Novel subdomains of the mouse olfactory bulb defined by molecular heterogeneity in the nascent external plexiform and glomerular layers
© Williams et al; licensee BioMed Central Ltd. 2007
Received: 18 January 2007
Accepted: 16 May 2007
Published: 16 May 2007
In the mouse olfactory system, the role of the olfactory bulb in guiding olfactory sensory neuron (OSN) axons to their targets is poorly understood. What cell types within the bulb are necessary for targeting is unknown. What genes are important for this process is also unknown. Although projection neurons are not required, other cell-types within the external plexiform and glomerular layers also form synapses with OSNs. We hypothesized that these cells are important for targeting, and express spatially differentially expressed guidance cues that act to guide OSN axons within the bulb.
We used laser microdissection and microarray analysis to find genes that are differentially expressed along the dorsal-ventral, medial-lateral, and anterior-posterior axes of the bulb. The expression patterns of these genes divide the bulb into previously unrecognized subdomains. Interestingly, some genes are expressed in both the medial and lateral bulb, showing for the first time the existence of symmetric expression along this axis. We use a regeneration paradigm to show that several of these genes are altered in expression in response to deafferentation, consistent with the interpretation that they are expressed in cells that interact with OSNs.
We demonstrate that the nascent external plexiform and glomerular layers of the bulb can be divided into multiple domains based on the expression of these genes, several of which are known to function in axon guidance, synaptogenesis, and angiogenesis. These genes represent candidate guidance cues that may act to guide OSN axons within the bulb during targeting.
In the mouse olfactory system, each olfactory sensory neuron (OSN) extends an axon from the epithelium to the olfactory bulb to form synapses within structures called glomeruli . The pattern of glomerular convergence is remarkably stereotyped  and is strongly dependent upon the specific odorant receptor expressed by the OSN [3, 4]. But despite the number of cues that have been found that are expressed by OSNs or along the path traveled by OSNs to the bulb, surprisingly little is known about the role of the olfactory bulb itself in this process . Transplantation and bulbectomy studies have observed the formation of glomerular-like structures by OSNs within cortical tissue . Genetic ablation of mitral and tufted projection neurons showed that topographically correct convergence by P2-expressing OSNs still occurred in the absence of these target cells . Disruption of the laminar nature of the mitral layer also did not affect convergence by P2-expressing OSNs . These results have suggested that the olfactory bulb plays a minimal role, if any, in targeting [9, 10].
On the other hand, studies ablating most of the bulb, as opposed to eliminating only projection neurons, have observed topographically incorrect convergence by P2-expressing OSNs, suggesting a role for the bulb in determining the ultimate location of OSNs . Tracing studies have long suggested the existence of zone-to-zone mapping of OSNs onto corresponding domains within the bulb . In Drosophila, Dscam expression in the antennal lobe is critical for proper olfactory receptor neuron targeting , providing the first data that the target structure is important for convergence. These experiments in Drosophila suggest that the mammalian bulb may also provide guidance information to OSNs . Part of the uncertainty surrounding the role of the bulb in guidance is that it is unclear what other bulbar cells in addition to mitral and tufted projection neurons could provide targeting information. While radial glia have long been suggested as one possibility [7, 15], cells in the external plexiform (EPL) and glomerular (GL) layers also possess many attractive characteristics. First, in addition to projection neurons, other cells in the EPL and GL form synapses with OSNs . Second, recent studies point to a diversity of cell-types within these layers , and electrophysiological studies have shown the presence of different types of interneurons within the EPL . Third, some glomerular cell populations migrate into the bulb much earlier than previously thought . And finally, it has been shown that external tufted cells indirectly connect glomeruli that are innervated by OSNs expressing a common receptor . Collectively, these data suggest that little is known about the different sub-types and developmental profiles of these cells. In one model, a subset of these EPL and/or GL cells could provide guidance information to OSNs. If this hypothesis is correct, then it should be possible to identify genes expressed by cells within these layers that are important for guidance and targeting.
Here we have undertaken a high resolution screen for genes that spatially define distinct regions of the EPL and GL. We used laser microdissection (LMD) coupled with RNA amplification, microarrays, bioinformatics, and in situ hybridization to search for genes that are differentially expressed by cells within these layers. We identified 23 such cDNAs and show that the superficial layers of the bulb can be subdivided into large regions based upon the various expression patterns. The patterns demonstrate the existence of gradients, localized expression domains, and medial-lateral symmetry within the olfactory bulb. Moreover, these patterns are maintained in cells that do not express Tbr1, consistent with the interpretation that they are expressed in non-projection neurons. We further show, using a chemical ablation paradigm, evidence consistent with the interpretation that these genes are expressed in cells that interact with OSNs.
Spatial reconstruction of gene expression using laser microdissection
We synthesized the pool of microarray data into a three-dimensional representation of gene expression within the bulb [see Methods and Additional File 1]. As a result, all 33,000 cDNAs on the array could be queried to identify those whose expression patterns indicate differential expression within the bulb. For example, we examined whether or not expression for any given gene was higher or lower in the VM aspect relative to the remainder of the bulb. Genes were grouped into four categories [see Additional File 1] and the top 12 from each were tested by in situ hybridization. Of the 48 cDNAs, nine did not have any detectable signal. Of the remaining 39, 20 were spatially differentially expressed (51%). In one category (VL/DL), 83% were spatially differentially expressed. These results speak to the sensitivity of our analysis, as all samples differ essentially only by position. We also examined a small subset of genes chosen based on annotation. These genes, which were all ranked within the top 50, were selected because they correspond to known, transmembrane-bound molecules. As such, they would be more likely to directly interact with OSNs and influence targeting. In total, 23 cDNAs were found to be spatially varied in expression. As expected, all identified cDNAs were differentially expressed within the nascent EPL/GL, as confirmed by DAPI labeling and double-label in situ hybridization (see below). Expression was not necessarily exclusive to these layers, and was often observed in the mitral layer as well.
Spatial differential expression of candidate cues at E17.5
We identified a diversity of patterns that are spatially differentially expressed within the nascent EPL/GL. We found that many of our positive cDNAs corresponded to genes that are either secreted or are transmembrane-bound proteins. Gicerin is an immunoglobulin superfamily member that is a homophilic cell-adhesion molecule . Gicerin is expressed differentially in the lateral aspect of the anterior bulb (Fig. 2C). However, not all cells in this lateral domain express Gicerin, as the pattern appears stippled. Expression is also observed in the nerve layer. Strikingly, in the middle bulb, expression of Gicerin appears symmetric in some areas about the ML and DV axes, with high levels observed in the DM and VL bulb (Fig. 2D). In more posterior bulb, these expression domains gradually become connected with one another along the DL surface, forming a crescent-shaped expression domain (data not shown). Expression in the VM aspect, however, remains low throughout the AP extent of the bulb.
Connective tissue growth factor, or CTGF , is a secreted protein that is a member of a small family of genes which interact with the extracellular matrix and growth factors, and are important for angiogenesis . In the anterior bulb, CTGF is localized to the lateral aspect (Fig. 2E). Like Gicerin, this expression is not uniform and significantly higher expression occurs within a subset of these cells. In the middle and posterior bulb, as with Gicerin, expression is symmetric as higher CTGF expression is detected along the DM and VL surfaces (Fig. 2F). CTGF is weaker in the DL aspect and noticeably reduced in the VM bulb. CTGF is also expressed in cartilage and connective tissue surrounding the bulb.
Zinc finger csl-type containing 3 (Zcsl3) is a protein of unknown function . Throughout the AP axis, Zcsl3 is expressed in a gradient-like pattern that extends from the DM to the VL bulb (Fig. 2G), although expression tends to shift more ventrally in the middle bulb (Fig. 2H).
Protocadherins have been shown to possess both homophilic and heterophilic adhesion properties . Protocadherin 17 (Pcdh17; ) is expressed throughout the bulb, but is enhanced in the anterior bulb along the VM aspect in a small subset of cells that extend into the EPL (Fig. 2I). In the middle bulb, a small subset of cells in the VL nerve layer expresses this gene (Fig. 2J). Protocadherin 7 (Pcdh7; ) is expressed anteriorly most strongly along the medial aspect, although some signal can be seen laterally (Fig. 2K). Just beyond the anterior bulb, this pattern is reversed and Pcdh7 becomes strongest laterally, although some expression can still be detected medially (Fig. 2L). This pattern is maintained throughout the posterior bulb (data not shown).
Jagged1, a ligand for Notch, has been shown to be required for proper endothelial branching during angiogenesis . In the anterior bulb, Jagged1 is symmetrically expressed about the ML axis (Fig. 2M). Expression of Jagged1 in these domains appears more even than that of Gicerin or CTGF, and no obvious high or low-expressing cells are observed within the expression domain. In the middle bulb, Jagged1 maintains this ML symmetry, but weaker expression connects the two subdomains (Fig. 2N). A similar pattern is observed in posterior bulb (data not shown).
Although many of our genes are expressed in the nascent mitral layer, we emphasize that at least some expressing cells are located within the developing EPL. All of our in situ hybridization experiments were counterstained with DAPI so that we could correlate the location of the observed signal with the position of the nascent mitral and external plexiform layers. Delta4, another ligand for Notch , is expressed primarily within the mitral layer, and does not extend into the EPL (Fig. 2O). This can be seen both by DAPI staining and by the relatively compact expression pattern for this gene (Fig. 2O). Jagged1, on the other hand, extends beyond the mitral layer into the EPL, and its expression domain is broader than that of Delta4 (Fig. 2P). We further demonstrate that these genes are expressed in the nascent EPL/GL using double-label experiments and temporal analysis of their expression patterns, as described below.
These expression patterns underscore the strong spatial heterogeneity within the external plexiform layer. Moreover, they emphasize that there is variation along all three axes of the bulb. They also show for the first time that the bulb does indeed express genes that appear symmetric about the ML axis.
A spatially differentially expressed putative microRNA
Bioinformatic analysis showed that all 17 cDNAs contained intronic or 3' untranslated sequence (Fig. 3G). Pair-wise alignment identified a 134 basepair common element that contained two microRNA (miRNA) sequences, miR-706 and miR-709 (Fig. 3G). The remainder of the 134 bp element was ~ 95% identical to sequence immediately upstream of the miR-706 sequence , corresponding to what we assume is part of the primary miRNA (pri-miRNA) sequence for miR-706 (Fig. 3H).
The expression pattern we observed may therefore reflect the expression of miR-706 pri-miRNA. We obtained locked nucleic acid (LNA)-modified oligos corresponding to miR-706 and miR-138 (a control miRNA known to be expressed in the bulb; ). We performed in situ hybridization with the LNA oligos as well as with the 134 basepair common element. Hybridization patterns using the common element were identical to that observed with our 17 cDNAs (compare Fig. 3A with 3D). Although the signal for the miR-706 probe was weaker than that of the common element, punctate signal was also observed in the nascent EPL/GL in a similar pattern as well (compare Fig. 3A with 3E). These results contrast with those for the miR-138 probe, which bound weakly and uniformly to the nerve layer and EPL (Fig. 3F). The miRNA in situ patterns that have been published in the mouse generally show uniform expression within a given layer or structure (e.g. miR-138), and/or are detected in specific cell-types. The pattern of the miR-706 probe suggests, however, that miRNAs may also be present selectively within subsets of cells in a spatially distinct fashion. Additional studies are required to determine the possible targets of this putative miRNA.
Three-dimensional reconstruction of gene expression
Our identified expression patterns show that the bulb can be subdivided into different domains that are distinguished molecularly by the expression of these candidate genes. These domains have not been previously described, as only a handful of genes are known to be differentially expressed within a given layer of the bulb [33–35]. We were surprised by the complexity of the observed patterns, as prior mapping studies had suggested the existence of DV cues in the bulb , and theoretical models had suggested the existence of symmetric cues to label medial and lateral bulb . Our observed expression patterns, however, do not readily fit any of these prior suggested models. For example, Gicerin symmetrically labels the DM and VL bulb, but symmetry about these axes has not been previously predicted. To obtain a more detailed understanding of the expression domains distinguished by these individual genes, we generated a three-dimensional reconstruction of gene expression for these genes within the bulb.
Temporally dynamic expression of candidate cues during development – E14
The first OSNs are known to extend an axon towards the developing bulb at E11–12, and "wait" at the periphery as the bulb develops . Sema3A expression at the periphery of the bulb, expressed by ensheathing glia, plays a role in maintaining OSNs outside of the bulb . It is known that mitral cell reorientation requires innervation by a small population of early OSNs  that enter the bulb in a medial-to-lateral progression . Why this innervation proceeds initially along the medial surface is not known. One possibility would be that additional cues, in conjunction with Sema3A expression, serve to guide OSN axons to their position within the bulb. We therefore examined the expression of our candidate cues at E14 for differential expression.
Several of the observed patterns at E14 mimic those seen at E17.5. For example, the putative miRNA is weakly expressed in the VM bulb at both stages. Similarly, the dual-layer expression of Pcdh17 in the VM bulb at E14 is reflected at E17.5. Jagged1 is ventrally expressed at both stages. On the other hand, other genes change in expression between E14 and E17.5. Gicerin, for example, is expressed most strongly at E14 in DL bulb, but at E17.5 is present more strongly in the DM and VL bulb. CTGF also shifts in expression along different axes from E14 to E17.5, increasing in expression laterally and anteriorly as development proceeds.
Temporally dynamic expression of candidate cues during development – P3
We next examined the expression of these various candidate cues at postnatal stages. While convergence has been shown to occur at E17.5 for OSNs expressing the P2 odorant receptor, other OSN populations do not converge until later stages, up to postnatal day 3 [40, 41]. The existence of relatively late converging OSNs prompted us to examine the expression of these genes at P3.
As with the data regarding expression at E14, these patterns indicate the temporally dynamic nature of expression by these various genes. Gicerin, for example, is expressed most highly in the DL bulb at E14, but is DM and VL in both the E17.5 and P3 bulb. Pcdh17 is VM and DL at E14, VM at E17, and uniform at P3. We note that, consistent with our initial observations at E17.5, Zcsl3, the putative miRNA, Jagged1, Pcdh17, and Pcdh7 expression can clearly be detected in the EPL, GL, and mitral layers.
Lack of precise colocalization with Tbr1
We have hypothesized that cells within the EPL express guidance cues for OSNs. Genetic studies using the Tbr1 mutant mouse resulted in the ablation of most projection neurons in the bulb . Projection neurons include mitral cells as well as internal and middle tufted cells . As targeting is unaffected in Tbr1 mutant mice, a major problem has been to determine what cell-types within the bulb remained that could be important for this process. If our genes are expressed in cells that provide guidance information to OSNs, we would predict that at least some cells that express our candidate genes do not express Tbr1, and would not be ablated in the mutant.
Expression of candidate cues are affected in chemically ablated mice
While we have shown the existence of a number of genes demarcating different domains within the bulb, it remains unknown whether or not they are functionally relevant for OSN guidance. Unfortunately, no in vitro system exists in the olfactory system to test this hypothesis, and genetic analysis has remained the primary means for assessing function. To more rapidly determine the potential role of these genes in targeting, we decided to employ a chemical ablation approach.
The olfactory epithelium is continuously being regenerated during adult life. Prior studies have shown that Tyrosine hydroxylase (TH) expression is drastically reduced following intranasal ablation with zinc sulfate . As regeneration proceeds, TH expression gradually recovers. Although TH is an intracellular enzyme, it has been proposed that this recovery is dependent upon afferent input from regenerating OSNs. This process, although poorly understood, is thought to be a result of transneuronal regulation of glomerular neurons by OSNs .
We found the bulb on the ablated side is smaller as previously reported  due to a significant reduction in the size of the nerve and glomerular layers (data not shown). Based on DAPI staining, the number of cells in the glomerular layer did not appear to be reduced (Fig. 8C–D), consistent with previous studies showing that little or no cell death results from the ablation process [43, 44]. We found that GluR1 expression appeared to be reduced in the glomerular layer and increased in the mitral layer at 24 days (Fig. 8E) but was equal in expression between ablated and control bulbs by 40 days (Fig. 8F). This is reminiscent of alterations in GluR1 expression in naris occluded animals . We confirmed that TH was absent from the ablated side at 24 days (Fig. 8G) but recovered by 40 days post-ablation (Fig. 8H).
We then tested a subset of our genes on these ablated animals. Interestingly, we found a variety of different responses to the ablation. CTGF is expressed in the adult in the glomerular and mitral layers (Fig. 8I). This expression is absent in bulbs on the ablated side at 24 days (Fig. 8I) but recovers by 40 days (Fig. 8J). Pcdh7 is expressed in the glomerular layer in adults, with weaker, more scattered expression in the mitral layer (Fig. 8K). Like CTGF, both layers are affected in the ablated side at 24 days (Fig. 8K), but expression recovers by 40 days post-ablation (Fig. 8L). Pcdh17 is also expressed in the glomerular and mitral layers (Fig. 8M). However, in the ablated side, only the glomerular layer expression of Pcdh17 is affected, while mitral layer expression remains constant (Fig. 8M). Expression in the glomerular layer returns to control levels by 40 days (Fig. 8N). We did not find any differences in expression for the putative miRNA or Zcsl3 in bulbs in ablated animals (data not shown). Based on prior studies showing TH regulation by OSN innervation, we conclude that several of our genes are likely to be expressed in cells that interact directly with OSNs.
Pcdh7 and Pcdh17are expressed in subsets of OSNs
Two of our genes, Pcdh7 and Pcdh17, are members of the cadherin superfamily. Protocadherins are known to possess weak homophilic adhesion properties in vitro [46, 47]. In one model, protocadherins expressed by OSNs would bind their cognate partner expressed by nascent EPL/GL cells within the bulb. The specificity of these interactions would contribute towards the final glomerular convergence pattern.
To our knowledge, no cell-type or gene expressed within the bulb has been functionally proven to provide spatial guidance information to OSNs. Sema3A, for example, has been genetically shown to affect OSN targeting, but this is due to its expression in ensheathing glia  which migrate to the nerve layer from the epithelium. A number of genes are uniformly expressed within cells derived from the bulb (e.g. Sema3A ), but it is difficult to understand how, given this uniform expression, these genes would guide OSN axons to different domains within the bulb. The lack of viable potential guidance cues has led instead to genetic approaches that ablated either whole classes of cell-types within the bulb (e.g. Tbr1; ) or eliminated much of the bulb entirely (e.g. Extratoes; ). The absence of suitable molecular candidates that could potentially guide OSN axons to subdomains of the bulb has strongly inhibited any study of the role of the bulb in guidance.
In contrast, studies manipulating odorant receptor expression have led to models where the odorant receptor plays a significant role in guidance [3, 4, 49]. The recent discovery that the Kirrel 2/3 and Ephrin/Eph A5 expression are controlled by odorant receptor activity provides new insight into how receptor choice influences glomerular positioning. However, in all of these receptor swap experiments, it is clear that the receptor alone does not mediate all aspects of glomerular targeting. The dorsal-ventral position of the OSN within the epithelium is thought to provide position-dependent information as well . But what genes comprise these position-dependent determinants, and what potential partners are bound by such determinants are still unknown. Moreover, why OSNs innervate both a medial and a lateral glomerulus has not been explained by studies manipulating odorant receptor expression. The presence of gradients  or of mirror-symmetric expression patterns within the bulb  have been suggested as potential explanations for how topographic OSN projection patterns are established. However, considerable disagreement exists over the role of the bulb in guidance .
The role of bulbar projection neurons in guiding OSN axons has been examined using different genetic paradigms [7, 8]. These studies have suggested that these cells are not necessary for topographic convergence of OSNs. While these analyses only examined OSNs that express the P2 odorant receptor, mitral cell tracing studies also do not necessarily support a role for these cells in guidance. Mitral cell dendritic arbors initially are distributed among several glomeruli, but are refined during the first week of postnatal development to a single glomerulus . It is difficult to understand, if mitral cells do indeed express the primary cues needed for OSN guidance, why such cues would be distributed among multiple glomeruli initially, and the dendritic arbor refined only after convergence has been completed. But despite the fact that these studies indicate that mitral cells may not play a primary role in guidance, genetic ablation studies removing the entire bulb  have shown that topographically correct convergence by P2-expressing OSNs does not occur. One interpretion of these results is that other cell-types within the bulb are important for providing guidance information to OSNs. We theorized that these cells correspond to those in the EPL and GL.
In this hypothesis-driven microarray-based experiment, we show that the nascent EPL/GL of the bulb can be subdivided into multiple domains based on the expression of our identified genes. At E17.5, the first time that convergence is observed , the anterior bulb can be distinguished molecularly both laterally (CTGF, Gicerin, Jagged1) and medially (Jagged1, Pcdh7). In the middle and posterior bulb, Zcsl3 is expressed in a gradient that extends from dorsal to ventrolateral, while a probe containing the miR-706 sequence is expressed in a subset of cells that span a similar region. Perhaps most interestingly, we find that Gicerin, CTGF, and Jagged1 are expressed in a symmetric manner. While it is well-known that OSNs project to medial and lateral glomeruli, the existence of mirror-symmetric cues has, to date, only been hypothesized . However, our patterns suggest that this medial-lateral symmetry is more complex than previously thought. Gicerin and CTGF are roughly symmetric about both the DV and ML axes, while Jagged1 appears symmetric only around the ML axis. Thus, each hemi-bulb can be further distinguished based on dorsal or ventral expression of Gicerin and CTGF.
Several lines of evidence are consistent with the expression of these candidate cues within cells of the EPL. First, we show by counterstaining with DAPI that expression of these genes does not colocalize precisely with the nascent mitral layer. Second, we show at P3 and in adult that many of these genes are expressed in the EPL and GL, consistent with their expression at E17.5 in the nascent EPL/GL. Finally, we show that there are populations of cells at E17.5 that express our candidate genes but do not express Tbr1, a marker for projection neurons. Thus, we have shown using a variety of approaches that, in addition to mitral layer expression, nascent EPL/GL cells also express these genes.
Interestingly, in addition to the fact that not all cells that express our genes express Tbr1, the cells that do not express Tbr1 are those that show most clearly the differential expression patterns observed with our genes. For example, Pcdh17 is strongly and selectively expressed in a subset of cells in the VM aspect of the anterior bulb. These same cells do not express Tbr1. Thus, in Tbr1 mutant animals, we predict that these cells would not only be present, but that the observed differential expression patterns would also still be present. These cells may therefore provide a potential solution to the problem of which cells within the bulb could express guidance cues important for OSN guidance.
Mitral cells are known to mature in a medial-to-lateral fashion from E14–E17 , and a rostral-caudal wave of development occurs during glomerular formation . However, these developmental events cannot readily explain our various observed patterns. For example, Gicerin and CTGF are expressed in subsets of cells in the lateral domain in anterior bulb, while Jagged1 and Pcdh7 are expressed medially. Several of our genes are both medially and laterally expressed. It seems unlikely that these patterns can be explained solely on the basis of neuronal maturation events during development.
Are any of our identified cDNAs involved in guiding OSNs to the bulb? An alternative interpretation would be that these genes may be important for establishing intrabulbar circuits. External tufted cells are known to indirectly connect glomeruli innervated by a common population of OSNs . Thus, these genes may act to establish these connections. Yet another potential function of these genes can be suggested based on their expression in the mitral layer. These genes may act to mediate connectivity between projection neurons and higher cortical synaptic partners. Finally, these genes may not be involved in axon guidance per se, but instead are important for events surrounding synaptogenesis.
To provide support for the hypothesis these genes are important for guiding OSN axons, we used a chemical ablation approach . Expression of Tyrosine hydroxylase (TH) is known to be down-regulated initially in ablated animals, but reappears as reinnervation of the bulb proceeds. Thus, TH expression is influenced by OSN innervation. Similarly, this approach has also been used to study Sema3A expression in axotomized bulbs . In these experiments, normal expression of Sema3A is greatly reduced in axotomized bulbs. Expression is instead upregulated near the site of axotomy. Two months after ablation, Sema3A expression reappeared at its normal location. These results indicate that cells which interact with OSNs can display molecular alterations in their expression profile upon deafferentation. We showed that CTGF, Pcdh17, and Pcdh7 were altered in expression in response to ablation. Based on prior studies of TH, we conclude that these genes are expressed in cells that interact with OSNs.
We note that several of our candidate cues correspond to genes known to be important for guiding axons elsewhere in the nervous system (Pcdh7, Pcdh17, Gicerin), or are involved in angiogenesis (Jagged1, CTGF). CTGF and Gicerin are known to mediate cell-cell adhesion in vitro. Cadherins have been shown to be play multiple roles during development, from neurite outgrowth to synaptogenesis . The changing patterns of expression of Pcdh7 and Pcdh17 during development may signal multiple roles as OSN innervation of the bulb proceeds. Pcdh7 and Pcdh17 may operate initially as guidance cues based on their differential expression within the bulb and epithelium. After growth of OSNs into the bulb, they may then act secondarily in synaptogenesis. While definitive proof requires genetic manipulation of the individual candidates, these genes provide novel targets that will enable new approaches towards determining the role of the bulb in guidance.
We have identified several genes whose expression patterns define various subdomains within the developing mouse olfactory bulb. We show that these genes are expressed in cells within the nascent EPL/GL and ML, and that some of these cells do not express Tbr1. Our regeneration paradigm suggests that these genes are expressed in cells that interact with OSNs. These genes therefore represent novel candidate guidance cues that may guide OSN axons to appropriate locations within the bulb.
Swiss-Webster mice were used for all experiments. The day a vaginal plug was observed was day P0.5. All protocols were approved by the Cornell IACUC.
Laser microdissection and RNA isolation
E17.5 embryos were embedded in 3% agarose and fresh frozen in isopentane/liquid nitrogen as described . LMD samples were collected into 30 μl of lysis buffer and RNA isolated using Gentra's PureScript kit. Total RNA yield was ~ 200 pg/LMD isolate using the Agilent PicoChip Bioanalyzer assay.
A modified T7 RNA polymerase amplification protocol was performed using an optimized T7 primer . First-round aRNA was tailed with poly(A) polymerase, providing a template for T7-d(T) primer and the generation of a second round of aRNA. For the arrays, 20% of this aRNA was amplified a third time, and a portion labeled by priming with amine-terminal random hexamers and incorporating a modified amino-allyl dUTP . Hybridizations were performed in 50% formamide, 3xSSC, 20 mM Tris 8.5, 0.1% SDS, 0.1 μg/μl salmon sperm DNA, and 0.1% BSA at 45°C using a Tecan HS400 machine and scanned on an Axon 4000B scanner.
PCR products from the NIA 15 k , 7.4 k , and 11 k BMAP (brainEST.eng.uiowa.edu) collections were resuspended in 1xSSC/0.005% sarkosyl and printed on Corning UltraGAPS slides using a custom-built microarrayer and Telechem 946 pins.
Experimental design. For each section, the four LMD samples were hybridized directly against one another using a loop design [see Additional File 1]. The directions of the loops (clockwise or counter-clockwise) were alternated along the AP axis to control for dye bias.
Background measurements were not subtracted to reduce variation in weakly expressing genes . Within slide normalization (print-tip loess) was applied  and a fixed effects linear model was employed to identify spatial differentially expressed genes . For a typical gene g, the intensity value corresponding to the four different sub-regions of the bulb is denoted by DM k ,DL k ,VM k ,VL k , where k represents the different sections taken along the AP-axis. The log transformation of these values is defined as dm k = log2 DM k , dl k = log2 DL k , vm k = log2 VM k and vl k = log2 VL k . To estimate the spatial gene expression for gene g, all of the spatial sections along the AP-axis are treated as replicates and used to fit four separate linear models for each of the comparisons of interest α1 = dm-dl, α2 = dm-vm, α3 = vl-dl and α4 = vm-vl . The linear model is written as
M ik = α i + ε ik
where i = 1,,4 representing the four main comparisons of interest; M ik is a vector of log-ratios and var(ε ik ) = σ i This model is equivalent to performing four separate series of one-sample comparisons on four separate sets of data. Estimates of the coefficient of variation, moderated t-statistics, and adjusted p-values were calculated using lmFit in the library limma . Comparison of this model with other analysis options, as well as gene lists for the individual groups, are presented as an additional file [see Additional File 1].
Probes were generated from the original clone printed on the array. Hybridization to sections was essentially as described . Fixed sections were treated with 10 μg/ml proteinase K prior to hybridization at 60°C. Double label in situ hybridization was performed with digoxigenin and biotin-labeled probes and detected using the Fast Red/HNPP substrate or the TSA Renaissance kit as per manufacturer's instructions. Digoxigenin-labeled locked nucleic acid (LNA)-oligos from Exiqon were hybridized at 50°C. Additional clones within the clonesets corresponding to Pcdh7 (clones H3060E02 and H3067F12) and Pcdh17 (clone AI843903) were used to confirm the observed pattern for these genes. Additional probes for other genes were produced by cloning various subregions: Jagged1 (nt716–1514 and 3747–4559), Zcsl3 (nt83–1004), Gicerin (nt23–1105, 904–2102, 1592–2249, 1993–2873), CTGF (nt1102–1697, 1686–2267), Tbr1 (nt521–1122 and nt2798–3398).
A solution of 5% zinc sulfate in PBS was injected using a 1 ml hypodermic syringe drawn out to produce a thin, flexible tube that can be easily inserted into the nares. Mice were anesthetized with isofluorane prior to instillation of 50–75 μl of zinc sulfate solution.
In situ hybridization data from E17.5 coronal sections were obtained for each gene every 60–80 μm along the AP extent of the bulb. The expression patterns for multiple bulbs were compared against one another, and were found to be similar. We therefore performed reconstructions using data from an individual bulb hybridized with any given probe. Images from hybridized sections were imported into Adobe Photoshop, and regions of high expression were selected and enhanced. These enhanced images were combined using ImageJ to create individual stacks of images corresponding to a given gene. A projection of the stacked images was then generated to produce a three-dimensional pattern of gene expression. These projections were then superimposed upon a cartoon depiction of the bulb to produce the final image.
DML was funded by the Sloan, Whitehall, and Beckman Foundations, and NIH grant DC07489. DML, GY, and JYY were supported by the CRCNS program of the NSF (0218521).
- Shepherd GM, Chen WR, Greer CA: Olfactory bulb. The Synaptic Organization of the Brain. Edited by: Shepherd GM. 2004, New York, Oxford University Press, 165-216. 5thView ArticleGoogle Scholar
- Buck LB: Olfactory receptors and odor coding in mammals. Nutr Rev. 2004, 62: S184-8; discussion S224-41. 10.1301/nr.2004.nov.S184-S188.View ArticlePubMedGoogle Scholar
- Imai T, Suzuki M, Sakano H: Odorant Receptor-Derived cAMP Signals Direct Axonal Targeting. Science. 2006Google Scholar
- Feinstein P, Mombaerts P: A contextual model for axonal sorting into glomeruli in the mouse olfactory system. Cell. 2004, 117: 817-831. 10.1016/j.cell.2004.05.011.View ArticlePubMedGoogle Scholar
- Nedelec S, Dubacq C, Trembleau A: Morphological and molecular features of the mammalian olfactory sensory neuron axons: What makes these axons so special?. J Neurocytol. 2005, 34: 49-64. 10.1007/s11068-005-5047-7.View ArticlePubMedGoogle Scholar
- Graziadei PP, Monti Graziadei GA: Neuronal changes in the forebrain of mice following penetration by regenerating olfactory axons. J Comp Neurol. 1986, 247: 344-356. 10.1002/cne.902470305.View ArticlePubMedGoogle Scholar
- Bulfone A, Wang F, Hevner R, Anderson S, Cutforth T, Chen S, Meneses J, Pedersen R, Axel R, Rubenstein JL: An olfactory sensory map develops in the absence of normal projection neurons or GABAergic interneurons. Neuron. 1998, 21: 1273-1282. 10.1016/S0896-6273(00)80647-9.View ArticlePubMedGoogle Scholar
- Royal SJ, Gambello MJ, Wynshaw-Boris A, Key B, Clarris HJ: Laminar disorganisation of mitral cells in the olfactory bulb does not affect topographic targeting of primary olfactory axons. Brain Res. 2002, 932: 1-9. 10.1016/S0006-8993(01)03384-4.View ArticlePubMedGoogle Scholar
- Mombaerts P: Axonal wiring in the mouse olfactory system. Annu Rev Cell Dev Biol. 2006, 22: 713-737. 10.1146/annurev.cellbio.21.012804.093915.View ArticlePubMedGoogle Scholar
- Graziadei PP, Monti Graziadei GA: Principles of organization of the vertebrate olfactory glomerulus: an hypothesis. Neuroscience. 1986, 19: 1025-1035. 10.1016/0306-4522(86)90119-3.View ArticlePubMedGoogle Scholar
- St John JA, Clarris HJ, McKeown S, Royal S, Key B: Sorting and convergence of primary olfactory axons are independent of the olfactory bulb. J Comp Neurol. 2003, 464: 131-140. 10.1002/cne.10777.View ArticlePubMedGoogle Scholar
- Saucier D, Astic L: Analysis of the topographical organization of olfactory epithelium projections in the rat. Brain Res Bull. 1986, 16: 455-462. 10.1016/0361-9230(86)90173-5.View ArticlePubMedGoogle Scholar
- Zhu H, Hummel T, Clemens JC, Berdnik D, Zipursky SL, Luo L: Dendritic patterning by Dscam and synaptic partner matching in the Drosophila antennal lobe. Nat Neurosci. 2006, 9: 349-355. 10.1038/nn1652.View ArticlePubMedGoogle Scholar
- Komiyama T, Luo L: Development of wiring specificity in the olfactory system. Curr Opin Neurobiol. 2006, 16: 67-73. 10.1016/j.conb.2005.12.002.View ArticlePubMedGoogle Scholar
- Lin DM, Ngai J: Development of the vertebrate main olfactory system. Curr Opin Neurobiol. 1999, 9: 74-78. 10.1016/S0959-4388(99)80009-9.View ArticlePubMedGoogle Scholar
- Kosaka K, Kosaka T: synaptic organization of the glomerulus in the main olfactory bulb: compartments of the glomerulus and heterogeneity of the periglomerular cells. Anat Sci Int. 2005, 80: 80-90. 10.1111/j.1447-073x.2005.00092.x.View ArticlePubMedGoogle Scholar
- Hamilton KA, Heinbockel T, Ennis M, Szabo G, Erdelyi F, Hayar A: Properties of external plexiform layer interneurons in mouse olfactory bulb slices. Neuroscience. 2005, 133: 819-829. 10.1016/j.neuroscience.2005.03.008.PubMed CentralView ArticlePubMedGoogle Scholar
- Tucker ES, Polleux F, Lamantia AS: Position and time specify the migration of a pioneering population of olfactory bulb interneurons. Dev Biol. 2006, 297: 387-401. 10.1016/j.ydbio.2006.05.009.View ArticlePubMedGoogle Scholar
- Lodovichi C, Belluscio L, Katz LC: Functional topography of connections linking mirror-symmetric maps in the mouse olfactory bulb. Neuron. 2003, 38: 265-276. 10.1016/S0896-6273(03)00194-6.View ArticlePubMedGoogle Scholar
- Royal SJ, Key B: Development of P2 olfactory glomeruli in P2-internal ribosome entry site-tau-LacZ transgenic mice. J Neurosci. 1999, 19: 9856-9864.PubMedGoogle Scholar
- Bailey MS, Puche AC, Shipley MT: Development of the olfactory bulb: evidence for glia-neuron interactions in glomerular formation. J Comp Neurol. 1999, 415: 423-448. 10.1002/(SICI)1096-9861(19991227)415:4<423::AID-CNE2>3.0.CO;2-G.View ArticlePubMedGoogle Scholar
- Verhaagen J, Zhang Y, Hamers FP, Gispen WH: Elevated expression of B-50 (GAP-43)-mRNA in a subpopulation of olfactory bulb mitral cells following axotomy. J Neurosci Res. 1993, 35: 162-169. 10.1002/jnr.490350206.View ArticlePubMedGoogle Scholar
- Taira E, Tsukamoto Y, Kohama K, Maeda M, Kiyama H, Miki N: Expression and involvement of gicerin, a cell adhesion molecule, in the development of chick optic tectum. J Neurochem. 2004, 88: 891-899.View ArticlePubMedGoogle Scholar
- Chen CC, Chen N, Lau LF: The angiogenic factors Cyr61 and connective tissue growth factor induce adhesive signaling in primary human skin fibroblasts. J Biol Chem. 2001, 276: 10443-10452. 10.1074/jbc.M008087200.View ArticlePubMedGoogle Scholar
- Perbal B: CCN proteins: multifunctional signalling regulators. Lancet. 2004, 363: 62-64. 10.1016/S0140-6736(03)15172-0.View ArticlePubMedGoogle Scholar
- Wada S, Hamada M, Satoh N: A genomewide analysis of genes for the heat shock protein 70 chaperone system in the ascidian Ciona intestinalis. Cell Stress Chaperones. 2006, 11: 23-33. 10.1379/CSC-137R.1.PubMed CentralView ArticlePubMedGoogle Scholar
- Yagi T: Diversity of the cadherin-related neuronal receptor/protocadherin family and possible DNA rearrangement in the brain. Genes Cells. 2003, 8: 1-8. 10.1046/j.1365-2443.2003.00614.x.View ArticlePubMedGoogle Scholar
- Redies C, Vanhalst K, Roy F: delta-Protocadherins: unique structures and functions. Cell Mol Life Sci. 2005, 62: 2840-2852. 10.1007/s00018-005-5320-z.View ArticlePubMedGoogle Scholar
- Buchler P, Gazdhar A, Schubert M, Giese N, Reber HA, Hines OJ, Giese T, Ceyhan GO, Muller M, Buchler MW, Friess H: The Notch signaling pathway is related to neurovascular progression of pancreatic cancer. Ann Surg. 2005, 242: 791-800, discussion 800-1. 10.1097/01.sla.0000189115.94847.f1.PubMed CentralView ArticlePubMedGoogle Scholar
- Shutter JR, Scully S, Fan W, Richards WG, Kitajewski J, Deblandre GA, Kintner CR, Stark KL: Dll4, a novel Notch ligand expressed in arterial endothelium. Genes Dev. 2000, 14: 1313-1318.PubMed CentralPubMedGoogle Scholar
- Mineno J, Okamoto S, Ando T, Sato M, Chono H, Izu H, Takayama M, Asada K, Mirochnitchenko O, Inouye M, Kato I: The expression profile of microRNAs in mouse embryos. Nucleic Acids Res. 2006, 34: 1765-1771. 10.1093/nar/gkl096.PubMed CentralView ArticlePubMedGoogle Scholar
- Obernosterer G, Leuschner PJ, Alenius M, Martinez J: Post-transcriptional regulation of microRNA expression. Rna. 2006, 12: 1161-1167. 10.1261/rna.2322506.PubMed CentralView ArticlePubMedGoogle Scholar
- Marillat V, Cases O, Nguyen-Ba-Charvet KT, Tessier-Lavigne M, Sotelo C, Chedotal A: Spatiotemporal expression patterns of slit and robo genes in the rat brain. J Comp Neurol. 2002, 442: 130-155. 10.1002/cne.10068.View ArticlePubMedGoogle Scholar
- Lin DM, Yang YH, Scolnick JA, Brunet LJ, Marsh H, Peng V, Okazaki Y, Hayashizaki Y, Speed TP, Ngai J: Spatial patterns of gene expression in the olfactory bulb. Proc Natl Acad Sci U S A. 2004, 101: 12718-12723. 10.1073/pnas.0404872101.PubMed CentralView ArticlePubMedGoogle Scholar
- Treloar HB, Gabeau D, Yoshihara Y, Mori K, Greer CA: Inverse expression of olfactory cell adhesion molecule in a subset of olfactory axons and a subset of mitral/tufted cells in the developing rat main olfactory bulb. J Comp Neurol. 2003, 458: 389-403. 10.1002/cne.10590.View ArticlePubMedGoogle Scholar
- Key B, St John J: Axon navigation in the mammalian primary olfactory pathway: where to next?. Chem Senses. 2002, 27: 245-260. 10.1093/chemse/27.3.245.View ArticlePubMedGoogle Scholar
- Valverde F, Santacana M, Heredia M: Formation of an olfactory glomerulus: morphological aspects of development and organization. Neuroscience. 1992, 49: 255-275. 10.1016/0306-4522(92)90094-I.View ArticlePubMedGoogle Scholar
- Schwarting GA, Kostek C, Ahmad N, Dibble C, Pays L, Puschel AW: Semaphorin 3A is required for guidance of olfactory axons in mice. J Neurosci. 2000, 20: 7691-7697.PubMedGoogle Scholar
- Blanchart A, De Carlos JA, Lopez-Mascaraque L: Time frame of mitral cell development in the mice olfactory bulb. J Comp Neurol. 2006, 496: 529-543. 10.1002/cne.20941.View ArticlePubMedGoogle Scholar
- Zheng C, Feinstein P, Bozza T, Rodriguez I, Mombaerts P: Peripheral olfactory projections are differentially affected in mice deficient in a cyclic nucleotide-gated channel subunit. Neuron. 2000, 26: 81-91. 10.1016/S0896-6273(00)81140-X.View ArticlePubMedGoogle Scholar
- Conzelmann S, Malun D, Breer H, Strotmann J: Brain targeting and glomerulus formation of two olfactory neuron populations expressing related receptor types. Eur J Neurosci. 2001, 14: 1623-1632. 10.1046/j.0953-816x.2001.01788.x.View ArticlePubMedGoogle Scholar
- Faedo A, Ficara F, Ghiani M, Aiuti A, Rubenstein JL, Bulfone A: Developmental expression of the T-box transcription factor T-bet/Tbx21 during mouse embryogenesis. Mech Dev. 2002, 116: 157-160. 10.1016/S0925-4773(02)00114-4.View ArticlePubMedGoogle Scholar
- Baker H, Kawano T, Margolis FL, Joh TH: Transneuronal regulation of tyrosine hydroxylase expression in olfactory bulb of mouse and rat. J Neurosci. 1983, 3: 69-78.PubMedGoogle Scholar
- Baker H, Kawano T, Albert V, Joh TH, Reis DJ, Margolis FL: Olfactory bulb dopamine neurons survive deafferentation-induced loss of tyrosine hydroxylase. Neuroscience. 1984, 11: 605-615. 10.1016/0306-4522(84)90047-2.View ArticlePubMedGoogle Scholar
- Hamilton KA, Coppola DM: Distribution of GluR1 is altered in the olfactory bulb following neonatal naris occlusion. J Neurobiol. 2003, 54: 326-336. 10.1002/neu.10182.View ArticlePubMedGoogle Scholar
- Sano K, Tanihara H, Heimark RL, Obata S, Davidson M, St John T, Taketani S, Suzuki S: Protocadherins: a large family of cadherin-related molecules in central nervous system. Embo J. 1993, 12: 2249-2256.PubMed CentralPubMedGoogle Scholar
- Yoshida K: Fibroblast cell shape and adhesion in vitro is altered by overexpression of the 7a and 7b isoforms of protocadherin 7, but not the 7c isoform. Cell Mol Biol Lett. 2003, 8: 735-741.PubMedGoogle Scholar
- Giger RJ, Pasterkamp RJ, Heijnen S, Holtmaat AJ, Verhaagen J: Anatomical distribution of the chemorepellent semaphorin III/collapsin-1 in the adult rat and human brain: predominant expression in structures of the olfactory-hippocampal pathway and the motor system. J Neurosci Res. 1998, 52: 27-42. 10.1002/(SICI)1097-4547(19980401)52:1<27::AID-JNR4>3.0.CO;2-M.View ArticlePubMedGoogle Scholar
- Wang F, Nemes A, Mendelsohn M, Axel R: Odorant receptors govern the formation of a precise topographic map. Cell. 1998, 93: 47-60. 10.1016/S0092-8674(00)81145-9.View ArticlePubMedGoogle Scholar
- Nagao H, Yoshihara Y, Mitsui S, Fujisawa H, Mori K: Two mirror-image sensory maps with domain organization in the mouse main olfactory bulb. Neuroreport. 2000, 11: 3023-3027.View ArticlePubMedGoogle Scholar
- Pasterkamp RJ, De Winter F, Holtmaat AJ, Verhaagen J: Evidence for a role of the chemorepellent semaphorin III and its receptor neuropilin-1 in the regeneration of primary olfactory axons. J Neurosci. 1998, 18: 9962-9976.PubMedGoogle Scholar
- Hirano S, Suzuki ST, Redies C: The cadherin superfamily in neural development: diversity, function and interaction with other molecules. Front Biosci. 2003, 8: d306-55. 10.2741/972.View ArticlePubMedGoogle Scholar
- Lin DM, Loveall B, Ewer J, Deitcher DL, Sucher NJ: Characterization of mRNA expression in single neurons. Methods in molecular biology. in press:
- Hughes TR, Mao M, Jones AR, Burchard J, Marton MJ, Shannon KW, Lefkowitz SM, Ziman M, Schelter JM, Meyer MR, Kobayashi S, Davis C, Dai H, He YD, Stephaniants SB, Cavet G, Walker WL, West A, Coffey E, Shoemaker DD, Stoughton R, Blanchard AP, Friend SH, Linsley PS: Expression profiling using microarrays fabricated by an ink-jet oligonucleotide synthesizer. Nat Biotechnol. 2001, 19: 342-347. 10.1038/86730.View ArticlePubMedGoogle Scholar
- Kargul GJ, Dudekula DB, Qian Y, Lim MK, Jaradat SA, Tanaka TS, Carter MG, Ko MS: Verification and initial annotation of the NIA mouse 15K cDNA clone set. Nat Genet. 2001, 28: 17-18. 10.1038/88206.PubMedGoogle Scholar
- VanBuren V, Piao Y, Dudekula DB, Qian Y, Carter MG, Martin PR, Stagg CA, Bassey UC, Aiba K, Hamatani T, Kargul GJ, Luo AG, Kelso J, Hide W, Ko MS: Assembly, verification, and initial annotation of the NIA mouse 7.4K cDNA clone set. Genome Res. 2002, 12: 1999-2003. 10.1101/gr.633802.PubMed CentralView ArticlePubMedGoogle Scholar
- Yang YH, Buckley MJ, Dudoit S, Speed TP: Comparison of methods for image analysis on cDNA microarray data. Journal of Computational and Graphical Statistics. 2002, 11: 108-136. 10.1198/106186002317375640.View ArticleGoogle Scholar
- Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, Speed TP: Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res. 2002, 30: e15-10.1093/nar/30.4.e15.PubMed CentralView ArticlePubMedGoogle Scholar
- Smyth GK: Linear models and empirical Bayes methods for assessing differential expression in microarray experiments. Statistical Applications in Genetics and Molecular Biology. 2004, 3: Article 3-10.2202/1544-6115.1027.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.