Species-Specific Differential AhR Expression Protects Human Neural Progenitor Cells against Developmental Neurotoxicity of PAHs

Background Because of their lipophilicity, persistent organic pollutants (POPs) cross the human placenta, possibly affecting central nervous system development. Most POPs are known aryl hydrocarbon receptor (AhR) ligands and activators of AhR signaling. Therefore, AhR activation has been suggested to cause developmental neurotoxicity (DNT). Objective We studied the effects of AhR ligands on basic processes of brain development in two comparative in vitro systems to determine whether AhR-activation is the underlying mechanism for reported DNT of POPs in humans. Methods We employed neurosphere cultures based on human neural progenitor cells (hNPCs) and wild-type and AhR-deficient mouse NPCs (mNPCs) and studied the effects of different AhR agonists [3-methylcholanthrene (3-MC), benzo(a)pyrene [B(a)P], and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)] and an antagonist [3′-methoxy-4′-nitroflavone (MNF)] on neurosphere development. Moreover, we analyzed expression of AhR and genes involved in AhR signaling. Results In contrast to wild-type mNPCs, hNPCs and AhR-deficient mNPCs were insensitive to AhR agonism or antagonism. Although AhR modulation attenuated wild-type mNPC proliferation and migration, hNPCs and AhR-deficient mNPCs remained unaffected. Results also suggest that species-specific differences resulted from nonfunctional AhR signaling in hNPCs. Conclusion Our findings suggest that in contrast to wild-type mNPCs, hNPCs were protected against polycyclic aromatic hydrocarbon–induced DNT because of an absence of AhR. This difference may contribute to species-specific differences in sensitivity to POPs.

Persistent organic pollutants (POPs) bio accumulate through the food chain and may cause adverse effects on human health and the environment. Main substance classes are poly cyclic aromatic hydrocarbons (PAHs), such as 3methylcholanthrene (3MC) and benzo(a) pyrene [B(a)P], dioxins such as 2,3,7,8tet rachlorodibenzopdioxin (TCDD), and polychlorinated biphenyls (PCBs). Because of their lipophilicity, POPs cross the human placenta, exposing the fetus to the contami nant body burden of the mother. This may result in adverse health outcomes, includ ing effects on central nervous system (CNS) develop ment (reviewed by Wormley et al. 2004). One potential source for in utero expo sure to POPs, especially PAHs, is maternal cigarette smoking during pregnancy (Polanska et al. 2009;Rodgman et al. 2000). An inverse relationship between maternal smoking dur ing pregnancy, offspring intelligence (intel ligence quotient), and cognitive ability has been described in numerous cohort studies (Batty et al. 2006;Fried et al. 1998;Olds et al. 1994). Additionally, behavioral problems and psychiatric disorders in offspring have been associated with smoking during pregnancy (Fergusson et al. 1998;Thapar et al. 2003;Weissman et al. 1999). In addition to smok ing, in utero exposure to PCBs and/or diox ins has been associated with cognitive deficits in children (reviewed by Grandjean and Landrigan 2006;Jacobson and Jacobson 1997;Patandin et al. 1999). It has been proposed that etiological mechanisms involve the POP activated aryl hydrocarbon receptor (AhR), an evolutionary highly conserved member of the basic helixloophelix/PerARNTSim (bHLH/PAS) family of transcription factors (reviewed by Kakeyama and Tohyama 2003). Ligand binding to this cytosolic receptor induces nuclear translocation (Gasiewicz and Bauman 1987) and hetero dimerization with another bHLH/PAS protein, ARNT (AhR nuclear translocator) (Reyes et al. 1992). The resulting complex recognizes and binds spe cific DNA sequences (i.e., dioxinresponsive elements) within gene promoter regions and modulates subsequent transcription of AhR dependent genes (FujisawaSehara et al. 1987). Many POPs, including PAHs and dioxins, are known AhR ligands and activators of AhR signaling. It has therefore been proposed that AhR activation causes develop mental neuro toxicity (DNT) (Williamson et al. 2005). This hypothesis is supported by studies in inverte brate (Caenorhabditis elegans) and vertebrate (zebrafish, chicken, rat, monkey) species in which dioxins and related compounds cause morphological abnormalities of the brain or deficits in cognition and/or behavior (Henshel et al. 1997;Hill et al. 2003; reviewed by Kakeyama and Tohyama 2003;Qin and PowellCoffman 2004). Whether AhR activation is the underlying mechanism for the reported DNT effects after POP exposure in humans is not known.
To investigate potential speciesspecific dif ferences, we employed comparative in vitro test systems for brain development based on neurosphere cultures from human and mouse neural progenitor cells (hNPCs and mNPCs, respectively). These threedimensional cell systems mirror basic processes of fetal brain development such as proliferation, migration, differentiation, and apoptosis. Moreover, they detect developmental neurotoxicants in vitro (Fritsche et al. 2005;Moors et al. 2007Moors et al. , 2009. We report here that, in contrast to wildtype mNPCs, hNPCs and AhRdeficient mNPCs were insensitive to AhR agonism or antagonism due to non functional AhR signaling, which suggests that humans are protected from AhR dependent, POPinduced DNT. Knowledge about such speciesspecific differences is of utmost importance regarding chemi cal testing and hazard assessment for humans.
Cell culture. hNPCs used in this study were purchased from Lonza Verviers SPRL (Verviers, Belgium). Data presented in this study are based on experiments conducted Background: Because of their lipophilicity, persistent organic pollutants (POPs) cross the human placenta, possibly affecting central nervous system development. Most POPs are known aryl hydrocarbon receptor (AhR) ligands and activators of AhR signaling. Therefore, AhR activation has been suggested to cause develop mental neurotoxicity (DNT). oBjective: We studied the effects of AhR ligands on basic processes of brain development in two comparative in vitro systems to determine whether AhR-activation is the underlying mechanism for reported DNT of POPs in humans. Methods: We employed neurosphere cultures based on human neural progenitor cells (hNPCs) and wild-type and AhR-deficient mouse NPCs (mNPCs) and studied the effects of different AhR agonists [3-methylcholanthrene (3-MC), benzo(a)pyrene [B(a)P], and 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD)] and an antagonist [3´-methoxy-4´-nitroflavone (MNF)] on neurosphere development. Moreover, we analyzed expression of AhR and genes involved in AhR signaling. results: In contrast to wild-type mNPCs, hNPCs and AhR-deficient mNPCs were insensitive to AhR agonism or antagonism. Although AhR modulation attenuated wild-type mNPC proliferation and migration, hNPCs and AhR-deficient mNPCs remained unaffected. Results also suggest that species-specific differences resulted from non functional AhR signaling in hNPCs. conclusion: Our findings suggest that in contrast to wild-type mNPCs, hNPCs were protected against polycyclic aromatic hydrocarbon-induced DNT because of an absence of AhR. This difference may contribute to species-specific differences in sensitivity to POPs.  (Bard et al. 1998). Brains of embryos were dissected, transferred to Dulbecco's modi fied Eagle's medium (DMEM) and mechani cally dissociated. Trypsin/EDTA solution was added, and the suspension was incubated for 30 min at 37°C in a humidified atmosphere. Afterward, the tissue suspension was tritu rated to obtain a singlecell suspension, which was centrifuged at 800 rpm for 5 min. Pellets were resuspended and plated in 10cm petri dishes. AhR deficiency of mice ) was confirmed by polymerase chain reaction (PCR). Data for wildtype mNPC were derived from four independent prepara tions, and data for AhRdeficient mNPCs were obtained from two different prepara tions. The animals were treated humanely and with regard for alleviation of suffering.
Cell viability. We measured cell viability using a lactate dehydrogenase (LDH) assay (CytoToxOne; Promega, Madison, WI, USA) according to the manufacturer's instruc tions. Briefly, supernatants of treated cells from the migration, mRNA expression, and proliferation assessments were collected at the respective time points and incubated 2:1 with the CytoToxOne reagent for 4 hr prior to detection of fluorescence (excitation, 540 nm; emission, 590 nm). Complete lysis of cells with the included lysis buffer for 2 hr at room temperature served as a positive control.
Proliferation assays, cell cycle analyses, and migration assay. Proliferation was assessed with a combination of CellTiterBlue Assay (Promega), which measures mitochondrial reductase activity, and microscopic determina tion of sphere diameter as described previously (by Moors et al. 2009). We analyzed cell cycle by fluorescenceactivated cell sorting (FACS). Therefore, neurospheres were exposed to the different chemicals for 48 hr, dis sociated to single cells with Accutase, fixed with paraform aldehyde, and stained with propidium iodine (Moors et al. 2009). Spheres cultured in pro liferation medium with 20 ng/mL EGF and FGF served as positive controls, and spheres cultured without these growth factors served as negative controls. Migration distance was measured microscopically after 48 hr as previ ously described by Moors et al. (2007Moors et al. ( , 2009). Spheres exposed to 1 µM MeHgCl for 48 hr served as positive controls.
Quantitative real-time PCR. NPCs were treated under differentiating conditions with 10 µM 3MC or 1 nM TCDD, or with 0.1% dimethyl sulfoxide (DMSO) as the solvent control. After indicated time points, RNA was prepared with the Absolutely RNA Microprep Kit (Stratagene, La Jolla, CA, USA) accord ing to the manufacturer's instructions. Real time reversetranscriptase PCR (RTPCR) was performed using LightCycler instru mentation (Roche, Mannheim, Germany) with QuantiTect SYBR green PCR Master Mix (Qiagen, Hilden, Germany) as pre viously described ). Conditions for PCR amplifications were as follows: initial 15 min at 95°C; 40 cycles of 15 sec at 94°C for denaturation, 25 sec of primer annealing, 30 sec at 72°C for elon gation; and 2 sec at 72°C for fluorescence detection. Intronspanning primers were designed using PRIMER 3 Software (Rozen and Skaletsky 2000) to reduce genomic amplification. Primer sequences for human ACTB (hACTB), hAhR, hAhR repressor (hAhRR), human cytochrome P450 1A1 (hCYP1A1), and hCYP1B1 were described by Fritsche et al. (2005). Other primer sequences are as follows: hARNT, forward (F), CCCTAGTCTCACCAATCGTGGAT; reverse (R), GTAGCTGT TGCTCT G A T C T C C C A G ; h C -M Y C , F , ACCAC CAGCAGCGACTCTGA; R, TCCAGCAGAAGGTGATCCAGACT (56°C annealing); mouse Actb (msActb), F, CTACAATGAGCTGCGTGTGG; R , TAGCTCTTCTCCAGGGAGGA ( 6 0°C a n n e a l i n g ) ; m s A h R , F , GACAGTTTTCCGGCTTCTTG; R, C G C T T C T G T A A A T G C T C T C G T ( 6 0°C a n n e a l i n g ) ; m s A r n t , F ,

T G C C T C A T C T G G T A C T G C T G ; R , G A A C A T G C T G C T C A C T G G A A
( 5 8°C a n n e a l i n g ) ; m s A h R R , F , ( 6 0°C a n n e a l i n g ) ; m s C y p 1 a 1 , F, G G C C A C T T T G A C C C T T A C A A ; R , CAGGTAACGGAGGACAGGAA ( 5 4°C a n n e a l i n g ) ; m s C y p 1 b 1 , F,

A C A T G A G T T T C A G T T A T G G C C ; R , T T C C A T T C A C T G C T G A G A G C
( 5 8°C a n n e a l i n g ) ; m s c -m y c , F , T G T C C A T T C A A G C A G A C G ; R , GCATTTTAATTCCAGCGCATAG (54°C annealing). Expression levels were normal ized to the expression of βactin. We evalu ated gene expression using the cycle threshold (Ct) value from each sample. Calculations are based on the ΔΔCt method (Livak and Schmittgen 2001). For determining abso lute copy numbers, we used productspecific standards amplified from cDNA to generate standard curves. A target gene was consid ered to be quantifiable if the ratio of the copy number target gene to the copy number for βactin × 1,000 exceeded 0.001.
Western blotting. Proteins were isolated from hNPC or HepG2 human hepato cellular liver carcinoma cells as described by Moors et al. (2007). Wholecell lysates (40 µg) were separated by 10% SDSPAGE and transferred to an Amersham HybondPmembrane (GE Healthcare, Chalfont St. Giles, UK). The membrane was blocked in Trisbuffered saline containing 0.01% Triton (vol/vol) and 5% non fat dry milk (wt/vol) for 1 hr at room temperature, followed by incubation in the same buffer containing a polyclonal anti AhR antibody (1:2,000; Enzo Life Sciences, Loerrach, Germany) or an antiGAPDH anti body (1:100,000; Biozol, Eching, Germany) overnight at 4°C. As secondary antibodies we used enhanced chemiluminescence (ECL) anti rabbit IgG or ECL antimouse IgG horseradish peroxidase-linked antibodies (both 1:5,000; GE Healthcare). Chemiluminescence signals were detected with Xray film.
Statistics. All results are presented as mean ± SEM of at least two independent experiments. We used analysis of variance combined with Dunnett's post hoc test for multi factor analyses (concentration effects), and Student's ttest for twogroup compari sons (treatment vs. control). Significance was set at p < 0.05 (Hays 1994).

Results
AhR agonism or antagonism does not cause cytotoxicity in hNPCs and mNPCs. Cells were exposed to the AhR agonists 3MC (10 µM; a synthetic model compound for AhRactivating metabolized PAH), B(a)P (10 µM; a metabo lized, environmentally relevant PAH), TCDD (1 nM; a non metabolized, environmentally relevant PAH), or the AhR antagonist MNF (1 µM; model compound). We treated hNPCs and mNPCs for 6, 12, 24, or 48 hr (differen tiation conditions). hNPCs were also treated for up to 14 days and mNPCs were treated for up to 7 days under proliferation conditions. Exposure times for cytotoxicity assays thereby correspond to the duration of treatments in the proliferation, migration, and gene expres sion assays. We chose concentrations based on known AhR activation/antagonism in other cell types. None of the exposures induced cyto toxicity (LDH release) in hNPCs or mNPCs [see Supplemental Material, Figure 1 (doi:10.1289/ehp.0901545)].
Proliferation of mNPCs, but not hNPCs, is inhibited by AhR blockage. Figure 1A and B shows quantification of the proliferation assays (sphere diameter and CellTiterBlue Assay, respectively) of human neurospheres after exposure to 1 and 10 µM 3MC, 1 nM TCDD, or 1 µM MNF. We found no signifi cant disparities in prolifera tion among cells with different exposures after 7 or 14 days. We verified this using hNPCs derived from a second individual (data not shown). The average diameter of the human neurospheres (excluding the negative controls cultured without growth factors) increased from 444 µm ± 15 µm to 653 µm ± 33 µm (at 7 days) and 829 µm ± 17 µm (at 14 days). The lack of exposure effects on proliferation was confirmed by FACS analyses for DNA content using propidium iodine staining ( Figure 1C). We found a cell cycle distribu tion typical of proliferating cells in approxi mately 2.5% of all NPCs, as described previously (Moors et al. 2009). After 48 hr of exposure, most proliferating cells were in G0/G1phase (53.5% ± 3.4%), whereas 28% ± 1.8% were in Sphase, and 18.4% ± 2.8% cells were in G2/Mphase, independent of exposure conditions.
We monitored proliferation of mNPCs over only 7 days because of a sizerestricted halt in sphere growth beyond that time. In contrast to the hNPCs, we observed significant inhibition of mNPC proliferation after 7 days of exposure to 1 µM MNF. Although con trol spheres and 3MC-and TCDDexposed spheres grew from an average diameter of 386 µm ± 23 µm to 551 µm ± 85 µm, MNF exposed spheres remained at 405 µm ± 78 µm after 7 days. This was due to AhR inhibition, because proliferation of AhRdeficient neuro spheres was not affected by MNF treatment (Figure 1D,E). Effects of MNF on mNPC proliferation were confirmed by FACS analy ses for DNA content using propidium iodine staining ( Figure 1F). As for hNPCs, we found a cell cycle distribution typical of proliferat ing cells in approximately 2.5% of all cells. In the positive control (solvent plus growth factors), 56.1% ± 4.6% of these cells were in G0/G1phase, 38.1% ± 4.3% were in Sphase, and 5.8% ± 0.4% cells were in G2/Mphase after 48 hr of exposure. 3MC-treated cells showed a similar distribution. However, MNF treatment reduced the number of cells in G2/Mphase to 2.0% ± 0.4%, whereas the distribution of cells in G0/G1 and Sphase remained relatively constant at 51.8% ± 0.2% and 46.3% ± 0.4%, respectively.

Neural cell migration is affected by AhR stimulation in mNPCs but not hNPCs.
Next, we investigated whether 3MC, B(a)P, TCDD, or MNF influences hNPC and mNPC migration. We exposed differentiating neurospheres to 1 and 10 µM 3MC, 10 µM B(a)P, 1 nM TCDD, 0.1 µM MNF, or 1 µM MNF for 48 hr; 1 µM MeHgCl served as a positive control for inhibition of migration (Moors et al. 2009). Independent of exposures, all human neurospheres adhered to the PDL/ laminin matrix and migrated radially out of the sphere with an average migration distance of 739 µm ± 61 µm after 48 hr (Figure 2A,C). This was verified using hNPCs from a second individual (data not shown).
In contrast to the results in hNPCs, AhR activation reduced migration distance in wild type mNPCs by 16% ± 5% (1 µM 3MC), 21% ± 13% (10 µM 3MC), and 32% ± 10% [10 µM B(a)P] compared with solvent con trols, whereas migration was comparable with controls after exposure to the AhR antagonist MNF ( Figure 2B,D). This effect can be attrib uted to AhR activation because migration of AhRdeficient neurospheres was not disturbed and WT and AhR-KO mouse (D, E) neuro spheres were measured; results are reported as percentages of corresponding measure ments in PCs. The results for each end point were plotted in separate x,y-diagrams, and the gradient of the linear regression curve was assessed; data represent the mean gradient ± SEM of two (AhR-KO) to four independent experiments (5-6 spheres/exposure). Cell cycle phase distributions (values are shown within or above bars) of dissociated, fixed, and propidium iodine-stained hNPCs (C) and mNPCs (F) analyzed by FACS. Data represent the mean ± SEM of three independent experiments after 48-hr exposure to 10 μM 3-MC or 1 μM MNF. *p < 0.05 compared with PCs.   Figure 2D). Interestingly, TCDD did not disrupt wildtype mNPC migration despite AhR activation ( Figure 2D). This might be due to the fact that, in con trast to 3MC and B(a)P, TCDD is hardly metabolized and thus does not produce reactive intermediates.
AhR-dependent gene transcription. AhR dependent gene transcription is inducible only in mNPCs, and not in hNPCs because of low abundance of AhR and ARNT tran scripts and absence of AhR protein in human cells. Because 3MC, B(a)P, and MNF did not influence hNPC viability, proliferation, or migration but did modulate proliferation or migration of mNPCs, we determined AhR and ARNT mRNA expression after exposure to these compounds in hNPCs and mNPCs under proliferating and differentiating condi tions. AhR and ARNT mRNAs were expressed at very low copy numbers in hNPCs (0.6-2.5 and 13/1,000 copies βactin, respectively) and higher copy numbers in mNPCs (20-63 and 716-1,045/1,000 copies βactin, respec tively) (Table 1). In addition, CYP1A1 expres sion was undetectable in untreated hNPCs. That βactin was in this case valid to use as a house keeping gene was demonstrated by normalization of proliferating versus differ entiating NPCs to three additional house keeping genes [RPL27, RPL30, OAZ1; see Supplemental Material, Figure 2 (doi:10.1289/ ehp.0901545)]. Expression of the AhR target genes AhRR, CYP1A1, CYP1B1, and c-myc was not significantly induced by 10 µM 3MC after 6, 12, 24, and 48 hr of differentiation in hNPCs ( Figure 3A). We obtained comparable results for hNPCs from a second individual after 6 hr of treatment (data not shown). In contrast, 6 hr of exposure to 10 µM 3MC significantly induced Cyp1a1 and Cyp1b1 mRNA in wildtype mNPCs to levels 6.6 ± 1.7 and 2.5 ± 0.25 times higher than con trols, respectively ( Figure 3C). Although 1 nM TCDD did not disturb neural migration, it induced AhR signaling in wildtype mNPCs, increasing Cyp1a1 expression 21 times relative to controls ( Figure 3C, inset).
Comparison of mRNA expression lev els between hNPCs and mNPCs showed that genes belonging to the AhR machin ery and AhRdependent genes were gener ally expressed in higher copy numbers/1,000 copies βactin in mNPCs than in hNPCs ( Table 1). Lack of AhR protein in hNPCs was confirmed by Western blot ( Figure 3B). These results demonstrate that AhR signaling pathway gene products mediate the effects of the AhR agonists 3MC and B(a)P and the AhR antagonist MNF on proliferation and migration in mNPCs, because AhRdeficient mNPCs are protected against these effects ( Figure 3C, inset).

Discussion
The development of cellbased, non animal test ing strategies for hazard assessment of chemicals is currently one of the most important tasks in toxicological research. In this regard, it is most important to choose appropriate model systems that are truly predictive for humans (Krewski et al. 2009;National Research Council 2007). Human tumor cell lines that are easily acces sible in large quantities bear the restriction that they do not represent cellular metabolism and signal transduction of normal cells. In con trast, primary cells are often obtained as ex vivo cultures from rodents. Such primary cultures are regarded as superior to tumorderived cells. However, speciesspecific differences limit their application. One example of how rodent primary cells can indeed mis classify hazards for humans is provided by this study, which shows mousederived primary cells to be more susceptible to AhR modulation than their human counter parts. With regard to chemical testing, it is critical to be aware of such differences in order to avoid over or under estimating hazards that chemicals pose to humans, and thereby protect human health and allow industry production and develop ment of chemicals at the same time.  Specifically, in this study we discovered that proliferation of hNPCs was not affected by AhR agonists (3MC, TCDD) or the AhR antagonist MNF ( Figure 1A-C), whereas pro liferation of wildtype mNPCs was completely blocked by AhR antagonism (Figure 1D-F). Results of the proliferation analysis in hNPCs are in contrast to previous findings obtained in human liver or neuro blastoma tumor cells that showed exogenous AhR activation by TCDD or other AhR ligands to inhibit cell proliferation and induce cell cycle arrest (Jin et al. 2004;Marlowe and Puga 2005). In con trast, high AhR content has been reported to promote proliferation in a human MCF breast cancer cell line that was blocked by the AhR antagonist MNF or selective AhR knockdown via small interfering RNA (Wong et al. 2009). In human umbilical vascular endothelial cells, AhR activation by 3MC also exerted anti proliferative effects, as in the tumor cell lines (Pang et al. 2008). These data demonstrate that effects of AhR activation on cell cycle progression versus cell cycle arrest vary among different cell types, as well as between tumor and non tumor cells.
In contrast to hNPCs, proliferation of mNPCs was completely blocked by MNF ( Figure 1D-F). Reiners et al. (1999) reported that cell proliferation was also inhibited in the murine hepatoma cell line 1c1c7 by fla vone and αnaphthoflavone treatment and in AhRKO cells. Moreover, mouse embryonic fibroblasts from AhRKO mice grew more slowly than did wildtype cells (Elizondo et al. 2000). In contrast, AhRKO animals show accelerated proliferation in different organs, such as skin, hair follicles, and liver blood vessels. Thus, as in humans, AhR effects on proliferation in rodent cells are also cell type dependent, and as our data indicate, cell type specificity is not consistent among different species.
Besides progenitor cell proliferation, migration is another essential process in brain develop ment. To address the role of the AhR in this process, we employed the neurosphere migration assay (Moors et al. 2007). AhR modulation by 3MC, B(a)P, TCDD, or MNF did not affect hNPC migration (Figure 2A,C). These results were in contrast to the findings of two previous studies of human tumor cells that showed increased motility and migra tion of MCF7 cells after TCDD and 3MC treatment based on a scratch assay (Diry et al. 2006) and a transwell migration assay (Seifert et al. 2009). In contrast, we showed that AhR stimulation reduced migration of mNPCs ( Figure 2B,D). This is supported by an earlier in vivo study in which prenatal exposure to the AhR ligand 7,12dimethylbenz[a]anthracene disrupted cerebellar cytoarchitecture in rats (Kellen et al. 1976). Conversely, immortalized mouse mammary fibro blasts from AhRnull mice had decreased migration capacities in culture and inhibition of signaling pathways that regu late cell migration, including focal adhesion kinase and mitogenactivated protein kinase ERK1 (extracellularsignal regulated kinase1) (MuleroNavarro et al. 2005). That ERKdependent pathways are also necessary for normal migration of hNPCs was recently shown by our group (Moors et al. 2007). However, AhR does not seem to determine ERKdependent migration in hNPCs, because migration distance did not change in the presence of AhR modulators (Figure 2A,C) but was impaired by ERK inhibition (Moors et al. 2007). Taken together, existing data on AhRdependent migration imply that, as for prolifera tion, modulation of cell migration dif fers among cell types and species.
To address the underlying reason for the observed speciesspecific differences in response to AhR modulation between hNPCs and mNPCs, we quantified copy numbers of genes that belong to the AhR machinery as well as genes that are AhR regulated. Both hNPCs and mNPCs express AhR, ARNT, AhRR, CYP1B1, and c-Myc ( Figure 3A,C). However, AhR and ARNT copy numbers were close to detection limit in hNPCs (and 8-100 times lower than in mNPCs), and hAhR protein was not detectable by Western blot; consequently, only mNPCs expressed quan tifiable amounts of Cyp1a1 and responded to 3MC treatment with a timerelated induction of Cyp1a1 and Cyp1b1 mRNA (Figure 3). Moreover, low human AhR ligand affin ity, which is approximately 10 times lower than C57/BL6 mouse AhRligand attraction (Harper et al. 1988), has been attributed to an amino acid substitution in the human AhR ligandbinding domain (Ema et al. 1994). This mutation is also responsible for divergent toxic potencies of TCDD between responsive Figure 3. Induction of mRNA expression of the indicated genes in hNPCs (A) and mNPCs (C) analyzed 6, 12, 24, or 48 hr after treatment with 10 μM 3-MC under differentiating conditions. The inset in C shows Cyp1a1 expression 6 hr after 1-nM TCDD or 10-μM 3-MC exposure in wild-type (WT) and AhR-KO mNPCs. Copy numbers of the respective genes were normalized to β-actin after and are expressed as relative induction compared with vehicle control (0.1% DMSO); data are mean ± SEM of two (AhR-KO) to four independent experiments; note the different y-axis scales in A and C. (B) hNPC lysates (40 μg protein/lane) analyzed for AhR expression by Western blot using an AhR-specific antibody and HepG2 cell lysates (10 μg protein/lane) as the positive control. A GAPDH-specific antibody was used as a loading control; a representative example is shown. *p < 0.05 compared with vehicle control. C57/BL6 and non responsive DBA mouse strains (Ema et al. 1994). The irresponsive ness of hNPCs toward AhR modulation was hence likely due to very low AhR expression combined with a human lowaffinity recep tor for ligand binding. Because several poly morphisms have been identified in the human AhR gene, inter individual differences in ligand responsiveness between hNPCs from different donors cannot be excluded. However, even though single AhR polymorphisms (e.g., at codon 517 or 570) found in people of African descent were under suspicion to affect the expression of single genes such as CYP1A1, it is widely accepted that none of the AhR polymorphisms described so far is of any func tional consequence regarding the overall out come of AhR response (Connor and Aylward 2009;Harper et al. 2002).
Species differences similar to those reported here were observed in a compara tive study of human and mouse palate organ cultures (Abbott et al. 1999). This earlier work was driven by the facts that TCDD induces cleft palate in mouse embryos and that the risk for humans to develop such malformations upon in utero exposure to TCDD or related compounds was not known. Abbott et al. (1999) concluded that it seems highly unlikely that human embryos are exposed to sufficient amounts of TCDD to cause interruption of palatal differentiation because human palate organ cultures expressed several hundred times less AhR mRNA than did the mouse cultures, and human palates required 200 times more TCDD to produce a cleft palate in vitro than did the respective mouse model. Our results support those findings in a different organ system, the developing brain, and suggest that studies on AhRdependent DNT in C57/BL6 mice in vivo over estimate the risk of distur bances of human brain development resulting from AhR activation. Our data also imply that the use of toxic equivalency factors, which cor respond to the relative potency of a chemical to generate AhRmediated effects compared with TCDD, based on data derived from rodents for risk assessment of POPs may not necessarily be useful for humans.
How can the epidemiological evidence for POPrelated DNT in humans be explained if the AhR is not involved in chemically induced DNT? Kodavanti (2005) comprehensively reviewed three alternative mechanisms for how POPs might interfere with human brain develop ment. First, POPinduced changes in neuro transmitters such as dopamine or sero tonin could affect learning, memory, and other functions. Because these end points extend beyond the basic processes of brain develop ment, the basic "neurosphere assay" (Breier et al. 2009) cannot detect such changes. Second, PCBs were found to alter intra cellular phospho kinase C (PKC) signaling and Ca 2+ homeo stasis in rodents. Because PKC mod ulation by the inhibitor BisI (bisindolylma leimide I) or the stimulator PMA (phorbol 12myristate 13acetate) causes inhibition and stimulation of hNPC migration, respectively (Moors et al. 2007), it is unlikely that the POPs used in this study affected PKC signaling. Third, effects on thyroid hormone (TH) bal ance might contribute to POPinduced DNT. Some POPs can elicit direct biological effects on target organs such as the brain by inter fering with cellular TH signaling in vivo and in vitro (Fritsche et al. 2005;Schreiber et al. 2010;Zoeller and Crofton 2000). Because the neurosphere assay is able to detect endo crine disruption of TH signaling by PCBs and polybrominated diphenyl ethers (Fritsche et al. 2005;Schreiber et al. 2010), this mecha nism is hence unlikely to be involved in AhR independent DNT of "classical" AhR ligands such as 3MC or TCDD in humans. The most likely mechanism through which POPs interfere with human brain development is via systemic effects on TH homeostasis result ing from direct effects on the thyroid gland and decreased synthesis of TH, reduced blood TH levels secondary to enhanced TH metabo lism, or displacement of natural ligand (thy roxine) binding to the TH plasma transport protein transthyretin. Evidence that TCDD also affects systemic TH balance in humans was reported by Baccarelli et al. (2008), who found that maternal exposure in the highly TCDDexposed Seveso cohort was associ ated with effects on neo natal thyroid function (Baccarelli et al. 2008). However, we cannot exclude the possibility that AhR expression in fetal brains increases with gestational age, resulting in AhRmediated toxicity of POPs in NPCs at later gestational stages.
In summary, we show that, in contrast to mNPCs, hNPCs are protected against PAH induced DNT and that this difference may be explained by the absence of AhR in hNPCs. An accumulating body of evidence now indi cates that human AhR signaling is less opera tive than AhR function in most laboratory animals. This knowledge should be taken into account for risk assessment of TCDD and related xenobiotics in humans.