|Home | About | Journals | Submit | Contact Us | Français|
A recent study proposed that differentiation of dopaminergic neurons requires a conserved “dopamine motif” (DA-motif) that functions as a binding site for ETS DNA binding domain transcription factors. In the mammalian olfactory bulb, the expression of a set of five genes (including tyrosine hydroxylase, Th) that are necessary for differentiation of dopaminergic neurons was suggested to be regulated by the ETS-domain transcription factor ER81 via the DA-motif. To investigate this putative regulatory role of ER81, expression levels of these five genes were compared both in olfactory bulbs of adult wild-type mice subjected to unilateral naris closure and in the olfactory bulbs of neonatal Er81 wild-type and mutant mice. These studies found that ER81 was necessary only for Th expression, and not the other cassette genes. ChIP and EMSA experiments showed that ER81 bound directly to a consensus binding site/DA-motif in the rodent Th proximal promoter. However, the ER81 binding site/DA-motif in the Th proximal promoter is poorly conserved in other mammals. Both ChIP assays with canine OB tissue and EMSA experiments with the human Th proximal promoter did not detect ER81 binding to the Th DA-motif from these species. These results suggest that regulation of Th expression by the direct binding of ER81 to the Th promoter is a species-specific mechanism. These findings indicate that ER81 is not necessary for expression of the OB dopaminergic gene cassette, and that the DA-motif is not involved in differentiation of the mammalian OB dopaminergic phenotype.
Dopaminergic neurons express a minimal set of 5 genes that are collectively necessary for the biosynthesis of dopamine [tyrosine hydroxylase (Th), GTP cyclohydrolase (Gtpch), and aromatic amino acid decarboxylase (Aadc)], as well as its uptake and storage [dopamine transporter (Dat) and vesicular monoamine transporter 2 (Vmat2)]. However, expression levels for each of these genes differs between the various groups of dopaminergic neurons in the mammalian brain (Weihe et al., 2006). Also, the dopaminergic groups differ in their spatial and temporal developmental origins, their molecular and morphological features as well as the neurological functions they mediate (Cave and Baker, 2008). These differences between dopamine cell groups raises the question of whether a common molecular mechanism regulates expression of this minimal set of 5 dopaminergic genes (or gene cassette).
Recent studies found that the ETS DNA binding domain transcription factor Ast-1 was necessary for expression of the dopaminergic gene cassette in C. elegans (Flames and Hobert, 2009). These studies also identified a putative ETS DNA binding site, the “dopamine motif” (DA-motif), near the transcription start site of each gene in the cassette. Although direct binding of Ast-1 to the DA-motif was not shown, the identification of DA-motifs near mammalian dopamine cassette genes suggested that ETS-domain protein/DA-motif regulatory logic is an evolutionarily conserved mechanism for mediating dopaminergic neuronal differentiation.
Dopaminergic interneurons in the mammalian olfactory bulb (OB) are located almost exclusively in the glomerular layer. They are distinct from other dopaminergic neurons in the mammalian brain since they are generated throughout life from neural stem cells in the subventricular zone (SVZ). Although the OB dopaminergic interneurons are initially derived from the embryonic lateral ganglionic eminence (LGE) (Anderson et al., 1999; Wichterle et al., 2001; Stenman et al., 2003), the vast majority of these neurons are generated in the post-natal SVZ (Luskin, 1993; Lois and Alvarez-Buylla, 1994) with a peak at approximately five days after birth (Hinds, 1968).
Er81 (Etv1) is co-expressed with Th in the mouse OB (Allen et al., 2007; Saino-Saito et al., 2007b), and has been proposed to be part of a “molecular code” that specifies OB dopaminergic interneurons (Allen et al., 2007). Since Er81 is an orthologue of C. elegansast-1, ER81 has been suggested to regulate differentiation of mammalian OB dopaminergic interneurons via the DA-motif (Flames and Hobert, 2009). Consistent with this possibility, the DA-motif sequence overlaps with the previously established ER81 consensus binding site (Brown and McKnight, 1992).
In the present study, the necessity of Er81 for dopaminergic gene cassette expression in the OB was examined in both neonatal Er81 mutant mice and adult wild-type mice subjected to a unilateral naris closure. Phylogenetic analyses of several mammalian Th proximal promoter sequences as well as both chromatin immunoprecipitation (ChIP) and electromobility shift assays (EMSA) were performed to establish whether ER81 regulated Th expression in the OB by directly binding to a DA-motif/ER81 consensus binding site in the Th proximal promoter.
Er81 mutant mice were obtained from Dr. Thomas Jessell (Arber et al., 2000). Mice were housed in humidity-controlled cages at 22°C under a 12:12 hour light:dark cycle and provided with food and water ad libitum. OB tissue from dogs was kindly provided by Dr. Paul Heerdt (Weill Cornell Medical College). All procedures were carried out under protocols approved by the Cornell University Institutional Animal Care and Use Committee and conformed with NIH guidelines.
One nostril of wild-type C57BL/6J mice (aged 6–8 weeks) was surgically closed using a spark-gap cautery under pentobarbital anesthesia. Naris occlusion was confirmed at 1 and 3 months post-operation. Details of the naris occlusion procedure have been previously published (Baker et al., 1993; Liu et al., 1999).
RNA was isolated from mouse OBs using an RNA mini-prep kit (Qiagen) following the manufacturer’s protocol. For non-quantitative analysis, RNA was collected from two adult wild-type C57BL/6J mice. For quantiative analysis of dopamine gene cassette expression levels, RNA was isolated from either four adult C57BL/6J mice subjected to unilateral naris occlusion or two sets of five P10 mice that were either wild-type C57BL/6J or homozygous Er81 mutant.
For the non-quantitative analysis, reverse-transcriptase and PCR amplification was performed using the SuperScript III One-Step kit (Invitrogen) following manufacturer’s protocol. The following primers were used: Th 5′-CACTCCCTGTCAGAGGAGCC-3′ and 5′-ATGAAGGGCAGGAGGAATGC-3′; Gtpch 5′-TAAACTTGCCAGGATTGTAGAAATC-3′ and 5′-CTTCAATCACTACTCCAACGCCA-3′; Aadc 5′-TTACATCCGAAAGCACGTGGAGCT-3′ and 5′-AAGCAGACCAACCCAAGAATGACT-3′; Vmat2 5′-TCATCGCTGCAGGCTCCATCT-3′ and 5′-AGCTGCCACTTTCGGGAACAC-3′; Dat 5′-ACTTCAGGGAAGGTGGTGTGGAT-3′ and 5′-GTAGAAGTCCACACTGAGGTATGC-3′.
For quantitative analysis, first strand reactions were conducted using SuperScript II first strand synthesis kit (Invitrogen), and the quantitative PCR reactions were performed on a 7500 Fast Real-time PCR System (Applied Biosystems). Expression levels for Th, Er81 and Gapdh were measured using Taqman Gene Expression Assay primer sets (Applied Biosystems) Mm00447557_m1, Mm00514804_m1 and Mm99999915_g1, respectively, with the TaqMan Universal PCR Master Mix (Applied Biosystems). For Aadc, Vmat2, and Dat the primers described above were used, and their expression levels were measured with the SYBR Green PCR Master Mix (Applied Biosystems). Expression levels for all genes were normalized to Gapdh levels, and reported as the mean with error bars representing the standard deviation. Data were analyzed using two-tailed Student T-tests for each gene, and differences were considered significant if p<0.01.
Localization of single antigens was performed as previously published (Saino-Saito et al., 2007a). Briefly, mice were anesthetized with an overdose of pentobarbital (100 mg/kg) and perfused transcardially with phosphate buffered 4% formaldehyde, post fixed over night and then cryoprotected in 30% sucrose. Cryostat sections were fixed for 5 minutes with phosphate-buffered, 4% formaldehyde and washed in phosphate-buffered saline (PBS) before being blocked with 1% bovine serum albumin in PBS and incubated overnight with primary antisera. Antigens were visualized by incubation with appropriate biotinylated secondary antiserum, the Vector Elite kit (Vector Laboratories) and 3,3′-diaminobenzidine (DAB, 0.05%) as chromogen with hydrogen peroxide (0.003%). Slides were dehydrated through a graded series of alcohols and cover slipped. The following primary antibodies and dilutions were used: rabbit anti-TH (lot 15-2, raised in our laboratory, 1:25,000); rabbit anti-ER81 (1:1000 for DAB immunohistochemistry, Covance; 1:1000 for immunofluorescence, kindly provided by Tom Jessell (Columbia Univ.)); goat anti-olfactory marker protein (OMP, 1:35,000, kindly provided by Frank Margolis, Univ. Maryland School of Medicine); hamster anti-ERM (1:1000 for DAB immunohistochemistry or 1:500 for immunofluorescence, kindly provided by Kenneth Murphy, Washington Univ. St. Louis/HHMI); rabbit anti-PEA3 (1:1,000, Santa Cruz); rabbit anti-cFOS (1:10,000, Calbiochem); rabbit anti-PAX6 (1:1,000, Covance).
TH labelled cells in the OB were counted in equivalent sections from three wild-type and four Er81 mutant mice. PAX6 labelled cells in the OB were counted within equivalent 1mm glomerular regions from three wild-type and three Er81 mutant mice. Cell counts are expressed as the mean relative to the wild-type genotype, with error bars representing the standard deviation. Data were analyzed by unpaired Student’s t-test, and results were considered significantly different at p<0.01.
For mice, two separate ChIP assays were performed using both bulbs in each assay. For dogs, each bulb from a single dog was separately used for two independent ChIP assays. OBs were dissected and fixed for 20 minutes in PBS and 1% formaldehyde, after which the tissue was rinsed in PBS and then placed in lysis buffer (20mM Tris pH 8.1, 150mM NaCl, 0.5% Triton X-100 and 0.1% SDS). Tissue in the lysate suspension was crushed with a Dounce Homogenizer before sonication with a Misonix 3000 sonicator (Misonix Inc). Following sonication, cellular debris was removed via centrifugation and lysate was then pre-cleared with Protein A/G Sepharose beads (Santa Cruz). The lysate was then divided into two equivalent samples before adding 5μg of either anti-Actin (Santa Cruz) or anti-ER81 (Covance). Antibody/lysate solutions were incubated with gentle rocking overnight at 4°C, before Protein A/G Sepharose was added to precipitate antibody-protein-DNA complexes. The Protein A/G Sepharose beads were then removed from the lysate via centrifugation and washed twice with lysis buffer, twice with wash buffer (20mM Tris pH 8.1, 150mM NaCl, 0.5% Triton X-100 and 0.1% SDS, 2mM EDTA), once with LiCl buffer (0.25M LiCl, 10mM Tris pH 8.1, 1mM EDTA, 1% NP-40, 1% deoxycholate), twice with TE (10mM Tris-EDTA pH 8.0, 1mM EDTA), and then placed in elution buffer (0.1M NaHCO3 pH 8.0, 1%SDS, 0.33M NaCl). The beads were incubated overnight in elution buffer at 65°C. DNA was isolated from the elution buffer using Qiaquick PCR clean-up spin kits (Qiagen). To establish whether specific regions of genomic DNA containing known or predicted ER81 binding sites were immunoprecipitated, PCR reactions were performed using the following primers: mouse Th, 5′-TGGATGCAATTAGATCTAATGGGACGGAGG-3′ and 5′-GCTCTGAGACGGCTCTTCTGAAGCCCTTGG-3′; mouse Mmp-13, 5′-CCCTCAGATTCTGCCACAAACCACAC-3′ and 5′-GGATAGCTGAATGCATGGTGCCCAGC-3′; dog Th, 5′-GTGATTCAGAGCGAGGGCCGCTG-3′ and 5′-GAGGCGGTGTTGGGAGTGGGCAT-3′; dog Mmp-13, 5′-TGTGCCTCCTTCACACACGTCCTG-3′ and 5′-GCTTGCCTGTCCAGTGTCTCAGCG-3′.
ER81 used in the EMSA experiments was generated by over-expression in recombinant pLysS E. coli bacteria containing a pET-28A expression plasmid (Novagen) containing an Er81 cDNA isolated from mouse OB RNA using the Oligotex mRNA mini-kit (Qiagen). ER81 expression was induced with 1mMIPTG in bacterial cultures and harvested 6 hours after induction. Cells were pelleted by centrifugation and then resuspended in 30 ml Tris buffered saline (pH 7.2). Resuspended cells were then sonicated, and cellular debris removed via centrifugation. Lysates were concentrated using centrifugal filtration to a final volume of 3 ml.
DNA oligos containing either the proximal or distal mouse consensus ER81 binding sites were synthesized so that there was a 4bp 5′ overhang containing one dT nucleotide. Oligos were labeled with 32P via fill-in reaction of the 5′ overhang using 32P-α dATP and Klenow fragment (New England Biolabs). Unincorporated nucleotides were removed using Bio-30 spin columns (Bio-Rad). The following oligo sequences were used: rodent (both rat and mouse) proximal site forward strand 5′-GGAGAGGATGCGCAGGA-3′, rodent proximal site reverse strand 5′-TCCTGCGCATCCTCTCCACGC-3′; human proximal site forward strand 5′-GGGGTGGGGGATGTAAG-3′, human proximal site reverse strand 5′-CCTCCTTACATCCCCCACCCC-3′.
Recombinant protein lysate and labeled DNA oligos were incubated for 30 minutes at room temperature. Binding reactions consisted of 1 μl labeled DNA, 2μl Herring sperm DNA (10mg/ml stock), 2 μl BSA (100mg/ml stock), 1 μl NP-40 and 5, 10, 20 or 30 μl of bacterial lysate. Binding reaction volumes were adjusted with Tris-buffered saline pH 7.2 so that the final volume was 36 μl. Protein-DNA complexes were resolved on 4.5% 29:1 acrylamide/bis-acrylamide gels at a running voltage of 215V for 2 hours. Gels were imaged using a BAS2500 phosphoimager (Fujifilm).
Mammalian Th promoter nucleotide sequences were obtained from genomic assemblies NCBIm37, Rnor3.4, GRCh37, Btau4.0 and CanFam2.0 for the rat, mouse, human, cow and dog, respectively, which are available from Ensembl (http://www.ensembl.org). Promoter regions were aligned using the MLAGAN alignment program available at http://lagan.stanford.edu (Brudno et al., 2003).
Individual genes within the dopaminergic gene cassette are expressed at substantially different basal levels in the OB. Of the five genes in the cassette, RT-PCR analysis revealed that both Th and Dat were expressed at the highest levels, whereas Aadc and Vmat2 were expressed at substantially lower levels, and Gtpch was nearly undetectable (Figure 1A). The extremely low Gtpch expression levels precluded further analysis of this gene in the present study. The different expression levels of the individual members in the dopaminergic cassette are consistent with reported immunohistological analyses of these genes in the OB (Baker et al., 1991; Peter et al., 1995; Revay et al., 1996; Hwang et al., 1998). Despite these dramatically different expression levels, dopamine production in the OB was previously confirmed (Baker et al., 1983).
To test whether the expression levels of the dopamine gene cassette members were dependent on odor-mediated sensory activity, the OB of mice subjected to unilateral naris occlusion were analyzed by quantitative RT-PCR (qRT-PCR). This analysis revealed that naris closure had no effect on the Aadc and Vmat2 expression, whereas Dat expression was slightly lowered and Th was strongly reduced (Figure 1B). Also, ER81 protein levels were substantially down-regulated in the OB ipsilateral to naris occlusion (Figure 2), and qRT-PCR revealed a 72 ± 22% decrease in Er81 mRNA expression levels. Thus, the non-uniform response in the expression of both Er81 and the gene cassette members to naris closure was inconsistent with the putative regulatory role of Er81 forthese genes in the OB.
To determine whether Er81 was necessary for expression of the dopaminergic gene cassette members in the OB, olfactory bulbs from homozygous Er81 mutant mice were analyzed by qRT-PCR. The mutant mouse strain used for this analysis contained an insertion in exon 11 of the Er81 gene that disrupts the ETS-DNA binding domain and prevents the mutant protein from functioning as a transcription factor (Arber et al., 2000). Mice used for this analysis were approximately 10 days old (P9-P11), rather than fetal or neonatal pups because the vast majority of OB interneurons are generated post-natally (Hinds, 1968). Examination of older mice was not possible because the homozygous Er81 mutant mice rarely lived past day P14.
The qRT-PCR analysis revealed that none of the genes in the dopaminergic cassette, except Th, were significantly down-regulated in the OB of homozygous mutant mice relative to wild-type litter-mate controls (Figure 3). These findings suggest that Er81 is only necessary for Th expression in the OB, but not for the other members of the dopamine gene cassette.
Er81 is a member of the PEA3 sub-family of ETS-domain transcription factors, which also contains Pea3 (Etv4) and Erm (Etv5). Since the amino acid sequences of the ETS DNA-binding domain of all PEA3 group proteins are nearly identical, these proteins are likely to have identical binding sites (de Launoit et al., 1997). In the absence of functional ER81, either PEA3 or ERM might maintain expression of the dopamine gene cassette, except for Th, within the OB. However, immunohistological analysis revealed that PEA3is not expressed in the OB, but only in the SVZ and RMS (Figure 4A–C). By contrast, immunofluorescence analysis found that ERM is expressed in the both the glomerular and mitral cell layers of the OB, but ERM is not co-expressed with ER81 in these layers (Figure 5). Also, quantitative RT-PCR analysis of Erm and Pea3 gene expression in the OB of wild-type and Er81 mutant mice revealed the neither of these genes are upregulated in the mutant OB (data not shown).
The results above clearly showed that Er81 was critical for proper expression of Th in the OB. Immunohistochemical analysis revealed a 70±9% reduction in the number of TH expressing cells in the OB of Er81 mutant mice (240±43 vs. 72±21 cells/OB for Er81+/+ and Er81−/−, respectively; Figure 6A and 6B). However, since TH expression in the OB is activity-dependent, the observed decrease in TH expression may have resulted from a disruption in afferent synaptic activity in the olfactory receptor neurons (ORNs). Immunostaining for Olfactory Marker Protein (OMP) confirmed that ORN axons ramify within the glomeruli of the Er81 mutant mice (Figure 7A and 7B). Furthermore, since c-FOS expression in the OB glomerular layer is activity-dependent (Guthrie et al., 1993; Klintsova et al., 1995; Jin et al., 1996; Liu et al., 1999), the presence of c-FOS expression in the OB glomerular layer of mutant mice indicated the loss of functional ER81 did not disrupt the synaptic activity of the ORNs (Figure 7C and 7D).
Since ER81 is expressed in the neurogenic regions of the SVZ and RMS, the mutation of Er81 may have disrupted the generation and/or migration of OB dopaminergic neuronal progenitors and reduced TH expression in the OB. However, immunohistochemical analysis of PAX6 (Figure 8A and 8B), which is co-expressed with Th in OB neurons (Dellovade et al., 1998), revealed that PAX6+ cell counts in the glomerular layer were unchanged in Er81 mutant mice (86±12 cells/mm2 for Er81+/+ vs. 89±16 cells/mm2 for Er81−/−). Together these findings indicate Er81 is critical for cell autonomous regulation of Th expression in the OB dopaminergic neurons.
Inspection of the 9kb upstream regions of mouse and rat Th genes identified two conserved consensus ER81 binding sites (Figure 9A). Previous studies have shown that the 4.5kb upstream region of the Th gene is sufficient to mediate brain region-specific reporter gene expression in vivo (Schimmel et al., 1999), which suggested that only the consensus ER81 binding site in the proximal promoter is critical for OB-specific Th expression. The proximal consensus binding site also overlaps with the DA-motif (Figure 9B) (Flames and Hobert, 2009). Chromatin immunoprecipitation (ChIP) assays using mouse OB tissue indicated that the proximal consensus binding site is bound by ER81 in vivo (Figure 9C). The previously identified and conserved ER81 binding site in the Mmp-13 gene (Mmp-1/collagenase; Supplemental Figure S5)(Bosc et al., 2001) was used as a positive control (Figure 9C). Since previous studies showed that the rodent proximal binding site is functional for mediating Th promoter activity in primary OB cells (Flames and Hobert, 2009), the ChIP findings in the current study suggest that ER81 regulates Th expression by directly binding the Th proximal promoter.
Although regulation of Th expression via ER81 binding to the DA-motif is suggested to be an evolutionarily conserved mechanism, analysis of the Th promoter from other mammals (humans, cows and dogs) revealed that the proximal ER81 binding site/DA-motif is not conserved (Figure 9B). The potential ER81 binding site inthe human Th promoter conforms to the DA-motif sequence, but the 5′-nucleotide adjacent to the core ETS recognition sequence is mutated and does not conform to the consensus ER81 binding site. Even though the core ETS DNA binding sequence is conserved in both the rodent and human sequences, electromobility shift assays revealed that ER81 was capable of only binding to the rodent, but not to the human, sequence (Figure 9D). As compared to humans, the equivalent regions in the cow and dog Th promoters are even more divergent from the rodent sequence, including mutations to the core ETS biding motif (Figure 9B). These observations suggest that it is unlikely that ER81 binds to either the cow or dog Th promoters.
ChIP assays were performed with dog OBs to test the whether ER81 binds the orthologous region in the dog Th proximal promoter. However, as shown in Figure 9C, ER81 binding detected with only the control binding site in the Mmp-13, and not with Th proximal promoter. Together, findings in this study suggested that the regulation of OB Th expression by the direct binding of ER81 to the Th promoter is a species-dependent molecular mechanism.
This study showed that Er81 is necessary for the expression of the mouse OB dopaminergic neuron phenotype through the regulation of Th expression. In naris occluded mice, both mRNA and protein expression levels of Er81 and Th showed strong activity-dependence. Th expression was also significantly reduced in Er81 mutant mice and this reduction was not due to either a loss of afferent ORN activity or diminished proliferation/migration of progenitor cells. The reduction of Th was also not attributed to an increase in apoptosis since immunohistochemical analysis in the SVZ, RMS and OB of wild-type and Er81 mutant litter mates revealed no differences in the number of cells containing activated Caspase-3 (data not shown). Together, these findings suggest that Er81 regulates Th expression through a cell autonomous mechanism.
The ChIP and EMSA experiments in this study support a molecular mechanism in which ER81 regulates Th expression in rodents by directly binding to a conserved consensus binding site in the Th proximal promoter. However, this consensus binding site in the rodents is not conserved in the Th proximal promoters of other mammals (human, dog and cow). In humans, the core ETS recognition sequence is present, but the 5′ nucleotide adjacent to the core motif is mutated and does not match the degenerate consensus binding site sequence. This mutation appears to render to the site non-functional since ER81 failed to bind the human sequence in EMSA experiments. The equivalent regions in the dog and cow Th promoter are even more divergent, and the ChIP assays with the dog OB tissue indicated that ER81 does not bind the dog Th promoter. Together, these findings suggest that ER81 does not regulate Th expression in these other mammals by direct binding to the proximal promoter. Thus, the molecular mechanism of ER81 directly binding the Th proximal promoter is likely species-dependent.
Although direct binding to the Th promoter may not be a conserved molecular mechanism, both the conservation of Er81 expression in the vertebrate brain (Chen et al., 1999; Zhu and Guthrie, 2002; Yoneshima et al., 2006; Langevin et al., 2007) and the co-expression of the Er81 orthologue ast-1 with Th in C. elegans (Flames and Hobert, 2009) suggest that regulation of the Th expression in the mammalian OB by Er81 is conserved. This conserved mechanism may involve ER81 binding to distal enhancer regions that are presently unidentified. Alternatively, ER81 may indirectly regulate Th by activating genes that encode for transcription factors which directly bind to either enhancer or promoter regions of Th.
Unlike Th, expression of the other dopamine gene cassette members in the OB did not require Er81. Although Er81 expression levels were strongly activity-dependent, Dat expression was only slightly activity-dependent and Aadc and Vmat2 levels were insensitive to naris closure. The results for Aadc were consistent with previous in situ studies with unilateral naris occluded rats (Stone et al., 1990). Analysis of Er81 mutant mice showed that neither Aadc, Vmat2 nor Dat expression levels were significantly decreased by the loss of functional ER81. In contrast to a recently proposed model for differentiation of dopaminergic neurons (Flames and Hobert, 2009), these findings indicate that Er81 is not necessary for the expression of Aadc, Vmat2 or Dat in the OB.
ER81 has been suggested to regulate expression of the dopaminergic cassette genes in the OB by direct binding to “dopamine (DA) motif” nucleotide sequences that are near the transcription start site in these genes (Flames and Hobert, 2009). Although the DA-motif sequence overlaps with the established ER81 consensus binding site, our study does not support a functional role for Er81 in the regulation of either Aadc, Vmat2 or Dat. Consistent with this finding, ChIP assays with mouse OB tissue suggest that ER81 does not bind identified DA-motifs in Vmat2, Dat and Gtpch (Supplemental Figure S6). Furthermore, the DA-motifs identified near the mouse Aadc, Vmat2, Dat and Gtpch genes are not conserved in other mammals (Supplementary Figures S1–S4), and ChIP assays with dog OB tissue do not immunoprecipitate the regions of dog dopaminergic cassette genes that are orthologous to the mouse genes that contain DA-motifs (Supplemental Figure S6). These results, together with the findings that ER81 is not necessary for expression of the entire dopaminergic gene cassette in the OB, do not support a significant role for the DA-motif in the differentiation of mammalian OB dopaminergic interneurons.
There is little or no evidence suggesting a role for the DA-motif in differentiation of dopaminergic neurons outside of the OB. The functionality of the DA-motif/ER81 consensus binding sites have not been demonstrated in dopaminergic neurons from other regions, such as the substantia nigra. The lack of conservation in the various mammalian dopaminergic gene cassette promoters also suggests that the DA-motif may not be significantfor dopamine neurons in general. The ER81 homologue, Erm (Etv5), has been suggested to mediate dopaminergic differentiation in mammalian non-OB dopaminergic neurons (Flames and Hobert, 2009), but there is no reported evidence that ERM is co-expressed with, or recruited to the regulatory regions of, the dopaminergic cassette genes.
In the OB glomerular layer, ER81 is co-expressed with both Calretinin and TH, but there is no co-expression of Calretinin and TH (Kosaka et al., 1995; Parrish-Aungst et al., 2007). These observations indicate that ER81 is necessary, but not sufficient for the OB dopaminergic phenotype. Previous studies have proposed that combinatorial molecular codes of transcription factors specify various glomerular layer interneuron phenotypes (Allen et al., 2007). These studies have suggested that a combination of ER81, PAX6 and MEIS2 expression forms a molecular code for OB dopaminergic interneurons since nearly all cells containing TH also express each of these factors. However, this current study revealed that the loss of ER81 did not completely eliminate TH expression in the OB (TH-immunoreactive cell counts were ~30% of wild-type controls). This observation suggests that there are other transcription factors that are redundant to ER81 in a subset of TH-expressing cells. However, paralogs of ER81, PEA3 and ERM, were not expressed in the proper OB cell types, and thus, do not account for this redundancy. Thus, there may be either other ETS DNA binding domain proteins or an entirely different class of transcription factors that are capable of replacing ER81 in a subset of OB dopaminergic neurons. An alternative, but not mutually exclusive, possibility is that the remaining TH-expressing cells in the OB of Er81 mutant may represent a sub-population of OB dopaminergic cells that have derived from an alternative origin. Recent studies have suggested that regions other than the LGE generate OB dopaminergic interneurons during embryonic development (Vergano-Vera et al., 2006; Inoue et al., 2007). The progenitors derived from these alternative origins may notdepend on ER81 for Th expression.
This complexity in the molecular-genetic pathways that regulate OB dopaminergic interneuron differentiation have also been reported for other OB interneuron phenotypes. Nearly all calretinin-expressing interneurons express SP8, but calretinin expression is lost from only about one-half of these cells in mice lacking SP8 (Waclaw et al., 2006). Together, these findings suggest that there may be multiple molecular transcription codes for individual glomerular interneuron phenotypes. Identification of the molecular transcription codes and elucidation of the mechanisms by which they regulate specific interneuron phenotypes remains a major challenge.
This work was supported by NIH R01DC008955 and BMRI