Connectivity of the neuronal network for contextual fear memory is disrupted in a mouse model of third-trimester binge-like ethanol exposure

Abstract

Background: In rodents, third-trimester-equivalent alcohol exposure (TTAE) produces significant deficits in hippocampal-dependent memory processes such as contextual fear conditioning (CFC). The present study sought to characterize changes in both behavior and Fos⁺ neurons following CFC in ethanol (EtOH)-treated versus saline-treated mice using TRAP2:Ai14 mice that permanently label Fos⁺ neurons following a tamoxifen injection. We hypothesized that TTAE would produce long-lasting disruptions to the networks engaged following CFC with a particular emphasis on the limbic memory system.

Methods: On postnatal day 7, mice received either two injections of saline or 2.5 g/kg EtOH spaced 2 h apart. The mice were left undisturbed until they reached adulthood, at which point they underwent CFC. After context exposure on day 2, mice received a tamoxifen injection. Brain tissue was harvested. Slides were automatically imaged using a Zeiss AxioScanner. Manual counts on a priori regions of interest were conducted. Automated counts were performed on the whole brain using the QUINT 2D stitching pipeline. Last, novel network analyses were applied to identify future regions of interest.

Results: TTAE reduced context recall on day 2 of CFC. Fos⁺ neural density increased in the CA1 and CA3. Fos⁺ counts were reduced in the anteroventral (AV) and anterodorsal thalamus. The limbic memory system showed significant hyperconnectivity in male TTAE mice, and the AV shifted affinity toward hippocampal subregions. Last, novel regions such as a subparafascicular area and basomedial amygdalar nucleus were implicated as important mediators.

Discussion: These results suggest that CFC is mediated by the limbic memory system and is disrupted following TTAE. Given the increase in CA1 and CA3 activity, a potential hypothesis is that TTAE causes disruptions to memory encoding following day 1 conditioning. Future studies will aim to determine whether this disruption specifically affects the encoding or retrieval of fear memories.

Keywords: c‐fos; engram; ethanol; fetal; learning; memory.

Figure 1.. TTAE produces significant decreases in freezing behavior following CFC.

A. Movement was captured by VideoFreeze and was centered on 4 shock timings. All shocks were pooled together and the mean was taken to assess if there were significant differences in animals’ response to shocks on day 1. The blue lightning bolt and yellow shading indicate time points when the shock was active. Lowercase a and lowercase b indicate the time points wherein our shock sensitivity metric was calculated. B. Shock sensitivity, measured as the ratio between a and b, was calculated for each sex and treatment condition per animal. A main effect of both sex and treatment were found with an additional interaction between treatment and sex. Values closer to 1 indicate that movement exclusively happened during the shock period, and a subsequent freeze occurred quickly after the shock ended. C. The percentage of the time the animal spent freezing on day 2 was assessed. A main effect of sex and treatment were detected, both lowered the expression of freezing. D.1, 2 No correlations between the percentages of time spent freezing or shock sensitivity was detected in females and males.

Figure 2.. TTAE produces increases in CA1 and CA3 Fos+ neural density following CFC.

A. Representative images show the ATN. B. Representative images show the hippocampal formation including the CA1, CA3, and DG. C. Representative images show the RSP. D. No main effect of either treatment or sex was detected in the ATN manual counts. E. A main effect of treatment, but not sex, was detected in the CA1. F. A main effect of treatment, but not sex, was detected in the CA3. G. No main effects of either sex or treatment was detected in the DG. H. No main effects of either sex or treatment was detected in the RSP.

Figure 3.. QUINT Further Splits ATN into AD and AV.

A. A Hartigan’s dip test confirmed bimodality in the manual counts of both female and B. male saline groups. C. QUINT counts separated the AV and AD that were previously combined as a singular ATN. A significant main effect of treatment was found in the AV wherein TTAE decreased QUINT FOS counts. D. A significant main effect of treatment was also found in the AD wherein TTAE decreased QUINT FOS counts.

Figure 4.. Functional connectivity of the limbic memory system shows alterations following TTAE and CFC.

A. The Pearson correlations of 13 brain regions part of or connected with the limbic memory system are shown for each sex and treatment condition. Warmer colors indicate higher, positive correlations whereas cooler colors indicate smaller, negative correlations. B. Hierarchical agglomerative clustering was performed on the 13 brain regions analyzed. The cluster cutoff refers to the height of the dendrogram wherein 0.5 indicates the halfway point. A rightward shift in EtOH treated animals is observed which suggests differences in clustering emerge in male, EtOH animals. C. The variation of information compares the EtOH and Saline clusters for male and female animals to determine at which point they are maximally different. D. The point of maximal difference between EtOH and Saline animals is the cluster cutoff point chosen to assign each brain region into its respective clusters. Critically, we see a shift in clustering between the AV and its neighboring anterior thalamic nuclei towards hippocampal sub regions. Abbreviations. ACA: Anterior cingulate cortex. RSP: Retrosplenial cortex. DG: Dentate Gyrus. SUB: Subiculum. AV: Anteroventral thalamic nucleus. AM: Anteromedial thalamic nucleus. AD: Anterodorsal thalamic nucleus. TM: Tuberomammillary body. PMd: Dorsal premammillary nucleus. PMv: Ventral premammillary nucleus.

Figure 5.. Brainwide hyperconnectivity may emerge following TTAE.

A. The Pearson correlations of 120 brain regions are shown. Identities of values can be found in Table 1. B. Hierarchical agglomerative clustering was performed. The cluster cutoff points represent the height of the dendrogram such that 0.5 indicates the number of clusters halfway up the dendrogram. C. The variation of information is a metric of how different the cluster identities are between Saline and EtOH conditions. D. The point of maximal variation of information is used to assign and plot final clusters. Qualitatively it is observed in the EtOH condition that distances shrink indicated by the presence of warmer colors.

Figure 6.. Broad increases in participation coefficients indicate hyperconnectivity with the parafasicular area of the thalamus being most commonly and greatly affected.

A. Within-Module Z-Scores (WMZs) were calculated for each brain region assessed and showed no overall difference between treatments. B. The difference between Saline and EtOH treatment WMZs was calculated for each brain region and sex. The 95th percentile was plotted for each sex with those in green being the highest change in Male WMZ, those in pink being the highest change in Female WMZs and those in gray being high in both. C. Participation coefficients were calculated for each sex and treatment condition with the females showing a higher increase following EtOH. D. The change in participation coefficient is calculated for each sex. Abbreviations: ARH: Arcuate hypothalamic nucleus. MPN: Medial preoptic nucleus. SCH: Suprachiasmatic nucleus. IMD: Intermediodorsal nucleus. VISam: Anteromedial visual area. Xi: Xiphoid thalamic nucleus. STR: Striatum. SSp: Primary somatosensory area. MEA: Medial amygdalar nucleus. AUDd: Dorsal auditory area. GP: Globus pallidus. VPM: Ventral posteromedial nucleus. STN: Subthalamic nucleus. SPF: Subparafascicular nucleus. LS: Lateral septal nucleus. TM: Tuberomamillary nucleus. PA: Posterior amygdalar nucleus. ME: Median eminence. SPA: Subparafascicular area. PF: Parafascicular nucleus. SubG: Subgeniculate nucleus. IGL: Intergeniculate leaflet of the lateral geniculate complex. FF: Fields of Forel.

Figure 7.. Stepwise linear regression suggests the BMA may be an additional, critical region.

A. All significant predictor variables and their covariance is shown. Largely, there are no covariant predictors with the exception of interactions. B. Coefficient beta values are depicted. Positive values indicate that more Fos+ neurons in that brain region is predictive of increases in freezing. Negative values indicate that more Fos+ neurons in that brain region is predictive of decreases in freezing. C. An individual conditional expectation plot is shown depicting the effects of BMA on freezing in both the Saline and EtOH conditions. BMA Fos+ neurons interact with treatment. D. BMA interacts with treatment to moderate freezing behavior in EtOH animals.