Prelimbic plasticity was increased by the overexpression of RGS14414 (Navarro-Lobato et al., 2021; Masmudi-Martín et al., 2019; Navarro-Lobato et al., 2022; Masmudi-Martín et al., 2020; López-Aranda et al., 2009; Figure 1A and B) and initially two behavioral experiments were performed using the Object Space task (Genzel et al., 2019). In this task, the condition Overlapping tests for the extraction of an underlying structure across five training trials, while the Stable condition tests for the simple memory of the last experience (Figure 1C). In both experiments, we examined the effect of an interference trial 24 hr after initial training, with object configurations violating previously trained rules, on a test trial 48 hr later (Figure 1D). The design was such that in the Overlapping condition remembering the training resulted in positive discrimination indices at test, in contrast the interference would lead to 0. In the control condition Stable remembering both the training or the interference would result in positive discrimination indices. Performance on the interference trial (Figure 1—figure supplement 1) tests the memory for the original training trials (Genzel et al., 2019). Both groups performed above chance for both conditions, evidence that later differences were not due to differences in the original memory but instead effects of the interference trial.

Figure 1 with 1 supplement see all

Download asset Open asset

At the test trial conducted after interference, RGS14-overexpressing animals exhibited higher discrimination indices in the Stable condition (Figure 1D, p<0.05) but lower in the Overlapping condition (p<0.01) in comparison to controls. The results in the Overlapping condition emphasize that in controls cumulative memory expression was protected from interference, but after increasing plasticity in the cortex interference effects were observed. Further, the simple memory in the Stable condition did not last until test in controls, however after increasing plasticity the memory lasted longer.

To show that these effects were a result of the interference, we conducted three control experiments. Firstly, we performed a behavioral control, in which animals did not experience the interference trial. Secondly, we performed a pharmacological control, in which animals did experience the interference, but any subsequent plasticity-related changes were inhibited in the cortex via the infusion of anisomycin, a protein synthesis inhibitor. In these experiments RGS14-overexpressing animals showed discrimination indices comparable to controls emphasizing the determinant role of the interference trial in producing opposite outcomes. Thirdly, we added an additional control, in which animals did not receive pretraining on the Object Space Task conditions, but instead were only exposed to an object configuration for one-trial. When tested 48 hr later, increased cortical plasticity led to an enhanced one-shot memory performance (Figure 1E, p<0.05).

Together, the behavioral and pharmacological results show that increasing cortical plasticity with RGS14-overexpression caused larger interference effects in a semantic-like memory task. Increased interference in this case was due to better memory for one-trial experiences and local changes in the prelimbic cortex. Thus, our experimental results verify for the first time the hypothesis that lower cortical plasticity is critical to protect previous knowledge from interference effects.

To characterize the build-up of a memory trace and its expression in the Object Space Task, we previously developed (Genzel et al., 2019) a computational model that progressively learns place–object associations and makes decisions about which proportion of time to spend exploring each object in order to minimize uncertainty about these place–object associations. The model employs two parameters: a learning rate α, which determines the balance between recent and remote memories, and a parameter β, which determines the balance between neophilic (preference for more novel object location) and neophobic (aversion for more novel object location) exploratory behaviors. Here, we fitted our model on the behavioral data-set to find for each individual subject the values of the model parameter set α and β that best fit the data. There was no difference in memory expression (β). However, RGS14-overexpressing animals had systematically higher learning rate (α) values (Figure 1F, p<0.05). This indicates that exploration behavior in RGS14-overexpressing animals was driven more by recent than remote memories in contrast to controls.

Thus, the modeling results show that increasing cortical plasticity with RGS14-overexpression caused larger interference effects in a semantic-like memory task due to a higher learning rate.

Sleep is supposedly the price the brain pays for plasticity (Tononi and Cirelli, 2014; Tononi and Cirelli, 2006). The idea is that during a waking episode, learning statistical regularities about the current environment requires strengthening connections throughout the brain. This increases cellular needs for energy and supplies, decreases signal-to-noise ratios, and saturates learning. Therefore, subsequently during sleep, previous waking activity would lead locally to larger delta-waves (1–4 Hz) and spontaneous activity during these oscillations in NonREM sleep should renormalize net synaptic strength and restore cellular homeostasis (Huber et al., 2004).

To test this, after viral-injection rats were implanted with hyperdrives containing 16 tetrodes targeting the hippocampus and prelimbic cortex (Figure 2). We recorded each day 7 hr of neural activity during training as well as sleep in the Object Space task (OS) and compared this to a home cage control (HC). Surprisingly, RGS14-overexpressing animals showed less NonREM sleep (p<0.05, Figure 2C combining all sleep periods in the 7 hr recording day), which can be attributed to shorter bout lengths (p<0.0001, Figure 2D). They also had slightly longer REM bout lengths (p=0.048, sleep stages over time Figure 2—figure supplement 1).

Figure 2 with 3 supplements see all

Download asset Open asset

NonREM sleep is dominated by alternations of on and off-periods of neural activity that can be detected in the LFP signal as delta-waves. We detected delta-waves in the cortex and hippocampus, surprisingly in both groups the hippocampus presented with a higher rate of delta-oscillations (p<0.0001, Figure 2E). In addition, RGS14 showed even higher rates than controls in the hippocampus (p<0.0001). But only in controls, we observed that cortical delta-waves occurred more after learning (OS vs HC, p<0.0001, Figure 2E) while their intrinsic frequency decreased (Figure 2—figure supplement 2). Interestingly, delta amplitudes presented with a different pattern. In controls, delta-waves were smaller in the hippocampus in contrast to the cortex (p<0.0001), but in RGS14-overexpressing animals cortical delta-waves were as small as their respective hippocampal delta-waves and significantly smaller than cortical delta-waves in controls (p<0.0001). In both groups and brain areas delta amplitude increased after learning (all OS vs HC p<0.0001).

Combining this analysis with analysis of neuronal activity (spikes) lets us determine more accurately the length of the respective on and off periods underlying the delta wave. RGS14-overexpressing animals presented with longer off periods than controls (p<0.01). However, known homeostatic changes, where off-period durations decrease over longer sleep periods, remained the same in both groups (Post Trial period X type interaction p<0.0001 but no interaction with group all p>0.8).

In sum, in controls we could confirm the proposition that learning and therefore plastic changes lead to a homeostatic response with increases in delta wave activity. However, we show that artificially increasing plasticity in the prelimbic cortex does not lead to a simple enhancement of this effect. Instead, we observed in RGS14-overexpressing animals that cortical delta-waves become as small as hippocampal delta-waves, off-periods become longer and less NonREM sleep is obtained.

Next, we focused on neuronal firing of individual neurons in the prelimbic cortex. We first determined the firing rate of each neuron during task performance and noticed that RGS14-overexpressing animals showed more neurons with low firing rates (p<0.0001, Figure 2H). When splitting neurons in controls into pentiles according to their firing rate and applying the same margins to RGS14-overexpressing animals, the two lowest firing rate groups represented 40% of the neurons in controls but 75% in RGS14 (p=0.0005, Figure 2I, same for split performed on wake in sleep box or sleep states see Figure 2—figure supplement 2).

Moving to neuronal firing during sleep, only faster-firing neurons showed decreases in firing rate across different wake and sleep states (Figure 2—figure supplement 2) and this was the same for both groups. Spikes were less phase-locked to the slow oscillations in RGS14-overexpressing animals, which was seen for all firing rate groups (Figure 2J, p<0.0001). However, faster-firing neurons contributed the most spiking activity to the upstate and with a decrease of these neurons in RGS14, these animals presented less spikes in the upstate (Figure 2J). This decrease in upstate activity likely underlies the decrease in delta-wave amplitude.

To ensure firing rate differences were not a result of differences in learning itself, firing rates during the period before the first trial was calculated. Most animals did not sleep in this period, thus we focused on wake firing rates and the same difference was present (Figure 2K, p<0.0001). Next, firing rates during the OS trials as well as during the periods directly after the trials (PT 1–8 wake and NREM) were normalized to pre-task rates (Figure 2L). In RGS14 there was a larger increase in firing rates from pre-task in comparison to controls (p<0.0001), further evidence that in RGS14 neurons were more plastic and changed more due to experience. Interestingly, RGS14 generally increased their firing rates over the day and presented with higher rates in comparison to pre-task during all time periods (one-sample t-test to 1 p=0.003–0001). In contrast, for Veh during task there was an increase (p=0.0004) but then during following NonREM a decrease in firing rates (p=0.0005) in comparison to pre-task. There was no change for PT wake. These effects were not seen on home cage days (Figure 2—figure supplement 2).

To summarize, RGS-overexpressing animals had more prelimbic neurons with slower firing rates. These results provide the first causal evidence that increasing synaptic plasticity shifts the neural firing towards the slow firing end of the neural firing spectrum. Furthermore, because it is the faster-firing neurons that dominate upstate spiking activity and therefore delta amplitude, the slowing of firing rates in the more plastic neurons is most likely the cause of the smaller delta waves seen in these animals.

Interactions between the hippocampus and cortex are critical during encoding as well as consolidation of memories (Buzsáki, 1989; Figure 3A). During wake as well as REM sleep these interactions take place in the theta domain and can be measured in theta coherence (Benchenane et al., 2010). In NonREM sleep, they can be captured in the coupling of cortical delta and spindle oscillations with hippocampal ripples (Genzel et al., 2014). Different types of interactions between these three oscillations have been reported; interactions between two oscillations such as delta followed by spindle (Mölle et al., 2002), delta followed by ripple (Peyrache et al., 2009), ripple followed by delta (Maingret et al., 2016), and spindles with a ripple in their troughs (Sirota et al., 2003), but also three-oscillation interactions such as delta followed by spindle with a ripple in the trough (Diekelmann and Born, 2010), delta followed by ripple then spindle (Genzel et al., 2014), ripple followed by delta and then spindle (Maingret et al., 2016). Interestingly, RGS14-overexpressing animals presented with higher hippocampal-cortical theta coherence during both task and REM sleep (Figure 3B p<0.001) as well as increased occurrences of all types of NonREM oscillatory coupling (Figure 3C p<0.001). In controls, we observed an experience-dependent increase (HC vs OS) in theta coherence (p<0.05) as well as NonREM oscillatory coupling rate (p<0.001, also single spindle rate Figure 2—figure supplement 2), confirming the proposed association of these events to learning. This experience-dependent change was absent in RGS14 and higher values were already present in home cage.

Figure 3

Download asset Open asset

Therefore, we could show that increased cortical plasticity led to increased cross-brain interactions during wake and sleep, emphasizing that decreased interactions and decoupling of the cortex from the hippocampus enables the protection of older, cortical memory representations. Further, these findings also highlight that hippocampal-cortical interactions can be regulated top-down by the cortex and not only by other neuromodulating brain areas or the hippocampus as previously assumed.

Next, we focussed on the hippocampal ripple oscillation in NonREM sleep, which is linked to memory reactivation, and therefore is suggested to be a key player in the memory consolidation process (Buzsáki, 2015; Girardeau et al., 2009; Grosmark and Buzsáki, 2016). Increased cortical plasticity led to more, larger and slower ripples (Figure 4A all p<0.0001). Further, learning in comparison to home cage led to a decrease in ripple amplitude and increase in frequency in both animals’ groups (each p<0.0001). In RGS14 ripples were less phase-locked to the slow oscillation (p<0.0001) and cortical delta power around ripples was decreased in comparison to controls (Figure 4B). Surprisingly, granger values around ripples measuring directional connectivity were higher in RGS14-overexpressing animals for Prl→Hpc in the delta frequency range (1–4 Hz) and in the theta/beta range (5–20 Hz) for Hpc→Prl (Figure 4B). The excitatory output of hippocampal ripples can lead to a neuronal response in the cortex (Wierzynski et al., 2009). This response was larger in controls in comparison to RGS14 (Figure 4C p<0.001) because the average activity of ripple responsive neurons was higher (p<0.05). Further, when split into firing-rate groups, it became noticeable that slow firing-neurons (G1) showed no ripple-responses (Figure 4D).

Figure 4 with 3 supplements see all

Download asset Open asset

In sum, increased cortical plasticity also influenced the hippocampal ripple oscillation, ripples became larger, more numerous and the corresponding information flow and cortical neural response was attenuated. Our results imply that there is an important top-down, cortical influence on this oscillation beyond the known local hippocampal mechanisms.

Next, we investigated the different Object Space Task conditions and focused on mechanisms previously proposed by the active systems consolidation theory (Diekelmann and Born, 2010; Genzel et al., 2014) to be key in the consolidation process: cortical response to ripples and oscillation interactions.

In controls Home Cage presented with less ripple-active neurons than the other OS conditions (Figure 5). Interestingly, Overlapping and Random, both conditions in which object-locations change between trials, had more ripple-modulated neurons than the simple-memory condition Stable (p=0.0007). In contrast, after increasing cortical plasticity in RGS14 Home Cage and Stable had the same number of ripple-modulated neurons as the other conditions.

Figure 5

Download asset Open asset

Analyzing the ripple responses revealed that especially activity before the ripple (120–40ms before the peak) differed between conditions and treatments. In controls Overlapping showed the highest activity but all OS conditions differed from home cage (all p<0.05), this was not the case for RGS14.

Next, we examined the fraction of ripples occurring alone or coupled to delta or spindles. In controls, more ripples occurred alone in both the Stable and Home Cage in contrast to Overlapping and Random (p<0.01). However, Overlapping and Random had more ripples following delta waves (large down states, p<0.001). In contrast, in RGS14 Stable and Home Cage had as many ripples following delta as the other conditions, significantly more than the same conditions in controls (p<0.001).

In sum, in controls both Overlapping and Random, conditions with moving locations in the different training trials, had more ripple-modulated neurons and more ripples followed delta-waves than in Stable (simple memory) or Home Cage. In contrast, in RGS14 the same effects were already present in Stable and Home Cage, emphasizing that with increased cortical plasticity all experiences lead to signatures of sleep-related consolidation.

The Object Space Task, applied here, allows to separate different types of learning. First off, the simple memory condition Stable, which has the different objects placed at the same two locations throughout training and at test one new location is used, is similar to classic reference memory such as the watermaze (Samanta et al., 2021; Genzel et al., 2017). Since the different trials have the same configurations, the animal can use both a cumulative memory across trials as well as a recency memory of the last trial to still perform above chance at test. In contrast, in the Overlapping and Random conditions the objects will be presented in different locations from trial to trial, this trial-by-trial novelty should encourage the animal to consolidate these experiences during sleep and abstract the underlying cumulative statistical distributions. In Overlapping this distribution is skewed, such that one location will always contain an object. Thus, we can test for the behavioral expression of the cumulative memory, by presenting the same spatial configuration at test as it was presented during the final trial. Now, only if the animal extracted the cumulative statistics during all training trials, will there be a preference for the less often shown location at test.

By contrasting these different conditions, we can extrapolate which effects are driven by which type of learning or experience. Changes induced by all OS conditions in comparison to Home Cage, should represent more general experience-dependent effects enabling simple memory consolidation or homeostasis (Figure 6 black arrows). Here, these were (1) increased hippocampal-cortical theta-coherence during task and REM, (2) increased delta-amplitude in NonREM and (3) increased spindle and delta rate leading to more oscillatory coupling events. In contrast, changes seen specifically after Overlapping and Random should represent semantic-like memory consolidation with the comparison and integration of new (current trial) and old (previous trials) information (Figure 6 Orange arrows). Here, these were (1) increased number of neurons responsive to ripples and (2) increased fraction of ripples following a delta-wave. Finally, contrasting Overlapping and Random would allow to isolate representations of the different statistical distributions. This would be applicable to for example decoding analysis including individual neural data. Regrettably, we did not record from enough neurons to allow for such an analysis here.

Figure 6

Download asset Open asset

The difference between a general-experience driven increase in delta and spindle as well as most other oscillatory-coupling rates and the specific increase of ripples following delta waves only after complex learning, is especially interesting finding brought to light by the Object Space Task. Different types of interactions between delta-waves, spindles and ripples have been reported and emphasized by different labs. These are interactions between two oscillations such as delta followed by spindle (Mölle et al., 2002), delta followed by ripple (Peyrache et al., 2009), ripple followed by delta (Maingret et al., 2016), and spindles with a ripple in their troughs (Sirota et al., 2003), but also three-oscillation interactions such as delta followed by spindle with a ripple in the trough (Diekelmann and Born, 2010), delta followed by ripple then spindle (Genzel et al., 2014), ripple followed by delta and then spindle (Maingret et al., 2016). Until now, experiments tended to report only on one type of oscillation, making it difficult to compare results and create a framework for potential different or common functions of these different oscillation-couplings.

We had previously already observed that the increase of delta and spindles and their coupling was linked to general experience (also seen after novelty exposure) (Aleman-Zapata et al., 2022) and delta-spindle coupling is most often reported by researchers investigating human subject and applying simple memory paradigms (Diekelmann and Born, 2010; Bastian et al., 2022). This fits to our current results, where general coupling increased after our simple memory condition Stable. Of note, this general increase of delta and spindles also led to an increase of all delta-spindle-ripple coupling types. In contrast ripples following delta coupling increased specifically after the complex learning conditions. Previously this coupling was described together with prelimbic reactivations when animals learn complex rules (Peyrache et al., 2009). It would be tempting to speculate that delta-spindle and spindle with ripple coupling would facilitate simple learning while delta-ripple would correspond to consolidation of complex memories that rely on the comparison of new memories to other previously encoded experiences.

Our hypothesis was that increasing cortical plasticity leads to “overlearning”. Instead of distinguishing salient information and selectively consolidating, all experiences would be treated equal, consolidated and retained on the long-term in the cortex. Therefore, signatures of learning should already be present in control conditions such as Home Cage. We confirmed this hypothesis, both experience and complex-memory dependent learning effects observed in controls were already present in Home Cage in RGS14 (Figure 6 light purple arrows). RGS14-overexpressing animals showed more hippocampal-cortical theta-coherence, increased oscillatory coupling rate and increased fraction of ripples following delta-waves. Thus, in RGS14 encoding and consolidation is already by default ‘turned-on’ instead of being experience or learning dependent. This overlearning would explain the increased interference effects seen behaviorally.

Interestingly, while most signatures of learning were increased, some seemed to show more complex interactions (Figure 6 dark purple arrows). In line with predictions by the SHY hypothesis (Tononi and Cirelli, 2014; Tononi and Cirelli, 2006; Vyazovskiy et al., 2009) RGS14-overexpressing animals showed longer off-periods during NonREM sleep. However, contrary to predictions by the SHY hypothesis (Tononi and Cirelli, 2014; Tononi and Cirelli, 2006), delta waves were smaller in RGS14-overexpressing animals. The cause is likely the general decrease in firing rates accompanying increased plasticity. Slower-firing neurons were less phase locked to the slow oscillation and less neuronal firing in the upstate led to smaller delta waves (Figure 2J).

It has been previously proposed that the brain shows log-normal distributions and slow-firing neurons are more plastic than fast-firing ones (Buzsáki and Mizuseki, 2014). Slow-firing neurons, due to potentially weaker synapses, would be more sensitive to plastic changes (Buzsáki and Mizuseki, 2014). The current evidence, however, was mainly correlational and was only presented for the hippocampus (Grosmark and Buzsáki, 2016). Our results provide the first causal evidence that increasing synaptic plasticity in the cortex shifts the neural firing towards the slow firing end of the neural firing spectrum. Due to insufficient yield of the hippocampal tetrodes, we could not compare cortical to hippocampal firing rates directly in our recordings. However, others have shown faster firing rates in the entorhinal cortex in comparison to the hippocampus, and within the hippocampus faster firing rates in CA1 in comparison to CA3 and DG (Buzsáki and Mizuseki, 2014), potentially mirroring expected differences in plasticity. Interestingly, the change induced in our experiment – slowing of firing rates in RGS14 by one log magnitude – matches the firing rates reported for CA1 in comparison to entorhinal cortex in Buzsáki and Mizuseki, 2014. Thus, our firing-rates would indicate we did what we set out to do: turning the slow learner (cortex) into a fast learner (hippocampal-like neuronal firing). Interestingly, this was also seen in delta-amplitudes, where in RGS14 cortical delta-waves became as small as hippocampal delta-waves. Thus, our results highlight the complex interaction of proposed plasticity-related mechanisms present in firing-rates and sleep oscillations. It seems that slight changes in plasticity – as one would expect to occur during normal learning and experiences – lead to shifts as proposed by the SHY hypothesis. However, larger plasticity differences as expected for different brain areas or induced with artificial manipulations as applied here, lead to much larger shifts – in the case of delta-waves even in the opposite direction as natural learning – and complex interactions.

Interestingly, the change in firing-rate also influenced neuronal response to ripples. Overall ripple response was less in RGS14, despite ripple response per firing-rate group being similar in controls and RGS14. The slowest firing neurons generally did not show large ripple responses in both groups and RGS14 had more of these neurons, which is likely the reason for the overall decreased ripple-response.

While RGS14-overexpressing animals had more ripples and cortical-hippocampal interactions, the response of cortical neurons to ripples remained lower than in controls. The current prevalent view on hippocampal ripple oscillations is that they are essential for memory consolidation. Furthermore, it is believed that they are generated mainly as an outcome of the local synaptic interactions within the hippocampus inducing changes in several cortical areas and supporting the broadcasting of information from the hippocampus to the cortex (Buzsáki, 2015). Our results imply that there is an important top-down, cortical influence on this oscillation beyond the known local hippocampal mechanisms. Such a top-down influence has previously been shown for hippocampal theta sequences and ripples during task execution by inhibiting the prefrontal cortex with DREADDs (Schmidt and Redish, 2021; Schmidt et al., 2019).

Furthermore, and in alignment with existing results, it shows that hippocampus doesn’t reactivate random incoming sensory information, instead it orients itself to the previous knowledge acquired by the cortex (Rothschild et al., 2017; Wang and Ikemoto, 2016). Fitting to this notion, we saw in controls that more hippocampal ripples followed cortical delta-oscillations after the Overlapping and Random condition in comparison to Stable or Home Cage. It has been proposed that the preceding delta-wave would carry the existing information that will then be updated by the incoming ripple information (Genzel et al., 2014). This would be needed in the Overlapping and Random condition to enable comparison of the current configuration to the previous configurations and thus the extraction of the statistical regularities. In RGS14 the control conditions (Home Cage and simple memory Stable) also showed this increase, further underlining the idea that in RGS14 all information is treated as salient independent of its relevance, which then would lead to increased interference effects.

Previous experiments using the plasticity-enhancer RGS14 in the peri-rhinal and visual cortex, showed increased conversion of short-term memory (40 min) to long-term memory (24 hr) (Navarro-Lobato et al., 2021; Masmudi-Martín et al., 2019). As in this study, treatment did not affect other behavioral parameters such as overall exploration times in a simple object exploration paradigm. They could show that overexpression of RGS14 led to increased BNDF at 60 min and 24 hr post training, which led to better long-term memory (Navarro-Lobato et al., 2021). The increase in BDNF in RGS14 resulted in more neurites and dendrite branching in pyramidal and non-pyramidal neurons (Masmudi-Martín et al., 2019). It would be tempting to speculate that the increased dendritic branching led to the decreases in firing rates. Increased branching with more synaptic connections would make it less likely for inputs to be synchronized and thus the neuron would be less likely to fire an action potential.

These previous experiments imply that the main effects of RGS14 on memory are the result of increased cellular plasticity and especially increased BDNF at encoding but also during consolidation. In the current results RGS14 overexpression had multiple effects in the brain, thus we cannot conclude if effects on behavior were directly due to the increase in cellular plasticity in the prelimbic cortex or indirectly induced by changing, for example, sleep-oscillations implicated in memory consolidation. This also is the case for the anisomycin treatment, which inhibits cellular consolidation via inhibition of protein synthesis. We did not record neural activity during the treatment; therefore, we cannot exclude that effects other than the inhibition of protein synthesis could negatively impacted consolidation after the interference trial and have contributed to the results. However, the main aim of the treatment was to impact the consolidation and thus retention of the interference memory, which was achieved.

Of note, with the current approach, we cannot disentangle if we see increased interference effects in RGS14 since they remember the interference-trial better than the original training and thus their behavior is more driven by the interference experience. Or if the effect is caused by the interference memory overwriting and thus eliminating the original training memory. Furthermore, the Granger analysis of directionality should be interpreted with caution since Granger can lead to spurious results especially in oscillating networks (Kispersky et al., 2011).

To conclude, we could show that increasing plasticity in the prelimbic cortex enhances the ability to retain one trial information but this negatively impacts abstracted, cumulative memory representations and the ability to distinguish learning experiences, confirming long-standing but unproven memory system theories. Furthermore, changes in cortical plasticity affect neuronal firing rates, hippocampal-cortical connectivity, cortical delta waves and hippocampal ripples.

A total of 72 rats were used in this experiment: 64 in the behavioral experiment and 8 in the electrophysiological recordings. In each case animals were first extensively handled for multiple days (at least 3) until they did not flinch when the experimenter touched them (see handling videos on https://www.genzellab.com/). Next, all animals underwent viral-injection surgery (see below), half the animals received RGS14₄₁₄-lentivirus while the other had a vehicle (empty) lentivirus. In addition, 32 animals also received pharmacological canula’s targeting prelimbic region during this surgery (16 vehicle, 16 RGS). These and the other 32 behavioral animals had a 2-week surgery recovery and then went on to do the habituation as well as training in the Object Space Task (all conditions counterbalanced within animal). The eight electrophysiology animals (four vehicle, four RGS) received a second surgery 3 weeks after the first one, for hyperdrive implantation. During 2–3 week surgery recovery, tetrodes were slowly lowered to target area before the animals also had habituation and training in the Object Space Task.

Three-month-old male Lister Hooded rats weighing between 300 and 350 g at the experiment start (Charles Rivers, Germany) were used in this study. Rats were pair-housed in conventional eurostandard type IV cages (Techniplastic, UK) in a temperature-controlled (20+2 °C) room following a 12 hr light/dark cycle with water and food provided ad libitum. After lentiviral surgical intervention, animals were single-housed for two days and paired with their cage mates after recovery in rat individually ventilated cages (IVC; Techniplastic, UK). Animals were maintained in their IVC in a barrier room for 14 days before downscaling them to conventional housing conditions. After hyperdrive implantation, rats were single-housed in until the end of the experiment. A total of 72 rats were used in this experiment: 64 in the behavioral experiment (n=16 per group, RGS14₄₁₄, vehicle, across two cohorts of 8–8 animals each; and n=16 per group RGS14₄₁₄-pharma, vehicle-pharma group across two cohorts of 8–8 animals each), and 8 in the electrophysiological recordings (n=4 per treatment, RGS14₄₁₄, and vehicle, split one animal per treatment across four cohorts). The behavioral experiments and electrophysiological recordings were performed during the light period (between 9:00-18:00).

All animal procedures were approved by the Central Commissie Dierproeven (CCD) and conducted according to the Experiments on Animals Act (protocol codes, 2016-014-020 and 2016-014-022).

The RGS14₄₁₄- and vehicle-lentivirus solutions (1.72x10⁷ CFU/ml and 2.75x10⁶ CFU/ml respectively) were prepared and provided by Dr. Zafaruddin Khan at the University of Malaga (Malaga, Spain) (Masmudi-Martín et al., 2019; López-Aranda et al., 2009; ). Briefly, the RGS14₄₁₄ gene (GenBank, AY987041) was cloned into the commercial vector p-LVX DsRed Monomer-C1 (Clontech, France) using DNA recombinant technology. Then, both non-replicant RGS14₄₁₄- and vehicle-lentivirus (empty vector) were prepared and titered using the Lenti-XTM (Clontech, France) according to the manufacturer’s instructions.

The animal’s procedures related to the non-replicant lentiviral solution were approved and carried out in compliance with institutional regulation.

A customized lightweight tetrode micro-drive was manufactured to implant 10 and 6 movable tetrodes in the prelimbic cortex and hippocampus (HPC), respectively (Brunetti et al., 2014; Kloosterman et al., 2009; Nguyen et al., 2009). Two separate bundles of #33 polyimide tubes (Professional Plastics,) were prepared: one of 2 columns x 5 rows for prelimbic cortex and 3X3 for HPC. The bundles were fixed first to the customized 3D printed cannula and then into the customized 3D printed body drive. The 3D printed cannula was designed according to the Rat Brain Atlas in Stereotaxic Coordinates Paxinos, 2013 for the correct placement of the bundles in the areas of interest. Inner tubes (#38 Polyimide tubes; Professional Plastics) were placed inside the outer tubes and glued to the shuttle, which moves through the body spokes thanks to an inox steel screw and a spring CBM011C 08E (Lee spring, Germany). A total of 16 tetrodes were built, twisting four 10 cm polyimide-insulated 12 μm Nickel-Chrome wires (80 turns forward and 40 turns reverse) (Kanthal Precision, Florida) and fused by heat. Tetrodes were loaded in the inner tubes, and their free ends were connected to a customized 64 channels, 24 mm round electrode interface board (EIB) using gold pins (Neuralynx). Previously, 2 NPD dual row 32 contact connectors (Omnetics) had been attached to the EIB. The tetrode tips were cut using fine sharp scissors (maximum length 3.5 mm and 3 mm for prelimbic cortex and HPC, respectively) and fixed to the inner tubes in the upper part. Tetrode tips were clean in distilled water and gold-plated (gold solution, Neuroalynx) using NanoZ software to lower their impedance to 100–200 kΩ and improve the signal-to-noise ratio. The tetrode tips were hidden at the same level as the bundle. The whole drive was covered with aluminum foil connected to the ground to reduce the electrostatic interference during the recordings. The bottom of the micro-drive was deepened in 70% ethanol for 12 hr before brain implantation.

Anisomycin (ANI) powder (Merck, Germany) was solved in 1 M HCl in 0.9% NaCl physiological serum, and the pH was adjusted with 10 μl of 5 M NaOH. Aliquots were prepared and stored at –20 °C until the moment of use. Immediately after the 24 hr test in the object space task, animals were infused with 3 µl/hemisphere of a 25.6 µg/µl ANI solution or the solvent as control at 300 nl/min. An infusion pump and two 10 μl Hamilton syringes connected through a PE10 tube (Plastic1;) to a customized 30 G bilateral internal cannula with a 2 mm projection (Plastic1;) was used for the infusion. The internal cannula was carefully removed after 3 min of diffusion time, and the dummy and cap were placed back. All the animals from Experiment 2 received the infusion of both ANI and vehicle across different weeks for each experimental paradigm.

The Object Space Task (OST) is a newly developed behavioral paradigm to study simple and semantic-like memories in rodents (Genzel et al., 2019). The task is based on the tendency of rodents to explore novel object-location in an open field across multiple trials. In these experiments, the OST took place as described previously (Genzel et al., 2019) at least 21 days after the viral infusion when the effect of the RGS14₄₁₄ protein is observed (Masmudi-Martín et al., 2019; López-Aranda et al., 2009; ). Briefly, animals were handled for 5 consecutive days before and after surgery recovery. Then, rats were accommodated in the experimental room and habituated to the open field across 5 sessions (one per day). In the first session, animals explored the open field with their cagemate for 30 min. In the rest of the session, each individual freely explored the open field for 10 min. Two Duplo objects were included in the open field center in the last two sessions to facilitate a better exploration time in the subsequent task.

The OST consists of two phases: a training phase of five training trials in which animals are exposed to two identical objects (different across trials) for 5 min (45–55 min intertrial time); and a interference/test phase consisting of a single 10 min' trial performed 24 hr and 72 hr after training. In the stable condition, both object locations were fixed during the training trials, and one object location was moved during the interference and tests sessions. In the overlapping condition, one object location was fixed, and the other one moved across training trials. In the interference and test sessions, the same object-location pattern from the last training trial was repeated.

The open field was a wooden square 75x75 x 60 cm. For the task, but not for the habituation, we placed 2D proximal cues on the open field walls and 3D distal cues above the open field. The cues were changed in each experimental session. In addition, the open field base colors changed across task sessions (white, blue, green, brown). The objects used vary in material (plastic, glass, wood, and metal), size, and colors. For electrophysiological recordings, not plastic objects were used to prevent static interference. All object bases were attached to 10 cm x 10 cm metal plates. Circular magnets were installed in the corners underneath the open field floor to prevent object movements during exploration. The open field and object surfaces were cleaned with 40% ethanol between trials to avoid odor biases.

Each trial was recorded using a camera above the open field. The object exploration time was manually scored online using the homemade software 'Scorer32'. The experimenter was blinded for treatment and experimental conditions at the moment of scoring. Object locations and experimental conditions were counterbalanced across treatments, individuals, and sessions.

For electrophysiological recordings, we run four cohorts of 2 animals each. Each cohort included one vehicle- and one RGS14₄₁₄-treated animal, which performed the identical condition sequences with the same object-location patterns. Object locations and experimental conditions were counterbalanced across cohorts. Electrophysiological recordings took place during trials and rest periods (45 min before and 3 h after both training and test; and 45 min intertrial time during the training). Therefore, two brown wooden sleep boxes (40x75 x 60 cm height) with bedding material were placed next to the open field. Animals had been accustomed to the sleep box for at least 3 hr in each open-field habituation session.

Rats involved in the electrophysiological recordings also performed two experimental control conditions: homecage and random. The random condition was carried out as described previously (Genzel et al., 2019), so there was a lack of repetitive object location patterns across different trials. In the homecage, the animal was recorded for 7 hr and 10 min in the sleep box (a whole training session recording), and the experimenter kept the rat awake for the equivalent trial times.

In vivo freely moving extracellular recordings were executed during the OST and the resting periods. One session per experimental condition (homecage, stable, overlapping, and random) was carried out per animal. The local field potential (LFP) and single-unit activity detected by the 64 channels were amplified, filtered, and digitized through two 32 channels chip amplifier headstages (InstanTechnology) connected through the Intan cables and a commutator into the Open Ephys acquisition box. The signal was visualized using the open-source Open Ephys GUI (sample rate 30 kHz). In addition, the headstage contains an accelerometer to record the movement of the animals.

After all the recording sessions, the tetrode-implanted animals received brain electrolytic lesions 48 hr before the transcardial perfusion to identify the electrode tips placement. Thus, a current of 8 µA for 10 s was applied in two wires per tetrode using the stimulator with the animal under isoflurane inhaled anesthesia.

Figures were generated in GraphPad Prism (truncated violin plots with high smoothing).

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

We thank Pelin Özsezer, Barbora Roldanus and Kopal Agarwal for data analysis, Mònica Siuraneta, Vera Mascarenhas Pombeiro Duarte Silva, Jay Chen, Alysha Maurmair, Jet van der Stoep, Joris van Hout, Koen van den Berg and Rian Kraan for performing behavioral experiments. Branco Weiss Fellowship – Society in Science (LG), VIDI NWO (LG), Fundación Alfonso Martín Escudero Fellowship (INL)

All animal procedures were approved by the Central Commissie Dierproeven (CCD) and conducted according to the Experiments on Animals Act (protocol codes, 2016-014-020 and 2016-014-022).

© 2023, Navarro Lobato et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.