Pyrrolizidine alkaloids in food: downstream contamination in the food chain caused by honey and pollen.

In recent years, there has been a steadily growing number of published data on pyrrolizidine alkaloids (PAs) in honey and pollen. This raises the question whether honey and/or pollen used as ingredients in food processing might provoke a downstream contamination in the food chain. Here we addressed two different facets in connection with PAs in honey and pollen. First, we analysed the PA content of several food types such as mead (n = 19), candy (n = 10), fennel honey (n = 9), soft drinks (n = 5), power bars and cereals (n = 7), jelly babies (n = 3), baby food (n = 3), supplements (n = 3) and fruit sauce (n = 1) that contained honey as an ingredient in the range of 5% to approximately 37%. Eight out of 60 retail samples were tested as being PA-positive, corresponding to 13%. Positive samples were found in mead, candy and fennel honey, and the average PA content was calculated to be 0.10 µg g(-1) retronecine equivalents (ranging from 0.010 to 0.484 µg g(-1)). Furthermore, we investigated the question whether and how PAs from PA pollen are transferred from pollen into honey. We conducted model experiments with floral pollen of Senecio vernalis and PA free honey and tested the influence of the quantity of PA pollen, contact time and a simulated honey filtration on the final PA content of honey. It could clearly be demonstrated that the PA content of honey was directly proportional to the amount of PA pollen in honey and that the transfer of PAs from pollen to honey was a rather quick process. Consequently, PA pollen represents a major source for the observed PA content in honey. On the other hand, a good portion remains in the pollen. This fraction is not detected by the common analytical methods, but will be ingested, and it represents an unknown amount of 'hidden' PAs. In addition, the results showed that a technically and legally possible honey filtration (including the removal of all pollen) would not be an option to reduce the PA level of the final product significantly.

Immunomodulatory effects in Litomosoides sigmodontis-infected Mastomys coucha.

The levels of parasite-specific IgG1 and IgG2 antibodies and mitogen-induced and parasite-specific proliferative T-cell responses were determined in Litomosoides sigmodontis-infected Mastomys coucha throughout an observation period of 400 days post infection (p.i.). These were compared with the respective reactions in animals that had been immunized with intrauterine stages/microfilariae of the parasite and in animals that had been challenged after immunization as determined at up to 60 days after challenge. IgG1 antibodies to adult antigen developed early and reached a plateau at 120 days p.i., whereas IgG2 antibodies were not found before day 60 p.i., increased with rising parasitemia, reached a plateau at 200 days p.i., and, in some animals, even became the predominant IgG subclass. Proliferative responses of spleen lymphocytes to concanavalin A (Con A) and lipopolysaccharides (LPS), but not Con-A-induced interleukin 2 (IL-2) production, were found to be suppressed in infected animals during patency as compared with uninfected controls. Spleen cells of infected animals showed a weak proliferative reaction to male antigen but were unresponsive to female and microfilarial antigen during prepatency and early patency. Subsequently, when microfilaremia decreased (200 days p.i.), continuously increasing responses to all antigens were observed. Immunized M. coucha developed specific IgG1 and IgG2 antibodies, and their spleen cells showed strong proliferative responses to the three L. sigmodontis antigens. Challenge infections down-regulated the proliferative responses of spleen cells to filarial antigens as early as during the prepatent phase of the challenge infection but supported existing IgG1 and IgG2 responses.