Senior Associate Dean
Rutgers Graduate School of Biomedical Sciences

Associate Professor
Department of Neuroscience and Cell Biology
RBHS, Robert Wood Johnson Medical School

Ph.D., 1993, Princeton University

Telephone: (848) 445-9852
Fax: (732) 235-4850

Autism, neural tube defects, and schizophrenia

My laboratory uses human, mouse, and molecular genetics to understand the biological basis of neurodevelopmental disorders such as autism, neural tube defects (NTDs) and schizophrenia.

Autism Spectrum Disorder (ASD)

Individuals diagnosed with ASD exhibit deficiencies in communication and reciprocal social interactions that are accompanied by rigid or repetitive interests and behaviors. ASD is considered to be a neurodevelopmental disorder that has a polygenic basis. Increased risk for the disorder is believed to be due to multiple genes interacting with each other as well as environmental factors.

My laboratory in collaboration with Linda Brzustowicz MD at Rutgers University (Department of Genetics) has demonstrated that the homeobox transcription factor gene, ENGRAILED 2 (EN2), is a likely ASD susceptibility gene (ASD [MIM 608636]; EN2 [MIM 131310]). Two intronic EN2 SNPs, rs1861972 and rs1861973, are significantly associated with ASD individually and as a haplotype. The common alleles for both SNPs (A-rs1861972; C-rs1861973) are over-transmitted to affected individuals and under-represented in unaffected siblings. These results were observed in an original dataset of 167 families (AGRE I, haplotype, narrow: P=.0009, broad: P=.0017) (Gharani et al 2004) and then replicated in two separate datasets (AGRE II, 222 families, haplotype, narrow: P=.0048, broad: P=.0016 and NIMH, 129 families, haplotype, narrow: P=.0463, broad: P=.0431). When all three datasets were combined additional evidence for association was observed (518 families, haplotype, narrow: P=.00000065; broad: P=.00000035)(Benayed et al 2005; Gharani et al 2004). Four other groups have also reported association for EN2 with autism in their datasets. These data are consistent with EN2 being a likely ASD susceptibility gene.

Further association analysis of 24 EN2 polymorphisms and LD mapping using Hapmap SNPs identified the associated haplotype as the most appropriate candidate for the EN2 risk allele. To test the functionality of A-C haplotype, luciferase (luc) assays were performed in 3 cell types: a non-neuronal cell line (HEK-293T), a neuronal cell line (PC12) and primary neuronal cultures of mouse cerebellar granule cells. Luc reporter constructs were generated with either a minimal SV40 promoter or the EN2 promoter (-5.5kb). In all cases the A-C haplotype resulted in a 15-20% increase in luc levels (P<.005). EMSAs demonstrated specific binding of nuclear proteins to both associated alleles (Benayed et al., 2009). These data indicate the associated A-C haplotype is functional and together with the previous association results support EN2 as an ASD susceptibility gene and the A-C haplotype as a possible risk allele.

We have now generated a series of mouse transgenics to determine the developmental ages (when) and cell types (where) in which the A-C haplotype is functional. These data will help develop hypotheses of how the A-C haplotype affects CNS development and will be important in order to translate our genetic findings into better therapies for ASD.

Neural Tube Detects (NTDs)

To understand the genetic and developmental basis of human NTDs, the lab has been studying the spontaneous mouse mutant called vacuolated lens (vl). NTDs affect neural tube closure and are the second most common congenital birth defect in humans. NTDs are considered to have a multi-factorial basis with both genetics and environment contributing to increased risk.

A single allele of the vacuolated lens mutation has arisen on the C3H/HeSnJ inbred background. Vl homozygotes display congenital cataracts and NTDs. We have positionally cloned the vl locus and determined that a mutation in the orphan G protein coupled receptor (GPCR), Gpr161, results in the vl disease phenotypes. In situ hybridizations demonstrated that Gpr161 is expressed in the lateral neural folds of the neural plate, the developing lens, retina, limb and CNS. Characterization of the vl mutation indicates that C terminal tail of Gpr161 is truncated, leading to multiple effects on the protein including reduced receptor-mediated endocytosis (Matteson et al., 2008)

Interestingly, the vl NTD phenotypes are also multi-genic. On the C3H background, 100% of vl/vl homozygotes display a NTD phenotype (50% spina bifida, 50% a closed NTD phenotype). In addition, 50% of the mutants die, which is likely to be due to the spina bifida phenotype. However, when the vl mutation was crossed onto different genetic backgrounds (C57BL6/J, Mus castaneus (CAST/Ei) and Mus molossinus (MOLF/Ei)) to map the genetic position of the locus, ~50% of the vl/vl homozygotes displayed no phenotype and lethality was not observed. This result indicates that unlinked genetic modifiers on these backgrounds can rescue the associated vl mutant phenotypes.

We have mapped 5 different vl modifier loci on these different genetic backgrounds (Modifiers of vacuolated lens, Modvl 1-5, LOD 3.7-5.0)(Matteson et al., 2008; Korstanje et al., 2008).

Modvl3 specifically affects the penetrance of the vl cataract phenotype (LOD=4.2). Bioinformatic analysis identified Foxe3, a winged helix-forkhead transcription factor, as a possible candidate gene for Modvl3. Mutations in Foxe3 cause cataracts in both mice and humans and a non-synonmous SNP exists between C3H and MOLF/Ei. Genetic and functional data have now determined that Foxe3 is a gene that contributes to the vl cataract modifying phenotype (Matteson et al., 2008).

To investigate whether the MOLF/Ei modifier, Modvl5 (Chr 18 LOD=5.0), is sufficient to rescue the vl phenoypes, we have generated a congenic line where the entire genome is C3H except for the region around Modvl5 on chromosome 18 which retained a MOLF/Ei haplotype. By crossing the Modvl5 modifier to vl mice, we have now demonstrated that Modvl5 is sufficient to rescue vl associated lethality. Ongoing experiments are examining whether Modvl5 rescues the spina bifida phenotype, whether candidate genes contribute to the Modvl5 modifying effect, and whether other modifiers are also sufficient to rescue the vl associated phenotypes.

In summary, because the different vl disease phenotypes (NTDs, cataracts) are due to multiple genetic loci and the vl locus encodes Gpr161 which binds to an unidentified endogenous ligand, our analysis has established an important mouse model to study the multi-factorial basis of both NTDs and cataracts. Future experiments are directed at determining whether Gpr161 and Modvl modifiers contribute to the human disease as we have determined for EN2 with ASD and at identifying the endogenous ligand for Gpr161.


Schizophrenia is a common neuropsychiatric disorder that afflicts ~1:100 individuals. Our collaborator Linda Brzustowicz MD (Rutgers University) has previously reported positive linkage and association results for the Nitric Oxide Synthase 1 Adaptor Protein (NOS1AP) gene (Brzustowicz et al., 2004; Brzustowicz et al., 2000). Molecular genetic experiments (luciferase assays, EMSAs) using human neural cell lines have demonstrated that one of the associated SNPs, rs12742393, is functional (Wratten et al., 2009). Future experiments are directed at identifying the proteins that mediate rs12742393 functionality and investigating whether other associated SNPs are functional in primary neuronal cultures.

Selected Publications

Genestine M, Lin L, Durens M, Yan Y, Jiang Y, Prem S, Bailoor K, Sonsalla PK, Matteson PG, Silverman J, Crawley JN, Millonig JH, DiCicco-Bloom E. (2015) Engrailed-2 (En2) deletion produces multiple neurodevelopmental defects in monoamine systems, forebrain structures and neurogenesis, and behavior. Hum. Mol. Genet. pii:ddv301 [Epub ahead of print]

Li B, Matteson PG, Ababon MR, Nato Jr AQ, Nanda V, Matise TC, Millonig JH. (2015) The orphan GPCR, Gpr161, regulates the retinoic acid and canonical Wnt pathways during neural fold apposition and fusion. Dev. Biol. 402:17-31

Rossman IT, Lin L, Morgan KM, Digiovine M, Van Buskirk EK, Kamdar S, Millonig JH, Dicicco-Bloom E. (2014) Engrailed2 modulates cerebellar granule neuron precursor proliferation, differentiation and insulin-like growth factor 1 signaling during postnatal development. Mol. Autism 5:9

Choi J, Ababon MR, Soliman M, Lin Y, Brzustowicz LM, Matteson PG, Millonig JH. (2014) Autism associated gene, ENGRAILED2, and flanking gene levels are altered in post-mortem cerebellum. PLoS ONE 9:e87208

Brielmaier J, Senerth JM, Silverman JL, Matteson PG, Millonig, JH, DiCicco-Bloom E, Crawley JN. (2014) Chronic desipramine treatment rescues depression-related, social and cognitive deficits in Engrailed-2 knockout mice. Genes Brain Behav. 13:286-98

Choi J, Ababon MR, Matteson PG, Millonig JH. (2012) Cut-like homeobox 1 and nuclear factor I/B mediate ENGRAILED2 autism spectrum disorder-associated haplotype function. Hum. Mol. Genet. 21:1566-80