The brain is able to create and maintain a map of our surroundings as we go about our daily activities. It is thought that some of this spatial information is first processed in a region of the brain called the visual cortex, the information is then relayed to a region called the hippocampus, where the map is reliably stored. However, researchers do not fully understand how the brain transfers spatial information between these two regions.

To explore what happens to the information, Haggerty and Ji recorded electrical signals from the brains of rats that were being trained to run back and forth along a C-shaped track. As they ran on the track, the rats gradually developed spatial maps of there surroundings. Electrodes were used to record brain signals from both the visual cortex and the hippocampus as this development took place.

Similar to previous studies, analysis of the recordings showed that a specific population of neurons in the hippocampus, called CA1 neurons, produced an electrical signal whenever the rat ran past specific locations on the track, such as corners. Building on this Haggerty and Ji showed a population of neurons in the visual cortex called V1 neurons also produced location-specific electrical signals at the same time. Moreover, the electrical signals of these two populations of neurons fluctuate together in a coordinated fashion.

The results of Haggerty and Ji support the idea that there exists a specific population of visual cortical neurons that communicate with hippocampal neurons in the development of a particular spatial map.

Spatial information is presumably encoded by hippocampal ‘place cells’, which fire predominantly when an animal is at one or a few specific places (place fields) of a given space (O'Keefe and Dostrovsky, 1971; McNaughton et al., 1983; Wilson and McNaughton, 1993). How these cells establish the location specificity has been a subject under intensive investigation. It is known that path integration of the information generated from self-motion is important for hippocampal place cell activity (Hafting et al., 2005; McNaughton et al., 2006). However, external sensory cues are equally important, if not more. In particular, visual cues play a pivotal role in place cell activities (Muller, 1996; Knierim, 2002; Colgin et al., 2008; Ravassard et al., 2013). For example, changing the proximal or distal visual cues alters firing rates and/or firing locations of place cells in a one- or two-dimensional space (Lee et al., 2004a, 2004b; Leutgeb et al., 2005), and the rotation of a salient cue card on the wall of a symmetric open arena evokes an equivalent rotation of place field locations (Muller and Kubie, 1987).

How visual cue information contributes to place cell activities is unknown. Visual cues are presumably encoded and processed in the visual cortex, which is connected to the hippocampus via a multi-synaptic pathway involving the associational cortices including the entorhinal cortex (Miller and Vogt, 1984; Vaudano et al., 1991; Lavenex and Amaral, 2000; Furtak et al., 2007). Computational models have shown that hippocampal cells can integrate input from cortical cells, in particular those encoding visual cues, to form a conjunctive response that is location-specific (de Araujo et al., 2001; Schonfeld and Wiskott, 2015). Alternatively, it is proposed that local visual cues influence place field location through entorhinal ‘border’ cells, whereas distal cues influence place field orientation via entorhinal ‘head direction’ cells (Knierim and Hamilton, 2011). Obviously, these models and others similar require the functional interaction between those visual cortical neurons encoding visual cues and corresponding hippocampal place cells. However, experimental data for such an interaction are very limited (Ji and Wilson, 2007), and the nature of this interaction during natural spatial behavior remains unexplored.

To probe the visual cortical—hippocampal interaction during spatial processing, we simultaneously recorded cells in the primary visual cortex (V1) and in the CA1 area of the hippocampus, while freely moving rats traversed back and forth on a track. We aimed to test a hypothesis that there exists a specific population of visual cortical neurons that responds to visual features of the track with specific activity patterns and functionally interacts with those hippocampal place cells representing the same track. We found that, like CA1 place cells, a large number of V1 neurons also displayed firing activities that were specific to particular locations on the track, especially around the visual landmarks. The location-specific activities of V1 neurons, on average, led those of CA1 place cells in both the spatial and temporal domains. For those V1 and CA1 cells that displayed spatially overlapping activity, their precise firing characteristics co-fluctuated every time the animal traveled through the track. These findings suggest the existence of specific V1 activities that participate in the encoding of external spaces. Given the critical role of the hippocampus in spatial memories (Scoville and Milner, 1957; Squire, 1992; Eichenbaum et al., 1999; Burgess and O’Keefe, 2003), these visual cells may also participate in the formation and storage of the spatial memories encoded by hippocampal place cells.

We used tetrodes to simultaneously record firing activities of V1 cells and CA1 place cells in 15 freely moving rats, while they were performing a track running task for food rewards. During the task, the food-deprived rats were free to run along a novel C-shaped track (Figure 1A). Each end of the track (food well) was baited with a milk reward which, if consumed, was not refilled until the rat had run to the opposite end of the track to consume the other reward. Over time rats became habituated to the task and repeatedly travelled back and forth along the two overlapping trajectories in order to maximize the number of rewards. The recording started on the very first day (Day 1) and continued for 2–14 days, with 20–60 min each day. Eight animals reached at least Day 7. We quantified the task performance of all the rats from Day 1 to Day 6, and a Day 7+, which included those days ≥ Day 7. Behavioral performance was measured by the running speed and the number of times (laps) per minute an animal traveled through each of the 2 trajectories. Both parameters increased rapidly during the early few days of running and reached a stable level on Day 3 (Figure 1B,C), indicating that much of the behavioral changes occurred during the first few days.

Figure 1

Download asset Open asset

A total of 852 V1 and 3627 CA1 cells were recorded across all recording days and from all the 15 animals (10 ± 1 V1 cells, 44 ± 2 CA1 cells per day per rat, N = 83 day × rats; see ‘Materials and methods’ for details). V1 cells were sampled in all layers (Figure 1D). In this study, we analyzed 776 V1 cells and 2033 CA1 cells that were active on at least one trajectory (mean firing rate >0.5 Hz). Putative CA1 interneurons (mean rate > 5 Hz) were excluded. For each cell active on a given trajectory, we computed the overall firing rate of the cell during the running of the trajectory. The median firing rate of V1 cells was 4.3 [2.0 9.9] Hz (median and [25% 75%] range values, N = 1501 cell × trajectories), whereas the median rate of CA1 cells was 1.6 [0.97 2.5] Hz (N = 2909).

Location-specific firing activities of V1 cells

To test our hypothesis, we asked whether V1 cells responded to particular locations on a given trajectory of the track in a manner similar to CA1 place cells, with the assumption that the V1 response might not be a spatial response per se, but possibly resulted from the visual cues. As expected, CA1 cells exhibited typical place-field firing characteristics on the track (Figure 2A). Many V1 cells also dramatically increased their firing rates at specific locations of a trajectory (Figure 2B–D), as we have shown previously (Ji and Wilson, 2007). The location-specific increase in firing activity of V1 cells was apparently stable during each lap of the trajectory. The firing rate curves, defined as the lap-averaged firing rate at every position of a trajectory, of these V1 cells displayed a few well-defined peaks, similar in character to the place fields of CA1 cells and to the multi-peak firing of medial entorhinal grid cells on linear tracks (Hafting et al., 2008). Henceforth, we refer to the locations corresponding to the rate curve peaks of a cell as its ‘firing fields’.

Figure 2 with 1 supplement see all

Download asset Open asset

We quantified the location-specificity of each individual V1 and CA1 cell active on a trajectory. To compare with the location-specificity arising randomly from chance, we randomly shuffled the cell's spiking activity by circularly shifting the spikes within each lap of the trajectory with a random time interval (Henriksen et al., 2010; Igarashi et al., 2014). First, we computed spatial information content (SIc), a measure of how much information (in bits per spike) a cell's spiking activity contained about the animal's location (Skaggs et al., 1993). Although the median SIc value of V1 cells (0.17 [0.085 0.34] bits/spike, N = 1501 cell × trajectories) was relatively small, compared with that of CA1 place cells (1.6 [1.1 2.2] bits/spike, N = 2909; p < 0.0001, ranksum test), it was significantly greater than that of the shuffled V1 cells (0.061 [0.025 0.13] bits/spike; p < 0.0001; Figure 2E). Second, we computed spatial information rate (SIr), which measures spatial information in bits per second. Similarly to SIc, the median SIr of V1 cells (0.70 [0.42 1.2] bits/s) was smaller than that of CA1 cells (2.4 [1.2 4.2] bits/s; p < 0.0001), but significantly greater than that of the shuffled V1 cells (0.22 [0.14 0.37] bits/spike; p < 0.0001; Figure 2F). Third, using a method modified from previous studies (Henriksen et al., 2010; Igarashi et al., 2014), we derived a normalized spatial modulation index (SMI). The reason for this additional measure was that SIc and SIr are affected by firing rate (Figure 2—figure supplement 1). Since V1 and CA1 cells had different firing rates, the SIc and SIr values between V1 and CA1 cells were not directly comparable. SMI was defined as the SIc (or equivalently SIr) of a cell relative to its chance-level distribution produced by the random shuffling (Figure 2—figure supplement 1). SMI does not directly quantify the location-specificity of a cell's firing activity, but provides a measure of the degree of location modulation relative to random spike trains with identical firing rate and temporal spiking patterns. SMI is therefore insensitive to firing rate. The chance-level of SMI for any given cell is zero. The median SMIs of both V1 (8.1 [3.4 15.4]) and CA1 cells (12.0 [3.3 23.2]) were much higher than zero (p < 0.0001, ranksum test; Figure 2G). Finally, we defined a cell with SMI >2.325 (99th percentile of the chance-level) as a ‘location-responsive’ V1 cell. We found that 81% of trajectory-active V1 cells were location-responsive on a trajectory and for comparison, 90% of the trajectory-active CA1 cells were location-responsive.

Next, we attempted to understand why location-responsive V1 cells had lower location-specificity than CA1 place cells. One obvious reason is that V1 cells, unlike CA1 place cells, often contained much baseline firing outside of a concentrated firing field (Figure 2B–D). A second observation is that V1 cells often fired at multiple firing fields on a trajectory (Figure 2B–D), whereas many CA1 cells only exhibited one or two fields on our track (Figure 2A). We quantified the number and average spatial length of firing fields for location-responsive V1 cells. On average, the location-responsive V1 cells displayed nearly twice the number of firing fields as the CA1 place cells (mean ± SE for V1: 2.9 ± 0.04 per trajectory, N = 1216 cell × trajectories; CA1: 1.5 ± 0.02 per trajectory, N = 2618; p < 0.0001, t-test; Figure 2H). On the other hand, the average field length of V1 cells was significantly smaller than that of CA1 cells (V1: 37.5 ± 0.3 cm, N = 3521 fields; CA1: 47.2 ± 0.4 cm, N = 3918; p < 0.0001; Figure 2I). The result shows that the relative low location-specificity in V1 cells was partially due to more firing fields, but not due to larger field sizes.

We also quantified the lap-by-lap stability of a V1 cell's location-specific activity as an animal traveled the same trajectory repeatedly. For each cell active on a trajectory, we computed its firing rate curve separately for each individual lap on the trajectory and then computed the correlation between the rate curves of any two laps (Cheng and Ji, 2013). Spatial stability was the average correlation among all combinations of laps. To understand how the stability compared with the chance level, we also computed the spatial stability of each cell after the random shifting of its rate curves lap by lap. Similarly as in spatial information measures, the median spatial stability of V1 cells (0.19 [0.06 0.31], N = 1501 cell × trajectories) was lower than that of CA1 cells (0.51 [0.28 0.70], N = 2909; p < 0.0001, ranksum test), but significantly greater than the chance level (0.0014 [-0.0074 0.0060]; p < 0.0001; Figure 2J), indicating that V1 cells responded reliably to same locations during each lap of track running.

Spatial correlation between V1 and CA1 firing activities

We next asked how V1 and CA1 firing activities interacted. First, we analyzed how the V1 and CA1 firing fields were distributed along the two trajectories on the track with opposite running directions. We found that V1 and CA1 fields were not equally distributed along the trajectories, but tended to increase in number close to the food wells, as shown previously for CA1 cells (Dupret et al., 2010). V1 and CA1 field numbers also increased before and after the animals traveled through the corners of the track, referred to as landmarks, and decreased as the animals traveled through the long segment between the two middle corners (Figure 3A,B). The numbers of V1 and CA1 fields along the trajectories fluctuated together in a highly correlated fashion for both trajectories (Figure 3C). We quantified this observation by computing the cross-correlation between the field distribution curves of V1 and CA1 cells on each trajectory. The cross-correlograms for both trajectories had peaks not exactly at position lag 0, but at a positive position (2 cm, p < 0.0001, Pearson's r) on one running direction and a negative position (−4 cm, p < 0.0001) on the opposite direction (Figure 3D,E), meaning that the number of V1 fields fluctuated ahead of the number of CA1 fields consistently on both running directions. Given the mean speed of 35 cm/s of our rats, the 2–4 cm leading distance is equivalent to 60–120 ms of leading time. While the higher numbers of firing fields around the food wells and corners might simply reflect more sensory information available around the landmarks, the significant correlation between the field distribution curves of V1 and CA1 cells with a spatial delay suggests an interaction between these groups of cells, possibly reflecting the propagation of the landmark-related sensory information from V1 to CA1.

Figure 3

Download asset Open asset

The bias of V1 and CA1 fields around the landmarks on both directions prompted us to examine the directionality of V1 and CA1 firing fields. We observed that some V1 cells fired in front of or behind the landmarks on two opposite running directions, which we termed ‘prospective’ or ‘retrospective’ firing as in a previous study (Battaglia et al., 2004). An example of V1 cell with prospective firing is shown in Figure 4A, which was quantified by a significant peak at a positive position lag (36 cm, p < 0.0001, Pearson's r) in the cross-correlogram between the cell's firing rate curves on the two trajectories. An example of CA1 cell with retrospective firing is shown in Figure 4B, with a significant peak at a negative position lag (−18 cm, p < 0.0001) in the cross-correlogram between its firing rate curves on the two trajectories. We found that among all trajectory-active cells, 9.7% of V1 cells (N = 75) and 10.4% of CA1 cells (N = 212) were bidirectional, defined as those with a significant peak within [−50, 50] cm in their rate curve cross-correlograms (see ‘Materials and methods’). The cross-correlogram peak positions of these bidirectional V1 cells display a bimodal distribution with two groups around 26 cm and −26 cm (Figure 4C), which corresponded to prospective and retrospective firing, respectively. Similarly, the peak positions of CA1 cells also show a bimodal distribution with two groups around 18 cm and −28 cm (Figure 4D). This result suggests that a group of V1 cells responded to the same landmarks on both directions by either ‘looking’ ahead or back to the landmarks.

Figure 4

Download asset Open asset

Temporal correlation between V1 and CA1 firing activities

We then analyzed the pair-wise cross-correlation between V1 location-responsive cells and CA1 place cells in the time domain. We used a normalized spike count cross-correlation (see ‘Materials and methods’), which is insensitive to the cells' firing rates, to quantify how the spiking activities of two cells were temporally correlated. Figure 5A–C shows an example pair of V1 and CA1 cells with both having well-defined firing fields on the same trajectory, which led to a prominent peak in its normalized cross-correlogram. The peak time quantified the temporal relationship of the two cells. For individual pairs of cells, this peak, especially that with a long peak time, might just passively reflect the fact that they had firing fields on the same trajectory, not necessarily reflecting any functional interaction. However, if a large number of such pairs are collected, the distribution of their peak times may inform the temporal relationship between the V1 and CA1 activities at the population level. Among 22,969 pairs of V1 location-responsive cells and CA1 place cells, we obtained 997 highly significantly correlated pairs (see ‘Materials and methods’ for definition) with peak correlation times within [−0.2, 0.2] s. The distribution of these peak times was significantly biased toward positive values (58% positive, 46% negative, p = 0.00017, binomial test; Figure 5D). The result indicates that, on average, V1 cells led CA1 cells in the time domain, suggesting the propagation of visual information from V1 to CA1. We also analyzed whether this direction of interaction was true to those V1-CA1 cell pairs with both showing bi-directional firing. Out of the 12 pairs with both the V1 and CA1 displaying prospective firing and 9 pairs with one displaying prospective firing while the other displaying retrospective firing (only 1 pair with both displaying retrospective firing), 9 (75%) and 6 (68%), respectively, showed positive peak times (Figure 5E), suggesting that the V1-CA1 interaction of these pairs followed the general trend at the population level.

Figure 5

Download asset Open asset

Lap-by-lap co-fluctuation of V1 and CA1 firing activities

As many pairs of V1 and CA1 cells displayed overlapping firing fields on a trajectory (Figure 6A; see more examples in Figure 6—figure supplement 1), we reasoned that these pairs of V1 and CA1 cells could be specifically interacting for integrating the visual information present at a location to the hippocampal place cell activity encoding the same location. Therefore, we next focused on the possible interaction between V1 and CA1 cells with overlapping firing fields. We observed that the firing rates and firing locations of such V1 and CA1 cells fluctuated from lap to lap within their respective firing fields, and interestingly, they often fluctuated together in a correlated fashion (Figure 6A,B, Figure 6—figure supplement 1). We quantified this observation by an analysis similar to the ‘noise’ correlation in previous studies on V1 cell correlation in primates (Zohary et al., 1994; Ecker et al., 2010; Hansen et al., 2012). For a pair of cells with overlapping firing fields, we identified each cell's spikes within its firing field for each individual lap, and computed two quantities of these spikes: the (within-field) firing rate and the center of mass (COM) of their firing locations. We then obtained the lap-by-lap fluctuations in within-field firing rate (Δrate - the difference between a lap's firing rate and the averaged firing rate across all laps) and in COM (ΔCOM) for each cell. Finally, we computed the Pearson correlation in Δrate and in ΔCOM between two cells, which measures how closely the two cells varied their precise firing rate or firing location each time the animal traveled through the overlapped firing fields. Indeed, the pair of cells shown in Figure 6A were significantly correlated in both Δrate and ΔCOM (Figure 6C; see more examples in Figure 6—figure supplement 1).

Figure 6 with 5 supplements see all

Download asset Open asset

In our data, we found 786 pairs with overlapping firing fields (see ‘Materials and methods’). We referred to these pairs as ‘overlapping pairs’. We found that 48% and 23% of the overlapping pairs were significantly correlated in Δrate and ΔCOM (p < 0.05, Pearson's r), respectively. As a group, the average correlation of all the overlapping pairs were significantly greater than 0, the chance-level correlation, for both Δrate (0.29 ± 0.01, p < 0.0001, t-test; Figure 6D) and ΔCOM (0.16 ± 0.01, p < 0.0001; Figure 6D). This result indicates that there was a precise lap-by-lap co-fluctuation in firing rate and COM between many pairs of V1 cells and CA1 place cells with overlapping firing fields.

We also performed the same analysis on two control groups of V1-CA1 cell pairs (see Figure 6—figure supplement 2 for examples of V1-CA1 pairs). The first group consisted of 6681 pairs, each made of a CA1 place cell and a V1 location-responsive cell that both exhibited firing fields on a trajectory, but that their firing fields were non-overlapping. In this case, the Δrate and ΔCOM were computed within their most dominant firing fields on the trajectory. This control group, referred to as ‘non-overlapping pairs’, allowed us to test whether the CA1-V1 co-fluctuation was spatially confined within the overlapped firing fields. The second group contained 238 pairs, each composed of a CA1 cell that had place field on a trajectory and an active V1 cell that was not location-responsive on the trajectory. In this case, the Δrate and ΔCOM fluctuations of the V1 cell were computed within a spatial interval that was overlapped with the dominant CA1 place field (see ‘Materials and methods’). We call this group ‘non-responsive’ pairs, which allowed us to test whether the co-fluctuation was specific to the location-responsive V1 and CA1 cells. We found that the overlapping pairs had significantly higher correlation in Δrate than both non-overlapping (correlation: 0.062 ± 0.003; p < 0.0001, t-test) and non-responsive pairs (−0.023 ± 0.015; p < 0.0001; Figure 6D). Similarly, the overlapping pairs had significantly higher correlation in ΔCOM than non-overlapping (correlation: 0.0005 ± 0.0028; p < 0.0001) and non-responsive pairs (correlation: 0.047 ± 0.020; p < 0.0001; Figure 6D). We also computed the Δrate and ΔCOM correlations for CA1-CA1 cell pairs and for V1-V1 cell pairs. The results were similar: Overlapping pairs within each of the two brain areas were significantly correlated in both Δrate and ΔCOM and had significantly higher correlation than non-overlapping pairs and non-responsive pairs (Figure 6—figure supplement 3). Taken together, the results above demonstrate a specific, precise co-fluctuation in the firing rates and COMs of V1 location-responsive cells and CA1 place cells with overlapping firing fields, suggesting a functional interaction between these cells. Remarkably, this long–range interaction between cells in the distal V1 and CA1 was qualitatively similar to the local interaction within CA1 place cells.

Next, we examined whether the co-fluctuation of activity in overlapping V1-CA1 cell pairs could be explained by the lap-by-lap behavioral fluctuations. As expected (Huxter et al., 2003; Saleem et al., 2013), the firing rate and COM were modulated by speed and head direction in many V1 and CA1 cells (Figure 6—figure supplement 4). We quantified the modulation of both speed and head direction on Δrate/ΔCOM by a multi-variant linear regression and then removed the modulation to obtain the modified lap-by-lap fluctuations in firing rate/COM (modified Δrate/ΔCOM), which were no longer correlated with speed or head direction (see ‘Materials and methods’). We computed the correlations in the modified Δrate/ΔCOM for overlapping V1-CA1 cell pairs. For the pair in Figure 6A, the correlations in their modified Δrate and ΔCOM remained unchanged (Figure 6E), although there was a modest reduction in other pairs (Figure 6—figure supplement 1). For the group of overlapping V1-CA1 pairs, the average correlation between the modified Δrate (0.14 ± 0.009) remained significantly greater than 0 (p < 0.0001, t-test), and was significantly higher than those of non-overlapping (0.018 ± 0.002, p < 0.0001) and non-responsive (−0.019 ± 0.014, p < 0.0001) V1-CA1 pairs (Figure 6F). Similarly, the average correlation between the modified ΔCOM (0.13 ± 0.009) was also significantly greater than 0 (p < 0.0001), and was significantly higher than those of non-overlapping (−0.003 ± 0.003, p < 0.0001) and non-responsive (0.029 ± 0.018, p < 0.0001) V1-CA1 pairs (Figure 6F). From these results, we conclude that, behavioral variations cannot fully account for the correlation in the firing rate and COM of overlapping V1-CA1 cell pairs.

The co-fluctuation of V1-CA1 cells suggests that those overlapping pairs of V1-CA1 cells with both displaying bidirectional firing should show the same type of directionality (prospective or retrospective firing), similarly as in previous reports on CA1 cells and entorhinal grid cells (De Almeida et al., 2012; Bieri et al., 2014). Our data yielded 4 such overlapping and 10 non-overlapping pairs of V1 and CA1 cells. We found that all the 4 overlapping pairs showed the same type of directionality, but only 5 of the 10 non-overlapping pairs did so whereas the other 5 showed the opposite directionality. Although the number of such bi-directional overlapping pairs is low (mainly due to the low percentage of bidirectional V1 and CA1 cells), the result is consistent with the correlated fluctuation in the firing rate and firing location between overlapping V1 and CA1 cells. In addition, CA1 cells are known to systematically shift their COMs backward lap by lap (Mehta et al., 1997, 2000), presumably as a result of synaptic plasticity (Ekstrom et al., 2001). To understand whether the observed co-fluctuation in COM between V1 and CA1 cells were related to this lap-by-lap plastic change, we examined the lap-by-lap shift in the COMs of V1 cells. Our analysis shows that, although there was a sign of backward shifting in V1 cells during the first 5 laps or so, the shift was not as robust as and much smaller than that of CA1 cells (Figure 6—figure supplement 5). The result suggests that V1 firing activities are less plastic than those of CA1 cells at this short time scale of laps (McClelland et al., 1995), and that the co-fluctuation between V1 and CA1 cells is not primarily driven by rapid plasticity in their firing activities.

Experience dependence of V1 firing activities

In our experiments, V1 and CA1 cells were recorded as the animals ran the same track for multiple days. We next analyzed how the activities of V1 location-responsive cells and CA1 place cells changed over the many days' experience of track running. We grouped the recording days to 3 different time points (T1–T3): T1 included Day 1 and 2, a novel condition to the animals, in which most of the behavioral changes occurred (Figure 1B,C), T3 included Day 6 to Day 7+, which we consider to be a familiar condition, and T2 included Days 3–5, an intermediate condition between novel and familiar. We examined 279 V1 and 628 CA1 cells at T1, 173 V1 and 519 CA1 cells at T2, and 212 V1 and 557 CA1 cells at T3 that were active on at least one trajectory. First, we found that there was a significant increase from T1 to T3 in the median overall firing rate of V1 cells (50% increase between T1 and T3; p < 0.0001, Kruskal–Wallis test including all data at T1, T2 and T3), but not in the median firing rate of CA1 cells (3% increase between T1 and T3; p = 0.079; Figure 7A). Second, there was a significant increase in the median SMI for both V1 (55% increase between T1 and T3; p < 0.0001) and CA1 cells (39% increase between T1 and T3; p < 0.0001; Figure 7B), suggesting an experience-dependent increase in the location-specificity of V1 cells at the time scale of days. Third, we examined the correlation in Δrate and ΔCOM between overlapping V1-CA1 cell pairs at different time points (N = 244 pairs at T1, 108 pairs at T2, 174 pairs at T3). For the Δrate correlation, there was no significant change from T1 to T3 (2% decrease between T1 and T3; p = 0.20, one-way ANOVA comparing all data at T1, T2 and T3; Figure 7C). This is also true for the modified Δrate correlation after the speed and head direction modulation was removed (19% increase between T1 and T3; p = 0.13; Figure 7D). For the ΔCOM correlation, there was a small, but statistically non-significant decrease from T1 to T3 (17% decrease between T1 and T3; p = 0.11, one-way ANOVA; Figure 7C). For the modified ΔCOM, the decrease was more prominent, but remained non-significant (33% decrease between T1 and T3; p = 0.088; Figure 7D). These data show that, although there was an indication of slightly decreased co-fluctuation between V1 and CA1 cells with experience, their functional interaction persisted as the animals became familiarized with the track. Taken together, day-to-day track experience was accompanied by an increase in the firing rate and location-specificity of V1 cells, which remained correlated in their lap-by-lap fluctuations with overlapping CA1 cells.

Figure 7

Download asset Open asset

Layer differences in V1 firing activities

Finally, we asked whether the firing properties of V1 neurons differed across layers. We determined the laminar locations of the recorded neurons on the final recording day by performing small electrolytic lesions and subsequently examining the Nissl and acetylcholinesterase (AChE) staining (Figure 1D). We estimated the locations of those neurons recorded on previous days by relative distances tetrodes had traveled. A neuron's location was assigned to either the superficial layers 2 and 3 (L2/3), layer 4 (L4), or the deep layers 5 and 6 (L5/6). We obtained 478 cells in L2/3, 120 in L4, and 83 in L5/6 that were active on a trajectory. A small number of V1 cells (95 out of 776, 12%) were not assigned to a layer due to failed AChE staining. We first analyzed the overall firing rate and SMI among the cells recorded in different layers (Figure 8A,B) on all days. There was a significant difference in firing rate across the layers (p < 0.0001, Kruskal–Wallis test). The post-hoc pair-wise comparisons revealed that L4 cells had significantly higher firing rates (10.2 [6.5 14.9] Hz) than both L2/3 (3.2 [1.7 7.0] Hz, p < 0.0001, ranksum test) and L5/6 (2.9 [1.8 6.3] Hz, p < 0.0001) cells, whereas the latter two were similar (p = 0.67). The analysis of SMI shows that 86% of L2/3, 78% of L4, and 70% of L5/6 cells were location-responsive. The results indicate that L4 cells had the highest firing rate, whereas L2/3 cells had the highest percentage of location-responsive cells. Second, we computed the correlations in Δrate and ΔCOM for overlapping pairs of CA1 cells with the cells in each of the three cortical locations (CA1-L2/3, CA1-L4, and CA1-L5/6). We found that L2/3, L4 and L5/6 cells were similarly correlated in Δrate and ΔCOM with CA1 cells (data not shown). Third, we have shown that V1 cells increased their firing rates and SMIs with track running experience (T1 to T3, Figure 7A,B). Here we examined whether this was the case for cells in all layers of V1. We found a significant increase in firing rate only in L2/3 cells (p = 0.00023, Kruskal–Wallis test comparing all data at T1, T2 and T3; 39% increase between T1 and T3), but not in L4 (p = 0.60; 2.5% increase between T1 and T3) or L5/6 cells (p = 0.10; 2.3% increase between T1 and T3; Figure 8C). However, there was a significant increase in SMI from T1 to T3 for cells in all layers (L2/3: p < 0.0001, 52% increase between T1 and T3; L4: p < 0.0001, 130% increase between T1 and T3; L5/6: p = 0.05, 54% increase between T1 and T3; Figure 8D).

Figure 8

Download asset Open asset

We have analyzed the firing activities of V1 and CA1 cells in freely moving rats as they ran back and forth on a C-shaped track for food reward. We found that a large number of cells in all layers of V1, like CA1 place cells, fire selectively and reliably at specific locations of the track. These location-responsive V1 cells have multiple, narrow firing fields on a given trajectory of the track. The number of V1 firing fields rises and falls as animals approach and move away from visual landmarks, leading the rise and fall in the number of CA1 place fields. The precise lap-by-lap fluctuations in the firing rate and firing location of V1 cells are correlated with those of CA1 cells with overlapping firing fields. Finally, V1 cells display an experience-dependent change in their location-specificity at the time scale of days. These results reveal that V1 cells have location-specific responses to a particular environment and are functionally coupled with the hippocampal place cells encoding the same environment.

Our data supports a functional interaction between V1 location-responsive cells and hippocampal place cells. Here we provide three pieces of evidence. First, the number of V1 firing fields fluctuates as the animal travels through a trajectory and this fluctuation is highly correlated with the change in the number of CA1 firing fields. More interestingly, we found that the V1 fields rise and fall earlier than the CA1 fields and this leading effect is consistent on both running directions. Second, the pair-wise cross-correlation analysis reveals a bias toward positive time lags, suggesting that, again, V1 location-specific activity appears earlier than the CA1 place field activity at the population level. Third, the lap-to-lap fluctuations in the firing rates and firing locations of individual V1 cells within their firing fields are significantly correlated with those of CA1 place cells with overlapping place fields. This is true even after removing the modulation of speed and head direction on firing rate and firing location. The correlated variation is specific to CA1-V1 cell pairs with the overlapping firing fields, suggesting a remarkably precise interaction at a fine temporal and spatial scale. These observed correlations are unlikely driven by non-specific common factors because of the following reasons. First, both the field distribution and pair-wise cross-correlation analyses show that V1 activities lead CA1 activities during trajectory running. Although the lead could be caused by CA1 and V1 cells independently responding to some unknown common factors with a delay, which we cannot rule out, the lead time of 60–120 ms, is consistent with our previous studies on the interaction between V1 and CA1 during slow-wave sleep (Ji and Wilson, 2007; Haggerty and Ji, 2014), when behavioral factors do not vary and the brain state is relatively uniform. Therefore, the delay likely reflects the propagation of activity from V1 to CA1. Second, the lap-by-lap co-fluctuation of overlapping pairs is much higher than that of non-responsive pairs. Since both types of pairs share overlapping locations, if common factors were mainly responsible, the non-responsive pairs would show correlated fluctuations comparable to those of overlapping pairs.

The leading of V1 activity over CA1 activity suggests a forward interaction, through which V1 location-responsive cells send visual information to CA1 place cells. The location-specific activity in V1 cells does not necessarily mean that V1 cells directly respond to places per se as hippocampal place cells do. Given what we know about the primary function of V1 neurons in behaving rodents (Yao et al., 2007; Niell and Stryker, 2010; Xu et al., 2012), they are likely driven by the visual cues associated with specific locations on the track, supported by our observation that V1 firing fields are more distributed around the landmarks. In particular, prospective firing may occur when V1 neurons ‘see’ landmarks in front of the animal. Retrospective firing may occur as V1 neurons ‘see’ landmarks behind the animal's head, which is possible because the side-facing eyes of rats have a wide visual field and can see visual stimuli behind or above the head (Wallace et al., 2013), although other mechanisms such as feedback from the higher cortical areas or the hippocampus may also be responsible. The landmark information represented by the V1 cells can reach the hippocampus via a multi-synaptic pathway, which includes the secondary visual cortex, temporal cortex, perirhinal/postrhinal cortex, and entorhinal cortex (Miller and Vogt, 1984; Vaudano et al., 1991; Lavenex and Amaral, 2000; Furtak et al., 2007). In this way, the V1 location-responsive cells may allow local and distal visual cues to influence CA1 place cell activities (Muller and Kubie, 1987; Lee et al., 2004a, 2004b; Leutgeb et al., 2005).

Our finding of location-responsive cells in the sensory area V1 raises the possibility that the distinction between what we call either a spatial or sensory response is not as clear as we thought. On one hand, given the apparent modulation of hippocampal place cells by visual cues (Muller and Kubie, 1987; Lee et al., 2004a, 2004b; Leutgeb et al., 2005), the response of a hippocampal cell to a particular place also contains a ‘sensory’ response to the visual cues associated with that place. In our data, the CA1 place fields are preferentially distributed around the landmarks, just as the V1 firing fields are, suggesting a bias of CA1 place cell activity by sensory input. On the other hand, the response of a V1 cell to visual clues at a particular place effectively establishes a ‘spatial’ response to the place, if the visual cues are stable in space. In this manner, the firing fields of V1 cells may appear to be similar to the CA1 place fields, with the important differences in that V1 cells exhibited less specificity by nature of non-zero baseline firing and multiple firing fields. The difference in V1 and CA1 location-specific activity is just that, because hippocampal place cell activities result from an integration of not just visual information, but also information from other sensory modalities and self-motion, hippocampal place fields are more specific and predictive of the animal's spatial location than the predominantly sensory-driven V1 firing fields. Therefore, it is possible that spatial representation in the brain may involve a gradual transformation along the anatomical pathway connecting the sensory cortex to the hippocampus, which converts a primarily sensory-driven, less reliable spatial response in the sensory cortex to a more precise and robust spatial response in the hippocampus.

Hippocampal place cells are believed not just to form a neural representation of space, but encode spatial memory and other type of memories (Eichenbaum et al., 1999; Burgess and O'Keefe, 2003; Ferbinteanu et al., 2006; Knierim at al., 2006; Moser et al., 2008). If so, then the gradual transformation of spatial representation also suggests that a spatial memory trace is not only encoded by hippocampal place cells, but also by the cortical cells involved in the transformation, which are presumably distributed in many cortical areas. This idea is an essential component in many modern theories of memory, including index theory (Teyler and Rudy, 2007), two-stage memory processing theory (Buzsaki, 1996), and the theory of two complementary memory systems (McClelland et al., 1995). All these theories propose the involvement of relevant cortical cells in long-term memory storage. Our previous study shows that activity patterns of V1 and CA1 cells during track running are coordinately replayed later during slow-wave sleep (Ji and Wilson, 2007), suggesting the involvement of V1 cells in memory consolidation. The current study provides another piece of key evidence that V1 cells and CA1 place cells functionally interact during the stage of memory formation as animals run a novel track for a first week or so. In our data, the lap-by-lap co-fluctuation between V1-CA1 cell pairs with overlapping firing fields is qualitatively similar to that between pairs of CA1-CA1 place cells with overlapping place fields. Since CA1 place cells with overlapping place fields are believed to constitute functional ‘cell assemblies’ that encode the spatial memory of the animal's environment (Harris et al., 2003; Dragoi and Buzsaki, 2006), it is possible that the overlapping V1 location-responsive cells and hippocampal place cells are part of the larger ‘cortical-hippocampal assemblies’ that encode and store the same spatial memory trace. Therefore, we propose that the V1 location-specific activities are a substrate for encoding and storing the visual component of long-term spatial memories, and possibly episodic memories.

1. 1. Battaglia FP
   2. Suttherland GR
   3. McNaughton BL

   (2004) Local sensory cues and place cell directionality: additional evidence of prospective codeing in the hippocampus

   The Journal of Neuroscience 24:4541–4550.

   https://doi.org/10.1523/JNEUROSCI.4896-03.2004

   • Google Scholar
2. 1. Bieri KW
   2. Bobbitt KN
   3. Colgin LL

   (2014) Slow and fast γ rhythms coodinate different spatial coding modes in hippocampal place cells

   Neuron 82:670–681.

   https://doi.org/10.1016/j.neuron.2014.03.013

   • Google Scholar
3. 1. Burgess N
   2. O'Keefe J

   (2003) Neural representations in human spatial memory

   Trends in Cognitive Sciences 7:517–519.

   https://doi.org/10.1016/j.tics.2003.10.014

   • Google Scholar
4. 1. Buzsaki G

   (1996) The hippocampo-neocortical dialogue

   Cerebral Cortex 6:81–92.

   https://doi.org/10.1093/cercor/6.2.81

   • Google Scholar
5. 1. Cheng J
   2. Ji D

   (2013) Rigid firing sequences undermine spatial memory codes in a neurodegenerative mouse model

   eLife 2:e00647.

   https://doi.org/10.7554/eLife.00647

   • Google Scholar
6. 1. Cheng S
   2. Frank LM

   (2008) New experiences enhance coordinated neural activity in the hippocampus

   Neuron 57:303–313.

   https://doi.org/10.1016/j.neuron.2007.11.035

   • Google Scholar
7. 1. Colgin LL
   2. Moser EI
   3. Moser MB

   (2008) Understanding memory through hippocampal remapping

   Trends in Neurosciences 31:469–477.

   https://doi.org/10.1016/j.tins.2008.06.008

   • Google Scholar
8. 1. De Almeida L
   2. Idiart M
   3. Villavicencio A
   4. Lisman J

   (2012) Alternating predictive and short-term memory modes of entorhinal grid cells

   Hippocampus 22:1647–1651.

   https://doi.org/10.1002/hipo.22030

   • Google Scholar
9. 1. de Araujo IE
   2. Rolles ET
   3. Stringer SM

   (2001) A view model which accounts for the spatial fields of the hippocampal primate spatial view cells and rat place cells

   Hippocampus 11:699–706.

   https://doi.org/10.1002/hipo.1085

   • Google Scholar
10. 1. Diba K
    2. Buzsaki G

    (2007) Forward and reverse hippocampal place-cell sequences during ripples

    Nature Neuroscience 10:1241–1242.

    https://doi.org/10.1038/nn1961

    • Google Scholar
11. 1. Dragoi G
    2. Buzsaki G

    (2006) Temporal encoding of place sequences by hippocampal cell assemblies

    Neuron 50:145–157.

    https://doi.org/10.1016/j.neuron.2006.02.023

    • Google Scholar
12. 1. Dupret D
    2. O'Neill J
    3. Pleydell-Bouverie B
    4. Csicsvari J

    (2010) The reorganization and reactivation of hippocampal maps predict spatial memory performance

    Nature Neuroscience 13:995–1002.

    https://doi.org/10.1038/nn.2599

    • Google Scholar
13. 1. Ecker AS
    2. Berens P
    3. Keliris GA
    4. Bethge M
    5. Logothetis NK
    6. Tolias AS

    (2010) Decorrelated neuronal firing in cortical microcircuits

    Science 327:584–587.

    https://doi.org/10.1126/science.1179867

    • Google Scholar
14. 1. Eichenbaum H
    2. Dudchenko P
    3. Wood E
    4. Shapiro M
    5. Tanila H

    (1999) The hippocampus, memory, and place cells: is it spatial memory or a memory space?

    Neuron 23:209–226.

    https://doi.org/10.1016/S0896-6273(00)80773-4

    • Google Scholar
15. 1. Ekstrom AD
    2. Meltzer J
    3. McNaughton BL
    4. Barnes CA

    (2001) NMDA receptor antagonism blocks experience-dependent expansion of hippocampal "place fields"

    Neuron 31:631–638.

    https://doi.org/10.1016/S0896-6273(01)00401-9

    • Google Scholar
16. 1. Ferbinteanu J
    2. Kennedy PJ
    3. Shapiro ML

    (2006) Episodic memory–from brain to mind

    Hippocampus 16:691–703.

    https://doi.org/10.1002/hipo.20204

    • Google Scholar
17. 1. Foster DJ
    2. Wilson MA

    (2006) Reverse replay of behavioural sequences in hippocampal place cells during the awake state

    Nature 440:680–683.

    https://doi.org/10.1038/nature04587

    • Google Scholar
18. 1. Furtak SC
    2. Wei SM
    3. Agster KL
    4. Burwell RD

    (2007) Functional neuroanatomy of the parahippocampal region in the rat: the perirhinal and postrhinal cortices

    Hippocampus 17:709–722.

    https://doi.org/10.1002/hipo.20314

    • Google Scholar
19. 1. Hafting T
    2. Fyhn M
    3. Bonnevie T
    4. Moser MB
    5. Moser EI

    (2008) Hippocampus-independent phase precession in entorhinal grid cells

    Nature 453:1248–1252.

    https://doi.org/10.1038/nature06957

    • Google Scholar
20. 1. Hafting T
    2. Fyhn M
    3. Molden S
    4. Moser MB
    5. Moser EI

    (2005) Microstructure of a spatial map in the entorhinal cortex

    Nature 436:801–806.

    https://doi.org/10.1038/nature03721

    • Google Scholar
21. 1. Haggerty DC
    2. Ji D

    (2014) Initiation of sleep-dependent cortical-hippocampal corrections at wakefulness-sleep transition

    Journal of Neurophysiology 112:1763–1774.

    https://doi.org/10.1152/jn.00783.2013

    • Google Scholar
22. 1. Hansen BJ
    2. Chelaru MI
    3. Dragoi V

    (2012) Correlated variability in laminar cortical circuits

    Neuron 76:590–602.

    https://doi.org/10.1016/j.neuron.2012.08.029

    • Google Scholar
23. 1. Harris KD
    2. Csicsvari J
    3. Hirase H
    4. Dragoi G
    5. Buzsaki G

    (2003) Organization of cell assemblies in the hippocampus

    Nature 424:552–556.

    https://doi.org/10.1038/nature01834

    • Google Scholar
24. 1. Henriksen EJ
    2. Colgin LL
    3. Barnes CA
    4. Witter MP
    5. Moser MB
    6. Moser EI

    (2010) Spatial representation along the proximodistal axis of CA1

    Neuron 68:127–137.

    https://doi.org/10.1016/j.neuron.2010.08.042

    • Google Scholar
25. 1. Huxter J
    2. Burgess N
    3. O'Keefe J

    (2003) Independent rate and temporal coding in hippocampal pyramidal cells

    Nature 425:828–832.

    https://doi.org/10.1038/nature02058

    • Google Scholar
26. 1. Igarashi KM
    2. Lu L
    3. Colgin LL
    4. Moser MB
    5. Moser EI

    (2014) Coordination of entorhinal-hippocampal ensemble activity during associative learning

    Nature 510:143–147.

    https://doi.org/10.1038/nature13162

    • Google Scholar
27. 1. Ji D
    2. Wilson MA

    (2007) Coordinated memory replay in the visual cortex and hippocampus during sleep

    Nature Neuroscience 10:100–107.

    https://doi.org/10.1038/nn1825

    • Google Scholar
28. 1. Ji D
    2. Wilson MA

    (2008) Firing rate dynamics in the hippocampus induced by trajectory learning

    The Journal of Neuroscience 28:4679–4689.

    https://doi.org/10.1523/JNEUROSCI.4597-07.2008

    • Google Scholar
29. 1. Karlsson MP
    2. Frank LM

    (2009) Awake replay of remote experiences in the hippocampus

    Nature Neuroscience 12:913–918.

    https://doi.org/10.1038/nn.2344

    • Google Scholar
30. 1. Knierim JJ

    (2002)

    Dynamic interactions between local surface cues, distal landmarks, and intrinsic circuitry in hippocampal place cells

    The Journal of Neuroscience 22:6254–6264.

    • Google Scholar
31. 1. Knierim JJ
    2. Hamilton DA

    (2011) Framing spatial cognition: neural representations of proximal and distal frames of reference and their roles in navigation

    Physiological Reviews 91:1245–1279.

    https://doi.org/10.1152/physrev.00021.2010

    • Google Scholar
32. 1. Knierim JJ
    2. Lee I
    3. Hargreaves EL

    (2006) Hippocampal place cells: parallel input streams, subregional processing, and implications for episodic memory

    Hippocampus 16:755–764.

    https://doi.org/10.1002/hipo.20203

    • Google Scholar
33. 1. Lavenex P
    2. Amaral DG

    (2000) Hippocampal-neocortical interaction: a hierarchy of associativity

    Hippocampus 10:420–430.

    https://doi.org/10.1002/1098-1063(2000)10:4<420::AID-HIPO8>3.0.CO;2-5

    • Google Scholar
34. 1. Lee I
    2. Rao G
    3. Knierim JJ

    (2004a) A double dissociation between hippocampal subfields: differential time course of CA3 and CA1 place cells for processing changed environments

    Neuron 42:803–815.

    https://doi.org/10.1016/j.neuron.2004.05.010

    • Google Scholar
35. 1. Lee I
    2. Yoganarasimha D
    3. Rao G
    4. Knierim JJ

    (2004b) Comparison of population coherence of place cells in hippocampal subfields CA1 and CA3

    Nature 430:456–459.

    https://doi.org/10.1038/nature02739

    • Google Scholar
36. 1. Leutgeb S
    2. Leutgeb JK
    3. Barnes CA
    4. Moser EI
    5. McNaughton BL
    6. Moser MB

    (2005) Independent codes for spatial and episodic memory in hippocampal neuronal ensembles

    Science 309:619–623.

    https://doi.org/10.1126/science.1114037

    • Google Scholar
37. 1. McClelland JL
    2. McNaughton BL
    3. O'Reilly RC

    (1995) Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory

    Psychological Review 102:419–457.

    https://doi.org/10.1037/0033-295X.102.3.419

    • Google Scholar
38. 1. McNaughton BL
    2. Barnes CA
    3. O'Keefe J

    (1983) The contributions of position, direction, and velocity to single unit activity in the hippocampus of freely-moving rats

    Experimental Brain Research 52:41–49.

    https://doi.org/10.1007/BF00237147

    • Google Scholar
39. 1. McNaughton BL
    2. Battaglia FP
    3. Jensen O
    4. Moser EI
    5. Moser MB

    (2006) Path integration and the neural basis of the ‘cognitive map’

    Nature Reviews Neuroscience 7:663–678.

    https://doi.org/10.1038/nrn1932

    • Google Scholar
40. 1. Mehta MR
    2. Barnes CA
    3. McNaughton BL

    (1997) Experience-dependent, asymmetric expansion of hippocampal place fields

    Proceedings of the National Academy of Sciences of USA 94:8918–8921.

    https://doi.org/10.1073/pnas.94.16.8918

    • Google Scholar
41. 1. Mehta MR
    2. Quirk MC
    3. Wilson MA

    (2000) Experience-dependent asymmetric shape of hippocampal receptive fields

    Neuron 25:707–715.

    https://doi.org/10.1016/S0896-6273(00)81072-7

    • Google Scholar
42. 1. Miller MW
    2. Vogt BA

    (1984) Direct connections of rat visual cortex with sensory, motor, and association cortices

    The Journal of Comparative Neurology 226:184–202.

    https://doi.org/10.1002/cne.902260204

    • Google Scholar
43. 1. Moser EI
    2. Kropff E
    3. Moser MB

    (2008) Place cells, grid cells, and the brain's spatial representation system

    Annual Review of Neuroscience 31:69–89.

    https://doi.org/10.1146/annurev.neuro.31.061307.090723

    • Google Scholar
44. 1. Muller R

    (1996) A quarter of a century of place cells

    Neuron 17:813–822.

    https://doi.org/10.1016/S0896-6273(00)80214-7

    • Google Scholar
45. 1. Muller RU
    2. Kubie JL

    (1987)

    The effects of changes in the environment on the spatial firing of hippocampal complex-spike cells

    The Journal of Neuroscience 7:1951–1968.

    • Google Scholar
46. 1. Niell CM
    2. Stryker MP

    (2010) Modulation of visual responses by behavioral state in mouse visual cortex

    Neuron 65:472–479.

    https://doi.org/10.1016/j.neuron.2010.01.033

    • Google Scholar
47. 1. O'Keefe J
    2. Dostrovsky J

    (1971) The hippocampus as a spatial map. Preliminary evidence from unit activity in the freely-moving rat

    Brain Research 34:171–175.

    https://doi.org/10.1016/0006-8993(71)90358-1

    • Google Scholar
48. Book

    1. Paxinos G
    2. Watson C

    (2007)

    The rat brain in sterotaxic coordinates

    Academic press.

    • Google Scholar
49. 1. Ravassard P
    2. Kees A
    3. Willers B
    4. Ho D
    5. Aharoni D
    6. Cushman J
    7. Aghajan ZM
    8. Mehta MR

    (2013) Multisensory control of hippocampal spatiotemporal selectivity

    Science 340:1342–1346.

    https://doi.org/10.1126/science.1232655

    • Google Scholar
50. 1. Saleem AB
    2. Ayaz A
    3. Jeffery KJ
    4. Harris KD
    5. Carandini M

    (2013) Integration of visual motion and locomotion in mouse visual cortex

    Nature Neuroscience 16:1864–1869.

    https://doi.org/10.1038/nn.3567

    • Google Scholar
51. 1. Schonfeld F
    2. Wiskott L

    (2015) Modeling place field activity with hierarchical slow feature analysis

    Frontiers in Computational Neuroscience 9:51.

    https://doi.org/10.3389/fncom.2015.00051

    • Google Scholar
52. 1. Scoville WB
    2. Milner B

    (1957) Loss of recent memory after bilateral hippocampal lesions

    Journal of Neurology, Neurosurgery, and Psychiatry 20:11–21.

    https://doi.org/10.1136/jnnp.20.1.11

    • Google Scholar
53. 1. Siapas AG
    2. Lubenov EV
    3. Wilson MA

    (2005) Prefrontal phase locking to hippocampal theta oscillations

    Neuron 46:141–151.

    https://doi.org/10.1016/j.neuron.2005.02.028

    • Google Scholar
54. 1. Sirota A
    2. Csicsvari J
    3. Buhl D
    4. Buzsaki G

    (2003) Communication between neocortex and hippocampus during sleep in rodents

    Proceedings of the National Academy of Sciences of USA 100:2065–2069.

    https://doi.org/10.1073/pnas.0437938100

    • Google Scholar
55. Book

    1. Skaggs WE
    2. McNaughton BL
    3. Gothard KM
    4. Markus EJ

    (1993)

    An information-theoretic approach to deciphering the hippocampal code

    In: Hanson SJ, Cowan JD, Giles CJ, editors. Advances in neural information processing systems. San Mateo: Morgan Kaufmann. pp. 1030–1037.

    • Google Scholar
56. 1. Squire LR

    (1992) Memory and the hippocampus: a synthesis from findings with rats, monkeys, and humans

    Psychological Review 99:195–231.

    https://doi.org/10.1037/0033-295X.99.2.195

    • Google Scholar
57. 1. Teyler TJ
    2. Rudy JW

    (2007) The hippocampal indexing theory and episodic memory: updating the index

    Hippocampus 17:1158–1169.

    https://doi.org/10.1002/hipo.20350

    • Google Scholar
58. 1. Vaudano E
    2. Legg CR
    3. Glickstein M

    (1991) Afferent and Efferent connections of temporal association cortex in the rat: a Horseradish Peroxidase study

    The European Journal of Neuroscience 3:317–330.

    https://doi.org/10.1111/j.1460-9568.1991.tb00818.x

    • Google Scholar
59. 1. Wallace DJ
    2. Greenberg DS
    3. Sawinski J
    4. Rulla S
    5. Notaro G
    6. Kerr JND

    (2013) Rats maintain an overhead binocular field at the expense of constant fusion

    Nature 498:65–69.

    https://doi.org/10.1038/nature12153

    • Google Scholar
60. 1. Wilson MA
    2. McNaughton BL

    (1993) Dynamics of the hippocampal ensemble code for space

    Science 261:1055–1058.

    https://doi.org/10.1126/science.8351520

    • Google Scholar
61. 1. Xu S
    2. Jiang W
    3. Poo MM
    4. Dan Y

    (2012) Activity recall in a visual cortical ensemble

    Nature Neuroscience 15:449–455.

    https://doi.org/10.1038/nn.3036

    • Google Scholar
62. 1. Yao H
    2. Shi L
    3. Han F
    4. Gao H
    5. Dan Y

    (2007) Rapid learning in cortical coding of visual scenes

    Nature Neuroscience 10:772–778.

    https://doi.org/10.1038/nn1895

    • Google Scholar
63. 1. Zilles K
    2. Wree A
    3. Schleicher A
    4. Divac I

    (1984) The monocular and binocular subfields of the rat's primary visual cortex: a quantitative morphological approach

    The Journal of Comparative Neurology 226:391–402.

    https://doi.org/10.1002/cne.902260308

    • Google Scholar
64. 1. Zohary E
    2. Shadlen MN
    3. Newsome WT

    (1994) Correlated neuronal discharge rate and its implications for psychophysical performance

    Nature 370:140–143.

    https://doi.org/10.1038/370140a0

    • Google Scholar

Author details

1. Daniel Christopher Haggerty

   Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, United States

   Contribution

   DCH, Acquisition of data, Analysis and interpretation of data, Drafting or revising the article

   Competing interests

   The authors declare that no competing interests exist.

2. Daoyun Ji

   1. Department of Molecular and Cellular Biology, Baylor College of Medicine, Houston, United States
   2. Department of Neuroscience, Baylor College of Medicine, Houston, United States

   Contribution

   DJ, Conception and design, Analysis and interpretation of data, Drafting or revising the article

   For correspondence

   dji@bcm.edu

   Competing interests

   The authors declare that no competing interests exist.

• 2,724

  views

• 654

  downloads

• 78

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.