Prenatal and adolescent alcohol exposure, neuroinflammation, and Alzheimer’s disease: a network meta analysis approach

Introduction

Alcohol is a substance commonly used in American society. About fourteen million Americans reach the clinical criteria for alcohol use disorder, with nine percent of alcohol-dependent patients being diagnosed with brain disorders. Alcohol-related cognitive impairment is the second leading cause of adult dementia in the United States, being exceeded only by AD [1]. While the association between alcohol and adult dementia is well studied [1], [2], [3], [4], there are few studies done about whether developmental exposure to alcohol promotes Alzheimer’s disease later in life. Fetal alcohol spectrum disorders (FASD) are caused by the exposure of the fetus to ethanol during pregnancy. Damage to the central nervous system is a common side effect of fetal alcohol exposure and is an important cause of developmental intellectual disability [5, 6]. Estimated rate of FASD is between 2.0 and 6.7 % and is the leading cause of mental retardation worldwide [7]. The neuropathology of FASD includes a reduced volume of the cerebral cortex, corpus callosum, cerebellum, and structures of the subcortex including the amygdala, basal ganglia, hippocampus, and thalamus [8–13]. FASD also causes memory and learning disorders, and problems with the processing of information [14], [15], [16]. Fetal alcohol exposure (FAE) has been demonstrated to induce specific biochemical and behavioral phenotypes that exhibit notable similarities to features observed in AD pathology [17]. In a rat study by Chaudhary and Sarkar, FAE increased levels of AD-related markers such as acetylcholinesterase, hyperphosphorylated-tau protein, β-amyloid, BACE1, and UNC5C proteins in the cerebral cortex and hippocampus and AD-related behavioral phenotypes like impairment of learning and memory functions [17]. These findings suggest that FAE could promote biochemical and cognitive traits akin to Alzheimer’s disease. In addition to FASD caused by ethanol exposure during gestation, alcohol use during the period of brain development in adolescence may also be harmful [8, 18, 19].

Adolescence is a stage of the developing brain during which changes in the structure and function of synaptic plasticity and the connections of neurons occur in various brain regions, such as the subcortical and cortical structures [20]. Alcohol exposure during adolescence period may lead to long-term cognitive and behavioral dysfunctions [21, 22]. Exposure to alcohol in the developing central nervous system (CNS) can alter the functions of neurons directly [23, 24] or via modification of the functions of glial cells such as astrocytes and microglia which maintain the homeostatic environment [10, 25, 26]. While these glial cells operate as the immune system in the CNS and defend against pathogens and other insults, their activation can also damage the CNS and lead to aberrant CNS function [27], [28], [29]. While the exact effects of ethanol in the developing brain are still not fully untangled, there is evidence that the response of the neuroimmune system is involved [26, 30, 31].

Current evidence from neuropathology and genetic studies suggests that AD may involve an amyloid cascade, which leads to the degeneration of neurons due to the accumulation of amyloid beta plaques in the brain [32–37]. Neurofibrillary tangles, which are intracellular aggregates of hyperphosphorylated tau protein, are another cause of AD [38], [39], [40]. The role of neuroinflammation in AD is supported by evidence that microglia are activated by cellular death or neurofibrillary tangles and cause an immune response mediated by danger or pathogen-associated molecular pattern receptors called DAMPS/PAMPS [41–50]. Mouse models have demonstrated that the selective elimination of microglia, achieved through pharmacological inhibition of the colony-stimulating factor 1 receptor (CSF1R) signaling, is associated with a reduction in neuronal loss and amelioration of memory impairments in AD [51]. The most relevant type of cellular death in AD is apoptosis, which is a type of regulated cell death. The intrinsic pathway of apoptosis, controlled by the pro-apoptotic proteins BAX and BAD from the family Bcl-2, is upregulated in the frontal cortex and hippocampus of AD mouse models [52, 53]. In this study, loss of Bad prevented neuronal apoptosis and degeneration and restored behavioral alterations and cognitive decline. These findings suggest that neuroinflammation and apoptosis are key processes in the pathogenesis of AD and may be important targets for therapeutic intervention. Finally, excessive glutamate released from damaged neuronal cells and microglia in AD leads to calcium influx through N-methyl-d-aspartate receptors (NMDARs), which can cause cellular damage [54], [55], [56], [57]. NMDARs are protein complexes with GluN1 and GluN2 subunits that modulate intracellular signaling pathways via Ca2+ influx through an ionic pore [55]. Overactivation of NMDARs leads to neuronal cell death through apoptosis involving excess calcium release, depolarization of mitochondria, and activation of cytochrome c and caspase-3 [56, 57]. Dysregulated amyloid precursor protein (APP) processing by microglia and excessive NMDA receptor (NMDAR) activation due to glutamate release from degenerating neurons have been implicated in the formation of neurofibrillary tangles, contributing to increased neuronal cell death [58].

Extensive research has focused on elucidating the role of neuroinflammation in AD development. Additionally, investigations have explored the influence of developmental alcohol exposure on the neuroinflammatory processes associated with AD. However, the specific effects of early-life alcohol exposure on the development of AD have yet to be determined. In this study, we used a meta-analysis approach to determine if prenatal alcohol exposure (PAE) and/or adolescent alcohol exposure (AAE) could activate key neuroinflammatory genes for an increased risk of developing AD later in life.

Methods

Collection of molecules affected by prenatal/adolescent alcohol exposure

Molecules affected by PAE [including fetal alcohol exposure (FAE), gestational alcohol exposure (GAE), neonatal alcohol exposure (NAE), fetal alcohol spectrum disorder (FASD)], and AAE were gathered from a literature search of MEDLINE via Pubmed (PubMed.gov) before October 20, 2022, which showed a total of 22,936 articles, including possible duplicates, for PAE and 6745 for AAE usage (Figure 1). Within these results, 2047 papers covered the effects of PAE and/or AAE exposure on neurotoxic or neuroinflammatory pathways, with 163 articles showing the effects of such pathways on gene expression. Out of these 163 articles describing gene expression of alcohol’s effects on developmental neuroinflammatory pathways, we selected 32 articles that best identified the differential expression of genes upon exposure to PAE/AAE in mouse, rat, and human subjects (Figure 2a). We also identified 38 datasets from the Gene Expression Omnibus database located at the NCBI website which demonstrated the effects of PAE or AAE on gene expression in the brains of mice, rats, and humans (Figure 2c). According to the IPA website, IPA software cannot interpret raw microarray data and this function can only be performed with a separate tool called CLC workbench. Out of these 38 RNA-seq datasets from mouse, rat, and human GEO entries, only nine were not based on microarrays or had array genetic data published in a publicly available study and demonstrated statistically significant effects of PAE and/or AAE on neuroinflammatory pathways above the threshold cutoff value.

Figure 1:

Schematic diagram showing neuroinflammatory pathways in Alzheimer’s disease leading to neuronal cell death. Abnormal APP processing by microglial cells leads to neuroinflammation and cell death. In a similar manner, NMDA receptor activation by excessive glutamate released from dying neurons leads to release of excessive calcium which activates BAD and depolarizes mitochondria, leading to cell death by apoptosis. Calcium also activates calpain which cleaves tau and creates neurofibrillary tangles, further leading to neuroinflammation and death of neuronal cells.

Figure 2:

Workflow to identify molecules affected by prenatal alcohol exposure (PAE) and adolescent alcohol exposure (AAE) from three different sources.

Due to the subject of this paper being PAE and AAE, only control and alcohol-exposed groups were selected from our GEO dataset search. If any duplicate categories of control or ethanol-exposed groups were found in a dataset, we calculated the average value for each group. Next, the gene expression fold change was calculated by dividing the gene expression values for ethanol-exposed groups by the value of the control group. The log2 of the fold change for each gene was then determined and an Excel spreadsheet was then placed into IPA for analysis (Figure 2c). The aforementioned steps were performed for each mouse, rat, and human dataset. Genes showing significant differences between PAE and AAE were selected to construct a pathway analysis map in IPA. Homologs were not converted between species. Data sets were analyzed under identical parameters and the cut-off was set from −1 to +1. Genes with reported significant expression level changes between the control values and PAE and/or AAE in the literature were identified as molecules affected by EtOH (p<0.05) and added to a pathway in IPA using its “Build, Add molecules/relationship” tool. Data sets were studied using IPA software to identify the neuroinflammatory genetic network. The analysis of GEO datasets was performed using the Canonical Pathway tool. In gene expression studies where thousands of genes are measured between two experimental conditions, the false discovery rate (FDR) is a concern. The False Discovery Rate (FDR) represents the risk of encountering false positive results when conducting multiple tests. In this study, we used IPA to employ a technique devised by Benjamini and Hochberg to identify and remove such false positives caused by FDR [62]. Each IPA analysis generated either a positive or negative z score for all canonical and individual molecules involved in each pathway which indicated its activation or inhibition respectively. The set of molecules that were found to be affected by literature and GEO datasets were then added to a pathway window in IPA. Connections between these molecules were then identified using the “Connect” tool using the focused results of QKB.

Results

IPA analysis of alcohol-exposed brain datasets using the Disease and Functional Analysis tool revealed significant associations between alcohol exposure and various functional categories and diseases. Notably, multiple datasets showed a strong association between alcohol exposure and the “Inflammatory Response” category (GSE184615, p=1.0e-07; GSE14714, p=1.0e-12; GSE154018, p=6.9e-03; GSE63561, p=5.0e-03; GSE133369, p=5.0e-03; GSE195489, p=1.0e-04; GSE60901, p=1.0e-04; GSE34469, p=1.0e-05; GSE115188, p=1.0e-05), indicating that alcohol exposure elicits an inflammatory response in the brain.

Additionally, alcohol exposure was significantly associated with “Nervous System Development and Function” in multiple datasets (GSE154018, p=1.0e-05; GSE184615, p=1.0e-14; GSE14714, p=1.0e-17; GSE195489, p=4.0e-03; GSE60901, p=3.0e-03; GSE34469, p=1.0e-05; GSE115188, p=1.0e-06), suggesting detrimental effects on the development and functioning of the nervous system.

Furthermore, alcohol exposure exhibited significant associations with the “Organismal Injury and Abnormalities” category (GSE184615, p=1.0e-08; GSE14714, p=1.0e-23; GSE133369, p=1.0e-05; GSE63561, p=1.0e-05; GSE195489, p=1.0e-14; GSE60901, p=1.0e-14; GSE34469, p=1.0e-04; GSE115188, p=1.0e-05), indicating potential adverse effects on overall organismal health.

The category “Cell Death and Survival” also showed significant associations with alcohol exposure (GSE133369, p=1.0e-05; GSE63561, p=5.0e-05; GSE154018, p=1.0e-04; GSE184615, p=1.0e-04; GSE195489, p=1.0e-04; GSE60901, p=1.0e-04; GSE34469, p=1.0e-07; GSE115188, p=1.0e-04), suggesting increased cell death in the brain due to alcohol exposure.

Moreover, the “Neurological Disease” category displayed significant associations with alcohol exposure (GSE115188, p=3.595e-04; GSE154018, p=1.0e-06; GSE184615, p=1.0e-05; GSE14714, p=1.0e-23; GSE133369, p=1.0e-04; GSE63561, p=1.0e-05; GSE195489, p=1.0e-06; GSE60901, p=1.0e-06; GSE34469, p=1.0e-04).

Furthermore, canonical pathway analysis revealed additional significant associations between alcohol exposure and specific pathways. The “Neuroinflammation Signaling Pathway” showed significant associations with alcohol exposure in multiple datasets (GSE14714, p=0.16; GSE34469, p=0.018; GSE60901, p=0.203; GSE63651, p=0.463). The “ERK/MAPK Signaling” pathway was significantly associated with alcohol exposure in several datasets (GSE14714, p=0.021; GSE60901, p=0.151; GSE63651, p=0.422; GSE115188, p=0.422). Additionally, the “Amyloid Processing” pathway showed significant associations with alcohol exposure in multiple datasets (GSE14714, p=0.012; GSE60901, p=0.176; GSE63651, p=0.512; GSE115188, p=0.341). The “IL-1 Signaling” pathway was significantly associated with alcohol exposure in several datasets (GSE14714, p=0.008; GSE34469, p=0.445; GSE60901, p=0.342; GSE115188, p=0.342). Finally, the “Calcium Signaling” pathway showed significant associations with alcohol exposure in multiple datasets (GSE14714, p=0.002; GSE34469, p=0.008; GSE60901, p=0.295; GSE63651, p=0.464; GSE115188, p=0.464).

In addition to these findings, the “Oxidative Phosphorylation” pathway was significantly associated with alcohol exposure in several datasets (GSE34469, p=0.016; GSE60901, p=0.322; GSE63651, p=0.422; GSE115188, p=0.422). These results suggest that alcohol exposure may impact oxidative phosphorylation processes in the brain.

In order to further explore the potential interactions and pathways involved in the neuroinflammatory processes associated with prenatal and adolescent alcohol exposure (PAE/AAE) and Alzheimer’s disease (AD), we conducted a network analysis using the Build/Grow and Connect tools. First, we identified molecules that were common to both the gene expression omnibus (GEO) datasets and the literature. These molecules were cross referenced with genes associated with neuroinflammation and AD obtained from the Qiagen Knowledge Base.

Figure 5 illustrates the results of this network analysis, highlighting the potential involvement of several cytokines and signaling molecules. These include TGF-β, TNF-β, IFN-β, CXCL8, IL1-β, B2M, BCL2, BDNF, CALB2, CASP1, CASP3, CASP8, CD86, CNTF, CX3CL1, CCL2, FOS, MYD88, GABRA1, GABRA2, GABRA4, GABRA5, GABRB3, NFKB, GABRD, GABRG1, GABRG2, GABRQ, GABRR1, BARR2, GRIN2A, GRIN2D, HMOX1, IL10, IL4, IL6, MAPK1, MAPK3, NOS2, TLR4, and TRAF6. These molecules may play a role in the PAE/AAE-associated neuroinflammatory signaling pathways and/or AD.

Discussion

Our study focused on investigating the effects of alcohol exposure on the brain, with a particular emphasis on neuroinflammation and its potential association with the development of neurological disorders, including AD. Through our analysis of disease and functional categories in alcohol-exposed brain datasets, we observed significant associations with neuroinflammation, nervous system development and function, organismal injury and abnormalities, cell death and survival, as well as neurological disease (Figure 3). These findings underscore the detrimental impact of alcohol exposure on the brain, disrupting normal brain functioning and potentially contributing to the development of neurological disorders (Figure 4).

Figure 3:

Showing the results of Disease and Functional Analysis performed on mRNA data sets extracted from the GEO accession database. The datasets were analyzed in IPA software to identify diseases and functions associated with differentially expressed genes in response to PAE and AAE. The disease and Functional Analysis tool was used to identify upregulated or downregulated canonical pathways in response to PAE/AAE. The pscore for each canonical pathway was calculated, and the significance for the association with each pathway was determined using benjamini–Hochberg-corrected Fisher’s exact test. The identified pathways are presented in the figure with their corresponding p-values. This graph is best viewed at larger size.

Figure 4:

This figure shows the results of Canonical Pathway Analysis performed on mRNA data sets extracted from the GEO accession database. The datasets were analyzed in IPA software to identify canonical pathways associated with differentially expressed genes in response to prenatal alcohol exposure (PAE). The Canonical Pathway tool was used to identify upregulated or downregulated canonical pathways in response to PAE. The p-score for each canonical pathway was calculated, and the significance for the association with each pathway was determined using a Benjamini–Hochberg-corrected Fisher’s exact test. The identified pathways are presented in the figure, with their corresponding p-values. This graph is best viewed at larger size.

Furthermore, our study utilized canonical pathway analysis to provide additional support for the involvement of alcohol exposure in neuroinflammatory processes and the dysregulation of critical signaling pathways. The Neuroinflammation Signaling pathway exhibited a significant association with alcohol exposure, suggesting that alcohol may induce neuroinflammatory responses in the brain. Dysregulated ERK/MAPK signaling, crucial for synaptic plasticity and cell survival, implies that alcohol exposure during critical stages of brain development may disrupt normal signaling processes, potentially leading to neurofunctional impairments [32].

We also observed the dysregulation of the Amyloid Processing pathway, which is implicated in the pathogenesis of Alzheimer’s disease [33], [34], [35], [36], [37]. This finding suggests a potential link between alcohol exposure, neuroinflammation, and the development of neurodegenerative disorders. The significant association between alcohol exposure and the IL-1 Signaling pathway, involved in inflammatory responses, further supports the involvement of inflammatory processes in alcohol-induced neurotoxicity. Additionally, the dysregulation of the Calcium Signaling pathway, critical for neuronal function and synaptic plasticity [56], highlights the potential impact of alcohol exposure on neurotoxicity and neurodevelopment.

Taking these findings together, our study provides valuable insights into the molecular and functional consequences of alcohol exposure on the brain. The identification of associations with neuroinflammation, dysregulated signaling pathways, and the involvement of critical pathways related to neurodegenerative disorders enhances our understanding of the mechanisms underlying the neurotoxic effects of alcohol. This knowledge is essential for comprehending the pathogenesis of neurological disorders, particularly Alzheimer’s disease, and may inform preventive strategies and interventions.

In the context of our study, Figure 5 depicts the results concerning the potential involvement of various cytokines and signaling molecules in the neuroinflammatory signaling pathways and/or Alzheimer’s disease associated with prenatal and adolescent alcohol exposure (PAE/AAE). Notably, several cytokines, including TGF-β, TNF-β, IFN-β, CXCL8, and IL1-β, were found to be associated with both neuroinflammation and AD. Prior research has implicated these cytokines in neuroinflammatory processes and demonstrated their upregulation in the brains of individuals with AD [63, 64]. The dysregulation of these cytokines in the context of PAE/AAE suggests their potential involvement in the neuroinflammatory response triggered by alcohol exposure, thereby potentially contributing to the development or progression of AD.

Figure 5:

Schematic diagram from IPA showing molecules interacting in the neuroinflammatory signaling pathway and Alzheimer’s disease. Lines highlighted in blue show connections between molecules and neuroinflammatory signaling pathways, black lines show connections between molecules and Alzheimer’s disease. This image was constructed using IPA software. The build option in My Pathway was used to add molecules common to both literature and GEO datasets. After the addition of all molecules, molecules associated with the neuroinflammatory pathway and Alzheimer’s disease identified from QKB were added to construct a network. In this diagram, numerous molecules have connections to the neuroinflammatory signaling pathway and Alzheimer’s disease. Molecules common to both GEO datasets and literature that were not shown to interact with genes relevant to neuroinflammation and/or Alzheimer’s disease were not included in this figure. This image shows that numerous genes associated with neuroinflammation and/or Alzheimer’s disease are activated by PAE/AAE exposure.

Furthermore, our study identified various signaling molecules, such as B2M, BCL2, BDNF, CALB2, CASP1, CASP3, CASP8, CD86, CNTF, CX3CL1, CCL2, FOS, MYD88, GABRA1, GABRA2, GABRA4, GABRA5, GABRB3, NFKB, GABRD, GABRG1, GABRG2, GABRQ, GABRR1, BARR2, GRIN2A, GRIN2D, HMOX1, IL10, IL4, IL6, MAPK1, MAPK3, NOS2, TLR4, and TRAF6, within the context of PAE/AAE, neuroinflammation, and AD. These molecules play diverse roles in various cellular processes, including apoptosis, neuroplasticity, immune responses, and synaptic functions [63–71]. PAE/AAE effects during developmental period may permanently change the normal brain development and functions and make the exposed brain vulnerable for the development of AD.

We found PAE/AAE altered expression levels of the neurotrophic factors BDNF and CNTF. Throughout the prenatal and adolescent phases, neurotrophic factors BDNF and CNTF play pivotal roles in nurturing neuronal vitality and fostering synaptic plasticity [66, 67]. BDNF and CNTF, critical during brain development, may influence AD vulnerability through their effects on early synaptic plasticity [68]. We also found changes in the GABA receptor subunits (GABRA1, GABRA2, GABRA4, GABRA5, GABRB3, GABRD, GABRG1, GABRG2, GABRQ, GABRR1) and the NMDA receptor subunit GRIN2A. Proper GABAergic function during developmental stages is crucial for establishing a balanced neural network, and disturbances could promote the AD-related neuronal imbalances [69, 70]. We detected changes in the apoptotic signaling molecules BCL2, CASP1, CASP3, and CASP8 due to PAE/AAE exposures. During the prenatal period, disturbances in BCL2, CASP3, and CASP8-mediated survival mechanisms could play pivotal roles in shaping the susceptibility to neuronal loss observed in AD [71], [72], [73]. PAE/AAE exposures also affected CD86, CCL2, TLR4, and MYD88 known to be involved in proinflammatory immune response [74, 75]. PAE/AAE could potentially prime the immune system towards a pro-inflammatory state, thereby contributing to the neuroinflammatory milieu inherent in AD [76]. Molecules like NFKB, FOS, and MAPKs (MAPK1 and MAPK3) are also affected by PAE/AAE. Disruptions in these molecules may compromise stress responses, synaptic plasticity, and cognitive function [77, 78], which is often observed in AD. Furthermore, HMOX1, IL10, IL4, and IL6, known to be involved in regulating oxidative stress and/or inflammation [79, 80], are also identified. PAE/AAE-induced alterations in these molecules might accentuate oxidative damage and spur neuroinflammatory cascades, thereby potentially heightening AD susceptibility. We identified PAE/AAE altered expression of CALB2 and CX3CL1 that are involved in calcium signaling and neuroprotection [81, 82]. Their roles could impact neuronal health and AD susceptibility through their effects on calcium homeostasis and neuroprotective mechanisms.

In summary, our study underscores the potential involvement of a wide range of cytokines and signaling molecules in PAE/AAE-associated neuroinflammatory signaling pathways and/or Alzheimer’s disease. These findings contribute to our understanding of the molecular mechanisms underlying the neurotoxic effects of alcohol exposure and their potential contribution to the development of neurological disorders, including AD. Further research focused on elucidating the specific roles of these molecules and their interactions within neuroinflammatory pathways is critical for a comprehensive understanding of AD’s pathogenesis and the development of targeted interventions to mitigate the detrimental effects of alcohol on the brain. Raising awareness about the adverse effects of alcohol exposure, particularly during critical periods of brain development, and promoting preventive measures are imperative for reducing the risk of neuroinflammation and associated neurological diseases.