WO2016097260A1

WO2016097260A1 - Methods for predicting or diagnosing autism spectrum disorders (asd)

Info

Publication number: WO2016097260A1
Application number: PCT/EP2015/080424
Authority: WO
Inventors: Laurence Colleaux; Claire Rougeulle; Lam Son NGUYEN; Mélanie MAKHLOUF; Yann HUMEAU
Original assignee: INSERM (Institut National de la Santé et de la Recherche Médicale); Assistance Publique-Hôpitaux De Paris (Aphp); Université Paris Descartes; Fondation Imagine; Centre National De La Recherche Scientifique (Cnrs)
Priority date: 2014-12-19
Filing date: 2015-12-18
Publication date: 2016-06-23

Abstract

The present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising a step of determining the expression levels of at least two miRs selected from the group consisting of miR-146a, miR- 221, miR-654-5p and miR-656 in a biological sample obtained from said subject.

Description

METHODS FOR PREDICTING OR DIAGNOSING

AUTISM SPECTRUM DISORDERS (ASD)

FIELD OF THE INVENTION:

The present invention relates generally to the field of neurology. More specifically, the present invention relates to methods and kits for predicting or diagnosing autism spectrum disorder (ASD) in a subject. BACKGROUND OF THE INVENTION:

Autism spectrum disorders (ASD) represent a group of childhood neurodevelopmental disorders characterized by deficits in verbal communication, social interaction, and restricted and repetitive patterns of interests and behaviors. These conditions are often considered as "connectopathies" or "synaptopathies", whereby neuronal circuits are mis-wired or dysfunctional. Much progress has been made in unravelling the genetics of ASD with large copy number variation¹, de novo mutations in a subset of genes² as well as inherited mutations³ being associated with ASD. Yet, the current consensus agrees that majority of ASD cases are caused by complex interaction of multiple genetic and environmental risk factors. Orchestrated by environmental factors, epigenetic modifications, such as DNA methylation, covalent modification of histones, or the activation or silencing of genes by miRNAs, may drive genetically predisposed individuals to develop autism and represent the crossroads between the environment and genetics in ASD. Accordingly, gene expression profiling of monozygotic twins discordant in diagnosis of autism has identified differentially expressed, neuro logically relevant genes⁴. Similarly, abnormal DNA methylation patterns have been detected at several candidate genes in ASD patients⁵. microRNAs (miRNAs) play key roles in neuronal development, synapse formation and fine-tuning of genes underlying synaptic plasticity and memory formation⁶. Several studies have identified aberrant miRNA expression in a wide range of neurological diseases⁷' ⁸, but a comprehensive understanding of miRNA networks deregulated in ASD has not been achieved. The expression of miRNAs in lymphoblastoid cell cultures of ASD patients has already been explored^9"12, but the relevance of this cell type is questionable in the context of a neurodevelopmental disorder. One study has identified deregulated miRNAs in the cerebellum of ASD patients , but it raises the issue of the investigated brain regions and the impact of cause of death and/or post-death handling of tissue on gene expression.

Olfactory mucosal stem cells (OMSCs) exhibit characteristics of ecto-mesenchymal stem cell and maintain neurogenic ability¹⁴; they could be considered as precursor of neuronal and glia-like cells. Thus, OMSC has been used successfully to study the pathology of schizophrenia¹⁵, Rett Syndrome¹⁶ and Parkinson's disease¹⁷ in which they displayed disease- specific alterations in gene expression and cellular functions in order to identify genes and pathways relevant for neurodevelopmental disorders.

However, OMSCs biopsied from living ASD patients have never been studied in order to identify a set of miRNAs which might serve as a sensitive and specific biomarker for ASD.

SUMMARY OF THE INVENTION:

In a first aspect, the present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising a step of determining the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in a biological sample obtained from said subject.

In a second aspect, the present invention relates to a kit suitable for performing the methods of the invention wherein said kit comprises means for measuring the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR- 654-5p and miR-656 in a biological sample obtained from the subject.

DETAILED DESCRIPTION OF THE INVENTION:

The present invention is based on the discovery that a particular combination of miRs

(i.e. at least miR-146a and miR-221, with preferably miR-654-5p and/or miR-656) is useful for predicting and diagnosing ASD in a subject with an optimal sensitivity and specificity.

Autism spectrum disorder (ASD) is an early childhood neurological condition whose etiology represents a complex interaction between inherited, de novo mutations and the environment. Using olfactory mucosal stem cells (OMSC) biopsied from living patients, as a source of primary neuronal-like cell type, giving a hint into the early neurogenesis process, the inventors identified a signature of microRNA (miR-146a up-regulated 2-fold; miR-221, miR-654-5p and miR-656 down-regulated ~1.3-fold) commonly deregulated in ASD. This signature is conserved in primary skin fibroblasts and allows discriminating between ASD and intellectual disability samples. Putative target genes of the differentially expressed miRNAs were enriched for pathways previously associated to ASD and altered levels of neuronal transcripts targeted by miR-146a, miR-221 and miR-656 were observed in patients' cells. Thus, the present results have strong diagnostic implications and emphasize the role of epigenetic deregulation in the etiology of ASD.

Definitions:

Throughout the specification, several terms are employed and are defined in the following paragraphs .

The term "miRNAs" (also called "miR") has its general meaning in the art and refers to microRNA molecules that are generally 21 to 22 nucleotides in length, even though lengths of 19 and up to 23 nucleotides have been reported. miRNAs are each processed from a longer precursor RNA molecule ("precursor miRNA"). Precursor miRNAs are transcribed from non- protein-encoding genes. The precursor miRNAs have two regions of complementarity that enables them to form a stem-loop- or fold-back-like structure, which is cleaved in animals by a ribonuclease Ill-like nuclease enzyme called Dicer. The processed miRNA is typically a portion of the stem. The processed miRNA (also referred to as "mature miRNA") become part of a large complex to down-regulate a particular target gene.

All the miRNAs pertaining to the invention are known per se and sequences of them are publicly available from the data base http://microrna.sanger.ac.uk/sequences/.

The human miRNAs of the invention are listed in Table A:

miRNA (Sequence ID) miRBase Accession number

(miR- 146a) (SEQ ID NO: 1) MI0000477

gagaacugaauuccauggguu

(miR-221) (SEQ ID NO: 2) MI0000298

accuggcauacaauguagauuu

(miR-654-5p) (SEQ ID NO: 3) MI0003676

uggugggccgcagaacaugugc

(miR-656) (SEQ ID NO: 4) MI0003678 agguugccugugagguguuca

Table A: list of the human miRNAs according to the invention

As used herein, the term "determining" encompasses detecting or quantifying. As used herein, "detecting" means determining if a miR (e.g. miR-146a or miR-221) is present or not in a biological sample and "quantifying" means determining the amount of a miR (e.g. miR- 146a or miR-221) in a biological sample.

As used herein, the term "autism spectrum disorder" (ASD) refers to complex neuro developmental disabilities characterized by stereotypic behaviors and deficits in communication and social interaction. The term "spectrum" refers to the wide range of symptoms, skills, and levels of impairment, or disability, that patients with ASD can have. ASD is generally diagnosed according to guidelines listed in the Diagnostic and Statistical Manual of Mental Disorders, Fith Edition (DSM-V). The manual currently defines five disorders, sometimes called pervasive developmental disorders (PDDs), as ASD, including Autistic disorder (classic autism), Asperger's disorder (Asperger syndrome), Pervasive developmental disorder not otherwise specified (PDD-NOS), Rett's disorder (Rett syndrome), and Childhood disintegrative disorder (CDD). Some patients are mildly impaired by their symptoms, but others are severely disabled. ASD encompasses a set of complex disorders with poorly defined etiologies, and no targeted cure.

Diagnostic methods:

In a first aspect, the present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising a step of determining the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in a biological sample obtained from said subject. In another aspect, the present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising a step of determining the expression levels of at least two miRs in an olfactory mucosa biopsy obtained from said subject. In one embodiment, the method further comprises a step of comparing the expression levels of said at least two miRs with their respective predetermined reference values; and determining whether the subject has or is at risk of having ASD when the determined expression levels of said at least two miRs is different (higher or lower) to their respective predetermined reference values.

In one embodiment, the present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising the following steps of:

i. providing a biological sample obtained from said subject;

ii. determining the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in said biological sample;

iii. comparing said at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 with their respective predetermined reference values; and

iv. determining whether the subject has or is at risk of having ASD, wherein an expression level of miR-146a higher than the predetermined reference value, an expression level of miR-221 lower than the predetermined reference value, an expression level of miR-654-5p lower than the predetermined reference value and an expression level of miR-656 lower than the predetermined reference value is indicative that the subject has, or is at risk of having ASD.

In one embodiment, the expression levels of at least miR-146a and miR-221 are determined. In another embodiment, the expression levels of miR-146a, miR-221 and miR- 654-5p are determined. In another embodiment, the expression levels of miR-146a, miR-221 and miR-656 are determined. In still another embodiment, the expression levels of miR-146a, miR-221 miR-654-5p and miR-656 are determined.

Accordingly, in one embodiment, the present invention relates to an in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising the following steps of:

i. providing a biological sample obtained from said subject; ii. determining the expression levels of miR-146a and miR-221 in said biological sample; iii. comparing said expression levels of miR-146a and miR-221 with their respective predetermined reference values; and

iv. determining whether the subject has or is at risk of having ASD , wherein an expression level of miR-146a higher than the predetermined reference value and an expression level of miR-221 lower than the predetermined reference value is indicative that the subject has, or is at risk of having ASD and/or intellectual disability.

In a particular embodiment, the method further comprises a step of determining the expression level of miR-654-5p.

Thus, the method comprises the following steps of:

i. providing a biological sample obtained from said subject;

ii. determining the expression levels of miR-146a, miR-221 and miR-654-5p in said biological sample;

iii. comparing said expression levels of miR-146a, miR-221 and miR-654-5p with their respective predetermined reference values; and

iv. determining whether the subject has or is at risk of having ASD wherein an expression level of miR-146a higher than the predetermined reference value, an expression level of miR-221 lower than the predetermined reference value and an expression level of miR-654-5p lower than the predetermined reference value is indicative that the subject has, or is at risk of having ASD and/or intellectual disability.

In a particular embodiment, the method further comprises a step of determining the expression level of miR-656.

Thus, the method comprises the following steps of:

i. providing a biological sample obtained from said subject;

ii. determining the expression levels of miR-146a, miR-221 and miR-656 in said biological sample;

iii. comparing said expression levels of miR-146a, miR-221 and miR-656 with their respective predetermined reference values; and iv. determining whether the subject has or is at risk of having ASD, wherein an expression level of miR-146a higher than the predetermined reference value, an expression level of miR-221 lower than the predetermined reference value and an expression level of miR-656 lower than the predetermined reference value is indicative that the subject has, or is at risk of having ASD.

In a preferred embodiment, the method further comprises a step of determining the expression levels of miR-654-5p and miR-656.

Thus, the method comprises the following steps of:

i. providing a biological sample obtained from said subject;

ii. determining the expression levels of miR-146a, miR-221, miR-654-5p and miR-656 in said biological sample;

iii. comparing said expression levels of miR-146a, miR-221, miR-654-5p and miR-656 with their respective predetermined reference values; and

In one embodiment, the autism spectrum disorder (ASD) is selected from the group consisting of Autism, Autistic Disorder, Asperger Syndrome, Childhood Disintegrative Disorder, Pervasive Developmental Disorder - Not Otherwise Specified (PDD-NOS), Fragile X Syndrome, and Rett Syndrome.

Any subject sample suspected of containing miRs may be tested according to the methods of the invention. By way of non-limiting examples, the biological sample may be tissue (e.g., an olfactory mucosa biopsy or a skin biopsy), blood or a fraction thereof (e.g., plasma or serum). An olfactory mucosa biopsy is preferably collected and use to isolate and test early passage olfatory mucosa stem cells. Conventional methods and reagents for isolating RNA from a sample comprise High Pure miRNA Isolation Kit (Roche), Trizol (Invitrogen), Guanidinium thiocyanate -phenol- chloroform extraction, PureLink™ miRNA isolation kit (Invitrogen), PureLink Micro-to- Midi Total RNA Purification System (invitrogen), RNeasy kit (Qiagen), miRNeasy kit (Qiagen), Oligotex kit (Qiagen), phenol extraction, phenol-chloroform extraction, TCA/acetone precipitation, ethanol precipitation, Column purification, Silica gel membrane purification, Pure Yield™ RNA Midiprep (Promega), PolyATtract System 1000 (Promega), Maxwell® 16 System (Promega), SV Total RNA Isolation (Promega), geneMAG-RNA / DNA kit (Chemicell), TRI Reagent® (Ambion), RNAqueous Kit (Ambion), ToTALLY RNA™ Kit (Ambion), Poly(A)Purist™ Kit (Ambion) and any other methods, commercially available or not, known to the skilled person.

The expression level of one or more miRNA such as miR-146a and miR-221 in the sample may be determined by any suitable method. Any reliable method for determining the level or amount of miRNA in a sample may be used. Generally, miRNA can be detected and quantified from a sample (including fractions thereof), such as samples of isolated RNA by various methods known for mRNA, including, for example, amplification-based methods (e.g., Polymerase Chain Reaction (PCR), Real-Time Polymerase Chain Reaction (RT-PCR), Quantitative Polymerase Chain Reaction (qPCR), rolling circle amplification, etc.), hybridization-based methods (e.g. , hybridization arrays (e.g. , microarrays), NanoString analysis, Northern Blot analysis, branched DNA (bDNA) signal amplification, in situ hybridization, etc.), and sequencing-based methods (e.g. , next- generation sequencing methods, for example, using the Illumina or IonTorrent platforms). Other exemplary techniques include ribonuclease protection assay (RPA) and mass spectroscopy.

In some embodiments, RNA is converted to DNA (cDNA) prior to analysis. cDNA can be generated by reverse transcription of isolated miRNA using conventional techniques. miRNA reverse transcription kits are known and commercially available. Examples of suitable kits include, but are not limited to the mirVana TaqMan® miRNA transcription kit (Ambion, Austin, TX), and the TaqMan® miRNA transcription kit (Applied Biosystems, Foster City, CA). Universal primers, or specific primers, including miRNA- specific stem- loop primers, are known and commercially available, for example, from Applied Biosystems. In some embodiments, miRNA is amplified prior to measurement. In some embodiments, the expression level of miRNA is measured during the amplification process. In some embodiments, the expression level of miRNA is not amplified prior to measurement. Some exemplary methods suitable for determining the expression level of miRNA in a sample are described in greater hereinafter. These methods are provided by way of illustration only, and it will be apparent to a skilled person that other suitable methods may likewise be used.

Many amplification-based methods exist for detecting the expression level of miRNA nucleic acid sequences, including, but not limited to, PCR, RT-PCR, qPCR, and rolling circle amplification. Other amplification-based techniques include, for example, ligase chain reaction, multiplex ligatable probe amplification, in vitro transcription (IVT), strand displacement amplification, transcription-mediated amplification, RNA (Eberwine) amplification, and other methods that are known to persons skilled in the art. A typical PCR reaction includes multiple steps, or cycles, that selectively amplify target nucleic acid species: a denaturing step, in which a target nucleic acid is denatured; an annealing step, in which a set of PCR primers (i.e., forward and reverse primers) anneal to complementary DNA strands, and an elongation step, in which a thermostable DNA polymerase elongates the primers. By repeating these steps multiple times, a DNA fragment is amplified to produce an amplicon, corresponding to the target sequence. Typical PCR reactions include 20 or more cycles of denaturation, annealing, and elongation. In many cases, the annealing and elongation steps can be performed concurrently, in which case the cycle contains only two steps. A reverse transcription reaction (which produces a cDNA sequence having complementarity to a miRNA) may be performed prior to PCR amplification. Reverse transcription reactions include the use of, e.g., a RNA -based DNA polymerase (reverse transcriptase) and a primer. Kits for quantitative real time PCR of miRNA are known, and are commercially available. Examples of suitable kits include, but are not limited to, the TaqMan® miRNA Assay (Applied Biosystems) and the mirVana™ qRT-PCR miRNA detection kit (Ambion). The miRNA can be ligated to a single stranded oligonucleotide containing universal primer sequences, a polyadenylated sequence, or adaptor sequence prior to reverse transcriptase and amplified using a primer complementary to the universal primer sequence, poly(T) primer, or primer comprising a sequence that is complementary to the adaptor sequence. In some embodiments, custom qRT-PCR assays can be developed for determination of miRNA levels. Custom qRT-PCR assays to measure miRNAs in a sample can be developed using, for example, methods that involve an extended reverse transcription primer and locked nucleic acid modified PCR. Custom miRNA assays can be tested by running the assay on a dilution series of chemically synthesized miRNA corresponding to the target sequence. This permits determination of the limit of detection and linear range of quantitation of each assay. Furthermore, when used as a standard curve, these data permit an estimate of the absolute abundance of miRNAs measured in the samples. Amplification curves may optionally be checked to verify that Ct values are assessed in the linear range of each amplification plot. Typically, the linear range spans several orders of magnitude. For each candidate miRNA assayed, a chemically synthesized version of the miRNA can be obtained and analyzed in a dilution series to determine the limit of sensitivity of the assay, and the linear range of quantitation. Relative expression levels may be determined, for example, according to the 2(^{~ΔΔ C(}-^T^) Method, as described by Livak et ah, Analysis of relative gene expression data using real-time quantitative PCR and the 2^{("ΔΔ C(T)}) Method. Methods (2001) Dec;25(4):402-8.

In some embodiments, two or more miRNAs are amplified in a single reaction volume. For example, multiplex q-PCR, such as qRT-PCR, enables simultaneous amplification and quantification of at least two miRNAs of interest in one reaction volume by using more than one pair of primers and/or more than one probe. The primer pairs comprise at least one amplification primer that specifically binds each miRNA, and the probes are labeled such that they are distinguishable from one another, thus allowing simultaneous quantification of multiple miRNAs. Rolling circle amplification is a DNA-polymerase driven reaction that can replicate circularized oligonucleotide probes with either linear or geometric kinetics under isothermal conditions (see, for example, Lizardi et al, Nat. Gen. (1998) 19(3):225-232; Gusev et al, Am. J. Pathol. (2001) 159(l):63-69; Nallur et al, Nucleic Acids Res. (2001) 29(23):E118). In the presence of two primers, one hybridizing to the (+) strand of DNA, and the other hybridizing to the (-) strand, a complex pattern of strand displacement results in the generation of over 10⁹ copies of each DNA molecule in 90 minutes or less. Tandemly linked copies of a closed circle DNA molecule may be formed by using a single primer. The process can also be performed using a matrix- associated DNA. The template used for rolling circle amplification may be reverse transcribed. This method can be used as a highly sensitive indicator of miRNA sequence and expression level at very low miRNA concentrations (see, for example, Cheng et al, Angew Chem. Int. Ed. Engl. (2009) 48(18):3268-72; Neubacher et al, Chembiochem. (2009) 10(8): 1289-91). miRNA may be detected using hybridization-based methods, including but not limited to hybridization arrays (e.g., microarrays), NanoString analysis, Northern Blot analysis, branched DNA (bDNA) signal amplification, and in situ hybridization. Microarrays can be used to measure the expression levels of large numbers of miRNAs simultaneously. Microarrays can be fabricated using a variety of technologies, including printing with fine-pointed pins onto glass slides, photolithography using pre- made masks, photolithography using dynamic micromirror devices, inkjet printing, or electrochemistry on microelectrode arrays. Also useful are micro fluidic TaqMan Low-Density Arrays, which are based on an array of micro fluidic qRT-PCR reactions, as well as related micro fluidic qRT-PCR based methods. In one example of microarray detection, various oligonucleotides (e.g., 200+ 5'- amino- modified-C6 oligos) corresponding to human sense miRNA sequences are spotted on three- dimensional CodeLink slides (GE Health/ Amersham Biosciences) at a final concentration of about 20 μMand processed according to manufacturer's recommendations. First strand cDNA synthesized from 20 μg TRIzol-purified total RNA is labeled with biotinylated ddUTP using the Enzo BioArray end labeling kit (Enzo Life Sciences Inc.). Hybridization, staining, and washing can be performed according to a modified Affymetrix Antisense genome array protocol. Axon B-4000 scanner and Gene-Pix Pro 4.0 software or other suitable software can be used to scan images. Non-positive spots after background subtraction, and outliers detected by the ESD procedure, are removed. The resulting signal intensity values are normalized to per-chip median values and then used to obtain geometric means and standard errors for each miRNA. Each miRNA signal can be transformed to log base 2, and a one-sample t test can be conducted. Independent hybridizations for each sample can be performed on chips with each miRNA spotted multiple times to increase the robustness of the data.

Microarrays can be used for the expression profiling of miRNAs. For example, RNA can be extracted from the sample and, optionally, the miRNAs are size- selected from total RNA. Oligonucleotide linkers can be attached to the 5' and 3' ends of the miRNAs and the resulting ligation products are used as templates for an RT-PCR reaction. The sense strand PCR primer can have a fluorophore attached to its 5' end, thereby labeling the sense strand of the PCR product. The PCR product is denatured and then hybridized to the microarray. A PCR product, referred to as the target nucleic acid that is complementary to the corresponding miRNA capture probe sequence on the array will hybridize, via base pairing, to the spot at which the, capture probes are affixed. The spot will then fluoresce when excited using a microarray laser scanner. The fiuorescence intensity of each spot is then evaluated in terms of the number of copies of a particular miRNA, using a number of positive and negative controls and array data normalization methods, which will result in assessment of the level of expression of a particular miRNA. Total RNA containing the miRNA extracted from the sample can also be used directly without size-selection of the miRNAs. For example, the RNA can be 3' end labeled using T4 RNA ligase and a fiuorophore-labeled short RNA linker. Fluorophore- labeled miRNAs complementary to the corresponding miRNA capture probe sequences on the array hybridize, via base pairing, to the spot at which the capture probes are affixed. The fiuorescence intensity of each spot is then evaluated in terms of the number of copies of a particular miRNA, using a number of positive and negative controls and array data normalization methods, which will result in assessment of the level of expression of a particular miRNA. Several types of microarrays can be employed including, but not limited to, spotted oligonucleotide microarrays, pre-fabricated oligonucleotide microarrays or spotted long oligonucleotide arrays. miRNAs can also be detected without amplification using the nCounter Analysis System (NanoString Technologies, Seattle, WA). This technology employs two nucleic acid- based probes that hybridize in solution (e.g., a reporter probe and a capture probe). After hybridization, excess probes are removed, and probe/target complexes are analyzed in accordance with the manufacturer's protocol. nCounter miRNA assay kits are available from NanoString Technologies, which are capable of distinguishing between highly similar miRNAs with great specificity. The basis of the nCounter® Analysis system is the unique code assigned to each nucleic acid target to be assayed (International Patent Application Publication No. WO 08/124847, U.S. Patent No. 8,415,102 and Geiss et al. Nature Biotechnology. 2008. 26(3): 317-325; the contents of which are each incorporated herein by reference in their entireties). The code is composed of an ordered series of colored fluorescent spots which create a unique barcode for each target to be assayed. A pair of probes is designed for each DNA or RNA target, a biotinylated capture probe and a reporter probe carrying the fluorescent barcode. This system is also referred to, herein, as the nanoreporter code system. Specific reporter and capture probes are synthesized for each target. The reporter probe can comprise at a least a first label attachment region to which are attached one or more label monomers that emit light constituting a first signal; at least a second label attachment region, which is non-over-lapping with the first label attachment region, to which are attached one or more label monomers that emit light constituting a second signal; and a first target- specific sequence. Preferably, each sequence specific reporter probe comprises a target specific sequence capable of hybridizing to no more than one gene and optionally comprises at least three, or at least four label attachment regions, said attachment regions comprising one or more label monomers that emit light, constituting at least a third signal, or at least a fourth signal, respectively. The capture probe can comprise a second target-specific sequence; and a first affinity tag. In some embodiments, the capture probe can also comprise one or more label attachment regions. Preferably, the first target- specific sequence of the reporter probe and the second target- specific sequence of the capture probe hybridize to different regions of the same gene to be detected. Reporter and capture probes are all pooled into a single hybridization mixture, the "probe library". The relative abundance of each target is measured in a single multiplexed hybridization reaction. The method comprises contacting the tumor sample with a probe library, such that the presence of the target in the sample creates a probe pair - target complex. The complex is then purified. More specifically, the sample is combined with the probe library, and hybridization occurs in solution. After hybridization, the tripartite hybridized complexes (probe pairs and target) are purified in a two-step procedure using magnetic beads linked to oligonucleotides complementary to universal sequences present on the capture and reporter probes. This dual purification process allows the hybridization reaction to be driven to completion with a large excess of target-specific probes, as they are ultimately removed, and, thus, do not interfere with binding and imaging of the sample. All post hybridization steps are handled robotically on a custom liquid-handling robot (Prep Station, NanoString Technologies). Purified reactions are typically deposited by the Prep Station into individual flow cells of a sample cartridge, bound to a streptavidin-coated surface via the capture probe, electrophoresed to elongate the reporter probes, and immobilized. After processing, the sample cartridge is transferred to a fully automated imaging and data collection device (Digital Analyzer, NanoString Technologies). The expression level of a target is measured by imaging each sample and counting the number of times the code for that target is detected. For each sample, typically 600 fields-of-view (FOV) are imaged (1376 X 1024 pixels) representing approximately 10 mm2 of the binding surface. Typical imaging density is 100- 1200 counted reporters per field of view depending on the degree of multiplexing, the amount of sample input, and overall target abundance. Data is output in simple spreadsheet format listing the number of counts per target, per sample. This system can be used along with nanoreporters. Additional disclosure regarding nanoreporters can be found in International Publication No. WO 07/076129 and WO07/076132, and US Patent Publication No. 2010/0015607 and 2010/0261026, the contents of which are incorporated herein in their entireties. Further, the term nucleic acid probes and nanoreporters can include the rationally designed (e.g. synthetic sequences) described in International Publication No. WO 2010/019826 and US Patent Publication No.2010/0047924, incorporated herein by reference in its entirety.

Mass spectroscopy can be used to quantify miRNA using RNase mapping. Isolated RNAs can be enzymatically digested with RNA endonucleases (RNases) having high specificity (e.g., RNase Tl, which cleaves at the 3'-side of all unmodified guanosine residues) prior to their analysis by MS or tandem MS (MS/MS) approaches. The first approach developed utilized the on-line chromatographic separation of endonuclease digests by reversed phase HPLC coupled directly to ESTMS. The presence of posttranscriptional modifications can be revealed by mass shifts from those expected based upon the RNA sequence. Ions of anomalous mass/charge values can then be isolated for tandem MS sequencing to locate the sequence placement of the posttranscriptionally modified nucleoside. Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) has also been used as an analytical approach for obtaining information about posttranscriptionally modified nucleosides. MALDI-based approaches can be differentiated from ESTbased approaches by the separation step. In MALDI-MS, the mass spectrometer is used to separate the miRNA. To analyze a limited quantity of intact miRNAs, a system of capillary LC coupled with nanoESI- MS can be employed, by using a linear ion trap-orbitrap hybrid mass spectrometer (LTQ Orbitrap XL, Thermo Fisher Scientific) or a tandem-quadrupole time-of-flight mass spectrometer (QSTAR® XL, Applied Biosystems) equipped with a custom-made nanospray ion source, a Nanovolume Valve (Valco Instruments), and a splitless nano HPLC system (DiNa, KYA Technologies). Analyte/TEAA is loaded onto a nano-LC trap column, desalted, and then concentrated. Intact miRNAs are eluted from the trap column and directly injected into a CI 8 capillary column, and chromatographed by RP-HPLC using a gradient of solvents of increasing polarity. The chromatographic eluent is sprayed from a sprayer tip attached to the capillary column, using an ionization voltage that allows ions to be scanned in the negative polarity mode.

Additional methods for miRNA detection and measurement include, for example, strand invasion assay (Third Wave Technologies, Inc.), surface plasmon resonance (SPR), cDNA, MTDNA (metallic DNA; Advance Technologies, Saskatoon, SK), and single- molecule methods such as the one developed by US Genomics. Multiple miRNAs can be detected in a microarray format using a novel approach that combines a surface enzyme reaction with nanoparticle- amplified SPR imaging (SPRI). The surface reaction of poly(A) polymerase creates poly(A) tails on miRNAs hybridized onto locked nucleic acid (LNA) microarrays. DNA-modified nanoparticles are then adsorbed onto the poly(A) tails and detected with SPRI. This ultrasensitive nanoparticle-amplified SPRI methodology can be used for miRNA profiling at attomole levels. miRNAs can also be detected using branched DNA (bDNA) signal amplification (see, for example, Urdea, Nature Biotechnology (1994), 12:926- 928). miRNA assays based on bDNA signal amplification are commercially available. One such assay is the QuantiGene® 2.0 miRNA Assay (Affymetrix, Santa Clara, CA). Northern Blot and in situ hybridization may also be used to detect miRNAs. Suitable methods for performing Northern Blot and in situ hybridization are known in the art. Advanced sequencing methods can likewise be used as available. For example, miRNAs can be detected using Illumina ® Next Generation Sequencing (e.g. Sequencing-By- Synthesis or TruSeq methods, using, for example, the HiSeq, HiScan, GenomeAnalyzer, or MiSeq systems (Illumina, Inc., San Diego, CA)). miRNAs can also be detected using Ion Torrent Sequencing (Ion Torrent Systems, Inc., Gulliford, CT), or other suitable methods of semiconductor sequencing. At step iii), the comparison step may be obtained by comparing the expression level in the biological sample from the subject with expression level in a biological sample from a healthy subject (or group of healthy subjects). A differential expression is indicative of that the subject has, or is at risk of having ASD. As used herein, a "higher expression level" consists of a an expression level value that is statistically (i.e. significantly) higher than the predetermined reference value (that may also be termed the "control" expression value or "control reference" values) that has been previously determined in the same biological sample from a healthy subject, e.g. an olfactory mucosa biopsy or a skin biopsy from a healthy subject.

As used herein, a "lower expression level" consists of a an expression level value that is statistically (i.e. significantly) lower than the predetermined reference value (that may also be termed the "control" expression value or "control reference" values) that has been previously determined in the same biological sample from a healthy subject, e.g. an olfactory mucosa biopsy or a skin biopsy from a healthy subject.

In a particular embodiment, the predetermined reference value is a threshold value or a cut-off value that can be determined experimentally, empirically, or theoretically. Typically, the predetermined reference value is a threshold value or a cut-off value. Typically, a "threshold value" or "cut-off value" can be determined experimentally, empirically, or theoretically. A threshold value can also be arbitrarily selected based upon the existing experimental and/or clinical conditions, as would be recognized by a person of ordinary skilled in the art. For example, retrospective measurement of the expression level of the selected miR A in properly banked historical subject samples may be used in establishing the predetermined reference value. The threshold value has to be determined in order to obtain the optimal sensitivity and specificity according to the function of the test and the benefit/risk balance (clinical consequences of false positive and false negative). Typically, the optimal sensitivity and specificity (and so the threshold value) can be determined using a Receiver Operating Characteristic (ROC) curve based on experimental data. For example, after determining the expression level of the selected miRNA in a group of reference, one can use algorithmic analysis for the statistic treatment of the expression levels determined in samples to be tested, and thus obtain a classification standard having significance for sample classification. The full name of ROC curve is receiver operator characteristic curve, which is also known as receiver operation characteristic curve. It is mainly used for clinical biochemical diagnostic tests. ROC curve is a comprehensive indicator that reflects the continuous variables of true positive rate (sensitivity) and false positive rate (1 -specificity). It reveals the relationship between sensitivity and specificity with the image composition method. A series of different cut-off values (thresholds or critical values, boundary values between normal and abnormal results of diagnostic test) are set as continuous variables to calculate a series of sensitivity and specificity values. Then sensitivity is used as the vertical coordinate and specificity is used as the horizontal coordinate to draw a curve. The higher the area under the curve (AUC), the higher the accuracy of diagnosis. On the ROC curve, the point closest to the far upper left of the coordinate diagram is a critical point having both high sensitivity and high specificity values. The AUC value of the ROC curve is between 1.0 and 0.5. When AUC>0.5, the diagnostic result gets better and better as AUC approaches 1. When AUC is between 0.5 and 0.7, the accuracy is low. When AUC is between 0.7 and 0.9, the accuracy is moderate. When AUC is higher than 0.9, the accuracy is high. This algorithmic method is preferably done with a computer. Existing software or systems in the art may be used for the drawing of the ROC curve, such as: MedCalc 9.2.0.1 medical statistical software, SPSS 9.0, ROCPO WER. S AS , DESIGNROC.FOR, MULTIREADER POWER.SAS, CREATE-ROC.SAS, GB STAT VIO.O (Dynamic Microsystems, Inc. Silver Spring, Md., USA), etc.

An accurate diagnosis should also be based on observation of the individual's communication, social interaction, and his or her activities and interests. In some embodiments, the method of the invention further comprises the step of assessing behavioral performance of the subject.

Behavioral performance can be measured and evaluated using various parameters and methods. For example, behavioral test can be conducted to determine the presence and/or extent of restricted repetitive behavior and/or stereotyped behavior patterns of the subject under test. In some embodiments, the Autism Behavior Checklist (ABC), Autism diagnostic Interview-Revised (ADI-R), childhood autism Rating Scale (CARS), and/or Pre -Linguistic Autism Diagnostic Observation Schedule (PL-ADOS) is used for the behavioral test. The behavioral test can include, but is not limited to, detecting the presence and/or extent of 1) preoccupation with one or more stereotyped and restricted patterns of interest that is abnormal in either intensity or focus, 2) inflexible adherence to specific, nonfunctional routines or rituals, c) stereotyped and repetitive motor mannerisms (such as hand flapping, finger flapping etc.), and/or d) persistent preoccupation with parts of objects. ...

Non- limiting examples of behavior that can be included in a behavioral test include: a) sensory behaviors, including poor use of visual discrimination when learning, seems not to hear, so that a hearing loss is suspected, sometimes shows no "startle response" to loud noise", sometimes painful stimuli such as bruises, cuts, and injections evoke no reaction, often will not blink when bright light is directed toward eyes, covers ears at many sounds, squints, frowns, or covers eyes when in the presence of natural light, frequently has no visual reaction to a "new" person, stares into space for long periods of time;

b) relating behaviors: frequently does not attend to social/environmental stimuli, has no social smile, does not reach out when reached for, non-responsive to other people's facial expressions/feelings, actively avoids eye contact, resists being touched or held, is flaccid when held in arms, is stiff and hard to held, does not imitate other children at play, has not developed any friendships, often frightened or very anxious, "looks through" people;

c) body and object use behaviors: whirls self for long periods of time, does not use toys appropriately, insists on keeping certain objects with him/her, rocks self for long periods of time, does a lot of lunging and darting, flaps hands, walks on toes, hurts self by banging head, biting hand, etc. . . . , twirls, spins, and bangs objects a lot, will feel, smell, and/or taste objects in the environment, gets involved in complicated "rituals" such as lining things up, etc. . . . , is very destructive; and

d) language behaviors: does not follow simple commands given once, has pronoun reversal, speech is atonal, does not respond to own name when called out among two others, seldom says "yes" or "I", does not follow simple commands involving prepositions, gets desired objects by gesturing, repeats phrases over and over, cannot point to more than five named objects, uses 0-5 spontaneous words per day to communicate wants and needs, repeats sounds or words over and over, echoes questions or statements made by others, uses at least 15 but less than 30 spontaneous phrases daily to communicate, learns a simple task but "forgets" quickly, strong reactions to changes in routine/environment, has "special abilities" in one area of development, which seems to rule out mental retardation, severe temper tantrums and/or frequent minor tantrums, hurts others by biting, hitting, kicking, etc. . . . , does not wait for needs to be met, difficulties with toileting, does not dress self without frequent help, frequently unaware of surroundings, and may be oblivious to dangerous situations, prefers to manipulate and be occupied with inanimate things, and a developmental delay was identified at or before 30 months of age.

One of ordinary skill in the art would appreciate that the attending physician would know how to identify a subject suffering from ASD.

Kits of the invention:

The invention also relates to a kit suitable for performing the methods of the invention, wherein said kit comprises means for measuring the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in a biological sample obtained from the subject. In one embodiment of the invention, said kit comprises means for measuring the expression levels of at least miR-146a and miR-221 in a biological sample obtained from the subject. Preferably, said kit further comprises means for measuring the expression level of miR-654-5p and/or miR-656 in a biological sample obtained from the subject.

The kits may include probes, primers, macroarrays or microarrays as described above. For example, the kit may comprise a set of miRNA probes as above defined, usually made of DNA, and that may be pre-labelled. Alternatively, probes may be unlabelled and the ingredients for labelling may be included in the kit in separate containers. The kit may further comprise hybridization reagents or other suitably packaged reagents and materials needed for the particular hybridization protocol, including solid-phase matrices, if applicable, and standards.

Alternatively the kits of the invention may comprise amplification primers (e.g. stem- loop primers) that may be pre-labelled or may contain an affinity purification or attachment moiety. The kit may further comprise amplification reagents and also other suitably packaged reagents and materials needed for the particular amplification protocol.

In some embodiments, labels, dyes, or labeled probes and/or primers are used to detect amplified or unamplified miRNAs. The skilled artisan will recognize which detection methods are appropriate based on the sensitivity of the detection method and the abundance of the target. Depending on the sensitivity of the detection method and the abundance of the target, amplification may or may not be required prior to detection. One skilled in the art will recognize the detection methods where miRNA amplification is preferred. A probe or primer may include standard (A, T or U, G and C) bases, or modified bases. Modified bases include, but are not limited to, the AEGIS bases (from Eragen Biosciences), which have been described, e.g., in U.S. Pat. Nos. 5,432,272, 5,965,364, and 6,001,983. In certain aspects, bases are joined by a natural phosphodiester bond or a different chemical linkage. Different chemical linkages include, but are not limited to, a peptide bond or a Locked Nucleic Acid (LNA) linkage, which is described, e.g., in U.S. Pat. No. 7,060,809. In a further aspect, oligonucleotide probes or primers present in an amplification reaction are suitable for monitoring the amount of amplification product produced as a function of time. In certain aspects, probes having different single stranded versus double stranded character are used to detect the nucleic acid. Probes include, but are not limited to, the 5'-exonuclease assay {e.g., TaqMan™) probes (see U.S. Pat. No.5, 538, 848), stem-loop molecular beacons (see, e.g., U.S. Pat. Nos. 6,103,476 and 5,925,517), stemless or linear beacons (see, e.g., WO 9921881, U.S. Pat. Nos. 6,485,901 and 6,649,349), peptide nucleic acid (PNA) Molecular Beacons (see, e.g., U.S. Pat. Nos. 6,355,421 and 6,593,091), linear PNA beacons (see, e.g. U.S. Pat. No. 6,329,144), non- FRET probes (see, e.g., U.S. Pat. No. 6,150,097), Sunrise™/AmplifluorB™ probes (see, e.g., U.S. Pat. No. 6,548,250), stem-loop and duplex Scorpion™ probes (see, e.g., U.S. Pat. No. 6,589,743), bulge loop probes (see, e.g., U.S. Pat. No. 6,590,091), pseudo knot probes (see, e.g., U.S. Pat. No. 6,548,250), cyclicons (see, e.g., U.S. Pat. No. 6,383,752), MGB Eclipse™ probe (Epoch Biosciences), hairpin probes (see, e.g. , U.S. Pat. No. 6,596,490), PNA light-up probes, antiprimer quench probes (Li et al, Clin. Chem.53:624-633 (2006)), self-assembled nanoparticle probes, and ferrocene-modified probes described, for example, in U.S. Pat. No. 6,485,901. In some embodiments, one or more of the primers in an amplification reaction can include a label. In yet further embodiments, different probes or primers comprise detectable labels that are distinguishable from one another. In some embodiments a nucleic acid, such as the probe or primer, may be labeled with two or more distinguishable labels. In some aspects, a label is attached to one or more probes and has one or more of the following properties: (i) provides a detectable signal; (ii) interacts with a second label to modify the detectable signal provided by the second label, e.g., FRET (Fluorescent Resonance Energy Transfer); (iii) stabilizes hybridization, e.g., duplex formation; and (iv) provides a member of a binding complex or affinity set, e.g. , affinity, antibody- antigen, ionic complexes, hapten-ligand (e.g. , biotin-avidin). In still other aspects, use of labels can be accomplished using any one of a large number of known techniques employing known labels, linkages, linking groups, reagents, reaction conditions, and analysis and purification methods. Labels include, but are not limited to: light-emitting, light- scattering, and light- absorbing compounds which generate or quench a detectable fluorescent, chemiluminescent, or bioluminescent signal (see, e.g. , Kricka, L., Nonisotopic DNA Probe Techniques, Academic Press, San Diego (1992) and Garman A., Non- Radioactive Labeling, Academic Press (1997).). A dual labeled fluorescent probe that includes a reporter fluorophore and a quencher fluorophore is used in some embodiments. It will be appreciated that pairs of fluorophores are chosen that have distinct emission spectra so that they can be easily distinguished. In some embodiments, labels are hybridization- stabilizing moieties which serve to enhance, stabilize, or influence hybridization of duplexes, e.g. , intercalators and intercalating dyes (including, but not limited to, ethidium bromide and SYBR-Green), minor-groove binders, and cross-linking functional groups (see, e.g. , Blackburn et al., eds. "DNA and R A Structure" in Nucleic Acids in Chemistry and Biology (1996)).

The invention will be further illustrated by the following figures and examples. However, these examples and figures should not be interpreted in any way as limiting the scope of the present invention.

FIGURES:

Figure 1: Identification of a conserved miRNA signature in ASD. (a) Overview of miRNA profiling. miRNA profiling was performed over 2 rounds, a discovery round using 667 miRNAs and one round of validation using 24 miRNAs which were identified as significantly deregulated in the discovery rounds (corrected P < 0.05) or whose assay failed in the first round. P values were calculated by Wilcoxon rank sum test and corrected using Bonferroni. All measurements were done in technical triplicates and biological duplicates, (b) 4 miRNAs: miR-146a, miR-221, miR-654-5p and miR-656 remained to be significantly deregulated (P < 0.05, controls - dark grey bar, patients - light grey bar) in ASD OMSC after the validation round, representing the miRNA deregulation signature of ASD. * P < 0.05 by Wilcoxon rank sum test, (c) Tissue and disease specificity of miRNA signature. Expression of miR-146a, miR-221, miR-654-5p and miR-656 were assessed in primary skin fibroblasts of ASD patients (n=5, light grey bar), ID patients (n=12, white bar) and controls (n=4, dark grey bar). Results were obtained from Taqman assays performed on Fluidigm array and analyzed using the same method and against the same reference miRNAs as described. * P < 0.05, *** P<0,001 by Student's paired 2-tailed t-test. (d) Predictive power of miRNA combination to distinguish between the groups in OMSCs. RT-qPCR results of 4 miRNAs of interest from the 3^rd round of miRNA profiling in OMSCs were analyzed by Principal Component Analysis. Principal component 1 (PCI) can explain 86% (proportion of variance) the difference between the two groups; this is mostly contributed by the expression difference observed in miR-146a (rotation value) and to a lesser extent miR-221. Other PCs or miRNA combinations do not offer the same predictive power as PCI . (e) Predictive power of miRNA combination to distinguish between ASD, ID and controls. Partition analysis indicates the relative expression of miR-221 and miR-654-5p combined could be used as markers to separate ASD from ID and the control groups. RT-qPCR results obtained from the Fluidigm chip was used for this analysis. Figure 2: miRNAs directly regulates neuronal relevant genes, (a) Gene ontology enrichment of predicted miR A targets by Ingenuity Pathway Analysis (IP A). About 1000 genes were predicted to be regulated by each miRNA by at least 3 different prediction programs and were included in the analyses. Only the top 5 enriched pathways are shown. P values were calculated by Fisher's Exact Test, red line represents correction threshold by Bonferroni Correction, (b) Mean expression (± s.d.) of known (in bold) and predicted targets of miRNAs in ASD (n=9, light grey bar) and control OMSCs (n=8, dark grey bar). Gene expression was measured using relative standard curve method, normalised against GPBP1 as reference gene. Results shown represent one of two independent repeats showing the same results, (c) Western blot showing down regulation of KCNK2 in ASD (n=3) with respect to control OMSCs (n=3). ACTB was used as loading control. Bottom panel displays densitometry of the bands using two images taken at low exposures. * P<0.05 by Student's paired 2-tailed t-test. (d) The 3'UTRs of GRIA3, KCNK2 and MAP2 are targeted by miRNAs. The 3'UTRs of GRIA3, KCNK2 and MAP2 were sub-cloned into the 3'UTR of Renilla Luciferase in the psiCheck2 plasmid and co-transfected into HEK293T with either plasmid overexpressing miR-146a, miR-221, miR-656 or empty plasmid. Ratio of Renilla/Firefly luciferase indicates the repression activity of miRNAs directly on the 3'UTR. Results are represented from one repeat of two showing the same results. *P<0,05, **P < 0.01, ***P<0,001 by Student's paired 2-tailed t-test.

EXAMPLE: Identification of a common set of microRNAs deregulated in ASD. Material & Methods Ethics statement: Human samples were obtained with informed consent of the patients, and studies were carried out under a protocol that was approved by the ethic committees of the Hopital Necker, Paris, and the Monteperrrin Hopital, Aix en Provence (CPP Marseille2). All the biological collections (blood, nucleic acid, tissues) are kept securely in one place and their management and the quality of their preservation follow the French regulations and ethical recommendations. Primary cortical neurons were extracted from mouse embryos (E15.5) following protocols approved by the local ethical committee of the Interdisciplinary Institute of Neuroscience of Bordeaux accordingly to the European Communities Council Directive (86/809/EEC). Patients' information: The patients included in this study have been clinically diagnosed as having ASD according to the criteria of DSM-V. We include here a brief clinical description of the patients (Supplementary Table 1). To investigate the genetic etiology, we extracted DNAs from their OMSCs and subjected to array comparative genomic hybridization and whole exome sequencing according to published protocol¹⁸. One patient (A3) carries two variants leading to premature termination in SCN2A and FAT1, which are both recurrent non- fully penetrant ASD genes; two patients (Al and A 10) were found to carry deletions at genomic region 22ql3.3 and 15ql3.2-ql3.3, respectively, which have been found to be associated with ASD. The genetic etiology of the remaining patients remained inconclusive.

Patients Controls

Code Age Sex Diagnosis Principal Symptoms Code Age

Infantile autism/

No verbal expression, very poor

Severe autism

Al 37 M social interactions, restricted C3 39

Important

interests and activities, passivity

dependence

Poor verbal expression (some

words and sentences), poor

Infantile autism/

social interactions, social

Severe autism

A2 22 F avoidance, restricted interests CI 21

Important

and activities, aimless motor

dependence

activity, motor stereotypes,

tantrums

Infantile autism/ Very poor verbal expression

Severe autism (few words), social avoidance

A3 35 M C6 35

Important and retrieval, restricted interests,

dependence self-injury

No major verbal delay, poor

Infantile autism/

gaze contact, inappropriate

Very severe

social interactions, cognitive

A5 22 F autism C9 34 disorders (dyslexia, dyscalculia),

Complete

social and professional

dependence

adaptation impairments

No verbal expression, relatively

Infantile

good verbal comprehension,

autism/Moderate

good visuo-spatial abilities, poor

A6 22 M to severe autism C5 18 social interactions, motor

Important

stereotypes, aggressive

dependence

behaviours

Infantile autism/ Very poor verbal expression

Moderate to (few words), poor social

A7 22 C7 18 severe autism interactions, restricted interests

Important and activities, gestural dependence stereotypes

No verbal expression, relatively

Severe autism good verbal comprehension,

A8 43 Important poor social interactions, body C8 44 dependence sway, some aggressive

behaviours

No significant verbal nor

Asperger intellectual delay, slow verbal

syndrome/ flow, gaze avoidance, poor

A9 22 Cll 32

Mild autism social contact, relatively good

Independence social and professional

adaptation

No verbal expression, very

Infantile autism/

restricted social interaction and

Profound autism

A10 38 activities, aimless motor activity, C2 40

Complete

motor stereotypes, self-injury,

dependence

incontinence

Very poor verbal expression

Infantile autism/ (few words), restricted verbal

Severe autism comprehension, gaze avoidance,

Al l 37 C4 34

Important restricted social interactions,

dependence motor stereotypes, aggressive

behaviours

Supplementary Table 1: Clinical descriptions of ASD patients whose OMSCs were used in this study. Maintaining of OMSC, human primary fibroblast and HEK293T: OMSC was biopsied and cultivated according to published protocol¹⁹' ²⁰. OMSC was maintained using DMEM/F-12 GlutaMAX™ (Life Technologies) supplemented with Penicillin-Steptomycin (100 U/ml) (Life Technologies) and fetal bovine serum. Human primary fibroblasts were extracted from skin punches with informed consents and maintained in RPMI 1640 Medium GlutaMAX™ (Life Technologies) supplemented with Penicillin-Steptomycin (100 U/ml) (Life Technologies) and fetal bovine serum. HEK293T cells were maintained in DMEM High Glucose Medium GlutaMAX™ (Life Technologies) supplemented with Penicillin- Steptomycin (100 U/ml) (Life Technologies) and fetal bovine serum. Both cell types were passaged first by detaching using 0.05% Trypsin-EDTA (IX) Phenol Red (Life Technologies) and re-plated in desired concentration.

Expression profiling of miRNA: miRNAs were extracted from frozen pellets using mirVana™ miRNA isolation kit (Life Technologies) according to manufacturer's instructions. Concentration was measured using Nanodrop 2000 (Thermo Scientific). During the discovery phase, miRNA profiling was performed across a two-card set of TaqMan® MicroRNA Arrays (Arrays A and B) (Life Technologies) for a total of 667 unique assays specific to human miRNAs (Sanger miRBase vlO). Analyses were done in technical triplicates and used 2 different miRNA extractions from 2 pellets. For validation, 24 miRNAs which were found either significantly deregulated in the first round or whose assays failed and the 7 reference miRNAs were assessed. Data were analyzed using both GenEX and RealTime StatMiner® softwares. Both integrate quality controls, selection of best endogenous controls, clustering, differential expression and 2-way ANOVA analysis. The samples included in miRNA profiling were: A2, A3, A5, A6, A7, A9, A10 and Al 1 vs. CI, C4, C5, C6, C7 and CIO. The miRNA profiling and analyses were performed as paid service by the IMAGIF Platform for high-throughput quantitative PCR, ICSN CNRS, Gif-sur-Yvette, France. To profile miRNAs in fibroblast samples, 11 miRNAs (miR-146a, miR-221, miR-654-5p, miR- 656 and 7 reference miRNAs) were assessed using Taqman assays on Fluidigm 48.48 array in technical triplicates using two different miRNA extractions. Sample preparation was done according to the manufacturer's protocol for Creating Custom RT and Preamplification Pools using Taqman® MicroRNA Assays (Life Technologies). This step was performed as paid service by Platform qPCR-HD-GPC, Ecole Normale Superieure, Paris, France. miRNA target prediction and pathway analysis: Target prediction was performed using mir-DIP which integrates predictions from multiple software. Only targets predicted by at least 3 different programs were included in pathway analysis. Gene ontology enrichment of predicted miRNA targets (-1000 targets per miRNA) was performed using Ingenuity Pathway Analysis (IP A).

Quantitative real time polymerase chain reaction: Total RNA was extracted using Triazol reagent (Life Technologies) and RNeasy Mini Kit (QIAGEN) with DNase I treatment step (QIAGEN). cDNA was reversed transcribed from extracted RNA using Superscript® II Reverse Transcriptase (Life Technologies). Gene expression was measured using SYBR Green Power Mix (Life Technologies) using primers specific for each gene. Relative standard-curve method was employed to calculate relative gene expression. Reactions were performed using the Step One Plus Real-Time PCR system (Applied Biosystems). Expression values were taken from mean of triplicates.

Cloning the 3'UTR of GRIA3, KCNK2 and MAP2: The entire or parts of the 3'UTR of GRIA3 (~2000bp), KCNK2 (~1300bp) and MAP2 (~1300bp) were amplified using specific primers, subcloned into TOPO 3.1 vector using TOPO® TA Cloning® Kit (Life Technologies) and transformed into One Shot® TOP 10 Chemically Competent E.Coli (Life Technologies) by heat shock method. Plasmid was extracted using Pure Yield™ Plasmid MiniPrep System (Promega), digested with Xhol and Notl-HF (New England Biolabs). Digested fragment was gel purified once more then cloned into cut psiCheck2 plasmid (Promega) using T4 DNA Ligase (New England Biolabs). For higher yield plasmid extraction, 100 ml of bacterial culture was subjected to Plasmid Midi Kit (QIAGEN). Cloned plasmids were sequenced using psiCheck2_hLucF and psiCheck2_hLucR for screening.

Dual Lucif erase Assay: Approximately 2 x 10⁵ HEK293T cells were plated in each well of 12 well plate the day before transfection. Cells were transfected with either plasmids over-expressing miRNAs, LentimiRa-GFP-mmu-mir-146a/221/656 (mml0082/mhl0296/mhl0968, ABM Good) or empty vector LentimiRa-GFP-empty (mOOl, ABM Good), together with psiCheck2_GRIA3_UTR, psiCheck2_KCNK2_UTR or psiCheck2_MAP2_UTR plasmids (Ratio 1 :3 to ensure good expression of luciferase). Transfection was performed using JetPRIME® Polyplus Transfection Reagent (Ozyme) followed manufacturer's instruction. Assay was performed using Dual-Luciferase® Reporter Assay System (Promega) 24 H after transfection. Ratio of Renilla Luciferase to Firefly Luciferase was taken as mean of technical triplicates.

Chromatin immunoprecipitation: ChIP experiments were carried out as previously described (Navarro et al. 2010). Briefly, 10 μg of sonicated chromatin (average length 200- 500bp) were incubated with 2 μg of antibody against H3K4me3 (04-745, Millipore), H3K27me3 (07-449, Millipore), H3Ac (06-599, Millipore) and H3K27Ac (39133, Active Motif). Real-time PCR was performed in triplicate on IP and input DNA using Power SYBR green PCR master mix on Viia7 Fast Real-Time PCR machines (Life Technologies). To assess the enrichment of each target genomic region after ChIP, we calculated the ratio between the average value obtained for the IP DNA and the corresponding input, both values being first normalized for dilution factors. Results obtained are presented as percentage of enrichment. In situ-hybridisation analysis:

Tissue preparation and processing. Male, 2-month old wild type mice were deeply anesthetized using pentobarbital and fixed by intracardiac perfusion with 4% paraformaldehyde in PBS. Brains were removed, post-fixed overnight at 4°C and cryo- protected using 2 baths of Tris-HCl buffered Saline (TBS) with 0,5M sucrose at 4°C during 48h. Brains were then frozen in bath of isopentane at -35-40 °C and stored at -80°C. 20μιη serial sections were done in a cryomold (Tissue-Tek) with a cryostat (Leica), mounted on Superfrost Plus glass slides (Therme Scientific) and preserved at -80°C. For embryos and early post-natal animals, intracardiac perfusion was replaced by direct 4% paraformaldehyde fixation overnight after decapitation

In situ hybridization procedure. Tissue sections mounted on glass slides were thawed and air dried for lh at room temperature, incubated in a solution containing 20 μg/ml proteinase K (Sigma) in TBS, pH 7,4 for 3 minutes, and then washed 5 minutes two times in TBS. Samples were fixed in 4% PFA for 10 minutes, washed with 2 mg/ml glycine in TBS, and twice 5 minutes in TBS. To remove residual phosphate from the TBS washes, slides processed with EDC fixation were incubated twice for 10 minutes in freshly prepared solution containing 0,13M 1-methylimidazole, 300mM NaCl, pH 8.0 adjusted with HCL. In the meantime, a solution of 0,16 M l-ethyl-3-(3-dimethylaminopropyl) carbodiimide (EDC, Sigma) is prepared by adding EDC into 1-methylimidazole and 300mM NaCl (pH 8.0) solution. The pH of the EDC solution is re-adjusted by adding 12M HC1 to pH 8.0. Slides are maintained in a humidified chamber for one hour at room temperature incubated in 500μ1 of EDC solution to each slide. The slides are then washed in 2 mg/ml glycine/TBS solution and then twice 5 minutes in TBS. For enzymatic inactivation, slides were acetylated by incubating for 30 minutes in solution freshly prepared 0,1M triethanolamine and 0,5% (v/v) acetic anhydride. Slides were then rinsed twice 5 minutes in TBS. For pre-hybridization, the tissue sections were covered with 500 μΐ of hybridization buffer containing 50% formamide, 5x SSC, 5x Denhardt's solution, 0,25 mg/ml yeast tRNA, 0,5 mg/ml salmon sperm DNA, 20 mg/ml blocking reagent (Roche), 1 mg/ml 3-((3-Cholaminodopropyl) dimethylammonio)-l- propanesulfate (CHAPs, Sigma), 0,5 % tween at room temperature for 2 hours in a humidified chamber. The hybridization buffer was removed by tilting the slide. For hybridization, 40 nM of DIG-labeled LNA probe (mmu-miR- 146a, mmu-miR-scr amble) diluted in hybridization buffer were applied per section, and covered with covers lips. The slides were incubated in a sealed humidified chamber for 16 h at 42°C.

Slides were immersed 3 minutes in 2x SSC at 52°C and then washed 5 minutes twice in lx SSC at 52°C and once in lx SSC at room temperature. Slides were washed again 5 minutes twice in 0,2x SSC at room temperature and once in TBS. In preparation for probe detection, 500μ1 of blocking solution containing 5 mg/ml blocking reagent (Roche), 10%> (v/v) normal goat serum, and 0,1 % (v/v) tween in TBS was applied to each slide for lh at room temperature in a humidified chamber. Antibody anti-DIG-FAB peroxidase (POD) (Roche) diluted 1 :500 in blocking solution was incubated overnight. Slides were then washed 3 times in TBS. Slides were incubated in freshly prepared NBT/BCIP substrate reagent containing 50 mM MgC12, 100 mM NaCl, 0,2mM Levamisole, 0,5 mg/ml NBT, 0,1875 mg/ml BCIP in lOOmM TBS pH9.5, and protected from light during the exposure time (approximately 6 h). Sides were then washed twice in water, then twice in TBS and incubated for 10 minutes in 4% PFA solution. Finally, slides were washed in water and mounted using 2 drops of mounting medium.

Immunohistochemistry and miRNA ISH co-staining: To assess cellular localization of miR-146a, brain sections were processed for both immunofluorescence staining and ISH staining. Slides were washed 3 times in PBS and then, incubated lh at room temperature in humidified chamber, with 500 μΐ of blocking solution containing 0,1% (v/v) triton-lOOX and 2% (v/v) normal donkey serum in PBS. The blocking solution was removed by tilting the slide. Slides were incubated overnight at room temperature with the same blocking solution containing polyclonal antibody against Fox3/NeuN (1 : 1000; Abeam, AB 104225) for specific neuronal staining and antibody against GFAP (1 :500; Abeam, ab7260). Slides were rinsed three times in PBS and incubated for 1 h 30 min at room temperature with donkey Alexa fluor 488-conjugated anti-rabbit (for NeuN), secondary antibody diluted in blocking solution. After being rinsed in PBS, then water, slides were mounted in mounting medium.

Microscopy and image processing: Images were captured on an upright epifluorescence microscope, Nikon Eclipse Ni-U (Nikon France S.A) using lOx objective CFI Plan Fluor NA 0,30 and 40x objective CFI Plan Fluor NA 0.75. For fluorescent imaging filters sets for GFP were used, and pictures acquired using a Zyla SCMOS camera (Andor Technology Ltd., Belfast, UK).

Primary neuronal cultures: Cultured mouse hippocampal neurons were prepared from El 6.5 embryos, grown on 18 mm glass coverslips coated with laminin/polylysin and maintained in Neurobasal B27-supplemented medium. Neurons were transfected using Lipofectamine 2000 on Days In Vitro 11 (DIV 11 ) with 800ng of plenti-III-miR-GFP-Blank or pLenti-III-miR146a-GFP and experiments were performed at DIV18.

Sholl analysis: Analysis of the dendritic morphology was done blind to the genotype. Sholl analysis was done using a sholl analysis pluggin on ImageJ, and determined the number of dendritic branch intersections with concentric circles of increasing radii (interval of 10 μιη) from the soma. All neurons were imaged at the same magnification and images were threshold, and to avoid false intersections because of noise, each threshold images were checked and compare to the original. The BI (branching index) was shown: The BI compares the difference in the number of intersections made in pairs of circles relative to the distance from the neuronal soma, following the following equation: BI=∑(Intersections circle n-Intersections circle n-l).rn according to (A new mathematical function to evaluate neuronal morphology using the Sholl analysis : Luis Miguel Garcia-Segura, Julio Perez- Mar quez). Results

Identification of miRNA deregulation signature in ASD: We harvested OMSCs from the lamina-propria layer of the nasal cavities of living ASD patients and age and sex matched controls (Supplementary Table 1) and profiled 667 miRNAs (Sanger miRBase vlO) in 8 patients and 6 controls using the TLDA high throughput real-time qPCR approach in 2 rounds (Fig. la). The genetic etiology of these patients is mostly unknown according to whole exome sequencing and array comparative genomic hybridisation analyses (Methods). In keeping with the current gold standard in miRNA analysis, miRNA relative quantity were calculated using ACt method using the geometrix mean of 7 house-keeping miRNAs : miR-let-7g, miR- 106a, miR-151-3p, miR-15b, miR-16, miR-17 and miR-99b, and the average of all controls for normalisation²¹' ¹¹. 24 miRNAs were included in the second round for validation (Fig. la). These miRNAs were either found significantly deregulated in the patient groups. These analyses identified 4 miRNAs, miR-146a (up-regulated 2-fold), miR-221, miR-654-5p and miR-656 (all three down-regulated -1.3 fold), significantly deregulated in the patients (P < 0.05) (Fig. lb). Notably, miR-654-5p and miR-656 are conserved only in the primate chain, suggesting that they could be important for higher cognitive function; by contrast, miR-146a and miR-221 sequences are 100% conserved in the mouse, suggesting a role in development.

Tissue- and disease specificity of the miRNA signature: We questioned the tissue specificity of this miRNA signature and tested their expression in more easily accessible tissue. For this, we used primary skin fibroblasts from 5 new ASD patients and 4 controls. This analysis showed a significant deregulation of miR-146a, miR-221 and miR-654-5p in the ASD group compared to controls in the same trend as in OMSC (Fig. lc) not only confirming the strength of the signature, but also demonstrating that such signature is conserved in other cell type. We then raised the issue of the disease specificity of this signature in related neurodevelopmental disorders and extended the analysis to primary skin fibroblasts from patients who have known causes of intellectual disability (ID) without autistic features (n=12). Interestingly, ID patients do not share the same signature as ASD (Fig. lc). Yet, miR- 146a, miR-221 and miR-656 were all significantly upregulated in the ID group, supporting the roles of these miRNAs in shared pathways affected in both ASD and ID. The principal component analyses on results obtained from round 3 of OMSC profiling, revealed that the combination of miR-146a and miR-221 explain 86% of the difference between the ASD and control groups (Fig. Id). Meanwhile, partition analysis on the results obtained from the primary skin fibroblasts indicated that the combination of miR-221 and miR-654-5p permitted clear distinction between the ASD, ID and the control groups (Fig. le).

ASD-deregulated miRNAs target neuronal relevant genes and pathways: Next, we searched for putative mRNA targets deregulated in patients' OMSCs. Using miR-DIP to integrate target prediction from multiple programs, we found that on average there are 1000 targets predicted for each miRNA by at least 3 different prediction programs. We integrated the prediction analyses with that of pathway enrichment using Ingenuity Pathway Analysis and identified several highly relevant pathways (Fig. 2a). These are neuronal pathways, including axonal guidance signalling, signalling by Rho Family GTPases, actin cytoskeleton signalling and synaptic long-term potentiation, as well as immunological pathways which include IL-8 signalling, CXCR4 signalling and macropinocytosis signalling (Fig. 2a).

We assayed by RT-qPCR transcript levels for several known and predicted miRNA targets involved in the identified pathways (Fig. 2b). Decreased levels of GRIA3 and KCNK2 transcripts and an increased level of MAP2 were found in patients' OMSCs (Fig. 2b). Decreased KCNK2 in the patient' cells was confirmed at the protein level by Western blot analysis (Fig. 2c). GRIA3 and MAP2 protein levels could however not be assessed as these are below the Western blot detection threshold. We thus used a dual luciferase reporter assay to confirm that these transcripts are direct targets of the identified miRNAs. Overexpressing either miR-146a, miR-221 or miR-656 significantly reduced the luciferase activity of reporter construct carrying the 3 '-untranslated region (3'-UTR) of GRIA3, KCNK2 and MAP2 (Fig. 2d).

Epigenetic modification underlines miR-146a deregulation in ASD: miR-146 is 100% conserved in the mouse, has already been reported as significantly upregulated in lymphoblastoid cell lines of another cohort of ASD patients⁹, and is also upregulated in patients with ID (Fig. lc). Collectively, these data suggest that it may play a key role in ASD and in brain development. We first investigated whether the deregulation of miR-146a expression in ASD patients could originate from an alteration of the transcriptional potential of its promoter region. Therefore, we examined the chromatin structure across the miR-146a assessing histone modifications associated with either active (H3K4me3, H3K27Ac and H3Ac) or inactive chromatin (H3K27me3). Transcriptionally active miR-146a promoters were devoid of repressive H3K27me3 marks in OMSC. On the other hand, miR-146a promoter was enriched for activating H3K27Ac, H3K4me3 and H3Ac marks. Notably, significantly higher levels of H3K27Ac and H3K4me3 were observed in ASD cells compared to controls (corrected P values < 0.05). These results are in agreement with the miR-146a overexpression observed in patients' cells and provide evidence for a common epigenetic deregulation of miR-146a in ASD patients. miR-146a is highly expressed in neurons and its overexpression leads to altered dendritic arborisation: In the mouse brain, miR-146a is expressed throughout the cortex, hippocampus and amygdala as evidenced by in-situ hybridisation. During development, miR- 146a expression exhibits an initially high and widespread expression that becomes restricted to some cellular layers in the above-cited postnatal brain regions. In situ hybridization associated with the immune-detection of cell-specific markers showed that in the adult mouse brain, miR-146a is essentially expressed in neurons, whereas very few labelling was detected in the glial lineage. Thus, miR-146a displays strong expression in neuronal cells in brain regions known to be important for high cognitive functions.

To explore the potential pathogenic consequences of miRNA deregulation, we examined the impact of miR-146a overexpression to neuronal cells morphology. Dendritic arborisation extent and complexity of primary neuronal cultures was assessed using Sholl analysis (Methods). When compared to GFP expressing cells transfected with a blank construct, miR-146a overexpressing neurons displayed shrivelled dendritic trees with branching points occurring more proximal than in control conditions. This suggests that the level miR-146a expression in an important determinant for neuronal development.

DISCUSSION:

We report here a comprehensive analysis of miRNA expression profiles in human stem cells from ASD patients. We identified a signature of four miRNAs (miR-146a, miR- 221, miR-654-5p and miR-656) commonly deregulated in ASD. Despite the limited number of samples tested, potentially restricting the statistical power, the observed conservation of the signature in primary skin fibroblasts from unrelated patients strongly supports its biological relevance. Moreover, our finding that ID patients do not share the same signature as ASD may provide strong diagnosis implications.

Target prediction analysis suggests that the transcripts regulated by these miRNAs code for proteins that participate to neurodevelopmental processes as well as to immune response and inflammation, both of which are relevant for pathology of ASD. It is now well established that the immune system has a tremendous impact on behaviour; many studies have demonstrated that immune imbalance impairs higher order brain functioning²³. Accordingly, early immune activation has been reported in individuals with ASD, with altered neuro-inflammatory processes and abnormal immune responses in adulthood²⁴. Moreover, DNA methylation analysis of the autistic brain have identified a very significant enrichment of differentially methylated region in genomic areas responsible for immune functions⁵. Along these lines, it is noteworthy that miR-146a was first identified as an immune system regulator and possibly plays an important role in the neuroinflammation²⁵'²⁶. Our results support the idea that altered miR146a expression may contribute to the changes in immune response genes observed in ASD.

miR-146a and miR-221 have also established roles in neurodevelopment. Validated targets of miR-146a include MAP1B, which regulates AMPA receptor endocytosis³⁰, meanwhile, direct target of miR-221 include FMR1, the Fragile-X gene which is the most frequent cause of syndromic intellectual disability (ID) and ASD³¹. Using target prediction tools combined to dual luciferase assays and Western Blot analysis, we report here on additional neuronal transcripts regulated by these miR As. We first showed in vitro that GRIA3 is a target transcript for both miR-146a and miR-221. GRIA3 encodes for a core subunit of the AMPA receptor and findings from various experimental systems implicate ionotropic GluR dysfunction in ASD³². In addition, mutations in this gene are associated with moderate cognitive impairment in humans³³. miR-146a up-regulation in ASD brain might thus alter AMPA receptor function through both impaired MAP IB-mediated endocytosis³⁰ and decreased amount of GR1A3. We also demonstrated that miRNAs expression deregulation correlates with a reduced level of KCNK2 protein in patients' OMSCs and that both miR- 146a and miR-221 directly target KCNK2 3'UTR. KCNK2 codes for a member of the potassium leak channel family, a group of proteins that are critical determinants of neuronal excitability in the cortex³⁴ and KCNK2 knockdown impairs neuronal migration in the developing mouse cerebral cortex³⁵. We thus propose that reduced amount of KCNK2 protein could contribute to the failure or delay in neuronal migration that has been observed in some cases of autism and ID. Additionally, we showed that MAP2 is a direct target for miR-656 and is overexpressed in patients' OMSCs. Altered MAP2 protein levels have already been described in neurodevelopmental disorders: in Rett syndrome there is general loss of MAP2 expression even in young children, whereas in Down's syndrome upregulation of this protein has been described. MAP2 deregulation may thus participate to the reduced neuronal plasticity underlying ASD. Lastly, previous studies also demonstrated that miR-146a inhibits the expression of neuron-specific targets Nlgnl, and Sytl, preventing glia from mistakenly adopting neuron-specific phenotypes³⁶. Abnormal expression of miR-146a during development might thus also impair astrocyte differentiation and function in ASD brain, contributing to the failure of neural connectivity and CNS homeostasis underlying autism. One prevalent theory of the pathogenesis of autism relates to a deficiency of plasticity of axonal sprouting and synaptic connectivity. Consistently, impaired (both reduced or increased) dendritic arborization has been observed in several models for ASD. There have been reports in autism cases of enlarged or abnormally oriented neurons, densely packed neuronal regions and isolated regions of atrophic neurons with reduced dendritic arborization³⁷' ³⁸. Increased spine density has been reported in fragile X syndrome³⁹. Conversely, neuronal atrophy and reduced dendritic spines is observed in Rett Syndrome⁴⁰. The observation that miR-146a overexpression alters neuronal dendritic arborisation suggests that the miRNA signature identified here may participate to the excess or reduced connectivity found in ASD.

Our findings support the role of epigenetic dysfunction in the etiology of ASD and may have important implication for development of future therapies. Indeed, unlike genetic changes, epigenetic modifications are reversible pointing toward the possibility to therapeutically revert epigenetic marks to re-establish prior gene expression patterns. Several classes of epigenetic drugs are being investigated. These include DNA methyltransferases (DNMTs) inhibitors, histone deacetylases (HDACs) inhibitors, histone methyltransferase (HMT) inhibitors and drugs targeting miRNAs. It was recently shown that the major antidepressant drug imipramine reverses the depressive state by altering an epigenetic mark (histone modification) in the brain-derived neurotrophic factor gene, Bdnf, in the hippocampus⁴¹. Similarly, valproic acid, a HDACs inhibitor, was found to normalize histone acetylation of genes in the hippocampus, and to suppress cognitive impairment by blocking aberrant neurogenesis⁴². These observations indicate that chemicals such as HDAC inhibitors that can alter epigenetic gene expression may be candidates for the treatment of neurodevelopmental and neurodegenerative diseases⁴³' ⁴⁴. Finally, in addition to chemical treatments, experiments have shown that appropriate environmental conditions (for example, providing toys that stimulate the brain) could ameliorate the neurological features of Mecp2 knockout mice by altering gene expression and synaptogenesis in the brain⁴⁵. These results suggest that a stimulating educational environment may potentially alter the epigenetic status. Thus, epigenetics may provide useful scientific information for the assessment of specific educational conditions in ASD patients.

In conclusion, our data present the identification of a conserved miRNA signature in ASD, and its discriminative power between ASD, ID and control groups. Further studies are now necessary to address when miRNA expression deregulation occurs during development of ASD brains and how it precisely impairs synaptogenesis and synaptic plasticity. Importantly, our findings raise the possibility of using miRNA signature as a biomarker for ASD, providing a mean for detecting early signs of the disease and allowing early intervention.

REFERENCES:

Throughout this application, various references describe the state of the art to which this invention pertains. The disclosures of these references are hereby incorporated by reference into the present disclosure.

1. Coe BP, Girirajan S, Eichler EE. The genetic variability and commonality of neurodevelopmental disease. Am J Med Genet C Semin Med Genet 2012; 160C(2): 118-129.

2. Ronemus M, Iossifov I, Levy D, Wigler M. The role of de novo mutations in the genetics of autism spectrum disorders. Nat Rev Genet 2014; 15(2): 133-141.

3. Gaugler T, K L, Sanders SJ, Bodea CA, Goldberg AP, Lee AB et al. Most genetic risk for autism resides with common variation. Nat Genet 2014; 46(8): 881-885.

4. Hu VW, Frank BC, Heine S, Lee NH, Quackenbush J. Gene expression profiling of lymphoblastoid cell lines from monozygotic twins discordant in severity of autism reveals differential regulation of neurologically relevant genes. BMC Genomics 2006; 7: 118.

5. Nardone S, Sams DS, Reuveni E, Getselter D, Oron O, Karpuj M et al. DNA methylation analysis of the autistic brain reveals multiple dysregulated biological pathways. Translational psychiatry 2014; 4: e433.

6. Wang W, Kwon EJ, Tsai LH. MicroRNAs in learning, memory, and neurological diseases. Learn Mem 2012; 19(9): 359-368.

7. Miller BH, Wahlestedt C. MicroRNA dysregulation in psychiatric disease. Brain Res 2010; 1338: 89-99.

8. Chan AW, Kocerha J. The Path to microRNA Therapeutics in Psychiatric and Neurodegenerative Disorders. Frontiers in genetics 2012; 3: 82. Talebizadeh Z, Butler MG, Theodora MF. Feasibility and relevance of examining lymphoblastoid cell lines to study role of microRNAs in autism. Autism research : official journal of the International Society for Autism Research 2008; 1(4): 240-250. Ghahramani Seno MM, Hu P, Gwadry FG, Pinto D, Marshall CR, Casallo G et al. Gene and miRNA expression profiles in autism spectrum disorders. Brain Res 2011; 1380: 85-97.

Sarachana T, Zhou R, Chen G, Manji HK, Hu VW. Investigation of post- transcriptional gene regulatory networks associated with autism spectrum disorders by microRNA expression profiling of lymphoblastoid cell lines. Genome Med 2010; 2(4): 23.

Abu-Elneel K, Liu T, Gazzaniga FS, Nishimura Y, Wall DP, Geschwind DH et al. Heterogeneous dysregulation of microRNAs across the autism spectrum. Neurogenetics 2008; 9(3): 153-161.

Okano H, Yamanaka S. iPS cell technologies: significance and applications to CNS regeneration and disease. Molecular brain 2014; 7: 22.

Delorme B, Nivet E, Gaillard J, Haupl T, Ringe J, Deveze A et al. The human nose harbors a niche of olfactory ectomesenchymal stem cells displaying neurogenic and osteogenic properties. Stem cells and development 2010; 19(6): 853-866.

Fan Y, Abrahamsen G, Mills R, Calderon CC, Tee JY, Leyton L et al. Focal Adhesion Dynamics Are Altered in Schizophrenia. Biol Psychiatry 2013.

Ronnett GV, Leopold D, Cai X, Hoffbuhr KC, Moses L, Hoffman EP et al. Olfactory biopsies demonstrate a defect in neuronal development in Rett's syndrome. Ann Neurol 2003; 54(2): 206-218.

Cook AL, Vitale AM, Ravishankar S, Matigian N, Sutherland GT, Shan J et al. NRF2 activation restores disease related metabolic deficiencies in olfactory neurosphere- derived cells from patients with sporadic Parkinson's disease. PLoS One 2011; 6(7): e21907.

Langouet M, Saadi A, Rieunier G, Moutton S, Siquier-Pernet K, Fernet M et al. Mutation in TTI2 reveals a role for triple T complex in human brain development. Hum Mutat 2013; 34(11): 1472-1476.

Murrell W, Feron F, Wetzig A, Cameron N, Splatt K, Bellette B et al. Multipotent stem cells from adult olfactory mucosa. Dev Dyn 2005; 233(2): 496-515. Girard SD, Deveze A, Nivet E, Gepner B, Roman FS, Feron F. Isolating nasal olfactory stem cells from rodents or humans. Journal of visualized experiments : JoVE 2011; (54).

Vandesompele J, De Preter K, Pattyn F, Poppe B, Van Roy N, De Paepe A et al. Accurate normalization of real-time quantitative RT-PCR data by geometric averaging of multiple internal control genes. Genome Biol 2002; 3(7): RESEARCH0034.

Mestdagh P, Van Vlierberghe P, De Weer A, Muth D, Westermann F, Speleman F et al. A novel and universal method for microRNA RT-qPCR data normalization. Genome Biol 2009; 10(6): R64.

Voineagu I, Eapen V. Converging Pathways in Autism Spectrum Disorders: Interplay between Synaptic Dysfunction and Immune Responses. Frontiers in human neuroscience 2013; 7: 738.

Depino AM. Peripheral and central inflammation in autism spectrum disorders. Mol Cell Neurosci 2013; 53: 69-76.

Taganov KD, Boldin MP, Chang KJ, Baltimore D. NF-kappaB-dependent induction of microRNA miR-146, an inhibitor targeted to signaling proteins of innate immune responses. Proc Natl Acad Sci U S A 2006; 103(33): 12481-12486.

Hou J, Wang P, Lin L, Liu X, Ma F, An H et al. MicroRNA-146a feedback inhibits RIG-I-dependent Type I IFN production in macrophages by targeting TRAF6, IRAKI, and IRAK2. J Immunol 2009; 183(3): 2150-2158.

Pauley KM, Satoh M, Chan AL, Bubb MR, Reeves WH, Chan EK. Upregulated miR- 146a expression in peripheral blood mononuclear cells from rheumatoid arthritis patients. Arthritis research & therapy 2008; 10(4): R101.

Cui JG, Li YY, Zhao Y, Bhattacharjee S, Lukiw WJ. Differential regulation of interleukin-1 receptor-associated kinase- 1 (IRAK-1) and IRAK-2 by microRNA- 146a and NF-kappaB in stressed human astroglial cells and in Alzheimer disease. / Biol Chem 2010; 285(50): 38951-38960.

Lukiw WJ, Zhao Y, Cui JG. An NF-kappaB-sensitive micro RNA-146a-mediated inflammatory circuit in Alzheimer disease and in stressed human brain cells. / Biol Chem 2008; 283(46): 31315-31322.

Chen YL, Shen CK. Modulation of mGluR-dependent MAP IB translation and AMPA receptor endocytosis by microRNA miR-146a-5p. / Neurosci 2013; 33(21): 9013- 9020. 31. Zongaro S, Hukema R, D'Antoni S, Davidovic L, Barbry P, Catania MV et al. The 3' UTR of FMR1 mRNA is a target of miR-101, miR-129-5p and miR-221 : implications for the molecular pathology of FXTAS at the synapse. Human Molecular Genetics 2013.

32. Essa MM, Braidy N, Vijayan KR, Subash S, Guillemin GJ. Excitotoxicity in the pathogenesis of autism. Neurotoxicity research 2013; 23(4): 393-400.

33. Wu Y, Arai AC, Rumbaugh G, Srivastava AK, Turner G, Hayashi T et al. Mutations in ionotropic AMPA receptor 3 alter channel properties and are associated with moderate cognitive impairment in humans. Proc Natl Acad Sci U S A 2007; 104(46): 18163-18168.

34. Goldstein SA, Bockenhauer D, O'Kelly I, Zilberberg N. Potassium leak channels and the KCNK family of two-P-domain subunits. Nat Rev Neurosci 2001; 2(3): 175-184.

35. Bando Y, Hirano T, Tagawa Y. Dysfunction of KCNK potassium channels impairs neuronal migration in the developing mouse cerebral cortex. Cereb Cortex 2014; 24(4): 1017-1029.

36. Jovicic A, Roshan R, Moisoi N, Pradervand S, Moser R, Pillai B et al. Comprehensive expression analyses of neural cell-type-specific miRNAs identify new determinants of the specification and maintenance of neuronal phenotypes. J Neurosci 2013; 33(12): 5127-5137.

37. Bauman ML, Kemper TL. Neuroanatomic observations of the brain in autism: a review and future directions. International journal of developmental neuroscience : the official journal of the International Society for Developmental Neuroscience 2005; 23(2-3): 183-187.

38. Murcia CL, Gulden F, Herrup K. A question of balance: a proposal for new mouse models of autism. International journal of developmental neuroscience : the official journal of the International Society for Developmental Neuroscience 2005; 23(2-3): 265-275.

39. Bagni C, Greenough WT. From mRNP trafficking to spine dysmorphogenesis: the roots of fragile X syndrome. Nat Rev Neurosci 2005; 6(5): 376-387.

40. Zoghbi HY. Postnatal neurodevelopmental disorders: meeting at the synapse? Science 2003; 302(5646): 826-830.

41. Tsankova NM, Berton O, Renthal W, Kumar A, Neve RL, Nestler EJ. Sustained hippocampal chromatin regulation in a mouse model of depression and antidepressant action. Nat Neurosci 2006; 9(4): 519-525. Jessberger S, Nakashima K, Clemenson GD, Jr., Mejia E, Mathews E, Ure K et al. Epigenetic modulation of seizure-induced neurogenesis and cognitive decline. / Neurosci 2007; 27(22): 5967-5975.

Renthal W, Nestler EJ. Epigenetic mechanisms in drug addiction. Trends Mol Med 2008; 14(8): 341-350.

Sleiman SF, Berlin J, Basso M, Karuppagounder SS, Rohr J, Ratan RR. Histone Deacetylase Inhibitors and Mithramycin A Impact a Similar Neuroprotective Pathway at a Crossroad between Cancer and Neurodegeneration. Pharmaceuticals 2011; 4(8): 1183-1195.

45. Lonetti G, Angelucci A, Morando L, Boggio EM, Giustetto M, Pizzorusso T. Early environmental enrichment moderates the behavioral and synaptic phenotype of MeCP2 null mice. Biol Psychiatry 2010; 67(7): 657-665.

Claims

CLAIMS:

1. An in vitro method for predicting or diagnosing autism spectrum disorder (ASD) in a subject, said method comprising a step of determining the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in a biological sample obtained from said subject.

2. The method according to claim 1, further comprising a step of comparing the expression levels of said at least two miRs with their respective predetermined reference values; and determining whether the subject has or is at risk of having ASD when the determined expression levels of said at least two miRs is different (higher or lower) to their respective predetermined reference values.

3. The method according to claim 1 or 2, wherein said biological sample is an olfactory mucosa biopsy or a skin biopsy obtained from said subject.

4. The method according to any one claims 1 to 3, wherein the expression levels of miR- 146a and miR-221 are determined.

5. The method according to claim 4, wherein the expression level of miR-656 is further determined.

6. The method according to claim 4 or 5, wherein the expression level of miR-654-5p is further determined.

7. A kit suitable for performing the methods according to any one claims 1 to 6, wherein said kit comprises means for measuring the expression levels of at least two miRs selected from the group consisting of miR-146a, miR-221, miR-654-5p and miR-656 in a biological sample obtained from the subject.

8. The kit according to claim 7, wherein said kit comprises means for measuring the expression levels of at least miR-146a and miR-221.

9. The kit according to claim 8, wherein said kit further comprises means for measuring the expression levels of miR-654-5p and/or miR-656.