We did not detect evidence of any known nicotine oxides with UPLC/ESI-QTOF-MS technology. We used third-instar larvae in contrast to the fifth-instar larvae used in the previous studies, which might differ physiologically from the first four growth-focused instars, as during the fifth instar, larvae prepare for pupation. Similar to the work of Self et al., we always fed the larvae immediately after eclosion on physiologically realistic concentrations of dietary nicotine, whereas Snyder et al. and Wink and Theile used the artificial diet-reared fifthinstar larvae, which had not been exposed to nicotine during their previous instars; these larvae were then abruptly exposed to high and physiologically unrealistic amounts of nicotine. We used freshly collected frass and hemolymph, whereas previous investigators used stored and dried frass in which nicotine oxidation may have occurred spontaneously. Lastly, we ruled out the possibility that nicotine was oxidized by the chemical constituents of diet, before it was ingested by the larvae, a possibility not examined in the previous work. None of the nicotine oxides were discharged from the hemolymph more rapidly than nicotine, clearly rejecting the long standing hypothesis that being more polar than nicotine, the oxides are more rapidly excreted. M. sexta larvae use pumps to purge toxic alkaloids to protect their nervous systems; these pumps function with structurally similar alkaloids like nicotine, morphine, atropine or even alkaline synthetic dyes. Previously, we proposed that MsCYP6B46 could be part of such a pump. The Waldbauer assays results were consistent with a ‘purge by excretion’ mechanism that functions with comparable efficiency between nicotine and its oxides. MsCYP6B46 transcripts were induced in larval midguts by the ingestion of all nicotine oxides, but no further oxidized products of nicotine, NNO or cotinine were found in the hemolymph or frass. Collectively, these results demonstrate either the absence of CYPmediated nicotine oxidation mechanisms in M. sexta, or that Manduca’s metabolism of nicotine is so rapid and efficient that no oxidative intermediates accumulate, or involves a novel pathway. All of the nicotine oxides had similar effects on larval growth as nicotine did and none deterred C. parallela, indicating that the oxidation of nicotine is not advantageous for M. sexta and hence cannot be considered either a detoxification or a co-option related process. The nicotine-elicited MsCYP6B46 clearly provides an ecological benefit for the larvae, due in large part to the volatility of nicotine. M. sexta larvae would clearly be at an ecological disadvantage if they converted nicotine to a less volatile and consequently, less spider-deterrent oxides. In contrast, for S. exigua, which is more sensitive to N. attenuata’s nicotine than M. sexta, oxidizing nicotine to cotinine, which is less lethal, represents true detoxification.

Many viruses, including MV, have developed strategies to alter antigen presentation or costimulatory properties of DCs in order to evade host immune responses, thereby causing immunosuppression and an increased susceptibility to opportunistic infections. Based on the present data, a similar mechanism is suggested for CDV-infection of dogs. The observed CDV-induced DC modulation might represent a mechanism to suppress protective immunity, which favors persistent infection in infected dogs. In agreement with this idea, it was reported that CDV-infected DC-like cells within lymphoid organs occur in advanced stages of canine distemper. Furthermore, the virtual lack of detectable cytopathogenic or lytic effects as determined by phase contrast microscopy, electron microscopy, and LDH assay might contribute to virus spread within the organism via circulating DCs as described for MV. Similarly, restricted viral infection of CNS cells together with prevention of cytolysis causes limited recognition by the immune surveillance, which favors viral persistence and transmission within the brain in canine distemper. In vitro experiments have demonstrated that virulent CDV strains exhibit conformational properties of the F protein, which limit cell-to-cell fusion activity and prevent cytopathogenicity, Thus, persistent infection with delayed production of infectious virus might be a consequence of reduced spread of less fusogenic viruses compared to cytolytic CDV strains. In addition, human cytomegalovirus, murine cytomegalovirus, and Epstein-Barr virus infect and manipulate DCs to circumvent cell death, which causes persistent infection. Besides this, disturbed function and prolonged survival of DC in canine distemper might compromise T cell maturation and selection, promoting the release of immature, potentially autoreactive cells as discussed for canine distemper. The present study shows, that CDV-infection down-regulates MHC class II and co-stimulatory molecules of DCs which could have the ability to impair T cell activation in affected dogs. Previous studies revealed an inhibition of antigen presenting cells in canine distemper as a consequence of reduced IL-1 production and increased prostaglandin E2 release. Thus, results of the present study further support the hypothesis that disturbed antigen presenting function contributes to reduced mitogen-induced lymphocyte proliferation observed in CDVinfected dogs. Moreover, diminished T helper cell function as a consequence of impaired antigen presentation in persistently infected dogs might lead to disturbed germinal center and plasma cell formation and reduced class switch from IgM to IgG in canine distemper. Dysregulation of antigen presenting properties is supposed to account for immunosuppression in human measles, by down-regulation of IL-12 by DCs, which leads to a failure to activate T cells. However, the precise role of DC infection in the pathogenesis of natural MVinfection.