vectors that carry single antibioticresistance determinants. Plasmid 29: 233–235.

Bergeland, M.E. (1972) Pathogenesis and immunity ofClostridium perfringens type C enteritis in swine. J Am VetMed Assoc 160: 568–571.

Chen, Y., McClane, B.A., Fisher, D.J., Rood, J.I., and Gupta,



P. (2005) Construction of an alpha toxin gene knockoutmutant of Clostridium perfringens type A by use of a mobilegroup II intron. Appl Environ Microbiol 71: 7542–7547.

Chen, Y., Caruso, L., McClane, B.A., Fisher, D.J., and Gupta,P. (2007) Disruption of a toxin gene by introduction of aforeign gene into the chromosome of Clostridium perfrin-gens using targetron-induced mutagenesis. Plasmid 58:182–189.

Fisher, D.J., Fernandez Miyakawa, M.F., Sayeed, S., Poon,R., Adams, V., Rood, J.I., et al. (2006) Dissecting the con-tributions of Clostridium perfringens Type C toxins to lethal-ity in the mouse intravenous injection model. Infect Immun74: 5200–5210.

Gibert, M., Jolivet-Renaud C., and Popoff M.R. (1997) Beta2toxin, a novel toxin produced by Clostridium perfringens.Gene 203: 65–73.

Ginter, A., Williamson, E.D., Dessy, F., Coppe, P., Bullifent,H., Howells, A., and Titball, R.W. (1996) Molecular varia-tion between the alpha-toxins from the type strain (NCTC8237) and clinical isolates of Clostridium perfringens asso-ciated with disease in man and animals. Microbiology 42:191–198.

Gui, L., Subramony, C., Fratkin, J., and Hughson, M.D.(2002) Fatal enteritis necroticans (Pigbel) in a diabeticadult. Mod Pathol 15: 66–70.

Hall, T.A. (1999) Bioedit: a user-friendly biological sequencealignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp Ser 41: 95–98.

Heap, J.T., Pennington, O.J., Cartman, S.T., Carter, G.P.,and Minton, N.P. (2007) The ClosTron: a universal geneknock-out system for the genus Clostridium. J MicrobiolMethods 70: 452–464.

Hughes, M.L., Poon, R., Adams, V., Sayeed, S., Saputo, J.,Uzal, F.A., et al. (2007) Epsilon toxin plasmids of Clo-stridium perfringens are conjugative. J Bacteriol 189: 7531–7538.

Hunter, S.E., Brown, J.E., Oyston, P.C., Sakurai, J., andTitball, R.W. (1993) Molecular genetic analysis of beta-toxin of Clostridium perfringens reveals sequence homol-ogy with alpha-toxin, gamma-toxin, and leukocidin ofStaphylococcus aureus. Infect Immun 61: 3958–3965.

Johnson, S., and Gerding, D.N. (1997) Enterotoxemic infec-tions. In The Clostridia. Molecular Biology and Pathogen-esis. Rood, J.I., McClane, B.A., Songer, J.G., and Titball,R.W. (eds). London: Academic press, pp. 117–140.

Keyburn, A.L., Sheedy, S.A., Ford, M.E., Williamson, M.M.,Awad, M.M., Rood, J.I., and Moore, R.J. (2006) Alpha-toxinof Clostridium perfringens is not an essential virulencefactor in necrotic enteritis in chickens. Infect Immun 74:6496–6500.

Kulkarni, R.R., Parreria, V.R., Sharif, S., and Prescott, J.F.(2007) Immunization of broiler chickens againstClostridium perfringens-induced necrotic enteritis. ClinVaccine Immunol 9: 1070–1077.

Lawrence, G.W. (1997) The pathogenesis of enteritis necroti-cans. In The Clostridia. Molecular Biology and Pathogen-esis. Rood, J.I., McClane, B.A., Songer, J.G., and Titball,R.W. (eds). London: Academic press, pp. 197–207.

Lee, H., Bueschel, D.M., and Nesheim, S.R. (2000) Enteritisnecroticans (Pigbel) in a diabetic child. N Engl J Med 342:1250–1253.

Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.J.(1951) Protein measurement with the Folin phenol reagent.J Biol Chem 193: 265–275.

Lyras, D., and Rood, J.I. (2006) Clostridial genetics. In Gram-Positive Pathogens. Fischetti, V.A., Novick, R.P., Ferretti,J.J., Portnoy, D.A., and Rood, J.I. (eds). Washington, DC:American Society for Microbiology Press, pp. 672–687.

McClane, B.A., Uzal, F.A., Miyakawa, M.F., Lyerly, D., andWilkins, T. (2004) The enterotoxic clostridia. In TheProkaryotes. Dworkin, M., Falkow, S., Rosenberg, E.,Schleifer, K.-H., and Stackebrandt, E. (eds). New York:Springer New York, pp. 698–752.

Matsuda, T., Okada, Y., Tanabe, Y., Shimizu, Y., Nagashima,K., Sakurai, J., et al. (2007) Enteritis necroticans ‘pigbel’ ina Japanese diabetic adult. Pathol Int 57: 622–626.

Matsuura, M., Saldanha, R., Ma, H., Wank, H., Yang, J.,Mohr, G., et al. (1997) A bacterial group II intron encodingreverse transcriptase, maturase, and DNA endonucleaseactivities: biochemical demonstration of maturase activityand insertion of new genetic information within the intron.Genes Dev 11: 2910–2924.

Miyamoto, K., Fisher, D.J., Li, J., Sayeed, S., Akimoto, S., andMcClane, B.A. (2006) Complete sequencing and diversityanalysis of the enterotoxin-encoding plasmids inClostridium perfringens Type A non-food-borne human gas-trointestinal disease isolates. J Bacteriol 188: 1585–1598.

Sarker, M.R., Carman, R.J., and McClane, B.A. (1999) Inac-tivation of the gene (cpe) encoding Clostridium perfringensenterotoxin eliminates the ability of two-cpe-positive C. per-fringens type A human gastrointestinal disease isolates toaffect rabbit ileal loops. Mol Microbiol 33: 946–958.

Smedley, J.G., 3rd, Fisher, D.J., Sayeed, S., Chakrabarti, G.,and McClane, B.A. (2004) The enteric toxins of Clostridiumperfringens. Rev Physiol Biochem Pharmacol 152: 183–204.

Sobel, J., Mixter, C.G., Kolhe, P., Gupta, A., Guarner, J.,Zaki, S., et al. (2005) Necrotizing enterocolitis associatedwith Clostridium perfringens type A in previously healthyNorth American adults. J Am Coll Surg 201: 48–56.

Songer, J.G. (1996) Clostridial enteric diseases of domesticanimals. Clin Microbiol Rev 9: 216–234.

Songer, J.G. (1997) Bacterial phospholipases and their rolein virulence. Trends Microbiol 5: 156–161.

Steinthorsdottir, V., Fridriksdottir, V., Gunnarsson, E., andAndresson, O.S. (1998) Site-directed mutagenesis ofClostridium perfringens beta-toxin: expression of wild-typeand mutant toxins in Bacillus subtilis. FEMS Microbiol Lett158: 17–23.

Titball, R.W. (1998) Bacterial phospholipases. Symp Ser SocAppl Microbiol 27: 127S–137S.

Tweten, R.K. (2001) Clostridium perfringens beta-toxin andClostridium septicum alpha toxin: their mechanisms andpossible role in pathogenesis. Vet Microbiol 82: 1–9.

Wen, Q., and McClane, B.A. (2004) Detection of enterotoxi-genic Clostridium perfringens type A isolates in Americanretail foods. Appl Environ Microbiol 70: 2685–2691.

Yao, J., Zhong, J., Fang, Y., Geisinger, E., Novick, R.P., andLambowitz, A.M. (2006) Use of targetrons to disrupt essen-tial and nonessential genes in Staphylococcus aureusreveals temperature sensitivity of Ll.LtrB group II intronsplicing. RNA 12: 1271–1281.



Supplementary material

This material is available as part of the online article from:http://www.blackwell-synergy.com/doi/abs/10.1111/j.1365-2958.2007.06007.x(This link will take you to the article abstract).

Please note: Blackwell Publishing is not responsible for thecontent or functionality of any supplementary materials sup-plied by the authors. Any queries (other than missing mate-rial) should be directed to the corresponding author for thearticle.



http://www.blackwell-synergy.com/doi/abs/10.1111

EspC translocation into epithelial cells byenteropathogenic Escherichia coli requiresa concerted participation of type V and IIIsecretion systems

Jorge E. Vidal and Fernando Navarro-García*Department of Cell Biology, Centro de Investigación yde Estudios Avanzados del IPN (CINVESTAV-IPN),Ap. Postal 14-740, 07000 México City, Mexico.

Summary

EspC is a non-locus of enterocyte effacement (LEE)-encoded autotransporter protein secreted by entero-pathogenic Escherichia coli (EPEC) that causes acytopathic effect on epithelial cells, including cytosk-eletal damage. EspC cytotoxicity depends on its inter-nalization and functional serine protease motif. Herewe show that during EPEC infection, EspC is secretedfrom the bacteria by the type V secretion system(T5SS) and then it is efficiently translocated into theepithelial cells through the type III secretion system(T3SS) translocon. By dissecting this mechanism, wefound that EspC internalization during EPEC–hostcell interaction occurs after 1 h, unlike purified EspC(8 h). LEE pathogenicity island is involved in specificEspC translocation as three espC-transformed at-taching and effacing (AE) pathogens translocatedEspC into the cells. A role for effectors and otherfactors involved in the intimate adherence encoded inLEE were discarded by using an exogenous EspCinternalization model. In this model, an isogenicEPEC DespC strain allows the efficient internalizationof purified EspC. Moreover, isogenic mutants in T3SSwere unable to translocate endogenous and exog-enous EspC into epithelial cells, as EspC–EspA inter-action is required. These data show for the first timethe efficient delivery of an autotransporter proteininside the epithelial cells by EPEC, through coopera-tion between T5SS and T3SS.

Introduction

Enteropathogenic Escherichia coli (EPEC) infection is aleading cause of infantile diarrhoea in developing coun-

tries which can be severe and lethal (Kaper et al., 2004).EPEC is the prototype of a family of pathogens that elicita histopathological lesion formed at the mucosal intestinalsurface that seems a pedestal-like structure, known asattaching and effacing (AE) lesion (Moon et al., 1983).The AE pathogens include enterohaemorrhagic E. coli(EHEC), rabbit EPEC (REPEC) and Citrobacter roden-tium, which infects mice. The genes responsible for theAE phenotype are located in a 35.6 kb pathogenicityisland (PAI) termed LEE (locus of enterocyte effacement)(McDaniel et al., 1995). LEE is organized into five poly-cistronic operons (LEE1 to LEE5). The LEE1, LEE2 andLEE3 operons encode the type III secretion system(T3SS) and a global regulator Ler (LEE-encodedregulator). LEE4 encodes T3SS-secreted proteins EspA,EspB and EspD (EPEC-secreted protein) that are alsocomponents of the translocation apparatus by which othereffector proteins are translocated into the cell. LEE5encodes intimin, Tir and the Tir-chaperone CesT (Elliottet al., 1998). Tir is injected by the translocation apparatusdirectly to the cell and is inserted in the membrane expos-ing an extracellular domain that is recognized by intimin(an EPEC membrane adhesin). Intimin–Tir interactionleads to intimate adherence and pedestal formationbeneath adherent bacteria (Kenny et al., 1997). Duringinfection, other effector proteins are translocated into thecell (EspF, EspG, EspH, Map and EspI), and they interferewith different aspects of the cell physiology (Elliott et al.,2001; McNamara et al., 2001). Notably, there are severalnewly identified non-LEE-encoded effectors in AE patho-gens that are translocated by the T3SS, including Cif,NleA/EspI, TccP/EspFU, EspJ, NleB, NleC and NleD(Marches et al., 2003; Campellone et al., 2004; Garmen-dia et al., 2004; Gruenheid et al., 2004; Mundy et al.,2004; Dahan et al., 2005; Marches et al., 2005; Kellyet al., 2006).

Thus, the mechanism of protein translocation into thecell plays a central role in EPEC virulence. Recently, wehave found that a 110 kDa protein, named EspC, causescytotoxic effects including cytoskeletal damage; theseeffects depend on the EspC internalization and on itsfunctional serine protease motif (Navarro-Garcia et al.,2004). EspC is a non-LEE-encoded protein that is

Received 21 February, 2008; revised 9 May, 2008; accepted 23 May,2008. *For correspondence. E-mail [email protected];Tel. (+52) 55 5747 3990. Fax (+52) 55 5747 3393.

Cellular Microbiology (2008) 10(10), 1975–1986 doi:10.1111/j.1462-5822.2008.01181.xFirst published online 8 July 2008

© 2008 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd

mailto:[email protected]

encoded in a second PAI of EPEC, and unlike proteinssecreted by the T3SS, EspC secretion is mediated by thetype V secretion system (T5SS). Proteins secreted by thissystem receive also the name of autotransporter proteinsbecause the pre-proteins promote their own secretionthrough the inner and outer membranes by using twoprocessing domains, the signal sequence (SS) and thetranslocation unit (TU) (Stein et al., 1996; Mellies et al.,2001; Henderson et al., 2004), to secrete the passengerdomain (PD). In the case of EspC, these domains com-prise from amino acids 1 to 53 (SS), 54 to 1030 (PD) and1031 to 1306 (TU). Interestingly, regulation of espC iscoupled to the global regulator Ler localized in LEE andthat controls virulence gene expression during EPECpathogenesis, including those encoding the T3SS-secreted Esp proteins, Tir and intimin (Mellies et al., 1999;Elliott et al., 2000). However, a deletion mutant in espC byallelic exchange seems to be indistinguishable from itsisogenic parent for adherence, invasion, actin rearrange-ment and Tir phosphorylation, events that are crucial forA/E lesion formation (Stein et al., 1996). Despite this, adirect role of EspC in EPEC pathogenesis and its relationwith other bacterial virulence factors during EPEC infec-tion have not been studied.

We have explored the EspC internalization mechanisminto epithelial cells, as this is a key event in inducing thecytoskeletal damage that might explain the cytopathiceffect caused by EPEC. We have shown that EspC, innon-physiological conditions (i.e. as a purified protein),is not efficiently internalized, because no receptor isinvolved in its uptake and no intracellular traffic isrequired. Additionally, the physiologically secreted EspCby EPEC, which is enhanced in tissue culture media andby cell contact, is efficiently internalized during EPEC andepithelial cell interaction (Vidal and Navarro-Garcia,2006). Here, we demonstrate that EspC translocation intoepithelial cells involves the T3SS translocon, strongly indi-cating cooperation between two secretion systems, T5SSand T3SS.

Results

Efficient EspC entry into epithelial cells depends onLEE-encoded factors

To investigate EspC internalization dependency duringthe EPEC infection, HEp-2 cells were infected withEPEC or other AE pathogens (carrying LEE locus) suchas EHEC and REPEC, which were transformed with theespC gene. Anti-EspC antibody detected EspC in thecytoplasmic fraction from cells infected for 4 h withEPEC, EHEC/espC or REPEC/espC but not in cellsinfected with wild-type EHEC, REPEC (as these strainsdo not have espC) or E. coli HB101, an avirulent labo-

ratory strain (Fig. 1A). These results suggest that EspCinternalization appears to be related to LEE-encodedfactors, thereby an isogenic EPEC mutant in escN,which is unable to secrete any of the T3SS-dependentsecreted proteins, was used to infect HEp-2 cells. After4 h of infection, the anti-EspC antibodies did not detecta clear EspC band in the cytoplasmic fraction ofEPECDescN-treated cells (Fig. 1B), despite that thismutant is able to secrete EspC when is grown inMinimum Essential Medium (MEM) culture medium(Fig. 1C, top) or during cell infection with EPECDescN(Fig. 1C, bottom). These data suggest that some of theEPEC factors encoded in LEE are required for the effi-cient internalization of EspC.

As EspC is an autotransporter protein, it must besecreted into the bacterial extracellular media; therefore,we used an exogenous EspC internalization model,which involved addition of purified EspC during epithelialcells infection by EPEC. To validate our model, HEp-2cells were incubated with various concentrations of puri-fied EspC for 4 h, and then cytoplasmic fractions wereobtained and probed by immunoblot using anti-EspCantibodies. Only a scarce EspC band was detected incells treated with 200 or 300 mg ml-1 EspC, whereasEspC was not detected in cells treated with 50 or

Fig. 1. LEE pathogenicity island is involved in efficient EspCtranslocation into epithelial cells. HEp-2 cells were infected withdifferent AE pathogens and the cytoplasmic fractions from theseinfected cells were analysed by immunoblot using anti-EspCantibodies.A. Infection with AE pathogens. HEp-2 cells were infected with thewild-type EPEC, EHEC or REPEC or with either of two last strainstransformed with espC (EHEC/espC or REPEC/espC).B. Infection with isogenic mutants. HEp-2 cells were infected withEPECDeae and EPECDescN for 4 h.C. Secretion of EspC by EPEC, EPECDeae or EPECDescN grownin MEM medium (top) or during infection of epithelial cells (bottom).Nitrocellulose membranes were also probed with anti-actinantibodies as protein loading control.

1976 J. E. Vidal and F. Navarro-García

© 2008 The AuthorsJournal compilation © 2008 Blackwell Publishing Ltd, Cellular Microbiology, 10, 1975–1986

100 mg ml-1 EspC (Fig. S1A), indicating no contamina-tion of the cytoplasmic fraction by EspC. On the otherhand, to avoid purified EspC internalization by pinocyto-sis (Vidal and Navarro-Garcia, 2006), HEp-2 cells wereincubated with 100 mg ml-1 EspC for 6 and 8 h and thenthe cytoplasmic fraction was probed with the anti-EspCantibodies. No clear EspC band was observed after 6 h,while after 8 h a burly EspC band was detected by theanti-EspC antibodies (Fig. S1B). These results allow usto establish that treatment with 100 mg ml-1 EspC for 4 havoid cellular fraction contamination and internalizationby pinocytosis. Additionally, there was not detection ofother EPEC proteins, which are not translocated into theeukaryotic cell, such as EspA that was detected in thesupernatant, but not in the cytoplasmic fraction, ofEPEC-infected cells (Fig. S2B). By using this experimen-tal model, an stronger band of EspC was detected in thecytoplasmic fraction from HEp-2 cells co-incubated withexogenous EspC and wild-type EPEC than in cellstreated with EPEC but without exogenous EspC(Fig. 2A). These data indicate that EPEC is able to inter-nalize exogenous and endogenous EspC at the same

time until saturation; compare data of 6 h versus 8 h(Fig. 2A). To estimate how much EspC is being translo-cated inside the cells, we performed the following con-sideration: after 6 h of infection the amount of EspC bydensitometry is similar to that of actin (actin is the mostabundant intracellular protein in an eukaryotic cell), sug-gesting that EspC can reach an intracellular concentra-tion of about 0.2 mM; as a typical cytosolic concentrationof actin in non-muscle cells is 0.5 mM, but solubleG-actin concentration in the cytosol is about 0.2 mM. Inaddition, cytoplasmic EspC band is of about 260 ng from15 mg of protein from the cytoplasmic fraction run inthe sodium dodecyl sulfate-polyacrilamyde gels (SDS-PAGE), as compared with 300 ng of purified EspCprotein used as control (Fig. S2A). Interestingly, anEPEC isogenic mutant in espC (EPECDespC) wascapable of internalizing EspC into the cytoplasm ofHEp-2 cells if those cells were co-incubated with exog-enous EspC (Fig. 2A). The same results were obtainedwhen HEp-2 cells were co-incubated with EHEC andexogenous EspC (Fig. 2B), but not in cells co-incubatedwith EPECDescN and exogenous EspC (Fig. 2C). Allthese data indicate that EspC, which is secreted byT5SS, is then translocated to the cytosol by the action ofLEE-encoded factor(s).

EspC translocation is a specific event and independentof LEE-encoded effector proteins

To determine the specificity of EspC translocation byEPEC, HEp-2 cells were co-incubated with EPECDespCand exogenous Pic, another autotransporter protein(homology of 34.6%) secreted by EAEC, or GST-fodrin,an irrelevant protein, both of which have a similarmolecular weight (109 kDa) to EspC (110 kDa). Unlikethe cytoplasmic fraction from cells co-incubated withEPECDespC and exogenous EspC (Fig. S3B), neitherPic (Fig. S3A) nor GST-fodrin (Fig. S3B) was translo-cated to the cytoplasm by EPECDespC; even though Picand EPECDespC co-incubation was prolonged for 5 or6 h (Fig. S3A). These results strongly suggest that themechanism of translocation used by EPEC is specific forEspC.

As the effector proteins injected by EPEC have diverseeffects on the host cells (Dean et al., 2005), we testedtheir role on EspC translocation. HEp-2 cells wereinfected with EPECDespC for sufficient times for injectingthe effector proteins into the cytosol (1, 2 or 3 h). At theend of these incubation times, the bacteria were killedwith tetracycline. After removing killed bacteria, the cellswere incubated with exogenous EspC for 4 h. EspC wasnot detected in the cytoplasmic fraction from these cells(Fig. S3C), suggesting that the effector proteins may notparticipate in the EspC internalization.

Fig. 2. Bacteria encoding a functional LEE translocate exogenousand endogenous EspC into the cells. HEp-2 cells were infectedwith EPEC, their derivative mutants or EHEC and co-incubated withpurified EspC (100 mg ml-1) for 4, 6 or 8 h. Cytoplasmic fractionswere obtained and processed as in Fig. 1.A. Cells treated with EPEC or EPECDespC alone for 4 h orco-incubated with exogenous EspC for various times.B. Cells treated with EHEC alone for 4 h or co-incubated withexogenous EspC for various times. EPEC was used as a positivecontrol.C. Cells treated with EPECDescN alone for 4 h or co-incubatedwith exogenous EspC for various times. After 8 h of incubation withEPECDescN, exogenous EspC is internalized by pinocytosis.Nitrocellulose membranes were also probed with anti-actinantibodies as protein loading control.

An autotransporter is translocated by the T3SS 1977


EspC translocation into epithelial cells is independentof intimate adherence

We have shown that EspC translocation occurs over thepedestals formation (Vidal and Navarro-Garcia, 2006),suggesting a role for the intimate adherence. To search forthis possibility, HEp-2 cells were infected with EPECisogenic mutants unable to produce intimin or Tir(EPECDeae or EPECDtir). Neither of these mutants wasable to translocate EspC into the cells (Figs 1B and 3A),although both mutants secreted EspC to the extracellularmedium, and during epithelial cell infection (Figs 1C and3B). However, by using the HEp-2 cells model ofco-incubation between either EPECDeae or EPECDtir andexogenous EspC, we found that both mutants were ableto translocate EspC into the cytoplasmic fraction (Fig. 3Aand C).

Improving the EspC translocation by EPECDeae andEPECDtir for adding exogenous EspC suggests a dimin-ished efficiency undetectable by cellular fractionation dueto the reduced adherence of these mutants. To explorethis possibility, HEp-2 cells were infected with wild-type

EPEC, EPECDeae or EPECDescN for 3 h and then thecells were immunostained with anti-EspC antibodies(green) and the actin cytoskeleton with rhodamine-phalloidin (red). We searched for cells with adherent bac-teria by Normarsky microscopy, and in these cells, EspCwas detected in the cytoplasm of cells infected with EPEC(Fig. 4G–I) or EPECDeae (Fig. 4D–F), as observed in amiddle section by confocal microscopy. However, EspCwas not detected into the cells infected with EPECDescN(Fig. 4B), despite these bacteria being adherent to theepithelial cells (Fig. 4C) and EspC being secreted bythese bacteria (Fig. 4A).

To further demonstrate that the inefficiency ofEPECDeae and EPECDtir for EspC translocation wasdue to the diminished adherence of these mutants toepithelial cells, the multiplicities of infection (moi) ofEPECDtir were increased until a bacterial adhesionsimilar to wild-type EPEC was reached. As previouslyestablished, an moi of 10 was used for the wild type,while for the EPECDtir moi of 50 and 200 were used.The cells were stained with rhodamine-phalloidin (red)and immunostained with anti-EspC (green) and anti-EPEC outer membrane (blue). As shown in Fig. 5, with20 times more EPECDtir bacteria (moi = 200) it was pos-sible to reach an amount of adherent bacteria compa-rable to the wild type and in that condition also a similaramount of EspC was translocated to the cytoplasm(Fig. 5A–D versus Fig. 5I–L). However, with five timesmore EPECDtir, cells did not show as many attachedbacteria as the wild-type EPEC and only trace amountsof EspC were detected by confocal microscopy in middlesections (Fig. 5E–H). Only in EPEC-infected cells, thethree markers colocalized because EPEC displaysintimate adherence and forms pedestals. Thus, eventhough EspC translocation is favoured by intimateadherence, it is independent of this infection step.

T3SS translocon of EPEC allows EspC translocationinto epithelial cells

As EPECDescN, lacking the bacterial cytoplasmic trans-locator, is unable to translocate EspC, we then used othertwo mutants in the T3SS, lacking extracellular transloca-tors, EPECDescF and EPECDespA, as well as a mutant ina translocator involved in pore formation on the hostmembrane, EPECDespB. First of all, we verified thatthese mutants secreted EspC during epithelial cell infec-tion at similar levels to the wild-type EPEC (Figs 3B and6B), confirming that EspC secretion is not mediated byT3SS.

HEp-2 cells were infected with EPECDescF,EPECDespA, EPECDespB or with any of the two mutantsthat are able to translocate exogenous EspC,EPECDespC and EPECDtir. Infected cells were

Fig. 3. EspC translocation is independent of the intimateattachment. HEp-2 cells were treated and processed as in Fig. 2.A. Cytoplasmic fractions from cells treated with EPECDtir alone for4 h or co-incubated with exogenous EspC for various times. EPECwas used as positive control.B. Secretion of EspC by EPEC, EPECDescF, EPECDespA orEPECDtir during infection of epithelial cells.C. Cytoplasmic fractions from cells treated with EPECDeae alonefor 4 h or co-incubated with exogenous EspC for various times.Nitrocellulose membranes were also probed with anti-actinantibodies as protein loading control.



co-incubated with exogenous EspC for 6 h. As shownabove, EPECDespC and EPECDtir were able to translo-cate exogenous EspC into the cytoplasm (Fig. 6A).However, EspC was not detected in the cytoplasmic frac-

tions of cells infected with EPECDescF, EPECDespA orEPECDespB (Fig. 6A and B). These data indicate thatEPEC uses the T3SS translocon for EspC translocationinto the epithelial cells. To corroborate these results,

Fig. 4. Reduced adherence to the cells byintimate attachment in the mutants decreasesthe EspC translocation. HEp-2 cells wereinfected with EPEC, EPECDescN orEPECDeae. Treated cells wereimmunostained with anti-EspC antibodies andthe actin cytoskeleton was detected withrhodamine-phalloidin.A–C. Cells treated with EPECDescN.D–F. Cells treated with EPECDeae.G–I. Cells treated with EPEC.Images from optical sections by confocalmicroscopy are indicated: surface sections(A, D and G), middle sections (B, E and H).Adherent bacteria were localized byNormarsky microscopy (C, F and I). Whiteand blue arrows show EspC secreted overthe cell surface or inside the cellsrespectively. Black arrows show microcoloniesof adhered cells.

Fig. 5. Low EspC translocation by intimateadherence mutants is related to theirdefective initial adhesion. HEp-2 cells wereinfected with EPEC or EPECDtir for 3 h.A–D. Cells infected with EPEC with an moi of10.E–H. Cells infected with EPECDtir with an moiof 50.I–L. Cells infected with EPECDtir with an moiof 200.Treated cells were immunostained withanti-EspC antibodies (green; C, G and K;middle sections) and anti-EPEC membranes(blue; B, F and J; surface sections), andthe actin cytoskeleton was detected withrhodamine-phalloidin (red; A, E and I;projection of optical sections). (D), (H) and (L)are the merge of the three channels. Whitearrows show pedestal formation; yellowarrows show microcolonies of adherentbacteria.



HEp-2 cells were infected with a high moi (200) of T3SSmutants (EPECDescN or EPECDespA). In this system,both mutants were capable for adhering to epithelial cellsat similar levels to the wild-type EPEC used at an moi = 10(Fig. 6D, H and L). However, only the wild-type EPEC wascapable of translocating EspC into the cells (Fig. 6Eversus Fig. 6I and M). Furthermore, complementation ofT3SS mutants re-established EspC translocation. Thus,complementation of EPECDescF with a plasmid contain-ing escF gene allowed EspC detection in the cytoplasmfraction, but it was not detected in the cytoplasm fractionfrom the isogenic mutant in escF gene (Fig. S2C). Allthese data clearly indicate that EspC secreted by T5SS istranslocated into epithelial cells by using the T3SS trans-locon of EPEC.

EspC interacts with EspA

To understand how EspC could be interacting with theT3SS, we first infected HEp-2 cells with EPECDespCco-incubated with exogenous EspC. After the co-incubation, the cells were extensively washed, fixed,

but not permeabilized, and the bacterial membraneand EspC were detected by immunofluorescence byusing anti-EPEC membrane (red) and anti-EspC (green)antibodies. By optical sections under confocal microscopy(Fig. 7), it was possible to detect EspC (green) betweenthe eukaryotic plasma membrane (black background)(Fig. 7A) and the bacterial membrane (red) (Fig. 7B) andto show partial localization in between them, but predomi-nantly the red on the green fluorescence by merging bothchannels (Fig. 7C). To determine whether EspC is inter-acting with T3SS proteins on the bacterial membrane, weperformed immunoprecipitation assays from the mem-brane fraction of EPEC incubated with EspC. After 1 h ofco-incubation, EPEC membrane mixed with EspC wassubjected to immunoprecipitation with anti-EspC. ProteinA-bound immunocomplexes were separated by SDS-PAGE. Besides EspC, five main proteins were resolved bythe SDS-PAGE, which were of 80, 75, 37, 35 and 26 kDa(Fig. 7D). Interestingly, when membrane from the espAmutant was used in the immunoprecipitation experimentsthese proteins were not seen (Fig. 7D, last lane). The lastthree proteins co-immunoprecipitated with the anti-EspC

Fig. 6. EPEC uses T3SS translocon to allowthe EspC translocation.A and B. EPECDescF, EPECDespA andEPECDespB are unable to translocateexogenous EspC. HEp-2 cells were infectedwith EPECDespC, EPECDescF, EPECDespA,EPECDtir or EPECDespB and co-incubatedwith purified EspC (100 mg ml-1) for 6 h.Cytoplasmic fractions were obtained andprocessed as in Fig. 1.C–N. No EspC translocation occurred in cellsinfected with higher moi of EPECDescN orEPECDespA. HEp-2 cells were infectedwith EPEC with an moi of 10 (C–F) orEPECDescN (G–J) or EPECDespA (K–N) withan moi of 200, during 3 h. Treated cells wereimmunostained with anti-EspC antibodies(green; C, G and K; middle sections) andanti-EPEC membranes (blue; B, F and J;surface sections), and the actin cytoskeletonwas detected with rhodamine-phalloidin(red; A, E and I; projection of confocal opticalsections). (D), (H) and (L) are the merge ofthe three channels. White arrows pointpedestal formation; yellow arrows showmicrocolonies of adhered bacteria.EPECDespB-sec in (B) indicates secretedproteins during cells infection.



antibody from EPEC membranes suggesting that theymay be EspD, EspB and EspA. To test this, we performedoverlay experiments by separating purified EspC, EspB orEspA, as well as EPEC lysates with SDS-PAGE (Fig. 7E),

and probing nitrocellulose membranes containing theseseparated proteins with purified EspB and EspA, and theinteraction among these proteins were visualized by usinganti-EspB (Fig. 7F) and anti-EspA (Fig. 7G) antibodies.

Fig. 7. EspC interacts with the translocator protein EspA on the bacterial membrane.A–C. Detection of EspC in between eukaryotic plasma membrane and bacterial membrane. HEp-2 cells were infected with EPECDespCco-incubated with exogenous EspC for 1 h. After infection, cells were extensively washed, fixed, but not permeabilized, and immunostained byusing Anti-EPEC membrane (A) and anti-EspC (B). The localization of exogenous EspC outside of the cells and its interaction with thebacterial outer membrane was analysed for merging images (C) by confocal microscopy.D. Immunoprecipitation of EspC-interacting proteins on bacterial outer membrane. Purified outer membrane from wild-type EPEC orEPECDespA (*) were incubated with purified EspC for 1 h. The mixture was subjected to immunoprecipitation with anti-EspC. As controls inthe SDS-PAGE was run immunoprecipitated EspC alone, anti-EspC and protein A incubation, the protein removed from the mixture bywashing, as well as purified EspC and purified EPEC membrane fraction.E–G. Overlay assay among EspC, EspA and EspB. Purified EspA, EspB and EspC, as well as EPEC lysates were subjected to SDS-PAGEanalysis (E). These separated proteins were transferred to nitrocellulose membrane and incubated with either purified EspB or purified EspA.The EspB or EspA interactions were developed by immunoblot by using anti-EspB (F) or anti-EspA (G) antibodies.H–K. EspC translocation appears to occur through channel-like structures underneath the adhered bacteria. HEp-2 cells were infected withEPEC (H), EPECDespC (I) or EPECDespC plus purified EspC (J and K) for 3 h. Treated cells were immunostained with anti-EspC antibodies(green) and anti-EPEC membranes (blue), and the actin cytoskeleton was detected with rhodamine-phalloidin (red). Images show confocalvertical sections (yz) of infected cells. Arrowheads in (H) and (J) show EspC fluorescent marks forming channel-like structure. In (K), yellow,white and green arrows show bacteria, actin pedestal and EspC.



The anti-EspB antibody detected mainly the purified EspBand a 26 kDa band in EPEC lysates at similar molecularweight to purified EspA; this purified protein slightly inter-acted with purified EspB, as detected with the anti-EspBantibody (Fig. 7F). The anti-EspA antibody detectedmainly the purified EspA, and the EspA interacting withpurified EspB and EspC, as well as with proteins of 39 and110 kDa in EPEC lysates, which might be EspD andEspC. In fact this latter protein was at the same molecularweight as purified EspC, which strongly interacted withEspA (Fig. 7G).

To further investigate the mechanism of EspC translo-cation, HEp-2 cells were infected with EPEC or with eitherEPECDespC alone or co-incubated with exogenousEspC. Confocal microscopy analysis of cells stained withrhodamine-phalloidin and immunostained with anti-EspCshowed that EspC is observed inside the cells infectedwith either EPEC (Fig. 7H) or EPECDespC co-incubatedwith exogenous EspC (Fig. 7J), but not in those cellsinfected with EPECDespC alone (Fig. 7I). Interestingly,confocal sections in the plane yz of these cells showedEspC fluorescent marks coming from the extracellularregion (Fig. 7H and J, arrows): these fluorescent markswere channel-like passing the cell membrane andreached the cytoplasm, suggesting that they come fromadhering bacteria. Indeed, in similar experiments but nowmarking also the bacteria by using anti-EPEC membraneantibodies, EspC (green) seemed to emerge from thebacteria (blue) and was detected forming these channel-like structures just beneath the bacteria and the pedestals(red) (Fig. 7K). Remarkably, even though a high concen-tration of exogenous EspC is present, the EspC translo-cation appears to occur only through these channel-likestructures under the adhered bacteria. Similar resultswere observed in cells infected only with the wild-typeEPEC, suggesting the T3SS translocon is involved.

Discussion

We have demonstrated that EspC is efficiently translo-cated to the cytoplasm of cells in culture infected withEPEC (Vidal and Navarro-Garcia, 2006). This transloca-tion event is related to the LEE PAI as at least two otherAE pathogens allowed EspC translocation. EspC is rec-ognized in a specific manner for these pathogens to betranslocated into the epithelial cells, as irrelevant orhomologous proteins are not translocated by thesepathogens. Furthermore, EspC translocation is not anevent induced by effector proteins injected by EPEC intothe target cell. Intimate adherence between EPEC andepithelial cells favours the EspC translocation but is notessential for such event. Finally, EspC is translocated bythe T3SS translocon of EPEC, but unlike the LEE-encoded proteins or the recently reported, non-LEE-

encoded proteins that use this T3SS translocon to beinjected from the bacterial cytoplasm to the eukaryoticcytoplasm, EspC is secreted to the milieu and then isincorporated to the T3SS translocon to enter into the cellsthrough its interaction with EspA.

Recently, we found that EspC is efficiently internalizedduring the pedestal formation (Vidal and Navarro-Garcia,2006), suggesting that EspC translocation might berelated to this lesion and thereby, to the LEE PAI. By usingother two AE pathogens, which harbour LEE PAI but notespC gene, we found that these espC-transformed patho-gens displayed an efficient EspC translocation. Moreover,an isogenic mutant (EPECDescN), lacking a cytoplasmicATPase translocator protein that provides energy tosecrete proteins through the T3SS encoded in LEE(Gauthier et al., 2003), was unable to translocate EspCinto the cells. These data also suggest that any factorsecreted by the T3SS might be helping EspC transloca-tion, as EspC is an autotransporter protein secreted bythe T5SS (Stein et al., 1996; Mellies et al., 2001; Navarro-Garcia et al., 2004). To explore this controversy, we con-ceived an EspC translocation model by infecting cells withbacteria harbouring LEE and co-incubated with exog-enous EspC. This model and the use of an isogenicmutant in espC (EPECDespC) allowed us to establish thatEPEC is able to translocate exogenous and endogenousEspC, indicating that EspC is translocated after itssecretion.

This EspC translocation event assisted by LEE-encoded factor(s) was specific for EspC, as other proteinsincluding another serine protease autotransporter fromEAEC, called Pic, that is not internalized into the cellsas a purified protein (our unpublished data) were nottranslocated. This suggests that EspC translocationdoes not occur by increased membrane permeability orpinocytic event as a result of EPEC infection but insteadinvolves a somewhat precise mechanism. The LEE con-tains 41 genes and encodes a T3SS, a regulator (Ler), anadhesin (intimin) and its receptor (Tir) responsible forintimate attachment, several secreted proteins and theirchaperones (Frankel et al., 1998; Clarke et al., 2003).Interestingly, EspC is also regulated by Ler, despite espCbeing encoded by another PAI (Elliott et al., 2000; Mellieset al., 2001). Chaperones might not be involved in EspCtranslocation, as they are bacterial cytoplasmic proteinsand EspC is secreted by the T5SS. Thereby, secretedproteins and/or intimate attachment or the complete T3SStranslocon may be involved in EspC translocation. Thesecreted proteins consist of effectors (Tir, EspG, EspF,Map and EspH) as well as translocators (EspA, EspD andEspB). Five LEE-encoded effectors have been identified,which are involved in modulating host cytoskeleton,including paracellular permeability (Clarke et al., 2003; Tuet al., 2003; Matsuzawa et al., 2005). However, the effec-



tors appear not to participate in the EspC internalizationas indicated in experiments injecting these effectors byusing EPECDespC, and exogenous EspC after killing thebacteria. Additionally, even though EspC translocationoccurs beneath adhered bacteria and over the pedestals(Vidal and Navarro-Garcia, 2006), the role of the intimateattachment in the EspC translocation was also discarded,as isogenic mutants in both proteins responsible for inti-mate attachment, intimin and Tir, were capable for EspCtranslocation into epithelial cells. However, it was neces-sary to overcome the reduced adherence observed inthese mutants (Shaw et al., 2005) by co-incubating themwith exogenous EspC or by increasing the moi at least 20times. Interestingly, by using these latter conditionsEPECDescN was still unable to translocate EspC, despitethose bacteria adhering to the epithelial cells and secret-ing EspC.

These last results lead us to suggest that the translo-cator proteins may participate in the EspC translocation,although these proteins (working as a T3SS translocon)only appear to translocate proteins from the bacterialcytoplasm to the eukaryotic cytosol. However, EspCtranslocation occurs when it has already been secretedto the intracellular milieu. Other findings also supportLEE–EspC association, including the fact that theLEE-encoded regulator (Ler) regulates the secretion oftranslocators as well as of EspC (Elliott et al., 2000). EspCis temporally the first protein secreted during the epithelialcell infection by EPEC (Kenny and Finlay, 1995), thus itmust be available for its translocation by the time thetranslocators are secreted. Remarkably, partially purifiedT3SS translocons, observed as injectosomes by electronmicroscopy, contain EspC as analysed by SDS-PAGE(Sekiya et al., 2001), even though this protein is secretedbefore the T3SS formation (Kenny and Finlay, 1995).Interestingly, EspC is separated from the injectosomesonly after (ultracentrifugation, sucrose cushion) a thirdpurification step on CsCl fractionation (Ogino et al., 2006).Supporting these isolated findings, we show here that thelack of the injectosome by using EPECDescN to infectcells prevented EspC translocation; also the lack of theneedle of the injectosome or of the filamentous extensionto the needle complex or translocators involved in poreformation on the membranes of the infected cells, byusing EPECDescF, EPECDespA or EPECDespB, respec-tively, also stopped EspC translocation. These data indi-cate that the T3SS translocon is required for allowingEspC translocation into epithelial cells, but also raise thequestion of how a milieu-secreted protein gains access tothe T3SS translocon to be injected into the host cells; isthe translocon a closed conduit? Furthermore, affinityexperiments using purified proteins showed that EspCclearly binds EspA, but not EspB (although EspC–EspDinteraction experiments were not performed). Additionally,

immunoprecipitation experiments from EPEC mem-branes incubated with EspC showed that EspCco-immunoprecipitates with EspA, and thereby with EspBand EspD (translocator proteins), as EspC does not inter-act with these proteins in membranes from EPECDespA.All these data suggest the T3SS might function asconduit/guiding filament by guiding proteins previouslysecreted from the bacteria but with affinity to the filament(EspA), as it was initially suggested for Pseudomonassyringae (Jin et al., 2001).

In summary, we demonstrated here that during EPECinfection, EspC is secreted by the T5SS and then it isefficiently translocated through the T3SS. This constitutestwo novel mechanisms of protein translocation into hostcells, which involves cooperation between two secretionsystems, T5SS and T3SS, as well as the uptake of aprotein from the culture/supernatant into the T3SS. Thiscooperation mechanism may be employed by otherpathogens encoding T3SS and autotransporter proteins,for instance, the T3SS and EspP of EHEC.

Experimental procedures

Bacterial strains and purification of recombinant protein

Characteristics of the strains used in this study are listed inTable S1. All strains were routinely grown in Luria–Bertani(LB) broth or MEM (without supplements) aerobically at 37°C.EPEC cultures were activated for 3 h as previously described(Rosenshine et al., 1996).

Strains HB101(pJLM174) or HB101(pPic1) were grown over-night in LB plus arabinose (0.2% w/v) and ampicillin (100 mg ml-1)or tetracycline (15 mg ml-1), respectively, at 37°C in shaking. Thesupernatants were obtained by centrifugation at 7000 g for15 min, filter sterilized through 0.22-mm-diameter filters (Corning,Cambridge, MA) and concentrated 100-fold in an ultrafree cen-trifugal filter device with a 100 kDa cut-off (Millipore, Bedford,MA). Recombinant proteins were filter sterilized again, aliquotedand quantified by the Bradford method (Bradford, 1976).

Tissue culture cells and cellular fractionation

The human epithelial cell line HEp-2 (ATCC-CCL23) was culturedin MEM supplemented with 10% fetal calf serum (FCS) (HyClone,Logan, UT), 1% non-essential amino acids, 5 mM L-glutamine,penicillin (100 U ml-1) and streptomycin (100 mg ml-1). Cells werenormally harvested with 10 mM EDTA and 0.25% trypsin (GibcoBRL, Grand Island, NY) in PBS (pH 7.4), re-suspended in theappropriated volume of supplemented DMEM and incubated at37°C in a humidified atmosphere of 5% CO2.

HEp-2 cells grown in 60 mm Petri dishes were infected withactivated cultures of EPEC or the indicated isogenic mutantalone or with either purified EspC or the indicated purified proteinduring the indicated time. EPEC-infected HEp-2 cells wereincubated in the presence of D-mannose (1%) and the appro-priated antibiotic. Cells were delicately washed three times withice-cold PBS, pH 7.4 and scraped in a buffer consisting of



Tris-HCl (0.25 M), pH 7.5, phenylmethylsulfonyl fluoride (PMSF)(50 mg ml-1), aprotinin (0.5 mg ml-1) and EDTA (0.5 mM). Thencells were lysed by three repeating freeze–thaw cycles (5 minincubation in a dry ice-ethanol bath and 3 min incubation in athermoblock at 37°C) (Taylor et al., 1998). The cell lysates wereultracentrifuged at 100 000 g for 1 h at 4°C, and the supernatantfraction containing soluble cytoplasmic proteins was obtained.Pellets containing HEp-2 cell membranes, adherent bacteria,nuclei and cytoskeletal proteins were washed with cold PBS andre-suspended in PBS. Protein concentrations were estimated bythe method of Bradford. Equivalent volumes were boiled for7 min, analysed on SDS-PAGE and electrotransferred to nitro-cellulose membranes for Western blot analyses essentially aspreviously described (Vidal and Navarro-Garcia, 2006). Theidentity of the cellular fractions was confirmed with a mousemonoclonal anti-actin antibody (a gift from Manuel Hernández,CINVESTAV, Mexico) and a polyclonal rabbit anti-pan-cadherin(Zymed laboratories).

Confocal microscopy

HEp-2 cells were seeded into eight-well LabTek slides (VWR,Bridgeport, NJ) at a density of 3 ¥ 104 cells per well. Beforeinfection with activated EPEC cultures, cells were washed threetimes with warmed PBS (pH 7.4) and incubated at 37°C in DMEM(without supplements) during 30 min. Infections were carried outat the indicated time in the presence of D-mannose (1%)(Research Organics). Infected HEp-2 cells were washed withPBS, fixed with 2% formalin–PBS, washed, permeabilized by theaddition of 0.1% Triton X-100–PBS, and stained with 0.05 mg oftetramethyl rhodamine isothiocyanate (TRITC)-phalloidin ml-1

and with a polyclonal rabbit anti-EspC as previously described(Navarro-Garcia et al., 2004), followed by an anti-rabbitfluorescein-labelled antibody. When indicated, the bacteria weredetected by using an antibody against membrane proteins fromEPEC (a gift from Angel Manjarrez, UNAM, Mexico). Slides weremounted on Gelvatol, covered with glass coverslips and exam-ined under a Leica TCS SP2 confocal microscope.

Immunoprecipitation and overlay assays

Immunoprecipitation assays were performed as a previous report(Navarro-Garcia et al., 2007). Briefly, outer membrane from wild-type EPEC or EPECDespA, obtained by ultrasonic treatment andextraction in 1% sarcosyl according to the methodology previ-ously described by Klingman and Murphy (1994), was incubatedwith purified EspC for 1 h at 37°C. This mixture wasre-suspended in 1 ml of cold lysis buffer and 500 mg of proteinwas centrifuged, and the supernatants were used for immuno-precipitation experiments. Supernatants were incubated withanti-EspC (2 mg) antibodies. Then, protein A-agarose suspension(Roche Diagnostics, Mannheim, Germany) was added for 3 h at4°C. The complexes were collected by centrifugation and thesupernatant was removed. The pellet was washed five times andre-suspended in loading buffer, and the immunocomplexes wereresolved by SDS-PAGE.

Overlay assays were performed as a previous report(Canizalez-Roman and Navarro-Garcia, 2003). Four or twomicrograms of the purified proteins (EspA, EspB or EspC) orEPEC lysates were separated by 10% SDS-PAGE. These sepa-

rated proteins were then transferred to nitrocellulose membrane,which was blocked overnight in blocking buffer at 4°C. The mem-branes were then incubated for 1 h in interacting buffer witheither 5 mg ml-1 EspA or 5 mg ml-1 EspB (kindly donated by DrAngel Cataldi, Institute of Biotechnology, Argentina). The mem-branes were washed and then incubated for 1 h in blocking bufferwith rabbit polyclonal antiserum raised against EspA (1:3000dilution) or rabbit polyclonal antiserum raised against EspB(1:3000 dilution). These antibodies were kindly donated by JorgeGirón (The University of Arizona). Following another wash step,the strips were incubated for 1 h in blocking buffer with a goatanti-rabbit AP-conjugated antibody (1:10 000 dilution) (KPL,Gaithersburg, MD). Following a final wash, binding was detectedusing one-step NBT/BCIP substrate (Pierce, Rockford, IL).

Acknowledgements

We thank Dr James Kaper and Bruce McClane for reviewing themanuscript and their valuable advices. We thank James Kaper,Michael Donnenberg, James Nataro, Brett Finlay, Gad Frankel,Eric Oswald and Akio Abe for providing the E2348/69 derivativesand AE pathogens. We also thank Rocio Huerta and JazminHuerta for their technical help.

References

Bradford, M.M. (1976) A rapid and sensitive method for thequantitation of microgram quantities of protein utilizing theprinciple of protein-dye binding. Anal Biochem 72: 248–254.

Campellone, K.G., Robbins, D., and Leong, J.M. (2004)EspFU is a translocated EHEC effector that interacts withTir and N-WASP and promotes Nck-independent actinassembly. Dev Cell 7: 217–228.

Canizalez-Roman, A., and Navarro-Garcia, F. (2003) FodrinCaM-binding domain cleavage by Pet from enteroaggrega-tive Escherichia coli leads to actin cytoskeletal disruption.Mol Microbiol 48: 947–958.

Clarke, S.C., Haigh, R.D., Freestone, P.P., and Williams,P.H. (2003) Virulence of enteropathogenic Escherichia coli,a global pathogen. Clin Microbiol Rev 16: 365–378.

Dahan, S., Wiles, S., La Ragione, R.M., Best, A., Woodward,M.J., Stevens, M.P., et al. (2005) EspJ is a prophage-carried type III effector protein of attaching and effacingpathogens that modulates infection dynamics. InfectImmun 73: 679–686.

Dean, P., Maresca, M., and Kenny, B. (2005) EPEC’sweapons of mass subversion. Curr Opin Microbiol 8:28–34.

Elliott, S.J., Wainwright, L.A., McDaniel, T.K., Jarvis, K.G.,Deng, Y.K., Lai, L.C., et al. (1998) The complete sequenceof the locus of enterocyte effacement (LEE) from entero-pathogenic Escherichia coli E2348/69. Mol Microbiol 28:1–4.

Elliott, S.J., Sperandio, V., Giron, J.A., Shin, S., Mellies, J.L.,Wainwright, L., et al. (2000) The locus of enterocyte efface-ment (LEE)-encoded regulator controls expression of bothLEE- and non-LEE-encoded virulence factors in entero-pathogenic and enterohemorrhagic Escherichia coli. InfectImmun 68: 6115–6126.



Elliott, S.J., Krejany, E.O., Mellies, J.L., Robins-Browne,R.M., Sasakawa, C., and Kaper, J.B. (2001) EspG, a noveltype III system-secreted protein from enteropathogenicEscherichia coli with similarities to VirA of Shigella flexneri.Infect Immun 69: 4027–4033.

Frankel, G., Phillips, A.D., Rosenshine, I., Dougan, G.,Kaper, J.B., and Knutton, S. (1998) Enteropathogenic andenterohaemorrhagic Escherichia coli: more subversiveelements. Mol Microbiol 30: 911–921.

Garmendia, J., Phillips, A.D., Carlier, M.F., Chong, Y.,Schuller, S., Marches, O., et al. (2004) TccP is an entero-haemorrhagic Escherichia coli O157:H7 type III effectorprotein that couples Tir to the actin-cytoskeleton. CellMicrobiol 6: 1167–1183.

Gauthier, A., Puente, J.L., and Finlay, B.B. (2003) Secretin ofthe enteropathogenic Escherichia coli type III secretionsystem requires components of the type III apparatus forassembly and localization. Infect Immun 71: 3310–3319.

Gruenheid, S., Sekirov, I., Thomas, N.A., Deng, W.,O’Donnell, P., Goode, D., et al. (2004) Identification andcharacterization of NleA, a non-LEE-encoded type III trans-located virulence factor of enterohaemorrhagic Escherichiacoli O157:H7. Mol Microbiol 51: 1233–1249.

Henderson, I.R., Navarro-García, F., Desvaux, M., Fernán-dez, R.C., and Ala’Aldeen, D. (2004) Type V protein secre-tion pathway: the autotransporter story. Microbiol Mol BiolRev 68: 692–744.

Jin, Q., Hu, W., Brown, I., McGhee, G., Hart, P., Jones, A.L.,and He, S.Y. (2001) Visualization of secreted Hrp andAvr proteins along the Hrp pilus during type III secretionin Erwinia amylovora and Pseudomonas syringae. MolMicrobiol 40: 1129–1139.

Kaper, J.B., Nataro, J.P., and Mobley, H.L. (2004) Patho-genic Escherichia coli. Nat Rev Microbiol 2: 123–140.

Kelly, M., Hart, E., Mundy, R., Marches, O., Wiles, S., Badea,L., et al. (2006) Essential role of the type III secretionsystem effector NleB in colonization of mice by Citrobacterrodentium. Infect Immun 74: 2328–2337.

Kenny, B., and Finlay, B.B. (1995) Protein secretion byenteropathogenic Escherichia coli is essential for transduc-ing signals to epithelial cells. Proc Natl Acad Sci USA 92:7991–7995.

Kenny, B., DeVinney, R., Stein, M., Reinscheid, D.J., Frey,E.A., and Finlay, B.B. (1997) Enteropathogenic E. coli(EPEC) transfers its receptor for intimate adherence intomammalian cells. Cell 91: 511–520.

Klingman, K.L., and Murphy, T.F. (1994) Purification andcharacterization of a high-molecular-weight outer mem-brane protein of Moraxella (Branhamella) catarrhalis. InfectImmun 62: 1150–1155.

McDaniel, T.K., Jarvis, K.G., Donnenberg, M.S., and Kaper,J.B. (1995) A genetic locus of enterocyte effacementconserved among diverse enterobacterial pathogens. ProcNatl Acad Sci USA 92: 1664–1668.

McNamara, B.P., Koutsouris, A., O’Connell, C.B., Nou-gayrede, J.P., Donnenberg, M.S., and Hecht, G. (2001)Translocated EspF protein from enteropathogenic Escheri-chia coli disrupts host intestinal barrier function. J ClinInvest 107: 621–629.

Marches, O., Ledger, T.N., Boury, M., Ohara, M., Tu, X.,Goffaux, F., et al. (2003) Enteropathogenic and entero-

haemorrhagic Escherichia coli deliver a novel effectorcalled Cif, which blocks cell cycle G2/M transition. MolMicrobiol 50: 1553–1567.

Marches, O., Wiles, S., Dziva, F., La Ragione, R.M., Schuller,S., Best, A., et al. (2005) Characterization of two non-locusof enterocyte effacement-encoded type III-translocatedeffectors, NleC and NleD, in attaching and effacingpathogens. Infect Immun 73: 8411–8417.

Matsuzawa, T., Kuwae, A., and Abe, A. (2005) Enteropatho-genic Escherichia coli type III effectors EspG and EspG2alter epithelial paracellular permeability. Infect Immun 73:6283–6289.

Mellies, J.L., Elliott, S.J., Sperandio, V., Donnenberg, M.S.,and Kaper, J.B. (1999) The Per regulon of enteropatho-genic Escherichia coli: identification of a regulatorycascade and a novel transcriptional activator, the locus ofenterocyte effacement (LEE)-encoded regulator (Ler). MolMicrobiol 33: 296–306.

Mellies, J.L., Navarro-Garcia, F., Okeke, I., Frederickson, J.,Nataro, J.P., and Kaper, J.B. (2001) espC pathogenicityisland of enteropathogenic Escherichia coli encodes anenterotoxin. Infect Immun 69: 315–324.

Moon, H.W., Whipp, S.C., Argenzio, R.A., Levine, M.M., andGiannella, R.A. (1983) Attaching and effacing activities ofrabbit and human enteropathogenic Escherichia coli in pigand rabbit intestines. Infect Immun 41: 1340–1351.

Mundy, R., Petrovska, L., Smollett, K., Simpson, N., Wilson,R.K., Yu, J., et al. (2004) Identification of a novel Citro-bacter rodentium type III secreted protein, EspI, and rolesof this and other secreted proteins in infection. InfectImmun 72: 2288–2302.

Navarro-Garcia, F., Canizalez-Roman, A., Bao Quan, S.,Nataro, J.P., and Azamar, Y. (2004) The serine proteasemotif of EspC from enteropathogenic Escherichia coli pro-duces epithelial damage by a different mechanism than Pettoxin from enteroaggregative E. coli. Infect Immun 72:3609–3621.

Navarro-Garcia, F., Canizalez-Roman, A., Burlingame, K.E.,Teter, K., and Vidal, J.E. (2007) Pet, a non-AB toxin, istransported and translocated into epithelial cells by a ret-rograde trafficking pathway. Infect Immun 75: 2101–2109.

Ogino, T., Ohno, R., Sekiya, K., Kuwae, A., Matsuzawa, T.,Nonaka, T., et al. (2006) Assembly of the type III secretionapparatus of enteropathogenic Escherichia coli. J Bacteriol188: 2801–2811.

Rosenshine, I., Ruschkowski, S., and Finlay, B.B. (1996)Expression of attaching/effacing activity by enteropatho-genic Escherichia coli depends on growth phase, tempera-ture, and protein synthesis upon contact with epithelialcells. Infect Immun 64: 966–973.

Sekiya, K., Ohishi, M., Ogino, T., Tamano, K., Sasakawa, C.,and Abe, A. (2001) Supermolecular structure of the entero-pathogenic Escherichia coli type III secretion system andits direct interaction with the EspA-sheath-like structure.Proc Natl Acad Sci USA 98: 11638–11643.

Shaw, R.K., Cleary, J., Murphy, M.S., Frankel, G., andKnutton, S. (2005) Interaction of enteropathogenic Escheri-chia coli with human intestinal mucosa: role of effectorproteins in brush border remodeling and formation ofattaching and effacing lesions. Infect Immun 73: 1243–1251.



Stein, M., Kenny, B., Stein, M.A., and Finlay, B.B. (1996)Characterization of EspC, a 110-kilodalton protein secretedby enteropathogenic Escherichia coli which is homologousto members of the immunoglobulin A protease-like family ofsecreted proteins. J Bacteriol 178: 6546–6554.

Taylor, K.A., O’Connell, C.B., Luther, P.W., and Donnenberg,M.S. (1998) The EspB protein of enteropathogenicEscherichia coli is targeted to the cytoplasm of infectedHeLa cells. Infect Immun 66: 5501–5507.

Tu, X., Nisan, I., Yona, C., Hanski, E., and Rosenshine, I.(2003) EspH, a new cytoskeleton-modulating effector ofenterohaemorrhagic and enteropathogenic Escherichiacoli. Mol Microbiol 47: 595–606.

Vidal, J.E., and Navarro-Garcia, F. (2006) Efficient transloca-tion of EspC into epithelial cells depends on enteropatho-genic Escherichia coli and host cell contact. Infect Immun74: 2293–2303.

Supplementary material

The following supplementary material is available for this articleonline:Table S1. Bacterial strains and plasmid used in this study.Fig. S1. Standardization of a model of exogenous EspCtranslocation.A. Concentration of purified EspC unable to contaminate thecytoplasmic fraction. HEp-2 cells were incubated with 50, 100,200 or 300 mg ml-1 EspC for 4 h. Cytoplasmic fractions of thesetreated cells were analysed by immunoblot by using anti-EspCantibodies. The first line shows cytoplasmic fraction obtained attime zero (addition of EspC, immediate cell washed and purifica-tion of the cytoplasmic fraction). The last line shows purifiedEspC used as positive reaction and molecular marker.B. Time-course of EspC internalization by pinocytosis detectedby cellular fractionation. HEp-2 cells were incubated with100 mg ml-1 EspC for 6 and 8 h. Untreated cells were used asnegative control.In both panels, nitrocellulose membranes were also probed withanti-actin antibodies as load control.

Fig. S2. Estimation of amount of translocated EspC, restorationof EspC translocation in complemented T3SS mutant and trans-location control of an EPEC protein that is not translocated to thecytoplasm.A. Densitometry measurement of EspC translocation versusintracellular actin in EPEC-infected cells, and purified EspC.B. Detection of EspA in supernatant and cytoplasm fractions fromEPEC-infected cells.C. Restoration of EspC translocation in EPECDescF comple-mented with a plasmid containing the escF gene.Fig. S3. EspC translocation is a specific event and effector pro-teins independent.A and B. Pic (A) and GST-fodrin (B) are not translocate by EPEC.HEp-2 cells were infected with EPECDespC and co-incubatedwith purified Pic (100 mg ml-1) for 4, 5 or 6 h or with GST-fodrin(100 mg ml-1) for 6 h. Cytoplasmic fractions of these treated cellswere analysed by immunoblot by using anti-Pic or anti-GSTantibodies. The first lines show cytoplasmic fraction from cellstreated only with EPECDespC for 4 h. Nitrocellulose membraneof (B) was also probed with anti-EspC antibodies; middle figure.The last lines show purified Pic or GST-fodrin used as positivereactions and molecular markers.C. Role of effectors on EspC translocation. HEp-2 cells wereinfected with EPECDespC for 1, 2 or 3 h and after this infectionthe bacteria were killed with tetracycline and then were incubatedwith purified EspC for 3 h. Cytoplasmic fraction was probed withanti-EspC antibodies.In all panels, nitrocellulose membranes were also probed withanti-actin antibodies as load control.

This material is available as part of the online article from:http://www.blackwell-synergy.com/doi/abs/10.1111/j.1462-5822.2008.01181.x

Please note: Blackwell Publishing is not responsible for thecontent or functionality of any supplementary materials suppliedby the authors. Any queries (other than missing material) shouldbe directed to the corresponding author for the article.



http://www.blackwell-synergy.com/doi/abs/10.1111

INFECTION AND IMMUNITY, Oct. 2008, p. 4396–4404 Vol. 76, No. 100019-9567/08/$08.00�0 doi:10.1128/IAI.00547-08Copyright © 2008, American Society for Microbiology. All Rights Reserved.

Effects of Clostridium perfringens Beta-Toxin on the Rabbit SmallIntestine and Colon�

Jorge E. Vidal,1 Bruce A. McClane,1 Juliann Saputo,2 Jaquelyn Parker,2 and Francisco A. Uzal2*Department of Microbiology and Molecular Genetics, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania,1 and

California Animal Health and Food Safety Laboratory System, San Bernardino Branch, School of Veterinary Medicine,University of California, Davis, California2

Received 5 May 2008/Returned for modification 10 June 2008/Accepted 5 July 2008

Clostridium perfringens type B and type C isolates, which produce beta-toxin (CPB), cause fatal diseasesoriginating in the intestines of humans or livestock. Our previous studies demonstrated that CPB is necessaryfor type C isolate CN3685 to cause bloody necrotic enteritis in a rabbit ileal loop model and also showed thatpurified CPB, in the presence of trypsin inhibitor (TI), can reproduce type C pathology in rabbit ileal loops.We report here a more complete characterization of the effects of purified CPB in the rabbit small and largeintestines. One microgram of purified CPB, in the presence of TI, was found to be sufficient to cause significantaccumulation of hemorrhagic luminal fluid in duodenal, jejunal, or ileal loops treated for 6 h with purifiedCPB, while no damage was observed in corresponding loops receiving CPB (no TI) or TI alone. In contrast tothe CPB sensitivity of the small intestine, the colon was not affected by 6 h of treatment with even 90 �g ofpurified CPB whether or not TI was present. Time course studies showed that purified CPB begins to inducesmall intestinal damage within 1 h, at which time the duodenum is less damaged than the jejunum or ileum.These observations help to explain why type B and C infections primarily involve the small intestine, establishCPB as a very potent and fast-acting toxin in the small intestines, and confirm a key role for intestinal trypsinas an innate intestinal defense mechanism against CPB-producing C. perfringens isolates.

Clostridium perfringens is an anaerobic, spore-forming, gram-positive pathogen of humans and domestic animals (10). Thevirulence of C. perfringens is largely attributable to its prolificsecretion of toxins. However, no single isolate produces all ofthe more than 15 toxins reported in the literature, providingthe basis for a commonly used classification scheme that as-signs C. perfringens isolates to one of five types (A to E),depending upon their production of four (alpha, beta, epsilon,and iota) typing toxins (10, 15). C. perfringens type B or Cisolates both produce alpha- and beta-toxins (CPA and CPB,respectively). In addition, type B isolates also produce epsilon-toxin, a potent neurotoxin listed as a class B select agent by theCenters for Disease Control and Prevention/U. S. Departmentof Agriculture.

Type B isolates cause an often-fatal hemorrhagic dysen-tery in sheep, and possibly in other species, that is accom-panied by sudden death or acute neurological signs (23).Intestinal lesions of those infected animals are characterizedby diffuse necrohemorrhagic enteritis, predominantly in theileum, with serosanguineous fluid in the abdominal cavity(27). There is currently limited information regarding thepathogenesis of type B-associated diseases, although someevidence indicates that both CPB and epsilon-toxin maycontribute to lethality (3).

C. perfringens type C isolates also cause fatal diseases rang-ing from enteritis to enterotoxemia, predominantly in newborn

animals of most livestock species. Infected animals typicallyshow necrohemorrhagic enteritis, which can result in death dueto direct intestinal damage or, probably more commonly, fromtoxemia after the absorption of toxins from the intestines intothe circulation (23, 24). Type C-associated diseases annuallyresult in serious economic losses for the agricultural industry.

In humans, C. perfringens type C isolates cause enteritisnecroticans (also known as Darmbrand or Pigbel), a diseasethat is endemic in much of Southeast Asia but particularlyPapua New Guinea (6). Although less common, this diseasealso occurs in diabetic patients from developed countries. Per-sons suffering from enteritis necroticans often survive less than48 h after the first appearance of symptoms (9, 16, 21, 26).Histologically, the disease is characterized by necrotic enteritisand the presence of numerous bacteria in the intestinal lumen(29). Immunohistochemistry studies using anti-CPB antibodiesshowed the presence of CPB on the necrotic intestinal epithe-lium of humans suffering from type C infection (9).

CPB is a 35-kDa protein that forms pores in the membraneof susceptible cell lines, which leads to swelling and cell lysis(12, 17, 25). CPB is also lethal for mice, with a calculated 50%lethal dose of 1.87 �g/per kg of body weight when administeredvia the intraperitoneal route. A relatively crude beta-toxoidwas shown to protect animals and humans against type Cinfection (5, 28), suggesting that CPB is important for thevirulence of type C isolates. However, the lack of a good smallanimal model and difficulties in producing C. perfringens mu-tants had prevented fulfilling molecular Koch’s postulates toestablish a definitive relationship between CPB and type Cvirulence. In response, we recently developed a rabbit ilealloop model for type C disease and improved mutagenesis tech-niques for C. perfringens. These advances were used to con-

* Corresponding author. Mailing address: California Animal Healthand Food Safety Laboratory System, San Bernardino Branch, Univer-sity of California-Davis, 105 West Central Avenue, San Bernardino,CA 92408. Phone: (909) 383-4287. Fax: (909) 884-5980. E-mail: [email protected].

� Published ahead of print on 14 July 2008.

4396

at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

struct a series of C. perfringens type C toxin-null mutants, whichdemonstrated that beta-toxin, but not perfringolysin O (PFO)or CPA, is necessary and sufficient for type C isolates to causedamage in rabbit ileal loops (19). Moreover, CPB is sufficientto damage ileal loops since purified CPB reproduced the nat-ural pathology of type C disease in rabbit ileal loops, providedthat trypsin inhibitor (TI) was present to prevent CPB degra-dation by endogenous trypsin (19).

This TI requirement to obtain CPB activity in rabbit ilealloops reflects natural type C disease in animals and humans.Risk factors for developing necrotizing enteritis from type Cisolates include low trypsin production due to a protein-poordiet or pancreatic disease and consumption of foods, such assweet potato, containing a high concentration of a TI. Also, thecolostrum ingested by newborn animals has powerful inhibi-tory properties against trypsin. These risk factors contribute tothe persistence of trypsin-sensitive CPB in the gastrointestinaltract during type C infection (10).

Despite these recent advances, the intestinal effects of CPBremain poorly characterized. Therefore, this study used therabbit intestinal loop model to investigate the pathologicaleffects of purified CPB in the colons, duodena, jejuna, and ileaof rabbits.

MATERIALS AND METHODS

Strain and bacterial culture media. C. perfringens type C strain CN3685 (plc�,pfoA�, cpb�, and tpeL�), which was isolated from the peritoneal fluid of a sheepwith struck (a type C infection of adult sheep), was used to purify beta-toxin, asdescribed below. The bacterial culture media used throughout the present studyincluded fluid thioglycolate medium (Difco Laboratories), TGY (3% tryptic soybroth [Becton Dickinson], 2% glucose [Sigma Aldrich], 1% yeast extract [BectonDickinson], 0.1% sodium thioglycolate [Sigma Aldrich]), and TSC agar medium(SFP agar [Difco Laboratories] supplemented with 0.04% D-cycloserine [SigmaAldrich]).

Purification of CPB protein. An isolated colony of CN3685 from a TSC agarplate was inoculated into fluid thioglycolate medium and grown overnight at37°C. An aliquot (0.1 ml) of this overnight culture was then transferred to 30 mlof TGY and grown at 37°C for �8 h. The 30-ml culture was transferred to 3 litersof fresh TGY and grown at 37°C for another �8 h. The culture was then chilledimmediately on ice for 10 min and centrifuged at 10,000 � g for 20 min. Proteinsin the culture supernatant were precipitated using 40% ammonium sulfate(Fisher Scientific), with constant stirring, at 4°C for �1 h. The precipitate wasthen collected by centrifugation at 10,000 � g for 30 min. The pellet resultingfrom the 40% saturation ammonium sulfate cut was resuspended in 40 ml of 30mM Tris-HCl buffer (pH 7.5) and dialyzed overnight against the same buffer (4l), with several changes, at 4°C. After the dialyzed solution was again centrifugedat 10,000 � g for 30 min, the supernatant was filtered through a 0.45-�m-pore-size filter (Millipore) and loaded onto a DEAE-CL6B Sepharose column(Sigma). This column was pre-equilibrated with 30 mM Tris-HCl buffer (pH 7.5)in an AKTA prime system (Amersham Bioscience). After loading of the sample,the DEAE-CL6B column was washed with 45 ml of 30 mM Tris-HCl buffer (pH7.5), and bound CPB was then eluted from the column using a gradient of NaCl(0 to100 mM) in 30 mM Tris-HCl buffer (pH 7.5). Fractions were assessed for thepresence of CPB by Western blotting using a mouse monoclonal anti-CPBantibody obtained from P. Hauer (Center for Veterinary Biologics, Ames, Iowa).Fractions containing the purified CPB were pooled and dialyzed with ice-coldphosphate-buffered saline (PBS [pH 7.4]) at 4°C overnight. Pooled fractionswere then concentrated by ultrafiltration using an Ultrafree 10-kDa cutoffcentrifugal filter device (Millipore) and stored at �80°C. The final concen-tration of purified CPB was estimated by Lowry assay, using bovine serumalbumin as the standard (8).

Analysis of the CPB toxin preparation purity. CPB purity was analyzed bysodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) anddensitometric analysis, which showed the preparation to be �95% homoge-neous. An aliquot (25 �l) of the ammonium sulfate supernatant-concentratedproteins or 1 �g of purified CPB was electrophoresed in a SDS–12% PAGE geland then either stained with Coomassie blue or electrophoresed and transferred

to nitrocellulose membrane. The membrane was then probed using a monoclonalanti-CPB antibody, a monoclonal anti-CPA antibody, or a rabbit polyclonalanti-PFO antibody. These analyses show no contamination in our CPB toxinpreparation with CPA or PFO (Fig. 1).

Effects of purified beta-toxin on rabbit intestinal loops. (i) Inoculum. In allexperiments an �1-ml aliquot of a Ringer’s solution containing specifiedamounts of purified CPB, without or with 150 �g of TI/ml, was injected into eachintestinal loop. For the dose-response experiments a mixture containing Ringer’ssolution with 1, 5, 10, or 20 �g of purified CPB and TI/ml was injected into eachloop. In time course experiments a mixture of Ringer’s solution with 10 �g ofpurified CPB and TI/ml was injected in each loop. Alternatively, some loopsreceived (i) an injection of Ringer’s solution containing only 20 �g of purifiedCPB (no TI)/ml or (ii) a mixture of Ringer’s solution containing TI and purifiedCPB (10 �g/ml) that had been preincubated for 45 min at room temperature ona rocking platform with a neutralizing anti-CPB monoclonal antibody (MAb; 200�g) or anti-CPA MAb (200 �g) (provided by P. Hauer). The amount of anti-CPBand anti-CPA used was the minimum amount that was protective in a mouseintravenous bioassay (unpublished observation). In all experiments, control loopsreceived a similar volume (�1 ml) of sterile Ringer’s solution containing 150 �gof TI/ml.

(ii) Rabbit loop model. Fasted young adult, male or female, New ZealandWhite rabbits (Charles River, California) were premedicated with acepromazine,xylazine, and burprenorphine. Anesthesia was then induced with ketamine andmaintained with inhalatory isofluorane. A laparotomy was performed via the midline, and the small intestine or colon was exposed. Lengths (�2 cm) of eachindividual intestinal section (duodenum, jejunum, ileum, or colon) were isolatedby ligation, leaving an empty segment of gut between the loops. Care was takento avoid overdistension of bowel loops and interference with the blood supply,eliminating a possible ischemic component to the toxin-induced damage. Thecolon content was washed with saline solution injected into the lumen, followedby a gentle massage before colonic loops were prepared. During surgery, theserosal surface of the loops was kept wet by frequent soaking with normal salinesolution. After injecting the inoculum, the abdominal incision was closed byseparate muscle and skin sutures, and the animals were kept deeply anesthetizedthroughout the experiment.

FIG. 1. Purity analysis of CPB preparations. To purify CPB, pro-teins in the supernatant of C. perfringens type C isolate CN3586 wereconcentrated by using 40% of ammonium sulfate saturation. CPB waspurified by anion-exchange chromatography as described in Materialsand Methods. (A) An aliquot of ammonium sulfate supernatant-con-centrated proteins (ASCP) and 1 �g of purified CPB (CPB) weresubjected to SDS–12% PAGE and stained with Coomassie blue orelectrophoresed and transferred to nitrocellulose membrane. Themembrane was immunoblotted with a mouse monoclonal anti-CPBantibody (B), a mouse monoclonal anti-CPA antibody (C), or a rabbitpolyclonal anti-PFO antibody (D). Bound antibody was detected witha horseradish peroxidase-conjugated secondary anti-mouse or -rabbitIgG antibody and incubation of blots with a chemiluminescent sub-strate. The numbers at the left are in kilodaltons.

VOL. 76, 2008 INTESTINAL EFFECTS OF BETA-TOXIN 4397

at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

(iii) Measurement of fluid accumulation and histological analyses. After a 6-h(dose-response experiment) or a 15-, 30-, or 60-min or 6-h (time course study)CPB treatment, the rabbits were euthanized by an overdose of sodium barbitu-rate (Beuthanasia, Schering-Plough Animal Health, Kenilworth, NJ). The ab-dominal cavity was then reopened, and the small intestinal loops were excised inthe same order that they had been inoculated. Loops were cut out and weighed,before and after the fluid was removed, and the length was measured. Fluidsecretion was expressed as the loop weight-to-length ratio (mg/cm).

For histological analysis, all tissues were fixed by immersion in 10% buffered(pH 7.4) formalin for 24 to 48 h, followed by dehydration through gradedalcohols to xylene before being embedded in paraffin wax. Sections (4 �m thick)were cut and stained with hematoxylin and eosin according to standard proce-dures. Tissue sections were examined by a pathologist in a blinded fashion, usinga quantitative scoring system as described previously (19). Briefly, the degree ofdamage was scored by using a scale of 1 to 5. On this scale, a “1” indicates nohistologic damage, while “2,” “3,” “4.” and “5” values indicate increasingly severedamage. Histologic parameters considered in this evaluation included mucosalnecrosis, desquamation of the epithelium, inflammation, villous blunting, edema,and hemorrhage. Sections of representative treated or control tissues were thenphotomicrographed using an Olympus microscope (Tokyo, Japan) at a 100� ora 200� final magnification. All procedures were reviewed and approved by theUniversity of California, Davis Committee for Animal Care and Use (permit04-11593).

Statistical analyses. To statistically validate our results, each experiment wasperformed with at least six repetitions in six different animals. All statisticalanalyses were done by using the Minitab 15 software. The fluid accumulationdata were analyzed by using two-way analysis of variance with the post hoc test.The histology data were analyzed by using the Friedman test.

RESULTS

Effects of purified CPB on fluid accumulation in rabbitintestinal loops. We recently demonstrated that purified CPBcan cause intestinal lesions in rabbit ileal loops (19). To char-acterize and compare the activity of CPB in various parts of theintestine, we first examined the effect of increasing doses ofCPB purified (Fig. 1) to near homogeneity (�95%), along withTI, in loops constructed in the rabbit jejunum, duodenum,ileum, or colon.

Compared to their corresponding control loops, 6-h treat-ment with the lowest dose of purified CPB tested (1 �g) in-duced a very conspicuous fluid accumulation in rabbit loopsmade in the duodenum, jejunum, or ileum (Fig. 2 and 3). Thefluid accumulation with all of the doses of CPB tested wasstatistically significant in CPB-treated loops versus their cor-responding control loops (P � 0.05). However, no statisticallysignificant difference was observed in fluid accumulation be-tween any of the small intestinal segments treated with any ofthe doses of CPB tested (Fig. 3). This indicated that a 6-htreatment with 1 �g of CPB was sufficient to induce near-maximal luminal fluid accumulation in the rabbit small intes-tine. The intestinal loop fluid, however, became progressivelybloodier as the CPB dose increased.

To confirm that the observed fluid accumulation and thepresence of luminal blood were induced by CPB rather than bya contaminant, the purified CPB (20 �g) was preincubatedwith a neutralizing anti-CPB MAb prior to its injection intoileal loops. This CPB MAb preincubation completely elimi-nated the ability of the toxin preparation to cause bloody fluidaccumulation in rabbit ileal loops (data not shown). A similarpreincubation of purified CPB with a neutralizing MAb againstalpha-toxin had no inhibitory effect on CPB ileal loop activity(data not shown).

In contrast to the observed effect of very small amounts ofpurified CPB plus TI on all three small intestinal regions, a 6-h

treatment with even 90 �g of purified CPB/ml, whether in-jected in the presence or absence of TI, did not induce fluidaccumulation or hemorrhage in the rabbit colon (Fig. 4A).These results indicated that purified CPB specifically elicitsabundant fluid accumulation and hemorrhage in the rabbitsmall intestine.

Evidence suggests that endogenous trypsin plays an impor-tant role as an innate intestinal defense mechanism againstCPB secreted by C. perfringens type C isolates during naturaldisease (5, 7). To further corroborate the protective role oftrypsin against CPB, small-intestine loops were treated withCPB in the absence of TI. Although rabbit small intestinalloops treated with 20 �g of CPB and TI showed abundantbloody fluid accumulation (Fig. 2), duodenal, jejunal, or ilealloops injected with 20 �g of CPB without TI exhibited noaccumulation or bloody fluid (Fig. 2 and 3). Moreover, fluidlevels present in loops injected with CPB without TI weresimilar to those found in negative control loops receiving onlyan injection of Ringer’s solution and TI (Fig. 2 and 3). Theseresults further support endogenous trypsin as playing a decisiveprotective role against CPB-induced small intestinal damage.

Since fluid accumulation had become prominent in small-intestine loops after 6 h of treatment with CPB and TI, a timecourse study was performed to evaluate when CPB begins toaffect the small intestine. Confirming Fig. 2 and 3, bloody fluidaccumulation was again observed after a 6-h treatment of theduodenum, jejunum, or ileum with 10 �g of CPB and TI, but

FIG. 2. Gross pathology of rabbit duodenal (top), jejunal (middle),or ileal (bottom) loops treated with Ringer’s solution containing spec-ified doses (1, 5, 10, or 20 �g) of CPB and TI (150 �g/ml); 20 �g ofpurified CPB with no TI (CPB/NTI); or 150 �g of TI/ml (control).After treatment, loops were incubated for 6 h. Note that as the dose ofCPB increased, the bloody fluid content in all loops receiving bothpurified CPB and TI also increased. No hemorrhagic fluid was ob-served in loops receiving an injection of CPB alone (no TI) or TI alone(no CPB). The data shown are representative of six repetitions.

4398 VIDAL ET AL. INFECT. IMMUN.

at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

no fluid accumulation beyond control levels (TI without CPB)was observed in any loops treated for 6 h with the same doseof CPB in the absence of TI (Fig. 5). Treatment of differentsmall intestinal segments with purified CPB and TI for 60 minconsistently caused some fluid accumulation over matchingcontrols, but this effect reached statistical significance (P �0.05) only in ileal loops (Fig. 5), which also contained traces ofblood in their luminal fluid by this treatment time (not shown).Significant fluid accumulation over matching controls was not

FIG. 3. Fluid accumulation of rabbit intestinal loops treated withdifferent doses of purified CPB. Rabbit duodenal (top), jejunal (mid-dle), or ileal (bottom) loops were treated with Ringer’s solution con-taining 1 (CPB 1), 5 (CPB 5), 10 (CPB 10), or 20 (CPB 20) �g ofpurified CPB and 150 �g of TI/ml; 20 �g of purified CPB with no TI(CPB20/NTI); or Ringer’s solution and 150 �g of TI/ml (control). Aftera 6-h treatment, fluid accumulation was recorded as described in Mate-rials and Methods. Fluid accumulation in all loops injected with any doseof CPB plus TI was statistically different from control loops (*, P � 0.05)or loops treated with CPB and no TI (**, P � 0.05). Every experimentwas independently performed six times; the data shown represent themean of these studies, and small bars represent standard errors.

FIG. 4. The rabbit colon is not affected by purified CPB. Rabbitcolonic loops were treated for 6 h with Ringer’s solution containing 90�g of purified CPB and TI (CPB) or Ringer’s solution and TI (con-trol). (A) Fluid accumulation in colonic loops was recorded as de-scribed in Materials and Methods. No statistically significant differencewas observed between CPB-treated and control colonic loops. (B) His-tology. Colonic loops treated with Ringer’s solution containing 90 �gof purified CPB and TI (CPB) or Ringer’s solution and TI (control)showed intact intestinal villi with a well-preserved epithelium andlamina propria. Tissues were then processed by histology using hema-toxylin and eosin stain. Preparations were photographed at 200� mag-nification. The data shown are representative of six experimentalrepetitions.


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

detected after 15 or 30 min of combined CPB and TI treatmentof any small intestinal region (data not shown). These resultsindicate that luminal fluid influx develops most quickly in theileum.

Histopathological damage induced by purified CPB in rab-bit intestinal loops. The results presented above demonstratedthat, in the presence of TI, CPB can induce bloody fluid accu-mulation in all small intestinal regions between ca. 1 and 6 h oftreatment. This observation suggested that purified CPB plusTI might be causing intestinal tissue damage. Therefore, his-tological damage in these loops was assessed. This analysisshowed that duodenal, jejunal, or ileal rabbit loops treatedwith any tested dose of purified CPB plus TI exhibited severetissue damage by 6 h (Table 1). In all small intestinal regions,the extent of these histological alterations in loops treated withCPB and TI was statistically different from control loops(Ringer’s solution plus TI but lacking CPB, Table 1).

The severity and extent of this CPB-induced 6-h intestinaldamage showed a statistically significant dose dependency induodenal loops (Fig. 6 and Table 1). Dose-dependent histo-logical damage was also observed after a 6-h treatment of thejejunum or ileum with CPB plus TI, although the dose-depen-dent differences in tissue damage did not reach statistical sig-nificance (Table 1). In general, the same dose of purified CPB(plus TI) induced more severe histological damage in the je-junum and ileum versus the duodenum. For example, necrosisof the epithelium or lamina propria, villous blunting, and in-flammation were more severe with any given dose of purifiedCPB plus TI injected into the jejunum or ileum versus theduodenum (Table 1), indicating that these two segments of thesmall intestine are more susceptible to CPB-induced damagethan is the duodenum.

Confirming that CPB was the active agent inducing the in-testinal damage in Fig. 6, preincubation of purified CPB with aneutralizing monoclonal anti-CPB antibody, but not with amonoclonal anti-CPA antibody, totally abolished the ability ofa CPB and TI mixture to cause histological damage (data notshown). In addition, rabbit duodenal, jejunal, or ileal loopstreated with purified CPB in the absence of TI exhibited nointestinal lesions, appearing histologically indistinguishablefrom negative control loops receiving only Ringer’s solutionand TI (Fig. 6 and 7 and Table 1).

No histologic damage was observed when the colon wastreated with CPB (90 �g) and the TI. For example, after 6 h oftreatment with CPB and TI, the colon appeared similar to thecontrol colon treated only with Ringer’s solution and TI (Fig.4B). Together, these results indicated that purified CPB, in areduced trypsin environment, causes histologic damage in therabbit small intestine but not in the colon, at least under theexperimental conditions used in the present study.

To help clarify whether the CPB-induced bloody fluid accu-mulation shown in Fig. 2 and 3 might be linked to the toxin’sability to induce severe tissue damage, histological lesions werealso assessed at early CPB treatment time points. As shown inFig. 8, normal histology was observed when the duodenum wastreated with 10 �g of purified CPB for 15 min in the presenceof TI. However, at the same time point, the jejunum and ileumwere already showing slight damage, including some destruc-tion of the villi tips. By 30 min or 1 h of CPB treatment in thepresence of TI, the extent of histological alterations in duode-

FIG. 5. Time course fluid accumulation in rabbit intestinal loops.Duodenal (top), jejunal (middle), or ileal (bottom) loops were treatedfor 1 or 6 h (as indicated) with Ringer’s solution containing 10 �g ofpurified CPB, along with TI (CPB) or Ringer’s solution with TI (con-trol). Other loops were treated for 6 h with Ringer’s solution contain-ing 10 �g of purified CPB but no TI (CPB 6 h/NTI). Fluid accumula-tion was recorded as described in Materials and Methods. Statisticallysignificant differences (*, P � 0.05) in fluid accumulation were ob-served in ileal loops treated for 1 h with CPB and TI versus controlloops (control 1 h). Loops (duodenal, jejunal, or ileal) treated for 6 hwith purified CPB and TI (CPB 6 h) also showed statistically significantdifferences (**, P � 0.05) from control loops (control 6 h) or loopstreated only with CPB (CPB 6 h NTI) (***, P � 0.05). Every exper-iment was performed independently six times; the data shown are themean values. Small bars represent standard errors.


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

nal, jejunal, or ileal loops treated with CPB and TI was statis-tically different from their corresponding control loops (Fig. 8and Table 2).

DISCUSSION

Type B and C diseases originate in the intestines but oftenlater involve sudden death or acute neurological signs resultingfrom enterotoxemia (absorption of toxins from the intestinesinto the circulation) (10, 23). In nearly all livestock and hu-mans the intestinal effects of C. perfringens type C isolatesinvolve necrotizing enteritis, which is clinically characterized byabundant bloody diarrhea, abdominal pain, and distension (13,23, 27). CPB has long been implicated in type B and type Cdisease, but its contribution to intestinal disease has only re-cently been demonstrated (19). Although type B and C isolatesusually produce between three and five different lethal toxins(3, 5), studies with isogenic toxin-null mutants have demon-strated that CPB is necessary to reproduce the intestinal pa-thology of type C isolate CN3685 in rabbit ileal loops (19).That study also showed, for the first time, that purified CPBalone is sufficient to cause bloody fluid accumulation and his-tologic damage in ileal loops (19). The key to demonstratingthe enteric activity of purified CPB in that study was the in-clusion of TI in the CPB treatment, which mimicked naturaltype C disease conditions, where high levels of TI and/or lowlevels of trypsin are present due to diet or disease.

The present study now significantly extends those initialfindings by characterizing more completely the intestinal ef-fects of purified CPB. A trypsin inhibition approach was againused to demonstrate that, in addition to the ileum, purified

FIG. 6. Purified CPB induces dose-dependent histologic damage induodenal loops. Duodenal loops were treated for 6 h with Ringer’s solu-tion containing: 1, 5, 10, or 20 �g of purified CPB along with TI or else TIalone (control). Other loops were treated for 6 h with Ringer’s solutioncontaining 20 �g of purified CPB without TI (CPB 20 �g NTI). Loopstreated with increasing doses of purified CPB and TI (CPB) showedprogressive tissue damage, which included necrosis and loss of epithelium,necrosis of lamina propria, blunting of the villi, hemorrhage of the mu-cosa, and diffuse neutrophilic infiltration of mucosa and submucosa (seeTable 1 for details). In contrast, duodenal loops injected with Ringer’ssolution and TI (control) or purified CPB without TI (CPB 20 �g NTI)showed intact intestinal villi with a well-preserved epithelium and laminapropria. Tissues were processed by histology using hematoxylin and eosinstain. Sections of treated or control tissues were then photomicrographedat 200� final magnification. Shown are representative photomicrographsof six repetitions for each condition.

TABLE 1. Rabbit loop pathology in 6-h incubation dose-response experiment

Site, treatment,and CPB dose

(�g)a

Rabbit loop pathology (mean degree of damage � SD)b

Desquamation ofepithelium

Necrosis ofepithelium

Necrosis oflamina propria Inflammation Edema Villous

bluntingOverallseverity

DuodenumCPB (1) 2.2 � 0.5a 2.1 � 0.6a 1.2 � 0.5 1.0 � 0.0 1.5 � 0.5a 1.9 � 0.4a 2.0 � 0.5a

CPB (5) 2.4 � 0.4a 2.3 � 0.4a 1.5 � 0.4a 1.3 � 0.5a 1.2 � 0.4 1.5 � 0.0a 2.3 � 0.2a

CPB (10) 3.0 � 0.3a 3.0 � 0.3a 2.0 � 0.0a 1.9 � 0.2a 1.8 � 0.4a 2.3 � 0.4a 3.0 � 0.3a

CPB (20) 3.6 � 0.7a,b 3.6 � 0.7a,b 2.8 � 0.8a,b 2.1 � 0.5a,b 2.2 � 0.4a,b 2.4 � 0.6a,b 3.6 � 0.6a,b

CPB (20)/NTI 1.2 � 0.4 1.2 � 0.4 1.0 � 0.0 1.0 � 0.0 1.2 � 0.4 1.2 � 0.4 1.4 � 0.4Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0

JejunumCPB (1) 3.3 � 1.2a 3.3 � 1.0a 3.3 � 1.0a 2.5 � 0.8a 2.6 � 0.7a 3.6 � 1.4a 3.3 � 1a

CPB (5) 3.6 � 0.7a 3.7 � 0.5a 3.3 � 1.2a 2.8 � 0.4a 2.8 � 0.4a 4.0 � 1.2a 3.6 � 0.6a

CPB (10) 3.9 � 0.2a 3.8 � 0.4a 3.7 � 0.8a 2.8 � 0.4a 2.8 � 0.4a 4.2 � 0.8a 3.9 � 0.2a


CPB (20)/NTI 2.2 � 1.3 2.2 � 1.3 1.6 � 0.5 1.6 � 0.5 1.8 � 0.8 1.5 � 0.5 2.0 � 0.8a

Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.3 � 0.4 1.0 � 0.0 1.0 � 0.0

IleumCPB (1) 3.1 � 1.1a 2.5 � 1.3a 2.8 � 1.4a 2.0 � 1.1a 2.5 � 0.5a 3.3 � 1.4a 3.1 � 1.1a

CPB (5) 3.8 � 0.4a 3.5 � 0.6a 3.8 � 0.4a 3.0 � 0.0a 3.0 � 0.0a 4.3 � 0.6a 3.8 � 0.4a

CPB (10) 3.8 � 0.4a 4.0 � 0.0a 3.8 � 0.4a 3.0 � 0.0a 3.0 � 0.0a 4.5 � 0.0a 3.9 � 0.2a


CPB (20)/NTI 1.3 � 0.4 1.2 � 0.3 1.0 � 0.0 1.0 � 0.0 1.5 � 0.4 1.5 � 0.6 1.3 � 0.2Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.1 � 0.2 1.0 � 0.0 1.0 � 0.0

a NTI, no TI; control, Ringer’s solution plus TI (no CPB).b A superscript “A” indicates a statistically significant difference (P � 0.05) relative to the control loop using the Friedman test. A superscript “B” indicates a

statistically significant difference (P � 0.05) relative to the loop treated with (CPB 20 �g) without TI using the Friedman test.


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

CPB produces bloody fluid accumulation and tissue damage inthe rabbit jejunum and duodenum. However, the duodenumwas found to be less CPB-sensitive than the jejunum or ileum.These regional CPB sensitivity differences provide one expla-nation for gastrointestinal aspects of natural type B or type Cdiseases, which primarily involve the jejuna and ilea of infectedhumans and animals (2, 9, 13, 27). The reduced CPB sensitivityof the duodenum in animal models and natural disease may beattributable, at least in part, to pancreatic secretion producinghigher intestinal trypsin levels in the duodenum, which thenreduces CPB activity (30).

Although purified CPB exhibited activity in all sections ofthe rabbit small intestines, the present study found that therabbit colon is not affected by even high CPB doses applied inthe presence of TI. These differences in CPB sensitivity be-tween the rabbit small intestine versus the colon provide oneexplanation for the predominant involvement of the small in-testine in type B and C infections of both humans and animals.

However, there are also now some emerging reports of colonicdamage in human type C disease (2, 21), which may involve (i)another type C toxin (not CPB) possessing colonic activity, (ii)species-specific differences in colonic CPB susceptibility, or(iii) CPB colonic damage requiring higher CPB doses or longertreatment times than used in the present study. Sorting outthese possibilities will require additional experiments, but it isnotable that another C. perfringens toxin, i.e., the enterotoxin,resembles CPB by also failing to damage rabbit colonic loopsand yet possessing some limited activity on the human colon invivo and ex vivo (1, 4, 11).

The present study also shows, for the first time, that CPBacts very quickly in the small intestine, with the duodenum,jejunum, or ileum all exhibiting visible tissue damage withinthe first hour of CPB treatment (Fig. 8). However, after 1 h ofCPB treatment, no fluid accumulation differences were ob-served between control versus the toxin-treated duodenum and

FIG. 7. CPB-induced histologic damage in the rabbit small intes-tine. Intestinal loops constructed in the duodenum (top row), jejunum(middle row), or ileum (bottom row) were treated with Ringer’s solu-tion containing 20 �g of purified CPB and TI (CPB), Ringer’s solutionwith TI (control) or Ringer’s solution with 20 �g of purified CPB butno TI (CPB/NTI). After 6 h, duodenal, jejunal, or ileal loops treatedwith purified CPB and TI (CPB) showed severe damage (see Table 1for details), but loops treated with TI but no CPB (control) or withpurified CPB but no TI (CPB/NTI) showed normal, full-length intes-tinal villi with a well-preserved epithelium and lamina propria. Intes-tinal tissues were processed by histology using a hematoxylin and eosinstain. Sections of treated or control tissues were then photomicro-graphed at 200� final magnification. Representative photomicro-graphs of six repetitions for each condition are shown.

FIG. 8. Time course development of histologic damage induced byCPB in the rabbit small intestine. Intestinal loops constructed in theduodenum (top row), jejunum (middle row), or ileum (bottom row)were treated with Ringer’s solution containing 10 �g of purified CPBand TI (CPB) or Ringer’s solution and TI (control). Loops were thenincubated for 15 min (CPB 15 min), 1 h (CPB 1 h), or 6 h (CPB 6 h).Note that after 15 min, jejunal or ileal loops treated with purified CPBand TI (CPB 15 min) showed destruction of the villus tips. Duodenalloops remained normal after 15 min of incubation with purified CPBand TI. Histologic damage then increased from moderate (1-h incu-bation period) to severe (6-h incubation period) in all loops treatedwith purified CPB and TI. Note the presence of necrosis and loss ofepithelium, necrosis of lamina propria, blunting of the villi, hemor-rhage of the mucosa, and diffuse neutrophilic infiltration of mucosaand submucosa. Intestinal loops injected with Ringer’s solution and TI(control) retained normal, full-length intestinal villi with a well-pre-served epithelium and lamina propria. Intestinal tissues were pro-cessed by histology using hematoxylin and eosin stain. Sections ofrepresentative treated or control tissues were photomicrographed at200� final magnification. Shown are representative photomicrographsof six repetitions for each condition.


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

jejunum, and only modest CPB-induced fluid accumulationwas detected in the ileum. Since luminal fluid accumulationdeveloped after the onset of CPB-induced damage and usuallyincluded the presence of blood, these results are consistentwith the accumulation of bloody luminal fluid in CPB-treatedsmall intestine resulting, at least in part, from severe mucosalnecrosis. CPB-induced inflammation may also contribute tothis intestinal bleeding since there is a well-established associ-ation between hemorrhage and inflammation for other enterictoxins and pathogens (20). Similar hemorrhaging and intestinalnecrosis as seen with purified CPB has also been reported forrabbit ileal loops injected with whole cultures of type C strainCN3685, but not with its isogenic CPB-null mutants (19), sup-porting the importance of CPB production in the ability of typeC (and possibly type B) isolates to cause enteric effects duringnatural disease.

The present study also demonstrates that CPB possessesconsiderable enteric potency. Histologic damage was observedin duodenal, jejunal, and ileal rabbits loops treated for 6 h, inthe presence of TI, with only 1 �g of purified CPB. A dose-dependent CPB effect on tissue damage was noted in therelatively CPB-insensitive duodenum. However, the dose de-pendency of CPB-induced tissue damage in the jejunum orileum was statistically insignificant, probably because near-maximal effects had already been produced with the lowestCPB dose tested in the present study. Future studies mightdetermine the minimal dose of CPB causing enteric effects inrabbit small-intestinal loops, but the current findings establishthat, in the presence of TI, CPB is considerably more potentin rabbit intestinal loops than is C. perfringens enterotoxin(18, 22).

The enteric potency of CPB demonstrated in the presentstudy could help explain, at least in part, the relatively rapidand often fatal progression of type C disease. For example, itis now clear that, in reduced trypsin conditions, even low dosesof CPB are sufficient to rapidly induce significant necrosis ofthe epithelium. Beyond its enteric potency now demonstrated

in the present study, CPB possesses the second lowest 50%lethal dose of all C. perfringens toxins when administered in-travenously to mice (5). This suggests that the absorption ofeven small amounts of CPB from a damaged small intestinemay be sufficient to cause death or damage to internal organs.It is notable in this respect that no rabbits died during thepresent experiments, which may suggest that longer treatmenttimes or higher CPB doses are required to cause systemiclethality, at least in the rabbit ileal loop model. The presentstudy’s observation of extensive CPB-induced enteric damagein the absence of lethality supports the general view that deathduring type B or type C diseases mainly results from toxemiarather than from intestinal pathology (23).

CPB is very sensitive to the action of trypsin, although themolecular basis for this sensitivity is unknown. Endogenousintestinal trypsin is known to play an important role as aninnate defense mechanism against type C infection (10). Law-rence and Cooke showed that C. perfringens type C could causea pigbel-like disease in guinea pigs fed with a persistent lowprotein diet combined with dietary protease inhibitors (7).Enteric lesions similar to those observed in human pigbel caseshave also been successfully reproduced by injecting a type Cculture into lambs (14) or rabbit ileal loops (19), but only whena TI was given as well. Attempts to reproduce type C disease inanimal models injecting CPB alone (no TI) have consistentlyfailed (23). In this report and a previous study (19), we havenow demonstrated that the presence of TI allows purified CPBto reproduce type C-like pathology in small intestinal loops.Since CPB is very sensitive to trypsin (5, 17), these resultssuggest that the presence of even small amounts of endoge-nous trypsin remaining in washed small intestinal loops can besufficient to inactivate CPB if a TI is not administered simul-taneously with the toxin.

The small-intestine histological alterations observed by us-ing purified CPB in the present study and our previous work(19) are very similar to those described in type C naturaldiseases. For example, type C isolates produce in sheep intes-

TABLE 2. Rabbit loop pathology in time-course experiments

Site and incubationtime in min (CPB 10 �g)

Rabbit loop pathology (mean degree of damage � SD)a

Desquamation ofepithelium

Necrosis ofepithelium

Necrosis oflamina propria Inflammation Edema Villous blunting Overall severity

Duodenum15 1.4 � 0.1a 1.4 � 0.1a 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.3 � 0.130 1.5 � 0.3a 1.5 � 0.3a 1.0 � 0.0 1.0 � 0.0 1.3 � 0.3 1.0 � 0.0 1.4 � 0.1a

60 1.8 � 0.3a 1.8 � 0.3a 1.2 � 0.4 1.0 � 0.0 1.3 � 0.3 1.0 � 0.0 1.6 � 0.2a

Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0

Jejunum15 2.0 � 0.5a 2.0 � 0.5a 1.5 � 0.5a 1.5 � 0.5 1.5 � 0.5 1.5 � 0.5 1.9 � 0.7a

30 2.5 � 0.3a 2.5 � 0.3a 1.8 � 0.3a 1.5 � 0.5 1.8 � 0.3a 2.3 � 0.3a 2.5 � 0.3a

60 2.8 � 0.6a 2.8 � 0.6a 2.2 � 0.7a 1.8 � 1.0 1.8 � 0.3a 2.3 � 0.3a 2.8 � 0.6a

Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0

Ileum15 1.6 � 0.2a 1.6 � 0.2a 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.5 � 0.3a

30 2.3 � 0.3a 2.3 � 0.3a 1.8 � 0.3a 1.5 � 0.5 1.8 � 0.3a 2.3 � 0.3a 2.3 � 0.3a

60 2.8 � 0.5a 2.8 � 0.5a 2.0 � 0.5a 1.8 � 0.8 2.1 � 0.7a 2.8 � 0.3a 2.8 � 0.5a

Control 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0 1.0 � 0.0

a A superscript “A” indicates a statistically significant difference (P � 0.05) relative to the control loop using the Friedman test.


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

tinal lesions consisting of diffuse or multifocal hemorrhagicand necrotizing enteritis, mainly in the ileum, with excess ofsanguineous serous fluid in the abdominal cavity (27). In pig-lets, type C infection also produces necrotizing enteritis withdeep mucosal necrosis and emphysema in small intestine,sometimes extending to the colon (24). The similarity noted inthis and our previous studies (3, 5) between intestinal lesionscaused by purified CPB and the lesions of natural type B andC disease, coupled with the inability of isogenic CPB-null mu-tants to cause intestinal pathology (19), further support CPB asa critical virulence factor for type C (and likely type B) disease.

In summary, this research shows that, in the presence of TI,CPB induce enteric effects in the duodenum, jejunum, or ileumthat resemble those in type C intestinal disease. However,colonic loops were found to be unaffected by similar treatment.Small-intestinal lesions developed quickly and required only asmall amount of the purified toxin, indicating that CPB ishighly active in the rabbit small intestine. These findings pro-vide new insights into the overall pathogenic mechanism offatal diseases induced by C. perfringens type B and C isolatesand support the importance of CPB immunity for vaccine-induced protection against type B or C disease.

ACKNOWLEDGMENTS

This study was supported by grant R01 AI056177-04 (B.A.M.) fromthe National Institute of Allergy and Infectious Diseases. J.E.V. re-ceived generous support from the Mexican National Council of Sci-ence and Technology (CONACyT).

We thank P. Hauer (Center for Veterinary Biologics, Ames, Iowa)for supplying monoclonal antibodies against CPB and CPA, RodTweten for supplying PFO antibody, and Jim Cravotta (University ofCalifornia at Davis) for his substantial assistance in rabbit surgery. Wealso thank Richard D. Day and The-Minh Luong of the BiostatisticsConsulting Service of the University of Pittsburgh for their assistancein statistical analyses.

REFERENCES

1. Bos, J., L. Smithee, B. McClane, R. F. Distefano, F. Uzal, J. G. Songer, S.Mallonee, and J. M. Crutcher. 2005. Fatal necrotizing colitis following afoodborne outbreak of enterotoxigenic Clostridium perfringens type A infec-tion. Clin. Infect. Dis. 40:e78–e83.

2. Farrant, J. M., Z. Traill, C. Conlon, B. Warren, N. Mortensen, F. V. Gleeson,and D. P. Jewell. 1996. Pigbel-like syndrome in a vegetarian in Oxford. Gut39:336–337.

3. Fernandez-Miyakawa, M. E., D. J. Fisher, R. Poon, S. Sayeed, V. Adams, J. I.Rood, B. A. McClane, and F. A. Uzal. 2007. Both epsilon-toxin and beta-toxinare important for the lethal properties of Clostridium perfringens type Bisolates in the mouse intravenous injection model. Infect. Immun. 75:1443–1452.

4. Fernandez-Miyakawa, M. E., V. Pistone-Creydt, F. A. Uzal, B. A. McClane,and C. Ibarra. 2005. Clostridium perfringens enterotoxin damages the humanintestine in vitro. Infect. Immun. 73:8407–8410.

5. Fisher, D. J., M. E. Fernandez-Miyakawa, S. Sayeed, R. Poon, V. Adams, J. I.Rood, F. A. Uzal, and B. A. McClane. 2006. Dissecting the contributions ofClostridium perfringens type C toxins to lethality in the mouse intravenousinjection model. Infect. Immun. 74:5200–5210.

6. Johnson, S., and D. N. Gerding. 1997. Enterotoxemic infections, p. 117–140.

In J. I. Rood, B. A. McClane, J. G. Songer, and R. W. Titball (ed.), Theclostridia: molecular biology and pathogenesis. Academic Press, London,United Kingdom.

7. Lawrence, G., and R. Cooke. 1980. Experimental pigbel: the production andpathology of necrotizing enteritis due to Clostridium welchii type C in theguinea pig. Br. J. Exp. Pathol. 61:261–271.

8. Lowry, O. H., N. J. Rosebrough, A. L. Farr, and R. J. Randall. 1951. Proteinmeasurement with the Folin phenol reagent. J. Biol. Chem. 193:265–275.

9. Matsuda, T., Y. Okada, E. Inagi, Y. Tanabe, Y. Shimizu, K. Nagashima, J.Sakurai, M. Nagahama, and S. Tanaka. 2007. Enteritis necroticans ‘pigbel’in a Japanese diabetic adult. Pathol. Int. 57:622–626.

10. McClane, B. A., F. A. Uzal, M. Fernandez-Miyakawa, D. Lyerly, and T. D.Wilkins. 2004. The enterotoxigenic clostridia, p. 698–752. In S. F. M. Dwor-kin, E. Rosenburg, K. F. Schleifer, and E. Stackebrandt (ed.), The pro-karyotes, vol. 4. Springer-Verlag, New York, NY.

11. McDonel, J. L., and G. W. Demers. 1982. In vivo effects of enterotoxin fromClostridium perfringens type A in the rabbit colon: binding versus biologicactivity. J. Infect. Dis. 145:490–494.

12. Nagahama, M., S. Hayashi, S. Morimitsu, and J. Sakurai. 2003. Biologicalactivities and pore formation of Clostridium perfringens beta toxin in HL 60cells. J. Biol. Chem. 278:36934–36941.

13. Niilo, L. 1988. Clostridium perfringens type C Enterotoxemia. Can. Vet. J.29:658–664.

14. Niilo, L. 1986. Experimental production of hemorrhagic enterotoxemia byClostridium perfringens type C in maturing lambs. Can. J. Vet. Res. 50:32–35.

15. Petit, L., M. Gibert, and M. R. Popoff. 1999. Clostridium perfringens: toxino-type and genotype. Trends. Microbiol. 7:104–110.

16. Petrillo, T. M., C. M. Beck-Sague, J. G. Songer, C. Abramowsky, J. D.Fortenberry, L. Meacham, A. G. Dean, H. Lee, D. M. Bueschel, and S. R.Nesheim. 2000. Enteritis necroticans (pigbel) in a diabetic child. N. Engl.J. Med. 342:1250–1253.

17. Sakurai, J., and C. L. Duncan. 1978. Some properties of beta-toxin producedby Clostridium perfringens type C. Infect. Immun. 21:678–680.

18. Sarker, M. R., R. J. Carman, and B. A. McClane. 1999. Inactivation of thegene (cpe) encoding Clostridium perfringens enterotoxin eliminates the abilityof two cpe-positive C. perfringens type A human gastrointestinal diseaseisolates to affect rabbit ileal loops. Mol. Microbiol. 33:946–958.

19. Sayeed, S., F. A. Uzal, D. J. Fisher, J. Saputo, J. E. Vidal, Y. Chen, P. Gupta,J. I. Rood, and B. A. McClane. 2008. Beta toxin is essential for the intestinalvirulence of Clostridium perfringens type C disease isolate CN3685 in a rabbitileal loop model. Mol. Microbiol. 67:15–30.

20. Sears, C. L., and J. B. Kaper. 1996. Enteric bacterial toxins: mechanisms ofaction and linkage to intestinal secretion. Microbiol. Rev. 60:167–215.

21. Severin, W. P., A. A. de la Fuente, and M. F. Stringer. 1984. Clostridiumperfringens type C causing necrotizing enteritis. J. Clin. Pathol. 37:942–944.

22. Sherman, S., E. Klein, and B. A. McClane. 1994. Clostridium perfringens typeA enterotoxin induces tissue damage and fluid accumulation in rabbit ileum.J. Diarrhoeal. Dis. Res. 12:200–207.

23. Songer, J. G. 1996. Clostridial enteric diseases of domestic animals. Clin.Microbiol. Rev. 9:216–234.

24. Songer, J. G., and F. A. Uzal. 2005. Clostridial enteric infections in pigs. J.Vet. Diagn. Investig. 17:528–536.

25. Steinthorsdottir, V., H. Halldorsson, and O. S. Andresson. 2000. Clostridiumperfringens beta-toxin forms multimeric transmembrane pores in human en-dothelial cells. Microb. Pathog. 28:45–50.

26. Tonnellier, M., E. Maury, J. Guglielminotti, and G. Offenstadt. 2001. A fatalsandwich. Lancet. Infect. Dis. 1:202.

27. Uzal, F. A. 2004. Diagnosis of Clostridium perfringens intestinal infections insheep and goats. Anaerobe 10:135–143.

28. Walker, P. D., W. H. Foster, P. A. Knight, D. S. Freestone, and G. Lawrence.1979. Development, preparation and safety testing of a Clostridium welchiitype C toxoid. I: preliminary observations in man in Papua New Guinea.J. Biol. Stand. 7:315–323.

29. Walker, P. D., T. G. Murrell, and L. K. Nagy. 1980. Scanning electronmicroscopy of the jejunum in enteritis necroticans. J. Med. Microbiol. 13:445–450.

30. Zeitlin, I. J., and W. Sircus. 1974. Factors influencing duodenal trypsin levelsfollowing a standard test meal as a test of pancreatic function. Gut 15:173–179.

Editor: S. R. Blanke


at University of P

ittsburgh HS

LS on S

eptember 18, 2008

iai.asm.org

Dow

nloaded from

http://iai.asm.org

university of pittsburgh - cinvestav eugenio... · 2013-11-11 · general de mis estudios es...

Documents