Autism Spectrum Disorder (ASD) is defined by the National Institute of Health (NIH) as a subclass of neurodevelopmental disorders that impairs social communication, often displaying repetitive and restrictive behavior (Autism spectrum disorder fact sheet, 2022). The prevalence of individuals diagnosed with ASD has increased over the last several decades. In the 1970s, the estimated prevalence of ASD in the United States was 1 in 10,000. In the year 2000, the estimation increased to 1 in 2000 and in 2016, the estimation was 1 in 54 (Han et al., 2022). The etiology of ASD is not fully understood, but research suggests the disruption of microglial cell activity contributes to the abnormal brain development in individuals with ASD. Microglial cells are the immune cells of the central nervous system and mediate the damage to cells caused by inflammation. Under normal physiological conditions, they shape the neuronal circuits of the brain through dendritic pruning (Wake et al., 2017). This activity is particularly high during early post-natal development because it is important for development of mature synapses and functional neuronal circuits. In cases of ASD, dysfunctional activity of microglia allows for larger dendritic spines, which results in functionally immature synapses and disrupted neuronal circuits. (Lee et al., 2017). One specific objective of current research is to find the source of the dysfunctional cells.

A suspected cause of this altered microglial activity is exposure of a developing brain to inflammation. One study used data mining of hospital records to show an increased risk of being diagnosed with ASD when an individual was exposed to an infection in utero (Al-Haddad et al., 2019). Multiple studies using animal models of ASD demonstrate that exposure to conditions that cause neuroinflammation, both prenatally and postnatally, result in altered microglial morphology and activity (Xu et al. 2020). Post-mortem studies of patients diagnosed with ASD reported similar results to animal studies, showing distorted morphology and activity of microglia in the cortex of the patients (Lee et al., 2017). Additionally, patients with ASD and animal models of ASD show altered dendritic spines overlapping with the areas affected by the distorted microglia (Hustler, Hong., 2010).

This literature review will focus on the correlation between decreased proper synaptic pruning, increased distortion of morphology and activity of microglia and dendrites, and exposure to neuroinflammation. The proposed hypothesis is that the high density of dendritic spines in individuals with ASD is due to a lack of functional microglia cells in response to inflammation during early post-natal developmental years, when high rates of synaptic pruning occur.  Understanding the cause and timing of neuroinflammation and its relationship to microglia’s synaptic pruning could lead to the development of preventative treatments for ASD.

Altered Microglia in Post-Mortem Patients diagnosed with ASD

Microglia are immune cells found in the central nervous system (CNS). They are regulators of synaptogenesis, inducing apoptosis of excess neurons and pruning of synapses (Matta et al., 2019).  Microglia are active both prenatally and postnatally and are quick to release cytokines and prostaglandins in response to pathological changes, further increasing inflammation in the CNS.

Microglia also adapt their own morphology in response to their environment. The stage of adaptation they are in indicates their function. Ramified microglia are resting microglia with typical morphology, while primed microglia are associated with higher levels of activity in response to inflammation (Lee et al., 2017). Primed microglia have larger soma bodies, thicker branches, with higher rates of activity (Matta et al., 2019).

In patients with ASD, the overall number of microglia does not differ in relation to typically developing patients, but the proportion of primed microglia is higher (Lee et al., 2017). In patients with ASD, the increased presence of primed microglia consistently appears across the cortex, such as the dorsolateral prefrontal cortex (Morgan et al., 2010), temporal cortex (Lee et al., 2017), and layer II of the parietal lobe (Hustler, Hong., 2010). It is possible that the altered state of the microglia in patients with ASD is a side-effect of the protective response to neuroinflammation, inducing altered dendritic spines in the cortex (Tetreault et al., 2012).

Colocalization of Dysfunctional Microglia and Altered Dendritic Spines

The increased presence of primed microglia with altered morphology and activity colocalizes with the altered morphology of dendritic spines in the cortex of patients with ASD and animal models of ASD. Histology studies show differences in the morphology of the cortex in patients with ASD and typically developing patients by examining the average spine densities found on individual pyramidal cells. Hustler and Hong (2010) found the areas of the cortex in ASD patients with increased presence of primed microglia colocalized with areas of altered dendritic spines. The dendrites were longer and more densely packed along the processes of the cell.

Xu et al. (2020) found that male mouse models of ASD had higher spine density in pyramidal neurons of medial prefrontal cortex layer 5 and hippocampal CA1 area, indicating that elevated microglial protein synthesis increases the density of excitatory synapses in those areas. A similar study by Gyoneve et al. (2019) with mouse models of ASD that demonstrated ASD-like behavior, such as repetitive behavior and decreased appropriate social interactions, also showed decreased presence of deramified microglia at 2 months old. Decreased presence of defamified microglia is associated with defective synaptic pruning, a finding similar to the human studies of Hustler and Hong (2010).

Isshiki et al. (2014) demonstrated that the early postnatal cortex in mouse models of ASD undergo altered turnover of excitatory synapses of spines associated with afferents from cortical neurons. Their study utilized in vivo spin imaging to analyze time lapse images of spine subtypes using an inhibitory postsynaptic marker, gephyrin. Their study suggested that an enhanced turnover of excitatory synapses in mouse models of ASD is responsible for the abnormal function of synapses in the cortex during development. While Isshiki et al. (2014) did not conclude if microglia were the instigator of this altered pruning activity, they came to similar conclusions as the previous research by Xu et al. (2020) and Hustler and Hong (2010). Altered dendritic spine morphology and activity is a likely contributor to ASD symptoms in both patients and animal models.

Altered Morphology and Activity of Microglia in ASD Following Inflammation Exposure

Microglia and inflammation of the CNS are a common theme across research studies on ASD. There are several methods of invoking inflammation in the CNS to observe if the chances of developing ASD-like symptoms and altered cortex morphology are linked. For example, injecting a serum called Poly (I:C) into the uterus of a pregnant mouse to induce ASD-like symptoms in the pups, or injections of ovalbumin to induce an asthma condition in mouse models. Saitoh et al. (2021) performed a study on ovalbumin induced mouse models of asthma to observe the effects of long-term asthma induced inflammation in comparison to short-term asthma induced inflammation and control groups. They found that long-term asthma induced a surplus of excitatory synapses and inflammatory microglia, along with ASD-like behavior. Additionally, they found that long-term asthma decreased the expression of glucocorticoid receptors on microglia. Research experiments of exposure of ASD models to inflammation in-utero showed similar results to the studies of post-natal inflammation exposure. Maternal immune activation (MIA) was found to disrupt the balance between pro-inflammatory and anti-inflammatory cytokines of the fetal brain, inducing long-lasting alterations on the neurodevelopmental process in research conducted by Han et al. (2022). Cytokines carry an important role in normal brain development and changes in their expression will alter synaptic connectivity and function. Ozaki et al. (2020) expanded on this research further by suggesting that the increased risk of ASD in response to inflammation was time dependent. In mouse models of ASD, they induced MIA at day 12 of gestation, for the human second semester equivalent estimation, and day 15 of gestation, for the third trimester estimation. They found that the timing of the MIA affected the microglial responses. Injection of the serum to induce MIA, Poly (I:C), on day 15 gestation increased postnatal microglial cytokine expression in the offspring and altered the morphology of the microglia, while injection at day 12 of gestation did not. Matta et al. (2019) found that prolonged disruption to the complex cytokine networks resulted in the disruption of functional neuronal circuits, hindering their role in cognition and synapse regulation in mouse models of ASD.

For patients diagnosed with ASD, data mining from hospital records of patients is utilized to compare rates of exposure to environments that invoke inflammatory responses and diagnoses of ASD later in life. Chen et al. (2014) related atopy diagnosis before the age of three to an elevated risk of ASD diagnosis. Atopic diseases were categorized as asthma, atopic dermatitis, allergic rhinitis, or allergic conjunctivitis. They followed the children for ten years and found a dose-dependent correlation between atopy and ASD. The more atopic comorbidities present (defined as the presence of more than one atopic disease), the greater the risk of being diagnosed with ASD. It is important to note that there are many possible factors that contribute to inflammation related to ASD. These include MIA, environmental factors such as infections in-utero, toxin exposure, stress of the mother, and maternal obesity (Bilbo et al., 2018).

Conclusion and Future Directions

This literature review focused on the correlation between decreased proper synaptic pruning, increased distortion of morphology and activity of microglia and dendrites, and exposure to neuroinflammation. The proposed hypothesis was that the high density of dendritic spines in individuals with ASD is due to a lack of functional microglia cells in response to inflammation during early post-natal developmental years, when high rates of synaptic pruning occur.  The task was to understand the cause and timing of neuroinflammation and its relationship to microglia’s synaptic pruning, potentially leading to the development of preventative treatments for ASD.

Across the literature, it was found that increased presence of altered microglia and the resulting defective pruning of dendritic spines are a potential pathogenic mechanism for ASD in both patients and mouse models alike (Saitoh et al., 2021). The altered glial morphology and function, along with cytokine dysfunction, were evident in ASD mouse models. Studies using mouse models of ASD found very similar altered morphology and activity of microglia and cortical tissue in comparison to the post-mortem cortical sections of patients with ASD (Lee et al., 2017). Additionally, prolonged exposure to inflammation in patients diagnosed with ASD and mouse models of ASD correlated with an increased risk of ASD, along with ASD-like behavior in the mouse models.

A common theme in several research articles was exposure to inflammatory conditions. Research also showed a time-dependent (Ozaki et al., 2020) and dose-dependent (Chen et al., 2014) nature to the development of ASD in both mouse models and patients after exposure to inflammation. Research of exposure to inflammation in-utero showed similar results to the studies of post-natal inflammation exposure. For example, in human studies, exposure to inflammatory conditions in-utero due to infection (Bilbo et al., 2018) and exposure to inflammatory conditions due to atopic diseases (Chen et al., 2014) increased the risk of being diagnosed with ASD later in life. In mouse models of ASD, induction of inflammatory conditions in-utero (Xu et al., 2020) and post-natal exposure to inflammation due to induced asthma (Saitoh et al. 2021) alike resulted altered microglia and ASD-like behavior. This brings into play another area of research, the gut-brain axis. The gut-brain axis is a potential non-invasive therapeutic target for ASD, due to the symbiotic relationship between the microbiome and cognitive function (de Theije et al., 2014). The intestinal system and nervous system have many negative-feedback loops that are under hormonal, neural, and immunological control. Research suggests that disruption of this network contributes to the dysregulation of microglial cell function in patients with ASD. Patients with ASD have higher rates of intestinal dysregulation than the average population, further suggesting a correlation between the two conditions (Matta et al., 2019). It is likely that the inclusion of the gut-brain axis to ASD research could help expose the underlying pathophysiology behind ASD. Research is already being done to determine if a diet that promotes a diverse microbiome or a fecal transplant from a control patient induces changes in the microbiome that would mitigate the inflammatory response of the microglia and improve ASD-like behavior in ASD patients (Kang et al., 2017). The continuation of this research could determine if intervention at the level of the gut microbiome would mediate the inflammation that induces the dysfunctional microglia and consequent ASD symptoms, potentially introducing an effective therapeutic for ASD.

[+] References