Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene Pogz

Margaret M Cunniff; Eirene Markenscoff-Papadimitriou; Julia Ostrowski; John LR Rubenstein; Vikaas Singh Sohal

doi:10.7554/eLife.54835

. 2020 Nov 6;9:e54835. doi: 10.7554/eLife.54835

Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene Pogz

Margaret M Cunniff ¹, Eirene Markenscoff-Papadimitriou ¹, Julia Ostrowski ¹, John LR Rubenstein ¹, Vikaas Singh Sohal ^1,^✉

Editors: Laura L Colgin², Laura L Colgin³

PMCID: PMC7682992 PMID: 33155545

Abstract

Many genes have been linked to autism. However, it remains unclear what long-term changes in neural circuitry result from disruptions in these genes, and how these circuit changes might contribute to abnormal behaviors. To address these questions, we studied behavior and physiology in mice heterozygous for Pogz, a high confidence autism gene. Pogz^+/- mice exhibit reduced anxiety-related avoidance in the elevated plus maze (EPM). Theta-frequency communication between the ventral hippocampus (vHPC) and medial prefrontal cortex (mPFC) is known to be necessary for normal avoidance in the EPM. We found deficient theta-frequency synchronization between the vHPC and mPFC in vivo. When we examined vHPC–mPFC communication at higher resolution, vHPC input onto prefrontal GABAergic interneurons was specifically disrupted, whereas input onto pyramidal neurons remained intact. These findings illustrate how the loss of a high confidence autism gene can impair long-range communication by causing inhibitory circuit dysfunction within pathways important for specific behaviors.

Research organism: Mouse

Introduction

Mutations in Pogz have been identified in over forty patients with autism spectrum disorder (ASD) (Fukai et al., 2015; Hashimoto et al., 2016; Iossifov et al., 2012; Iossifov et al., 2014; Neale et al., 2012; Stessman et al., 2016; Zhao et al., 2019), intellectual disability (Dentici et al., 2017; Fitzgerald et al., 2015; Gilissen et al., 2014; Tan et al., 2016; White et al., 2016; Ye et al., 2015), and schizophrenia (Fromer et al., 2014; Gulsuner et al., 2013). Most of these are de novo mutations presumed to cause loss of function. Such de novo loss of function mutations are exceedingly rare in controls, ranking Pogz among the highest confidence genes for ASD (FDR < 0.01) (Sanders et al., 2015). POGZ is known to play a role in chromatin regulation, mitotic progression, and chromosome segregation (Nozawa et al., 2010). ASD associated mutations have been shown to disrupt POGZ’s DNA-binding activity (Matsumura et al., 2016) and reduce neurite outgrowth in vitro (Hashimoto et al., 2016; Zhao et al., 2019).

Among the highest confidence ASD-associated genes, there is a striking enrichment for genes which, like Pogz, are involved in chromatin remodeling (Cotney et al., 2015; De Rubeis et al., 2014; Krumm et al., 2014; Sanders et al., 2015). One hypothesis is that this enrichment reflects the developmental complexity of the nervous system, which renders the brain more vulnerable than other systems to regulatory disruptions (Ronan et al., 2013; Suliman et al., 2014). This hypothesis is supported by the convergent expression of genes associated with neurodevelopmental disease at specific developmental timepoints (Gulsuner et al., 2013; Willsey et al., 2013). Despite this progress in identifying ASD-associated genes and their convergence onto specific developmental processes, we do not yet understand how these genetic disruptions cause behavioral phenotypes, nor what mechanisms in the developed brain might be targeted to normalize behavior. This is because it remains unclear what long-term changes in neural circuitry result from these genetic disruptions, and how they might contribute to the abnormal functioning of the developed brain.

In order to further understand the nature of neural network dysfunction that results from genetic disruptions and altered development, we characterized behavior and physiology in adult Pogz heterozygous loss of function (Pogz^+/-) mice. We found that these mice exhibit altered behavior in a well-studied assay of anxiety-related avoidance, the elevated plus maze (EPM). Interestingly, this is similar to a recently published study which found evidence for decreased anxiety in Pogz mutants using the open field test (Matsumura et al., 2020). We then studied communication between the ventral hippocampus (vHPC) and medial prefrontal cortex (mPFC), which is known to be necessary for normal anxiety-related avoidance in the EPM. We found that theta-frequency synchronization between the vHPC and mPFC is decreased in vivo. In vitro, we found a specific loss of excitatory synaptic drive from the vHPC onto prefrontal GABAergic interneurons. Notably excitatory input from the vHPC onto prefrontal pyramidal neurons was spared.

Two major hypotheses about the pathophysiology of ASD are that developmental disruptions can lead to (1) persistent dysfunction of cortical GABAergic circuits (Nelson and Valakh, 2015; Rubenstein and Merzenich, 2003; Sohal and Rubenstein, 2019), and (2) impairments in long-range communication (Kana et al., 2014). Our findings illustrate a case in which these two mechanisms may be linked following the heterozygous loss of a high confidence ASD gene – specifically, deficient long-range communication is associated with an impairment in inhibitory circuits. Our findings also suggest that feedforward inhibition (not just feedforward excitation) in the vHPC-mPFC pathway may play an important role in anxiety-related avoidance behavior.

Results

Pogz^+/- mice have decreased anxiety-related avoidance in the EPM

To characterize their behavioral phenotypes, we tested Pogz^+/- mice using a battery of standard behavioral assays. We found a reduction in anxiety-related avoidance in the EPM (Figure 1A,B). Rodents typically avoid the center and open arms of the EPM, because they are exposed, brightly lit, and raised off the ground, and instead spend the bulk of their time in the closed arms. However, Pogz^+/- mice spent significantly more time exploring the open arms and center region of the EPM compared to their wildtype littermates (Figure 1C,E; ratio of open vs. closed arm time: p=0.003; open time: p=0.001, center time: p=0.02. Wilcoxon rank sum, WT N = 18, Het N = 27). The total distance traveled during the assay was not different between genotypes, suggesting that this increase in open arm exploration is not simply an artefact related to changes in overall exploratory behavior (Figure 1D, p=0.35, Wilcoxon rank sum, WT N = 16, Het N = 23). Pogz heterozygotes also made more head dips in the EPM than their wildtype littermates, consistent with the interpretation that their phenotype reflects a decrease in anxiety-related behavior and a corresponding increase in active exploration (Figure 1F, p=0.03, Wilcoxon rank sum, WT N = 14, Het N = 14). There was no difference in the number of open arm entries between genotypes, but individual open arms visits were longer in duration in Pogz^+/- mice (Figure 1G,H; number of entries: p=0.32; duration of entries: p=0.047, Wilcoxon rank sum, WT N = 16, Het N = 23). We also confirmed that increases in open arm exploration and head dips were not driven simply by sex differences (Figure 1—figure supplement 1). The performance of Pogz heterozygotes did not differ from that of wild-type mice on cognitive tests including an odor-texture rule shifting task (Cho et al., 2015; Ellwood et al., 2017) and a T-maze based delayed nonmatch-to-sample task (Spellman et al., 2015; Tamura et al., 2017). This indicates that their altered behavior in the EPM was not related to nonspecific impairments in spatial cognition or learning (Figure 1—figure supplement 2).

(**A, B**) Occupancy plot for a 15 min EPM session for a representative wildtype (A) and *Pogz*^+/- (B) mouse. (C) Ratio of time spent in open vs. closed arms of the EPM. Wilcoxon rank-sum test, U = −2.8857, p=0.003, WT N = 18, Het N = 27. (D) Total distance traveled during EPM sessions. Wilcoxon rank-sum test U = −1.9434, p=0.35, WT N = 16, Het N = 23. (E) Total time spent in exposed areas of EPM, open arms: statistic = −3.0753, p=0.001, center: U = −2.2112, p=0.02. Wilcoxon rank-sum test, WT N = 18, Het N = 27. (F) Total number of head dips for each mouse, U = −1.9434, p=0.03. Wilcoxon rank-sum test, WT N = 14, Het N = 14. (G) Number of open arm entries, U = −0.9993, p=0.32. Wilcoxon rank-sum test, WT N = 16, Het N = 23. (H) Average duration of each open arm visit, U = −1.984, p=0.047. Wilcoxon rank-sum test, WT N = 16, Het N = 23.

Figure 1—source data 1. Source data for Figure 1.

elife-54835-fig1-data1.csv^{(11.5KB, csv)}

Figure 1—figure supplement 1. — (**A, B**) Occupancy plot for a 15 min EPM session for a representative wildtype (A) and *Pogz*^+/- (B) mouse. (C) Ratio of time spent in open vs. closed arms of the EPM. Wilcoxon rank-sum test, U = −2.8857, p=0.003, WT N = 18, Het N = 27. (D) Total distance traveled during EPM sessions. Wilcoxon rank-sum test U = −1.9434, p=0.35, WT N = 16, Het N = 23. (E) Total time spent in exposed areas of EPM, open arms: statistic = −3.0753, p=0.001, center: U = −2.2112, p=0.02. Wilcoxon rank-sum test, WT N = 18, Het N = 27. (F) Total number of head dips for each mouse, U = −1.9434, p=0.03. Wilcoxon rank-sum test, WT N = 14, Het N = 14. (G) Number of open arm entries, U = −0.9993, p=0.32. Wilcoxon rank-sum test, WT N = 16, Het N = 23. (H) Average duration of each open arm visit, U = −1.984, p=0.047. Wilcoxon rank-sum test, WT N = 16, Het N = 23.

Figure 1—source data 1. Source data for Figure 1.

elife-54835-fig1-data1.csv^{(11.5KB, csv)}

Pogz^+/- mice have reduced hippocampal-prefrontal theta synchrony

Many studies, including work from our lab, have shown that communication between the vHPC and medial prefrontal cortex (mPFC), is necessary for anxiety-related avoidance in the EPM, and that theta-frequency synchronization between these structures can serve as a biomarker for this communication (Adhikari et al., 2010; Adhikari et al., 2011; Jacinto et al., 2016; Kjaerby et al., 2016; Lee et al., 2019; Padilla-Coreano et al., 2016; Padilla-Coreano et al., 2019). Based on this, we recorded local field potentials from the mPFC and vHPC to assess hippocampal-prefrontal theta synchrony in Pogz^+/- mice (Figure 2A). Because we were specifically interested in hippocampal-theta prefrontal synchrony, we limited analysis to mice which had clearly visible theta-frequency peaks in vHPC power spectra recorded during periods of locomotion, and had electrodes located within the mPFC and vHPC (based on post-hoc histology). Previous work has shown that vHPC-mPFC theta synchrony is dynamically modulated in different compartments of the EPM (Adhikari et al., 2010; Jacinto et al., 2016). Consistent with these earlier findings, in wild-type mice, vHPC-mPFC theta synchrony increased as mice approached the center of the EPM. This has previously been interpreted to reflect movement from a less-anxiogenic to more anxiogenic location, as well as the approach to a choice point where mice must decide whether to avoid or explore the open arms (Adhikari et al., 2010; Jacinto et al., 2016). We measured synchrony between the vHPC and mPFC using the weighted-phase locking index (WPLI) (Vinck et al., 2011) and found that the increase in theta synchrony, which normally occurs as mice approach the center of the EPM, was conspicuously absent in Pogz heterozygous mice, (Figure 2B; p=0.00039 for genotype X timepoint as a fixed factor in a linear mixed model; difference in theta synchrony at the time of center approach: p=0.001). Pogz^+/- mice also had overall reduced vHPC-mPFC theta synchrony while in the EPM, as compared to wild-type littermates (Figure 2C; 2-way ANOVA with genotype and open vs. closed arms as factors, significant effect of genotype, p=0.03). In fact, Pogz^+/- mice had reduced theta synchrony at baseline, in the home cage (Figure 2D; p=0.03). There were no differences in power in the vHPC or mPFC between Pogz^+/- mice and wildtypes, suggesting that this change in synchrony reflects altered communication between these brain regions, not just reduced activity in one or both structures (Figure 2—figure supplement 1).

(A) Recording schematic and examples of raw local field potential traces. (B) Z-scored theta band weighted-phase locking index (WPLI) as mice approach the center of the EPM. Linear mixed effects model using timepoint (−3,–1.5, 0, and +1.5 s relative to center zone entry), genotype, mouse, and timepoint X genotype interaction as fixed factors and individual run as a random factor, p=0.00039 for timepoint X genotype interaction, t-statistic = −3.55, DF = 2355 for fixed factors, n = 274 and 316 closed-center runs from 7 WT mice and 6 Het mice, respectively. Wilcoxon rank-sum test for t = 0: U = 3.3738, p=0.0007, for t = 1.5: U = 2.0275, p=0.043 (n = 274 closed-center runs from 7 WT mice and 316 from 6 Het mice). (C) Average theta band WPLI in the open vs. closed arms of the EPM. Two-way ANOVA including arm and genotype as factors - significant effect of genotype: p=0.03 (d.f. = 1, N = 6 WT and 7 Het mice, F = 5.66). (D) Theta band WPLI for mice in their homecages: U = 2.2417, p=0.031 (Wilcoxon rank-sum with N = 6 WT and 7 Het mice).

Figure 2—source data 1. Source data for Figure 2.

elife-54835-fig2-data1.csv^{(25.2KB, csv)}

Figure 2—figure supplement 1. — (A) Recording schematic and examples of raw local field potential traces. (B) Z-scored theta band weighted-phase locking index (WPLI) as mice approach the center of the EPM. Linear mixed effects model using timepoint (−3,–1.5, 0, and +1.5 s relative to center zone entry), genotype, mouse, and timepoint X genotype interaction as fixed factors and individual run as a random factor, p=0.00039 for timepoint X genotype interaction, t-statistic = −3.55, DF = 2355 for fixed factors, n = 274 and 316 closed-center runs from 7 WT mice and 6 Het mice, respectively. Wilcoxon rank-sum test for t = 0: U = 3.3738, p=0.0007, for t = 1.5: U = 2.0275, p=0.043 (n = 274 closed-center runs from 7 WT mice and 316 from 6 Het mice). (C) Average theta band WPLI in the open vs. closed arms of the EPM. Two-way ANOVA including arm and genotype as factors - significant effect of genotype: p=0.03 (d.f. = 1, N = 6 WT and 7 Het mice, F = 5.66). (D) Theta band WPLI for mice in their homecages: U = 2.2417, p=0.031 (Wilcoxon rank-sum with N = 6 WT and 7 Het mice).

Figure 2—source data 1. Source data for Figure 2.

elife-54835-fig2-data1.csv^{(25.2KB, csv)}

Notably, WPLI is unsigned, that is, it measures phase locking using the magnitudes of the imaginary component of the phase difference. However, when we examined the signs of these phase differences, we found that when mice were in the open arms, for 5/6 wild-type mice and 6/6 Pogz heterozygous mice, the imaginary component of the phase difference was above the x-axis in the complex plane, indicating that hippocampal activity tends to lead prefrontal activity.

An unbiased, data-driven approach to examine the significance of vHPC-mPFC theta synchrony for normal behavior and Pogz^+/- mice

As noted earlier, many studies have focused on vHPC-mPFC theta synchrony as a potential biomarker for vHPC-mPFC communication that is relevant to anxiety-related behaviors. As described above, we found deficits in vHPC-mPFC theta synchrony that correlate with deficits in anxiety-related avoidance behaviors in Pogz^+/- mice. However, perhaps this is simply a case of the streetlight effect. I.e., perhaps there are alternative patterns of activity within the hippocampal-prefrontal circuit that are also engaged during EPM exploration, but which remain largely intact in Pogz^+/- mice. In this context, multiple studies from the Dzirasa laboratory and one from ours have shown that data-driven approaches can uncover patterns of rhythmic activity across limbic networks (‘electomes’ or ‘intrinsic coherence networks’) which correlate with, and potentially predict, aspects of emotional behaviors (Hultman et al., 2016; Hultman et al., 2018; Kirkby et al., 2018). Can this kind of data-driven approach identify hippocampal-prefrontal networks that are engaged by EPM exploration, and if so, would these be intact or deficient in Pogz^+/- mice?

To address this question, we took a data-driven approach to identify salient features within LFP recordings, relate these to EPM behavior, and assess them in Pogz^+/- mice. A combination of principal components analysis (PCA) (to compute dimensionality) and independent components analysis (ICA) (to reduce dimensionality) was applied (Methods) to a broad list of potential LFP features for all mice (Table 1, Table 2). These features comprise power (within each region), synchrony (between regions), and cross-frequency coupling (within or between regions), across multiple frequency bands. Each independent component (ICs) discovered in this way was defined by a set of weights for each feature (Figure 3A; 80 total ICs derived from 13 mice). To identify similar ICs that were conserved across mice and thus likely to be biologically meaningful, we calculated the correlation coefficient between all pairs of ICs (Figure 3B), then applied a threshold to this pairwise correlation matrix to identify pairs of highly similar ICs (Figure 3C). We then performed clustering on this dataset (Methods) to identify characteristic ICs that appear repeatedly across mice. One such cluster was characterized by strong weights for cross-frequency (phase-amplitude) coupling between hippocampal-theta and higher-frequency activity in either the vHPC or mPFC (Figure 3D). In other words, this cluster corresponds to a ‘network’ that is conserved across mice. When activity in this network goes up, it means that the hippocampal-theta rhythm more strongly modulates the amplitude of beta and gamma-frequency activity in both the hippocampus and prefrontal cortex.

Table 1. Single frequency LFP measures used as features in PCA/ICA analysis.

Measure	Region	Frequencies
Power	HPC	Theta (4–12 Hz)
		Beta (13–30 Hz)
		Low Gamma (30–55 Hz)
		High Gamma (65–100 Hz)
	PFC	Theta (4–12 Hz)
		Beta (13–30 Hz)
		Low Gamma (30–55 Hz)
		High Gamma (65–100 Hz)
Amplitude Covariation	HPC-PFC	Theta (4–12 Hz)
		Beta (13–30 Hz)
		Low Gamma (30–55 Hz)
		High Gamma (65–100 Hz)
Weighted-Phase Locking	HPC-PFC	Theta (4–12 Hz)
		Beta (13–30 Hz)
		Low Gamma (30–55 Hz)

Open in a new tab

High Gamma (65–100 Hz).

Table 2. Multiple frequency LFP measures used as features in PCA/ICA analysis.

Measure	Regions	Frequencies
Cross-Frequency Coupling	HPC (low) → PFC (high)	Theta (2–6 Hz) → Beta (13–30 Hz)
		Theta (2–6 Hz) → Low Gamma (30–55 Hz)
		Theta (2–6 Hz) → High Gamma (65–100 Hz)
		Alpha (6–10 Hz) → Beta (13–30 Hz)
		Alpha (6–10 Hz) → Low Gamma (30–55 Hz)
		Alpha (6–10 Hz) → High Gamma (65–100 Hz)
	PFC (low) → HPC (high)	Theta (2–6 Hz) → Beta (13–30 Hz)
		Theta (2–6 Hz) → Low Gamma (30–55 Hz)
		Theta (2–6 Hz) → High Gamma (65–100 Hz)
		Alpha (6–10 Hz) → Beta (13–30 Hz)
		Alpha (6–10 Hz) → Low Gamma (30–55 Hz)
		Alpha (6–10 Hz) → High Gamma (65–100 Hz)
	HPC (low) → HPC (high)	Theta (2–6 Hz) → Beta (13–30 Hz)
		Theta (2–6 Hz) → Low Gamma (30–55 Hz)
		Theta (2–6 Hz) → High Gamma (65–100 Hz)
		Alpha (6–10 Hz) → Beta (13–30 Hz)
		Alpha (6–10 Hz) → Low Gamma (30–55 Hz)
		Alpha (6–10 Hz) → High Gamma (65–100 Hz)
	PFC (low) → PFC (high)	Theta (2–6 Hz) → Beta (13–30 Hz)
		Theta (2–6 Hz) → Low Gamma (30–55 Hz)
		Theta (2–6 Hz) → High Gamma (65–100 Hz)
		Alpha (6–10 Hz) → Beta (13–30 Hz)
		Alpha (6–10 Hz) → Low Gamma (30–55 Hz)

Open in a new tab

Alpha (6–10 Hz) → High Gamma (65–100 Hz).

Figure 3. — (A) Example weight vectors showing how various LFP features (x-axis) contribute to different independent components (ICs) in one mouse. The y-axis shows the weight of each feature. (B) Correlation matrix showing the similarity of weight vectors corresponding to different ICs, from all mice. (C) Binarized version of the correlation matrix showing pairs of ICs that have a correlation coefficient > 0.7. (D) Example weights vectors (light, colored traces) for ICs from one cluster. This cluster is characterized by strong weights for cross-frequency coupling between vHPC theta activity and higher-frequency activity in either vHPC or mPFC. The bold black trace shows the average of these weight vector. (F) The projection of network activity onto the characteristic (averaged) weight vector (from E) as a function of time during approaches to the center of the EPM, for wild-type or *Pogz*^+/- mice. As mice approach the center, activity in this characteristic IC rises sharply and reaches a peak in WT mice, but this is absent in *Pogz*^+/- mice. Linear mixed effects model using timepoints (t = 0 vs. baseline based on the average of the first/last points), mouse, genotype, and timepoint X genotype interaction as fixed factors, and individual runs as random factors, timepoint X genotype interaction p=0.01, DF = 147, t-statistic = 2.60; Wilcoxon rank-sum test for t = 0: p=0.007, U = 2.6864; n = 39 closed-center-open runs from 6 WT mice and 37 runs from 7 Het mice.

Figure 3—figure supplement 1. — (A) Example weight vectors showing how various LFP features (x-axis) contribute to different independent components (ICs) in one mouse. The y-axis shows the weight of each feature. (B) Correlation matrix showing the similarity of weight vectors corresponding to different ICs, from all mice. (C) Binarized version of the correlation matrix showing pairs of ICs that have a correlation coefficient > 0.7. (D) Example weights vectors (light, colored traces) for ICs from one cluster. This cluster is characterized by strong weights for cross-frequency coupling between vHPC theta activity and higher-frequency activity in either vHPC or mPFC. The bold black trace shows the average of these weight vector. (F) The projection of network activity onto the characteristic (averaged) weight vector (from E) as a function of time during approaches to the center of the EPM, for wild-type or *Pogz*^+/- mice. As mice approach the center, activity in this characteristic IC rises sharply and reaches a peak in WT mice, but this is absent in *Pogz*^+/- mice. Linear mixed effects model using timepoints (t = 0 vs. baseline based on the average of the first/last points), mouse, genotype, and timepoint X genotype interaction as fixed factors, and individual runs as random factors, timepoint X genotype interaction p=0.01, DF = 147, t-statistic = 2.60; Wilcoxon rank-sum test for t = 0: p=0.007, U = 2.6864; n = 39 closed-center-open runs from 6 WT mice and 37 runs from 7 Het mice.

For each mouse, we could calculate the time-varying activity of this IC by convolving the weights of this IC (averaged across mice) with the time series of each feature. When mice approach the center of the EPM during closed arm-center-open arm runs, the activity of this IC shows a pattern similar to what we previously observed for vHPC-mPFC theta synchrony. Specifically, in wild-type mice, the activity of this IC increased as mice ran approached the center zone. Strikingly, this behavioral modulation of network activity was once again absent in Pogz heterozygotes (Figure 3E). Thus, this unbiased approach validated the general finding we made earlier, when we focused on a specific metric of vHPC–mPFC theta synchrony. Again we found a pattern of activity, related to theta-frequency synchronization across the hippocampal-prefrontal circuit (in this case, measured by the modulation of higher-frequency activity), normally correlates with entries into more anxiogenic regions of the EPM, but this relationship is abolished in Pogz heterozygotes.

We found a total of three clusters, corresponding to three characteristic ICs that were conserved across mice. The average weights for each of these characteristic ICs, as well as the pattern of activity for each during approaches to the center of the EPM, are plotted in Figure 3—figure supplement 1. As described above, one characteristic IC represents coupling between vHPC theta phase and the amplitude of higher-frequency vHPC or mPFC activity. Another characteristic IC represents coupling between mPFC theta phase and the amplitude of higher-frequency vHPC or mPFC activity. Notably, activity in the latter characteristic IC was not appreciably modulated during approaches to the center of the EPM. The third characteristic IC represents broadband vHPC and mPFC power. We also did not find a characteristic IC corresponding to coupling between alpha phase and higher-frequency activities. These observations support our finding that theta-frequency communication between the hippocampus and downstream structures such as the PFC is behaviorally modulated, and that the normal pattern of modulation is disrupted in Pogz mutant mice. Notably, this finding is specific for both frequency band and anatomical pathway, as we did not find conserved clusters of ICs corresponding to cross-frequency coupling outside the theta band and did not observe behavioral modulation for the IC which represents coupling of vHPC activity to mPFC theta.

vHPC excitation of mPFC interneurons is deficient in Pogz^+/- mice

Impaired synchrony suggests a deficit in the transmission of neural activity from the vHPC to mPFC. This could reflect local deficits within these structures, and/or altered synaptic connections between them. To explore potential factors underlying this impaired synchrony, we made patch clamp recordings from neurons in the prefrontal cortex. The resting membrane potential, input resistance, and action potential properties of pyramidal cells and interneurons were not grossly different between Pogz^+/- mice and wild-type littermates (Figure 4—figure supplement 1, Figure 5—figure supplement 1). To assess synaptic communication between the vHPC and mPFC, we injected virus encoding CamKII-ChR2-EYFP into the vHPC, then, after waiting 8 weeks for viral expression, recorded optically-evoked responses in the mPFC. Optogenetic stimulation was delivered at 8 Hz, to specifically focus on theta-frequency transmission. We recorded both excitatory currents and optically-evoked spikes (Figure 4).

(**A, B**) Representative examples of optically-evoked excitatory post-synaptic currents (oEPSCs) recorded from prefrontal FSINs in wildtype (A) or *Pogz*^+/- mice (B). (**C, D**) Representative traces of optically-evoked excitatory post-synaptic potentials (oEPSPs) and action potentials recorded from FSINs in wildtype (C) or *Pogz*^+/- mice (D). (E) The total oEPSC charge in FSINs is reduced in *Pogz*^+/- mice, U = 2.7652, p=0.006. (F) The paired- pulse ratio (PPR) for oEPSCs is reduced in *Pogz*^+/- FSINs, U = 2.128, p=0.03. (G) The latency of the first optically-evoked action potential is increased in *Pogz*^+/- FSINs, U = −2.490, p=0.013. (H) The number of action potentials elicited by oEPSPs is non-significantly altered, U = 1.766, p=0.08. In **E-H**, different hues correspond to specific mice, and squares indicate datapoints from cells that were used for the representative traces shown in **A-D**. All p-values from Wilcoxon rank sum, WT N = 6 animals, n = 11 cells. Het N = 3 animals, n = 7 cells.

Figure 4—source data 1. Source data for Figure 4.

elife-54835-fig4-data1.csv^{(4.2KB, csv)}

Figure 4—figure supplement 1. — (**A, B**) Representative examples of optically-evoked excitatory post-synaptic currents (oEPSCs) recorded from prefrontal FSINs in wildtype (A) or *Pogz*^+/- mice (B). (**C, D**) Representative traces of optically-evoked excitatory post-synaptic potentials (oEPSPs) and action potentials recorded from FSINs in wildtype (C) or *Pogz*^+/- mice (D). (E) The total oEPSC charge in FSINs is reduced in *Pogz*^+/- mice, U = 2.7652, p=0.006. (F) The paired- pulse ratio (PPR) for oEPSCs is reduced in *Pogz*^+/- FSINs, U = 2.128, p=0.03. (G) The latency of the first optically-evoked action potential is increased in *Pogz*^+/- FSINs, U = −2.490, p=0.013. (H) The number of action potentials elicited by oEPSPs is non-significantly altered, U = 1.766, p=0.08. In **E-H**, different hues correspond to specific mice, and squares indicate datapoints from cells that were used for the representative traces shown in **A-D**. All p-values from Wilcoxon rank sum, WT N = 6 animals, n = 11 cells. Het N = 3 animals, n = 7 cells.

Figure 4—source data 1. Source data for Figure 4.

elife-54835-fig4-data1.csv^{(4.2KB, csv)}

Fast-spiking interneurons (FSINs) in Pogz heterozygotes showed a marked reduction in excitatory synaptic input from vHPC projections, including a ~ 50% reduction in total charge (Figure 4, p=0.006, Wilcoxon rank sum WT N = 6, n = 11, Het N = 3, n = 7). Short term plasticity of these excitatory synapses onto FSINs also exhibited a shift toward greater depression as evidenced by a decrease in the paired-pulse ratio (PPR) (Figure 4, p=0.03, Wilcoxon rank sum WT N = 6, n = 11, Het N = 3, n = 7). In current clamp recordings, these FSINs exhibited a much longer latency to spike following each light flash (Figure 4, p=0.01, Wilcoxon rank sum WT N = 6, n = 11, Het N = 3, n = 7). There was a trend toward an overall reduction in spiking which did not reach statistical significance (Figure 4, p=0.08, Wilcoxon rank sum WT N = 6, n = 11, Het N = 3, n = 7). Notably, all of these changes were specific to FSINs. In recordings from pyramidal neurons, we did not observe any changes in the size or PPR of optogenetically evoked synaptic currents, nor in the latency or number of optogenetically evoked spikes (Figure 5).

(**A, B**) Representative examples of optically-evoked excitatory post-synaptic currents (oEPSCs) recorded from prefrontal pyramidal neurons in wildtype (A) or *Pogz*^+/- mice (B). (**C, D**) Optically-evoked excitatory post-synaptic potentials (oEPSPs) and action potentials in wildtype (C) or *Pogz*^+/- (D) pyramidal neurons. (E) Total oEPSC charge in pyramidal neurons, U = 1.0736, p=0.28. (F) Paired-pulse ratio for oEPSCs in pyramidal neurons, U = 1.4347, p=0.15 (G) Latency to first optically-evoked action potential in pyramidal neurons, U = −0.305, p=0.76. (H) Number of action potentials elicited by oEPSPs in pyramidal neurons, U = 0.2822, p=0.78. In **E-H**, different hues correspond to specific mice, and squares indicate datapoints from cells that were used for the representative traces shown in **A-D**. All p-values from Wilcoxon rank sum, WT N = 13 animals, n = 17 cells. Het N = 8 animals, n = 11 cells.

Figure 5—source data 1. Source data for Figure 5.

elife-54835-fig5-data1.csv^{(7.3KB, csv)}

Figure 5—figure supplement 1. — (**A, B**) Representative examples of optically-evoked excitatory post-synaptic currents (oEPSCs) recorded from prefrontal pyramidal neurons in wildtype (A) or *Pogz*^+/- mice (B). (**C, D**) Optically-evoked excitatory post-synaptic potentials (oEPSPs) and action potentials in wildtype (C) or *Pogz*^+/- (D) pyramidal neurons. (E) Total oEPSC charge in pyramidal neurons, U = 1.0736, p=0.28. (F) Paired-pulse ratio for oEPSCs in pyramidal neurons, U = 1.4347, p=0.15 (G) Latency to first optically-evoked action potential in pyramidal neurons, U = −0.305, p=0.76. (H) Number of action potentials elicited by oEPSPs in pyramidal neurons, U = 0.2822, p=0.78. In **E-H**, different hues correspond to specific mice, and squares indicate datapoints from cells that were used for the representative traces shown in **A-D**. All p-values from Wilcoxon rank sum, WT N = 13 animals, n = 17 cells. Het N = 8 animals, n = 11 cells.

Figure 5—source data 1. Source data for Figure 5.

elife-54835-fig5-data1.csv^{(7.3KB, csv)}

Deficient FSIN excitation impairs information transmission across vHPC-mPFC circuits

Excitatory and inhibitory postsynaptic currents are major contributors to LFPs (Buzsáki et al., 2012). Thus, a major deficit in synaptic currents evoked by hippocampal inputs could explain the reductions in synchronization between vHPC and mPFC LFPs that we observed. But how might this synaptic deficit in Pogz^+/- mice explain their decreased avoidance of the open arms in the EPM? As discussed above, the transmission of information from the vHPC to mPFC is necessary for open arm avoidance. We hypothesized that a decrease in excitatory drive onto FSINs could impair the PFC’s ability to appropriately filter information, reducing the transmission of information from the vHPC to mPFC, and resulting in the decreased open arm avoidance seen in Pogz heterozygotes. Specifically, we hypothesized that because ventral hippocampal input to the mPFC is rhythmically modulated, feedforward inhibition might preferentially suppress the responses of prefrontal neurons to out-of-phase ‘noise’ while sparing hippocampally-driven responses.

To test the plausibility of this hypothesis, we constructed a simple computational model composed of 2 integrate-and-fire neurons – a FSIN and an output neuron (i.e. a pyramidal cell). Both cells received the same two sources of synaptic input – ‘noise,’ generated by a Poisson process with constant rate, and ‘hippocampal input,’ which was modeled as a Poisson process whose rate varied according to the theta rhythm, that is, was modulated at 8 Hz (Figure 6A). Both cells had the same thresholds and membrane time constants, and we set the time constants of decay for EPSPs and IPSPs to 8 and 20 msec, respectively, to reflect the typically longer timescales for synaptic inhibition. The rate of hippocampal inputs varied sinusoidally between 0 and 100 Hz, and the rate of noise inputs was constant at the midpoint of this distribution (50 Hz). Pyramidal neuron spiking ranged from ~0 to 50 Hz, whereas FSIN spiking ranged from ~0 to 150 Hz. Finally, we explored how varying the strength of excitatory input from both hippocampal and noise inputs onto FSINs affected the transmission of information from the vHPC to mPFC. Specifically, we quantified the correlation between hippocampal input and mPFC output spikes, as well as between the noise input and mPFC output spikes, while varying a single parameter which represents the EPSP amplitude that each hippocampal or noise spike elicits in the FSIN.

Figure 6—figure supplement 1. — (A) Computational model schematic. Both a model pyramidal neuron (triangle) and a model FSIN (circle) receive simulated hippocampal input (which is rhythmically modulated at 8 Hz), and additional input which represents noise. (B) The correlation between the pyramidal neuron output spike rate and the rate of either noise inputs (dark blue) or hippocampal spikes (turquoise), as functions of a single parameter which represents how strongly hippocampal and noise inputs excite the model FSIN. (C) The spike rate of the model pyramidal neuron (turquoise) and FSIN (dark blue) as functions of a single parameter representing how strongly hippocampal and noise inputs excite the model FSIN. (D) The ratio of the correlation between pyramidal neuron output spikes and either hippocampal input or noise input.

As expected, as excitatory drive to the FSIN decreases, the rate of FSIN spiking falls while that of the pyramidal cell goes up (Figure 6C). When we examined the correlation between pyramidal cell spikes and either noise or hippocampal input, we found that decreasing FSIN excitatory drive also decreases the correlation between pyramidal cell output and hippocampal input (Figure 6B), causing a drop in the signal-to-noise ratio (Figure 6D). This occurs because as the strength of FSIN excitation increases, feedforward inhibition preferentially filters noise inputs, while hippocampal inputs are spared (due to their rhythmicity) (Figure 6B). Thus, when FSIN excitation is weak, there is minimal FSIN spiking and minimal pyramidal cell inhibition. Under these conditions, weak input is sufficient to excite the pyramidal cell, and the circuit fails to distinguish between the rhythmically occurring hippocampal signal and the (nonrhythmic) noise. As the level of FSIN excitation increases, it reaches an optimal level at which FSINs generate inhibition that suffices to filter out weak inputs. As a result, isolated noise inputs fail to elicit pyramidal cell spikes, whereas rhythmic bursts of hippocampal input provide a strong drive that allows them to be reliably transmitted via pyramidal cell spiking. Finally we note that while an extensive exploration of all possible inhibitory-disinhibitory circuit motifs is beyond the scope of this study, adding a simple form of disinhibition, in which a simulated interneuron-selective interneuron receives feedforward excitation and inhibits other interneurons, does not change our basic finding that there is an optimal level of feedforward excitation onto interneurons, below which the transmission of hippocampal input is degraded (Figure 6—figure supplement 1). The strength of this enhancement of hippocampal input over noise is dependent on hippocampal input frequency and best for intermediate, theta range values (Figure 6—figure supplement 2).

Discussion

We identified a specific behavioral deficit in mice with heterozygous loss of function of a high confidence ASD gene, then found associated deficits in biomarkers and pathways that we and others have previous linked to this behavior. Pogz^+/- mice show reduced anxiety-related avoidance in the EPM. Communication between the vHPC and mPFC is known to be necessary for this avoidance (Kjaerby et al., 2016; Padilla-Coreano et al., 2016), theta synchrony between LFPs recorded from the vHPC and mPFC is a biomarker for this communication (Padilla-Coreano et al., 2016), and vHPC-mPFC theta synchrony normally increases when mice approach the center of the EPM (Adhikari et al., 2010; Lee et al., 2019). In Pogz^+/- mice, both baseline vHPC-mPFC theta synchrony and its task-dependent modulation in the EPM are reduced. Notably, we confirmed this specific deficit in behaviorally-modulated theta-frequency vHPC–mPFC communication using an unbiased, data-driven approach. Furthermore, by directly examining vHPC-mPFC connections in brain slices, we found reduced excitatory drive from vHPC onto FSINs. This synaptic abnormality could plausibly contribute to the abnormalities we found in both avoidance behavior and LFP synchrony. Specifically, synaptic potentials and inhibitory activity are major drivers of LFP signals (Buzsáki et al., 2012; Haider et al., 2016; Teleńczuk et al., 2017). Thus, the deficit in vHPC excitation of mPFC interneurons we found should reduce the component of mPFC inhibitory synaptic activity that is driven by, and synchronized with, vHPC. Furthermore, we found that in a computational model, weakening feedforward excitation of inhibitory interneurons impairs the transmission of signals from the vHPC to mPFC.

Notably, a previous study found that during a working memory task, inhibiting mPFC PV interneurons did not affect vHPC-mPFC theta synchrony (Abbas et al., 2018). However, a recent study has found that distinct populations of vHPC pyramidal neurons project to different classes of mPFC neurons (Sánchez-Bellot and MacAskill, 2019). This study found that the population which specifically innervates PV interneurons also drives open arm avoidance in the EPM, whereas a distinct population of vHPC-mPFC projection neurons drives exploratory behavior. This shows that different populations of vHPC-mPFC projection neurons, which innervate distinct mPFC targets, are active during different behaviors. Thus, mPFC PV interneurons might contribute to vHPC theta synchrony during EPM behavior but not during working memory. The deficits we found in vHPC inputs to PV interneurons may also affect other interneuron populations which contribute to theta synchrony.

Interestingly, another study characterizing mice with heterozygous disruptions of Pogz was recently published (Matsumura et al., 2020). These Pogz mutant mice spent more time in the center of an open field, less time sniffing novel mice, and more time grooming, compared to wild-type mice. These Pogz mutants also exhibited an increased frequency of miniature excitatory post-synaptic currents (mEPSCs) in anterior cingulate cortex neurons. Based on the latter observation, the authors hypothesized that these Pogz mutants exhibit a shift in the balance of excitation and inhibition (E-I balance) toward excitation, and found that systemic treatment with an AMPA receptor antagonist increases social interaction in Pogz mutants. Their finding of increased time spent in the center of an open field is similar in nature to our finding that Pogz mutants spent increased time in the open arms of the EPM. Furthermore, we too find evidence of an alteration in E-I balance, although as elaborated below, we find that this reflects deficits in specific excitatory synapses onto inhibitory interneurons. Unlike Matsumara et al., we did not find a social deficit. However, the social assays used in the two studies were very different. Specifically, Matsumara et al. measured interaction over 60 min between the subject mouse and a novel adult mouse in an open field, whereas we measured interaction in the home cage with a junvenile mouse over 5 min. Furthermore, it is worth noting that we studied a mouse in which one copy of Pogz has a premature stop codon, whereas Matsumara et al. studied mice heterozygous for a patient-derived mutant allele of Pogz.

vHPC–mPFC communication and anxiety

A growing body of work shows that vHPC-prefrontal communication is important for anxiety-related behavior. The vHPC, unlike other portions of the hippocampus, projects directly to prefrontal cortex (Parent et al., 2010), and both structures are necessary for normal anxiety-related behavior (Kjelstrup et al., 2002; Shah and Treit, 2003). Theta-frequency synchronization between activity in the vHPC and mPFC normally increases in anxiety-provoking environments such as the EPM (Adhikari et al., 2010). Furthermore, single units in the mPFC that encode anxiety-related information phase-lock to the hippocampal-theta rhythm more strongly than other mPFC units (Adhikari et al., 2011). This suggests that these anxiety-encoding prefrontal units preferentially receive theta-modulated hippocampal input. Optogenetically manipulating vHPC–mPFC projections can also bidirectionally modulate anxiety-related avoidance (Padilla-Coreano et al., 2016; Padilla-Coreano et al., 2019). In particular, suppressing vHPC input to the mPFC reduces vHPC–mPFC theta synchrony, avoidance behavior, and the encoding of anxiety-related information by mPFC neurons. In previous work, we similarly found that pharmacologically suppressing vHPC–mPFC connections reduces open arm avoidance in the EPM (Kjaerby et al., 2016). Our present results build on these prior findings, while also extending them in a new direction.

In particular, because inhibiting all projections from the vHPC to mPFC reduced the firing rate in the preferred arm type for mPFC neurons which prefer either the open or closed arm, a previous study concluded that the predominant effects of vHPC input to mPFC are excitatory and serve to increase firing rates in each mPFC neuron’s preferred arm (Padilla-Coreano et al., 2016). By contrast, in Pogz mutant mice, the theta coordination of vHPC–mPFC activity and open arm avoidance are both impaired even though vHPC input to mPFC pyramidal neurons remains intact. This raises the possibility that feedforward inhibition may be important for vHPC input to transmit anxiety-related information to the mPFC, and that deficits in feedforward inhibition may contribute to abnormal avoidance behavior in Pogz mutant mice. By showing how vHPC input to interneurons and feedforward inhibition may also play an important role, our results contrast with/add to the model suggested by previous studies, in which main role of vHPC input is to provide excitation that drives mPFC neuron firing in specific locations. It is not currently possible to selectively inhibit input from one presynaptic source onto one postsynaptic cell-type (e.g., vHPC input to interneurons) using optogenetic or chemogenetic manipulations. Therefore, while imperfect, genetic models, for example, Pogz^+/- mice, can reveal behavioral phenotypes that may result from physiological alterations that cannot be readily modeled using optogenetics or chemogenetics.

Hippocampal-prefrontal communication is important for other behaviors, besides open arm avoidance, most notably tasks that measure spatial working memory using the T-maze (Sigurdsson et al., 2010; Spellman et al., 2015). We did not find deficits in delayed alternation in Pogz^+/- mice. This may reflect the fact that our task used a very short delay (4 s) for which prefrontal circuits may not be necessary (Bolkan et al., 2017), because other forms of synchronization may compensate for deficits in vHPC–mPFC theta synchrony (Tamura et al., 2017), or because the deficits in feedforward inhibition that we found in Pogz^+/- mice might involve classes of prefrontal interneurons that are not required for spatial working memory (Abbas et al., 2018).

The other recently published study which examined mice heterozygous for a missense mutation in Pogz found they had smaller brains. We did not observe smaller brains, but if there were anatomical differences between our Pogz^+/- mice and WT mice, these could have caused mistargeting of the vHPC in mutants, thereby contributing to the abnormalities we observed when measuring vHPC–mPFC synchrony. We do not believe this was the case, because we verified electrode placement both histologically (by visually examining the anatomical location of the electrode track) and electrophysiologically (by confirming the presence of a prominent theta- frequency peak in the vHPC LFP power spectrum). Importantly, the fraction of experiments excluded due to the absence of a clear theta-frequency peak in hippocampal recordings, was not different between WT and mutant mice. This suggests there was not systemic mistargeting in Pogz mutant mice as a result of anatomical differences.

Whereas previous studies (including our own), have taken a hypothesis-driven approach to evaluating the role of theta-frequency vHPC-mPFC communication in approach-avoidance decisions, here we also explored a data-driven approach, using ICA to identify biomarkers associated with these decisions. This approach yielded an IC which measures synchrony between theta-frequency vHPC activity and mPFC activity, and which exhibits modulation as mice approach decision points (the center zone). Thus, this IC represents a data-driven metric that shows how theta-frequency communication between the vHPC and mPFC (phase-amplitude coupling between mPFC gamma and vHPC theta) correlates with approach-avoidance decisions. Finding that this metric, like theta-frequency WPLI, is altered in Pogz^+/- mice during closed-center transitions, thus provides strong confirmation that theta-frequency hippocampal-prefrontal communication related to approach-avoidance decisions is disrupted in Pogz^+/- mice.

However, it is interesting to note theta phase synchrony was not itself ‘pulled out’ by the ICA. This presumably reflects the fact that over the entirety of the task, theta phase synchrony is being influenced by different factors than this IC, even though theta phase synchrony and this IC both evolve in parallel specifically during closed arm-center zone approaches. In other words, during closed-center runs, both theta phase synchrony and the IC both exhibit a sharp rise followed by a return to baseline. However, during the rest of the task, there must be other behaviors that differentially recruit these two measures. Future studies might identify these behaviors using approaches such as MoSeq and DeepLabCut (Mathis et al., 2018; Wiltschko et al., 2015).

A final note is that while we have measured vHPC-mPFC synchronization at the level of field potentials, an important future direction is measuring the synchronization of specific cell types, which could be done using electrophysiology or genetically encoded voltage indicators (Cho et al., 2020).

Excitatory-inhibitory (E-I) balance in anxiety and autism

Another recently published study from our laboratory showed that inhibiting vasoactive intestinal polypeptide (VIP)-expressing interneurons in the mPFC causes a similar behavioral phenotype, that is, reduced open arm avoidance in the EPM (Lee et al., 2019). That study found VIP interneurons normally facilitate the transmission of anxiety-related information from the vHPC to mPFC by disinhibiting prefrontal responses to vHPC input. As a result, when VIP interneurons are inhibited, information about anxiety is not transmitted properly, causing mice to spend more time exploring the open arms. Since VIP interneurons inhibit other GABAergic interneurons, the effect of inhibiting VIP interneurons is to increase feedforward inhibition. In this context, it may seem paradoxical that the present study finds a similar phenotype (increased open arm exploration) in Pogz^+/- mice when mPFC inhibition evoked by vHPC input is impaired. Together, these two studies underscore the importance of properly balanced cortical circuit inhibition.

In the context of approach-avoidance behaviors, the PFC is believed to play a key role by evaluating information from multiple sources in order to make a decision about whether to approach or avoid a potentially anxiogenic region (Calhoon and Tye, 2015). As illustrated by the computational model depicted in Figure 6, circuit inhibition is critical for this process. When levels of inhibition are too low, the firing of simulated mPFC output neurons is driven mainly by noise, that is, inputs unrelated to anxiety signals. This could prevent the mPFC from properly representing anxiety-related information, and/or cause the inappropriate transmission of signals related to exploratory behavior. Higher levels of inhibition can filter out the noise, allowing hippocampal inputs to be preferentially transmitted. As described in our earlier study, the ability of rhythmic hippocampal inputs to periodically recruit VIP interneuron-mediated phasic disinhibition could further promote the preferential transmission of hippocampally-driven activity. Thus, multiple classes of interneurons may work together to inhibit and filter out non-hippocampal inputs while optimizing the responsiveness to hippocampal input, potentially facilitating the transmission of anxiety-related information across hippocampal-prefrontal circuits. In this way, appropriately balanced inhibition may be indispensable for proper action selection related to approach and avoidance behaviors.

Disruptions in the balance between cortical excitation and inhibition (E-I balance) have long been hypothesized to play a role in ASD (Lee et al., 2017; Rubenstein and Merzenich, 2003). Numerous studies have identified examples of altered E-I balance related to autism. These reflect changes in the relative levels of synaptic excitation and inhibition and can be secondary to a variety of different factors, including alterations in synaptic plasticity, homeostasis, and regulatory feedback loops (Bourgeron, 2015; Mullins et al., 2016; Nelson and Valakh, 2015; Sohal and Rubenstein, 2019; Toro et al., 2010; Wondolowski and Dickman, 2013).

Deficits in long-range communication in autism

In addition to the hypothesis that E-I balance is disturbed in autism, another hypothesis is that autism (and altered E-I balance) may reflect changes in long-range connectivity (Just et al., 2012). While early work focused mainly on a theory of under-connectivity in autism (Just et al., 2004), evidence for both hypo- and hyper-connectivity has been identified using a range of methods, including functional magnetic resonance imaging (fMRI) (Müller et al., 2011; Redcay et al., 2013), electroencephalography (EEG) (Coben et al., 2014; Zeng et al., 2017), magnetoencephalography (MEG) (Buard et al., 2013), and structural imaging (Mueller et al., 2013; Nair et al., 2013). Changes in long-range connectivity have been identified in a number of other disorders, including schizophrenia (Guo et al., 2015; Wang et al., 2014), generalized anxiety disorder (Andreescu et al., 2014; Xing et al., 2017), and bipolar disorder (Kam et al., 2013; Wang et al., 2017), suggesting that altered connectivity may be common to a range of neurodevelopmental and psychiatric disorders. Here we find disturbed long-range connectivity (as measured by LFP synchrony) which, when examined at a finer scale, is associated with a selective deficit in the recruitment of inhibitory interneurons. This reveals a specific mechanism – impaired feedforward inhibition – that could potentially link together two prominent hypotheses about the neurobiology of autism in a way that could contribute to behavioral abnormalities.

It should be noted that the changes we observed are not necessarily static. Connectivity abnormalities in ASD have been shown to be age-(Keehn et al., 2013; Padmanabhan et al., 2013) and state-dependent (You et al., 2013). Our study focuses on the outcome of developmental disruptions in the adult brain but does not establish a direct mechanism tracing changes in Pogz expression to network-level changes. It is possible that these changes in connectivity would be different in juvenile mice, and/or that the changes we see reflect a compensatory response to changes at an earlier timepoint.

Possible relevance of Pogz behavioral phenotypes to autism

This study focuses on a phenotype whereby Pogz^+/- mice exhibit reduced avoidance of the open arms of the EPM. The EPM is often regarded as an assay that measures anxiety-related behavior. In this framework, reduced open arm avoidance is interpreted to reflect reduced anxiety. Reduced anxiety is not typically associated with autism, raising a question about the relevance of our findings for the clinical condition.

On the one hand, relying on face validity to determine which mouse behavioral phenotypes are relevant to human autism can be problematic for multiple reasons. First, mouse assays measure only the most rudimentary aspects of social behavior – typically social preference and/or preference for social novelty. In many individuals with autism, social preference and preference for social novelty are intact, but social functioning is disrupted in other ways. In particular, the largest study of individuals with disruptions in Pogz found ‘in many cases, a seemingly contrary overly social and overly friendly demeanor’ (Stessman et al., 2016). Thus, it is questionable how well face valid mouse assays of social behavior capture the more nuanced and heterogeneous phenotypes characteristic of clinical autism. On the other hand, we do not want to assert that any behavioral phenotype observed in mice with disruptions in an autism-associated gene will automatically be relevant to autism.

In this context, a logical approach is to focus on brain regions and networks that have consistently been implicated in autism. While specific behaviors may not be well conserved across species, we hypothesize that general principles underlying the function of limbic circuits, for example, hippocampal-prefrontal interactions, will be more likely to translate. In this context, we found that prefrontal circuits fail to use limbic input to appropriately guide decisions about approach vs. avoidance behavior. This is notable as a recent review hypothesized that deficits in the ability of the prefrontal cortex to appropriate guide approach/avoidance decisions plays a key role in autism (Pfaff and Barbas, 2019).

Conclusion

We characterized behavior and network-level physiology in mice with heterozygous loss of function in Pogz, a high confidence autism gene. Pogz^+/- mice show reduced avoidance behavior in the EPM and altered vHPC-PFC synchrony, consistent with recent work characterizing the role of the vHPC-mPFC circuit in anxiety behavior. Additionally, in slice experiments, we found reduced excitatory drive from the hippocampus to prefrontal FSINs, suggesting an impairment in ability to properly filter incoming hippocampal input. This work elucidates the nature of a network-level phenotype linking genetic and developmental perturbations with specific behavioral and physiological changes in the adult brain.

Materials and methods

Key resources table.

Reagent type (species) or resource	Designation	Source or reference	Identifiers	Additional information
Strain, strain background (Mus. Musculus)	C57BL6/J	Jackson Labs	Stock No: 000664
Genetic reagent (Mus. Musculus)	PogZ+/-	Rubenstein Lab
Recombinant DNA reagent	AAV5-CaMKIIa-hChR2(H134R)-EYFP	UNC Vector Core	RRID:Addgene_26969
Recombinant DNA reagent	AAV5-DlxI12b-mCherry	Virovek, Sohal lab
Software, algorithm	Sirenia Acquisition	Pinnacle	RRID:SCR_016183
Software, algorithm	ANY-maze tracking software	ANY-maze	RRID:SCR_014289
Software, algorithm	Python	Python	RRID:SCR_008394	Packages: Numpy, Scipy, Matplotlib, Seaborn
Software, algorithm	MATLAB	Mathworks	RRID:SCR_001622	Signal Processing Toolbox
Software, algorithm	PClamp	Molecular Devices	RRID:SCR_011323

Open in a new tab

Subjects and behavioral assays

All experiments were conducted in accordance with procedures established by the Administrative Panels on Laboratory Animal Care at the University of California, San Francisco. Male and female mice > 4 weeks old were used in all experiments. All mice were Pogz heterozygotes or wild-type littermates. Gene expression changes in these mice are characterized in a related publication (Markenscoff-Papadimitriou et al., in preparation). Briefly, these mice were generated by CRISPR-Cas9 and sgRNAs targeting exons 1 and 6, a 10 kb span, which generated a premature stop codon. Reduced POGZ expression in Pogz^+/- cortex at P28 was verified by Western blot.

Unless otherwise noted, experiments were performed under ambient light and mice were group housed with littermates. Mice were habituated to the behavioral testing area for >30 min at the beginning of all sessions. LFP recording during behavior was done in a separate cohort from the mice used to establish behavioral phenotypes. For LFP experiments, mice were habituated to the head tether in their home cage for 15 min daily for 3 days. ANY-maze (Stoelting) was used to track the position of the mouse during assays using a USB webcam. Experimenter was blinded to each mouse’s genotype during behavioral assessment. Note: The overall design for our behavioral studies was to perform an initial screen using multiple behavioral assays. This initial screen revealed altered behavior in the EPM, but not for many other behavioral assays, for example, for social interaction. Therefore, we then we validated the EPM finding using additional mice. For this reason, the N is larger for the EPM than for other social and cognitive assays. In addition, in some cases it was not possible to perform all possible analyses on every mouse run on a particular behavioral assay, e.g., because some analyses were performed at later times and the original data had not been recorded/stored in a manner that was suitable for a specific analysis. This explains why the Ns sometimes differ for multiple analyses of data from the same assay. Importantly, no animals were excluded from specific analyses post-hoc.

Elevated plus maze

Mice were exposed to the EPM for a single 15 min session. All mice were placed in the center of the maze facing an open arm. Time spent in zones, distance traveled, and number of entries were scored with ANY-maze; head dips were manually scored by a blinded observer.

Social/novel assay

Mice were exposed to a conspecific juvenile followed by a novel object in their home cage for 10 min each. Active interaction time was scored by a blinded observer.

Marble burying

Marble burying was performed as previously described (Angoa-Pérez et al., 2013). Mice were placed in a larger housing cage for 20 min with 20 marbles arranged in a 4 × 5 grid. After 20 min, the number of fully buried marbles was counted.

Cognitive tasks

Mice were singly housed and placed on a reverse light-dark cycle for the duration of testing. Mice received 3 days of restricted food intake to reach a goal weight of ~80% free-feeding weight in order to sufficiently motivate them. In each task, this period was used to habituate mice to testing apparatus and basic task mechanics (location of food reward, trial structure, etc.). Water was freely available during the entire period. All testing was done under red light.

Rule shifting

An odor/texture rule shifting task was performed as previously described (Cho et al., 2015; Ellwood et al., 2017). Briefly, mice were presented with two bowls containing either sand (Mosser Lee White Sand) or bicarbonate-free cat x (1% by volume) with either ground coriander (McCormick) or garlic powder (McCormick), as well as finely chopped peanut butter chips to mask scent of food reward. Each trial contained one of two possible combinations of media: sand and garlic paired with litter and coriander, respectively, or sand and coriander paired with litter and garlic, respectively. In the initial association phase of the task, mice had to learn that a single cue (e.g. sand) signaled the location of a reward. Once mice learned this rule (8 out of 10 previous trials correct), there was an un-cued extradimensional rule shift such that a different type of cue (e.g. garlic) now signaled the reward.

Delayed match to sample task

A delayed match to sample T-maze task was performed as previously described (Spellman et al., 2015; Tamura et al., 2017). Briefly, mice were placed at the base of a T-shaped maze at the start of each trial. During the sample phase, one of the two choice arms of the T was blocked off such that mice were forced to one arm. After reaching the end of the arm, mice then had to return to the start point, where a sliding door held them for a variable delay phase (all data presented here from a 4 s delay). Following the delay was a choice phase – the door was removed, allowing the mice to run down the arms and choose which to enter. Mice had to learn to go to the opposite arm from the sample phase (e.g. if they entered the right arm during the sample phase, a food reward would be present in the left arm).

Local field potential recordings

All surgeries were done under isofluorane anesthesia in a stereotaxic frame (Kopf). Standard-tip 0.5 MΩ-impedance stainless steel electrodes (Microprobes, SS30030.5A10) were inserted into the vHPC and mPFC. The coordinates for vHPC and mPFC were as follows: vHPC, −3.25 (AP), 3.1(ML), −4.1 (DV); mPFC, 1.7 (AP), 0.3 (ML), −2.75 (DV). A common reference screw was implanted into the cerebellum (0.5 mm posterior to lambda) and a silver ground wire was placed underneath the left lateral scalp. After affixing the electrodes in place using Metabond, connections were made to the headstage of a multi-channel recording system (Pinnacle). All channels shared a common reference (cerebellum). Data was collected at 2000 Hz and band-pass filtered 1–200 Hz at the pre-amp. Electrode placement was verified histologically. We also examined the power spectra from all electrodes; only animals with vHPC power spectra that exhibited a visible peak in the theta- frequency range as judged by a blinded observer were used for further analysis. 4 mice of each genotype were excluded due to lack of a visible theta peak (these mice were excluded from further workflow including histology).

Analysis of LFP data was facilitated using custom MATLAB code. The LFP signals were FIR-filtered (filter length 3x period corresponding to minimum frequency of frequency band) and Hilbert transformed to yield the instantaneous amplitudes (magnitude) and phases (angle). Bulk measures were calculated using data from the entire recording period; dynamic measures were calculated using a 2.5 s window, at 1.5 s intervals from 7.5 s before to 7.5 s after the animal entered the center of the EPM. Dynamic measurements were quantified as z-scores calculated relative to the rest of the run (7.5 s before to 7.5 s after the animal entered the center).

Power was quantified using Welch’s power spectral density estimate with nonoverlapping segments. Synchrony between vHPC and mPFC was measured by taking the Hilbert transform of band-passed data and either comparing the instantaneous phase using the weighted-phase locking index (Vinck et al., 2011) or instantaneous amplitude using amplitude covariation. These measures were computed across four frequency bands: theta (4–12 Hz), beta (13–30 Hz), low gamma (30–55 Hz), and high gamma (65–100 Hz).

Cross-frequency coupling was calculated by comparing the instantaneous phase in a low frequency band with the instantaneous amplitude in a high frequency band. Specifically, instantaneous phase and amplitude were obtained using the Hilbert transform (using the Matlab function hilbert). At each point in time, this phase and amplitude were combined to yield a vector in the complex plane. We combined vectors from successive timepoints, and the amplitude of the vector sum was normalized to the sum of all the amplitudes to quantify the strength of cross-frequency coupling. Low frequency bands were theta (2–6 Hz) and alpha (6–10 Hz). High frequency bands were beta (13–30 Hz), low gamma (30–55 Hz), and high gamma (65–100 Hz). Cross- frequency coupling was calculated for all possible combinations of a single low and single high frequency band in all combinations of brain regions (PFC low/HPC high, HPC high/PFC low, PFC low/PFC high, HPC low/HPC high).

These features (Table 1) were all used as input for the ICA based on methods outlined in previous work (Kirkby et al., 2018). First, all features were calculated for each subject and PCA was performed for dimensionality reduction and orthogonalization and the number of significant components was calculated using the threshold set by the Marchenko-Pastur Law (Lopes-dos-Santos et al., 2013). ICA was used on the significant PCs to separate the signal mixtures into independent sources using the fastICA algorithm (Hyvärinen and Oja, 2000). Similarity of ICs across mice was calculated using the Pearson correlation coefficient. Significant clusters were isolated by selecting for ICs that had a correlation coefficient of >0.7 with at least one other IC and using MATLAB’s graph function to identify groups of highly similar ICs. Characteristic ICs were found by averaging groups of ICs with members from at least three different animals. The projection of these characteristic ICs onto behavior was found by multiplying the vector of Z-scored features in each point in time by the weight in the characteristic IC and summing all values.

Whole cell patch clamp recordings

Mice were injected with 750 nL of AAV5-CaMKIIa-hChR2(H134R)-EYFP (UNC Vector Core) into the vHPC (DV: −4, AP: −3.3, ML: −3.2) to label excitatory projections from the vHPC to the mPFC. A subset of mice were also injected with 500 nL AAV-DlxI12b-mCherry in the mPFC (DV: −2.75, AP: 1.7, ML: 0.3) to label MGE-derived interneurons (Potter et al., 2009). We waited ~8 weeks from virus injection to slice experiments. Whole cell patch recordings were obtained from 250 μm coronal slices. Cells were identified using differential contrast video microscopy on an upright microscope (BX51W1, Olympus) and recordings were made using a Multiclamp 700A (Molecular Devices). Data was collected using pClamp (Molecular Devices) software and analyzed using custom MATLAB code. Patch electrodes were filled with the following (in mM): 130 K-gluconate, 10 KCl, 10 HEPES, 10 EGTA, 2 MgCl, 2 MgATP, and 0.3 NaGTP (pH adjusted to 7.3 with KOH). All recordings were at 32.0 ± 1°C. Series resistance was usually 10–20 MΩ, and experiments were discontinued above 25 MΩ. For voltage clamp recordings, cells were held at −70 mV and +10 mV to isolate EPSCs and IPSCs, respectively. An LED engine (Lumencor) was used for optogenetic stimulation of terminals from vHPC projections. We used ~1–3 mW of 470 nm light in 5 ms pulses to stimulate ChR2-infected fibers. The light was delivered to the slice via a 40x objective (Olympus) which illuminated the full field.

Computational model of the role of feedforward inhibition

The effects of changing the strength of excitatory drive onto interneurons was modeled using two integrate-and-fire neurons – an output cell, representing a pyramidal cell, and an interneuron that targeted the output cell, representing a FSIN. Each cell received noise input and theta-patterned ‘hippocampal’ input. Initial values were selected such that the inhibitory neuron would spike at ~20 Hz and the output neuron would spike at ~25 Hz and ~50 Hz in the presence and absence of inhibition. All values were held constant except for the strength of excitatory input onto the output-targeting interneuron, adjusting either just the hippocampal strength or adjusting the hippocampal and noise strength in parallel. Input spikes were modeled as a Poisson process. Correlation between the input sources was calculated by comparing binned spike times for input spikes (from the Poisson train) and output spikes (when the output cell’s membrane potential cleared a threshold). The relative contributions of the two input sources was calculated by comparing the ratio of the correlation between the output spikes and the noise input or hippocampal input. Correlation values were based on 1000 iterations of a 1 s spike train.

Statistics, data analysis, and data and code availability

Unless otherwise specified, non-parametric tests were used for all statistical comparisons and all tests are two-sided. Statistics were calculated using MATLAB or Python’s SciPy package. Linear mixed models were evaluated using the ‘fitlme’ function in Matlab. Sample sizes were based on prior studies. All Ns indicate biological replication, that is, data from different samples (different cells or different animals), rather than technical replication (multiple measurements of the same sample). Details of p-values, Ns and statistical tests for all comparisons performed in this study are given in Table 3. Raw data related to this study has been deposited in Dryad (doi:10.7272/Q6ZP44B9). All custom written analysis code is available on Github (Cunniff, 2020) (https://github.com/mcunniff/PogZ_paper); Cunniff, 2020 copy archived at swh:1:rev:189f9c500bdeaddeb69d3eef8b604949c2936d19.

Table 3. Details of all statistical tests N indicates biological replicates for example individual cells or behavior trials.

Figure	Data	Test	P val	WT Animals	Het Animals	WT n	Het n
Figure 1C	Zone occupancy	Wilcoxon rank sum	0.003	18	27
Figure 1D	EPM Distance	Wilcoxon rank sum	0.35	16	23
Figure 1E	Open time	Wilcoxon rank sum	0.001	18	27
Figure 1E	Center time	Wilcoxon rank sum	0.02	18	27
Figure 1F	Head dips	Wilcoxon rank sum	0.03	14	14
Figure 1G	Open entries	Wilcoxon rank sum	0.32	16	23
Figure 1H	Open visit	Wilcoxon rank sum	0.047	16	23
Figure 2B	WPLI, t = 0	Wilcoxon rank sum	0.0007	6	7	274	316
Figure 2B	WPLI, t = 1.5	Wilcoxon rank sum	0.043	6	7	274	316
Figure 2B	WPLI, t = −3,–1.5, 0, +1.5 during closed-center runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.0026 0.47 0.059 0.0004	6	7	274	316
Figure 2C	Avg zone WPLI, genotype	Two-way ANOVA	0.03	6	7
Figure 2C	Avg zone WPLI, zone	Two-way ANOVA	0.063	6	7
Figure 2C	Avg zone WPLI, interaction	Two-way ANOVA	0.98	6	7
Figure 2D	Theta WPLI	Wilcoxon rank sum	0.031	6	7
Figure 3E	IC zone projection, t = 0	Wilcoxon rank sum	0.007	6	7	39	37
Figure 3E	ICA zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.085 0.16 0.0044 0.010	6	7	39	37
Figure 4E	FSIN charge	Wilcoxon rank sum	0.006	6	3	11	7
Figure 4F	FSIN PPR	Wilcoxon rank sum	0.03	6	3	11	7
Figure 4G	FSIN latency	Wilcoxon rank sum	0.013	6	3	11	7
Figure 4H	FSIN # spikes	Wilcoxon rank sum	0.08	6	3	11	7
Figure 5E	Pyr charge	Wilcoxon rank sum	0.28	13	8	17	11
Figure 5F	Pyr PPR	Wilcoxon rank sum	0.15	13	8	17	11
Figure 5G	Pyr latency	Wilcoxon rank sum	0.76	13	8	17	11
Figure 5H	Pyr # spikes	Wilcoxon rank sum	0.78	13	8	17	11
Figure 1—figure supplement 1	Sex-corrected zone occupancy	Wilcoxon rank sum	0.013	18	27
Figure 1—figure supplement 1A	Zone occupancy for Het M vs. F	Wilcoxon rank sum	0.60		M: 10, F: 17
Figure 1—figure supplement 1B	Sex-corrected EPM distance	Wilcoxon rank sum	0.79	16	23
Figure 1—figure supplement 1C	Head dips: genotype	2-way ANOVA	0.02	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips: sex	2-way ANOVA	0.81	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips: genotype X sex	2-way ANOVA	0.36	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips for Het M vs. F	Wilcoxon rank sum	0.54		M: 8, F: 6
Figure 1—figure supplement 1D	Open arm entries: genotype	2-way ANOVA	0.22	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1D	Open arm entries: sex	2-way ANOVA	0.61	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1D	Open entries: genotype X sex	2-way ANOVA	0.32	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: genotype	2-way ANOVA	0.22	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: sex	2-way ANOVA	0.49	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: genotype X sex	2-way ANOVA	0.76	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1F	Sex-corrected open arm time	Wilcoxon rank sum	0.013	16	23
Figure 1—figure supplement 1F	Open arm time: Het M vs. F	Wilcoxon rank sum	0.61		M: 10, F: 17
Figure 1—figure supplement 1G	Center time: genotype	2-way ANOVA	0.087	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 1G	Center time: sex	2-way ANOVA	0.29	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 1G	Center time: genotype X sex	2-way ANOVA	0.48	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 2A	Social interaction	Wilcoxon rank sum	0.34	7	7
Figure 1—figure supplement 2B	Novel objection	Wilcoxon rank sum	0.95	7	7
Figure 1—figure supplement 2C	Marble burying	Wilcoxon rank sum	0.45	8	7
Figure 1—figure supplement 2D	OF distance	Wilcoxon rank sum	0.15	14	17
Figure 1—figure supplement 2F	T-maze trials	Wilcoxon rank sum	0.6	5	5
Figure 1—figure supplement 2H	Rule shift IA	Wilcoxon rank sum	0.89	4	4
Figure 1—figure supplement 2H	Rule shift RS	Wilcoxon rank sum	0.89	4	4
Figure 2—figure supplement 1A	PFC theta power	Wilcoxon rank sum	0.91	6	7
Figure 2—figure supplement 1A	PFC beta power	Wilcoxon rank sum	0.94	6	7
Figure 2—figure supplement 1A	PFC low gamma power	Wilcoxon rank sum	0.47	6	7
Figure 2—figure supplement 1A	PFC high gamma power	Wilcoxon rank sum	0.8	6	7
Figure 2—figure supplement 1B	HPC Theta power	Wilcoxon rank sum	0.23	6	7
Figure 2—figure supplement 1B	HPC Beta power	Wilcoxon rank sum	0.093	6	7
Figure 2—figure supplement 1B	HPC low gamma power	Wilcoxon rank sum	0.17	6	7
Figure 2—figure supplement 1B	HPC high gamma power	Wilcoxon rank sum	0.94	6	7
Figure 2—figure supplement 1C	PFC closed theta power	Wilcoxon rank sum	0.88	6	7
Figure 2—figure supplement 1C	PFC closed beta power	Wilcoxon rank sum	1	6	7
Figure 2—figure supplement 1C	PFC closed LG power	Wilcoxon rank sum	0.29	6	7
Figure 2—figure supplement 1C	PFC closed HG power	Wilcoxon rank sum	0.1	6	7
Figure 2—figure supplement 1D	HPC closed theta power	Wilcoxon rank sum	0.25	6	7
Figure 2—figure supplement 1D	HPC closed beta power	Wilcoxon rank sum	0.15	6	7
Figure 2—figure supplement 1D	HPC closed LG power	Wilcoxon rank sum	0.48	6	7
Figure 2—figure supplement 1D	HPC closed HG power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open theta power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open beta power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open LG power	Wilcoxon rank sum	1	6	7
Figure 2—figure supplement 1E	PFC open HG power	Wilcoxon rank sum	0.15	6	7
Figure 2—figure supplement 1F	HPC open theta power	Wilcoxon rank sum	0.32	6	7
Figure 2—figure supplement 1F	HPC open beta power	Wilcoxon rank sum	0.2	6	7
Figure 2—figure supplement 1F	HPC open LG power	Wilcoxon rank sum	0.25	6	7
Figure 2—figure supplement 1F	HPC open HG power	Wilcoxon rank sum	1	6	7
Figure 3—figure supplement 1D	IC #1 zone projection	Wilcoxon rank sum	0.007	6	7	39	37
Figure 3—figure supplement 1D	IC #1 zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.085 0.16 0.0044 0.010	6	7	39	37
Figure 3—figure supplement 1F	IC #3 zone projection	Wilcoxon rank sum	0.015	6	7	39	37
Figure 3—figure supplement 1F	IC #3 zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.0094 0.50 0.026 0.052	6	7	39	37
Figure 4—figure supplement 1C	FSIN resting potential	Wilcoxon rank sum	0.50	6	3	11	7
Figure 4—figure supplement 1D	FSIN input resistance	Wilcoxon rank sum	0.44	6	3	11	7
Figure 4—figure supplement 1E	FSIN halfwidth	Wilcoxon rank sum	0.47	6	3	11	7
Figure 4—figure supplement 1F	FSIN max firing rate	Wilcoxon rank sum	0.50	6	3	11	7
Figure 5—figure supplement 1C	Pyr resting potential	Wilcoxon rank sum	0.94	13	8	17	11
Figure 5—figure supplement 1D	Pyr input resistance	Wilcoxon rank sum	0.80	13	8	17	11
Figure 5—figure supplement 1E	Pyr halfwidth	Wilcoxon rank sum	0.46	13	8	17	11
Figure 5—figure supplement 1F	Pyr max firing rate	Wilcoxon rank sum	0.93	13	8	17	11

Open in a new tab

Acknowledgements

We acknowledge funding from the Simons Foundation Autism Research Initiative (Grants # 399853 and 514438), NIMH (R56MH117961 and R01MH117961), and a Trailblazer Award from the Weill Institute for Neurosciences.

Funding Statement

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

Contributor Information

Vikaas Singh Sohal, Email: vikaas.sohal@ucsf.edu.

Laura L Colgin, University of Texas at Austin, United States.

Funding Information

This paper was supported by the following grants:

Simons Foundation 399853 to Margaret M Cunniff, Vikaas Singh Sohal.
National Institute of Mental Health R56MH117961 to Margaret M Cunniff, Vikaas Singh Sohal.
Weill Institute for Neurosciences to Margaret M Cunniff, Vikaas Singh Sohal.
Simons Foundation 514438 to Eirene Markenscoff-Papadimitriou, John LR Rubenstein.
National Institute of Mental Health R01MH117961 to Margaret M Cunniff, Vikaas Singh Sohal.

Additional information

Competing interests

No competing interests declared.

JLRR is cofounder, stockholder, and currently on the scientific board of Neurona, a company studying the potential therapeutic use of interneuron transplantation.

VSS has received research funding from Neurona therapeutics and is on the scientific board of Empathic Therapeutics.

Author contributions

Conceptualization, Formal analysis, Investigation, Visualization, Writing - original draft.

Resources, Methodology.

Software, Formal analysis, Visualization.

Resources, Supervision, Funding acquisition, Project administration, Writing - review and editing.

Conceptualization, Software, Supervision, Funding acquisition, Methodology, Project administration, Writing - review and editing.

Ethics

Animal experimentation: All experiments were performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee (IACUC) protocols of the University of California, San Francisco (IACUC protocol #AN170116).

Additional files

Transparent reporting form

elife-54835-transrepform.docx^{(246.4KB, docx)}

Data availability

All data has been deposited in Dryad, DOI: doi:10.7272/Q6ZP44B9 All code has been deposited in GitHub: https://github.com/mcunniff/PogZ_paper (copy archived at https://archive.softwareheritage.org/swh:1:rev:189f9c500bdeaddeb69d3eef8b604949c2936d19/).

The following dataset was generated:

Sohal VS. 2020. Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene PogZ. Dryad Digital Repository.

References

Abbas AI, Sundiang MJM, Henoch B, Morton MP, Bolkan SS, Park AJ, Harris AZ, Kellendonk C, Gordon JA. Somatostatin interneurons facilitate Hippocampal-Prefrontal synchrony and prefrontal spatial encoding. Neuron. 2018;100:926–939. doi: 10.1016/j.neuron.2018.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]
Adhikari A, Topiwala MA, Gordon JA. Synchronized activity between the ventral Hippocampus and the medial prefrontal cortex during anxiety. Neuron. 2010;65:257–269. doi: 10.1016/j.neuron.2009.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
Adhikari A, Topiwala MA, Gordon JA. Single units in the medial prefrontal cortex with anxiety-related firing patterns are preferentially influenced by ventral hippocampal activity. Neuron. 2011;71:898–910. doi: 10.1016/j.neuron.2011.07.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
Andreescu C, Sheu LK, Tudorascu D, Walker S, Aizenstein H. The ages of anxiety--differences across the lifespan in the default mode network functional connectivity in generalized anxiety disorder. International Journal of Geriatric Psychiatry. 2014;29:704–712. doi: 10.1002/gps.4051. [DOI] [PMC free article] [PubMed] [Google Scholar]
Angoa-Pérez M, Kane MJ, Briggs DI, Francescutti DM, Kuhn DM. Marble burying and nestlet shredding as tests of repetitive, Compulsive-like behaviors in mice. Journal of Visualized Experiments. 2013;82:50978. doi: 10.3791/50978. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bolkan SS, Stujenske JM, Parnaudeau S, Spellman TJ, Rauffenbart C, Abbas AI, Harris AZ, Gordon JA, Kellendonk C. Thalamic projections sustain prefrontal activity during working memory maintenance. Nature Neuroscience. 2017;20:987–996. doi: 10.1038/nn.4568. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bourgeron T. From the genetic architecture to synaptic plasticity in autism spectrum disorder. Nature Reviews Neuroscience. 2015;16:551–563. doi: 10.1038/nrn3992. [DOI] [PubMed] [Google Scholar]
Buard I, Rogers SJ, Hepburn S, Kronberg E, Rojas DC. Altered oscillation patterns and connectivity during picture naming in autism. Frontiers in Human Neuroscience. 2013;7:742. doi: 10.3389/fnhum.2013.00742. [DOI] [PMC free article] [PubMed] [Google Scholar]
Buzsáki G, Anastassiou CA, Koch C. The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nature Reviews Neuroscience. 2012;13:407–420. doi: 10.1038/nrn3241. [DOI] [PMC free article] [PubMed] [Google Scholar]
Calhoon GG, Tye KM. Resolving the neural circuits of anxiety. Nature Neuroscience. 2015;18:1394–1404. doi: 10.1038/nn.4101. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cho KK, Hoch R, Lee AT, Patel T, Rubenstein JL, Sohal VS. Gamma rhythms link prefrontal interneuron dysfunction with cognitive inflexibility in Dlx5/6+/− mice. Neuron. 2015;85:1332–1343. doi: 10.1016/j.neuron.2015.02.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cho KKA, Davidson TJ, Bouvier G, Marshall JD, Schnitzer MJ, Sohal VS. Cross-hemispheric gamma synchrony between prefrontal parvalbumin interneurons supports behavioral adaptation during rule shift learning. Nature Neuroscience. 2020;23:892–902. doi: 10.1038/s41593-020-0647-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
Coben R, Mohammad-Rezazadeh I, Cannon RL. Using quantitative and analytic EEG methods in the understanding of connectivity in autism spectrum disorders: a theory of mixed over- and under-connectivity. Frontiers in Human Neuroscience. 2014;8:045. doi: 10.3389/fnhum.2014.00045. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cotney J, Muhle RA, Sanders SJ, Liu L, Willsey AJ, Niu W, Liu W, Klei L, Lei J, Yin J, Reilly SK, Tebbenkamp AT, Bichsel C, Pletikos M, Sestan N, Roeder K, State MW, Devlin B, Noonan JP. The autism-associated chromatin modifier CHD8 regulates other autism risk genes during human neurodevelopment. Nature Communications. 2015;6:ncomms7404. doi: 10.1038/ncomms7404. [DOI] [PMC free article] [PubMed] [Google Scholar]
Cunniff MM. PogZ_paper. 189f9c5GitHub. 2020 https://github.com/mcunniff/PogZ_paper
De Rubeis S, He X, Goldberg AP, Poultney CS, Samocha K, Cicek AE, Kou Y, Liu L, Fromer M, Walker S, Singh T, Klei L, Kosmicki J, Shih-Chen F, Aleksic B, Biscaldi M, Bolton PF, Brownfeld JM, Cai J, Campbell NG, Carracedo A, Chahrour MH, Chiocchetti AG, Coon H, Crawford EL, Curran SR, Dawson G, Duketis E, Fernandez BA, Gallagher L, Geller E, Guter SJ, Hill RS, Ionita-Laza J, Jimenz Gonzalez P, Kilpinen H, Klauck SM, Kolevzon A, Lee I, Lei I, Lei J, Lehtimäki T, Lin CF, Ma'ayan A, Marshall CR, McInnes AL, Neale B, Owen MJ, Ozaki N, Parellada M, Parr JR, Purcell S, Puura K, Rajagopalan D, Rehnström K, Reichenberg A, Sabo A, Sachse M, Sanders SJ, Schafer C, Schulte-Rüther M, Skuse D, Stevens C, Szatmari P, Tammimies K, Valladares O, Voran A, Li-San W, Weiss LA, Willsey AJ, Yu TW, Yuen RK, Cook EH, Freitag CM, Gill M, Hultman CM, Lehner T, Palotie A, Schellenberg GD, Sklar P, State MW, Sutcliffe JS, Walsh CA, Scherer SW, Zwick ME, Barett JC, Cutler DJ, Roeder K, Devlin B, Daly MJ, Buxbaum JD, DDD Study. Homozygosity Mapping Collaborative for Autism. UK10K Consortium Synaptic, transcriptional and chromatin genes disrupted in autism. Nature. 2014;515:209–215. doi: 10.1038/nature13772. [DOI] [PMC free article] [PubMed] [Google Scholar]
Dentici ML, Niceta M, Pantaleoni F, Barresi S, Bencivenga P, Dallapiccola B, Digilio MC, Tartaglia M. Expanding the phenotypic spectrum of truncating POGZ mutations: association with CNS malformations, skeletal abnormalities, and distinctive facial dysmorphism. American Journal of Medical Genetics Part A. 2017;173:1965–1969. doi: 10.1002/ajmg.a.38255. [DOI] [PubMed] [Google Scholar]
Ellwood IT, Patel T, Wadia V, Lee AT, Liptak AT, Bender KJ, Sohal VS. Tonic or phasic stimulation of dopaminergic projections to prefrontal cortex causes mice to maintain or deviate from previously learned behavioral strategies. The Journal of Neuroscience. 2017;37:8315–8329. doi: 10.1523/JNEUROSCI.1221-17.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fitzgerald TW, Gerety SS, Jones WD, Van Kogelenberg M, King DA, McRae J, Morley K, Parthiban V, Al-Turki S, Ambridge K, Barrett DM, Bayzetinova T, Clayton S, Coomber EL, Gribble S, Jones P, Krishnappa N, Mason LE, Middleton A, Hurles ME, Deciphering Developmental Disorders Study Large-scale discovery of novel genetic causes of developmental disorders. Nature. 2015;519:223–228. doi: 10.1038/nature14135. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fromer M, Pocklington AJ, Kavanagh DH, Williams HJ, Dwyer S, Gormley P, Georgieva L, Rees E, Palta P, Ruderfer DM, Carrera N, Humphreys I, Johnson JS, Roussos P, Barker DD, Banks E, Milanova V, Grant SG, Hannon E, Rose SA, Chambert K, Mahajan M, Scolnick EM, Moran JL, Kirov G, Palotie A, McCarroll SA, Holmans P, Sklar P, Owen MJ, Purcell SM, O'Donovan MC. De novo mutations in schizophrenia implicate synaptic networks. Nature. 2014;506:179–184. doi: 10.1038/nature12929. [DOI] [PMC free article] [PubMed] [Google Scholar]
Fukai R, Hiraki Y, Yofune H, Tsurusaki Y, Nakashima M, Saitsu H, Tanaka F, Miyake N, Matsumoto N. A case of autism spectrum disorder arising from a de novo missense mutation in POGZ. Journal of Human Genetics. 2015;60:277–279. doi: 10.1038/jhg.2015.13. [DOI] [PubMed] [Google Scholar]
Gilissen C, Hehir-Kwa JY, Thung DT, van de Vorst M, van Bon BW, Willemsen MH, Kwint M, Janssen IM, Hoischen A, Schenck A, Leach R, Klein R, Tearle R, Bo T, Pfundt R, Yntema HG, de Vries BB, Kleefstra T, Brunner HG, Vissers LE, Veltman JA. Genome sequencing identifies major causes of severe intellectual disability. Nature. 2014;511:344–347. doi: 10.1038/nature13394. [DOI] [PubMed] [Google Scholar]
Gulsuner S, Walsh T, Watts AC, Lee MK, Thornton AM, Casadei S, Rippey C, Shahin H, Nimgaonkar VL, Go RC, Savage RM, Swerdlow NR, Gur RE, Braff DL, King MC, McClellan JM, Consortium on the Genetics of Schizophrenia (COGS) PAARTNERS Study Group Spatial and temporal mapping of de novo mutations in schizophrenia to a fetal prefrontal cortical network. Cell. 2013;154:518–529. doi: 10.1016/j.cell.2013.06.049. [DOI] [PMC free article] [PubMed] [Google Scholar]
Guo W, Liu F, Xiao C, Liu J, Yu M, Zhang Z, Zhang J, Zhao J. Increased short-range and long-range functional connectivity in first-episode, medication-naive schizophrenia at rest. Schizophrenia Research. 2015;166:144–150. doi: 10.1016/j.schres.2015.04.034. [DOI] [PubMed] [Google Scholar]
Haider B, Schulz DP, Häusser M, Carandini M. Millisecond coupling of local field potentials to synaptic currents in the awake visual cortex. Neuron. 2016;90:35–42. doi: 10.1016/j.neuron.2016.02.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hashimoto R, Nakazawa T, Tsurusaki Y, Yasuda Y, Nagayasu K, Matsumura K, Kawashima H, Yamamori H, Fujimoto M, Ohi K, Umeda-Yano S, Fukunaga M, Fujino H, Kasai A, Hayata-Takano A, Shintani N, Takeda M, Matsumoto N, Hashimoto H. Whole-exome sequencing and neurite outgrowth analysis in autism spectrum disorder. Journal of Human Genetics. 2016;61:199–206. doi: 10.1038/jhg.2015.141. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hultman R, Mague SD, Li Q, Katz BM, Michel N, Lin L, Wang J, David LK, Blount C, Chandy R, Carlson D, Ulrich K, Carin L, Dunson D, Kumar S, Deisseroth K, Moore SD, Dzirasa K. Dysregulation of prefrontal Cortex-Mediated Slow-Evolving limbic dynamics drives Stress-Induced emotional pathology. Neuron. 2016;91:439–452. doi: 10.1016/j.neuron.2016.05.038. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hultman R, Ulrich K, Sachs BD, Blount C, Carlson DE, Ndubuizu N, Bagot RC, Parise EM, Vu MT, Gallagher NM, Wang J, Silva AJ, Deisseroth K, Mague SD, Caron MG, Nestler EJ, Carin L, Dzirasa K. Brain-wide electrical spatiotemporal dynamics encode depression vulnerability. Cell. 2018;173:166–180. doi: 10.1016/j.cell.2018.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
Hyvärinen A, Oja E. Independent component analysis: algorithms and applications. Neural Networks. 2000;13:411–430. doi: 10.1016/S0893-6080(00)00026-5. [DOI] [PubMed] [Google Scholar]
Iossifov I, Ronemus M, Levy D, Wang Z, Hakker I, Rosenbaum J, Yamrom B, Lee YH, Narzisi G, Leotta A, Kendall J, Grabowska E, Ma B, Marks S, Rodgers L, Stepansky A, Troge J, Andrews P, Bekritsky M, Pradhan K, Ghiban E, Kramer M, Parla J, Demeter R, Fulton LL, Fulton RS, Magrini VJ, Ye K, Darnell JC, Darnell RB, Mardis ER, Wilson RK, Schatz MC, McCombie WR, Wigler M. De novo gene disruptions in children on the autistic spectrum. Neuron. 2012;74:285–299. doi: 10.1016/j.neuron.2012.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
Iossifov I, O'Roak BJ, Sanders SJ, Ronemus M, Krumm N, Levy D, Stessman HA, Witherspoon KT, Vives L, Patterson KE, Smith JD, Paeper B, Nickerson DA, Dea J, Dong S, Gonzalez LE, Mandell JD, Mane SM, Murtha MT, Sullivan CA, Walker MF, Waqar Z, Wei L, Willsey AJ, Yamrom B, Lee YH, Grabowska E, Dalkic E, Wang Z, Marks S, Andrews P, Leotta A, Kendall J, Hakker I, Rosenbaum J, Ma B, Rodgers L, Troge J, Narzisi G, Yoon S, Schatz MC, Ye K, McCombie WR, Shendure J, Eichler EE, State MW, Wigler M. The contribution of de novo coding mutations to autism spectrum disorder. Nature. 2014;515:216–221. doi: 10.1038/nature13908. [DOI] [PMC free article] [PubMed] [Google Scholar]
Jacinto LR, Cerqueira JJ, Sousa N. Patterns of theta activity in limbic anxiety circuit preceding exploratory behavior in Approach-Avoidance conflict. Frontiers in Behavioral Neuroscience. 2016;10:171. doi: 10.3389/fnbeh.2016.00171. [DOI] [PMC free article] [PubMed] [Google Scholar]
Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity. Brain. 2004;127:1811–1821. doi: 10.1093/brain/awh199. [DOI] [PubMed] [Google Scholar]
Just MA, Keller TA, Malave VL, Kana RK, Varma S. Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity. Neuroscience & Biobehavioral Reviews. 2012;36:1292–1313. doi: 10.1016/j.neubiorev.2012.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kam JW, Bolbecker AR, O'Donnell BF, Hetrick WP, Brenner CA. Resting state EEG power and coherence abnormalities in bipolar disorder and schizophrenia. Journal of Psychiatric Research. 2013;47:1893–1901. doi: 10.1016/j.jpsychires.2013.09.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kana RK, Uddin LQ, Kenet T, Chugani D, Müller RA. Brain connectivity in autism. Frontiers in Human Neuroscience. 2014;8:349. doi: 10.3389/fnhum.2014.00349. [DOI] [PMC free article] [PubMed] [Google Scholar]
Keehn B, Wagner JB, Tager-Flusberg H, Nelson CA. Functional connectivity in the first year of life in infants at-risk for autism: a preliminary near-infrared spectroscopy study. Frontiers in Human Neuroscience. 2013;7:444. doi: 10.3389/fnhum.2013.00444. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kirkby LA, Luongo FJ, Lee MB, Nahum M, Van Vleet TM, Rao VR, Dawes HE, Chang EF, Sohal VS. An Amygdala-Hippocampus subnetwork that encodes variation in human mood. Cell. 2018;175:1688–1700. doi: 10.1016/j.cell.2018.10.005. [DOI] [PubMed] [Google Scholar]
Kjaerby C, Athilingam J, Robinson SE, Iafrati J, Sohal VS. Serotonin 1B receptors regulate prefrontal function by gating callosal and hippocampal inputs. Cell Reports. 2016;17:2882–2890. doi: 10.1016/j.celrep.2016.11.036. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kjelstrup KG, Tuvnes FA, Steffenach HA, Murison R, Moser EI, Moser MB. Reduced fear expression after lesions of the ventral Hippocampus. PNAS. 2002;99:10825–10830. doi: 10.1073/pnas.152112399. [DOI] [PMC free article] [PubMed] [Google Scholar]
Krumm N, O'Roak BJ, Shendure J, Eichler EE. A de novo convergence of autism genetics and molecular neuroscience. Trends in Neurosciences. 2014;37:95–105. doi: 10.1016/j.tins.2013.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lee E, Lee J, Kim E. Excitation/Inhibition imbalance in animal models of autism spectrum disorders. Biological Psychiatry. 2017;81:838–847. doi: 10.1016/j.biopsych.2016.05.011. [DOI] [PubMed] [Google Scholar]
Lee AT, Cunniff MM, See JZ, Wilke SA, Luongo FJ, Ellwood IT, Ponnavolu S, Sohal VS. VIP interneurons contribute to avoidance behavior by regulating information flow across Hippocampal-Prefrontal networks. Neuron. 2019;102:1223–1234. doi: 10.1016/j.neuron.2019.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS, Brockway KS, Byrnes EJ, Chen L, Chen L, Chen TM, Chin MC, Chong J, Crook BE, Czaplinska A, Dang CN, Datta S, Dee NR, Desaki AL, Desta T, Diep E, Dolbeare TA, Donelan MJ, Dong HW, Dougherty JG, Duncan BJ, Ebbert AJ, Eichele G, Estin LK, Faber C, Facer BA, Fields R, Fischer SR, Fliss TP, Frensley C, Gates SN, Glattfelder KJ, Halverson KR, Hart MR, Hohmann JG, Howell MP, Jeung DP, Johnson RA, Karr PT, Kawal R, Kidney JM, Knapik RH, Kuan CL, Lake JH, Laramee AR, Larsen KD, Lau C, Lemon TA, Liang AJ, Liu Y, Luong LT, Michaels J, Morgan JJ, Morgan RJ, Mortrud MT, Mosqueda NF, Ng LL, Ng R, Orta GJ, Overly CC, Pak TH, Parry SE, Pathak SD, Pearson OC, Puchalski RB, Riley ZL, Rockett HR, Rowland SA, Royall JJ, Ruiz MJ, Sarno NR, Schaffnit K, Shapovalova NV, Sivisay T, Slaughterbeck CR, Smith SC, Smith KA, Smith BI, Sodt AJ, Stewart NN, Stumpf KR, Sunkin SM, Sutram M, Tam A, Teemer CD, Thaller C, Thompson CL, Varnam LR, Visel A, Whitlock RM, Wohnoutka PE, Wolkey CK, Wong VY, Wood M, Yaylaoglu MB, Young RC, Youngstrom BL, Yuan XF, Zhang B, Zwingman TA, Jones AR. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 2007;445:168–176. doi: 10.1038/nature05453. [DOI] [PubMed] [Google Scholar]
Lopes-dos-Santos V, Ribeiro S, Tort AB. Detecting cell assemblies in large neuronal populations. Journal of Neuroscience Methods. 2013;220:149–166. doi: 10.1016/j.jneumeth.2013.04.010. [DOI] [PubMed] [Google Scholar]
Mathis A, Mamidanna P, Cury KM, Abe T, Murthy VN, Mathis MW, Bethge M. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nature Neuroscience. 2018;21:1281–1289. doi: 10.1038/s41593-018-0209-y. [DOI] [PubMed] [Google Scholar]
Matsumura K, Nakazawa T, Nagayasu K, Gotoda-Nishimura N, Kasai A, Hayata-Takano A, Shintani N, Yamamori H, Yasuda Y, Hashimoto R, Hashimoto H. De novo POGZ mutations in sporadic autism disrupt the DNA-binding activity of POGZ. Journal of Molecular Psychiatry. 2016;4:1. doi: 10.1186/s40303-016-0016-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Matsumura K, Seiriki K, Okada S, Nagase M, Ayabe S, Yamada I, Furuse T, Shibuya H, Yasuda Y, Yamamori H, Fujimoto M, Nagayasu K, Yamamoto K, Kitagawa K, Miura H, Gotoda-Nishimura N, Igarashi H, Hayashida M, Baba M, Kondo M, Hasebe S, Ueshima K, Kasai A, Ago Y, Hayata-Takano A, Shintani N, Iguchi T, Sato M, Yamaguchi S, Tamura M, Wakana S, Yoshiki A, Watabe AM, Okano H, Takuma K, Hashimoto R, Hashimoto H, Nakazawa T. Pathogenic POGZ mutation causes impaired cortical development and reversible autism-like phenotypes. Nature Communications. 2020;11:1–16. doi: 10.1038/s41467-020-14697-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mueller S, Keeser D, Samson AC, Kirsch V, Blautzik J, Grothe M, Erat O, Hegenloh M, Coates U, Reiser MF, Hennig-Fast K, Meindl T. Convergent findings of altered functional and structural brain connectivity in individuals with high functioning autism: a multimodal MRI study. PLOS ONE. 2013;8:e67329. doi: 10.1371/journal.pone.0067329. [DOI] [PMC free article] [PubMed] [Google Scholar]
Müller RA, Shih P, Keehn B, Deyoe JR, Leyden KM, Shukla DK. Underconnected, but how? A survey of functional connectivity MRI studies in autism spectrum disorders. Cerebral Cortex. 2011;21:2233–2243. doi: 10.1093/cercor/bhq296. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mullins C, Fishell G, Tsien RW. Unifying views of autism spectrum disorders: a consideration of autoregulatory feedback loops. Neuron. 2016;89:1131–1156. doi: 10.1016/j.neuron.2016.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nair A, Treiber JM, Shukla DK, Shih P, Müller RA. Impaired thalamocortical connectivity in autism spectrum disorder: a study of functional and anatomical connectivity. Brain. 2013;136:1942–1955. doi: 10.1093/brain/awt079. [DOI] [PMC free article] [PubMed] [Google Scholar]
Neale BM, Kou Y, Liu L, Ma'ayan A, Samocha KE, Sabo A, Lin CF, Stevens C, Wang LS, Makarov V, Polak P, Yoon S, Maguire J, Crawford EL, Campbell NG, Geller ET, Valladares O, Schafer C, Liu H, Zhao T, Cai G, Lihm J, Dannenfelser R, Jabado O, Peralta Z, Nagaswamy U, Muzny D, Reid JG, Newsham I, Wu Y, Lewis L, Han Y, Voight BF, Lim E, Rossin E, Kirby A, Flannick J, Fromer M, Shakir K, Fennell T, Garimella K, Banks E, Poplin R, Gabriel S, DePristo M, Wimbish JR, Boone BE, Levy SE, Betancur C, Sunyaev S, Boerwinkle E, Buxbaum JD, Cook EH, Devlin B, Gibbs RA, Roeder K, Schellenberg GD, Sutcliffe JS, Daly MJ. Patterns and rates of exonic de novo mutations in autism spectrum disorders. Nature. 2012;485:242–245. doi: 10.1038/nature11011. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nelson SB, Valakh V. Excitatory/Inhibitory balance and circuit homeostasis in autism spectrum disorders. Neuron. 2015;87:684–698. doi: 10.1016/j.neuron.2015.07.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
Nozawa RS, Nagao K, Masuda HT, Iwasaki O, Hirota T, Nozaki N, Kimura H, Obuse C. Human POGZ modulates dissociation of HP1alpha from mitotic chromosome arms through aurora B activation. Nature Cell Biology. 2010;12:719–727. doi: 10.1038/ncb2075. [DOI] [PubMed] [Google Scholar]
Padilla-Coreano N, Bolkan SS, Pierce GM, Blackman DR, Hardin WD, Garcia-Garcia AL, Spellman TJ, Gordon JA. Direct ventral Hippocampal-Prefrontal input is required for Anxiety-Related neural activity and behavior. Neuron. 2016;89:857–866. doi: 10.1016/j.neuron.2016.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
Padilla-Coreano N, Canetta S, Mikofsky RM, Alway E, Passecker J, Myroshnychenko MV, Garcia-Garcia AL, Warren R, Teboul E, Blackman DR, Morton MP, Hupalo S, Tye KM, Kellendonk C, Kupferschmidt DA, Gordon JA. Hippocampal-Prefrontal theta transmission regulates avoidance behavior. Neuron. 2019;104:601–610. doi: 10.1016/j.neuron.2019.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
Padmanabhan A, Lynn A, Foran W, Luna B, O'Hearn K. Age related changes in striatal resting state functional connectivity in autism. Frontiers in Human Neuroscience. 2013;7:814. doi: 10.3389/fnhum.2013.00814. [DOI] [PMC free article] [PubMed] [Google Scholar]
Parent MA, Wang L, Su J, Netoff T, Yuan LL. Identification of the hippocampal input to medial prefrontal cortex in vitro. Cerebral Cortex. 2010;20:393–403. doi: 10.1093/cercor/bhp108. [DOI] [PMC free article] [PubMed] [Google Scholar]
Pfaff D, Barbas H. Mechanisms for the approach/Avoidance decision applied to autism. Trends in Neurosciences. 2019;42:448–457. doi: 10.1016/j.tins.2019.05.002. [DOI] [PubMed] [Google Scholar]
Potter GB, Petryniak MA, Shevchenko E, McKinsey GL, Ekker M, Rubenstein JL. Generation of Cre-transgenic mice using Dlx1/Dlx2 enhancers and their characterization in GABAergic interneurons. Molecular and Cellular Neuroscience. 2009;40:167–186. doi: 10.1016/j.mcn.2008.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
Redcay E, Moran JM, Mavros PL, Tager-Flusberg H, Gabrieli JD, Whitfield-Gabrieli S. Intrinsic functional network organization in high-functioning adolescents with autism spectrum disorder. Frontiers in Human Neuroscience. 2013;7:573. doi: 10.3389/fnhum.2013.00573. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ronan JL, Wu W, Crabtree GR. From neural development to cognition: unexpected roles for chromatin. Nature Reviews Genetics. 2013;14:347–359. doi: 10.1038/nrg3413. [DOI] [PMC free article] [PubMed] [Google Scholar]
Rubenstein JL, Merzenich MM. Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes, Brain and Behavior. 2003;2:255–267. doi: 10.1034/j.1601-183X.2003.00037.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sánchez-Bellot C, MacAskill A. Push-pull regulation of exploratory behavior by two opposing hippocampal to prefrontal cortex pathways. bioRxiv. 2019 doi: 10.1101/2019.12.18.880831. [DOI]
Sanders SJ, He X, Willsey AJ, Ercan-Sencicek AG, Samocha KE, Cicek AE, Murtha MT, Bal VH, Bishop SL, Dong S, Goldberg AP, Jinlu C, Keaney JF, Klei L, Mandell JD, Moreno-De-Luca D, Poultney CS, Robinson EB, Smith L, Solli-Nowlan T, Su MY, Teran NA, Walker MF, Werling DM, Beaudet AL, Cantor RM, Fombonne E, Geschwind DH, Grice DE, Lord C, Lowe JK, Mane SM, Martin DM, Morrow EM, Talkowski ME, Sutcliffe JS, Walsh CA, Yu TW, Ledbetter DH, Martin CL, Cook EH, Buxbaum JD, Daly MJ, Devlin B, Roeder K, State MW, Autism Sequencing Consortium Insights into autism spectrum disorder genomic architecture and biology from 71 risk loci. Neuron. 2015;87:1215–1233. doi: 10.1016/j.neuron.2015.09.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
Shah AA, Treit D. Excitotoxic lesions of the medial prefrontal cortex attenuate fear responses in the elevated-plus maze, social interaction and shock probe burying tests. Brain Research. 2003;969:183–194. doi: 10.1016/S0006-8993(03)02299-6. [DOI] [PubMed] [Google Scholar]
Sigurdsson T, Stark KL, Karayiorgou M, Gogos JA, Gordon JA. Impaired hippocampal-prefrontal synchrony in a genetic mouse model of schizophrenia. Nature. 2010;464:763–767. doi: 10.1038/nature08855. [DOI] [PMC free article] [PubMed] [Google Scholar]
Sohal VS, Rubenstein JLR. Excitation-inhibition balance as a framework for investigating mechanisms in neuropsychiatric disorders. Molecular Psychiatry. 2019;24:1248–1257. doi: 10.1038/s41380-019-0426-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Spellman T, Rigotti M, Ahmari SE, Fusi S, Gogos JA, Gordon JA. Hippocampal-prefrontal input supports spatial encoding in working memory. Nature. 2015;522:309–314. doi: 10.1038/nature14445. [DOI] [PMC free article] [PubMed] [Google Scholar]
Stessman HAF, Willemsen MH, Fenckova M, Penn O, Hoischen A, Xiong B, Wang T, Hoekzema K, Vives L, Vogel I, Brunner HG, van der Burgt I, Ockeloen CW, Schuurs-Hoeijmakers JH, Klein Wassink-Ruiter JS, Stumpel C, Stevens SJC, Vles HS, Marcelis CM, van Bokhoven H, Cantagrel V, Colleaux L, Nicouleau M, Lyonnet S, Bernier RA, Gerdts J, Coe BP, Romano C, Alberti A, Grillo L, Scuderi C, Nordenskjöld M, Kvarnung M, Guo H, Xia K, Piton A, Gerard B, Genevieve D, Delobel B, Lehalle D, Perrin L, Prieur F, Thevenon J, Gecz J, Shaw M, Pfundt R, Keren B, Jacquette A, Schenck A, Eichler EE, Kleefstra T. Disruption of POGZ is associated with intellectual disability and autism spectrum disorders. The American Journal of Human Genetics. 2016;98:541–552. doi: 10.1016/j.ajhg.2016.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
Suliman R, Ben-David E, Shifman S. Chromatin regulators, phenotypic robustness, and autism risk. Frontiers in Genetics. 2014;5:81. doi: 10.3389/fgene.2014.00081. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tamura M, Spellman TJ, Rosen AM, Gogos JA, Gordon JA. Hippocampal-prefrontal theta-gamma coupling during performance of a spatial working memory task. Nature Communications. 2017;8:9. doi: 10.1038/s41467-017-02108-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
Tan B, Zou Y, Zhang Y, Zhang R, Ou J, Shen Y, Zhao J, Luo X, Guo J, Zeng L, Hu Y, Zheng Y, Pan Q, Liang D, Wu L. A novel de novo POGZ mutation in a patient with intellectual disability. Journal of Human Genetics. 2016;61:357–359. doi: 10.1038/jhg.2015.156. [DOI] [PubMed] [Google Scholar]
Teleńczuk B, Dehghani N, Le Van Quyen M, Cash SS, Halgren E, Hatsopoulos NG, Destexhe A. Local field potentials primarily reflect inhibitory neuron activity in human and monkey cortex. Scientific Reports. 2017;7:40211. doi: 10.1038/srep40211. [DOI] [PMC free article] [PubMed] [Google Scholar]
Toro R, Konyukh M, Delorme R, Leblond C, Chaste P, Fauchereau F, Coleman M, Leboyer M, Gillberg C, Bourgeron T. Key role for gene dosage and synaptic homeostasis in autism spectrum disorders. Trends in Genetics. 2010;26:363–372. doi: 10.1016/j.tig.2010.05.007. [DOI] [PubMed] [Google Scholar]
Vinck M, Oostenveld R, van Wingerden M, Battaglia F, Pennartz CM. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size Bias. NeuroImage. 2011;55:1548–1565. doi: 10.1016/j.neuroimage.2011.01.055. [DOI] [PubMed] [Google Scholar]
Wang X, Xia M, Lai Y, Dai Z, Cao Q, Cheng Z, Han X, Yang L, Yuan Y, Zhang Y, Li K, Ma H, Shi C, Hong N, Szeszko P, Yu X, He Y. Disrupted resting-state functional connectivity in minimally treated chronic schizophrenia. Schizophrenia Research. 2014;156:150–156. doi: 10.1016/j.schres.2014.03.033. [DOI] [PubMed] [Google Scholar]
Wang Y, Wang J, Jia Y, Zhong S, Niu M, Sun Y, Qi Z, Zhao L, Huang L, Huang R. Shared and specific intrinsic functional connectivity patterns in unmedicated bipolar disorder and major depressive disorder. Scientific Reports. 2017;7:8. doi: 10.1038/s41598-017-03777-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
White J, Beck CR, Harel T, Posey JE, Jhangiani SN, Tang S, Farwell KD, Powis Z, Mendelsohn NJ, Baker JA, Pollack L, Mason KJ, Wierenga KJ, Arrington DK, Hall M, Psychogios A, Fairbrother L, Walkiewicz M, Person RE, Niu Z, Zhang J, Rosenfeld JA, Muzny DM, Eng C, Beaudet AL, Lupski JR, Boerwinkle E, Gibbs RA, Yang Y, Xia F, Sutton VR. POGZ truncating alleles cause syndromic intellectual disability. Genome Medicine. 2016;8:3. doi: 10.1186/s13073-015-0253-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
Willsey AJ, Sanders SJ, Li M, Dong S, Tebbenkamp AT, Muhle RA, Reilly SK, Lin L, Fertuzinhos S, Miller JA, Murtha MT, Bichsel C, Niu W, Cotney J, Ercan-Sencicek AG, Gockley J, Gupta AR, Han W, He X, Hoffman EJ, Klei L, Lei J, Liu W, Liu L, Lu C, Xu X, Zhu Y, Mane SM, Lein ES, Wei L, Noonan JP, Roeder K, Devlin B, Sestan N, State MW. Coexpression networks implicate human midfetal deep cortical projection neurons in the pathogenesis of autism. Cell. 2013;155:997–1007. doi: 10.1016/j.cell.2013.10.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wiltschko AB, Johnson MJ, Iurilli G, Peterson RE, Katon JM, Pashkovski SL, Abraira VE, Adams RP, Datta SR. Mapping Sub-Second structure in mouse behavior. Neuron. 2015;88:1121–1135. doi: 10.1016/j.neuron.2015.11.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
Wondolowski J, Dickman D. Emerging links between homeostatic synaptic plasticity and neurological disease. Frontiers in Cellular Neuroscience. 2013;7:223. doi: 10.3389/fncel.2013.00223. [DOI] [PMC free article] [PubMed] [Google Scholar]
Xing M, Tadayonnejad R, MacNamara A, Ajilore O, DiGangi J, Phan KL, Leow A, Klumpp H. Resting-state theta band connectivity and graph analysis in generalized social anxiety disorder. NeuroImage: Clinical. 2017;13:24–32. doi: 10.1016/j.nicl.2016.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
Ye Y, Cho MT, Retterer K, Alexander N, Ben-Omran T, Al-Mureikhi M, Cristian I, Wheeler PG, Crain C, Zand D, Weinstein V, Vernon HJ, McClellan R, Krishnamurthy V, Vitazka P, Millan F, Chung WK. De novo POGZ mutations are associated with neurodevelopmental disorders and microcephaly. Molecular Case Studies. 2015;1:a000455. doi: 10.1101/mcs.a000455. [DOI] [PMC free article] [PubMed] [Google Scholar]
You X, Norr M, Murphy E, Kuschner ES, Bal E, Gaillard WD, Kenworthy L, Vaidya CJ. Atypical modulation of distant functional connectivity by cognitive state in children with autism spectrum disorders. Frontiers in Human Neuroscience. 2013;7:00482. doi: 10.3389/fnhum.2013.00482. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zeng K, Kang J, Ouyang G, Li J, Han J, Wang Y, Sokhadze EM, Casanova MF, Li X. Disrupted brain network in children with autism spectrum disorder. Scientific Reports. 2017;7:16440. doi: 10.1038/s41598-017-16440-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
Zhao W, Tan J, Zhu T, Ou J, Li Y, Shen L, Wu H, Han L, Liu Y, Jia X, Bai T, Li H, Ke X, Zhao J, Zou X, Hu Z, Guo H, Xia K. Rare inherited missense variants of POGZ associate with autism risk and disrupt neuronal development. Journal of Genetics and Genomics. 2019;46:247–257. doi: 10.1016/j.jgg.2019.04.002. [DOI] [PubMed] [Google Scholar]

eLife. doi: 10.7554/eLife.54835.sa1

Decision letter

Editor: Laura L Colgin¹

Reviewed by: Alex Harris²

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Acceptance summary:

The authors find that mice deficient for the ASD-linked gene POGZ show decreased anxiety, impaired theta synchrony during anxiety assays, and decreased hippocampal input to mPFC cortical fast-spiking cells. These findings will be of interest to those working as autism, as POGZ is a high confidence autism gene. In combination with other similar studies, these results could provide a means for assessing whether potential new treatments can alleviate deficits in animal models.

Decision letter after peer review:

Thank you for submitting your article "Altered hippocampal-prefrontal communication and anxiety-related behavior in mice deficient for the ASD-linked gene POGZ" for consideration by eLife. Your article has been reviewed by three peer reviewers, and the evaluation has been overseen by Laura Colgin as the Reviewing and Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Alex Harris (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). Specifically, when editors judge that a submitted work as a whole belongs in eLife but that some conclusions require a modest amount of additional new data, as they do with your paper, we are asking that the manuscript be revised to either limit claims to those supported by data in hand, or to explicitly state that the relevant conclusions require additional supporting data.

In the latter case, our expectation is that the authors will eventually carry out the additional experiments and report on how they affect the relevant conclusions either in a preprint on bioRxiv or medRxiv, or if appropriate, as a Research Advance in eLife, either of which would be linked to the original paper. If the claims in question are instead removed, additional experiments would not be expected.

Summary:

Cunniff et al. provide a comprehensive account of vHPC/mPFC-dependent behavior and neurophysiological abnormalities in mice harboring heterozygous loss-of-function mutations in the ASD-associated Pogz chromatin remodelling gene (Pogz^+/- mice). Their interpretation of results is supported using a computational model, predicting that reduced excitatory drive onto mPFC FSINs decreases the correlation between mPFC pyramidal cell output and hippocampal input. The impressive breadth of techniques constitutes a logical and complementary approach to delineating vHPC-mPFC abnormalities in Pogz mice.

However, reviewers agreed that some of the major claims made by the authors were not well supported by their data. Moreover, reviewers agreed that it is unclear whether the deficits these Pogz mice display are relevant to autism and felt that the translational claims made in the paper should be tempered. Additionally, there are some few presentation issues that need to be addressed.

Revisions for this paper:

Reviewers felt that the authors cannot claim any specificity of the decreased hippocampal input to mPFC cortical fast-spiking cells for the reported behavioral and theta synchrony findings, as one would intuit brain-wide changes as a result of disrupting a gene implicated in such universally important processes. Also, the translational claims and links to autism were viewed as overstated. There were other presentation and analysis issues that should also be addressed. See Major Comments below for details.

Revisions expected in follow-up work:

If authors choose to keep the overstated causal relationship between physiological and behavioral findings in the paper, and the unsupported specificity claims, then appropriate rescue experiments would be expected in the future.

Major Comments:

1) The authors find that Pogz^+/- mice show: decreased anxiety, impaired theta synchrony during anxiety assays and decreased hippocampal input to mPFC cortical fast-spiking cells. Overall, these findings are interesting and relevant to autism, as Pogz is a high confidence autism gene. The three findings may interact with each other to produce the anxiety behavioral deficit Pogz mice show. That is the narrative the authors favor. However, it is important to consider that as the authors state, Pogz "is known to play a role in chromatin regulation, mitotic progression, and chromosome segregation ". These are ubiquitous processes that likely affect a myriad of cell types and countless circuits in the brain. Thus, from the data shown, it is not possible to support the conclusions the authors claim.

For example, the theta synchrony deficit may be related to altered hippocampal input to cortical interneurons as the authors state. But it is also entirely possible that the observed theta synchrony deficit is due to changes in some neuromodulatory input to cortex in Pogz mice, or due to some other change induced by Pogz. The point is that Pogz affects so many developmental processes that it is not possible to state with any reasonable confidence that input to fast spiking interneurons causes the observed that synchrony deficit. If Pogz only selectively affected the development of hippocampal input to cortical interneurons, then the authors would be justified in their claims. Similarly, the behavioral deficit may be related to alterations in theta synchrony, but it may also be due to some other completely unrelated function of Pogz during development.

To truly support the claims, the authors would have to show that somehow rescuing or normalizing hippocampal input to fast spiking interneurons would both normalize theta synchrony deficits and behavioral symptoms displayed by Pogz mice. As it is not possible to perform this experiment, the authors should explicitly and clearly write that their findings provide a plausible, possible explanation for the behavioral deficit. But these findings do not prove that hippocampal inputs to fast spiking interneurons in cortex cause lower anxiety in Pogz mice. Unsubstantiated claims such as "We found deficient theta-frequency synchronization between the vHPC and mPFC in vivo. Furthermore, this involves a specific deficit in excitatory input from vHPC onto prefrontal GABAergic interneurons" should be removed from the paper. The authors do not show that the theta synchronization deficit involves a deficit in vhpc input to interneurons. They only show these two deficits exist, not that one is involved in the occurrence of the other. Similarly, the claim "we are able to show that the theta coordination of vHPC-mPFC activity and open arm avoidance can be disrupted simply by suppressing vHPC input to interneurons" is unsupported. The authors did not show a mechanistic connection between their 3 main findings. They only show 3 differences between Pogz and WT mice. On a positive note, later the authors state "we found reduced excitatory drive from vHPC onto fast-spiking interneurons. This synaptic abnormality could plausibly contribute to the abnormalities we found in both avoidance behavior and LFP synchrony". This is the appropriate interpretation of the results, and this is the tone that should be used throughout the paper.

The alternative argument that Pogz functions as a tool to study selective disruption of excitatory input to PFC interneurons is intriguing but would change the focus of the paper to establishing that claim by providing a) the mechanism for such selectivity and b) demonstrating that it exists in vivo.

2) Of note, a very recent Matsumura et al., 2020, paper generates a different line of Pogz^+/- mice, also reporting that these mice spend more time in the central zone of an open field, alongside cortical thinning and hyperactivation of anterior cingulate cortex during social behavior. Importantly, Matsumara et al. rescue abnormalities using the AMPA receptor negative allosteric modulator perampanel. It would be of great interest to see if this same approach could alleviate the deficits in Cunniff et al.'s model. Reference to the Matsumura et al., 2020, and relevant comparisons with the present work will have to be added in Introduction and Discussion.

3) The authors should delete or tone down the "Clinical and therapeutical implications" section. It is unfeasible at this point to begin to plan to implement a therapy for autism based on these findings. There are no known methods that can selectively control hippocampal inputs to fast spiking cortical neurons in human patients. The authors summarize this state of affairs as "it is not immediately obvious how one would translate our findings into new treatments", making this whole section highly speculative.

4) Related to the above points, face validity may be an imperfect metric of translational utility, but that doesn't mean that the disruption of any behavior involving the prefrontal cortex has relevance for autism/intellectual disability. An alternative explanation is that the gene functions differently in mice and people. This is particularly the case since tests for the expected deficits (social and intellectual functioning) appeared intact. However, it may be premature to make this claim: these latter tests are underpowered (1/3 to 1/2 of the N used for EPM), and a recent paper (Matsumura et al., 2020) found social and cognitive deficits in Pogz missense mutation mice. To be translationally relevant, the authors would need to show that the long-range synchrony deficits underlie social or cognitive behavioral deficits. The Pogz mice in this study show only a very selective behavioral deficit in lower anxiety, which is not a typical autism symptom. These data raise questions on the validity of using Pogz mice for modeling autism phenotypes. This issue must be discussed clearly in the Discussion.

5) Materials and methods state that LFP electrode placements were checked histologically. Were any mice excluded on these grounds? It becomes important to show these data in light of a related recent paper by Matsumura et al., 2020, showing smaller brains in a different line of Pogz^+/- mice. It is therefore feasible that electrode placement was systematically different between WT and+/- mice, contributing to WPLI and other neurophysiological differences.

6) How is it possible to have hundreds of biological replicates of the weighted phase locking measure with only ~20 mice?

7) Given that WPLI was reduced in both homecage and in EPM, presumably it is also reduced on the T-maze, yet Pogz^+/- delayed alternation was apparently normal. Why is this, in light of all the work from the Gordon lab? Why would the phenotype preferentially manifest on the EPM?

8) Figure 3: Why is this particular IC highlighted? (i) How is the appropriate threshold for significant ICs calculated? (ii) How do the significant ICs in 3C relate to the points in the clusters in 3D? (iii) What is this justification for focus on the cluster that happens to correspond to theta-related to cross-frequency (phase-amplitude) coupling? And (iv) how does this relate to theta WPLI?

9) The authors argue based on their in vitro data and modeling that decreased vHPC excitatory input to FSIN are responsible for the Pogz deficits in theta synchrony. However, this model would predict mPFC spike phase-locking with vHPC theta (which they don't show) not theta phase synchrony or cross-frequency coupling. They should either show decreased spike phase locking or a model that accounts for their synchrony observations. It would be powerful to use the computational model to explore different frequency ranges of HPC input: does the model predict specific effects on theta-frequency inputs?

10) Does homecage WPLI predict EPM behavior in individual animals?

11) This model is proposed as an alternative to the suggestion that vHPC provides direct excitatory input to pyramidal neurons, but it doesn't address Padilla-Coreano et al., 2016 findings that blocking excitatory input did not affect overall firing rates of either pyramidal or interneurons but decreased firing rates in the preferred arm. By their model, shouldn't firing rates increase in the non-preferred arm?

12) Figure 2: Was the directionality of the HPC-PFC WPLI measure consistent with an effect on HPC drive to PFC?

13) Figure 3 needs to be expanded and clarified:

a) How was cross-frequency coupling calculated? The Materials and methods state that "Cross-frequency coupling was calculated by comparing the instantaneous phase in a high frequency band with the instantaneous amplitude in a low frequency band". Usually the converse would be true, but still more details are required – was an established method like the Modulation Index used?

b) Although it is possible to understand the results in Figure 3, it is not very reader-friendly. The authors should add some panels explaining each step in this analysis, showing visually how they go from raw LFP recordings to the plots shown in Figure 3.

c) There is insufficient information given about the empirical salient feature analysis. Figures 3A-D are not labeled in a way that readers can decipher the identities of features. The authors should show how all the features fared throughout the pipeline. Moreover, why doesn't this data analysis strategy pull out the known finding (theta phase synchrony)?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for submitting your article "Altered hippocampal-prefrontal communication and anxiety-related behavior in mice deficient for the ASD-linked gene POGZ" for consideration by eLife. Your revised article has been reviewed by three peer reviewers, and the evaluation has been overseen by Laura Colgin as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Alex Harris (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

As the editors have judged that your manuscript is of interest, but as described below that additional analyses are required before a final decision can be made.

We would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

Summary:

Many positive points were noted about the revised submission. This resubmission addressed key concerns by reducing claims of causality and translational relevance. The authors make a reasonable case justifying the relevance of their behavioral and circuit finding. The authors give a reasonable explanation for why delayed alternation might be spared in the T-maze. They have done analyses to address the impact of homecage WLPI and sex on their findings. They analyze the directionality of the synchrony effect. They have added new modeling showing frequency specific effects. They have rewritten the text to clarify several of the points. However, some points were incompletely addressed, and some new concerns were raised by the responses to the first set of critiques. Reviewers agreed that robust evidence for the paper's conclusions requires additional analyses, and thus the ultimate decision on the manuscript will depend on the outcome of these analyses.

Essential revisions:

1) Even though it occurred in equal numbers across genotype, a concern was raised that in a paper focused on theta, 38% (8/21) of the mice were excluded for lacking a clear theta peak (despite having good histology). Is there any explanation for why theta is aberrant in such a large fraction of the data? What would happen to the findings if these mice were included? The authors should include any mice that don't have a technical problem with the electrodes or location. An alternative approach to point 1 would be to analyze the 4+4 “non theta” mice separately in a supplement.

2) In general, it seems that people tend to treat units recorded from the same animal as independent but consider LFP data recorded with the same electrode as a single biological replicate (and trials as experimental replicates). The authors should use a repeated measures design, use the average LFP per mouse, or some other statistically valid approach for their analyses.

3) The explanation of Figure 3 is improved, but questions remain about the procedure. It seems that ICs generated from a combination of WT and POGZ mice (should the number of mice be 13?). How was consistency across mice measured? By pairwise correlations? How did this procedure measure consistency across all mice? Was there any consideration given to whether ICs were consistent within or across genotype?

4) What percent of high correlation ICs contained a measure of theta synchrony? Similarly, what percent of ICs were task-modulated? Without these measures, it is hard to interpret the significance of a given IC's task modulation and genotype disruption.

5) The explanation of why ICA did not yield theta phase synchrony was unclear. Isn't theta synchrony a stable “network” conserved across mice? Doesn't that serve as an important positive control of this empiric method?

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your article "Altered hippocampal-prefrontal communication and anxiety-related behavior in mice deficient for the ASD-linked gene POGZ" for consideration by eLife. Your revised article has been reviewed by two peer reviewers, and the evaluation has been overseen by Laura Colgin as the Senior Editor and Reviewing Editor. The following individual involved in review of your submission has agreed to reveal their identity: Alex Harris (Reviewer #3).

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The authors have adequately addressed the discussion points. However, regarding the results, reviewers remain concerned about the analysis and agree that analysis concerns that were raised previously were not adequately addressed in the revised manuscript. Revisions that are considered essential for acceptance are explained below.

Essential revisions:

1) Reviewers remain concerned about the authors' decision to ignore the recommendation to use a repeated measures analysis for some of the reported results, in which different trials from the same mouse were treated as independent. Based on the statistics currently used, it is not convincing to draw the conclusion that theta band WLPI is task-modulated in WT, but not PogZ^+/- mice. If there happened to be a dependency of the trials due to the mouse, then z-scoring wouldn't remove it. In other words, there could be some unknown reason why impairments are strong in some animals and not others (e.g., some interaction between the genotype and individual experiences during development), and z-scoring would not take care of that. Reviewers agree that a repeated measures analysis should be used. Also, for the effects in some frequency bands and not others, again, a repeated measures analysis would be best, as power in different frequency bands within the same recordings is also not independent. It is important to remember that, at eLife, the decision letter containing the consensus review and recommendations is published along with the article. So, we think that readers would wonder why authors did not follow the suggestion to apply a repeated measures analysis, and this could lower confidence in the results and diminish the paper's impact.

2) The authors should also include the exact statistics and degrees of freedom, APA style, not just p-values when reporting effects. Also, reviewers think it would be best to request that all the figure legends list N, name of statistical test, and p value, so that it is easier for readers to understand how data were analyzed. Currently some figures (such as Figure 1) list only p values.

3) The final N for Figure 3E and Figure 3—figure supplement 1D-F remains unclear. Are these N also based on z-scored trial data? If so, they are subject to the same statistical concern mentioned above.

eLife. 2020 Nov 6;9:e54835. doi: 10.7554/eLife.54835.sa2

Author response

Major Comments:

1) The authors find that Pogz^+/- mice show: decreased anxiety, impaired theta synchrony during anxiety assays and decreased hippocampal input to mPFC cortical fast-spiking cells. Overall, these findings are interesting and relevant to autism, as Pogz is a high confidence autism gene. The three findings may interact with each other to produce the anxiety behavioral deficit Pogz mice show. That is the narrative the authors favor. However, it is important to consider that as the authors state, Pogz "is known to play a role in chromatin regulation, mitotic progression, and chromosome segregation ". These are ubiquitous processes that likely affect a myriad of cell types and countless circuits in the brain. Thus, from the data shown, it is not possible to support the conclusions the authors claim.

For example, the theta synchrony deficit may be related to altered hippocampal input to cortical interneurons as the authors state. But it is also entirely possible that the observed theta synchrony deficit is due to changes in some neuromodulatory input to cortex in Pogz mice, or due to some other change induced by Pogz. The point is that Pogz affects so many developmental processes that it is not possible to state with any reasonable confidence that input to fast spiking interneurons causes the observed that synchrony deficit. If Pogz only selectively affected the development of hippocampal input to cortical interneurons, then the authors would be justified in their claims. Similarly, the behavioral deficit may be related to alterations in theta synchrony, but it may also be due to some other completely unrelated function of Pogz during development.

To truly support the claims, the authors would have to show that somehow rescuing or normalizing hippocampal input to fast spiking interneurons would both normalize theta synchrony deficits and behavioral symptoms displayed by Pogz mice. As it is not possible to perform this experiment, the authors should explicitly and clearly write that their findings provide a plausible, possible explanation for the behavioral deficit. But these findings do not prove that hippocampal inputs to fast spiking interneurons in cortex cause lower anxiety in Pogz mice. Unsubstantiated claims such as "We found deficient theta-frequency synchronization between the vHPC and mPFC in vivo. Furthermore, this involves a specific deficit in excitatory input from vHPC onto prefrontal GABAergic interneurons" should be removed from the paper. The authors do not show that the theta synchronization deficit involves a deficit in vhpc input to interneurons. They only show these two deficits exist, not that one is involved in the occurrence of the other. Similarly, the claim "we are able to show that the theta coordination of vHPC-mPFC activity and open arm avoidance can be disrupted simply by suppressing vHPC input to interneurons" is unsupported. The authors did not show a mechanistic connection between their 3 main findings. They only show 3 differences between Pogz and WT mice. On a positive note, later the authors state "we found reduced excitatory drive from vHPC onto fast-spiking interneurons. This synaptic abnormality could plausibly contribute to the abnormalities we found in both avoidance behavior and LFP synchrony". This is the appropriate interpretation of the results, and this is the tone that should be used throughout the paper.

We agree with the point made by the Reviewers: we did not directly test causal relationships between observed physiological and behavioral abnormalities. Rather, the physiological abnormalities occur in a pathway that is known to influence the behaviors we measured, and thus constitute a plausible mechanism underlying behavioral abnormalities. However, Pogz influences many developmental processes, so we cannot rule out that Pogz may elicit behavioral abnormalities through alterations in other processes which we have not examined. We had tried to be circumspect about our claims in the original manuscript, but the reviewers’ feedback makes clear that in many cases the exact meaning of our text was ambiguous. Therefore, we have followed the reviewers’ recommendation to revise our text to make the nature of our claims more clear.

Specifically, we revised the last three sentences of the Abstract as follows:

“We found deficient theta-frequency synchronization between the vHPC and mPFC in vivo. When we examined vHPC-mPFC communication at higher resolution, vHPC input onto prefrontal GABAergic interneurons was specifically disrupted, whereas input onto pyramidal neurons remained intact. These findings illustrate how the loss of a high confidence autism gene can impair long-range communication by causing inhibitory circuit dysfunction within pathways important for specific behaviors.”

The alternative argument that Pogz functions as a tool to study selective disruption of excitatory input to PFC interneurons is intriguing but would change the focus of the paper to establishing that claim by providing a) the mechanism for such selectivity and b) demonstrating that it exists in vivo.

2) Of note, a very recent Matsumura et al., 2020, paper generates a different line of Pogz^+/- mice, also reporting that these mice spend more time in the central zone of an open field, alongside cortical thinning and hyperactivation of anterior cingulate cortex during social behavior. Importantly, Matsumara et al. rescue abnormalities using the AMPA receptor negative allosteric modulator perampanel. It would be of great interest to see if this same approach could alleviate the deficits in Cunniff et al.'s model. Reference to the Matsumura et al,. 2020, and relevant comparisons with the present work will have to be added in Introduction and Discussion.

We have added a sentence in the Introduction describing key similarities between that study and ours, and a paragraph to the Discussion more thoroughly describing the Matsumara study.

3) The authors should delete or tone down the "Clinical and therapeutical implications" section. It is unfeasible at this point to begin to plan to implement a therapy for autism based on these findings. There are no known methods that can selectively control hippocampal inputs to fast spiking cortical neurons in human patients. The authors summarize this state of affairs as "it is not immediately obvious how one would translate our findings into new treatments", making this whole section highly speculative.

We have removed the “Clinical and therapeutic implications” section from the Discussion.

4) Related to the above points, face validity may be an imperfect metric of translational utility, but that doesn't mean that the disruption of any behavior involving the prefrontal cortex has relevance for autism/intellectual disability. An alternative explanation is that the gene functions differently in mice and people. This is particularly the case since tests for the expected deficits (social and intellectual functioning) appeared intact. However, it may be premature to make this claim: these latter tests are underpowered (1/3 to 1/2 of the N used for EPM), and a recent paper (Matsumura et al., 2020) found social and cognitive deficits in Pogz missense mutation mice. To be translationally relevant, the authors would need to show that the long-range synchrony deficits underlie social or cognitive behavioral deficits. The Pogz mice in this study show only a very selective behavioral deficit in lower anxiety, which is not a typical autism symptom. These data raise questions on the validity of using Pogz mice for modeling autism phenotypes. This issue must be discussed clearly in the Discussion.

We agree with the basic premise raised by the reviewers: simply finding a behavior that is abnormal when an autism gene is disrupted in mice does not mean that the underlying mechanisms will be relevant to autism. We also agree that face validity is not completely useless. However, there are two issues with face validity. First, social deficits in autism are complex and heterogenous, whereas mouse assays measure only the most rudimentary aspects of social behavior – social preference and/or preference for social novelty. In many individuals with autism, social preference and preference for social novelty are intact, but social functioning is disrupted in other ways. In particular, the largest study of individuals with disruptions in POGZ (25 individuals) found “in many cases, a seemingly contrary overly social and overly friendly demeanor” (Stessman et al., 2016). Furthermore, disruptions in genes of large effect size (like POGZ) elicit heterogeneous behavioral phenotypes. Thus, while lower anxiety may not be a general feature of autism, it may be relevant to individuals with disruptions in POGZ, because they have unusual phenotypes, including aggression, self-injury and irritability (White et al., 2016).

For the reasons outlined above, existing assays for mouse social behavior do not seem to possess even face validity for the unusual phenotypes found in patients with disruptions in POGZ. In this context, how should we study mechanisms that are likely to be relevant to the loss of POGZ? A logical approach is focusing on brain regions and networks that are broadly implicated in autism. While specific behaviors, e.g., social approach, may not be well conserved across species, there is a widespread belief that understanding general principles underlying the function of specific limbic circuits, e.g., interactions between the hippocampus and prefrontal cortex, will yield insights that translate across species. We are not asserting that what we found is definitively relevant to clinical autism. Rather we are arguing that studies of autism pathophysiology using mouse models should be driven by circuits, not behaviors. In this context, it would be premature to assert that social deficits are more likely to be relevant to autism than the phenotype we found – an inability of the PFC to use input to appropriately guide approach/avoidance decisions. Indeed, one hypothesis is that deficits in the ability of the PFC to appropriate guide approach/avoidance decisions is at the core of autism (Pfaff and Barbas, 2019). We have added a discussion of these issues to the Discussion under “Possible relevance of Pogz behavioral phenotypes to autism” (this replaces the “Clinical and therapeutic implications” section which has been removed).

Re: the comment that our studies of social behavior may have been underpowered. The design for our behavioral studies was to perform an initial screen using a large number of assays. This initial screen showed altered behavior in the EPM, but not in other behavioral assays, e.g., for social interaction. Then we validated the EPM deficit using additional mice. This is why the N is larger for the EPM than for other social and cognitive assays. We have added a note about this design to the Materials and methods to make this clear.

5) Materials and methods state that LFP electrode placements were checked histologically. Were any mice excluded on these grounds? It becomes important to show these data in light of a related recent paper by Matsumura et al., 2020, showing smaller brains in a different line of Pogz^+/- mice. It is therefore feasible that electrode placement was systematically different between WT and+/- mice, contributing to WPLI and other neurophysiological differences.

We verified electrode placement both histologically (by visually examining the anatomical location of the electrode track) and electrophysiologically (by confirming the presence of a prominent theta frequency peak in the LFP power spectrum). Importantly, as noted before, the fraction of experiments excluded due to the absence of a clear theta-frequency peak, was not different between WT and mutant mice (5 mice of each genotype were excluded) suggesting there was not systemic mistargeting in Pogz mutant mice as a result of anatomical differences. We have added text to the Discussion to acknowledge this possibility and discuss the reasons why we do not believe it was the case.

6) How is it possible to have hundreds of biological replicates of the weighted phase locking measure with only ~20 mice?

Table 3 specifies the Ns for this measurement – there were 274 closed arm-center runs from 6 WT mice, and 316 closed arm-center runs from 7 mutant mice. Each datapoint was converted to a z-score using the other points from the same run. This means the z-scored value on each run was statistically independent of z-scored values on other runs from the same mouse. We apologize that this was not clear in the original manuscript and have added text to the Materials and methods to make this clear.

7) Given that WPLI was reduced in both homecage and in EPM, presumably it is also reduced on the T-maze, yet Pogz^+/- delayed alternation was apparently normal. Why is this, in light of all the work from the Gordon lab? Why would the phenotype preferentially manifest on the EPM?

We agree that the physiological deficits we found are likely to impact other behaviors that also depend on hippocampal-prefrontal communication. So why didn’t we observe deficits in delayed alternation? There are three possible reasons. First, delayed alternation is much simpler than true T-maze based working memory tasks because the delay is much shorter (only 4 sec). Previous studies have shown that performance at such short delays does not require the same circuitry as longer delays (Bolkan et al., 2017). Second, the Gordon lab has previously shown how compensation can enable mutant mice to perform normally during spatially working memory tasks (Tamura et al., 2017). Specifically, other forms of synchronization may enable Pogz mutants to compensate for reduced theta-frequency phase synchronization. Third, we found that feedforward inhibition from the vHPC to mPFC is deficient. However, feedforward inhibition is mediated by multiple prefrontal cell-types including somatostatin (SST) and parvalbumin (PV) interneurons. The Gordon lab has recently shown that in mPFC, SST but not PV interneurons are required for spatial working memory (Abbas et al., 2018). Thus, it is possible that the deficits in feedforward inhibition we found spare some aspects of inhibition, and thus do not deleteriously impact spatial working memory. We have added a short paragraph discussing these possibilities in the Discussion.

8) Figure 3: Why is this particular IC highlighted? (i) How is the appropriate threshold for significant ICs calculated? (ii) How do the significant ICs in 3C relate to the points in the clusters in 3D? (iii) What is this justification for focus on the cluster that happens to correspond to theta-related to cross-frequency (phase-amplitude) coupling? And (iv) how does this relate to theta WPLI?

i) We compared the eigenvalues obtained using principal components analysis (PCA) to the Marchenko-Pastur distribution. The Marchenko-Pastur distribution yields the magnitude of eigenvalues expected by chance. Thus, the number of eigenvalues which exceed this threshold can be used to compute the number of significant dimensions. We then ran ICA to find this number of ICs.

ii) For the original figure all the significant ICs in panel 3C which are correlated with other ICs are shown in panel 3D (ICs which were not correlated with other ICs were omitted/not shown). Note: based on reviewer feedback, this figure has now been significantly altered (see comment 13).

iii-iv) We focused on this particular IC for two reasons. First, it measures the synchronization of theta-frequency hippocampal activity with activity in the prefrontal cortex. Thus, it measures theta-frequency communication/interaction between these two structures. Second, this IC exhibits modulation as mice approach the decision points in the EPM (namely the center zone). Thus, this IC represents a data-driven metric (i.e., a quantity discovered by unsupervised methods) that shows how theta-frequency communication between the vHPC and mPFC (phase-amplitude coupling between mPFC gamma and vHPC theta) correlates with approach-avoidance decisions (timepoints when the mouse enters the center zone). This makes it an almost perfect analog of theta-frequency WPLI, however, it was discovered in a data-driven manner rather than chosen a priori. Finding that this metric, like theta-frequency WPLI, is altered in Pogz^+/- mice during closed arm-center zone transitions thus provides strong confirmation that theta-frequency hippocampal-prefrontal communication related to approach-avoidance decisions is disrupted in Pogz^+/- mice.

9) The authors argue based on their in vitro data and modeling that decreased vHPC excitatory input to FSIN are responsible for the Pogz deficits in theta synchrony. However, this model would predict mPFC spike phase-locking with vHPC theta (which they don't show) not theta phase synchrony or cross-frequency coupling. They should either show decreased spike phase locking or a model that accounts for their synchrony observations. It would be powerful to use the computational model to explore different frequency ranges of HPC input: does the model predict specific effects on theta-frequency inputs?

First, as requested, we have simulated the responses of the model for different frequencies of HPC input. As now shown in Figure 6—figure supplement 2, our finding that feedforward inhibition enhances the signal-to-noise ration is frequency-specific. In particular, SNR is enhanced for delta and theta frequency inputs, but not for higher or lower frequency inputs.

We agree that decreased vHPC excitatory input to FSINs should decrease spike phase-locking with vHPC theta (at least for FSINs) as stated by the reviewers. However, we disagree with the next part of this statement. Specifically, LFP synchrony (either theta phase synchrony or cross frequency coupling) between the vHPC and mPFC reflects synchronization, not of spiking, but rather of local fields which are driven largely by synaptic currents (Buzsáki, Anastassiou and Koch, 2012; Haider et al., 2016). This is because synaptic currents are more extensive in both space (e.g., extending over larger membrane surface areas) and time compared to spikes. In the awake cortex, local field potentials seem to be particularly dominated by inhibitory activity (Teleńczuk et al., 2017). Reduced feedforward drive from the vHPC onto mPFC interneurons should reduce the component of mPFC inhibitory synaptic activity that is driven by (and synchronized with) vHPC. In this way, the deficit in vHPC excitation of mPFC interneurons that we found is consistent with the reduction in vHPC-mPFC LFP synchrony we observed. We apologize that this logic was not more explicit in the original manuscript and have added text to expand on this point in the Discussion of the revised manuscript. We do agree that looking at the synchronization between specific prefrontal cell types and the hippocampal theta rhythm is an important future direction and now mention this in the Discussion.

10) Does homecage WPLI predict EPM behavior in individual animals?

At the level of the entire population (WT + Het), there is a non-significant trend towards a predictive relationship, but this is driven entirely by the difference between the WT and Het groups. We have included Author response image 1 showing this.

11) This model is proposed as an alternative to the suggestion that vHPC provides direct excitatory input to pyramidal neurons, but it doesn't address Padilla-Coreano et al., 2016 findings that blocking excitatory input did not affect overall firing rates of either pyramidal or interneurons but decreased firing rates in the preferred arm. By their model, shouldn't firing rates increase in the non-preferred arm?

Our model is that vHPC-mPFC input recruits feedforward inhibition, and this feedforward inhibition suppresses the firing of mPFC neurons, specifically out-of-phase firing that is driven by “noise” input. vHPC-mPFC input also excites pyramidal neurons. Thus, firing rates of excitatory neurons reflect a combination of feedforward excitation and feedforward vHPC-mPFC inhibition, as well as excitation and inhibition from other sources. We hypothesize that when vHPC input is suppressed, total circuit E/I remains relatively constant but that inputs from other sources now become the primary drivers of inhibition. As a result, overall firing rates may not change. However, if the differential firing in the preferred arm relative to the non-preferred arm is driven mainly by vHPC input, then this difference should be suppressed, which is what was observed. In other words, our model emphasizes that feedforward inhibition from the vHPC to mPFC plays an important role in enabling the mPFC to respond appropriately to feedforward excitation in this pathway. However, our model still presumes a key role for vHPC-mPFC feedforward excitation as well.

12) Figure 2: Was the directionality of the HPC-PFC WPLI measure consistent with an effect on HPC drive to PFC?

WPLI is unsigned, i.e., it measures phase locking using the magnitudes of the imaginary component of the phase difference. However, when we examined the signs of these phase differences, we found that when mice were in the open arms, for 5/6 wild-type mice and 6/6 Pogz heterozygous mice, the imaginary component was above the x-axis in the complex plane, suggesting that hippocampal activity leads prefrontal activity. We have now included this in the revised text, subsection “Pogz^+/- mice have reduced hippocampal-prefrontal theta synchrony”.

13) Figure 3 needs to be expanded and clarified:

a) How was cross-frequency coupling calculated? The Materials and methods state that "Cross-frequency coupling was calculated by comparing the instantaneous phase in a high frequency band with the instantaneous amplitude in a low frequency band". Usually the converse would be true, but still more details are required – was an established method like the Modulation Index used?

The reviewers are correct – there was a typo in terms of the phase/amplitude for high/low frequency activity. We did compute cross-frequency coupling via a standard method and have corrected and expanded our description of this as follows: “Cross-frequency coupling was calculated by comparing the instantaneous phase in a low frequency band with the instantaneous amplitude in a high frequency band. Specifically, instantaneous phase and amplitude were obtained using the Hilbert transform (using the Matlab function hilbert). At each point in time, this phase and amplitude were combined to yield a vector in the complex plane. We combined vectors from successive timepoints, and the amplitude of the vector sum was normalized to the sum of all the amplitudes to quantify the strength of cross-frequency coupling.”

b) Although it is possible to understand the results in Figure 3, it is not very reader-friendly. The authors should add some panels explaining each step in this analysis, showing visually how they go from raw LFP recordings to the plots shown in Figure 3.

We have substantially reworked Figure 3 and tried to indicate the workflow with arrows and more thorough descriptions. We believe the new version is more reader-friendly.

c) There is insufficient information given about the empirical salient feature analysis. Figures 3A-D are not labeled in a way that readers can decipher the identities of features. The authors should show how all the features fared throughout the pipeline. Moreover, why doesn't this data analysis strategy pull out the known finding (theta theta phase synchrony)?

First, we have tried to make the features more clear in the revised version of Figure 3. All of the features are listed in Tables 1 and 2. If there is a specific piece of information that the reviewers would like us to provide or plot, we would be happy to do so.

Second, the reviewer raises a good point re: theta phase synchrony. Neither theta phase synchrony, nor theta amplitude covariation is “pulled out” by the ICA. This reflects the fact that over the entirety of the task, theta phase synchrony is being influenced by different factors than this IC, even though theta phase synchrony and this IC both evolve in parallel specifically during closed arm-center zone approaches. In other words, during closed-center runs, both theta phase synchrony and the IC both exhibit a sharp rise followed by a return to baseline. However, during the rest of the task, these two measures must diverge. I.e., closed arm-center approaches recruit both theta phase synchrony and the IC, but there must be other EPM behaviors which differentially recruit these two measures. As we identify more behaviors within the EPM (e.g., behavioral motifs identified via MoSeq, DeepLabCut, etc.), presumably we will identify behaviors for which theta synchrony and this IC diverge. This is an important and ongoing area of work in the lab. We have added text related to all of these issues to the Discussion.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Essential revisions:

1) Even though it occurred in equal numbers across genotype, a concern was raised that in a paper focused on theta, 38% (8/21) of the mice were excluded for lacking a clear theta peak (despite having good histology). Is there any explanation for why theta is aberrant in such a large fraction of the data? What would happen to the findings if these mice were included? The authors should include any mice that don't have a technical problem with the electrodes or location. An alternative approach to point 1 would be to analyze the 4+4 “non theta” mice separately in a supplement.

We apologize – we obviously were not clear about the details and this led to understandable confusion on the part of the reviewers. We did not exclude mice which lacked a clear theta peak despite their having good histology. Rather we excluded mice once they lacked a clear theta peak and did not include these mice subsequently for histology. The ventral hippocampus is a deep structure, so mistargeting is not uncommon. In tetrode recording experiments electrodes are typically lowered progressively until electrophysiological markers of the pyramidal cell layer are observed. In the case of fixed electrodes this was not an option, so we simply excluded mice which did not have a clear theta peak during periods of locomotion from the subsequent workflow, which included histology.

2) In general, it seems that people tend to treat units recorded from the same animal as independent but consider LFP data recorded with the same electrode as a single biological replicate (and trials as experimental replicates). The authors should use a repeated measures design, use the average LFP per mouse, or some other statistically valid approach for their analyses.

We agree that appropriate statistical comparisons are critical. That being said, we think statistical validity should be determined by mathematical considerations rather than what researchers in the field tend to do. In this context, we are not clear on the specific reason why the reviewers think this comparison was problematic. Specifically, the reviewers suggested that we use a repeated measures design. A repeated measures design accounts for statistical dependencies between groups (caused by one set of variables) in order to compare the effect of a different variable. In this case, measurements from a single mouse are correlated. Therefore, we computed the z-score of each measurement relative to other measurements from the same run (in the same mouse). This removes any statistical dependencies related to the fact that multiple observations come from the same mouse. This seems to be what the reviewers are asking for, and we don’t understand why this would not be statistically valid. We agree that simply taking all the LFP measurements from one group of mice and comparing them to another group of mice would not be valid, but that is not what we did.

We understand what the reviewers are saying – that most LFP studies do not use multiple LFP measurements as separate samples. This is because in those LFP studies there is no way to remove the statistical dependency between different LFP measurements from the same mouse without subtracting off the mean value for that mouse and thus eliminating any between group differences.

Put another way, in most LFP studies, the source of statistical dependencies within each group (mouse by mouse variation) cannot be separated from a variable driving statistical differences between different groups of mice. However, we are not directly comparing an LFP-derived measurement between groups. Rather we are comparing the modulation of this LFP-derived measurement during each run between groups. As such the relevant unit of analysis is a run (not a mouse). Within-run variation is distinct from within-mouse statistical dependencies. Thus, in contrast to most other LFP studies, we are able to remove within-mouse statistical dependencies without affecting between-group differences in our measure of interest.

3) The explanation of Figure 3 is improved, but questions remain about the procedure. It seems that ICs generated from a combination of WT and POGZ mice (should the number of mice be 13?).

Correct – ICs were identified in both WT and Het mice, and the reviewer is correct that the total number of mice should be 13 (we apologize for the typo).

How was consistency across mice measured? By pairwise correlations?

As stated in the text: “To identify similar ICs that were conserved across mice and thus likely to be biologically meaningful, we calculated the correlation coefficient between all pairs of ICs (Figure 3B), then applied a threshold to this pairwise correlation matrix to identify pairs of highly similar ICs (Figure 3C). We then performed clustering on this dataset (Materials and methods) to identify characteristic ICs that appear repeatedly across mice (Figure 3D).”

How did this procedure measure consistency across all mice?

Correlation measures the similarity (i.e., the consistency) between ICs in two different mice. Clusters correspond to groups of ICs which all have high correlation (correlation above a threshold) and thus are all similar (consistent).

Was there any consideration given to whether ICs were consistent within or across genotype?

No. One could have done a different analysis and identified ICs that were consistent within a genotype. However, we did not do this because of the following concern. Suppose there was no meaningful biological difference between two groups of mice. If you identified ICs in one group (e.g., WT mice), then compared the activity of these ICs between the two groups, you might be more prone to find “false positive” differences than if you had derived the ICs using data from both groups. This is because the ICs would be “overfit” to variation within one group. Thus, variation in that group would tend to occur along each IC axis. However, variation in the other group would be less well aligned with each IC axis. In principle, this would lead to behaviorally-driven changes in IC activity being more pronounced in the group used to identify the IC than in the other group, even if these were not really two different groups. Deriving the ICs using mice from both groups should mitigate this potential issue.

4) What percent of high correlation ICs contained a measure of theta synchrony? Similarly, what percent of ICs were task-modulated? Without these measures, it is hard to interpret the significance of a given IC's task modulation and genotype disruption.

We agree that this information is important and should be included. There were three clusters of ICs that were strongly correlated across different mice. The first one, which we originally described, corresponds to cross-frequency phase-amplitude coupling between hippocampal theta and beta or gamma activity in the hippocampus or PFC. Activity of this IC exhibits a marked rise during center approaches which is absent/deficient in Pogz mutant mice. The second one corresponds to cross-frequency phase-amplitude coupling between prefrontal theta and beta or gamma activity in the hippocampus or PFC. Activity of this characteristic IC was not clearly modulated during center approaches and was not significantly altered in Pogz mutant mice. The third cluster corresponds to broadband power across all frequency bands in the hippocampus and PFC. Similar to IC #1, activity of this IC normally increased during center approaches, but this increase was absent/deficient in Pogz mutant mice.

These observations are now presented in Figure 3—figure supplement 1. Taken together, they support our central finding that theta-frequency communication between the hippocampus and downstream structures such as the PFC is behaviorally modulated, and that the normal pattern of modulation is disrupted in Pogz mutant mice. Characteristic IC networks are not observed for cross-frequency coupling in other frequency bands (e.g., phase amplitude coupling between alpha-frequency activity and beta or gamma oscillations), and behavioral modulation is observed in the HPC->PFC direction but not in the opposite direction. This shows that our findings are specific for both frequency-band and anatomical pathway. We have added text describing this in the revised manuscript.

5) The explanation of why ICA did not yield theta phase synchrony was unclear. Isn't theta synchrony a stable “network” conserved across mice? Doesn't that serve as an important positive control of this empiric method?

We appreciate that this is complicated, but hopefully this explanation is more clear. An IC is a group of measures that co-vary. That means they co-vary in time within mice. Furthermore, we found that the same set of measures compose an IC across mice. Thus, consistently across mice these features co-vary in time.

Now, suppose that during center approaches/entries, the hippocampal theta rhythm exerts a particularly strong influence on the rest of the circuit. As a result, multiple measures related to this rhythm will evolve in concert during center approaches/entries. I.e., vHPC-mPFC phase synchrony and coupling between the phase of vHPC theta and the amplitude of higher frequency exhibit a similar pattern of increases and decreases, as we observed. However, most of the time, the mouse is doing things other than approaching the center. And during these other periods of time, these measures can diverge, presumably because different factors are driving their respective evolution (e.g., perhaps the amplitude of gamma activity is largely determined by something other than theta phase during these periods). In this manner, vHPC-mPFC theta phase synchrony need not be a part of the same IC as other theta-related measures.

Here is an analogy. Senators Bernie Sanders (pre-2016) and Rand Paul may have occasionally voted in concert because of their shared isolationist and anti-gun control views. If you looked only at bills related to gun control and foreign military intervention, their pattern of voting would appear to be strongly correlated. However, when viewed in totality, their voting patterns would not be well aligned. So too, theta phase synchrony can be aligned with the rest of the IC during center approaches/entries without actually being part of the IC (because IC membership is mainly determined by other periods, during which the mouse is not approaching the center).

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Essential revisions:

1) Reviewers remain concerned about the authors' decision to ignore the recommendation to use a repeated measures analysis for some of the reported results, in which different trials from the same mouse were treated as independent. Based on the statistics currently used, it is not convincing to draw the conclusion that theta band WLPI is task-modulated in WT, but not PogZ^+/- mice. If there happened to be a dependency of the trials due to the mouse, then z-scoring wouldn't remove it. In other words, there could be some unknown reason why impairments are strong in some animals and not others (e.g., some interaction between the genotype and individual experiences during development), and z-scoring would not take care of that. Reviewers agree that a repeated measures analysis should be used. Also, for the effects in some frequency bands and not others, again, a repeated measures analysis would be best, as power in different frequency bands within the same recordings is also not independent. It is important to remember that, at eLife, the decision letter containing the consensus review and recommendations is published along with the article. So, we think that readers would wonder why authors did not follow the suggestion to apply a repeated measures analysis, and this could lower confidence in the results and diminish the paper's impact.

We apologize and were not trying to skirt this issue. The statistical test we had done captured some elements of this statistical design (basically comparing one timepoint to the average of all the other timepoints and looking for a genotype effect on this difference), but now we have a better understanding of what exactly the reviewers are looking for and agree that a more standard test would be beneficial. We have used a linear mixed effects model (specifically the fitlme function in Matlab) as specifically suggested in the helpful correspondence. We now include the results of this linear mixed effects model statistical analysis (using mouse, genotype, timepoint, and genotype X timepoint as fixed factors and run as a random factor) in the legend for Figure 2B and Table 3. Specifically, the p-value for the genotype X timepoint interaction was 0.00039.

We have done a similar analysis for the IC projections and report these results in the legends for Figures 3E (IC #1) and 3-1F (IC #3). For IC #1 the p-value for the interaction term is 0.01. In the case of IC #3, the p-value for the interaction term is 0.052, which is just above the threshold for significance – we do not make any claims about whether or not IC #3 differs across genotypes and have simply provided this information as previously requested by the reviewers, to provide context for our finding re: IC #1.

Re: the WPLI when mice are the home cage – in this case, there is a significant difference between home cage theta band WPLI for WT vs. Het mice. However, when we did an ANOVA, we did not find a significant genotype X frequency band difference. We could remove this data. However, we specifically focused on theta-band synchrony because this has previously been implicated in avoidance behavior in the EPM in multiple studies from multiple labs. Given that we found that theta band WPLI is deficient in Het mice during approaches to the center of the EPM, and that theta band WPLI is also deficient in Het mice in the EPM overall (i.e., irrespective of whether mice are in the open or closed arms), it would be natural for readers to be interested in whether there is also a deficit in theta band WPLI in the home cage. Therefore, we have re-arranged this figure (Figure 2) and now simply present the home cage theta band WPLI, after presenting the data about theta band WPLI in the EPM.

2) The authors should also include the exact statistics and degrees of freedom, APA style, not just p-values when reporting effects. Also, reviewers think it would be best to request that all the figure legends list N, name of statistical test, and p value, so that it is easier for readers to understand how data were analyzed. Currently some figures (such as Figure 1) list only p values.

We have added information on test name, exact statistic, Ns, and degrees of freedom to the figure legends (for Wilcoxon rank sum tests, we simply report the N’s as there is no specific degrees of freedom apart from the N).

3) The final N for Figure 3E and Figure 3—figure supplement 1D-F remains unclear. Are these N also based on z-scored trial data? If so, they are subject to the same statistical concern mentioned above.

These N’s are based on runs – we have made this more clear in the revised figure legends and Table 3, and have performed a linear mixed effects model analysis, as described above in our response to point 1.

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Citations

Sohal VS. 2020. Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene PogZ. Dryad Digital Repository. [DOI] [PMC free article] [PubMed]

Supplementary Materials

Figure 1—source data 1. Source data for Figure 1.

elife-54835-fig1-data1.csv^{(11.5KB, csv)}

Figure 2—source data 1. Source data for Figure 2.

elife-54835-fig2-data1.csv^{(25.2KB, csv)}

Figure 4—source data 1. Source data for Figure 4.

elife-54835-fig4-data1.csv^{(4.2KB, csv)}

Figure 5—source data 1. Source data for Figure 5.

elife-54835-fig5-data1.csv^{(7.3KB, csv)}

Transparent reporting form

elife-54835-transrepform.docx^{(246.4KB, docx)}

Data Availability Statement

Table 3. Details of all statistical tests N indicates biological replicates for example individual cells or behavior trials.

Figure	Data	Test	P val	WT Animals	Het Animals	WT n	Het n
Figure 1C	Zone occupancy	Wilcoxon rank sum	0.003	18	27
Figure 1D	EPM Distance	Wilcoxon rank sum	0.35	16	23
Figure 1E	Open time	Wilcoxon rank sum	0.001	18	27
Figure 1E	Center time	Wilcoxon rank sum	0.02	18	27
Figure 1F	Head dips	Wilcoxon rank sum	0.03	14	14
Figure 1G	Open entries	Wilcoxon rank sum	0.32	16	23
Figure 1H	Open visit	Wilcoxon rank sum	0.047	16	23
Figure 2B	WPLI, t = 0	Wilcoxon rank sum	0.0007	6	7	274	316
Figure 2B	WPLI, t = 1.5	Wilcoxon rank sum	0.043	6	7	274	316
Figure 2B	WPLI, t = −3,–1.5, 0, +1.5 during closed-center runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.0026 0.47 0.059 0.0004	6	7	274	316
Figure 2C	Avg zone WPLI, genotype	Two-way ANOVA	0.03	6	7
Figure 2C	Avg zone WPLI, zone	Two-way ANOVA	0.063	6	7
Figure 2C	Avg zone WPLI, interaction	Two-way ANOVA	0.98	6	7
Figure 2D	Theta WPLI	Wilcoxon rank sum	0.031	6	7
Figure 3E	IC zone projection, t = 0	Wilcoxon rank sum	0.007	6	7	39	37
Figure 3E	ICA zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.085 0.16 0.0044 0.010	6	7	39	37
Figure 4E	FSIN charge	Wilcoxon rank sum	0.006	6	3	11	7
Figure 4F	FSIN PPR	Wilcoxon rank sum	0.03	6	3	11	7
Figure 4G	FSIN latency	Wilcoxon rank sum	0.013	6	3	11	7
Figure 4H	FSIN # spikes	Wilcoxon rank sum	0.08	6	3	11	7
Figure 5E	Pyr charge	Wilcoxon rank sum	0.28	13	8	17	11
Figure 5F	Pyr PPR	Wilcoxon rank sum	0.15	13	8	17	11
Figure 5G	Pyr latency	Wilcoxon rank sum	0.76	13	8	17	11
Figure 5H	Pyr # spikes	Wilcoxon rank sum	0.78	13	8	17	11
Figure 1—figure supplement 1	Sex-corrected zone occupancy	Wilcoxon rank sum	0.013	18	27
Figure 1—figure supplement 1A	Zone occupancy for Het M vs. F	Wilcoxon rank sum	0.60		M: 10, F: 17
Figure 1—figure supplement 1B	Sex-corrected EPM distance	Wilcoxon rank sum	0.79	16	23
Figure 1—figure supplement 1C	Head dips: genotype	2-way ANOVA	0.02	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips: sex	2-way ANOVA	0.81	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips: genotype X sex	2-way ANOVA	0.36	M: 10, F: 4	M: 8, F: 6
Figure 1—figure supplement 1C	Head dips for Het M vs. F	Wilcoxon rank sum	0.54		M: 8, F: 6
Figure 1—figure supplement 1D	Open arm entries: genotype	2-way ANOVA	0.22	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1D	Open arm entries: sex	2-way ANOVA	0.61	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1D	Open entries: genotype X sex	2-way ANOVA	0.32	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: genotype	2-way ANOVA	0.22	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: sex	2-way ANOVA	0.49	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1E	Open visit length: genotype X sex	2-way ANOVA	0.76	M: 12, F: 4	M: 9, F: 14
Figure 1—figure supplement 1F	Sex-corrected open arm time	Wilcoxon rank sum	0.013	16	23
Figure 1—figure supplement 1F	Open arm time: Het M vs. F	Wilcoxon rank sum	0.61		M: 10, F: 17
Figure 1—figure supplement 1G	Center time: genotype	2-way ANOVA	0.087	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 1G	Center time: sex	2-way ANOVA	0.29	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 1G	Center time: genotype X sex	2-way ANOVA	0.48	M: 12, F: 6	M: 10, F: 17
Figure 1—figure supplement 2A	Social interaction	Wilcoxon rank sum	0.34	7	7
Figure 1—figure supplement 2B	Novel objection	Wilcoxon rank sum	0.95	7	7
Figure 1—figure supplement 2C	Marble burying	Wilcoxon rank sum	0.45	8	7
Figure 1—figure supplement 2D	OF distance	Wilcoxon rank sum	0.15	14	17
Figure 1—figure supplement 2F	T-maze trials	Wilcoxon rank sum	0.6	5	5
Figure 1—figure supplement 2H	Rule shift IA	Wilcoxon rank sum	0.89	4	4
Figure 1—figure supplement 2H	Rule shift RS	Wilcoxon rank sum	0.89	4	4
Figure 2—figure supplement 1A	PFC theta power	Wilcoxon rank sum	0.91	6	7
Figure 2—figure supplement 1A	PFC beta power	Wilcoxon rank sum	0.94	6	7
Figure 2—figure supplement 1A	PFC low gamma power	Wilcoxon rank sum	0.47	6	7
Figure 2—figure supplement 1A	PFC high gamma power	Wilcoxon rank sum	0.8	6	7
Figure 2—figure supplement 1B	HPC Theta power	Wilcoxon rank sum	0.23	6	7
Figure 2—figure supplement 1B	HPC Beta power	Wilcoxon rank sum	0.093	6	7
Figure 2—figure supplement 1B	HPC low gamma power	Wilcoxon rank sum	0.17	6	7
Figure 2—figure supplement 1B	HPC high gamma power	Wilcoxon rank sum	0.94	6	7
Figure 2—figure supplement 1C	PFC closed theta power	Wilcoxon rank sum	0.88	6	7
Figure 2—figure supplement 1C	PFC closed beta power	Wilcoxon rank sum	1	6	7
Figure 2—figure supplement 1C	PFC closed LG power	Wilcoxon rank sum	0.29	6	7
Figure 2—figure supplement 1C	PFC closed HG power	Wilcoxon rank sum	0.1	6	7
Figure 2—figure supplement 1D	HPC closed theta power	Wilcoxon rank sum	0.25	6	7
Figure 2—figure supplement 1D	HPC closed beta power	Wilcoxon rank sum	0.15	6	7
Figure 2—figure supplement 1D	HPC closed LG power	Wilcoxon rank sum	0.48	6	7
Figure 2—figure supplement 1D	HPC closed HG power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open theta power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open beta power	Wilcoxon rank sum	0.89	6	7
Figure 2—figure supplement 1E	PFC open LG power	Wilcoxon rank sum	1	6	7
Figure 2—figure supplement 1E	PFC open HG power	Wilcoxon rank sum	0.15	6	7
Figure 2—figure supplement 1F	HPC open theta power	Wilcoxon rank sum	0.32	6	7
Figure 2—figure supplement 1F	HPC open beta power	Wilcoxon rank sum	0.2	6	7
Figure 2—figure supplement 1F	HPC open LG power	Wilcoxon rank sum	0.25	6	7
Figure 2—figure supplement 1F	HPC open HG power	Wilcoxon rank sum	1	6	7
Figure 3—figure supplement 1D	IC #1 zone projection	Wilcoxon rank sum	0.007	6	7	39	37
Figure 3—figure supplement 1D	IC #1 zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.085 0.16 0.0044 0.010	6	7	39	37
Figure 3—figure supplement 1F	IC #3 zone projection	Wilcoxon rank sum	0.015	6	7	39	37
Figure 3—figure supplement 1F	IC #3 zone projection t = 0 vs. baseline (average of first and last timepoints) during closed-center-open runs	Linear mixed effects model timepoint mouse genotype timept X genotype	0.0094 0.50 0.026 0.052	6	7	39	37
Figure 4—figure supplement 1C	FSIN resting potential	Wilcoxon rank sum	0.50	6	3	11	7
Figure 4—figure supplement 1D	FSIN input resistance	Wilcoxon rank sum	0.44	6	3	11	7
Figure 4—figure supplement 1E	FSIN halfwidth	Wilcoxon rank sum	0.47	6	3	11	7
Figure 4—figure supplement 1F	FSIN max firing rate	Wilcoxon rank sum	0.50	6	3	11	7
Figure 5—figure supplement 1C	Pyr resting potential	Wilcoxon rank sum	0.94	13	8	17	11
Figure 5—figure supplement 1D	Pyr input resistance	Wilcoxon rank sum	0.80	13	8	17	11
Figure 5—figure supplement 1E	Pyr halfwidth	Wilcoxon rank sum	0.46	13	8	17	11
Figure 5—figure supplement 1F	Pyr max firing rate	Wilcoxon rank sum	0.93	13	8	17	11

Open in a new tab

The following dataset was generated:

Sohal VS. 2020. Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene PogZ. Dryad Digital Repository.

[bib1] Abbas AI, Sundiang MJM, Henoch B, Morton MP, Bolkan SS, Park AJ, Harris AZ, Kellendonk C, Gordon JA. Somatostatin interneurons facilitate Hippocampal-Prefrontal synchrony and prefrontal spatial encoding. Neuron. 2018;100:926–939. doi: 10.1016/j.neuron.2018.09.029. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib2] Adhikari A, Topiwala MA, Gordon JA. Synchronized activity between the ventral Hippocampus and the medial prefrontal cortex during anxiety. Neuron. 2010;65:257–269. doi: 10.1016/j.neuron.2009.12.002. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib3] Adhikari A, Topiwala MA, Gordon JA. Single units in the medial prefrontal cortex with anxiety-related firing patterns are preferentially influenced by ventral hippocampal activity. Neuron. 2011;71:898–910. doi: 10.1016/j.neuron.2011.07.027. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib4] Andreescu C, Sheu LK, Tudorascu D, Walker S, Aizenstein H. The ages of anxiety--differences across the lifespan in the default mode network functional connectivity in generalized anxiety disorder. International Journal of Geriatric Psychiatry. 2014;29:704–712. doi: 10.1002/gps.4051. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib5] Angoa-Pérez M, Kane MJ, Briggs DI, Francescutti DM, Kuhn DM. Marble burying and nestlet shredding as tests of repetitive, Compulsive-like behaviors in mice. Journal of Visualized Experiments. 2013;82:50978. doi: 10.3791/50978. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib6] Bolkan SS, Stujenske JM, Parnaudeau S, Spellman TJ, Rauffenbart C, Abbas AI, Harris AZ, Gordon JA, Kellendonk C. Thalamic projections sustain prefrontal activity during working memory maintenance. Nature Neuroscience. 2017;20:987–996. doi: 10.1038/nn.4568. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib7] Bourgeron T. From the genetic architecture to synaptic plasticity in autism spectrum disorder. Nature Reviews Neuroscience. 2015;16:551–563. doi: 10.1038/nrn3992. [DOI] [PubMed] [Google Scholar]

[bib8] Buard I, Rogers SJ, Hepburn S, Kronberg E, Rojas DC. Altered oscillation patterns and connectivity during picture naming in autism. Frontiers in Human Neuroscience. 2013;7:742. doi: 10.3389/fnhum.2013.00742. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib9] Buzsáki G, Anastassiou CA, Koch C. The origin of extracellular fields and currents--EEG, ECoG, LFP and spikes. Nature Reviews Neuroscience. 2012;13:407–420. doi: 10.1038/nrn3241. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib10] Calhoon GG, Tye KM. Resolving the neural circuits of anxiety. Nature Neuroscience. 2015;18:1394–1404. doi: 10.1038/nn.4101. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib11] Cho KK, Hoch R, Lee AT, Patel T, Rubenstein JL, Sohal VS. Gamma rhythms link prefrontal interneuron dysfunction with cognitive inflexibility in Dlx5/6+/− mice. Neuron. 2015;85:1332–1343. doi: 10.1016/j.neuron.2015.02.019. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib12] Cho KKA, Davidson TJ, Bouvier G, Marshall JD, Schnitzer MJ, Sohal VS. Cross-hemispheric gamma synchrony between prefrontal parvalbumin interneurons supports behavioral adaptation during rule shift learning. Nature Neuroscience. 2020;23:892–902. doi: 10.1038/s41593-020-0647-1. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib13] Coben R, Mohammad-Rezazadeh I, Cannon RL. Using quantitative and analytic EEG methods in the understanding of connectivity in autism spectrum disorders: a theory of mixed over- and under-connectivity. Frontiers in Human Neuroscience. 2014;8:045. doi: 10.3389/fnhum.2014.00045. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib14] Cotney J, Muhle RA, Sanders SJ, Liu L, Willsey AJ, Niu W, Liu W, Klei L, Lei J, Yin J, Reilly SK, Tebbenkamp AT, Bichsel C, Pletikos M, Sestan N, Roeder K, State MW, Devlin B, Noonan JP. The autism-associated chromatin modifier CHD8 regulates other autism risk genes during human neurodevelopment. Nature Communications. 2015;6:ncomms7404. doi: 10.1038/ncomms7404. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib15] Cunniff MM. PogZ_paper. 189f9c5GitHub. 2020 https://github.com/mcunniff/PogZ_paper

[bib16] De Rubeis S, He X, Goldberg AP, Poultney CS, Samocha K, Cicek AE, Kou Y, Liu L, Fromer M, Walker S, Singh T, Klei L, Kosmicki J, Shih-Chen F, Aleksic B, Biscaldi M, Bolton PF, Brownfeld JM, Cai J, Campbell NG, Carracedo A, Chahrour MH, Chiocchetti AG, Coon H, Crawford EL, Curran SR, Dawson G, Duketis E, Fernandez BA, Gallagher L, Geller E, Guter SJ, Hill RS, Ionita-Laza J, Jimenz Gonzalez P, Kilpinen H, Klauck SM, Kolevzon A, Lee I, Lei I, Lei J, Lehtimäki T, Lin CF, Ma'ayan A, Marshall CR, McInnes AL, Neale B, Owen MJ, Ozaki N, Parellada M, Parr JR, Purcell S, Puura K, Rajagopalan D, Rehnström K, Reichenberg A, Sabo A, Sachse M, Sanders SJ, Schafer C, Schulte-Rüther M, Skuse D, Stevens C, Szatmari P, Tammimies K, Valladares O, Voran A, Li-San W, Weiss LA, Willsey AJ, Yu TW, Yuen RK, Cook EH, Freitag CM, Gill M, Hultman CM, Lehner T, Palotie A, Schellenberg GD, Sklar P, State MW, Sutcliffe JS, Walsh CA, Scherer SW, Zwick ME, Barett JC, Cutler DJ, Roeder K, Devlin B, Daly MJ, Buxbaum JD, DDD Study. Homozygosity Mapping Collaborative for Autism. UK10K Consortium Synaptic, transcriptional and chromatin genes disrupted in autism. Nature. 2014;515:209–215. doi: 10.1038/nature13772. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib17] Dentici ML, Niceta M, Pantaleoni F, Barresi S, Bencivenga P, Dallapiccola B, Digilio MC, Tartaglia M. Expanding the phenotypic spectrum of truncating POGZ mutations: association with CNS malformations, skeletal abnormalities, and distinctive facial dysmorphism. American Journal of Medical Genetics Part A. 2017;173:1965–1969. doi: 10.1002/ajmg.a.38255. [DOI] [PubMed] [Google Scholar]

[bib18] Ellwood IT, Patel T, Wadia V, Lee AT, Liptak AT, Bender KJ, Sohal VS. Tonic or phasic stimulation of dopaminergic projections to prefrontal cortex causes mice to maintain or deviate from previously learned behavioral strategies. The Journal of Neuroscience. 2017;37:8315–8329. doi: 10.1523/JNEUROSCI.1221-17.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib19] Fitzgerald TW, Gerety SS, Jones WD, Van Kogelenberg M, King DA, McRae J, Morley K, Parthiban V, Al-Turki S, Ambridge K, Barrett DM, Bayzetinova T, Clayton S, Coomber EL, Gribble S, Jones P, Krishnappa N, Mason LE, Middleton A, Hurles ME, Deciphering Developmental Disorders Study Large-scale discovery of novel genetic causes of developmental disorders. Nature. 2015;519:223–228. doi: 10.1038/nature14135. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib20] Fromer M, Pocklington AJ, Kavanagh DH, Williams HJ, Dwyer S, Gormley P, Georgieva L, Rees E, Palta P, Ruderfer DM, Carrera N, Humphreys I, Johnson JS, Roussos P, Barker DD, Banks E, Milanova V, Grant SG, Hannon E, Rose SA, Chambert K, Mahajan M, Scolnick EM, Moran JL, Kirov G, Palotie A, McCarroll SA, Holmans P, Sklar P, Owen MJ, Purcell SM, O'Donovan MC. De novo mutations in schizophrenia implicate synaptic networks. Nature. 2014;506:179–184. doi: 10.1038/nature12929. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib21] Fukai R, Hiraki Y, Yofune H, Tsurusaki Y, Nakashima M, Saitsu H, Tanaka F, Miyake N, Matsumoto N. A case of autism spectrum disorder arising from a de novo missense mutation in POGZ. Journal of Human Genetics. 2015;60:277–279. doi: 10.1038/jhg.2015.13. [DOI] [PubMed] [Google Scholar]

[bib22] Gilissen C, Hehir-Kwa JY, Thung DT, van de Vorst M, van Bon BW, Willemsen MH, Kwint M, Janssen IM, Hoischen A, Schenck A, Leach R, Klein R, Tearle R, Bo T, Pfundt R, Yntema HG, de Vries BB, Kleefstra T, Brunner HG, Vissers LE, Veltman JA. Genome sequencing identifies major causes of severe intellectual disability. Nature. 2014;511:344–347. doi: 10.1038/nature13394. [DOI] [PubMed] [Google Scholar]

[bib23] Gulsuner S, Walsh T, Watts AC, Lee MK, Thornton AM, Casadei S, Rippey C, Shahin H, Nimgaonkar VL, Go RC, Savage RM, Swerdlow NR, Gur RE, Braff DL, King MC, McClellan JM, Consortium on the Genetics of Schizophrenia (COGS) PAARTNERS Study Group Spatial and temporal mapping of de novo mutations in schizophrenia to a fetal prefrontal cortical network. Cell. 2013;154:518–529. doi: 10.1016/j.cell.2013.06.049. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib24] Guo W, Liu F, Xiao C, Liu J, Yu M, Zhang Z, Zhang J, Zhao J. Increased short-range and long-range functional connectivity in first-episode, medication-naive schizophrenia at rest. Schizophrenia Research. 2015;166:144–150. doi: 10.1016/j.schres.2015.04.034. [DOI] [PubMed] [Google Scholar]

[bib25] Haider B, Schulz DP, Häusser M, Carandini M. Millisecond coupling of local field potentials to synaptic currents in the awake visual cortex. Neuron. 2016;90:35–42. doi: 10.1016/j.neuron.2016.02.034. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib26] Hashimoto R, Nakazawa T, Tsurusaki Y, Yasuda Y, Nagayasu K, Matsumura K, Kawashima H, Yamamori H, Fujimoto M, Ohi K, Umeda-Yano S, Fukunaga M, Fujino H, Kasai A, Hayata-Takano A, Shintani N, Takeda M, Matsumoto N, Hashimoto H. Whole-exome sequencing and neurite outgrowth analysis in autism spectrum disorder. Journal of Human Genetics. 2016;61:199–206. doi: 10.1038/jhg.2015.141. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib27] Hultman R, Mague SD, Li Q, Katz BM, Michel N, Lin L, Wang J, David LK, Blount C, Chandy R, Carlson D, Ulrich K, Carin L, Dunson D, Kumar S, Deisseroth K, Moore SD, Dzirasa K. Dysregulation of prefrontal Cortex-Mediated Slow-Evolving limbic dynamics drives Stress-Induced emotional pathology. Neuron. 2016;91:439–452. doi: 10.1016/j.neuron.2016.05.038. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib28] Hultman R, Ulrich K, Sachs BD, Blount C, Carlson DE, Ndubuizu N, Bagot RC, Parise EM, Vu MT, Gallagher NM, Wang J, Silva AJ, Deisseroth K, Mague SD, Caron MG, Nestler EJ, Carin L, Dzirasa K. Brain-wide electrical spatiotemporal dynamics encode depression vulnerability. Cell. 2018;173:166–180. doi: 10.1016/j.cell.2018.02.012. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib29] Hyvärinen A, Oja E. Independent component analysis: algorithms and applications. Neural Networks. 2000;13:411–430. doi: 10.1016/S0893-6080(00)00026-5. [DOI] [PubMed] [Google Scholar]

[bib30] Iossifov I, Ronemus M, Levy D, Wang Z, Hakker I, Rosenbaum J, Yamrom B, Lee YH, Narzisi G, Leotta A, Kendall J, Grabowska E, Ma B, Marks S, Rodgers L, Stepansky A, Troge J, Andrews P, Bekritsky M, Pradhan K, Ghiban E, Kramer M, Parla J, Demeter R, Fulton LL, Fulton RS, Magrini VJ, Ye K, Darnell JC, Darnell RB, Mardis ER, Wilson RK, Schatz MC, McCombie WR, Wigler M. De novo gene disruptions in children on the autistic spectrum. Neuron. 2012;74:285–299. doi: 10.1016/j.neuron.2012.04.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib31] Iossifov I, O'Roak BJ, Sanders SJ, Ronemus M, Krumm N, Levy D, Stessman HA, Witherspoon KT, Vives L, Patterson KE, Smith JD, Paeper B, Nickerson DA, Dea J, Dong S, Gonzalez LE, Mandell JD, Mane SM, Murtha MT, Sullivan CA, Walker MF, Waqar Z, Wei L, Willsey AJ, Yamrom B, Lee YH, Grabowska E, Dalkic E, Wang Z, Marks S, Andrews P, Leotta A, Kendall J, Hakker I, Rosenbaum J, Ma B, Rodgers L, Troge J, Narzisi G, Yoon S, Schatz MC, Ye K, McCombie WR, Shendure J, Eichler EE, State MW, Wigler M. The contribution of de novo coding mutations to autism spectrum disorder. Nature. 2014;515:216–221. doi: 10.1038/nature13908. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib32] Jacinto LR, Cerqueira JJ, Sousa N. Patterns of theta activity in limbic anxiety circuit preceding exploratory behavior in Approach-Avoidance conflict. Frontiers in Behavioral Neuroscience. 2016;10:171. doi: 10.3389/fnbeh.2016.00171. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib33] Just MA, Cherkassky VL, Keller TA, Minshew NJ. Cortical activation and synchronization during sentence comprehension in high-functioning autism: evidence of underconnectivity. Brain. 2004;127:1811–1821. doi: 10.1093/brain/awh199. [DOI] [PubMed] [Google Scholar]

[bib34] Just MA, Keller TA, Malave VL, Kana RK, Varma S. Autism as a neural systems disorder: a theory of frontal-posterior underconnectivity. Neuroscience & Biobehavioral Reviews. 2012;36:1292–1313. doi: 10.1016/j.neubiorev.2012.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib35] Kam JW, Bolbecker AR, O'Donnell BF, Hetrick WP, Brenner CA. Resting state EEG power and coherence abnormalities in bipolar disorder and schizophrenia. Journal of Psychiatric Research. 2013;47:1893–1901. doi: 10.1016/j.jpsychires.2013.09.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib36] Kana RK, Uddin LQ, Kenet T, Chugani D, Müller RA. Brain connectivity in autism. Frontiers in Human Neuroscience. 2014;8:349. doi: 10.3389/fnhum.2014.00349. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib37] Keehn B, Wagner JB, Tager-Flusberg H, Nelson CA. Functional connectivity in the first year of life in infants at-risk for autism: a preliminary near-infrared spectroscopy study. Frontiers in Human Neuroscience. 2013;7:444. doi: 10.3389/fnhum.2013.00444. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib38] Kirkby LA, Luongo FJ, Lee MB, Nahum M, Van Vleet TM, Rao VR, Dawes HE, Chang EF, Sohal VS. An Amygdala-Hippocampus subnetwork that encodes variation in human mood. Cell. 2018;175:1688–1700. doi: 10.1016/j.cell.2018.10.005. [DOI] [PubMed] [Google Scholar]

[bib39] Kjaerby C, Athilingam J, Robinson SE, Iafrati J, Sohal VS. Serotonin 1B receptors regulate prefrontal function by gating callosal and hippocampal inputs. Cell Reports. 2016;17:2882–2890. doi: 10.1016/j.celrep.2016.11.036. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib40] Kjelstrup KG, Tuvnes FA, Steffenach HA, Murison R, Moser EI, Moser MB. Reduced fear expression after lesions of the ventral Hippocampus. PNAS. 2002;99:10825–10830. doi: 10.1073/pnas.152112399. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib41] Krumm N, O'Roak BJ, Shendure J, Eichler EE. A de novo convergence of autism genetics and molecular neuroscience. Trends in Neurosciences. 2014;37:95–105. doi: 10.1016/j.tins.2013.11.005. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib42] Lee E, Lee J, Kim E. Excitation/Inhibition imbalance in animal models of autism spectrum disorders. Biological Psychiatry. 2017;81:838–847. doi: 10.1016/j.biopsych.2016.05.011. [DOI] [PubMed] [Google Scholar]

[bib43] Lee AT, Cunniff MM, See JZ, Wilke SA, Luongo FJ, Ellwood IT, Ponnavolu S, Sohal VS. VIP interneurons contribute to avoidance behavior by regulating information flow across Hippocampal-Prefrontal networks. Neuron. 2019;102:1223–1234. doi: 10.1016/j.neuron.2019.04.001. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib44] Lein ES, Hawrylycz MJ, Ao N, Ayres M, Bensinger A, Bernard A, Boe AF, Boguski MS, Brockway KS, Byrnes EJ, Chen L, Chen L, Chen TM, Chin MC, Chong J, Crook BE, Czaplinska A, Dang CN, Datta S, Dee NR, Desaki AL, Desta T, Diep E, Dolbeare TA, Donelan MJ, Dong HW, Dougherty JG, Duncan BJ, Ebbert AJ, Eichele G, Estin LK, Faber C, Facer BA, Fields R, Fischer SR, Fliss TP, Frensley C, Gates SN, Glattfelder KJ, Halverson KR, Hart MR, Hohmann JG, Howell MP, Jeung DP, Johnson RA, Karr PT, Kawal R, Kidney JM, Knapik RH, Kuan CL, Lake JH, Laramee AR, Larsen KD, Lau C, Lemon TA, Liang AJ, Liu Y, Luong LT, Michaels J, Morgan JJ, Morgan RJ, Mortrud MT, Mosqueda NF, Ng LL, Ng R, Orta GJ, Overly CC, Pak TH, Parry SE, Pathak SD, Pearson OC, Puchalski RB, Riley ZL, Rockett HR, Rowland SA, Royall JJ, Ruiz MJ, Sarno NR, Schaffnit K, Shapovalova NV, Sivisay T, Slaughterbeck CR, Smith SC, Smith KA, Smith BI, Sodt AJ, Stewart NN, Stumpf KR, Sunkin SM, Sutram M, Tam A, Teemer CD, Thaller C, Thompson CL, Varnam LR, Visel A, Whitlock RM, Wohnoutka PE, Wolkey CK, Wong VY, Wood M, Yaylaoglu MB, Young RC, Youngstrom BL, Yuan XF, Zhang B, Zwingman TA, Jones AR. Genome-wide atlas of gene expression in the adult mouse brain. Nature. 2007;445:168–176. doi: 10.1038/nature05453. [DOI] [PubMed] [Google Scholar]

[bib45] Lopes-dos-Santos V, Ribeiro S, Tort AB. Detecting cell assemblies in large neuronal populations. Journal of Neuroscience Methods. 2013;220:149–166. doi: 10.1016/j.jneumeth.2013.04.010. [DOI] [PubMed] [Google Scholar]

[bib46] Mathis A, Mamidanna P, Cury KM, Abe T, Murthy VN, Mathis MW, Bethge M. DeepLabCut: markerless pose estimation of user-defined body parts with deep learning. Nature Neuroscience. 2018;21:1281–1289. doi: 10.1038/s41593-018-0209-y. [DOI] [PubMed] [Google Scholar]

[bib47] Matsumura K, Nakazawa T, Nagayasu K, Gotoda-Nishimura N, Kasai A, Hayata-Takano A, Shintani N, Yamamori H, Yasuda Y, Hashimoto R, Hashimoto H. De novo POGZ mutations in sporadic autism disrupt the DNA-binding activity of POGZ. Journal of Molecular Psychiatry. 2016;4:1. doi: 10.1186/s40303-016-0016-x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib48] Matsumura K, Seiriki K, Okada S, Nagase M, Ayabe S, Yamada I, Furuse T, Shibuya H, Yasuda Y, Yamamori H, Fujimoto M, Nagayasu K, Yamamoto K, Kitagawa K, Miura H, Gotoda-Nishimura N, Igarashi H, Hayashida M, Baba M, Kondo M, Hasebe S, Ueshima K, Kasai A, Ago Y, Hayata-Takano A, Shintani N, Iguchi T, Sato M, Yamaguchi S, Tamura M, Wakana S, Yoshiki A, Watabe AM, Okano H, Takuma K, Hashimoto R, Hashimoto H, Nakazawa T. Pathogenic POGZ mutation causes impaired cortical development and reversible autism-like phenotypes. Nature Communications. 2020;11:1–16. doi: 10.1038/s41467-020-14697-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib49] Mueller S, Keeser D, Samson AC, Kirsch V, Blautzik J, Grothe M, Erat O, Hegenloh M, Coates U, Reiser MF, Hennig-Fast K, Meindl T. Convergent findings of altered functional and structural brain connectivity in individuals with high functioning autism: a multimodal MRI study. PLOS ONE. 2013;8:e67329. doi: 10.1371/journal.pone.0067329. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib50] Müller RA, Shih P, Keehn B, Deyoe JR, Leyden KM, Shukla DK. Underconnected, but how? A survey of functional connectivity MRI studies in autism spectrum disorders. Cerebral Cortex. 2011;21:2233–2243. doi: 10.1093/cercor/bhq296. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib51] Mullins C, Fishell G, Tsien RW. Unifying views of autism spectrum disorders: a consideration of autoregulatory feedback loops. Neuron. 2016;89:1131–1156. doi: 10.1016/j.neuron.2016.02.017. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib52] Nair A, Treiber JM, Shukla DK, Shih P, Müller RA. Impaired thalamocortical connectivity in autism spectrum disorder: a study of functional and anatomical connectivity. Brain. 2013;136:1942–1955. doi: 10.1093/brain/awt079. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib53] Neale BM, Kou Y, Liu L, Ma'ayan A, Samocha KE, Sabo A, Lin CF, Stevens C, Wang LS, Makarov V, Polak P, Yoon S, Maguire J, Crawford EL, Campbell NG, Geller ET, Valladares O, Schafer C, Liu H, Zhao T, Cai G, Lihm J, Dannenfelser R, Jabado O, Peralta Z, Nagaswamy U, Muzny D, Reid JG, Newsham I, Wu Y, Lewis L, Han Y, Voight BF, Lim E, Rossin E, Kirby A, Flannick J, Fromer M, Shakir K, Fennell T, Garimella K, Banks E, Poplin R, Gabriel S, DePristo M, Wimbish JR, Boone BE, Levy SE, Betancur C, Sunyaev S, Boerwinkle E, Buxbaum JD, Cook EH, Devlin B, Gibbs RA, Roeder K, Schellenberg GD, Sutcliffe JS, Daly MJ. Patterns and rates of exonic de novo mutations in autism spectrum disorders. Nature. 2012;485:242–245. doi: 10.1038/nature11011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib54] Nelson SB, Valakh V. Excitatory/Inhibitory balance and circuit homeostasis in autism spectrum disorders. Neuron. 2015;87:684–698. doi: 10.1016/j.neuron.2015.07.033. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib55] Nozawa RS, Nagao K, Masuda HT, Iwasaki O, Hirota T, Nozaki N, Kimura H, Obuse C. Human POGZ modulates dissociation of HP1alpha from mitotic chromosome arms through aurora B activation. Nature Cell Biology. 2010;12:719–727. doi: 10.1038/ncb2075. [DOI] [PubMed] [Google Scholar]

[bib56] Padilla-Coreano N, Bolkan SS, Pierce GM, Blackman DR, Hardin WD, Garcia-Garcia AL, Spellman TJ, Gordon JA. Direct ventral Hippocampal-Prefrontal input is required for Anxiety-Related neural activity and behavior. Neuron. 2016;89:857–866. doi: 10.1016/j.neuron.2016.01.011. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib57] Padilla-Coreano N, Canetta S, Mikofsky RM, Alway E, Passecker J, Myroshnychenko MV, Garcia-Garcia AL, Warren R, Teboul E, Blackman DR, Morton MP, Hupalo S, Tye KM, Kellendonk C, Kupferschmidt DA, Gordon JA. Hippocampal-Prefrontal theta transmission regulates avoidance behavior. Neuron. 2019;104:601–610. doi: 10.1016/j.neuron.2019.08.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib58] Padmanabhan A, Lynn A, Foran W, Luna B, O'Hearn K. Age related changes in striatal resting state functional connectivity in autism. Frontiers in Human Neuroscience. 2013;7:814. doi: 10.3389/fnhum.2013.00814. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib59] Parent MA, Wang L, Su J, Netoff T, Yuan LL. Identification of the hippocampal input to medial prefrontal cortex in vitro. Cerebral Cortex. 2010;20:393–403. doi: 10.1093/cercor/bhp108. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib60] Pfaff D, Barbas H. Mechanisms for the approach/Avoidance decision applied to autism. Trends in Neurosciences. 2019;42:448–457. doi: 10.1016/j.tins.2019.05.002. [DOI] [PubMed] [Google Scholar]

[bib61] Potter GB, Petryniak MA, Shevchenko E, McKinsey GL, Ekker M, Rubenstein JL. Generation of Cre-transgenic mice using Dlx1/Dlx2 enhancers and their characterization in GABAergic interneurons. Molecular and Cellular Neuroscience. 2009;40:167–186. doi: 10.1016/j.mcn.2008.10.003. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib62] Redcay E, Moran JM, Mavros PL, Tager-Flusberg H, Gabrieli JD, Whitfield-Gabrieli S. Intrinsic functional network organization in high-functioning adolescents with autism spectrum disorder. Frontiers in Human Neuroscience. 2013;7:573. doi: 10.3389/fnhum.2013.00573. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib63] Ronan JL, Wu W, Crabtree GR. From neural development to cognition: unexpected roles for chromatin. Nature Reviews Genetics. 2013;14:347–359. doi: 10.1038/nrg3413. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib64] Rubenstein JL, Merzenich MM. Model of autism: increased ratio of excitation/inhibition in key neural systems. Genes, Brain and Behavior. 2003;2:255–267. doi: 10.1034/j.1601-183X.2003.00037.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib65] Sánchez-Bellot C, MacAskill A. Push-pull regulation of exploratory behavior by two opposing hippocampal to prefrontal cortex pathways. bioRxiv. 2019 doi: 10.1101/2019.12.18.880831. [DOI]

[bib66] Sanders SJ, He X, Willsey AJ, Ercan-Sencicek AG, Samocha KE, Cicek AE, Murtha MT, Bal VH, Bishop SL, Dong S, Goldberg AP, Jinlu C, Keaney JF, Klei L, Mandell JD, Moreno-De-Luca D, Poultney CS, Robinson EB, Smith L, Solli-Nowlan T, Su MY, Teran NA, Walker MF, Werling DM, Beaudet AL, Cantor RM, Fombonne E, Geschwind DH, Grice DE, Lord C, Lowe JK, Mane SM, Martin DM, Morrow EM, Talkowski ME, Sutcliffe JS, Walsh CA, Yu TW, Ledbetter DH, Martin CL, Cook EH, Buxbaum JD, Daly MJ, Devlin B, Roeder K, State MW, Autism Sequencing Consortium Insights into autism spectrum disorder genomic architecture and biology from 71 risk loci. Neuron. 2015;87:1215–1233. doi: 10.1016/j.neuron.2015.09.016. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib67] Shah AA, Treit D. Excitotoxic lesions of the medial prefrontal cortex attenuate fear responses in the elevated-plus maze, social interaction and shock probe burying tests. Brain Research. 2003;969:183–194. doi: 10.1016/S0006-8993(03)02299-6. [DOI] [PubMed] [Google Scholar]

[bib68] Sigurdsson T, Stark KL, Karayiorgou M, Gogos JA, Gordon JA. Impaired hippocampal-prefrontal synchrony in a genetic mouse model of schizophrenia. Nature. 2010;464:763–767. doi: 10.1038/nature08855. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib69] Sohal VS, Rubenstein JLR. Excitation-inhibition balance as a framework for investigating mechanisms in neuropsychiatric disorders. Molecular Psychiatry. 2019;24:1248–1257. doi: 10.1038/s41380-019-0426-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib70] Spellman T, Rigotti M, Ahmari SE, Fusi S, Gogos JA, Gordon JA. Hippocampal-prefrontal input supports spatial encoding in working memory. Nature. 2015;522:309–314. doi: 10.1038/nature14445. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib71] Stessman HAF, Willemsen MH, Fenckova M, Penn O, Hoischen A, Xiong B, Wang T, Hoekzema K, Vives L, Vogel I, Brunner HG, van der Burgt I, Ockeloen CW, Schuurs-Hoeijmakers JH, Klein Wassink-Ruiter JS, Stumpel C, Stevens SJC, Vles HS, Marcelis CM, van Bokhoven H, Cantagrel V, Colleaux L, Nicouleau M, Lyonnet S, Bernier RA, Gerdts J, Coe BP, Romano C, Alberti A, Grillo L, Scuderi C, Nordenskjöld M, Kvarnung M, Guo H, Xia K, Piton A, Gerard B, Genevieve D, Delobel B, Lehalle D, Perrin L, Prieur F, Thevenon J, Gecz J, Shaw M, Pfundt R, Keren B, Jacquette A, Schenck A, Eichler EE, Kleefstra T. Disruption of POGZ is associated with intellectual disability and autism spectrum disorders. The American Journal of Human Genetics. 2016;98:541–552. doi: 10.1016/j.ajhg.2016.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib72] Suliman R, Ben-David E, Shifman S. Chromatin regulators, phenotypic robustness, and autism risk. Frontiers in Genetics. 2014;5:81. doi: 10.3389/fgene.2014.00081. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib73] Tamura M, Spellman TJ, Rosen AM, Gogos JA, Gordon JA. Hippocampal-prefrontal theta-gamma coupling during performance of a spatial working memory task. Nature Communications. 2017;8:9. doi: 10.1038/s41467-017-02108-9. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib74] Tan B, Zou Y, Zhang Y, Zhang R, Ou J, Shen Y, Zhao J, Luo X, Guo J, Zeng L, Hu Y, Zheng Y, Pan Q, Liang D, Wu L. A novel de novo POGZ mutation in a patient with intellectual disability. Journal of Human Genetics. 2016;61:357–359. doi: 10.1038/jhg.2015.156. [DOI] [PubMed] [Google Scholar]

[bib75] Teleńczuk B, Dehghani N, Le Van Quyen M, Cash SS, Halgren E, Hatsopoulos NG, Destexhe A. Local field potentials primarily reflect inhibitory neuron activity in human and monkey cortex. Scientific Reports. 2017;7:40211. doi: 10.1038/srep40211. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib76] Toro R, Konyukh M, Delorme R, Leblond C, Chaste P, Fauchereau F, Coleman M, Leboyer M, Gillberg C, Bourgeron T. Key role for gene dosage and synaptic homeostasis in autism spectrum disorders. Trends in Genetics. 2010;26:363–372. doi: 10.1016/j.tig.2010.05.007. [DOI] [PubMed] [Google Scholar]

[bib77] Vinck M, Oostenveld R, van Wingerden M, Battaglia F, Pennartz CM. An improved index of phase-synchronization for electrophysiological data in the presence of volume-conduction, noise and sample-size Bias. NeuroImage. 2011;55:1548–1565. doi: 10.1016/j.neuroimage.2011.01.055. [DOI] [PubMed] [Google Scholar]

[bib78] Wang X, Xia M, Lai Y, Dai Z, Cao Q, Cheng Z, Han X, Yang L, Yuan Y, Zhang Y, Li K, Ma H, Shi C, Hong N, Szeszko P, Yu X, He Y. Disrupted resting-state functional connectivity in minimally treated chronic schizophrenia. Schizophrenia Research. 2014;156:150–156. doi: 10.1016/j.schres.2014.03.033. [DOI] [PubMed] [Google Scholar]

[bib79] Wang Y, Wang J, Jia Y, Zhong S, Niu M, Sun Y, Qi Z, Zhao L, Huang L, Huang R. Shared and specific intrinsic functional connectivity patterns in unmedicated bipolar disorder and major depressive disorder. Scientific Reports. 2017;7:8. doi: 10.1038/s41598-017-03777-8. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib80] White J, Beck CR, Harel T, Posey JE, Jhangiani SN, Tang S, Farwell KD, Powis Z, Mendelsohn NJ, Baker JA, Pollack L, Mason KJ, Wierenga KJ, Arrington DK, Hall M, Psychogios A, Fairbrother L, Walkiewicz M, Person RE, Niu Z, Zhang J, Rosenfeld JA, Muzny DM, Eng C, Beaudet AL, Lupski JR, Boerwinkle E, Gibbs RA, Yang Y, Xia F, Sutton VR. POGZ truncating alleles cause syndromic intellectual disability. Genome Medicine. 2016;8:3. doi: 10.1186/s13073-015-0253-0. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib81] Willsey AJ, Sanders SJ, Li M, Dong S, Tebbenkamp AT, Muhle RA, Reilly SK, Lin L, Fertuzinhos S, Miller JA, Murtha MT, Bichsel C, Niu W, Cotney J, Ercan-Sencicek AG, Gockley J, Gupta AR, Han W, He X, Hoffman EJ, Klei L, Lei J, Liu W, Liu L, Lu C, Xu X, Zhu Y, Mane SM, Lein ES, Wei L, Noonan JP, Roeder K, Devlin B, Sestan N, State MW. Coexpression networks implicate human midfetal deep cortical projection neurons in the pathogenesis of autism. Cell. 2013;155:997–1007. doi: 10.1016/j.cell.2013.10.020. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib82] Wiltschko AB, Johnson MJ, Iurilli G, Peterson RE, Katon JM, Pashkovski SL, Abraira VE, Adams RP, Datta SR. Mapping Sub-Second structure in mouse behavior. Neuron. 2015;88:1121–1135. doi: 10.1016/j.neuron.2015.11.031. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib83] Wondolowski J, Dickman D. Emerging links between homeostatic synaptic plasticity and neurological disease. Frontiers in Cellular Neuroscience. 2013;7:223. doi: 10.3389/fncel.2013.00223. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib84] Xing M, Tadayonnejad R, MacNamara A, Ajilore O, DiGangi J, Phan KL, Leow A, Klumpp H. Resting-state theta band connectivity and graph analysis in generalized social anxiety disorder. NeuroImage: Clinical. 2017;13:24–32. doi: 10.1016/j.nicl.2016.11.009. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib85] Ye Y, Cho MT, Retterer K, Alexander N, Ben-Omran T, Al-Mureikhi M, Cristian I, Wheeler PG, Crain C, Zand D, Weinstein V, Vernon HJ, McClellan R, Krishnamurthy V, Vitazka P, Millan F, Chung WK. De novo POGZ mutations are associated with neurodevelopmental disorders and microcephaly. Molecular Case Studies. 2015;1:a000455. doi: 10.1101/mcs.a000455. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib86] You X, Norr M, Murphy E, Kuschner ES, Bal E, Gaillard WD, Kenworthy L, Vaidya CJ. Atypical modulation of distant functional connectivity by cognitive state in children with autism spectrum disorders. Frontiers in Human Neuroscience. 2013;7:00482. doi: 10.3389/fnhum.2013.00482. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib87] Zeng K, Kang J, Ouyang G, Li J, Han J, Wang Y, Sokhadze EM, Casanova MF, Li X. Disrupted brain network in children with autism spectrum disorder. Scientific Reports. 2017;7:16440. doi: 10.1038/s41598-017-16440-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib88] Zhao W, Tan J, Zhu T, Ou J, Li Y, Shen L, Wu H, Han L, Liu Y, Jia X, Bai T, Li H, Ke X, Zhao J, Zou X, Hu Z, Guo H, Xia K. Rare inherited missense variants of POGZ associate with autism risk and disrupt neuronal development. Journal of Genetics and Genomics. 2019;46:247–257. doi: 10.1016/j.jgg.2019.04.002. [DOI] [PubMed] [Google Scholar]

PERMALINK

Altered hippocampal-prefrontal communication during anxiety-related avoidance in mice deficient for the autism-associated gene Pogz

Margaret M Cunniff

Eirene Markenscoff-Papadimitriou

Julia Ostrowski

John LR Rubenstein

Vikaas Singh Sohal

Roles

Abstract

Introduction

Results

Pogz+/- mice have decreased anxiety-related avoidance in the EPM

Figure 1. Pogz+/- mice exhibit reduced avoidance in the elevated plus maze (EPM).

Figure 1—figure supplement 1. Sex differences do not account for elevated-plus maze (EPM) phenotypes.

Figure 1—figure supplement 2. Other behavioral assays in Pogz+/- mice.

Figure 1—figure supplement 3. Distributions of sex and age for WT and Pogz Het mice used in all experiments.

Pogz+/- mice have reduced hippocampal-prefrontal theta synchrony

Figure 2. Pogz+/- mice have reduced vHPC-PFC theta synchrony both at baseline and in the elevated-plus maze (EPM).

Figure 2—figure supplement 1. LFP power in various frequency bands in the vHPC and mPFC is not changed in Pogz+/- mice.

Figure 2—figure supplement 2. Location of LFP electrodes (A–C) mPFC electrode locations.

An unbiased, data-driven approach to examine the significance of vHPC-mPFC theta synchrony for normal behavior and Pogz+/- mice

Table 1. Single frequency LFP measures used as features in PCA/ICA analysis.

Table 2. Multiple frequency LFP measures used as features in PCA/ICA analysis.

Figure 3. An unbiased, data-driven approach confirms that theta-frequency vHPC-mPFC communication is behaviorally-relevant and deficient in Pogz+/- mice.

Figure 3—figure supplement 1. Activity in conserved independent components (ICs) during approaches to the center of the EPM.

vHPC excitation of mPFC interneurons is deficient in Pogz+/- mice

Figure 4. Excitatory hippocampal input to prefrontal fast-spiking interneurons (FSINs) is reduced in Pogz mutants.

Figure 4—figure supplement 1. Intrinsic properties of prefrontal FSIN are not changed in Pogz+/- mice.

Figure 5. Excitatory hippocampal input to prefrontal pyramidal neurons is not changed in Pogz mutants.

Figure 5—figure supplement 1. Pyramidal cell properties are not changed in Pogz+/-mice.

Deficient FSIN excitation impairs information transmission across vHPC-mPFC circuits

Figure 6. Reducing the excitatory drive onto prefrontal FSINs impairs the transmission of hippocampal inputs.

Figure 6—figure supplement 1. Adding feedforward disinhibition does not change the relationship between inhibitory strength and hippocampal correlation.

Figure 6—figure supplement 2. The effect of reducing inhibition on the transmission of signals across hippocampal-prefrontal synapses depends on the frequency of hippocampal input.

Discussion

vHPC–mPFC communication and anxiety

Excitatory-inhibitory (E-I) balance in anxiety and autism

Deficits in long-range communication in autism

Possible relevance of Pogz behavioral phenotypes to autism

Conclusion

Materials and methods

Key resources table.

Subjects and behavioral assays

Elevated plus maze

Social/novel assay

Marble burying

Cognitive tasks

Rule shifting

Delayed match to sample task

Local field potential recordings

Whole cell patch clamp recordings

Computational model of the role of feedforward inhibition

Statistics, data analysis, and data and code availability

Table 3. Details of all statistical tests N indicates biological replicates for example individual cells or behavior trials.

Acknowledgements

Funding Statement

Contributor Information

Funding Information

Additional information

Competing interests

Author contributions

Ethics

Additional files

Data availability

References

Decision letter

Roles

Author response

Author response image 1.

Associated Data

Data Citations

Supplementary Materials

Data Availability Statement

Table 3. Details of all statistical tests N indicates biological replicates for example individual cells or behavior trials.

ACTIONS

PERMALINK

RESOURCES

Similar articles

Cited by other articles

Links to NCBI Databases

Pogz^+/- mice have decreased anxiety-related avoidance in the EPM

Figure 1. Pogz^+/- mice exhibit reduced avoidance in the elevated plus maze (EPM).

Figure 1—figure supplement 2. Other behavioral assays in Pogz^+/- mice.

Pogz^+/- mice have reduced hippocampal-prefrontal theta synchrony

Figure 2. Pogz^+/- mice have reduced vHPC-PFC theta synchrony both at baseline and in the elevated-plus maze (EPM).

Figure 2—figure supplement 1. LFP power in various frequency bands in the vHPC and mPFC is not changed in Pogz^+/- mice.

An unbiased, data-driven approach to examine the significance of vHPC-mPFC theta synchrony for normal behavior and Pogz^+/- mice

Figure 3. An unbiased, data-driven approach confirms that theta-frequency vHPC-mPFC communication is behaviorally-relevant and deficient in Pogz^+/- mice.

vHPC excitation of mPFC interneurons is deficient in Pogz^+/- mice

Figure 4—figure supplement 1. Intrinsic properties of prefrontal FSIN are not changed in Pogz^+/- mice.

Figure 5—figure supplement 1. Pyramidal cell properties are not changed in Pogz^+/-mice.