Skip to main content

Psychoactive pharmaceuticals at environmental concentrations induce in vitro gene expression associated with neurological disorders



A number of researchers have speculated that neurological disorders are mostly due to the interaction of common susceptibility genes with environmental, epigenetic and stochastic factors. Genetic factors such as mutations, insertions, deletions and copy number variations (CNVs) are responsible for only a small subset of cases, suggesting unknown environmental contaminants play a role in triggering neurological disorders like idiopathic autism. Psychoactive pharmaceuticals have been considered as potential environmental contaminants as they are detected in the drinking water at very low concentrations. Preliminary studies in our laboratory identified gene sets associated with neuronal systems and human neurological disorders that were significantly enriched after treating fish brains with psychoactive pharmaceuticals at environmental concentrations. These gene expression inductions were associated with changes in fish behavior. Here, we tested the hypothesis that similar treatments would alter in vitro gene expression associated with neurological disorders (including autism) in human neuronal cells. We differentiated and treated human SK-N-SH neuroblastoma cells with a mixture (fluoxetine, carbamazepine and venlafaxine) and valproate (used as a positive control to induce autism-associated profiles), followed by transcriptome analysis with RNA-Seq approach.


We found that psychoactive pharmaceuticals and valproate significantly altered neuronal gene sets associated with human neurological disorders (including autism-associated sets). Moreover, we observed that altered expression profiles in human cells were similar to gene expression profiles previously identified in fish brains.


Psychoactive pharmaceuticals at environmental concentrations altered in vitro gene expression profiles of neuronal growth, development and regulation. These expression patterns were associated with potential neurological disorders including autism, suggested psychoactive pharmaceuticals at environmental concentrations might mimic, aggravate, or induce neurological disorders.


There are approximately 3000 synthetic chemicals known to interact with humans through air, water, and food [1, 2]. These chemicals might serve as environmental contaminants and trigger neurological disorders, such as autism spectrum disorders (ASD), in genetically susceptible individuals [35]. Among this diverse group of environmental contaminants, we focused on pharmaceuticals and personal care products (PPCPs) due to their tendency to contaminate environmental water systems [2]. PPCPs include psychoactive pharmaceuticals that are highly prescribed in the United States and other chemicals like bis-phenol A (BPA) in plastics, phthalates in cosmetics and teratogenic chemicals [1].

Psychoactive pharmaceuticals like fluoxetine, venlafaxine and carbamazepine, have been detected in the drinking water at very low concentrations [2, 6, 7]. These pharmaceuticals, some of which are metabolically active and have relatively long half-lives for over a month, reach waste-water treatment plants (WWTP) through excretion by clinical patients [7, 8]. Due to the chemical properties of these drugs and inefficient filtration of WWTP, these drugs end up mixing up with the ground water, and thus reach drinking water at low concentrations [7, 8].

We previously hypothesized that psychoactive pharmaceuticals as environmental contaminants alter neuronal gene expression associated with neurological disorders like ASD. To determine this, our lab treated juvenile fathead minnows (Pimephales promelas) with psychoactive pharmaceuticals (fluoxetine, venlafaxine and carbamazepine) individually and in mixtures at environmentally relevant concentrations [2, 6]. After treating them for 15 days, we extracted the brains and carried out microarray analysis. Using gene set enrichment analysis (GSEA) [9], we identified enrichment (up- or down-regulation) of gene sets associated with neuronal growth, regulation and development in the juvenile minnow brains in response to psychoactive drug exposure [2, 6]. We also identified altered neuronal gene sets associated with neurological disorders, including ASD [6]. Moreover, fish exposed to psychoactive pharmaceuticals exposed had an altered behavioral phenotype [2, 10].

In the present study, we hypothesized that psychoactive pharmaceuticals (fluoxetine, venlafaxine and carbamazepine) at environmental concentrations would alter in vitro human neuronal gene expression that is 1) similar to gene expression profile in fathead minnow, and 2) associated with neuro-developmental disorders. To determine if altered gene expression profile was associated with idiopathic autism, we also treated human neuronal cells with valproic acid, which is known to induce autism-like phenotypes in mice [11]. Identifying altered gene expression profiles in treated cells and finding a similar pattern with valproic acid induced gene expression would reveal the extent to which psychoactive pharmaceuticals at very low concentrations could induce gene expression associated with potential neurological disorders like ASD.


Cell culture and differentiation

Human SK-N-SH cell line was obtained from American Type Culture Collection (ATCC #HTB-11). Cells were cultured in Eagle’s Minimum Essential Medium (EMEM; ATCC). This media was supplemented with 1 % penicillin-streptomycin-neomycin (Sigma) and 10 % (v/v) fetal bovine serum (FBS). Retinoic acid (RA; Sigma) was used to induce SK-N-SH cells [12] to differentiate into more neuron-like cells [13] because this cell line is a mixture of different cell types. Approximately 15,000 cells were cultured in T-75 flask (Corning) supplemented with EMEM media for two days, and then Retinoic acid (10 μM) was added. Cells were treated with RA for two weeks and media was replaced every three-four days [12]. Cultures were monitored visually using light microscopy for morphological changes, and evaluated for neuronal cell markers (NeuN, PSD95 and NCAM) during the differentiation process [12, 14].

Pharmaceuticals treatments

Stock solutions (10 mM) of fluoxetine (FLX; Sigma F133; Active metabolite), venlafaxine (VNX; Sigma D2069; ​Active metabolite) and carbamazepine (CBZ; Sigma C4206; ​Active metabolite), and 1 mM stock solution of valproic acid (VPA; Santa Cruz Biotechnology sc211393; Active metabolite) were prepared in dimethyl sulfoxide (DMSO). After differentiation for two weeks, cells were treated with a mixture composed of (MIX: FLX 10 μg/l; VNX 50 μg/l; CBZ 100 μg/l) [2], and Valproate (VPA: 4.9 mg/l) [15]. Control cells were treated with DMSO (vehicle) only, and the final concentration of DMSO in the cultures was 0.05 %. To examine if psychoactive pharmaceuticals in drinking water at environmental concentrations could induce neuronal gene expression, we chose to treat neuronal cells with the mixture of three pharmaceuticals (fluoxetine, carbamazepine and venlafaxine). Cells were treated in three replicates with the pharmaceuticals for 48 h in EMEM media without FBS to avoid any binding of pharmaceuticals with the serum proteins. All of the treatments were shown not to affect overall cell viability with respect to control (no treatment), based on the adherent nature of the monolayers and the result from CyQuant Viability Assay [16]. After treating cells for 48 h, they were collected with Versene solution (Gibco).

RNA extraction, cDNA synthesis and sequencing

After 48 h of exposure, cells were collected with trypsin, centrifuged, and RNA was extracted using Qiagen RNeasy Plus Mini Kit (74134) according to the manufacturer’s protocol. Total RNA concentration was determined using NanoDrop (Thermo Scientific), and the RNA integrity value (RIN) was analyzed on Agilent 2100 Bioanalyzer (Agilent, Santa Clara, CA). RNA was quantified with Qubit spectrophotometer (Life technologies). Using Illumina Tru-seq stranded total RNA kit, cDNA library was prepared in following steps: RiboZero depletion and fragmentation, first and second strand cDNA synthesis, adenylation of 3′ ends, adapter ligation, PCR amplification, library validation using qPCR – Kapa Biosystems Library Quantification Kit, and normalization and pooling in preparation for cluster generation on MiSeq. Samples were then loaded onto flow cell (MiSeq Reagent kit v3 150 cycle) and sequenced on Illumina MiSeq 2 according to the manufacturer’s instructions.

Bioinformatics data analysis

Quality control, alignment, and read counting

In total, three treatments (mixture, valproate, and control) with three replicates each were sequenced with five flow cells using Illumina MiSeq, which generated more than 10 M paired-end reads for each replicate (refer Additional file 1). The raw sequences in FASTQ files underwent quality control analysis using FastQC ( All the samples were sequenced at 75b read length and they all passed the quality check. We aligned the quality checked reads to human genome hg19 using TopHat version 2.0.11 [17]. Most of the reads (~90 %) had average base quality above Q30. The reference genome sequence, the bowtie index files for the reference genome sequence, and the gene annotation file were downloaded from Illumina iGenomes project ( The number of reads that map to human known genes were counted by summerizeOverlaps function in GenomicAlignments R package [18]. Only the genes that have at least one read for all replicates were retained for downstream analysis.

Analysis of differential gene expression

Differentially expressed genes were identified using DESeq2 version 1.6.1 [19]. Differentially expressed genes from 2 groups (Mixture and Valproate) with respect to the control treatment were identified.

Gene set enrichment analysis

We used GAGE version 2.14.4 [20] for gene set enrichment analysis. We used GAGE to analyze gene sets that were analyzed previously in Fish brains [2, 6]. These gene sets are grouped as Neuronal development, regulation, growth; Neurological Disorders (ND); Autism spectrum disorders (ASD) [2, 6]. For extensive analysis, we also analyzed gene sets from MSigDB ‘C2’ and ‘C5’ (refer Additional file 2). Enrichment analyses were carried out using gage function using non-parametric Kolmogorov-Smirnov tests. We used GAGE package to identify significantly enriched (significantly up- and/or down-regulated; P-value < 0.01 and Q-value < 0.1) gene sets within mixture and valproate treatments.


Patterns of psychoactive pharmaceuticals – induced in vitro gene expression in neuronal development, growth and regulation

We postulated that psychoactive pharmaceuticals at environmental concentrations would alter gene expression of neuronal systems. Support for this hypothesis would suggest that dysregulation of neuronal systems would result in altered neuronal circuits and may result in fewer neuronal connections. To address this question we used differentiated SK-N-SH neuroblastoma cells and treated them with the mixture (MIX: FLX-10 μg/l; VNX-50 μg/l; CBZ-100 μg/l), and valproate (VPA: 4.9 mg/l) in replicates of three samples for each treatment.

For the development collection, we observed six enriched gene sets (significantly down-regulated, P-value < 0.01, Q-value < 0.1) by MIX treatment. VPA treatment enriched the expression of seven gene sets (four gene sets down- and three gene sets up-regulated, Additional file 3). AXONOGENESIS, REGULATION OF NEUROGENESIS and SYNAPSE PART gene sets were enriched in both VPA and MIX treatments. And, SYNAPSE PART gene set was also enriched (up-regulated) in fish brains.

For the regulation collection, we observed 14 enriched gene sets (13 gene sets down- and one gene set up-regulated, P-value < 0.01, Q-value < 0.1) by MIX treatment. VPA treatment enriched the expression of eight gene sets (up-regulated, Additional file 3). All eight up-regulated gene sets in VPA treatment were down-regulated in MIX treatment. Among those sets, two gene sets (NEUROTRANSMITTER_BINDING and SYNAPSE) were also enriched in fish brains.

For the growth collection, we observed eight enriched gene sets (all down-regulated, P-value < 0.01, Q-value < 0.1) by MIX treatment. VPA treatment enriched the expression of five gene sets (up-regulated, Additional file 3). All five up-regulated gene sets in VPA treatment were down-regulated in MIX treatment. Two gene sets (AXON and NEURON_PROJECTION) were also enriched in fish brains.

Patterns of psychoactive pharmaceuticals – induced in vitro gene expression in neurological disorders (ND) and ASD groups

We then sought to determine if altered in vitro gene expression was associated with neurological disorders and ASD. To accomplish this, we analyzed and compared the gene expression of already published ND and ASD gene sets in MIX and VPA treatments.

For the ND collection, we observed six enriched gene sets (four down- and two up-regulated, P-value < 0.01, Q-value < 0.1) by MIX treatment. VPA treatment enriched the expression of three gene sets (two gene sets down- and one gene set up-regulated, Additional file 3). AUTISM_IDIOPATHIC and PARKINSONS gene sets were enriched in both VPA and MIX treatments as well as in fish brains.

For the ASD collection, we observed two enriched gene sets (both up-regulated, P-value < 0.01, Q-value < 0.1) by MIX treatment. VPA treatment enriched the expression of five gene sets (down-regulated, Additional file 3). ASD_MILD gene set was enriched in both VPA and MIX treatments as well as in fish brains.

Ranked gene lists from the mixture and valproate treatment

We sorted all genes from gene expression profiles in human neuronal cells treated with the MIX (FLX, VNX, CBZ) and valproate (VPA). In each treatment, we sequenced ~ 17,000 genes and analyzed their expression. We then ranked them based on their fold change expression within each treatment. Tables 1 and 2 show, the 50 most strongly up-and down-regulated genes. We also reported the fold-change score of genes which were presented on the fish microarray chip.

Table 1 Ranked gene list from the mixture treatment (50 most up- and down-regulated)
Table 2 Ranked gene list from the valproate treatment (50 most up- and down-regulated)


Comparison between human MIX treatment and fish gene expression patterns

The results partially supported our first hypothesis that the MIX treatment on human SK-N-SH cell line induced gene sets enrichment patterns similar to that of the fish experiment following mixture treatment, although the degree of similarity was not high.

Among the neural circuit development gene sets, all the six significantly enriched gene sets in human neuronal cells following MIX treatment were enriched in a down-regulated manner; whereas in the fish experiment, the two significantly enriched sets were both up-regulated. We do not know why the direction was opposite, but we do notice that, one gene set, SYNAPSE PART, was enriched in both treatments [2]. Previous studies have found that altered expression of NCAM, IRX3 and NKX6.1 genes in SYNAPSE PART change the fate and position of neurons generated in the chick neural tube [21, 22]. Also within those down-regulated gene sets contains gene PSD95 (DLG4) and GABA, which have recently been found to be associated with neurological disorders like autism by altering synaptic assembly [14, 23, 24].

In the growth group, we observed similar patterns where all enriched sets in MIX treatment were down-regulated and all those in the fish experiment were up-regulated. Among those sets, two were enriched in both treatments, AXON and NEURON PROJECTION. Other studies have found that genes within these two gene sets modulate the fate, lineage, and timing of neuronal development by playing a critical role in the formation and maturation of neural circuits [23, 25, 26].

In the regulation group, we observed that NEUROTRANSMITTER BINDING gene set was down-regulated in fish brains as well as in human cells [2]. This could be possible due to the therapeutic effect of fluoxetine (SSRI) in the mixture treatment [2]. Fluoxetine is known to reduce the re-uptake of serotonin by inhibiting monoamine transporters on the pre-synaptic neuronal membrane [27, 28]. Due to the longer availability of neurotransmitters in the synaptic cleft, the expression of serotonin receptors is down-regulated, thus a decrease in neurotransmitter binding [29]. Another gene set SYNAPSE was significantly up-regulated in fish brains [2], but down-regulated in treated human cells. This gene set was responsible for modulating wiring of neuronal circuits by controlling the number of synapse as well as organization of synaptic assembly and specificity [26, 30]. Altered synaptogenesis has been strongly considered as a potential mechanism in ASD pathogenesis [31, 32].

Comparison between human MIX and VPA gene expression patterns

We used valproate (an anticonvulsant) to treat human neuronal cells as a positive control because prenatal exposure of valproate has been found to be strongly associated with autism [3] and valproate is known to induce autism-like phenotypes in mice [11]. Similarly, carbamazepine (presented in the mixture treatment) is a mood stabilizer and anticonvulsant [11, 33] and it also inhibits the epileptic effects in the brain by blocking sodium channels [11, 33]. By and large, the results support our second hypothesis that the MIX and VPA treatments change the RNA expression profile in similar ways, albeit in an interesting fashion, since the two treatments often enrich the same gene sets but in different directions.

In the development group of VPA treatment, we found four gene sets significantly down-regulated and three gene sets significantly up-regulated in human neuronal cells. Three gene sets (AXONOGENESIS, REGULATION OF NEUROGENESIS and SYNAPSE PART) were enriched in both mixture and valproate treatments, but in the opposite direction. Similar to fish gene expression, VPA treatment up-regulated SYNAPSE PART gene set [2]. This states that VPA exposure might be associated with disturbed neuronal fate and position [21, 22] as well as synaptic assembly [1, 22, 34].

In the growth group, we observed that five gene sets were up-regulated in the VPA treatment. Interestingly, the same five sets were all down-regulated in the MIX treatment. This suggested that both treatments disturbed the human gene expression in similar pathways despite the different directions. One gene set NEURON PROJECTION was up-regulated similar to fish brains [2]. In the regulation group, we noticed similar results. Eight gene sets were up-regulated in valproate treatment, but down-regulated in the mixture treatment. From this repeating phenomenon of opposite direction enrichment of the same gene set, we postulate a general pattern that the three pharmaceuticals (fluoxetine, venlafaxine and carbamazepine) and valproate dysregulate the same pathways in different directions. However, more investigation is needed to confirm or disprove this postulation. We also plotted multi-dimensional scaling using edgeR [35] and observed that mixture treatment samples were somewhat closer to valproate samples than control samples (refer Additional file 1). However, MIX and VPA samples were still apart, confirming that these two treatments did not exhibit similar response.

Association of human MIX and VPA gene expression patterns with neurological disorders

To determine the extent to which altered gene expression in both MIX and VPA treatments were associated with neurological disorders (including ASD), we examined the expression of the already-published gene sets from neurological disorders (ND) and ASD groups in both treatments. We also compared these to their corresponding expression pattern in fish brains. For ND group, MIX and VPA treatments altered PARKINSONS and Autism_Idiopathic gene sets significantly but in different directions. Interestingly, mixture treatment of human cells and fish brains up-regulated Autism_Idiopathic gene set [6]. For another ASD group, MIX treatment up-regulated two gene sets (ASD_2Class and ASD_Mild) similar to fish brains [6]. On the other side, VPA treatment down-regulated three gene sets (ASD_2Class, ASD_Mild, ASD_Savant) in a different direction compared to the expression in fish brains [6]. These expression patterns stated that VPA and MIX treatments of human cells exhibited a similar response to neurological disorders (including ASD), suggesting a common induction effect.

Insights into important genes in both MIX and VPA treatments

We sought to identify important, or novel genes that were significantly up- and down-regulated in human neuronal cells treated with MIX and VPA. We generated ranked lists of genes based on their fold change, and tabulated the most 50 strongly up- and down-regulated genes (Table 1 and 2). We also compared genes from ranked lists with the ones from fish microarray data. For MIX treatment, we noticed four genes (NUPR1, RTL1, THBS1 and HTR2B) which could be considered important and novel. The thrombospondin (THBS1) gene plays an important role in synaptogenesis in the developing brain [36]. Recent association studies have found that both rare and common variants of this gene are associated with autism [36]. Although this gene was found to be down-regulated by two-fold, it was up-regulated in fish brains under similar mixture treatment [2]. Another important gene, HTR2B, which codes for serotonin receptor 2B, was down-regulated by ~ two-fold in human cells. Similar serotonin receptor genes were also found to be down-regulated in fish brains. Moreover, recent protein studies in our lab showed that HTR2B protein was down-regulated in the same mixture (FLX, VNX, CBZ) treatment. This mechanism is explained by the drug effect of fluoxetine (SSRI), which provides more neurotransmitter in the synaptic cleft, thus reducing serotonin receptors [29].

In the ranked list of genes by valproate (VPA) treatment, we found three genes (VSNL1, PTER and OXTR) of particular importance. VSNL1 gene, which encodes for visinin-like protein 1 in humans, modulates neuronal morphology by controlling the key signaling pathways in CNS. This gene was up-regulated by three-fold in the valproate treatment. We observed similar up-regulation of this gene in fish brains following mixture (FLX, VNX, CBZ) treatment. Other studies have recently found the association of single-nucleotide polymorphisms (SNPs) in the VSNL1 gene with neurological disorders like schizophrenia [37]. Another important gene, oxytocin receptor (OXTR) was found to be down-regulated by two-fold in human cells treated with valproate (Table 2). However, this gene was up-regulated in fish brains exposed to the mixture of psychoactive pharmaceuticals [2]. Moreover, recent protein studies in our lab showed an increased expression of OXTR in the same cells treated with carbamazepine and the mixture (FLX, VNX, CBZ). Other studies have shown that OXTR serves as an anxiolytic agent by modulating serotonin release in serotonergic neurons of the raphe nuclei [38].


To investigate the environmental trigger for idiopathic autism, we focused on psychoactive pharmaceuticals, potential environmental contaminants that have been detected in drinking water. We found that psychoactive pharmaceuticals altered the gene expression of neuronal systems in vitro at environmental concentrations. These altered gene expressions are associated with potential neurological disorders by playing a key role in the formation, growth and regulation of neurons. Our data suggests that psychoactive pharmaceuticals might disrupt neuronal connections by altering the gene expression associated with neuronal growth, development and regulation.


  1. 1.

    Landrigan PJ. What causes autism? Exploring the environmental contribution. Curr Opin Pediatr. 2010;22(2):219–25.

    Article  PubMed  Google Scholar 

  2. 2.

    Thomas MA, Joshi PP, Klaper RD. Gene-class analysis of expression patterns induced by psychoactive pharmaceutical exposure in fathead minnow (Pimephales promelas) indicates induction of neuronal systems. Comp Biochem Physiol C-Toxicol Pharmacol. 2012;155(1):109–20.

    CAS  Article  PubMed  Google Scholar 

  3. 3.

    Dufour-Rainfray D, Vourc'h P, Tourlet S, Guilloteau D, Chalon S, Andres CR. Fetal exposure to teratogens: Evidence of genes involved in autism. Neurosci Biobehav Rev. 2011;35(5):1254–65.

    CAS  Article  PubMed  Google Scholar 

  4. 4.

    State MW, Levitt P. The conundrums of understanding genetic risks for autism spectrum disorders. Nat Neurosci. 2011;14(12):1499–506.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  5. 5.

    Kaushik G, Zarbalis KS. Prenatal Neurogenesis in Autism Spectrum Disorders. Front Chem. 2016;4.

  6. 6.

    Thomas MA, Klaper RD: Psychoactive Pharmaceuticals Induce Fish Gene Expression Profiles Associated with Human Idiopathic Autism. Plos One. 2012;7(6).

  7. 7.

    Calisto V, Esteves VI. Psychiatric pharmaceuticals in the environment. Chemosphere. 2009;77(10):1257–74.

    CAS  Article  PubMed  Google Scholar 

  8. 8.

    Kaushik G, Huber DP, Aho K, Finney B, Bearden S, Zarbalis KS, Thomas MA. Maternal exposure to carbamazepine at environmental concentrations can cross intestinal and placental barriers. Biochem Biophys Res Commun. 2016;474(2):291–5.

    CAS  Article  PubMed  Google Scholar 

  9. 9.

    Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, Paulovich A, Pomeroy SL, Golub TR, Lander ES, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A. 2005;102(43):15545–50.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  10. 10.

    Kaushik G, Thomas MA, Aho KA. Psychoactive pharmaceuticals as environmental contaminants may disrupt highly inter-connected nodes in an Autism-associated protein-protein interaction network. BMC Bioinformatics. 2015;16.

  11. 11.

    Harden CL. In Utero Valproate Exposure and Autism: Long Suspected, Finally Proven. Epilepsy Curr. 2013;13(6):282−+.

    Article  PubMed  PubMed Central  Google Scholar 

  12. 12.

    Jain P, Cerone MA, LeBlanc AC, Autexier C. Telomerase and neuronal marker status of differentiated NT2 and SK-N-SH human neuronal cells and primary human neurons. J Neurosci Res. 2007;85(1):83–9.

    CAS  Article  PubMed  Google Scholar 

  13. 13.

    Preis PN, Saya H, Nadasdi L, Hochhaus G, Levin V, Sadee W. Neuronal cell-differentiation of human neuro-blastoma cells by retinoic acid plus herbimycin-A. Cancer Res. 1988;48(22):6530–4.

    CAS  PubMed  Google Scholar 

  14. 14.

    Pizzi M, Boroni F, Bianchetti A, Moraitis C, Sarnico I, Benarese M, Goffi F, Valerio A, Spano P. Expression of functional NR1/NR2B-type NMDA receptors in neuronally differentiated SK-N-SH human cell line. Eur J Neurosci. 2002;16(12):2342–50.

    Article  PubMed  Google Scholar 

  15. 15.

    Kiang TKL, Teng XW, Karagiozov S, Surendradoss J, Chang TKH, Abbott FS. Role of Oxidative Metabolism in the Effect of Valproic Acid on Markers of Cell Viability, Necrosis, and Oxidative Stress in Sandwich-Cultured Rat Hepatocytes. Toxicol Sci. 2010;118(2):501–9.

    CAS  Article  PubMed  Google Scholar 

  16. 16.

    Jones LJ, Gray M, Yue ST, Haugland RP, Singer VL. Sensitive determination of cell number using the CyQUANT (R) cell proliferation assay. J Immunol Methods. 2001;254(1–2):85–98.

    CAS  Article  PubMed  Google Scholar 

  17. 17.

    Kim D, Pertea G, Trapnell C, Pimentel H, Kelley R, Salzberg SL. TopHat2: accurate alignment of transcriptomes in the presence of insertions, deletions and gene fusions. Genome Biol. 2013;14(4).

  18. 18.

    Team RC. R: A Language and Environment for Statistical Computing. In.; 2012: R Foundation for Statistical Computing.

  19. 19.

    Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-Seq data with DESeq2. 2014. In. bioRxiv.

    Google Scholar 

  20. 20.

    Luo W, Friedman MS, Shedden K, Hankenson KD, Woolf PJ. GAGE: generally applicable gene set enrichment for pathway analysis. BMC Bioinformatics. 2009;10.

  21. 21.

    Briscoe J, Pierani A, Jessell TM, Ericson J. A homeodomain protein code specifies progenitor cell identity and neuronal fate in the ventral neural tube. Cell. 2000;101(4):435–45.

    CAS  Article  PubMed  Google Scholar 

  22. 22.

    Jessell TM. Neuronal specification in the spinal cord: Inductive signals and transcriptional codes. Nat Rev Genet. 2000;1(1):20–9.

    CAS  Article  PubMed  Google Scholar 

  23. 23.

    Toro R, Konyukh M, Delorme R, Leblond C, Chaste P, Fauchereau F, Coleman M, Leboyer M, Gillberg C, Bourgeron T. Key role for gene dosage and synaptic homeostasis in autism spectrum disorders. Trends Genet. 2010;26(8):363–72.

    CAS  Article  PubMed  Google Scholar 

  24. 24.

    Ghosh A, Michalon A, Lindemann L, Fontoura P, Santarelli L. Drug discovery for autism spectrum disorder: challenges and opportunities. Nat Rev Drug Discov. 2013;12(10):777–90.

    CAS  Article  PubMed  Google Scholar 

  25. 25.

    Benson DL, Colman DR, Huntley GW. Molecules, maps and synapse specificity. Nat Rev Neurosci. 2001;2(12):899–909.

    CAS  Article  PubMed  Google Scholar 

  26. 26.

    Goda Y, Davis GW. Mechanisms of synapse assembly and disassembly. Neuron. 2003;40(2):243–64.

    CAS  Article  PubMed  Google Scholar 

  27. 27.

    Airhart MJ, Lee DH, Wilson TD, Miller BE, Miller MN, Skalko RG. Movement disorders and neurochemical changes in zebrafish larvae after bath exposure to fluoxetine (PROZAC). Neurotoxicol Teratol. 2007;29(6):652–64.

    CAS  Article  PubMed  Google Scholar 

  28. 28.

    Hiemke C, Hartter S. Pharmacokinetics of selective serotonin reuptake inhibitors. Pharmacol Ther. 2000;85(1):11–28.

    CAS  Article  PubMed  Google Scholar 

  29. 29.

    Lopez-Figueroa AL, Norton CS, Lopez-Figueroa MO, Armellini-Dodel D, Burke S, Akil H, Lopez JF, Watson SJ. Serotonin 5-HT1A, 5-HT1B, and 5-HT2A receptor mRNA expression in subjects with major depression, bipolar disorder, and schizophrenia. Biol Psychiatry. 2004;55(3):225–33.

    CAS  Article  PubMed  Google Scholar 

  30. 30.

    Margeta MA, Shen K. Molecular mechanisms of synaptic specificity. Mol Cell Neurosci. 2010;43(3):261–7.

    CAS  Article  PubMed  Google Scholar 

  31. 31.

    Bourgeron T. A synaptic trek to autism. Curr Opin Neurobiol. 2009;19(2):231–4.

    CAS  Article  PubMed  Google Scholar 

  32. 32.

    Baudouin SJ, Gaudias J, Gerharz S, Hatstatt L, Zhou K, Punnakkal P, Tanaka KF, Spooren W, Hen R, De Zeeuw CI, et al. Shared synaptic pathophysiology in syndromic and nonsyndromic rodent models of autism. Science. 2012;338:128–32.

    CAS  Article  PubMed  Google Scholar 

  33. 33.

    Hough CJ, Irwin RP, Gao XM, Rogawski MA, Chuang DM. Carbamazepine inhibition of N-methyl-D-aspartate-evoked calcium influx in rat cerebellar granule cells. J Pharmacol Exp Ther. 1996;276(1):143–9.

    CAS  PubMed  Google Scholar 

  34. 34.

    Aluru N, Deak KL, Jenny MJ, Hahn ME. Developmental exposure to valproic acid alters the expression of microRNAs involved in neurodevelopment in zebrafish. Neurotoxicol Teratol. 2013;40:46–58.

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  35. 35.

    Cox MAA, Cox TF. Multidimensional Scaling. In: Handbook of Data Visualization. Springer Berlin Heidelberg; 2008: 315–347.

  36. 36.

    Lu L, Guo H, Peng Y, Xun G, Liu Y, Xiong Z, Tian D, Liu Y, Li W, Xu X, et al. Common and rare variants of the THBS1 gene associated with the risk of autism. In.: Psychiatric Genetics; 2014.

  37. 37.

    Braunewell KH, Dwary AD, Richter F, Trappe K, Zhao C, Giegling I, et al. Association of VSNL1 with schizophrenia, frontal cortical function, and biological significance for its gene product as a modulator of cAMP levels and neuronal morphology. Transl Psychiatry. 2011;1.

  38. 38.

    Yoshida M, Takayanagi Y, Inoue K, Kimura T, Young LJ, Onaka T, Nishimori K. Evidence That Oxytocin Exerts Anxiolytic Effects via Oxytocin Receptor Expressed in Serotonergic Neurons in Mice. J Neurosci. 2009;29(7):2259–71.

    CAS  Article  PubMed  Google Scholar 

Download references


The authors thank Late Dr. Chris Cretekos for providing the guidance for using neuronal cells for our study. We also thank Dr. Jean Pfau for providing her laboratory space and reagents for neuronal cell culture.


The publication costs for this article were funded by an Institutional Development Award (IDeA) from the National Institute of General Medical Sciences of the National Institutes of Health under Grant #P20GM103408.

This article has been published as part of BMC Genomics Volume 17 Supplement 3, 2016: Selected articles from the 12th Annual Biotechnology and Bioinformatics Symposium: genomics. The full contents of the supplement are available online at

Availability of data and materials

The RNA-Seq dataset for this research article is available in the Gene Expression Omnibus (GEO) repository, [GSE80635 and].

Authors’ contributions

GK, YX and MAT designed the experiments, and GK performed them. MAT provided the direction and guidance for the research. YX and LY carried out analyses using R-programming and GK wrote the manuscript. All authors have read and approved the final manuscript.

Competing interests

The authors declare that they have no competing interests.

Consent for publication

Not applicable.

Ethics approval and consent to participate

Not applicable.

Author information



Corresponding author

Correspondence to Gaurav Kaushik.

Additional files

Additional file 1:

R-programming code for RNA-Seq analysis and multi-dimensional scaling (MDS) function. The file contains R-code for analysis of RNA-Seq data for mixture and valproate treatments. This file also includes the code for plotMDS function for using multi-dimensional scaling. (DOCX 61 kb)

Additional file 2:

Complete data of gene sets analyzed. Excel file containing a complete data of all gene sets (neuronal development, growth, regulation; neurological disorders and ASD group) that were enriched (up- and down-regulated) in mixture (carbamazepine, fluoxetine and venlafaxine) and valproate treatments. We also reported P- values and Q-values for individual gene sets. (XLSX 1621 kb)

Additional file 3:

Descriptions of gene sets and results of analyses. This file contains table that represents the analyses of gene sets within neuronal systems (a. Development, b. Regulation, c. Growth), (d) Neurological Disorders, and (e) ASD groups. Source indicates the database from where the gene set was derived, Gene Ontology (GO) or Molecular Signatures Database (MSigDB). Size represents the number of genes in each gene set. MIX_UP, MIX_DOWN, VPA_UP and VPA_DOWN indicates gene sets which were found up-regulated in the mixture, down-regulated in the mixture, up-regulated in the valproate, and down-regulated in the valproate treatments, respectively. Gene sets with P-value < 0.01 and Q-value < 0.1 were considered as statistically significantly enriched. For non-significant gene sets, refer Additional file 2 for P-values and Q-values. Enriched gene sets in human cells and corresponding scores are marked in bolds. Enriched gene sets in fish brains are italicized and marked in parentheses. (DOCX 124 kb)

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (, which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver ( applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Kaushik, G., Xia, Y., Yang, L. et al. Psychoactive pharmaceuticals at environmental concentrations induce in vitro gene expression associated with neurological disorders. BMC Genomics 17, 435 (2016).

Download citation


  • Autism Spectrum Disorder
  • Autism Spectrum Disorder
  • Fluoxetine
  • Valproate
  • Venlafaxine