This study is to my knowledge the rst attempt at understanding the function of primary sensory cortices in social contexts and their representation of stimuli with social relevance on the cellular level. Behavioral monitoring at dierent timescales and the recording of neuronal activity in a considerable population of single units from barrel cortex were the mainstays of this study. In addition, I gathered data on the estrus state of female rats. This multi-faceted approach allowed a wider view on the relation between behavior and physiology, and this, critically, in freely-moving animals which performed a complex and natural behavior. The near-naturalistic paradigm, which avoided any extrinsic motivations through rewards, and where the animals were free to interact as often as they wanted, for as long as they wanted, and with all whiskers present, is a distinctive feature of this study. At the same time, it had the consequence that experimental control was limited, and many factors made a characterization of sensory input on temporal and spatial scales common in sensory neuroscience impossible.
I addressed this, among others, by a transformation of interactions into a space centered on the subject rat nose, but the main limitation remains that the precise sequence and timing of whisker touches during each interaction remains unknown.
On the physiological level, I have taken into consideration the dierences which might
11It has been questioned, whether barbering really is an expression of dominance (Sarna et al., 2000), but whatever its cause, it is a behavior repeatedly reported in the literature (Sarna et al., 2000; Kalue et al., 2006;
Nicholson et al., 2009), and thus seems a normal constituent of mouse social behavior, at least in laboratory
arise in dierent layers and dierent cell types, as far as the methodology allowed. How-ever, the assignments to layers and cell types are partly incomplete or simplifying, and other functionally relevant distinctions like the dierences between barrels and septa could not be considered. The short overview of barrel cortex connectivity given in 1.3 serves to further highlight its complexity and the care that should be exerted at interpreting results from even seemingly large neuronal populations.
That said, the results still strongly support the view that neurons in barrel cortex - and, by extension, in primary sensory cortices in general - can show responses which extend beyond simple stimulus properties, as typically ascribed to them. As any study, it leaves many questions open, and in particular it is largely restricted to the correlational rather than the mechanistic level. Nevertheless, it suggests the study of complex social as well as non-social stimuli and the associated responses in primary sensory cortices as a eld, which can yield new insights on sensory computation in real-world settings, and thus, ultimately, a better understanding of how the brain performs the complex tasks of real life, a question which we are just beginning to address.
5 Summary/Zusammenfassung
5.1 English
Rats use their sti facial hairs (whiskers) for somatosensation, and the pathway from the whiskers to the primary somatosensory cortex (barrel cortex) is well known. Rats also show diverse social behaviors, one of which is touch of conspecics with their whiskers. Nothing is known, however, about the representation of these social touch signals in the brain, and whether it diers from the touch of non-social stimuli. Thus, the present study aimed at char-acterizing the neuronal representation of social touch signals in barrel cortex and comparing them with other somatosensory stimulation.
Using extracellular single-cell recordings in freely-moving rats, I show that a large part of neurons in barrel cortex modulates their activity during social touch. Here, the reported fraction of 40% of neurons is a lower-bound estimate. Responses were typically excitatory, but there were also examples of inhibition. The pattern of ring rates during interactions diered between cortical layers, with layer 5B being both the most active and the most strongly modulated. Layer 2/3 neurons had very low ring rates even during social touch, conrming earlier reports from patch-clamp recordings employing non-social stimuli. Regular- and fast-spikers could be distinguished based on their spike shapes, with fast-fast-spikers showing more reliable but less strongly modulated responses.
In behavioral experiments rats preferred interactions with alive conspecics over inani-mate stimuli. Whisking strategies also diered substantially. Thus, inaniinani-mate stimuli were whisked at with regular palpating movements from protracted set angles. In contrast, social interactions were marked by irregular, small-amplitude whisking. Neuronal responses were also dierent during touch of alive rats and inanimate stimuli. Thus, objects elicited slightly but consistently weaker responses than alive rats. There was also a trend towards weaker responses when touching a stued rat.
I observed sex-specic dierences in neuronal responses. The most prominent one was stronger modulation by social touch in regular-spikers recorded from males. This dierence could not be explained by behavioral measures as whisking power and set angle, possibly indicating a neural origin of this dierence. Regular-spikers from females red much more weakly when the female was in the estrus phase of the estrus cycle. Furthermore, female-female interactions were associated with neuronal inhibition, when the subject female-female was in estrus. The latter eect was observed in very low-ring neurons only.
In summary, this is the rst study which investigated social signals in a primary sensory area of freely-moving animals at the cellular level. It suggests that representations in sensory cortices might be less stimulus-driven and more top-down modulated than previously thought.