The goal of the present study was to characterize the representations of complex social stimuli in rat primary somatosensory cortex, considering the subject and stimulus animal sex. The low-level hypothesis was that social touch would be associated with reliable response changes in neuronal activity. This was addressed by comparing ring rates during interactions with baseline ring rates, as well as by recording from whisker-trimmed animals. Starting from this, the questions to be addressed were: How do responses to social touch compare to the touch of other stimuli? Does the sex of the subject and/or the stimulus rat have an inuence on neuronal response patterns during interactions? Does the estrus state play a role in how female rats repesent social touch? And if dierences are found, are they due to dierential whisking behavior and can thus be explained by diering mechanical inputs? Or are there indications of additional inuences on neuronal activity? Although this was not the initial focus of the study, it was later attempted to go beyond pooling of all neurons, and towards investigating how the response patterns vary between neurons of dierent type and layer, as well as neurons with dierent RFs.
2 Methods
2.1 Subjects, paradigms, and stimuli
2.1.1 Experimental setup and denition of interaction events
During social interactions rats were placed on metal platforms within a Faraday cage (Fig.
2.1). The elevation of the platforms from the table surface was approximately 20 cm, sucient to avoid the rats from jumping down. In fact, even those rats which fell down once, were apparently too afraid to actively jump down afterwards. The platforms were 30 cm long and 25 cm wide, and were surrounded by walls of 35 cm height. Only rarely did rats jump onto the walls. The typical distance between the subject and stimulus platforms was 20 cm, although it was varied between 16 cm and 22 cm according to animal size. The platforms and walls were covered with soft black foam mats, which provided a dark background for video recordings and helped minimize mechanical and electrical artifacts when the subject rat touched a wall with the headstage. Experiments took place in near-complete darkness, as the room was darkened and the Faraday cage additionally covered with blackout curtains. However, a low level of stray light from computer screens and recording equipment could not be excluded.
In early experiments, two LEDs were attached to the headstage for head tracking, in which case the head was brightly illuminated. This was stopped later to avoid illumination and because the tether wire tended to twist around the LEDs. Illumination for observation and video-taping was provided by two infrared projectors (Abus TV6830, wavelength 880 nm), which were positioned ca. 60 cm over the setup.
~200 mm
~800 mm
250 mm
IR Lights High-speed
camera
Subject rat Stimulus rats
Gap Low-speed
camera
Figure 2.1: Behavioral setup. Depicted is the chooser paradigm, where two stimulus rats were presented in parallel. The scene is illuminated by infrared (IR) lights and recorded continuously with a low-speed camera, as well as with a high-speed camera during interacions.
Behavior was continuously recorded using a camera with either 25 or 30 frames per second with a visible light-blocking lter. These low-speed videos were used to determine the be-havioral events relative to which neuronal responses were analyzed: The time of rst whisker overlap was dened as the interaction start, and the time of last whisker overlap as the inter-action end. Most interinter-actions included very close approaches with the heads of the two rats touching each other, which was also recorded as a behavioral event. The end of head touch was not recorded, but typically occurred within 100 ms before the end of whisker overlap.
To be included in the analysis, an interaction had to last at least 400 ms, and be separated from the next interaction by minimally 400 ms. When the approach was strongly lateralized, which was only rarely the case, the whisker overlap was determined for the subject rat's whiskers contralateral to the implanted hemisphere. In these cases, rst whisker overlap and rst head touch times could be identical. To ensure sucient sampling, each stimulus rat had to be interacted with at least three times to enter the analysis based on single stimulus animals. The same restriction applied to the touch of objects, as well as the stued and the anesthetized rat.
Stimulus rats were presented either in a chooser setting or alone. In the chooser paradigm (see Fig. 2.1, as well as Wolfe et al., 2011), which was the majority of experiments, two stimulus rats were placed on the platform, separated by a wall. This separation wall was made from thick black styrofoam, covered by black cardboard, and extended beyond the edge of the stimulus platform, so that facial interactions between the stimulus rats were precluded.
Nevertheless, the stimulus rats showed behaviors indicative of strong interest in each other, namely biting the wall, sning at the wall's edges, and sometimes attempts to climb the walls. Most stimulus pairs were of mixed sex, but in a subset of experiments stimulus rats of the same sex were presented.
Recordings consisted of blocks of 5-10 minutes each. After each block, the light was turned on, the stimulus animals were removed and the mats exchanged to minimize olfactory cues.
There were 3-8 blocks in one recording session.
2.1.2 Subject and stimulus rats
Subject and stimulus animals were adult Wistar rats aged P50 to P110. Prior to surgery, the subject animals were handled for two to three days to get them used to the experimenter and to being touched and lifted. On two to four further days they were put onto the platforms in the setup for ca. 20 min for habituation. The same procedure applied to the stimulus animals. Extracellular recordings were performed in eight female and six male Wistar rats.
Stimulus animals were in the same age range as the subject animal or slightly older. After surgery, subject animals were housed singly, while stimulus animals remained housed in same-sex groups of two to three per cage. No contingent record was kept of whether the subject rat had been in one cage with the stimulus animals previously, or whether they were even littermates. Thus, stimulus animals cannot be categorized by their novelty. However, even if known, this would only apply for a small subset of recordings, as the stimulus animals were
reused over days and were thus nearly always known from day two onwards.
All animals were kept on a 12:12 reversed light/dark cycle with lights o at 8 a.m. and had free access to food and water.
All experimental procedures were performed according to German guidelines on animal welfare under the supervision of local ethics committees according to animal experimentation permit G0259/09.
2.1.3 Whisker trimming
In a subset of experiments, the whiskers were trimmed. Trimming took place in two ways.
First, in three animals, it was possible to trim the whiskers during a recording session and acquire interaction data before and after trimming from the same units. This was only possible if the animal was very well trained and remained calm during the trimming procedure. In this trimming, macrovibrissae stumps of <2 mm and the microvibrissae remained uncut. At a later stage in these three, as well as in two further animals, the macrovibrissae and larger microvibrissae were trimmed under isourane anesthesia so that no stumps remained. For trimmed animals, behavioral events analogous to the whisker overlap start and end, as used for untrimmed rats, were dened as the time points when the whiskers would have presumably started or stopped overlapping, if they had been present. Naturally, these virtual events can only be approximations. However, in a vast majority of cases there was also a head touch event, so that purely virtual interactions were an exception.
2.1.4 Stued rat as stimulus
In both purely behavioral and physiological experiments, a stued rat was included as a control stimulus. The stued rat was a taxidermized adult male rat. The main tactile dierences to alive rats were the stiness of the body and skin, and the angle of the whiskers. While alive rats either whisked actively or held their whiskers only slightly retracted, the stued rat had its whiskers in a very retracted position (ca. -45°, with 0° being perpendicular to head midline). The stued rat did not have any distinctive smell to the human observer, but clearly lacked the strong smell of adult male rats.
The positioning of the stued rat was such that the nose was approximately in the middle of the gap between the platforms, and could always comfortably be reached by the subject rat. As the subject rats occasionally bit the stued rat and pulled it down from the platform, in later experiments the stued rat was additionally xed on the stimulus platform.
2.1.5 Objects and anesthetized rat as stimuli
In physiological experiments, objects and an anesthetized rat were presented as further con-trol stimuli. The objects presented were most often cuboidal, such as blocks made from polystyrene, cardboard, wood, or a mix of cardboard and paper (i.e., books). In some cases, cylindrical objects as plastic bottles or toilet paper rolls were presented. The objects were positioned on the platform opposite to the subject animal in a fashion similar to the stued
rat, such that the object extended approximately to the center of the gap between platforms.
Objects changed over the days of one experiment to increase interest in and the number of interactions with them. In some cases, the subject rats tried to climb onto the objects and over onto the opposite platform, so that interactions had to be either interrupted, or the presentation of the object in question had to be stopped altogether.
The anesthetized rat was an adult male rat of ca. 350 g weight. It was anesthetized with ketamine and xylazine, similar to the subject rat at the beginning of surgery (see 2.3.1), and was then presented to the subject rat in a location similar to the stued rat and objects, i.e., in the middle of the gap between platforms. However, it was hand-held by the experimenter, and in some cases the snout or anogenital region of the anesthetized rat were brought close to the subject rat to rouse interest. This typically resulted in active whisking onto the anesthetized rat, and a bout of interactions even after the anesthetized rat had been brought back to its normal position over the gap.
2.1.6 Comparison of stued with alive rat interactions
From here on, the term 'alive rat' denominates an awake, behaving rat, as opposed to other, immobile stimuli, including the anesthetized rat, although this was, of course, also alive.
Paradigms A stued rat was presented either alongside alive rats in a chooser paradigm, or alone. In behavioral experiments explicitly aimed at comparing interactions with stued and alive rats, these stimuli were presented together in a chooser setting. On each day, two recordings of ve minutes length were performed for each subject rat, which could be either male or female. Between these two recordings, the alive stimulus rat, which was always male, was changed. The position of the stimuli on the platform was pseudo-randomized.
In physiological experiments, to maximize the number of interactions with the stued rat while minimizing the presentation time, the stued rat was presented alone in nearly all cases. In addition, presentation blocks could either be shortened when the subject rat showed no interest in the stued rat at all, or lengthened to increase the number of interactions.
Furthermore, stued rat presentations were not distributed evenly within recordings, with an above-average probability of being presented amongst the last blocks on a given recording day. These adjustments to other experimental needs made interaction times with the stued rat from experiments with neuronal recordings dicult to compare with alive rat interaction times. This limitation did not apply to the purely behavioral experiments, however.
Quantication of behavior in chooser settings The behavior during purely behavioral stued vs. alive chooser experiments was quantied by scoring the subject rat behavior in the low-speed videos with a custom-written Matlab script. The scoring was applied in intervals of 0.67 s (corresponding to 20 frames), with ve possible behavioral states: (1) subject rat interacting with alive rat, (2) subject rat in the proximity of the alive rat, but not touching it (typically over the gap, but sometimes orienting towards the stimulus while remaining
completely over the platform), (3) no orientation towards either stimulus, (4) subject rat in the proximity of the stued rat, and (5) subject rat interacting with the stued rat.
2.2 Receptive elds and whisker set angles