Journal of Medical Microbiology (2005), 54, 155–157

DOI 10.1099/jmm.0.45808-0

Short Communication
Correspondence R. Alonso ralonso.hgugm@salud. madrid.org

An improved protocol for pulsed-field gel electrophoresis typing of Clostridium difficile
´ ´ R. Alonso, A. Martın, T. Pelaez, M. Marın, M. Rodrıguez-Creixems and E. Bouza ´ ´ ´
Department of Clinical Microbiology and Infectious Diseases, Hospital General Universitario ‘‘Gregorio Maranon’’, C/Doctor Esquerdo, 46, 28007 Madrid, Spain ˜´

Received 2 July 2004 Accepted 24 October 2004

Pulsed-field gel electrophoresis (PFGE) is the ‘gold standard’ technique for bacterial typing and has proved to be discriminatory and reproducible for typing Clostridium difficile. Nevertheless, a high proportion of strains are non-typable by this technique due to the degradation of the DNA during the process. The introduction of several modifications in the PFGE standard procedure increased typability from 40 % (90 isolates) to 100 % (220 isolates) while maintaining the high degree of discrimination and reproducibility of the technique.

Introduction
Clostridium difficile is the main aetiologic agent of nosocomial diarrhoea (Barbut et al., 1996) and is responsible for an important increase in hospital stays, with high healthcare and economic repercussions (Wilcox et al., 1996). Typing of the micro-organism can be useful in the detection and control of epidemic outbreaks and endemic situations, and in the characterization of recurrences (Alonso et al., 2001; Kato et al., 1996). Different approaches have been used for typing bacteria, although PFGE has traditionally been the gold standard (Finney, 1993). PFGE has proved to be discriminatory and reproducible for typing C. difficile; however, a considerable proportion of strains are non-typable by this technique due to degradation of the DNA during the procedure, making uninterpretable gel smears (Kato et al., 1994; Bidet et al., 2000). Different modifications of the typing procedure have been proposed (Corkill et al., 2000), but they have led to variable results with only partial improvement (Klaassen et al., 2002; Fawley & Wilcox, 2002). The aim of this study was to develop a new PFGE protocol which can offer universal typability for all our C. difficile strains.

Standard protocol. Briefly, the recommended standard protocol was

as follows. A bacterial culture was inoculated into LB broth with agitation and grown to an OD600 of 0.8–1.0. Then 5 3 108 bacteria were removed for each millilitre of agarose plug to be made and centrifuged to obtain the bacterial pellet. The supernatant was discarded and the pellet resuspended in 500 ìl of cell suspension buffer. The bacterial suspension was warmed to 50 8C. The 2 % CleanCut agarose was melted and equilibrated to 50 8C. Five hundred millilitres of agarose was added to the bacterial suspension and mixed thoroughly. The mixture was transferred to plug moulds and the agarose was allowed to solidify at 4 8C for 10–15 min. The solidified plugs were incubated in 1 mg mlÀ1 lysozyme solution (100 ìl 25 mg mlÀ1 lysozyme stock plus 2.5 ml lysozyme buffer) for 2 h at 37 8C. The lysozyme was removed and the plugs were rinsed with sterile water. The plugs were incubated with .20 U mlÀ1 proteinase K solution (100 ìl of .600 U mlÀ1 proteinase K stock in 2.5 ml of proteinase K reaction buffer) at 50 8C for 24–96 h. The plugs were washed four times in 13 wash buffer for 1 h each at room temperature with gentle agitation. The second or third wash contained 1 mM PMSF to inactivate residual proteinase K. Prior to endonuclease restriction, each plug was washed for 1 h in 1 ml 0.13 wash buffer and then rinsed with fresh 0.13 wash buffer. The plug was incubated for 1 h with 1 ml of 13 restriction enzyme buffer at room temperature and then overnight in 300 ìl of 13 restriction enzyme buffer containing 30–50 U of the restriction enzyme at the appropriate temperature. After overnight digestion, the buffer was removed and the plug was incubated for 30 min in 13 wash buffer and then equilibrated in the appropriate concentration of gel-running buffer (e.g. 0.53 TBE). The manufacturer recommended loading one-quarter to one-third of a plug per well for electrophoresis. Although the manufacturer does not give any recommendation about the electrophoresis itself, most of the standard published series use 1 % agarose gels in 0.53 TBE buffer and the electrophoresis parameters vary widely depending on the restriction endonuclease used.
Modified protocol. Our protocol was based upon the standard

Methods
Samples and tests. Two hundred and twenty toxigenic C. difficile

isolates were included in this study. We obtained isolates from the diarrhoeic stool samples of patients admitted to our hospital over a period of 6 months. Cultures, identification and toxin detection were carried out using standard procedures. The CHEF bacterial DNA plug kit (BIO-RAD) was used for PFGE typing the isolates. The standard manufacturer’s recommended technique and our modified protocol were performed and compared.

commercial procedure as described, with the following modifications.
This paper was presented at the First International Clostridium difficile Symposium, Kranjska Gora, Slovenia, 5–7 May 2004.

Fresh 24 h plate cultures were always used. When unfreezing strains, they were grown three times in CCFA or Brucella agar before starting. 155

45808 & 2005 SGM Printed in Great Britain

. was included in every run and was always correctly typed). always freshly prepared and kept in darkness until use. In conclusion. 2001. Alonso and others BHI tubes were inoculated and incubated for 24 h to obtain a highdensity inoculum (OD600 . difficile using (a) standard procedures and (b) the modified protocol. ribotyping scheme (Bidet et al. Incubations of more than 24 h were not performed. Some other authors have reported the same phenomenon with PFGE and C. Secondly.R. as a result of which we do not know whether our non-typable strains belong to those reported major groups. a considerable proportion (32.. Lysozyme and proteinase K (Sigma) were prepared ‘in-house’ and used at high concentrations (2 mg mlÀ1 and 75 U mlÀ1 . Journal of Medical Microbiology 54 . Some strains may produce fewer nucleases. Bidet et al.g. According to this. Endonuclease digestions were performed with high concentrations of enzyme (SmaI. 1). 25 %) were from several others (data not shown). The concentration of thiourea is controversial. others report concentrations of up to 200 ìM in both the gel and the buffer (Fawley & Wilcox. 2000). 2001.1. the micro-organism has very potent endonucleases that may degrade the extracted DNA inside the agarose plug during the extended process (Spigaglia et al. Samore et al. non-typable strains belonged to serogroup G and ribotype 1 (O’Neill et al.. Proteinase K incubation was no longer than 18 h.. Firstly. 2000). 1995). we think it is extremely important to avoid spore formation. C. 156 ´ ˜ This paper has been supported in part by ‘Red Espanola de Investigacion ´ en Patologıa Infecciosa’ (REIPI – CO3/14) and by ‘Fondo de Investigaciones Sanitarias’ (FIS 02/1049).. Thiourea (Sigma). the strains belonging to our major ribotype could be more likely to produce spores. difficile isolates to be characterized by PFGE (Fig. almost 60 % of the assayed isolates (130) remained untyped by PFGE (Fig. enabling all our C... although others have also reported the problem in different strains (Kato et al. difficile by PFGE. we did not serotype our strains and we used a different Acknowledgements Fig. Cultures with a high proportion of spores (over 40 %) were ruled out. 75 %) belonged to our most prevalent ribotype (R1). 60 U in 300 ìl volumes). 2002). In most cases. 1. Electrophoresis was run at 195 V for 20 h with pulse times starting at 4 s and ending at 30 s.g. Example of PFGE patterns belonging to five representative clinical isolates of C. Although most of our non-typable isolates (98. Hypothetically.2). Washing steps were reduced to 30 min. We also recommend using high lysozyme and proteinase K concentrations to facilitate spore lysis. respectively). 2000). Electrophoresis. 1997). In our proposed protocol. it has been reported that a nucleolytic peracid derivative of Tris may form at the anode during electrophoresis and this chemical nucleolysis could be inhibited by the addition of thiourea to the running buffer (Ray et al. 1996) (some strains from this serogroup are typable. We removed 109 bacteria for each millilitre of agarose plug to be made. the reference strain. The high concentrations of proteinase K used could also help to digest nucleases more efficiently. we shortened incubation and washing times in order to avoid or reduce such degradation. which could explain both the difficulty in PFGE typing and a favourable spread in the hospital... 2000). visualized under UV transillumination and recorded. Thirdly. Spigaglia et al. difficile at different frequencies. which could be removed or degraded during nucleic acid extraction. The remaining strains offered good PFGE patterns that were reproduced on every occasion (e. This effect has also been described for other micro-organisms ¨ ¨ (e. With the described modifications to the protocol we increased the typability of the technique. 1b). 2001). was included in the gel and running buffer in high concentrations (200 ìM). Pseudomonas aeruginosa. Results and Discussion Using the standard procedure. We found that 50 ìM in the buffer had a reduced effect whereas the best results were achieved with 200 ìM in gel and buffer... Gels were stained with ethidium bromide. Romling & Tummler. difficile ATCC 9689. Unfortunately. many outbreaks have been reported to be caused by PFGE nontypable strains (Kato et al. Several factors have been proposed to explain the lack of typability in C. We thank Thomas O’Boyle for his help with the correction of the English version of this manuscript. A fresh preparation of the BHI culture was observed by microscopy at 31000 magnification to check for the presence of spores. 2001). Cultures with a high rate of sporulated forms probably exhibit poorer lysis and produce less DNA. 1993. and while some authors defend that 50 ìM is enough (Corkill et al. Kato et al. For this reason we recommend working with fresh cultures and screening for the presence of spores by microscopy before starting the DNA-extraction process. the above-mentioned modifications enabled us to achieve complete typability (increasing from 40 % to 100 %) within the assayed isolates while maintaining both a high degree of discrimination and the reproducibility of the technique.

& Horrevorts. 3546–3547... J Med Microbiol 50. 220–223. Corkill. Y. O’Neill. Hart. K. Bowman. K. M. P. Bidet. (2002).org 157 . 1391–1395. Pelaez.. B.. & Petit. Financial burden of hospital-acquired Clostridium difficile infection. H. Barbut.. J Clin Microbiol 38. J Clin Microbiol 32. difficile strains by pulsed-field gel electrophoresis and PCR-ribotyping. B. Y. R. T. difficile References Alonso. N. P. arbitrarily primed PCR.. & Stubbs. Pulsed-field gel strand scission during electrophoresis. Delmee. A. H. Cunniffe. Gros. 23–30. Prevalence and pathogenicity of Clostridium difficile in hospitalized patients. M.. R. Garcia-de-Viedma.. Killgore. can yield DNA fingerprints of degradation-susceptible Clostridium difficile strains. T. E. Pulsed-Field Gel Electrophoresis. animals and their environments. Tris-dependent oxidative DNA ribotyping. Samore. J Hosp Infect 34.. & Wilcox. van Haren. A molecular characterization of Clostridium difficile isolates in humans. M.. T. S. Kato. J Clin Microbiol 39. 86–92.. M. C. (1993). S. Cardines. ¨ Romling. 1233–1238. Graham. C. C. Ray.. A.. 2484–2487. S. H. Barbut. Arch Intern Med 156. 464–465. 407–414. G.. Salauze. Pulsed-field gel electrophoresis Pseudomonas aeruginosa by pulsed-field gel electrophoresis. Wilcox. RodriguezCreixems. Avesani. V. 888–894.. Kato. F. Epidemiol Infect 111. H. M. K. J.. 1449–1454. Sakata. H. T. J Clin Microbiol 38. & 13 other authors (1996). V. & Mastrantonio. Curr Microbiol 33. C. 101–104. 2067–2070....Improved protocol for PFGE typing of C. & Bouza.. Rossi.. Molecular typing and long-term comparison of Clostridium to the study of Clostridium difficile in a neonatal intensive care unit. & Tummler. K. Hashira. A. Kato. Watanabe. G.. (1996). G. (1994). Lalande. Watanabe. P. R.. New York: Greene- Wiley. J. Comparison of PCR- fingerprinting of Clostridium difficile isolates: pulsed-field gel electrophoresis versus amplified fragment length polymorphism. J. http://jmm.. 2791–2792. N. Watanabe. J Infect Dis 176. H. Menozzi. and pulsed-field gel electrophoresis for typing Clostridium difficile. (2000). J Hosp Infect 48. G. & 8 other authors (1997). Ueno. W. (2001).. A French multicenter study.. & Dyson. J Clin Microbiol 38. C. G. N. H. Analysis of Clostridium difficile isolates from nosocomial outbreaks at three hospitals in diverse areas of Japan. Klaassen.. (2001). K. Spigaglia. E. J Clin Microbiol 40. Mills. Charpak.sgmjournals. Corthier. (2000).... N. A. U. B. Kato.. P. (1996). M. J. & 7 other authors (2001). Achieving 100 % typeability of ¨ electrophoresis of degradation-sensitive DNAs from Clostridium difficile PCR ribotype 1 strains.. A. R.. (1993). V. S. K. Kato. (2002). Molecular infection in HIV-positive patients suffering from recurrent Clostridium difficile associated diarrhoea. Finney. M. J Clin Microbiol 40. Adams. E. Electrophoresis 16. (1995). 257–264. H. Burghoffer.. Ushijima. Relapses or reinfections: analysis of a case of Clostridium difficileassociated colitis by two typing systems.. A. Johnson. Application of typing by pulsed-field gel electrophoresis Multicenter typing comparison of sporadic and outbreak Clostridium difficile isolates from geographically diverse hospitals. Rossier. & Redpath. & Fujita. & Riley. F. (2000). Trundle. D. & Abe... Molecular analysis of relapse vs re- Kato. Fawley. S. Ueno. M.