Ethanol sustains phosphorylated tau protein in the cultured neonatal rat hippocampus: Implications for fetal alcohol spectrum disorders

Abstract

Fetal Alcohol Spectrum Disorders (FASDs) are comprised of developmental, behavioral, and cognitive abnormalities caused by prenatal alcohol exposure, affecting an estimated 2%–5% of children and costing $4 billion annually in the United States. While some behavioral therapies help, the neurobiological mechanisms that underpin FASDs need further elucidation for development of effective pharmacotherapeutics. The role of the tau protein in the hippocampus is likely to be involved. Tau catalyzes microtubule polymerization in developing neurons. However, this function can become disrupted by hyperphosphorylation. Many of the cognitive deficits observed in neurodegenerative tauopathies overlap to some degree with what is observed in juvenile developmental disabilities, such as FASDs (e.g., selective memory, executive dysfunction). Thus, tau protein phosphorylation may be one important mechanism of dysfunction in FASDs. The purpose of this study is to provide an empirical basis for a tauopathic characterization of FASDs. To do so, hippocampal slices were extracted from rats at postnatal day 10 (PND10); hippocampal slices were then exposed to 5 days of 50-mM ethanol between 6 days in vitro (DIV) and 11DIV. Immunoblots were taken for Total and p-Tau (Threonine231) at 12DIV and 24DIV. Immunohistochemical fluorescent images were taken for p-Tau (Threonine231) at 12DIV and 24DIV. Separate p-Tau measures were taken for the cornu ammonis 1 (CA1), CA3, and dentate gyrus (DG). Total Tau protein expression remained unchanged between 12DIV and 24DIV regardless of ethanol condition. In the control group, longer DIV was associated with decreased p-Tau. However, in the ethanol-exposed group, p-Tau was sustained across DIV. This is the first study to show that ethanol exposure sustains tau Threonine231 phosphorylation in the perinatal hippocampus regardless of Total Tau expression. These findings could lead to innovative pharmacotherapeutic targets for the treatment of cognitive deficits seen in FASDs.

Introduction

Fetal alcohol spectrum disorders (FASDs) result from teratogenic prenatal ethanol exposure and include individuals who are diagnosed with partial or full Fetal Alcohol Syndrome (pFAS; FAS), alcohol-related neurodevelopmental disorder (ARND), alcohol-related birth defects (ARBD), and neurobehavioral disorders associated with prenatal alcohol exposure (ND-PAE). These conditions are on a continuum of severity, with FAS patients presenting a greater magnitude of physical (lower birth weight, thin vermilion, smooth philtrum, flat midface, shortened palpebral fissure) and neurodevelopmental (microcephaly, cerebellar hypoplasia) abnormalities compared to other FASDs (Bertrand et al., 2004; de la Ferreira & Cruz, 2017; Jones & Smith, 1973; Kully-Martens, Denys, Treit, Tamana, & Rasmussen, 2012). Prenatal ethanol exposure is the most common cause of preventable developmental disability: an estimated 2–5% of children are affected by FASDs, annually costing up to $4 billion in the United States (Lupton, Burd, & Harwood, 2004; May et al., 2009). Although early behavioral intervention somewhat improves behavioral and cognitive outcomes for children diagnosed with FASDs (for a comprehensive analysis, see Bertrand & Interventions for Children with Fetal Alcohol Spectrum Disorders Research Consortium, 2009), there are currently no pharmacological treatments specifically for FASDs. In order to develop such treatments and the current standard of care for patients, the neurobiological mechanisms that underpin these disorders need further elucidation.

The rodent CNS is an excellent animal model of human CNS development, and also shows susceptibility to ethanol insult. During perinatal rodent CNS development, ethanol is capable of inducing aberrant neonatal epigenetic modification (for a review, see Basavarajappa & Subbanna, 2016) and disrupting proliferation, axonal growth, white matter production, and synaptic communication (Lindsley, Kerlin, & Rising, 2003; Mathews, Dewees, Diaz, & Favero, 2021; Redila et al., 2006; Sadrian, Subbanna, Wilson, Basavarajappa, & Saito, 2012). Additionally, ethanol can produce deficits observed in paradigms thought to depend upon proper function of the rodent hippocampus, such as performance in Y-maze, radial arm maze, and object recognition tasks (Berman & Hannigan, 2000; Goodlett & Johnson, 1997; Johnson & Goodlett, 2002; Subbanna & Basavarajappa, 2014; Wagner, Zhou, & Goodlett, 2014). The deleterious effects that ethanol has upon the axonal development in the hippocampus may partially be explained by its action on microtubules, microtubule-associated proteins (MAPs) (Smith, Butler, & Prendergast, 2013), and their kinases, which have the capacity to alter the function of MAPs by way of regulatory phosphorylation (Ahluwalia, Ahmad, Adeyiga, Wesley, & Rajguru, 2000). Of the ethanol-sensitive MAPs, the tau protein plays an integral role in ubiquitous axonal development and structural maintenance (Gendron, McCartney, Causevic, Ko, & Yen, 2008). However, the tau protein can become pathological under dysregulated or untimely hyperphosphorylation, resulting in a battery of cognitive deficits and disorders known as tauopathies (for a review, see Orr, Sullivan, & Frost, 2017).

The most carefully studied tauopathies are neurodegenerative disorders. There are likely to be similarities and differences in the mechanisms and dysfunctions of hippocampal hyperphosphorylated tau (p-Tau) dysfunction in neurodegenerative tauopathies compared to FASDs. Specifically, the effects of p-Tau are functionally distinct at different stages in brain development (Yoshida & Ihara, 1993; Yu et al., 2009). For instance, during early CNS development, tau proteins are involved in catalyzing microtubule polymerization, synaptogenesis, neural migration, axonal transport, and cytoskeletal support (Cleveland, Hwo, & Kirschner, 1977; Dayanandan et al., 1999; Gendron et al., 2008; Mandelkow, Stamer, Vogel, Thies, & Mandelkow, 2003; Saito et al., 2010; Wang & Liu, 2008) – functions that are critical for proper development of the fetal and perinatal CNS (Tau & Peterson, 2010).

Interestingly, the tau protein exists in a highly phosphorylated state during fetal and perinatal development in both humans and rodents (Brion, Smith, Couck, Gallo, & Anderton, 1993; Kenessey & Yen, 1993), and this phosphorylation state dynamically changes across different amino acid residues on the tau protein and across different developmental periods and subregions in the rat hippocampus (Yu et al., 2009). While it has been shown that the tau protein is sensitive to ethanol exposure in humans and rats (Ahluwalia et al., 2000; Gendron et al., 2008) and that tau phosphorylation changes across different residues at different periods of neonatal development in the rat hippocampus (Yu et al., 2009), the effects that ethanol exerts on the constitutive changes to tau phosphorylation in the developing neonatal rat hippocampus have yet to be established.

In this study, we investigate the distinct changes to the phosphorylation state of the tau Threonine 231 (Thr231) residue. In some cases, the developmental changes in the tau Thr231 phosphorylation state are inconclusive (Yu et al., 2009); however, assuming there are mechanistic similarities between neurodegenerative tauopathies and the neurodevelopmental tauopathy we aim to establish here (FASDs), we selected the Thr231 residue for three reasons: 1) Thr231 phosphorylation disrupts microtubule polymerization by tau (Lin et al., 2007); 2) Thr231 phosphorylation has been established as a pathological residue in neurodegenerative diseases involving the hippocampus (Chen, 2005); and 3) Thr231 has a distinct relationship with and is often hyperphosphorylated by the kinase glycogen synthase kinase-3β (GSK-3β) (Lin et al., 2007), the activity of which is disinhibited by ethanol exposure (Chen et al., 2009; Liu et al., 2009; Luo, 2009). Taken together, we hypothesize that perinatal ethanol exposure may sustain Thr231 tau phosphorylation across time.

This is the first study to use the organotypic hippocampal slice culture (OHSC) model to document the effect of ethanol in the phosphorylated state of the tau Thr231 residue across time. In this report, we show the effects of ethanol exposure on the phosphorylation state of the tau Thr231 residue out to 24 days in vitro (DIV) in the cultured rat hippocampus.

Materials and methods

Organotypic hippocampal slice culture

All experiments complied with the National Research Council’s Guide for the Care and Use of Laboratory Animals and were approved by the University of Kentucky Institutional Animal Care and Use Committee. Thirty-one male (n = 21) and female (n = 10) Sprague Dawley pups (Harlan Laboratories; Indianapolis, Indiana, United States) were humanely euthanized at postnatal day 10 (PND10). The rodent CNS at PND10 is roughly equivalent to gestational termination in the third trimester human (for a review, see Semple, Blomgren, Gimlin, Ferriero, & Noble-Haeusslein, 2013; Wagner et al., 2014). Additionally, rat hippocampi undergo accelerated development between PND0ePND21 (Lee, Kim, Cho, Kim, & Park, 2017), and therefore, this age and model were selected to investigate the impact of ethanol on rapid neonatal hippocampal development and were intended to inform research investigating the impact of prenatal ethanol exposure on rodent hippocampal development when modeling the human condition in FASDs.

Full brains were aseptically removed and placed into chilled dissecting media solution containing: 97.19% (v/v) Minimum Essential Media (MEM; Invitrogen; Carlsbad, California, United States), 0.024 M 4-(2-Hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES; Sigma; St. Louis, Missouri, United States), 0.97% streptomycin/penicillin (Invitrogen), and 1.94% amphotericin B solution (Sigma). Brains were then midsagittally transected, and hippocampi were removed. Any excess tissue that was attached to the hippocampi following harvest was removed using a No. 15 scalpel. These hippocampi were then flatly positioned upon a plastic platform and sectioned into 200-μm slices using the McIllwain Tissue Chopper (Mickle Laboratory Engineering Co. Ltd.; Gomshall, United Kingdom). We found limited evidence of dorsalventral functional distinction of the hippocampus at PND10 (Lee et al., 2017), and therefore, we generated no hypotheses and did not investigate or control for hippocampal slices along the dorsalventral axis. Slices that retained the CA1, CA3, and dentate gyrus (DG) (Fig. 1) were then selected to be plated upon Millicell biopore membranes (0.4 mm, 30-mm diameter; MilliporeSigma; Burlington, Massachusetts, United States); each membrane supported four hippocampal slices. One advantage of the OHSC is that it minimizes animal resources by multiplying viable tissue per one animal sacrificed; as such, each animal generated several 200-μm hippocampal slices. Slices from each animal were maintained in culture media in an animal- and sex-labeled petri dish and segregated by animal sex. Upon plating the slices onto the culture inserts, the petri-dish labels were removed, and slices were randomly plated by sex using a trimmed transfer pipette to pull from different petri dishes within sex. While the number of pups by sex varies in each litter, a single animal may have only contributed to 0–2 slices per culture inserts, which were later randomly assigned to the experimental conditions so that the animals were represented in each experimental group. This was meant to mitigate the impact that one rat may have had on any given experimental group.

Fig. 1.

a) Representative image of 12DIV control. b) Representative image of 12DIV ethanol. c) Representative image of 24DIV control. d) Representative image of 24DIV ethanol. p-Thr231 fluorescence significantly decreased in control slices between 12 (a) and 24DIV (c); p-Thr231 fluorescence was sustained in ethanol-exposed slices between 12 (b) and 24DIV (d) except in the CA3 region.

Slices upon membranes were inserted into 6-well plates (5 mL, 9.5 cm² area; VWR; Radnor, Pennsylvania, United States) containing 1.5 mL of culturing media and placed in humidity-controlled incubation at 37 °C, 5% CO₂, and 95% air. Culturing media contained: 49.26% dissecting media (detailed above), 22.17% double distilled water, 24.63% heat-inactivated horse serum (HIHS; Sigma), 2.46% Hanks balanced salt solution (HBSS; Invitrogen), 0.49% penicillin/streptomycin (Invitrogen), and 0.99% amphotericin B solution (Sigma). Any excess culturing media that pooled on top of the membrane or around any of the hippocampal slices was removed using a micropipette to improve slice adherence and ensure oxygen exposure. Slices were permitted 6 days of uninterrupted incubation to heal from removal and to securely anchor into the membrane prior to any manipulations.

Ethanol treatment

Control slices received unadulterated culture media for the entire experimental duration. Experimental slices received 5 days of 50-mM ethanol exposure in culture media from 6DIV to 11DIV, and then received plain culture media for the remainder of the experimental timeline; this was meant to mimic a human late third-trimester equivalent maternal ethanol binge. Atmospheric treatment was employed to prevent ethanol evaporation during the 5-day exposure following previously published procedures (Prendergast et al., 2004). Briefly, ethanol-exposed slices in culture plates that received 50-mM ethanol media between 6 and 11DIV were placed into a Tupperware^® container with a solution of 50-mM ethanol in water, placed into a Ziploc^® bag filled with 5% CO₂ and 95% air, and then sealed and placed into the incubator. Similarly, culture plates containing control slices were placed into a Tupperware^® container with 50 mL of distilled water and placed into a Ziploc^® bag, filled with 5% CO₂ and 95% air, and then sealed and placed into the incubator.

Experiment 1. Effect of ethanol exposure on hippocampal Total and p-Tau at 12 and 24DIV

Cell lysis procedure

We assessed Total and p-Tau using the western blot technique at 12DIV and 24DIV. Hippocampal slices (n = 96; 48 females) were evenly distributed across control and ethanol groups at 12DIV and 24DIV. In order to amplify the signal, slices from membranes within respective conditions and sex were grouped and scraped from the membrane into chilled 1X PBS buffer containing: protease inhibitor cocktail III (1:200) (MilliporeSigma) and 10-mM sodium fluoride (1:100) (Sigma). Tissue was then lysed in lysis buffer containing: 62.5-mM tris base (Fisher Scientific; Hampton, New Hampshire, United State), 6-M urea (Sigma), 10% glycerol (Fisher Scientific), 2% SDS (Bio-Rad; Hercules, California), protease inhibitor cocktail III (1:200) (MilliporeSigma), and 10-mM sodium fluoride (1:100) (Sigma). Samples were placed in ice for 30 min. Following dissociation, samples were centrifuged at 13,000 × g for 10 min at 4 °C. Supernatant was separated and stored at −80 °C for later use.

Protein quantification and western blot procedure for Total and p-Tau

Proteins from hippocampal slice cultures were extracted as described above, equalized using the Pierce BCA Assay (Thermo Fisher), and loaded onto 5–20% gradient gels for SDS PAGE (Bio-Rad). Proteins were transferred onto nitrocellulose membranes using the Trans-Blot Turbo Transfer System (Bio-Rad) and blocked for 1 h in 5% bovine serum albumin in 1X TBS containing 0.05% Tween-20 (Fisher Scientific) (TBST) at room temperature. Membranes were incubated with Total Tau Monoclonal Antibody (TAU-5) (Thermo Fisher; Cat# AHB0042; RRID: AB_2536235; 1:500), Phospho-Tau (Thr231) Monoclonal Antibody (AT180) (Thermo Fisher; Cat# MN1040; RRID: AB_223649; 1:500), and anti-GAPDH antibody (SigmaeAldrich; Cat# G9545, RRID: AB_796208; 1:20,000). Total and p-Tau primary antibodies were independently co-incubated with GAPDH antibody, diluted in blocking solution (1X TBST, 0.05% Tween 20, 5% bovine serum albumin) overnight at 4 °C.

Membranes were washed with 1X TBST prior to secondary antibody incubation. The secondary antibodies used in these experiments were: IRDye 680RD goat anti-Rabbit IgG (Li-Cor; Lincoln, Nebraska, United States; Cat# 925–68071: RRID: AB_2721181; 1:5000) and IRDye 800CW goat anti-mouse IgG (Li-Cor; Cat# 925–32210; RRID: AB_2687825; 1:1000). IRDye 800CW goat anti-mouse IgG secondary against Total and p-Tau primary antibodies were independently co-incubated with the IRDye 680RD goat anti-rabbit IgG against GAPDH primary antibody and placed on a rocker for 1 h at room temperature. Membranes were washed twice with 1X TBST and once with 1X TBS immediately prior to fluorescent imaging. Fluorescent images were acquired using the Li-Cor Odyssey CLx (Li-Cor) and quantified using optical densitometry on Image Studio (version 5.1; Li-Cor). A within-image capture measure of background fluorescent signal was removed from each lane to obtain an accurate fluorescent signal. Both Total and p-Tau signals were lane-corrected using respective GAPDH signals.

Statistical analysis for Total and p-Tau signal in western blot measures

Prior to analyses, the dependent variables (p-Tau signal and Total Tau signal) were examined for outliers ±3SD from the mean. There were no outliers within ±3SD from the mean for either p-Tau signal or Total Tau signal, as measured in arbitrary units [AU]. The variables that were included in the analyses were sex (male = 1, female = 0), ethanol exposure (ethanol = 1, control = 0), days in vitro (12DIV = 0,24DIV = 1), and an interaction between ethanol exposure and days in vitro. Models for western blots were estimated hierarchically, with each variable added in a new step. After estimating models for Total and p-Tau, it was determined that the assumption of homoscedasticity was violated for Total Tau signal. A natural log transformation of the dependent variable (Total Tau signal [AU]) improved this issue and was therefore used in the analysis for Total Tau. Results below are from models with p-Tau Signal [AU] and the transformed LN (Total Tau [AU]) dependent variables. Significant interactions indicating different associations between days in vitro and p-Tau signal for control vs. ethanol groups were probed by fitting the model again with ethanol group as the reference category, following the probing procedure from IHC assessments. This was only conducted for p-Tau signal [AU], as no interaction in LN (Total Tau [AU]) warranted further statistical probing.

Experiment 2: Identifying regional differences in the effect of ethanol exposure on hippocampal tau phosphorylation at 12 and 24DIV

We further assessed the effect of ethanol exposure by identifying regional tau phosphorylation in hippocampal slices (n = 169; 71 female) using immunohistochemistry (IHC) in the CA1, CA3, and DG of the hippocampus at 12DIV and 24DIV. For assessment of tau phosphorylation using IHC, all slices were first fixed in 10% formalin. Then the slices were washed twice in 1X PBS solution and placed into an IHC buffer solution containing: 1X PBS, Triton detergent (1:1000), and 760-nM bovine serum albumin for 30 min. The primary antibody used in these experiments was Phospho-Tau (Thr231) Monoclonal Antibody (AT180) (Thermo Fisher; Waltham, Massachusetts; Cat# MN1040; RRID: AB_223649). The primary antibody was diluted 1:40 in IHC buffer solution and administered onto slices for 24 h at 4 °C; the slices were then washed twice in 1X PBS. The secondary antibody used in these experiments was goat anti-mouse IgG (H + L), TRITC (Thermo Fisher; Cat# A16071; RRID: AB_2534744). The secondary antibody was diluted 1:200 in IHC buffer solution and administered onto slices for 24 h at 4 °C; the slices were then washed twice in 1X PBS immediately prior to fluorescent imaging. Images were acquired through a 5X objective with a Leica DMIRB microscope (W. Nuhsbaum, Inc.; McHenry, Illinois, United States) using a SPOT 7.2 color mosaic camera (W. Nuhsbaum, Inc.) on SPOT software for Windows (advanced version 4.0.2; W. Nuhsbaum, Inc.). Fluorescent intensity was measured group-blinded using optical densitometry in imageJ (National Institutes of Health; Bethesda, Maryland, United States). A within-image background measure of fluorescent intensity was removed from each CA1, CA3, and DG measurement to obtain accurate fluorescence of these hippocampal subregions.

Statistical analysis for p-Tau fluorescent intensity in IHC measures

Statistical analyses for IHC data were conducted to determine the effect of age and ethanol exposure on the fluorescent intensity of phosphorylated tau proteins in the different regions of the hippocampus. Prior to analyses, the dependent variable (p-Tau fluorescent intensity) was examined for outliers ±3SD from the mean. While there were three total outliers using the ±3SD criteria, exclusion of all three outliers did not affect model significance, and therefore the outliers were retained in the analyses in order to preserve sample size; no outliers existed below 3SD of the mean. Preliminary analyses included means, standard deviations, and bivariate correlations. Regression analyses were conducted in SPSS Version 26 (IBM Corporation; Armonk, New York, United States). Models were estimated separately for each region. The variables that were included in the analyses were sex (male = 1, female 0), ethanol exposure (ethanol = 1, control = 0), days in vitro (12DIV = 0, 24DIV = 1), and an interaction between ethanol exposure and days in vitro. Models were estimated hierarchically, with each variable added in a new step. After estimating models, it was determined that the assumption of normality of residuals was violated. A square root transformation of the dependent variable (p-Tau fluorescent intensity) improved this issue and was therefore used. IHC results reflect models with the transformed dependent variable. Significant interactions indicating different associations between days in vitro and p-Tau fluorescent intensity for control vs. ethanol groups were probed by fitting the model again with ethanol group as the reference category. In other words, the initial interaction model was fit using CTRL as the reference group in the “ethanol” dummy variable (ethanol = 1, CTRL = 0), and therefore the regression coefficient for DIV indicates the effect of DIV for the CTRL group; we further probed the interaction by creating a “CTRL” dummy variable (CTRL = 1, ethanol = 0) where we used the ethanol group as the reference group, making the regression coefficient for DIV now indicate the effect of DIV for the ethanol group. This revealed the nature of the interaction.

Results

Preliminary analyses for western blot p-Tau measures

As shown in Table 1, the observations for p-Tau signal are evenly distributed across sex and experimental conditions. Assignment to ethanol condition was unrelated to sex or p-Tau signal. Greater days in vitro was associated with lower p-Tau signal, r = −.50, p < .05.

Table 1.

Means, standard deviations, and bivariate correlations among western blot p-Tau experimental variables.

	1	2	3	4
1. Male
2. Days in vitro	.00
3. Ethanol	.00	.00
4. p-Tau signal	.24	−.50*	−.10
M	.50	.50	.50	18916.17
SD	.51	.51	.51	3126.36

Note: Male and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group);

^*p < .05.

Results for western blot p-Tau signal

As shown in Table 2, there was a significant interaction between ethanol and days in vitro, β = .63, p < .05, accounting for 13% of the variance in p-Tau signal beyond Male and the main effect of days in vitro. For the control group, p-Tau signal significantly decreased across days in vitro, β = −.86, p < .01. For the ethanol group, there was no significant association between days in vitro and p-Tau signal, β = −.14, n.s. The bar graph for these data is shown in Fig. 2.

Table 2.

Results from the regression model predicting western blot p-Tau signal in the hippocampus.

	β	ΔR2
Male	0.24	0.56
Days in vitro	−.86**	.25*
Ethanol	−.46	.01
Ethanol × Days in vitro	.63*	.13*
Total R2		.45*

Note: Coefficients are standardized. Male and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group).

^*p < .05;

^**p < .01.

Fig. 2.

Days in vitro by mean p-Tau Signal in the hippocampus. a) In CTRL samples, mean p-Thr231 signal significantly decreased between 12 and 24DIV. b) In samples exposed to ethanol, mean p-Thr231 signal was sustained between 12 and 24DIV.

Preliminary analyses for western blot Total Tau measures

As shown in Table 3, the observations for Total Tau signal are evenly distributed across sex and experimental conditions. Assignment to ethanol condition was unrelated to sex or Total Tau signal. There was no association between days in vitro and any other experimental variable.

Table 3.

Means, standard deviations, and bivariate correlations among western blot Total Tau experimental variables.

	1	2	3	4
1. Male
2. Days in vitro	.00
3. Ethanol	.00	.00
4. LN (Total Tau signal)	.01	−.06	−.21
M	.50	.50	.50	9.72
SD	.51	.51	.51	.21

Note: Male and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group).

Results for western blot Total Tau signal

As shown in Table 4, there was no significant interaction between ethanol and days in vitro, β = .48, n.s. The bar graph for these data is shown in Fig. 3. Photographic renderings of western blot protein bands are displayed in Fig. 4.

Table 4.

Results from the regression model predicting western blot Total Tau signal in the hippocampus.

	β	ΔR2
Male	.01	.00
Days in vitro	.22	.00
Ethanol	.49	.04
Ethanol × Days in vitro	−.48	.08
Total R2		.12

Note: Coefficients are standardized. Male and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group).

Fig. 3.

Days in vitro by LN (mean p-Tau Signal) in the hippocampus. a) In CTRL samples, mean Total Tau signal was unchanged between 12 and 24DIV. b) In samples exposed to ethanol, mean Total Tau signal was unchanged between 12 and 24DIV.

Fig. 4.

Representative image of protein bands for a) p-Tau Thr231 and b) Total Tau Signal. Bands were lane-corrected for GAPDH. Full blots are provided in Supplementary Fig. 1 (p-Tau) and Supplementary Fig. 2 (Total Tau).

Preliminary analyses for IHC measures

As shown in Table 5, the observations are evenly distributed across the three regions, sex, and experimental conditions. Assignment to ethanol condition was related to days in vitro (r = −.17, p < .01), but unrelated to region or sex. The fluorescent intensity was higher in the CA1 region compared to the other regions, r = .22, p < .01. The fluorescent intensity was lower in DG, r = −.17, p < .01, compared to the other regions. Across all regions, greater days in vitro were associated with lower fluorescent intensity, r = −.25, p < .01. Also across all regions, observations in the Male (r = .09, p < .05) and ethanol (r = .14, p < .01) conditions had significantly higher fluorescent intensity.

Table 5.

Means, standard deviations, and bivariate correlations among study variables in Experiment 2.

	1	2	3	4	5	6	7
1. CA1
2. CA3	−.33**
3. Dentate gyrus	−.33**	−.33**
4. Male	.00	.00	.00
5. Days in vitro	.00	.00	.00	−.07
6. Ethanol	.00	.00	.00	.00	−.17**
7. SQRT (Intensity)	.22**	−.08*	−.17**	.09*	−.21**	.14**
M	0.25	0.25	0.25	0.58	0.58	0.54	4.31
SD	0.43	0.43	0.43	0.49	0.50	0.50	0.74

Note: CA1, CA3, Dentate gyrus, Male, and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group);

^*p < .05,

^**p < .01.

Results for the CA1 p-Tau fluorescent intensity

As shown in Table 6, there was a significant interaction between ethanol and days in vitro, β = .44, p < .05, accounting for 4% of the variance in fluorescent intensity beyond Male and the main effects in the CA1. For the control group, fluorescent intensity significantly decreased across days in vitro, β = −.29, p < .05. For the ethanol group, there was no significant association between days in vitro and fluorescent intensity, β = .11, n.s. The bar graph for these data is shown in Fig. 5.

Table 6.

Results of regression models predicting p-Tau fluorescent intensity by region.

Variable	Dependent variables
	CA1	CA3	Dentate gyrus
Male	.07	.09	−.06
R2	.01	.01	.00
Days in vitro	.28*	−.38**	−.25***
ΔR2	.01	.07***	.03*
Ethanol	−.31	−.11	.03
ΔR2	.00	.00	.00
Ethanol × Days in vitro	.44*	.17	.25*
ΔR2	.04*	.01	.03*
Total R2	.06^	.11**	.06*

Note: Three models were fit, one for each region of the hippocampus. Coefficientsare standardized. Male and Ethanol are dummy variables in which 1 = membership in that category and 0 = not a member of that category (e.g., female, control group);

^{^}p = .05;

^*p < .05,

^**p < .01,

^***p < .001.

Fig. 5.

Days in vitro by SQRT (p-Tau) fluorescent intensity in the hippocampal CA1. a) In CTRL samples, p-Thr231 fluorescent intensity significantly decreased between 12 and 24DIV. b) In samples exposed to ethanol, p-Thr231 fluorescent intensity was sustained between 12 and 24DIV.

Results for the CA3 p-Tau fluorescent intensity

See Table 6. There was no significant interaction between ethanol and days in vitro, β = .17, n.s. For the control group, fluorescent intensity significantly decreased across days in vitro, β = −.38, p < .01. For the ethanol group, fluorescent intensity significantly decreased across days in vitro, β = −.23, p < .05. The bar graph for these data is shown in Fig. 6.

Fig. 6.

Days in vitro by SQRT (p-Tau) fluorescent intensity in the hippocampal CA3. a) In CTRL samples, p-Thr231 fluorescent intensity significantly decreased between 12 and 24DIV. b) In samples exposed to ethanol p-Thr231, fluorescent intensity significantly decreased between 12 and 24DIV. Decrement in tau phosphorylation was unaffected by ethanol exposure only in the CA3 region.

Results for the dentate gyrus p-Tau fluorescent intensity

As shown in Table 6, there was a significant interaction between ethanol and days in vitro, β = .25, p < .05, accounting for 3% of the variance in fluorescent intensity beyond Male and the main effects in the DG. For the control group, fluorescent intensity significantly decreased across days in vitro, β = −.34, p < .01. For the ethanol group, there was no significant association between days in vitro and fluorescent intensity, β = .01, n.s. The bar graph for these data is shown in Fig. 7.

Fig. 7.

Days in vitro by SQRT (p-Tau) fluorescent intensity in the hippocampal DG. a) In CTRL samples, p-Thr231 fluorescent intensity significantly decreased between 12 and 24DIV. b) In samples exposed to ethanol, p-Thr231 fluorescent intensity was sustained between 12 and 24DIV.

Discussion

To our knowledge, this is the first study to investigate the impact of ethanol exposure on tau protein Thr231 phosphorylation across neonatal hippocampal culture. In Experiment 1, we demonstrated that ethanol exposure had no effect on Total Tau expression between 12 and 24DIV. The phosphorylated state of the tau Thr231 residue, however, was sensitive to ethanol exposure. We demonstrated that under control conditions, increased DIV was associated with significantly lower phosphorylation state of the tau protein at the Thr231 residue, but this association was eliminated in the ethanol samples. In both Experiments 1 and 2, tau phosphorylation was statistically equivalent at 12 and 24DIV, indicating that ethanol disrupted the age-dependent reduction in tau phosphorylation. In Experiment 2, we demonstrated that this effect is specific to the CA1 and DG, not the CA3. In the CA3, ethanol had no effect on age-dependent reductions in tau phosphorylation. No differences based on sex were observed. Overall, our hypothesis regarding the temporally sustained nature of tau phosphorylation in the ethanol-treated slices was supported. It is important to note that Tan, Abel, and Berman (1993) found no differences in tau phosphorylation between ethanol and control groups across PND28 in Long Evans offspring. However, several methodological differences may account for divergent data reported here, primarily the analysis procedure, alcohol exposure regimen, differences in tau antibodies, and the use of only male samples.

We did not observe changes in Total Tau expression regardless of ethanol exposure in our samples, as has been reported elsewhere in in vitro human neuroblastoma cells (Gendron et al., 2008), mouse Alzheimer’s models (Hoffman et al., 2019), and human alcohol-related Wernicke’s encephalopathy reports (Matsushita et al., 2008). Similar findings in a comparable timeline of Total Tau expression absent of ethanol exposure have been previously reported (Yu et al., 2009), but our report likely diverges from the literature due to procedural differences among models investigating Total Tau expression in the context of ethanol exposure. Because Total Tau was unaffected by time or experimental condition, we propose that the changes in phosphorylation presently are under kinase/phosphatase jurisdiction as opposed to increased tau phosphorylation on newly expressed tau proteins.

Evidence suggests that both GSK-3β and PP2A are involved in biochemical responses to ethanol exposure (Liangpunsakul et al., 2008; Luo, 2009). Specifically, ethanol administration in cultured mouse Neuro-2a cells reduces neurite outgrowth by activation of GSK-3β and subsequent phosphorylation of the Ser396 residue on tau proteins (Chen et al., 2009). While Chen et al. (2009) did not investigate p-Thr231, it is possible that the phosphorylation of the Thr231 residue was also elevated in this experiment, considering that both Thr231 and Ser396 are targets of increased GSK-3β activity (Martin et al., 2013). In contrast, the fate of tau dephosphorylation by PP2A in response to ethanol is not as clear. However, it appears that the inhibition of PP2A may be as or more important than tau kinases in sustaining tau phosphorylation. For example, in mice with starvation-induced tau hyperphosphorylation, the sustained nature of tau phosphorylation in the hippocampus was unrelated to GSK-3β activity, suggesting that tau dephosphorylation by PP2A was dysfunctional, resulting in sustained tau phosphorylation under these experimental conditions (Planel, Yasutake, Fujita, & Ishiguro, 2001). While it is likely that ethanol implicates GSK-3β and PP2A in sustained p-Tau, additional research is needed to clarify whether GSK-3β and PP2A are directly influenced by ethanol or whether the relationship derives from compounding allosteric influence (e.g., neuroinflammation, insulin signaling, and/or mTORC) resulting from ethanol exposure in this model. Careful delineation would bolster the tauopathic characterization of this model.

Reduction of tau phosphorylation in the CA3 region of the cultured hippocampus was robust to ethanol exposure. The CA3 region is implicated in cases of traditional tauopathies in humans such as Alzheimer’s (Blazquez-Llorca, Garcia-Marin, Merino-Serrais, Ávila, & DeFelipe, 2011) and dementia (Lace et al., 2009), as well as different rodent models of Alzheimer’s (Härtig et al., 2005; Kazim et al., 2017; Torres, Jara, Olesen, & Tapia-Rojas, 2021). These findings, however, accumulate from models of neurodegenerative tauopathies. While we anticipate that FASDs will have overlapping mechanisms with neurodegenerative tauopathies, we also assume that there will be distinctions, perhaps in the hippocampal CA3, between FASDs and traditional neurodegenerative tauopathies.

There are limitations to this study. First, DIV is a categorical variable, indicating distinct increments of measurement: 12DIV and 24DIV. Given the nature of IHC and western blot techniques, tissue collected and measured at 12DIV cannot be measured again at 24DIV. Unfortunately, this is an artifact of the design, prohibiting a within-sample repeated measure of Total or p-Tau across DIV. Second, some benefits of the OHSC model are also its limitations: the tissue is divorced from the rest of the CNS and peripheral nervous system (PNS), which are elements in conjunction with environmental factors that influence structural and functional hippocampal development (Baroncelli et al., 2010; Mohammed et al., 2002; Van Praag, Kempermann, & Gage, 2000). Because the animals included in this study were sacrificed at PND10, very little can be inferred about the effect of environmental factors on the hippocampus at the time of sacrifice. Finally, while our findings demonstrate that tau protein phosphorylation remains elevated in the hippocampus following ethanol exposure, these findings do not indicate that sustained tau phosphorylation confers any behavioral or cognitive abnormalities observed in rodent models of FASDs or human patients with FASDs. The diagnostic criteria for FASDs almost solely rely on clinical observations of cognitive and behavioral divergences in children prenatally exposed to ethanol (Mattson, Crocker, & Nguyen, 2011). Therefore, connecting p-Tau to cognitive and behavioral outcomes is necessary to better characterize FASDs as a neurodevelopmental tauopathy in order to proceed with any meaningful pharmacotherapeutic pursuits.

Significance and conclusions

Phosphorylated tau protein plays an important role in neurodegenerative diseases such as Alzheimer’s, but its role in neurodevelopmental disorders such as FASDs has received little experimental or clinical attention. This is despite the similarities in cognitive disruptions seen between neurodevelopmental and neurodegenerative disorders. We began addressing this gap in research by demonstrating that, under control conditions, the phosphorylated state of tau protein was diminished in the hippocampus out to 24DIV; however, in hippocampal tissue that was exposed to ethanol, tau phosphorylation was sustained. Total Tau protein expression was unchanged between 12 and 24DIV, confirming that the sustained tau phosphorylation occurred independently of Total Tau protein expression following ethanol exposure. Together, this study suggests that hippocampal p-Tau is sensitive to perinatal ethanol exposure and may provide a unique tauopathic lens for investigating potential pharmacotherapeutic targets in the treatment of FASDs.