Effects of postnatal alcohol exposure on hippocampal gene expression and learning in adult mice

Fetal alcohol syndrome (FAS) is a condition resulting from excessive drinking by pregnant women. Symptoms of FAS include abnormal facial features, stunted growth, intellectual deficits and attentional dysfunction. Many studies have investigated FAS, but its underlying mechanisms remain unknown. This study evaluated the relationship between alcohol exposure during the synaptogenesis period in postnatal mice and subsequent cognitive function in adult mice. We delivered two injections, separated by 2 h, of ethanol (3 g/kg, ethanol/saline, 20% v/v) to ICR mice on postnatal day 7. After 10 weeks, we conducted a behavioral test, sacrificed the animals, harvested brain tissue and analyzed hippocampal gene expression using a microarray. In ethanol-treated mice, there was a reduction in brain size and decreased neuronal cell number in the cortex, and also cognitive impairment. cDNA microarray results indicated that 1,548 genes showed a > 2-fold decrease in expression relative to control, whereas 974 genes showed a > 2-fold increase in expression relative to control. Many of these genes were related to signal transduction, synaptogenesis and cell membrane formation, which are highlighted in our findings.


INTRODUCTION
Fetal alcohol syndrome (FAS) is a condition caused by alcohol exposure during pregnancy.Symptoms of FAS include low cognitive function, stunted growth, facial deformities and central nervous system dysfunction (Streissguth et al., 1991(Streissguth et al., , 1994;;Astley et al., 2000;Streissguth and O'Malley, 2000;Barr and Streissguth, 2001).Neurobehavioral deficits associated with FAS include reduced intelligence quotient (IQ) and multiple impaired neurodevelopmental aspects, including attention, executive function, visuospatial and mathematical abilities, reaction time, fine motor skills, memory, language and social interactions (Streissguth et al., 1990(Streissguth et al., , 1991(Streissguth et al., , 1994;;Riley and McGee, 2005).While alcohol has been known to be a teratogen since the 1970s, many children are still affected by prenatal alcohol exposure (May et al., 2009).This may be due to women not realizing that they are pregnant during the first trimester, the most significant period of organ formation, and thereby continuing to drink excessively.In industrialized countries, the asso-ciated cost of prenatal alcohol exposure is burdensome for both parents and governments (Stade et al., 2009).
Inhibition of N-methyl-D-aspartate receptors and hyperactivation of γ-aminobutyric acid receptors are the two main mechanisms of alcohol toxicity (Hoffman et al., 1989;Lovinger et al., 1989;Harris et al., 1995).Extensive neuronal death and subsequent proliferation of astrocytes (astrogliosis) are histological markers of alcohol toxicity (Shin et al., 2007;Jeon et al., 2009aJeon et al., , 2009b)).There have been many cytological and molecular biological studies on the toxic effects of maternal alcohol abuse during pregnancy, but a complete understanding remains elusive.
In Mus musculus (in particular, the ICR mouse), neuronal development occurs during two periods.From post-fertilization day 8 until birth, there is growth in neuron numbers (neurogenesis period); afterwards, from birth until postnatal day 14, neurons make connections (synaptogenesis period) (Belnoue et al., 2013).Unlike mice, the synaptogenic period in humans starts from the sixth month of pregnancy.If the fetus is exposed to large amounts of alcohol during this synaptogenic period, extensive apoptosis and neuronal degeneration often occur (Ikonomidou et al., 2000;Dikranian et al., 2001).
To investigate the mechanism of fetal alcohol toxicity, we exposed ICR mice to ethanol via injection during the synaptogenic period.After the mice were fully grown, they underwent behavioral testing.We then harvested hippocampal tissue, extracted RNA and evaluated changes in gene expression.Our findings may explain how fetal alcohol exposure contributes to cognitive dysfunction.

Animals
Pregnant mice were purchased from Central Lab Animal Inc. (Seoul, Korea).Animals were kept on a 12-h circadian cycle and at 22 ± 3°C ambient temperature with free access to food and water.On postnatal day 7, mice were given two 3 g/kg ethanol (20% v/v) hypodermic injections separated by 2 h.On postnatal day 21, the mice were divided by gender; only the males were used in this study.The Morris water maze test was performed 10 weeks after injection.All animal procedures adhered to the animal care guidelines of Gyeongsang National University (Approval No. GLA-110512-M0029).

Tissue preparation
For histological analysis, all the mice were administered Zoletil (5 mg/kg, ip, Virbac Laboratories, Carros, France) and transcardially perfused with heparinized saline followed by 4% paraformaldehyde in 0.1 M phosphate buffered saline (PBS).The brains were removed and then post-fixed in 4% paraformaldehyde for 6 h.After post-fixation, brains were sequentially immersed in 0.1 M PBS containing 15% and then 30% sucrose until they sank.Brains were subsequently cut into 30-μm coronal sections and stained with cresyl violet, and the slides were evaluated by microscopy.Brain size was measured after transcardial perfusion and removal from the skull.Since the olfactory bulb and brain stem sides varied during removal, we took photographs and measured the diameter of the brain at the levels of the posterior margin of the optic tract and the hypothalamus-midbrain junction.The mean diameters were compared at each level.
Behavioral analysis Seven days before the Morris water maze test, we performed the elevated plus maze (EPM) test.The EPM has two open arms and two closed arms (30 × 7 cm each) and a connecting central platform (7 × 7 cm) mounted 50 cm above the floor.Tested mice were placed in the center of the maze facing an open arm, and behavior was recorded for 5 min.Arm entry was scored when the mouse moved into one of the arms.Seven days after the EPM, the Morris water maze test was performed as previously described (Jeon et al., 2012) with the following modifications.In brief, mice (n = 8 per group) were first trained to find a randomly positioned invisible platform in a 120-cm diameter swimming pool, maintained at 24 ± 1°C and 200 Lux.All mice were subjected to four trials per day for six consecutive days.The escape latency and swimming distance required to find the platform were recorded by a video-tracking program (Noldus EthoVision XT7, Noldus Information Technology, Wageningen, Netherlands).On day 7, the platform was removed, and time spent in the target quadrant, where the platform had been located during training, was analyzed.The starting position was changed with each trial.
Total RNA isolation We sacrificed the mice, and isolated the hippocampi and froze them in liquid nitrogen.Total RNA was extracted with Trizol (Invitrogen, Carlsbad, CA, USA).The purity and quantity of the RNA was measured by spectrophotometry (Ultrospec2100, Amersham, Piscataway, NJ, USA) at 260 nm.
cDNA microarray analysis After finishing the behavior tests, animals were kept for more than one week to stabilize.Hippocampal gene expression after behavioral testing was analyzed using an Agilent mouse oligo microarray kit (Digital Genomics, Seoul, Korea).In each group, eight mice and four cDNA chips were used (hippocampal total RNAs from two animals were pooled and applied to one chip).Scanned images were analyzed using GenePix Pro 6.0 software (Axon Instruments, Union City, CA, USA) to obtain gene expression ratios.The transformed data were normalized by LOWESS regression and analyzed using the GeneSpring GX 7.3 software program (Agilent Technologies, Santa Clara, CA, USA).Differentially expressed genes were then identified by fold changes and P values, which were obtained via the Student's t-test.Threshold values for upor down-regulated genes were fold changes of > 2 or > 5, respectively.P values < 0.05 were deemed statistically significant.

Statistical analysis
We performed one-way analysis of variance (ANOVA) and pairwise multiple comparisons (Tukey's test) for multiple-group comparisons.For between-group comparisons, we used the unpaired t-test.Statistical analysis was conducted using Prism software (GraphPad Prism, San Diego, CA, USA).

RESULTS
We injected ethanol (3 g/kg) into 1-week-old mice (during the synaptogenesis period) to evaluate changes in brain size, histology, cognition and hippocampal gene expression at adult stage (11 weeks old).
Ten weeks after ethanol injections, we sacrificed the animals, extracted brain tissues and measured brain size.We used two metrics to compare brain size: the diameter (at the coronal plane) at the level of the posterior margin of the optic tract (diameter a) and the diam-eter at the boundary of the hypothalamus and midbrain (diameter b) (Fig. 1).When the control group brain size was normalized to a value of 100, the relative values in the ethanol-treated group for diameters a and b were 92.44 and 94.36, respectively.This result clearly indicates that a binge level of ethanol treatment during the synaptogenesis period evoked smaller brain volume, as did ethanol treatment during neurogenesis (Mandal et al., 2015).
We stained the cerebral cortex with cresyl violet to identify neurons (Fig. 2) and counted the number of neurons in the motor cortex (MC) and cingulate gyrus  (CG).There were fewer neurons in the ethanol-treated group than in the control group.Neuron counts for the ethanol-treated animals in the MC and CG (control counts normalized to 100) were 75.23 and 81.07, respectively.We noted that there were many vacuoles in both the neurons and the intercellular space in the MCs of ethanol-treated mice.EPM tests and body weight profiles showed that there was no difference in anxiety or locomotor ability between saline-and ethanol-exposed groups (Fig. 3).
Ten weeks after ethanol injections, we performed Morris water maze tests to assess spatial cognition, learning and memory (Fig. 4).After the third day of training, there was a significant difference between the saline-and eth-anol-treated groups.These results suggest that alcohol exposure during the synaptogenic period decreases learning and memory in adulthood.
We next performed cDNA microarray analysis to investigate changes in gene expression in the hippocampus, a brain region well known to be critical for learning and memory.A total of 1,548 genes showed decreased expression (> 2-fold) in the ethanol-injected group relative to the control group.Table 1 lists down-regulated hippocampal genes related to cell membrane formation, signal transduction and signaling pathways with > 5-fold changes.There were 974 genes that showed increased expression (> 2-fold) in the ethanol-injected group relative to the control group.Table 2 lists up-regulated hip-Fig.3. Change of body weight and results of EPM tests.After weaning, body weights of the two groups were measured weekly (A).One week before the Morris water maze test, the elevated plus maze test was performed (B-E).Data are presented as mean ± SD.
pocampal genes related to cell membrane formation, signal transduction and Wnt signal pathways (which regulate body axis patterning, cell fate specification, cell proliferation and cell migration) with a > 3-fold increase in expression.

DISCUSSION
Intellectual deficits associated with FAS persist throughout life.Cognitive dysfunctions include lack of goal-oriented activities, attentional problems and failure in temporal planning.These kinds of dysfunction can Fig. 4. Results of the Morris water maze test.The ethanol-treated group showed impaired learning and memory.A, There were significant differences in escape latency for the two experimental groups, and learning and memory ability, represented as the daily slope, were lower in the ethanol-exposed group.B, Comparison of time spent in the target quadrant (southwest (SW), where the hidden platform was located during 6-day training) after removal of the platform on day 7. Data are presented as mean ± SD.One-way ANOVA and pairwise multiple comparisons (Tukey's test), #, $ P < 0.01 compared with day 1.Student's t-test between saline-and ethanoltreated groups, *P < 0.05, **P < 0.01.result in few achievements in work and academic settings (Klintsova et al., 2007;Whitcher and Klintsova, 2008).We explored changes in gene expression that might explain cognitive dysfunction in adults previously exposed to alcohol during the synaptogenic period.We confirmed that in mice, alcohol exposure during the synaptogenic period can impair learning and memory in adulthood.FAS is the most severe disorder in the fetal alcohol spectrum disorders (FASD) resulting from prenatal alcohol exposure.There are many FASD studies utilizing diverse alcohol injection methods and timings of injections; nevertheless, these animal models exhibit symptoms of FASD (Oshiro et al., 2014;Patten et al., 2014).Our results with the Morris water maze test to evaluate cognition in our model of FAS indicate that both control and ethanol-treated mice had learning and memory ability, but the ethanol-treated group had fewer achievements.
The most widely accepted theory of memory formation is that it occurs during synaptogenesis (Arnsten and Cai, 1993;Cattabeni, 1997;Wilson and Tonegawa, 1997;Abel and Lattal, 2001;Lim et al., 2014).In humans, the synaptogenic period occurs in the latter half of the prenatal period, whereas in the mouse it is from birth to postnatal day 14.Exposure of mice to large amounts of ethanol during this period can lead to neuronal cell death in various regions of the cerebral cortex within 24 h (Han et al., 2006;Allan et al., 2014).Therefore, we injected ethanol at postnatal day 7 to perturb synapse formation during the synaptogenic period.We examined whether or not the effects of the known early neuronal death within 24 h in the cortex persisted at ten weeks after injection, through cresyl violet staining.Ethanol exposure led to a smaller number of neurons in the cortex.Furthermore, the sizes of ethanol-exposed mouse brains were smaller than those of control brains.These results imply that alcohol exposure even at a late stage of pregnancy, synaptogenesis, in humans may cause permanent anatomical and histological changes in offspring.Ten weeks after injection, the ethanol-exposed group displayed fewer learning achievements in the Morris water maze test, suggesting that our model can replicate cognitive deficits in human adults with FASD.
We next tried to correlate the lowered cognitive function of ethanol-exposed mice with gene expression, for which purpose cDNA microarray analysis was adopted using RNAs from the hippocampus, a brain region well known to be critical for learning and memory.Among the genes with significantly altered expression levels were those related to synapse formation, cell membrane plasticity and signal transduction (Tables 1 and 2).One noteworthy result is that many genes for olfactory receptors showed alterations in expression level.Ansoleaga et al. (2013) reported that olfactory receptors 2, 51 and 52 were down-regulated in the dorsolateral prefrontal cortex of schizophrenia patients.Changes in the gene expression of olfactory receptors in schizophrenia and dementia patients have been reported in many previous studies; these changes correlated highly with changes in cognitive function in these patients (Brewer and Pantelis, 2010;Ansoleaga et al., 2013).Interestingly, our microarray results also suggest that in alcohol-treated mice, there were attempts at removing debris, presumably resulting from the death of neurons, and recovery of cognitive function.One recent study found that apoptosis antagonizing transcription factor (Aatf) prevents Ats fibril formation and alleviates symptoms of dementia.Aatf can also prevent pathological plaque formation after neuronal cell death.We found that hippocampal Aatf expression increased in ethanoltreated mice (Table 2).Zhu et al. (2013) reported that decreased synaptojanin1 (Synj1) expression facilitates Aβ fibril removal and alleviates cognitive dysfunction in an Alzheimer's disease mouse model.Indeed, Synj1 expression decreased in ethanol-treated mice (Table 1).A recent study reported that leptin, whose expression increased > 5-fold in treated mice (Table 2), is important for regulating neuronal plasticity, and is essential to restoring cognitive function in affected mice (Davis et al., 2014).Our results also indicate that expression of genes related to cell division and migration increased in ethanol-treated mice.Abl-interactor 2 (Abi2, Table 2) is important for the growth of neuronal dendrites (Courtney et al., 2000); in the present study, expression of Abi2 was up-regulated more than 5-fold relative to control.Wnt signal-related proteins promote long-term potentiation, synapse formation and axon guidance via three Wnt subpathways (Fortress and Frick, 2015).These three functions are crucial for restoring memory function of the affected brain.To summarize, in the adult mouse brain, 10 weeks after ethanol exposure, there appears to be evidence of attempts to clear debris from the death of neurons to restore cognitive function.Despite indications of compensation, galanin receptor 3 (Galr3, Table 1), which improves cognitive function, is still down-regulated (Brewer and Pantelis, 2010), which may explain the persistence of cognitive impairments in FAS patients.
Changes in expression of individual genes cannot fully explain the cognitive deficits in ethanol-treated mice.Neuronal structure can also affect cognitive function.For example, the dendritic spine density has been found to be decreased in the frontal cortices of mice treated with ethanol during the synaptogenesis period, without changes in dendritic spine morphology (Whitcher and Klintsova, 2008).Additionally, alcohol can damage stem cells and progenitor cells in the hippocampus, which can inhibit neurogenesis in the adult mouse brain (Ieraci and Herrera, 2007).These findings may account for cognitive dysfunction not explained by our microarray data.Thus, alcohol may induce cognitive deficits not only by affecting individual genes but also by a process involving elements from the level of dendritic spine formation to adult stem cell division and migration (Guizzetti et al., 2014).
We believe that changes in gene expression reported herein reflect compensatory mechanisms in response to the loss of neuronal connections following ethanolinduced cell death.These findings may provide evidence of neuronal stem cell division, proliferation and/or migration directed at replacement of dead neurons (Jeon et al., 2009b).Reduction in brain size resulting from ethanol exposure during the synaptogenesis period may also highlight permanent neuronal damage and lower-than-normal cognitive function.Our results imply that changes in hippocampal gene expression may underlie the lifetime cognitive deficits of FAS patients.However, there were changes in the expression of many genes associated with signal transduction.Further studies and analyses are required for a complete interpretation of our present results and of the relationship between gene expression and cognition.
This research was supported by a grant from the Institute of Health Sciences (IHS GNU-2013-04).

Fig. 1 .
Fig. 1.Diameters of brains of saline-and ethanol-treated mice.A, Method of measurement, where a indicates the posterior margin of the optic tract, and b indicates the hypothalamus-midbrain junction.B, Representative photographs of brains of saline-and ethanol-treated mice.C, Ethanol-treated mice had smaller smaller brains than mice in the saline-treated group.Data are presented as mean (mm) ± SD. *P < 0.05, **P < 0.01.

Fig. 2 .
Fig. 2. Cresyl violet staining showing neurons in the cortex.A, The cortex was divided into the motor cortex (MC) and cingulate gyrus (CG).The cortices of ethanol-treated mice had fewer neurons than those of the saline-treated group.B, Higher-resolution images showing many vacuoles (arrowheads) in the MC of an ethanol-treated mouse (right; left, salinetreated MC).Scale bars, 100 μm.*P < 0.05, **P < 0.01.

Table 1 .
Down-regulated hippocampal genes related to cell membrane formation, signal transduction and signaling pathways (> 5fold change 10 weeks after ethanol injection)

Table 2 .
Up-regulated hippocampal genes related to cell membrane formation, signal transduction and Wnt signaling pathways (> 3fold change 10 weeks after ethanol injection)