Introduction
Autism spectrum disorder (ASD) is a prevailing neurodevelopmental disorder, primarily diagnosed by a core of symptoms including social impairments, communication difficulties, restricted interests and repetitive behaviors [
1]. ASD patients often show cognitive and mental deficits comorbid with other neuropsychiatric disorders, such as attention-deficit/hyperactivity disorder (ADHD), anxiety, and bipolar disorder [
2]. A system-level analysis of brain transcriptome has pointed out that the patterns of gene expression in schizophrenia, bipolar disorder and ASD significantly overlap, and that neurons/synapses are susceptible targets of polygenic modulations in all cases [
3]. The common genetic variants and phenotypic traits among these disorders indicate shared neuropathology in the cell signaling pathways.
The dopamine (DA) system is an intriguing candidate. In the brain, tyrosine hydroxylase (TH) catalyzes the hydroxylation of tyrosine to L-DOPA, which is further converted to the modulatory neurotransmitter DA. DA binds to a large family of G-protein coupled receptors that are classified into two subgroups: D1-like (D1 & D5) and D2-like (D2-D4) receptors. Dopaminergic neurons primarily originate from substantia nigra pars compacta (SNc) and the ventral tegmental area (VTA) in the midbrain. Projections from SNc to the striatum (STR) form the nigrostriatal pathway that is important for controlling voluntary movement. Projections from VTA to the nucleus accumbens (NAc) and the frontal cortex make up the mesocorticolimbic pathway to regulate memory, reward, motivation and emotion [
4]. Similar to ADHD [
5], bipolar disorder [
6] and schizophrenia [
7], DA dysfunction is linked to ASD [
8]. The STR and the frontal cortex that receive dopaminergic inputs are altered in human ASD [
9‐
11] as well as in animal models with autistic-like behaviors induced by environmental factors [
12]. In vivo imaging data demonstrate presynaptic alterations of DA synthesis and DA transporter (DAT) in the striatal and frontal cortical regions [
13]. Interestingly, drugs involved in DA actions, such as risperidone, clozapine, haloperidol and methylphenidate, have yielded beneficial effects in ASD patients [
14], although none of them acts selectively on the DA system.
To investigate how the dopaminergic pathways are modified and whether application of DA can have a therapeutic effect in ASD, we employed two distinct mouse models for ASD, i.e., Black and Tan BRachyury T
+Itpr3
tf/J (BTBR) and Fragile X Mental Retardation 1 knockout (
Fmr1-KO) mice. The BTBR strain is a phenotypic model for idiopathic ASD, which exhibits impaired sociability, altered ultrasonic vocalization and increased self-grooming behaviors, simulating the main symptoms of human ASD [
15]. We also found that BTBR mice display cognitive and emotional abnormalities akin to the psychiatric comorbidity of ASD [
16,
17]. In addition, similar neuroanatomical changes between the BTBR model and ASD subpopulations are reported [
18,
19]. As to the DA system, BTBR mice show reduced D2, but not D1, receptor-mediated neurotransmission [
20]. On the other hand,
Fmr1-KO mice are a genetically defined model for Fragile X syndrome (FXS) [
21]. FXS is the result of transcriptional silencing of
Fmr1 gene and loss-of-function of its product, FMR protein (FMRP). Given that FXS is a leading inherited form of mental retardation and autism [
22],
Fmr1-KO animals are widely used for ASD-relevant studies. Characterizations of the
Fmr1-KO mouse line have revealed a decreased number of SNc cells [
23], compromised extracellular DA release [
24] and disrupted D1 receptor-mediated synaptic transmission in the prefrontal cortex [
25,
26].
In comparison with wild type (WT) control mice, we used biochemistry, immunohistochemistry and imaging methods to analyze dopaminergic, glutamatergic and GABAergic neurons in the DA pathways in BTBR and Fmr1-KO animals, with respective antibodies against TH, vesicular glutamate transporter 1 (VGLUT1) and glutamic acid decarboxylase 67 (GAD67). Fractal analysis of TH-positive axons in the STR was applied to reveal morphological changes of the dopaminergic projections and their spatial relationships with VGLUT1-immunoreactive nerve terminals. Moreover, we evaluated the effects of intranasal application of DA on the behavior and protein expression in BTBR and Fmr1-KO mice. Our results indicate that the DA system is altered differently yet intranasal treatment with DA improves the behavioral deficits in both mouse models, presenting a potential therapy for ASD.
Methods
Subjects
BTBR (stock # 002282),
Fmr1-KO (stock # 003025) and C57BL/6 J (stock # 000664) mice were purchased from Jackson Laboratory (Bar Harbor, ME) and were housed in a facility accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care. All mice were kept under a 12-h light-dark cycle (light on from 07:00 to 19:00) and reared 3–5 per cage with food and water ad libitum. Male and female mice were maintained in the same room. All procedures were approved by the Institutional Animal Care and Use Committee and the Institutional Biosafety Committee of University of Minnesota, in accordance with the National Institutes of Health guidelines. C57BL/6 J mice served as control because they are the most commonly adopted control for BTBR mice [
27], and share the same genetic background with
Fmr1-KO mice (
https://www.jax.org/strain/003025). Unless specified, male mice (2–4 months old) were used for experiments due to the male-dominant prevalence of ASD [
28].
Immunohistochemistry
Mice (n = 3 mice/group for each set of experiments) were anaesthetized with ketamine (100 mg/kg, i.p.) and xylazine (10 mg/kg, i.p.), and transcardially perfused with phosphate-buffered saline (PBS, pH 7.4), followed by 4% paraformaldehyde. Brains were removed, post-fixed in 4% paraformaldehyde overnight at 4 °C, then immersed in 30% sucrose solution and stored at 4 °C until they sank. 50 μm-thick coronal sections were made on a microtome (Leica VT1200 S, Buffalo Grove, IL). They were maintained in 0.3% H2O2 solution for 10 min, rinsed with PBS and incubated in blocking solution containing 2% goat serum and 0.2% Triton-X 100 at 37 °C for 30 min. Sections were labelled with a rabbit TH antibody (1:800; Abcam ab112, Cambridge, MA) at 4 °C overnight. After being washed with 0.2% Triton-X 100 PBS solution, they were incubated in biotinylated anti-rabbit blocking solution for 2 h (1:200; Vector Lab. BA-1000, Burlingame, CA). They were then transferred to an ABC reagent (Vector Lab. PK-6100), stained with 3,3′-Diaminobenzidine (Vector Lab. SK-4100), mounted on slides and cover-slipped with Vectashield (Vector Lab. H-5000). To add fluorescence, after incubation in the primary antibody, sections were incubated in goat anti-rabbit Alexa Fluor 555 (1:1000; Invitrogen A21428, Waltham, MA). 3 h later, they were washed in PBS and cover-slipped with Vectashield (Vector Lab. H-1500). Immunostaining procedures for VGLUT1 and GAD67 were the same, except that primary antibodies guinea pig anti-VGLUT1 (1:5000; Millipore AB5905, St Louis, MO) and mouse anti-GAD67 (1:1000; Millipore MAB5406); and secondary antibodies goat anti-guinea pig Alexa 488 (1:1000; Invitrogen A11073) and goat anti-mouse Alexa 555 (1:1000; Invitrogen A32727) were used.
Imaging analysis
Images were taken with a Zeiss LSM 710 confocal microscope with 20x and 63x oil immersion objectives or a Leica DMi8 light microscope with a 10x lens. Experimenters, who were blind to the design, analyzed the data using NIH ImageJ software (
https://imagej.nih.gov/ij/). A region of interest (ROI) was selected based on a mouse brain atlas [
29]. Areas of anti-TH, −VGLUT1 and -GAD67 signals were manually circled and summated with ImageJ. For each brain region, 2–3 images were sequentially taken with an inter-section interval of 200 μm. For each image, the intensity of ROI was subtracted from its background. The values of ROI intensity were then normalized to the average values of WT controls. Synaptic boutons were defined within the size of 0.2–3 μm
2 of VGLUT1-positive signals. Co-localization of TH- and VGLUT1-positive boutons were identified with overlaps of the two fluorescent signals.
Fractal analysis of neuronal morphology is a validated methodology [
30]. Striatal TH-positive axons were imaged with the 63x lens (6 Z-planes, 0.35 μm steps, field size: 30 × 30 μm
2, pixel size: 1024 × 1024). Images were stacked to 2D and processed through deconvolution with ImageJ. A plugin FracLac (
https://imagej.nih.gov/ij/plugins/fraclac/FLHelp/Installation.htm) with an implemented box counting method [
31] was applied to calculate box-counting dimension (Db) and lacunarity. Db and lacunarity are fractal dimensional parameters used for quantifying complexity and inhomogeneity of digital spatial patterns, respectively.
Western blotting
The striatum (
n = 4–6 mice/group) was dissected bilaterally at 4 °C and stored at − 80 °C until processed. As described earlier [
32], tissues were homogenized in ice-cold lysis buffer, centrifuged at 14,000 rpm for 10 min and the supernatants were collected for determination of protein concentration. Aliquots of protein (20–30 μg) were subjected to 10% SDS-PAGE and transferred to Immun-Blot PVDF Membrane (Bio-Rad, Hercules, CA). Membranes were incubated in 10% dry milk in PBST solution at room temperature for 1.5 h and then in primary antibodies at 4 °C overnight. Subsequently, they were washed with PBST for 30 min and incubated in secondary anti-mouse (1:5000; Abcam ab205719) and anti-rabbit (1:5000; Abcam ab205718) antibodies at room temperature for 1.5 h. Following a rinse with PBST for 30 min, they were reacted with enhanced chemiluminescent reagent (GE Healthcare Life Sciences, Buckinghamshire, UK) and imaged with Odyssey Fc Imaging System (LI-COR Biosciences, Lincoln, NE). Blot intensity was quantified using ImageJ and normalized to β-actin.
Primary antibodies included rabbit anti-TH (1:500; Abcam ab112), guinea pig anti-VGLUT1 (1:1000; Millipore AB5905), mouse anti-GAD67 (1:5000; Millipore MAB5406), mouse anti-DAT (1:1000; NovusBio mAb16, Littleton, CO) and mouse anti-β-actin (1:5000; Sigma-Aldrich A5441, St. Louis, MO).
Dopamine administration
Mice were handled once a day for at least three consecutive days before drug administration. Dopamine hydrochloride (Sigma-Aldrich H8502) was suspended in a viscous castor oil-based formulation (MetP Pharma, Emmetten, Switzerland). It was freshly prepared in a dose of 3 mg/kg in a volume of 10 μl and kept on ice with protection from light. DA or vehicle suspension was applied 5 μl per nostril for ~ 30 s with a pipette (Eppendorf North America, Hauppauge, NY). 10 min after administration, animals underwent elevated plus maze, open field or object-based attention tests. Because of their respective deficit in social approaching or social novelty [
16,
33], in the three-chamber test, BTBR mice received DA before the habituation trial, whereas
Fmr1-KO mice received DA before the sociability trial. As BTBR mice have intact recognition of social novelty [
16,
17], the social novelty trial was excluded here to avoid excessive administration of DA in the same subjects within a short time. The dosage and timing of DA administration were based on a previous study in mice [
34] .
Behavioral testing
BTBR mice (
n = 12; 8 males and 4 females) were tested using a within-subject design to minimize individual differences. Choice of sex was based on our earlier report that both male and female BTBR mice show autistic-like behaviors [
17]. Half of the animals were treated with DA and the other half with vehicle. The behavioral tests were conducted in the order: elevated plus maze, open field, object-based attention and three-chamber social test, with an inter-test interval of 24–48 h. One week after the final test, the previous DA group received vehicle and vice versa, followed by the same behavioral testing.
Fmr1-KO mice were randomly divided into vehicle (
n = 7 males) or DA (
n = 9 males) group and tested using a between-subject design. Choice of sex was based on the location of
Fmr1 gene on the X chromosome with male-dominant occurrence [
21]. The order of behavioral tests and the inter-test interval were the same as for BTBR mice. Each
Fmr1-KO animal was exposed to each test only once.
Behavioral testing was done in a quiet room (< 40 dB) during 10:00–16:00. LED light provided dim illumination (~ 50 lx). Male mice were tested before female ones. Apparatus was cleaned with 70% ethanol between animals to remove pheromones. A camera was connected to a computer for tracking animals and recording videos (Stoelting ANY-maze, Wood Dale, IL).
Elevated plus maze
This test was used to assess anxiety-like behaviors [
35]. The maze has two open arms (30 × 5 cm), two closed arms (30 × 5 cm) and a central platform (5 × 5 cm) and is elevated at 30 cm height. The subject was placed on the central platform facing the open arms and given 5 min to travel freely. Entries to and time spent in the center, open and closed arms, and head-dips, were counted.
Open field
This test was used to assess locomotion and exploration behaviors. The subject was placed in a polyvinyl chloride box (40 × 40 × 30 cm) for 15 min. Distance traveled, thigmotaxis (distance travelled along the walls), duration of grooming, time spent in the center (virtual central square 13.3 × 13.3 cm) and counts/duration of rearing were analyzed for 15 min in three 5-min bins. Rearing is a sign for non-selective attention [
36]. Self-grooming is a measure of repetitive behavior [
37]. Thigmotaxis is an index for sensorimotor function and/or anxiety [
38,
39]. Time spent in the center indicates the anxiety level [
40]. One BTBR mouse was excluded from the analysis due to lack of rearing.
Object-based attention test
This test was used to assess attention-associated processes and/or short-term working memory [
16]. Objects of different materials (plastic or glass), textures (smooth or rough), sizes (diameter 7–9 cm, height 14–17 cm) and shapes (column or irregular) were placed in the open field. Objects weighed enough to prevent being displaced by the animal. Assignment of the objects was counterbalanced to minimize a potential bias for their identity or location. Prior to testing, the animal was habituated to the open field without objects. The actual test consisted of a learning trial (5 min) and a test trial (5 min) without time delay in between. In the learning trial, the animal was introduced to the field which contained two distinct objects. In the test trial, the objects were replaced by a new one and a copy of either of the explored ones in the same location. During the replacement, the animal remained inside the arena. Object exploration was defined as physical contacts with the objects by the animal’s nose, head and forepaws, but not by the body or tail. Climbing or sitting aside the objects was not included. Animals that explored the objects for < 10 s in either trial were excluded from the analysis. Here, the index = [time spent on the novel object - time spent on the old object] / total time spent on both objects, with a positive value representing intact performance.
Three-chamber sociability and social novelty test
This test was used to assess sociable behavior and recognition of social novelty [
41], in a polyvinyl chloride apparatus composed of three chambers (20 × 40 × 30 cm each) with passages (5 × 5 cm) dividing the chambers. The test included three sessions: habituation, sociability and social novelty (9 min each). In the habituation trial, the subject was placed into the middle chamber and allowed to freely explore the whole apparatus. In the sociability trial, a gender- and age-matched WT mouse that had never been contacted by the subject was put underneath a metal grid cup (diameter 10 cm, height 12 cm) in one of the side chambers. Another identical cup was put in the opposite side chamber. The locations for placing the stranger mouse and the empty cup were counterbalanced between subjects. In the social novelty trial, another stranger mouse was placed underneath the previously empty cup. The same strangers were used between subjects. Physical contacts around the cups by the subject’s nose, head and forelimbs were defined as explorative behaviors. Sociable index = [time for exploring the stranger mouse - time for exploring the empty cup] / total exploration time. Social novelty index = [time for exploring the novel mouse - time for exploring the familiar mouse] / total exploration time. Positive values represent intact sociability and social novelty preference.
Statistics
Repeated two-way ANOVAs with “within-subject” factors (treatment, interval or object) were used in the analyses of BTBR behaviors. Mixed two-way ANOVAs with a “between-subject” factor (group) and a “within-subject” factor (interval or object) were used in the analyses of
Fmr1-KO behaviors. One-way ANOVAs, paired
t-tests and one-sample
t-tests were applied when appropriate. Independent
t-tests (if allowed by results of one-way ANOVAs) or Mann-Whitney U tests (in case of lack of homogeneity or normality of variance) were used for analyzing immunohistochemistry and Western blotting data. Data were expressed as mean ± standard error of mean (SEM). Statistical significance was set as
p < 0.05. All tests were two-tailed tests. For imaging analyses,
n denoted the number of samples from 3 mice per group. Otherwise,
n represented the number of mice per group. Sample sizes were determined on the basis of previous studies using similar experimental protocols [
16,
17,
33].
Discussion
In this comparative study, we have unraveled distinct alterations and common phenotypes in the DA pathways of two widely adopted mouse models for ASD. BTBR mice showed a hypofunction of the DA system, as indicated by the low expression of TH in several DA centers (Figs.
1,
2,
3), in line with previous studies showing compromised DA-mediated responses in these mice [
20]. Moreover, they exhibited decreased motivation for social and food rewards in operant conditioning tasks [
50] and less social conditioned place preference [
51], which reinforces the perspective of dysfunctional DA system in this model. As TH is an enzyme for synthesis of both DA and norepinephrine, future studies are required to differentiate their roles in the BTBR brain. For
Fmr1-KO mice, the TH level did not change significantly (Fig.
3), largely consistent with other reports [
23,
24,
49]. Yet, fractal analysis revealed unusual arborization of TH-positive axons in their STR (Fig.
2), strengthening an essential role of FMRP in axon formation [
52]. For instance, a loss of FMRP homologue dFMR1 in
Drosophila generated aberrant extensions and branches of axons [
53,
54], while overexpression of dFMR1 led to abridged axonal arbors [
55]. In cultured rat cortical neurons, FMRP overexpression attenuated the axon complexity [
56]. Furthermore, the axon integrity was altered in the cortex of
Fmr1-KO mice [
57] and in the dSTR of FXS patients [
58]. Despite of the controversy [
59], FMRP also plays a role in the development of axon myelination [
60,
61]. Whether defective myelination could contribute to the abnormal morphology of TH-positive axons in the
Fmr1-KO STR remains unclear. The differences in TH expression between BTBR and
Fmr1-KO mice could relate to their individual genetic background as well.
Co-labeling VGLUT1 with TH showed an increased number of VGLUT1-containing nerve terminals in close spatial relationship with TH-positive axons, indicating an enhanced interaction between the cortical afferents and released DA in the STR of the two ASD mouse lines (Fig.
2). The molecular underpinnings of DA modulation are complex, for example, depending on the subtypes of DA receptors [
62]. Whether DA facilitates or attenuates glutamatergic neurotransmission will be subject to further investigations. Another commonality between the two models was the downregulation of striatal DAT (Fig.
3). DAT is critical for maintaining DA homeostasis by recycling DA from the synaptic cleft to the cytosol. Whole-exome sequencing has identified a DAT mutation in ASD families [
8]. Transgenic mice with DAT deficiency showed hyperactivity [
63]. Administration of amphetamine, which causes DAT-mediated DA efflux, alleviated self-grooming in BTBR mice [
64] and facilitated object recognition in
Fmr1-KO mice [
48]. More studies are needed to elucidate the mechanisms and consequences of the DAT defect.
The causes for the protein regulation by intranasal DA in the ASD models may be diverse. The bidirectional modulation of the quantity of TH in the BTBR and
Fmr1-KO brain (Fig.
4) likely depends on the different changes in their endogenous DA system (Fig.
3). DA administration in BTBR mice may increase the extracellular DA concentration and DA availability [
65], which presumably increases the TH activity in the STR [
66]. In
Fmr1-KO mice, a lack of FMRP may play a role in reducing the level of TH, considering the interaction between FMRP and DA signaling [
25,
67]. This possibility is supported by an observation that intranasal DA did not change TH protein in normal rats [
68]. As to DAT, increased extracellular DA could affect its binding activity [
69]. In spite of the evidence for striatal DAT deficiency in the two strains (Fig.
3), intranasal delivery of DA did not restore DAT expression (Fig.
4).
Intranasal application of DA efficiently rescued the cognitive and social deficits of the BTBR (Fig.
5) and
Fmr1-KO (Fig.
6) models. It should be noted that the behavioral assessments of these functions can be confounded by other factors. For instance, hyper- or hypo-locomotor activity may influence animals’ performance in the object-based attention test and three-chamber social test. However, this unlikely compromised the effects of DA on cognition and social interaction as the overall motor and exploratory behaviors were comparable between the DA and vehicle groups (Figs.
5 &
6). In rats treated with intranasal DA, an elevated concentration of DA was found in the cerebrospinal fluid and in the brain [
70], including the dSTR and NAc [
65]. Moreover, intranasal DA had antidepressant-like effects [
71], attenuated fear responses [
72], compensated behavioral asymmetries in a Parkinsonism model [
73], and alleviated cognitive deficits in aged rats [
74] as well as in animal models for schizophrenia [
75] and ADHD [
76]. Therefore, the behavioral rescues by intranasal DA likely stem from brain-wide actions. Here, we focused on the STR because it is a key neuronal correlate to social behaviors [
77] and a disparate fronto-striatal circuit has been specified in ASD patients [
78,
79]. Mice with deletion of ASD-relevant genes, e.g.,
Shank3, displayed social impairments, along with decreased corticostriatal neurotransmission, increased morphological complexity of medium spiny neurons, and reduced glutamate receptors in the STR [
77]. Although we (Fig.
3) and others did not note dramatic changes in TH expression in
Fmr1-KO animals [
23,
24], DA release and uptake were blunted in their STR [
24]. The syndromic ASD model, BTBR mice with polygenetic mutations [
27], presented severe detriments in production and reuptake of DA as manifested by reduced striatal TH and DAT proteins, respectively (Fig.
3). These findings suggest that the defective STR and its connected circuitry are major features of ASD. Furthermore, the susceptibility of striatal TH protein to intranasal DA treatment in BTBR and
Fmr1-KO mice (Fig.
4) indicate that the STR can be an effective target for therapeutic interventions of ASD. Our results provide not only empirical evidence for the DA hypothesis of ASD [
80], but also a proof of principle for developing clinical treatments for the disorder.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.