Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=iwbp20

Download by: [New York University] Date: 04 November 2015, At: 22:01

The World Journal of Biological Psychiatry

ISSN: 1562-2975 (Print) 1814-1412 (Online) Journal homepage: http://www.tandfonline.com/loi/iwbp20

Excitatory/inhibitory imbalance in autismspectrum disorders: Implications for interventionsand therapeutics

Genoveva Uzunova, Stefano Pallanti & Eric Hollander

To cite this article: Genoveva Uzunova, Stefano Pallanti & Eric Hollander (2015): Excitatory/inhibitory imbalance in autism spectrum disorders: Implications for interventions andtherapeutics, The World Journal of Biological Psychiatry, DOI: 10.3109/15622975.2015.1085597

To link to this article: http://dx.doi.org/10.3109/15622975.2015.1085597

Published online: 15 Oct 2015.

Submit your article to this journal

Article views: 124

View related articles

View Crossmark data

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY, 2015http://dx.doi.org/10.3109/15622975.2015.1085597

REVIEW ARTICLE

Excitatory/inhibitory imbalance in autism spectrum disorders: Implications forinterventions and therapeutics

Genoveva Uzunova1, Stefano Pallanti1,2,3,4 and Eric Hollander1

1Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY, USA, 2Psychiatry and Behavioural Sciences, UC DavisHealth System, CA, USA, 3Department Psychiatry, University of Florence, Florence, Italy, and 4Icahn School of Medicine at Mount Sinai,New York, NY, USA

ABSTRACTObjectives: Imbalance between excitation and inhibition and increased excitatory-inhibitory (E-I)ratio is a common mechanism in autism spectrum disorders (ASD) that is responsible for thelearning and memory, cognitive, sensory, motor deficits, and seizures occurring in these disorders.ASD are very heterogeneous and better understanding of E-I imbalance in brain will lead to betterdiagnosis and treatments. Methods: We perform a critical literature review of the causes andpresentations of E-I imbalance in ASD. Results: E-I imbalance in ASD is due primarily to abnormalglutamatergic and GABAergic neurotransmission in key brain regions such as neocortex,hippocampus, amygdala, and cerebellum. Other causes are due to dysfunction of neuropeptides(oxytocin), synaptic proteins (neuroligins), and immune system molecules (cytokines). At theneuropathological level E-I imbalance in ASD is presented as a ‘‘minicolumnopathy’’. E-I imbalancealters the manner by which the brain processes information and regulates behaviour. Newdevelopments for investigating E-I imbalance such as optogenetics and transcranial magneticstimulation (TMS) are presented. Non-invasive brain stimulation methods such as TMS fortreatment of the core symptoms of ASD are discussed. Conclusions: Understanding E-I imbalancehas important implications for developing better pharmacological and behavioural treatments forASD, including TMS, new drugs, biomarkers and patient stratification.

ARTICLE HISTORYReceived 14 October 2014Revised 17 July 2015Accepted 18 August 2015

KEYWORDSAutistic disorder; transcranialmagnetic stimulation; elec-troencephalography; bio-markers; translationalmedical research

I. Introduction. A heuristic model for develop-ment of interventions and therapeutics

A widely-accepted hypothesis on the aetiology of autism

spectrum disorders (ASD) proposes that there is excita-

tory-inhibitory imbalance (E-I imbalance) in brain neural

circuits (Rubenstein and Merzenich 2003). This imbal-

ance underlies the social, behavioural, emotional, cog-

nitive, sensory and motor control abnormalities.

Increased E-I ratio in prefrontal cortex is shown to

result in behavioural and social impairments character-

istic of ASD using optogenetics (Yizhar et al. 2011).

Decreased E-I ratio is found in Rett syndrome (RS), a

form of ASD (Eichler and Meier 2008). We will refer to

these mechanisms collectively as E-I imbalance. Several

reviews highlight the significance of E-I imbalance in

ASD (Rubenstein and Merzenich 2003; Polleux and

Lauder 2004; Eichler and Meier 2008; Baroncelli et al.

2011; Ecker et al. 2013; Zikopoulos and Barbas 2013).

Studies show local hyperconnectivity and long-range

hypoconnectivity and disconnection between neural

circuits in ASD (Courchesne and Pierce 2005; Peters et al.

2013; Tye and Bolton 2013). The literature on ASD has

greatly expanded and we consider it is necessary to

write a review highlighting new advances on E-I imbal-

ance in ASD such as the role of the immune system, new

methods to study E-I, biomarkers, patient stratification

and selection of appropriate treatments. We will focus

on developments with significance for neuroscience-

informed and research-based clinical therapeutic

approaches.

Factors controlling the formation and functioning of

excitatory and inhibitory synapses affect the E-I ratio. The

earliest time point when E-I imbalance occurs in ASD is

important to identify in order to develop early thera-

peutic interventions. There are challenges to studying E-I

imbalance in human brain and much of our knowledge

derives from animal models. Although this knowledge

CONTACT Dr Eric Hollander eholland@montefiore.org Clinical Professor of Psychiatry and Behavioural Sciences and Director – Autism and OCD Spectrum& Anxiety and Depression Research Programs, Bronx, NY, USA.

� 2015 Taylor & Francis

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

largely overlaps with humans, there are differences,

and it is necessary to uncover the mechanisms in human

ASD with techniques such as transcranial magnetic

stimulation (TMS), electroencephalography (EEG), mag-

netoencephalograpgy (MEG) and post-mortem brain

studies.

II. Mechanisms and presentation of E-Iimbalance

1. Mechanisms by which E-I imbalance impacts

the clinical presentation of ASD

ASD are neurodevelopmental disorders with symptoms

in two core domains – social-communication deficits

and repetitive behaviours and restricted interests.

Associated symptoms are irritability, hyperactivity,

movement abnormalities, language delay, sensory

issues, seizures, learning disabilities, developmental

regression and gastrointestinal symptoms. Most ASD

present without an apparent cause (i.e. idiopathic). One

group, ‘‘syndromic forms’’, result from single-gene

defects such as Tuberous Sclerosis Complex 1 and 2

(TSC1 and 2), Fragile X syndrome (FXS), RS and

Angelman syndrome. Toxins, infections, immunological,

nutritional, epigenetic factors and stress may play roles

in ASD.

E-I imbalance may be due to increase in glutama-

tergic or decrease in GABAergic signalling. The E-I

ratio in neocortex is determined by the activity of

pyramidal glutamatergic neurones and inhibitory

GABAergic parvalbumin (PV)-positive interneurons, and

is modulated by minicolumns. They consist of self-

contained neuronal aggregations and their afferent,

efferent and interneuronal connections, mediating the

interactions of neuronal microcircuits, and are patho-

logically altered in ASD (Casanova 2007; Opris and

Casanova 2014). In ASD there is activation of microglia,

patches of disorganised cortex, focal cortical dysplasia

(smaller pyramidal neurones, reduced numbers of inter-

neurons), and altered radial migration of neurones

resulting in heterotopias, i.e. normal cells in an abnor-

mal location (Casanova et al. 2013). Subependymal and

subcortical heterotopias may cause seizures, develop-

mental delay, dyslexia, obsessive–compulsive disorder

(OCD) and ASD. ASD is proposed to be a minicolumno-

pathy with increased numbers of minicolumns and

decreased minicolumn width (Casanova 2006; Casanova

et al. 2002, 2003a, 2003b, 2010). The sheath around the

minicolumns provides inhibitory control and may be

decreased due to dense packing (Casanova 2008). These

alterations cause inhibitory deficit, sensory noise over-

load and lack of stimulus discrimination evidenced by

evoked gamma EEG power at frontal cortical sites.

Importantly, after repetitive TMS, ASD individuals show

significant improvement in cortical discrimination of

stimuli (Sokhadze et al. 2009).

The E-I ratio of synaptic inputs in developing and

adult cortical neurones is dynamically regulated (Zhang

et al. 2011). One hypothesis underscores the importance

of altered connectivity of excitatory and inhibitory

cortical circuits in ASD (Zikopoulos and Barbas 2013).

E-I imbalance occurs due to altered neuronal migration,

development and network formation. This may result

from lack of the extracellular matrix protein reelin

supporting GABAergic neuronal development (Folsom

and Fatemi 2013).

Brain synapses may be modified by immune proteins

such as Major Histocompatibility Class I (MHCI) antigens

that are present on axons and dendrites of cortical

neurons during early development, negatively regulate

cortical synapse density, establishment of neuronal

connections, strength of excitatory but not inhibitory

synapses, and control the E-I balance onto cortical

neurons (Glynn et al. 2011). Through MHCI the immune

system may regulate the E-I balance in ASD. The immune

system may modulate E-I balance through proinflamma-

tory cytokines such as IL-1b, IL-6 and TNFa which play

roles in synaptic plasticity (Beattie et al. 2002), trafficking

of AMPA (Beattie et al. 2010), GABAA receptors and

learning (Miller and Fahey 1994; Lai et al. 2006;

Boulanger 2009; Besedovsky and del Rey 2011; del Rey

et al. 2013; Wei et al. 2011, 2012). Mice with elevated

brain IL-6 display autistic features and E-I imbalance.

TNFa modulates AMPA receptor trafficking in neurons

and glia (Beattie et al. 2010; He et al. 2012).

The E-I ratio depends on levels of ionotropic and

metabotropic glutamate, GABA receptors (Banerjee et al.

2013; Billingslea et al. 2014), extracellular glutamate and

GABA concentrations (Bejjani et al. 2012), oxytocin,

reelin, Arc, neuronal translation (Gkogas et al. 2013;

Gkogas and Sonenberg 2013), signal transduction, cell

adhesion molecules and PSD proteins such as neuroli-

gins (Bourgeron 2007; Kohl et al. 2013). Neuroendocrine

factors such as oestrogens may cause E-I imbalance,

evidenced in reeler mice (Macri et al. 2010).

E-I imbalance in brain may give rise to:

(a) altered synaptic plasticity, learning and memory

discussed in Sections II.2 (Bateup et al. 2013) and

III.2.b (Oberman et al. 2010);

(b) seizures (Moavero et al. 2010) due to increase in

glutamatergic or decrease in GABAergic neuronal

activity, frequent in syndromic ASD such as TSC, RS

and FXS;

(c) neural network oscillatory abnormalities such as

abnormal gamma oscillations (Grice et al. 2001;

Orekhova et al. 2008; Gandal et al. 2010). Gamma

2 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

oscillations are instrumental for the synchronisation

of neuronal discharges in cortical networks and

sensory processing. Research suggests that they

may be linked to cognitive functions such as

selective attention, short- and long-term memory

and multisensory integration;

(d) visual system abnormalities such as atypical visual

perception (sensitivity to bright light, atypical face

perception), attributed to E-I imbalance in occipital,

parietal and frontal cortex (Milne et al. 2009).

Gamma oscillations (30–90 Hz) depend on E-I

balance and can be recorded using EEG.

Measurement of the gamma response to contextual

modulation of visual stimuli in control and ASD

subjects has shown that controls exhibit a gamma

response in the 60-Hz range influenced by context-

ual modulation. ASD subjects show a smaller

gamma response not influenced by contextual

modulation (Snijders et al. 2013). These findings

indicate that E-I imbalance in ASD is due to

GABAergic deficit, extending earlier studies showing

atypical gamma response to visual stimuli in ASD

(Grice et al. 2001; Sun et al. 2012; Wright et al. 2012).

Gamma oscillations may be a biomarker to stratify

patients and follow the responses to therapy.

Another measurement of E-I imbalance is binocular

rivalry in which perception reflects the contrasting

effects of excitation and inhibition in cortex (Laing and

Chow 2002; Said et al. 2013). Adult high-functioning ASD

subjects show a slower rate of binocular rivalry than

controls, longer duration of mixed percepts and

increased likelihood to revert to the previously perceived

object when exiting a mixed percept (Robertson et al.

2013). The atypical dynamics of binocular rivalry in ASD

predict the autistic symptoms measured with the Autism

Diagnostic and Observation Schedule (ADOS), suggest-

ing that binocular rivalry may be a clinical biomarker.

Another study found normal binocular rivalry in ASD

subjects in comparison with controls and no corres-

pondence between binocular rivalry and ASD severity

(Said et al. 2013), indicating minimal E-I imbalance in

visual system.

Children with high-functioning ASD exhibit early-

stage visual processing abnormalities manifested by

augmented evoked potentials elicited by task-irrelevant

stimuli which disrupt stimulus discrimination (Baruth

et al. 2010a). These abnormalities result from E-I imbal-

ance within parietal-occipital and frontal cortical regions

and can be reversed using TMS.

(e) general dyspraxia (impaired performance of skilled

gestures) due to abnormalities in frontal-parietal-

subcortical neural circuits important for learning of

sensory representations of movement and/or the

motor sequences necessary to execute them

(Mostofsky et al. 2006; Dowell et al. 2009; MacNeil

and Mostofsky 2012). This reflects an E-I imbalance in

neural circuits within parietal, premotor and motor

cortices. Praxis is associated with social, communica-

tive and behavioural impairments suggesting that

dyspraxia may be a biomarker of neurological

abnormalities in ASD (Dziuk et al. 2007);

(f) behavioural changes and social dysfunction such as

repetitive and disruptive behaviours, irritability,

social avoidance/withdrawal.

Notably, E-I imbalance is not ASD-specific and is found

in other related neurodevelopmental conditions such as

attention deficit hyperactivity disorder (ADHD) and OCD

(Zimmermann et al. 2014). The specific changes and

brain localisation in the E-I imbalance are likely charac-

teristic for each disorder.

2. Syndromic forms of ASD with marked E-I

imbalance

TSC is a genetic disorder due to inactivating mutations

in the TSC1 or 2 genes. It is associated with learning

abnormalities, intellectual disabilities, developmental

delay, autism, tubers, and epilepsy (Jeste et al. 2008,

2014). The pathological processes result from loss of the

TSC1 or TSC2 proteins regulating the mammalian target

of rapamycin (mTOR) pathway: abnormal protein syn-

thesis, synaptic plasticity, reduced neuronal connectivity

and CNS myelination, and E-I imbalance (Han and Sahin

2011; Tsai and Sahin 2011; Peters et al. 2012; Julich and

Sahin 2014). Loss of Tsc1 in mice leads to upregulation of

mTOR, abnormal protein synthesis and enhanced exci-

tatory synaptic function that alter circuit information

processing (Bateup et al. 2011). Loss of Tsc1 in mice is

accompanied with weakened inhibition resulting from

insensitivity of the glutamatergic neurons to signals

from GABAergic neurons due to mTOR deficiency

(Bateup et al. 2013), suggesting that drugs enhancing

GABA functions may not be effective in TSC. Tsc2

heterozygous knockout mice show deficiency in CA1

hippocampal long-term potentiation (LTP) as result of

mTOR dysregulation and learning and memory deficits

that are corrected using the mTOR inhibitor rapamycin

(Ehninger et al. 2008, 2009). TSC1/2 are important for

axon specification, guidance and regeneration, and have

effects on neuronal connectivity and E-I balance (Choi

et al. 2008; Tsai and Sahin 2011). The deficits in mTOR,

GABAergic neurotransmission and E-I imbalance in TSC

may be overcome by the neuropeptide oxytocin or

mTOR inhibitors (Julich and Sahin 2014).

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 3

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

FXS is a genetic disorder associated with ASD

affecting primarily boys. It results from loss-of-function

of the X-linked gene FMR1 encoding the RNA-binding

protein FMRP (Ronesi and Huber 2008). Associated

symptoms are anxiety, sensory abnormalities, hyper-

activity, irritability, seizures (Tranfaglia 2012). The E-I

imbalance changes in FXS are brain region-specific.

Major mechanism is enhanced synaptic protein synthesis

due to overactivation of group I metabotropic glutamate

receptors (gpI mGluRs; Dolen and Bear 2008; D’Antoni

et al. 2014). The functions of the affected synaptic

proteins are abnormal leading to changes in excitation

and inhibition (Ronesi et al. 2012). Proteins with

enhanced translation are PSD-95, Arc, GluA1 and GluA2

receptors (Muddashetty et al. 2007), microtubule-asso-

ciated protein 1B, striatal-enriched protein tyrosine

phosphatase, calcium-calmodulin kinase II (Luscher and

Huber 2010). There are alterations in synaptic plasticity

(Huber et al. 2002), activity-dependent formation and

refinement of neuronal circuits (Paluskiewitz et al. 2011;

Doll and Broadie 2014) and learning. There is enhanced

gpI mGluR-LTD expressed with AMPA receptor internal-

isation. Besides gpI mGluRs, M1 muscarinic acetylcholine

Gq-coupled receptors trigger LTD expressed with AMPA

receptor internalisation which is enhanced in FXS (Volk

et al. 2007). There are GABAergic deficits in cortex,

hippocampus, amygdala, striatum, and subiculum,

decreased GABAA receptors and alterations in GABA

production and metabolism in the Fmr1 KO mouse

(Paluskiewitz et al. 2011). Human ASD post-mortem

brain studies report decreased GABAA and GABAB

receptors in anterior cingulate cortex (Oblak et al.

2009, 2010). There is hyperexcitability in layer 4 of

somatosensory neocortex of the Fmr1 KO mouse (Gibson

et al. 2008). The expression of NMDA receptors is

increased in PFC and hippocampus of Fmr1 KO mice

(Krueger et al. 2011; Uzunova et al. 2014). These findings

provided basis for development of drugs decreasing gpI

mGluR activity such as MTEP, MPEP, CTEP (Michalon

et al. 2012), AFQ056 (Pop et al. 2014; Jacquemont et al.

2011), increasing GABAergic functions such as

Arbaclofen (Berry-Kravis et al. 2012), and decreasing

NMDA receptor functions such as memantine. They aim

to restore the E-I imbalance and target the core

symptoms of ASD. Clinical trials with these drugs have

yielded mixed results. This may be due to the hetero-

geneity of E-I imbalance and lack of appropriate

outcome measures.

RS is due to loss-of-function mutations in the X-linked

gene for the methyl-CpG binding protein 2 (MeCP2), a

transcriptional activator and repressor. RS is associated

with autism, regression of language, cognitive functions,

social and motor skills, stereotypies, seizures, breathing

difficulties, affecting primarily girls. Similarly to TSC and

FXS, RS is a disorder of synaptic protein synthesis. There is

disrupted development and functions of neuronal circuits

and E-I imbalance. Using transgenic mice it is established

that MeCP2 is critical for normal function of GABA-

releasing neurons (Chao et al. 2010). Human RS post-

mortem brain studies show abnormal minicolumns

(Casanova et al. 2003a, 2003b). Experimental therapies

aiming to restore the E-I imbalance are Insulin-like growth

factor 1 (Tropea et al. 2009) for correcting the synaptic

plasticity deficits and the ampakine CX546 (Kwaja and

Sahin 2011) for correcting the protein synthesis deficits

and increasing Brain-derived neurotrophic factor.

III. Methods to interrogate E-I imbalance inautistic brain

1. Preclinical (animal models)

(a) Optogenetics allows manipulating the activity of

neurons and neuronal networks (Fenno et al. 2011).

It is an invasive method in which light-sensitive ion

channels (opsins) are introduced in neurons by viral

(AAV or lentiviral) transfection or genetic means.

Using light the neurons are activated or inhibited

with temporal and spatial precision and the effects

of these manipulations on behaviour of living

animals is observed (Tye and Deisseroth 2012;

Yizhar 2012). Optogenetic manipulation of principal

excitatory or inhibitory parvalbumin-positive inter-

neurons in the mouse medial prefrontal cortex

(mPFC) shows that elevated (by activating the

principal neurons) and not reduced (by activating

the inhibitory neurons) E-I balance in the mPFC

impairs social behaviour and conditioning (Yizhar

et al. 2011). The elevation of the E-I ratio is

accompanied by elevated baseline gamma oscilla-

tions, a clinical ASD biomarker (Sohal 2012).

The social deficits can be reduced by increasing

the interneuron inhibitory tone. This study provides

support for E-I imbalance in ASD and indicates that

the changes are reversible.

2. Clinical

(a) EEG and MEG may be used to determine E-I

imbalance in ASD alone, or together with TMS,

and provide useful biomarkers and outcome meas-

ures. Characteristic EEG changes manifesting as

excess oscillations in the high frequency range are

found in ASD children (Cantor et al. 1986; Orekhova

et al. 2007). The EEG findings alone in ASD can be

non-specific and resemble those of toddlers indicat-

ing maturational lag (Cantor et al. 1986).

4 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

MEG studies find abnormal oscillations in ASD children

and adolescents suggesting E-I imbalance (Wilson et al.

2007; Cornew et al. 2012). Children and adolescents with

ASD exhibit reduced left hemispheric gamma oscilla-

tions (Wilson et al. 2007). ASD children exhibit elevations

in delta (1–4 Hz), theta (4–8 Hz), alpha (8–12 Hz) and

high frequency (20–120 Hz) power (Cornew et al. 2012).

The increased alpha power in temporal and parietal

regions correlates with the severity of ASD symptoms

and is of particular interest because inhibitory inter-

neurons play role in its maintenance. MEG studies report

that gamma activation underlying face processing is

abnormal in ASD (Wright et al. 2012). This is likely due to

E-I imbalance in occipital regions. MEG oscillations may

depend on medications (SSRIs, methylphenidate), state of

arousal, anxiety, eyes open or closed condition and high-

or low-functioning ASD. Nevertheless, the findings of

abnormal brain activity using MEG and EEG partially

correspond and may have clinical significance for

biomarkers.

(b) TMS is a technique for non-invasive brain stimula-

tion (NIBS) that can be applied to study and

modulate cortical excitability and synaptic plasticity

in vivo in humans (Pascual-Leone et al. 2011; Vicario

and Nitsche 2013). TMS consists of application of

rapidly changing magnetic fields to discrete brain

regions using a coil positioned above the targeted

region. The magnetic fields penetrate into the brain

and induce electric currents depolarising groups of

neurons. If applied to motor cortex, the depolarisa-

tion can produce contralateral muscle contraction,

motor evoked potential (MEP), measured using

electromyography (EMG). If applied to non-motor

cortical regions, TMS may evoke field potentials

which are recorded using EEG. In addition to effects

in the targeted neurons, TMS may produce transsy-

naptic distributed network effects that can be

detected using neuroimaging. TMS may be com-

bined with EEG or functional magnetic resonance

imaging (fMRI; Oberman et al. 2010). TMS is

particularly useful because it detects changes in

synaptic plasticity (Oberman et al. 2013). TMS is

subdivided into single-pulse (spTMS), paired-pulse

(ppTMS) or repetitive (rTMS). SpTMS is a measure of

cortical excitability when applied together with EMG

or EEG. ppTMS can be used to study the E-I ratio

(Kobayashi and Pascual-Leone 2003) whereas theta

burst stimulation (TBS) may be used to study LTP,

LTD and metaplasticity (Huang et al. 2005). TBS can

measure synaptic plasticity in cortex determined by

GABA and glutamate receptors. Using ppTMS it is

possible to measure intracortical inhibition

mediated by GABA(A) or GABA(B) receptors

(Oberman et al. 2010). TMS studies in ASD and FXS

subjects and controls show that ASD patients have

increased cortical excitability and that ASD and FXS

patients have altered synaptic plasticity. The find-

ings of hyperplasticity in ASD are consistent

irrespective of the aetiology, suggesting that this

may be a biomarker (Oberman et al. 2010). The

findings of altered synaptic plasticity in ASD and FXS

patients are consistent with animal models.

Combination of TMS and EEG shows that in ASD

there are robust patterns of over- and under

connectivity in frontal and temporal cortical regions

in the eyes closed resting state (Murias et al. 2007).

In the theta frequency range (3–6 Hz) there is locally

elevated coherence in frontal and temporal cortex

within the left hemisphere, and in the lower alpha

range (8–10 Hz) there is reduced coherence within

frontal regions in adults with ASD (Murias et al.

2007). The TMS studies need to be expanded with

different brain areas and larger patient populations.

TMS may be used to assess physiological biomarkers

of E-I imbalance in ASD. It is possible to use

common measures of E-I imbalance such as som-

atosensory evoked potentials (SEP) at baseline to

stratify patients and understand responses to ther-

apy. TMS may be used in ASD to induce acute and

long-lasting changes in brain cortical excitability and

functions depending on the protocol. Examples of

use of TMS in ASD are discussed below.

Induction of paired association (PAS) using TMS in

high-functioning ASD patients is impaired implying

presence an LTP-like deficit (Jung et al. 2013). PAS

consists of peripheral electrical stimulation paired with

TMS applied to the contralateral motor cortex and

depends on intact sensorimotor integration. The short

intracortical interval inhibition (SICI) dependent on

GABAergic transmission was normal and did not correl-

ate with the PAS effect. These findings suggest that in

patients with high-functioning ASD there is reduced

excitatory synaptic connectivity and deficits in sensory-

motor integration (Jung et al. 2013). The motor learning

deficits are likely due to decreased connectivity within

basal ganglia, thalamus, primary motor cortex (M1) and

SMA (Mostofsky et al. 2009).

Low-frequency rTMS has been applied to modulate

the abnormal gamma oscillations resulting from sen-

sory-perceptual deficits in ASD (Baruth et al. 2010b). ASD

individuals show augmented gamma band activity

indiscriminative of stimulus type during the early

stages of visual processing, a consequence of cortical

E-I imbalance. Twelve sessions of bilateral slow rTMS to

the DLPFC in ASD individuals significantly improved the

discriminatory gamma band activity between relevant

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 5

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

and irrelevant stimuli. This was associated with improve-

ment on the responses in the behavioural questionnaires

with respect to irritability and repetitive behaviour.

Therefore, it is possible to use EEG measurement of

stimulus discrimination in ASD as a physiological

biomarker of E-I imbalance before and after therapy.

This study shows that rTMS can be useful to treat the

core symptoms of ASD without major side effects.

In another study low frequency rTMS is applied to the

bilateral DLPFC in children with ASD to determine if this

will improve the error monitoring and correction deficit

(Sokhadze et al. 2012). There was beneficial effect of

rTMS and the study showed that the ERPs associated

with responses to errors and behavioural performance

measures may be used as functional outcome measures.

The brain regions affected by E-I imbalance in human

ASD determined in the studies discussed above are

indicated in Figure 1.

IV. Interventions for restoration of E-I balancein ASD

1. Pharmacological therapeutics

(a) Glutamatergic drugs. Group I mGlu (mGlu5) antag-

onists have not proven very effective in human ASD.

Ampakines (enhancing the activity of AMPA recep-

tors) have shown partial effectiveness in FXS.

Memantine, NMDA receptor antagonist, has been

tested in ASD treatment trials (Ghaleiha et al. 2013;

Uzunova et al. 2014). A multi-site double-blind clinical

trial with memantine tested the safety, tolerability and

efficacy in a paediatric ASD patient group

(NCT01592747). In the memantine study, similarly to

clinical trials with mGlu5 antagonists, results may be

different if patients were stratified on the basis of E-I

imbalance. This may be done using gamma oscillations

or SEPs as biomarkers. Biomarkers can identify patients

with greater E-I imbalance and select a subgroup likely

to have a favourable treatment response. Gamma

oscillations and SEPs can also be used as outcome

measures assessing the effects of therapy. Lack of

patient stratification may explain why single-site mem-

antine studies in which the patient population is

homogenous have shown positive results, not confirmed

in ASD multi-site studies.

(b) GABA agonists. The drug AZD7325 (AstraZeneca) is

a selective modulator of the GABAA2/3 receptor, is

well-tolerated and without tendency to induce cog-

nitive deficits. AZD7325 is tested in a Fast-Fail Trial

(FAST) as ASD treatment by restoring the E-I imbal-

ance. An EEG biomarker is used as patient selection

tool and means to assess the drug action, and side

effects, attention and learning are measured.

Arbaclofen, a GABA(B) agonist, has shown promise in

treatment of FXS and ASD (Berry-Kravis et al. 2012;

Henderson et al. 2012). A clinical trial with STX209

(Arbaclofen) on the neurobehavioral functions of children

and adults with FXS has not shown difference from

Figure 1. Brain regions affected by E-I imbalance in ASD targeted for therapeutic purposes using TMS. (A) Lateral view of the humanbrain. Areas targeted by TMS are: (1) dorsolateral prefrontal cortex (DLPFC) important for social cognition and cognition in general; (2)inferior frontal gyrus (IFG), important for language and verbal comprehension; (3) supplementary motor area (SMA), important forcontrol of movements and coordinating the temporal sequence of actions; (4) premotor cortex (Prem Ctx) and 5) primary motorcortex (Motor Ctx). Other regions such as parietal, temporal, occipital cortex (Occ Ctx) and cerebellum (Crb) can also be affected by E-Iimbalance in ASD. It is possible to target them with TMS in the future as this technique is further developed. Among them, thecerebellum is of particular interest. (B) Medial view of the human brain. Areas affected by E-I imbalance currently targeted by TMS areindicated: (1) dorsomedial prefrontal cortex (DMPFC) that contains mirror neurones important for social interactions; (2) the anteriorcingulate cortex (ACC) involved in empathy, emotion and impulse control; and (3) the SMA. Other regions that are affected by E-Iimbalance in ASD are cerebellum, thalamus (Th), basal ganglia, hippocampus and amygdala.

6 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

placebo on the primary endpoint measure – ABC-I

(Aberrant Behaviour Checklist-Irritability) (Berry-Kravis

et al. 2012). However, analysis using ABC – Social

Avoidance Scale showed significant beneficial treatment

effect. Subsequent study of STX209 in ASD subjects ages

5–21 (ClinicalTrials.gov Identifier: NCT01706523) was

terminated due to absence of effects on the primary

outcome measure – ABC. Future studies will benefit from

patient stratification.

(c) mTOR inhibitors. Rapamycin and its analogue

everolimus used for management of the tumours

and epilepsy in TSC (Franz 2011, Franz et al. 2013)

may be beneficial for ameliorating neurocognitive

manifestations in TSC and ASD (de Vries 2010),

based on the hypothesis that mTOR overactivation

is responsible for neurocognitive symptoms.

Rapamycin has improved measures of learning and

memory (immediate and delayed recall) by 20%

from baseline in 5 of 8 TSC subjects (de Vries 2010).

(d) Oxytocin (OT) has improved ASD symptoms in

clinical trials (Soorya et al. 2008; Green and

Hollander 2010; Anagnostou et al. 2012). OT has

advantages over mTOR inhibitors such as lower

toxicity, lack of immunosuppression, effects on the

phosphatidylinositol 3-kinase (PI3K) pathway and

enhancement of GABAergic neurotransmission. In

humans OT modulates a variety of behaviours

relevant to ASD including anxiety, stress, social

memory, recognition, bonding (affiliation), sexual

and aggressive behaviours. In CA1 neurons OT

stimulates protein synthesis and enhances LTP (Lin

et al. 2012). There is relationship between OT and

GABAergic neurotransmission in brain. GABAergic

abnormalities are established in ASD (Yip et al. 2007;

Fu et al. 2012; Mori et al. 2013). OT activates

hippocampal interneurons (Ogier et al. 2008) and

increases inhibitory GABAergic synapses in brain

(Theodosis et al. 2006). OT receptor knockout mice

exhibit reduced hippocampal GABAergic synapses,

increased ratio glutamatergic/GABAergic synapses,

increased seizure susceptibility, impaired social

behaviour, increased aggression, reduced cognitive

flexibility, and are a neurobehavioral model of ASD

(Sala et al. 2011). Intracerebral administration of OT

and vasopressin rescues these deficits.

(e) Implication of E-I imbalance for development of

pharmacologic drugs for treatment of ASD. It is

difficult to predict before treatment which patients

will respond favourably. To predict the treatment

response it is possible to administer a test drug dose

and monitor for early efficacy biomarker effect to

demonstrate target engagement of the drug. If

there is positive effect, the drug can be administered

for a longer period of time. If there is no effect on

early efficacy biomarker or if there are side effects,

the drug will not be administered.

2. Clinical somatic treatments

TMS can be applied as a therapeutic modality in ASD to

restore the imbalance of excitation and inhibition,

modulate synaptic plasticity and gamma oscillations. In

future it will be possible to target additional brain

regions affected by E-I imbalance such as cerebellum

(Demirtas-Tatlidede et al. 2013). Pharmacological, behav-

ioural or neuromodulation treatments that reverse the

altered E-I balance may be developed. Various TMS

protocols in terms of location of stimulation (i.e. dorso-

lateral prefrontal cortex, DLPFC, right operculus, supple-

mentary motor area, SMA), intensity of stimulation (high

frequency to activate – 20 Hz, and low frequency to

inhibit – 1 Hz) have been tested in ASD clinical trials. With

respect to the optimal brain area to target, no consensus

is reached. In summary, the results of TMS in ASD are

promising but inconclusive. Knowledge of the underlying

mechanisms will allow clinicians to design TMS protocols

so that the pathological changes in excitation, inhibition

and neural connectivity are reversed. TMS may be

preferred in patients who cannot take medications, are

not responsive to therapies, and may be combined with

medications, neurofeedback or behavioural challenges.

Another NIBS technique that has potential for

inducing changes in neuronal activity and management

of ASD is transcranial direct current stimulation (tDCS;

D’Urso et al. 2015). tDCS is still under development and

not FDA-approved method.

Evidence supports the gut–brain connection and roles

of inflammation and immunity ASD (Hsiao et al. 2013).

This is relevant to E-I imbalance because proinflamma-

tory cytokines released during gut–brain pathological

processes may affect trafficking of AMPA, GABA recep-

tors and synaptic plasticity. Based on this knowledge

novel treatments for ASD such as Trichuris suis ova

are currently tested by Dr. Eric Hollander’s team at

Montefiore Medical Centre in a clinical trial with

promising effects on repetitive and disruptive behav-

iours in ASD (Hollander et al. 2013).

3. Early detection using event-related potentials

(ERPs)

Clinical studies in the USA (Luyster et al. 2011) and UK

(Elsabbagh et al. 2012) evaluate early detection of ASD

by measurement of neural correlates of E-I imbalance in

infant brain as at 6–12 months of age using sensor nets

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 7

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

to detect ERPs. ERPs are changes in brain electrical

activity resulting from simultaneous activation of a

group of neurons that can be time-locked to a stimulus.

They are often recorded in response to a face-processing

stimulus and are useful to test in infants because they do

not require verbal abilities. The ERP components N290

and P400 (occipito-temporal region components) are

precursors of the adult N170 component, a marker of

detection of faces. The Nc component (negative central,

observed over frontal regions) is a marker of attention.

The group of Luyster examined the N290, P400 and Nc

components in response to familiar and unfamiliar faces

in 12-month-old infants at high vs. low ASD risk. They

found that the N290 and P400 did not differ among the

high- and low-risk infants. However, there were differ-

ences in the Nc responses among groups. The group of

Elsabbagh determined in infants 6–10 months of age the

ERP components P1, N290 and P400 as measures of the

neural responses to dynamic eye gaze. The infant’s

sensitivity to eye gaze is a precursor to social and

communicative skills. The researchers hypothesised that

infants at risk for ASD have abnormal neural responses

to eye gaze correlating with the behavioural

impairments observed later and can be used for early

diagnosis. These studies found that face-related ERP

distinguish infants at risk for ASD from the controls and

that atypical brain function precedes the onset of

behavioural signs and overt ASD symptoms. Other

studies show that toddlers at risk for ASD, infants

whose siblings have ASD and whose siblings do not

have ASD exhibit variability in the early trajectories of

development measured using the ADOS and the Autism

Diagnostic Interview – Revised (ADI-R), suggesting that

there is need for early identification, regular monitoring

and standardised assessments of young children sus-

pected of having ASD (Lord et al. 2012). Studies on

development of toddlers with early and later diagnosis

of ASD show that ASD may involve developmental

arrest, slowing or even regression (Landa et al. 2007). By

early detection (3–6 months) it is possible to achieve

better therapeutic results. For early ASD management it

is feasible to apply a neurofeedback approach to retrain

the E-I imbalance (Pineda et al. 2012). The various

mechanism and clinical diagnostic and treatment

aspects of E-I imbalance in ASD discussed are schemat-

ically presented in Figure 2.

Figure 2. Mechanisms and clinical aspects of E-I imbalance in ASD. The figure summarises the various aspects of E-I imbalance asdiscussed in the review. Central to ASD is E-I imbalance in key neural circuits in brain. In the lower panel are shown in the left box themajor mechanisms currently thought to lead to E-I imbalance (Causative Mechanisms). In turn, they may give rise to variousfunctional changes shown in the lower right box such as altered synaptic plasticity, learning and memory, seizures, etc. (FunctionalConsequences). These functional changes can measured for research and diagnostic purposes with different techniques shown in thetop left box such as optogenetics and clinical methods such as EEG, MEG and NIBS (Methods to Interrogate). These latter methodscan also be suitable ASD biomarkers and outcome measures. Based on the knowledge from the three components (causativemechanisms, functional consequences and methods to interrogate E-I imbalance) new therapeutics may be developed that aresummarised in the top right box (Interventions for Reversal).

8 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

V. Summary, conclusions and future directions

Can a single model of E-I imbalance consistent with the

various studies be proposed in ASD? ASD are very

heterogeneous (Hollander et al. 2011) and there is a

need to stratify patients based on clinical features and

biomarkers. Measurement of E-I imbalance is influenced

by the experimental method, subtype of ASD (e.g. FXS,

TSC, high-functioning vs. low-functioning) and develop-

mental period. We propose that E-I imbalance is a ‘‘final

common pathway’’ in ASD but the different patient

subgroups may have different mechanisms of E-I imbal-

ance and respond to different treatments.

Acknowledgements

None

Statement of interest

Dr Hollander has received research grants from Roche, Forest,Coronado, and Brainsway, and consulted to Roche, andCoronado.

References

Anagnostou E, Soorya L, Chaplin W, Bartz J, Halpern D,Wasserman S, Hollander E. 2012. Intranasal oxytocin versusplacebo in the treatment of adults with autism spectrumdisorders: a randomized controlled trial. Mol Autism. 3:16.doi: 10.1186/2040-2392-3-16

Banerjee A, Garcia-Oscos F, Roychowdhury S, Galindo LC, Hall S,Kilgard MP, et al. 2013. Impairment of cortical GABAergicsynaptic transmission in an environmental rat model ofautism. Int J Neuropsychopharmacol. 16:1309–1318.

Baroncelli L, Braschi C, Spolidoro M, Begenisic T, Maffei L,Sale A. 2011. Brain plasticity and disease: a matter ofinhibition. Neural Plast. 286073. doi: 10.1155/2011/286073

Baruth JM, Casanova MF, Sears L, Sokhadze E. 2010a. Early-stage visual processing abnormalities in high-functioningautism spectrum disorder (ASD). Transl Neurosci. 1:177–187.

Baruth JM, Casanova MF, El-Baz A, Horrell T, Mathai G, Sears L,et al. 2010b. Low-Frequency Repetitive TranscranialMagnetic Stimulation (rTMS) Modulates Evoked-GammaFrequency Oscillations in Autism Spectrum Disorder (ASD).J Neurother. 14:179–194.

Bateup HS, Takasaki KT, Saulnier JL, Denefrio CL, Sabatini BL.2011. Loss of Tsc1 in vivo impairs hippocampal mGluR-LTDand increases excitatory synaptic function. J Neurosci.31:8862–8869.

Bateup HS, Denefrio CL, Johnson CA, Saulnier JL, Sabatini BL.2013. Excitatory/inhibitory synaptic imbalance leads tohippocampal hyperexcitability in mouse models of tuberoussclerosis. Neuron. 78:510–522.

Beattie EC, Stellwagen D, Morishita W, Bresnahan JC, Ha BK,von Zastrow M, et al. 2002. Control of synaptic strength byglial TNFalpha. Science. 295:2282–2285.

Beattie MS, Ferguson AR, Bresnahan JC. 2010. AMPA-receptortrafficking and injury-induced cell death. Eur J Neurosci.32:290–297.

Bejjani A, O’Neill J, Kim JA, Frew AJ, Yee VW, Ly R, et al. 2012.

Elevated glutamatergic compounds in pregenual anteriorcingulate in pediatric autism spectrum disorder demon-

strated by 1H MRS and 1H MRSI. PloS One. 7:e38786. doi:10.1371/journal.pone.0038786

Berry-Kravis EM, Hessl D, Rathmell B, Zarevich P, Cherubini M,

Walton-Bowen K, et al. 2012. Effects of STX209 (Arbaclofen)on neurobehavioral function in children and adults with

fragile X syndrome: a randomized, controlled, phase 2 trial.Sci Transl Med. 4:152ra127.

Besedovsky HO, del Rey A. 2011. Central and peripheralcytokines mediate immune-brain connectivity. NeurochemRes. 36:1–6.

Billingslea EN, Tatard-Leitman VN, Anguiano J, Jutzeler CR,Suh J, Saunders JA, et al. 2014. Parvalbumin cell ablation

of NMDA-R1 causes increased resting network excitabilitywith associated social and self-care deficits.

Neuropsychopharmacology39:1603–1613.Boulanger LM. 2009. Immune proteins in brain development

and synaptic plasticity. Neuron. 64:93–109.Bourgeron T. 2007. The possible interplay of synaptic and clock

genes in autism spectrum disorders. Cold Spring Harbor

Symp Quant Biol. 72:645–54. doi: 10.1101/sqb.2007.72.020Cantor DS, Thatcher RW, Hrybyk M, Kaye H. 1986.

Computerized EEG analyses of autistic children. J Autism

Dev Disord. 16:169–187.Casanova MF, Buxhoeveden DP, Brown C. 2002. Clinical and

macroscopic correlates of minicolumnar pathology inautism. J Child Neurol. 17:692–695.

Casanova MF, Buxhoeveden D, Switala A, Roy E. 2003a.

Rett syndrome as a minicolumnopathy. Clin Neuropathol.22:163–168.

Casanova MF, Buxhoeveden D, Gomez J. 2003b. Disruption inthe inhibitory architecture of the cell minicolumn: implica-

tions for autism. Neuroscientist. 9:496–507.Casanova MF. 2006. Neuropathological and genetic findings in

autism: the significance of a putative minicolumnopathy.

Neuroscientist. 12:435–441.Casanova MF. 2007. The neuropathology of autism. Brain

Pathol. 17:422–433.Casanova MF. 2008. The minicolumnopathy of autism: A link

between migraine and gastrointestinal symptoms. MedHypotheses. 70:73–80.

Casanova MF, El-Baz A, Vanbogaert A, Narahari P, Switala A.2010. A topographic study of minicolumnar core width by

lamina comparison between autistic subjects and controls:possible minicolumnar disruption due to an anatomical

element in-common to multiple laminae. Brain Pathol.20:451–458.

Casanova MF, El-Baz ES, Kamat SS, Dombroski BA, Khalifa F,Elnakib A, et al. 2013. Focal cortical dysplasias in autism

spectrum disorders. Acta Neuropathol Commun. 1:67. doi:10.1186/2051-5960-1-67

Chao HT, Chen H, Samaco RC, Xue M, Chahrour M, Yoo J, et al.2010. Dysfunction in GABA signalling mediates autism-likestereotypies and Rett syndrome phenotypes. Nature.

468:263–269.Choi YJ, Di Nardo A, Kramvis A, Meikle L, Kwiatkowski DJ, Sahin

M, et al. 2008. Tuberous sclerosis complex proteins controlaxon formation. Genes Dev. 22:2485–2495.

Cornew L, Roberts TP, Blaskey L, Edgar JC. 2012. Resting-stateoscillatory activity in autism spectrum disorders. J Autism

Dev Disord. 42:1884–1894.

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 9

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

Courchesne E, Pierce K. 2005. Why the frontal cortex in autismmight be talking only to itself: local over-connectivity butlong-distance disconnection. Curr Opin Neurobiol. 15:225–230.

D’Antoni S, Spatuzza M, Bonaccorso CM, Musumeci SA,Ciranna L, Nicoletti F, et al. 2014. Dysregulation of group Imetabotropic glutamate (mGlu) receptor signalling in dis-orders associated with Intellectual Disability and Autism.Neurosci Biobehav Rev. 46:228–241. doi: 10.1016/j.neubiorev.2014.02.003

Del Rey A, Balschun D, Wetsel W, Randolf A, Besedovsky HO.2013. A cytokine network involving brain-borne IL-1b, IL-1ra,IL-18, IL-6, and TNFa operates during long-term potentiationand learning. Brain Behav Immunol. 33:15–23.

Demirtas-Tatlidede A, Vahabzadeh-Hagh AM, Pascual-Leone A.2013. Can noninvasive brain stimulation enhance cogni-tion in neuropsychiatric disorders? Neuropharmacology.64:566–578.

De Vries PJ. 2010. Targeted treatments for cognitive andneurodevelopmental disorders in tuberous sclerosis com-plex. Neurotherapeutics. 7:275–282.

Dolen G, Bear MF. 2008. Fragile x syndrome and autism: fromdisease model to therapeutic targets. J Neurodev Disord.1:133–140.

Doll CA, Broadie K. 2014. Impaired activity-dependent neuralcircuit assembly and refinement in autism spectrum disordergenetic models. Front Cell Neurosci. 8:30. doi: 10.3389/fncel.2014.00030

Dowell LR, Mahone EM, Mostofsky SH. 2009. Associationsof postural knowledge and basic motor skill with dyspraxiain autism: implication for abnormalities in distributedconnectivity and motor learning. Neuropsychology.23:563–570.

D’Urso G, Bruzesse D, Ferrucci R, Priori A, Pascotto A, GalderisiS, et al. 2015. Transcranial direct current stimulation forhyperactivity and noncompliance in autistic disorder.World J Biol Psychiatry. 16:361–366. doi: 10.3109/15622975.2015.1014411

Dziuk MA, Gidley Larson JC, Apostu A, Mahone AM, Denkla MB,Mostofsky SH. 2007. Dyspraxia in autism: association withmotor, social, and communicative deficits. Dev Med ChildNeurol. 49:734–739.

Ecker C, Spooren W, Murphy DG. 2013. Translationalapproaches to the biology of Autism: false dawn or a newera? Mol Psychiatry. 18:435–442.

Ehninger D, Han S, Shilyanski C, Zhou Y, Li W, Kwiatkowski DJ,et al. 2008. Reversal of learning deficits in a Tsc2+/- mousemodel of tuberous sclerosis. Nat Med. 14:843–848.

Ehninger D, de Vries PJ, Silva AJ. 2009. From mTOR tocognition: molecular and cellular mechanisms of cognitiveimpairments in tuberous sclerosis. J Intellect Disabil Res.53:838–851.

Eichler SA, Meier JC. 2008. E-I balance in human diseases – frommolecules to networking. Front Mol Neurosci. 1:2. doi:10.3389/neuro.02.002

Elsabbagh M, Mercure E, Hudry K, Chandler S, Pasco G,Charman T, et al. 2012. Infant Neural Sensitivity to DynamicEye Gaze is Associated with Later Emerging Autism. CurrBiology. 22:338–342.

Fenno L, Yizhar O, Deisseroth K. 2011. The development andapplication of optogenetics. Annu Rev Neurosci. 34:389–412.

Folsom TD, Fatemi SH. 2013. The involvement of reelin inneurodevelopmental disorders. Neuropharmacology.68:122–135.

Franz DN. 2011. Everolimus: an mTOR inhibitor for thetreatment of tuberous sclerosis. Expert Rev AnticancerTher. 11:1181–1192. doi: 10.1586/era.11.93

Franz DN, Belousova E, Sparagana S, Bebin EM, Frost M,Kuerman R, et al. 2013. Efficacy and safety of everolimusfor subependymal giant cell astrocytomas associatedwith tuberous sclerosis complex (EXIST-1): a multicentre,randomised, placebo-controlled phase 3 trial. Lancet. 12;381:125–132. doi: 10.1016/S0140-6736(12)61134-9

Fu C, Cawthon B, Clinkscales W, Bruce A, Winzenburger P, EssKC. 2012. GABAergic interneuron development and functionis modulated by the Tsc1 gene. Cereb Cortex. 22:2111–2119.

Gandal MJ, Edgar JC, Erlichman RS, Mehta M, Roberts TP, SiegelSJ. 2010. Validating g oscillations and delayed auditoryresponses as translational biomarkers of autism. BiolPsychiatry. 68:1100–1106.

Ghaleiha A, Asabadi M, Mohammadi MR, Shahei M, Tabrizi M,Hadjiaghaee R, et al. 2013. Memantine as adjunctivetreatment to risperidone in children with autistic disorder:a randomized, double-blind, placebo-controlled trial. Int JNeuropsychopharmacol. 16:783–789.

Gibson JR, Bartley AF, Hays SA, Huber KM. 2008. Imbalance ofneocortical excitation and inhibition and altered UP statesreflect network hyperexcitability in the mouse model offragile X syndrome. J Neurophysiol. 100:2615–2626.

Glynn MW, Elmer BM, Garay PA, Liu XB, Needleman LA, ElSabeawy F, et al. 2011. MHCI negatively regulates synapsedensity during the establishment of cortical connections. NatNeurosci. 14:442–451.

Green JJ, Hollander E. 2010. Autism and oxytocin: newdevelopments in translational approaches to therapeutics.Neurotherapeutics. 7:250–257. doi: 10.1016/j.nurt.2010.05.006

Gkogas CG, Sonenberg N. 2013. Translational control andautism-like behaviors. Cell Logist. 3:e24551.

Gkogas CG, Khoutorsky A, Ran I, Rampakakis E, Nevarko T,Weatherill DB, et al. 2013. Autism-related deficits viadysregulated eIF4E-dependent translational control. Nature.493:371–377.

Grice SJ, Spratling MW, Karmiloff-Smith A, Halit H, Csibra G, deHaan M, et al. 2001. Disordered visual processing andoscillatory brain activity in autism and Williams syndrome.Neuroreport. 12:2697–2700.

Han JM, Sahin M. 2011. TSC1/TSC2 signaling in the CNS. FEBSLett. 585:973–980.

He P, Liu Q, Wu J, Shen W. 2012. Genetic deletion of TNFreceptor suppresses excitatory synaptic transmission viareducing AMPA receptor localization in cortical neurons.FASEB J. 26:334–345. doi: 10.1096/fj.11-192716

Henderson C, Wijetunje L, Kinoshita MN, Shumway M,Hammond RS, Postma FR, et al. 2012. Reversal of disease-related pathologies in the fragile x mouse model by selectiveactivation of GABAB receptors with arbaclofen. Sci TranslMed. 4:152ra128.

Hollander E, Kolevzon A, Coyle JT. 2011. Textbook of AutismSpectrum Disorders. Washington, DC. American PsychiatricPublishing, Inc., 603 pp.

Hollander E, Ferretti CJ, Taylor BP, Noone R, Kirsch J, Racine E.2013. Trichuris Suis Ova (TSO) as an Immune-InflammatoryTreatment for Repetitive Behaviors in Autism SpectrumDisorders (ASD). Poster presented at the American College ofNeuropsychopharmacology Annual Meeting, Dec., 2013;Florida.

10 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

Hsiao EY, McBride SW, Hsien S, Sharon G, Hyde ER, McCue T,et al. 2013. Microbiota modulate behavioral and physio-logical abnormalities associated with neurodevelopmentaldisorders. Cell. 155:1451–1463.

Huang YZ, Edwards MJ, Rounis E, Bhatia KP, Rothwell JC. 2005.Theta burst stimulation of the human motor cortex. Neuron.45:201–206.

Huber KM, Gallagher SM, Warren ST, Bear MF. 2002. Alteredsynaptic plasticity in a mouse model of fragile x mentalretardation. Proc Natl Acad Sci USA. 99:7746–7750.

Jacquemont S, Curie A, des Portes V, Torrioli MG, Berry-Kravis E,Hagerman RJ, et al. 2011. Epigenetic modification of theFMR1 gene in Fragile X syndrome is associated withdifferential response to the mGluR5 antagonist AFQ056. SciTransl Med. 3:64ra1. doi: 10.1126/scitranslmed.3001708

Jeste SS, Sahin M, Bolton P, Ploubidis JB, Humphrey A. 2008.Characterization of autism in young children with tuberoussclerosis complex. J Child Neurol. 23:520–525.

Jeste SS, Wu JY, Senturk D, Varcin K, Ko J, McCarthy B, et al.2014. Early developmental trajectories associated with ASDin infants with tuberous sclerosis complex. Neurology.83:160–168.

Julich K, Sahin M. 2014. Mechanism-based treatment intuberous sclerosis complex. Pediatr Neurol. 50:290–296.

Jung NH, Janzanik WG, Delvendahl I, Munchau A, Biscaldi M,Mainberger F, et al. 2013. Impaired induction of long-termpotentiation-like plasticity in patients with high-functioningautism and Asperger syndrome. Dev Med Child Neurol.55:83–89.

Kobayashi M, Pascual-Leone A. 2003. Transcranial magneticstimulation in neurology. Lancet Neurol. 2:145–156.

Kohl C, Riccio O, Grosse J, Zanoletti O, Fournier C, Schmidt MV.2013. Hippocampal neuroligin-2 overexpression leads toreduced aggression and inhibited novelty reactivity in rats.PLoS One. 8:e56871. doi: 10.1371/journal.pone.0056871

Krueger DD, Osterweil EK, Chen SP, Tye LD, Bear MF. 2011.Cognitive dysfunction and prefrontal synaptic abnormalitiesin a mouse model of Fragile X syndrome. Proc Natl Acad SciUSA. 108:2587–2592.

Kwaja OS, Sahin M. 2011. Translational research: Rett syn-drome and Tuberous sclerosis complex. Curr Opin Pediatr.23:633–639.

Lai AY, Swayze RD, El-Husseini A, Song C. 2006. Interleukin-1beta modulates AMPA receptor expression and phosphoryl-ation in hippocampal neurons. J Neuroimmunol. 175:97–106.

Laing CR, Chow CC. 2002. A spiking neuron model for binocularrivalry. J Comput Neurosci. 12:39–53.

Landa RJ, Holman KC, Garrett-Meyer E. 2007. Social andcommunication development in toddlers with earlyand later diagnosis of autism spectrum disorders. Arch GenPsychiatry. 64:853–864.

Lin YT, Huang CC, Hsu KS. 2012. Oxytocin promotes long-termpotentiation by enhancing epidermal growth factorreceptor-mediated local translation of protein kinase Mzfactor mediated translation protein MJ Neurosci. 32:15476–15488.

Lord C, Luyster R, Guthrie W, Pickles A. 2012. Patterns ofdevelopmental trajectories in toddlers with autism spec-trumdisorder. J Consult Clin Psychol. 80:477–489.

Luscher C, Huber KM. 2010. Group 1 mGluR-dependentsynaptic long-termdepression: mechanisms and implica-tions for circuitry and disease and for and Neuron.65:445–459.

Luyster RJ, Wagner JB, Vogel-Farley V, Tager-Flusberg H, Nelson

CA 3rd. 2011. Neural correlates of familiar and unfamiliar

face processing in infants at risk for autism spectrum

disorders. Brain Topogr. 24:220–228.MacNeil LK, Mostofsky SH. 2012. Specificity of dyspraxia in

children with autism. Neuropsychology. 26:165–171.Macri S, Biamonte F, Romano E, Marino R, Keller F, Laviola G.

2010. Perseverative responding and neuroanatomical

alterations in adult heterozygous reeler mice are

mitigated by neonatal estrogen administration.

Psychoneuroendocrinology. 35:1374–1387.Michalon A, Sidorov M, Ballard TM, Ozmen L, Spooren W,

Wettstein JG, et al. 2012. Chronic pharmacological mGlu5

inhibition corrects Fragile X in adult mice. Neuron. 74:49–56.Miller LG, Fahey JM. 1994. Interleukin-1 modulates GABAergic

and glutamatergic function in brain. Ann NY Acad Sci.

739:292–298.Milne E, Scope A, Pascalis O, Buckley D, Makeig S. 2009.

Independent component analysis reveals atypical electro-

encephalographic activity during visual perception in indi-

viduals with autism. Biol Psychiatry. 65:22–30.Moavero R, Cerminara C, Curatolo P. 2010. Epilepsy secondary

to tuberous sclerosis: lessons learned and current chal-

lenges. Childs Nerv Syst. 26:1495–1504.Mori T, Mori K, Fujii E, Toda Y, Miyazaki M, Harada M, et al. 2013.

Evaluation of the GABAergic nervous system in autistic brain:

(123)I-iomazenil SPECT study. Brain Dev. 34:648–654.Mostofsky SH, Dubey P, Jerath VK, Jansiewitz EM, Goldberg MC,

Denkla MB. 2006. Developmental dyspraxia is not limited to

imitation in children with autism spectrum disorders. J Int

Neuropsychol Soc. 12:314–326.Mostofsky SH, Powell SK, Simmonds DJ, Goldberg MC, Caffo B,

Pekar JJ. 2009. Decreased connectivity and cerebellar

activity in autism during motor task performance. Brain.

132:2413–2425.Muddashetty RS, Kelic S, Gross C, Xu M, Bassell GJ. 2007.

Dysregulated metabotropic glutamate receptor-dependent

translation of AMPA receptor and postsynaptic density-95

mRNAs at synapses in a mouse model of Fragile X syndrome.

J Neurosci. 27:5338–5348.Murias M, Webb SJ, Greenson J, Dawson G. 2007. Resting state

cortical connectivity reflected in EEG coherence in individ-

uals with autism. Biol Psychiatry. 62:270–273.Oberman L, Ifert-Miller F, Najib U, Bashir S, Woolacott I,

Gonzalez-Heydrich J, et al. 2010. Transcranial magnetic

stimulation provides means to assess cortical plasticity

and excitability in humans with fragile x syndrome and

autism spectrum disorder. Front Synaptic Neurosci. 2:26. doi:

10.3389/fnsyn.2010.00026Oberman LM, Rotenberg A, Pascual-Leone A. 2013. Use of

Transcranial Magnetic Stimulation in Autism Spectrum

Disorders. J Autism Dev Disord. doi:10.1007/s10803-013-

1960-2Oblak A, Gibbs TT, Blatt GJ. 2009. Decreased GABAA receptors

and benzodiazepine binding sites in the anterior cingulate

cortex in autism. Autism Res. 2:205–219.Oblak A, Gibbs TT, Blatt GJ. 2010. Decreased GABA(B) receptors

in the cingulate cortex and fusiform gyrus in autism.

J Neurochem. 114:1414–1423.Ogier R, Wrobel LJ, Raggenbass M. 2008. Action of tachykinins

in the hippocampus: facilitation of inhibitory drive to

GABAergic interneurons. Neuroscience. 156:527–536.

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 11

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

Opris I, Casanova MF. 2014. Prefrontal cortical minicolumn:from executive control to disrupted cognitive processing.Brain. 137:1863–1875.

Orekhova EV, Stroganova TA, Nygren G, Tsetlin MM, PosikeraIN, Gillberg C, et al. 2007. Excess of high frequencyelectroencephalogram oscillations in boys with autism. BiolPsychiatry. 62:1022–1029.

Orekhova EV, Stroganova TA, Prokofyev AO, Nygren G, GillbergC, Elam N. 2008. Sensory gating in young children withautism:relation to age, IQ and EEG gamma oscillations.Neurosci Lett. 434:218-23. doi: 10.1016/j.neulet.2008.01.066

Paluskiewitz SM, Martin BS, Huntsman MM. 2011. Fragile Xsyndrome: the GABAergic system and circuit dysfunction.Dev Neurosci. 33:349–364.

Pascual-Leone A, Freitas C, Oberman L, Horvath JC, Haiko M,Eldaief M, et al. 2011. Characterizing brain cortical plasticityand network dynamics across the age-span in health anddisease with TMS-EEG and TMS-fMRI. Brain Topogr. 24:302–315.

Peters JM, Sahin M, Vogel-Farley VK, Jeste SS, Nelson CA 3rd,Gregas MC, et al. 2012. Loss of white matter microstruc-tural integrity is associated with adverse neurologicaloutcome in tuberous sclerosis complex. Acad RadiolRadiol.19:17–25.

Peters JM, Taquet M, Vega C, Jeste SS, Fernandez IS, Tan J, et al.2013. Brain functional networks in syndromic and non-syndromic autism: a graph theoretical study of EEGconnectivity. BMC Med. 11:54. doi: 10.1186/1741-7015-11-54

Polleux F, Lauder JM. 2004. Toward a developmental neuro-biology of autism. Ment Retard Dev Disabil Res Rev. 10:303–317.

Pineda JA, Juavinett A, Datko M. 2012. Self-regulation of brainoscillations as a treatment for aberrant brain connections inchildren with autism. Med Hypotheses. 79:790–798.

Pop AS, Levenga J, de Esch CE, Buijsen RA, Niewvenhuisen IM,Li T, et al. 2014. Rescue of dendritic spine phenotype in Fmr1KO mice with the mGluR5 antagonist AFQ056/Mavoglurant.Psychopharmacology. 231:1227–1235.

Robertson CE, Kravitz DJ, Freyberg J, Baron-Cohen S, Baker CI.2013. Slower rate of binocular rivalry in autism. J Neurosci.33:16983–91. doi: 10.1523/JNEUROSCI.0448-13.2013

Ronesi JA, Huber KM. 2008. Metabotropic glutamate receptorsand fragile X mental retardation protein: partners in trans-lational regulation at the synapse. Sci Signal. 1:pe6.doi:10.1126/stke

Ronesi JA, Collins KA, Hays SA, Tsai NP, Guo W, Birnbaum SG,et al. 2012. Disrupted Homer scaffolds mediate abnormalmGluR5 function in a mouse model of Fragile X syndrome.Nat Neurosci. 15:431–440.

Rubenstein JLR, Merzenich MM. 2003. Model of autism:Increased ratio of excitation/inhibition in key neural systems.Genes Brain Behav. 2:255–267.

Said CP, Egan RD, Minshew NJ, Behrmann M, Heeger DJ. 2013.Normal binocular rivalry in autism: implications for theexcitation/inhibition imbalance hypothesis. Vision Res.77:59–66.

Sala M, Braida D, Lentini D, Busnelli M, Bulgheroni E, Capurro V,et al. 2011. Pharmacologic rescue of impaired cognitiveflexibility, social deficits, increased aggression, and seizuresusceptibility in oxytocin receptor null mice: a neurobeha-vioral model of autism. Biol Psychiatry. 69:875–882.

Snijders TM, Milivojevic B, Kemner C. 2013. Atypical excitation-inhibition balance in autism captured by the gamma

response to contextual modulation. Neuroimage Clin. 3:65-

72. doi: 10.1016/j.nicl.2013.06.015Sohal VS. 2012. Parvalbumin neurons and gamma rhythms

enhance cortical circuit performance. Biol Psychiatry.

71:1039-45. doi: 10.1016/j.biopsych.2012.01.024Sokhadze EM, El-Baz A, Baruth J, Mathai G, Sears L, Casanova

MF. 2009. Effects of low frequency repetitive transcranial

magnetic stimulation (rTMS) on gamma frequency oscilla-

tions and event-related potentials during processing of

illusory figures in autism. J Autism Dev Disord. 39:619–634.

doi: 10.1007/s10803-008-0662-7

Sokhadze EM, Baruth JM, Sears L, Sokhadze GE, El-Baz AS,

Casanova MF. 2012. Prefrontal neuromodulation using rTMS

improves error monitoring and correction function in autism.

Appl Psychophysiol Biofeedback. 37:91–102. doi: 10.1007/

s10484-012-9182-5Soorya L, Kiarashi J, Hollander E. 2008. Psychopharmacologic

interventions for repetitive behaviors in autism spectrum

disorders.in spectrum. Child Adolesc Psychiatr Clin N Am.

17:753–771.Sun L, Grutzner C, Bolte S, Wibral M, Tozman T, Schlitt S, et al.

2012. Impaired gamma-band activity during perceptual

organization in adults with autism spectrum disorders:

evidence for dysfunctional network activity in frontal-

posterior cortices. J Neurosci. 32:9563–9573.Theodosis DT, Koksma JJ, Trailin A, Langle SL, Piet R, Lodder JC,

et al. 2006. Oxytocin and estrogen promote rapid formation

of functional GABA synapses in the adult supraoptic nucleus.

Mol Cell Neurosci. 31:785–794.Tranfaglia MR. 2012. Fragile X Syndrome: a psychiatric per-

spective. Results Probl Cell Differ. 54:281–295.Tropea D, Giacometti E, Wilson NR, Beard C, McCurry C, Fu DD,

et al. 2009. Partial reversal of Rett Syndrome-like symptoms

in MeCP2 mutant mice. Proc Natl Acad Sci USA. 106:2029–

2034.Tsai P, Sahin M. 2011. Mechanisms of neurocognitive dysfunc-

tion and therapeutic considerations in tuberous sclerosis

complex. Curr Opin Neurol. 24:106–113. doi: 10.1097/

WCO.0b013e32834451c4Tye C, Bolton P. 2013. Neural connectivity abnormalities in

autism: insights from the Tuberous sclerosis model. BMC

Med. 11:55. doi: 10.1186/1741-7015-11-55Tye KM, Deisseroth K. 2012. Optogenetic investigation of

neural circuits underlying brain disease in animal models.

Nat Rev Neurosci. 13:251–266.Uzunova G, Hollander E, Shepherd J. 2014. The Role of

Ionotropic Glutamate Receptors in Childhood

Neurodevelopmental Disorders: Autism Spectrum Disorders

and Fragile X Syndrome. Curr Neuropharmacol. 12:71–98.Vicario CM, Nitsche MA. 2013. Non-invasive brain stimulation

for the treatment of brain diseases in childhood and

adolescence: state of the art, current limits and future

challenges. Front Syst Neurosci. 7:94. doi: 10.3389/

fnsys.2013.00094Volk LJ, Pfeiffer BE, Gibson JR, Huber KM. 2007. Multiple Gq-

coupled receptors converge on a common protein synthe-

sis-dependent long-term depression that is affected in

fragile X syndrome mental retardation. J Neurosci.

27:11624–11634.Wei H, Zou H, Sheikh AM, Malik M, Dobkin C, Brown WT, et al.

2011. IL-6 is increased in the cerebellum of autistic brain

and alters neural cell adhesion, migration and synaptic

12 G. UZUNOVA ET AL.

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015

formation. J Neuroinflammation. 8:52. doi: 10.1186/1742-2094-8-52

Wei H, Chadman KK, McCloskey DP, Sheikh AM, Malik M, BrownWT, et al. 2012. Brain IL-6 elevation causes neuronal circuitryimbalances and mediates autism-like behaviors. Biochim.Biophys. ActaBiophys. 1822:831–842.

Wilson TW, Rojas DC, Reite ML, Teale PD, Rojers SJ. 2007.Children and adolescents with autism exhibit reducedMEG steady-state gamma responses. Biol Psychiatry.62:192–197.

Wright B, Alderson-Day B, Prendergast G, Bennett S, Jordan J,Whitton C, et al. 2012. Gamma activation in youngpeople with autism spectrum disorders and typically-developing controls when viewing emotions on faces.PLoS One. 2012;7(7):e41326. doi: 10.1371/journal.pone.0041326

Yip J, Soghomonian JJ, Blatt GJ. 2007. Decreased GAD67mRNA levels in cerebellar Purkinje cells in autism: patho-physiological implications. Acta Neuropathol. 113:559–568.

Yizhar O, Fenno LE, Prigge M, Schneider F, Davidson TJ, O’SheaDJ, et al. 2011. Neocortical excitation/inhibition balance ininformation processing and social dysfunction. Nature.477:171–178.

Yizhar O. 2012. Optogenetic insights into social behaviorfunction.function. Biol Psychiatry. 71:1075–1080.

Zhang Z, Jiao YY, Sun QQ. 2011. Developmental maturationof excitation and inhibition balance in principal neuronsacross four layers of somatosensory cortex. Neuroscience.174:10–25.

Zikopoulos B, Barbas H. 2013. Altered neural connectivityin excitatory and inhibitory cortical circuits in autism.Front Hum Neurosci. 7:609. doi: 10.3389/fnhum.2013.00609

Zimmermann AM, Jene T, Wolf M, Gorlich A, Gurniack CB,Sassoe-Pognetto M, et al. 2014. Attention-deficit/hyperactiv-ity disorder-like phenotype in a mouse model with impairedactin dynamics. Biol Psychiatry. 78:95–106. doi: 10.1016/j.biopsych.2014.03.011

THE WORLD JOURNAL OF BIOLOGICAL PSYCHIATRY 13

Dow

nloa

ded

by [

New

Yor

k U

nive

rsity

] at

22:

01 0

4 N

ovem

ber

2015