Methylmercury Alters Eph and Ephrin Expression During Neuronal Differentiation of P19 Embryonal Carcinoma Cells D.T

NeuroToxicology 26 (2005) 661–674

Methylmercury Alters Eph and Ephrin Expression During Neuronal Differentiation of P19 Embryonal Carcinoma Cells D.T. Wilson 1, M.A. Polunas 1, R. Zhou 1,2, A.K. Halladay 1,3, H.E. Lowndes 1,3, K.R. Reuhl 1,3,* 1 Joint Graduate Program in Toxicology, Ernest Mario School of Pharmacy, Rutgers University, Piscataway, NJ 08854, USA 2 Department Chemical Biology, Ernest Mario School of Pharmacy, Rutgers University, Piscataway, NJ 08854, USA 3 Department Pharmacology and Toxicology, Ernest Mario School of Pharmacy, Rutgers University, Piscataway, NJ 08854, USA

Received 17 June 2004; accepted 15 January 2005 Available online 28 June 2005

Abstract

Developmental exposure to methylmercury (MeHg) induces a spectrum of neurological impairment characterized by cognitive disturbance, sensory/motor deficit, and diffuse structural abnormalities of the brain. These alterations may arise from neural path-finding errors during brain development, resulting from disturbances in the function of morphoregulatory guidance molecules. The Eph family of tyrosine kinase receptors and their ligands, the ephrins, guide neuronal migration and neurite pathfinding mainly via repulsive intercellular interactions. The present study examined the effects of MeHg on mRNA and protein expression profiles of Ephs and ephrins in the P19 embryonal carcinoma (EC) cell line and its neuronal derivatives. Undifferentiated control P19 cells displayed low- to undetectable levels of mRNA for ephrins or Ephs, with the sole exception of EphA2 which was highly expressed. Upon differentiation into neurons, the ephrin expression increased progressively through day 10. Similarly, expression of the Ephs, including EphsA3, -A4, -A8, -B2, -B3, -B4, and -B6, increased significantly. In contrast, EphA2 expression decreased in day 2, 6 and 10 control neurons. Treatment with MeHg did not affect the expression of mRNA for ephrins or Ephs in undifferentiated P19 cells. However, treatment of differentiating neurons with MeHg for 24 h caused consistent increases in ligand mRNA expression, particularly ephrin-A5, -A6, -B1, and -B2. Similarly, MeHg induced variable increases in mRNA expression of receptors EphA2, -A3, -B3, and -B6. A trend toward a concentration–response relationship was observed for the alterations in Eph receptor mRNA expression although increases at the low and mid concentrations did not reach statistical significance. Immunoblots for ligand and receptor proteins mirrored the increases in the mRNA levels at the 0.5 and 1.5 mM MeHg concentrations but showed decreased protein levels compared to controls at the 3.0 mM concentration. Alterations in the Eph/ephrin family of repulsion molecules may represent an important mechanism in developmental MeHg neurotoxicity. # 2005 Elsevier Inc. All rights reserved.

Keywords: Methylmercury; Embryonal carcinoma; Neurons; Ephrin; Eph; Tyrosine kinase receptors

INTRODUCTION causes defects in brain architecture, many of which may be attributable to perturbation of neuronal migration and Studies in human victims and experimental animals defective formation of post-migratory neural pathways. have shown that pre- and perinatal exposure to MeHg Persistence of fetal cytoarchitecture into postnatal life, formation of neural-glial heterotopias and thinning of * Corresponding author. Tel.: +1 732 445 6909. major commissural ﬁber pathways have been reported in E-mail address: [email protected] (K.R. Reuhl). humans (Matsumoto et al., 1965; Takeuchi, 1968; Choi,

1989; Choi et al., 1978, 1989), nonhuman primates functional impairments (Halladay et al., 2004; Gerlai (Burbacher et al., 1990) and rodents (Barone et al., et al., 1999). 1998; Choi,1986) followingcongenital MeHgexposure. The P19 embryonal carcinoma cell culture model It has been postulated that MeHg causes its effects, at has been widely used to study developmental processes least in part, by disruption of the regulated expression of during neuronal maturation (Graff et al., 1997; McBur- adhesive and repulsive morphoregulatory molecules ney et al., 1988). Following induction with retinoic involved in neuronal migration and neurite outgrowth acid, EC cells differentiate into neurons and glia, during development (Lagunowich et al., 1994; Reuhl recapitulating many of the molecular events occurring et al., 1994). in the developing brain in vivo (Bain et al., 1994; Potential targets of MeHg include the erythropoie- McBurney et al., 1988). During differentiation, cells tin-producing hepatocellular carcinoma cell-derived acquire a neuronal phenotype and begin to express (Eph) tyrosine kinase receptors, whose 14 known classical neuronal markers such as HNK-1, neurotrans- members comprise the largest subfamily of receptor mitters (McBurney et al., 1988) and neurofilament tyrosine kinases (Zhou, 1998). The Ephs are expressed proteins (Pyle et al., 2001). Our laboratory, as well on developing neurons (Flanagan and Vanderhaeghen, as others, has used this cell line to study the effects of 1998; Murai and Pasquale, 2004) and interact with various xenobiotics on neuronal differentiation, neurite their ligands, the ephrins, to guide cell migration and outgrowth, and cytoskeletal behavior (Cadrin et al., neurite extension primarily by repulsive mechanisms 1988; Graff et al., 1997; Wasteneys et al., 1988). (Zhou, 1998). The current study was performed in order to identify The ephrin ligands are divided into two classes, the pattern of Eph and ephrin expression in undiffer- types A and B. The A-ephrins are bound to the cell entiated P19 cells and neurons derived from these cells membrane by a glycophosphatidyl-inositol linkage and to determine the effects of MeHg treatment on the while the B-ephrins have a transmembrane domain expression profiles during neuronal differentiation and and an intracellular tail. The structure of the B-ephrins maturation. not only allows them to trigger signaling cascades in Eph-expressing cells but also to transmit signals back to the ligand-bearing cell. The A-ephrins can also MATERIALS AND METHODS transmit signals back into the ligand-bearing cell via a yet unknown adaptor protein (Knoll and Drescher, Cell Culture Protocol 2002). The Ephs are classified according to sequence homology and specificity of ligand binding. In general, Undifferentiated P19 embryonal carcinoma cells EphAs bind to ephrin-As while EphBs bind to ligands were maintained with slight modification of previously of the B class, but there is considerable promiscuity of described methods (McBurney et al., 1988). Briefly, binding between the subclasses (Gale et al., 1996; undifferentiated EC cells were grown in 10 cm cell Himanen et al., 2004). Recent studies revealed that culture dishes containing alpha-modified Eagle med- the Eph/ephrin family members play key roles in ium (a-MEM) with 10% fetal bovine serum (Gibco axonal guidance and establishment of CNS structure Invitrogen Corp., Grand Island, NY) and 1% antibiotic/ (Chen et al., 2004; Gao et al., 1998b, 2001; Flanagan antimycotic (penicillin G/streptomycin/amphotericin and Vanderhaeghen, 1998; Holder et al., 1998;Hu B; Gibco) in 5% CO2 at 37 8C. Once cultures et al., 2000; Klein, 2001; Nakamoto, 2000; O’Leary approached confluence, the medium was replaced with and Wilkinson, 1999; Wilkinson, 2000a, 2000b, 2001; a-MEM containing 1.0 mM retinoic acid to induce Zhou, 1998). Transgenic mice in which specific Ephs neuronal differentiation (day 0). On day 2, a-MEM/ or ephrins have been genetically altered display mal- retinoic acid was removed. Plates were rinsed with formations of CNS cytoarchitecture, demonstrating Dulbeco’s modified eagle medium (DMEM) (Gibco that disruption of these guidance molecules can lead Invitrogen Corp., Grand Island, NY) and cells were to path-finding errors (Chen et al., 2004; Coonan et al., maintained in serum-free medium (SFM) for the 2001; Hu et al., 2003; Kullander et al., 2001; Lyckman remainder of the study. Neuronal cultures were treated et al., 2001; Orioli et al., 1996; Park et al., 1997; Yue with 0.5 mg/ml cytosine arabinoside in SFM for 16 h et al., 2002). Alterations in the function of EphA on day 3 to suppress non-neuronal cell proliferation, receptors produce changes in motor and cognitive after which time the media was removed and replaced functioning in rodents, indicating that perturbations with SFM. Fresh SFM was supplied daily thereafter in the expression or activity of these molecules result in until the cells were collected. Cells were treated on day D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 663

1, 5, or 9 with 0.0, 0.5, 1.5, or 3.0 mM methylmercuric of each ligand or receptor band to the L32 house- chloride (MeHg) in the culture medium (a-MEM for keeping RNA band were compared for each sample. undifferentiated cells and day 1 neurons; SFM on days 5 and 9) for 24 h. Following MeHg treatment, the Protein Isolation/Immunoblotting plates were rinsed once with D-MEM, photographed, and cells harvested in 0.5 ml Tri-Reagent LS (Mole- Protein was isolated from the frozen organic phase cular Research Center; Cincinnati, OH). Cell suspen- collected after the initial homogenization in Tri- sions were stored at À80 8C until RNA isolation. Reagent and centrifugation as per manufacturer’s Undifferentiated cells were exposed to MeHg for instructions (Molecular Research Center Inc., Cin- 24 h as the cultures approached confluence and cells cinnati, OH). Isolated protein samples were stored at were collected after treatment. À80 8C until used for immunoblotting. Samples were quantified using the BCA assay (Pierce, Rockford, RNA Isolation and Quantitation IL) modified for a COBAS FARA II enzyme analyzer (Roche Diagnostics, Nutley, NJ). Ten to twenty-five All procedures involving RNA isolation were per- micrograms of total protein for each sample was formed under RNAse-free conditions using RNAse- separated by SDS-PAGE using 10% polyacrylamide free reagents and equipment. Cell samples stored gels in a Bio-Rad mini-protean II system (Bio-Rad, frozen in Tri-Reagent were thawed and homogenized Melville, NY). Proteins were transferred overnight using a Kiematica PCU11 polytron with a PTA7 gen- onto nitrocellulose membranes as described by Tow- erator. Fifty microlitres of 1-bromo-3-chloropropane bin et al. (1979). Blots were rinsed and incubated (BCP) and 50 ml RNAse-free water were mixed with overnight with primary antibody in blocking buffer. each sample and total RNA was isolated by precipita- Primary polyclonal goat anti-mouse antibodies to tion with isopropanol. RNA samples were stored at EphA3, -A4, -A5, -A8, -B2, -B3, -B6, ephrin-A4 À80 8C until quantitation and ribonuclease protection and -B1 were obtained from R&D Systems (Minnea- assay. polis, MN) and used at 1:1000 dilution. EphA2 monoclonal primary antibody (1:2000) was from Ribonuclease Protection Assay Probe Synthesis Sigma (St. Louis, MO). Ephrin-B3 polyclonal anti- and Hybridization body (1:500) was from Zymed (San Francisco, CA), ephrin-A5 polyclonal (1:500) from Aviva (San Diego, Radiolabeled nucleotide probes for EphA1, -A2, - CA), GAPDH monoclonal (1:5000) from Novus Bio- A3, -A4, -A5, -A6, -A7, -A8, -B2, -B3, -B4, -B6 and logical (Littleton, CO) and actin monoclonal (1:500) ephrins-B1, -B2, -B3, -A1, -A3, -A4, -A5, -A6 were from Santa Cruz (Santa Cruz, CA). Following incu- synthesized as described in the RiboQuant1 Multi- bation in primary antibody, blots were rinsed with Probe RNAse Protection Assay System instruction PBS-Tween buffer, incubated for 1 h with the appro- manual (BD Biosciences, San Diego, CA, 6th Ed., priate horseradish peroxidase-linked secondary anti- 1999). Briefly, radiolabeled anti-sense probes were body in blocking buffer, rinsed, and visualized using synthesized by T7 polymerase using cDNA template a electrochemoluminescence (ECL Plus) detection sets for each receptor or ligand. Ten to twenty micro- system on Hyperfilm ECL autoradiographic film grams of total RNAwas used for hybridization with the (Amersham, Arlington Heights, IL). Rabbit anti-goat, probes. Hybridization was carried out overnight (12– goat anti-rabbit, and goat anti-mouse HRP-linked 16 h) in a heat block. The initial temperature of the secondary antibodies were obtained from Southern block was 90 8C and was reduced to 56 8C immediately Biotech (Birmingham, AL) and were used at 1:3000 upon placing the hybridization mixtures in the block. dilution. Developed films were analyzed by densito- Following hybridization, unhybridized probe and RNA metry using ImagePro software. were digested with RNAse A solution and the probe- protected RNA complexes isolated. Samples were Statistical Analysis prepared for gel electrophoresis and separated on a 5% SDS-PAGE gel. The gels were dried, exposed to Differences in mRNA and protein expression autoradiographic film for 16–72 h, developed, and the between control and MeHg exposed cultures were developed films were analyzed by densitometry using analyzed using a univariate analysis of variance ImagePro1 software. To compensate for variations in (ANOVA) with MeHg dose serving as independent the final amount of RNA loaded on the gels, the ratios variable and receptor or ligand as dependent variable. 664 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674

When appropriate, post-hoc analysis was conducted using a Fisher’s PLSD using a p < 0.05.

RESULTS

Developmental Expression of Eph and Ephrins in P19 Cells and Their Neuronal Derivatives

Results of the ribonuclease protection assays showed a distinct time-course of Eph mRNA expression after neuronal induction of undifferentiated EC cells (Figs. 1 and 3a). Undifferentiated cells expressed low levels of most Eph receptors, with the exception of EphA2 which was heavily expressed. Two days following retinoic acid induction, the differentiating neurons showed increased mRNA expression of B-class receptors, while expression of A-class receptors remained low. EphA2 mRNA decreased dramatically

Fig. 2. RNAse protection assay results showing the expression time-course of Ephrin mRNA following retinoic acid treatment. Undifferentiated P19 cells were treated with retinoic acid (day 0) and allowed to form neurons. Low levels of ligand expression were observed in undifferentiated cells. Expression levels increased as differentiation and maturation progressed.

with retinoic acid treatment and was already markedly reduced by day 2. By day 6 post-retinoic acid, most A- and B-class receptors were expressed in the neuronal cultures. The receptors continued to be highly expressed on day 10. mRNA for both A- and B-class ephrins displayed a similar time-course of expression (Figs. 2 and 3b). Ephrin expression was virtually absent in undifferentiated cells and increased during the process of neuronal differentiation. Most ligands were detectable as early as day 2. Ephrins-A5 and-B1 were up-regulated the most on day 2, while mRNA expression of ephrins- A3, -B2 and -B3 was up-regulated at day 6. Expression of several ligands, including ephrins-B1, -B2, -B3, and -A6, was maximal at day 6, while the other ephrin-A class ligands reached maximal expression on day 10 Fig. 1. RNAse protection assay results showing the expression time-course of Eph mRNA following retinoic acid treatment. Undifferentiated P19 cells (Fig. 3b). were treated with retinoic acid (day 0) and allowed to form neurons. Low Expression levels of proteins in undifferentiated EC levels of receptor expression were observed in undifferentiated cells. cells and neurons derived from these cells followed Expression levels increased as differentiation and maturation progressed. EphB receptors were elevated as early as day 2 and A receptors showed patterns consistent with their mRNA levels. EphA2 increased expression levels by day 6. protein decreased as neuronal maturation progressed, D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 665

Fig. 3. (a and b) Graphs summarizing Eph (top) and ephrin (bottom) expression time-course in undifferentiated cells and neurons derived from these cells. EphBs are up-regulated as early as day 2 while A receptors are increased by day 6. A distinct pattern was not observed for A vs. B ligands with most being elevated by day 2. while many of the other receptors and ligands showed at 1.5 and 3.0 mM MeHg and EphA4 was increased at trends of increasing protein expression as neurons all concentrations in day 2 neuron cultures compared to matured (data not shown). control (Fig. 5a). Ephrins-A4, -A5, A-6, -B1, -B2, and - B3 were all signiﬁcantly elevated in day 2 neurons at Methylmercury Effects on Eph and Ephrin 3.0 mM compared to controls (Fig. 5b). Expression in P19 Cells and Their Neuronal Day 6 neurons showed the most widespread altera- Derivatives tions in Eph and ephrin mRNA levels following MeHg treatment (Fig. 6a and b). mRNA for four of the ten Eph MeHg treatment did not affect Eph or ephrin mRNA receptors and four of the eight ligands were increased levels in undifferentiated P19 cells at any concentration compared to control. A concentration–response rela- examined (Fig. 4a and b). However, MeHg did affect tionship was observed for increased Eph expression in Eph and ephrin RNA levels following retinoic acid- these cells, with increased levels of receptor expression induced neuronal differentiation. EphA2 was increased at the lower concentrations of MeHg and signiﬁcant 666 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674

Fig. 4. (a and b) Graphs showing RNAse protection assay autoradiography results for Eph receptor (top) and ephrin ligand (bottom) mRNA following MeHg exposure in undifferentiated P19 cells. No significant changes were observed for either receptor or ligand expression. At the highest dose of 3.0 mM MeHg, cells were not viable. N = 3–5 per group. (*) p < 0.05 compared to cells treated with culture medium only using Fisher’s PLSD. elevations in EphA2, -A3, -B3, and -B6 at the 3.0 mM was consistent with mRNA results (Figs. 8a and 9b). level. The increase in EphA4 mRNA at 3.0 mM MeHg On the other hand, protein levels for most receptors and approached significance ( p = 0.06). mRNA for ligands examined were decreased well below control ligands-B1, -B2, -A5 and -A6 also showed an increas- levels at 3.0 mM MeHg in the absence of a decrease in ing concentration response to MeHg treatment in day 6 mRNA. Protein levels of Eph-A3 and ephrin-B2 at neurons. 3.0 mM were decreased, in contrast to the increase in Day 10 neurons, although less extensively affected mRNA expression (Figs. 8b and 10b). Immunoblots of by MeHg than day 6 neurons, also showed elevations in two housekeeping genes, GAPDH and actin, showed several receptor and ligand mRNA levels compared to that GAPDH was not affected by MeHg treatment control after MeHg exposure at the 1.5 and 3.0 mM while actin levels were decreased only at the highest concentrations (Fig. 7a and b). Eph-B2 was signifi- concentration (Fig. 11). cantly elevated, while expression of Eph-B6 was reduced following MeHg exposure. All ligands were elevated following the highest dose of MeHg (3.0 mM) DISCUSSION with the exception of ephrins-A1 and -A3. A positive concentration–response was seen for ligand expression Although expression of several Ephs and ephrins has levels in day 10 neurons. been documented previously in P19 cells (Bouillet Immunobloting revealed an increase in ephrin-B1 et al., 1995; Stein et al., 1998), a systematic character- and -A5 protein in day 6 neurons at 1.5 mM MeHg that ization of the expression of these molecules during D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 667

Fig. 5. (a and b) Graphs showing RNAse protection assay autoradiography results for Eph receptor (top) and ephrin ligand (bottom) mRNA following MeHg exposure in day 2 neurons. The highest dose of MeHg (3.0 mM) led to an increase in mRNA in all Ephrin ligands. N = 3–4 per group. (*) p < 0.05 compared to cells treated with culture medium only using Fisher’s PLSD. differentiation and neurite formation has been lacking. ferentiation, falling to less than 50% of undifferen- In this study, the mRNA expression proﬁles of the Ephs tiated cell levels by day 2 of differentiation, while and ephrins in P19 cells were determined from the mRNA levels of the other receptors and ligands gen- undifferentiated state to 10 days following neural erally increased as neurons matured. The level of induction. Neither receptors nor ligands were EphA2 expression may thus serve as a marker of expressed at appreciable levels in undifferentiated differentiation state of tumor cells. cells, with the exception of EphA2 which is highly The time course for expression of the Ephs after expressed in many tumor cell lines such as prostate, retinoic acid treatment is noteworthy. All detectable B- colon, and breast cancers and, thus not surprisingly also class receptors are expressed by day 2 after retinoic in embryonal carcinoma cells (Andres et al., 1994; acid treatment, whereas the A-class receptors do not Kinch and Carles-Kinch, 2003; Sulman et al., 1997; appear until day 6 post-treatment. This suggests that Walker-Daniels et al., 1999). The level of EphA2 the different classes of Eph receptors may serve dif- decreased rapidly following retinoic acid-induced different roles during stages of cellular differentiation and 668 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674

Fig. 6. (a and b) Graphs showing RNAse protection assay autoradiography results for Eph receptor (top) and ephrin ligand (bottom) mRNA following MeHg exposure in day 6 neurons. Increases in several Eph receptors and ephrin ligands were observed following MeHg treatment. N = 3–4 per group. (*) p < 0.05 compared to cells treated with culture medium only using Fisher’s PLSD. maturation. The B-class receptors may be involved in of early neuronal differentiation. They are expressed at early differentiation processes, particularly events sur- a time prior to the appearance of classic neuronal rounding lineage commitment and early neurite for- markers such as HNK-1, which are ﬁrst seen 3 days mation, while the A-class receptors are expressed at after retinoic acid induction (Bain et al., 1994; McBur- later times to stabilize maturing cytoarchitecture. Dis- ney et al., 1988). There have been other reports of tinct roles for A- and B-receptors have been documen- guidance molecules and receptor tyrosine kinases ted during structuring of synapses during learning and being up-regulated in P19 cells following retinoic acid memory in vivo (Gao et al., 1998a; Gerlai, 2001; induction. For example, Wu and Adamson (1993) Ghosh, 2002; Murai and Pasquale, 2002). Utilizing reported an up-regulation of epidermal growth factor the P19 cell culture model, it may be possible to receptors and Husmann et al. (1989) showed an distinguish the distinct roles each class of receptor increase in the polysialylated form of the neural cell plays in neurodevelopmental processes. adhesion molecule (NCAM) following retinoic acid The temporal expression of the EphB receptors treatment. The widespread up-regulation and coopera- suggests that these molecules may be useful markers tive expression of these proteins is suggestive of their D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 669

Fig. 7. (a and b) Graphs showing RNAse protection assay autoradiography results for Eph receptor (top) and ephrin ligand (bottom) mRNA following MeHg exposure in day 10 neurons. Increases in several Eph receptors and ephrin ligands were observed following MeHg treatment. N = 3–4 per group. (*) p < 0.05 compared to cells treated with culture medium only using Fisher’s PLSD. co-regulation. Several of these proteins involved in the sion. It is unclear what roles the A-ligands play in early highly orchestrated formation of the brain such as neuronal commitment, since their receptors are not NCAM (Dey et al., 1994), trk receptors (Barone expressed at this early time point. However, it is et al., 1998, 2004), and Ephs and ephrins, are affected possible that some of the A-ligands such as ephrin- by MeHg exposure. The combined disruption of sev- A5 may interact with B-receptors (Himanen et al., eral individual guidance molecules, even to a small 2004). While B-ligands may interact with B-receptors degree, may cause additive effects with adverse con- on day 2, A-ligands may activate signaling pathways sequences on morphogenesis and brain function. unrelated to Eph receptors or may modulate expression Similar to the Ephs, ephrin levels increased follow- of other developmental genes. Further investigation is ing retinoic acid treatment by day 2 and continued to required to determine how ephrin-As participate in show increasing expression on days 6 and 10. There early neuronal differentiation of P19 cells. was no class-speciﬁc pattern of expression for the In the current studies, Ephs and ephrins in undiffer- ligands as was seen with the receptors; both ephrin- entiated cells were largely insensitive to MeHg expo- As and -Bs showed early and constant mRNA expres- sure. In contrast, Eph and ephrin expression in 670 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674

Fig. 9. (a and b) Immunoblots for Ephrin-B3 (top) and A5 (bottom) in day 6 neurons exposed to MeHg. Protein levels increase at low dose levels but are reduced at 3.0 mM MeHg. Lanes 1 and 2: control; lanes 3 and 4: 0.5 mM; lanes 5 and 6: 1.5 mM; lanes 7 and 8: 3.0 mM. (*) p < 0.05 compared to cells treated with culture medium only using Fisher’s PLSD. Fig. 8. (a and b) Immunoblots for Ephrin-B1 (top) and B2 (bottom) in day 6 neurons exposed to MeHg. Protein levels increase at low dose levels but are reduced at 3.0 mM MeHg. Lanes 1 and 2: control; lanes 3 and 4: 0.5 mM; remained increased at the 3.0 mM level; however, lanes 5 and 6: 1.5 mM; lanes 7 and 8: 3.0 mM. (*) p < 0.05 compared to cells protein levels were decreased. This suggests that MeHg treated with culture medium only using Fisher’s PLSD. at this concentration affects the cell’s translational machinery directly but continues to stimulate transcrip- P19-derived neurons was responsive to MeHg expo- tion. This is consistent with several reports that MeHg sure. Day 6 neurons showed the most widespread increases protein expression at low concentrations but alterations in Eph and ephrin levels, perhaps due to decreases protein levels at higher concentrations (Bru- the active growth and developmental events occurring baker et al., 1973; Cavanagh and Chen, 1971; Omata in the cells at this time point, such as extension of et al., 1980; Saraﬁan and Verity, 1990a, 1990b). MeHg neurites and formation of synaptic connections, two has been reported to disperse polysomal aggregates and processes known to be Eph/ephrin-dependent. Days 2 to remove ribosomes from the endoplasmic reticulum and 10 neurons were less sensitive to MeHg treatment (Brown and Yoshida, 1965); however, these effects than day 6 cultures but did show alterations in Eph and were noted at MeHg concentrations substantially ephrin mRNA levels. This pattern of sensitivity sup- higher than those used in our studies. GAPDH and ports the speciﬁc nature of the MeHg-induced altera- actin, two housekeeping proteins, were not affected at tions of Ephs and ephrins. the 0.5 and 1.5 mM MeHg levels. GAPDH remained Protein and mRNA levels of many Eph receptors unchanged at the high MeHg concentration but actin and ephrin ligands were elevated following 0.5 and was modestly decreased. The general non-responsive- 1.5 mM MeHg treatment in day 6 neurons. RNA levels ness of these housekeeping genes to MeHg gives D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 671

Fig. 11. (a and b) Immunoblots for GAPDH (top) and actin (bottom) in day Fig. 10. Immunoblots for EphA4 (top) and EphA3 (bottom) in day 6 6 neurons exposed to MeHg. GAPDH protein level was unaffected by MeHg neurons exposed to MeHg. EphA4 protein levels increased at low dose exposure. Actin was unaffected at low concentrations but was decreased at levels but decreased at 3.0 mM MeHg. EphA3 levels decreased at the mid the 3.0 mM level. Lanes 1 and 2: Control; lanes 3 and 4: 0.5 mM; lanes 7 and and high dose levels. Lanes 1 and 2: control; lanes 3 and 4, 0.5 mM; lanes 5 8: 3.0 mM. (*) p < 0.05 compared to cells treated with culture medium only and 6: 1.5 mM; lanes 7 and 8: 3.0 mM. (*) p < 0.05 compared to cells treated using Fisher’s PLSD. with culture medium only using Fisher’s PLSD. baker et al. (1973) reported that reduction of protein synthesis activity in the liver occurs at tissue mercury further support to the theory that MeHg specifically levels lower than those required to suppress RNA targets the Ephs and ephrins, and is not simply altering synthesis. Our experiments seem to support this obser- overall protein synthesis within the cell. vation since all three concentrations of MeHg caused The mechanism by which MeHg alters mRNA and stimulation of RNA synthesis, but only the lower protein levels is not known, although several molecular concentrations caused a concomitant alteration in pro- processes have been implicated. Frenkel et al. (1985) tein level. reported that RNA polymerase II transcribes DNA at a These data support the notion that MeHg indirectly higher rate when exposed to MeHg and demonstrated stimulates protein synthesis by first increasing tran- that the increase in transcriptional efficiency is due to a scription. It is also possible that MeHg enhances RNA three- to four-fold lower Km of the enzyme for MeHg- stability and thereby mRNA levels. Both mRNA and treated DNA. Increases in protein levels in the brain protein levels were increased at the lower MeHg con- following MeHg treatment have also been reported centrations in our experiments. Conversely, the data (Brubaker et al., 1973; Slotkin et al., 1985). In contrast, gathered from the 3.0 mM concentration imply that Cheung and Verity (1985) reported decreased protein MeHg affects protein synthesis directly since mRNA synthesis following MeHg treatment and attributed the levels remained stimulated while protein levels fell. inhibition to defective aminoacylation of tRNA. Bru- Regardless of the effects of MeHg exposure on Eph and 672 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 ephrin protein levels, the changes in expression ACKNOWLEDGEMENTS (increase or decrease) of morphoregulatory molecules could influence differentiation, migration, and synapse The authors thank Dr. Suzie Chen and Kathy formation in the brain. Increasing or decreasing gui- Roberts for their technical assistance. This work was dance molecule levels can disrupt the delicate balance supported by NIH grants ES07148, EPA R829391, between repulsion and adhesion, and cause abnormal ES11256 and ES05022. brain structuring (Edelman, 1992; Garner et al., 2002; Knoll and Drescher, 2002). The results of these experiments support our hypoth- REFERENCES esis that alterations of Eph and ephrin expression in the developing CNS by MeHg may contribute to distur- Andres A, Reid H, Zurcher G, Blaschke R, Albrecht D, Ziemiecki bances of brain development. MeHg causes various A. Expression of two novel eph-related receptor protein tyrosine morphological changes specifically in several brain kinases in mammary gland development and carcinogenesis. structures where Ephs and ephrins are known to be Oncogene 1994;9:1461–7. key elements in proper topographical mapping. For Bain G, Ray W, Yao M, Gottlieb D. From embryonal carcinoma example, EphB3 and EphA5 are involved in corpus cells to neurons: the P19 pathway. BioEssays 1994;16:343–8. Barone S, Haykal-Coates N, Parran D, Tilson H. Gestational callosum formation (Hu et al., 2003; Orioli et al., 1996; exposure to methylmercury alters the developmental pattern Zhou, 1998), and MeHg has been shown to cause of trk-like immunoreactivity in the rat brain and results in thinning of this structure (Takeuchi, 1977). EphA2, - cortical dysmorphology. Brain Res Dev Brain Res 1998; A3, and -A4 are expressed in the ventricular and 109:13–31. subventricular zones during development (Conover Bouillet P, Oulad-Abelghani M, Vicaire S, Garnier J, Schuhbaur B, et al., 2000; Zhou, 1998); MeHg has been shown to Dolle P, et al. Efficient cloning of cDNAs of retinoic acid- responsive genes in P19 embryonal carcinoma cells and char- cause defects in the ventricular walls of exposed infants acterization of a novel mouse gene, Stra1 (mouse LERK-2/ (Matsumoto et al., 1988). EphA4 is involved in devel- Eplg2). Dev Biol 1995;170:420–33. opment of the cerebellum, another brain region Brown W, Yoshida N. Organic mercurial encephalopathy B. An adversely affected by MeHg (Rogers et al., 1999). experimental electron microscopic study. Adv Neurol Sci 1965; Lastly, ephrin-A4 is expressed widely in the embryonic 9:34–42. Brubaker P, Klein R, Herman S, Lucier G, Alexander L, Long M. cortex (Zhou, 1998), another region where effects of DNA, RNA, and protein synthesis in brain, liver, and kidneys of MeHg exposure can be detected (Choi et al., 1978; asymptomatic methylmercury treated rats. Exp Mol Pathol Reuhl and Chang, 1979). 1973;18:263–80. Since few studies have examined the responses of Burbacher TM, Rodier PM, Weiss B. MeHg developmental neu- Ephs and ephrins to toxicants, it is difficult to speculate rotoxicity: a comparison of effects in humans and animals. if results similar to those obtained in these experiments Neurotoxicol Teratol 1990;12:191–202. Cadrin M, Wasteneys G, Jones-Villeneuve E, Brown D, Reuhl K. would occur after exposure to other neurotoxic com- Effects of methylmercury on retinoic acid-induced neuroecto- pounds. Halladay et al. (2000) demonstrated a dose- dermal derivatives of embryonal carcinoma cells. Cell Biol dependent increase in EphB1 expression in the cortex, Toxicol 1988;4:61–80. striatum, and substantia nigra of neonatal mice exposed Cavanagh J, Chen F. Amino acid incorporation in protein during the pre- or postnatally to cocaine and suggest that the silent phase before organo-mercury and p-bromophenylacety- lurea neuropathy in the rat. Acta Neuropathol 1971;19:216–24. expression of this receptor is mediated by dopamine Chen Z, Sun C, Reuhl K, Bergemann A, Hankemeyer M, Zhou R. signaling. Yue et al. (1999) demonstrated that ephrin- Abnormal hippocampal bundling in EphB receptor mutant B2, the preferred ligand for EphB1, was up-regulated mice. J Neurosci 2004;24:2366–74. following cocaine administration in adult mice. Mor- Cheung M, Verity A. Experimental methyl mercury neurotoxicity: eno-Flores and Wandosell (1999) documented an locus of mercurial inhibition of brain protein synthesis in vivo increase in EphA4, B2, and A5 mRNA levels in the and in vitro. Neurochemistry 1985;44:1799–808. Choi B, Lapham L, Amin-Zaki L, Saleem T. Abnormal neuronal adult rat hippocampus after excitotoxic injury by kai- migration, deranged cerebral cortical organization, and diffuse nate. Preliminary results from a separate experiment in white matter astrocytosis of human fetal brain: a major effect of our laboratory showed increased EphA2 mRNA and methylmercury poisoning in utero. J Neuropathol Exp Neurol decreased EphA3 mRNA in brains from adult mice 1978;37:719–33. treated with trimethyltin (data not shown). Future Choi B. Methylmercury poisoning of the developing nervous system: I. Pattern of neuronal migration in the cerebral cortex. studies will determine whether alteration of Eph/ephrin Neurotoxicology 1986;7:591–600. expression represents a common mechanism of action Choi B. The effects of methylmercury on the developing brain. of neurotoxic chemicals. Prog Neurobiol 1989;32:447–70. D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674 673

Conover J, Doetsch F, Garcia-Verdugo J, Gale N, Yancopoulos G, Hu Z, Yue X, Shi G, Yue Y, Crockett DP, Blair-Flynn J, et al. Alvarez-Buylla A. Disruption of Eph/ephrin signaling affects Corpus callosum deficiency in transgenic mice express- migration and proliferation in the adult subventricular zone. Nat ing a truncated Eprin-A receptor. J Neurosci 2003;23: Neurosci 2000;3:1091–7. 10963–70. Coonan J, Greferath U, Messenger J, Hartley L, Murphy M, Boyd Husmann M, Gorgen I, Weisgerber C, Bitter-Suermann D. Up- A, et al. Development and reorganization of corticospinal regulation of embryonic NCAM in an EC cell line by retinoic projections in EphA4 deficient mice. J Comp Neurobiol acid. Dev Biol 1989;136:194–200. 2001;436:248–62. Kinch M, Carles-Kinch K. Overexpression functional alterations of Dey P, Graff R, Lagunowich L, Reuhl K. Selective loss of the 180- the EphA2 tyrosine kinase in cancer. Clin Exp Metastasis kDa form of the neural cell adhesion molecule in hippocampus 2003;20:59–68. and cerebellum of the adult mouse following trimethyltin Klein R. Excitatory Eph receptors and adhesive ephrin ligands. administration. Toxicol Appl Pharmacol 1994;126:69–74. Curr Opin Cell Biol 2001;13:196–203. Edelman G. Morphoregulation. Dev Dynam 1992;193:2–10. Knoll B, Drescher U. Ephrin—as receptors in topographic projec- Flanagan J, Vanderhaeghen P. The ephrins and Eph receptors in tions. Trends Neurosci 2002;25:145–9. neural development. Ann Rev Neurosci 1998;21:309–45. Kullander K, Croll S, Zimmer M, Pan L, McClain J, Hughes V, et Frenkel G, Cain R, Chao E. Exposure of DNA to methyl mercury al. Ephrin-B3 is the midline barrier that prevents corticospinal results in an increase in the rate of its transcription by RNA tract axons from recrossing, allowing for unilateral motor polymerase II. Biochem Biophys Res Comm 1985;127:849– control. Genes Dev 2001a;15:877–88. 56. Kullander K, Mather N, Diella F, Dottori M, Boyd A, Klein R. Gale N, Holland S, Valenzuela D, Flenniken A, Pan L, Ryan T, et Kinase-dependent and kinase-independent functions of EphA4 al. Eph receptors and ligands comprise two major specificity receptors in major axon tract formation in vivo. Neuron subclasses and are reciprocally compartmentalized during 2001b;29:73–84. embryogenesis. Neuron 1996;17:9–19. Lagunowich L, Stein A, Reuhl K. N-Cadherin in normal and Gao W-Q, Shinsky N, Armanini M, Moran P, Zheng J, Mendoza- abnormal brain development. Neurotoxicology 1994;15:123– Ramirez J, et al. Regulation of hippocampal synaptic plasticity 32. by the tyrosine kinase receptor, REK7/EphA5, and its ligand, Lyckman A, Jhaveri S, Feldheim D, Vanderhaeghen P, Flanagan AL-1/ephrin-A5. Mol Cell Neurosci 1998a;11:247–59. J, Sur M. Enhanced plasticity of retinothalamic projections in Gao P, Yue Y, Zhang J, Cerretti D, Levitt P, Zhou R. Regulation of an ephrin-A2/A5 double mutant. Neuroscience 2001;21: thalamic neurite outgrowth by the Eph ligand ephrin-A5: 7684–90. implications in the development of thalamocortical projections. Matsumoto H, Koya G, Takeuchi T. Fetal Minamata disease. A Proc Natl Acad Sci USA 1998b;95:5329–34. neuropathological study of two cases of intrauterine intoxica- GaoX,BianW,YangJ,TangK,KitaniH,AtsumiT,etal.Arole tion by methylmercury compound. J Neuropath Exp Neurol of N-cadherin in neuronal differentiation of embryonic 1965;24:563–74. carcinoma P19 cells. Biochem Biophys Res Comm 2001; Matsumoto SC, Okajima T, Inayoshi S, Ueno H. Minamata disease 284:1098–103. demonstrated by computed tomography. Neuroradiology Garner C, Zhai R, Gundelfinger E, Ziv N. Molecular mechanisms 1988;30:42–6. of CNS synaptogenesis. Trends Neurosci 2002;25:243–50. McBurney M, Reuhl K, Ally A, Nasipuri S, Bell J, Craig J. Gerlai R, Shinsky N, Shih A, Williams P, Winer J, Armanini M, et Differentiation and maturation of embryonal carcinoma-derived al. Regulation of learning by EphA5 receptors: a protein neurons in cell culture. J Neurosci 1988;8:1063–73. targeting study. J Neurosci 1999;19(21):9538–49. Moreno-Flores M, Wandosell F. Up-regulation of Eph tyrosine Gerlai R. Eph receptors and neural plasticity. Nat Rev Neurosci kinase receptors after excitotoxic injury in adult hippocampus. 2001;2:205–9. Neuroscience 1999;91:193–201. Ghosh A. Learning more about NMDA receptor regulation. Murai K, Pasquale E. Can Eph receptors stimulate the mind?. Science 2002;295:449–51. Neuron 2002;33:159–62. Graff R, Falconer M, Brown D, Reuhl K. Altered sensitivity of Murai K, Pasquale E. Eph receptors, ephrins, and synaptic function. posttranslationally modified microtubules to methylmercury in Neuroscientist 2004;10:304–14. differentiating embryonal carcinoma-derived neurons. Toxicol Nakamoto M. Eph receptors and ephrins. Int J Biochem Cell Biol Appl Pharmacol 1997;144:215–24. 2000;32:7–12. Halladay A, Yue Y, Michna L, Widmer D, Wagner G, Zhou R. O’Leary D, Wilkinson D. Eph receptors and ephrins in neural Regulation of EphB1 expression by dopamine signaling. Mol development. Curr Opin Neurobiol 1999;9:65–73. Brain Res 2000;85:171–8. Omata S, Horigome T, Momose Y, Kambayashi M, Mochizuki M, Halladay A, Tessarollo L, Zhou R, Wagner G. Neurochemical and Sugano H. Effect of methylmercury chloride on the in vivo rate behavioral deficits consequent to expression of a dominant of protein synthesis in the brain of the rat: examination with the negative EphA5 receptor. Mol Brain Res 2004;123:104–11. injection of a large quantity of [14C] valine. Toxicol Appl Himanen J, Chumley M, Lackmann M, Li C, Barton W, Jeffrey P, Pharmacol 1980;56:207–15. et al. Repelling class discrimination: ephrin-A5 binds to and Orioli D, Henkemeyer M, Lemke G, Klein R, Pawson T. Sek4 and activates EphB2 receptor signaling. Nat Neurosci 2004;7: Nuk receptors cooperate in guidance of commissural axons and 501–9. in palate formation. EMBO J 1996;15:6035–49. Holder N, Cooke J, Brennan C. The Eph receptor tyrosine kinases Park S, Frisen J, Barbacid M. Aberrant axonal projections in mice and the ephrins B roles in nervous system development. Neu- lacking EphA8 (Eek) tyrosine protein kinase receptors. EMBO J roscience 1998;10:405–8. 1997;16:3106–14. 674 D.T. Wilson et al. / NeuroToxicology 26 (2005) 661–674

Pyle S, Roberts K, Reuhl K. Delayed expression of the NFH subunit Takeuchi T. Pathology of fetal Minamata disease–the effect of in differentiating P19 cells. Dev Brain Res 2001;132:103–6. methylmercury on human intrauterine life. Pediatrics 1977; Reuhl K, Chang L. Effects of methylmercury on the development 6:69–87. of the nervous system: a review. Neurotoxicology 1979;1:21– Towbin M, Staelin T, Gorden J. Electrophoretic transfer of proteins 55. from polyacrylamide gels to nitrocellulose sheets: procedures Reuhl K, Lagunowich L, Brown D. Cytoskeleton and cell adhesion and some applications. Proc Natl Acad Sci USA 1979;76: molecules: critical targets of toxic agents. Neurotoxicology 4350–9. 1994;15:133–45. Walker-Daniels J, Coffman K, Azimi M, Rhim J, Bostwick D, Rogers J, Ciossek T, Menzel P, Pasquale E. Eph receptors and Snyder P, et al. Overexpression of the EphA2 tyrosine kinase in ephrins demarcate cerebellar lobules before and during their prostate cancer. Prostate 1999;41:275–80. formation. Mech Dev 1999;87:119–28. Wasteneys G, Cadrin M, Reuhl K, Brown D. The effects of Sarafian T, Verity A. Methyl mercury stimulates protein 32P methylmercury on the cytoskeleton of murine embryonal car- phospholabeling in cerebellar granule cell culture. J Neurochem cinoma cells. Cell Biol Toxicol 1988;4:41–60. 1990a;55:913–21. Wilkinson D. Topographic mapping: organising by repulsion and Sarafian T, Verity A. Altered patterns of protein phosphorylation competition?. Curr Biol 2000a;10:R447–51. and synthesis caused by methyl mercury in cerebellar granule Wilkinson D. Eph receptors and ephrins: regulators of guidance and cells. J Neurochem 1990b;55:922–9. assembly. Int Rev Cytol 2000b;196:177–244. Slotkin T, Pachman S, Kavlock R, Bartolome J. Effects of neonatal Wilkinson D. Multiple roles of Eph receptors and ephrins in neural methylmercury exposure on development of nucleic acids and development. Nat Rev Neurosci 2001;2:155–64. proteins in rat brain: regional specificity. Brain Res Bull Wu J, Adamson E. Inhibition of differentiation in P19 embryonal 1985;14:397–400. carcinoma cells by the expression of vectors encoding truncated Stein E, Lane A, Cerretti D, Schoecklmann H, Schroff A, Van Etten or antisense EGF receptor. Dev Biol 1993;159:208–22. R, et al. Eph receptors discriminate specific ligand oligomers to Yue Y, Chen Z, Gale N, Blair-Flynn J, Hu T, Yue X, et al. determine alternative signaling complexes, attachment, and Mistargeting hippocampal axons by expression of a truncated assembly responses. Genes Dev 1998;12:667–78. Eph receptor. Proc Natl Acad Sci USA 2002;99:10777–82. Sulman E, Tang X, Allen C, Biegel J, Pleasure D, Brodeur G, et al. Yue Y, Widmer D, Halladay A, Cerretti D, Wagner G, Dreyer J, et a human Eph-related gene, maps to 1p36.1, a common region of al. Specification of distinct dopaminergic neural pathways: roles alteration in human cancers. Genomics 1997;40:371–4. of the Eph family receptor EphB1 and ligand ephrin-B2. Takeuchi, T. Pathology of Minamata disease. From Minamata Neuroscience 1999;19:2090–101. disease (organic mercury poisoning) study group of Minamata Zhou R. The Eph family receptors and ligands. Pharmacol Ther disease. Japan: Kumamoto University; 1968. 1998;77:151–81.