The Role Of Dorsal Striatum In Reward And Goal-directed Actions

Learning and memory from a startle event and behavior resulted as reward can help animals and humans to live better which is a essential skill. The striatum is the major input nucleus of the basal ganglia which we well known as a crucial region which related to the movement disease. The recent evidence suggested that striatum may play a role in learning and cognitive behavior.Most investigations which using the high density multiple channel electrodes are focus on rodents'hippocampus and in striatum of rat and primates. The question what we concern is what is the function of striatum in cerebral hemisphere. Specifically, we want to know are:1, what is the basic firing characteristic of striatal neurons and how these neurons represent discrete startle event.2, How the striatal neural network encode the goal-directed behavior which lead to reward including the action selection.3, How the striatal neural network represent an associated event based on reward or aversive stimuli. Our investigation use multiple microdrive electrodes to in vivo recording striatal neurons in mice when they perform in plus maze and classic conditioning in an attempt to unveil the underlying mechanisms of striatal neurons in learning and cognitive behavior.1, The basic feature of striatal neurons firing and local field potential There are about 96% medium spiny neurons in striatum and they are GBABergic, most of other neurons are interneurons. Our results showed there are two kinds of waveforms can be recorded in dorsal striatum. First kind of waveform has short duration, sharp wave trough, and symmetrical waveform. The second kind has long duration, asymmetrical waveform. The firing rates of first kind of waveforms are much higher than the second waveforms but have lower density.2, Reprogrammed of striatal neuron network in goal-directed behavior which lead to rewardBy using our multiple channel technique to recording neural activity when the mouse perform the plus maze. The mouse need use spatial cue to find the fixed location reward from fixed start spot. The striatal network will have dynamic change during progress.First, we random gave the food pellet to the mouse and record the neural response when the animal biting the pellet after food habituation. Using the mouse biting moment as time zero, the results showed approximate 60% neurons changed their firing pattern, most of them increased firing rate transiently and the other responsive neurons'firing were inhibited transiently.Recent evidence showed the dorsal striatum and hippocampus are involved in mediating the goal-directed responding based on the spatial cue to lead the reward. After over trained, there will be a shift from goal-directed behavior to response habit and there is a dissociation role in sub region of dorsal striatum. Before the habit formation, the dorsolmedial striatum mediate the integration of spatial information to lead the behavior. After the habit formation, the dorsolateral striatum will force the animal to use response strategy. The results showed the striatal network had dynamic change when the mouse performs the plus maze day by day.There are five groups of neurons according to their response to reward:Prolong duration increased firing; Short duration increased firing; Two peak increased firing; Prolong duration decreased firing; Short duration decreased firing. The groups of neural network will change firing pattern when the mouse arrived the end of the goal arm without reward.1, The increased groups will shrink the duration or no response; 2, The decreased groups will have short inhibition or disinhibition. The interesting thing is the Two peak increased firing group' s second peak will shift forward and merge to the first peak.The firing rate relative to the place in plus maze will change during the progress (1st,4th,7th day). There is no firing specific to the place in the first day training, but it will change to hence the firing rate during the path from the start arm to baited arm and finally focus on the start spot, before turn and arrived the end of baited arm. The Cr.osscorrelograms analysis showed much more higher when the mouse perform in the plus maze.To make sure the firing pattern changed is base on the reward rather than the cup, we did the comparision experiment including orange (control) cup, blue circle plastic cup, white triangle paper cup, clearcircle glass cup. The results showed there was no cup specific neurons.In the 8th day probe test, we relocated the start arm to the opposite arm to test the mouse would use "place" or "response" strategy and the results showed 75.5±8.03% would be "place" strategy.3, Encode the classic conditioning in striatal neural networkTo further investigate the striatal neural activity during the conditioning followed with two opposite results which are reward or aversive stimuli, we using the same tecnique to record the neural activity in dorsal striatum. The results from the operant conditioning showed that after several lever press and food drop pairing training, mouse would learn to press lever more frequently to get the reward. The neurons from dorsolmedial striatum changed their firing rate when the mouse pressing lever including increased and decreased. The interesting thing is the start point of the change is before they press the lever. Also the firing rate related to the location analysis showed that after training, the high firing rate was concentrated by the lever and the food dish.The classic conditioning results showed before paired the conditioning stimuli (8k hz tone) to unconditioning stimuli (reward), there was no response to the tone; after paired training, the neurons would response to the tone only. When gave the same tone stimuli but no reward in a novel enviroment, there also had firing changed. It indicated that the striatal neurons would connect the tone and the reward. Furthermore, we paired another frequence tone to another unconditioning stimuli (Earthquake shaking), we found these neurons would response to the Shaking before paired and would response to both after paired, also there were few neurons would response to tone only after over trained. There also may have stimuli generalization when deliver a novel frequence tone in the shaking box but without shaking, some neurons will still firing changed. In the memory recall, the conditioning stimuli can evoke some neurons which response to the shaking response to tone only, it may indicates that the striatal neurons can be reactivated by conditioning stimuli. In the memory extinction test, the neurons would response to the tone stimuli but extinct at the first test. 4, Activity of striatal neurons response to startle event stimuliWe deliver the startle event randomly including earthquake shaking, elevator free drop, acoustic and air puff. We using 400ms duraion and 3000rpm shaking speed in our earthquake shaking; 20cm high in drop; 78db sound in acoustic and 500ms duration and 20psi in air puff experiments. There are 38.7% neurons would increase firing rate transiently and 19.4% would prolong increased,1.4%neurons decreased transiently and 2.8% neurons prolong decreased in shaking startle event. There are 66.8% neurons would increase firing rate transiently and 19.8% would prolong increased,0.9%neurons decreased transiently and 1.4% neurons prolong decreased in elevate drop startle event. There are 62.6% neurons would increase firing rate transiently and 4.1% would prolong increased,1.6%neurons decreased transiently in acoustic startle event. There are 57.7% neurons would increase firing rate transiently and 13% would prolong increased,0.8%neurons decreased transiently and 2.84% neurons prolong decreased in air puff startle event.Our studies investigated the striatal neural network pattern in goal-directed behavior lead to reward and classic conditioning related to reward and aversive stimuli. The resultes gave us some ideas of striatal network's function in integration of representations of spatial cues for the purpose of reaching a desired reward.