Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988

Contents lists available at ScienceDirect Neuroscience and Biobehavioral Reviews

journal homepage: www.elsevier.com/locate/neubiorev

Review A novel NMDA receptor glycine-site partial agonist, GLYX-13, has therapeutic potential for the treatment of autism

Joseph R. Moskal a,∗, Jeffrey Burgdorf a, Roger A. Kroes a, Stefan M. Brudzynski b, Jaak Panksepp a,c a Falk Center for Molecular Therapeutics, Department of Biomedical Engineering, Northwestern University, Evanston, IL 60201, United States b Department of Psychology, Brock University, St. Catharines, Ontario, L2S 3A1 Canada c Department of Veterinary Comparative Anatomy, Physiology, and Pharmacology, College of Veterinary Medicine Washington State University, Pullman, WA 99163, United States article info abstract

Article history: Deficits in social approach behavior, rough-and-tumble play, and speech abnormalities are core features Received 18 November 2010 of autism that can be modeled in laboratory rats. Human twin studies show that autism has a strong Received in revised form 8 June 2011 genetic component, and a recent review has identified 99 that are dysregulated in human autism. Accepted 10 June 2011 Bioinformatic analysis of these 99 genes identified the NMDA receptor complex as a significant interaction d We dedicate this paper to Ole Ivar Lovaas hub based on -protein interactions. The NMDA receptor glycine site partial agonist -cycloserine (May 8, 1927−August 2, 2010), a pioneer in has been shown to treat the core symptom of social withdrawal in autistic children. Here, we show that the field of autism. rats selectively bred for low rates of play-induced pro-social ultrasonic vocalizations (USVs) can be used to model certain core symptoms of autism. Low-line animals engage in less social contact time with Keywords: conspecifics, show lower rates of play induced pro-social USVs, and show an increased proportion of Autism non-frequency modulated (i.e. monotonous) ultrasonic vocalizations, compared to non-selectively bred NMDA random-line animals. expression patterns in the low-line animals show significant enrichment in Vocalizations autism-associated genes and the NMDA receptor family was identified as a significant hub. Treatment GLYX-13 of low-line animals with the NMDAR glycine site partial agonist GLYX-13 rescued the deficits in play- Rough-and-tumble play induced pro-social 50-kHz and reduced monotonous USVs. Thus, the NMDA receptor has been shown to Microarray play a functional role in autism, and GLYX-13 shows promise for the treatment of autism. We dedicate Rats this paper to Ole Ivar Lovaas (May 8, 1927–August 2, 2010), a pioneer in the field of autism. © 2011 Elsevier Ltd. All rights reserved.

Contents

1. Introduction ...... 1982 2. Brain opioids and autism ...... 1983 3. NMDA receptors play a functional role in autism ...... 1984 4. Low-line rats show an autism-like phenotype and gene expression patterns...... 1984 5. GLYX-13 reverses the autistic-like symptoms of low-line animals ...... 1984 6. Summary and conclusions ...... 1987 Acknowledgements...... 1987 References ...... 1987

1. Introduction patterns of behavior (Wing and Gould, 1979). Autism is now considered to be a developmental disorder and is highly inherita- Autism is characterized by impairments in social interac- ble, with presently unknown environmental exacerbating vectors. tion, communication, and restricted, repetitive, and stereotyped Twin studies show strong genetic heritability for autism. The con- cordance for autism is 60–91% in monozygotic twins compared to 0–10% in dizygotic twins, and 0.3% in the general population ∗ (CDC, 2009; Ritvo et al., 1985; Rosenberg et al., 2009). To date, 99 Corresponding author at: Falk Center for Molecular Therapeutics, Northwestern autism candidate genes have been identified by University, Department of Biomedical Engineering, 1801 Maple Ave, Suite 4300, Evanston, IL 60201, United States. Tel.: +1 847 491 4802; fax: +1 847 491 4810. linking studies and human gene expression profiling by microar- E-mail address: [email protected] (J.R. Moskal). ray (Lee et al., 2009). These include synaptic development genes,

0149-7634/$ – see front matter © 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.neubiorev.2011.06.006 J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988 1983 neurotransmission genes, and neurotrophic genes, among others. autism, namely the brain opioid dysregulation theory of impaired Like other psychiatric disorders, the individual autism candidate social behavior and affect. genes have “weak penetrance” with only a subset of autistic chil- dren showing an abnormality in any individual candidate gene (Meyer-Lindenberg and Weinberger, 2006). A key challenge is to 2. Brain opioids and autism identify which of these autism candidate genes are functionally linked to autism and could therefore be pursued as a therapeutic The increasing opportunities that new strategies for gene dis- target. For instance, bioinformatics tools can now be used to iden- covery in psychiatric disorders can now be combined with theory tify autism candidate genes that have been shown to have a greater driven preclinical models of autism spectrum disorders that focus number of protein–protein interactions with other autism candi- on the potential neurochemical changes that may mediate autistic date genes, called hub genes. It has been hypothesized that these symptoms. Because of recent advances in affective neuroscience, hub genes can serve as “master switches” that control biological the core psychological deficits found in autistic disorders can now function (Borneman et al., 2006). be considered in conjunction with the more easily studied behav- Pharmacological-based therapeutics have had limited success ioral features (Panksepp, 1998). This strategy led to the recognition and substantial side effects. Currently, behavioral modification that brain opioids mediate social bonding, with the recognition that techniques have been the most powerful positive interventions for naltrexone may promote social motivation in a subset of autistic autism spectrum disorders (Lovaas, 1987; Lovaas and Simmons, children (Elchaar et al., 2006). 1969). The antipsychotic risperidone is the only FDA approved Prior to the advent of modern neuro-genetic approaches, pre- medication for autism, and has been reported to (i) significantly clincial models had to rely exclusively on simulation of symptoms. decrease hyperactivity and irritability symptoms, (ii) have little to The first animal model of autism that yielded a positive therapeu- no effect on inappropriate speech or social withdrawal, and (iii) tic effect in humans, followed the dictum that autism reflected a show significant weight gain and sedative side effects (Jesner et al., neural deficit in the neurochemical mechanisms of social related- 2007). ness (Panksepp, 1981). Thirty years ago, it was clear that modest Perhaps the next most effective treatment in facilitating pro- doses of opioid agonists could yield autistic symptoms in animals, social responsivity in a subset of autistic children is the off-label including diminished pain, social motivation, communication and use of the opiate receptor antagonist naltrexone (Elchaar et al., attachment, with some elevation of stereotyped behavioral ten- 2006; Panksepp et al., 1991; Rossignol, 2009). This notion is dencies (from Moles et al., 2004; Panksepp et al., 1980). Social based on the first animal model of autism ever developed which affect and communication were clearly modulated by endogenous was derived from emerging understanding of the opioid neuro- opioids (Keverne et al., 1989; Wohr et al., 2011). The ability of chemistries that regulate social affect and motivation (Bouvard opioid antagonists to increase certain aspects of social engage- et al., 1995; Brambilla et al., 1997; Cazzullo et al., 1999; Gillberg ment and communication suggested the possibility of beneficially et al., 1985; Leboyer et al., 1999; Tordjman et al., 1997). More recent up-regulating social motivation with low doses of orally available work on other neurochemical mechanisms that regulate social opiate receptor antagonists such as naltrexone (Panksepp, 1979; processes has identified oxytocin as a promising candidate for Panksepp and Sahley, 1987). beneficial modulation of autistic symptoms (Green and Hollander, Small open-trials were promising, with optimal prosocial 2010; Panksepp, 1992). Genetic analysis has highlighted one oxy- effects at quite low doses (e.g., ∼0.25 mg/kg p.o. given every other tocin secretary allele that leads to reduced plasma oxytocin, and in day, with higher doses needed for regulating self-injurious behav- such individuals intranasal oxytocin spray “showed relief of symp- iors), suggesting some type of beneficial rebound effects following toms” – for instance, one 23-year old receiving twice daily oxytocin short periods of endogenous opioid blockade (Panksepp et al., showed immediate symptomatic improvements, in which he was 1991). High doses of naltrexone have been especially effective in no longer boisterous after awaking early in the morning. This qui- controlling self-injurious behavior in autism (Elchaar et al., 2006; eting on awaking has been maintained for more than 12 months. Parikh et al., 2008; Symons et al., 2004). Indeed, the benefits of He showed improvements in eye contact behavior with smiling, naltrexone in the self-injurious behaviors in autistic individuals and answering to yes/no questions in his daily life. There were no have been linked to abnormal release patterns of POMC (proopi- adverse effects (Munesue et al., 2010, p. 189). omelanocortin) of the precursor beta-endorphin (Sandman et al., The relevance of endogenous opioids and oxytocin was first 2000). Pilot observations led to promising double-blind studies in identified through animal research. It is now generally accepted young children (Bouvard et al., 1995; Kolmen et al., 1997). Overall, that preclinical animal models are of critical importance for the these results suggested that low dose naltrexone was regulating understanding of brain imbalances that lead to psychiatric dis- the endogenous opioid systems in beneficial ways, perhaps with orders (Hyman, 2007), with models that deal directly with brain benefits being due to the many rebound effects of temporary emotional endophenotypes potentially leading the way (Panksepp, opioid receptor blockade (Brown and Panksepp, 2009). According 2006). At present, the availability of many inbred genetic mouse to Rossignol (2009) the Grade A off-label treatments for autism lines provides abundant allelic variability that can guide thinking spectrum disorders include melatonin, acetylcholinesterase about the sources of the classic triad of core autistic symptoms: inhibitors, naltrexone, and music therapy. (i) communication impairments, (ii) deficits in social motivation, Considering the profound social and communicative deficits and (iii) maladaptive repetitive motor stereotypes, with the affec- of most autistic children, many brain neurochemical imbalances tive substrates best studied through their emotional ultrasonic are bound to mediate the patho-phenotypic expressions. A recent vocalizations (Panksepp et al., 2007; Panksepp and Lahvis, 2007). review by Silverman et al. (2010), underscores this point in a sum- Currently there are many examples of autism-relevant behaviors mary of the recent advances in the identification of candidate in genetic mouse models of autism spectrum disorders (Silverman autism-associated genes and the variety of murine model systems et al., 2010, pp. 492–493). Also, gene knock-out models may pro- that have emerged to characterize them. For instance, deficits in vide functional linkages to production of autistic-like behaviors attachment behaviors have been observed in mice lacking mu- in animals (Nakatani et al., 2009). Before proceeding to our novel opioid receptor genes (Moles et al., 2004). These data underscore glutamate-modulation approach for treating autism, we will briefly the need for continued pre-clinical work to identify effective opi- summarize the only other theory-driven neurobiological approach oid modulating interventions, even as we focus on other relevant to autism that has already brought some benefits for treating genetic, endophenotypic, neurochemical and functional analyses. 1984 J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988

One such possibility has emerged from our ongoing work on pre- NR2A subunit of the NMDA receptor; P < .0001 one tailed t-test]. clinical depression models. We now summarize how our work on As a negative control, we also analyzed cortical microarray data glutamatergic control of the relevant social-affective brain pro- from rats that received social defeat (Burgdorf et al., 2010a), which cesses has revealed a new therapeutic target for ameliorating is not thought to model autism, and found that neither the autism autistic symptoms. candidate genes were upregulated in this gene set, nor was the NMDAR identified as a hub gene (all P’s > .05). 3. NMDA receptors play a functional role in autism 5. GLYX-13 reverses the autistic-like symptoms of low-line Recently, the glutamatergic receptor family, particularly N- animals methyl-d-aspartate receptors (NMDAR), have become of interest as targets for the development of autism therapy. Based on GLYX-13 is an amidated tetrapeptide (threonine–proline– neuroimaging and neuroanatomical studies together with the proline–threonine) derived from one of the hypervariable regions observation that NMDAR antagonists produce autistic-like behav- of a monoclonal antibody, B6B21, shown to be a glycine-site partial iors in healthy subjects, Carlsson has proposed that infantile autism agonist of the NMDAR (Burgdorf et al., 2011; Haring et al., 1991; is a hypoglutamatergic disorder (Carlsson, 1998). Post-mortem Moskal et al., 2005; Stanton et al., 2009; Thompson et al., 1992; studies have also shown abnormal mRNA and Wood et al., 2008; Zhang et al., 2008). GLYX-13 also behaves as an protein levels in the cerebella of autistic individuals (Purcell et al., NR2B subunit preferring NMDAR glycine-site partial agonist (Zhang 2001). The NMDA receptor antagonists, PCP and ketamine, given et al., 2008). To date, GLYX-13 has been reported to: (1) enhance acutely have been shown to mimic the symptoms of autism in the magnitude of long-term potentiation of synaptic transmission humans (Carlsson, 1998). A recent study with autistic individuals while reducing long-term depression (Zhang et al., 2008); (2) sig- showed that daily doses of d-cycloserine (DCS), an NMDAR glycine- nificantly increase learning in a variety of hippocampus-dependent site partial agonist, significantly improved social withdrawal (Posey learning tasks including trace eyeblink conditioning and the Morris et al., 2004). Further, two studies have reported that daily doses water maze in both young adult and learning-impaired aging rats of the NMDA receptor noncompetitive antagonist, amantadine (Burgdorf et al., 2011); (3) have anti-nociceptive properties in both (memantine) reduced some of the negative symptoms of autism the rat formalin model of tonic pain and the rat constriction nerve such as hyperactivity (Chez et al., 2007; King et al., 2001). injury model of neuropathic pain at doses that do not induce ataxia (Wood et al., 2008); (4) markedly reduce delayed (24 h) CA1 pyra- 4. Low-line rats show an autism-like phenotype and gene midal neuronal cell death produced by bilateral carotid occlusion expression patterns in Mongolian gerbils when administered up to 5 h post-ischemia (Stanton et al., 2009); and produced robust antidepressant-like The core symptoms of autism are: (1) deficits in social emo- effects in the rat Porsolt test (Burgdorf et al., 2010b). Recently, a tions and motivations, (2) communication problems, and (3) a high Phase 1 clinical trial was conducted that showed that GLYX-13 did incidence of motor abnormalities, especially repetitive behaviors not produce any ketamine-like dissociative side effects. (Wing and Gould, 1979). These symptoms can be modeled in part in There are several ways to define a partial agonist. For our pur- animals using the following behavioral measures, respectively: (1) poses, a partial agonist is a compound that when added in the reduced time spent in social contact; (2) reduced rates of pro-social presence of a low concentration of a full agonist produces further vocalizations, and (3) disrupted ultrasonic vocalization patterns activation of the NMDA receptor/ complex. However, at (Silverman et al., 2010). high concentrations of a full agonist, the addition of a partial agonist In order to create an animal model that displays these core will inhibit the NMDA receptor/ion channel complex. social and communicative symptoms of autism, we selectively bred That GLYX-13 is a partial agonist at the glycine site of the NMDAR for rats that show low rates of pro-social ultrasonic vocalizations comes from a variety of data. First, GLYX-13 was derived from one (i.e., frequency modulated 50-kHz USVs (Burgdorf et al., 2008)in of the hypervariable regions of the light chain of a monoclonal anti- response to rough and tumble play behavior (Burgdorf et al., 2005, body shown to act as an NMDA receptor glycine site partial agonist. 2009). Our low line animals engage in less social contact time with GLYX-13 was identified by measuring 3H-MK801 binding to rat conspecifics, show lower rates of play induced pro-social ultrasonic hippocampal membrane preparations in the presence of saturat- vocalizations, and show an increased proportion of monotonous ing concentrations of the NMDAR glycine-site specific competitive ultrasonic vocalizations compared to randomly bred animals when antagonist, 7-chloro-kyneurenic acid. GLYX-13 was found to be tested as adults (Fig. 1 and Burgdorf et al., 2009). able to override this block in a concentration-dependent manner We also examined gene expression changes in low-line ani- and in a similar fashion to d-cycloserine, a well established NMDAR mals by microarray using methods as previously described in glycine site partial agonist (Moskal et al., 2005). Second, GLYX-13 Burgdorf et al. (2010a), Kroes et al. (2006) in two brain regions was examined for its partial agonist properties by direct current (medial prefrontal cortex and nucleus accumbens) that have been analyses performed using NMDARs expressed in frog oocytes. It shown to be functionally linked to play-induced pro-social vocal- was seen that GLYX-13 activated these currents in the absence izations (Burgdorf et al., 2007). The list of significantly changed of glycine but in the presence of the coagonist glutamate, inhib- genes in both brain regions in adult low-line animals compared to ited currents in the presence of optimal concentrations of glycine, non-selectively bred adult random-line animals appears in Table 1. competed with 7-chloro-kyneurenic acid all in a dose dependent There is statistically significant overlap between the human autism fashion (Moskal et al., 2005) – exactly as expected for a glycine site associated genes and genes differentially expressed in low-line ani- specific partial agonist. Third, GLYX-13 shows similar behavior mals (8 changed autism candidate genes/101 total changed genes using rat hippocampal slices and measuring both NMDAR cur- representing a 3.5 fold enrichment compared to the chip set bias of rents and the enhancement of long-term potentiation (Zhang 25/1108; P < .05; Chi-square test, two tailed). et al., 2008). Fourth, behavioral studies have shown that low The NMDA receptor was identified as a hub in the genes concentrations of GLYX-13 enhance learning in a variety of differentially expressed in the low-line animals [mean ± SEM hippocampus-dependent learning tasks – clearly an agonist (0.985 ± 0.004) average connections per gene differentially phenomenon – and inhibition of neuropathic pain – an antagonist- expressed in low-line animals compared to 3 connections for the dependent phenomenon (Wood et al., 2008). And fifth, using whole J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988 1985

Table 1 Enrichment of autism candidate genes following gene expression profiling in rats selective bred for low rates of play-induced pro-social vocalizations. Gene expression profiling was conducted by microarray as described in Burgdorf et al. (2010a) in the accumbens and medial prefrontal cortex of adult male Long Evans rats selectively bred for low rates of 50-kHz USVs (Burgdorf et al., 2005, 2009) and non-selectively bred random-line controls. n = 6 per group. The false discovery rate was set at 5%. Low-line animals showed a significant enrichment in human autism-associated genes (2.9 fold enrichment vs. chip set bias, P < .05; Yang and Gill, 2007).

Gene name Medial prefrontal cortex Nucleus accumbens Gene ID Fold change Fold change

Autism candidate genes ABAT 0.80 1.16 4-Aminobutyrate aminotransferase APOE 0.81 1.19 Apolipoprotein E CHRNA4 0.91 0.89 Cholinergic receptor, nicotinic, alpha 4 GABRA5 0.67 1.18 Gamma-aminobutyric acid (GABA) A receptor, alpha 5 GFAP 0.89 1.15 Glial fibrillary acidic protein GRIN2A 0.86 0.91 Glutamate receptor, ionotropic, N-methyl d-aspartate 2A PDYN 0.90 0.92 Prodynorphin PENK 0.80 1.13 Proenkephalin 1 Glutamate receptors GRIA2 0.61 0.66 Glutamate receptor, ionotropic, AMPA 2 GRIA3 0.73 0.77 Glutamate receptor, ionotropic, AMPA 3 GRIA4 0.82 1.27 Glutamate receptor, ionotropic, AMPA4 GRIK4 0.89 1.11 Glutamate receptor, ionotropic, kainate 4 GRM1 0.92 1.23 Glutamate receptor, metabotropic 1 Potassium channels KCNA4 0.72 1.38 Potassium voltage-gated channel, shaker-related subfamily, member 4 KCNC2 0.94 1.08 Potassium voltage gated channel, Shaw-related subfamily, member 2 KCND2 0.82 1.22 Potassium voltage gated channel, Shal-related family, member 2 KCNJ1 0.90 0.82 Potassium inwardly rectifying channel, subfamily J, member 1 KCNK3 0.92 1.07 Potassium channel, subfamily K, member 3 KCNMA1 1.09 1.17 Potassium large conductance calcium-activated channel, subfamily M, alpha member 1 Insulin like growth factor related genes IGF1 1.10 1.10 Insulin-like growth factor 1 IGF2 0.83 1.06 Insulin-like growth factor 2 IGFBP1 1.27 1.10 Insulin-like growth factor binding protein 1 IGFBP3 1.13 0.89 Insulin-like growth factor binding protein 3 Phosphoproteins ATP2A2 0.87 0.92 ATPase, Ca2+ transporting, cardiac muscle, slow twitch 2 ATP2B1 0.90 1.08 ATPase, Ca2+ transporting, plasma membrane 1 C3 1.19 1.13 Complement component 3 CACNA1C 0.87 1.16 Calcium channel, voltage-dependent, L type, alpha 1C subunit CALM1 0.90 0.78 pseudogene 2; calmodulin 3; calmodulin 2; calmodulin 1 CASP3 1.11 1.14 Caspase 3, apoptosis related cysteine protease CKB 0.77 0.89 Creatine kinase, brain CNP 0.92 1.11 2,3-Cyclic nucleotide 3 phosphodiesterase DLGAP4 0.93 0.93 Discs, large homolog-associated protein 4 (Drosophila) FAS 0.93 0.92 Fas (TNF receptor superfamily, member 6) FYN 0.86 1.09 FYN oncogene related to SRC, FGR, YES GABRA4 0.89 1.09 Gamma-aminobutyric acid (GABA) A receptor, alpha 4 GAP43 0.85 0.85 Growth associated protein 43 GNAS 0.72 0.86 GNAS complex H3F3B 1.23 1.20 Similar to H3 histone, family 3B HES1 1.35 1.09 Hairy and enhancer of split 1 (Drosophila) HK2 1.19 1.18 Hexokinase 2 HMBS 1.33 1.19 Hydroxymethylbilane synthase HSPB1 1.16 1.08 Heat shock protein 1 D1 1.34 0.86 Inhibitor of DNA binding 1 ITGB7 1.09 1.07 Integrin, beta 7 ITPR1 0.84 0.82 Inositol 1,4,5-triphosphate receptor, type 1 LMNB1 1.52 1.13 Lamin B1 maP2 0.72 0.86 Microtubule-associated protein 2 MAPK3 1.08 1.13 Mitogen activated protein kinase 3 MBP 0.86 0.90 Myelin basic protein MIF 1.11 0.91 Macrophage migration inhibitory factor MPZ 1.16 0.91 Myelin protein zero NOS2 1.08 0.92 Nitric oxide synthase 2, inducible NR4A1 1.18 1.10 Nuclear receptor subfamily 4, group A, member 1 NTRK2 0.81 1.20 Neurotrophic tyrosine kinase, receptor, type 2 NTRK3 0.86 0.86 Neurotrophic tyrosine kinase, receptor, type 3 P2RX7 0.92 1.29 P2X, ligand-gated ion channel, 7 PDGFRA 0.95 0.95 Platelet derived growth factor receptor, alpha polypeptide PLCB1 0.88 1.11 Phospholipase C, beta 1 (phosphoinositide-specific) PPAP2B 0.83 0.86 Phosphatidic acid phosphatase type 2B PRKCB 0.73 0.80 Protein kinase C, beta SCNN1G 0.95 0.88 Sodium channel, nonvoltage-gated 1 gamma SLC6A17 1.08 0.92 Solute carrier family 6 (neurotransmitter transporter), member 17 SMN1 1.10 1.07 Survival motor neuron 1 SNAP25 0.82 0.83 Synaptosomal-associated protein 25 ST13 1.18 1.12 Similar to suppression oftumorigenicity 13; suppression oftumorigenicity 13 STAT3 1.12 1.14 Signal transducer and activator of transcription 3 STX12 0.95 0.95 12 1986 J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988

Table 1 (Continued)

Gene name Medial prefrontal cortex Nucleus accumbens Gene ID Fold change Fold change

STX1B 0.86 1.09 Syntaxin 1B SYNE1 0.71 0.74 Spectrin repeat containing, nuclear envelope 1 SYT9 0.90 1.15 IX VIM 1.24 1.45 Vimentin Other ATP1B3 1.13 1.07 ATPase, Na+/K+ transporting, beta 3 polypeptide B2M 1.18 1.11 Beta-2 microglobulin BMP1 1.10 0.87 Bone morphogenetic protein 1 CALB1 0.68 0.86 1 CALB2 0.78 1.11 Calbindin 2 CALCB 1.22 0.84 Calcitonin-related polypeptide, beta CASP1 1.15 0.83 Caspase 1 CD48 1.23 1.12 Cd48 molecule CITED2 1.12 1.11 Cbp/p300-interacting transactivator, with Glu/Asp-rich carboxy-terminal domain, 2 CPE 0.84 0.74 Carboxypeptidase E GABRD 0.74 1.14 Gamma-aminobutyric acid (GABA) A receptor, delta GFRA2 0.93 0.91 GDNF family receptor alpha 2 GLRA1 0.86 1.09 , alpha 1 GNA01 0.84 0.93 Guanine nucleotide binding protein (G protein), alpha activating activity polypeptide O HK3 1.35 1.13 Hexokinase 3 (white cell) HPRT1 1.14 1.11 Hypoxanthine phosphoribosyltransferase 1 LTP4 1.24 1.15 Non-specific lip id-transfer protein 4 MGAT2 1.18 0.82 Mannosyl (alpha-1,6-)-glycoprotein beta-1,2-N-acetylglucosaminyltransferase NCAN 0.92 1.14 Neurocan PDYN 0.90 0.92 Prodynorphin Penk 0.80 1.13 Proenkephalin 1 PLP1 0.70 0.72 Proteolipid protein 1 POU3F3 0.90 1.24 POLJ class 3 homeobox 3 PTN 0.91 1.11 Pleiotrophin RHEB 0.94 0.93 Ras homolog enriched in brain SERPINE1 1.20 1.08 Serine (or cysteine) peptidase inhibitor, clade E, member 1 SLC32A1 0.83 1.16 Solute carrier family 32 (GABA vesicular transporter), member 1 SST 0.82 0.89 Somatostatin STX5 1.12 1.09 Syntaxin 5 THRB 0.94 0.91 Thyroid hormone receptor beta VCAM1 0.90 0.90 Vascular cell adhesion molecule 1 WNT4 1.14 0.85 Wingless-type MMTV integration site family, member 4 cell patch clamp technology to measure NMDAR currents in hip- also revealed that GLYX-13 has approximately 25% of the intrinsic pocampal slices and following the Stephenson protocol for the efficacy as d-serine as well (Burgdorf et al., 2010b). Thus GLYX- identification of partial agonists, it was found that again GLYX-13 13 in pharmacological, physiological and behavioral studies shows acted as a partial agonist by facilitating currents in the presence typical partial agonist properties at the glycine site of the NMDAR. of low concentrations of d-serine and inhibiting currents at high In Fig. 2, we show that GLYX-13 (50 mg/kg s.c.) significantly concentrations of d-serine (Burgdorf et al., 2010b). These studies increased rates of play-induced pro-social USVs and significantly

A400 B 200 C 45 180 * 350 40 160 35 300 140 30 250 * 120 25 200 100 20 80 * 150 15 60 100 USVs Monotonous %

Social Contact Time (sec) 10 40 Mean Pro-Social USVs / 2 min / 2 USVs Pro-Social Mean 50 20 5

0 0 0 Random Line Low Line Random Line Low Line Random Line Low Line

Fig. 1. Selective breeding for low rates of play-induced pro-social USVs in rats produces an autism-like phenotype. Mean ± SEM (A) time spent in direct social contact in adult male Long Evans rats selectively bred for low or random rates of play-induced pro-social USVs. Animals were tested in pairs with a conspecific of the same selectively bred line. (B) Rates of conspecific play-induced pro-social USVs (frequency-modulated 50 kHz calls) in adult male low- and random-line animals. (C) Proportion of total conspecific play-induced USVs that are monotonous (i.e., bandwidth less than 7 kHz; Burgdorf et al., 2008). Data adapted and reanalyzed from Burgdorf et al. (2009). n = 14–20 per group. *P < .05 Fisher PLSD post hoc test, 2 tailed. J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988 1987 A 200 * B 60 190 50 180 s 2 min

170 V 40

160 30 * 150

140 20 % Monotonous US Monotonous % an Pro-Social USVs / an Pro-Social USVs

e 130

M 10 120

110 0 Vehicle GLYX-13 Vehicle GLYX-13

Fig. 2. GLYX-13 reverses the low-line autism-like phenotype. Mean ± SEM (A) rates of conspecific play-induced pro-social USVs (frequency-modulated 50 kHz calls) in adolescent male low-line rats pretreated with vehicle (1 mg/ml sterile saline s.c.) or GLYX-13 (50 mg/kg s.c.) 15 min before the start of testing using a within-subjects design. (B) Proportion of total conspecific play-induced USVs that were monotonous (i.e., bandwidth less than 7 kHz; Burgdorf et al., 2008). n = 9 per group. *P < .05 Fisher’s PLSD post hoc test, 2 tailed. decreased the proportion of total USVs that are monotonous (i.e. nificantly influence and shape our thinking. Today we seem to have pure tone whistles without any detectible frequency modulation as actually done what we could only dream about 30 years ago, create described in Burgdorf et al. (2008). Whereas, frequency modulated novel potential therapeutics for treating a variety of CNS disorders 50-kHz calls have been shown to be associated with pro-social pos- including depression, anxiety and perhaps even make an impact itive affective states (Burgdorf and Panksepp, 2006), monotonous on autism spectrum disorder. It has been a remarkable journey and calls have been associated with communication deficits (Ciucci now it appears that the real work is now beginning. This research et al., 2009). These results show that GLYX-13 can reverse the was supported by the Hope for Depression Research Foundation, social/emotional communication deficit in low-line rats. Further New York, NY (JSB, JB, JP and JRM), and The Dr. Ralph and Marian studies examining the effects of GLYX-13 on other behavioral mea- Falk Medical Research Trust, Chicago IL (JRM). We thank Mary sures relevant to autism (e.g. social contact time) are warranted. Schmidt for her expert technical assistance, and the Northwestern University Behavioral Phenotyping Core for its assistance. 6. Summary and conclusions References Convergent evidence that the NMDA receptor complex plays a pivotal role in autism comes from three sources: (1) clinical tri- Borneman, A.R., Leigh-Bell, J.A., Yu, H., Bertone, P., Gerstein, M., Snyder, M., 2006. als with NMDAR modulators; (2) the ability of GLYX-13 to reverse Target hub serve as master regulators of development in yeast. Genes Dev. 20, 435–448. the autistic-like phenotype of rats selectively bred for low rates Bouvard, M.P., Leboyer, M., Launay, J.M., Recasens, C., Plumet, M.H., Waller-Perotte, of play-induced pro-social vocalizations at doses that produce D., Tabuteau, F., Bondoux, D., Dugas, M., Lensing, P., et al., 1995. Low-dose antidepressive-like and anxiolytic-like effects; and (3) the identifi- naltrexone effects on plasma chemistries and clinical symptoms in autism: a double-blind, placebo-controlled study. Psychiatry Res. 58, 191–201. cation of the NMDAR family as a gene hub in autism. Thus NMDAR Brambilla, F., Guareschi-Cazzullo, A., Tacchini, C., Musetti, C., Panerai, A.E., Sacerdote, modulating drugs, particularly glycine-site partial agonists such P., 1997. Beta-endorphin and cholecystokinin 8 concentrations in peripheral as GLYX-13, may have therapeutic potential for the treatment of blood mononuclear cells of autistic children. Neuropsychobiology 35, 1–4. Brown, N., Panksepp, J., 2009. Low-dose naltrexone for disease prevention and qual- autism spectrum disorders. ity of life. Med. Hypotheses 72, 333–337. Burgdorf, J., Kroes, R.A., Beinfeld, M.C., Panksepp, J., Moskal, J.R., 2010a. Uncovering the molecular basis of positive affect using rough-and-tumble play in rats: a role Acknowledgements for insulin-like growth factor I. Neuroscience 168, 769–777. Burgdorf, J., Kroes, R.A., Moskal, J.R., Pfaus, J.G., Brudzynski, S.M., Panksepp, J., Personal note from JRM: Jaak Panksepp has played a much 2008. Ultrasonic vocalizations of rats (Rattus norvegicus) during mating, play, and aggression: Behavioral concomitants, relationship to reward, and self- larger and more vital role in the evolution of the research and administration of playback. J. Comp. Psychol. 122, 357–367. program development than reflected in this manuscript. He has Burgdorf, J., Panksepp, J., 2006. The neurobiology of positive emotions. Neurosci. been a mentor and a friend, an enthusiastic supporter and critical Biobehav. Rev. 30, 173–187. Burgdorf, J., Panksepp, J., Brudzynski, S.M., Beinfeld, M.C., Cromwell, H.C., Kroes, R.A., thinker, and a remarkably sharing individual-both in terms of his Moskal, J.R., 2009. The effects of selective breeding for differential rates of 50- time, his ideas and willingness to help. I have been significantly kHz ultrasonic vocalizations on emotional behavior in rats. Dev. Psychobiol. 51, influenced by Jaak since my graduate school days in biochemistry 34–46. starting in the 1970s when I would make the trek to Bowling Green Burgdorf, J., Panksepp, J., Brudzynski, S.M., Kroes, R., Moskal, J.R., 2005. Breeding for 50-kHz positive affective vocalization in rats. Behav. Genet. 35, 67–72. State University for the privilege and pleasure of spending free Burgdorf, J., Westrich, L., Sprouse, J., Moskal, J.R., 2010b. The antidepressant and weekends working with Jaak. His influence was instrumental in anxiolytic properties of GLYX-13: a novel NMDA receptor glycine site functional helping me craft a Ph.D. in chemistry into a life’s work trying to partial agonist. Soc. Neurosci. Annu. Meet. 572., 523. Burgdorf, J., Wood, P.L., Kroes, R.A., Moskal, J.R., Panksepp, J., 2007. Neurobiology of identify and connect molecular mechanisms to human behavior. 50-kHz ultrasonic vocalizations in rats: electrode mapping, lesion, and pharma- When the opportunity to create the Falk Center for Molecular cology studies. Behav. Brain Res. 182, 274–283. Therapeutics arose, the time had finally come to put all of this Burgdorf, J., Zhang, X.L., Weiss, C., Matthews, E., Disterhoft, J.F., Stanton, P.K., Moskal, J.R., 2011. The N-methyl-d-aspartate receptor modulator GLYX-13 enhances together and Jaak enthusiastically and most graciously agreed to learning and memory, in young adult and learning impaired aging rats. Neu- be our Director of Affective Neuroscience and thus continue to sig- robiol. Aging 23, 3698–3706. 1988 J.R. Moskal et al. / Neuroscience and Biobehavioral Reviews 35 (2011) 1982–1988

Carlsson, M.L., 1998. Hypothesis: is infantile autism a hypoglutamatergic disorder? Tanda, K., Takao, K., Miyakawa, T., Bradley, A., Takumi, T., 2009. Abnormal behav- Relevance of glutamate – serotonin interactions for pharmacotherapy. J. Neural ior in a -engineered mouse model for human 15q11-13 duplication Transm. 105, 525–535. seen in autism. Cell 137, 1235–1246. Cazzullo, A.G., Musetti, M.C., Musetti, L., Bajo, S., Sacerdote, P., Panerai, A., 1999. Panksepp, J., 1979. A neurochemical theory of autism. Trends Neurosci. 2, 174–177. Beta-endorphin levels in peripheral blood mononuclear cells and long-term nal- Panksepp, J., 1981. Brain opioids: a neurochemical substrate for narcotic and social trexone treatment in autistic children. Eur. Neuropsychopharmacol. 9, 361–366. dependence. In: Cooper (Ed.), Progress in Theory in Psychopharmacology. Aca- CDC, 2009. Prevalence of autism spectrum disorders – Autism and Developmental demic Press, London, pp. 149–175. Disabilities Monitoring Network, United States, 2006. MMWR Surveill Summ. Panksepp, J., 1992. Oxytocin effects on emotional processes: separation distress, 58, 1–20. social bonding, and relationships to psychiatric disorders. Ann. N. Y. Acad. Sci. Chez, M.G., Burton, Q., Dowling, T., Chang, M., Khanna, P., Kramer, C., 2007. 652, 243–252. Memantine as adjunctive therapy in children diagnosed with autistic spectrum Panksepp, J., 1998. Affective Neuroscience: The Foundations of Human and Animal disorders: an observation of initial clinical response and maintenance tolerabil- Emotions. Oxford University Press, New York. ity. J. Child Neurol. 22, 574–579. Panksepp, J., 2006. Emotional endophenotypes in evolutionary psychiatry. Prog. Ciucci, M.R., Ahrens, A.M., Ma, S.T., Kane, J.R., Windham, E.B., Woodlee, M.T., Schallert, Neuropsychopharmacol. Biol. Psychiatry 30, 774–784. T., 2009. Reduction of dopamine synaptic activity: degradation of 50-kHz ultra- Panksepp, J., Herman, B.H., Vilberg, T., Bishop, P., DeEskinazi, F.G., 1980. Endogenous sonic vocalization in rats. Behav. Neurosci. 123, 328–336. opioids and social behavior. Neurosci. Biobehav. Rev. 4, 473–487. Elchaar, G.M., Maisch, N.M., Augusto, L.M., Wehring, H.J., 2006. Efficacy and safety of Panksepp, J., Lensing, P., Leboyer, M., Bouvard, M.P., 1991. Naltrexone and other naltrexone use in pediatric patients with autistic disorder. Ann. Pharmacother. potential new pharmacological treatments of autism. Brain Dysfunction 4, 40, 1086–1095. 281–300. Gillberg, C., Terenius, L., Lonnerholm, G., 1985. Endorphin activity in childhood psy- Panksepp, J., Sahley, T., 1987. Possible brain opioid involvement in disrupted social chosis. Spinal fluid levels in 24 cases. Arch. Gen. Psychiatry 42, 780–783. intent and language development of autism. In: Schopler, E., Mesibov, G. (Eds.), Green, J.J., Hollander, E., 2010. Autism and oxytocin: new developments in transla- Neurobiological Issues in Autism. Plenum Press, New York. tional approaches to therapeutics. Neurotherapeutics 7, 250–257. Panksepp, J.B., Jochman, K.A., Kim, J.U., Koy, J.J., Wilson, E.D., Chen, Q., Wilson, C.R., Haring, R., Stanton, P.K., Scheideler, M.A., Moskal, J.R., 1991. Glycine-like modula- Lahvis, G.P., 2007. Affiliative behavior, ultrasonic communication and social tion of N-methyl-d-aspartate receptors by a monoclonal antibody that enhances reward are influenced by genetic variation in adolescent mice. PLoS One 2, e351. long-term potentiation. J. Neurochem. 57, 323–332. Panksepp, J.B., Lahvis, G.P., 2007. Social reward among juvenile mice. Genes Brain Hyman, S.E., 2007. Can neuroscience be integrated into the DSM-V? Nat. Rev. Neu- Behav. 6, 661–671. rosci. 8, 725–732. Parikh, M.S., Kolevzon, A., Hollander, E., 2008. Psychopharmacology of aggression in Jesner, O.S., Aref-Adib, M., Coren, E., 2007. Risperidone for autism spectrum disorder. children and adolescents with autism: a critical review of efficacy and tolerabil- Cochrane Database Syst. Rev., CD005040. ity. J. Child Adolesc. Psychopharmacol. 18, 157–178. Keverne, E.B., Martensz, N.D., Tuite, B., 1989. Beta-endorphin concentrations in Posey, D.J., Kem, D.L., Swiezy, N.B., Sweeten, T.L., Wiegand, R.E., McDougle, C.J., 2004. cerebrospinal fluid of monkeys are influenced by grooming relationships. Psy- A pilot study of d-cycloserine in subjects with autistic disorder. Am. J. Psychiatry choneuroendocrinology 14, 155–161. 161, 2115–2117. King, B.H., Wright, D.M., Handen, B.L., Sikich, L., Zimmerman, A.W., McMahon, W., Purcell, A.E., Jeon, O.H., Zimmerman, A.W., Blue, M.E., Pevsner, J., 2001. Postmortem Cantwell, E., Davanzo, P.A., Dourish, C.T., Dykens, E.M., Hooper, S.R., Jaselskis, brain abnormalities of the glutamate neurotransmitter system in autism. Neu- C.A., Leventhal, B.L., Levitt, J., Lord, C., Lubetsky, M.J., Myers, S.M., Ozonoff, S., rology 57, 1618–1628. Shah, B.G., Snape, M., Shernoff, E.W., Williamson, K., Cook Jr., E.H., 2001. Double- Ritvo, E.R., Freeman, B.J., Mason-Brothers, A., Mo, A., Ritvo, A.M., 1985. Concordance blind, placebo-controlled study of amantadine hydrochloride in the treatment for the syndrome of autism in 40 pairs of afflicted twins. Am. J. Psychiatry 142, of children with autistic disorder. J. Am. Acad. Child Adolesc. Psychiatry 40, 74–77. 658–665. Rosenberg, R.E., Law, J.K., Yenokyan, G., McGready, J., Kaufmann, W.E., Law, P.A., Kolmen, B.K., Feldman, H.M., Handen, B.L., Janosky, J.E., 1997. Naltrexone in young 2009. Characteristics and concordance of autism spectrum disorders among 277 autistic children: replication study and learning measures. J. Am. Acad. Child twin pairs. Arch. Pediatr. Adolesc. Med. 163, 907–914. Adolesc. Psychiatry 36, 1570–1578. Rossignol, D.A., 2009. Novel and emerging treatments for autism spectrum disor- Kroes, R.A., Panksepp, J., Burgdorf, J., Otto, N.J., Moskal, J.R., 2006. Modeling depres- ders: a systematic review. Ann. Clin. Psychiatry 21, 213–236. sion: social dominance-submission gene expression patterns in rat neocortex. Sandman, C.A., Hetrick, W., Talyor, D., Marion, S., Chicz-DeMet, A., 2000. Uncoupling Neuroscience 137, 37–49. of proopiomelanocortin (POMC) fragments is related to self-injury. Peptides 21, Leboyer, M., Philippe, A., Bouvard, M., Guilloud-Bataille, M., Bondoux, D., Tabuteau, 785–791. F., Feingold, J., Mouren-Simeoni, M.C., Launay, J.M., 1999. Whole blood serotonin Silverman, J.L., Yang, M., Lord, C., Crawley, J.N., 2010. Behavioural phenotyping assays and plasma beta-endorphin in autistic probands and their first-degree relatives. for mouse models of autism. Nat. Rev. Neurosci. 11, 490–502. Biol. Psychiatry 45, 158–163. Stanton, P.K., Potter, P.E., Aguilar, J., Decandia, M., Moskal, J.R., 2009. Neuroprotection Lee, S.A., Chan, C.H., Chen, T.C., Yang, C.Y., Huang, K.C., Tsai, C.H., Lai, J.M., Wang, by a novel NMDAR functional glycine site partial agonist, GLYX-13. Neuroreport F.S., Kao, C.Y., Huang, C.Y., 2009. POINeT: protein interactome with sub-network 20, 1193–1197. analysis and hub prioritization. BMC Bioinform. 10, 114. Symons, F.J., Thompson, A., Rodriguez, M.C., 2004. Self-injurious behavior and the Lovaas, O.I., 1987. Behavioral treatment and normal educational and intellectual efficacy of naltrexone treatment: a quantitative synthesis. Ment. Retard. Dev. functioning in young autistic children. J. Consult. Clin. Psychol. 55, 3–9. Disabil. Res. Rev. 10, 193–200. Lovaas, O.I., Simmons, J.Q., 1969. Manipulation of self-destruction in three retarded Thompson, L.T., Moskal, J.R., Disterhoft, J.F., 1992. Hippocampus-dependent learning children. J. Appl. Behav. Anal. 2, 143–157. facilitated by a monoclonal antibody or d-cycloserine. 359, 638–641. Meyer-Lindenberg, A., Weinberger, D.R., 2006. Intermediate phenotypes and genetic Tordjman, S., Anderson, G.M., McBride, P.A., Hertzig, M.E., Snow, M.E., Hall, L.M., mechanisms of psychiatric disorders. Nat. Rev. Neurosci. 7, 818–827. Thompson, S.M., Ferrari, P., Cohen, D.J., 1997. Plasma beta-endorphin, adreno- Moles, A., Kieffer, B.L., D’Amato, F.R., 2004. Deficit in attachment behavior in mice corticotropin hormone, and cortisol in autism. J. Child Psychol. Psychiatry 38, lacking the mu-opioid receptor gene. Science 304, 1983–1986. 705–715. Moskal, J.R., Kuo, A.G., Weiss, C., Wood, P.L., O’Connor Hanson, A., Kelso, S., Harris, Wing, L., Gould, J., 1979. Severe impairments of social interaction and associated R.B., Disterhoft, J.F., 2005. GLYX-13: a monoclonal antibody-derived peptide that abnormalities in children: epidemiology and classification. J. Autism Dev. Disord. acts as an N-methyl-d-aspartate receptor modulator. Neuropharmacology 49, 9, 11–29. 1077–1087. Wohr, M., Moles, A., Schwarting, R.K., DAmato, F.R., 2011. Lack of social exploratory Munesue, T., Yokoyama, S., Nakamura, K., Anitha, A., Yamada, K., Hayashi, K., Asaka, activation in male mu-opioid receptor KO mice in response to playback of female T., Liu, H.X., Jin, D., Koizumi, K., Islam, M.S., Huang, J.J., Ma, W.J., Kim, U.H., Kim, ultrasonic vocalizations. Soc. Neurosci. 6, 76–87. S.J., Park, K., Kim, D., Kikuchi, M., Ono, Y., Nakatani, H., Suda, S., Miyachi, T., Hirai, Wood, P.L., Mahmood, S.A., Moskal, J.R., 2008. Antinociceptive action of GLYX-13: H., Salmina, A., Pichugina, Y.A., Soumarokov, A.A., Takei, N., Mori, N., Tsujii, M., an N-methyl-d-aspartate receptor glycine site partial agonist. Neuroreport 19, Sugiyama, T., Yagi, K., Yamagishi, M., Sasaki, T., Yamasue, H., Kato, N., Hashimoto, 1059–1061. R., Taniike, M., Hayashi, Y., Hamada, J., Suzuki, S., Ooi, A., Noda, M., Kamiyama, Yang, M.S., Gill, M., 2007. A review of gene linkage, association and expression stud- Y., Kido, M.A., Lopatina, O., Hashii, M., Amina, S., Malavasi, F., Huang, E.J., Zhang, ies in autism and an assessment of convergent evidence. Int. J. Dev. Neurosci. J., Shimizu, N., Yoshikawa, T., Matsushima, A., Minabe, Y., Higashida, H., 2010. 25, 69–85. Two genetic variants of CD38 in subjects with autism spectrum disorder and Zhang, X.L., Sullivan, J.A., Moskal, J.R., Stanton, P.K., 2008. A NMDA receptor glycine controls. Neurosci. Res. 67, 181–191. site partial agonist GLYX-13, simultaneously enhances LTP and reduces LTD Nakatani, J., Tamada, K., Hatanaka, F., Ise, S., Ohta, H., Inoue, K., Tomonaga, S., Watan- at Schaffer collateral-CA1 synapses in hippocampus. Neuropharmacology 55, abe, Y., Chung, Y.J., Banerjee, R., Iwamoto, K., Kato, T., Okazawa, M., Yamauchi, K., 1238–1250.