Altered meningeal immunity contributing to the autism-like behavior of BTBR T⁺Itpr3^tf/J mice

Abstract

Autism spectrum disorder (ASD) is a complicated neurodevelopmental disorder, which is categorized by deficiency of social contact and communication, and stereotyped forms of performance. Meningeal immunity conditions the immune reflection and immune defense in the meningeal area involving meningeal lymphatic organization, glymphatic structure, immune cells, and cytokines. The development of meningeal immunity dysfunction might be the leading cause for many neural diseases including ASD. The inbred mouse strain BTBRT + Itpr3tf/J (BTBR) shows multiple ASD-like behavioral phenotypes, thus making this strain a widely used animal model for ASD. In our previous study, we reported an altered peripheral immune profile in BTBR mice. Herein, we are investigating immunological and neural interactions associated with the aberrant behavior of BTBR mice. BTBR mice have an increased level of immune cell deposition in the meninges along with a higher level of CD4⁺ T cells expressing CD25 and of B and myeloid cells expressing more MHCII than C57BL/6 (B6) mice, which have normal behaviors. BTBR mice also have higher levels of autoantibodies to dsDNA, Aquaporin-4, NMDAR1, Pentraxin/SAP and Caspr2 than B6 mice, which may affect neural functions. Interestingly, the T regulatory (Treg) cell population and their function was significantly reduced in the meninges and brain draining lymph nodes, which may explain the increased level of activated B and T cells in the meninges of BTBR mice. A low level of Treg cells, less IL-10 production by Treg, and activated T and B cells in meninges together with higher autoantibody levels might contribute to the development of autism-like behavior through neuroinflammation, which is known to be increased in BTBR mice.

Highlights

• •BTBR mice have higher level of immune cell meningeal deposition compared to C57BL/6 (B6) mice.
• •Meningeal T cells and B cells of BTBR mice express a higher level of CD25 and MHCII, respectively, than B6 mice.
• •BTBR mice have a higher level of serum autoantibodies to dsDNA, Aquaporin-4, NMDAR1, Pentraxin/SAP and Caspr2 than B6 mice.
• •BTBR mice have reduced level of T regulatory (Treg) cells in the meninges and brain draining lymph nodes and produce less IL-10.
• •Fewer Treg cells, more activated meningeal T and B cells, and higher autoantibody levels contribute to the autism-like development of BTBR mice.

1. Introduction

Autism spectrum disorder (ASD) is characterized by the existence of numerous different indicators such as verbal impairment, deficiency in social interaction, and monotonous pattern of behavioral. ASD is a lifelong neurodevelopmental condition with a strong involvement of dysfunctional or altered immune system. Currently, in USA, 1 out of 68 children aged 8 years have been diagnosed with ASD (Christensen et al., 2016; Eshraghi et al., 2018; Gładysz et al., 2018). The pathogenesis/progression of ASD involves a dysfunctional immune system including activation of both innate and adaptive immune cells. Dysregulated T cell subsets such as Th1, Th2, Th17 and Treg have been associated with ASD (Ahmad et al., 2017; Nadeem et al., 2020; Shmarina et al., 2020; Wong and Hoeffer, 2018). Increased levels of antinuclear and brain-specific antibodies have been described in children with autism (Edmiston et al., 2017; Rossignol and Frye, 2021). Although our prior studies also suggested an autoimmune profile in a mouse model of ASD (the BTBR T ⁺ Itpr3^tf/J (BTBR) strain) (Heo et al., 2011; Zhang et al., 2013; Uddin et al., 2020a, 2020b), we investigated a closer connection of the autoimmunity to central nervous system (CNS) involvement with inclusion of meningeal and autoantibody (autoAb) analysis.

Meningeal immunity is important for neuronal homeostasis and for neuronal activity through neuro-modulatory cytokines affecting neuronal signaling, animal behavior, senses, and thought (Alves de Lima et al., 2020a, 2020b; Rua and McGavern, 2018). The meninges neighboring the brain are occupied by a diversity of immune cell types, which not only offer immune observation but also affect brain function (Kipnis, 2016). Recently, it has been reported that the meningeal lymphatic system can regulate neuronal lymphatic drainage and neuroinflammation (Louveau et al., 2018). Meningeal inflammation caused by various agents can influence neurological disorders and play a key role in governing immunity in the central nervous system (Da Mesquita et al., 2018). Recent investigation of COVID-19 has been suggesting neural malfunctions, which further connects neuroimmune and vascular functions (Spudich and Nath, 2022; Satarker and Nampoothiri, 2020; Gonçalves de Andrade et al., 2021).

The meningeal lymphatic system is connected to the peripheral space and could sample and drain T cells, B cells, myeloid cells, and cerebrospinal fluid (CSF) contents directly from the deep cervical lymph nodes (Aspelund et al., 2015; Louveau, 2018). Meningeal lymphatic system ensures metabolic homeostasis between the parenchyma and peripheral tissue and plays a role in regulating immune surveillance and immune responses in CNS (Hu et al., 2020; Mestre et al., 2020). Several brain functions such as spatial learning, short-term memory sensory responses and hippocampal neurogenesis are controlled by meningeal T cells through the secretion of cytokines. (Brombacher et al., 2017; Radjavi et al., 2014; Ribeiro et al., 2019; Oetjen et al., 2017). IL-4 and IL-13 secreted by CD4⁺T cells can promote astrocytes to express brain derived neurotrophic factor (BDNF) (Brombacher et al., 2017; Ron-Harel et al., 2011), and IL-4 stimulates microglia to produce BDNF, IGF-1, and TGF-β to affect neuronal activity (Fenn et al., 2014; Zhao et al., 2015). Additionally, meningeal macrophages are reactive to the condition of the surrounding situation and can control the immune responses by their plasticity of anti- or pro-inflammatory phenotypes (Brendecke and Prinz, 2015). Meningeal dendritic cells (DCs) can sense and carry any antigens (Ags) to peripheral T cells (Pashenkov et al., 2003). In addition, meningeal DCs might also induce peripheral tolerance by restraining T follicular helper (Tfh) and T follicular regulatory (Tfr) cell differentiation (Sage et al., 2018). Moreover, impairment of meningeal lymphatics could cause weakened drainage of brain Ags, accrual of metabolic wastes and encourage immune cells to enter the brain region, unsettling the neuronal connections and leading to irregular behaviors (Bolte et al., 2020; Gordleeva et al., 2020; Louveau et al., 2015; Rasmussen et al., 2018; Sun et al., 2018). Cervical lymph nodes are the main draining lymph nodes of meningeal lymphatic system (das Neves et al., 2021). The meningeal lymphatic vessels oversee draining immune cells, small molecules, and additional fluid from the central nervous system into the cervical lymph nodes (das Neves et al., 2021; Papadopoulos et al., 2020).

T regulatory (Treg) cells are defined as a CD4⁺CD25⁺ T cell population expressing transcription factor forkhead box P3 (FoxP3) whose deficiency is linked to the development of severe autoimmunity (Fontenot et al., 2003; Hori et al., 2003; Sakaguchi et al., 1995). Both soluble factors like cytokines and direct contact by cell-surface molecules between cells could possibly function as suppression molecules in Treg cell–mediated immune regulation (Josefowicz et al., 2012). Treg cells can prevent neuroinflammation through the regulation of Th17 cell effector functions by limiting access of Th17 cells to antigen presenting cells (APCs) and suppression of Th17 (Othy et al., 2020).

The BTBR strain is an inbred strain with naturally occurring vigorous and selective social shortfalls, lack of social communication, increased level of monotonous self-grooming, and nominal vocalization in social interaction. BTBR mice have been increasingly used to study the underlying mechanisms for the development of ASD and to explore the therapeutic potential to treat the symptoms (McFarlane et al., 2008; Silverman et al., 2010). The Th2-like immune profile and constitutive neuroinflammation have been observed in BTBR mice, which could impact their abnormal behaviors (Heo et al., 2011). It has been reported that B6 progeny developed in BTBR dams had weakened social interaction like BTBR mice, whereas BTBR progeny developed in B6 dams had enhanced social interaction (Zhang et al., 2013). BTBR mice show numerous behavioral and immune aberrations detected in children with autism (Li et al., 2009). Altered immune structure of Th1, Th2, Th17, and T regulatory cells along with cytokine, chemokine and transcriptional signaling was observed in the BTBR mice (Ahmad et al., 2017; Bakheet et al., 2016).

Previously we reported higher levels of serum IgG and anti-brain antibodies (Abs), higher expression of some cytokines in the periphery and brain of BTBR mice (Heo et al., 2011) and maternal environment importantly maternal anti-brain autoantibodies (autoAbs) influence the development of ASD-like behavior (Zhang et al., 2013). An increase of Tfh cells and antibody producing plasma cells in BTBR mice was also reported (Uddin et al., 2020a). In the present study, we are investigating profile of meningeal immune cells such as T cells, B cells, and antigen presenting cells (APCs), and their function and interaction with neural cells contributing to the development of autism-like behavior of BTBR mice. Additionally, the level of autoantibodies specific to brain antigens, anti-nuclear autoantibodies, and the frequency and function of immune suppressing Treg cells were also investigated.

2. Results

2.1. Immune cell profile in the meninges

In autism research, meningeal immunity is a poorly studied aspect of neuroimmune interactions. In our previous studies (Heo et al., 2011; Zhang et al., 2013; Uddin et al., 2020a, 2020b), we observed altered immune profiles in peripheral blood and lymphoid organs. Herein, meningeal immunity of BTBR and C57BL/6J (B6) mice is compared. To investigate the immune profiles of meninges in B6 and BTBR mice, we analyzed all CD45⁺ immune cells, CD3⁺ T cells, CD19⁺ B cells, and CD11b⁺ myeloid cells by flow cytometry. The frequency of CD45⁺ cells in the meninges of BTBR mice was higher than that of B6 mice (Fig. 1A); statistical analysis indicated the deposition of immune cells in BTBR meninges was significantly higher (Fig. 1D, P = 0.005). On average, the number of CD45⁺ cells increased 2-fold from ∼11% in B6 to ∼23% in BTBR mice. When we further investigated the types of immune cells in meninges, the frequency of CD3⁺ T cells was significantly higher in BTBR meninges compared to B6 meninges (Fig. 1 B & E, P = 0.0013). Similarly, the frequency of CD19⁺ B cells was higher in BTBR meninges (Fig. 1 C & F (left graph), P = 0.0003), whereas the frequency of CD11b⁺ myeloid cells was not different in between B6 and BTBR mice (Fig. 1 C & F (right graph), P = 0.49). The percentage of CD3⁺ cells of total CD45⁺ cells was ∼10% and >20% for B6 and BTBR mice, respectively. For CD19⁺ B cells, the percentage increased from ∼20% in B6 mice to ∼25% in BTBR mice. The higher deposition of total immune cells and higher frequency of T cell and B cell populations may shape the immune profile activities of the meninges in BTBR mice.

Fig. 1.

BTBR mice have significantly higher deposition of immune cells in meninges than B6 mice. Representative flow cytometric analysis of CD45⁺ cells (A), CD3⁺ T cells (B), CD19⁺ B cells and CD11b⁺ myeloid cells (C) populations in the meninges of B6 and BTBR mice. Frequency of CD45⁺ immune cells, CD3⁺ T cells, CD19⁺ B cells and CD11b⁺ myeloid cells (D) in the meninges of B6 and BTBR mice. Data are representative of three or four independent experiments with 4–6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.2. Activation status of meningeal immune cells

To assess the function or activation conditions of the T and B cell populations that were increased in the BTBR mice, we evaluated the expression of CD25 by T cells and MHCII expression by B cells and myeloid cells. The expression of CD25 (Fig. 2A) was significantly higher (P < 0.0001) in BTBR mice indicating more activated T cells in the meninges or possibly more Treg cells. The geometric mean fluorescent intensity (GMFI) of CD25 by T cells increased from ∼2400 in B6 to ∼2800 in BTBR mice. The expression level of MHCII also was higher by B cells (Fig. 2B, P = 0.0006) and CD11b⁺ myeloid cells (Fig. 2C, P < 0.0001) of BTBR mice compared with B6 mice. The GMFI of MHCII by CD19⁺ B cells increased from ∼2200 in B6 to ∼4400 in BTBR mice. In case of CD11b⁺ myeloid cells, the GMFI of MHCII was 3-4-fold higher in BTBR mice (∼12000) compared to B6 mice (∼3200). Higher level of MHCII expression in APC suggests greater action and higher Ag presenting capability of B and myeloid cells in the meninges of BTBR mice compared to B6 mice.

Fig. 2.

Increased expression of CD25 and MHCII on meningeal CD45⁺ cells of BTBR mice. Representative flow cytometric analysis of CD25 expression on CD3⁺ T cells (A), MHCII expression on CD19⁺ B cells (B) and CD11b⁺ myeloid cells (C) in the meninges of B6 and BTBR mice. Data are representative of three or four independent experiments with 4–6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.3. Immune cell frequency in the cervical lymph nodes

Considering the high connection with the meningeal lymphatic system, we also investigated the immune profile of cervical lymph nodes. Like the meningeal system, we assayed the frequency of CD3⁺ T cell, CD19⁺ B cell and CD11b⁺ myeloid cell populations in cervical lymph nodes (Fig. 3A &B). The frequency of CD3⁺ T cells were significantly higher in the cervical lymph of BTBR mice compared to B6 mice (Fig. 3 A & C, P = 0.002). The percentage of CD3⁺ T cells was ∼61% in B6 mice and ∼72% in BTBR mice. A similar profile was observed in BTBR meninges; whereas the frequency of CD19⁺ B cells in cervical lymph nodes (Fig. 3 A & C, P = 0.0001) was reduced in BTBR mice compared to B6 mice, which is opposite of meninges where it was increased. The percentage of CD19⁺ B cells was ∼32% in B6 mice and ∼20% in BTBR mice. Like meninges, the frequency of CD11b⁺ myeloid cells (Fig. 3 B & D, P = 0.76) were not different in the cervical lymph nodes of B6 and BTBR mice.

Fig. 3.

BTBR mice have significantly higher frequency of T cells and lower frequency of B cells in the cervical lymph nodes than B6 mice. Representative flow cytometric analysis and frequency of CD3⁺ T cells (A), CD19⁺ B cells (A), and CD11b⁺ (B) myeloid cells in the cervical lymph nodes of B6 and BTBR mice. Data are representative of three or four independent experiments with 4–6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.4. Plasma cell and B cells in the cervical lymph nodes

While B cell frequency was reduced, the number of plasma cells was significantly elevated in BTBR spleens (Heo et al., 2011) suggesting the lower frequency of B cells in the cervical lymph nodes may also be due to increased differentiation to plasma cells. Therefore, the frequency of plasma cells in cervical lymph nodes was investigated. The analysis of plasma cells was accomplished by quantification of CD138⁺ cells in the CD3⁻CD19⁻ population by flow cytometry. The frequency of CD138⁺ plasma cells was significantly higher in the cervical lymph nodes of BTBR mice (Fig. 4A, P = 0.01) compared to B6 mice. The frequency of plasma cells was about 2-fold higher in BTBR mice compared to B6 mice. The expression of CD40 on B cells is essential for the generation of germinal centers, isotype switching, and stable antibody secretion. To investigate B cell activity, CD40⁺ B cells and the expression level of CD40 on B cells was assessed. Both, the frequency of CD40⁺ B cells and the expression levels of CD40 (Fig. 4B, P < 0.0001) were higher on BTBR B cells compared to those of B6 mice. The percentage of CD40⁺ B cells was 40% in B6 and 80% in BTBR mice. The GMFI of CD40 expression by B cells also was higher at ∼700 in B6 to ∼900 in BTBR mice. Higher frequency of plasma cell and more functionally active B cells in BTBR mice supports the higher level of an adaptive or autoimmune condition.

Fig. 4.

Plasma cells and activated B cells are increased in BTBR mice. Representative flow cytometric analysis and frequency of CD138⁺ plasma cells (A) and CD40 expression on B cells (B) in the cervical lymph nodes of B6 and BTBR mice. The geometric mean fluorescent intensity (GeoMFI) of CD40 for expression level on B cells in the cervical lymph nodes of B6 and BTBR mice is also shown (B). Data are representative of three independent experiments with 4–6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.5. Serum IgG autoAb levels

Since BTBR spleens and cervical lymph nodes had a higher number of plasma cells, we further assayed autoAbs for some specificities of Ags suggested to affect brain functions. The Abs are considered autoAbs since they are constitutively produced; the sera autoAbs levels are presented as optical density (OD) values. Aquaporin (AQP)-4 is a major membrane water channel in the central nervous system. The IgG autoAb levels to AQP-4 were significantly higher in the sera of BTBR mice compared to B6 mice (Fig. 5A, P < 0.0001). The amount of anti-AQP-4 was 3-fold higher in BTBR sera compared to B6 sera. Similarly, the autoAb IgG levels to N-methyl-D-aspartate Receptor (NMDAR), Contactin Associated Protein 2 (CASPR2), a membrane protein complexed with the neuronal potassium channel (VGKC), and Pentraxin2/SAP were assayed. The ELISA data indicates that the levels of IgG autoAbs to NMDR (Fig. 5B, P < 0.0001), CASPR2 (Fig. 5C, P < 0.0001), and pentraxin2/SAP (Fig. 5D, P < 0.0001) were also significantly higher in BTBR sera compared to B6 sera. IgG ant-NMDAR was 2 -fold higher in BTBR sera compared to B6 sera. IgG anti-CASPR2 and anti-pentraxin2 were about 8–10 fold higher in BTBR sera compared to B6 sera. Higher level of autoAb to neuronal associated antigens suggests autoimmunity in BTBR mice.

Fig. 5.

Increased level of serum IgG autoAbs to brain associated antigens in BTBR compared to B6 mice. ELISA of BTBR and B6 serum IgG Abs to Aquaporin-4 (A), NMDAR1 (B) Pentraxin (C) and Caspr2 (D). Data are representative of four independent experiments with 6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.6. AutoAb IgG to double-stranded (ds) DNA

Analysis of serum anti-dsDNA Ab levels is a routine clinical assay for some autoimmune diseases. Thus, BTBR vs B6 sera were measured for IgG autoAbs to dsDNA. The ELISA data demonstrated that the IgG anti-dsDNA levels were higher in BTBR sera compared to B6 sera (Fig. 6, P < 0.0001). The IgG anti-dsDNA level was about 3-fold higher in BTBR sera than B6 sera. The higher level of anti-dsDNA further supports the autoimmune condition in BTBR mice.

Fig. 6.

Increased serum anti-dsDNA levels in BTBR than B6 mice with ELISA analyses. Data are representative of four independent experiments with 6 pairs of mice in each experiment. The p values were determined by unpaired two-tailed Student's t-test, and p < 0.05 is considered as significantly different. Error bar indicates mean ± SEM.

2.7. T regulatory (Treg) cells in the meninges and lymphoid organs

Since the activation status of T and B cells, the abundance of plasma cells, and autoAb levels are higher in BTBR mice, it was important to assess if this could be accounted for by a Treg deficiency. We investigated the Treg cell population in the meninges, cervical lymph nodes, and spleen based on the expression of CD25 and FoxP3 in the CD4⁺ T cell population. Flow cytometric analysis revealed the level of Treg cells in the meninges (Fig. 7A, P = 0.001) and cervical lymph nodes (Fig. 7B, P = 0.0005) were significantly lower in BTBR mice than B6 mice; however, the Treg frequency in the BTBR spleens was higher (Fig. 7C, P = 0.0006). The percentage of Treg cells in meninges was ∼22% in B6 and ∼10% in BTBR mice; in cervical lymph nodes, the percentage was ∼11% in B6 and ∼7% in BTBR mice, which suggests the higher number of Treg in spleen failed to control the autoimmune condition in BTBR mice.

Fig. 7.

Decreased frequency of Treg cells in the meninges and cervical lymph nodes of BTBR mice. Representative flow cytometric analysis of CD25⁺FoxP3⁺ Treg cells in the meninges (A), cervical lymph nodes (B) and spleen (C) of B6 and BTBR mice. Frequencies of CD25⁺FoxP3⁺ T regulatory cells (Treg) cells in the meninges (A), cervical lymph nodes (B) and spleen (C) of B6 and BTBR mice are shown. Data are representative of three or four independent experiments with 6 pairs of mice in each experiment. P values determined by unpaired two-tailed Student's t-test are indicated and p < 0.05 is considered as significantly different. Error bars indicate mean ± SEM.

2.8. Cytokine production by Treg cell

Since there were decreased levels of Treg cells in meninges and cervical lymph nodes but and increased number in the BTBR's spleen, Treg function based on IL-10 expression was also necessary. To investigate Treg function, cells were stimulated in vitro and assayed for production of the immunosuppressive cytokine IL-10. The frequency of Treg cells producing IL-10 and IL-10 expression level were significantly lower in the cells from meninges (Fig. 8A, P < 0.0001), cervical lymph nodes (Fig. 8B, P < 0.0001 & P = 0.003) and spleens (Fig. 8C, P < 0.0006 & P = 0.032) of BTBR mice compared to B6 mice. The percentage of IL-10 producing cells was 3-5-fold lower meningeal, lymph node and splenic cells of BTBR mice compared to B6 mice. GMFI of IL-10 also was significantly lower for all tissues with greatest reduction for meningeal cells (∼1500 in B6 and ∼600 in BTBR). Although the spleen showed higher frequency of Treg based on percentage of CD4⁺ T cells expressing CD25 and FoxP3, after stimulation in vitro the splenic cultures had fewer CD4⁺ cells expressing IL-10 (Fig. 9A, P < 0.001) and their IL-10 expression was significantly lower for cultured BTBR cells than those from B6 mice (Fig. 9, P = 0.01). Both the number of IL-10 expressing cells and GMFI of IL-10 was 50% of the B6 values suggesting a lower number and activity of Treg cells fails to control the autoimmunity of BTBR mice.

Fig. 8.

Decreased production of IL-10 by Treg cells in BTBR mice. Representative flow cytometric analysis of IL-10 production by CD25⁺FoxP3⁺ T regulatory cells (Treg) cells in the meninges (A), cervical lymph nodes (B) and spleen (C) of B6 and BTBR mice. Frequencies of IL-10⁺ Treg cells and geometric mean fluorescent intensity (Geo MFI) of IL-10 in Treg cells from meninges (A), cervical lymph nodes (B) and spleen (C) of B6 and BTBR mice are shown. Data are representative of three independent experiments with 4–6 pairs of mice in each experiment. P values determined by unpaired two-tailed Student's t-test are indicated and p < 0.05 is considered as significantly different. Error bars indicate mean ± SEM.

Fig. 9.

Decreased production of IL-10 by CD4⁺ cells in BTBR mice. Frequencies of CD4⁺ IL-10⁺ cells (A) and geometric mean fluorescent intensity (Geo MFI) of IL-10 (B) in CD4⁺ cells from spleens of B6 and BTBR mice. Data are representative of three independent experiments with 4–6 pairs of mice in each experiment. P values determined by unpaired two-tailed Student's t-test are indicated and p < 0.05 is considered as significantly different. Error bars indicate mean ± SEM.

3. Discussion

This study examined the meningeal immune profile (immune cell types, cell activation status, and cell function) of the BTBR mouse model of ASD and autoAbs that might have an influence on neurodevelopment and neurofunctions. Mechanisms involved in autoimmunity also were considered. BTBR mice serve as a widely recognized mouse model as they display unusual behavior including lack of social interaction and restricted repetitive actions that bear a resemblance to ASD (Uddin et al., 2020a; Silverman et al., 2010). Extensive modifications of immune functionality have been observed in individuals with ASD, such as neuroinflammation, higher proinflammatory cytokines in the brain and peripheral blood, brain Ag-specific autoAbs and altered immune profiles (Bakheet et al., 2016). Moreover, these dysfunctional immune responses relate to increased weakening of behaviors and deficits in social relations and communication, suggesting that the immune system plays a key role in the development of ASD (Uddin et al., 2020a, 2020b). AutoAbs to neuronal Ags have been reported in numerous autoimmune neuronal disorders including ASD (Wöhr et al., 2011; Hughes et al., 2018).

Immune cells such as T cells, B cells, innate lymphoid cells (ILCs), macrophages, dendritic cells, mast cells, and neutrophils can circulate through the meningeal lymphatic system under usual and pathological situations. Their activation, functions and signals are important for brain homeostatic conditions and animal behavior (Alves de Lima et al., 2020a; Masi et al., 2017). Autoimmune T cells and B cells in meninges can lead to neuronal inflammation and motor or intellectual dysfunction (Onore et al., 2012a, 2012b). A recent study proposed that meningeal immunity might control the neuroinflammatory response in autoimmune disease including multiple sclerosis (MS) (Rua and McGavern, 2018). Autoreactive CD4⁺ T cells can enter the CNS through the meningeal blood vessels, where they are restimulated by APCs, which can promote the development of neuroinflammation by producing inflammatory cytokines (Wills et al., 2007). An inflammatory condition can induce a tertiary lymphoid structure in meningeal space, which can attract and activate T cells, B cells and APCs leading to a pathogenic condition in neuronal autoimmune diseases (Ma et al., 2021).

An increased proportion of T and B cell populations are present in the meninges of BTBR mice compared to B6 mice. Moreover, the T cells in BTBR meninges are activated, indicated by higher levels of CD25 expression. Higher level of MHCII molecule expression on B cells and myeloid cells also indicated greater activation in BTBR meninges. BTBR mice have higher levels of serum IgG and autoAbs to brain associated Ags, which bound to different brain regions and fractions of brains and neuronal cell lines such as MN9D and CATH.a (Fontenot et al., 2003; Othy et al., 2020). Activated B cell (CD19⁺CD40⁺) and CD138⁺ plasma cell population may contribute to the autoimmune condition in BTBR mice.

BTBR mice had higher serum IgG levels to AQP-4, NMDAR1, Caspr2, and pentraxin. Anti-AQP-4 Abs have been linked to an inflammatory Neuromyelitis Optica Spectrum Disorder (NMOSD) that specifically disrupts optic nerves and spinal cord (Dobson and Giovannoni, 2019). AutoAb to NMDAR is known to cause dysfunctional glutamate neurotransmission in the brain that manifests as psychiatric symptoms (Reali et al., 2020). Brimberg et al. (Codarri et al., 2011) reported an Ab binding to Contactin Associated Protein 2 (CASPR2), a membrane protein complexed with the neuronal potassium channel, is abundant in a mother of ASD child who displayed abnormal neuronal development as well as weakening social interaction, flexible learning and monotonous behavior. AutoAbs to pentraxins have been observed in systemic lupus erythematosus (SLE) and many other autoimmune diseases (Pikor et al., 2015). SAP can suppress development of experimental autoimmune encephalomyelitis (Rosenthal et al., 2020). Thus, BTBR Abs have multiple specificities that may be responsible for the ASD-like behaviors. Brain-reactive Abs are higher in mothers of ASD individuals and are suggested to connect with autoimmunity (Tzang et al., 2019). The BTBR mice also have a high ratio of antinuclear immunoglobulins. Anti-dsDNA or anti-nucleosome Abs are present in 34 and 47% of individuals with ASD (Brimberg et al., 2016) and SLE patients. (Brilland et al., 2021).

The low functionality of Treg cells and higher number of splenic Tfh cells (Heo et al., 2011) in the BTBR mice allow the differentiation of more B cells to become plasma cells generating the autoAbs interfering with brain functions. Treg cells as a primary mediator of peripheral immune tolerance, protecting from autoimmune diseases, and restricting chronic inflammatory diseases (Ji et al., 2012) were dysfunctional in the BTBR mice. FOXP3 expressing Treg cells are important for immune homeostasis and a reduction of this population is connected to up-regulation of many cell types leading to inflammatory neuronal disorder (Sage et al., 2018; Rasmussen et al., 2018). Anti-inflammatory cytokine IL-10 produced by Treg cells plays an important role in suppressing the immune response (Brimberg et al., 2013). BTBR mice having fewer Treg cells and less IL-10 in the meninges and cervical lymph nodes allows more inflammation. With more neuroinflammation, there may be enhanced leakage of brain Ags causing more damage associated molecular patterns (DAMPS) to stimulate microglia and other myeloid cells to release cytokines and chemokines attracting more T and B cells to the meninges or even plasma cells into the brain parenchyma. A similar outcome may occur in individuals with ASD and in experimental mouse model of ASD (Ahmad et al., 2017; Silverman et al., 2010). Although BTBR mice have an elevated number of Treg cells in spleen, their production of IL- 10 was low like that of the Treg cells in BTBR meninges and cervical lymph nodes. Bone marrow derived macrophages from BTBR mice also produce a lower level of IL-10 than B6 mice (Li et al., 2009). Moreover, IL-10 suppresses expression of MHC-II molecule and co-stimulatory molecules on antigen presenting cells (APCs) such as DCs and macrophages to control the proliferation of antigen-specific CD4⁺ T cells (Mostafa et al., 2014). Overall, lower level of Treg cells and their decreased functional activity is unable to prevent the autoimmune-like phenotype of BTBR mice and could possibly contribute to the autism-like behavior.

Male BCF1 mice showed higher anti-brain Ab sera levels than those of B6 mice, however, they had lower levels than the BTBR mice, had less IgG in the brain and no significance differences from B6 mice were observed in social interaction and behavioral index (Heo et al., 2011). Although adult BTBR progeny that developed in B6 dams had similar adult levels of serum anti-brain Abs and IgG in the brain as BTBR mice, they had improved behavior (Zhang et al., 2013). The influence of neonatal experience in the maternal environment is important, and apparently anti-brain Abs alone is only partially responsible for the aberrant behaviors. However, serum BTBR IgG given to B6 dams on gestational days 13–18 was able to significantly lower the normal behavior of the adult B6 offspring (Zhang et al., 2013). The pathogenesis of autoimmune diseases is connected to genetic disposition and epigenetic modifications. Damage in central and peripheral immune tolerance results in autoimmune diseases including systemic lupus erythematosus, type 1 diabetes, rheumatoid arthritis, and primary biliary cirrhosis (Kavanaugh et al., 2000). Genes such as AIRE, Foxp3, CTLA4 and FAS are associated with loss of immune tolerance in autoimmunity (Vignali et al., 2008). Additionally, epigenetics changes such as DNA methylation, histone modification and alteration in microRNAs expression are possibly accountable for the failure of immune tolerance in autoimmune disorders (Vignali et al., 2008). Since the maternal environment plays a significant role in the early development affecting later abnormal behaviors, investigating maternal nuclear and mitochondrial inheritance genes as well as early maternal epigenetic changes during fetal development could potentially decipher the mechanism of autism development. The reduced levels of Treg cells and IL-10 (Hedrich and Bream, 2010; Mittal and Roche, 2015) and their effects on the T and B cell populations are likely implicated in the development of autism-like behavior with production of autoAbs (Hewagama and Richardson, 2009; Zhang and Lu, 2018); however, other immune cell types and brain neurotrophic factors might be involved. The lower production of IL-10 and higher levels of inflammatory cytokines such as IL-1β, IL-18 and IL-33, which are higher in BTBR mice (Uddin et al., 2020b) may account for the neurodevelopment disorder.

4. Materials and methods

4.1. Animals

Two-to three-month-old C57BL/6J (B6) and BTBR T ⁺ Itpr3^tf/J (BTBR) mice were obtained from Jackson Laboratories (Bar Harbor, ME, USA). Both genders were used for all experiments. Findings were combined across sex because of no differences between males and females were obtained. Mouse colonies were maintained in the AAALAC-approved Wadsworth Center Animal Facility under ideal conditions of humidity and temperature in a 12h light-dark cycle (7AM-7PM). All procedures were approved by Wadsworth Center's IACUC.

4.2. Cell and serum preparation

Spleens, cervical lymph nodes and brain meninges were harvested after perfusion with cold phosphate buffered saline (PBS). Single cell suspensions were prepared from spleens and cervical lymph nodes by pressing through the cell strainer (Fisherband, Sterile Cell strainer, 100 μM Cat No: 22363549, Fisher Scientific). Red blood cells (RBC) were lysed with RBC lysis buffer (Cat; 00-4333-57, eBioscience/Invitrogen) before staining.

Cells from the meninges were collected according to previously published protocols with few modifications (Hedrich and Bream, 2010; DiSano et al., 2020). Briefly, using sharp forceps the meninges were collected under dissecting microscope and transfer into Petri dish containing RPMI 1640 (Thermo Fisher, Cat: 11875093) medium supplemented with 25 mM HEPES. Meninges were digested with Collagenase D (Cat:11088858001, Roche, Germany) and DNase I (Cat:10104159001, Roche, Germany) for 1 h at 37 °C on rotator. The digested meninges solution passed through the nylon mesh strainer in a 50 mL conical tube with the help of a 5 ml syringe plunger. The strainer was washed with an additional 2–3 mL of RPMI/HEPES media to pass the cells through the strainer. To get the suitable cell numbers, meninges from 3 to 4 animals were pooled together. Cells were counted with an automated cell counter (Countess II FL, Invitrogen) after staining with viability dye, and 1 x 10⁶ cells were used for staining.

Blood samples were collected into centrifuge tubes and kept at room temperature for 30–60 min. Sera were separated by centrifugation at 14,000×g for 10 min. Aliquoted serum was stored at −80 °C until use.

4.3. In vitro stimulation for cytokine production

For in vitro cytokine production assay, cells from spleen, cervical lymph nodes and meninges were placed in 96-well plates, at a concentration of 1 × 10⁶ cells per 100 μl in RPMI 1640 (Thermo Fisher, Cat: 11875093) medium supplemented with 10%FBS and stimulated for overnight with Cell activation cocktail (Cat:423301, Biolegend) at 37 °C incubator. Brefeldin A solution (Cat: 420601, Biolegend) and Monensin solution (Cat: 420701, Biolegend) were added to the cells for the final 3 h. Cells were collected, stained for surface markers and intracellularly for cytokines, and analyzed by flow cytometry. Data were obtained from three independent experiments with 4–6 pairs of mice in each experiment. In each experiment, equal number (2/3 males and 2/3 females from each group) of male and female mice were used.

4.4. Surface and intracellular/intranuclear staining

Cells from the spleen, cervical lymph nodes and meninges were transferred into FACS tube (100 μl, 1X10⁶ cell), blocked with FC blocker (anti-CD16/32, Cat:553142, BD Pharmingen) and stained with surface antibodies. Cells were then washed, resuspended with FACS buffer and acquired by Flow Cytometer. For intracellular cytokine staining, cells were stained for surface markers in the presence of Brefeldin A solution (Cat: 420601, Biolegend) and Monensin solution (Cat: 420701, Biolegend). After washing cells were fixed and permeabilized with BD Cytofix/Cytoperm solution (Cat: 51-2090KZ, BD Bioscience), and washed with BD Perm/Wash solution (Cat: 51-2091KZ, BD Bioscience). After resuspension in Perm/Wash solution, cells were stained for intracellular cytokine, washed, resuspended in FACS buffer and acquired by FACSCanto flow cytometer. For intranuclear staining (FoxP3), cells were stained with surface antibody and then fixed and permeabilized with fixation/permeabilization buffer (Fixation/Permeabilization buffer set, Cat: 00-5123-43, eBioscience). Intranuclear staining and washing were performed in the presence of permeabilization buffer (Permeabilization buffer, Cat:00-8333-56, eBioscience). After final washing cells were resuspended in FACS buffer and data were acquired by FACSCanto flow cytometer (BD Biosciences). Data were obtained from three to four independent experiments with 4–6 pairs of mice in each experiment. In each experiment, equal number (2/3 males and 2/3 females from each group) of male and female mice were used.

4.5. Flow cytometry

The cells were stained for surface and intracellular/intranuclear, and analysis by FACS Canto flow cytometer (BD Biosciences). The following fluorochrome conjugated Abs: PerCP Cy5.5-CD45, APC Cy7-CD45, PerCP Cy5.5-CD4, PE-CD4, FITC-CD4, FITC-CD3, APC-CD3e, PerCP-CD3e, APC Cy7-CD3, PE-CD11b,PE Cy7-CD11b, FITC-MHC-II, PerCP Cy5.5-MHCII, PE-CD19, FITC-CD19, PE Cy7-CD19, APC-CD25, PE Cy7-CD25, PerCp Cy5.5-CD25, APC-CD138, FITC-CD40, PE-FoxP3, FITC-IL10 and anti-CD16/32 (Fc block) were purchased from BD Pharmingen (San Diego, CA), Biolegend (San Diego, CA), or eBiosciences (Thermo Fisher Scientific, MA, USA). Frequencies and numbers of populations in the meninges, lymph nodes and spleens of B6 and BTBR mice were gated based on FSC-A and SSC-A, followed by gating in single cells and finally gated on CD45⁺ cells. Acquired FCS files were analyzed by Flow Jo-V10. Treg cell population were identified based on CD25⁺FoxP3⁺ cell in CD3⁺CD4⁺ population. To observe the Treg cell population, CD4⁺ cells were analyzed for the expression of CD25 and FoxP3 and gated on the double positive population. Data were obtained from three to four independent experiments with 4–6 pairs of mice in each experiment. In each experiment, equal number (2/3 males and 2/3 females from each group) of male and female mice were used.

4.6. ELISA for IgG anti-dsDNA in mouse serum

The level of total anti-dsDNA in serum was determined with a sandwich ELISA. Costar assay plate (96-well, flat bottom, High binding. Corning incorporated, Ref: 3369) were U–V treated and coated with poly dA-dT (Sigma D4522, Lot: SLCD4909, 7.5 μg/ml) and HRP-conjugated goat anti-mouse IgG Fc (dilution 1:10000, Cat: 115035008, Jackson immune research) was used as the detection Ab; TMB (3,3′,5,5′-tetramethylbenzidine) was used as substrate (Sigma). Plates were washed with an automated plate washer (BioTek ELx405Select CW, Winooski, VT) and read the absorbance at OD₄₅₀ with an ELISA plate reader (BioTek EL808). Sera were diluted 1:100 with 10% FBS in PBS. Data were obtained from four independent experiments with 6 pairs of mice in each experiment. In each experiment, equal number (3 males and 3 females from each group) of male and female mice were used.

4.7. ELISA for IgG autoAbs

Rabbit polyclonal anti-AQP4 (SAB5200112-100UL, Lot:PA187526, Millipore Sigma), rabbit polyclonal anti-Pentraxin2 or SAP (R&D systems, Catalog # AF2558), rabbit polyclonal anti-Caspr2/CNTNAP2 (Cat: ab218048, abcam), and rabbit monoclonal anti-NMDAR1 (Clone # 54.1, Invitrogen, Catalog # 32–0500, Lot: WA314671) were used to coat 96-well plate (0.5 μg/well) for capturing Ags known to be expressed in brains. The plates were incubated for 2 h at room temperature. After 3X washing, the plates were blocked with 2.5% BSA-PBST for overnight at 4 °C. After incubation and washing, SCID mouse whole brain lysate (10 μg/100μl/well) was added into the wells and incubated for 2 h at room temperature. After 3X washing, sera from B6 or BTBR mice (dilution 1/10 or 1/100) were added into the wells and incubated for 2 h at room temperature. After 6X washing, HRP-goat anti-mouse IgG Ab (dilution 1:10000, Cat: 115035008, Jackson immune research) was added and incubated for another 2 h at room temperature. The plates were washed 6X, and the TMB (3,3′,5,5′-tetramethylbenzidine, Sigma) was used as a substrate solution for color development. Absorbance was measured at 450 nm by ELISA analyzer (BioTek EL808). Data were obtained from four independent experiments with 6 pairs of mice in each experiment. In each experiment, equal number (3 males and 3 females from each group) of male and female mice were used.

4.8. Statistical analysis

Data are presented as mean ± SEM, unpaired two-tailed Student's t-test were used to determine p values and p < 0.05 measured as statistically significant difference.

Funding

The work reported in this manuscript was supported by a NIH grant (R01 ES025584) to DAL.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Data will be made available on request.