Rett syndrome, a postnatal neurodevelopmental disorder, is characterized by a period of normal development lasting until 6–18 months of life, followed by a period of regression and development of motor deficits, autonomic dysfunction, and intellectual disability (Chahrour and Zoghbi, 2007). Rett syndrome is caused by loss-of- function mutations in the X-linked gene MECP2, which encodes Methyl-CpG Binding Protein 2, a protein proposed to regulate chromatin states (Amir et al., 1999). Rett syndrome is primarily seen in heterozygous females (Chahrour and Zoghbi, 2007). Accordingly, female mice heterozygous for the deletion of Mecp2 (Mecp2^+/-; hereafter referred to as RTT mice) have been used to model the syndrome (Guy et al., 2001). RTT mice recapitulate many features of the human disorder, including the cognitive phenotypes (Guy et al., 2001). For example, RTT mice display deficits in contextual fear memory (Samaco et al., 2013) and spatial reference memory (Hao et al., 2015). However, how changes in the learning and memory neural circuits in RTT mice give rise to these deficits in vivo is not understood.

The hippocampus, a key area in the brain’s learning and memory circuits, is critical for both contextual fear memory (Phillips and LeDoux, 1992; Fanselow, 2000) and spatial memory (Morris et al., 1982). In particular, spatial memory is thought to be encoded by hippocampal ‘place cells’, which fire when the animal enters one or a few locations (place fields) in an environment (O'Keefe and Dostrovsky, 1971; Wilson and McNaughton, 1993). Place field locations are mostly stable over time and become more refined as the animal’s experience in the environment increases (Thompson and Best, 1990; Frank et al., 2004; Cacucci et al., 2007). This experience-dependent refinement of spatial memory code is thought to involve memory consolidation in rest periods and sleep (Diekelmann and Born, 2010; Walker and Stickgold, 2004), during which place cells are activated synchronously within short (~100 ms) time windows of high-frequency (100–250 Hz) oscillations, called sharp-wave ripples, in the local-field potentials (LFPs) of the hippocampal CA1 area (Buzsáki et al., 1992; Csicsvari et al., 2000; Buzsáki, 1989; Wilson and McNaughton, 1994). The synchronous firing in ripples is believed to induce synaptic plasticity, which transforms short-term memories to long-term memories (Buzsáki, 1989; Wilson and McNaughton, 1994; Tatsuno et al., 2007; Ji and Wilson, 2007; Sirota et al., 2003). Consistent with this idea, disrupting ripples is shown to impair experience-dependent learning (Girardeau et al., 2009; Ego-Stengel and Wilson, 2010). Despite the importance of place cells and their activities during ripple oscillations in hippocampus-dependent learning and memory, how these cells contribute to cognitive deficits in animal models of neurodevelopmental disorders is unknown.

To understand the neural circuit basis of memory deficits in RTT mice, we set out to study how hippocampal place cell activities are altered in RTT mice. In particular, our previous study shows that hippocampal cells in RTT mice tend to display an exaggerated level of synchrony (hypersynchrony) in vitro and in vivo (Lu et al., 2016). We therefore focused on the impact of hypersynchrony on spatial memory codes in freely behaving RTT mice. Specifically, we set out to test the hypothesis that hypersynchrony interferes with ripple-dependent memory processing and consequently impairs the long-term refinement of spatial memory codes in RTT mice.

To test the hypothesis we recorded place cells and LFPs using tetrodes (Cheng and Ji, 2013), from the hippocampal CA1 of adult RTT mice and their wildtype (WT) littermates that were at least 3 months of age (Figure 1A). The recordings were made as animals ran back and forth (two trajectories) for food rewards on either a novel or familiar linear track, repeatedly for two sessions (Run 1, Run 2). There was also a rest session before (Pre-run) and after (Post-run) the first run session in a majority of the recordings (Figure 1B). Each of these run and rest sessions was 10–15 min. Animals (N = 10 RTT, 9 WT mice) were first trained to perform the task on the familiar track in a familiar room for at least 2 weeks before the recording on the same track began. Then, a subset of these animals (N = 9 RTT, 7 WT mice) was placed in a new room and recorded following the same procedure on a novel track, which had a similar shape to the familiar track. The animals had never been exposed to the novel room or the novel track before. During the recording on both the familiar and novel tracks, RTT and WT mice displayed similar performance, as measured by the number of times (laps) an animal traversed a trajectory in a run session and the animal’s average speed of running (Figure 1C,D). We did not observe any seizures behaviorally and did not detect any epileptic-like activities in any of the recorded LFPs.

Figure 1

Download asset Open asset

Place cells in RTT mice contained less spatial information in the familiar, but not the novel, environment

We first analyzed place cell activities on the novel track. A total of 47 cells from 9 RTT mice (RTT cells) and 93 cells from 7 WT mice (WT cells) were recorded on the novel track. Of these, 39 RTT cells and 84 WT cells were classified as putative pyramidal cells (with mean firing rate <7 Hz; Table 1). We defined a run-active cell as a putative pyramidal cell active (with mean firing rate >0.5 Hz) on at least one trajectory during at least one run session. The percentage of run-active cells among all putative pyramidal cells in RTT mice (79%, N = 31) did not significantly differ from that in WT (63%, N = 53) mice (p=0.069, binomial test). The median firing rate of the run-active cells in RTT mice (median [25–75%] range: 1.8 [0.71 3.2] Hz, N = 31) was also not significantly different from that in WT mice (1.1 [0.57 2.3] Hz, N = 53; p=0.10, Wilcoxon ranksum test) on the novel track.

Table 1

Genotype	Novel track cells		Familiar track cells
	Putative pyramidal	Active	Putative pyramidal	Active
WT	23	12	36	27
WT	-	-	1	1
WT	21	15	8	5
WT	8	6	6	4
WT	10	6	26	11
WT	1	1	7	6
WT	9	4	18	11
WT	12	9	x	x
RTT	-	-	36	29
RTT	5	5	5	4
RTT	6	5	8	6
RTT	6	5	34	27
RTT	7	3	19	14
RTT	10	8	30	23
RTT	1	1	3	0
RTT	1	1	0	0
RTT	0	0	x	x
RTT	3	3	x	x
WT total	84	53	102	65
RTT total	39	31	135	103

To examine place fields of run-active cells on the novel track, we linearized an animal’s locations on a track trajectory. For each cell active on a trajectory in a run session, we plotted the cell’s spike raster over the linearized locations during every running lap and computed a firing rate curve (average firing rates over all laps versus linearized locations of the trajectory). On the novel track, raster plots and rate curves show that RTT and WT cells fired spikes at specific locations (place fields) along a trajectory (Figure 2A). To quantify firing specificity, we computed a cell’s spatial information (Skaggs et al., 1993), which measures, in bits per spike, the amount of information the cell’s firing provides about the animal’s location on a trajectory. We found that the spatial information of RTT cells was similar to that of WT cells (Figure 2B) on the novel track. We then quantified how consistent a cell’s firing on a trajectory was between Run 1 and Run 2 by firing stability, which is the Pearson correlation between the rate curves of a cell in the two sessions (Ciupek et al., 2015). We found that the stability of RTT cells on the novel track was significantly reduced from that of WT cells (Figure 2B). We then identified individual place fields of each cell and determined whether place field properties were altered in RTT mice. We found a greater number of place fields per cell per active trajectory in RTT (median [25–75%] range: 2 [1 3]; N = 31 cells) than that in WT cells (1 [1 2], N = 53; p=0.049, Wilcoxon ranksum test), but the place field length in RTT mice was shorter (Figure 2B). The median peak firing rate within place fields was comparable between RTT and WT cells on the novel track (Figure 2B). These quantifications show that place cells in RTT mice contained similar information on the novel track as those in WT mice, but encoded space by more place fields with shorter lengths.

Figure 2 with 2 supplements see all

Download asset Open asset

We next performed the same analyses on RTT and WT cells active on the familiar track (Figure 2C), where we recorded 171 cells from 8 RTT mice and 136 cells from 7 WT mice. Among these, 135 RTT cells and 102 WT cells were putative pyramidal cells (Table 1). The percentage of run-active cells among all putative pyramidal neurons in RTT mice (77%, N = 103) was greater than that in WT (64%, N = 65) mice (p=0.035). The median firing rate of these run-active neurons in RTT (1.8 [1.1 3.0] Hz, N = 103) mice was significantly increased from that in WT mice (1.1 [0.6 2.1] Hz, N = 65; p=0.0012). However, the spatial information of run-active RTT cells was significantly reduced (by 63%), compared to that of WT run-active cells, while the stability was similar (Figure 2D). The number of place fields per cell was similar between RTT (2 [1 2], N = 103 cells) and WT (1 [1 2], N = 65) cells (p=0.16, Wilcoxon ranksum test) on the familiar track, but the place field length of RTT cells was significantly larger than that of WT cells (Figure 2D). The median peak firing rate within place fields was comparable between RTT and WT cells (Figure 2D). The results suggest that the reduced specificity of RTT cells on the familiar track was due to larger place fields.

To gain more understanding of this abnormality, we compared the firing specificity and place field properties of WT or RTT cells between the novel and familiar tracks. In WT mice, the median spatial information of place cells was greatly increased (300%) from the novel to familiar track (p=3.3×10⁻¹⁶, Wilcoxon ranksum test), indicating an experience-dependent refinement of firing specificity in WT cells. The number of place fields remained similar between the novel and familiar track, but the field length became significantly smaller (p=0.00056) and median peak firing rate within place fields was significantly increased (p=1.5×10⁻⁷). The result suggests that experience refines place cells by narrowing place field sizes and increasing firing rates. In RTT mice, the spatial information of RTT cells was only modestly increased between novel and familiar tracks (48%, p=0.008), indicating a reduction in experience-dependent refinement of firing specificity in RTT mice, compared to that in WT mice. The number of place fields was similar between the novel and familiar track and the peak rate increased (p=4.5×10⁻⁵), as in WT mice. However, place fields in RTT mice were relatively narrow on the novel track and became significantly widened on the familiar track (p=2.4×10⁻⁸). The comparison indicates that, although place cells in RTT mice still increased spatial information with experience, likely due to higher peak firing rates, this increase was much less than that in WT mice mainly because of larger place fields.

We need to point out that there were fewer cells recorded on the novel track than on the familiar track. The smaller sample size could render a true effect involving the novel track statistically insignificant. Therefore, we performed a downsampling analysis, in which we re-sampled down the number of cells active on the familiar track in both WT and RTT mice, as well as the number of WT cells active on the novel track, to the same number of RTT cells (31) on the novel track. We repeated the downsampling 200 iterations and examined a key effect involving place cells on the familiar track, spatial information, for each iteration. We found that 42% of the iterations produced a significant difference in spatial information between WT and RTT cells on the familiar track and 3% in spatial information on the novel track (Figure 2—figure supplement 1). The numbers suggest that, although the difference on the familiar track was more frequently observed than that on the novel track, the downsampling did reduce the statistical power for identifying a significant difference between WT and RTT cells on the familiar track. On the other hand, 98% of the iterations produced a significant difference in spatial information of WT cells between the novel and familiar tracks, but only 25% in RTT cells, suggesting that the experience-dependent refinement was robust in WT mice, but largely disappeared in RTT mice, with the downsampled cells. Therefore, although the downsampling led to less statistical power, the reduced experience-dependent refinement in RTT cells, compared to that in WT cells, was reproduced even with the downsampled cells.

We then examined the dynamic change in spatial information at a shorter time scale of within a recording day. We computed the lap-by-lap spatial information during both Run 1 and Run 2 on the novel and familiar track (Figure 2—figure supplement 2). On the novel track, WT and RTT cells showed similar spatial information in early and late laps during Run 1 and throughout Run 2. On the familiar track, WT cells started both Run 1 and Run 2 with already high spatial information in early laps, but RTT cells started with a low level of spatial information similar to that on the novel track. This suggests that place cells in RTT mice could produce specific firing from ongoing sensory experience on the track. However, the specificity was not maintained between running sessions or across days.

Finally, to understand whether place fields of RTT cells were impaired at a fine level, we examined theta phase precession, which establishes a fine correlation between place cell firing and location (O'Keefe and Recce, 1993). We found that theta phases of both RTT and WT place cells precessed as animals passed through their place fields (Figure 3A). We quantified theta phase precession by an optimal linear regression (O'Keefe and Recce, 1993) and a circular-linear correlation (Ravassard et al., 2013) between spike theta phases and positions within a place field. The linear correlation and the slope of the linear regression did not significantly differ between WT and RTT cells on either the novel or the familiar track (Figure 3B). However, there was a small, yet significant, reduction in the circular-linear correlation of RTT cells in the familiar environment (11%), compared to that of WT cells, but not in the novel environment (Figure 3C). The analysis suggests that the fine level of correlation between firing activity and location within place fields was present in RTT mice, but could be slightly weaker in the familiar environment, possibly related to their larger place fields.

Figure 3

Download asset Open asset

Place cell synchrony failed to increase after novel experience in RTT mice

Next, we examined ripples in the rest session before (Pre-run) and after (Post-run) the first spatial experience (Run 1) on a given day in either the novel or familiar room. For each animal and each rest session, we identified individual ripple events in LFPs (Figure 4A) and quantified ripple properties by occurrence rate (number of ripples per second), duration, amplitude, and frequency. Each of these parameters was averaged over all ripple events in a rest session. In the novel environment, ANOVA analysis shows that there was no significant effect of genotype or significant interaction between genotype and rest session in ripple rate, duration, amplitude, or frequency (Figure 4B). Similar findings were made in the familiar environment (Figure 4C). The data suggest that ripples in RTT and WT mice were qualitatively similar in rest sessions.

Figure 4 with 3 supplements see all

Download asset Open asset

We then examined firing rates of CA1 cells within ripple events (Figure 4—figure supplement 1). In the novel room, the firing rates of all putative pyramidal cells or run-active cells were not significantly different between WT and RTT in either Pre- or Post-run. In the familiar room, the median firing rate of all putative pyramidal cells in RTT mice was similar to that in WT mice in both Pre-run and Post-run, but run-active cells had higher firing rate in RTT mice than in WT mice, especially in Pre-run. This firing rate increase suggests that run-active cells might fire in ripples more often in RTT mice than in WT mice in the familiar room. Indeed, run-active cells in RTT mice participated in ripples with a higher rate and fired more spikes per ripple event in the familiar room, but not in the novel room (Figure 4—figure supplement 2). In contrast, run-inactive cells did not differ between WT and RTT mice in either of these parameters in either the novel or familiar room (Figure 4—figure supplement 3), providing a control that the difference in ripple participation between WT and RTT mice was not due to differences in resting behavior or ripple events per se.

It is known that place cells active in a new spatial experience tend to fire synchronously during ripples immediately after the experience more frequently than during ripples before (Wilson and McNaughton, 1994; Skaggs and McNaughton, 1996). We therefore asked whether this experience-dependent increase in firing synchrony was altered in RTT mice. We examined firing activities of run-active cells within ripples in Pre- and Post-run in the novel room. In WT mice, run-active cells rarely fired together during ripples in Pre-run, but did so more frequently in Post-run. However, in RTT mice run-active cells already fired together in Pre-run and they appeared not to do so more frequently in Post-run (Figure 5A). To quantify this, we computed a normalized pair-wise cross-correlogram (Sirota et al., 2003; Ciupek et al., 2015) within ripples in Pre-/Post-run for each pair of run-active cells that were also recorded in Pre-run and Post-run. Cross-correlograms of these cell pairs in WT mice showed low correlation around the time lag 0 in Pre-run and relatively high correlation in Post-run (Figure 5B). However, cross-correlograms in RTT mice showed high correlation around the time lag 0 already in Pre-run, without further increase in Post-run (Figure 5B). We used the average correlation value around time lag 0 (within [−50 50] ms) as a measure of ‘coactivity’ for each pair of cells (Wilson and McNaughton, 1994). We then compared the coactivity values of all run-active pairs between RTT and WT mice to evaluate their difference in firing synchrony at the cell population level. We found that the coactivity in WT mice was significantly increased from Pre- to Post-run (Figure 5C), indicating an experience-dependent increase in synchrony following running on the novel track, as expected (Wilson and McNaughton, 1994). In RTT mice, however, the coactivity was not significantly different between Pre-run and Post-run, indicating a lack of increase in synchrony after the novel track experience (Figure 5C). Moreover, the Pre-run coactivity was significantly higher in RTT mice than that in WT mice, whereas the Post-run coactivity was not significantly different. Therefore, run-active cells in RTT mice failed to increase their synchrony following a novel experience and this failure appeared to be due to their higher than normal synchrony (hypersynchrony) prior to the experience, consistent with our previous study (Lu et al., 2016).

Figure 5 with 3 supplements see all

Download asset Open asset

The finding of hypersynchrony among run-active cells in Pre-run leads to the question of whether a general hypersynchrony within ripples existed among all putative pyramidal cells, unrelated to any track-running experience. We first re-analyzed the correlograms in Pre-run before the novel-track running, but for all pairs of putative pyramidal cells in RTT and WT mice. Indeed, we found that the coactivity in RTT mice was significantly higher than that in WT mice (Figure 5—figure supplement 1). Second, we analyzed the coactivity of run-active cells in the familiar environment. We found a similar hypersynchrony among run-active cells in RTT mice, relative to those in WT mice, within ripples in both Pre- and Post-run (Figure 5—figure supplement 2), although the coactivity of RTT cells in Post-run now became significantly higher than that in Pre-run, likely because the repetitive experience on the familiar track eventually further increased the cells’ coactivity (see Discussion). Finally, we computed the coactivity of run-silent cells within ripples (Pre- and Post-run in novel and familiar environments combined) and found that the coactivity was higher in RTT mice than that in WT mice (Figure 5—figure supplement 3). The results suggest a general hypersynchrony among CA1 neurons in RTT mice.

RTT mice displayed normal contextual fear memory at short, but not at long, time scales

Our analysis so far indicates that place cell synchrony failed to increase after novel experience, likely due to baseline hypersynchrony, and place cell firing specificity was impaired on the familiar track. These results suggest that impaired memory consolidation contributes to memory deficits in RTT mice. We attempted to provide behavioral evidence for this hypothesis. Previous studies have shown that RTT mice are impaired in contextual fear memory when they are tested 3 or 24 hr after learning (Samaco et al., 2013; Hao et al., 2015). Here we examined the contextual fear memory of our RTT and WT mice at a shorter time scale, 20 min after learning, when presumably only minimal consolidation had occurred and memory should be relatively normal, according to our hypothesis. After exploring a novel fear-conditioning box and receiving two mild foot shocks, animals were returned to their home cages. We measured their freezing level 20 min later, at which time the animals were placed back to the conditioning box for 5 min. We found that the freezing level at 20 min was comparable between RTT and WT mice (Figure 6A). We also examined the freezing level of the mice at 1 and 24 hr later after the shocks and used the 24 hr freezing level as a measure of long-term memory, when presumably substantial ripple events and thus memory consolidation had occurred. ANOVA analysis revealed a significant effect of genotype and post-hoc Wilcoxon ranksum test showed that there was a significant difference at 24 hr, but not at 20 min, between WT and RTT mice. The memory of a subset of mice (N = 12 WT, 14 RTT) was examined at all three time points. However, ANOVA analysis for these mice failed to identify a significant interaction of genotype by time (F(2,72) = 0.5; p=0.61), probably due to memory extinction of these animals that were tested repeatedly. Nevertheless, our data did show that the contextual memory at a short time scale (20 min) was relatively normal in RTT mice. This result, together with previous studies (Samaco et al., 2013; Hao et al., 2015) showing contextual fear memory deficits at longer time scales, is consistent with a memory consolidation impairment in RTT mice.

Figure 6

Download asset Open asset

To understand how genetic deficiencies underlying Rett syndrome alter the function of memory circuits in vivo, we have examined CA1 place cells in freely behaving RTT and WT mice as they ran a linear track and as they rested in both a novel and familiar environment. We found that the firing specificity of place cells in RTT mice does not increase from the novel to the familiar environment as much as that of place cells in WT mice (Figure 6B). In parallel to this dynamic change of place cells, contextual fear memory in RTT mice is normal at a short time scale (20 min), but is impaired at longer time scales (Figure 6A). These deficits point to a possible impairment in memory consolidation in RTT mice. Indeed, we found that run-active cells fail to increase their firing synchrony within ripples following the novel-track running (Figure 6C), likely due to an already high baseline synchrony (hypersynchrony) among CA1 cells in RTT mice. These results are consistent with our hypothesis that hippocampal spatial memory codes are abnormally refined with experience, due to impaired memory consolidation. Our finding may provide a neural circuit mechanism for memory deficits of Rett syndrome.

It is believed that spatial memories are initially encoded by a population of place cells active in a novel environment, driven by sensory input from the environment (O'Keefe and Burgess, 1996; Knierim and Hamilton, 2011). As novel environments become familiar, spatial memories are consolidated and their memory codes become refined, which reflect not only ongoing sensory input from the familiar environments, but also reflect the consolidated memories (Frank et al., 2004; Cacucci et al., 2007; Lever et al., 2002). Our data show that CA1 place cells in RTT mice contain a similar amount of spatial information as those in WT mice on the novel track, but form more place fields with relatively shorter field lengths. This suggests that, although certain aspects of the encoding are changed, ongoing sensory input is capable of driving place cells to encode novel spatial experience in RTT mice. Indeed, our data show that contextual fear memory in RTT mice is normal shortly after learning, presumably because place cells in RTT mice can initially encode the spatial context of a conditioning box. Our data also show that place cells in WT mice are more specific on the familiar track with narrower place fields and increased peak firing rates, compared to those on the novel track. In contrast, place cells in RTT mice increase their spatial specificity from the novel to the familiar track much less than those in WT mice. Although the smaller number of recorded cells on the novel track might not produce sufficient statistical power to identify possible significant differences between WT and RTT cells in parameters such as firing rate or ripple participation rate, our key finding of impaired increase in spatial information of RTT cells in the familiar environment is robust even with downsampled cells. Consistent with this observation, RTT mice are impaired in contextual fear memories tested 3 or 24 hr after learning in previous studies (Samaco et al., 2013; Hao et al., 2015). Therefore, abnormal spatial memory code refinement may underlie behavioral deficits in RTT mice.

It is known that ripple evens are important in memory processing (Girardeau et al., 2009; Ego-Stengel and Wilson, 2010; Wu et al., 2017). Within ripple events during Pre-run prior to a novel spatial experience, our data show an enhanced correlation among place cells in RTT mice. This hypersynchrony may be related to the greater number of place fields or shorter field size seen on the novel track, but appears not sufficient to impact overall spatial information or the behavioral performance. Importantly, following the running on the novel track, run-active cells in RTT mice fail to increase their synchrony within ripples. This is in contrast to the known phenomenon that place cells active in a novel experience normally tend to fire together within ripples after the experience, as shown in previous studies (Wilson and McNaughton, 1994; Ji and Wilson, 2007) and in our data from WT mice. This experience-dependent increase in firing synchrony is hypothesized to be a neural mechanism for memory consolidation, because it likely results from synaptic plasticity during the novel experience and leads to further experience-specific synaptic plasticity for memory consolidation (Buzsáki, 1989; Wilson and McNaughton, 1994), via precise ‘replay’ of the activity patterns associated with the experience (Skaggs and McNaughton, 1996; Lee and Wilson, 2002). Our finding that this experience-dependent increase in synchrony is reduced in RTT mice suggests an impairment in memory consolidation in these animals. In addition, we also found that place fields are relatively unstable between two running sessions of the novel track, providing further evidence that spatial memory codes fail to consolidate following the first running session. Given that there is hypersynchony during Pre-run in RTT mice, it is likely that this baseline hypersynchrony occludes further experience-dependent increases in firing synchrony and thus prevents the experience-specific synaptic plasticity underlying memory consolidation, possibly due to saturation or other homeostatic mechanisms. Our previous study reported the hypersynchrony in hippocampal slices and when animals are quietly awake on an elevated resting platform (Lu et al., 2016). Our previous study also showed that the hypersynchrony, as well as memory deficits, in RTT mice can be rescued by deep-brain stimulation that promotes synaptic plasticity (Hao et al., 2015). Taken together with these previous findings, our data support a model that, although sensory input can drive the initial formation of spatial memory codes in novel environments, the baseline hypersynchrony leads to reduced experience-dependent increase in firing synchrony within ripples and consequently impairs memory consolidation and memory code refinement.

One issue that arises from our data is how to understand the change in firing synchrony within ripples in familiar environment. In WT mice, although there is still an increase in firing synchrony after running on a familiar track, the increase is less dramatic than that following running on a novel track. This is likely because experience-induced firing synchrony among run-active cells gets weaker when memories no longer need to be consolidated. In RTT mice, although hypersynchrony is still present in the rest session before running on the familiar track, the hypersynchrony among run-active cells becomes less prominent than that before running on the novel track. Also, running on the familiar track induces a larger increase in the firing synchrony than running on the novel track. It seems that run-active cells in RTT mice exhibit a change following the running on the familiar track similar to the cells in WT mice following the running on the novel track. One possibility is that this is caused by the repetitive overtraining experience on the familiar track, which may induce some long-term synaptic plasticity that ‘breaks’ the baseline hypersynchrony. The low baseline may now permit some experience-dependent increase in firing synchrony within ripples.

Our notion of abnormal memory consolidation caused by abnormal firing synchrony within ripples in RTT mice is consistent with a previous study on a mouse model of schizophrenia, which found that stronger correlations of place cell pairs within ripples lead to impaired reactivation of place cell patterns following an experience (Suh et al., 2013). However, in our study the abnormally high correlation occurs prior to any experience, and therefore the abnormal memory consolidation appears to be caused by a baseline hypersynchrony within ripples during resting. In addition, our idea that abnormal place field refinement in RTT mice is linked to abnormal neuronal activities within ripples is supported by a recent study that silencing place cells within ripples impairs refinement of their place fields (Roux et al., 2017). Finally, our finding of unstable place fields on the novel track in the face of hypersynchrony within ripples is consistent with several papers that study how manipulating neuronal activities within ripples affects place field stability. Silencing place cells within ripples leads to unstable place fields of these cells (Roux et al., 2017). Disrupting ripples altogether did not impact place field stability in one study (Kovács et al., 2016), but did reduce stability of a portion of place cells in another study (van de Ven et al., 2016). In any case, place cell silencing or ripple disruption is a much more severe alteration than the hypersynchrony in RTT mice, yet the hypersynchrony in RTT mice seems sufficient to impact place field stability.

Although we need to be cautious when interpreting data from animal studies, our study may provide insights into the cognitive deficits in human patients. The RTT mouse model shares the same genetic deficiencies that cause Rett syndrome and recapitulates key behavioral deficits in Rett patients, including impaired learning and memory (Samaco et al., 2013). It is likely that the in vivo neural circuit deficits identified here in RTT mice may also occur in human patients. Indeed, there is evidence that the increased synchrony described here is paralleled by human studies of Rett patients using fMRI or EEG recordings, which have shown that hypersynchronous neural activity at baseline is a hallmark of the disease (Naidu et al., 2001; Garofalo et al., 1988; Glaze, 2005). Our data suggest that baseline hypersynchrony in RTT mice reduces experience-dependent increase in firing synchrony for memory consolidation and consequently leads to impaired spatial memory codes. Our animal study on RTT mice thus suggests that abnormal memory codes caused by impaired memory consolidation underlie the cognitive deficits in Rett syndrome.

Data generated and analyzed for this study have been deposited to Dryad.

1. 1. Sara E Kee
   2. Xiang Mou
   3. Huda Y Zoghbi
   4. Daoyun Ji

   (2018) Data from: Impaired spatial memory codes in a mouse model of Rett syndrome

   Available at Dryad Digital Repository under a CC0 Public Domain Dedication.

   http://dx.doi.org/10.5061/dryad.h752n4q

1. 1. Amir RE
   2. Van den Veyver IB
   3. Wan M
   4. Tran CQ
   5. Francke U
   6. Zoghbi HY

   (1999) Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2

   Nature Genetics 23:185–188.

   https://doi.org/10.1038/13810

   • PubMed
   • Google Scholar
2. 1. Buzsáki G
   2. Horváth Z
   3. Urioste R
   4. Hetke J
   5. Wise K

   (1992) High-frequency network oscillation in the Hippocampus

   Science 256:1025–1027.

   https://doi.org/10.1126/science.1589772

   • PubMed
   • Google Scholar
3. 1. Buzsáki G

   (1989) Two-stage model of memory trace formation: a role for ‘noisy’ brain states

   Neuroscience 31:551–570.

   https://doi.org/10.1016/0306-4522(89)90423-5

   • PubMed
   • Google Scholar
4. 1. Cacucci F
   2. Wills TJ
   3. Lever C
   4. Giese KP
   5. O'Keefe J

   (2007) Experience-dependent increase in CA1 place cell spatial information, but not spatial reproducibility, is dependent on the autophosphorylation of the alpha-isoform of the calcium/calmodulin-dependent protein kinase II

   Journal of Neuroscience 27:7854–7859.

   https://doi.org/10.1523/JNEUROSCI.1704-07.2007

   • PubMed
   • Google Scholar
5. 1. Chahrour M
   2. Zoghbi HY

   (2007) The story of Rett syndrome: from clinic to neurobiology

   Neuron 56:422–437.

   https://doi.org/10.1016/j.neuron.2007.10.001

   • PubMed
   • Google Scholar
6. 1. Ciupek SM
   2. Cheng J
   3. Ali YO
   4. Lu HC
   5. Ji D

   (2015) Progressive functional impairments of hippocampal neurons in a tauopathy mouse model

   Journal of Neuroscience 35:8118–8131.

   https://doi.org/10.1523/JNEUROSCI.3130-14.2015

   • PubMed
   • Google Scholar
7. 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

   • PubMed
   • Google Scholar
8. 1. Csicsvari J
   2. Hirase H
   3. Mamiya A
   4. Buzsáki G

   (2000) Ensemble patterns of hippocampal CA3-CA1 neurons during sharp wave-associated population events

   Neuron 28:585–594.

   https://doi.org/10.1016/S0896-6273(00)00135-5

   • PubMed
   • Google Scholar
9. 1. Diba K
   2. Amarasingham A
   3. Mizuseki K
   4. Buzsáki G

   (2014) Millisecond timescale synchrony among hippocampal neurons

   Journal of Neuroscience 34:14984–14994.

   https://doi.org/10.1523/JNEUROSCI.1091-14.2014

   • PubMed
   • Google Scholar
10. 1. Diekelmann S
    2. Born J

    (2010) The memory function of sleep

    Nature Reviews Neuroscience 11:114–126.

    https://doi.org/10.1038/nrn2762

    • PubMed
    • Google Scholar
11. 1. Garofalo EA
    2. Drury I
    3. Goldstein GW

    (1988) EEG abnormalities aid diagnosis of rett syndrome

    Pediatric Neurology 4:350–353.

    https://doi.org/10.1016/0887-8994(88)90081-1

    • PubMed
    • Google Scholar
12. 1. Ego-Stengel V
    2. Wilson MA

    (2010) Disruption of ripple-associated hippocampal activity during rest impairs spatial learning in the rat

    Hippocampus 20:1–10.

    https://doi.org/10.1002/hipo.20707

    • PubMed
    • Google Scholar
13. 1. Fanselow MS

    (2000) Contextual fear, gestalt memories, and the hippocampus

    Behavioural Brain Research 110:73–81.

    https://doi.org/10.1016/S0166-4328(99)00186-2

    • PubMed
    • Google Scholar
14. 1. Frank LM
    2. Stanley GB
    3. Brown EN

    (2004) Hippocampal plasticity across multiple days of exposure to novel environments

    Journal of Neuroscience 24:7681–7689.

    https://doi.org/10.1523/JNEUROSCI.1958-04.2004

    • PubMed
    • Google Scholar
15. 1. Girardeau G
    2. Benchenane K
    3. Wiener SI
    4. Buzsáki G
    5. Zugaro MB

    (2009) Selective suppression of hippocampal ripples impairs spatial memory

    Nature Neuroscience 12:1222–1223.

    https://doi.org/10.1038/nn.2384

    • PubMed
    • Google Scholar
16. 1. Glaze DG

    (2005) Neurophysiology of Rett syndrome

    Journal of Child Neurology 20:740–746.

    https://doi.org/10.1177/08830738050200090801

    • PubMed
    • Google Scholar
17. 1. Guy J
    2. Hendrich B
    3. Holmes M
    4. Martin JE
    5. Bird A

    (2001) A mouse Mecp2-null mutation causes neurological symptoms that mimic Rett syndrome

    Nature Genetics 27:322–326.

    https://doi.org/10.1038/85899

    • PubMed
    • Google Scholar
18. 1. Hao S
    2. Tang B
    3. Wu Z
    4. Ure K
    5. Sun Y
    6. Tao H
    7. Gao Y
    8. Patel AJ
    9. Curry DJ
    10. Samaco RC
    11. Zoghbi HY
    12. Tang J

    (2015) Forniceal deep brain stimulation rescues hippocampal memory in Rett syndrome mice

    Nature 526:430–434.

    https://doi.org/10.1038/nature15694

    • PubMed
    • Google Scholar
19. 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

    • PubMed
    • Google Scholar
20. Software

    1. Ji D

    (2018) DataManager, version f64fcdd

    Github.

    https://github.com/DaoyunJiLab/DataManager
21. 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

    • PubMed
    • Google Scholar
22. 1. Kovács KA
    2. O'Neill J
    3. Schoenenberger P
    4. Penttonen M
    5. Ranguel Guerrero DK
    6. Csicsvari J

    (2016) Optogenetically blocking sharp wave ripple events in sleep does not interfere with the formation of stable spatial representation in the CA1 area of the Hippocampus

    PLoS One 11:e0164675.

    https://doi.org/10.1371/journal.pone.0164675

    • PubMed
    • Google Scholar
23. 1. Lee AK
    2. Wilson MA

    (2002) Memory of sequential experience in the hippocampus during slow wave sleep

    Neuron 36:1183–1194.

    https://doi.org/10.1016/S0896-6273(02)01096-6

    • PubMed
    • Google Scholar
24. 1. Lever C
    2. Wills T
    3. Cacucci F
    4. Burgess N
    5. O'Keefe J

    (2002) Long-term plasticity in hippocampal place-cell representation of environmental geometry

    Nature 416:90–94.

    https://doi.org/10.1038/416090a

    • PubMed
    • Google Scholar
25. 1. Lu H
    2. Ash RT
    3. He L
    4. Kee SE
    5. Wang W
    6. Yu D
    7. Hao S
    8. Meng X
    9. Ure K
    10. Ito-Ishida A
    11. Tang B
    12. Sun Y
    13. Ji D
    14. Tang J
    15. Arenkiel BR
    16. Smirnakis SM
    17. Zoghbi HY

    (2016) Loss and gain of MeCP2 cause similar hippocampal circuit dysfunction that is rescued by deep brain stimulation in a rett syndrome mouse model

    Neuron 91:739–747.

    https://doi.org/10.1016/j.neuron.2016.07.018

    • PubMed
    • Google Scholar
26. 1. Mehta MR
    2. Lee AK
    3. Wilson MA

    (2002) Role of experience and oscillations in transforming a rate code into a temporal code

    Nature 417:741–746.

    https://doi.org/10.1038/nature00807

    • PubMed
    • Google Scholar
27. 1. Morris RG
    2. Garrud P
    3. Rawlins JN
    4. O'Keefe J

    (1982) Place navigation impaired in rats with hippocampal lesions

    Nature 297:681–683.

    https://doi.org/10.1038/297681a0

    • PubMed
    • Google Scholar
28. 1. Naidu S
    2. Kaufmann WE
    3. Abrams MT
    4. Pearlson GD
    5. Lanham DC
    6. Fredericksen KA
    7. Barker PB
    8. Horska A
    9. Golay X
    10. Mori S
    11. Wong DF
    12. Yablonski M
    13. Moser HW
    14. Johnston MV

    (2001) Neuroimaging studies in Rett syndrome

    Brain and Development 23:S62–S71.

    https://doi.org/10.1016/S0387-7604(01)00381-3

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

    (1996) Geometric determinants of the place fields of hippocampal neurons

    Nature 381:425–428.

    https://doi.org/10.1038/381425a0

    • PubMed
    • Google Scholar
30. 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

    • PubMed
    • Google Scholar
31. 1. O'Keefe J
    2. Recce ML

    (1993) Phase relationship between hippocampal place units and the EEG theta rhythm

    Hippocampus 3:317–330.

    https://doi.org/10.1002/hipo.450030307

    • PubMed
    • Google Scholar
32. 1. Phillips RG
    2. LeDoux JE

    (1992) Differential contribution of amygdala and hippocampus to cued and contextual fear conditioning

    Behavioral Neuroscience 106:274–285.

    https://doi.org/10.1037/0735-7044.106.2.274

    • PubMed
    • Google Scholar
33. 1. Ravassard P
    2. Kees A
    3. Willers B
    4. Ho D
    5. Aharoni DA
    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

    • PubMed
    • Google Scholar
34. 1. Roux L
    2. Hu B
    3. Eichler R
    4. Stark E
    5. Buzsáki G

    (2017) Sharp wave ripples during learning stabilize the hippocampal spatial map

    Nature Neuroscience 20:845–853.

    https://doi.org/10.1038/nn.4543

    • PubMed
    • Google Scholar
35. 1. Samaco RC
    2. McGraw CM
    3. Ward CS
    4. Sun Y
    5. Neul JL
    6. Zoghbi HY

    (2013) Female Mecp2(+/-) mice display robust behavioral deficits on two different genetic backgrounds providing a framework for pre-clinical studies

    Human Molecular Genetics 22:96–109.

    https://doi.org/10.1093/hmg/dds406

    • PubMed
    • Google Scholar
36. 1. Sirota A
    2. Csicsvari J
    3. Buhl D
    4. Buzsáki G

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

    PNAS 100:2065–2069.

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

    • PubMed
    • Google Scholar
37. Book

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

    (1993)

    Advances in Neural Information Processing Systems

    Hanson S. J, Cowan J. D, Giles C. J, editors. Morgan Kaufmann.

    • Google Scholar
38. 1. Skaggs WE
    2. McNaughton BL

    (1996) Replay of neuronal firing sequences in rat hippocampus during sleep following spatial experience

    Science 271:1870–1873.

    https://doi.org/10.1126/science.271.5257.1870

    • PubMed
    • Google Scholar
39. 1. Suh J
    2. Foster DJ
    3. Davoudi H
    4. Wilson MA
    5. Tonegawa S

    (2013) Impaired hippocampal ripple-associated replay in a mouse model of schizophrenia

    Neuron 80:484–493.

    https://doi.org/10.1016/j.neuron.2013.09.014

    • PubMed
    • Google Scholar
40. 1. Tatsuno M
    2. Tatsuno M
    3. McNaughton BL

    (2007) Fast-forward playback of recent memory sequences in prefrontal cortex during sleep

    Science 318:1147–1150.

    https://doi.org/10.1126/science.1148979

    • PubMed
    • Google Scholar
41. 1. Terada S
    2. Sakurai Y
    3. Nakahara H
    4. Fujisawa S

    (2017) Temporal and rate coding for discrete event sequences in the Hippocampus

    Neuron 94:1248–1262.

    https://doi.org/10.1016/j.neuron.2017.05.024

    • PubMed
    • Google Scholar
42. 1. Thompson LT
    2. Best PJ

    (1990) Long-term stability of the place-field activity of single units recorded from the dorsal hippocampus of freely behaving rats

    Brain Research 509:299–308.

    https://doi.org/10.1016/0006-8993(90)90555-P

    • PubMed
    • Google Scholar
43. 1. van de Ven GM
    2. Trouche S
    3. McNamara CG
    4. Allen K
    5. Dupret D

    (2016) Hippocampal offline reactivation consolidates recently formed cell assembly patterns during sharp Wave-Ripples

    Neuron 92:968–974.

    https://doi.org/10.1016/j.neuron.2016.10.020

    • PubMed
    • Google Scholar
44. 1. Walker MP
    2. Stickgold R

    (2004) Sleep-dependent learning and memory consolidation

    Neuron 44:121–133.

    https://doi.org/10.1016/j.neuron.2004.08.031

    • PubMed
    • Google Scholar
45. Software

    1. Wilson M

    (2008) XClust

    Github.

    https://github.com/wilsonlab/mwsoft64/tree/master/src/xclust
46. 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

    • PubMed
    • Google Scholar
47. 1. Wilson MA
    2. McNaughton BL

    (1994) Reactivation of hippocampal ensemble memories during sleep

    Science 265:676–679.

    https://doi.org/10.1126/science.8036517

    • PubMed
    • Google Scholar
48. 1. Wu CT
    2. Haggerty D
    3. Kemere C
    4. Ji D

    (2017) Hippocampal awake replay in fear memory retrieval

    Nature Neuroscience 20:571–580.

    https://doi.org/10.1038/nn.4507

    • PubMed
    • Google Scholar

Author details

1. Sara E Kee

   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

   Conceptualization, Data curation, Formal analysis, Methodology, Writing—original draft, Writing—review and editing

   Competing interests

   No competing interests declared

2. Xiang Mou

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

   Contribution

   Software, Formal analysis, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8579-7316

3. Huda Y Zoghbi

   1. Department of Neuroscience, Baylor College of Medicine, Houston, United States
   2. Department of Molecular and Human Genetics, Baylor College of Medicine, Houston, United States
   3. Department of Neurology, Baylor College of Medicine, Houston, United States
   4. Department of Pediatrics, Baylor College of Medicine, Houston, United States
   5. Howard Hughes Medical Institute, Baylor College of Medicine, Houston, United States
   6. Jan and Dan Duncan Neurological Research Institute, Texas Children’s Hospital, Houston, United States

   Contribution

   Conceptualization, Supervision, Funding acquisition, Methodology, Project administration, Writing - original draft, Writing—review and editing

   For correspondence

   hzoghbi@bcm.edu

   Competing interests

   Senior editor, eLife.

   "This ORCID iD identifies the author of this article:" 0000-0002-0700-3349

4. 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

   Conceptualization, Supervision, Funding acquisition, Methodology, Project administration, Wriitng - original draft, Writing—review and editing

   For correspondence

   dji@bcm.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-4115-5888

• 1,286

  views

• 206

  downloads

• 21

  citations

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