Publication Date



David Bodznick


Neuroscience and Behavior




The principal electrosensory neurons (AENs) in the medulla of little skates extract relevant signals from a noisy electrosensory background
generated by the animal’s own behaviors (reafference). Research supports the existence of an adaptive filter in the cerebellar-like electrosensory nuclei of skates. Internal reference signals (e.g. motor commands and sensory feedback) in the molecular layer carry information pertaining to ongoing behaviors and supply AENs with cancellation signals (i.e. inverse predictions of electrosensory input), which when added to the actual electrosensory input, negate reafference. Thus the removal of predictable electrosensory input and preservation of novel signals in AENs are achieved.

In this study, predictable electrosensory input was experimentally
created by repetitive pairings of a peripheral excitatory electrosensory stimulus with either ventilatory motor commands or passive fin movements, both of which are known to be represented among the internal reference signals of the molecular layer. In both cases, the development of a cancellation signal that suppresses predictable electrosensory input was apparent in AEN responses: the response to the electrosensory stimulus declined during repeated pairings, and removal of the electrosensory stimulus revealed suppressed activity time-locked to the period of the suddenly absent electrosensory stimulus. Using this coupling paradigm, the functional properties of multiple cancellation signals within a single AEN were examined. First, multiple cancellation signals associated with distinct internal reference signals in the molecular layer are shown to exist within a single AEN and are modified independently, suggesting that separate cancellation signals for distinct behaviors may also exist, and can be stored and modified without affecting one another. Second, overlapping cancellation signals associated with distinct internal reference signals are additive in a subset of AENs, although not all AENs had combined cancellation signals that were significant or stronger than singular cancellation signals. These results support the idea that a single behavior may have several component cancellation signals that are associated with different aspects of the behavior. Together, the component cancellation signals may work in conjunction to more
effectively suppress reafference associated with that behavior. We propose that a functional role for multiple cancellation signals, shown in this study to be independent and additive, may be to provide sensory predictions that suppress dynamic, complex patterns of reafference that result from continuously, intermittently or concurrently performed behaviors.



© Copyright is owned by author of this document