Helsingin yliopisto


Helsingin yliopiston verkkojulkaisut

University of Helsinki, Helsinki 2006

Clostridium botulinum in honey production with respect to infant botulism

Mari Nevas

Doctoral dissertation, April 2006.
University of Helsinki, Faculty of Veterinary Medicine, Department of Food and Environmental Hygiene.

Infant botulism is a rare illness caused by spores of Clostridium botulinum, which are ubiquitous in nature and harmless for healthy adults. Due to the immaturity of infantile intestinal microflora, the spores carried into the intestinal canal may germinate and produce toxin in vivo. The course of the infant botulism can vary from a prolonged constipation to sudden death of an infant, but several child-dependent predisposing factors are known to influence susceptibility to infection. Honey was the only food item associated with infant botulism until 2005, when a suspected case of infant milk formula-derived infant botulism was described. This thesis is based on the previously recognized association between honey and infant botulism.

The autopsy specimens of a suddenly deceased 11-week-old boy were studied with multiplex polymerase chain reaction (PCR), and Clostridium botulinum group I type B was detected in the intestinal contents and in the vacuum cleaner dust of the patient's household. Genetic similarity of these isolates was demonstrated with pulsed-field gel electrophoresis (PFGE) typing and randomly amplified polymorphic DNA (RAPD) analysis, confirming that dust can act as a vehicle for infant botulism, resulting in sudden death. This was the first reported case of infant botulism in Finland.

To evaluate the prevalence of C. botulinum in honey, a test protocol for reliable PCR-based detection of spores was developed. The inhibiting effects of honey were overcome by using a supernatant filtration method of in the preparation of samples before enrichment and PCR.

PFGE analysis was optimized to study C. botulinum group I isolates, and applied to analyse the diversity of 55 strains of C. botulinum types A, AB, B and F. A total of nine restriction enzymes were tested, from which SacII, SmaI and XhoI were chosen for the cluster analysis of the 55 strains. Of these three enzymes, SacII was shown to be superior for analysis of C. botulinum group I isolates. C. botulinum group I was found to be heterogeneous, and in the majority of cases, PFGE enabled discrimination between individual strains of types A and B. However, certain clustering of type B strains was observed, and type F strains were located in a single cluster.

A total of 190 samples from honey sold in Finland were studied by PCR to determine whether C. botulinum spores of types A and B were present. Spores of C. botulinum were detected in 8 of the 114 Finnish (7%) and in 12 of the 76 imported honey samples (16%). In addition, 294 honey samples produced in Denmark, Norway and Sweden were studied using multiplex-PCR for the presence of C. botulinum type A, B, E and F spores. The prevalence of C. botulinum showed a significant variation between Denmark, Norway and Sweden, the proportions of positive samples in each country being 26%, 10% and 2%, respectively. Type B was most frequently detected. The 24 strains isolated were confirmed to be proteolytic and were analysed with PFGE using restriction enzyme SacII. Twelve different PFGE patterns were produced. Three clearly distinguishable clusters containing more than three strains were noted, one of which included strains isolated in both from Denmark and Norway.

Factors influencing C. botulinum contamination in the honey production environment were evaluated in a three-year survey. A total of 1168 samples from 100 apiaries and related facilities in Finland were analysed for the presence of C. botulinum types A, B, E and F using multiplex-PCR. Production methods and environmental factors were registered using a questionnaire and personal observation. C. botulinum was shown to be common throughout the whole honey production chain, with the type most frequently detected being group I type B. A PFGE analysis of 202 group I type B isolates suggested the existence of different genetic lineages. Only six PFGE profiles were observed, two of which predominated, covering 96% of the isolates. The high prevalence of C. botulinum in soil and in samples associated with beeswax suggests the accumulation of soil-derived botulinal spores in wax. In addition, based on the most significant bivariate Spearman's rank order correlations, production hygiene and extent of production were shown to have significant influences on the prevalence of botulinal spores in extracted honey. According to a logistic multivariable regression model, the presence of hand-washing facilities in the extraction room had a significant impact on decreasing the number of spores.

Based on the results it is concluded that the number of spores in honey may be diminished. However, due to the observed wide existence of C. botulinum spores in the environment, it is not possible to produce honey totally free from spores.

The title page of the publication

This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.

© University of Helsinki 2006

Last updated 28.03.2006

Yhteystiedot, Contact information E-thesis Helsingin yliopisto, University of Helsinki