Introduction

ASDs are a group of neurodevelopmental disorders manifested by social and communication deficits, repetitive behavior, and lack of motivation and interest (Liu et al. 2019). According to WHO, 7 November 2019, the global prevalence is estimated to be 1 in 160 children, thus ASDs impose significant economic burden on education, health, and social service systems.

A huge debate concerning the etiology of ASDs and the degree of contribution of genetic and/or environmental factors is still going on. Only about 25% of the cases have a strong genetic element as a rare genetic mutation or carrying an inheritable single gene (Devlin and Scherer 2012; Howlin et al. 2000; Ozonoff et al. 2010; Seltzer et al. 2004). Other cases are likely exposed to, in utero or in the early neonatal life, environmental factors such as toxins, drugs as valproate, pesticides, infections, and heavy metals (Berko et al. 2014; Bescoby-chambers et al. 2001; Lamb et al. 2000; Rasalam et al. 2005; St-Hilaire et al. 2012; Tordjman et al. 2014; Volk et al. 2013). ASDs relationship to cadmium, arsenic, and lead exposure were studied; however, great focus was applied on the role of mercury (Filipek 1995).

Thimerosal (THIM; sodium ethylmercurithiosalicylate, 49% mercury by weight) is a well-established antimicrobial preservative reagent used in vaccines and some medicinal preparations since 1930s (Magos 2001; Pless and Risher 2000). It is accused for numerous neurodevelopmental disorders as attention deficit hyperactivity disorder (ADHD); speech/language difficulties, and ASDs (Ball et al. 2001). However, the casual relation between THIM exposure and ASDs is still a debatable issue since 1999 (Adamson et al. 2011; Gallagher and Goodman 2016; Mcfarlane et al. 2008). Some studies considered THIM as a significant risk factor (Gallagher and Goodman 2010, 2016; Geier and Geier 2003; Geier et al. 2014, 2018; Geier et al. 2015a, b; Young et al. 2008). However, other studies mitigated that theory (Andrews et al. 2015; Price et al. 2010; Rimland 2004; Schechter and Grether 2008). Thus, the link of THIM to neonatal ASDs remains an open query.

The neurotransmitter systems (Buitelaar and Willemsen-swinkels 2000; Purcell et al. 2001; Whitaker-azmitia 2001), especially the cholinergic system, (Lee et al. 2002; Perry et al. 2001) were reported to be involved in ASDs pathophysiology (Tsai 1999). Literatures described the loss of neuronal α7nAChRs in cerebral and cerebellar cortex of autistic subjects (Deutsch et al. 2010; Perry et al. 2001). Furthermore, their activation has been shown to suppress the inflammatory processes, involved in numerous neurological disorders, via targeting NF-κB signaling (Kalkman and Feuerbach 2016). Thus, positive allosteric modulation of α7nAChRs is expected to have a promising therapeutic role in ASDs.

Mitochondrial dysfunction-induced overproduction of reactive oxygen species (ROS) and neuronal apoptosis were estimated in 5–80% of autistic children (Rego and Oliveira 2003; Wu et al. 2017). ROS stimulate the release of proinflammatory lipid mediators, prostaglandins, and leukotrienes (LTs) from activated glial cells (Chiurchiù and MacCarrone 2011; Kyrkanides et al. 2002) which potentiate neuronal inflammation and excitotoxicity (Ding et al. 2006; Sanico et al. 1998). Drugs that suppress LTs release or antagonize their receptors could be of value in ameliorating cognitive and/or social deficits of autistic kids.

Montelukast is a selective cysteinyl leukotriene receptor 1 (cysLT1R) antagonist used in allergic rhinitis, asthma, and other inflammatory conditions (Ilarraza et al. 2012). It has been proved to be protective against ischemic brain damage (Saad et al. 2014) and some neurological complications of Alzheimer disease (AD) (Rahman et al. 2019). Alexiou et al. (2018) presented significant correlations between protein markers of AD and ASDs. The AD-linked amyloid-β (Aβ) precursor protein-α has been raised in cases of severe autism (Zeidán-Chuliá et al. 2014). Sokol et al. (2019) discussed its possible association to white brain enlargement in ASDs. Montelukast has been proved to be effective in treating dementia and memory deficits (Rozin 2017); however, its potential ability to ameliorate social and cognitive impairment in autistic children has not been studied yet.

The hypothesis tested in this study is the likely ameliorative role of montelukast over THIM-induced neuronal damage as a mice model of ASDs. Its potential mitigating effect was investigated through evaluating neuronal inflammatory, apoptotic, and brain injury markers expression.

Materials and Methods

Animals and Experimental Design

The experiments were carried out with pups (P7) weighing 5–7 g purchased from the animal house of The Faculty of Medicine, Assuit University. They were kept with their mothers in stainless steel cages. Their mothers were allowed water and food (laboratory chow) ad libitum. The pups were randomly assigned as follows:

Group (1) Cont. group (n = 10) received saline injections for 5 weeks.

Group (2) THIM group (n = 10) received THIM (Advent Co (INDIA)), dissolved in saline, in a volume of 50 μl, IM into the glutei maximi at dose of 3000 μg Hg/kg per injection on postnatal days 7, 9, 11, and 15.

Group (3) Monte group (n = 10) received montelukast sodium (Sigma-Aldrich (USA), dissolved in saline, at a dose of 10 mg/kg/day IP for 3 weeks, one week after the last dose of THIM.

All the animal procedures were performed with approval from the Institutional Animal Care and Use Committee of The Faculty of Medicine, Assuit University and according to the National Institutes of Health Guidelines for the Human and treatment of Animals.

Developmental Assessments

Serial Weights

To characterize the postnatal developmental patterns (Zhang et al. 2014), Cont. and THIM groups were observed daily starting from the first day of THIM administration (P7) till P21 for weight changes. Starting from P22 (one week after the last dose of THIM), half of THIM-treated mice were given montelukast for extra 3 weeks. Serial body weights were observed for the three groups starting from the first day of montelukast treatment (P22) till P45. Self-righting and air-righting (midair reflex) were fully developed before starting THIM administration.

Behavioral Studies

Battery of behavioral paradigms was performed in the following sequence: Sociability test, Open field test, and Morris Water Maze (MWM) test.

Three-Chamber Test (Social Interaction Test)

The three-chamber test was carried out on mice for evaluating their social behaviors (Kumar and Sharma 2015). The apparatus is rectangular in shape (70 × 40 × 22 cm) and separated into three equal chambers with dividing walls. The walls are made from clear Plexiglass, with small rectangular doors allowing free access to each chamber. Two equal wire cups are put in the 2 lateral chambers; one holds a single mouse, while the other one is kept empty. Mice were left to habituate in the central compartment for 10 min. During the test, mice were allowed to move freely through the adjacent chambers. The test mouse was free to spend time with a stranger mouse versus staying solely in the empty chamber. The time spent in each chamber was recorded and parameters such as sociability, social index (time spent in the stranger side/time spent in the empty side), and sniffing time were calculated. The chambers and the wire cups were wiped between trials with a 70% ethyl alcohol solution.

Open Field Test

The test was carried out according to the protocols of Brown et al. (1999) for assessment of the degree of anxiety and locomotor activity. It was implemented in a quiet dim-lighted room. Mice were, at first, acclimatized to the test chamber for 5 min. The test chamber is white, square shaped, with dimensions of 40 × 40 cm. Its floor was divided into 16 squares (8 × 8 cm/each). Mice activity was recorded for 5 min by digital camera. The time spent in the central area and the locomotor activity were observed and calculated. The test chamber was wiped between trials with a 70% ethyl alcohol solution.

Morris Water Maze (MWM) Test

MWM test is a commonly used test for assessing memory in animals. MWM apparatus consists of a large rounded pool (45 cm height, 150 cm in diameter, filled to a depth of 30 cm with water at 28 °C). Milk was added to opacify water. The tank was alienated into four equal quadrants and an underwater platform (10 cm2) was kept 1 cm beneath the surface of the water within the target quadrant of the pool. The location of the platform was maintained unchanged during the training period. Every mouse was exposed to four successive trials daily with a 5 min interval. The mouse was quietly placed in the water between quadrants. The falling position varied for each trial and the test mouse was permitted 90 s to trace the underwater platform. After finding the platform, the mouse was allowed to reside on the platform for 20 s. If the mouse failed to reach the platform within 90 s, it was directed tenderly onto the platform and left to stay there for 20 s. The platform was taken away on the sixth day and mice were allowed to search the pool for 90 s. All trials were conducted during the light cycle, i.e., between 09.00 and 18.00 h (Gupta and Sharma 2014; Singh et al. 2015).

Enzyme-Linked Immunosorbent Assay (ELISA)

The animal groups were sacrificed after carrying out all behavioral testing and their brains were carefully isolated. Neuronal NF-κB p65 was quantitatively detected in brain homogenates using an ELISA kit, product of Chongqing Biospes Co., Ltd, according to the manufacturer’s instructions. After the enzyme–substrate reaction was terminated, the optical density (OD) was measured at a wavelength of 450 nm using a microplate reader, then the concentrations of NF-κB p65 were calculated by comparing the OD of the samples to the standard curve.

Histopathology Experiments

The animal groups were sacrificed after carrying out all behavioral testing. The mice were anesthetized with thiopental sodium (50 mg/kg) IP. Their hearts were exposed and perfused with saline till getting clear flow. The perfusion was finalized with 10% formalin. Brains were carefully isolated from the skull and immersed in 10% neutral formalin to be processed to get 5 µm paraffin sections which were stained with Hematoxylin and Eosin (Hx&E) for the left hippocampus examination by light microscopy (Bancroft 2008).

Immunohistochemistry Studies

The isolated brains were fixed in 10% neutral formalin for two days followed by dehydration, clearing, and embedding. Paraffin sections were cut at 4 µm and stained with the modified avidin–biotin–peroxidase technique. Deparaffinization and rehydration were done and the slides were incubated in 0.3% H2O2 for 10 min to block the endogenous peroxidase activity. The tissue sections were incubated at 4 °C overnight with the following primary antibodies (CHRNA7, Chongqing Biospes Co., Ltd, 1:50 dilution), (Bax, Biogenex, 1:50 dilution), (anti-S100b-rabbit antiserum, Sigma-Aldrich Inc., St Louis, MO, USA, 1:200 dilution), (monoclonal mouse antisynaptophysin (MAB5258; Chemicon, Temecula, CA, USA), 1:200 dilution in 0.5% TX, and 5% normal donkey serum). The slides were washed with PBS and incubated with biotinylated secondary antibodies and then with avidin–biotin enzyme complex. The visualization of immunoreactivity was performed using 3,30-diaminobenzidine hydrogen peroxide as chromogen sections were finally counterstained with Mayer's hematoxylin (Thermo Scientific Inc.Waltham, USA) and mounted with water-based mounting medium. HRP secondary antibody kit (Anti-polyvalent HRP, Labvision Corp, Fremont, CA, USA) was used as a secondary antibody and AEC kit (Labvision Corp, Fremont, CA, USA) was used for the chromogen. Negative (−ve) control slides were done with omission of the primary antibodies using a non-immunized goat sera (Kiernan 2001).

Morphometric Studies

Morphometric studies were done using a Java-based open source image processing package, image J. The required parameters were measured in three non-overlapped fields/five sections/3 mice from each group. The measured parameters were (1) the thickness of dentate gyrus (DG) granule cells and pyramidal neurons in CA1 fields using X40 lens in Hx&E-stained sections, (2) number of dark cells in DG granule cells and pyramidal neurons in CA1 fields using X40 lens in Hx&E-stained sections, (3) the number of α7nAChRs immunostained cells counted in 3 non-overlapping fields in each section using X40 lens, (4) the number of positive (+ve) Bax immunostained cells counted in 3 non-overlapping fields in each section using X40 lens, (5) the number of +ve S100b protein immunostained cells counted in 3 non-overlapping fields in each section using X40 lens and (6) the area percentage of synaptophysin immune +ve reaction estimated in 3 non-overlapping fields in each section using X40 lens.

Electron Microscopy Studies

Mice were intracardially perfused with 4% cold glutaraldehyde in cacodylate buffer (pH 7.4). Specimens (1 mm size) were taken from hippocampi and processed. Ultrathin sections (500A) were cut from selected areas in semithin sections (Gupta 1983). Each section was studied with the transmission electron microscope, JEOL (J.E.M.-100 CXII), and photographed at 80 kV in Assiut University / Electron Microscope Unit.

Statistical Analysis

Data were expressed as mean ± standard error of the mean (SEM). Statistical analysis was performed by multiple t-test, one-way and two-way repeated measures ANOVA followed by post hoc Tukey's test when appropriate. All statistical tests were directed using GraphPad Prism 7 software. The difference among groups was considered significant for p < 0.05.

Results

The Effects of THIM and Montelukast on Growth Development

THIM group presented significant growth delay throughout the whole experiment period compared to Cont. group (p < 0.05, P10; p < 0.0001, P11 till P20; p < 0.001, P21; Fig. 1a and p < 0.0001, P24, P39, P45; p < 0.01, P36; Fig. 1b). No significant body weight changes have been observed between THIM and Monte groups (Fig. 1b) which indicates that montelukast has no ameliorative potential over THIM-induced growth delay in mice.

Fig. 1
figure 1

The effects of THIM and montelukast on growth development. a Serial body weights of THIM versus Cont. (P7 till P21), (*p < 0.05, P10; ****p < 0.0001, P11 till P20; ***p < 0.001, P21). b Serial body weights of THIM, Monte versus Cont. (P21 till P45), (****p < 0.0001, P24, P39, P45; **p < 0.01, P36). Data presented as mean ± SEM

Full size image

The Effects of THIM and Montelukast on Behavior Outcomes

Three-Chamber Assay (Social Interaction Test)

THIM group showed less sociability compared to Cont. group as THIM mice spent significantly less time in the stranger mouse side than that in the empty side (object side) (p < 0.05) (Fig. 2a). Monte group displayed more social preference compared to THIM group (p < 0.05) (Fig. 2a). Social index of THIM group was significantly lower compared to Cont. group (p < 0.05); however, Monte group showed insignificantly higher social index versus THIM group (Fig. 2b). THIM group spent less time sniffing the stranger mouse compared to Cont. group (p < 0.01) (Fig. 2c). On the contrary, Monte group spent significantly more time sniffing the stranger mouse compared to THIM group (p < 0.001). Monte group behavioral response was similar to that induced by saline in the Cont. group (Fig. 2c). These results confirmed the hypothesis that THIM induced social deficit which could be ameliorated through montelukast treatment.

Fig. 2
figure 2

The effects of THIM and montelukast on behavior outcomes in the social interaction, Three-Chamber Assay Test. a Sociability, the preference for social stimulus measured by cumulative time spent in either a stranger mouse side or an empty cage side. b Social index, mathematical calculation measured as time spent in the stranger mouse side divided on that spent in the empty side. c Sniffing time reflects the motivation to interact socially. *p, #p < 0.05, **p < 0.01, ###p < 0.001. Data presented as mean ± SEM

Full size image

Open Field Test

The exploratory, locomotor, and anxiety behavior of the studied groups were studied using open field test. The scored behavior included central square duration (measure of anxiety), total path length, and speed (measure of exploratory and locomotor activity). One way ANOVA analysis showed that THIM induced a significant increase in the central activity compared to Cont. and Monte groups (p < 0.001) (Fig. 3a). Total path length was significantly longer in THIM compared to Cont. and Monte groups (p < 0.01) as well (Fig. 3b). Thus, THIM affected the total movements and the traveled distance within the central area of the open field. THIM group had significantly higher exploration speed compared to Cont. group (p < 0.05) (Fig. 3c). THIM aggravated the locomotor, exploratory, and the anxiety behavior that were mitigated through montelukast administration.

Fig. 3
figure 3

The effects of THIM and montelukast administration on the exploratory, locomotor, and anxiety behavior of mice in the Open Field test. The scored behavior included the following: a Central zone duration, b Total path length, and c Speed. *p < 0.05, **p, ##p < 0.01, ***p, ###p < 0.001. Data presented as mean ± SEM

Full size image

Morris Water Maze (MWM) Test

Spatial learning and memory were assessed by MWM test using the acquisition and the probe trials, respectively (Fig. 4). THIM group showed longer escape latencies (Fig. 4a) and higher swimming speed (Fig. 4b); however, these changes were insignificant (except for the escape latency of day 1 in THIM group). In the probe test, THIM group results revealed insignificantly increased latency to the target quadrant (Fig. 4c) and insignificantly decreased time spent there (Fig. 4d). Significance may probably be attained with larger sample sizes. Montelukast may have the ability to rescue the learning and the memory deficit induced by THIM.

Fig. 4
figure 4

The effects of THIM and montelukast administration on acquisition and probe trials in MWM test. a Escape latency (s). b Swimming speed (m/s). c Time to reach the target quadrant (s) (that previously contained the platform). d Time spent in the target quadrant (s). *p < 0.05. Data presented as mean ± SEM

Full size image

Assay of NF-κB p65 Levels

Using ELISA, NF-κB p65 expression levels were measured in the cerebrum of the different studied groups. NF-κB p65 expression was significantly increased in THIM group (p < 0.01) (Fig. 5). However, Monte treatment dramatically reduced the levels of NF-κB p65 expression (p < 0.05) (Fig. 5). Montelukast may have a promising therapeutic role in ameliorating neuroinflmmation frequently detected in ASDs cases.

Fig. 5
figure 5

NF-κB p65 concentrations in the cerebrum of THIM versus Monte groups. *p < 0.05, **p < 0.01. Data presented as mean ± SEM

Full size image

Hippocampus Histopathology

Sections from brain Cont. group stained with Hx&E showed normal hippocampus structure which consists of DG and Ammon's horn (AH). Cont. AH is formed of CA1, CA2, and CA3. Cont. DG is a V-shaped structure which consists of inner blade between CA1 and CA3 and outer blade opposite to CA3 (Fig. 6a). The DG of the Cont. group is formed of the outer molecular layer, the inner polymorphic layer, and the middle granule cell layer which is the principal layer and is composed of several rows of closely packed granule cells. The granule cell layer is formed of rounded or oval vesicular nuclei and is surrounded by a thin rim of cytoplasm (Fig. 6b). Hx&E-stained sections from THIM DG revealed dark cells with deeply stained nuclei and irregular outlines especially in deeper layers leaving empty spaces (Fig. 6c). On the contrary, most of granular cells appeared normal in Monte group (Fig. 6d). Morphometric results verified significant reduction in the DG mean thickness of THIM group (p < 0.01); however, significant increase in the mean thickness (p < 0.001) was observed after treatment with montelukast (Fig. 6e). There was a significant increase in the number of dark cells in THIM group (p < 0.0001); however, Monte group displayed a significant decrease of the number of dark cells (p < 0.01) compared to THIM group (Fig. 6f).

Fig. 6
figure 6

Mice hippocampus (DG field) of the studied groups. a Cont. hippocampus; fields of AH: CA1, CA2, CA3, and DG with its inner and outer blades Ib&Ob (Hx&E×40). b Cont. DG; the granular layer which is closely packed with rows of rounded vesicular granular cells (↑) Hx&E × 400. c THIM DG; most of the granular cells have deeply stained irregular nuclei (arrow head), while few of them are light vesicular (↑) × 400, Inset: Higher magnification showed light cells (↑) and dark cells (arrow) × 1000. d Monte DG; most of the granular cells are light vesicular (↑) × 400, Inset: Higher magnification showed light cells (↑) and dark cells (arrow) × 1000. e Statistical analysis of the mean thickness of DG among the studied groups. f Statistical analysis of the number of dark cells in DG region among the studied groups. **p < 0.01, ***p < 0.001, ****p < 0.0001. Data presented as mean ± SEM

Full size image

The CA1 field of Cont. AH is formed of pyramidal neurons which are the principal cell type. These neurons are medium sized with triangular perikaryal and are closely located in stratum pyamidale (Fig. 7a). They have rounded vesicular nuclei and basophilic cytoplasm (Fig. 7a). Blood capillaries were also seen (Fig. 7a). CA1 area of THIM group revealed dark shrunken pyramidal neurons bounded by empty spaces (Fig. 7b). Empty areas were seen around the blood capillaries as well (Fig. 7b). After treatment with montelukast, most of the pyramidal cells appeared pale stained with vesicular nuclei, while few of them appeared darkly stained (Fig. 7c). Morphometric results showed a significant decrease in the mean thickness of stratum pyamidale and significant rise of dark cells in THIM group (p < 0.01 and 0.0001) (Fig. 7d, e). Monte group revealed a significant increase of stratum pyamidale thickness and a significant decrease in the number of dark cells (p < 0.001 and 0.0001), respectively, compared to THIM group (Fig. 7d, e).

Fig. 7
figure 7

Mice hippocampus (CA1 field) of the studied groups. a Cont. CA1; medium-sized pyramidal neurons with rounded vesicular nuclei (arrow heads). Small blood capillaries can be seen (*) Hx&E × 400. b THIM CA1; most of the neurons are deeply stained and shrunken (arrow heads) leaving empty spaces. Dilated blood capillaries can be seen (*) Hx&E × 400, Inset: Higher magnification showed light cells (↑) and dark cells (arrow) × 1000. c Monte CA1; multiple pale stained neurons (↑) and few dark shrunken neurons leaving empty spaces (arrow heads) Hx&E × 400, Inset: Higher magnification showed light cells (↑) and dark cells (arrow) × 1000. d Statistical analysis of the mean thickness of CA1 region among the studied groups. e Statistical analysis of the number of dark cells in CA1 regions among the studied groups. **p, ++p < 0.01, ###p < 0.001, ****p, ####p < 0.0001. Data presented as mean ± SEM

Full size image

Immunohistochemical Studies

α7nAchRs Expression

Immunostaining of Cont. group hippocampus with CHRNA7 demonstrated multiple +ve immunostained cells in both the DG and CA1 fields (Fig. 8a, d). THIM group revealed decreased number of immunostained cells in both fields (Fig. 8b, e). After treatment with montelukast, there was an increase in the number of immunostained cells in DG and CA1 fields compared to THIM group (Fig. 8c, f). Morphometric results have shown significant decrease of α7nAchRs immunostained +ve cells in THIM group in both DG and CA1 regions (p < 0.0001 and 0.01) respectively; however, Monte group showed significant increase of the immunostained +ve cells in both fields (p < 0.01 and 0.01) compared to THIM group (Fig. 8g, h).

Fig. 8
figure 8

α7nAchRs expression in the mice hippocampus of the studied groups. a Cont. DG; multiple +ve immunostained cells (↑). b THIM DG; few +ve immunostained cells (↑). c Monte DG; relative increase of +ve immunostained cells (↑). d Cont. CA1; multiple +ve immunostained cells (↑). e THIM CA1; few +ve immunostained cells (↑). f Monte CA1; multiple +ve immunostained cells (↑) × 400. g Statistical analysis of the mean number of α7nAchRs immune +ve cells in DG regions among the studied groups. h Statistical analysis of mean number of α7nAchRs immune +ve cells in CA1 regions among the studied groups. **p, ##p < 0.01, +++p < 0.001, ****p < 0.0001. Data presented as mean ± SEM

Full size image

Bax Expression

Immunostaining of Cont. group hippocampus with Bax revealed few +ve cells in the DG and CA1 fields (Fig. 9a, d). THIM group revealed increased immunopositive cells in DG, especially in deeper layers, and CA1 fields (Fig. 9b, e). Treatment with Montelukast decreased the number of immunostained +ve cells in both fields compared to THIM group (Fig. 9c, f). Morphometric results showed significant increase of Bax immunostained +ve cells in THIM group in both DG and CA1 regions (p < 0.001 and 0.0001), respectively, compared to Cont. group (Fig. 9g, h). Monte group showed a significant decrease of Bax immunostained +ve cells in DG and CA1 fields (p < 0.001 and 0.001) compared to THIM group (Fig. 9g, h).

Fig. 9
figure 9

Bax expression in mice hippocampus of the studied groups. a Cont. DG; few +ve immunostained cells (↑). b THIM DG; increase of +ve immunostained cells (↑). c Monte DG; few +ve immunostained cells (↑). d Cont. CA1; few +ve immunostained cells (↑). e THIM CA1; multiple +ve immunostained cells (↑). f Monte CA1; few +ve immunostained cells (↑) × 400. g Statistical analysis of the mean number of Bax immune +ve cells in DG regions among the studied groups. h Statistical analysis of the mean number of Bax immune +ve cells in CA1 regions among the studied groups. +p < 0.05, ***p, ###p < 0.001, ****p < 0.0001. Data presented as mean ± SEM

Full size image

S100b Proteins Expression

Cont. group hippocampus demonstrated few immunostained +ve cells for S100b protein in both DG and CA1 fields (Figs. 10a, 11a). Higher magnification showed short processed glial cells in DG field (Fig. 10b). THIM group revealed an increase of S100b +ve cells in all layers of DG and CA1 fields compared to Cont. group (Figs. 10c, 11b). Higher magnification revealed long and branched processed glial cells in both DG and CA1 fields (Figs. 10d, 11c). Monte group showed multiple S100b immunostained +ve cells in DG and CA1 fields compared to Cont. group (Figs. 10e, 11d). Higher magnification showed short processed glial cells (Figs. 10f, 11e) in both DG and CA1 fields. Morphometric results showed a significant increase of S100b protein immunostained +ve cells in THIM group in both DG and CA1 fields (p < 0.0001 and 0.0001), respectively. Monte group showed an insignificant decrease in its expression in both DG and CA1 fields compared to THIM group and significant increase of S100b protein immunostained +ve cells in both DG and CA1 regions (p < 0.0001 and 0.0001) compared to Cont. group (Figs. 10g and 11f).

Fig. 10
figure 10

S100b expression in mice hippocampus (DG field) of the studied groups. a Cont.; few +ve immunostained cells (↑). b Higher magnification showing short processed glial cells (curved arrows) × 1000. c THIM; multiple +ve immunostained cells (↑). d Higher magnification showing branched long processed glial cells (curved arrows) × 1000. e Monte; multiple +ve immunostained cells (↑). f Higher magnification showing few short processed (curved arrows) glial cells × 1000. g Statistical analysis of the mean number of S100b protein immune +ve cells in DG regions among the studied groups.****p < 0.0001. Data presented as mean ± SEM

Full size image
Fig. 11
figure 11

S100b expression in mice hippocampus (CA1 field) of the studied groups. a Cont.; few +ve immunostained cells (↑). b THIM; multiple +ve immunostained cells (↑). c Higher magnification showing long branched processed glial cells (curved arrows) × 1000. d Monte; multiple +ve immunostained cells (↑) × 400. e Higher magnification showing short processes of glial cells (curved arrows) × 1000. f Statistical analysis of the mean number of S100b protein immune +ve cells in CA1 regions among the studied groups. ****p < 0.0001. Data presented as mean ± SEM

Full size image

Synaptophysin Proteins Expression

Immunostaining of hippocampus Cont. group with synaptophysin revealed intense expression in both DG and CA1 fields (Fig. 12a, d). Decreased expression of synaptophysin in both DG and CA1 fields was noticed in THIM group (Fig. 12b, e). Treatment with Montelukast induced intense staining in both DG and CA1 fields (Fig. 12c, f). Morphometric results showed a significant decrease of area % of synaptophysin reaction in THIM group in both DG and CA1 fields (p < 0.0001 and 0.0001), respectively (Fig. 12i, j). Montelukast administration induced significant increase of area % of synaptophysin in DG and CA1 fields (p < 0.001 and 0.0001) compared to THIM group (Figs. 12i, j).

Fig. 12
figure 12

Synaptophysin expression in mice hippocampus of the studied groups. a Cont. DG; reaction for synaptophysin (↑) × 400, Inset: higher magnification showing intense staining for synaptophysin among granular cells (arrow heads) × 1000. b THIM DG; +ve reaction for synaptophysin (↑) × 400, Inset: higher magnification showing decreased intensity for synaptophysin (arrow heads) × 1000. c Monte DG; +ve reaction for synaptophysin (↑), Inset: Higher magnification showing intense staining for synaptophysin (arrow heads). d Cont. CA1; intense staining for synaptophysin (↑), +ve reaction for synaptophysin (↑), Inset: higher magnifications showing intense reaction for synaptophysin (arrow heads). e THIM CA1; staining for synaptophysin (↑), Inset: Higher magnification showing decreased reaction for synaptophysin (arrow heads) × 1000. f Monte CA1; +ve staining for synaptophysin (↑) × 400, Inset: Higher magnification showing intense reaction for synaptophysin (arrow heads) × 1000. g (−ve) Control; Control mice hippocampus showed negative reaction for synaptophysin immunoreactivity × 100. h (+ve) Control; Control mice cerebellum showed marked synaptophysin-positive immunoreactivity in the synaptic connection in the granular layer (↑) × 400. i Statistical analysis of area % for synaptophysin immunostaining distribution in DG fields in the studied groups. j Statistical analysis of area % for synaptophysin immunostaining distribution in CA1 fields in the studied groups. ++p < 0.01, +++p, ###p < 0.001, ****p, ####p < 0.0001. Data presented as mean ± SEM

Full size image

Electron Microscopic Examination

As shown in Fig. 13, the granular neurons appeared rounded euchromatic with minimal intercellular tissue in Cont. DG (Fig. 13a). The small amount of cytoplasm contained rounded nuclei, mitochondria, Golgi cisternae, free ribosomes, and short strands of the rough endoplasmic reticulum (rER) (Fig. 13a). Ultrastructure of THIM DG revealed multiple degenerated neurons with electron dense cytoplasm and ill-defined organelles (Fig. 13b). After treatment with montelukast, few degenerated cells were noticed in DG field (Fig. 13c). Cont. CA1 field showed the normal ultrastructure of the principal pyramidal neurons (Fig. 13d). They appeared large oval or rounded with euchromatic nucleus and prominent nucleolus (Fig. 13d). It showed normal amount of cytoplasm containing mitochondria, rER, and multiple ribosomes (Fig. 13d). THIM CA1 field showed degenerated cells with electron dense cytoplasm and ill-defined organelles, while others appeared electrolucent with rarified cytoplasm (Fig. 13e). Monte group revealed electrolucent pyramidal cells with large rounded nuclei and rarified cytoplasm (Fig. 13f).

Fig. 13
figure 13

Transmission electron micrographs (TEM) of mice hippocampus of the studied groups. a Control DG; granular neurons (G) with large rounded euchromatic nuclei (N) and prominent nucleolus × 3600. b THIM DG; granular neurons (G) with largely rounded nuclei (N), while others are degenerated with electron dense cytoplasm and ill-defined organelles (arrow heads) × 4800. c Monte DG; most of granular neurons (G) with rounded euchromatic nuclei (N), one degenerated cell can be seen (arrow head) × 4800. d Control CA1; pyramidal neuron (P) with a large rounded euchromatic nucleus (N) and large cytoplasm contains rER, mitochondria (m), and multiple ribosomes × 4800. e THIM CA1; degenerated pyramidal neurons (arrow heads) with electron dense cytoplasm and ill-defined organelles surrounding large pyramidal neuron (P) with rarified cytoplasm × 4800. f Monte CA1; pyramidal cells (P) with large rounded nuclei (N), but with rarified cytoplasm (*) × 4800

Full size image

Discussion

Our study confirmed the role of THIM in mediating ASDs like behavioral alterations when given to neonatal mice in the early postnatal period. It is the first to investigate the potential therapeutic role of montelukast in alleviating autistic manifestations induced by THIM through upregulating α7nAchRs expression and downregulating the expression of NF-κB p65, Bax, and markers of brain injury.

Autism is a lifelong neurodevelopmental impairment that typically occurs before the fourth year of life (Barger et al. 2013). It is characterized by social impairment, cognitive shortage, and stereotyped behavior with restricted interests (Nicolini and Fahnestock 2018). The ASDs prevalence in USA has climbed markedly during the past decades from 0.67% in 2000 to 1.46% in 2012 which could not be elucidated by enhanced people awareness or improvement in diagnostic procedures (Christensen et al. 2016; Xu et al. 2018; Zablotsky et al. 2015).

Non-genetic factors have a significant role in the etiology of ASDs as submitted by many chromosomal studies which reported that approximately 80% of the affected children have a normal genome (Shen et al. 2010). Based on epidemiological evidence; methylmercury, lead, arsenic, toluene, and polychlorinated biphenyls “industrial elements” were recognized to have a deleterious impact on developing brains (Plunkett 2007). Numerous studies have correlated their exposure to the severity of symptoms in autistic patients (Adams et al. 2009; Elsheshtawy et al. 2011; Nataf et al. 2008; Ozonoff et al. 2010; Priya and Geetha 2011).

Majority of studies reported mercury as a predominant risk factor (Kern et al. 2012). Kern et al. (2016a, b) reported that autistic symptoms were worse in kids with higher blood and/or nail mercury levels. The more severely affected the child was, the lower the levels of hair mercury that were detected as explained by the “poor excretory theory” in ASDs kids (Kern et al. 2016a, b). Thimerosal (THIM, sodium ethylmercurithiosalicylate; 49% mercury by weight) has been extensively used as a preservative in many vaccines and other pharmaceutical products (Rosenblatt and Stein 2015). It is used as an adjuvant to vaccines to enhance the immune response to an antigen (Rosenblatt and Stein 2015). Although THIM has been forbidden in mandatory childhood vaccines in USA, it is still used in some developing countries for the benefit of the multiuse of the THIM-preserved vaccine vials (Geier et al. 2015a, b).

THIM is rapidly metabolized into ethyl mercury and thiosalycylic acid (Reader and Lines 1983; Tan and Parkin 2000) and subsequently into inorganic mercury which is deposited for many years in vital organs especially the brain (Dórea et al. 2012; Kern et al. 2016a, b). After exposure, wide deleterious reactions usually affect the nervous system due to its high sensitivity and the maldevelopment of the neonatal blood–brain barrier (Filipek 1995). Mercury toxicity may lead to autoimmunity and disruption of the antioxidant system due to extensive changes in proteins’ three dimensional structures (Filipek 1995). On the contrary, it was stated that low-dose exposure of THIM via vaccination did not significantly influence behavior because administration of THIM-containing vaccine to infant rhesus macaques did not induce autism-like behavior (Gadad et al. 2015). Thus, so far, the link of THIM to neonatal ASDs is still an open question. Animal models have been developed to evaluate the possible neurotoxicity induced by THIM mimicking that occurs through inoculation of THIM vaccines in children (Berman et al. 2008).

ASDs are multifaceted diseases that comprise many disrupted mechanisms as neuroinflammation and neurotransmitters dysregulation (Morgan et al. 2010; Vargas et al. 2004; Young et al. 2011). Chemical neurotransmitters, such as serotonergic (Whitaker-azmitia 2001), dopaminergic (Buitelaar and Willemsen-swinkels 2000), GABAergic (Fatemi et al. 2002), glutamatergic (Purcell et al. 2001), and cholinergic (Lee et al. 2002; Perry et al. 2001) systems were considered to be involved in autism (Tsai 1999). Chronic neuroinflammation has a key role in ASDs pathophysiology as it mediates ROS overproduction and stimulates activated glial cells to release of proinflammatory lipid mediators such as LTs (Chiurchiù and MacCarrone 2011; Kyrkanides et al. 2002). Antagonism of CysLT1 receptors provides benefits against neuroinflammation and apoptosis, by suppressing activation of NF-κB, tumor necrosis factor alpha (TNF-α), interleukin 1 beta (IL-1β), and caspase-3 in the hippocampus and cortex regions (Lai et al. 2014).

Montelukast, a selective CysLT1R antagonist, has been studied to provide neuroprotective and anti-apoptotic effects mediated by decreasing ROS generation, suppressing brain lipid peroxidation (Erşahin et al. 2012) and increasing brain superoxide dismutase (SOD) activity (Çevik et al. 2015). It improved learning and memory impairment induced by chronic cerebral hypoperfusion through ameliorating brain oxidative stress and the associated cholinergic dysfunction (Singh and Sharma 2016). Montelukast ameliorated Aβ-induced cognitive deficit (Lai et al. 2014) and offered neuroprotection against kainic acid-induced impairment of cognitive function (Kumar et al. 2012).

Considering neuroinflammation as an important factor in both etiologies of autism and the neurochemical processes of mercury-mediated toxicity (Duszczyk-Budhathoki et al. 2012), we have designed our current work to study the potential therapeutic utility of montelukast in mitigating THIM-induced autism-like behavior in mice model and to study the possible underlying signaling mechanisms.

Our study established that THIM mice showed retarded weight gain in consistant with the previous results showing THIM-induced growth retardation in autoimmune disease-sensitive SJL/J mice (Hornig et al. 2004). THIM administration produced neurodevelopmental delay through dysregulating the synaptic function and the endocrine system (Lai et al. 2014). Previous literature reported that injection of infant monkeys with one dose of THIM-containing hepatitis B vaccine predisposed them to neurodevelopmental and survival impairments (Burbacher et al. 2005). In our study, THIM induced growth delay in mice that could not be ameliorated with montelukast administration. On the contrary, montelukast helped young mice to gain weight after induction of growth retardation by cranial irradiation (CIR) to their brains possibly due to its anti-inflammatory action (Eriksson et al. 2018).

Social deficit is a major symptom of ASDs (Frye 2018). THIM group displayed the typical autism-like phenotype and failed to show significant preference towards the stranger mouse compared with the object. In consistence with our work, Namvarpour et al. (2018) reported that THIM-exposed mice showed reduced social activity and social preference. Poor sociability could be credited to fear, anxiety, or impaired understanding of communicating signals from the stranger mouse (Wu et al. 2017). It could also be explained by THIM-induced neurodegeneration as well (Araghi-niknam and Fatemi 2003) as it was administered at a critical time for the development of neural pathways and synthesis of synapses (Hafidi and Hillman 1997). In our study, montelukast ameliorated the social impairment induced by early neonatal exposure to THIM. This finding is considered novel as it reveals the ability of montelukast to improve a major THIM-induced behavioral change.

THIM group exhibited a significant increase in the central movement in the open field test which is a strong indication of the anxious behavior commonly detected in autistic kids. Significant increase in locomotor and exploratory movements with enhanced speed was observed in THIM group compared to Cont. group as well. These behavioral impairments were mitigated through montelukast administration. On the contrary, motor deficit has been induced by mercury species (especially methylmercury) in humans, non-human primates, and laboratory rodents and pigeons (Gandhi and Dinesh 2014). Similarly, Luciana et al. (2005) and Rebane et al. (2014) described impaired motor activity and decreased motor performance induced by disturbances in the cerebellar free radical scavenging system in mice after exposure to methylmercury. These results could be elucidated by species differences as within SJL mice, traveling distance was reduced with THIM, however, no significant effects were noted in C57BL/6J or BALB/cJ mice (Hornig et al. 2004).

Cognitive function and memory in animals were commonly evaluated by MWM test (Wu et al. 2017). Early postnatal THIM administration insignificantly disabled acquisition and retrieval of spatial learning and memory. These results agree with Hornig et al. (2004) who stated that THIM does not affect acquisition or retention of a simple spatial memory task. Montelukast ameliorated cognitive deficit induced by kainic acid, middle cerebral artery, or bilateral common carotid artery occlusion (Kumar et al. 2012; Lai et al. 2014; Singh and Sharma 2016). It also enhanced cognitive function in aged rats through reduced microglia activation, reduced neuroinflammation, elevated neurogenesis, and through restoration of blood–brain barrier integrity (Marschallinger et al. 2015).

Inflammation is reported as a pivotal factor in ASDs pathogenesis; autopsy brain samples of autistic subjects showed enhanced expression of inflammatory markers in cerebrospinal fluid (Young et al. 2011). NF-κB is a protein present in almost all cells and possesses immune-regulatory functions by inducing expression of inflammatory cytokines, chemokines, and apoptosis markers (Young et al. 2011). The inactive form of NF-κB is confined to the cytoplasm and comprises 3 subunits: DNA binding; p50 and p65 subunits; and the inhibitory subunit; inhibitor of NF-κB (IκBα) (Vermeulen et al. 2002). Phosphorylation of IκBα induced its ubiquitination and degradation and then nuclear translocation of the active NF-κB dimmers (Vermeulen et al. 2002). The crucial regulatory step involves activation of a high-molecular weight I Kappa B kinase (IKK) complex which comprises three tightly associated IKK subunits. IKKα and IKKβ act as the catalytic subunits of the kinase (Malik et al. 2011). In our study, NF-κB p65 expression was significantly increased in THIM group cerebral homogenate. Young et al. (2011) reported that NF-κB has been unreasonably expressed in the orbitofrontal cortex of autistic children brains. Another study did not find aberrant NF-κB in the brains of autistic children; however, increased expression of IKKα kinase in the cerebellum was observed (Malik et al. 2011). In our research, montelukast treatment induced downregulation of NF-κB p65 upregulated by THIM treatment. Montelukast rescued neurons against Aβ1–42-induced neuronal inflammation and apoptosis through downregulating NF-κB (Lai et al. 2014).

Olczak et al. (2010a, b) stated that about 70% of the brains of THIM-exposed mice showed dark neurons in the prefrontal cortex, hippocampus, and temporal cortex. Increased tendency towards cell death was recognized in the frontal cortex of many autistic patients as well (Olczak et al. 2010a, b). Aragão et al. (2018) reported that THIM induced neuronal death by overproduction of ROS and nitrogen species and/or by impairment of the antioxidant defense system in the hippocampus of rats after exposure to methyl mercury. In our study, THIM group demonstrated deeply stained shrunken granular and pyramidal cells in DG and CA1 fields associated with pericellular vacuolation. Electron microscopic results showed multiple degenerated neuronal cells with ill-defined organelles or with marked rarified cytoplasm. These results are consistent with the previous results (Chen et al. 2013; Hornig et al. 2004) that proved marked reduction in the number of hippocampal neurons and changes in nerve cells after administration of THIM to P1 and P20 rats. Similar microscopic results were achieved by Wu et al. (2016) after methyl mercury administration to P20 rats. Monte group demonstrated low levels of dark cells with amelioration of most of degenerative changes compared to THIM group. These results supported montelukast neuroprotective effects proved by Yu et al. (2005) against ischemic/reperfusion injury (Liu et al. 2015). Montelukast promoted hippocampal neurogenesis specially progenitor cells proliferation resulting in increase in the number of generated neurons (Marschallinger et al. 2015). Its effect was ascribed to its antioxidant and anti-inflammatory effects through inhibiting microglial activation (Erşahin et al. 2012; Marschallinger et al. 2015).

Methyl mercury induced cytotoxicity, cell cycle arrest, and induction of cell death via both p53-dependent and p53-independent pathways (Gribble et al. 2005). Similarly, exposure to methyl mercury inhibited DNA synthesis with reduction of the size of hippocampus credited to the apoptotic changes induced in DG granular cells (Falluel-morel et al. 2007). Early postnatal THIM administration blocked neuronal glutamate transporter receptors and increased glutamate levels which enhanced neuronal pro-apoptotic proteins (Bax and Caspase-3) and decreased anti-apoptotic protein BCl2 (Shinohe et al. 2006). Our results revealed significant increase of Bax-positive immunostained cells after THIM administration. Monte group presented significant decrease of Bax immunopositive cells which indicates montelukast anti-apoptotic effect.

Current literature supported the existence of cholinergic dysfunction in ASDs (Deutsch et al. 2010, 2014; Perry et al. 2001) mainly due to loss of nAChRs in cerebral and cerebellar cortex. nAChRs are expressed on various cell types both peripherally and centrally and have diverse functional roles in neuroprotection, inflammation, and immunity (Benowitz 2009). They are pentameric neurotransmitter-gated ion channel receptors composed of nine α (α2–α10) and three β (β2–β4) subunits (Dani and Bertrand 2007; Yakel 2013). The predominant subtypes functionally expressed in the brain are categorized as α7 subunit-containing receptors (either homo- or heteromeric) (Colombo et al. 2013; Jones and Yakel 1997; Jones et al. 1999; Khiroug et al. 2002; Sudweeks and Yakel 2000). They are supposed to participate in many neuroprotective mechanisms (Parri et al. 2011; Shen and Yakel 2009) and considered presynaptic modulators responsible for release of neurotransmitters such as glutamate, γ-aminobutyric acid, dopamine, and norepinephrine (Wessler and Kirkpatrick 2008). In addition, α7nAChRs are expressed in non-neuronal cells including astrocytes, microglia, oligodendrocyte precursor cells, endothelial cells (Hawkins et al. 2005), and immune cells and can regulate the release of inflammatory mediators (Gahring et al. 2013). Downregulation of α7nAChRs is associated with multiple neurological disorders including schizophrenia, learning disability, ADHD, AD, epilepsy, and ASDs (Helbig et al. 2009). Thus, α7nAChRs upregulation could be beneficial in alleviating autism-associated neuroinflammation (Stigler et al. 2011). Our research has detected significant downregulation of α7nAChRs in THIM group compared to Cont. group in both DG and CA1 fields. These results are consistent with Ray et al. (2005) who reported loss of α7nAChRs immunoreactive neurons in the paraventricular nucleus in autism. However, Martin-ruiz et al. (2004) found no change in the expression of the α7 subunit; however, they reported a 40–50% decline in the expression of the α3, α4, and β2 nAChR subunits in the adult post-mortem autistic cerebral cortex. Montelukast treatment induced upregulation of α7nAChRs in both DG and CA1 fields which could have ameliorative potential over autism-associated neuroinflammation and cognitive dysfunction.

S100b is a group of proteins that regulate the cytoskeleton and proliferation of astrocytes and glial cells which provide neurons with trophic and metabolic support (Sedaghat and Notopoulos 2008). It is vital for brain cytoarchitecture and synaptogenesis (Verkhratsky and Nedergaard 2014). S100b proteins are concentrated in astrocytes and are considered markers of glial activation following brain damage (Guloksuz et al. 2017). An elevated level of S100b proteins is a common finding in autistic children and is significantly linked to disease severity (Al-ayadhi and Mostafa 2012). Our results revealed increased immunoreactivity for S100b protein in DG and CA1 fields in THIM group. However, Monte group showed insignificant decrease in immunoreactivity compared to THIM group, but showed significant increased immunoreactivity compared to Cont. group indicating enhanced glial cell activation.

Synaptophysin is an abundant synaptic vesicle protein that participates in exocytosis of synaptic vesicles (Tarsa and Goda 2002). Decreased synaptophysin expression may result in suppression of neurotransmitters release and synaptic functions downregulation (Janz et al. 1999) with subsequent cognitive deficit as previously detected in AD mice models (Tampellini et al. 2010). Marschallinger et al. (2015) described significant reduction of synaptophysin expression in the hippocampus of aged rats. Our results showed a significantly decreased intensity of synaptophysin in the hippocampus of THIM mice. Similarly, a marked reduction of synaptophysin expression in hippocampal tissues in THIM rats was observed (Olczak et al. 2010a, b). Montelukast treatment increased synaptophysin-positive immuonostaining. On the contrary, the synaptic density of the hippocampus of old rats could not be restored by montelukast despite improvement of cognitive function (Marschallinger et al. 2015).

In conclusion, the current study has offered further experimental evidence signifying that THIM exposure induced abnormal behavioral symptoms consistent with those observed in ASDs. Furthermore, it was observed as a novelty that these abnormalities could be ameliorated by montelukast treatment. Finally, given the ability of THIM to cause neurotoxicity in the developing brain, every effort should be made to exclude this dangerous component from vaccines and other medicinal products as its risk outweighs its benefit.

Limitations of the study

There are several limitations in this study. First, we did not consider the genetic alteration that may occur due to the possible DNA damage induced by THIM; however, we will take it in our consideration in the future studies. Second, the insignificance of the cognitive impairment results may be related to the relatively small sample size.