Dysregulation of synaptic pruning as a possible link between intestinal microbiota dysbiosis and neuropsychiatric disorders

The prenatal and early postnatal stages represent a critical time window for human brain development. Interestingly, this window partly overlaps with the maturation of the intestinal flora (microbiota) that play a critical role in the bidirectional communication between the central and the enteric nervous systems (microbiota‐gut‐brain axis). The microbial composition has important influences on general health and the development of several organ systems, such as the gastrointestinal tract, the immune system, and also the brain. Clinical studies have shown that microbiota alterations are associated with a wide range of neuropsychiatric disorders including autism spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, and bipolar disorder. In this review, we dissect the link between these neuropsychiatric disorders and the intestinal microbiota by focusing on their effect on synaptic pruning, a vital process in the maturation and establishing efficient functioning of the brain. We discuss in detail how synaptic pruning is dysregulated differently in the aforementioned neuropsychiatric disorders and how it can be influenced by dysbiosis and/or changes in the intestinal microbiota composition. We also review that the improvement in the intestinal microbiota composition by a change in diet, probiotics, prebiotics, or fecal microbiota transplantation may play a role in improving neuropsychiatric functioning, which can be at least partly explained via the optimization of synaptic pruning and neuronal connections. Altogether, the demonstration of the microbiota's influence on brain function via microglial‐induced synaptic pruning addresses the possibility that the manipulation of microbiota‐immune crosstalk represents a promising strategy for treating neuropsychiatric disorders.


| INTRODUC TI ON
At the most fundamental level, brain function is based mainly on computations performed by synapses. Perturbations in physiological synaptic structure and function and dysregulated synaptic formation, elimination and plasticity have been hypothesized to underlie altered neuronal function in complex neuropsychiatric disorders, such as autism spectrum disorder (ASD) and schizophrenia (SZ) (Wang, Christian, Song, & Ming, 2018). At the early stages of life, synapse formation (synaptogenesis) exceeds elimination, yielding excessive synapses essential for the assembly of neural networks (Bruer, 1999). Subsequently, synaptic elimination/pruning outpaces synaptogenesis, providing selection and maturation of synapses and neural circuits from childhood through adolescence (Tang et al., 2014). Different patterns of dysregulated synaptic pruning have been linked to various neuropsychiatric phenotypes, confirming the importance of the balanced synaptic formation and pruning in normal brain function. Recent studies have pointed to a key role for microglia, the innate immune cells of the central nervous system (CNS), in synaptic pruning by purging the brain of infrequently used synapses (Weinhard et al., 2018).
In mammals, microglial activation and function during developmentally sensitive periods can be modulated by the microbiota (Erny et al., 2015), the different resident phyla and bacterial species in the gastrointestinal (GI) system (Dickerson, Severance, & Yolken, 2017;Fond et al., 2015;Nemani, Hosseini Ghomi, McCormick, & Fan, 2015). Many causes can alter the well-being of the microbiota including administration of antibiotics or non-steroidal antiinflammatory medicines, herbicides, ingredients present in food (sugar or gluten) or in water (chlorine) (Larroya-Garcia, Navas-Carrillo, & Orenes-Pinero, 2019). In turn, the imbalance in microbiota composition (dysbiosis) can affect the function of neuronal circuits via synaptic pruning alteration (Tognini, 2017). Indeed, the current data indicate that various neuropsychiatric disorders are associated with microbiota alterations (Cenit, Sanz, & Codoñer-Franch, 2017;Kim & Shin, 2018). Hence, a better understanding of the effect of intestinal microbiota dysbiosis on synaptic pruning can pave the way to enhance the treatment outcomes of neuropsychiatric disorders.

| SYNAP TOG ENE S IS AND SYNAP TIC PRUNING DURING B R AIN DE VELOPMENT
Synaptogenesis is a complex multifactorial developmental process which enables the formation of synapses between neurons. Synapse formation is essential for all nervous system functions including establishing neural circuits and ultimately expressing complex behavior (Hong & Park, 2016). Across mammalian species, neurons present at birth undergo a period of overproduction of their arborization and synaptic contacts to increase synaptic density (Semple, Blomgren, Gimlin, Ferriero, & Noble-Haeusslein, 2013). In humans, the thickness of the cortex typically increases in the first few years of life as a result of excessive synapse formation (Tau & Peterson, 2010), with different cortical regions showing their peaks of synapse formation periods (Huttenlocher & Dabholkar, 1997). For example, synaptic density in the primary visual cortex reaches its peak between the ages of 4 and 12 months (Tau & Peterson, 2010). Synaptogenesis in the prefrontal cortex that requires remodeling to achieve fully mature and complex behavior begins about the same time as in the visual cortex, but it continues to reach its peak through the second and third year of life (Huttenlocher, de Courten, Garey, & Van der Loos, 1982;Huttenlocher & Dabholkar, 1997;Kostović, Judaš, Petanjek, & Šimić, 1995;Lenroot & Giedd, 2006).
Later in life, at the time of early adolescence, cortical thickness decreases by pruning weak and redundant synaptic connections, and strengthening the remaining synapses (Sowell, Thompson, & Toga, 2004;Wang et al., 2018). Synaptic pruning is a crucial process to enhance neuronal transmission and to establish the finely tuned circuitry by eliminating ineffective synapses and strengthening the vital neuronal connections, which allows for more efficient processing of adult cognition. In mammals, axonal and dendritic processes constitute approximately 60% of cortical volume (Tau & Peterson, 2010), and pruning of these processes may represent the source of cortical thinning (Paus, Keshavan, & Giedd, 2008).
Synaptic pruning begins in late gestation and becomes increasingly active postnatally (Tau & Peterson, 2010). The time course for pruning differs across brain regions, with sensory and motor cortices undergoing dramatic fine-tuning after birth, followed by association cortices and the corpus callosum, and later by regions that subserve higher cognitive functions (Levitt, 2003). In early childhood (2 and 7 years of age), neuronal density in layer III of the prefrontal cortex decreases from 55% to approximately 10% above adult levels (Huttenlocher, 1979). During later childhood (7-15 years of age), synaptic density in the frontal cortex decreases by approximately 40% (Lidow, Goldman-Rakic, Gallager, & Rakic, 1991). These synaptic changes occur in the absence of any significant neuronal loss and are accompanied by a reduction in the expression of genes involved in axonal and synaptic functions (Colantuoni et al., 2011).
The continuous cortical thinning via synaptic pruning throughout

Significance
The association between the intestinal microbiota and brain function is not fully understood. In this review, we propose synaptic pruning dysregulation as a possible link between microbiota dysbiosis and neuropsychiatric disorders including autism, schizophrenia, bipolar and attention deficit hyperactivity disorders. To this end, the alleviation of neuropsychiatric symptoms via improving the intestinal microbiota composition might be partly explained by the modulation of microglial function, leading to a modification in neuronal connections. Therefore, the microglial activity and its effect on synaptic pruning may be good markers for testing the efficacy of probiotics and prebiotics as supportive therapeutic approaches for neuropsychiatric disorders.
zoaffective disorder and BD, whereas general grey matter loss was observed for SZ (Gogtay, 2008;Mattai et al., 2011). In contrast, acceleration in brain growth of all regions except occipital grey matter in the early years of patients with ASD was observed (Schumann et al., 2010).

| Autism spectrum disorder
ASD is a complex neurodevelopmental disorder with a strong genetic component, which refers to a constellation of clinical conditions with two main phenotypic characteristics: impairment in social communication and patterns of repetitive restrictive behavior (Berkel et al., 2018). Remarkably, the time when ASD symptoms appear or become more apparent, at around 2-3 years old, coincides with the window for the initial generalized synaptic pruning event ( Figure 1).

F I G U R E 1
The onset of neuropsychiatric disorders symptoms in relation to synaptic pruning and microbiota development in the intestine. Autism spectrum disorder and attention deficit hyperactivity disorder symptoms appear concurrent with the start of the synaptic pruning process. In contrast, schizophrenia and bipolar disorder symptoms are parallel with the end of the synaptic pruning process. The microbiota composition is continuously changed throughout the lifespan with major changes concurrent with the synaptic pruning process. The association of microbiota alteration to neuropsychiatric disorders via synaptic pruning is an important example of the signaling between the central and the enteric nervous systems (microbiota-gut-brain axis) [Color figure can be viewed at wileyonlinelibrary.com] Some children diagnosed with ASD feature a significant excess in brain size and weight in the first year of life due to a high acceleration rate of brain growth (Courchesne, Carper, & Akshoomoff, 2003). This involves an enlargement in different brain regions with an expansion in both grey and white matter volume as observed by magnetic resonance imaging (MRI) studies (Courchesne et al., 2003;Gaffney, Kuperman, Tsai, & Minchin, 1989;Hardan, Muddasani, Vemulapalli, Keshavan, & Minshew, 2006;Piven et al., 1992Piven et al., , 1995. Moreover, diffusion tensor images from the brains of ASD children have shown an increase in axons and myelination between neighboring areas of the brain compared with more distal connections, suggesting an increase in connectivity . On the microstructural scale, synaptic densities of pyramidal neurons in the temporal lobe are observed in postmortem studies to be increased in the brains of children and adults with ASD (Hutsler & Zhang, 2010;Tang et al., 2014). Moreover, the reduction in cortical spine density that is observed in brains of typically developing adolescents is diminished in individuals with ASD (Tang et al., 2014), suggesting a deficit in pruning, at least at this later age. Besides, mice carrying rare, penetrant mutations that are found in individuals with ASD show elevated spine densities in the temporal cortex and cerebellum (Kim et al., 2017;Piochon et al., 2014;Tang et al., 2014) and deficient adolescent pruning (Tang et al., 2014). Altogether, children with ASD are believed to have excess synapses and synaptic connections in the brain due to deficits in synaptic pruning during early brain development (Tang et al., 2014), which may account for the abnormal patterns of brain connectivity in ASD (Belmonte et al., 2004).
The idea that enhanced connectivity may have disadvantageous effects on brain function and cognition was supported by mouse models. Mice with excessive synaptic connections due to a failure in synaptic pruning were primarily able to learn spatial locations but unable to re-learn new locations (Afroz, Parato, Shen, & Smith, 2016). This indicates that too many brain connections may put limitations on the learning potential. Moreover, by impairing synaptic pruning in mice, they exhibited ASD-like phenotypes including social interaction deficits and repetitive behavior (Fernandez de Cossio, Guzman, van der Veldt, & Luheshi, 2017;Kim et al., 2017). A similar process to the methodology in the aforementioned mouse studies is suggested to play a role in the impairment of synaptic pruning in humans diagnosed with ASD. Therefore, there is evidence from mouse studies to support the claim that children diagnosed with ASD are believed to have increased synaptic connections in the brain due to deficits in synaptic pruning during early brain development (Tang et al., 2014).
However, caution should be taken when translating mice finding to humans. Although a large variety of mouse behavioral tests are currently used and showed considerable face validity in testing the ASD core symptoms, the lack of a "human-specific" read-out resulting from complex gene-environment interactions occurring during early postnatal stages and adolescence is the main limitation (Pasciuto et al., 2015). Therefore, the validity of drawing solid conclusions from mice to humans is still limited.

| Attention deficit hyperactivity disorder
ADHD is a complex brain disorder marked by an ongoing pattern of inattention, hyperactivity, and impulsivity that significantly impacts many aspects of behavior as well as cognitive performance (Singh, Yeh, Verma, & Das, 2015). Structural MRI studies on people with ADHD have revealed a subtle but significant grey and white matter loss , which was not progressive (Castellanos et al., 2002). Moreover, cross-sectional MRI studies have shown a reduction in the size of cortico-striatal brain regions that are known to develop late in adolescence (Berger, Slobodin, Aboud, Melamed, & Cassuto, 2013;Krain & Castellanos, 2006), in the volumes of the right and left inferior-posterior cerebellar lobes (Mackie et al., 2007), and in the thickness of cerebellar brain region . However, the structural development of almost all cortical regions in ADHD children was similar to non-psychiatric control subjects . To this end, it is suggested that ADHD children suffer from a maturational delay due to a lag in synaptic pruning (Rubia, 2007;Shaw et al., 2007;Vaidya, 2012), which is supported by the fact that 80% of children grow out of ADHD in adulthood (Faraone et al., 2000). The cortical maturation delay in ADHD was most prominent in the lateral prefrontal cortex, which supports the ability to suppress inappropriate responses and thoughts, executive control of attention, evaluation of reward contingencies, and working memory (Shaw et al., 2006. In contrast, only the motor cortex had a maturation peak 4 months ahead in children diagnosed with ADHD compared to control children, which may account for the impulsivity in people with ADHD .

| Schizophrenia
SZ is a complex, mental disorder characterized by an array of symptoms including delusions, hallucinations, disorganized speech or behavior, lack of motivation, and impaired cognitive ability (Patel, Cherian, Gohil, & Atkinson, 2014). Unlike ASD, the onset of the symptoms of SZ typically occurs between the ages of 15 and 25 and coincides with later stages of synaptic pruning in the adolescent prefrontal cortex (Selemon & Zecevic, 2015).
The association between synaptic pruning dysregulation in adolescence and SZ was first hypothesized by Feinberg in 1982(Feinberg, 1982. This hypothesis was revisited in 1994 by analyzing postmortem brains which showed that excessive synaptic pruning in the excitatory glutamatergic neurons in the prefrontal cortico-cortical and -subcortical areas was seen in SZ neuropathology (Keshavan, Anderson, & Pettegrew, 1994). Synaptic pruning ends at the age of onset for SZ  Figure 1), which further links the dysregulation of synaptic pruning to the SZ pathophysiology. In one study, the grey matter volume, as a measure of the extent of synaptic pruning (Sowell, Thompson, Tessner, & Toga, 2001), was reduced in both males and females diagnosed with SZ and correlated with reduced cognitive performance (Gur, Turetsky, Bilker, & Gur, 1999). This finding was replicated in another study which showed a fourfold excess of permanent grey matter loss in SZ compared to control subjects evaluated prospectively over 5 years . A meta-analysis study including over 18,000 subjects revealed that intracranial and total brain volume in patients with SZ was significantly decreased (Haijma et al., 2013).
Patients with SZ revealed a specifically reduced grey matter volume in the prefrontal cortex , which may account for their disturbed behavioral inhibition.
Neuropathologic studies indicated that the grey matter reduction in the brains of individuals with SZ is due to cortical thinning, with the greatest severity in the frontal lobes due to primarily shrinkage in neuropil combined with a decrease in neuronal size (Andreasen et al., 2011;Berdenis van Berlekom et al., 2019;Osimo, Beck, Reis Marques, & Howes, 2019). Moreover, the locus of complement factor C4A that, among other functions, regulates synaptic pruning is one of the genetic loci significantly associated with SZ (Sekar et al., 2016). C4A has also been shown to be expressed during the postnatal neurodevelopmental stage in proportion to the allelic risk association with SZ (Sekar et al., 2016). One of the functions of activated complement is the opsonization of synapses to facilitate phagocytosis by microglia, hence leading to enhanced pruning. In addition, neuroimaging studies confirmed the enhanced synaptic pruning in individuals with clinical high risk for developing SZ (Cannon et al., 2015). Based on the aforementioned studies, individuals with SZ are suggested to have fewer synapses due to excessive synaptic pruning and suboptimal fine-tuning of neural circuits mediating motor,

| Bipolar disorder
BD is a chronic mental health condition that is characterized by depressive and (hypo)manic mood episodes, as well as an impairment in cognitive ability (Gondalia, Parkinson, Stough, & Scholey, 2019). In BD, serial MRI scanning of adolescent patients has shown significant grey matter reduction in some areas of the brain including the bilateral anterior and subgenual cingulate cortex (Gogtay et al., 2007). In 2012, another study has shown that individuals diagnosed with BD suffer from a disruption of the emotional control networks during development linked with synaptic pruning dysfunction, which leads to abnormal ventral prefrontal-limbic modulation causing the onset of mania (Strakowski et al., 2012). In addition, the age of onset of BD and the monoaminergic synaptic density measured with PET measures were found to be interrelated (Zubieta et al., 1998).

| THE INFLUEN CE OF MI CROG LIA AC TIVATION ON SYNAP TIC PRUNING AND NEURONAL FUN C TION
Microglia are the innate immune cells of the CNS that account for 10%-15% of all cells found within the brain (Lawson, Perry, & Gordon, 1992). Several studies have identified a set of critical signaling pathways between microglia and neurons (for a review, see [Neniskyte & Gross, 2017]). Importantly, microglia have been shown to play a major role in the synaptic pruning process by purging the brain of infrequently used synapses (Boksa, 2012;Paolicelli et al., 2011;Schafer & Stevens, 2013;Stephan, Barres, & Stevens, 2012;Trapp et al., 2007). The first indication that microglia are involved in synaptic pruning was demonstrated by a study that showed that the large-scale axonal remodeling in embryonic and early postnatal development in cats was accompanied by a phagocytic activity of microglia and astrocytes, which were suggested to contribute to axon elimination (Berbel & Innocenti, 1988). Other studies in diverse model systems and circuits, ranging from peripheral synapses in the neuromuscular junctions to central synapses in the cortex, hippocampus, thalamus, and cerebellum strengthened the role of microglia in synaptic fine-tuning (Darabid, Perez-Gonzalez, & Robitaille, 2014;Hoshiko, Arnoux, Avignone, Yamamoto, & Audinat, 2012;Ichikawa et al., 2011;Paolicelli et al., 2011;Sasaki et al., 2014aSasaki et al., , 2014bSchafer et al., 2012;Zhan et al., 2014).
Many factors including genetic predisposition, head trauma, and infection can cause microglial overactivation, hence affecting the normal brain function by influencing synaptic plasticity and synaptic elimination (Boulanger, 2009;Goshen et al., 2007;Khairova, Machado-Vieira, Du, & Manji, 2009;Lui et al., 2016;Murray & Lynch, 1998). One genetic predisposition is for the complement factor C4A gene which had a strong risk association with SZ. The phenotype of C4A shows increased synapse engulfment and thus excessive synaptic pruning (Sellgren et al., 2019;Wang, Zhang, & Gage, 2019), and the "omic" studies revealed a link between C4A and SZ pathogenesis (Birnbaum & Weinberger, 2019;van Mierlo, Schot, Boks, & de Witte, 2019). A recent meta-analysis on postmortem brain studies for the immune involvement in the pathogenesis of SZ showed a significant increase in TA B L E 1 Clinical studies on intestinal microbiota dysbiosis in patients with autism spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, or bipolar disorder
• Healthy sibling group harbored the lowest levels of Bacteroides of all the subject groups.
• GI problems were associated with high levels of Clostridia.
• An association between high Clostridial counts and individuals consuming probiotics NA Sandler et al.

TA B L E 1 (Continued)
the density of microglia, mostly in the temporal cortex, and on the molecular level an overall increase in expression of pro-inflammatory genes on both transcript and protein levels (van Kesteren et al., 2017) together with an increase in microglial markers (Barichello, Simoes, Quevedo, & Zhang, 2019). Interestingly, in live human subjects with SZ or BD, C4A mRNA expression in peripheral blood mononuclear cells predicts the presence and severity of delusions (Melbourne, Rosen, Feiner, & Sharma, 2018).
For ASD, the increase in brain volume in some children with ASD manifestations is frequently associated with excessive activation of microglia in regions with an overabundance of cortical neurons and connections (Morgan et al., 2010;Redcay & Courchesne, 2005;Sacco, Gabriele, & Persico, 2015;Suzuki et al., 2013;Walker et al., 2012).
Genes associated with the functioning of microglia have shown higher expression levels in brain samples from patients with ASD in comparison to controls , and gene co-expression network analysis of postmortem ASD brain tissue identified an upregulation of glial markers in the cortex, along with a downregulation of a module containing synaptic genes, compared with typically developing individuals (Voineagu et al., 2011). Moreover, genes that regulate the development of microglia are more activated in males (Werling, Parikshak, & Geschwind, 2016) who suffer from ASD more frequently compared to females with a ratio of 4:1 (Fombonne, 2009;Gillberg, Cederlund, Lamberg, & Zeijlon, 2006). In addition, the DNA methylation of genes known to be the modulators of the microglial activity or implicated in synaptic pruning has been found to be dysregulated in patients with ASD (Nardone et al., 2014;Prinz & Priller, 2014).
Autophagy also appears to be an essential component of microglia-

| THE LINK B E T WEEN INTE S TINAL MI CROB I OTA DYS B I OS IS AND MI CROG LIAL DYS FUN C TION IN NEUROPSYCHIATRIC DISORDER S
The bidirectional communication between microbiota, the different phyla and bacterial species in the GI system, and the brain has drawn much attention in recent years. Several studies in mice showed significant effects of the different microbiota compositions on early life control of emotions like anxiety, motor activity, and cognitive functions (Clarke et al., 2013;Desbonnet, Clarke, Shanahan, Dinan, & Cryan, 2014;Diaz Heijtz, 2016;Neufeld, Kang, Bienenstock, & Foster, 2011), confirming a functional connection between the microbiota and brain. Studies mostly from the microbiota-devoid germ-free mice or mice treated with broadspectrum antibiotics have shown that specific microbiota can impact brain physiology and neurochemistry and exhibit structural changes in the brain (Dinan & Cryan, 2017;Fung, Olson, & Hsiao, 2017;Martin & Mayer, 2017;Principi & Esposito, 2016;Zhang et al., 2015) and neurological deficiencies in learning, memory, recognition, and emotional behaviors (Foster, Rinaman, & Cryan, 2017;Gareau et al., 2011;Smith, 2015), along with less social behaviors (Mayer, Tillisch, & Gupta, 2015;Schumann & Amaral, 2006;Vuong, Yano, Fung, & Hsiao, 2017).
In the general population, it is believed that brain development is influenced by the intestinal microbiota via several immunological and signaling pathways (for a review, see [Ma et al., 2019]).
Moreover, the intestinal microbiota can modulate neurogenesis in the brain as demonstrated by the promotion of fetal neural development by some regulators from gut bacteria, which have a potential impact on cognitive function during adulthood (Humann et al., 2016;Rolls et al., 2007). The blood-brain barrier and vagus nerve actively participate in the bidirectional interactions between the intestinal microbiota and brain to maintain their homeostasis (Bonaz, Sinniger, & Pellissier, 2017;Braniste et al., 2014;Forsythe, Bienenstock, & Kunze, 2014).
Recently, the microbiota have shown an impact on the properties and function of microglia. For instance, with the absence of microbiota, microglia in germ-free mice displayed alteration in their morphological characteristics and gene expression profiles, accompanied by inhibition in their maturation state in the brain cortex (Erny et al., 2015). This can indicate that the intestinal microbiota contribute directly to the maturation progress of naïve microglia (Ma et al., 2019).
In another study, it has been shown that microglia respond to microbiota change in a sex-and time-dependent manner from prenatal stages (Thion et al., 2018). In a very recent study, the manipulation of microbiota in antibiotic-treated or germ-free adult mice resulted in significant deficits in fear extinction learning combined by an immature state and a change in gene expression in microglia (Chu et al., 2019). These microglial differentially expressed genes were found to be enriched in pathways related to synapse organization and synapse assembly, suggesting that deliberate manipulation of the microbiota may alter microglia-mediated synaptic pruning and disrupt dendritic spine remodeling, causing behavioral abnormalities. By recolonizing germ-free mice with a complete microbiota from healthy control mice immediately after birth, but not after weaning, the extinction learning ability was restored, indicating that the extinction learning and learning-related plasticity require microbiota-derived signals during a critical developmental period before weaning ( In humans, the intestinal microbiota have been shown to modulate the microglial activation and function during developmentally sensitive periods (Erny et al., 2015). The change in microglial activity can have a direct effect on neuronal circuits function via synaptic pruning alteration (Tognini, 2017). It is probably not a coincidence that synaptic pruning occurs during the same time as the maturation of the intestinal microbiota (Figure 1; Agans et al., 2011). Indeed, in a recent study in rats using diffusion tensor imaging, the intestinal microbiota have shown an association with structure-specific changes in white matter architecture in the brain via modulation of synaptic pruning (Ong et al., 2018), which can influence the brain development and function. Moreover, several studies found evidence that the development of the early postnatal nervous system and brain plasticity is influenced by the intestinal microbial status Yolken & Dickerson, 2017; for a total summary of the aforementioned neuropsychiatric disorders, see Table 1).

| IMPROVING THE INTE S TINAL MI CROB I OTA COMP OS ITI ON A S A SUPP ORTIVE THER APEUTIC APPROACH FOR NEUROPSYCHIATRIC DISORDER S
Since the intestinal microbiota have been associated with microglial modulation and neuronal development, the modification of the microbiota composition represents a promising therapeutic target for patients with neuropsychiatric disorders (Genedi, Janmaat,

| Dietary intervention
Rapid changes in the microbiota composition can be driven by a change in the diet (David et al., 2014;Turnbaugh et al., 2009;Wu et al., 2011), mainly by altering the quality and quantity of dietary fat, fibers, and carbohydrates (Fava et al., 2013;Flint, Duncan, Scott, & Louis, 2007;Sonnenburg et al., 2010). Therefore, modifying the diet may prove an easy way to improve the condition and  anti-inflammatory nutrients can be used as an adjunct to antipsychotic medication in patients with SZ (for a meta-analysis, see [Arroll, Wilder, & Neil, 2014]; Table 2 Table 2).

| Probiotics and prebiotics
Another way to modify the intestinal microbiota composition is mediated by the supplementation of probiotics, the living microorganisms which can provide a benefit to the host when administered in adequate amounts (Butel, 2014). The main bacterial genera used as probiotics in both animal and human studies are the Lactobacillus and Bifidobacterium genera (Genedi et al., 2019). Probiotics have been shown to boost the brain-derived neurotrophic factor (BDNF; Jeong, Kim, Hwang, Han, & Kim, 2016;Ranuh et al., 2019) that promotes the survival of existent neurons and enhance the neurogenesis (Liu & Nusslock, 2018;Numakawa, Odaka, & Adachi, 2017, 2018Scharfman et al., 2005), thus playing an essential role in the normal neurological development (Larroya-Garcia et al., 2019).
Moreover, two studies were published targeting individuals with BD (Dickerson et al., 2018;Reininghaus et al., 2018; for an overview of clinical studies with probiotic intervention in SZ and BD, see [Genedi et al., 2019]; for a summary, see Table 3).
In different mouse models of ASD, autistic-like symptoms were improved by the administration of probiotics (Buffington et al., 2016;Hsiao et al., 2013;Sgritta et al., Table 3). In the field, 95 out of 500 physicians reported using probiotics in treating children with ASD (Golnik & Ireland, 2009).
In individuals with ADHD, the probiotic supplementation combined with additional nutritional supplements were shown to be just as effective at improving symptoms as treatment with methylphenidate (Harding, Judah, & Gant, 2003). Another randomized clinical trial supports the beneficial role of probiotics in alleviating ADHD symptoms (Partty, Kalliomaki, Wacklin, Salminen, & Isolauri, 2015; for a summary of clinical studies, see Table 3). Nowadays, there are other ongoing clinical studies to test the efficacy of probiotic on improving ADHD symptoms (for more information, check https://clini caltr ials.gov/ct2/resul ts?cond=Adhd&term=probi otic&cntry =&state =&city=&dist=).
The use of prebiotics, the non-digestible plant-based carbohydrates that serve as nutrition for resident bacteria, increases the beneficial microbiota and attenuates stress behaviors in rodents (Burokas et al., 2017;Mika et al., 2017). In humans, a randomized trial revealed an improvement in the emotional affect and modulation of stress response following galacto-oligosaccharide supplementation in healthy volunteers (Schmidt et al., 2015). Combining diet restriction and a 6-week Bimuno ® galacto-oligosaccharide (B-GOS ® ) prebiotic intervention in autistic children have revealed significantly lower scores of abdominal pain and improvement on bowel movement in addition to observed improvements in social behavior (Grimaldi et al., 2018).

| Fecal microbiota transplantation
Although not as widely used as diet, probiotics and prebiotics, fecal microbiota transplantation (FMT) is another procedure to modify the dysbiotic intestinal microbiota toward a balanced one (Evrensel & Ceylan, 2016). A large number of commensal microbes can be introduced orally or through enemas or colonoscopy into recipient patients. FMT is commonly used in the treatment of GI diseases such as Clostridium difficile infection, Crohn's disease, and ulcerative colitis (Anderson, Edney, & Whelan, 2012;Aroniadis & Brandt, 2013;Xu et al., 2015). A study reported that FMT from depressed individuals causes depression-like behaviors in mice (Gareau et al., 2011;Huo et al., 2017). Moreover, obese-type intestinal microbiota induce neurobehavioral changes in the absence of obesity which is linked to the prevalence of mental illness, particularly depression and dementia (Bruce-Keller et al., 2015). According to a preliminary study including six anxiety patients aged between 36 and 83, FMT has shown a 70% improvement in anxiety (Zamudio-Tiburcio et al., 2017). In an open-label study of FMT combined with antibiotics, a bowel cleanse and a stomach-acid suppressant, significant improvements in GI symptoms, ASD-related symptoms, and intestinal microbiota were observed (Kang et al., 2017). The follow-up on the same participants after 2 years of treatment revealed maintenance of the improvements in GI symptoms and ASD-related symptoms with significant increases in bacterial diversity and relative abundances of Bifidobacteria and Prevotella . Although these results are promising, caution needs to be taken as this trial was not placebo-controlled, blinded, nor randomized; and further trials are • ADI-R, ABC-T, CBCL, SRS, and SNAP-IV-T questionnaires • No statistically significant differences in behavioral scores were detected between probiotics and placebo control groups • Post hoc subgroup analysis showed a small but statistically significant improvement in SNAP-IV-T total score (p = 0.02) and opposition/defiance subscale score (p = 0.03) in probiotics groups size, decreased process length, and increased major projection along with increased activated microglia, leading to synaptic stripping, impaired hippocampal plasticity, decreased dendritic spine density, as well as decreased synaptic protein such as postsynaptic density protein 95, synaptophysin, and spinophilin along with cognitive impairment (Bocarsly et al., 2015;Chunchai et al., 2018;Hao, Dey, Yu, & Stranahan, 2016). The morphological abnormalities in microglia caused by high-fat diet consumption were rescued by probiotic and prebiotic supplementation concurrent with amelioration in microglial activation and cognitive function (Chunchai et al., 2018).
In summary, changing the composition of the intestinal microbiota via diet, probiotics, prebiotics, and FMT showed a trend in improving the psychiatric conditions and/or their concurrent GI complaints. To this end, we hypothesize that the effectiveness of these dietary and probiotic interventions in improving symptoms in the aforementioned studies can at least partly be explained via interaction with microglia function and perhaps modifying synaptic pruning and neuronal connections, leading to a change in synaptic density.
For these reasons, these interventions could form a beneficial supportive treatment in neuropsychiatric disorders.

| CON CLUS ION
Different patterns of synaptic pruning in neuropsychiatric disorders may suggest a common pathogenic pathway (de Silva, 2018).
Modification of synaptic pruning can be a promising target for treating or preventing these neuropsychiatric symptoms. As the intestinal microbiota may have an influence on synaptic pruning via regulation of microglial activation, the optimization of the microbiota composition might be an easy way to modify the disturbed synaptic pruning in neuropsychiatric disorders during specific developmental stages, which, in turn, can alleviate some symptoms ( Figure 2

TA B L E 3 (Continued)
therapeutic interventions on the cortical thickness, as an indicator of synaptic pruning, should be tested by means of MRI. Moreover, the microglial activation and microglial gene expression assessed in cerebrospinal fluid may be used as markers for testing the efficacy of probiotic and prebiotic administration in different neuropsychiatric disorders and for determining the optimum duration of treatment.

K E Y P OINTS
• Normal brain function is associated with an initial formation of excessive synapses that have to be removed in a controlled and timely manner in a process called synaptic pruning.
• Various patterns of synaptic pruning dysregulation play an important role in the development of different neuropsychiatric disorders.
• The intestinal microbiota may affect synaptic pruning through microglial activation.
• Current clinical research on using probiotic and prebiotic supplementations in addition to changing the diet and fecal microbiota transplantation demonstrate biological and clinical efficacy in ASD, ADHD, SZ, and BD, which might be partly explained by a modification of synaptic pruning.

DECL AR ATION OF TR ANS PAREN C Y
The authors, reviewers, and editors affirm that in accordance with the policies set by the Journal of Neuroscience Research, this manuscript presents an accurate and transparent account of the study being reported and that all critical details describing the methods and results are present.

ACK N OWLED G M ENTS
We thank Shaimaa Madbouly, Abdelrahman Darwish and Dr. Rolf Sprengel for their assistance and support.

F I G U R E 2
Hypothesis model of the effect of the intestinal microbiota modulation on alleviating neuropsychiatric symptoms. The dysregulation of intestinal microbiota is associated with a dysregulation in microglial activity, causing an imbalance in the synaptic pruning process. Different patterns of dysregulated synaptic pruning have been linked to various neuropsychiatric phenotypes including autism spectrum disorder, attention deficit hyperactivity disorder, schizophrenia, and bipolar disorder. The modulation of the intestinal microbiota via diet change, probiotic, prebiotic, and fecal microbiota transplantation might balance the microglial activity and synaptic pruning, leading to normal brain function and alleviation of neuropsychiatric symptoms [Color figure can be viewed at wileyonlinelibrary.com]