SEARCH

SEARCH BY CITATION

Keywords:

  • Gbx1;
  • Gbx2;
  • Lbx1;
  • homeodomain transcription factor;
  • GABA;
  • spinal cord;
  • dorsal horn;
  • mouse

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The dorsal spinal cord processes somatosensory information and relays it to higher brain centers and to motoneurons in the ventral spinal horn. These functions reside in a large number of distinct sensory interneurons that are organized in specific laminae within the dorsal spinal horn. Homeodomain and bHLH transcription factors can control the development of neuronal cell types in the dorsal horn. Here, we demonstrate that the murine homeodomain transcription factor Gbx1 is expressed specifically in a subset of Lbx1+ (class B) neurons in the dorsal horn. Expression of Gbx1 in the dorsal spinal cord depends on Lbx1 function. Immunohistological analyses revealed that Gbx1 identifies a distinct population of late-born, Lhx1/5+, Pax2+ neurons. In the perinatal period as well as in the adult spinal cord, Gbx1 marks a subpopulation of GABAergic neurons. The expression of Gbx1 suggests that it controls development of a specific subset of GABAergic neurons in the dorsal horn of the spinal cord. Developmental Dynamics 234:767–771, 2005. © 2005 Wiley-Liss, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The dorsal spinal cord processes somatosensory information and relays it to motor neurons located in the ventral horn and to higher brain centers, like the brainstem, thalamus, and cerebellum. These functions reside in a large number of distinct interneuron types that are arranged in a highly organized laminar structure, the dorsal horn (Rexed, 1952; Brown, 1981). Such interneurons participate in spinal reflexes and are major targets of descending neuronal systems that can modulate incoming somatosensory information, for example pain stimuli (Gillespie and Walker, 2001; Julius and Basbaum, 2001). The functional architecture of the mature dorsal horn is the result of developmental processes that involve cell-type specification and differentiation as well as migration of neurons that were generated in the dorsal neural tube (Lee and Jessell, 1999; Caspary and Anderson, 2003).

Recent studies demonstrate that homeodomain transcription factors play a central role during development of neurons in the dorsal horn (for recent reviews see Goulding et al., 2002; Helms and Johnson, 2003). Sensory interneurons of the superficial dorsal horn are generated from neurons that are born during a late phase of neurogenesis (E11.5–E14.0 in mice) in the alar plate. At the time of their birth, these neurons are characterized by the expression of the homeodomain transcription factor Lbx1 and are designated as class B neurons. Correct specification of late-born class B neurons depends on Lbx1 function. In mice that carry a null mutation of the Lbx1 gene, the neurons that arise in the dorsal spinal cord assume aberrant molecular characteristics and are subsequently eliminated by apoptosis, resulting in a loss of the superficial dorsal horn (Gross et al., 2002; Muller et al., 2002). Late-born class B neurons comprise two neuron types (dILA and dILB neurons). The two neurons types can be distinguished by the expression of different sets of homeodomain transcription factors: dILB neurons express the homeodomain factors, Lmx1b and Tlx3 (Rnx), whereas dILA neurons are identified by the expression of Lhx1/5 and Pax2. These homeodomain factors control distinct steps during the further differentiation of these neuron types. Tlx3 and the closely related Tlx1 determine a glutamatergic cell fate in postmitotic dorsal neurons and in the absence of these genes the neurons take on an aberrant GABAergic fate (Cheng et al., 2004). Lmx1b and a further paired homeodomain factor, Drg11, are both essential for the differentiation of these neurons in the superficial dorsal horn. In addition, the innervation of the spinal cord by sensory afferents does not occur correctly in Tlx1/3, Lmx1b, and Drg11 mutant mice (Chen et al., 2001; Ding et al., 2004). Pax2 is required for GABAergic differentiation of dILA cells (Cheng et al., 2004). The molecular mechanisms that control subtype diversification and terminal differentiation of dorsal spinal interneurons are incompletely understood, and molecular markers that allow distinguishing the diverse neuronal subtypes in the dorsal horn are still limited.

Here, we show that the murine homeodomain factor Gbx1 is expressed specifically in late-born neurons that subsequently populate the superficial dorsal horn. Gbx1+ spinal neurons co-express Lbx1, Lhx1/5, and Pax2. Gbx1+ cells are not generated in Lbx1 mutant mice. Gbx1 expression identifies a specific subpopulation of GABAergic neurons in the dorsal spinal cord. The distinct expression of Gbx1 suggests a function in the maturation of a particular GABAergic neuronal subtype of the dorsal horn.

RESULTS AND DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

Identification and Isolation of Gbx1 cDNA

In order to identify genes differentially expressed in the dorsal spinal cord at E13.5, a microarray-based genome wide expression analysis was performed (Britsch and Birchmeier, unpublished data). Among the genes specifically expressed in the developing dorsal neural tube, we identified an EST sequence (Affymetrix-ID 116752_at, corresponding to GenBank AI464517). Computational analysis with CELERA DISCOVERY SYSTEM indicated that the EST mapped to the 3′-UTR of the murine Gbx1 gene. To isolate the complete Gbx1 cDNA, we screened a cDNA library generated from E13.5 mouse spinal cord with a probe derived from the EST. We isolated two independent cDNA clones. One contained the entire ORF of the murine Gbx1 gene including 5′ and 3′ untranslated sequences.

Expression Analysis of Gbx1 and Gbx2 During Spinal Cord Development

We determined the expression of the Gbx1 gene in the developing mouse spinal cord by in situ hybridization. Gbx1 transcripts were first detected at E11.5 in the ventricular zone of the spinal cord (Fig. 1A). At E12.5–13.5, Gbx1 was broadly expressed in the mantle zone of the dorsal spinal cord (Fig. 1B,C). With the appearance of a discernable dorsal horn around E14, Gbx1 expression became more restricted (Fig. 1D). Perinatally, Gbx1 expression was observed in a narrow layer in the superficial dorsal horn (Fig. 1F). Gbx1 expression was still detectable in a small number of neurons in the superficial dorsal horn in the adult (not shown).

thumbnail image

Figure 1. Expression analysis of Gbx1 and Gbx2 in the developing spinal cord. Vibratome sections from mouse cervical spinal cords at E11.5–E19.5 (AJ). Spinal cord tissues were hybridized with probes specific for Gbx1 (A–F) or Gbx2 (G–J). Note that both genes are transiently expressed in overlapping regions within the alar plate (B and H). ap, alar plate; vz, ventricular zone; mz, mantle zone.

Download figure to PowerPoint

Gbx2 is closely related to Gbx1 and has been reported to be expressed in the neural tube (Rhinn et al., 2004). Functional analyses have demonstrated an essential role of Gbx2 during development of the mid-hindbrain boundary (Wassarman et al., 1997; for review, see Joyner et al., 2000). As yet, a detailed analysis of the expression of Gbx2 in the spinal cord has not been reported. We compared expression of Gbx1 and Gbx2 during spinal cord development. At E11.5, Gbx2 expression was detected in the dorsal ventricular zone and in the mantle zone within the ventral half of the neural tube (Fig. 1G). Strong expression of Gbx2 was observed at E12.5 in the ventricular zone and the mantle zone of the entire dorsal spinal cord. At this developmental stage, the expression domains of Gbx1 and Gbx2 overlapped (Fig. 1B and H). After E12.5, Gbx2 expression was rapidly downregulated (Fig. 1I,J). Gbx1 and Gbx2 appear thus to be transiently co-expressed in neurons of the dorsal spinal cord. In addition to its dorsal expression, Gbx2 is also expressed in the ventral spinal cord. In this region, no Gbx1 co-expression was detectable. Ventral expression of Gbx2 was observed between E11.5 and E14.5. In this part of the spinal cord, Gbx2 may identify a distinct population of ventrally located spinal interneurons.

Characterization of Gbx1+ Neurons in the Dorsal Spinal Cord

Dorsal interneurons are generated during two distinct waves of neurogenesis in the spinal cord (Gross et al., 2002; Muller et al., 2002, 2005). During the second phase of neurogenesis (E11.5–14.0), the majority of neurons generated from dorsal progenitors are class B neurons that express the homeodomain transcription factor Lbx1. During further development, these late-born neurons colonize the superficial dorsal horn (Gross et al., 2002; Muller et al., 2002). In situ hybridization indicated that Gbx1 is expressed in the neurons generated during the second phase of neurogenesis. To define this neuronal population, anti-Gbx1 antibodies were generated. Immunohistological analyses of the spinal cord at E12.5 demonstrated that Gbx1 protein was detected in cells located in the alar plate, but not in the ventricular zone or in the ventral spinal cord. In addition, when Gbx1 in situ hybridization and Gbx1 immunohistology were carried out on identical sections, both signals were found to co-localize. Together, this indicates that the antibody specifically recognizes Gbx1 protein (compare Fig. 1B,H and Fig. 2A,B; and not shown). Further analysis revealed that Gbx1+ cells co-express Lbx1 at E12.5. Gbx1+ cells correspond thus to late-born class B neurons (Fig. 2A). Lbx1 is a key determinant for the specification of class B neurons (Gross et al., 2002; Muller et al., 2002). We, therefore, tested if Gbx1 expression is detectable in the dorsal spinal cord of Lbx1 mutant mice, and found that Gbx1 was absent at E12.5 (Fig. 2C). In Lbx1 mutants, miss-specified neurons that arise in the dorsal spinal cord are eliminated by apoptosis during further development (Gross et al., 2002; Muller et al., 2002). At E12.5 Gbx1+ wildtype spinal neurons co-express the homeodomain factor Lhx1/5 (Fig. 2B). In Lbx1 mutants, Lhx1/5+ neurons are still detectable in similar numbers in the dorsal spinal cord (Fig. 2D), indicating that the lack of Gbx1 expression is not caused by the apoptotic loss of dorsal neurons in Lbx1 mutants. We conclude that Gbx1 expression depends on Lbx1 function and that the Gbx1+ neuronal cell population is not generated in the dorsal horn of Lbx1 mutant animals.

thumbnail image

Figure 2. Lbx1 controls expression of Gbx1 in the dorsal spinal cord. Immunofluorescence analysis of the developing cervical spinal cord from wildtype (A, B) and Lbx1−/− (C, D) embryos at E12.5 using antibodies directed against Gbx1 (A–D), Lbx1 (A, C), and Lhx1/5 (B, D). Within the alar plate of control embryos, Gbx1+ cells co-express Lbx1 and Lhx1/5 (A, B). In homozygous mutants, expression of either Lbx1 or Gbx1 is no longer detectable, demonstrating that Gbx1 expression in dorsal spinal neurons depends on Lbx1 function (C, D). Lhx1/5 expressing cells are still present on E12.5 in homozygous mutants, indicating that the loss of Gbx1 expression is not caused by a depletion of the corresponding cells.

Download figure to PowerPoint

Late-born class B neurons comprise initially two neuron populations, dILA and dILB, which are born in an apparent salt and pepper pattern in the dorsal spinal cord. dILA neurons express Lbx1, Pax2, and Lhx1/5, whereas dILB cells express Lbx1, Lmx1b, and Tlx3 (Muller et al., 2002). At E12.5, only a subpopulation of the Lbx1+ cells co-expressed Gbx1 (Fig. 2A). We, therefore, characterized the Gbx1+ neuronal population further. At E12.5 and 14.5, Gbx1+ neurons co-express the transcription factors Lhx1/5 and Pax2, but are negative for Lmx1b and Tlx3. This indicates that Gbx1+ neurons correspond to the dILA neuronal subtype (Figs. 2B, 3A–D, and not shown). We observed at no developmental stage that all Pax2+ or Lhx1/5+ cells expressed Gbx1, and many Pax2+ or Lhx1/5+ cells that do not express Gbx1 were detectable. Thus, Gbx1 expression distinguishes a subpopulation of dILA cells (Figs. 2B, 3A,C, and not shown). We analyzed Gbx1-expressing cells also during the maturation of the dorsal spinal cord. At E16.5, most Gbx1+ cells were located in the superficial dorsal horn (Fig. 3E). At this stage, only few Gbx1+ neurons co-expressed Lbx1; however, most of the Gbx1+ cells co-expressed Pax2 or Lhx1/5 (Fig. 3E, and not shown). It has been shown previously that dILA neurons undergo GABAergic differentiation (Cheng et al., 2004). We have, therefore, tested if markers for GABAergic differentiation are expressed in Gbx1+ cells. When GABA+ or Gad67+ neurons are first detectable in the dorsal spinal cord, the GABA+ or Gad67+ neurons co-express Gbx1. Gbx1+ cells can thus differentiate into GABAergic neurons (Fig. 3F, and not shown). Recent studies demonstrated that Pax2 is essential for GABAergic differentiation in the dorsal spinal cord, since GABAergic differentiation does not occur correctly in Pax2 mutant mice (Cheng et al., 2004). Gain-of-function studies demonstrated that Pax2, when over-expressed in the chick spinal cord, is however not sufficient to induce GABAergic differentiation (Cheng et al., 2004). This suggests that additional factors might be involved in this process.

thumbnail image

Figure 3. Characterization of Gbx1 expressing dorsal spinal neurons. Co-expression analysis of Gbx1+ dorsal spinal neurons at E12.5 (A, B), E14.5 (C, D), E16.5 (E, F), and P28 (G, H). At E12.5–E14.5, Gbx1+ cells co-express Pax2 (A, C) but exclude expression of Lmx1b and Tlx3 (B, D) indicating that Gbx1 expressing dorsal neurons share molecular characteristics of dILA cells. Note that only part of Pax2+ cells co-express Gbx1. At E16.5, Gbx1+ neurons are detectable in the superficial dorsal horn and only few of them co-express Lbx1 (E). With the onset of GABA expression, both GABA and Gbx1 are co-expressed in the superficial dorsal horn (F). At postnatal day 28 (P28), Gbx1-expressing cells are distributed throughout the entire superficial dorsal horn (G), and most of the Gbx1+ cells co-express GABA (G, H).

Download figure to PowerPoint

In order to further determine the fate of Gbx1+ cells, we have analyzed Gbx1 expression in the early postnatal (P7) and in the adult spinal cord (Fig. 3G and H). Here, Gbx1 identifies a small population of interneurons that are scattered throughout the entire superficial dorsal horn. Very few Gbx1+ cells are located in the deeper layers of dorsal horn, close to the central canal (Fig. 3G and not shown). GABA expression was found to be abundant in the adult superficial dorsal horn. Within this region, the great majority of Gbx1+ neurons co-express GABA, further indicating that Gbx1 comprises a distinct subpopulation of GABAergic dorsal spinal neurons (Fig. 3G and H).

We have analyzed here the expression pattern of the homeodomain transcription factor Gbx1 during development of the mouse spinal cord. We show that Gbx1 expression is restricted to the dorsal spinal cord and that it coincides with the generation of a late neurogenic cell population. Immunohistological analyses identified Gbx1+ neurons as a subpopulation of dILA neurons that require the homeodomain transcription factor Lbx1 for development. As development proceeds, Gbx1+ interneurons differentiate into a subpopulation of GABAergic neurons and colonize the superficial layers of the dorsal horn. Physiological experiments have indicated that GABAergic neurons are not homogeneous, but GABAergic subpopulations are ill defined on a molecular level (Heinke et al., 2004). Further work is required to define the functional properties of the Gbx1+ population of GABAergic neurons, and to determine the role of Gbx1 in the development of this neuronal lineage.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

In Situ Hybridization

Whole mount in situ hybridization was performed as previously described (Britsch et al., 1998). In brief, spinal cords were micro-dissected from mouse embryos at E12.5–19.5 and fixed overnight with 4% PFA. After whole mount in situ hybridization, cervical spinal cords that were matched for their axial levels were embedded in 20% gelatin and post-fixed for several days. Thirty μm vibratome sections were examined on a Zeiss Axiophot microscope. For generation of Gbx1 specific riboprobes a SalI fragment of a plasmid that contained the entire Gbx1 cds was used. A plasmid for the generation of Gbx2 specific probes was obtained from A. Joyner (Skirball Institute, NY).

Generation of Anti-Gbx1 Antisera

A 741-bp fragment from the murine Gbx1 cDNA, corresponding to aa 61–308 was amplified by PCR with the following primers: Gbx1upper 5′-CATATGCAGAGAGCGGCAGGCGGCGGC and Gbx1lower 5′-GGATCCCCTGTGGCCACTGGTGTCCCCTCCTC. The PCR fragment was cloned into the bacterial expression vector pET-14b (Novagen), which provided coding sequences for a His6-tag. His6-Gbx1 was propagated in BL21(DE3)pLysS cells, affinity purified on TALON metal resin (BD Biosciences), and injected into rabbits and guinea pigs (Sequence Laboratories, Göttingen, Germany).

Immunohistology

For immunofluorescence staining, embryonic mouse tissue was fixed with 4% PFA in 0.1 M sodium phosphate buffer (pH 7.4). After cryoprotection tissue was embedded into OCT compound (Sakura) and 12 μm cryosections were made from matched cervical parts (C5/6) of spinal cords on a Microm cryostat. Stained sections were examined on a confocal microscope (Pascal, Zeiss). The following antibodies were used: rabbit and guinea pig anti-Gbx1, rabbit and guinea pig anti-Lbx1 (Muller et al., 2002), rabbit anti-Tlx3 (Muller et al., 2005), guinea pig anti-Lmx1b (gift from T. Jessel, New York), rabbit anti-Pax2 (Chemicon, Temecula, CA), rabbit anti-GABA (Sigma, St. Louis, MO), rabbit anti-Gad67 (Sigma), mouse monoclonal anti-Lhx1/5 (Developmental Studies Hybridoma Bank, University of Iowa), and fluorophore-conjugated secondary antibodies (Dianova).

Animals

Heterozygous Lbx1 mutant mice were intercrossed to obtain homozygous mutant embryos. Genotyping was performed as previously described (Brohmann et al., 2000).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES

The authors thank Verena Sasse and Karin Gottschling for excellent technical assistance, Carmen Birchmeier, Thomas Müller, and Martin Sieber for critical reading of the manuscript and for helpful discussions. We gratefully acknowledge the following scientists for gifts of antibodies and plasmids: Alexandra Joyner (New York), Thomas Jessell (New York), and Thomas Müller (Berlin). This work has been supported by a grant from the DFG (SFB 665-A1) to S.B.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS AND DISCUSSION
  5. EXPERIMENTAL PROCEDURES
  6. Acknowledgements
  7. REFERENCES