Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Aug 29.
Published in final edited form as: Somatosens Mot Res. 2012 Mar 7;29(1):1–12. doi: 10.3109/08990220.2011.650869

The transcription factor, Lmx1b, promotes a neuronal glutamate phenotype and suppresses a GABA one in the embryonic trigeminal brainstem complex

Chuan-Xi Xiang a,d, Kai-Hua Zhang a,d, Randy L Johnson b, Mark F Jacquin c, Zhou-Feng Chen a
PMCID: PMC3756596  NIHMSID: NIHMS495945  PMID: 22397680

Abstract

Achieving an appropriate balance between inhibitory and excitatory neuronal fate is critical for development of effective synaptic transmission. However, the molecular mechanisms dictating such phenotypic outcomes are not well understood, especially in the whisker-to-barrel cortex neuraxis, an oft-used model system for revealing developmental mechanisms. In trigeminal nucleus principalis (PrV), the brainstem link in the whisker-barrel pathway, the transcription factor Lmx1b marks glutamatergic cells. In PrV of Lmx1b knockout mice (−/−), initial specification of glutamatergic versus GABAergic cell fate is normal until embryonic day 14.5. Subsequently until the day of birth, glutamatergic markers (e.g. VGLUT2) stain significantly fewer PrV neurons, whereas, GABAergic markers (Pax2 and Gad1) stain significantly more PrV cells, notably in Lmx1b-null PrV cells. These changes also occurred in Lmx1b/Bax double −/− mice, where PrV cells are rescued from Lmx1b−/−-induced apoptosis; thus, effects upon excitatory/inhibitory cell ratios do not reflect a cell death confound. Electroporation-induced ectopic expression of Lmx1b in an array of sites decreases numbers of neurons that express GABAergic markers, but increases VGLUT2+ cell numbers or stain intensity. Thus, Lmx1b is not involved in the initial specification of glutamatergic cell fate, but is essential for maintaining a glutamatergic phenotype. Other experiments suggest that Lmx1b acts to suppress Pax2, a promoter of GABAergic cell fate, in a cell-autonomous manner, which may be a mechanism for maintaining a functional balance of glutamatergic and GABAergic cell types in development.

Keywords: Principalis, Whiskers, Barrels, Neurotransmission, Bax, Homeodomain

INTRODUCTION

Trigeminal (V) primary afferent neurons project to several brainstem nuclei that integrate somatotopic information from the face and convey it to the thalamus and somatosensory (S1) cortex (reviewed in Woolsey, 1990; Henderson and Jacquin, 1995). These inputs are processed and relayed by excitatory or inhibitory neurons in the V brainstem complex which is comprised of the V nucleus principalis (PrV) and the spinal trigeminal subnuclei (SpV). PrV originates the V lemniscal pathway that is the essential conduit for transmission of a whisker-related pattern to thalamic barreloids in development (Killackey and Fleming, 1985; Ding et al., 2003). PrV is also known to consist mostly of glutamatergic cells, most of which project to the contralateral thalamus, and a smaller number of GABAergic local circuit neurons (Magnusson et al., 1987; Kaneko et al., 1987; Haring et al., 1990; Clements and Beitz, 1991; Bae et al., 2000). Recent studies have begun to reveal the molecular mechanisms that govern the development of the PrV-based lemniscal pathway (reviewed in Erzurumlu et al., 2006; Jacquin et al., 2008; Xiang et al., 2010; da Silva et al., 2011), but the molecular machinery that controls the determination of a glutamatergic versus GABAergic phenotype in developing PrV neurons is unclear. This has important system implications because inhibitory local circuits have a profound influence upon the response properties of whisker-sensitive PrV cells (Minnery et al., 2003; Furuta et al., 2008).

In the mammalian central nervous system, specification of a glutamatergic versus GABAergic cell fate has been shown to occur at multiple molecular levels. In the retina, cerebellum, forebrain and dorsal spinal cord, the transcription factors Ngn1, Ascl1, Gsh1/2, Otx2 and Ptf1a are expressed in neuronal progenitor cells and influence whether a neuron becomes glutamatergic or GABAergic (Glasgow et al., 2005; Mizuguchi et al., 2006; Hoshino et al., 2005; Puelles et al., 2006; Dullin et al., 2007). In postmitotic cells, other transcription factors have also been shown to be necessary for the specification of neurotransmitter phenotype. In the developing midbrain, Gata2 has been shown to be a necessary post-mitotic selector gene for GABAergic neurons (Kala et al., 2009). In the developing dorsal spinal cord, Tlx1 and Tlx3 serve as key selectors in defining the fate of glutamatergic neurons, whereas Lbx1 and Pax2 may function as selector genes to specify a GABAergic fate (Cheng et al., 2004, 2005; Batista and Lewis, 2008). While Tlx3 may function to antagonize Lbx1 to promote a glutamatergic cell fate, it is unclear whether a similar factor is required to suppress Pax2 and, consequently, a GABAergic phenotype. In this regard, Lhx1 and Lhx5 are known to activate and maintain Pax2 expression which is necessary for the differentiation of GABAergic cells in the spinal cord (Pillai et al., 2007). Lmx1b, a LIM homeodomain transcription factor whose expression is restricted to glutamatergic cells and not colocalized with Pax2 in the dorsal horn, may have a role analogous to the Tlx genes noted above. To date, however, there is no evidence in support of Pax2 upregulation in the dorsal horn of Lmx1b−/− mice, nor is there evidence of upregulation of glutamatergic markers in the dorsal horn of Pax2−/− mutants (Cheng et al., 2004; Ding et al., 2004).

The present study addresses the function of Lmx1b in determining neurotransmitter phenotype in the PrV of embryonic mice. We (Xiang et al., 2010) have recently reported that Lmx1b is necessary for whisker-related pattern formation in PrV. Here, Lmx1b and Pax2 expression in the PrV were found to be mutually exclusive. Loss-of-function and gain-of function approaches were then utilized to test the hypotheses that Lmx1b promotes a neuronal glutamatergic phenotype and suppresses a GABAergic phenotype and that Lmx1b accomplishes this by suppressing Pax2 expression.

MATERIAL AND METHODS

Animals

The generation and genotyping of Lmx1b−/− and Lmx1b+/− mice were performed as previously described (Chen et al., 1998; Xiang et al., 2010). Lmx1b/Bax double null mice were generated by breeding Lmx1b+/− and Bax−/− mice (Jacquin et al., 2008); Bax mutants were on a C57BL/6 background. Lmx1b+/− mice were maintained in the mouse facility accordingto protocols approved by the Division of Comparative Medicineat Washington University.

BrdU/Lmx1b double labeling

To assess the mitotic status of Lmx1b expressing cells, BrdU pulse-labeling was followed by Lmx1b immunohistochemistry. BrdU was injected into timed mated female mice on embryonic day (E) 11.5 and embryos were dissected out after 2 hrs for immunocytochemical analysis. The processing of the embryos and double staining for BrdU/Lmx1b were performed as previously described (Ding et al., 2004).

In situ hybridization and immunocytochemistry

In situ hybridization and immunocytochemical staining were performed as previously described (Chen et al., 2001; Ding et al., 2003; Ding et al., 2004). Sections were incubated with rabbit anti-Lmx1b (1:1500), goat anti--Lmx1b (1:200) or rabbit anti-Pax2 (1:300) in phosphate buffered saline (PBS) containing 2% normal donkey serum and 0.3% TritonX-100 overnight, followed by incubation in biotinylated donkey anti-goat or anti-rabbit (Jackson Immuno Research, 1:200) for 2 hrs, followed by FITC or cy3-conjugated streptavidin (Molecular Probes, 1:1000) for 1 hr. After washing with PBS, slides were coverslipped and observed under an Olympus fluorescent microscope (BX51). For immunostaining after in situ hybridization, protocols for incubation in primary antibodies and biotinylated donkey anti-rabbit IgG were identical to those described above. Following incubation with secondary antibody, ABC reagents (Vector) or TSA reagents were used. Controls for antibody and probe specificity were run in tandem with initial experiments and they included deletion of the primary antibody and substitution of a sense probe, respectively.

In utero electroporation

To introduce exogenous Lmx1b into the embryonic central nervous system, in utero electroporation was performed as previously described (Ding et al., 2004). Briefly, the uterine horns were exposed and 2–4 μl of plasmid solution was injected through the uterus into the lateral or third ventricles of the embryonic brain, or central canal of the embryonic spinal cord, with a glass micropipette. Embryos were then placed between the electrodes of an ECM830 electroporator, and five 50-msec pulses of 25–40 volts were delivered. The uterus was then placed back into the repaired abdominal cavity and embryonic development continued until appropriate later stages for phenotypic analyses.

Cell counting and statistical analysis

Changes in total numbers of labeled PrV cells were estimated in consecutive series of sections taken through the pons in each case. Four to six animals from each experimental and wild-type (or plasma negative) control group provided such data. The V motor nucleus and spinal tract served as cytoarchitectural landmarks to ascertain the PrV and adjoining structures, such as the supratrigeminal nucleus (SuV) and the SpV subnucleus interpolaris (SpVi). Images used in these analyses were digitally stored. For cell counting after electroporation, at least 4 slides deemed to have the highest electroporation efficiency from positive embryos were selected for analysis. Data are presented as means ±S.E.M, and Student’s independent t tests wereused for statistical analysis with p < .05 considered significant.

RESULTS

Lmx1b marks glutamatergic cells and Pax2 marks GABAergic cells in the PrV

Expression of Lmx1b and VGLUT2, a vesicular glutamate transporter that is considered to mark all glutamatergic neurons (Fremeau et al., 2001), were assessed. During development, Lmx1b is abundantly expressed in the PrV (Fig. 1a). Expression of Lmx1b was also detected to a lesser degree in the SpVi and SuV (Fig. 1a; data not shown). When considering all brainstem regions receiving primary afferent inputs, VGLUT2 was found to be most robustly expressed in the PrV and SuV (Fig. 1b). To assess the extent to which Lmx1b and VGLUT2 are coexpressed, in situ hybridization of VGLUT2 followed by Lmx1b immunocytochemistry were performed. Double labeling for Lmx1b and VGLUT2 showed that the vast majority (95%) of Lmx1b+ cells are VGLUT2+ in the PrV (Fig. 1d, h), indicating that Lmx1b expression is largely restricted to glutamatergic PrV cells.

Figure 1.

Figure 1

Open in a new tab

Expression of molecular markers in transverse sections through the PrV. (a–c) On the day of birth (P0), Lmx1b (a) and VGLUT2 (b) are abundantly expressed in the PrV, whereas Gad1 staining occurs in a smaller number of PrV cells (c). (d) Double staining of VGLUT2 (blue) and Lmx1b (yellow) reveals robust colocalization (arrows) of Lmx1b and VGLUT2 in P0 PrV cells. (e) Lmx1b (red cells, arrowhead) is not colocalized with Pax2 (green cells, arrow) in PrV on E14.5. (f) A majority of Pax2+ cells (red) also express Gad1 (green, arrows indicate double labeled cells) in PrV on E14.5, while almost all Gad1+ cells co-express Pax2. (g) Lmx1b (red, arrowheads) is not colocalized with BrdU staining (green, arrows). (h) 95.2% of Lmx1b+ cells coexpress VGLUT2, while 78.5% of Pax2+ cells coexpress Gad1. Mo: V motor nucleus, t: spinal V tract, SuV: supratrigeminal nucleus. Dotted lines outline the PrV. Scale bars: 100 μm (a–c) or 50 μm (d–f).

Similar staining procedures were utilized to assess potential Lmx1b colocalization with Pax2, a transcription factor marking GABAergic neurons in the dorsal spinal cord (Cheng et al., 2004). On E14.5, Pax2+ cells were most numerous in the lateral regions of the PrV (Fig. 1e). Double staining for Lmx1b and Pax2 failed to reveal coexpression in PrV cells. Mutual exclusion of Lmx1b and Pax2 labeling in PrV cells was the case at all embryonic timepoints examined (Fig. 1e; data not shown). To determine whether Pax2 reliably marks GABAergic cells in the PrV, double-staining of Pax2 and Gad1, the rate-limiting enzyme that is essential for the synthesis of GABA (Mugnaini and Oertel, 1985; Erlander et al., 1991), was performed. Although Pax2+ cells were much more numerous than Gad1+ cells prior to E13.5 (data not shown), from E14.5 onwards, Gad1+ and Pax2+ staining usually occurred in the same cells (78%, Fig. 1f, h), indicating that Pax2 is a fairly reliable marker for GABAergic cells in the PrV during the last prenatal week. To determine whether Lmx1b expression is restricted to postmitotic cells, BrdU labeling was examined at E11.5 after 2 hours of BrdU exposure. No BrdU+ and Lmx1b+ cells were colocalized (Fig. 1g). Thus, in the hindbrain region where PrV cells are generated, Lmx1b is only expressed in postmitotic cells.

In Lmx1b mutant mice, glutamatergic and GABAergic markers are normally expressed until E14

An evaluation of the expression of neurotransmitter markers in Lmx1b−/− mice failed to reveal a difference in VGLUT2 expression between Lmx1b−/− and wild-type mice until E14.5 (Fig. 2a, b;data not shown). Similarly, there was no apparent effect of the mutation upon expression of the GABAergic markers Pax2 and Gad1 in the PrV of Lmx1b−/− mice up to E14.5 (Fig. 2c–f). Thus, Lmx1b is not required for initial specification of glutamatergic versus GABAergic cells in the PrV when it is undergoing extensive mitosis.

Figure 2.

Figure 2

Open in a new tab

Expression of VGLUT2, Gad1 and Pax2 in transverse sections through the PrV of wild-type and Lmx1b−/− mice at E14.5. (a, b) normal expression of VGLUT2 in Lmx1b−/− mice. (c,d) Gad1 expression in Lmx1b−/− mice is similar to that of wild-type controls. (e, f) Pax2 expression is not altered in Lmx1b−/− mice. WT: wild-type. Scale bars: 100 μm (a–f).

After E14.5, in Lmx1b mutants, glutamatergic PrV cells are fewer than normal and GABAergic PrV cells are more numerous

Post-mitotic (post-E14.5) PrV cells are significantly less likely to express VGLUT2 in the absence of Lmx1b as compared to wild-type controls. For example, on P0, Figures 3a and b illustrate marked downregulation of VGLUT2 relative to that which occurs in the wild-type PrV. This conclusion was supported by estimates of labeled cell numbers. (Fig. 3e). Staining for another glutamatergic neuron-related marker, Gria2, which encodes the AMPA-class receptor for glutamate (Kerr et al., 1998), was completely absent in the PrV from Lmx1b−/− mice (Fig. 3c, d). These data suggest that the development of glutamatergic cells is severely compromised in the PrV of Lmx1b−/− mice in the last week of embryonic development.

Figure 3.

Figure 3

Open in a new tab

Expression of glutamatergic markers is reduced or absent in the Lmx1b−/− PrV at P0. (a,b) VGLUT2 is down-regulated in the mutant (b), as compared to controls (a). (c,d) Gria2 staining is absent in the Lmx1b−/− PrV (d). (e) Estimates of total cell numbers revealed a significant reduction of the number of VGLUT2+ cells in the Lmx1b−/− PrV (201.60±10.78), as compared to wild-type controls (459.83±22.73) (p < 0.001). Scale bar: 100 μm (a–d).

Conversely, the development of GABAergic PrV cells was significantly facilitated in the absence of Lmx1b after E14.5 when Gad1+ cells were more numerous in the PrV of Lmx1b−/−mice relative to that of wild-type controls (Fig. 4a–c). Expression of another GABAergic marker, VGAT, which encodes the vesicular GABA transporter (Chaudry et al., 1998), was also significantly upregulated in Lmx1b−/− mice when compared with wild-type controls (Fig. 4d–f). A final example pertains to Pax2, which marks GABAergic PrV cells at late embryonic times. In Lmx1b−/− mice, Pax2 staining was normal prior to E14.5, but gradually increased to label about five times greater than normal numbers of PrV cells by the day of birth (Figs. 4g–i, Figs. 5a and b; data not shown). Taken together, these results are consistent with the notion that a significant Lmx1b−/− phenotype is a switch from glutamatergic to GABAergic fate in PrV neurons. The latter also appears to have occurred in the nearby SupV and SpVi subnuclei (Fig. 4; data not shown).

Figure 4.

Figure 4

Open in a new tab

Increased numbers of PrV cells expressing GABAergic markers in Lmx1b−/−. Expression of Gad1 increased in Lmx1b−/− mice (b) relative to wild-type control (a). Gad1+cells were more numerous in the PrV of Lmx1b−/− mice compared with the control (c, p < 0.001). (d–f) VGAT is similarly up-regulated in the Lmx1b−/− PrV (p < 0.001). (g–i) A significant increase in the number of Pax2+ cells in the Lmx1b−/− PrV, as compared to controls (P < 0.001). WT: wild-type, Mo: V motor nucleus, t: V spinal tract, SuV: supratrigeminal nucleus. Scale bar: 100 μm (a, b, d, e, g, h).

Figure 5.

Figure 5

Open in a new tab

Pax2 upregulation in the PrV of Lmx1b−/− mice is cell autonomous. (a) Pax2 (red, arrow in the frame) is not colocalized with Lmx1b mRNA (green, arrowhead in the frame) in the wild-type PrV. (b) Many Pax2/Neo double labeled cells (arrow in the frame) in Lmx1b−/− mice are present among numerous Neo+ only cells (green, arrowhead in the frame). (c) Estimates of labeled cell numbers indicate that 11.6% of Pax2+ cells in the Lmx1b−/− PrV are Neo+, whereas only 3.4% of Pax2+ cells in wild-type mice are Lmx1b+, indicating that a disproportionate number of Lmx1b+ cells become Pax2+ in the PrV of Lmx1b−/− mice. t: V spinal tract. Scale bar: 100 μm.

Increased Pax2 staining occurs in those PrV cells lacking Lmx1b

The increased number of Pax2+ cells in the PrV of Lmx1b−/− mice prompted an examination of the genotype of GABAergic PrV cells; namely, does Pax2 expression occur only in cells lacking Lmx1b? In the wild-type PrV on E17.5, Pax2+ cells were intermingled with Lmx1b+ cells, but they were rarely colocalized (Fig. 5a). To identify Lmx1b−/− cells in the PrV, the neo gene was introduced into the Lmx1b locus as a positive selection marker (Chen et al., 1998) that mimics the endogenous Lmx1b expression, therein permitting double staining of Pax2 and neo in the PrV of Lmx1b−/− mice. Because neo expression in Lmx1b+/− heterozygotes was very weak, Lmx1b/Pax2 staining in wild-type mice served as the control group. Relative to Lmx1b+/Pax2+ cells in wild-type control, the number of neo+/Pax2+ cells was significantly increased in the PrV of Lmx1b−/− mice (Fig. 5b.c). Moreover, the increased number of neo+/Pax2+ cells in the mutants was comparable to the increased number of Pax2+ cells in the PrV of Lmx1b−/− mice. These data suggest that the increased number of Pax2+ cells reflects Pax2 staining in cells lacking Lmx1b. However, Pax2 staining was not observed in all Lmx1b−/− cells, suggesting that other factors also regulate Pax2 expression in Lmx1b−/− PrV cells. Nevertheless, it is clear that Lmx1b functions in a cell-autonomous manner in the regulation of Pax2 expression in the PrV.

Altered neurotransmitter phenotypic ratios in the Lmx1b−/− do not reflect a mutation-induced PrV cell death confound

It has been documented that approximately half of all PrV neurons die by the day of birth in Lmx1b mutants (Xiang et al., 2010). The present study’s main observation, that there are increased numbers of GABAergic cells and decreased numbers of glutamatergic cells in the Lmx1b−/−, could simply reflect a less than interesting disproportionate cell death in the glutamatergic PrV cell class, as opposed to Lmx1b-induced specification of neurotransmitter phenotype. This possibility was tested by performing the exact same set of experiments in Lmx1b/Bax double knockouts, where currently studied mutants were bred with mice lacking the apoptosis promoting gene, Bax, therein rendering them resistant to apoptosis (see Jacquin et al., 2008, and Xiang et al., 2010, for details and baseline data). Should the same set of results be obtained in the Lmx1b/Bax double knockouts, then the presently described shift in transmitter phenotype cannot be attributed to a mutation-induced cell death confound. Indeed, this is exactly what was observed. As illustrated in Figure 6, the reduction in numbers of glutamatergic PrV cells observed in Lmx1b single null mutants was also observed in the Lmx1b/Bax double knockouts. Similarly, the increase in numbers of GABAergic PrV cells observed in Lmx1b null mutants was also observed in the Lmx1b/Bax double knockouts. For estimates of total numbers of labeled PrV cells in both of these staining paradigms, single and double null mutants were statistically indistinguishable.

Figure 6.

Figure 6

Open in a new tab

Relative to wild-type (WT) controls (A, E), reduced numbers of glutamatergic, and increased numbers of GABAergic, PrV cells occur in both the Lmx1b null mutant and the Lmx1b/Bax double null mutant, as revealed by staining for VGLUT2 (B, C) and VGAT (F, G) respectively. Estimates of total numbers of labeled PrV cells (ordinate) in these two staining conditions are shown in D and H. Bars containing an X indicate statistically significant differences from controls, at the p < .001 level. Arrows highlight equivalent locations in PrV in transverse sections from different cases.

Lmx1b actions upon transmitter phenotype are not unique to the PrV

To test the generality of above described effects of Lmx1b deletion upon the developing PrV, additional gain-of-function experiments were carried out in the developing spinal dorsal horn, cerebral cortex and superior colliculus, where exogenous Lmx1b is conveniently introduced by electroporation, which is not the case for the the PrV embedded deep within the ventral brainstem.

The findings that Lmx1b and Pax2 expression are mutually exclusive in the PrV, and that Pax2 expression was promoted in Lmx1b−/− cells, raise the possibility that Lmx1b may suffice to suppress Pax2 expression. Because of the technical difficulties in introducing exogenous Lmx1binto the developing PrV by in utero electroporation, misexpression studies were carried out in the dorsal horn of the spinal cord. This site was chosen because of prior indications of mutual exclusivity in Lmx1b and Pax2 expression there (Ding et al., 2004; Gross et al., 2002). Moreover, Pax2 is abundantly expressed in the dorsal horn, which makes for more likely incorporation of exogenous Lmx1b into Pax2+ cells. Exogenous Lmx1b-EGFP plasmids (expressing Lmx1b and EGFP separately via IRES cassette insertion) were introduced into the dorsal horn at E12.5 and dorsal horn phenotype was analyzed a few days later. In the dorsal horn electroporated with EGFP plasmids only, EGFP/Lmx1b+ double labeled cells were observed (Fig. 7b, d). By contrast, when Lmx1b-EGFP plasmids were introduced into the dorsal horn, no EGFP+/Pax2+ co-stained cells were observed (Fig. 7a, c). While it is possible that Lmx1b-EGFP might have been incorporated by Pax2+ cells, this is likely to be rare because EGFP/Pax2 double positive cells were always seen in electroporation control cases (Fig. 7b, d). These data suggest that exogenous Lmx1b is necessary and sufficient to suppress Pax2 expression in the developing dorsal horn.

Figure 7.

Figure 7

Open in a new tab

Electroporation of exogenous Lmx1b suppresses endogenous Pax2 expression in the dorsal spinal cord. (a,b) Transverse sections through the right E18.5 mouse spinal cord electroporated on E12.5 with Lmx1b-EGFP (which expressed Lmx1b and EGFP separately via insertion of an IRES cassette) or EGFP only plasmids. (c,d) EGFP (green) and Pax2 (red) double staining in the same sections as in a,b. The dorsal horn electroporated with Lmx1b-EGFP rarely displays Pax2/EGFP double labeling, whereas many cells do so in the EGFP plasmid control cases. Examples of single and double labeled cells are shown in the insets in c and d. Scale bar: 50 μm applies to all panels.

Whether Lmx1b is inherently capable of promoting a glutamatergic phenotype was then assessed by ectopic expression of Lmx1b-EGFP plasmids in the developing cerebral cortex, where numerous glutamatergic neurons exist and transcription factors are known to impact their genesis (Hevner et al., 2006). An increased expression of VGLUT2 was observed in cortices electroporated with Lmx1b-EGFP plasmids, as compared to normal levels of VGLUT2 staining in the contralateral cortices into which no plasmids were introduced (Fig. 8a, c). By contrast, VGLUT2 expression was normal in cortices electroporated with EGFP only (Fig. 8b, d). Increased VGLUT2 expression could reflect an increased number of VGLUT2+ cells or increased levels of VGLUT2 expression. Insofar as there was no difference in numbers of VGLUT2 labeled cells between Lmx1b electroporated cortices and contralateral controls (Fig. 8f), and EGFP/VGLUT2 double-staining revealed many single- and double-labeled cells in cortices electroporated with Lmx1b-EGFP plasmids (Fig. 8e), Lmx1b alone did not activate cortical VGLUT2 expression in an all-or-none manner. However, examination of the optical density of gene expression revealed that the quantity of VGLUT2 expression increased relative to the control cortices (Fig. 8g). Similar results were obtained when Lmx1b-EGFP plasmids were introduced into the developing hippocampus (data not shown). Thus, exogenous Lmx1b does not appear to be sufficient to induce ectopic VGLUT2 expression, yet it is sufficient to increase the level of expression of VGLUT2.

Figure 8.

Figure 8

Open in a new tab

Ectopic expression of Lmx1b increases the intensity of VGLUT2 expression in transverse sections of the embryonic cerebral cortex. (a,b) Lmx1b-EGFP and EGFP only plasmids were electroporated into the forebrain at E12.5 and analyzed here at E15.5. (c,d) Expression of VGLUT2 in the cortex; EGFP immunocytochemistry followed in situ hybridization with a VGLUT2 probe. (e). Higher magnification epifluorescence of the frame shown in c. Arrows indicate EGFP and VGLUT2 double-stained cells; arrowhead highlights EGFP only cell. (f) The number of VGLUT2+ cells in the Lmx1b electroporated cortex does not differ from the control side, whereas ectopic Lmx1b produces a statistically significant increase in staining intensity in VGLUT2+ cells (g). EP: electroporation; Con: contralateral side. Scale bar: 100 μm, applies to panels A–D.

Lastly, whether Lmx1b is sufficient to suppress expression of GABAergic markers was tested by ectopic expression of Lmx1b in the superior colliculus. The latter contains many GABAergic neurons (Tsunekawa et al., 2005) and no endogenous Lmx1b expression. At E19.5, Gad1 staining revealed numerous GABAergic cells, with the most intense staining in the intermediate zone (Fig. 9a, d). In colliculi electroporated with Lmx1b-EGFP at E14.5, Gad1 expression was markedly reduced relative to the contralateral side into which no Lmx1b-EGFP plasmids were introduced (Fig. 9b, e). An estimate of total numbers of labeled cells revealed a significant reduction in the number of Gad1+ cells as compared to the control side (Fig. 9c). By contrast, the colliculi electroporated with EGFP only showed no change in Gad1+ expression (data not shown). To assess possible coexpression of Gad1 and Lmx1b, Gad1 and EGFP (Lmx1b positive) double staining failed to reveal double labeled cells in the electroporated in the electroporated colliculi (Fig. 9f). These data suggest that ectopic Lmx1b will suppress Gad1 expression in the superior colliculus.

Figure 9.

Figure 9

Open in a new tab

Ectopic expression of Lmx1b suppresses Gad1 expression in transverse sections through the E19.5 superior colliculus. (a,b) a is the control side and b shows the side where Coronal sections through the midbrain at superior colliculus level; a indicates the contralateral side, whereas b indicates the side where Lmx1b-EGFP plasmids were electroporated into the midbrain at E14.5. (c) Total numbers of Gad1+ cells in the colliculus was significantly reduced after Lmx1b electroporation. (d,e) Gad1 expression in adjacent sections as revealed by in situ hybridization and brightfield optics. (f) EGFP immunostaining was performed following Gad1 in situ hybridization. f shows with epifluorescence and at higher magnification the field boxed in e; EGFP (green, arrows) did not occur in Gad1 labeled cells (red, arrowheads). Scale bar: 100 μm.

DISCUSSION

The present study offers data in support of the following conclusions. Lmx1b marks glutamatergic cells, but not GABAergic cells, in the developing PrV. In the absence of Lmx1b, initial specification of glutamatergic versus GABAergic fate appears normal until E14.5. Over the next week, glutamatergic cell number decreases in the PrV of Lmx1b−/− embryos. Conversely, the number of GABAergic PrV cells increases, consistent with the notion that, in the absence of Lmx1b, many cells initially specified as glutamatergic become GABAergic. In addition to such a fate alteration in Lmx1b−/− cells, additional evidence supports the idea that Lmx1b is both necessary and sufficient for suppressing expression of a GABAergic phenotype. Namely, misexpression of Lmx1b in the superior colliculus reduces the number of Gad1+ cells there, and exogenous Lmx1b marked by EGFP is capable of suppressing Pax2 expression in the dorsal spinal cord. Together with the observation that Pax2 expression is upregulated in Lmx1b−/− cells in the PrV, these data strongly suggest that Lmx1b also functions to suppress the GABAergic phenotype.

How does Lmx1b maintain transcriptional suppression of GABAergic-specific gene expression? One possible mechanism is that Lmx1b suppresses Pax2, whose expression is mutually exclusive with that of Lmx1b. This conclusion is prompted by the following observations: 1) Lmx1b deletion produces an upregulation of Pax2 in the PrV in a cell-autonomous manner; 2) Lmx1b misexpression in the dorsal horn suppresses Pax2 expression; 3) upregulation of Pax2 also occurs in other brainstem V regions, such as the SpVi and SuV, in Lmx1b−/− mice. These results strongly support the notion that Pax2/Lmx1b mutual exclusivity is a general transcriptional mechanism used in multiple developing hindbrain and spinal cord regions to maintain an effective ratio of glutamatergic versus GABAergic cell fates.

The present study also suggests that not all glutamatergic PrV cells adopt a GABAergic cell fate in the absence of Lmx1b because Gad1 expression was suppressed in only a minority of Lmx1b−/− cells. Such an impression of selective suppression may reflect the use of only a limited number of GABAergic markers in examining Lmx1b−/− cells, therein resulting in our failure to visualize more GABAergic cells. Alternatively, other yet to be revealed GABAergic suppressors may have retained their function in these cells. The latter interpretation is favored in light of the observation here that, prior to E14.5, Lmx1b deletion had no effect upon the ratio of stained neurotransmitter phenotypes in PrV. Other unknown transcription factors may serve to maintain the glutamatergic phenotype at particular embryonic timepoints. Because the timing of VGLUT2 downregulation begins at E15.5, coinciding with the onset of GABAergic marker upregulation, it may also be the case that GABAergic phenotype suppression is the default signaling pathway, though further study is required to test this idea.

An important conclusion derived from the present study is that Lmx1b is essential for maintaining or strengthening the glutamatergic phenotype. In the PrV of Lmx1b−/− mice, VGLUT2 staining is significantly reduced during the final embryonic week. Conversely, introducing Lmx1b to the late prenatal cerebral cortex enhances the intensity of VGLUT2 staining. These results demonstrate that Lmx1b imposes maintenance or promotion of the glutamatergic phenotype during development. However, because ectopic induction of Lmx1b expression in the cortex does not induce glutamatergic expression, but rather enhances such in specific locations, Lmx1b must work in concert with other existing glutamatergic-specific factors to sculpt excitatory circuits in the developing cortex. Moreover, it remains unclear whether Lmx1b regulates VGLUT2 expression directly or indirectly. Other transcription factors may also be involved in the specification and/or maintenance of the glutamatergic fate decision. One likely candidate is Tlx3 which has been shown to be expressed in glutamatergic cells in the dorsal spinal cord (Cheng et al., 2004; Xu et al., 2008). Inasmuch as Tlx3 is also expressed in the developing PrV (data not shown), it may function similarly there, perhaps acting synergistically with Lmx1b.

Prior studies have revealed the Tlx1 and Tlx3 genes as postmitotic determinants of the glutamatergic phenotype, possibly via selective suppression of Lbx1 function (Cheng et al., 2004, 2005). Here, Lmx1b is revealed to do the same in PrV, possibly via Pax2, a promoter of GABAergic cell fate, in a cell-autonomous manner. Yet, one important distinction between Tlx1/3 and Lmx1b is that the latter is not required for the initial specification of glutamatergic versus GABAergic cell fate, which is also the case for Lmx1b in the dorsal horn (Ding et al., 2004. And, Lmx1b is unique in its duality, facilitating glutamatergic phenotype at the expense of the GABAergic phenotype. Taken together, information to date suggests that Lmx1b is a third class of transcription factors that is neither expressed in the progenitor cells nor essential for the initial specification of glutamatergic cell fate. This class of transcription factors is essential for maintaining glutamatergic phenotype at late embryonic stages and does so by suppressing aGABAergic phenotype in nascent glutamatergic cells. Thus, Lmx1b is a gene that is necessary for development of excitatory and inhibitory circuits in the developing brainstem.

A final implication of the present study pertains to a potential mechanism underlying Lmx1b’s essential role in PrV pattern formation. PrV normally develops sufficiently by the day of birth to view a whisker-related barrel-like pattern there, thus obviating the experimental limitations imposed by perinatal lethality in the Lmx1b−/−. We (Xiang et al., 2010) have previously shown that PrV patterning of whiskers (barrelettes) fails to form in the newborn Lmx1b−/−. While there are compelling data (Henderson et al., 1992) that fail to support a necessary role for neuronal impulse activity in the development of such patterns in PrV, there are indications (reviewed in Erzurumlu et al., 2010) that excitatory synaptic transmission is necessary for barrelette formation in PrV. If the latter is the case, then the dramatic diminution of glutamatergic cells in the Lmx1b−/− PrV would be sufficient to disrupt the formation of this pattern. Selective transcriptional suppression of Pax2 by Lmx1b may, by switching the balance of PrV circuitry in favor of requisite excitatory neurotransmission, be a patterning mechanism in the developing V system. The receptive field and electrophysiological properties of PrV cells in Lmx1b−/−s are, therefore, important topics for future study.

Acknowledgments

We thank J. Yin for genotyping the Lmx1b mutant mice and T. Jessell and Y.Q. Ding for the anti-Lmx1b antibody. This project was supported by NIH Grant P01-NS049048 to Z.F.C and M.J.

References

  1. Bae YC, Ihn HJ, Park MJ, Ottersen OP, Moritani M, Yoshida A, Shigenaga Y. Identification of signal substances in synapses made between primary afferents and their associated axon terminals in the rat trigeminal sensory nuclei. J Comp Neurol. 2000;418:299–309. [PubMed] [Google Scholar]
  2. Batista MF, Lewis KE. Pax2/8 act redundantly to specify glycinergic and GABAergic fates of multiple spinal interneurons. Dev Biol. 2008;323:88–97. doi: 10.1016/j.ydbio.2008.08.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chaudhry FA, Reimer RJ, Bellocchio EE, Danbolt NC, Osen KK, Edwards RH, Storm-Mathisen J. The vesicular GABA transporter, VGAT, localizes to synaptic vesicles in sets of glycinergic as well as GABAergic neurons. J Neurosci. 1998;18:9733–50. doi: 10.1523/JNEUROSCI.18-23-09733.1998. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen H, Lun Y, Ovchinnikov D, Kokubo H, Oberg KC, Pepicelli CV, Gan L, Lee B, Johnson RL. Limb and kidney defects in lmx1b mutant mice suggest an involvement of LMX1B in human nail patella syndrome. Nat Genet. 1998;19:51–55. doi: 10.1038/ng0598-51. [DOI] [PubMed] [Google Scholar]
  5. Chen ZF, Rebelo S, White F, Malmberg AB, Baba H, Lima D, Woolf CJ, Basbaum AI, Anderson DJ. The paired homeodomain protein DRG11 is required for the projection of cutaneous sensory afferent fibers to the dorsal spinal cord. Neuron. 2001;31:59–73. doi: 10.1016/s0896-6273(01)00341-5. [DOI] [PubMed] [Google Scholar]
  6. Cheng L, Arata A, Mizuguchi R, Qian Y, Karunaratne A, Gray PA, Arata S, Shirasawa S, Bouchard M, Luo P, Chen CL, Busslinger M, Goulding M, Onimaru H, Ma Q. Tlx3 and Tlx1 are post-mitotic selector genes determining glutamatergic over GABAergic cell fates. Nat Neurosci. 2004;7:510–517. doi: 10.1038/nn1221. [DOI] [PubMed] [Google Scholar]
  7. Cheng L, Samad OA, Xu Y, Mizuguchi R, Luo P, Shirasawa S, Goulding M, Ma Q. Lbx1 and Tlx3 are opposing switches in determining GABAergic versus glutamatergic transmitter phenotypes. Nat Neurosci. 2005;8:1510–1515. doi: 10.1038/nn1569. [DOI] [PubMed] [Google Scholar]
  8. Clements JR, Beitz AJ. An electron microscopic description of glutamate-like immunoreactive axon terminals in the rat principal sensory and spinal trigeminal nuclei. J Comp Neurol. 1991;309:271–280. doi: 10.1002/cne.903090208. [DOI] [PubMed] [Google Scholar]
  9. da Silva S, Hasegawa H, Scott A, Zhou X, Wagner AK, Han BX, Wang F. Proper formation of whisker barrelettes requires periphery-derived Smad4-dependent TGF-β signaling. P N A S. 2011;108:3395–3400. doi: 10.1073/pnas.1014411108. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Ding YQ, Yin J, Xu HM, Jacquin MF, Chen ZF. Formation of whisker-related principal sensory nucleus-based lemniscal pathway requires a paired homeodomain transcription factor, Drg11. J Neurosci. 2003;23:7246–7254. doi: 10.1523/JNEUROSCI.23-19-07246.2003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Ding YQ, Yin J, Kania A, Zhao ZQ, Johnson RL, Chen ZF. Lmx1b controls the differentiation and migration of the superficial dorsal horn neurons of the spinal cord. Development. 2004;131:3693–3703. doi: 10.1242/dev.01250. [DOI] [PubMed] [Google Scholar]
  12. Dullin JP, Locker M, Robach M, Henningfeld KA, Parain K, Afelik S, Pieler T, Perron M. Ptf1a triggers GABAergic neuronal cell fates in the retina. BMC Dev Biol. 2007;7:110. doi: 10.1186/1471-213X-7-110. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Erlander MG, Tillakaratne NJ, Feldblum S, Patel N, Tobin AJ. Two genes encode distinct glutamate decarboxylases. Neuron. 1991;7:91–100. doi: 10.1016/0896-6273(91)90077-d. [DOI] [PubMed] [Google Scholar]
  14. Erzurumlu RS, Chen ZF, Jacquin MF. Molecular determinants of the face map development in the trigeminal brainstem. Anat Rec. 2006;288:121–34. doi: 10.1002/ar.a.20285. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Erzurumlu RS, Murakami Y, Rijli FM. Mapping the face in the somatosensory brainstem. Nat Rev Neurosci. 2010;11:252–263. doi: 10.1038/nrn2804. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Fremeau RT, Troyer MD, Pahner I, Nygaard GO, Tran CH, Reimer RJ, Bellocchio EE, Fortin D, Storm-Mathisen J, Edwards RH. The expression of vesicular glutamate transporters defines two classes of excitatory synapse. Neuron. 2001;31:247–260. doi: 10.1016/s0896-6273(01)00344-0. [DOI] [PubMed] [Google Scholar]
  17. Furuta T, Timofeeva E, Nakamura K, Okamato-Furuta K, Togo M, Kaneko T, Deschenes M. Inhibitory gating of vibrissal inputs in the brainstem. J Neurosci. 2008;28:1789–1797. doi: 10.1523/JNEUROSCI.4627-07.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Glasgow SM, Henke RM, Macdonald RJ, Wright CV, Johnson JE. Ptf1a determines GABAergic over glutamatergic neuronal cell fate in the spinal cord dorsal horn. Development. 2005;132:5461–5469. doi: 10.1242/dev.02167. [DOI] [PubMed] [Google Scholar]
  19. Gross MK, Dottori M, Goulding M. Lbx1 specifies somatosensory association interneurons in the dorsal spinal cord. Neuron. 2002;34:535–549. doi: 10.1016/s0896-6273(02)00690-6. [DOI] [PubMed] [Google Scholar]
  20. Haring JH, Henderson TA, Jacquin MF. Principalis- or parabrachial-projecting spinal trigeminal neurons do not stain for GAD or GABA. Somatosens Mot Res. 1990;7:391–398. doi: 10.3109/08990229009144715. [DOI] [PubMed] [Google Scholar]
  21. Henderson TA, Woolsey TA, Jacquin MF. Infraorbital nerve blockade from birth does not disrupt central trigeminal pattern formation in the rat. Dev Brain Res. 1992;66:146–152. doi: 10.1016/0165-3806(92)90152-m. [DOI] [PubMed] [Google Scholar]
  22. Henderson TA, Jacquin MF. What makes subcortical barrels? Requisite trigeminal circuitry and developmental mechanisms. In: Jones EG, Diamond IT, editors. Barrel Cortex. Vol. 11. Cerebral Cortex; Plenum, New York: 1995. pp. 123–187. [Google Scholar]
  23. Hevner RF, Hodge RD, Daza RA, Englund C. Transcription factors in glutamatergic neurogenesis: conserved programs in neocortex, cerebellum, and adult hippocampus. Neurosci Res. 2006;55:223–233. doi: 10.1016/j.neures.2006.03.004. [DOI] [PubMed] [Google Scholar]
  24. Hoshino M, Nakamura S, Mori K, Kawauchi T, Terao M, Nishimura YV, Fukuda A, Fuse T, Matsuo N, Sone M, Watanabe M, Bito H, Terashima T, Wright CVE, Kawaguchi Y, Nakao K, Nabeshima Y. Ptf1a, a bHLH transcriptional gene, defines GABAergic neuronal fates in cerebellum. Neuron. 2005;47:201–213. doi: 10.1016/j.neuron.2005.06.007. [DOI] [PubMed] [Google Scholar]
  25. Jacquin MF, Arends JJA, Xiang C, Shapiro L, Ribak CE, Chen ZF. In Drg11 knockout mice, trigeminal cell death is extensive and does not account for failed brainstem patterning. J Neurosci. 2008;28:3577–3585. doi: 10.1523/JNEUROSCI.4203-07.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kala K, Haugas M, Lilleväli K, Guimera J, Wurst W, Salminen M, Partanen J. Gata2 is a tissue-specific post-mitotic selector gene for midbrain GABAergic neurons. Development. 2009;136:253–262. doi: 10.1242/dev.029900. [DOI] [PubMed] [Google Scholar]
  27. Kaneko T, Itoh K, Shigemoto R, Mizuno N. Glutaminase-like immunoreactivity in the lower brainstem and cerebellum of the adult rat. Neurosci. 1987;32:79–98. doi: 10.1016/0306-4522(89)90109-7. [DOI] [PubMed] [Google Scholar]
  28. Kerr RC, Maxwell DJ, Todd AJ. GluR1 and GluR2/3 subunits of the AMPA-type glutamate receptor are associated with particular types of neurons in laminae I-III of the spinal dorsal horn of the rat. Eur J Neurosci. 1998;10:324–33. doi: 10.1046/j.1460-9568.1998.00048.x. [DOI] [PubMed] [Google Scholar]
  29. Killackey HP, Fleming K. The role of the principal sensory nucleus in central trigeminal pattern formation. Brain Res. 1985;354:141–145. doi: 10.1016/0165-3806(85)90077-x. [DOI] [PubMed] [Google Scholar]
  30. Magnusson KR, Clements JR, Larson AA, Madl JE, Beitz AJ. Localization of glutamate in trigeminothalamic projection neurons: a combined retrograde transport-immunohistochemical study. Somatosensory Res. 1987;4:177–90. doi: 10.3109/07367228709144605. [DOI] [PubMed] [Google Scholar]
  31. Minnery BS, Bruno RM, Simons DJ. Response transformation and receptive field synthesis in the lemniscal trigeminothalamic circuit. J Neurophysiol. 2003;90:1556–1570. doi: 10.1152/jn.00111.2003. [DOI] [PubMed] [Google Scholar]
  32. Mizuguchi R, Kriks S, Cordes R, Gossler A, Ma Q, Goulding M. Ascl1 and Gsh1/2 control inhibitory and excitatory cell fate in spinal sensory interneurons. Nat Neurosci. 2006;9:770–778. doi: 10.1038/nn1706. [DOI] [PubMed] [Google Scholar]
  33. Mugnaini E, Oertel WH. An atlas of the distribution of GABA-ergic neurons and terminals in the rat CNS as revealed by GAD immuno-histochemistry. In: Bjorklund A, Hokfelt T, editors. Handbook of Chemical Neuroanatomy. Elservier; Amsterdam: 1985. pp. 436–608. [Google Scholar]
  34. 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. doi: 10.1242/dev.02717. [DOI] [PubMed] [Google Scholar]
  35. Puelles E, Acampora D, Gogoi R, Tuorto F, Papalia A, Guillemot F, Ang SL, Simeone A. Otx2 controls identity and fate of glutamatergic progenitors of the thalamus by repressing GABAergic differentiation. J Neurosci. 2006;26:5955–5964. doi: 10.1523/JNEUROSCI.1097-06.2006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Schuurmans C, Armant O, Nieto M, Stenman JM, Britz O, Klenin N, Brown C, Langevin LM, Seibt J, Tang H, Cunningham JM, Dyck R, Walsh C, Campbell K, Polleux F, Guillemot F. Sequential phases of cortical specification involve Neurogenin-dependent and -independent pathways. EMBO J. 2004;23:2892–902. doi: 10.1038/sj.emboj.7600278. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Tsunekawa N, Yanagawa Y, Obata K. Development of GABAergic neurons from the ventricular zone in the superior colliculus of the mouse. Neurosci Res. 2005;51:243–51. doi: 10.1016/j.neures.2004.11.011. [DOI] [PubMed] [Google Scholar]
  38. Woolsey TA. Peripheral alteration and somatosensory development. In: Coleman JR, editor. Development of Sensory Systems in Mammals. Wiley & Sons; New York: 1990. pp. 461–516. [Google Scholar]
  39. Xiang C, Zhang KH, Yin J, Arends JJA, Erzurumlu RS, Jacquin MF, Chen ZF. The transcription factor, Lmx1b, is necessary for the development of the principal trigeminal nucleus-based lemniscal pathway. Mol Cell Neurosci. 2010;44:394–403. doi: 10.1016/j.mcn.2010.05.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Xu Y, Lopes C, Qian Y, Liu Y, Cheng L, Goulding M, Turner EE, Lima D, Ma Q. Tlx1 and Tlx3 coordinate specification of dorsal horn pain-modulatory peptidergic neurons. J Neurosci. 2008;28:4037–4046. doi: 10.1523/JNEUROSCI.4126-07.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES