Transcriptional control of axonal guidance and sorting in dorsal interneurons by the Lim-HD proteins Lhx9 and Lhx1
- Oshri Avraham†1,
- Yoav Hadas†1,
- Lilach Vald1,
- Sophie Zisman1,
- Adi Schejter1,
- Axel Visel2 and
- Avihu Klar1Email author
© Avraham et al.; licensee BioMed Central Ltd. 2009
Received: 13 January 2009
Accepted: 19 June 2009
Published: 19 June 2009
Lim-HD proteins control crucial aspects of neuronal differentiation, including subtype identity and axonal guidance. The Lim-HD proteins Lhx2/9 and Lhx1/5 are expressed in the dorsal spinal interneuron populations dI1 and dI2, respectively. While they are not required for cell fate acquisition, their role in patterning the axonal trajectory of dI1 and dI2 neurons remains incompletely understood.
Using newly identified dI1- and dI2-specific enhancers to trace axonal trajectories originating from these interneurons, we found that each population is subdivided into several distinct groups according to their axonal pathways. dI1 neurons project axons rostrally, either ipsi- or contra-laterally, while dI2 are mostly commissural neurons that project their axons rostrally and caudally. The longitudinal axonal tracks of each neuronal population self-fasciculate to form dI1- and dI2-specific bundles. The dI1 bundles are spatially located ventral relative to dI2 bundles. To examine the functional contribution of Lim-HD proteins to establishment of dI axonal projections, the Lim-HD code of dI neurons was altered by cell-specific ectopic expression. Expression of Lhx1 in dI1 neurons caused a repression of Lhx2/9 and imposed caudal projection to the caudal commissural dI1 neurons. Complementarily, when expressed in dI2 neurons, Lhx9 repressed Lhx1/5 and triggered a bias toward rostral projection in otherwise caudally projecting dI2 neurons, and ventral shift of the longitudinal axonal fascicule.
The Lim-HD proteins Lhx9 and Lhx1 serve as a binary switch in controlling the rostral versus caudal longitudinal turning of the caudal commissural axons. Lhx1 determines caudal turning and Lhx9 triggers rostral turning.
The diverse functions of the vertebrate nervous system depend on synaptic connections between specific classes of neurons and their targets. Neurons differ from each other by their type of afferent input, cell body positioning along the body axis, axonal trajectory and axonal target. The projection of axons to their targets occurs in a stepwise manner, under the control of guidance cues arrayed at discrete locations along the pathway of axonal growth. A specific axonal pathway of a neuron, governed by a transcriptional code, is manifested by the expression of receptors for guidance molecules that interpret the guidance cues en route and at their putative target [1, 2].
In vertebrates, the coordinated development of neurons and their targets has been well documented in the context of the peripheral projections of spinal motor neurons. Motor neurons innervate many different muscle targets, and the location of motor neurons within the spinal cord is linked to target position. Lim-HD proteins control aspects of neuronal differentiation, such as subtype identity and axonal guidance (reviewed in ). The broad repertoire of specification by Lim-HD factors is exemplified in the development of motor neurons [4–9]. While the early expression of Isl1 is required for the differentiation of all the motor neurons , later in development, Isl1 confers LMCm subtype identity to motor neurons and directs LMCm axons to the ventral limb. In a contrasting and complementary manner, Lhx1 confers LMCl subtype identity and directs LMCl axons to the dorsal limb [4, 5].
The uncertainty about the role of Lim-HD proteins in the control of motor axon pathfinding stems from the fact that many genes of this class control earlier developmental decisions – the regulation of neural pattern, cell specification, and cell survival . A replacement of the Lim-HD code of LMC neurons, via ectopic expression of Isl1 or Lhx1, causes a binary switch in cell fate, where ectopic Isl1-expressing motor neurons adopt LMCm subtype identity, and ectopic Lhx1-expressing motor neurons become LMCl neurons . Similarly, the LIM homeobox genes Lhx3 (Lim3) and Lhx4 (Gsh4) are transiently expressed by spinal motor neurons but appear to specify neuronal subtype identity and migratory behaviour, indirectly influencing the position at which motor axons emerge from the spinal cord . Nevertheless, studies in Drosophila have shown that Lim-HD proteins direct motor axon projections without influencing neuronal fate [11, 12], suggesting that some of their vertebrate counterparts may have similar roles.
Spinal sensory neurons are derived from several populations of dorsal interneurons (dI1-6) in the embryonic dorsal spinal cord that are distinguished by a transcriptional code and differentiated cell body positions. dI1-3 neurons differentiate from distinct groups of ventricular zone progenitor cells that express the basic helix loop helix (bHLH) transcription factors Atoh1, Ngn1/2 or Mash1, respectively. As the dI1-3 neurons differentiate, Lim-HD transcription factors are expressed: Lhx2 and Lhx9 in dI1, Lhx1 and Lhx5 in dI2, and Isl1 in dI3 [13, 14]. Gene targeting and transgenesis in mice have revealed that dI1 neurons project their axons ipsi- and contra-laterally toward the brain [15, 16], and dI2 neurons project their axons contra-laterally . However, the precise en route axonal pathway, as well the topographic organization of dI axons within the neural tube, is not known.
In this study we used genetic assays in chick embryos to address the basis of the selection of interneuron axonal trajectory within the developing neural tube. Initially, taking advantage of novel enhancer elements, we mapped the axonal trajectories of dI1 and dI2 neurons. Each dI has a unique pattern of axonal projections. dI1 neurons project their axons rostrally along two pathways: either ipsi- or contra-laterally. dI2 are mostly commissural neurons. After crossing the floor plate the rostral dI2 axons turn rostrally, while the caudal dI2 axons turn caudally. To begin to understand the possible role of Lim-HD in patterning the axonal trajectories of spinal interneurons, the Lim-HD code of dI1 and dI2 neurons was altered by cell type-specific ectopic expression. We found that Lhx1, ectopically expressed in dI1 neurons, confers caudal projection to the otherwise rostrally projecting commissural dI1 axons; while Lhx9, expressed in dI2 neurons, causes a rostral bias to the caudally projecting dI2 axons. Thus, Lim-HD proteins control the longitudinal axonal choice of dI1 and dI2 neurons.
Employment of enhancer elements to drive expression of reporter genes in neurons is a widely used paradigm for tracking axonal projection. For tracking axonal projection of spinal interneurons in vertebrates, germ line-targeted reporter genes yield bilaterally symmetric labelling [15, 17, 18]. Therefore, it is hard to distinguish between the ipsi- and contra-laterally projecting axons. Unilateral electroporation into the chick neural tube provides a useful means to restrict expression of a reporter gene to one side of the central nervous system, and to follow axonal projection on both sides [19, 20]. Mouse enhancer elements are appropriately active in the chick neural tube. Thus, Atoh1, HB9, and HoxA1 enhancer elements drive expression in dI1, motor neurons and floor plate cells, respectively [20–23].
To further characterize the cell type specificity of the enhancers, the embryos were co-electroporated along with a Cre-dependent nuclear GFP (nGFP). To determine the identity of reporter-expressing cells, embryos were analyzed at stage 23 to 24 by co-staining with dI-specific antibodies to Lhx2/9 (dI1), Lhx1/5 (dI2, dI4, dI6, V0, V1), Isl1 (dI3), Pax2 (dI4, dI6, V0, V1) and Engrailed1 (V1) (Figure 2I). nGFP expression under the control of #284 is restricted to dI1 neurons, as indicated by co-staining with Lhx2/9 Ab (Figure 1B, C) and the segregation from Lhx1- (Figure 1C) and Isl1-positive neurons (Figure 1B). Of the nGFP-positive neurons, 95.7% (n = 139) are Lhx2/9, and 4.3% (6 of 139) were nGFP-positive but negative to all the above interneuron markers. This minor population may represent progenitors of dI1 that have not upregulated the expression of Lhx2 and Lhx9 yet. The #284 enhancer is herein indicated as EdI1.
Of the neurons that express nGFP under the control of #169, 56% (n = 166) are dI2, as indicated by the co-localization to the dorsal Lhx1/5+ (Figure 2B) and Lhx1/5+/Pax2- cells (Figure 2D) and the segregation from Lhx2/9+ and Isl1+ neurons (Figure 2B, C); 35.5% are V1 neurons as indicated by the localization to the medial Pax+/En1+/Lhx1/5+ neurons (Figure 2D, E). Of the nGFP neurons, 8.5% are presumed to be progenitors of dI2 and V1, since no expression of any cell fate marker was scored. Its proximity to the dI2/V1-specific gene Foxd3 suggests that #169 is a dI2/V1-specific enhancer of Foxd3 (herein indicated as EdI2/V1).
For further focusing on the axonal projection pattern of dI2 neurons, the dI2-specific enhancer element (13G) of the Ngn1 gene was studied in the chick neural tube , utilizing the Cre-dependent nGFP system. Expression of nGFP was detected in dorsal/lateral interneurons that express either Lhx2/9 or Lhx1/5 (Additional file 1). Numerous dorsal interneurons located between the ventricular and marginal zones express nGFP. These neurons are presumed to be progenitors of dI1 and dI2 neurons. The expression of Ngn1 in progenitor neurons supports this assumption. The leakage in dI1 neurons while using the Ngn1-13G enhancer , versus the entire Ngn1 enhancer , suggests that cis elements that are required for repression of expression in dI1 neurons are absent in the 13G enhancer. Hence, in the chick, the Ngn1 13G enhancer is a dI1/2-specific enhancer (herein indicated as Ed1/2).
Axonal projection of dI1 axons
dI1 neurons give rise to two subpopulations that differ in their cell position, axonal projection and transcription of the Lim-HD proteins: the dI1comm population, located at the dorsal neural tube and more ventral/medially, which projects axons toward and across the floor plate; and the dI1ipsi population, located in a ventral/lateral position, which projects axons ipsi-laterally. The division into two subpopulations is also evident in the transcription of the Lim-HD proteins Lhx2 and Lhx9. dI1comm neurons express Lhx2high and Lhx9low, while dI1ipsi neurons express Lhx9 [16, 28].
Summary of the axonal phenotypes.
dI1 axonal patterning
dI2 commissural axonal patterning
C to R
dI1+dI2 in LF
Number of embryos
dI2 only (intersection)
The positions of the longitudinal dI1ipsi and dI1comm fascicules at the LF along the dorsal/ventral axis appear similar. Hence, it is conceivable that dI1comm axons from one side of the neural tube, and dI1ipsi axons from the other side, fasciculate together. To test this hypothesis, GFP or taumyc were expressed in the two halves of the neural tube, respectively (see Materials and methods). Projection of dI1comm GFP-positive axons toward the dI1ipsi taumyc axons was inspected (Figure 3C, D, E). dI1comm axons turned diagonally toward the dI1ipsi bundle. As they contacted the dI1ipsi bundle, dI1comm axons fasciculated with dI1ipsi and turned rostrally (Figure 3D). Thus, homophilic interaction between dI1comm and dI1ipsi may facilitate axonal turning of dI1comm at the LF.
Axonal projection of dI2 neurons
The Ngn1 enhancer was utilized previously for labelling the axons of dI2 neurons in transgenic mice. Cross-sections and open-book preparation demonstrated that dI2 neurons project their axons toward and across the floor plate [17, 20, 27]. However, the bi-symmetrical expression of the reporter gene precluded detailed mapping of dI2 axonal trajectories.
The axonal cues of dI2 axons were studied utilizing three paradigms: the EdI2/V1 enhancer – V1 neurons project their axons only ipsi-laterally [29, 30] and, thus, the EdI2/V1 enhancer can be used for studying the contra-lateral projection pattern of dI2 neurons (six embryos; Table 1); the Ed1/2 enhancer – divergence from the dI1 axonal pattern, when employing the dI1/2 enhancer, can be attributed to dI2 neurons (four embryos; Table 1); and molecular intersection of the EdI2/V1 and EdI1/2 enhancers – we have designed a method that enables labelling of neurons that co-express the above enhancers (four embryos; Table 1).
The co-expression in dI2 plus V1 neurons or dI2 plus dI1 utilizing the EdI2/V1 and EdI1/2 enhancers, respectively, precludes the identification of ipsi-laterally projecting axons. For labelling dI2 neurons solely, an enhancer intersection technique was adopted. The EdI2/V1 and EdI1/2 enhancers are not exclusive to dI2 neurons; however, their intersection occurs in dI2 neurons. In order to label neurons that co-express EdI2/V1 and EdI1/2 enhancers, we combined the Cre/LoxP and the Gal4/UAS systems. Cre was expressed under the EdI1/2 enhancer, and Gal4 under the EdI2/V1 enhancer. The reporter plasmid contains GFP under a dual control of Gal4 and Cre (UAS-LoxP-STOP-LoxP-GFP; for more details, see Materials and methods). GFP-expressing neurons in which the intersection of EdI1/2 and EdI2/V1 enhancers occurs are Lhx1/5+/Pax2- (100%, n = 22; Figure 2F), Pax2-/En1- (100%, n = 18; Figure 2G) and Lhx2/9-/Isl1- (100%, n = 29; Figure 2H). Hence, they are dI2 neurons.
At the ipsi-lateral side, few longitudinally projecting axons are seen (Figure 5A, yellow arrow in A5; Additional file 2A). The majority of the axons project circumferentially toward the floor plate (Figure 5A, A6; Additional file 3). No longitudinal tracks are observed at either the VF or the LF. To estimate the ratio between the ipsi- and contra-lateral axonal choice of dI2 neurons, the extent of ipsi-/contra-lateral axons at the cervical level was scored. At this level, no cell bodies were labeled on the ipsi-lateral side. Only in one neural tube (n = 4) were longitudinally projecting axons visible at the cervical ipsi-lateral side (Additional file 3). The ratio between ipsi- to contra-lateral axons is 8.6% in that neural tube. Hence, dI2 neurons have mainly commissural axons that elongate longitudinally either rostrally or caudally, depending on their position along the longitudinal axis.
The expression of Lhx2 following Lhx9 ectopic expression was used to study dI1 cell fate acquisition. Neurons expressing Lhx9 down-regulate the expression of Lhx2 (Figure 7C, D), suggesting that the segregation to dI1ipsi neurons, expressing Lhx9, and dI1comm neurons, expressing Lhx2high/Lhx9low, is mediated by cross-repression between Lhx9 and Lhx2. This hypothesis is supported by the reciprocal downregulation of Lhx9 via ectopic expression of Lhx2 (YH and AS, data not shown). Atoh1 is expressed in dI1 progenitor neurons, and its expression is down-regulated in differentiated post-mitotic dI1 neurons. Ectopic expression of Lhx9 causes down-regulation of Atoh1 (Figure 7E, F), suggesting that Lhx9, when up-regulated in post-mitotic dI1 neurons, is a repressor of Atoh1. Barhl1/2 genes are expressed in rodent dI1 neurons . However, the orthologous genes are not present in the chick genome. Due to the lack of post-mitotic dI1 markers, the possible repression of dI2 cell fate was studied following Lhx9 ectopic expression. Foxd3 is not repressed in Lhx9-expressing neurons (Figure 7G, H). Lhx9 is thus insufficient for repressing dI2 cell fate. Thus, Lhx9 and Lhx1 cross-repress each other without changing the complete range of cell fate identity. Lhx9 and Lhx1 may nevertheless control some features of differentiated dI neurons. The role of Lhx9 and Lhx1 in axon guidance was studied in the following experiments.
Changing the Lim-HD code of dI1 and dI2 neurons by ectopic expression – general considerations
The repression of endogenous Lim-HD following ectopic expression of a reciprocal Lim-HD gene results in replacement of the Lim-HD code. To study the role of the Lim-HD code in the assignment of the axonal projection pattern of dI1 and dI2 neurons, their Lim-HD code was alternated. The following considerations were taken into account in the subsequent ectopic expression experiments. First, to study cell autonomous effects, Lim-HD proteins were expressed specifically in the reciprocal dI neurons utilizing EdI enhancers (Lhx9 in dI2, and Lhx1 in dI1). Second, to follow the axonal trajectories of the manipulated neurons, taumyc or GFP were co-expressed with the ectopic Lim-HD protein from the same plasmid. Third, ectopic expression of Lim-HD may lead to a change in cell properties and, subsequently, to its own down-regulation. For example, Lhx9 expressed in dI2 utilizing the EdI2/V1 enhancer may up-regulate certain dI1 characteristics, ultimately leading to down-regulation of the EdI2/V1 enhancer. The stable Cre/Lox systems were used to stabilize the ectopic expression. Fourth, ectopic expression may result in high, non-physiological levels of exogenous protein levels. The levels of ectopic Lhx9 were compared to the endogenous levels of Lhx2 and Lhx9 (in the non-electroporated side of the neural tube). Utilizing the Cre/Lox system, the exogenous and endogenous levels of Lhx9 were similar (Additional file 4).
Lhx1 controls caudal turning
The commissural dI1 and dI2 axonal patterning differ in two aspects (Figures 3, 4 and 5): in the lumbosacral neural tube all dI2caud axons project caudally, while only the caudal sacral dI1ipsi axons project caudally; and at the caudal third of the thoracic level, dI2 axons turn either rostrally or caudally, forming a 'crisscross' axonal pattern at the contra-lateral side, while dI1comm axons turn only rostrally. The consequence of ectopic Lhx1 expression in dI1 neurons (dI1Lhx1) was studied, focusing on the above features (Figure 8). At the lumbosacral level dI1Lhx1 axons turn caudally (Figure 8A, A1). At the caudal thoracic level a crisscross pattern of axons turning either rostrally or caudally is evident at the contra-lateral side of the neural tube (Figure 8A, A2). Hence, all the dI2caud axonal features are assumed by the commissural dI1Lhx1 neurons (Figure 8B).
Longitudinal axonal tracks of dI1Lhx1 are present at the ipsi-lateral side (Figure 8A). Hence, Lhx1 does not suppress the ipsi-lateral projection of dI1ipsi neurons. However, a dI1Lhx1 fascicule is observed at the ipsi-VF and the ipsi-LF, while dI1ipsi axons form only an ipsi-LF bundle. The ipsi-VF is a characteristic of V1 axons  (Figure 4A), which also express Lhx1. Thus, Lhx1 is sufficient to impose dI2-like and V1-like axonal trajectories to dI1comm and dI1ipsi neurons, respectively.
Lhx9 controls rostral turning
Lhx9 controls the dorsoventral position at which axons turn into the longitudinal plane
The relative position at the LF of dI2Lhx9 axons was compared to dI1 axons. Taumyc was expressed under the control of the EdI1 enhancer (EdI1::Cre + pCAGG-LoxP-STOP-LoxP-taumyc), together with ectopic Lhx9 in dI2 neurons (EdI2/V1::Gal4 + UAS::Lhx9_UAS::GFP). dI2Lhx9 axons turn rostrally at the LF, together with dI1comm axons (Figure 10B, C). The specific bundle of dI2 axons at the LF was not formed, and dI2 and dI1comm axons intermingled and formed one fascicle as they turned longitudinally at the LF (Figure 10E, Table 1). Thus, Lhx9 may control the homophilic interaction with dI1 axons along their longitudinal projection toward the brain. Alternatively, Lim-HD code may control the position of the LF along the dorsal/ventral axis, where Lhx9 directs a more ventral position than Lhx1, and the mis-expression of Lhx9 in dI2 axons shifts the dorsoventral position at which dI2 axons turn into the longitudinal plane.
Cell fate acquisition is manifested by the activation of transcription factors. Many features define the function of a neuron, including cell body positioning, dendritic tree morphology, axonal projection, neurotransmitters specificity and excitatory or inhibitory output. The specification of neurons might be governed by linear sequential activation of transcription factors, or by activation of a parallel pathway, each one driving a specific neuronal characteristic. In the current study we have combined molecular and morphological tools to follow the development and axonal patterning of molecularly defined groups of dorsal spinal interneurons. We provide evidence that the Lim-HD proteins Lhx1 and Lhx9 are sufficient to influence axonal patterning without affecting neuronal fate.
Diversity of dI axonal projections
The combination of specific enhancers, augmentation of expression levels utilizing the Cre/LoxP and the Gal4/UAS systems, and chick electroporation provide quick and efficient tools for deciphering axonal pathways of a genetically defined group of neurons. The emerging picture is of a complex divergence of axonal cues that arises from dI1 and dI2 subpopulations. dI1 and dI2 give rise to two subpopulations each, defined by the direction of their axonal projections. The simultaneous molecular and spatially restricted labelling of two neuronal populations, dI1 + dI2 and dI1ipsi+ dI1comm, underscores the axonal architecture of dI1 and dI2 axons: dI1ipsi and dI1comm fasciculate together at the LF; dI1 and dI2 axonal tracks at the LF are segregated.
The axonal pathways of spinal internerons were mapped previously utilizing diI injection [32–34]. Kadison and Kaprielian described four main axonal projections of decussating axons: intermediate longitudinal commissural (ILC), medial longitudinal commissural (MLC), bifurcating longitudinal commissural (BLC), and forked transverse commissural (FTC) . ILC axons travelled rostrally in an arcuate manner, extending into VF regions of the spinal cord before executing a second turn into the longitudinal plane at the LF. The contra-laterally projecting dI1 and dI2 neurons project their axons in an ILC pattern. MLC axons, which extend along the floor plate boundary at the VF for distances greater than 100 mm, BLC axons, which bifurcate to rostral and caudal projections, and FTC axons, which form a trident-shaped or forked projection, were not identified in the current study. The labelling of multiple neurons achieved using the electroporation paradigm in the current study may obscure these projection patterns. A moderate number (10%) of decussated axons was observed to extend in the caudal direction following DiI injection . However, our studies point to a larger quantity of caudally projecting neurons. At the sacral level dI1ipsi and all dI2caud axons extend caudally. At the lumbar region and the caudal third of thoracic levels, about half of dI2 axons project caudally. Hence, a rostral to caudal stepwise increase in caudal projection is evident. Injection of DiI into neurons at the sacral level may reveal more caudally projecting axons.
Transcriptional control of axonal guidance
The divergence in axonal growth along the ipsi/contra and the caudal/rostral axes may stem from a cell type-specific expression of transcription factors. Namely, dI1comm and dI2comm genes could theoretically be expressed in the contra-laterally projecting dI1comm and dI2 neurons, respectively. A possible candidate for such a mechanism is Lhx2, which is expressed only in dI1comm neurons. However, gene-targeting experiments of Lhx2 and Lhx9 have shown that a Lim-HD code does not control ipsi- versus contra-lateral axonal projection . Similarly, Lhx1 is expressed in the ipsi-only population V1 and the contra-mostly population dI2. Therefore, Lhx1 is probably not implicated in controlling of the contra-lateral projection of dI2 neurons. Alternatively, common dIcomm and dIipsi genes might be expressed in all the dINcomm and dINipsi neurons, respectively. Transcription factors such as Unc4 and NSCL1, which are expressed in all interneurons [35, 36] in an overlapping pattern to the commissural-only genes TAG1 and Robo3, are candidates for controlling commissural guidance choice of dIs.
A similar transcriptional mechanism may account for the caudal versus rostral axonal choice. A transcriptional code may discriminate between the longitudinal levels. Hence, the combination of dIcaudal, expressed at the caudal neural tube, and Lhx1 may confer caudal projection. Potential dIcaudal and dIrostral factors may be the Hox proteins. A Hox code determines the rostral/caudal identity of motor neurons, and the combination of Hox and Lim-HD codes determines the subclassification of motor neuron pools [37, 38]. The caudally expressing Hox10 and Hox11 genes may confer caudal turning to the lumbosacral dI2 neurons, while the more rostral Hox, Hox8 and beyond, may confer rostral turning. The caudal thoracic crisscross pattern might be controlled by Hox9+/Hox8- code. The rostral to caudal change in axonal projection, following Lhx1 ectopic expression, may place Lhx1 as a positive activator of the Hoxcaudal genes Hox10 and Hox11. Complementarily, Lhx9 may suppress caudal turning by suppressing Lhx1 and/or Hox10 and Hox11 expression. Alternatively, Lhx9 may suppress Hox10 and Hox11 expression, and Lhx1 may reveal Hox10 and Hox11 activity by suppressing their suppressor, Lhx9.
What are the axonal cues that may govern caudal turning? Axons may turn in different directions due to different axonal cues or differential responsiveness to common cues. The differential cues theory is not supported by our data. At the caudal thoracic levels dI2 axons, at the same rostro/caudal level, turn either rostrally or caudally. The conversion in axonal directionality may be governed cell autonomously by receptors or signalling molecules that convert attraction to repulsion. The rostral turning of commissural neurons along the floor plate is mediated by increasing caudal-to-rostral levels of Wnt proteins , which attract axons; and decreasing caudal-to-rostral levels of Shh , which repel axons. Caudally turning neurons may express receptors or signalling molecules that convert Wnt attraction to repulsion and/or Shh repulsion to attraction. In vitro assays with caudal dI2 neurons challenged with Wnts and Shh should clarify whether a cell autonomous change in responsiveness governs dorsal and caudal turning.
Role of Lim-HD in cell fate determination
The emergence of interneuron divisions is marked by a mutual exclusion in the expression profile of bHLH proteins and Lim-HD proteins. Progenitor dI1/2 neurons express Atoh1 and Ngn1/2, respectively . Loss and gain of function experiments have demonstrated that these proteins cross repress each other, and are both required and sufficient for the differentiation of dI1/2 neurons [18, 41]. Therefore, in the absence of Atoh1, dI1 neurons fail to differentiate, and are converted to dI2 neurons . Lim-HD genes, expressed in the post-mitotic dI1/2 neurons, are probably activated by the bHLH proteins. Our ectopic expression experiments demonstrate that Lim-HD proteins also cross-repress each other in dI1 and dI2 neurons. Thus, the distinct identity of adjacent neurons is guaranteed at the mitotic and post-mitotic stages by cross-repression of bHLH and Lim-HD proteins, respectively. Loss of function experiments have demonstrated that in the absence of Lhx2/9 or Lhx1/5, the fate of dI1 and dI2 neurons is not altered [16, 42]. In the Lhx2/9 double knockout mouse, dI1 cells express dI1-specific genes, and the Lim-HD code is not changed to Lhx1/5. It is conceivable that Atoh1, which acts upstream to Lhx2/9, is repressing Ngn1/2 and thus indirectly prevents the activation of Lhx1/5. It is also possible that bHLH proteins control dI1/2 cell fate by activating Lim-HD proteins and, in addition, in a feed forward mechanism, directly control cell fate. Thus, the elimination of Lim-HD can be compensated for by bHLH proteins. Ectopic Lim-HD proteins may play a dominant role in repressing other Lim-HD proteins and in repressing bHLH protein activity. This assumption is supported by the observation that ectopically expressed Lhx1 suppresses the expression of Atoh1 (YH and OA, unpublished results). The Lim-HD proteins Isl1 and Lhx1 play a similar role in determining the fate of LMC neurons. Retinoic acid induces LMCl neurons by activating Lhx1 and repressing Isl1 expression. In the absence of Lhx1, LMCl neurons differentiate, settle at the lateral LMC column and do not upregulate Isl1 expression. Thus, like bHLH proteins in dI1/2 neurons, retinoic acid is sufficient to confer LMCl identity, probably by bypassing Lhx1 signalling in a feed-forward mechanism [4, 5].
The emergence of interneuron divisions is marked by mutual exclusion in the expression profile of bHLH and Lim-HD proteins . Loss of function experiments have demonstrated that in the absence of Lhx2/9 or Lhx1/5, the fate of dI1 and dI2 neurons is not altered [16, 42]. We have used targeted ectopic expression to explore the role of the Lim-HD proteins Lhx9 and Lhx1 in patterning the axonal trajectories of dI1 and dl2 neurons. Our results point to a new role of Lim-HD proteins in controlling the longitudinal turning choice and axonal sorting of dI1 and dI2 neurons.
Materials and methods
In ovo electroporations
Fertilized white Leghorn chicken eggs were incubated at 38.5 to 39°C. A DNA solution of 5 mg/ml was injected into the lumen of the neural tube at either HH stage 12 to 14 (cytomegalovirus (CMV) enhancer in pCAGG plasmid) or stage 17 to 18 (EdI1 and EdI2/V1 enhancers). For double-sided electroporation, a 1-h interval interceded between the first and second electroporations.
Electroporation was performed using three 50 ms pulses at 25V, applied across the embryo using a 0.5 mm Tungsten wire and a BTX electroporator (ECM 830). Embryos were incubated for 2 to 3 days prior to analysis.
Strategies for cell type specific expression
Testing enhancer specificity
Co-expression of a plasmid containing an enhancer driving the expression of Cre recombinase and a reporter plasmid in which a floxed mCherry gene was inserted between the CAGG enhancer/promoter module and the GFP gene (pCAGG-LoxP-mCherry-LoxP-GFP) was performed. Cells that do not express Cre (general expression) will express mCherry, while cells that express Cre under the control of the specific enhancer will express GFP. The CAGG enhancer is not restricted either spatially or temporally, while expression from the specific enhancer is initiated in post-mitotic cells. Thus, residual expression of mCherry is observed in the GFP-positive cells.
Testing cell type specificity of an enhancer
Co-expression of a plasmid containing an enhancer driving the expression of Cre recombinase and a reporter plasmid in which a transcriptional STOP module was inserted between the CAGG enhancer/promoter module and the nuclearGFP gene (pCAGG-LoxP-STOP-LoxP-nGFP) was performed. Embryos were electroporated at stage 16 since earlier electroporation may yield non-specific expression . Embryos were analyzed at stage 23 to 24.
Mapping axonal trajectories using the Cre/Lox system
Conditional GFP (pCAGG-LoxP-STOP-LoxP-GFP) or taumyc (pCAGG-LoxP-STOP-LoxP-taumyc) plasmids were electroprated along with enhancer::Cre plasmid. The entire spinal cord was excised at E6 and was prepared as an open-book for further analyses.
Mapping axonal trajectories using the Gal4/UAS system
The Gal4 DNA binding domain fused to an activation domain  was cloned downstream of the dI specific enhancers. The enhancer::Gal4 plasmid was co-electroporated with a UAS::GFP plasmid.
Enhancer intersection technique
The expression of the reporter gene GFP is dependent on both Gal4 and Cre. A floxed STOP cassette was inserted between UAS and GFP (UAS-LoxP-STOP-LoxP-GFP). Hence, removal of the STOP cassette by Cre recombinase, and activation of transcription by Gal4 are required for GFP expression. The intersection between two expression patterns is attained by electroporation of three plasmids: Enhancer1::Cre, Enhancer2::Gal4, and UAS-LoxP-STOP-LoxP-GFP.
Spinal cord open-book preparation
E6 electroporated chick spinal cord tissues were prepared as an open-book preparation by making a longitudinal incision along the roof plate with a sharp tungsten microneedle from the hindbrain down to the tail. The dorsal root ganglia (DRGs) were then separated from the spinal cord, leaving the floor plate intact. The hind and forelimb were marked with charcoal powder, and then the spinal cord was detached from the body and fixed in 4% paraformaldehyde in phosphate-buffered saline for 1 h at room temperature, after which the tissue was spread open to produce flat-mount preparations.
Embryos were fixed overnight at 4°C in 4% paraformaldehyde/0.1 M phosphate buffer, washed twice with phosphate-buffered saline, incubated in 30% sucrose/phosphate-buffered saline for 24 h, and embedded in Optimal Cutting Temperature solution (OCT). Cryostat sections (14 μm) were collected on Superfrost Plus slides and kept at -70°C. Antigen retrieval was used for paraffin sections. Sections were treated with boiled 10 mM citric acid, pH 6, for 10 minutes in the microwave. The following antibodies were used: rabbit polyclonal GFP antibody (Molecular Probes, Eugene, Oregon, USA)), Pax2 (Abcam, Cambridge, MA, USA)), myc (9E10), Isl1 (4D5), Lhx1/5 (4F2), Engrailed, Lhx2/9 (rabbit serum) (all provided by T Jessell, Columbia University, New York, NY, USA). Cy2, RRX and cy5 were used as fluorochromes. Images were taken under a microscope (Axioscope 2; Zeiss) with a digital camera (DP70; Olympus) or confocal microscope (FV1000; Olympus).
The EdI1 enhancer element was amplified by PCR from a genomic mouse DNA utilizing the primers [ATGAGCTCATCCCTTTTTGCTCCCTCAC] and [ATGCTAGCGGTGTTGTGGTTGACAGCAG]. EdI2/V1 was amplified utilizing the primers [ATGAGCTCGCTCTCTCTGCCTACCTCAGC] and [ATGCTAGCAACCTAGTGCCCTTGCACAC]. The enhancers were cloned into 5'Sac I/3'Nhe I sites of the appropriate Cre and Gal4 plasmids . The 13G Ngn1 enhancer (EdI1/2) was generated by PCR from a genomic mouse DNA utilizing the primers [ATTGCGGCCGCATCAGGCGCCGGATCACTTTG] and [GATCTAGACCTTCACCATCGTTAACACTGG] and cloned into 5'Not I/3'Xba I sites of the Cre plasmid. Chick Lhx2 and Lhx9 were obtained from Tom Jessell. Chick Lhx1 was obtained from Artur Kania. Chick Foxd3 and Atoh1 were obtained from the chick expressed sequence tag sequencing project .
In situ hybridization
Antisense digoxigenin-labeled probes were prepared by in vitro transcription (Roche, Nutley, NJ, USA)). In situ hybridization was performed on electroporated sections and was combined with immunohistochemistry. Prior to the in situ hybridization, sections were incubated with primary GFP antibody, and then a standard in situ protocol was applied, followed by a secondary fluorescent antibody treatment. Alternatively, adjacent sections were collected on different slides. One set of slides was used for in situ hybridization, and the other for immuno-detection of GFP.
basic helix loop helix
bifurcating longitudinal commissural
forked transverse commissural
green fluorescent protein
intermediate longitudinal commissural
medial longitudinal commissural
The authors thank Thomas Jessell for Isl1, the Lhx1/5 monoclonal antibody, and Lhx2/9 and Lhx1/5 rabbit antibodies and the Lhx2 and Lhx9 chick genes, Artur Kania for chick Lhx1 gene, and Len Pennacchio for help with initial characterization of enhancers in mice. We are also grateful to Artur Kania and Sara Wilson for comments on the manuscript. This work was supported by grants to AK from the Israel Science Foundation, and DFG (German Research Foundation). AV was supported by grants from the National Human Genome Research Institute and from the National Institute for Neurological Disorders and Stroke.
- Tessier Lavigne M, Goodman CS: The molecular biology of axon guidance. Science. 1996, 274: 1123-1133. 10.1126/science.274.5290.1123.View ArticlePubMedGoogle Scholar
- Yu TW, Bargmann CI: Dynamic regulation of axon guidance. Nat Neurosci. 2001, 4 (Suppl): 1169-1176. 10.1038/nn748.View ArticlePubMedGoogle Scholar
- Shirasaki R, Pfaff SL: Transcriptional codes and the control of neuronal identity. Annu Rev Neurosci. 2002, 25: 251-281. 10.1146/annurev.neuro.25.112701.142916.View ArticlePubMedGoogle Scholar
- Kania A, Jessell TM: Topographic motor projections in the limb imposed by LIM homeodomain protein regulation of ephrin-A:EphA interactions. Neuron. 2003, 38: 581-596. 10.1016/S0896-6273(03)00292-7.View ArticlePubMedGoogle Scholar
- Kania A, Johnson RL, Jessell TM: Coordinate roles for LIM homeobox genes in directing the dorsoventral trajectory of motor axons in the vertebrate limb. Cell. 2000, 102: 161-173. 10.1016/S0092-8674(00)00022-2.View ArticlePubMedGoogle Scholar
- Pfaff SL, Mendelsohn M, Stewart CL, Edlund T, Jessell TM: Requirement for LIM homeobox gene Isl1 in motor neuron generation reveals a motor neuron-dependent step in interneuron differentiation. Cell. 1996, 84: 309-320. 10.1016/S0092-8674(00)80985-X.View ArticlePubMedGoogle Scholar
- Sharma K, Sheng HZ, Lettieri K, Li H, Karavanov A, Potter S, Westphal H, Pfaff SL: LIM homeodomain factors Lhx3 and Lhx4 assign subtype identities for motor neurons. Cell. 1998, 95: 817-828. 10.1016/S0092-8674(00)81704-3.View ArticlePubMedGoogle Scholar
- Thaler JP, Koo SJ, Kania A, Lettieri K, Andrews S, Cox C, Jessell TM, Pfaff SL: A postmitotic role for Isl-class LIM homeodomain proteins in the assignment of visceral spinal motor neuron identity. Neuron. 2004, 41: 337-350. 10.1016/S0896-6273(04)00011-X.View ArticlePubMedGoogle Scholar
- Thaler JP, Lee SK, Jurata LW, Gill GN, Pfaff SL: LIM factor Lhx3 contributes to the specification of motor neuron and interneuron identity through cell-type-specific protein-protein interactions. Cell. 2002, 110: 237-249. 10.1016/S0092-8674(02)00823-1.View ArticlePubMedGoogle Scholar
- Hobert O, Westphal H: Functions of LIM-homeobox genes. Trends Genet. 2000, 16: 75-83. 10.1016/S0168-9525(99)01883-1.View ArticlePubMedGoogle Scholar
- Thor S, Andersson SG, Tomlinson A, Thomas JB: A LIM-homeodomain combinatorial code for motor-neuron pathway selection. Nature. 1999, 397: 76-80. 10.1038/16275.View ArticlePubMedGoogle Scholar
- Thor S, Thomas JB: The Drosophila islet gene governs axon pathfinding and neurotransmitter identity. Neuron. 1997, 18: 397-409. 10.1016/S0896-6273(00)81241-6.View ArticlePubMedGoogle Scholar
- Helms AW, Johnson JE: Specification of dorsal spinal cord interneurons. Curr Opin Neurobiol. 2003, 13: 42-49. 10.1016/S0959-4388(03)00010-2.View ArticlePubMedGoogle Scholar
- Lee KJ, Jessell TM: The specification of dorsal cell fates in the vertebrate central nervous system. Annu Rev Neurosci. 1999, 22: 261-294. 10.1146/annurev.neuro.22.1.261.View ArticlePubMedGoogle Scholar
- Helms AW, Johnson JE: Progenitors of dorsal commissural interneurons are defined by MATH1 expression. Development. 1998, 125: 919-928.PubMedGoogle Scholar
- Wilson SI, Shafer B, Lee KJ, Dodd J: A molecular program for contralateral trajectory: Rig-1 control by LIM homeodomain transcription factors. Neuron. 2008, 59: 413-424. 10.1016/j.neuron.2008.07.020.View ArticlePubMedGoogle Scholar
- Gowan K, Helms AW, Hunsaker TL, Collisson T, Ebert PJ, Odom R, Johnson JE: Crossinhibitory activities of Ngn1 and Math1 allow specification of distinct dorsal interneurons. Neuron. 2001, 31: 219-232. 10.1016/S0896-6273(01)00367-1.View ArticlePubMedGoogle Scholar
- Bermingham NA, Hassan BA, Wang VY, Fernandez M, Banfi S, Bellen HJ, Fritzsch B, Zoghbi HY: Proprioceptor pathway development is dependent on Math1. Neuron. 2001, 30: 411-422. 10.1016/S0896-6273(01)00305-1.View ArticlePubMedGoogle Scholar
- Arakawa T, Iwashita M, Matsuzaki F, Suzuki T, Yamamoto T: Paths, elongation, and projections of ascending chick embryonic spinal commissural neurons after crossing the floor plate. Brain Res. 2008, 1223: 25-33. 10.1016/j.brainres.2008.05.062.View ArticlePubMedGoogle Scholar
- Reeber SL, Sakai N, Nakada Y, Dumas J, Dobrenis K, Johnson JE, Kaprielian Z: Manipulating Robo expression in vivo perturbs commissural axon pathfinding in the chick spinal cord. J Neurosci. 2008, 28: 8698-8708. 10.1523/JNEUROSCI.1479-08.2008.PubMed CentralView ArticlePubMedGoogle Scholar
- Lee SK, Pfaff SL: Synchronization of neurogenesis and motor neuron specification by direct coupling of bHLH and homeodomain transcription factors. Neuron. 2003, 38: 731-745. 10.1016/S0896-6273(03)00296-4.View ArticlePubMedGoogle Scholar
- Timmer J, Johnson J, Niswander L: The use of in ovo electroporation for the rapid analysis of neural-specific murine enhancers. Genesis. 2001, 29: 123-132. 10.1002/gene.1015.View ArticlePubMedGoogle Scholar
- Zisman S, Marom K, Avraham O, Rinsky-Halivni L, Gai U, Kligun G, Tzarfaty-Majar V, Suzuki T, Klar A: Proteolysis and membrane capture of F-spondin generates combinatorial guidance cues from a single molecule. J Cell Biol. 2007, 178: 1237-1249. 10.1083/jcb.200702184.PubMed CentralView ArticlePubMedGoogle Scholar
- Pennacchio LA, Ahituv N, Moses AM, Prabhakar S, Nobrega MA, Shoukry M, Minovitsky S, Dubchak I, Holt A, Lewis KD: In vivo enhancer analysis of human conserved non-coding sequences. Nature. 2006, 444: 499-502. 10.1038/nature05295.View ArticlePubMedGoogle Scholar
- Visel A, Minovitsky S, Dubchak I, Pennacchio LA: VISTA Enhancer Browser – a database of tissue-specific human enhancers. Nucleic Acids Res. 2007, 35: D88-D92. 10.1093/nar/gkl822.PubMed CentralView ArticlePubMedGoogle Scholar
- Visel A, Prabhakar S, Akiyama JA, Shoukry M, Lewis KD, Holt A, Plajzer-Frick I, Afzal V, Rubin EM, Pennacchio LA: Ultraconservation identifies a small subset of extremely constrained developmental enhancers. Nat Genet. 2008, 40: 158-160. 10.1038/ng.2007.55.PubMed CentralView ArticlePubMedGoogle Scholar
- Nakada Y, Parab P, Simmons A, Omer-Abdalla A, Johnson JE: Separable enhancer sequences regulate the expression of the neural bHLH transcription factor neurogenin 1. Dev Biol. 2004, 271: 479-487. 10.1016/j.ydbio.2004.04.021.View ArticlePubMedGoogle Scholar
- Lee KJ, Mendelsohn M, Jessell TM: Neuronal patterning by BMPs: a requirement for GDF7 in the generation of a discrete class of commissural interneurons in the mouse spinal cord. Genes Dev. 1998, 12: 3394-3407. 10.1101/gad.12.21.3394.PubMed CentralView ArticlePubMedGoogle Scholar
- Saueressig H, Burrill J, Goulding M: Engrailed-1 and netrin-1 regulate axon pathfinding by association interneurons that project to motor neurons. Development. 1999, 126: 4201-4212.PubMedGoogle Scholar
- Alvarez FJ, Jonas PC, Sapir T, Hartley R, Berrocal MC, Geiman EJ, Todd AJ, Goulding M: Postnatal phenotype and localization of spinal cord V1 derived interneurons. J Comp Neurol. 2005, 493: 177-192. 10.1002/cne.20711.PubMed CentralView ArticlePubMedGoogle Scholar
- Saba R, Nakatsuji N, Saito T: Mammalian BarH1 confers commissural neuron identity on dorsal cells in the spinal cord. J Neurosci. 2003, 23: 1987-1991.PubMedGoogle Scholar
- Bovolenta P, Dodd J: Guidance of commissural growth cones at the floor plate in the embryonic rat spinal cord. Development. 1990, 109: 435-447.PubMedGoogle Scholar
- Imondi R, Kaprielian Z: Commissural axon pathfinding on the contralateral side of the floor plate: a role for B-class ephrins in specifying the dorsoventral position of longitudinally projecting commissural axons. Development. 2001, 128: 4859-4871.PubMedGoogle Scholar
- Kadison SR, Kaprielian Z: Diversity of contralateral commissural projections in the embryonic rodent spinal cord. J Comp Neurol. 2004, 472: 411-422. 10.1002/cne.20086.View ArticlePubMedGoogle Scholar
- Mansouri A, Yokota Y, Wehr R, Copeland NG, Jenkins NA, Gruss P: Paired-related murine homeobox gene expressed in the developing sclerotome, kidney, and nervous system. Dev Dyn. 1997, 210: 53-65. 10.1002/(SICI)1097-0177(199709)210:1<53::AID-AJA6>3.0.CO;2-0.View ArticlePubMedGoogle Scholar
- Theodorakis K, Kyriakopoulou K, Wassef M, Karagogeos D: Novel sites of expression of the bHLH gene NSCL1 in the developing nervous system. Gene Expr Patterns. 2002, 2: 105-108. 10.1016/S0925-4773(02)00297-6.View ArticlePubMedGoogle Scholar
- Dasen JS, Liu JP, Jessell TM: Motor neuron columnar fate imposed by sequential phases of Hox-c activity. Nature. 2003, 425: 926-933. 10.1038/nature02051.View ArticlePubMedGoogle Scholar
- Dasen JS, Tice BC, Brenner-Morton S, Jessell TM: A Hox regulatory network establishes motor neuron pool identity and target-muscle connectivity. Cell. 2005, 123: 477-491. 10.1016/j.cell.2005.09.009.View ArticlePubMedGoogle Scholar
- Lyuksyutova AI, Lu CC, Milanesio N, King LA, Guo N, Wang Y, Nathans J, Tessier-Lavigne M, Zou Y: Anterior-posterior guidance of commissural axons by Wnt-frizzled signaling. Science. 2003, 302: 1984-1988. 10.1126/science.1089610.View ArticlePubMedGoogle Scholar
- Bourikas D, Pekarik V, Baeriswyl T, Grunditz A, Sadhu R, Nardo M, Stoeckli ET: Sonic hedgehog guides commissural axons along the longitudinal axis of the spinal cord. Nat Neurosci. 2005, 8: 297-304. 10.1038/nn1396.View ArticlePubMedGoogle Scholar
- Helms AW, Battiste J, Henke RM, Nakada Y, Simplicio N, Guillemot F, Johnson JE: Sequential roles for Mash1 and Ngn2 in the generation of dorsal spinal cord interneurons. Development. 2005, 132: 2709-2719. 10.1242/dev.01859.PubMed CentralView ArticlePubMedGoogle Scholar
- Pillai A, Mansouri A, Behringer R, Westphal H, Goulding M: Lhx1 and Lhx5 maintain the inhibitory-neurotransmitter status of interneurons in the dorsal spinal cord. Development. 2007, 134: 357-366. 10.1242/dev.02717.View ArticlePubMedGoogle Scholar
- Ma J, Ptashne M: A new class of yeast transcriptional activators. Cell. 1987, 51: 113-119. 10.1016/0092-8674(87)90015-8.View ArticlePubMedGoogle Scholar
- BBSRC ChickEST Database. [http://www.chick.manchester.ac.uk/]
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.