Analytical validation of an Enzyme Immunoassay for the measurement of urinary oxytocin in dogs and wolves

Assessing changes in oxytocin (OT) levels in response to a variety of social stimuli has become of major interest in the field of behavioral endocrinology. OT is involved in the regulation of various aspects of social behavior such as tolerance, and the formation and maintenance of social bonds but also the regulation of stress. All of these aspects have been identified as potential targets of selection during the domestication process. Therefore, comparing the role of the oxytocinergic system in various aspects of dog and wolf social behavior, might help to understand whether this system was involved in the domestication process. Studies assessing OT levels in dogs and wolves have used invasively collected plasma and serum samples and non-invasively collected urine samples. However, when using an assay system on a new species a careful and complete validation of the method is of crucial importance, and to date no proper validation, to assess urinary OT levels in dogs and wolves, has been reported. We therefore conducted an analytical validation of an Enzyme Immunoassay (EIA) for the measurement of OT in urine of dogs and wolves, using a commercially available EIA. Stability tests revealed that OT levels degrade over time when stored at 4 °C, but are little affected by repeated thawing. In addition, our results indicate that the variance in OT levels is slightly lower when phosphoric acid is added following collection to prevent OT degradation. Long term storage tests revealed that urinary OT levels are least variable when stored as extracts in ethanol at −20 °C, rather than as unextracted urine samples. Validation results were acceptable with regard to parallelism, but values for accuracy and extraction efficiency were not meeting the standard criteria usually applied to steroid EIAs, especially when assessed for the lower range of the assay. The results of this study highlight the importance of an analytical assay validation, since even if validation parameters are not optimal, if published, they allow readers to estimate the relevance of studies using the validated method