Skip to main content

Gene expression profiling of lymphoblastoid cell lines from monozygotic twins discordant in severity of autism reveals differential regulation of neurologically relevant genes



The autism spectrum encompasses a set of complex multigenic developmental disorders that severely impact the development of language, non-verbal communication, and social skills, and are associated with odd, stereotyped, repetitive behavior and restricted interests. To date, diagnosis of these neurologically based disorders relies predominantly upon behavioral observations often prompted by delayed speech or aberrant behavior, and there are no known genes that can serve as definitive biomarkers for the disorders.


Here we demonstrate, for the first time, that lymphoblastoid cell lines from monozygotic twins discordant with respect to severity of autism and/or language impairment exhibit differential gene expression patterns on DNA microarrays. Furthermore, we show that genes important to the development, structure, and/or function of the nervous system are among the most differentially expressed genes, and that many of these genes map closely in silico to chromosomal regions containing previously reported autism candidate genes or quantitative trait loci.


Our results provide evidence that novel candidate genes for autism may be differentially expressed in lymphoid cell lines from individuals with autism spectrum disorders. This finding further suggests the possibility of developing a molecular screen for autism based on expressed biomarkers in peripheral blood lymphocytes, an easily accessible tissue. In addition, gene networks are identified that may play a role in the pathophysiology of autism.


Autism and related autism spectrum disorders (including Asperger's Syndrome and pervasive developmental disorder-not otherwise specified (PDD-NOS)) are considered to be among the most devastating psychiatric illnesses affecting children. The three core symptoms of autism spectrum disorders (ASD) are: 1) deficits in social interactions and understanding, 2) aberrant communication and/or language development, and 3) restricted interests and repetitive, stereotyped behaviors [1]. To date, there are no definitive molecular or genetic markers that allow unequivocal diagnosis of ASD, with the exceptions of tuberous sclerosis, Rett's Syndrome, and Fragile X Syndrome [212]. Together, these genetically defined mutations are present in only a minority of individuals (<10%) within the broad autism spectrum. The majority of diagnoses are dependent on behavioral characteristics, according to DSM-IV guidelines, using questionnaires such as the Autism Diagnostic Interview-Revised (ADI-R) [13] or the Autism Diagnostic Observation Schedule (ADOS) [14], which are structured to evaluate children who are approximately 2 or older in mental age. Although the guidelines are relatively clear, the individual rater's (eg., parents, teachers, clinicians, therapists) perception of the evaluated behavior leaves much room for ambiguity. Moreover, with the more mildly affected individuals (eg., with Asperger's Syndrome), diagnosis is often not made until well after the child starts school and, even then, the child is often diagnosed with other more common disorders (such as attention deficit disorder or learning disability) before Asperger's Syndrome is considered, which delays appropriate intervention and effective educational programming. Thus, there is a great need to identify biomarkers that can be used consistently in a clinical setting to diagnose ASD. Furthermore, it is important to identify biological processes that are associated with distinct ASD phenotypes in order to design effective drug therapies targeted to specific individuals.

Although genetic linkage analyses have identified numerous candidate genes for autism [15], there is little consistent data that would support the use of any (or a combination) of these as diagnostic biomarkers for ASD. Furthermore, each candidate gene alone lends little insight into the pathophysiology of these disorders, which are believed to arise from dysregulation of multiple genes. Recently, attention has turned to transcriptional profiling approaches [1619], which involve simultaneous, large-scale expression analysis of thousands of genes on a cDNA (or oligonucleotide) microarray slide, to unravel complex psychiatric disorders. The advantage of transcriptional profiling using microarrays is the ability to study multiple genes in the context of functional gene networks within a living cell, as opposed to forward genetic approaches.

So far, application of microarrays to the study of autism has been described in just one study on post-mortem brain tissue from autistic subjects and matched tissue controls [20]. Thirty genes were identified as being differentially expressed in autistic brain samples relative to matched tissue controls on a combination of 2 separate array platforms containing 588 or 9374 cDNA probes, indicating that autism is associated with multiple disturbances in gene expression. Of this list, only a few genes related specifically to neurological functions and, of these, the glutamate receptor system was targeted for further study. In a similar vein, a recent bioinformatics analysis of autism positional candidate genes using biological databases and computational gene network prediction software demonstrates that the often disparate results from genetic studies implicating a multitude of different genes can be coalesced into interconnected but distinct pathways centered on a specific gene or genes (e.g., FOS and TP53), or on a particular biological theme, e.g., apoptosis [15]. Both of these studies suggest the involvement of multiple genes not previously associated with autism and illustrate the power of using a global approach to study this complex disorder.

The experimental strategy used in the study reported here was designed to tease out differences in gene expression among genetically identical individuals with ASD which might relate to observed differences in the degree of expression of autistic symptoms. Inasmuch as natural variations in gene expression are especially low for monozygotic twins [21, 22] (up to 1.76% in the latter study involving 10,000 genes in 5 pairs of monozygotic twins, compared to ~14% in unrelated individuals), such a strategy has been shown to be useful in identifying candidate genes for bipolar disorder [23] and osteoporosis [24]. Moreover, lymphoblastoid cell lines (LCL) derived from blood cells of autistic individuals were used in this study to explore the possibility that biomarkers for autism could be expressed in easily accessible peripheral cells. Indeed, it has been reported previously that LCL from individuals with bipolar disorder displayed altered gene expression in both postmortem brain tissue and lymphoblasts, although one of the genes, LIM, was altered in the opposite direction in LCL [25]. Follow-up genetic association analyses of this gene demonstrated association of a single nucleotide polymorphism with bipolar disorder [26], indicating the usefulness of LCL and DNA microarray analyses in identifying potential biomarkers of a complex neurological disease.

While studies of gene expression in brain tissue may lead to a better understanding of the mechanistic basis for ASD, it is not an appropriate target for diagnostic assays. Ideally, diagnostic assays should use easily obtained patient samples such as blood, although there is no evidence that gene expression or other markers exist in the peripheral blood of ASD patients. However, one may hypothesize that ASD might arise, in part, through dysregulation of expression of specific neuronal genes and that expression differences between affected and unaffected individuals might be present in tissues other than brain. As a test of this hypothesis, we chose to use DNA microarray analysis to examine gene expression in LCL derived from peripheral blood lymphocytes.

Here we report the first study using a genome-scale approach to identify biomarkers for autism. We demonstrate by gene expression profiling on DNA microarrays that: 1) LCL derived from five monozygotic twin pairs discordant for diagnosed autism and/or language impairment show differential gene expression; 2) a number of the most differentially expressed genes are present in pathways critical to the development and function of the nervous system; 3) there appears to be a quantitative relationship between the severity of the autistic phenotype exhibited by the twins and the expression level of certain genes relative to that of the respective genes in cell lines from non-affected siblings; and 4) approximately half of the most highly differentially expressed genes map in silico to previously reported chromosomal regions containing autism susceptibility genes or quantitative trait loci.


Differential gene expression in lymphoblastoid cell lines from monozygotic twins discordant for classic autism

To determine whether LCL derived from individuals with autism exhibit patterns of gene expression that may be relevant to autism spectrum disorders (ASD), gene expression profiling was performed on LCL derived from 3 sets of male monozygotic twins, one of each pair who met standard diagnostic criteria for autism based on the ADI-R. In each case, the other twin, while not clinically autistic, exhibited autistic traits and was classified either as "broad spectrum" or "not quite autistic" according to guidelines described by the Autism Genetic Resource Exchange (AGRE) repository. Two of the three twin pairs had an unaffected sibling and these were also used for comparison with their respective twin siblings. All of these assays employed an experimental design in which RNA from twin siblings were cohybridized on two-color spotted microarrays containing 39,936 human cDNA elements. Each microarray experiment involved dye-reversal replicates, and was performed in duplicate or, in one case, triplicate for the different sets of twins. The mean log2ratios of each gene from the dye-reversal replicates were used for statistical analyses of the biological replicates.

Principal components analysis (PCA) of the combined microarray data with respect to samples from the 3 discordant twin sets showed that genotype is responsible for the major portion of the variation in differential gene expression, reflecting the expected transcriptome heterogeneity among unrelated individuals (see Additional file 1). The microarray data from the 3 sets of discordant twins was analyzed using SAM in order to identify genes that were significantly different from log2 = 0 across the biological replicates (n = 3). Twelve hundred genes were identified as significant with an FDR of 26%. Twenty-five genes were found to be up-regulated at least 1.5-fold in the more severely affected twin relative to the other twin (log2(ratio) = 0.58) and 19 genes were down-regulated by at least 1.5-fold (Table 1). Of these, eight of the 26 known genes (representing six unique genes) correspond to genes involved in neurological development, function, or disease. Because of this surprising finding, we used quantitative RT-PCR (qPCR) to confirm the differential expression of these specific genes (identified by boldtype in Table 1), as well as several with no known neuronal functions. As shown in Table 1, qPCR confirmed the relative expression levels of all but one of the tested genes, including neurologically relevant argininosuccinate synthetase (ASS), cell death-associated kinase (DAPK1), 5-lipoxygenase-activating protein (ALOX5AP or FLAP), interleukin-6 signal transducer (IL6ST), and Roundabout homolog 1 precursor (ROBO-1), as well as 3 of the "non-neuronal" genes including clotting factor XIIIa (F13A1), eukaryotic translation initiation factor 2C (EIF2C,2), and SAM domain, SH3 domain and nuclear localization signals,1 (SAMSN1).

Table 1 Significant up- and down-regulated genes from SAM analysis of microarray experiments on 3 sets of monozygotic twins discordant for autism diagnosis with log2(ratio) ≥ ± 0.58

Moreover, when the expression profile of cells from the autistic twin was directly compared against that of his respective normal sibling in a dye-reversal microarray experiment, neurologically relevant genes represented 3 of the top 5 most differentially expressed genes (Table 2). Interestingly, the mean log2(ratio) of each of these, ASS, CHL1, and FLAP, are higher for the autistic twin than for the more mildly affected twin when each is compared against their respective normal sibling, suggesting a quantitative relationship between differential gene expression (relative to normal individuals) and severity of autistic symptoms, at least for these specific genes. These quantitative differences have also been confirmed by qPCR analyses.

Table 2 Expression of ASS, CHL1, and FLAP in 2 sets of discordant monozygotic twins relative to expression levels in their respective normal siblings

Network prediction analysis shows interconnected pathways involving differentially expressed, neurologically relevant genes centered around inflammatory mediators

Network prediction analysis using Ingenuity Pathways Analysis Software of the 1200 significant genes from the SAM analysis further revealed that 25 out of 58 network focus genes exhibiting a differential expression = ± 1.5-fold (i.e., log2(ratio) greater than ± 0.58) in at least one discordant twin set are involved in neurological function or disease (Table 3). This expression cutoff was selected because of reports in the literature suggesting that 1.47-fold increases or decreases in gene expression are generally reproducible when Lowess normalization is used [27], and our own ability to confirm expression changes of at least 1.5-fold by qPCR. Of particular note is the gene network that is derived from pathway analysis of the mean expression values (with log2 ratio ≥ ± 0.58) across 3 sets of discordant twins which shows that the majority of significantly differentially expressed genes are part of an extended network centered on TNF and other inflammatory mediators (Fig. 1). The neurological functions of 8 of the genes in this network are described in Table 3, which lists all of the network focus genes that exceed the differential expression cutoff (1.5-fold increase or decrease) in at least 1 pair of twins. In comparing this list to that of Table 1, it is clear that, aside from overlapping differentially expressed genes across the 3 sets of twins, there are network genes that are uniquely up- or down-regulated within each of the twin pairs, as illustrated by Additional file 2. The collective data from the above-mentioned microarray and pathway analyses suggested a short list of novel candidate autism susceptibility genes with reported neurological functions for further evaluation.

Table 3 Network focus genes from Ingenuity Pathways Analysis meeting differential expression cutoff of log2(ratio) ≥ ± 0.58 (1.5-fold increase or decrease) in at least 1 set of twins.
Figure 1

Gene networks showing inter-relationship between differentially expressed genes in LCL from 3 discordant autistic twin sets using Ingenuity Pathways Analysis software. The over-expressed (red) and under-expressed (green) genes were identified as significant using SAM analysis (FDR = 26.4%) of microarray data across 3 twin pairs. The log2 expression ratio cutoff was set at ± 0.58 and was based upon the mean values for each gene. Genes within this network that have a reported role in nervous system development and function are marked with a "#" symbol and include: ASS, ALOX5AP (FLAP), DAPK1, F13A1, IL6ST, NAGLU, PTGS2, and ROBO1. Gray genes are present but do not meet expression cutoff.

Differential expression of autism candidate genes in "concordant" autistic twins

Expression analysis of autism candidate genes in LCL from two sets of twins in which both individuals were diagnosed as autistic surprisingly showed differential expression of several of the candidate genes (Table 4). However, in each case, the "concordant" autistic twin siblings were found to be discordant with respect to severity of language impairment based on each twin's scores on the Peabody Picture Vocabulary Test (PPVT) (see Additional file 3 for profile of subjects studied). Thus, when microarray data from the more language-impaired twin (lower percentile PPVT score) was compared relative to the less impaired twin, the differential gene expression profile overlapped with that which was obtained from the "discordant" twin sets. This result underscores the importance of considering autistic phenotype and/or severity as a means of reducing heterogeneity of gene expression in the search for biomarkers of autism. Interestingly, as shown in Table 4, expression analysis of the candidate genes in cells from monozygotic nonautistic twins demonstrated that two of the genes, CHL1 and possibly ROBO1, were differentially expressed. However, it is worth noting that this set of "normal" twins has two autistic siblings. Thus, it is not clear to what extent their gene expression profiles are "normal" since autism exhibits strong genetic influences. It is therefore possible that the differential expression of these two neurologically relevant genes is not coincidental, but does not, by itself, meet the threshold for association with an autistic phenotype. Alternatively, this result might suggest that these genes are not involved in autism. Clearly, this observation on only one set of normal twins warrants future investigation preferably with twins with no autism in their family background. At this time it is difficult to obtain normal monozygotic twins with no autistic siblings from the AGRE repository which focuses exclusively on collecting samples from pedigrees with familial autism.

Table 4 Relative expression of candidate genes in monozygotic "concordant" twin pairs with differential language impairment (PPVT percentile scores) and in normal twins

The serotonin transporter (5-HTT) gene is also differentially expressed in lymphoblastoid cells from monozygotic twins discordant in severity of autism and/or language impairment

To evaluate whether differential expression of the serotonin transporter (5-HTT), which is strongly implicated in autism, can be detected in LCL from the autistic and nonautistic twins, qPCR analyses were performed, as 5-HTT is not represented on the microarray platform. Results indicated that, while there is no difference in 5-HTT expression between the nonautistic twins, there is a significant decrease in expression in the more severely affected twin in all of the autistic twin pairs studied, as shown in Table 4. Reduced expression of 5-HTT in blood-derived cells may explain hyperserotonemia in a subset of autistic individuals [28]. It should also be noted that a polymorphism in the promoter region of 5-HTT which results in reduced transcription of 5-HTT is a factor in anxiety-related traits [29, 30], common in autism. The present finding suggests that LCL, or their precursor blood lymphocytes, may be useful as reporter cells to evaluate neurologically relevant gene expression differences between autistic and normal individuals.

Network and global functional analyses of the pooled microarray data on monozygotic twins with autism highlight genes involved in nervous system development and function

Because of the observed relationship between severity of symptoms and differential expression of candidate genes across the 5 autistic twin pairs studied, SAM was applied to microarray data from all 5 sets of twins to identify genes that were significantly up- or down-regulated across all twin pairs, each pair of which exhibited differential severity with respect to language ability (Table 5). Once again, pathway analysis of the differentially expressed significant genes revealed an extended network centered on TNF and other cytokines (including IL1B, IL4, and IL6, which was highly expressed in the brain tissues of autistic individuals [34]), connecting a number of neurologically relevant genes (Fig. 2). Global functional analysis of the 1281 significant differentially expressed genes from 5 pairs of twins further shows that genes related to "nervous system development and function" are among the most statistically significant, enriched genes across the 5 sets of twins (Table 6).

Table 5 Significant genes exceeding expression cutoff across 5 sets of twins with ASD
Table 6 Global functional analysis: Enrichment of high level functions represented in datasets of differentially expressed genes across 5 sets of monozygotic twins
Figure 2

Gene networks showing inter-relationship between differentially expressed genes in lymphoblastoid cell lines from monozygotic twins discordant in severity of autism spectrum disorder and/or language impairment. The over-expressed (red) and under-expressed (green) genes were identified as significant using SAM analysis (FDR = 15.6%) of microarray data across 5 twin pairs. The log2 expression ratio cutoff was set at ± 0.58 and was based upon the mean values for each gene. Differentially expressed genes within this network that have a reported role in nervous system development and function are marked with a "#" symbol and include: ALOX5AP (FLAP), CD44, CHL1, EGR2, F13A1, FLT1, IL6ST, ITGB7, and NAGLU. Gray genes are present but do not meet expression cutoff.

In silico mapping of differentially expressed genes to chromosomal regions containing autism candidate genes or quantitative trait loci (QTL)

Although most of the differentially expressed genes identified in this study are novel candidate genes with respect to autism, Table 7 shows that 6 out of 8 of the candidate genes listed in Table 4 [and approximately half (55%) of the differentially expressed genes listed in Tables 3 and 5 (see Additional file 4)] map within or close to chromosomal regions containing previously reported autism candidate genes (ACG) or recently identified QTL for language and nonverbal communication. Interestingly, as observed in Additional File 4 and Table 7 respectively, 32–50% of the genes mapped in silico to autism susceptibility or quantitative trait loci are located within language QTL recently identified by Alarcon et al [59], perhaps reflecting the differential severity of language impairment between the co-twins in this study. This observation suggests that the overlay of expression data onto genetic data allows one to focus further genetic and functional analyses on expressed, neurologically relevant genes that may relate to the behavioral phenotype. Taken together with the network and global functional analyses described above, these results suggest that blood-derived cells may be useful as surrogates to screen for biomarkers for autism.

Table 7 Differentially expressed candidate genes from microarray experiments mapped in silico to autism susceptibility genes and QTL


These studies provide a novel approach to determining candidate genes for autism through the use of peripheral cell lines derived from individuals with ASD. The observations represent a model for the development of a diagnostic screen for autism based on biomarker detection in blood, which is an easily accessible tissue.

In this study, DNA microarrays containing ~40 K human cDNA probes were utilized to examine differences in gene expression profiles in LCL derived from 5 pairs of monozygotic twins with ASD. Three sets of twins were discordant with respect to clinical diagnosis of autism, and 2 sets (with both co-twins diagnosed as autistic) differed with respect to severity of language impairment. We specifically chose this experimental model (direct comparison of identical twins) because differential gene expression in blood leukocytes from monozygotic twins has been reported to be much more restricted than between unrelated controls and, furthermore, the differentially expressed genes exhibited "random variations", showing no specific preference for any functional class [22]. The most remarkable finding of this study is that global functional analysis of the significant differentially expressed genes in LCL from these 5 sets of twins identifies "Nervous system development and function" as a top "high level function" that is significantly enriched across the 5 gene expression datasets (Table 6). Moreover, in silico mapping of our most differentially expressed genes across as well as within the twin sets demonstrates that many of these genes are located in or close to chromosomal regions previously identified as autism susceptibility loci by genetic analyses (Table 7 and Additional file 4). Quantitative RT-PCR analysis has further confirmed the differential expression of a subset of our novel candidate genes in the majority of twin sets studied.

Several of these candidate genes and their associated gene networks may provide insight into potential mechanisms involved in the autistic phenotype(s). One of the striking results of the pathway analyses is that a relatively large number of the differentially expressed, neurologically relevant genes are linked in networks that are centered on genes involved in inflammation (see Figs. 1 and 2). The network genes with reported neurological functions include the proteins ASS, ALOX5AP (FLAP), CD44, CHL1, DAPK1, EGR2, F13A1, FLT1, IL6ST, NAGLU, PTGS2, and ROBO1 (See Table 3). The protein ASS regulates the rate-limiting step involved in nitric oxide (NO) production through regeneration of arginine from citrulline, a byproduct of the nitric oxide synthetase (NOS) reaction [31]. Since NO is a major signaling molecule in the brain that has been implicated in several psychiatric disorders, including autism [32], the increased expression of ASS may be of potential relevance to the autistic phenotype. ASS has also been shown to be induced in a rat model of brain inflammation [33], which would be consistent with the hypothesis that neural inflammation may play a role in autism [34]. DAPK1, a cell death-associated serine/threonine kinase which is involved in suppression of integrin activity and disruption of matrix survival signals [35], is also induced by inflammation [36]. Interestingly, the expression of FLT1 (VEGF receptor 1) is also regulated by inflammatory cytokines as well as by NO [37]. Furthermore, the fact that IL6ST (gp130) is increased in LCL from the more severely affected twin, may complement previous observations that IL-6 is the most elevated inflammatory cytokine in the middle frontal gyrus and anterior cingulate gyrus of brain autopsy tissue from autistic individuals [34]. While upregulation of ASS, DAPK1, FLT1, and IL6ST may be responses to inflammation, ALOX5AP (FLAP) and PTGS2 (COX-2) mediate inflammation through the production of leukotrienes [38] and prostaglandins [39]. Interestingly, 5-lipoxygenase, the target of FLAP activation, has been implicated in aging and neurodegenerative diseases [40], as well as other psychiatric disorders [41], including anxiety and depression, which are frequently co-morbid conditions of autism, while a COX-2 inhibitor, celecoxib, has been shown to have therapeutic effects in major depression [42], further suggesting a role for inflammatory processes in psychiatric disease. Collectively, the potential involvement of these specific genes that are associated with neurological function and disease and their presence in pathways regulated by inflammatory mediators lend further support to the neural inflammation model for autism [34], which may be also manifested by immune dysfunctions commonly observed in autism [43].

In addition to the possible role of genes involved in inflammation, a review of the gene list in Table 3 suggests several additional recurring biological themes among the differentially expressed genes with neurological functions: neuronal survival, neurite extension/guidance, and myelination. In this regard, altered expression of EGR2, the most down-regulated gene across 5 twin sets, may be particularly significant (See Table 5 and Additional file 2). EGR2 (Krox-20) is a transcription factor involved in the development of the brain and peripheral nervous system, routing of axons, and myelination [44, 45]. Some of these functions may be related to EGR2-mediated regulation of ROBO1, which is involved in neuronal differentiation and axon guidance [46, 47], and integrin beta-7 (ITGB7) which has been implicated in chronic demyelinating disease [48]. The expression levels of all three of these genes are relatively reduced with increased severity of autism or language impairment (Table 4). The involvement of cell migration in the pathophysiology of autism is also implicated by the altered expression of CHL1, a novel neural cell adhesion molecule that is involved in neurite migration, outgrowth, connectivity, and survival. Deficiency in CHL1 has been shown to be associated with mental and motor impairments as well as with alterations in exploratory and emotional behavior in mice [49, 50], characteristics that are often associated with autism. However, the effect of CHL1 overexpression, which we observe to be associated with the more severe phenotype, has yet to be determined. While the function of such neurologically relevant genes in lymphoblastoid cell lines is unknown, there is growing evidence that gene expression is under genetic control in LCL, as well as in other cells, with one study showing that 31% of the differential expression in LCL among unrelated individuals was heritable [21]. Thus, it is reasonable to postulate that hereditary factors that are responsible for the development of the autistic brain might also be manifested in the LCL as differentially expressed genes. If expression of these genes can be shown to be consistently altered in LCL in case-control studies on a larger sample of unrelated individuals, these cells, and by inference their precursor blood lymphocytes, can potentially be used as reporter cells for diagnosis of ASD. While we have focused on differentially expressed genes of neurological relevance in this study, it should be noted that the biomarkers for autism in LCL or lymphocytes need not have specific neurological functions (as we have also detected and confirmed differential expression of "non-neuronal" genes). Given that ASD is most probably a multigene disorder of varying etiology, a biomarker screen for ASD would likely include a panel of genes consistently associated with ASD phenotypes, in which diagnosis for the disorder will depend upon differential expression of a defined percentage of genes within the consensus set.

The observed relationship between differential gene expression and severity of ASD between monozygotic twins suggests a role for epigenetic factors in ASD. A recent report on normal monozygotic twins indicates that epigenetic differences arise over time, increasing with age and with physical separation from each other after birth [45]. Indeed, epigenetic differences between monozygotic twins have been examined as possible causes for discordancy in schizophrenia as well as bipolar disorder [5153]. Possible epigenetic mechanisms leading to differences in gene expression include differential methylation, differences in histone acetylation, and micro RNA, although there is no available evidence linking any of these to autism at this time. On the other hand, a mutation in a methyl-CpG binding protein, X-linked MeCP2, has been identified as being involved in 80% of all cases of Rett Syndrome [54], a developmental disorder which overlaps ASD, thus implicating the importance of methylation-dependent gene expression in at least this related disorder. Interestingly, though ubiquitously expressed [55], mutated MeCP2 induces a specific neuronal dysfunction, i.e., Rett Syndrome. One could therefore postulate that differential methylation or differential histone acetylation might give rise to differential expression in LCL from monozygotic twins with ASD and test for global changes in methylation or histone acetylation as done by Fraga et al [45], or for specific changes within a given candidate gene. Such epigenetic modifications in turn could be in response to environmental factors, stochastic processes, or immortalization procedures, which can persist even after the modifying stimulus (eg., inflammation) is removed [56]. If present, these differences could be further tested by evaluation of the methylation/acetylation patterns of DNA/histones in primary lymphocytes from monozygotic twins discordant in severity of autism or language impairment within autism which, while interesting, is beyond the scope of this study.

Regardless of origin, the gene expression differences between monozygotic twins who present with differential severity along the autism spectrum or within a specific behavioral domain (eg., language) are potentially useful, not only as biomarkers for ASD, but also as indicators of genes or metabolic/signaling pathways that may contribute to the autistic phenotype. While our short list of candidate genes (Table 4) focuses on genes with known neurological functions that are similarly up- or down-regulated across twin sets affected by ASD, the set of differentially expressed, neurologically relevant genes that are unique to a given twin set may also be important to the determination of a specific autistic phenotype. Indeed, comparison of the pathways represented in the respective datasets of individual twin pairs reveals not only overlapping genes but also neurologically relevant genes that are differentially expressed in only one of the twin pairs (see Additional file 2). Inasmuch as our microarray analyses directly compared genetically matched individuals who differ only in degree of expression of autistic symptoms, it is likely that other genes, not identified in our study, also play a role in the pathophysiology of autism. This experimental design possibly explains why the candidate genes identified here are different from those reported by an earlier genomic study [20] which compared autopsy brain tissues from autistic and normal (nonautistic) controls (i.e., case-control studies). On the other hand, it is interesting that many of our novel genes map closely to genetically identified autism susceptibility genes/loci or QTL (Table 7 and Additional file 4).

Aside from identifying novel candidate genes for autism, our study also illustrates the need for phenotype definition or subgrouping according to severity along a specific behavioral domain for biological studies of autism. Specifically, the results show that the differential gene expression profiles of concordantly autistic twins with differential severity of language impairment mirror some of the differences in gene expression which are observed in the twins with discordant diagnosis of autism, who also exhibit differential language deficits. Thus, for case-control studies in which individuals from the general population are compared against unrelated controls, subgrouping the autistic individuals by phenotype or stratifying them according to severity of symptoms may provide more clarity in analyzing biological data. Towards this goal, we have used several different clustering methods commonly used in DNA microarray analyses to divide over 1300 autistic individuals into endophenotypic subgroups (eg., language, nonverbal communication, and savant skills) based on item scores on the ADIR questionnaire (manuscript in preparation). Based on these methods, the twin siblings analyzed in this study, including those who were both diagnosed as autistic, each fall into different phenotypic clusters (unpublished data), with the exception of one set of twins who were discordant in the diagnosis of autism. These "endophenotypic" differences may therefore account for some of the differences in gene expression profiles between the twin siblings (i.e., co-twins) as well as among the different sets of twins. To test the ability of our clustering algorithm to restrict phenotypic and biological heterogeneity, we evaluated the short list of candidate genes in Table 4 by qPCR in an additional set of "concordant" autistic twins in which both co-twins exhibit the severely language-impaired phenotype. Results showed that, for this twin pair, there are no differences in expression of the candidate genes exceeding a log2ratio of ± 0.58 (unpublished data).


In summary, our study indicates that LCL from genetically identical autistic individuals who differ in severity of autistic symptoms and/or language impairment exhibit differential expression of genes relevant to neurological development, structure, and function. Many of these genes map to chromosomal regions previously identified by genetic analyses as harboring autism susceptibility genes or QTL, thus demonstrating the potential of combined genomic-genetic analyses to prioritize autism candidate genes for further genetic and functional analyses. In addition, a quantitative relationship is seen between severity of symptoms and expression of several autism candidate genes when twins with classic autism or with milder autistic traits are compared against their respective normal siblings. The finding that gene expression differences were also observed in cells from twins who were both diagnosed as autistic, but who differed in severity in language deficits, strongly suggests that autistic phenotype as well as severity of symptoms must be considered in gene expression studies on autistic individuals in order to reduce biological heterogeneity due to these factors. Collectively, these studies provide proof-of-principle that LCL (and possibly their precursor peripheral blood cells) may exhibit biomarkers relevant to autism, and further suggest their potential usefulness as reporter cells in developing a diagnostic screen for autism. While it is unlikely that microarray studies on LCL will identify the etiology(ies) of autism, this global approach to gene expression analyses is expected to highlight molecular or pathway defects related to the pathophysiology of the condition which, in turn, can be targeted for drug therapies. Moreover, as opposed to fixed autopsy tissues in which RNA may have degraded, a live cell model can also be used to examine the functional consequences of the genomic alteration(s) and the efficacy of different pharmacological agents in ameliorating the impaired function.


Cell lines and culture conditions

Lymphoblastoid cell lines (LCL) derived from lymphocytes of 5 pairs of monozygotic twins with ASD were obtained from the Autism Genetic Resource Exchange (AGRE; Los Angeles, CA) and cultured in DMEM with 15% fetal bovine serum and 1% penicillin-streptomycin. Cell lines from normal siblings of 2 sets of twins were also obtained for comparison of gene expression profile with that of their respective autistic siblings. In addition, cell lines from a set of non-autistic monozygotic twins were also studied. All LCL were constructed and maintained by the Cell Repository of the Department of Genetics at Rutgers University under contract from AGRE. To minimize differences in gene expression due to culture and sample workup conditions, all samples that underwent direct comparison of gene expression profile were cultured and harvested at the same time (3 days after passaging) using the same medium preparation and RNA isolation reagents.

Description of individual donors of cell lines

Additional file 3 provides a case description of all of the subjects included in this study. In brief, all of the twin pairs and normal siblings, with the exception of 1 set of twins, were Caucasian males between the ages of 6 and 16 at the time that blood was drawn. The remaining set of twins (age 12) was of mixed race (black, Hispanic) but had the same mother as one of the Caucasian pairs of autistic twins. For 3 sets of twins (designated "discordant" twins), one twin of each pair met standard diagnostic criteria for autism based on the Autism Diagnostic Interview-Revised (ADIR) [13]. In each case, his co-twin, while not clinically autistic, exhibited autistic traits and could be considered to be on the autism spectrum. These co-twins were described either as "Broad spectrum" or "Not quite autistic (NQA)" by the AGRE repository according to criteria established on the basis of ADI-R scores [79]. Gene expression in cell lines from two of these twin pairs were also directly compared against the gene expression profile in cell lines from their respective "normal" sibling. Two of the 5 sets of twins with ASD (designated "concordant" twins) were examples in which both co-twins were diagnosed with autism, but who were discordant in severity of language impairment, as indicated by their respective percentile scores on the Peabody Picture Vocabulary Test (PPVT). The Autism Diagnostic Observation Schedule (ADOS) [14] was used to diagnose one of these sets of twins. None of the individuals whose cells were used presented with any co-morbid condition or mental retardation. All of the phenotypic data were obtained through the AGRE databases [79].

DNA microarray analyses

RNA was isolated from the LCL using TRIzol Reagent (Invitrogen, CA) according to the manufacturer's protocol. The RNA was further purified using Centricon YM-30 columns and tested for purity on RNA 6000 NanoChips using the Agilent 2100 Bioanalyzer. Labeled cDNA was obtained by incorporation of 5-(3-aminoallyl)-2'deoxyuridine-5'-triphosphate (Ambion, TX) during first-strand synthesis, followed by coupling to the ester of cyanine (Cy-3 or Cy-5) (Molecular Probes, OR) as appropriate according to Standard Operating Protocol (SOP) M004 on The Institute for Genomic Research (TIGR) website [80]. For two-color microarray analyses, the Cy5- and Cy3-labeled cDNA from each pair of twins (or twin and normal sib) were co-hybridized using TIGR SOP M005 to spotted microarrays (TIGR 40 K Human Set) containing 39,936 human cDNA probes which were obtained from Research Genetics. Dye reversal (flip-dye) replicates were included in all analyses, and at least 2 sets of replicates were carried out for each pair of monozygotic twins. Gene expression levels were derived from the scanned hybridized arrays using a combination of TIGR SpotFinder, MIDAS, and MeV analysis programs which are all part of the TM4 Microarray Analysis Software Package available at the above-cited website. These programs have all been previously described in detail [57]. Data analyses included normalization using local LOWESS followed by standard deviation regularization across individual subarrays, and flip-dye consistency checking for dye reversal replicates as implemented in MIDAS [27, 58]. The SAM (Significance Analysis of Microarrays) module within MeV was used to determine statistical significance of differential gene expression and false discovery rate (FDR) which corrects for multiple testing.

Quantitative RT-PCR

Total RNA (same preparations used in microarray analyses) was reverse transcribed into cDNA using the iScript cDNA Synthesis Kit (Bio-Rad, Hercules, CA). Briefly, 2 μg of RNA were added to a 40 μl reaction mix containing reaction buffer, magnesium chloride, dNTPs, an optimized blend of random primers and oligo(dT), an RNase inhibitor, and a MMLV RNase H+ reverse transcriptase. The reaction was incubated at 25°C for 5 minutes followed by 42°C for 30 minutes and ending with 85°C for 5 minutes. The cDNA reactions were then diluted to a volume of 100 μl with water. Real-time PCR was carried out either on a 7900 HT Sequence Detection System from Applied Biosystems using the iTaq SYBR Green Supermix with ROX (Bio-Rad, Hercules, CA) or an Applied Biosystems 7300 Real-Time PCR system using Invitrogen's Superscript III Platinum SYBR Green Two-step qRT-PCR Kit with ROX. Gene-specific primers at a final concentration of 200 nM and 1 μl of cDNA templates were combined into 20 μl reaction mixes and run through 40 cycles of PCR. Quantitation was performed using the Universal 18S rRNA primers (Ambion, Austin, TX) with samples normalized to their 18S rRNA standard curves. Forward and reverse primers are described in Additional file 5.

Network prediction analyses

Lists of differentially expressed genes identified as "significant" by SAM analysis of microarray data across different sets of twins were analyzed using Ingenuity Pathways Analysis (Ingenuity Systems, Inc.), a web-delivered application that enables biologists to discover, visualize and explore therapeutically relevant networks significant to their specific experimental results (e.g., gene expression array data sets). Specifically, a data set containing gene identifiers (in this case, GenBank Accessions) and their corresponding expression values were uploaded as an Excel spreadsheet using the template provided in the application. Each gene identifier was mapped to its corresponding gene object in the Ingenuity Pathways Knowledge Base. The gene list was filtered prior to analysis with Ingenuity by using a log2(ratio) cutoff of 0.58. These genes were then used as the starting point for generating biological networks. The networks are displayed graphically as nodes (genes/gene products) and edges (the biological relationships between the nodes). Human, mouse, and rat orthologs of a gene are stored as separate objects in the knowledge base, but are represented as a single node in the network. The intensity of the node color indicates the degree of up- (red) or down- (green) regulation. When networks from different samples are merged (as in Additional file 2), yellow node color denotes overlapping differentially expressed genes from two or more samples. Nodes are displayed using various shapes that represent the functional class of the gene product, as described on Ingenuity's website [81].

Global functional analyses

Biological functions were assigned to the overall analysis (across data from 5 monozygotic twin pairs) by using the findings that have been extracted from the scientific literature and stored in the Ingenuity Pathways Knowledge Base [81]. The biological functions assigned to the analysis are ranked according to the significance of that biological function to the analysis. A Fisher's exact test is used to calculate a p-value determining the probability that the biological function assigned to the analysis is explained by chance alone. Again, a differential expression cutoff of log2(ratio) ≥ ± 0.58 was used.

In silico mapping of differentially expressed genes

The physical locations of each of the significant differentially expressed genes with log2(ratio) ≥ ± 0.58 were obtained using TIGR's Resourcerer Gene Annotation Software [80]. These locations were then compared manually to those of autism candidate genes (ACG) or quantitative trait loci (QTL) identified on the basis of published genetic linkage and association studies [5978]. For comparison, using Human Genome Build 35 and defining genes as the collection annotated by Ensembl (a total of 24974), we have found that the set of non-redundant genes mapping to the large, often overlapping autism susceptibility or QTL regions is 3838, or about 15.4% of the total.

Submission of microarray data to GEO repository

The GEO accession number for all of the microarray data is GSE4187.



autism candidate genes


Autism Diagnostic Interview-Revised


Autism Diagnostic Observation Schedule


Autism Genetic Resource Exchange


autism spectrum disorders


argininosuccinate synthetase


cell adhesion molecule with homology to L1CAM


death-associated protein kinase 1


early growth response 2 protein (Krox-20)


5-lipoxygenase activating protein


Fms-related tyrosine kinase 1 (VEGF receptor)


5-hydroxytryptamine (serotonin) transporter


integrin beta-7


lymphoblastoid cell lines


Peabody Picture Vocabulary Test


prostaglandin-endoperoxide synthase 2 (cyclooxygenase-2, COX-2)


quantitative trait loci


roundabout, axon guidance receptor


Significance Analysis of Microarrays


  1. 1.

    American Psychiatric Association., American Psychiatric Association. Task Force on DSM-IV.: Diagnostic and statistical manual of mental disorders : DSM-IV. 1994, Washington, D.C. , American Psychiatric Association, xxvii, 886 p.-4th

    Google Scholar 

  2. 2.

    Crino PB, Henske EP: New developments in the neurobiology of the tuberous sclerosis complex. Neurology. 1999, 53 (7): 1384-1390.

    PubMed  CAS  Article  Google Scholar 

  3. 3.

    Bolton PF: Neuroepileptic correlates of autistic symptomatology in tuberous sclerosis. Ment Retard Dev Disabil Res Rev. 2004, 10 (2): 126-131. 10.1002/mrdd.20024.

    PubMed  Article  Google Scholar 

  4. 4.

    Kandt RS, Haines JL, Smith M, Northrup H, Gardner RJ, Short MP, Dumars K, Roach ES, Steingold S, Wall S, Blanton SH, Flodman P, Kwiatkowski DJ, Jewell A, Weber JL, Roses AD, Pericak-Vance MA: Linkage of an important gene locus for tuberous sclerosis to a chromosome 16 marker for polycystic kidney disease. Nat Genet. 1992, 2 (1): 37-41. 10.1038/ng0992-37.

    PubMed  CAS  Article  Google Scholar 

  5. 5.

    Identification and characterization of the tuberous sclerosis gene on chromosome 16. The European Chromosome 16 Tuberous Sclerosis Consortium. Cell. 1993, 75 (7): 1305-1315.

  6. 6.

    Kandt RS: Tuberous sclerosis: the next step. J Child Neurol. 1993, 8 (2): 107-10; discussion 110-1.

    PubMed  CAS  Article  Google Scholar 

  7. 7.

    Mulley JC, Sutherland GR: Integrating maps of chromosome 16. Curr Opin Genet Dev. 1993, 3 (3): 425-431. 10.1016/0959-437X(93)90116-7.

    PubMed  CAS  Article  Google Scholar 

  8. 8.

    van Slegtenhorst M, de Hoogt R, Hermans C, Nellist M, Janssen B, Verhoef S, Lindhout D, van den Ouweland A, Halley D, Young J, Burley M, Jeremiah S, Woodward K, Nahmias J, Fox M, Ekong R, Osborne J, Wolfe J, Povey S, Snell RG, Cheadle JP, Jones AC, Tachataki M, Ravine D, Sampson JR, Reeve MP, Richardson P, Wilmer F, Munro C, Hawkins TL, Sepp T, Ali JB, Ward S, Green AJ, Yates JR, Kwiatkowska J, Henske EP, Short MP, Haines JH, Jozwiak S, Kwiatkowski DJ: Identification of the tuberous sclerosis gene TSC1 on chromosome 9q34. Science. 1997, 277 (5327): 805-808. 10.1126/science.277.5327.805.

    PubMed  CAS  Article  Google Scholar 

  9. 9.

    Henske EP, Ozelius L, Gusella JF, Haines JL, Kwiatkowski DJ: A high-resolution linkage map of human 9q34.1. Genomics. 1993, 17 (3): 587-591. 10.1006/geno.1993.1376.

    PubMed  CAS  Article  Google Scholar 

  10. 10.

    [Epigenetic study of Rett's syndrome as an adequate model for autistic disorders]. Zh Nevrol Psikhiatr Im S S Korsakova. 2005, 105 (7): 4-11.

  11. 11.

    Gedeon AK, Baker E, Robinson H, Partington MW, Gross B, Manca A, Korn B, Poustka A, Yu S, Sutherland GR, Mulley JC: Fragile X syndrome without CCG amplification has an FMR1 deletion. Nat Genet. 1992, 1 (5): 341-344. 10.1038/ng0892-341.

    PubMed  CAS  Article  Google Scholar 

  12. 12.

    Hansen RS, Gartler SM, Scott CR, Chen SH, Laird CD: Methylation analysis of CGG sites in the CpG island of the human FMR1 gene. Hum Mol Genet. 1992, 1 (8): 571-578.

    PubMed  CAS  Article  Google Scholar 

  13. 13.

    Lord C, Rutter M, Le Couteur A: Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord. 1994, 24 (5): 659-685. 10.1007/BF02172145.

    PubMed  CAS  Article  Google Scholar 

  14. 14.

    Lord C, Rutter M, Goode S, Heemsbergen J, Jordan H, Mawhood L, Schopler E: Autism diagnostic observation schedule: a standardized observation of communicative and social behavior. J Autism Dev Disord. 1989, 19 (2): 185-212. 10.1007/BF02211841.

    PubMed  CAS  Article  Google Scholar 

  15. 15.

    Yonan AL, Palmer AA, Smith KC, Feldman I, Lee HK, Yonan JM, Fischer SG, Pavlidis P, Gilliam TC: Bioinformatic analysis of autism positional candidate genes using biological databases and computational gene network prediction. Genes Brain Behav. 2003, 2 (5): 303-320. 10.1034/j.1601-183X.2003.00041.x.

    PubMed  CAS  Article  Google Scholar 

  16. 16.

    Geschwind DH: DNA microarrays: translation of the genome from laboratory to clinic. Lancet Neurol. 2003, 2 (5): 275-282. 10.1016/S1474-4422(03)00379-X.

    PubMed  CAS  Article  Google Scholar 

  17. 17.

    Schena M, Shalon D, Davis RW, Brown PO: Quantitative monitoring of gene expression patterns with a complementary DNA microarray. Science. 1995, 270 (5235): 467-470.

    PubMed  CAS  Article  Google Scholar 

  18. 18.

    Lipshutz RJ, Fodor SP, Gingeras TR, Lockhart DJ: High density synthetic oligonucleotide arrays. Nat Genet. 1999, 21 (1 Suppl): 20-24. 10.1038/4447.

    PubMed  CAS  Article  Google Scholar 

  19. 19.

    Pongrac J, Middleton FA, Lewis DA, Levitt P, Mirnics K: Gene expression profiling with DNA microarrays: advancing our understanding of psychiatric disorders. Neurochem Res. 2002, 27 (10): 1049-1063. 10.1023/A:1020904821237.

    PubMed  CAS  Article  Google Scholar 

  20. 20.

    Purcell AE, Jeon OH, Zimmerman AW, Blue ME, Pevsner J: Postmortem brain abnormalities of the glutamate neurotransmitter system in autism. Neurology. 2001, 57 (9): 1618-1628.

    PubMed  CAS  Article  Google Scholar 

  21. 21.

    Cheung VG, Conlin LK, Weber TM, Arcaro M, Jen KY, Morley M, Spielman RS: Natural variation in human gene expression assessed in lymphoblastoid cells. Nat Genet. 2003, 33 (3): 422-425. 10.1038/ng1094.

    PubMed  CAS  Article  Google Scholar 

  22. 22.

    Sharma A, Sharma VK, Horn-Saban S, Lancet D, Ramachandran S, Brahmachari SK: Assessing natural variations in gene expression in humans by comparing with monozygotic twins using microarrays. Physiol Genomics. 2005, 21 (1): 117-123. 10.1152/physiolgenomics.00228.2003.

    PubMed  CAS  Article  Google Scholar 

  23. 23.

    Kakiuchi C, Iwamoto K, Ishiwata M, Bundo M, Kasahara T, Kusumi I, Tsujita T, Okazaki Y, Nanko S, Kunugi H, Sasaki T, Kato T: Impaired feedback regulation of XBP1 as a genetic risk factor for bipolar disorder. Nat Genet. 2003, 35 (2): 171-175. 10.1038/ng1235.

    PubMed  CAS  Article  Google Scholar 

  24. 24.

    Mak YT, Hampson G, Beresford JN, Spector TD: Variations in genome-wide gene expression in identical twins - a study of primary osteoblast-like culture from female twins discordant for osteoporosis. BMC Genet. 2004, 5 (1): 14-10.1186/1471-2156-5-14.

    PubMed  PubMed Central  Article  Google Scholar 

  25. 25.

    Iwamoto K, Kakiuchi C, Bundo M, Ikeda K, Kato T: Molecular characterization of bipolar disorder by comparing gene expression profiles of postmortem brains of major mental disorders. Mol Psychiatry. 2004, 9 (4): 406-416. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  26. 26.

    Kato T, Iwayama Y, Kakiuchi C, Iwamoto K, Yamada K, Minabe Y, Nakamura K, Mori N, Fujii K, Nanko S, Yoshikawa T: Gene expression and association analyses of LIM (PDLIM5) in bipolar disorder and schizophrenia. Mol Psychiatry. 2005

    Google Scholar 

  27. 27.

    Yang IV, Chen E, Hasseman JP, Liang W, Frank BC, Wang S, Sharov V, Saeed AI, White J, Li J, Lee NH, Yeatman TJ, Quackenbush J: Within the fold: assessing differential expression measures and reproducibility in microarray assays. Genome Biol. 2002, 3 (11): research0062-

    PubMed  PubMed Central  Google Scholar 

  28. 28.

    Coutinho AM, Oliveira G, Morgadinho T, Fesel C, Macedo TR, Bento C, Marques C, Ataide A, Miguel T, Borges L, Vicente AM: Variants of the serotonin transporter gene (SLC6A4) significantly contribute to hyperserotonemia in autism. Mol Psychiatry. 2004, 9 (3): 264-271. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  29. 29.

    Bradley SL, Dodelzon K, Sandhu HK, Philibert RA: Relationship of serotonin transporter gene polymorphisms and haplotypes to mRNA transcription. Am J Med Genet B Neuropsychiatr Genet. 2005, 136 (1): 58-61.

    Article  Google Scholar 

  30. 30.

    Lesch KP, Bengel D, Heils A, Sabol SZ, Greenberg BD, Petri S, Benjamin J, Muller CR, Hamer DH, Murphy DL: Association of anxiety-related traits with a polymorphism in the serotonin transporter gene regulatory region. Science. 1996, 274 (5292): 1527-1531. 10.1126/science.274.5292.1527.

    PubMed  CAS  Article  Google Scholar 

  31. 31.

    Hattori Y, Campbell EB, Gross SS: Argininosuccinate synthetase mRNA and activity are induced by immunostimulants in vascular smooth muscle. Role in the regeneration or arginine for nitric oxide synthesis. J Biol Chem. 1994, 269 (13): 9405-9408.

    PubMed  CAS  Google Scholar 

  32. 32.

    Akyol O, Zoroglu SS, Armutcu F, Sahin S, Gurel A: Nitric oxide as a physiopathological factor in neuropsychiatric disorders. In Vivo. 2004, 18 (3): 377-390.

    PubMed  CAS  Google Scholar 

  33. 33.

    Heneka MT, Schmidlin A, Wiesinger H: Induction of argininosuccinate synthetase in rat brain glial cells after striatal microinjection of immunostimulants. J Cereb Blood Flow Metab. 1999, 19 (8): 898-907. 10.1097/00004647-199908000-00009.

    PubMed  CAS  Article  Google Scholar 

  34. 34.

    Vargas DL, Nascimbene C, Krishnan C, Zimmerman AW, Pardo CA: Neuroglial activation and neuroinflammation in the brain of patients with autism. Ann Neurol. 2005, 57 (1): 67-81. 10.1002/ana.20315.

    PubMed  CAS  Article  Google Scholar 

  35. 35.

    Wang WJ, Kuo JC, Yao CC, Chen RH: DAP-kinase induces apoptosis by suppressing integrin activity and disrupting matrix survival signals. J Cell Biol. 2002, 159 (1): 169-179. 10.1083/jcb.200204050.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  36. 36.

    Gupta S: A decision between life and death during TNF-alpha-induced signaling. J Clin Immunol. 2002, 22 (4): 185-194. 10.1023/A:1016089607548.

    PubMed  CAS  Article  Google Scholar 

  37. 37.

    Frank S, Stallmeyer B, Kampfer H, Schaffner C, Pfeilschifter J: Differential regulation of vascular endothelial growth factor and its receptor fms-like-tyrosine kinase is mediated by nitric oxide in rat renal mesangial cells. Biochem J. 1999, 338 ( Pt 2): 367-374. 10.1042/0264-6021:3380367.

    CAS  Article  Google Scholar 

  38. 38.

    Byrum RS, Goulet JL, Griffiths RJ, Koller BH: Role of the 5-lipoxygenase-activating protein (FLAP) in murine acute inflammatory responses. J Exp Med. 1997, 185 (6): 1065-1075. 10.1084/jem.185.6.1065.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  39. 39.

    Pokharel YR, Yang JW, Kim JY, Oh HW, Jeong HG, Woo ER, Kang KW: Potent inhibition of the inductions of inducible nitric oxide synthase and cyclooxygenase-2 by taiwaniaflavone. Nitric Oxide. 2006

    Google Scholar 

  40. 40.

    Manev H, Uz T, Sugaya K, Qu T: Putative role of neuronal 5-lipoxygenase in an aging brain. Faseb J. 2000, 14 (10): 1464-1469. 10.1096/fj.14.10.1464.

    PubMed  CAS  Article  Google Scholar 

  41. 41.

    Manev R, Manev H: 5-Lipoxygenase as a putative link between cardiovascular and psychiatric disorders. Crit Rev Neurobiol. 2004, 16 (1-2): 181-186. 10.1615/CritRevNeurobiol.v16.i12.190.

    PubMed  CAS  Article  Google Scholar 

  42. 42.

    Muller N, Schwarz MJ, Dehning S, Douhe A, Cerovecki A, Goldstein-Muller B, Spellmann I, Hetzel G, Maino K, Kleindienst N, Moller HJ, Arolt V, Riedel M: The cyclooxygenase-2 inhibitor celecoxib has therapeutic effects in major depression: results of a double-blind, randomized, placebo controlled, add-on pilot study to reboxetine. Mol Psychiatry. 2006

    Google Scholar 

  43. 43.

    Cohly HH, Panja A: Immunological findings in autism. Int Rev Neurobiol. 2005, 71: 317-341.

    PubMed  CAS  Article  Google Scholar 

  44. 44.

    Kamholz J, Awatramani R, Menichella D, Jiang H, Xu W, Shy M: Regulation of myelin-specific gene expression. Relevance to CMT1. Ann N Y Acad Sci. 1999, 883: 91-108. 10.1111/j.1749-6632.1999.tb08572.x.

    PubMed  CAS  Article  Google Scholar 

  45. 45.

    LeBlanc SE, Jang SW, Ward RM, Wrabetz L, Svaren J: Direct regulation of myelin protein zero expression by the Egr2 transactivator. J Biol Chem. 2006, 281 (9): 5453-5460. 10.1074/jbc.M512159200.

    PubMed  CAS  Article  Google Scholar 

  46. 46.

    Connor RM, Key B: Expression and role of Roundabout-1 in embryonic Xenopus forebrain. Dev Dyn. 2002, 225 (1): 22-34. 10.1002/dvdy.10130.

    PubMed  CAS  Article  Google Scholar 

  47. 47.

    Lee JS, Ray R, Chien CB: Cloning and expression of three zebrafish roundabout homologs suggest roles in axon guidance and cell migration. Dev Dyn. 2001, 221 (2): 216-230. 10.1002/dvdy.1136.

    PubMed  CAS  Article  Google Scholar 

  48. 48.

    Kanwar JR, Harrison JE, Wang D, Leung E, Mueller W, Wagner N, Krissansen GW: Beta7 integrins contribute to demyelinating disease of the central nervous system. J Neuroimmunol. 2000, 103 (2): 146-152. 10.1016/S0165-5728(99)00245-3.

    PubMed  CAS  Article  Google Scholar 

  49. 49.

    Buhusi M, Midkiff BR, Gates AM, Richter M, Schachner M, Maness PF: Close homolog of L1 is an enhancer of integrin-mediated cell migration. J Biol Chem. 2003, 278 (27): 25024-25031. 10.1074/jbc.M303084200.

    PubMed  CAS  Article  Google Scholar 

  50. 50.

    Naus S, Richter M, Wildeboer D, Moss M, Schachner M, Bartsch JW: Ectodomain shedding of the neural recognition molecule CHL1 by the metalloprotease-disintegrin ADAM8 promotes neurite outgrowth and suppresses neuronal cell death. J Biol Chem. 2004, 279 (16): 16083-16090. 10.1074/jbc.M400560200.

    PubMed  CAS  Article  Google Scholar 

  51. 51.

    Abdolmaleky HM, Thiagalingam S, Wilcox M: Genetics and epigenetics in major psychiatric disorders: dilemmas, achievements, applications, and future scope. Am J Pharmacogenomics. 2005, 5 (3): 149-160. 10.2165/00129785-200505030-00002.

    PubMed  CAS  Article  Google Scholar 

  52. 52.

    Petronis A, Gottesman II, Kan P, Kennedy JL, Basile VS, Paterson AD, Popendikyte V: Monozygotic twins exhibit numerous epigenetic differences: clues to twin discordance?. Schizophr Bull. 2003, 29 (1): 169-178.

    PubMed  Article  Google Scholar 

  53. 53.

    Petronis A: Epigenetics and bipolar disorder: new opportunities and challenges. Am J Med Genet C Semin Med Genet. 2003, 123 (1): 65-75. 10.1002/ajmg.c.20015.

    Article  Google Scholar 

  54. 54.

    Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY: Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet. 1999, 23 (2): 185-188. 10.1038/13810.

    PubMed  CAS  Article  Google Scholar 

  55. 55.

    Nan X, Ng HH, Johnson CA, Laherty CD, Turner BM, Eisenman RN, Bird A: Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature. 1998, 393 (6683): 386-389. 10.1038/30764.

    PubMed  CAS  Article  Google Scholar 

  56. 56.

    Jablonka E, Lamb MJ: The changing concept of epigenetics. Ann N Y Acad Sci. 2002, 981: 82-96.

    PubMed  Article  Google Scholar 

  57. 57.

    Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N, Braisted J, Klapa M, Currier T, Thiagarajan M, Sturn A, Snuffin M, Rezantsev A, Popov D, Ryltsov A, Kostukovich E, Borisovsky I, Liu Z, Vinsavich A, Trush V, Quackenbush J: TM4: a free, open-source system for microarray data management and analysis. Biotechniques. 2003, 34 (2): 374-378.

    PubMed  CAS  Google Scholar 

  58. 58.

    Quackenbush J: Microarray data normalization and transformation. Nat Genet. 2002, 32 Suppl: 496-501. 10.1038/ng1032.

    PubMed  Article  Google Scholar 

  59. 59.

    Smith M, Filipek PA, Wu C, Bocian M, Hakim S, Modahl C, Spence MA: Analysis of a 1-megabase deletion in 15q22-q23 in an autistic patient: identification of candidate genes for autism and of homologous DNA segments in 15q22-q23 and 15q11-q13. Am J Med Genet. 2000, 96 (6): 765-770. 10.1002/1096-8628(20001204)96:6<765::AID-AJMG13>3.0.CO;2-L.

    PubMed  CAS  Article  Google Scholar 

  60. 60.

    Bartlett CW, Goedken R, Vieland VJ: Effects of updating linkage evidence across subsets of data: reanalysis of the autism genetic resource exchange data set. Am J Hum Genet. 2005, 76 (4): 688-695. 10.1086/429345.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  61. 61.

    Shao Y, Cuccaro ML, Hauser ER, Raiford KL, Menold MM, Wolpert CM, Ravan SA, Elston L, Decena K, Donnelly SL, Abramson RK, Wright HH, DeLong GR, Gilbert JR, Pericak-Vance MA: Fine mapping of autistic disorder to chromosome 15q11-q13 by use of phenotypic subtypes. Am J Hum Genet. 2003, 72 (3): 539-548. 10.1086/367846.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  62. 62.

    A full genome screen for autism with evidence for linkage to a region on chromosome 7q. International Molecular Genetic Study of Autism Consortium. Hum Mol Genet. 1998, 7 (3): 571-578. 10.1093/hmg/7.3.571.

  63. 63.

    Risch N, Spiker D, Lotspeich L, Nouri N, Hinds D, Hallmayer J, Kalaydjieva L, McCague P, Dimiceli S, Pitts T, Nguyen L, Yang J, Harper C, Thorpe D, Vermeer S, Young H, Hebert J, Lin A, Ferguson J, Chiotti C, Wiese-Slater S, Rogers T, Salmon B, Nicholas P, Petersen PB, Pingree C, McMahon W, Wong DL, Cavalli-Sforza LL, Kraemer HC, Myers RM: A genomic screen of autism: evidence for a multilocus etiology. Am J Hum Genet. 1999, 65 (2): 493-507. 10.1086/302497.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  64. 64.

    Serajee FJ, Zhong H, Nabi R, Huq AH: The metabotropic glutamate receptor 8 gene at 7q31: partial duplication and possible association with autism. J Med Genet. 2003, 40 (4): e42-10.1136/jmg.40.4.e42.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  65. 65.

    Gharani N, Benayed R, Mancuso V, Brzustowicz LM, Millonig JH: Association of the homeobox transcription factor, ENGRAILED 2, 3, with autism spectrum disorder. Mol Psychiatry. 2004, 9 (5): 474-484. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  66. 66.

    Bacchelli E, Blasi F, Biondolillo M, Lamb JA, Bonora E, Barnby G, Parr J, Beyer KS, Klauck SM, Poustka A, Bailey AJ, Monaco AP, Maestrini E: Screening of nine candidate genes for autism on chromosome 2q reveals rare nonsynonymous variants in the cAMP-GEFII gene. Mol Psychiatry. 2003, 8 (11): 916-924. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  67. 67.

    Cantor RM, Kono N, Duvall JA, Alvarez-Retuerto A, Stone JL, Alarcon M, Nelson SF, Geschwind DH: Replication of autism linkage: fine-mapping peak at 17q21. Am J Hum Genet. 2005, 76 (6): 1050-1056. 10.1086/430278.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  68. 68.

    Sultana R, Yu CE, Yu J, Munson J, Chen D, Hua W, Estes A, Cortes F, de la Barra F, Yu D, Haider ST, Trask BJ, Green ED, Raskind WH, Disteche CM, Wijsman E, Dawson G, Storm DR, Schellenberg GD, Villacres EC: Identification of a novel gene on chromosome 7q11.2 interrupted by a translocation breakpoint in a pair of autistic twins. Genomics. 2002, 80 (2): 129-134. 10.1006/geno.2002.6810.

    PubMed  CAS  Article  Google Scholar 

  69. 69.

    Wassink TH, Brzustowicz LM, Bartlett CW, Szatmari P: The search for autism disease genes. Ment Retard Dev Disabil Res Rev. 2004, 10 (4): 272-283. 10.1002/mrdd.20041.

    PubMed  Article  Google Scholar 

  70. 70.

    Persico AM, D'Agruma L, Zelante L, Militerni R, Bravaccio C, Schneider C, Melmed R, Trillo S, Montecchi F, Elia M, Palermo M, Rabinowitz D, Pascucci T, Puglisi-Allegra S, Reichelt KL, Muscarella L, Guarnieri V, Melgari JM, Conciatori M, Keller F: Enhanced APOE2 transmission rates in families with autistic probands. Psychiatr Genet. 2004, 14 (2): 73-82. 10.1097/01.ypg.0000128768.37838.17.

    PubMed  CAS  Article  Google Scholar 

  71. 71.

    Veenstra-VanderWeele J, Kim SJ, Lord C, Courchesne R, Akshoomoff N, Leventhal BL, Courchesne E, Cook EHJ: Transmission disequilibrium studies of the serotonin 5-HT2A receptor gene (HTR2A) in autism. Am J Med Genet. 2002, 114 (3): 277-283. 10.1002/ajmg.10192.

    PubMed  Article  Google Scholar 

  72. 72.

    Bradford Y, Haines J, Hutcheson H, Gardiner M, Braun T, Sheffield V, Cassavant T, Huang W, Wang K, Vieland V, Folstein S, Santangelo S, Piven J: Incorporating language phenotypes strengthens evidence of linkage to autism. Am J Med Genet. 2001, 105 (6): 539-547. 10.1002/ajmg.1497.

    PubMed  CAS  Article  Google Scholar 

  73. 73.

    Chen GK, Kono N, Geschwind DH, Cantor RM: Quantitative trait locus analysis of nonverbal communication in autism spectrum disorder. Mol Psychiatry. 2005

    Google Scholar 

  74. 74.

    Persico AM, D'Agruma L, Maiorano N, Totaro A, Militerni R, Bravaccio C, Wassink TH, Schneider C, Melmed R, Trillo S, Montecchi F, Palermo M, Pascucci T, Puglisi-Allegra S, Reichelt KL, Conciatori M, Marino R, Quattrocchi CC, Baldi A, Zelante L, Gasparini P, Keller F: Reelin gene alleles and haplotypes as a factor predisposing to autistic disorder. Mol Psychiatry. 2001, 6 (2): 150-159. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  75. 75.

    A genomewide screen for autism: strong evidence for linkage to chromosomes 2q, 7q, and 16p. Am J Hum Genet. 2001, 69 (3): 570-581. 10.1086/323264.

  76. 76.

    Alarcon M, Yonan AL, Gilliam TC, Cantor RM, Geschwind DH: Quantitative genome scan and Ordered-Subsets Analysis of autism endophenotypes support language QTLs. Mol Psychiatry. 2005, 10 (8): 747-757. 10.1038/

    PubMed  CAS  Article  Google Scholar 

  77. 77.

    Robinson PD, Schutz CK, Macciardi F, White BN, Holden JJ: Genetically determined low maternal serum dopamine beta-hydroxylase levels and the etiology of autism spectrum disorders. Am J Med Genet. 2001, 100 (1): 30-36. 10.1002/ajmg.1187.

    PubMed  CAS  Article  Google Scholar 

  78. 78.

    Alarcon M, Cantor RM, Liu J, Gilliam TC, Geschwind DH: Evidence for a language quantitative trait locus on chromosome 7q in multiplex autism families. Am J Hum Genet. 2002, 70 (1): 60-71. 10.1086/338241.

    PubMed  CAS  PubMed Central  Article  Google Scholar 

  79. 79.


  80. 80.


  81. 81.


Download references


We gratefully acknowledge the resources provided by the Autism Genetic Resource Exchange (AGRE) Consortium* and the participating AGRE families. The Autism Genetic Resource Exchange is a program of Cure Autism Now and is supported, in part, by grant MH64547 from the National Institute of Mental Health to Daniel H. Geschwind (PI).

*The AGRE Consortium:

Dan Geschwind, M.D., Ph.D., UCLA, Los Angeles, CA;

Maja Bucan, Ph.D., University of Pennsylvania, Philadelphia, PA;

W.Ted Brown, M.D., Ph.D., F.A.C.M.G., N.Y.S. Institute for Basic Research in Developmental Disabilities, Long Island, NY;

Rita M. Cantor, Ph.D., UCLA School of Medicine, Los Angeles, CA;

John N. Constantino, M.D., Washington University School of Medicine, St. Louis, MO;

T.Conrad Gilliam, Ph.D., University of Chicago, Chicago, IL;

Clara Lajonchere, Ph.D, Cure Autism Now, Los Angeles, CA;

David H. Ledbetter, Ph.D., Emory University, Atlanta, GA;

Christa Lese-Martin, Ph.D., Emory University, Atlanta, GA;

Janet Miller, J.D., Ph.D., Cure Autism Now, Los Angeles, CA;

Stanley F. Nelson, M.D., UCLA School of Medicine, Los Angeles, CA;

Gerard D. Schellenberg, Ph.D., University of Washington, Seattle, WA;

Carol A. Samango-Sprouse, Ed.D., George Washington University, Washington, D.C.;

Sarah Spence, M.D., Ph.D., UCLA, Los Angeles, CA;

Matthew State, M.D., Ph.D., Yale University, New Haven, CT.

Rudolph E. Tanzi, Ph.D., Massachusetts General Hospital, Boston, MA.

We also thank Dr. Vlad Kustanovich of AGRE for providing us with a comprehensive list of reported genetic markers for autism and Dr. Timothy McCaffrey for helpful discussions. This study was supported by a pilot grant from the Catharine Birch McCormick Genomics Center (VWH).

Author information



Corresponding author

Correspondence to Valerie W Hu.

Additional information

Authors' contributions

VWH conceived of the study and experimental design using monozygotic twins discordant in autism. She also carried out the sample preparations, DNA microarray hybridizations, data analyses, and manuscript preparation. JQ and NHL provided helpful advice with regard to microarray data analyses and also contributed to manuscript preparation. BCF and SH were responsible for the confirmatory qRT-PCR analyses. All authors read and approved the final manuscript.

Electronic supplementary material

Additional File 1: Principal components analysis of microarray data from the 5 sets of monozygotic twins with ASD, with each color representing a separate pair of twins. This figure shows that genotype is a major contributor to variations in overall gene expression profile. Each point on the graph represents a dye-reversal experiment on a given twin pair. Note that even the 2 pairs of twins who share the same mother but have different fathers (pink and yellow points) are distinguishable from each other. (JPEG 4 KB)

A representative gene network showing overlap of some neurologically relevant genes among 3 discordant autistic twin sets using Ingenuity Pathways Analysis software

Additional File 2: . Genes shown in yellow represent overlap of differentially expressed genes in 2 or more sets of twins, whereas the red and green nodes correspond to genes that are up- or down-regulated, respectively, in only 1 twin set. The expression cutoff was set at log2(ratio) = ± 0.58 for each twin set. The 12 genes marked by "#" are known to be involved in nervous system development and function. (JPEG 76 KB)

Additional File 3: Case description of subjects from whom LCL were derived and used in this study. (Self-explanatory) (PDF 21 KB)

Differentially expressed genes within or across twin sets mapped within or close to autism candidate genes or quantitative trait loci

Additional File 4: . This table shows that many of the differentially expressed genes map in silico to autism susceptibility loci or quantitative trait loci identified by genetic analyses. (PDF 25 KB)

Additional File 5: Primers used for quantitative RT-PCR analyses. (Self-explanatory) (PDF 12 KB)

Authors’ original submitted files for images

Below are the links to the authors’ original submitted files for images.

Authors’ original file for figure 1

Authors’ original file for figure 2

Rights and permissions

This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Reprints and Permissions

About this article

Cite this article

Hu, V.W., Frank, B.C., Heine, S. et al. Gene expression profiling of lymphoblastoid cell lines from monozygotic twins discordant in severity of autism reveals differential regulation of neurologically relevant genes. BMC Genomics 7, 118 (2006).

Download citation


  • Quantitative Trait Locus
  • Autism Spectrum Disorder
  • Autism Spectrum Disorder
  • Twin Pair
  • Monozygotic Twin