Characterization of Clostridium botulinum Strains Associated with an Infant Botulism Case in the United Kingdom
http://www.100md.com
微生物临床杂志 2005年第6期
Department of Food Microbiology and Toxicology, Food Research Institute, University of Wisconsin, 1925 Willow Drive, Madison, Wisconsin 53706-1187
Carpenders Close, Harpenden, Herts AL5 3HN United Kingdom
Foodborne Diseases Branch, Microbiological Safety Division, Food Standards Agency, Aviation House, 125 Kingsway, London WC2B 6NH, United Kingdom
Wyeth, Huntercombe Lane South, Taplow, Maidenhead, Berkshire SL6 0PH, and Faculty of Medicine, Imperial College, London W2 1NY, United Kingdom
ABSTRACT
The sixth case of infant botulism in the United Kingdom was reported in 2001. The case was caused by a type B strain of Clostridium botulinum. Strains of C. botulinum were isolated from the baby's feces and from foodstuffs in the household in an attempt to document transmission. The aims of this study were to characterize the strains of C. botulinum associated with the botulism case. This was performed using a variety of techniques, including C. botulinum culture phenotypic properties, neurotoxin characterization, and pulsed-field gel electrophoresis (PFGE) banding patterns. Cultures associated with this case as well as isolates from stored and historical samples were analyzed and compared. C. botulinum type B PFGE patterns from the infant and from an opened container of infant formula were indistinguishable, while the PFGE profile of a strain presumably isolated from an unopened archival container was unique. The results suggest that the unopened brand of formula was not the source for transmission of spores to the infant and that the strain was distinct from previous botulism cases in the United Kingdom. Since environmental testing was not performed, it is not possible to deduce other sources of transmission.
INTRODUCTION
In the United Kingdom infant botulism has been particularly rare, with only five previous confirmed single cases occurring in 1978 (1, 45), 1987 (2, 43), 1989 (29, 41), 1993 (3, 16), and 1994 (4, 5, 17). A sixth case of infant botulism, the subject of the present study, was confirmed in 2001 (6, 7). An isolated case of infant botulism in a predominantly breast-fed weaning 5-month-old infant was reported to the Public Health Laboratory Service (PHLS) (London) (6, 7). There were no obvious predisposing factors within the household, and there was no proximity to a farm, although the case occurred near a new house under construction and disruption of soil and formation of dust could lead to spread of Clostridium botulinum spores. Strains of C. botulinum type B were isolated from the baby's feces and from an opened container of powdered infant formula in the household by researchers in the United Kingdom. A C. botulinum strain was also presumptively isolated from an unopened container of powdered formula. This container was from a manufacturing lot of 122,388 cans in October 1998 and was therefore near the 3-year expiration date. This lot had been distributed throughout the United Kingdom. To our knowledge, the lot had not led to previously reported infant botulism cases. Five sealed retained archival containers of infant formula were submitted for examination by the manufacturer, one of which was reported to contain Clostridium botulinum type B. The case was suspected to be associated with the consumption of this source of infant formula. The linking of this case at the time with powdered infant formula led to considerable publicity as well as withdrawal and recall of product and of "sister" batches of the same lot from the shelves. It also spurred the company to conduct a thorough review of manufacturing procedures and quality control.
This infant botulism case was investigated in a collaborative effort by two independent laboratories utilizing microbial culture and molecular techniques for identification of the organism causing the incident and its suspected relation to infant formula. A blinded investigation into the genetic relatedness of the strains from the infant, the opened container and the closed container, as well as historical isolates, was performed using two typing methods: (i) amplified fragment length polymorphism (AFLP) performed at the PHLS in London (unpublished data) and (ii) pulsed-field gel electrophoresis (PFGE) at the University of Wisconsin—Madison in the United States. This communication presents analyses of the various C. botulinum strains isolated during the case by PHLS and from retrospective clinical and food isolates using PFGE. Finally, potential measures to investigate and prevent future cases of infant botulism are described.
MATERIALS AND METHODS
Bacterial strains. Twelve lyophilized samples (no. 1 to 12) received blinded from the PHLS were aseptically rehydrated in sterile TPGY media (50 g/liter trypticase peptone, 5 g/liter Bacto peptone, 4 g/liter D-glucose, 20 g/liter yeast extract, 1 g/liter cysteine-HCl, pH 7.4). The TPGY medium was sparged with nitrogen gas prior to autoclaving, and resazurin (2 mg per liter) was used to ascertain anaerobic conditions. These cultures were then grown in TPGY broth and in TPGY supplemented with cooked meat particles (75 g/liter) (TPGY/CM). Cultures were grown at both 30° and 37°C. Since all cultures were determined to be proteolytic, they were grown at 37°C in subsequent experiments. All bacterial media components were purchased from Becton Dickinson Microbiology Systems, Becton Dickinson and Company, Sparks, MD.
Cultural properties of Clostridium botulinum isolates. Bacterial cultures were examined microscopically to evaluate their purity. All cultures were also grown on TPGY agar plates and incubated aerobically and anaerobically in an anaerobic chamber (with a gas mix containing 80% nitrogen-10% carbon dioxide-10% hydrogen) to determine their purity. Mixed cultures (samples 1, 6, 10, 11, and 12) were heat shocked and/or ethanol treated and then plated on Clostridium botulinum isolation media (CBI; 40 g/liter trypticase peptone, 5 g/liter Na2HPO4, 2 g/liter NaCl, 2 g/liter D-glucose, 5 g/liter yeast extract, 20 g/liter agar, 100 ml/liter egg yolk suspension, 50% 250 mg/liter cycloserine, 76 mg/liter sulfamethoxazole, and 4 mg/liter trimethoprim) in an attempt to isolate pure cultures. Colonies that were determined to contain rod-shaped bacteria were reisolated from the anaerobic plates and grown in TPGY media for further analyses. Stock cultures were prepared from all purified strains. Proteolytic properties of the pure cultures were determined by examination of degradation of cooked meat particles during growth. After 5 days of growth in TPGY the culture pH was measured.
Botulinum neurotoxin A or B detection and serotyping. Neurotoxin titers in culture supernatants were determined using an intravenous injection method as previously described (32). Culture supernatants obtained from 5 days of growth in TPGY were diluted 1:2 in gel-phosphate buffer (30 mM sodium phosphate buffer, pH 6.3, 0.2% gelatin). One-hundred microliters of each sample were injected into the lateral tail veins of five mice (female ICR mice, 18 to 22 g [Sprague Dawley, Madison, WI]). The animals were observed for characteristic signs of botulism, and the time to death was recorded. The average time to death (ATTD) for 5 animals was converted to intraperitoneal (i.p.) 50% lethal dose (LD50)/ml using the formula i.p. LD50/ml = 6.9 x 106 – 3.4 x 105(ATTD) + 5.7 x 103(ATTD)2 – 31.9(ATTD)3. This assay has been demonstrated in our laboratory to correlate with the mouse i.p. bioassay (32).
Toxin serotype was determined by enzyme-linked immunosorbent assay (ELISA) using cell-free culture supernatants. Cultures were tested with C. botulinum type A and type B antitoxins. Five-hundred microliters of culture samples were centrifuged for 5 min at 5,000 x g to pellet cells, and the supernatants were used for assays of neurotoxin protein. Ninety-six-well microtiter plates (Corning International, Corning, NY) were coated overnight at 4°C with 100 μl of a 5-μg/ml solution of antibodies raised against highly purified BoNT/A or BoNT/B toxoid. Antibodies were purified by affinity chromatography as previously described (22), diluted in 100 mM sodium carbonate buffer, pH 9.6, and used for coating plates. Plates were washed with TBST (10 mM Tris-HCl, 0.1% bovine serum albumin [BSA; Sigma], 150 mM NaCl, 0.05% Tween 20, pH 7.4) to remove unbound capture antibody, and then treated with 300 μl of blocking solution (5% vitamin-free casein [Research Organics, Inc., Cleveland, OH] in 10 mM Tris-HCl, 0.9% NaCl, 6 mM NaOH, pH 7.6) for 1 h at room temperature. Blocking solution was removed, 100 μl of serial dilutions of culture supernatants and a botulinum neurotoxin type A and B complex standard in blocking solution were added to the microtiter plate wells, and the plates were incubated for 2 h at room temperature. Plates were washed with TBST, 100 μl of a 2-μg/ml solution of biotinylated botulinum neurotoxin type A or B affinity purified immunoglobulin G (IgG) in blocking buffer was added to the microtiter plate wells, and the plates were incubated for 1 h at room temperature. Plates were washed with TBST to remove unbound secondary antibody, 100 μl of a 1:10,000 dilution of an avidin alkaline phosphatase conjugate (ExtrAvidin; Sigma) in blocking buffer was added to the microtiter plate wells, and the plates were incubated for 1 h at room temperature. Plates were again washed with TBST and developed by addition of Gibco BRL ELISA Amplification System reagents (Life Technologies, Rockville, MD) according to the manufacturer's instructions. Plates were read at 490 nm in a SpectraFluor plate reader (Tecan U.S., Inc., Durham, NC). All analyses were conducted in triplicate.
PFGE. PFGE was performed to compare the genomic DNA patterns of the 12 PHLS isolates and well-characterized C. botulinum type A1 strains 62A, ATCC 3502, Hall A-hyper; type A2 strains Kyoto F and FRI-H1A2; and type B strains Okra B (proteolytic) and 17B (nonproteolytic) from our laboratory collection. Of particular importance to this study, strains Kyoto F and FRI-H1A2 are A2 isolates associated with infant botulism cases in Japan and the United States, respectively.
Bacterial cultures were grown anaerobically in Bellco anaerobic tubes in 10 ml of TPGY media until the optical density at 600 nm (OD600) reached approximately 0.6. Bacterial cells were fixed with formaldehyde to inhibit DNase activity (21). One milliliter of formaldehyde solution was injected through the septum into the 10-ml bacterial cultures. The contents were mixed, and tubes were incubated on ice for 30 min. The fixed bacterial cultures were then transferred to 15-ml centrifuge tubes, and cells were precipitated by centrifugation. Cell pellets were washed twice with 10 ml of 0.85% NaCl solution, resuspended in 1.5 ml of solution containing 10 mM Tris-HCl, pH 8.0, 1 M NaCl, and 100 mM EDTA and was incubated for 10 min in a 50°C water bath. Equal volumes of the cell suspension and 1.5% InCert agarose (BioWhitaker Molecular Applications, Rockland, ME) were mixed, and the cell-agarose mixtures were loaded into plug-forming molds. The gel molds were chilled in a refrigerator for 15 min, and agarose plugs were removed and incubated in 4 volumes of lysis buffer (10 mM Tris-HCl, pH 8.0, 1 M NaCl, 100 mM EDTA, 0.5% Brij 58, 0.2% deoxycholate, 0.5% sarcosyl, 100 μg/ml RNase A, 5 mg/ml lysozyme, 40 U/ml mutanolysin) at 37°C for 12 h. The lysis buffer was removed, and the plugs were rinsed with 50 mM EDTA solution to remove lysozyme, RNase and mutanolysin. Two volumes of PK buffer (10 mM Tris-HCl, pH 8.0, 0.5 M EDTA, 1% sarcosyl, and 1 mg/ml Proteinase K) were added, and the agarose plugs incubated for 48 h with gentle agitation in a 50°C water bath. Following Proteinase K treatment, the plugs were washed with TE buffer and incubated for 2 h at 37°C with shaking in TE buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF) to inactivate Proteinase K. The agarose plugs were washed three times for 2 h each with 10 volumes of TE buffer to remove PMSF. Restriction digests of PFGE plugs were performed with four restriction endonucleases (NruI, SacII, SmaI, and XhoI), and samples were separated using a clamped homogenous electric field system (CHEF-DRII; Bio-Rad, Hercules, CA) following the manufacturer's instructions.
RESULTS
Culture purification. Twelve lyophilized culture samples (no. 1 to 12) received blinded from the PHLS were aseptically rehydrated with sterile TPGY and grown in TPGY/CM medium. Cultures were plated on TPGY and incubated aerobically and anaerobically to assess their purity. Cultures grown in liquid media and on plates were examined microscopically to also evaluate purity. Five of the cultures, nos. 1, 6, 10, 11, and 12, were mixed cultures and contained rods and cocci. In order to isolate pure strains of these bacterial cultures, two approaches were used. The cultures were heat shocked or treated with 50% ethanol (23). The cultures were plated on CBI plates, on which C. botulinum colonies form a characteristic opaque zone surrounding them, indicating lipase activity. Colonies that contained rod-shaped bacteria were isolated from the anaerobic plates and grown in TPGY media for further analyses. Stock cultures were prepared from all purified strains.
C. botulinum growth characteristics. The C. botulinum strains were analyzed for phenotypic properties. These properties are summarized in Table 1. The purified cultures all had a rod morphology, which varied in length and width dependent on the strain. Certain cultures produced high spore titers, while others sporulated infrequently. The spores varied in size and morphology depending on the strain. After 5 days of growth in TPGY, the final pH values ranged from 5.9 to 6.6 due to the production of organic acids during fermentation. The cultures were all proteolytic but varied markedly in proteolysis, as evaluated by digestion of meat particles in TPGY/CM. All strains showed lipase activity on CBI. These results indicate that there was considerable variation among the strains in cultural properties.
Botulinal neurotoxin production and typing. The strains were evaluated for botulinum neurotoxin production and toxin serotype by ELISA, mouse bioassay, and neutralization of mouse toxicity by serotype-specific antitoxins (23). Except for culture 7 (nontoxinogenic control), all the strains produced either A or B botulinum neurotoxin. None of the toxic supernatants reacted with both A and B antitoxins. Biological toxicity was determined using the mouse LD50 bioassay. Cultures 2 and 5 produced type A neurotoxin, with titers of 1.3 x 105 and 5.6 x 105 mouse LD50/ml, respectively. The remaining nine toxic cultures produced type B neurotoxin, and the titers varied from 3 x 103 to 2.6 x 105 mouse LD50/ml.
PFGE analyses. PFGE was used to assess genetic relatedness among the PHLS isolates and well-characterized C. botulinum type A1 strains 62A, ATCC 3502, Hall A-hyper; type A2 strains Kyoto F and FRI-H1A2; and type B strains Okra B (proteolytic) and 17B (nonproteolytic) from our laboratory collection. Restriction digests of PFGE plugs were performed with four restriction endonucleases: NruI, SacII, SmaI, and XhoI. The digests were separated using a clamped homogenous electric field system. PFGE revealed that some of the isolates showed banding similarities, and we were able to assign them to arbitrary groups (Table 1, Fig. 1). PFGE group 1 included the PHLS isolates 2 and 5, which revealed identical banding patterns in all four digests. Interestingly, these PFGE patterns were identical to the patterns from the infant type A2 botulism-associated strains Kyoto F and FRI-H1A2 (with minor differences in NruI banding pattern). PFGE group 2 included PHLS isolates 1, 6, and 10, which yielded similar banding patterns. The SacII and NruI digestion patterns were identical for these three isolates, while isolates 1 and 10 showed identical SmaI digestion patterns and isolate 6 had a slightly different pattern. All three isolates yielded very similar patterns by XhoI, with minor differences in some band sizes. The banding pattern of these isolates (1, 6, and 10) did not reveal similarity to any of the type A or B C. botulinum strains analyzed. PFGE group 3 included PHLS isolates 4 and 8, which had identical banding patterns with all four restriction endonucleases. The patterns did not show any similarity to other C. botulinum strains examined. The PHLS isolates 3, 7, 9, 11, and 12 exhibited unique PFGE patterns with no similarity to the other PHLS isolates or C. botulinum type A and B strains from our collection used for comparison.
After these cultural, toxin, and PFGE analyses were completed, the strain origins were unblinded (Table 2). It was revealed that strains 1, 6, 10, 11, and 12 were isolated during the investigation of the infant botulism case in 2001. All of these strains produced type B botulinum neurotoxin. Isolates 1 (from the baby's feces) and isolates 6 and 10 (from opened containers of baby milk) had the same PFGE pattern (Group 2). However, isolate 6 differed from isolates 1 and 10 in cell morphology, spore titers, and toxin production in culture. Isolate 11 (from the baby's feces) produced type B toxin but had a unique PFGE pattern which did not show similarity to the other strains examined. Strain 12 (presumptively isolated from an unopened infant formula) also produced low levels of type B toxin, was strongly proteolytic in comparison to strains 1, 6, 10, and 11, and importantly had a unique PFGE pattern with no similarities to the other strains. Interestingly, isolates 2 and 5 (from different United Kingdom infant botulism cases), Kyoto F associated with an infant botulism case in Japan, and FRI-HIA2 from honey associated with an infant botulism case in the United States all produced type A toxin and had identical PFGE patterns. These results indicate that cultural, toxin, and PFGE analyses can clearly distinguish groups of C. botulinum, even from widely different geographic regions of the world.
DISCUSSION
Since the discovery of infant botulism in 1976 (34, 38), infant botulism has been detected very rarely in the United Kingdom. This study describes an investigation of the sixth case that occurred in 2001. Infant botulism is caused by the ingestion of spores of C. botulinum followed by gut colonization by the organism and absorption of the toxin into the blood (19, 34, 38). Epidemiological risk factors for infants include rural residence where agricultural practices are carried out and urban areas where building or earth moving for road construction are frequent (10). Ingestion of honey (10) or possibly corn syrup (36) are also risk factors. Although the predominance of cases of infant botulism recorded to date have occurred in the state of California, infant botulism has also been reported in other regions of the United States (8, 35), Italy (20), Denmark (12), Japan (30), Venezuela (46), Australia (33), and various other countries (10). Some reports suggest a link with sudden infant death syndrome (13) but others do not (15), and the possible relationship of sudden infant death syndrome and infant botulism remains controversial. Although infant formula feeding has occasionally been implicated in the transmission of infections such as salmonellosis and Enterobacter sakazakii (9, 11, 18), it has not been epidemiologically linked to infant botulism (10). Large numbers of samples of infant formula were tested for C. botulinum spores in the 1980s, and no evidence of contamination with spores of C. botulinum was found (S. Arnon, personal communication).
This investigation is unique in that independent studies were performed by two different laboratories, one in the United Kingdom at the PHLS and the other in the United States at the University of Wisconsin—Madison. The two laboratories used distinct genetic approaches to characterize C. botulinum strains. The United Kingdom study, which will be published elsewhere (unpublished data), used AFLP to analyze strains isolated during the 2001 infant botulism case. Our study used cultural properties, toxicity, and PFGE analyses to characterize the 2001 isolates as well as other PHLS isolates, some of which were associated with previous food and infant botulism incidents. These strains were received blinded in our laboratory from the PHLS. In addition to the PHLS cultures, several C. botulinum type A and B strains from our collection, including the A2 isolates Kyoto F and FRI-H1A2 previously associated with infant botulism, were included in the PFGE analyses.
Various methods have been used for molecular typing of C. botulinum, including ribotyping, PCR, repetitive element sequence-based PCR (rep-PCR), random amplified polymorphic DNA analysis (RAPD), amplified fragment length polymorphisms (AFLP), and PFGE (14, 25, 26, 27, 28, 40, 42). In our laboratory, we have found PFGE to be a rapid and accurate method of determining genomic relationships among groups and strains of C. botulinum (28, 31). AFLP and ribotyping are less time-consuming and are useful for typing of strains for which satisfactory high-molecular-weight chromosomal DNA preparations and restriction fragment digests cannot be obtained. In our experience as well as in certain other laboratories, PFGE analysis of C. botulinum strains has been the most discriminating technique (24, 25, 28). Computer programs have been developed to accurately analyze PFGE data and for molecular typing and disease surveillance for other bacterial pathogens in the PulseNet system (44).
In this study, the two isolates from the infant's fecal specimens differed in PFGE patterns. One isolate from the opened container matched an isolate from the infant's fecal specimen, and one isolate from the unopened container had a unique PFGE pattern. The fact that the strain presumptively isolated from the unopened container differed from those isolated from infant feces and the opened container of infant formula indicates that the source from which the infant was infected is not proven. It is not clear how the opened can was contaminated and the child was exposed, particularly since the home and environs near to the home were not sampled for C. botulinum during the investigation.
Since the commercial source of the baby formula was initially suspected to cause the infant botulism case, a large recall was undertaken on the basis that a single strain was presumptively isolated from an unopened container, suggesting an apparent link to a clinical case. Since the batch of this infant formula consisted of 122,388 containers manufactured in October 1998 and fed to around 30,000 infants within the United Kingdom through the 3-year shelf life, if contamination of the batch had been widespread it is likely that numerous cases of infant botulism would have occurred in 1998, but this was not the case. Although proprietary, the general ingredients of the infant formula consisted of reduced mineral whey, vegetable oils, skimmed-milk powder, and lactose. Due to the ubiquity of bacterial spores, these products could be contaminated. One of the processing steps was heat treatment at 90 to 95°C for 20 to 25 s, which would not be sufficient to inactivate heat-resistant bacterial spores. Therefore, it is possible that C. botulinum spores could survive the manufacturing process of the infant formula, but this would require a separate study and sampling of a statistically meaningful number of containers to show their presence in infant formula. Although the infant botulism case did not appear to be associated with the commercial "unopened" source of infant formula, nonetheless it stimulated the industry to consider various precautionary measures and actions to prevent infant botulism from powdered formula. Most importantly, measures should be taken to ensure that viable C. botulinum spores are not present in baby formula.
This study describes an investigation of the association of C. botulinum with an infant botulism case. The investigation utilized various methods, including microbiological analyses, toxin testing and typing, and molecular typing for epidemiological characterization and strain identification of isolates associated with infant botulism. It was evident from this study that typing methods can provide new insights into causation and transmission of infectious diseases, including infant botulism.
ACKNOWLEDGMENTS
We thank Eric Bolton, Moira Brett, Jim McLauchlin, and their colleagues for sending the strains and for the collaborative study.
This study was supported by sponsors of the Food Research Institute and the College of Agriculture and Life Sciences, University of Wisconsin, Madison.
REFERENCES
Anonymous. 1978. Communicable Disease Report no. 5.
Anonymous. 1987. Communicable Disease Report no. 38.
Anonymous. 1993. A case of infant botulism. Communicable Disease Report no. 29.
Anonymous. 1994. A case of infant botulism. Communicable Disease Report no. 12.
Anonymous. 1994. Quarterly communicable disease review January to March 1994. PHLS Communicable Disease Surveillance Centre. J. Public Health Med. 16:357-364.
Anonymous. 2001. A case of infant botulism. Communicable Disease Report 28 June.
Anonymous. 2001. Infant botulism: update. Communicable Disease Report 16 August.
Anonymous. 2003. Infant botulism-New York City, 2001-2002. Morb. Mortal. Wkly. Rep. 52:21-24.
Anonymous. 2004. Joint FAO/W.H.O. Workshop on Enterobacter sakazakii and other microorganisms in powdered infant formula, Geneva, Switzerland, 2 to 5 February 2004. [Online.] http://www.who.int/foodsafety/publications/micro/feb2004/en/.
Arnon, S. S. 2004. Infant botulism, p. 1758-1766. In R. D. Feigen and J. D. Cherry (ed.), Textbook of pediatric infectious diseases, 5th ed. W. B. Saunders, Philadelphia, Pa.
Baker, R. D. 2002. Infant formula safety. Pediatrics 110:833-835.
Balsev, T., E. Ostergaard, I. K. Madsen, and D. A. Wandall. 1997. Infant botulism. The first culture-confirmed Danish case. Neuropediatrics 28:287-288.
Bohnel, H., S. Behrens, P. Loch, K. Lube, and F. Gessler. 2001. Is there a link between infant botulism and sudden infant death Bacteriological results obtained in central Germany. Eur. J. Pediatr. 160:623-628.
Broda, D. M., J. A. Boerema, and R. G. Bell. 1998. A PCR survey of psychrotrophic Clostridium botulinum-like isolates for the presence of BoNT genes. Lett. Appl. Microbiol. 27:219-223.
Byard, R. W., L. Moore, A. J. Bourne, A. J. Lawrence, and P. N. Goldwater. 1992. Clostridium botulinum and sudden infant death syndrome: a ten-year prospective study. J. Paediatr. Child Health 28:156-157.
Communicable Disease Surveilllance Centre. 1993. A case of infant botulism. Commun. Dis. Rep. CDR Wkly. 3:129.
Communicable Disease Surveillance Centre. 1994. A case of infant botulism. Commun. Dis. Rep. CDR Wkly. 4:53.
Estrada, B. 2002. Infections associated with powdered infant formula. Infect. Med. 19:350.
Faucheux, R. C., A. K. Shetty, and G. S. Cowan. 1997. Infant botulism. Clin. Pediatr. 36:591-593.
Fenicia, L., L. Da Dalt, F. Anniballi, G. Franciosa, S. Zanconato, and P. Aureli. 2003. A case of infant botulism due to neurotoxigenic Clostridium butyricum type E associated with Clostridium difficile colitis. Eur. J. Clin. Microbiol. Infect. Dis. 21:736-738.
Gibson, J. R., K. Sutherland, and R. J. Owen. 1994. Inhibition of DNAse activity in PFGE analysis of DNA from Campylobacter jejuni. Lett. Appl. Microbiol. 19:357-358.
Goodnough, M. C., B. Hammer, H. Sugiyama, and E. A. Johnson. 1993. Colony immunoblot assay of botulinal toxin. Appl. Environ. Microbiol. 59:2339-2342.
Hatheway, C. L., and E. A. Johnson. 1998. Clostridium: the spore-bearing anaerobes, p. 731-782. In L. Collier, A. Balows, and M. Sussman (ed.), Topley and Wilson's microbiology and microbial infections, 9th ed., vol 2: systematic bacteriology. Arnold, London, United Kingdom.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1998. Genomic analysis of Clostridium botulinum group II by pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 64:703-708.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1998. Prevalence of Clostridium botulinum in Finnish trout farms: pulsed-field gel electrophoresis typing reveals extensive genetic diversity among type E isolates. Appl. Environ. Microbiol. 64:4161-4167.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1999. Ribotyping as an identification tool for Clostridium botulinum strains causing human botulism. Int. J. Food Microbiol. 47:121-131.
Hyyti, E., J. Bjrkroth, S. Hielm, and H. Korkeala. 1999. Characterization of Clostridium botulinum group I and II by randomly amplified polymorphic DNA analysis and repetitive element sequence-based PCR. Int. J. Food Microbiol. 48:179-189.
Johnson, E. A., and M. Bradshaw. 2001. Clostridium botulinum and its neurotoxins: a metabolic and cellular perspective. Toxicon 39:1703-1722.
Jones, S., Z. Huma, C. A. Haugh, Y. Young, F. Starer, and L. Sinclair. 1990. Central nervous system involvement in infantile botulism. Lancet i:228.
Kakinuma, H., H. Maruyama, H. Takahashi, K. Yamakawa, and S. Nakamura. 1996. The first case of type B infant botulism in Japan. Acta Paediatr. Japonica 38:541-543.
Lin, W.-J., and E. A. Johnson. 1995. Genome analysis of Clostridium botulinum by a pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 61:4441-4447.
Malizio, C. J., M. C. Goodnough, and E. A. Johnson. 2000. Purification of Clostridium botulinum type A neurotoxin. Methods Mol. Biol. 145:27-39.
May, M., M. Coulthard, G. Delbridge, J. McEniery, C. Burke, and M. Nissen. 2002. Difficulties in the diagnosis and management of infant botulism. J. Paediatr. Child Health 38:425-426.
Midura, T. F., and S. S. Arnon. 1976. Infant botulism. Identification of Clostridium botulinum and its toxins in faeces. Lancet ii:934-936.
Morris, J. G., Jr., J. D. Snyder, R. Wilson, and R. A. Feldman. 1983. Infant botulism in the United States: an epidemiology study of cases occurring outside of California. Am. J. Public Health 73:1385-1388.
Olsen, S. J., and D. L. Swerdlow. 2000. Risk of infant botulism from corn syrup. Pediatr. Infect. Dis. J. 19:584-585.
O'Mahoney, M., E. Mitchell, R. J. Gilbert, D. N. Hutchinson, N. T. Begg, J. C. Rodhouse, and J. E. Morris. 1990. An outbreak of foodborne botulism associated with contaminated hazelnut yoghurt. Epidemiol. Infect. 104:389-395.
Picket, J., B. Berg, E. Chaplin, and M. Brunstetter-Shafer. 1976. Syndrome of Clostridium botulinum in infancy: a clinical and electrophysiologic study. N. Engl. J. Med. 295:770-772.
Roberts, E., J. M. Wales, M. M. Brett, and P. Bradding. 1998. Vomiting with cranial nerve palsies. Lancet 352:1674.
Santos-Buelga, J. A., M. D. Collins, and A. K. East. 1998. Characterization of the genes encoding the botulinum neurotoxin complex in a strain of Clostridium botulinum producing type B and F neurotoxins. Curr. Microbiol. 37:312-318.
Sinclair, L., S. T. Jones, Y. Young, and C. A. Haugh. 1990. Infant botulism. Communicable Disease Report no. 05.
Skinner, G. E., S. M. Gendel, G. A. Fingerhut, H. A. Solomon, and J. Ulaszek. 2000. Differentiation between types and strains of Clostridium botulinum by ribotyping. J. Food Protection. 63:1374-1382.
Smith, G. E., F. Hinde, D. Westmoreland, P. R. Berry, and R. J. Gilbert. 1989. Infantile botulism. Arch. Dis. Childhood 64:871-872.
Swaminathan, B., T. J. Barrett, S. B. Hunter, R. V. Tauxe, and the CDC PulseNet Task Force. 2001. PulseNet: the molecular subtyping network for foodborne bacterial disease surveilance, United States. Emerg. Infect. Dis. 7:382-389.
Turner, H. D., E. M. Brett, R. J. Gilbert, A. C. Ghosh, and H. J. Leibeschuetz. 1978. Infant botulism in England. Lancet i:1277-1278.
Urdaneta-Carruyo, E., A. Suranyi, and M. Milano. 2000. Infantile botulism: clinical and laboratory observations of a rare neuroparalytic disease. J. Paediatr. Child Health 36:193-195.(Eric A. Johnson, William )
Carpenders Close, Harpenden, Herts AL5 3HN United Kingdom
Foodborne Diseases Branch, Microbiological Safety Division, Food Standards Agency, Aviation House, 125 Kingsway, London WC2B 6NH, United Kingdom
Wyeth, Huntercombe Lane South, Taplow, Maidenhead, Berkshire SL6 0PH, and Faculty of Medicine, Imperial College, London W2 1NY, United Kingdom
ABSTRACT
The sixth case of infant botulism in the United Kingdom was reported in 2001. The case was caused by a type B strain of Clostridium botulinum. Strains of C. botulinum were isolated from the baby's feces and from foodstuffs in the household in an attempt to document transmission. The aims of this study were to characterize the strains of C. botulinum associated with the botulism case. This was performed using a variety of techniques, including C. botulinum culture phenotypic properties, neurotoxin characterization, and pulsed-field gel electrophoresis (PFGE) banding patterns. Cultures associated with this case as well as isolates from stored and historical samples were analyzed and compared. C. botulinum type B PFGE patterns from the infant and from an opened container of infant formula were indistinguishable, while the PFGE profile of a strain presumably isolated from an unopened archival container was unique. The results suggest that the unopened brand of formula was not the source for transmission of spores to the infant and that the strain was distinct from previous botulism cases in the United Kingdom. Since environmental testing was not performed, it is not possible to deduce other sources of transmission.
INTRODUCTION
In the United Kingdom infant botulism has been particularly rare, with only five previous confirmed single cases occurring in 1978 (1, 45), 1987 (2, 43), 1989 (29, 41), 1993 (3, 16), and 1994 (4, 5, 17). A sixth case of infant botulism, the subject of the present study, was confirmed in 2001 (6, 7). An isolated case of infant botulism in a predominantly breast-fed weaning 5-month-old infant was reported to the Public Health Laboratory Service (PHLS) (London) (6, 7). There were no obvious predisposing factors within the household, and there was no proximity to a farm, although the case occurred near a new house under construction and disruption of soil and formation of dust could lead to spread of Clostridium botulinum spores. Strains of C. botulinum type B were isolated from the baby's feces and from an opened container of powdered infant formula in the household by researchers in the United Kingdom. A C. botulinum strain was also presumptively isolated from an unopened container of powdered formula. This container was from a manufacturing lot of 122,388 cans in October 1998 and was therefore near the 3-year expiration date. This lot had been distributed throughout the United Kingdom. To our knowledge, the lot had not led to previously reported infant botulism cases. Five sealed retained archival containers of infant formula were submitted for examination by the manufacturer, one of which was reported to contain Clostridium botulinum type B. The case was suspected to be associated with the consumption of this source of infant formula. The linking of this case at the time with powdered infant formula led to considerable publicity as well as withdrawal and recall of product and of "sister" batches of the same lot from the shelves. It also spurred the company to conduct a thorough review of manufacturing procedures and quality control.
This infant botulism case was investigated in a collaborative effort by two independent laboratories utilizing microbial culture and molecular techniques for identification of the organism causing the incident and its suspected relation to infant formula. A blinded investigation into the genetic relatedness of the strains from the infant, the opened container and the closed container, as well as historical isolates, was performed using two typing methods: (i) amplified fragment length polymorphism (AFLP) performed at the PHLS in London (unpublished data) and (ii) pulsed-field gel electrophoresis (PFGE) at the University of Wisconsin—Madison in the United States. This communication presents analyses of the various C. botulinum strains isolated during the case by PHLS and from retrospective clinical and food isolates using PFGE. Finally, potential measures to investigate and prevent future cases of infant botulism are described.
MATERIALS AND METHODS
Bacterial strains. Twelve lyophilized samples (no. 1 to 12) received blinded from the PHLS were aseptically rehydrated in sterile TPGY media (50 g/liter trypticase peptone, 5 g/liter Bacto peptone, 4 g/liter D-glucose, 20 g/liter yeast extract, 1 g/liter cysteine-HCl, pH 7.4). The TPGY medium was sparged with nitrogen gas prior to autoclaving, and resazurin (2 mg per liter) was used to ascertain anaerobic conditions. These cultures were then grown in TPGY broth and in TPGY supplemented with cooked meat particles (75 g/liter) (TPGY/CM). Cultures were grown at both 30° and 37°C. Since all cultures were determined to be proteolytic, they were grown at 37°C in subsequent experiments. All bacterial media components were purchased from Becton Dickinson Microbiology Systems, Becton Dickinson and Company, Sparks, MD.
Cultural properties of Clostridium botulinum isolates. Bacterial cultures were examined microscopically to evaluate their purity. All cultures were also grown on TPGY agar plates and incubated aerobically and anaerobically in an anaerobic chamber (with a gas mix containing 80% nitrogen-10% carbon dioxide-10% hydrogen) to determine their purity. Mixed cultures (samples 1, 6, 10, 11, and 12) were heat shocked and/or ethanol treated and then plated on Clostridium botulinum isolation media (CBI; 40 g/liter trypticase peptone, 5 g/liter Na2HPO4, 2 g/liter NaCl, 2 g/liter D-glucose, 5 g/liter yeast extract, 20 g/liter agar, 100 ml/liter egg yolk suspension, 50% 250 mg/liter cycloserine, 76 mg/liter sulfamethoxazole, and 4 mg/liter trimethoprim) in an attempt to isolate pure cultures. Colonies that were determined to contain rod-shaped bacteria were reisolated from the anaerobic plates and grown in TPGY media for further analyses. Stock cultures were prepared from all purified strains. Proteolytic properties of the pure cultures were determined by examination of degradation of cooked meat particles during growth. After 5 days of growth in TPGY the culture pH was measured.
Botulinum neurotoxin A or B detection and serotyping. Neurotoxin titers in culture supernatants were determined using an intravenous injection method as previously described (32). Culture supernatants obtained from 5 days of growth in TPGY were diluted 1:2 in gel-phosphate buffer (30 mM sodium phosphate buffer, pH 6.3, 0.2% gelatin). One-hundred microliters of each sample were injected into the lateral tail veins of five mice (female ICR mice, 18 to 22 g [Sprague Dawley, Madison, WI]). The animals were observed for characteristic signs of botulism, and the time to death was recorded. The average time to death (ATTD) for 5 animals was converted to intraperitoneal (i.p.) 50% lethal dose (LD50)/ml using the formula i.p. LD50/ml = 6.9 x 106 – 3.4 x 105(ATTD) + 5.7 x 103(ATTD)2 – 31.9(ATTD)3. This assay has been demonstrated in our laboratory to correlate with the mouse i.p. bioassay (32).
Toxin serotype was determined by enzyme-linked immunosorbent assay (ELISA) using cell-free culture supernatants. Cultures were tested with C. botulinum type A and type B antitoxins. Five-hundred microliters of culture samples were centrifuged for 5 min at 5,000 x g to pellet cells, and the supernatants were used for assays of neurotoxin protein. Ninety-six-well microtiter plates (Corning International, Corning, NY) were coated overnight at 4°C with 100 μl of a 5-μg/ml solution of antibodies raised against highly purified BoNT/A or BoNT/B toxoid. Antibodies were purified by affinity chromatography as previously described (22), diluted in 100 mM sodium carbonate buffer, pH 9.6, and used for coating plates. Plates were washed with TBST (10 mM Tris-HCl, 0.1% bovine serum albumin [BSA; Sigma], 150 mM NaCl, 0.05% Tween 20, pH 7.4) to remove unbound capture antibody, and then treated with 300 μl of blocking solution (5% vitamin-free casein [Research Organics, Inc., Cleveland, OH] in 10 mM Tris-HCl, 0.9% NaCl, 6 mM NaOH, pH 7.6) for 1 h at room temperature. Blocking solution was removed, 100 μl of serial dilutions of culture supernatants and a botulinum neurotoxin type A and B complex standard in blocking solution were added to the microtiter plate wells, and the plates were incubated for 2 h at room temperature. Plates were washed with TBST, 100 μl of a 2-μg/ml solution of biotinylated botulinum neurotoxin type A or B affinity purified immunoglobulin G (IgG) in blocking buffer was added to the microtiter plate wells, and the plates were incubated for 1 h at room temperature. Plates were washed with TBST to remove unbound secondary antibody, 100 μl of a 1:10,000 dilution of an avidin alkaline phosphatase conjugate (ExtrAvidin; Sigma) in blocking buffer was added to the microtiter plate wells, and the plates were incubated for 1 h at room temperature. Plates were again washed with TBST and developed by addition of Gibco BRL ELISA Amplification System reagents (Life Technologies, Rockville, MD) according to the manufacturer's instructions. Plates were read at 490 nm in a SpectraFluor plate reader (Tecan U.S., Inc., Durham, NC). All analyses were conducted in triplicate.
PFGE. PFGE was performed to compare the genomic DNA patterns of the 12 PHLS isolates and well-characterized C. botulinum type A1 strains 62A, ATCC 3502, Hall A-hyper; type A2 strains Kyoto F and FRI-H1A2; and type B strains Okra B (proteolytic) and 17B (nonproteolytic) from our laboratory collection. Of particular importance to this study, strains Kyoto F and FRI-H1A2 are A2 isolates associated with infant botulism cases in Japan and the United States, respectively.
Bacterial cultures were grown anaerobically in Bellco anaerobic tubes in 10 ml of TPGY media until the optical density at 600 nm (OD600) reached approximately 0.6. Bacterial cells were fixed with formaldehyde to inhibit DNase activity (21). One milliliter of formaldehyde solution was injected through the septum into the 10-ml bacterial cultures. The contents were mixed, and tubes were incubated on ice for 30 min. The fixed bacterial cultures were then transferred to 15-ml centrifuge tubes, and cells were precipitated by centrifugation. Cell pellets were washed twice with 10 ml of 0.85% NaCl solution, resuspended in 1.5 ml of solution containing 10 mM Tris-HCl, pH 8.0, 1 M NaCl, and 100 mM EDTA and was incubated for 10 min in a 50°C water bath. Equal volumes of the cell suspension and 1.5% InCert agarose (BioWhitaker Molecular Applications, Rockland, ME) were mixed, and the cell-agarose mixtures were loaded into plug-forming molds. The gel molds were chilled in a refrigerator for 15 min, and agarose plugs were removed and incubated in 4 volumes of lysis buffer (10 mM Tris-HCl, pH 8.0, 1 M NaCl, 100 mM EDTA, 0.5% Brij 58, 0.2% deoxycholate, 0.5% sarcosyl, 100 μg/ml RNase A, 5 mg/ml lysozyme, 40 U/ml mutanolysin) at 37°C for 12 h. The lysis buffer was removed, and the plugs were rinsed with 50 mM EDTA solution to remove lysozyme, RNase and mutanolysin. Two volumes of PK buffer (10 mM Tris-HCl, pH 8.0, 0.5 M EDTA, 1% sarcosyl, and 1 mg/ml Proteinase K) were added, and the agarose plugs incubated for 48 h with gentle agitation in a 50°C water bath. Following Proteinase K treatment, the plugs were washed with TE buffer and incubated for 2 h at 37°C with shaking in TE buffer containing 1 mM phenylmethylsulfonyl fluoride (PMSF) to inactivate Proteinase K. The agarose plugs were washed three times for 2 h each with 10 volumes of TE buffer to remove PMSF. Restriction digests of PFGE plugs were performed with four restriction endonucleases (NruI, SacII, SmaI, and XhoI), and samples were separated using a clamped homogenous electric field system (CHEF-DRII; Bio-Rad, Hercules, CA) following the manufacturer's instructions.
RESULTS
Culture purification. Twelve lyophilized culture samples (no. 1 to 12) received blinded from the PHLS were aseptically rehydrated with sterile TPGY and grown in TPGY/CM medium. Cultures were plated on TPGY and incubated aerobically and anaerobically to assess their purity. Cultures grown in liquid media and on plates were examined microscopically to also evaluate purity. Five of the cultures, nos. 1, 6, 10, 11, and 12, were mixed cultures and contained rods and cocci. In order to isolate pure strains of these bacterial cultures, two approaches were used. The cultures were heat shocked or treated with 50% ethanol (23). The cultures were plated on CBI plates, on which C. botulinum colonies form a characteristic opaque zone surrounding them, indicating lipase activity. Colonies that contained rod-shaped bacteria were isolated from the anaerobic plates and grown in TPGY media for further analyses. Stock cultures were prepared from all purified strains.
C. botulinum growth characteristics. The C. botulinum strains were analyzed for phenotypic properties. These properties are summarized in Table 1. The purified cultures all had a rod morphology, which varied in length and width dependent on the strain. Certain cultures produced high spore titers, while others sporulated infrequently. The spores varied in size and morphology depending on the strain. After 5 days of growth in TPGY, the final pH values ranged from 5.9 to 6.6 due to the production of organic acids during fermentation. The cultures were all proteolytic but varied markedly in proteolysis, as evaluated by digestion of meat particles in TPGY/CM. All strains showed lipase activity on CBI. These results indicate that there was considerable variation among the strains in cultural properties.
Botulinal neurotoxin production and typing. The strains were evaluated for botulinum neurotoxin production and toxin serotype by ELISA, mouse bioassay, and neutralization of mouse toxicity by serotype-specific antitoxins (23). Except for culture 7 (nontoxinogenic control), all the strains produced either A or B botulinum neurotoxin. None of the toxic supernatants reacted with both A and B antitoxins. Biological toxicity was determined using the mouse LD50 bioassay. Cultures 2 and 5 produced type A neurotoxin, with titers of 1.3 x 105 and 5.6 x 105 mouse LD50/ml, respectively. The remaining nine toxic cultures produced type B neurotoxin, and the titers varied from 3 x 103 to 2.6 x 105 mouse LD50/ml.
PFGE analyses. PFGE was used to assess genetic relatedness among the PHLS isolates and well-characterized C. botulinum type A1 strains 62A, ATCC 3502, Hall A-hyper; type A2 strains Kyoto F and FRI-H1A2; and type B strains Okra B (proteolytic) and 17B (nonproteolytic) from our laboratory collection. Restriction digests of PFGE plugs were performed with four restriction endonucleases: NruI, SacII, SmaI, and XhoI. The digests were separated using a clamped homogenous electric field system. PFGE revealed that some of the isolates showed banding similarities, and we were able to assign them to arbitrary groups (Table 1, Fig. 1). PFGE group 1 included the PHLS isolates 2 and 5, which revealed identical banding patterns in all four digests. Interestingly, these PFGE patterns were identical to the patterns from the infant type A2 botulism-associated strains Kyoto F and FRI-H1A2 (with minor differences in NruI banding pattern). PFGE group 2 included PHLS isolates 1, 6, and 10, which yielded similar banding patterns. The SacII and NruI digestion patterns were identical for these three isolates, while isolates 1 and 10 showed identical SmaI digestion patterns and isolate 6 had a slightly different pattern. All three isolates yielded very similar patterns by XhoI, with minor differences in some band sizes. The banding pattern of these isolates (1, 6, and 10) did not reveal similarity to any of the type A or B C. botulinum strains analyzed. PFGE group 3 included PHLS isolates 4 and 8, which had identical banding patterns with all four restriction endonucleases. The patterns did not show any similarity to other C. botulinum strains examined. The PHLS isolates 3, 7, 9, 11, and 12 exhibited unique PFGE patterns with no similarity to the other PHLS isolates or C. botulinum type A and B strains from our collection used for comparison.
After these cultural, toxin, and PFGE analyses were completed, the strain origins were unblinded (Table 2). It was revealed that strains 1, 6, 10, 11, and 12 were isolated during the investigation of the infant botulism case in 2001. All of these strains produced type B botulinum neurotoxin. Isolates 1 (from the baby's feces) and isolates 6 and 10 (from opened containers of baby milk) had the same PFGE pattern (Group 2). However, isolate 6 differed from isolates 1 and 10 in cell morphology, spore titers, and toxin production in culture. Isolate 11 (from the baby's feces) produced type B toxin but had a unique PFGE pattern which did not show similarity to the other strains examined. Strain 12 (presumptively isolated from an unopened infant formula) also produced low levels of type B toxin, was strongly proteolytic in comparison to strains 1, 6, 10, and 11, and importantly had a unique PFGE pattern with no similarities to the other strains. Interestingly, isolates 2 and 5 (from different United Kingdom infant botulism cases), Kyoto F associated with an infant botulism case in Japan, and FRI-HIA2 from honey associated with an infant botulism case in the United States all produced type A toxin and had identical PFGE patterns. These results indicate that cultural, toxin, and PFGE analyses can clearly distinguish groups of C. botulinum, even from widely different geographic regions of the world.
DISCUSSION
Since the discovery of infant botulism in 1976 (34, 38), infant botulism has been detected very rarely in the United Kingdom. This study describes an investigation of the sixth case that occurred in 2001. Infant botulism is caused by the ingestion of spores of C. botulinum followed by gut colonization by the organism and absorption of the toxin into the blood (19, 34, 38). Epidemiological risk factors for infants include rural residence where agricultural practices are carried out and urban areas where building or earth moving for road construction are frequent (10). Ingestion of honey (10) or possibly corn syrup (36) are also risk factors. Although the predominance of cases of infant botulism recorded to date have occurred in the state of California, infant botulism has also been reported in other regions of the United States (8, 35), Italy (20), Denmark (12), Japan (30), Venezuela (46), Australia (33), and various other countries (10). Some reports suggest a link with sudden infant death syndrome (13) but others do not (15), and the possible relationship of sudden infant death syndrome and infant botulism remains controversial. Although infant formula feeding has occasionally been implicated in the transmission of infections such as salmonellosis and Enterobacter sakazakii (9, 11, 18), it has not been epidemiologically linked to infant botulism (10). Large numbers of samples of infant formula were tested for C. botulinum spores in the 1980s, and no evidence of contamination with spores of C. botulinum was found (S. Arnon, personal communication).
This investigation is unique in that independent studies were performed by two different laboratories, one in the United Kingdom at the PHLS and the other in the United States at the University of Wisconsin—Madison. The two laboratories used distinct genetic approaches to characterize C. botulinum strains. The United Kingdom study, which will be published elsewhere (unpublished data), used AFLP to analyze strains isolated during the 2001 infant botulism case. Our study used cultural properties, toxicity, and PFGE analyses to characterize the 2001 isolates as well as other PHLS isolates, some of which were associated with previous food and infant botulism incidents. These strains were received blinded in our laboratory from the PHLS. In addition to the PHLS cultures, several C. botulinum type A and B strains from our collection, including the A2 isolates Kyoto F and FRI-H1A2 previously associated with infant botulism, were included in the PFGE analyses.
Various methods have been used for molecular typing of C. botulinum, including ribotyping, PCR, repetitive element sequence-based PCR (rep-PCR), random amplified polymorphic DNA analysis (RAPD), amplified fragment length polymorphisms (AFLP), and PFGE (14, 25, 26, 27, 28, 40, 42). In our laboratory, we have found PFGE to be a rapid and accurate method of determining genomic relationships among groups and strains of C. botulinum (28, 31). AFLP and ribotyping are less time-consuming and are useful for typing of strains for which satisfactory high-molecular-weight chromosomal DNA preparations and restriction fragment digests cannot be obtained. In our experience as well as in certain other laboratories, PFGE analysis of C. botulinum strains has been the most discriminating technique (24, 25, 28). Computer programs have been developed to accurately analyze PFGE data and for molecular typing and disease surveillance for other bacterial pathogens in the PulseNet system (44).
In this study, the two isolates from the infant's fecal specimens differed in PFGE patterns. One isolate from the opened container matched an isolate from the infant's fecal specimen, and one isolate from the unopened container had a unique PFGE pattern. The fact that the strain presumptively isolated from the unopened container differed from those isolated from infant feces and the opened container of infant formula indicates that the source from which the infant was infected is not proven. It is not clear how the opened can was contaminated and the child was exposed, particularly since the home and environs near to the home were not sampled for C. botulinum during the investigation.
Since the commercial source of the baby formula was initially suspected to cause the infant botulism case, a large recall was undertaken on the basis that a single strain was presumptively isolated from an unopened container, suggesting an apparent link to a clinical case. Since the batch of this infant formula consisted of 122,388 containers manufactured in October 1998 and fed to around 30,000 infants within the United Kingdom through the 3-year shelf life, if contamination of the batch had been widespread it is likely that numerous cases of infant botulism would have occurred in 1998, but this was not the case. Although proprietary, the general ingredients of the infant formula consisted of reduced mineral whey, vegetable oils, skimmed-milk powder, and lactose. Due to the ubiquity of bacterial spores, these products could be contaminated. One of the processing steps was heat treatment at 90 to 95°C for 20 to 25 s, which would not be sufficient to inactivate heat-resistant bacterial spores. Therefore, it is possible that C. botulinum spores could survive the manufacturing process of the infant formula, but this would require a separate study and sampling of a statistically meaningful number of containers to show their presence in infant formula. Although the infant botulism case did not appear to be associated with the commercial "unopened" source of infant formula, nonetheless it stimulated the industry to consider various precautionary measures and actions to prevent infant botulism from powdered formula. Most importantly, measures should be taken to ensure that viable C. botulinum spores are not present in baby formula.
This study describes an investigation of the association of C. botulinum with an infant botulism case. The investigation utilized various methods, including microbiological analyses, toxin testing and typing, and molecular typing for epidemiological characterization and strain identification of isolates associated with infant botulism. It was evident from this study that typing methods can provide new insights into causation and transmission of infectious diseases, including infant botulism.
ACKNOWLEDGMENTS
We thank Eric Bolton, Moira Brett, Jim McLauchlin, and their colleagues for sending the strains and for the collaborative study.
This study was supported by sponsors of the Food Research Institute and the College of Agriculture and Life Sciences, University of Wisconsin, Madison.
REFERENCES
Anonymous. 1978. Communicable Disease Report no. 5.
Anonymous. 1987. Communicable Disease Report no. 38.
Anonymous. 1993. A case of infant botulism. Communicable Disease Report no. 29.
Anonymous. 1994. A case of infant botulism. Communicable Disease Report no. 12.
Anonymous. 1994. Quarterly communicable disease review January to March 1994. PHLS Communicable Disease Surveillance Centre. J. Public Health Med. 16:357-364.
Anonymous. 2001. A case of infant botulism. Communicable Disease Report 28 June.
Anonymous. 2001. Infant botulism: update. Communicable Disease Report 16 August.
Anonymous. 2003. Infant botulism-New York City, 2001-2002. Morb. Mortal. Wkly. Rep. 52:21-24.
Anonymous. 2004. Joint FAO/W.H.O. Workshop on Enterobacter sakazakii and other microorganisms in powdered infant formula, Geneva, Switzerland, 2 to 5 February 2004. [Online.] http://www.who.int/foodsafety/publications/micro/feb2004/en/.
Arnon, S. S. 2004. Infant botulism, p. 1758-1766. In R. D. Feigen and J. D. Cherry (ed.), Textbook of pediatric infectious diseases, 5th ed. W. B. Saunders, Philadelphia, Pa.
Baker, R. D. 2002. Infant formula safety. Pediatrics 110:833-835.
Balsev, T., E. Ostergaard, I. K. Madsen, and D. A. Wandall. 1997. Infant botulism. The first culture-confirmed Danish case. Neuropediatrics 28:287-288.
Bohnel, H., S. Behrens, P. Loch, K. Lube, and F. Gessler. 2001. Is there a link between infant botulism and sudden infant death Bacteriological results obtained in central Germany. Eur. J. Pediatr. 160:623-628.
Broda, D. M., J. A. Boerema, and R. G. Bell. 1998. A PCR survey of psychrotrophic Clostridium botulinum-like isolates for the presence of BoNT genes. Lett. Appl. Microbiol. 27:219-223.
Byard, R. W., L. Moore, A. J. Bourne, A. J. Lawrence, and P. N. Goldwater. 1992. Clostridium botulinum and sudden infant death syndrome: a ten-year prospective study. J. Paediatr. Child Health 28:156-157.
Communicable Disease Surveilllance Centre. 1993. A case of infant botulism. Commun. Dis. Rep. CDR Wkly. 3:129.
Communicable Disease Surveillance Centre. 1994. A case of infant botulism. Commun. Dis. Rep. CDR Wkly. 4:53.
Estrada, B. 2002. Infections associated with powdered infant formula. Infect. Med. 19:350.
Faucheux, R. C., A. K. Shetty, and G. S. Cowan. 1997. Infant botulism. Clin. Pediatr. 36:591-593.
Fenicia, L., L. Da Dalt, F. Anniballi, G. Franciosa, S. Zanconato, and P. Aureli. 2003. A case of infant botulism due to neurotoxigenic Clostridium butyricum type E associated with Clostridium difficile colitis. Eur. J. Clin. Microbiol. Infect. Dis. 21:736-738.
Gibson, J. R., K. Sutherland, and R. J. Owen. 1994. Inhibition of DNAse activity in PFGE analysis of DNA from Campylobacter jejuni. Lett. Appl. Microbiol. 19:357-358.
Goodnough, M. C., B. Hammer, H. Sugiyama, and E. A. Johnson. 1993. Colony immunoblot assay of botulinal toxin. Appl. Environ. Microbiol. 59:2339-2342.
Hatheway, C. L., and E. A. Johnson. 1998. Clostridium: the spore-bearing anaerobes, p. 731-782. In L. Collier, A. Balows, and M. Sussman (ed.), Topley and Wilson's microbiology and microbial infections, 9th ed., vol 2: systematic bacteriology. Arnold, London, United Kingdom.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1998. Genomic analysis of Clostridium botulinum group II by pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 64:703-708.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1998. Prevalence of Clostridium botulinum in Finnish trout farms: pulsed-field gel electrophoresis typing reveals extensive genetic diversity among type E isolates. Appl. Environ. Microbiol. 64:4161-4167.
Hielm, S., J. Bjrkroth, E. Hyyti, and H. Korkeala. 1999. Ribotyping as an identification tool for Clostridium botulinum strains causing human botulism. Int. J. Food Microbiol. 47:121-131.
Hyyti, E., J. Bjrkroth, S. Hielm, and H. Korkeala. 1999. Characterization of Clostridium botulinum group I and II by randomly amplified polymorphic DNA analysis and repetitive element sequence-based PCR. Int. J. Food Microbiol. 48:179-189.
Johnson, E. A., and M. Bradshaw. 2001. Clostridium botulinum and its neurotoxins: a metabolic and cellular perspective. Toxicon 39:1703-1722.
Jones, S., Z. Huma, C. A. Haugh, Y. Young, F. Starer, and L. Sinclair. 1990. Central nervous system involvement in infantile botulism. Lancet i:228.
Kakinuma, H., H. Maruyama, H. Takahashi, K. Yamakawa, and S. Nakamura. 1996. The first case of type B infant botulism in Japan. Acta Paediatr. Japonica 38:541-543.
Lin, W.-J., and E. A. Johnson. 1995. Genome analysis of Clostridium botulinum by a pulsed-field gel electrophoresis. Appl. Environ. Microbiol. 61:4441-4447.
Malizio, C. J., M. C. Goodnough, and E. A. Johnson. 2000. Purification of Clostridium botulinum type A neurotoxin. Methods Mol. Biol. 145:27-39.
May, M., M. Coulthard, G. Delbridge, J. McEniery, C. Burke, and M. Nissen. 2002. Difficulties in the diagnosis and management of infant botulism. J. Paediatr. Child Health 38:425-426.
Midura, T. F., and S. S. Arnon. 1976. Infant botulism. Identification of Clostridium botulinum and its toxins in faeces. Lancet ii:934-936.
Morris, J. G., Jr., J. D. Snyder, R. Wilson, and R. A. Feldman. 1983. Infant botulism in the United States: an epidemiology study of cases occurring outside of California. Am. J. Public Health 73:1385-1388.
Olsen, S. J., and D. L. Swerdlow. 2000. Risk of infant botulism from corn syrup. Pediatr. Infect. Dis. J. 19:584-585.
O'Mahoney, M., E. Mitchell, R. J. Gilbert, D. N. Hutchinson, N. T. Begg, J. C. Rodhouse, and J. E. Morris. 1990. An outbreak of foodborne botulism associated with contaminated hazelnut yoghurt. Epidemiol. Infect. 104:389-395.
Picket, J., B. Berg, E. Chaplin, and M. Brunstetter-Shafer. 1976. Syndrome of Clostridium botulinum in infancy: a clinical and electrophysiologic study. N. Engl. J. Med. 295:770-772.
Roberts, E., J. M. Wales, M. M. Brett, and P. Bradding. 1998. Vomiting with cranial nerve palsies. Lancet 352:1674.
Santos-Buelga, J. A., M. D. Collins, and A. K. East. 1998. Characterization of the genes encoding the botulinum neurotoxin complex in a strain of Clostridium botulinum producing type B and F neurotoxins. Curr. Microbiol. 37:312-318.
Sinclair, L., S. T. Jones, Y. Young, and C. A. Haugh. 1990. Infant botulism. Communicable Disease Report no. 05.
Skinner, G. E., S. M. Gendel, G. A. Fingerhut, H. A. Solomon, and J. Ulaszek. 2000. Differentiation between types and strains of Clostridium botulinum by ribotyping. J. Food Protection. 63:1374-1382.
Smith, G. E., F. Hinde, D. Westmoreland, P. R. Berry, and R. J. Gilbert. 1989. Infantile botulism. Arch. Dis. Childhood 64:871-872.
Swaminathan, B., T. J. Barrett, S. B. Hunter, R. V. Tauxe, and the CDC PulseNet Task Force. 2001. PulseNet: the molecular subtyping network for foodborne bacterial disease surveilance, United States. Emerg. Infect. Dis. 7:382-389.
Turner, H. D., E. M. Brett, R. J. Gilbert, A. C. Ghosh, and H. J. Leibeschuetz. 1978. Infant botulism in England. Lancet i:1277-1278.
Urdaneta-Carruyo, E., A. Suranyi, and M. Milano. 2000. Infantile botulism: clinical and laboratory observations of a rare neuroparalytic disease. J. Paediatr. Child Health 36:193-195.(Eric A. Johnson, William )