Altered Expression and In Vivo Activity of mGlu5 Variant a Receptors in the Striatum of BTBR Mice: Novel Insights Into the Pathophysiology of Adult Idiopathic Forms of Autism Spectrum Disorders

Abstract

Background

mGlu5 metabotropic glutamate receptors are considered as candidate drug targets in the treatment of “monogenic” forms of autism spectrum disorders (ASD), such as Fragile-X syndrome (FXS). However, despite promising preclinical data, clinical trials using mGlu5 receptor antagonists to treat FXS showed no beneficial effects.

Objective

Here, we studied the expression and function of mGlu5 receptors in the striatum of adult BTBR mice, which model idiopathic forms of ASD, and behavioral phenotype.

Methods

Behavioral tests were associated with biochemistry analysis including qPCR and western blot for mRNA and protein expression. In vivo analysis of polyphosphoinositides hydrolysis was performed to study the mGlu5-mediated intracellular signaling in the striatum of adult BTBR mice under basal conditions and after MTEP exposure.

Results

Expression of mGlu5 receptors and mGlu5 receptor-mediated polyphosphoinositides hydrolysis were considerably high in the striatum of BTBR mice, sensitive to MTEP treatment. Changes in the expression of genes encoding for proteins involved in excitatory and inhibitory neurotransmission and synaptic plasticity, including Fmr1, Dlg4, Shank3, Brd4, bdnf-exon IX, Mef2c, and Arc, GriA2, Glun1, Nr2A, and Grm1, Grm2, GriA1, and Gad1 were also found. Behaviorally, BTBR mice showed high repetitive stereotypical behaviors, including self-grooming and deficits in social interactions. Acute or repeated injections with MTEP reversed the stereotyped behavior and the social interaction deficit. Similar effects were observed with the NMDA receptor blockers MK-801 or ketamine.

Conclusion

These findings support a pivotal role of mGlu5 receptor abnormal expression and function in idiopathic ASD adult forms and unveil novel potential targets for therapy.

1. INTRODUCTION

Current FDA-approved medications for autism spectrum disorder (ASD) belong to the class of antipsychotics (e.g., risperidone and aripiprazole) that mainly target irritability and aggressive behaviors in children with ASD. Their clinical use is limited by undesirable adverse effects, such as weight gain and sedation, responsible for treatment discontinuation in many cases. In addition, these drugs fail to treat core symptoms of ASD, which include deficits in communication skills and sociability and repetitive stereotypical behaviors [1-3]. Sociability and stereotyped behaviors are the most common behavioral features studied in ASD-like animal models [4, 5]. Interest in exploiting the role of metabotropic glutamate (mGlu)5 receptors in ASD significantly increased since the evidence that mGlu5 receptor-dependent LTD was amplified in the hippocampus of Fmr1 knockout mice modeling Fragile-X syndrome (FXS) [6]. Indeed, this suggests that mGlu5 receptor inhibitors might be useful as novel therapeutic interventions in ASD [7, 8].

The mGlu5 receptor, a Gα_q-coupled receptor, is involved in learning and memory processes and synaptic plasticity, ion channel modulation, and calcium-mediated excitotoxicity [9,10]. The primary transduction mechanism of mGlu5 receptors is the phospholipase-Cb-mediated hydrolysis of phosphatidylinositol-4,5-bisphosphate into inositol-1,4,5-trisphosphate (InsP₃) and diacylglycerol (DAG). InsP₃ releases Ca²⁺ from intracellular stores, whereas DAG activates protein kinase C. This, in turn, activates downstream intracellular transduction pathways, such as the MAP kinase and phosphatidylinositol-3-kinase (PI3K)-Akt pathways, leading to mTOR activation, new protein synthesis, and increased expression of early inducible genes, including Egr1, Arc, Mef2, and Bdnf. In addition, mGlu5 receptors interact with NMDA receptors post-synaptically to promote Ca²⁺ influx and LTP/LTD. Both ASD subjects and autism-like animal models show alterations in synaptic plasticity and abnormal GABAergic/glutamatergic neurotransmission [11-14]. Moving from studies of mGlu5 receptors in an FXS animal model, characterized by mutation of the Fmr1 gene and loss of the fragile X mental retardation protein (FMRP) [8], the selective mGlu5 receptor antagonist, MTEP, has been successfully tested and found to improve social behaviors in BTBR T+tf/J (BTBR) mice [15,16]. Although a phase 2 clinical trial using the mGlu5 receptor negative allosteric modulator (NAM), basimglurant, did not improve over placebo in FXS subjects [17], inhibition of the mGlu5 receptor remains a suitable target for idiopathic ASD. This mainly results from its relevant role in synaptic plasticity, learning and memory processes, and mGlu5/InsP₃calcium-mediated excitotoxicity in cases of hyperactivation.

ASD pathophysiology is complex, with most cases being idiopathic with an unknown cause, complicating the development of new treatment strategies [18, 19]. Thus, to develop targeted therapies identifying subtypes characterized by specific neurobiological phenotypes is required. BTBR mice are a commercially available inbred strain showing behavioral phenotypes resembling the main clinical features of ASD, including deficits in social behavior, repetitive behaviors or stereotypes, and deficits in communication [4, 20].

Although ASD is considered a neurodevelopmental condition, we focused on the behavioral phenotype of BTBR mice as a model to study idiopathic forms of ASD at adult age. Further, we studied (i) the expression of mGlu5 receptors (both splice variants a and b) in the striatum, a brain region involved in programming social behaviors and motor programming [21,22]; (ii) the mGlu5 receptor-mediated intracellular canonical signaling pathway by measuring in vivo poly phosphoinositide hydrolysis; (iii) the expression of mGlu1, -2, -3, and -7 receptors and ionotropic glutamate NMDA and AMPA receptor subunits; (iv) the expression of early inducible genes relevant for synaptic plasticity, including Mef2c, Arc, Egr1, and Bdnf; and (v) the behavioral effects induced by the mGlu5 receptor allosteric antagonist, MTEP. Our results suggest that the highly abnormal expression of the developmental mGlu5 receptor splice variant a form in the striatum of adult BTBR mice associated with hyperactive intracellular signal transduction and consequent alterations in synaptic plasticity may play a significant role in the pathogenesis of behavioral deficits and constitute a tool to identify adult idiopathic ASD subtypes.

2. MATERIALS AND METHODS

2.1. Animals

All procedures were performed according to National Institutes of Health guidelines for animal research and were approved by the Animal Care Committee of the University of Illinois at Chicago. Male C57BL/6J (referred to as B6) and BTBR mice were obtained from Jackson Laboratories. BTBR T+ tf/J (BTBR) mice represent an inbred mouse strain that shows robust behavioral phenotypes resembling the major diagnostic symptoms of ASD (deficits in social interaction, deficits in communication skills, and repetitive/ restricted behaviors). We chose the B6 mice (commonly used for comparison with BTBR mice), which do not display ASD-like characteristics, as relative control compared to BTBR mice (as reported in McFarlane et al., 2008; Pobbe et al., 2010; Nizan Uddin et al., 2020). After arrival, the mice were left undisturbed for one week before behavioral testing. Experiments were performed with PND 80-90 mice by researchers blind from treatments. The mice were maintained under a 12-h dark/light cycle (light on at 06:00 a.m. and off at 6:00 p.m.), provided food and water ad libitum, and housed in groups of 5. The vivarium temperature was 24°C, and the humidity was 65%. Depending on the study, the number of mice for each experimental group varied from 6 to 8 (see figure legends). They were killed by decapitation. Experimental protocols were approved by the Office and Animal Care and Institutional Biosafety Committee and the Office of the Vice-Chancellor for Research of the University of Illinois at Chicago.

2.2. Drug Treatments

BTBR and control B6 mice were injected with 3-((2-Methyl-4-thiazolyl)ethynyl) pyridine hydrochloride (MTEP, Tocris Bioscience) or vehicle. MTEP (10mg/kg, i.p., in saline) was injected 30 minutes before the behavioral test for an acute experiment, or twice a day for five days for sub-chronic experiments, with the last injection 24h before the test. The MTEP dose was selected based on previous reports [23]. MK-801 and ketamine (Sigma-Aldrich) (0.1 mg/Kg i.p. in saline) were injected 30 minutes before the behavioral test.

2.3. Behavioral Tests

2.3.1. Locomotor Activity

VersaMax software (AccuScan Instruments, Columbus, OH, USA) assisted a computerized Animal Activity Monitoring System to quantify and track locomotor activity in mice [24]. Horizontal activity and vertical activity were determined over 15 minutes. Each activity cage consisted of a 20 cm × 20 cm × 20 cm Perspex box surrounded by horizontal and vertical infrared sensor beams. Horizontal sensors’ beam interruptions were taken as a measure of horizontal activity. The activity of the mice was recorded for 10 min.

2.3.2. Self-grooming

Stereotypic behavior (self-grooming) in BTBR and B6 mice was measured (4) as grooming behavior, including head washing, body grooming, genital and tail grooming, and paw/leg licking, and recorded for 10 minutes [25] (JWatcher software, Dan Blumstein's Lab, University of California Los Angeles & The Animal Behaviour Lab, Macquarie University, Sydney).

2.3.3. Social Interaction

The social approach was measured using a three-chamber apparatus and a rectangular, transparent, three-chambered box [26]. The small openings in the clear Plexiglas walls (10 cm wide x 5cm high) divided the center compartment from the side compartments. Two identical wire cups were placed in the left and right chambers, one for enclosing a stranger (novel) mouse and one for a control object. Social interaction was defined as the ratio of the sniffing time of the cup-enclosed stranger mouse vs. the empty cup.

2.3.4. Elevated Plus Maze Test (EPM)

The mice were placed individually on the central platform, and their behavior was monitored for 5 minutes. It consisted of two open and two closed arms made of Plexiglas. The closed arms had transparent Plexiglas walls at the sides and end. The floor was made of black Plexiglas and elevated to a height of 50 cm above the floor. At the start of each test, mice were placed individually on the central platform, and a computer monitored their behavior for 5 minutes. The number of entries for each arm and the time spent on each arm were recorded and analyzed. The percentage of open arm entries and the time spent in open arms indexed anxiety-like behavior. The number of closed arm entries is represented as the general activity of mice [27].

2.3.5. Novel Object Recognition Test (NOR)

The test was performed as previously described [28]. Briefly, the test consists of a phase of habituation, in which animals are free to explore the apparatus for 10 minutes for 3 consecutive days. Then, during the acquisition trial, the animals were allowed to explore two identical objects (e.g., A1 and A2) for 10 minutes. This was followed by a 24-hour inter-trial interval when the animals were returned to the home cage. During the retention trial, the animals were allowed to explore a familiar object (e.g., A) from the acquisition trial and a novel object (e.g., B). Each object's exploration time (s) in each trial was recorded manually using two stopwatches. The discrimination index [(time spent exploring the novel object - time spent exploring the familiar object)/total exploration time] was then calculated for retention trials.

2.4. Real-time Polymerase Chain Reaction

Quantitative real-time PCR was performed using Applied Biosystems Real-Time PCR System (Stratagene MX3005P: Agilent Technologies, Santa Clara, California) with an SYBR green master mix (Maxima SYBR Green/ROX qPCR Master Mix: ThermoScientific, Waltham, Maryland). Total RNA was isolated using RNeasy Mini Kit (Qiagen, Valencia, California). The primer sequences for the genes analyzed are summarized in Table 1. The mGlu5 receptors constitute two splice variants, a and b, that differ by inserting 32 amino acids in the C-terminus region in variant b. To measure mRNA levels of the two mGluR5 variants, we designed two different sets of primers: one common to both variants (here termed as tot = a + b) (NM_001081414.2, CDS: Fw primer 2706-2725; Rw primer 2885 – 2904) and one specific for the b variant (NM_01143834.1, CDS: Fw primer 3194 – 3211, Rw primer 3346 – 3366 which falls into the insertion portion of the sequence). We used Gapdh as a housekeeping gene for RNA normalization.

Table 1.

Primer sequences for mRNA expression. Primer sequences used for qPCR mRNA measurements in BTBR mice and controls.

Gene	Primer Sequences for qPCR
Grm5(tot)	F: ACGAAGACCAACCGTATTGC	R: AGACTTCTCGGATGCTTGGA
Grm5b	F: AGCCAAACCGGAGAGAAA	R: CCAGTCTCCTGTCTTTGTAC
Grm1	F: CATACGGAAAGGGGAAGTGA	R: AAAAGGCGATGGCTATGATG
Grm2	F: TGGACCGCATCAACCGCGAC	R: CCACGGCTGAGTGAGGCACG
Grm3	F: GGTGGGGCGCTCCAACATCC	R: AAGCATTCACGCGGCTGGCT
Grm7	F: GAACGCAAAGACAGGACCAT	R: CTGACCCAACCCAGAGAAAA
Fmr1	F: AGAGGAGGAGGCTTCAAAGG	R: CTACGCTGTCTGGCTTTTCC
Shank3	F: GAAGTCACCAGAGGACAAGAAG	R: GGTGGCATGCACAACAATAAG
Bdnf-IX	F: GCAGCTGGAGTGGATCAGTAA	R: CATTCACGCTCTCCACAGTCCC
S100b	F: GGTTGCCCTCATTGATGTCT	R: CGTCTCCATCACTTTGTCCA
Mecp2	F: GGTTGTCTCCACTGCTACTTAC	R: GCTAACTTGGGTGCTGATCT
Tamalin	F: CGTTCTCAGGCTGGAAACTC	R: ATGCTTGGGTCTTTCACCAC
Mef2c	F: CACCTACATAACATGCCGCC	R: TGGTGGTACGGTCTCTAGGA
Brd4	F: AGCAACAGCAATGTGAGCAA	R: GCCTAGCTTCTCACCAGGAA
Arc	F: TCCAGAGGGGTATGGAAGTG	R: TATTCAGGCTGGGTCCTGTC
Preso-I	F: CCCACCTTGTCAAAGCAAAT	R: ACTGAGCGCTTCTCTGC
Gria1	F: GCTAGAAAAGAACGGCATCG	R: AGCGTCACTTGTCCTCCACT
Gria2	F: TCGACAATTTGGAGGTAGCC	R: GTTGGGAAGCTTGGTGTGAT
Avpr1a	F: TGCAACCCGTGGATCTACAT	R: TTTGTTGGGCTTCGGTTGTT
Glun1	F: ACTCCCAACGACCACTTCAC	R: GTAGACGCGCATCATCTCAA
Nr2B	F: ACAGTTTCATCCCTGAGCCC	R: CTCCCACTTCCTCTCCTTGT
Nr2A	F: CCTGCTCTACTGCTCCAAGG	R: TGAGCACCGATGGAAACTCT
Grid1	F: GCCACTGGGTCTTTGTGAAT	R: TCCTGTGGATCACAGAGCAG
Gad1	F: GAGGAGAGCGGGCC AAGA	R: GTGCCGCTCCACACGCC
Reln	F: GGGCGGCGGGCCCCGAGG	R: AGAGACCGACGGGCTGCC
Psd-95	F: GGTGACGACCCATCCATCTTTATC	R:CGGACATCCACTTCATTGACAAAC
Homer1a	F: TCTTCAGTCTCCTTTGACACCA	R: CATGATTGCTGAATTGAATGTG
PanHomer1	F: TGGACTGGGATTCTCCTCTG	R: TGTGTCACATCGGGTGTTCT
NeuN	F: GGTCGTGTATCAGGATGGATTT	R: GCGGCATAGACTCTACCATAAC
ActB	F:TGTGATGGTGGGAATGGGTCAGAA	R:TGTGGTGCCAGATCTTCTCCATGT
Gpdh	F: CAATGTGTCCGTGGATCT	R: GTCCTCAGTGTAGCCCAAGATG

2.5. Western Blot Analysis

Extraction was performed with BioBasic Membrane, Nuclear & Cytoplasmic Protein Extraction Kit at 4°C (BioBasic, Amherst, NY). Total protein concentration was determined by the bicinchoninic acid assay method (Pierce BCA Protein Assay Kit: ThermoFisher Scientific, Waltham, MA). Samples were diluted in Laemmli buffer, run on 4%-12% Bis-Tris gel (Thermo Fisher Scientific), blotted onto polyvinylidene difluoride membranes (Millipore-Sigma), incubated at RT for 1 hour with anti-mGlu5 antibody (1:5000; Abcam) [29]. Immunocomplexes were visualized by enhanced chemiluminescence with an Immobilon Western Chemiluminescent HP Substrate Kit (Merck, Burlington, MA) by using the LI-COR (Lincoln, NE) Odyssey System and analyzed with Image Studio 5.2 (iS5.2) software (LI-COR).

2.6. In Vivo Measurement of Polyphosphoinositide (PI) Hydrolysis

Measurements of endogenous inositol monophosphate (InsP, a dephosphorylated metabolite of InsP₃) were performed [30] in the striatum extracted from mice treated i.p. with lithium chloride (LiCl) 100 mg/kg [lithium ions inhibit InsP conversion into free inositol] administered one hour before i.p. injection of MTEP (10 mg/kg) or saline. We used MTEP to identify the component of basal PI hydrolysis mediated by endogenous activation of mGlu5 receptors. Tissue was sonicated in 10 µl/mg Tris-HCl buffer (100mM; pH 7.5) containing 150mM NaCl, 5mM EDTA, 1% Triton X-100, 1% SDS. Homogenates were diluted at 1:50, and InsP levels were assessed by IP-One ELISA kits (Cisbio, Codolet, France).

2.7. Statistical Analyses

The results were analyzed using a two-tailed unpaired t-test for Fig. (1), Fig. (3A-C - 3A-C), and Fig. (4a). Two-way ANOVA was followed by Turkey’s multiple comparison test for Fig. (2A-D) and Fig. (4b). Multiple unpaired t-tests with Welch correction and desired FDR 5.00% have been used to analyze genes in Table 3 (Volcano plot in Fig. 5). The F-value was determined by ANOVA using GraphPad Prism 8.2.1. Values are expressed as the means ± SEM of 6–8 mice for the biochemical analysis and the behavioral tests. Significance was set at *p < 0.05. The specific number of values for each experiment is indicated in the respective results sections.

Fig. (1).

Behavioral features of BTBR mice compared to C57BL/6J (B6) mice used as controls. In A, a statistically significant increase in self-grooming activity observed in BTBR mice compared to B6 controls (two-tailed unpaired Student’s t-test, t₍₁₄₎=5.487; *p<.0001). In B, no changes in locomotor activity between BTBR and B6 mice. In C, social interaction activity of BTBR mice compared to B6 mice (two-tailed unpaired Student’s t-test, t₍₁₄₎=4.466; *p<.0005). In D, elevated plus maze (EPM) significant increase in time spent in open arm and open arms entries in BTBR mice compared to B6 mice (unpaired t test, two-tailed t₍₁₄₎=3.718, *p<.002; t₍₁₄₎=3.910, *p<.001, respectively. N = 8).

Fig. (3).

In (A), mGlu5 receptor mRNA expression levels in the frontal cortex (FC), hippocampus (HP), striatum (STR), and cerebellum (CB) of BTBR mice. Statistically significant decrease in mGlu5 was found in HP (two-tailed unpaired Student’s t-test *p 0.03, t₍₁₀₎ = 2.511). In FC, no changes of mGlu1, 3, 5b were found (two-tailed unpaired Student’s t-test: mGlu1: p 0.24, t₍₁₀₎ = 1.23; mGlu3: p 0.84, t₍₁₀₎ = 0.20; mGlu5b: p 0.17, t₍₁₀₎ = 1.45) whereas a significant decrease in mGlu2 was found (*p 0.03, t₍₁₀₎ = 2.47). In (B) the mGlu1, 2, 3, 5, 5b, and 7 receptor subtypes in the striatum of BTBR mice. Note the increase mGlu5 receptor (total mRNA) without changes in the mGlu5 receptor variant b mRNA levels. Decreased levels of mGlu1 and mGlu2 receptors were found in the striatum of BTBR mice, whereas no changes were found in mGlu3 and mGlu7 mRNA levels. Values are means ± S.E.M. of 6 controls (B6) and 6 BTBR mice groups used (n = 6) for all the experiments performed for this paper. **p<0.01 vs controls (unpaired Student’s t-test). The mGlu5 receptor variant a might reflect the difference in the mRNA expression between the mGlu5 total expression and the variant b. In (C), a representative immunoblots of striatal samples of BTBR mice and controls for mGlu5 receptor protein and Gpdh used as normalizer (B6). Densitometric values are means ± S.E.M. of 6 B6 and 6 BTBR mice. #p<0.05 vs controls (unpaired Student’s t-test). Values are means ± S.E.M. of 6 B6 and 6 BTBR. ^ap<0.005 vs the controls and ^bp<0.05 vs the controls (unpaired Student’s t-test).

Fig. (4).

In (A), measurements of endogenous InsP levels in the striatum of BTBR mice and controls (B6) treated i.p. with lithium ions (100 mg/kg) for basal levels measurements (two-tailed unpaired Student’s t-test, t = 2.56, *p = 0.03; N = 5). In B, endogenous InsP levels in the striatum of BTBR mice pretreated with lithium ions (100 mg/kg) and after one hour with a single MTEP injection (two-way ANOVA followed by a Tukey post-hoc test). Values are means ± S.E.M. of 6-8 animals per experimental group. ^ap<0.01 vs Lithium B6 and ^bp<0.005 vs MTEP BTBR. In (C), a schematic representation of the canonical pathway activated by mGlu5 receptor leading to increased calcium mobilization through inositol-triphosphate. Lithium blocks free inositol production.

Fig. (2).

In (A) the effect of a single injection of mGlu5 receptors negative allosteric modulator MTEP (10mg/kg, ip) on self-grooming activity in BTBR mice and relative controls administered 30min before the behavioral test (two-way ANOVA Tukey’s multiple comparisons: F _(3,16) = 31.05, *p = .0001 (BTBR saline vs B6 saline). Similar effect was observed in the social interaction test (two-way ANOVA Tukey’s multiple comparisons: F_{(3, 17)} = 9.32; *p = .0007. In (B) sub-chronic (5 days) injections of MTEP (10 mg/kg, i.p.) reversed the grooming behavior of BTBR mice and social interaction deficits [two-way ANOVA Tukey’s multiple comparisons: F_(3,18) = 19.50, *p = .0001; or F_{(3, 16)} =11.63; *p = .0003, respectively (BTBR saline vs B6 saline)]. Values are means ± S.E.M. of 6-8 BTBR mice and controls (B6) (two-way ANOVA analysis *p<0.05 vs B6 Sal) (# p<0.05 vs BTBR Sal). In C and D, the effect on stereotyped behavior (self-grooming) induced by a single administration of the NMDA receptor channel blockers MK-801 (C) and ketamine (D) (0.1mg/kg, ip). A marked decrease in self-grooming is shown with both compounds in BTBR mice (two-way ANOVA Tukey’s multiple comparisons: F_(3,15) = 25.28, p = .0001 or F _(3,15) = 32.67, p = .0001, respectively).

Table 3.

Measurements of mRNA levels of genes related to ASD, synaptic plasticity and early inducible genes related to stress-response and activity-dependent synaptic plasticity. Changes in Fmr1, Psd-95 and Shank-3 were found in the striatum of BTBR mice associated with altered expression of mRNA levels of AMPAR subtype Gria-1, Gria2 and NMDA-subtype Glun1 and NR2A. Values are means ± S.E.M. of 6-8 animals per experimental group. *p<0.05 vs B6 (Unpaired t-test student-two tailed). Reduced expression of Gad1 and increased expression of Brd-4 and Bdnf-ix in the striatum of BTBR mice, associated with altered expression of gene involved in stress-response and activity-dependent synaptic plasticity including Arc and Mef2c. Values are means ± S.E.M. of 6-8 animals per experimental group. *p<0.05 vs B6 (Unpaired t-test student-two tailed).

Gene	B6	Btbr	t Ratio, p Values
Fmr1	1.0 ± 0.10	1.81 ± 0.20(*)	t = 3.62; p = 0.004
Psd-95	0.99 ± 0.09	1.81 ± 0.20(*)	t = 3.73; p = 0.04
Shank3	0.99 ± 0.08	1.52 ± 0.11(*)	t = 3.89; p = 0.001
Gria1	1.0 ± 0.10	0.64 ± 0.07(*)	t = 2.94; p = 0.01
Gria2	1.0 ± 0.21	1.11 ± 0.22 (*)	t = 2.92; p = 0.01
Glun1	1.0 ± 0.17	1.19 ± 0.12 (*)	t = 3.89; p = 0.001
Nr2A	1.0 ± 0.19	4.24 ± 0.95(*)	t = 3.34; p = 0.01
Nr2B	1.0 ± 0.34	0.85 ± 0.19	t = 0.358; p = 0.73
Mecp2	1.0 ± 0.09	0.64 ± 0.2	t = 1.64; p = 0.13
Bdnf-IX	1.0 ± 0.21	3.48 ± 0.37(*)	t = 5.821; p = 0.0001
Gad1	1.0 ± 0.12	0.3 ± 0.03(*)	t = 5.65; p = 0.0005
Mef2c	1.0 ± 0.28	5.3 ± 0.78(*)	t = 5.182; p = 0.0006
Brd4	0.87 ± 0.09	1.28 ± 0.13(*)	t =2.59; p = 0.02
Arc	0.89 ± 0.12	1.65 ± 0.27(*)	t =2.57; p = 0.02
Egr1	1.0 ± 0.10	0.67 ± 0.27	t =1.14; p = 0.28
Preso-I	1.0 ± 0.09	0.81 ± 0.08	t = 1.57; p = 0.13
Tamalin	0.74 ± 0.17	0.93 ± 0.18	t = 0.86; p = 0.40
Homer1a	1.0 ± 0.20	1.65 ± 0.41	t = 1.42; p = 0.18
Grid1	0.99 ± 0.05	1.12 ± 0.14	t = 0.87; p = 0.40

Fig. (5).

In (A), a volcano plot showing the main differences in the expression levels of the genes summarized in Table 3 and analyzed using the multiple unpaired t tests. In the Table 3 the relative p values and t ratios. Data were analyzed with multiple unpaired t test with Welch correction, using FDR 5% and two-stage step up (Benjamini, Krieger, and Yekutieli) method [73]. In (B), A schematic modeling of mGlu5 receptors increased expression and activity leading to altered AMPA receptor-mediated synaptic plasticity in the striatum of BTBR mice. In (C), hierarchical gene networks. The schematic representation reflects our hypothesis that the triad represented by (i) a decreased GABA synteshsis, (ii) a decreased expression of mGlu2 receptors, and (iii) an increased expression of mGlu5 receptors leads to a GABA-glutamate imbalance. The imbalance might be corrected by mGlu2 receptor PAMs o mGlu5 receptor NAMs. Both categories of drugs are currently under clinical development. The schematic model also explores the complex network that links mGlu5 receptors and genes involved in postsynaptic density, dendritic spine maturation and formation, regulation in synaptic plasticity and neurotransmission mediated by AMPA receptors. Based on our results, a combination of presynaptic decreased mGlu2 receptors and postsynaptic increased mGlu5 receptors expression associated with decreased levels of Gad1 suggest a potential imbalance of glutamatergic transmission with overactive intracellular signaling activated by excessive glutamate release, which then activate both mGlu5 receptors and NMDA receptors at post-synaptic level. These effects might be explained by the reversal behavioral effects induced by both MK801 and ketamine in stereotyped behaviors in BTBR mice. We chose to measure genes related to ASD such as Fmr1, Shank3, genes related to AMPA receptor-mediated glutamate transmission such as Gria1, genes related to the mGlu5 receptor signaling and post-synaptic density and synaptic plasticity including Psd-95, Arc, Mef2c, Egr-1 and Bdnf. The model indicates a predicted, physical interactions or co-expression of several genes related to mGlu5 receptor signaling associated with Arc and ionotropic glutamate receptor subunits (see Table 3 for the significance of each gene measured). Genetic interactions, co-expression and physical interactions network of altered genes measured in the striatum of BTBR mice (Table 3). Interactions with histone deacethylation enzymes such as Hdac7 and Arc is also noted which support future studies to explore epigenetic mechanisms potentially involved in mGlu5 receptor variant a and Arc overexpression in BTBR mice. The gene network model highlights interactions of several genes involved in ASD-like phenotype including Grm5 with Arc, Mef2c, Fmr1 and supports the hypothesis of abnormal excitatory/inhibitory (E/I) neurotransmission and alterations in genes involved in synaptic plasticity and elimination mediated by AMPA receptors linked to mGlu5 receptors intracellular signaling via Mef2c/Arc. Further studies are needed to better understand the complexity of the mGlu5 receptor signaling and its implications in synapses formation and elimination during early stages of development and adult forms of ASD-like phenotypes.

3. RESULTS

3.1. Behavioral Study of BTBR Mice

BTBR mice showed significantly higher repetitive self-grooming activity than controls (B6 mice) (Fig. 1A) (two-tailed unpaired Student’s t-test, t₍₁₄₎=5.487; *p<.0001). BTBR showed no difference in the locomotor activity test (two-tailed unpaired Student’s t-test horizontal activity: t₍₁₄₎=.006; p = 0.99; vertical activity: t₍₁₄₎=2.139; p=0.05) (Fig. 1B). BTBR mice showed a marked decrease in sociability as measured as time spent with the intruder in the 3-chamber social interaction test compared to controls (B6) (two-tailed unpaired Student’s t-test, t₍₁₄₎=4.466; *p<.0005) (Fig. 1C). The elevated plus maze (EPM) test showed no difference in basic movements between BTBR and B6 mice (two-tailed unpaired Student’s t-test p = 0.16; t = 1.56 (data not shown)), whereas it showed a significant difference in total time spent in the open arms (two-tailed unpaired Student’s t-test, t₍₁₄₎=3.718; *p<.002) and open entries (two-tailed unpaired Student’s t-test, t₍₁₄₎=3.910; *p<.0016) between BTBR mice and controls (Fig. 1D). In the Novel Object Recognition test (NOR), BTBR mice showed no exploration activity because of the high time spent on self-grooming activity during the retention trial. Therefore, we did not proceed further with this test. This behavioral effect is in line with previous observation [31], in which no differences were observed in the time spent sniffing the novel object compared to familiar object in BTBR mice during both the acquisition and retention trial. Based on these basic behavioral characteristics, we tested the effects of the mGlu5 receptors allosteric antagonist MTEP on stereotyped behavior (self-grooming) and sociability (social interaction test).

3.2. Behavioral Effects of the mGlu5 Receptors Allosteric Antagonist MTEP

A single administration of the mGlu5 receptor allosteric antagonist, MTEP (10mg/kg, i.p., 30 min before test) markedly reversed the BTBR mice's high level of self-grooming activity (two-way ANOVA Tukey’s multiple comparisons test F_(3,16)=31.05; *p=.0001, N=6-8) and the social deficits in BTBR mice (F_(3,17) = 9.32, *p=.0007, N=6-8) compared to controls (Fig. 2A). Further studies are needed to examine enduring effects. A similar effect was reached with a 5-day MTEP administration with a marked decrease in self-grooming (two-way ANOVA F_(3,18)=19.5; *p=.0001, N=6-8) and a significant reversal effect on the deficit in social interaction in BTBR mice (two-way ANOVA F_3,16=11.63; *p=.0003, N=6-8) (Fig. 2B), suggesting the absence of tolerance effects induced by repeated drug administration under our experimental conditions. We tested two NMDA receptor blockers because of the cooperativity between mGlu5 and NMDA receptors. A single dose of the NMDA receptor blockers MK-801 and ketamine-induced a marked reduction in the self-grooming activity of BTBR mice (two-way ANOVA, F_3,15=25.28, *p<.0001 and F_3,15=32.67; *p<.0001 (respectively); N=6) (Fig. 2C, D) unveiling a potential additional molecular target to address subgroups of the population characterized by stereotyped behaviors mediated by increased mGlu5 receptor activity.

3.3. Alterations in the Expression of mGlu5 Receptors in the Brain Regions of BTBR Mice

In Fig. 3A, we measured the mGlu5 receptors mRNA levels in frontal cortex (FC), hippocampus (HP), striatum (STR) and cerebellum (CB) of BTBR mice compared to controls (B6) mice. In Table 2, we then measured mRNA expression of mGlu receptors 1, 2, 3, 5, and 7 in the striatum of BTBR mice and controls (B6 mice). Interestingly, we first noticed a marked increase in the mGlu5 receptor mRNA levels in the striatum of BTBR mice at PND 85. In particular, in the striatum of BTBR mice at PND 85, we observed a marked increase in total mGlu5 mRNA levels [variant tot (a + b) two-tailed *p = 0.01, (t₍₁₀₎ = 3.50] compared to B6 mice, whereas levels of the b variant were unchanged (p = 0.15, (t₍₁₀₎ = 1.54) (Table 2). We can speculate that the increase in mGlu5 receptor mRNA in the striatum of BTBR mice is primarily due to the overexpression of variant a (variant b:tot = 18% in BTBR mice vs 100% in B6). BTBR mice also showed decreased mGlu1 (t₍₁₀₎ = 3.659; p = .004) and mGlu2 (t₍₉₎ = 2.411; p = .04) receptor mRNA levels in the striatum (Fig. 3B, Table 2). No changes were found in mGlu3 (t₍₁₀₎ = 0.3316; p = .74) and mGlu7 (t₍₉₎ = 2.411; p = .25) receptor mRNA levels in the striatum of BTBR mice. A slight reduction of mGlu5 receptor mRNA was found in Hip (t = 2.511; *p = 0.03) (Fig. 3A). In FC, no changes of mGlu1, 3, 5b were found (two-tailed unpaired Student’s t-test: mGlu1: p = 0.24, t₍₁₀₎ = 1.23; mGlu3: p= 0.84, t₍₁₀₎ = 0.20; mGlu5b: p = 0.17, t₍₁₀₎ = 1.45) whereas a significant decrease in mGlu2 was found (*p = 0.03, t₍₁₀₎ = 2.47) (data not shown in graph). In HP, both mGlu2 and mGlu5 receptors mRNA were decreased (two-tailed unpaired Student’s t-test: mGlu2: *p = 0.02, t₍₁₀₎ = 2.77; mGlu5_(tot): *p= 0.03, t₍₁₀₎ = 2.51). No changes in mGlu5b variant (p= 0.43, t₍₁₀₎ = 0.82). No changes in mGlu receptor subtypes were also found in cerebellum (data not shown).

Table 2.

Measurements of mRNA levels of mGlu receptors subtypes in the striatum of BTBR mice. Changes in mGlu1, 2, 3, 5 (a+b), mGlu5b, and mGlu7 mRNA levels (gene symbols: Grm). Values are means ± S.E.M. of 6-8 animals per experimental group. *p<0.05 vs B6 (two-tailed Unpaired t-test student).

Values are Expressed as Fold of Controls Mean ± SEM. (*) p < 0.05
Gene	B6 mice	BTBR mice	t values, p values
GRM1	1.0 ± 0.10	0.45 ± 0.10(*)	t 10 = 3.659; p = 0.04
GRM2	1.0 ± 0.06	0.72 ± 0.08(*)	t 10 = 2.268; p = 0.02
GRM3	1.0 ± 0.12	1.00 ± 0.15(*)	t 10 = 0.331; p = 0.74
GRM5	1.0 ± 0.30	3.9 ± 0.8(*)	t 10 = 3.228; p = 0.009
GRM5b	1.0 ± 0.15	0.64 ± 0.17	t 10 = 1.548; p = 0.15
GRM7	1.0 ± 0.13	0.8 ± 0.06	t 10 = 1.222; p = 0.24

We decided to focus our further analysis on the mGlu5 receptor protein and in vivo activity on the striatum for its role in stereotyped behaviors and dyskinesia [32, 33].

Increased mGlu5 receptor protein expression was found in the striatum of BTBR mice (t₍₉₎ = 2.602; p = 0.02) (Fig. 3C). The antibody used for western blot experiments recognizes both a and b splice variants of mGlu5 receptors (no variant-specific antibodies are available to our knowledge). The densitometric analysis combined the two bands corresponding to receptor monomers and dimers.

3.4. In Vivo Measurement of mGlu5 Receptor-mediated Polyphosphoinositide (PI) Hydrolysis in the Striatum of BTBR Mice

We carried out an in vivo analysis of the canonical intracellular pathway associated with mGlu5 receptor activation as previously described [30]. Using the Cisbio IP-One ELISA kit, which detects endogenous InsP levels (a degradation product of InsP3), we measured InsP levels in the striatum of BTBR mice and controls after systemic treatment with lithium, which prevents the conversion of InsP into free inositol [32]. Interestingly, under basal conditions, InsP levels were significantly higher in BTBR mice compared to controls (two-tailed unpaired Student’s t-test, t₍₈₎=2.56; *p<.003, N = 5) (Fig. 4A). We then repeated the analysis in a larger group of animals treated with the mGlu5 receptors allosteric antagonist, MTEP. The difference between BTBR and control mice was abolished after treatment with a single injection MTEP (10 mg/kg, i.p.), which statistically reduced striatal InsP levels (two-way ANOVA F_{(3, 15)} = 7.893; p = .002, N =8) (Fig. 4B). This suggests that the PI response to endogenous activation of mGu5 receptors was amplified in the striatum of BTBR mice.

3.5. Changes in the Expression of ASD-related Genes in the Striatum of BTBR Mice

In Fig. (5a), we reported a volcano plot showing the differences in the expression of specific genes analyzed in the striatum of BTBR mice. Table 3 reports mRNA levels of the genes measured, some related to syndromic forms of ASD, including Fmr1, Psd-95, and Shank-3. In the striatum of BTBR mice, Fmr1, Psd95, Shank3, Gria1, Gria2, Glun1, Nr2A were altered (see Table 3 for p values and t ratios). We also measured the transcript of genes related to neurotransmission, synaptic plasticity, and regulation of gene expression, such as glutamic acid decarboxylase (Gad)1 (encoding the 67 kDa isoform of GAD), Mecp2, Bdnf exon-IX, Mef2c, Brd4, Arc, Egr1, Preso-I, Tamalin, Homer1a, Grid1 (Table 3).

The transcripts of Gad1, Mef2c, Arc, Brd4, and Bdnf-IX were also changed in the striatum of BRBT mice compared to controls, whereas the transcripts of Egr1, Mecp2, Preso-I, Tamalin, Homer1a, Grid1, and Avpr1A were unchanged (Table 3). We then measured specific mGlu5 receptor signaling intracellular target genes such as Arc, Mef2c, and BDNF after a single injection of MTEP. The two-way ANOVA showed no differences in the striatum of BTBR mice (data not shown).

3.6. Changes in the Expression of Ionotropic Glutamate NMDA and AMPA Receptor Subunits in the Striatum of BTBR Mice

We extended the analysis to Gria1, Gria2, Grin1, Grin2A, and Grin2B genes encoding the GluA1 and GluA2 AMPA receptor subunits and GluN1, GluN2A, and GluN2B NMDA receptor subunits, respectively. We found a significant decrease in GluA1 (t₍₈₎=2.680, p=.02) and an increase in GluN2A (t₍₆₎=3.332, p=.01) in the striatum of BTBR mice as compared to control mice. No changes were found in the expression of Gria2, Grin1, and GriN2B genes (Table 3).

4. DISCUSSION

A large body of evidence links mGlu5 receptors to monogenic ASD, such as FXS (see the introduction and references therein), tuberous sclerosis [34-36], Rett’s syndrome [37], and Angelman syndrome [10]. Behavioral and electrophysiological studies in BTBR mice raised the possibility that mGlu5 receptors are linked to non-syndromic idiopathic forms of ASD and are candidate targets for therapeutic intervention [38, 39]. We showed that pharmacological blockade of mGlu5 receptors with the allosteric antagonist MTEP reduced repetitive behaviors and rescued social deficits in BTBR mice. Stereotypic (self-grooming) activity is a well-recognized behavioral feature of BTBR mice and is associated with strength in glutamate release [31, 37]. In addition, we found that MTEP maintained its behavioral effects after 5 days of treatment, indicating no development of tolerance. This gave us the impetus to study whether changes in the expression and/or activity of mGlu5 receptors were associated with the ASD-like phenotype of BTBR mice and whether similar changes could be observed in the expression of additional receptors or intracellular proteins that are involved in mGlu5-dependent mechanisms of synaptic plasticity. We focused on the striatum, a region that encodes habit memory and, therefore, plays a key role in the programming of repetitive, stereotyped behaviors.

Importantly, we found that mGlu5 receptor mRNA and protein levels were significantly higher in the striatum of BTBR mice than in controls and that at least the increase in the transcript was receptor subtype- and region-specific. We could also demonstrate that receptor-coupled signal transduction (i.e., the component of constitutive PI hydrolysis mediated by endogenous activation of mGlu5 receptors) was amplified in the striatum of BTBR mice. This is the first report of hyperactivity in striatal mGlu5 receptors in any mouse model of ASD. The techniques used in our study (immunoblotting, PCR, and in vivo measurements of PI hydrolysis) do not allow the identification of the cellular source of changes in expression and signaling of mGlu5 receptors. In addition, mGlu5 receptors are also found in intracellular compartments, including the nuclear membrane, where their precise function remains undetermined. In addition, mGlu5 receptors are widely expressed in both neurons and glial cells as well as at the intracellular level in the substantia nigra (SN), which may contribute to the intracellular signaling activation [40, 41]. This represents an obvious limitation in the interpretation of our findings.

Transcript analysis of mGlu5 receptors revealed that the splice variant a was specifically up-regulated in the striatum of BTBR mice, suggesting a persistent abnormal expression of this variant in adult age BTBR mice, as reported in Romano et al., 1996. mGlu5b differs from mGlu5a in the presence of 32 amino acids inserted in the C-terminus domain and the 50 residues downstream of the 7th TM domain [42, 43]. The two splice variants display identical function and pharmacological profiles in heterologous expression systems. However, a remarkable difference is that the a variant is highly expressed in the early postnatal development when a high PI response to mGlu5 receptor agonists is observed [42, 43], whereas the b variant predominates in the adult brain [43-46]. In addition, the two variants differentially regulate neuronal development, with mGlu5a hindering neuronal maturation and mGlu5b fostering neurite formation and elongation [46]. Our data indicate that the developmental variant of mGlu5 receptors (i.e., mGlu5a) persists in the striatum of adult BTBR mice, explaining the >2-fold increase in the mGlu5a receptor-mediated PI hydrolysis found in these mice without changes in mGlu5b. Also, the additional sequence expressed in mGluR5 variant b, which contains additional putative phosphorylation sites, may play a specific role in signaling transduction mechanisms that need further investigation [43].

mGlu5 receptors have been extensively studied in rodent and primate models of parkinsonism and are candidate drug targets for treating Parkinson’s disease and L-DOPA-induced dyskinesias [47, 48]. In striatal projection neurons of the indirect pathway, mGlu5 receptors colocalize with dopamine D2 and adenosine A_2A receptors, and their activation restrains the inhibitory control exerted by D2 receptors on the pathway [49,46]. mGlu5 receptors are also found in projection neurons of the direct pathway, where they contribute to the development of a maladaptive form of activity-dependent synaptic plasticity underlying stereotyped movements [50]. mGlu5 and D1 receptors may form heteromeric complexes in the striatum and synergize in the activation of PI hydrolysis and ERK pathway leading to the induction of dyskinesias [51]. Our data, combined with the evidence that D1 receptors are unchanged in the striatum of BTBR mice [52], suggest that heteromeric complexes in these mice are enriched by mGlu5a receptors. This might result in abnormal motor programming leading to stereotyped behavior.

We also found a significant increase in the transcript encoding the GluN2A subunit of NMDA receptors in the striatum of BTBR mice. Synaptic GluN2A-containing NMDA receptors mediate activity-dependent synaptic plasticity processes as opposed to extra-synaptic NMDA receptors, which support neuronal survival [53]. NMDA and mGlu5 receptors are linked by a chain of interacting proteins, such as PSD-95, Skank, and Homer, and functionally interact, potentiating each other [54-61]. The transcripts of at least PSD-95 and Shank-3 were also higher in the striatum of BTBR mice than in the controls. mGlu5 and NMDA receptors are partners in mechanisms of synaptic plasticity, and, in the striatum, both receptors are essential for the induction of long-term potentiation (LTP) at synapses between cortico-striatal glutamatergic neurons and projection neurons of the direct pathway [62, 63]. In BTBR mice, a high expression and activity of mGlu5 receptors might reinforce the functional crosstalk with NMDA receptors resulting in hyperactivity of the direct pathway and stereotyped behaviors. In support of this hypothesis, the NMDA receptor blockers, MK-801 or ketamine, were able to reverse the high self-grooming activity of BTBR mice.

A battery of genes involved in mechanisms of synaptic plasticities, such as Fmr1, Egr1, Arc, Mef-2, Brd-4, and Bdnf, were overexpressed in the striatum of BTBR mice. Some of the genes are interconnected (Fig. 5). For example, Arc is induced by BDNF, Egr-1, and Mef-2c [63,64]. In CA1 hippocampal neurons, activation of mGlu5 receptors promotes synapse elimination and long-term depression by stimulating the translation of Mef-2-induced Arc mRNA [65-70].

Arc is an activity-dependent protein representing an activation marker of striatal medium spiny projection neurons [66]. Interestingly, Arc is up-regulated in neurons of the direct pathway in a rat model of L-DOPA-induced dyskinesias [71]. We hypothesize that overactivation of the mGlu5/Arc axis in striatal projection neurons of the direct pathway causes abnormal motor programming that ultimately results in stereotyped behaviors in BTBR mice. Kumar et al. [70] have shown that Arc can be induced by activating intracellular mGlu5 receptors in striatal neurons. Our experimental approach does not allow us to discard this hypothesis.

We were intrigued by the findings that the transcripts of Grm1, Grm2, Gria1, and Gad1 were reduced in the striatum of BTBR mice. mGlu1 receptors must also be activated to induce LTP in the striatum [64-66], but their role in striatal motor programming is less prominent in mGlu5 receptors. Perhaps Grm1 expression is reduced in the striatum of BTBR mice as an attempt to compensate for the overexpression of mGlu5 receptors. The reduced expression of Grm2 fits with the hypothesis of increased direct pathway activity because mGlu2 receptors negatively modulate glutamate release [71]. Many cells in the striatum and striatal projection regions use GABA as a neurotransmitter and contain glutamic acid decarboxylase (GAD) as a rate-limiting enzyme for GABA synthesis [72]. Importantly, in this study, we showed that GAD1 mRNA expression is markedly lower in the striatum of BTBR mice than in controls, which may play a role in the underlying imbalance between GABAergic and glutamatergic neurotransmission in the striatum.

Reductions in the transcripts encoding for GluA1 and Gad1 are difficult to interpret considering the widespread expression of these two proteins in different neuronal populations within the striatum. Further characterization of these changes at the cellular level is needed to better understand the relationship between an imbalance between excitatory and inhibitory neurotransmission and behavioral abnormalities in BTBR mice.

Finally, an important question is whether the increased expression and activity of striatal mGlu5 receptors underlies the deficit in social interaction found in BTBR mice, considering that treatment with MTEP was able to reverse this deficit and that at least the transcript of mGlu5 receptors was unchanged in the cerebral cortex and reduced in the hippocampus. One possible explanation is that the balance between self-and hetero-directed behavior in BTBR mice is shifted towards self-directed behavior because of the hyperactivity of striatal mGlu5 receptors, resulting in stereotyped self-rewarding behavior. This hypothesis warrants further investigation.

CONCLUSION

Our data demonstrate for the first time that mGlu5 receptors splice variant a is overexpressed and hyperactive in the striatum of an adult mouse model of non-syndromic ASD forms. This alteration might lie at the core of the increased stereotyped and self-directed behavior, which is a hallmark of ASD. The increase in mGlu5 receptor activity explains the robust behavioral effect of mGlu5 receptor NAMs in BTBR mice and suggests that these drugs may be valuable in treating high mGlu5-expressing non-syndromic ASD.

HIGHLIGHTS

• BTBR mice show behavioral abnormalities such as stereotypic behavior (self-grooming) or social interaction deficits characteristic of autism spectrum disorder.

• The striatum of BTBR mice showed higher expression and function of mGlu5 receptors (splice variant a) associated with in vivo hyperactivity as measured by higher levels of inositol monophosphate formation.

• The alterations of mGlu5 receptors were associated with changes in the RNA expression of ASD-related genes, including the early inducible genes Mef2c and Arc and genes related to synaptic plasticities such as Gria1, Gria2, Glun1, and Nr2A.

• Targeting the mGlu5 receptor using the allosteric antagonist MTEP reverses the pathological behavioral phenotype, suggesting that mGlu5 receptors represent a suitable molecular target for adult non-syndromic forms of ASD.

AUTHORS’ CONTRIBUTIONS

F.M. was involved in experimental planning, performing experiments, analyzing and interpreting data, writing the manuscript, and graphic designing. V.L. performed experiments, and E.D. handled preliminary experiments. Moreover, F.N. was also involved in experimental planning, scientific supervision, supporting F.M., and manuscript writing. D.R.G. contributed to experimental planning and editing of the manuscript, and A.G. was involved in providing scientific support, interpreting data, and experimental planning. The authors have nothing to disclose.

ETHICS APPROVAL AND CONSENT TO PARTICIPATE

This study was approved by the Office and animal Care and Institutional Biosafety Committee and the Office of the Vice-Chancellor for Research of the University of Illinois at Chicago.

HUMAN AND ANIMAL RIGHTS

All procedures were performed according to National Institutes of Health guidelines for animal research and were approved by the Animal Care Committee of the University of Illinois at Chicago.

CONSENT FOR PUBLICATION

Not applicable.

AVAILABILITY OF DATA AND MATERIALS

Not applicable.

FUNDING

None.

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.