Effects of ethanol exposure in neonatal mice on retinoic acid signaling in forebrain neurons and astrocytes Authors: Mariko Saito, Jungann Park, Anusha Nalluri, Brandon Marino, Colin RO Williams, Donald A Wilson, Bhaskar C Das, John F Smiley Published: 2026 ABSTRACT AbstractToxicity of prenatal ethanol leading to fetal alcohol spectrum disorders (FASDs) has been linked to disturbances in retinoic acid (RA) signaling necessary for embryonic development. While ethanol exposure in the postnatal day 7 (P7) mice, which induces immediate neurodegeneration and long-lasting GABAergic cell loss and behavioral deficits, has been used for the third trimester FASD model, involvement of RA signaling in the process has not been well explored. Using RARE-LacZ reporter mice that express β-galactosidase (β-Gal) under the control of retinoic acid response element (RARE), we examined RA signaling activity of the forebrains of P8 and P30 mice with or without P7 ethanol treatment. In all experimental groups, β-Gal was expressed mainly in the hippocampus with the strongest expression in the granule cell layer of dentate gyrus. In addition, β-Gal was expressed in pyramidal neurons and parvalbumin (PV) neurons in CA1–3 pyramidal layer and in astrocytes scattered around the CA1–3 region although PV neurons were only examined at P30 because of the low PV expression at P8. β-Gal was also expressed in the anteroventral/anteromedial (AV/AM) thalamus and the retrosplenial (Rs) and Tbr1-positive (+) layer 6 cortices. β-Gal-expressing PV neurons were also found in the cortex such as Rs, while β-Gal was barely detected in somatostatin neurons in any brain regions examined. Such region and cell specific β-Gal expression was significantly higher in P8 brains than P30 brains in various brain regions. P7 ethanol reduced β-Gal expression in the CA1–3 pyramidal layer, Tbr1 + cortical layer 6, and the AV/AM thalamus at P8 or P30 or both. Although P7 ethanol decreased PV cells in CA2–3 pyramidal layers as reported, it decreased β-Gal+ PV cells more drastically. The active RA signaling found in PV neurons and the effects of P7 ethanol on the signaling suggest that reduced RA signaling by P7 ethanol may disturb PV cell maturation and enhance long-lasting brain abnormalities.Keywords: Neonatal ethanol, Retinoic acid signaling, RARE-LacZ reporter mice, PV neurons, astrocytes, FASD Introduction Prenatal alcohol exposure perturbs fetal development and can lead to FASDs. Specifically, FASD neuropathology, such as neuronal loss and volume reduction in the corpus callosum, is associated with cognitive and behavioral abnormalities (Riley and McGee, 2005, Astley et al., 2009, May et al., 2009, Norman et al., 2009, Riley et al., 2011, May et al., 2014). The incidence of FASD is approximately 2–5 in 100 births in Europe and North America, and FASD is the leading cause of intellectual disability in the United States (Riley and McGee, 2005, May et al., 2009, May et al., 2014, May et al., 2018). While various factors can be involved in the pathogenesis of FASD, RA signaling has been considered important, based on clinical and preclinical studies (Duester, 1991, Johnson et al., 2007, Sapin, 2008, Kane et al., 2010, Young et al., 2014, Fainsod et al., 2020). RA, the active form of vitamin A, regulates transcription of RA-responsive genes via nuclear RA receptor (RAR)-α,β,ɣ and retinoid X receptor (RXR)-α,β,ɣ activation and plays essential roles in development of embryos and their CNS as well as in adult tissue homeostasis (Mey and Mccaffery, 2004, Niederreither and Dollé, 2008, Crandall et al., 2011, Rhinn and Dollé, 2012, Shearer et al., 2012, Haushalter et al., 2017, Dowling, 2020). The content of RA in each brain region seems precisely controlled by expression patterns of synthesizing/metabolizing enzymes (Niederreither and Dollé, 2008, Rhinn and Dollé, 2012), and insufficient or excess RA causes teratogenic and/or toxic effects, which mimic the ethanol effects on the developing brain (Johnson et al., 2007, Fainsod et al., 2020, Wolf, 2010). It has been shown that ethanol reduces RA levels due to the competition between ethanol clearance and RA biosynthesis (Duester, 1991, Fainsod et al., 2020), and RA signaling reduction induces deficits similar to FASD in both patients and animal models, and addition of RA or RA synthesizing enzymes restores the deficits in adult or embryonic brains (Johnson et al., 2007, Yelin et al., 2005, Marrs et al., 2010, Shabtai et al., 2018). Ethanol exposure in neonatal rodents induces immediate neurodegeneration and long-lasting neuroanatomical and behavioral deficits, providing a third trimester binge drinking FASD model. Specifically, ethanol exposure in P7 mice induces robust neurodegeneration within 24 h in various brain regions such as the cingulate (Cg) and Rs cortex and the thalamus (Olney et al., 2002a, Saito et al., 2007). The acute neurodegeneration is followed by activation of microglia and astrocytes and long-lasting GABAergic cell deficits (Coleman et al., 2012, Saito et al., 2015, Saito et al., 2019, Smiley et al., 2015, Smiley et al., 2019, Bird et al., 2018, Bird et al., 2020). The deficits observed in GABAergic neurons such as PV positive (+) and somatostatin (SST)+ cells are more prominent than the total neuron deficits and may cause neurobehavioral abnormalities (Smiley et al., 2015, Sadrian et al., 2012, Sadrian et al., 2014, Wilson et al., 2016, Lewin et al., 2018). However, mechanisms of this relatively specific long-lasting GABAergic cell loss are unclear. As described above, disturbed RA signaling appears to be involved in ethanol toxicity in the developing brain (Duester, 1991, Johnson et al., 2007, Sapin, 2008, Kane et al., 2010, Young et al., 2014, Fainsod et al., 2020, Yelin et al., 2005, Marrs et al., 2010, Shabtai et al., 2018, Alfos et al., 2001, McCaffery et al., 2004), and our previous studies indicated that RARα modulation affected P7 ethanol-induced neurodegeneration and neuroinflammation (Saito et al., 2023). In addition, it has been reported that disturbed RA signaling influences PV cell maturation in early postnatal mouse prefrontal cortex (Larsen et al., 2019). These studies suggest the involvement of RA signaling in P7 ethanol-induced neurodegeneration/inflammation. However, RA signaling activity in the neonatal brain and the effects of ethanol on the activity have been an understudied subject. Here, using RARE-LacZ reporter mice that express β-Gal (LacZ) under the control of RARE (Rossant et al., 1991), we examined RA signaling activity in the P8 and P30 forebrains and the effects of P7 ethanol on the activity to test our hypothesis that P7 ethanol induces aberrant RA signaling in forebrains, which may be associated with P7 ethanol-induced brain abnormalities. We examined the regional/cellular RA signaling activity measured by β-Gal expression in the control P8 and P30 RARE-LacZ mouse forebrains and then examined the effects of P7 ethanol on the β-Gal expression. Materials and methods Animals RARE-LacZ reporter mice [STOCK Tg(RARE-Hspa1b/lacZ)12Jrt/J (RARE-hsp68LacZ), JAX:008477] and C57BL/6 J [JAX: 000664] mice were obtained from Jackson laboratory (Bar Harbor, ME) and bred at the Nathan Kline Institute animal facility in a temperature controlled room (22°C ± 1°C) under the dark/light cycle (12/12 h) with ad lib food (rodent laboratory chow 5001, LabDiet, St. Louis, MO) and water at all times. Mice were kept in plastic cages (19x29x13 cm) with nestlet (Ancare, Bellmore, NY). RARE-LacZ reporter mice and C57BL/6 J mice were maintained as homozygotes. However, genotyping of RARE-LacZ mice was occasionally carried out using a PCR/gel electrophoreses protocol recommended by Jackson laboratory. All procedures were approved by the Nathan Kline Institute IACUC and were in accordance with NIH guidelines for the proper treatment of animals. Treatments of RARE-LacZ reporter mice with ethanol P7 mouse pups (both males and females) were injected subcutaneously with saline or ethanol (2.5 g/kg) twice at a 2 h interval as previously described (Olney et al., 2002a, Saito et al., 2007). Pups were returned to their home cage immediately following injections. Pups were weaned at P28 into group cages of littermates. Same-sex mice were housed together in cages in numbers between two and four per cage. At P8 (24 h after the first ethanol injection) or P30, mice were perfusion-fixed for immunofluorescence histochemistry. Since significant sex differences in the effects of P7 ethanol have not been previously observed (Saito et al., 2019, Wilson et al., 2016), nor were any significant differences observed between sexes here, the data from males and females were combined. Immunofluorescence and cell counting In each experiment, 4–8 mice per group derived from 4 different litters were used. Mice were anesthetized by intraperitoneal administration of ketamine/xylazine, perfused with a solution containing 4% paraformaldehyde and 4% sucrose in cacodylate buffer (pH 7.2), and the heads were removed and further fixed in the perfusion solution overnight. Then brains were removed, transferred to phosphate buffered saline (PBS) solution, and kept at 4°C for 2–5 days until cut with a vibratome into 50μm thick coronal sections. For immunofluorescence, the free-floating sections were rinsed in PBS, permeabilized in PBS containing 0.3% Triton X−100 and 1% BSA for 5 min, and incubated for 30 min in blocking solution (PBS containing 5% BSA and 0.1% Triton X−100), followed by incubation overnight at 4°C with antibodies against parvalbumin (PV) (mouse, PV235, Swant AG, Kehrsatz, Switzerland), somatostatin (SST) (rabbit polyclonal antibody AP33464SU-N, OriGene, Rockville, MD), GFAP (D1F4Q) (Rabbit mAb #12389, Cell Signaling, Danvers, MA), GFAP (GA5) (mouse mAb #3670, Cell Signaling), CB (calbindin)38a (rabbit, Swant), S100B (rabbit, AB52642, Abcam, Waltham, MA), NeuN (mouse, #MAB377, Sigma, St. Louis, MO), TBR1(rabbit mAb, #49661, Cell Signaling) or ALDH1L1 (E712Q) (rabbit mAb, #85828, Cell Signaling) in PBS containing 3% BSA and 0.1% Triton X−100. Then, brain sections were rinsed in 0.1% Triton X−100 in PBS three times and incubated with (E.coli) β-galactosidase (chicken, #ab9361, Abcam) antibody for 2 h at r.t., followed by incubation with Alexa Fluoro488 goat anti-rabbit (or mouse) IgG (Life Technologies, Grand Island, NY) and Alexa Fluoro594 goat anti-chicken IgY (Life Technologies) in 0.1% Triton X−100 in PBS containing 1% BSA for 1 h at r.t. Sections were finally rinsed in PBS three times, mounted, and coverslipped using ProLong Gold Antifade Reagent (Life Technologies). All photomicrographs were taken through a 4X, 10X, or 20X objective with a Nikon Eclipse TE2000 inverted microscope attached to a Basler acA5472–17um camera. The PV, SST, GFAP, and β-Gal+ cell number of each area of interest (AOI), total dimensions of each AOI, mean fluorescence intensities of each AOI, and overlaid image analyses were done using ImageJ 1.53o. The AOIs were defined according to the Atlas of mouse brain (Paxinos and Franklin, 2004) or Atlas of the Developing Mouse Brain at E17.5, P0, and P6 (Paxinos et al., 2007). The cell density (calculated as the mean cell number per square millimeter) or mean fluorescence intensities of each AOI were obtained from 4–8 mice (derived from 4 different litters) using 4–6 sections around bregma −1.34 to −2.06 mm for total hippocampus, granule cell layer of dentate gyrus, CA1–3 pyramidal layers, cortical Tbr1 +layer 6, and Rs cortex, and around bregma −0.94 to −1.22 for AV/AM thalamus according to Atlas of mouse brain (Paxinos and Franklin, 2004). However, these methods are considered semiquantitative, which is a limitation of our current studies. Antibody validation PV235 (Swant) validated by knockout mice has been used in many published studies. Immunofluorescence staining by somatostatin (SST) antibody (AP33464SU-N, OriGene) showed brain regional distribution expected for SST neurons. Also, our studies using Nkx2.1-Cre;Ai9 mice, in which major GABAergic neurons (including SST neurons) and their progenitors in the cortex are labeled with constitutively expressed tdTomato fluorescent protein, indicated that SST neurons identified by this antibody were primarily co-localized with tdTomato+ cells in the cortex (Saito et al., 2023). GFAP(GA5) and GFAP (D1F4Q) antibodies (Cell Signaling) have been widely used (1235 and 324 citations, respectively) and orthogonal validation was done for GFAP (GA5). Calbindin (CB) 38a (Swant) antibody is validated using calbindin knockout mice. S100B (AB52642, Abcam) antibody is considered validated with 745 citations and showed astrocyte morphology in immunohistochemistry. NeuN antibody (mouse, MAB377, Sigma) has been highly cited (7632) and validated. TBR1(D6C6X) (Cell Signaling) antibody recognizes endogenous levels of total TBR1 protein specifically although low levels of nuclear staining of unknown specificity have been observed in mouse small intestine, spleen, and pancreas (Cell Signaling). Our immunostaining of the cortex using this antibody showed expected brain regional distribution of TBR1 + neurons. ALDH1L1 (E712Q) (Cell Signaling) has been recognized as an astrocyte marker. Our immunofluorescence indicated that Aldh1l1 + cells showed astrocyte morphology and those cells were dual-labeled with antibodies against other astrocyte markers, GFAP and S100B. As shown in Fig. S1, β-galactosidase (chicken, ab9361, Abcam) expression was observed in RARE-lacZ reporter mice, but not in C57BL/6 J mice. Blood alcohol level measurement At 0.5, 2, and 6 h following the second ethanol injection, the truncal blood was collected in heparinized tubes and centrifuged at 2700xg for 20 min. Ethanol concentrations in the plasma were measured using Alcohol Reagent Set (Pointe Scientific, Canton, MI). Two µL of the plasma diluted 1:10 with saline, or 2 µL of standard alcohol (Pointe Scientific) with concentrations of 100, 50, 25, or 12.5 mg/dL were mixed with 200 µL of alcohol reagent in each well of microtiter plates, incubated at 30°C for 10 min, and the optical density was measured at 340 nm with a microplate reader (Molecular Devices, Sunnyvale, CA). Statistics Statistical analyses were done by independent-samples T-test or 2 way-ANOVA (2 ×2) after confirming normality of each data set using a Shapiro-Wilk test by SPSS statistics software (version 24). For simple main effects analyses in 2 × 2 ANOVA, Bonferroni multiple comparison was used. For all analyses, p < 0.05 was considered statistically significant. Values were expressed as mean ± SD obtained from 4 to 8 animals per group. The animal number used in each experiment was described in figure legends. Those animals were derived from 4 litters per group, and there were no more than 2 animals derived from the same litter in each group to minimize litter effects. While both sexes were combined for data analyses as described above (Method 2.2), males and females were equally distributed in each group. Results β-Gal expression in the forebrain of P8 and P30 mice First, we examined β-Gal+ (RA signaling active) regions in the forebrains of control mice at P8 and P30. P8 mice were chosen because P7 ethanol-induced apoptotic neurodegeneration is the maximum at P8, and P30 (adolescent) mice were chosen because long-lasting effects of P7 ethanol, especially GABA neuron loss has been described in P30 mice (Saito et al., 2023). Fig. 1A and B show representative images of the control brain sections labeled using β-Gal antibody. In the P8 brain, β-Gal expression was the strongest in the dentate gyrus (DG) granule cell layer (GrDG). In addition, β-Gal expression was observed in the pyramidal cell layer (Py) of the dorsal hippocampus and the Rs cortex near the cingulum (cg). Also, β-Gal was expressed in AV/AM thalamus (Fig. 1B). β-Gal+ astrocyte-like cells (some of which are indicated by arrows in Fig. 1A) were also found in molecular dentate gyrus (Mol), lacunosum moleculare (LMol), stratum radiatum (Rad) and oriens layer (Or) of the dorsal hippocampus. In the P30 brain, while β-Gal expression was also the strongest in the GrDG, expression levels of other regions, such as CA1-CA3 pyramidal cell layer (CA1–3Py), AV/AM thalamus, and Rs cortex were lower compared to those of the P8 brain, which was consistent with the notion that RA-signaling activity decreases as the brain matures and the activity becomes restricted to specific regions such as the DG (Luo et al., 2004). Quantitative comparisons between P8 and P30 samples were made by measuring the mean fluorescence intensity in each area of interest (AOI) of GrDG, CA1–3Py, and AV/AM (Fig. 1C). The results indicated that differences were not significant in the GrDG, but significant in CA1–3Py and AV/AM thalamus regions. β-Gal+ cells in the cortex were concentrated in Rs regions near cg in both P8 and P30 brains. However, while β-Gal+ cells were mainly found in layer 6 in the P8 cortex, β-Gal+ cells in the P30 cortex were spread more widely in the cortex (Fig. 1A). Thus, the quantitative comparison between P8 and P30 in these regions was difficult and omitted. As described in Materials and methods, we combined males and females for analyses because significant sex differences in the effects of P7 ethanol have not been previously observed (Saito et al., 2019, Wilson et al., 2016). Fig. 1D shows that there are also no significant differences in the mean fluorescence densities between control P8 males and females in GrDG, CA1–3Py, and AV/AM thalamus regions. A: RARE-LacZ mice (control) were perfusion-fixed at P8 or P30, and forebrain coronal sections were fluorescent-labeled using anti-β-Gal (LacZ) and Alexa594 anti-chicken antibody. The representative images of dorsal parts of hippocampus and cortex show strong β-Gal expression in the hippocampus, especially in GrDG. Scale bar: 500 µm. Mol: molecular dentate gyrus, LMol: lacunosum moleculare, Rad: stratum radiatum, Or: oriens layer, Rs: retrosplenial cortex, cg: cingulum, AV/AM: anteroventral/anteromedial thalamus. Arrows indicate some of the astrocyte-like cells. B: Representative images including AV/AM thalamus were obtained as described in A. Scale bar: 500 µm. C: β-Gal expression levels in GrDG, CA1−3Py, and AV/AM thalamus were compared between P8 and P30 brains by measuring the mean fluorescence intensities of the regions. Data shown as mean ± SD were obtained from 8 P8 mice and 5 (for GrDG and AV/AM) or 6 (for CA1−3Py) P30 mice from 4 litters each and tested by independent-samples T-test: GrDG [t(11) = 1.555, p = .148], CA1−3Py [t(12) = 7.247, *p < .0005] and AV/AM [t(11) = 4.964, *p < .0005]. Males and females were equally distributed in each group. D: The mean fluorescence intensities of P8 control male and female mice (shown as mean ± SD) gave no significant differences in GrDG, CA1−3Py, and AV/AM thalamus regions when analyzed by independent-samples T-test: GrDG [t(6) = .104, p = .920], CA1−3Py [t(6) = .660, p = .534] and AV/AM [t(6) = .255, p = .807]. Four male and 4 female mice from 4 litters were used. β-Gal antibody did not react with any brain regions of wild-type C57BL/6 J mice. Fig. S1 shows representative images of brain sections from P8 RARE-LacZ and C57BL/6 J mice fluorescent-labeled using β-Gal antibody and NeuN antibody, indicating the specificity of the E. coli β-Gal (LacZ) antibody used here. Cell types showing RA signaling activity DG granule neurons Brain sections of saline-treated P8 and P30 RARE-lacZ mice were immunofluorescent-labeled using calbindin-D28k (CB) antibody as well as β-Gal antibody. CB is known to be expressed in mature DG granule neurons in addition to various other neuronal populations in the rodent brain (Baimbridge and Miller, 1982, Celio et al., 1990). As shown in Fig. 2, many CB+ neurons in GrDG were β-Gal+ in both P8 and P30 brains, which is consistent with a study showing that about 40% of CB+ GrDG are β-Gal positive in adult RARE-LacZ mice (Yeo et al., 2024). Except for CB+ mature DG granule cells and a few cells in the pyramidal layer of hippocampus, CB+ cells found widely in other brain regions such as the cortex were β-Gal negative in both P8 and P30 brains (Fig. 2). P8 (upper panels) or P30 (lower panels) RARE-LacZ mice (control) were perfusion-fixed, and brain sections were dual-labeled by anti-β-Gal (LacZ) antibody (red) and anti-calbindin (CB) antibody (green). Right panels showing DG regions are enlarged images of the left panels which present dorsal hippocampus and cerebral cortex. Many CB+ mature GrDG expressed β-Gal (yellow) in both P8 and P30. At P8, substantial amounts of β-Gal+CB- cells were found in the immature hilus/SGZ regions. These results were reproduced in images obtained from 2 other sets of RARE-LacZ mice. There is a thin layer of β-Gal-positive, but CB-negative (β-Gal+CB-) cells in the subgranular zone (SGZ) of P30 brain (Fig. 2). These cells may be neural stem and progenitor cells (NSPCs) and immature granule neurons since it has been indicated that RA is present in the hippocampal NSPCs with a role in developmental and adult neurogenesis (Maden, 2007, Jacobs et al., 2006, Mishra et al., 2018). In contrast, the layer of β-Gal+CB- cells in the immature SGZ/hilus region of P8 brain were thicker than that of P30. These cells may also consist of NSPCs and immature granule neurons because many GCs are generated at this developmental period (Bayer and Altman, 1974, Nicola et al., 2015, Berg et al., 2019, Radic et al., 2017). Further studies are necessary to identify these β-Gal+CB- cell types. PV neurons in the hippocampus and the cortex As shown in Fig. 3A, a significant portion of PV+ cells in the pyramidal layer of the dorsal hippocampus were β-Gal+ at P30. The right panel shows an enlarged image of the white square region in the left panel. Images of P8 brains were not shown because PV neurons were still immature and hardly expressed PV at P8 as described previously (Saito et al., 2019, Itami et al., 2007). Some PV neurons in the P30 Rs cortex also expressed β-Gal (Fig. 3B). However, somatostatin (SST)+ neurons, another major GABAergic neuron type in the cortex and hippocampus, rarely expressed β-Gal both at P8 and P30 (Fig. S2). Also, as stated in the previous section, CB neurons in the cortex which are mainly GABAergic interneurons (Baimbridge and Miller, 1982, Celio et al., 1990) were β-Gal-negative (Fig. 2). These results suggest specific RA signaling activity in PV neurons among GABA neurons in the P30 cortex and hippocampus. A: The brain sections of P30 RARE-LacZ mice were dual-labeled by anti-β-Gal (red) and anti-PV (green) antibody. The representative image shows dorsal hippocampal CA1−3 region. The right panel is an enlarged image of the region indicated by a white square in the left panel. The image indicates that some PV neurons in the CA2 +CA3 pyramidal layer were β-Gal+ cells. These results were reproduced in images obtained from other RARE-LacZ mice. Quantitative results about the presence of PV+β-Gal+ cells in the CA2 +CA3 pyramidal layer were also shown in Fig. 7B. B: The representative image shows the Rs cortex area. Some PV neurons in Rs were β-Gal+ cells. These results were reproduced in images obtained from other RARE-LacZ mice. Tbr1 + layer 6 neurons As shown in Fig. 1 and Fig. 4, a cortex region including a part of Rs and cortical layer 6 (traced by a dotted line in Fig. 4) expressed β-Gal especially at P8. These cells overlapped with Tbr1 +layer 6 neurons both in P8 and P30 brains. Brain sections of P8 and P30 RARE-LacZ mice were dual-labeled with anti-β-Gal (red) and anti-Tbr1 (Green) antibody. The images showed that the cortex region including a part of Rs and cortical layer 6 (traced by white dashed line) expressed β-Gal highly especially at P8. These β-Gal + cells were overlayed with layer 6 Tbr1 + neurons. These results were reproduced in images obtained from 2 other sets of RARE-LacZ mice. Scale bars in the left panels are 500 µm and the furthest right panels are 200 µm. Astrocytes with β-Gal expression Astrocyte-like cells expressing β-Gal in Mol, LMol, Rad, and Or of the hippocampus were generally GFAP-positive in P30 mice (Fig. 5A) These cells were also positive for S100B (Fig. 5B) and Aldh1l1 (Fig. 5C), which are considered astrocyte markers, indicating that these cells are indeed astrocytes. The β-Gal+GFAP+ astrocytes were also present in the P8 hippocampus (Fig. 5D). At P8, the morphology of these astrocytes appeared to be less complex and showed smaller domain sizes compared to adult astrocytes as reported (Felix et al., 2021). It has been shown that astrogliosis, which starts during the early postnatal period, persists in adult DG (Schneider et al., 2022). The origin and the characteristics of β-Gal+ astrocytes remain to be elucidated. A: Brain sections of P30 RARE-LacZ mice were dual-labeled by anti-β-Gal antibody (red) and anti-GFAP antibody (green). The representative image shows hippocampus area. The right panel shows an enlarged image of the region indicated by the white square in the left panel. B,C: β-Gal+ astrocyte-like cells also expressed S100B (B) and Aldh1l1(C), both considered astrocyte markers. D: β-Gal+ astrocyte-like cells were also in P8 mouse hippocampus. Scale Bars in B, C, and D: 100 µm. Red arrows indicate S100B+β-Gal+ astrocytes (B) and Aldh1l1 + β-Gal+ astrocytes (C). The quantification of the density of β-Gal+ astrocytes in the P8 and P30 hippocampus with saline or ethanol treatment is shown in Fig. 8. The effects of P7 ethanol on β-Gal expression RARE-LacZ mice were treated with saline or ethanol at P7 as described in Materials and methods. Our previous studies showed that this P7 ethanol treatment in C57BL/6By mice induced a peak blood alcohol level of 0.45 g/dL when truncal blood was collected at 0.5, 1, 3, and 6 h following the second ethanol injection (Saito et al., 2007). Similarly, the same ethanol treatment in P7 RARE-LacZ mice induced blood alcohol levels of 0.43 ± 0.05 (mean±SD, n = 3), 0.37 ± 0.02 (n = 3), and 0.32 ± 0.05 (n = 3) g/dL when truncal blood was collected 0.5, 2, and 6 h following the second ethanol injection and analyzed as described in Materials and methods. RARE-LacZ mice treated with saline or ethanol were perfusion-fixed at P8 or P30, and brain sections were fluorescent-labeled using β-Gal antibody. Fig. 6A shows images of brain sections of dorsal hippocampal and the cortex area. β-Gal expression intensity was attenuated in some regions of ethanol-treated samples in both P8 and P30 brain. When the mean intensities of β-Gal in the GrDG were analyzed by 2-way ANOVA, there was no significant main effect of treatment (saline/ethanol), while there was a significant main effect of time (P8 and P30) although simple main effects of time did not reach the significant level (Fig. 6B) as also indicated in Fig. 1C. However, β-Gal expression intensities in AV/AM thalamus and cortical layer 6 (the area indicated by dotted line in Fig. 4) were significantly reduced by P7 ethanol at P8 (Fig. 6C). Since β-Gal+ cell expression in AV/AM and cortical layer 6 regions of P30 ethanol samples were very low, the measurement was omitted. A: β-Gal expression of brain sections of P8 and P30 RARE-LacZ mice treated with saline or ethanol (EtOH) at P7. Scale bar: 500 µm. B: Mean intensities of fluorescence in the GrDG of P8 and P30 mice treated with saline or ethanol were presented as mean ± SD. A two-way ANOVA to examine the effects of time (P8/P30) and treatment (saline/EtOH) indicated main effects of time, F(1,21)= 4.6, p = .043, although simple main effects of time did not reach the significant level (p = .228 in saline and p = .087 in EtOH group). However, there is no significant main effects of treatment, F(1,21)= .628, p = .437. We used 8 saline and 7 ethanol-treated P8 mice and 5 saline and 5 ethanol-treated P30 mice derived from 4 litters in each group. C: Mean intensities of fluorescence in the AV/AM thalamus and cortical layer 6 were compared between P8 mice treated with saline and ethanol. Independent-samples T-test analyses showed [t(7) = 3.069, *p = .018] for AV/AM and [t(12) = 5.972, **p < .0005] for layer 6. We used 5 saline and 4 ethanol-treated P8 mice for AV/AM and 7 saline and 7 ethanol-treated P8 mice for layer 6 derived form 4 litters of saline-treated and 4 litters of ethanol treated groups. Because P7 ethanol induces cell death in many neuronal populations including thalamic and Rs/Cg cortical neurons (Olney et al., 2002a, Saito et al., 2007), reduction in the β-Gal intensity by P7 ethanol may be due to the neuronal cell loss. Therefore, we examined densities of β-Gal+ cells, PV+ cells, and PV+β-Gal+ cells in the CA2/CA3 Py, because some of the β-Gal+ cells in the CA2/CA3 Py region were PV+ (Fig. 3), and P7 ethanol induces cell death of immature PV neurons as reported previously (Saito et al., 2019, Smiley et al., 2015, Smiley et al., 2019, Sadrian et al., 2014). Fig. 7A shows that densities of β-Gal+ cells in the CA2/CA3 Py were significantly reduced in P7 ethanol-treated mice both at P8 and P30. Fig. 7B indicated that both total PV cell density and PV+β-Gal+ cell density were reduced by P7 ethanol. However, PV+β-Gal cell density was more affected by ethanol, and while 55% of PV neurons were β-Gal+ in the control group, only 22% of PV neurons were β-Gal+ in the ethanol group (Fig. 7C). It suggests that P7 ethanol affects β-Gal expression (RA signaling activity) of surviving immature PV neurons, although it is also possible that immature PV neurons with active RA signaling are more vulnerable to P7 ethanol treatment. A: Brain sections of P8 and P30 RARE-LacZ mice treated with saline or ethanol at P7 were immunolabeled with anti-β-Gal antibody. β-Gal+ cell densities were measured in the CA2/CA3 pyramidal layer and presented as mean ± SD. Two-way ANOVA indicated that there are significant main effects of time (P8 and P30) [F(1,15)= 27.234, p < .0005] as well as treatment (saline and ethanol) [F(1,15)+32.761, p < .0005]. Simple main effects of time (analyzed by Bonferroni multiple comparisons) are significant in both treatment groups (both p = .002) and simple main effects of treatment are significant in both P8 and P30 samples (both *p = .001). We used 4 saline and 6 ethanol-treated P8 mice and 5 saline and 4 ethanol-treated P30 mice derived from 4 litters per group. B: Brain sections of P30 RARE-LacZ mice treated with saline or ethanol at P7 were immunolabeled with anti-β-Gal and anti-PV antibody. The cell densities of PV+ cells and PV+β-Gal+ cells were measured in the CA2/CA3 Py. Two-way ANOVA indicated that there are significant main effects of treatment (saline and ethanol) [F(1,12)= 37.121, p < .0005] as well as cell types (PV+ and PV+β-Gal+) [F(1,12)= 124.0, p < .0005]. Simple main effects of treatment are significant in both cell types (PV+ and PV+ β-Gal+) (both *p = .001) and simple main effects of cell types are significant in both saline and ethanol samples (both *p < .0005). We used 4 saline and 6 ethanol-treated P8 mice and 5 saline and 4 ethanol-treated P30 mice from 4 litters per group. C: The ratio (%) of (PV+β-Gal+ cell number)/(total PV+ cell number) was compared in CA2/CA3 pyramidal layer regions between saline and EtOH-treated P30 mice. Independent-samples T-test analyses showed that [t(7) = 7.228, **p < .0005], indicating a significance difference. Five saline-treated and 4 ethanol-treated P30 mice were used from 4 litters per group. As shown in Fig. 5, some astrocytes in the hippocampus which were recognized by GFAP, S100B, and Aldh1l1 antibodies were β-Gal positive. We examined the effects of P7 ethanol on the densities of these β-Gal+ astrocytes in the hippocampus at P8 and P30 (Fig. 8). Although 2-way ANOVA indicated significant interaction between time (P8 and P30) and treatment (saline and ethanol), there were no significant simple main effects of ethanol treatment either at P8 or P30. We have shown previously that P7 ethanol induces reactive astrogliosis identified by higher levels of GFAP and phospho-NFκB at P30 and P60 in the hippocampus (Saito et al., 2023). However, in the present study, the density of β-Gal+ astrocytes in the hippocampus did not increase by P7 ethanol at P30, suggesting that β-Gal+ astrocytes may not be P7 ethanol-induced reactive astrocytes observed in our previous study (Saito et al., 2023). Densities of β-Gal+ astrocytes in the hippocampus were compared between saline and ethanol groups at P8 and P30. Two-way ANOVA indicated that there was significant interaction between time (P8 and P30) and treatment (saline and EtOH) [F(1,21)= 6.492, p = 0.019]. Analyses of simple main effects indicated that there were significant differences in astrocyte densities between ethanol treated P8 and P30 samples (p = 0.006). However, there were no significant simple main effects of treatment (p = .108 at P8 and p = .069 at P30). Seven saline-treated and 7 ethanol-treated P8 mice, and 6 saline-treated and 5 ethanol-treated P30 mice derived from 4 litters per group were used. Discussion It is well recognized that RA signaling is necessary for development of embryos, and the perturbed synthesis of RA by prenatal ethanol would induce defects in development of embryos and their brains, which may be related to FASD pathogenesis. Although RA signaling is also important in the postnatal mouse brain, RA signaling activity in the neonatal and adolescent brains and the effects of neonatal ethanol on the activity have not been well explored. Here, using RARE-LacZ reporter mice, in which ongoing or recent RA signaling can be assessed by E. coli β-Gal (LacZ) protein expression levels (Rossant et al., 1991), we studied canonical RA signaling activity regulated by nuclear receptors in P8 and P30 forebrains and the effects of P7 ethanol on it, although non-canonical posttranscriptional retinoid signaling (Piazza et al., 2024) may also be involved. We found similar distributions of β-Gal expression in both P8 and P30 RARE-LacZ reporter mice although the intensity of β-Gal expression was significantly higher in P8 mice (Fig. 1). It has been well documented that one of the roles of RA in the developing brain is in neuronal differentiation (Niederreither and Dollé, 2008, Rhinn and Dollé, 2012, Haushalter et al., 2017, Maden, 2007). Therefore, it is assumed that RA signaling is more active in the developing P8 brain, where many types of neurons are still differentiating. In P30 brains, β-Gal expression is more restricted to DG, where adult neurogenesis continues and RA is required for NSPC proliferation (Mishra et al., 2018), and RA in adult granule cells is involved in regulating hippocampal neuroplasticity and cognitive functions (Yeo et al., 2024). We observed that β-Gal+ cells in P30 brains were predominantly DG granule cells (Fig. 1) as previously reported (Yeo et al., 2024). Some β-Gal+CB- cells in SGZ may also be NSPCs, progenitor cells, or neuroblasts (Nicola et al., 2015, Berg et al., 2019, Radic et al., 2017). More widely distributed β-Gal+CB- cells found in the developing DG of P8 mice (Fig. 2) may also consist of NSPCs and immature granule neurons because many GCs are generated at this developmental period (Bayer and Altman, 1974, Nicola et al., 2015, Berg et al., 2019, Radic et al., 2017). β-Gal+ cells were also observed in other brain regions in both P8 and P30 brains, including the hippocampal Py cell layer, thalamus (AV/AM), and the Rs cortex (Fig. 1), which agreed with a previous report using 7-week-old RARE-LacZ mice (Luo et al., 2004). At the cellular level, we found β-Gal+ PV neurons, Tbr1 + layer 6 neurons, and astrocytes in the hippocampus and/or cortex in addition to DG granule neurons. Interestingly, while 55% of PV neurons in the CA2–3 Py cell layer were β-Gal+ (Figs. 3A,7C), and PV+β-Gal+ cells were also found in the Rs cortex (Fig. 3B), SST neurons, another major GABA cell type in the cortex and hippocampus, were β-Gal-negative. Also, CB+ neurons, which contain GABA neurons, were β-Gal-negative except for CB+ mature DG granule cells. The rather specific RA signaling in PV cells among major GABA neurons in the cortex and hippocampus may indicate important roles of RA signaling in PV neuron development and functions as previously suggested (Larsen et al., 2019). In this work (Larsen et al., 2019), the authors indicated that RA signaling in neonatal PV neurons in the prefrontal cortex controls PV neuron maturation. Active RA signaling may be found specifically in PV neurons because PV neuron maturation is slower than other GABAergic neurons (Saito et al., 2019, Itami et al., 2007, Moissidis et al., 2025). β-Gal was also expressed in a population of astrocytes in the hippocampus (Fig. 5). These astrocytes were positive for GFAP, S100B, and Aldh1l1, which are considered astrocyte markers. The morphology and the levels of GFAP expression suggest that these cells are not typical reactive astrocytes. It has been shown that astrocyte differentiation persists till around P30, and astrogliosis continues in the DG of the adult brain (Rossant et al., 1991). Therefore, β-Gal+GFAP+ cells may be newly generated astrocytes. However, further studies are necessary to characterize the β-Gal+ astrocytes. In the cortex, in addition to Rs, we found β-Gal+ cells in layer 6 near the cg region (Fig. 1, Fig. 3, Fig. 4). These β-Gal+ cells in the cortical layer 6 were well overlapped with Tbr1 + neurons (Fig. 4). Tbr1 + layer 6 neurons appear to be important in the development of thalamocortical connectivity (Briggs, 2010). Besides hippocampus and cortex, β-Gal expression was also detected in AV/AM thalamus especially at P8 (Fig. 1). The expression is predominant in AV/AM thalamus compared to other thalamic nuclei. RA signaling activity detected in Tbr1 +layer 6 neurons and AV/AM thalamus may be related to a previous study indicating that the thalamocortical connectivity regulates RA signaling and PV neuron development in the prefrontal cortex of neonatal brain (Larsen et al., 2019). Then, we examined the effects of P7 ethanol on β-Gal expression in the RARE-LacZ mice. As shown in Fig. 6, ethanol generally decreased the intensities of β-Gal expression 24 h after P7 ethanol injection. Although mean β-Gal expression (measured by mean fluorescence intensities in AOI) of DG granule neurons did not significantly change after P7 ethanol, β-Gal expression in the AV/AM thalamus and layer 6 near cg region significantly decreased (Fig. 6). However, our previous studies demonstrated that P7 ethanol induced severe apoptosis with loss of 30–60% of neurons in the anterior thalamic nuclei (Smiley et al., 2023), which may cause the reduction in β-Gal expression. Therefore, we compared the effects of P7 ethanol on PV neurons and β-Gal+ PV neurons in P30 CA2–3 Py. Results indicated that PV+β-Gal+ cells were more severely affected by ethanol (Fig. 7). It suggests that P7 ethanol not only induced cell death but also affected β-Gal expression of PV neurons. How P7 ethanol decreases RA signaling in PV neurons is presently unknown. Immature PV neurons with the high RA signaling activity at P7 may be more vulnerable to P7 ethanol-induced apoptosis. It is also possible that P7 ethanol affects β-Gal expression (RA signaling activity) of surviving immature PV cells by reducing RA levels. It has been shown that ethanol reduces RA levels due to the competition between ethanol clearance and RA biosynthesis (Duester, 1991, Fainsod et al., 2020), and many studies indicate that RA reduction triggers prenatal ethanol-induced abnormalities found in FASD (Johnson et al., 2007, Yelin et al., 2005, Marrs et al., 2010, Shabtai et al., 2018) including a mouse study showing that acute RA deficiency during the early gastrulation induces FAS-like sentinel craniofacial features (Petrelli et al., 2023). The similar ethanol/RA relationship may exist in the third trimester FASD model used in this study because our previous studies indicated that BT75, an RARα agonist, attenuated PV cell reduction induced by P7 ethanol (Saito et al., 2023). Thus, our current hypothesis is that P7 ethanol not only induces apoptosis in a set of immature PV neurons but also reduces RA signaling activity and disturbs PV cell maturation, which may enhance the long-lasting cognitive and behavioral abnormalities induced by developmental ethanol exposure. Further studies using RAR agonists/antagonists may be important. Conclusions To explore the effects of developmental ethanol on RA signaling activity, P7 RARE-LacZ reporter mice were treated with saline/ethanol, and β-Gal expression (RA signaling activity) was examined by immunofluorescence histochemistry in P8 and P30 forebrains. β-Gal expression was observed in limited cell types, such as DG granule neurons, AV/AM thalamic neurons, Tbr1 + layer 6 neurons, PV neurons, and astrocytes. P7 ethanol treatment reduced β-Gal expression in AV/AM thalamic neurons, Tbr1 + layer 6 neurons, and PV neurons. While P7 ethanol decreased total PV neuron numbers, PV+ β-Gal+ cell numbers were more severely reduced. As indicated in the prenatal FASD models, reduced RA levels by metabolic competition with ethanol may affect PV cell maturation and enhance cognitive and behavioral abnormalities. We used this third trimester mouse binge drinking model because it provides precise temporal control to help isolate effects of ethanol on neuroapoptosis and glial activation, which may be related to RA signaling activity. Furthermore, this model provides long-lasting brain deficits, cognitive dysfunctions, and behavioral abnormalities, which are often seen in FASD patients. However, this model not only exposes pups with high concentrations of ethanol but also lacks the effects of placenta, which are limitations of this study. CRediT authorship contribution statement Das Bhaskar C: Writing – review & editing, Resources, Funding acquisition, Conceptualization. Williams Colin R.O.: Methodology, Data curation. Wilson Donald A: Conceptualization, Funding acquisition, Writing – review & editing. Smiley John F: Writing – review & editing, Supervision, Funding acquisition, Formal analysis, Conceptualization. Jungann Park: Methodology, Investigation, Data curation. Brandon Marino: Methodology, Data curation. Anusha Nalluri: Validation, Methodology, Investigation, Data curation. Mariko Saito: Writing – original draft, Supervision, Project administration, Investigation, Funding acquisition, Data curation, Conceptualization. Ethical statement I have read and have abided by the statement of ethical standards for manuscripts submitted to IBRO Neuroscience Reports. All procedures of animal studies were reviewed and approved by the Nathan Kline Institute Institutional Animal Care and Use Committee IACUC) and were in accordance with NIH guidelines for the proper treatment of animals. This Manuscript is not under consideration for publication elsewhere and its publication is approved by all authors. If accepted, it will not be published elsewhere in the same form in English or in any other language, including electronically without the written consent of the copyright-holder. Funding This work was supported by grants from National Institute on Alcohol Abuse and Alcoholism (R21AA027374 to BD and MS; R01AA023181 to MS, DW, and JS). Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements None Footnotes Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ibneur.2026.05.006. Contributor Information Mariko Saito, Email: mariko.saito@nki.rfmh.org. Jungann Park, Email: jungannpark@gmail.com. Anusha Nalluri, Email: anusha.nalluri@nki.rfmh.org. Brandon Marino, Email: brandon.marino@nki.rfmh.org. Colin R.O. Williams, Email: colin.williams@nki.rfmh.org. Donald A. Wilson, Email: donald.wilson@nyulangone.org. Bhaskar C. Das, Email: baskard@bufallo.edu. John F. Smiley, Email: john.smiley@nki.rfmh.org. Supplementary material Supplementary material References Associated Data Supplementary Materials Supplementary material