Advances in the pathogenesis of Rett syndrome using cell models

Abstract Rett syndrome (RTT) is a progressive neurodevelopmental disorder that occurs mainly in girls with a range of typical symptoms of autism spectrum disorders. MeCP2 protein loss‐of‐function in neural lineage cells is the main cause of RTT pathogenicity. As it is still hard to understand the mechanism of RTT on the basis of only clinical patients or animal models, cell models cultured in vitro play indispensable roles. Here we reviewed the research progress in the pathogenesis of RTT at the cellular level, summarized the preclinical‐research‐related applications, and prospected potential future development.


| G ENER AL INFORMATI ON ON RT T
Mutations of the MECP2 gene lead to MeCP2 protein loss of function in part or whole, which affects the methylation binding ability and regulatory function on gene expression, resulting in the phenotypes of typical RTT. Most patients with MECP2 mutant RTT are female, with a prevalence of approximately 1/10 000-1/15 000. 1 Most of the symptoms of patients with RTT occur in the central nervous system, including smaller brain volume and thinner cortical layer, which specifically present as smaller cell bodies, reduced spinous process density and complexity, and significantly lower overall neuronal maturity. [10][11][12] These findings indicate that cellularlevel changes play an important role in RTT onset. We summarize the abnormal physiological processes, cell types, and pathological phenotypes affected by MECP2 mutations in Figure 1. In addition, deficiency of this protein outside the nervous system can lead to lesions in the corresponding organs, such as cardiac, liver, and digestive tract, etc., 13 indicating that mutations of MECP2 have complex functions throughout the body. Current research has mainly concerned damages in the nervous system.
As a transcriptional regulator, MeCP2 has a dual regulatory function, that is, transcriptional activation or inhibition. The severity of the RTT phenotype is related to the type of mutation. 14,15 Clinically, most mutation sites are located at the 2 functional domains of transcriptional repression domain (TRD) and methyl CpG binding domain (MBD), 16  assists in DNA bending and chromatin remodeling. [16][17][18][19][20]

| A B RIEF OVERVIE W OF THE RE S E ARCH PROG RE SS OF RT T ANIMAL MODEL S
Rodent and nonhuman primate models are commonly used to study the disease progress and pathogenic mechanism of RTT. In 2001, Mecp2-knockout mice were first reported, 21,22 which exhibit phenotypes resembling some of the symptoms of patients with RTT. In the conditional knockout mice, loss of Mecp2 in inhibitory neurons impaired the GABA signaling pathway, exhibiting autistic stereotypical behavior and severe phenotypes 23 ; loss of Mecp2 in cholinergic circuits of basal forebrain and the striatum recapitulated some phenotypes. 24 In addition, knockdown of Mecp2 in different brain regions of mice displayed different neuropathological phenotypes, suggesting a region-specific effect. 16 Mecp2-deficient rat models generated in 2016 showed Rett-like behavioral and motor deficits. 25,26 Subsequently, nonhuman primate models of RTT 27 were constructed in 2017. Monkey models showed unique advantages in mimicking abnormal phenotypes of RTT in advanced cognitive, social behavior, and movement activity. They were also used to monitor brain development by neuroimaging. 27 The abnormal development of white matter (WM) microstructure and network topological organization of monkey models may cause the RTT behavioral phenotypes. 28 The above models have made a great contribution in tracking the disease progression and understanding the phenotypes of RTT. However, for the studies aimed to elucidate the mechanisms in living cells, or to carry out functional verification and pathogenesis exploration more conveniently and comprehensively, cell models are essential.

| RE S E ARCH PROG RE SS ON THE PATHOG ENE S IS OF RT T US ING CELL MODEL S
Most cells used in laboratory are usually derived from patients with RTT, animal models, or gene-modified cells. The advent of drug studies based on induced pluripotent stem cells (iPSCs) is a milestone in nonclinical trials. Although only relatively few studies have used iPSCs derived from patients with RTT, there have been very important findings and research progress on pathogenesis and preclinical trials, as listed in Table 1.

| Research progress on RTT neurons
Neurons differentiated from iPSCs derived from patients with RTT show specific pathological phenotypes, that is, smaller neuron cell bodies, decreased synapses and spine density, and abnormal calcium F I G U R E 1 The effects of MECP2 mutations on RTT. Multiple abnormal biological processes were regulated by MECP2 mutations, which further lead to cellular physiological deficits in neurons and glial cells. Ohashi et al. 107 Female patients iPSCs 705delG, X487W The reduction in dendritic complexity of RTT neurons may be due to activation of the p53 pathway, or be associated with aging signaling and electrophysiological function, reflecting important changes in the morphological structures and functions of RTT. 29,30 Disruption of the excitatory/inhibitory activity balance between synapses in different brain regions and circuits leads to an imbalance in microenvironmental homeostasis, which may lead to abnormal brain firing, resulting in epilepsy or other symptoms. 31 Wild-type iPSC-derived neurons typically express high levels of synaptic adhesion molecule GluD1. 32 MeCP2 deficiency caused changes in the action potential of glutamate neurons and a decrease in the number of synapses on glutamate neurons, 33 implying that the fate of neural differentiation may shift to inhibitory neurons, which manifests itself in an increase in inhibitory synapses and a decrease in excitatory synaptic structures.
This result may be caused by the downregulated expression of neuron-specific membrane transporter K + /Cl + cotransporter (KCC2) mediated by RE1-silencing transcription factor (REST), a neuronal gene inhibitor in RTT, which is essential for maintaining excitatory balance in the brain. 34

TA B L E 1 (Continued)
By analyzing and comparing the different stages of neural differentiation of RTT-iPSCs, researchers revealed some changes and mechanisms at the molecular and cellular level. Transcriptome analysis showed that MeCP2 began to modulate before neural differentiation. 38 During subsequent differentiation, forebrain neurons derived from several human RTT stem cell lines showed a reduction in the expression level of cAMP-response element binding protein (CREB) and phosphorylated CREB, which could lead to functional defects in neurons. 39 In RTT human iPSCs, neural progenitor cells, and cortical neurons, the expression of genes related to the dysregulation of mTOR signaling pathway and the ubiquitin pathway alters the structure of neurons, leading to defects in cell structure. 40 Proteomic analysis revealed that both dendritic morphology and synaptogenesis-related proteins were altered during RTT iPSCderived neuronal progenitors, that is, in early neuronal differentiation. 41 Another study found that the LIN28A gene may participate in the regulation of neuronal differentiation in RTT-iPSCs. 42 Therefore, dysregulation of the expression of various genes and proteins during this early phase of neuronal differentiation may be an important reason for the progression of RTT.
In addition, MeCP2 also regulates microRNA (miRNA). The expression of miR-199 and miR-214 was found to be most significantly affected by the MeCP2 mutation. Restoration of miRNA expression in patients with RTT and MeCP2-deficient neural stem/precursor cells can relieve the pathological phenotype of RTT neurons. 43,44 These results suggest that MeCP2 has a wide range of roles that may not only alter the substance transport process involved in organelles, but also adversely affect the formation and/or maintenance of neural processes by influencing the transcription. Although the effects of MeCP2 on normal brain development is not fully understood, there is no doubt that the mutation of MeCP2 disrupts the expression regulation of a large number of genes and the homeostasis of their microenvironment, which is an important premise of the RTT neuropathological phenotype.

| Research progress on glial cells
Previous  found in both Mecp2-null mice and their primary glial cells after lipopolysaccharide treatment. 56 Likewise, persistent dysfunction of neurons or other glial cells also enhances the immune response of RTT microglia, which may further exacerbate the disease process. 57 Higher levels of glutamate were detected in RTT microglia-enriched conditioned medium, and addition of the medium to normal cultures also resulted in damage to dendrites and synapses in neurons. 58 Mecp2 deficiency leads to overexpression of glutamine transporter (SNAT1), resulting in the production of large amounts of glutamine in mitochondria for metabolism and the formation of glutamate, which may be responsible for mitochondrial dysfunction and neurotoxicity. 59 Other studies have shown that the involvement of miRNA in the regulation of the MECP2-STAT3 axis or the modification of MECP2 phosphorylation may also be the reason for the inflammatory response of microglia. 60,61 In mice, Mecp2 deficiency in the oligodendrocyte lineage also plays a unique role in the disease process of RTT. 62  The point of treatment now is to improve the growth and development of neurons or restore their damaged neurites and synapses.
Brain-derived neurotrophic factor (BDNF) is a neurotrophic factor that plays an important role in neuronal survival and plasticity, and its expression is also regulated by MeCP2. 72 The expression level was reduced in Mecp2-deficient male mice, and when a certain level of expression was restored, the symptoms and lifespan of the diseased mice could be reversed, 73 suggesting that treatments targeting the MeCP2-BDNF axis in RTT could alleviate some symptoms and are potential therapeutic options for RTT. Protein tyrosine phosphatase-1B (PTP1B) is a receptor for BDNF, and its pharmacological inhibition ameliorated the effects of MECP2 disruption in RTT mice. 74 When insulin-like growth factor 1 (IGF-1) and low concentrations of gentamicin were administered to RTT neurons, the morphology and function of damaged neurons could also be restored. 75 Histone deacetylase 6 (HDAC6)-selective inhibitors showed great application and therapeutic prospects to reverse the decreased microtubule acetylation in neurons of RTT. 76,77 Overexpression of L1 retrotransposon can partially restore neurite growth during RTT-iPSC differentiation. 78,79 MeCP2's regulation of the multi-subunit protein complex BLOC-1 may also be a therapeutic target for synaptic dysfunction. 80 The mutation impairs the neuronal AKT/mTOR pathway and mitochondrial function. 81 Administration of the above pathways may improve the pathological phenotype of RTT neurons.
Restoring normal expression of Mecp2 in the medial prefrontal cortex can improve behavioral deficits in mice. 82  Long-term culture requires the support of scaffolds, but their components are unclear; Difficult to directed differentiation; Less repeatability than 2D cells in MeCP2-related transcriptional regulation processes could also become therapeutic targets. 83

| SUMMARY AND PROS PEC TI ON
In-depth study of RTT and MeCP2 has given us an understanding of MeCP2's multifunction: widely involved in the transcription regulation of genes, self-translational modification in response to neuronal activity, and promotion of chromatin central aggregation, 91,92 etc., indicating its importance for individual neurodevelopment. Owing to the difficulties in pre-onset data and sample collections from patients with RTT, the detailed mechanism of RTT in the early stages of postnatal development is still unknown. The emergence of iPSCs has brought great application prospects. 93 To date, iPSCs provide a reusable, versatile, and consistent source from patients for in vitro studies of RTT. Although cell models are a powerful tool to investigate potential regulation mechanisms in RTT, neurodevelopment is a temporary and spatially related complex progress. Whether the results of in vitro and in vivo studies are highly consistent remains to be further demonstrated.
The understanding of RTT should not be limited to the severe consequences of its mutations; the function of MeCP2 in the nervous system and even in early ontogeny also needs to be understood. At present, 2D neural differentiation and 3D brain organoid culture technologies have been developed rapidly. [93][94][95][96][97][98] The combination of RTT-iPSCs with the above technologies can be better applied to further research on RTT pathogenesis, drug screening, gene repair, and cell therapy. Taken together, the rational use of cells and other models 99-102 for research will help us to understand the pathogenic mechanism and develop new treatments for RTT in the future.

AUTH O R CO NTR I B UTI O N S
Sijia Lu conceived and wrote the original draft of the manuscript.
Zhengbo Wang and Yongchang Chen revised the manuscript. All authors critically read and contributed to the manuscript, and approved its final version.

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest regarding the publication of this article.