Whereas the novel bacteria diet per se had no Madecassic-acid effect on anxiety- and depression-like endophenotypes of the animals, the stable isotope 15N repeatedly decreased depressionlike 8-O-Acetylharpagide behavior in high anxious animals, irrespective of the diet used. We are not aware of any other study showing that a stable isotope can affect behavioral characteristics of labeled animals. A possible explanation for this phenomenon could be altered enzymatic activities of 15N-labeled proteins due to increased chemical bond strength, which might affect pathways involved in the pathobiology of anxiety/depression. Indeed, Ditzen et al. have recently described altered enzyme kinetics likely contributing to the phenotype of HAB animals. Altogether, the phenomenon of alterations in depression-like behavior due to 15N labeling, with the anxiety-related behavior remaining unchanged, might allow us to carry out further analyses to detect critical, so far unstudied, mechanisms exclusively involved in the depression-like behavior of HAB mice. Additionally, these results clearly demonstrate the necessity to confirm the behavioral phenotype of a mouse model fed with a 15N diet. Metabolic labeling of non-mammalian organisms with 15N for comparing two proteomes by MS analysis is a valuable technique and routinely used in quantitative proteomic approaches. Labeling rats with a 15N-enriched blue-green algae diet has demonstrated that the method is applicable to rodents. However, studying psychopathologically relevant proteomic parameters requires the examination of any diet effect per se onto relevant behaviors. For 15N-labeled rats fed with a blue-green algaebased diet, no obvious behavioral changes were reported. However, no thorough phenotypic analysis of the animals was reported. Furthermore, the usage of a restricted feeding paradigm does a priori not allow the exclusion of an impact on the animals�� stress-related behaviors. This was solved by using blue-green algae powder incorporated in normal pellets fed ad libitum.To be able to correlate psychopathological behavioral parameters with protein levels as early as PND5 via ultrasound vocalization, it was necessary to use a feeding protocol, in which the pregnant dams were already fed with the respective diet.