Abstract: Though sharp-wave ripples have been recorded in the EEG data of the hippocampus of mammals for years, it remains unclear how ripples can contribute to memory for different behaviors.. Sharp wave ripples are one of the most synchronous patterns in the mammalian brain. These waves are most common during non-REM sleep, although they can also be associated with consummatory behaviors. In EEG recordings, these occurrences can be seen as large amplitude negative polarity deflections (40–100 ms) in CA1 stratum radiatum that are associated with a short-lived fast oscillatory pattern of the LFP in the CA1 pyramidal layer, known as “ripples.” The purpose of this study was to investigate the distinction between sleep and awake ripples along with the connection between sharp-wave ripples and specific mammalian behaviors during memory tasks. The hypothesis tested was that SPW-Rs occur when the animal has an experience that will help guide subsequent successful task completion that results in obtaining a desired reward. To conduct the experiment electrophysiological signals were collected from a rat’s hippocampus during various tasks. The data were then analyzed using Neuroscope and compared to a visual recording of the rat’s actions. The data suggest that sharp wave ripples are more likely to occur close to a reward, most often before the reward, and do not have a higher tendency to occur early or late in learning. Future research can further clarify these results and investigate the process by which these ripples occur.
I. Introduction
The thoroughly investigated hippocampus, a region of the brain, is shown to have a pivotal role in learning and consolidation of memory (Bartsch and Wulff, 2015). This complex, elongated structure is embedded deep in the medial temporal lobe, forming part of the limbic system, and known to regulate emotional responses (Anand and Dhikav, 2012; Knierim, 2015). The hippocampus is a plastic structure that may be damaged by several stimuli (Anand and Dhikav, 2012). It can be distinguished externally with a layer of densely packed neurons that curl into a S-shaped structure on the edge of the temporal lobe (Anand and Dhikav, 2012). The hippocampus consists of two parts: Cornu ammonis, or hippocampus proper, and dentate gyrus (DG) (Anand and Dhikav, 2012). These two parts are separated by the hippocampus proper and curve into one another (Anand and Dhikav, 2012). The Cornu ammonis or hippocampus proper is divided into CA1, CA2, CA3, and CA4 (Anand and Dhikav, 2012). The hippocampus is part of the allocortex, or archicortex, and is separated from the neocortex (Anand and Dhikav, 2012). In rodents, the hippocampus is a relatively large, cashew-shaped structure that lies beneath the neocortex (Knierim, 2015). The cross-section of its long axis reveals the hippocampal anatomical connectivity, or the trisynaptic loop (Knierim, 2015). This loop can be described as follows: the entorhinal cortex, composed of two distinct brain regions in rats, provides major cortical input to the hippocampus with strong projections from the performant path to the DG region; the DG region projects to the CA3 region via the mossy fiber pathway; CA3 projects to the CA1 region via the Schaffer Collateral pathway; CA1 projects back to the previously described entorhinal cortex (Knierim, 2015). It should be noted that the connectivity within the transverse axis of the hippocampus is complex, with multiple parallel processing and feedback circuits: the entorhinal complex also directly projects to the CA3 and CA1 regions; CA3 provides a feedback projection to the DG through excitatory mossy cells of the dentate hilus, proving that the hippocampal processing is not exclusively unidirectional (Knierim, 2015). The CA2 unit has its own functions and is regarded as a distinct computational unit similar to the CA3 and CA1 (Knierim, 2015). A copious amount of information is known about the neurophysiology of the hippocampus. The most studied cell of hippocampal neural activity is the place cell (Knierim, 2015). The pyramidal cells, a type of multipolar neuron, of the CA1, CA2, CA3 regions, and the granule cells in the DG, are selectively fired when rats inhabit more than one specific location in an environment, which is the ‘firing field’ or ‘place field’ (Knierim, 2015). Discovering these cells prompted the theory that the hippocampus forms a cognitive map of the environment (Knierim, 2015).
Sharp-wave ripples have been observed in the hippocampus of every species investigated so far including humans (Bragin et al., 1999; Le Van Quyen et al., 2010). These waves are most common during non-REM sleep, although they are also associated with consummatory behaviors during wakefulness. Furthermore, they are the most synchronous events in the mammalian brain and are thus associated with short, impermanent excitability in the hippocampus. The synchronous population events from SPW-Rs are especially significant because they can lead to interictal epileptic discharges if altered erroneously (Suzuki and Smith, 1988; Buzsaki et al., 1989) and the fast ripples are often used as markers for epileptic propensity (Bragin et al., 1999). Additionally, the spike content of SPW-Rs represents sequentially organized neurons similar to those in the walking animal.
While extensive research has covered SPW-Rs in the past, the timing of these events during different stages of learning and in different behavioral tasks remains poorly understood. Therefore, the goal of this project is to evaluate the role of sharp wave ripples in learning and discover the relationship between these ripples and actions of mammals in behavioral tasks. To do this, EEG recordings were collected and analyzed from rats during various behavioral tasks. The hypothesis tested in this study is that SPW-Rs occur when the animal has an experience that will help guide subsequent successful task completion that results in obtaining a desired reward. This research can help fill in the gaps of knowledge regarding mechanisms of learning and memory in diverse mammalian behaviors.
II. Materials and Methods
The data collected from each training session were stored and opened in Neuroscope to be analyzed. NeuroScope is a viewer for displaying various physiological and behavioral data and it allows comparison of analyzed data with the original recordings. NeuroScope allowed the researcher to mark specific occurrences of the sharp wave ripple and then compare those occurrences to the video file to determine the relationship between the actions of the animals and the sharp wave ripple occurrences. To open the files in Neuroscope correctly, input the correct number of channels, sampling rate, and amplification (the amplification can be modified later).
Once the physiological recordings are imported into Neuroscope the first step to analyzing the SPW-Rs is isolating specific channels. Look for channels with ripples and waves to easily see the artifacts. Extraneous channels can be hidden from the “Units” section and important channels can be moved into separate groups to better organize the channels. Next, with the duration set to about 1000 milliseconds, the researcher can begin scanning the data for SPW-Rs. The measure tool can be used to calculate the period between the peaks of the ripples and the duration of the ripples. After creating or loading an event file, new events can be marked for each artifact.
III. Results
After collecting the recordings from a sleeping rat, the data were imported into Neuroscope to be analyzed. The specific channels to be viewed were moved to a separate group and the channels were organized so that the ripples could be seen vertically above or below the sharp waves. In the sleep state, SPW-Rs are much clearer to observe in the channels and more common to find, especially during non-REM sleep. These ripples can be measured with a period of five to seven milliseconds from peak to peak (using the measuring tool), last between 20-100 milliseconds, and are seen adjacent to sharp waves in the lower channels. There were 110 minutes of data collection for the sleep state and the various SPW-Rs throughout the were marked in the event file. During a period of ten minutes during non-REM sleep, thirty SPW-R occurrences were marked. This resulted in a calculated rate of events value of 3.0 SPW-Rs per minute by dividing the number of occurrences over the number of minutes that the ripples spanned over.
IV. Discussion
The first result of this study found that SPW-Rs are more common to find during non-REM sleep states than the awake state in rats. According to the results, there were about 3 occurrences per minute for the sleep state while the rate was only 1.2 for the awake state, suggesting that SPW-Rs occur more often during sleep. Only one trial was analyzed for the sleep state rate, however, this finding conforms to previously believed ideas about the ripples in non-REM sleep.
The next part of this study analyzed the SPW-Rs in relation to a reward during the cheeseboard task. Data was collected from four files and the SPW-Rs were detected and compared to the video of the rat. The data suggest that SPW-Rs are slightly more likely to occur close to a reward (immediately before or after they found and ate the food in the cheese board task) than at other times during the task. Calculating the proportion of ripples in each condition found that ripples occurred close to the reward 56% of the time, just slightly more than ripples not close to the reward. Further research is necessary to draw definitive conclusions from this result however, since many of the events that were considered not close to a reward occurred at moments the rat was not on the cheeseboard and out of view from the camera angle. Thus, further research should be conducted taking into consideration the actions of the rats between the learning tasks when the rats were off the board.
Finally, this study looked at the difference between the number of SPW-Rs in the first training session compared to the second training session to see whether SPW-Rs occurred more in the early or late stages of learning. The data collected show that they are more likely to occur in the first training session, with 57% of ripples occurring in the first session. However, a closer look at the data shows that one file seemed to be an outlier, with 34 occurrences, thus, it is more likely that SPW-Rs occur at the same tendency in the first session as the second session. By eliminating the first trial, the proportion of ripples in the first session averages out to just 52% which is much closer to an equal number of ripples for both sessions. Thus, the null hypothesis fails to be rejected and there is no statistically significant evidence to suggest that SPW-Rs occur more often earlier in learning than later.
V. Conclusion
The purpose of this research was to study the role of sharp wave ripples in learning and discover the relationship between these ripples and the actions of mammals in behavioral tasks. After collecting data from rats during various training sessions, the study was able to support four important findings: (1) SPW-Rs are more common to find during non-REM sleep states than the awake state in rats. (2) SPW-Rs are slightly more likely to occur close to a reward than at other times during the task. (3) SPW-Rs are marginally more likely to occur before a reward is found rather than after the reward. (4) SPW-Rs do not occur more often earlier in learning than later.
Although this study was deliberately planned, every experiment faces some inevitable limitations. Some of these limitations from this experiment include the small sample size, noise from around the silicon probe affecting the data recordings, and the human error involved in analyzing hours of data collection. Inevitably, some events may have been missed. While these limitations are important to address, it is important to note that they are unlikely to have a significant effect on the results of this study.
Future research will be important to confirm the results of this study and further analyze the role of SPW-Rs in memory consolidation. Further research could include learning about the specific biological mechanisms by which SPW-Rs form, the reasoning behind the visual difference between sleep and awake state ripples, and the process by which SPW-Rs spread across the brain from the hippocampus.
References
NYCSEA