SEARCH

SEARCH BY CITATION

Keywords:

  • Myf5;
  • MyoD;
  • neurotrophin-3;
  • motor neurons;
  • sensory neurons;
  • spinal cord;
  • brainstem;
  • skeletal muscle;
  • mouse embryo

Abstract

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

We examined the effects of a single injection of exogenous NT-3, administered at embryonic day (E) 13.5, on the survival of two populations of motor neurons and two populations of sensory neurons. Both wild-type and double knockout, Myf5−/−:MyoD−/−, mutant embryos were examined to determine the effects of the aforementioned neurotrophin on motor and sensory neuron survival in the presence and absence, respectively, of skeletal muscle. We found that, although NT-3 rescues select populations of motor neurons in the absence of muscles, there is a lack of increase in neuron survival when skeletal muscle is present. Additionally, NT-3 was found to rescue a select population of proprioceptive sensory neurons in the absence of target tissue, while, at times, exacerbating neuron cell death when target tissues are present. Lastly, we found that neurons in the spinal cord and brainstem show both a regional and functional specificity in their response to the administration of NT-3 in utero. Our results indicate the possibility that different pathways are involved in the survival of neurons during naturally occurring programmed cell death and during excessively occurring programmed cell death. Developmental Dynamics 236:1193–1202, 2007. © 2007 Wiley-Liss, Inc.


INTRODUCTION

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

It has been well established in the literature that several neurotrophic agents are capable of promoting the proliferation, elongation, and arborization of motor neurons (MNs) of the central nervous system (CNS), and proprioceptive sensory neurons of the peripheral nervous system (PNS). Furthermore, neurotrophins have proven to be critical for the survival of muscle-associated neurons (MANs) of the CNS and PNS during the developmental period of naturally occurring programmed cell death (NPCD), both in vitro and in vivo, in avian and mammalian species (Oppenheim et al.,1993). Several studies have demonstrated that, during development, neurotrophins are not only present, and exerting a local effect, in the spinal cord (SC), associated ganglia, and cranial regions, including the brainstem, but many of these factors are also present, and exerting an effect, within the target tissues, specifically in the skeletal muscles of mammalian and avian species, as well as in the somatic muscle of zebrafish (reviewed in Homma et al.,2003).

Previous studies using the same model as in this study, in which myogenesis does not occur thereby resulting in a complete absence of skeletal muscle (i.e., target tissue) due to null mutations of both the Myf5 and MyoD myogenic regulatory transcription factors (MRFs; Myf5−/−:MyoD−/−, denoted double mutants [DM]), noted a near complete abolishment of MNs in the SC (a 91% decrease in neuron numbers) by embryonic day (E) 14.5 (Kablar and Rudnicki,1999). The facial motor nucleus (FMN) in the brainstem of DMs also demonstrates a near complete absence of large neurons. Along with this, a marked decrease in the number of proprioceptive neurons, in conjunction with a large increase in the number of terminal deoxynucleotidyl transferase–mediated deoxyuridinetriphosphate nick end-labeling (TUNEL)-labeled apoptotic bodies, is observed in the SC-associated dorsal root ganglia (DRG) at E17.5. One might expect such an obliteration of numbers because the identity of a sensory neuron is defined by its central and, importantly, peripheral projections, which in turn play a fundamental role in functional modality, namely the control of motor behavior (Arber et al.,2000; Oakley et al.,2000). The loss of neurons in all regions, due to lack of target tissues, is far more severe than during NPCD; neuronal losses in both lumbar and thoracic regions peak 1 day sooner than during normal wild-type (WT) development (E13.5 in DM vs. E14.5 in WT), and losses in the thoracic region occur more rapidly (E13.5–E14.5 in DM vs. E13.5–17.5 in WT). This demonstrated neuronal loss can be likened to neurodegenerative diseases in humans in which there is an observed excessive programmed cell death (EPCD), leading to extreme losses of neurons in the brain and SC, which ultimately lead to partial and complete losses of function.

More recent studies have shown that, while myogenic precursor cells (MPCs) are present in the periphery of DM embryos, they do not contain neurotrophic factors, namely neurotrophin-3 (NT-3), brain-derived neurotrophic factor (BDNF), neurotrophin-4/5 (NT-4/5), and glial cell line-derived neurotrophic factor (GDNF), in the absence of myogenic specification (Kablar and Belliveau,2005). Interestingly, in a separate study, we showed that these same neurotrophins (with the exception of GDNF, which was not looked at) are present in the brachial and upper thoracic SC, and corresponding DRG, of DMs at E12.5 and E13.5, just before the peak of neuron cell death, indicating that their expression is independent of myogenesis (Stephens et al.,2005). However, unlike BDNF, the tropomysin receptor kinase (Trk) B, TrkC, and Islet-1/2–positive staining neurons, which are completely absent in the SC of DMs at E17.5, and NT-3 and NT-4/5 remain in the existing cells of the SC at this later stage. Similarly, while BDNF and TrkC are completely absent and Islet-1/2 large proprioceptive cells are nearly completely absent in the DRG of DMs at E17.5, NT-3 and TrkB are clearly present in the remaining DRG cells of DMs at E17.5.

Strong evidence exists that the peripheral targets of sensory neurons can influence their phenotype during development. For example, a DRG that does not contain any proprioceptive afferents that is allowed to innervate limb muscle will produce many muscle spindle afferents that will, in turn, form synapses with the appropriate MNs (reviewed in Oakley et al.,2000). Limb bud ablation (i.e., lack of target tissue) and deafferentation experiments, using developing chicks, have shown a rescue of MANs through administration of exogenous neurotrophin (Houenou et al.,1994; Qin-Wei et al.,1994). Similar studies have also been carried out using adult mammals showing a rescue of almost all MNs in the L4 SC after axotomy of the sciatic nerve (Houenou et al.,1994; Li et al.,1994). However, a recent study published by our laboratory was the first to complete an in vivo investigation into the effects of a single neurotrophin, BDNF, on the survival of subpopulations of MANs during EPCD in developing mammals. The observed effects of exogenous BDNF, on the survival of motor and proprioceptive neurons, served as compelling evidence for the role of this neurotrophin in sustaining neuron life during both NPCD and EPCD in accordance with the neurotrophic hypothesis which states that the survival of a neuron during this sensitive period is directly related to the ability of that neuron to assimilate target secreted trophic support (Geddes et al.,2006). In the current study, we will address a similar scenario but, instead of BDNF, we will look at the effects of NT-3 on MAN survival during both NPCD and EPCD.

Neurotrophin-3 was selected for this study as it is a neurotrophin that is naturally expressed in skeletal muscle. It is known to act as a chemoattractant in axon guidance and aid in the proliferation and survival of neuron precursors and axon collateral branching in target fields. Neurotrophin, and its respective receptors, knockout studies have led to a common belief that NT-3 is a crucial component in promoting the survival of proprioceptive afferents (Barbacid,1994; Ernfors et al.,1994a; Snider,1994; Tessarollo et al.,1994; Farinas et al.,1996; Snider and Silos-Santiago,1996; Liu and Jaenisch,2000; Huang and Reichardt,2001). Furthermore, the induction of NT-3 secretion in NT-3−/− mice serves to rescue approximately 20% of the lost proprioceptive neurons (Wright et al.,1997). Several in vitro studies have also confirmed the effects of NT-3 on proprioceptive neuron survival (reviewed in Oakley et al.,2000). It is likely that NT-3 also plays a major role in MN survival, although, due to functional redundancy in the requirements of MNs of the CNS, the effect is not easily observed in neurotrophin knockout studies (Farinas et al.,1994).

Neurotrophin-3, like other neurotrophins of the same family, binds to the low affinity p75NTR receptor protein (where NTR = neurotrophin receptor; Yan et al.,1993). The high-affinity receptor most commonly associated with NT-3 is the tyrosine kinase TrkC receptor, although NT-3 is known to interact with both TrkA and TrkB in certain cell culture systems (Barbacid,1994; Snider,1994). Supporting this in vitro data is the observation that the phenotype of TrkC mutants is less severe than that of NT-3 mutant mice (Farinas et al.,1994), indicating alternate pathways for NT-3 action. It should be noted that recent evidence suggests that it may no longer be accurate to use the terms high- and low-affinity receptors due to the possibility that the two receptor types interact in forming a tertiary complex (Barker,2007). TrkC is the only Trk receptor expressed in the large proprioceptive neurons of the DRG by E9.5 and is expressed throughout the CNS by E11.5 (Barbacid,1994) and, along with TrkB, is expressed in SC MNs by E13.5 (Yan et al.,1993). Importantly, retrograde labeling has enabled investigators to confirm that both adult and developing MNs transport NT-3 from the target muscle to the cell body of respective neurons, in vivo (Yan et al.,1993, Chalazonitis et al.,2001). The results of several studies suggest that both neurotrophin and Trk receptor are transported, by means of endocytosis, to the cell body and that chemical signals may in fact be exerting an effect before the neurotrophin (or NT/receptor complex) arrives to the cell body. Furthermore, neurotrophins, including NT-3, are capable of passing through the blood–brain barrier after intravenous bolus injection to the jugular vein of mice (Pan et al.,1998).

Relatively recent studies in the chick have demonstrated the ability of exogenous NT-3 to promote the survival and differentiation of motor, and particularly proprioceptive, neurons during the developmental period of NPCD (Oppenheim et al.,1993; Oakley et al.,1995). Others have demonstrated that application of exogenous NT-3 serves to rescue both SC and cranial motor and proprioceptive neurons from axotomy-induced and deafferentation cell death in developing chicks and in adult mice (Yan et al.,1993; Li et al.,1994). In the current study, we examine the ability of NT-3 to rescue neurons and increase neuron numbers during EPCD and NPCD, respectively. It is our goal to determine the extent to which NT-3 will have an effect on MAN survival in the SC, associated sensory ganglia, and brainstem when injected into the amniotic cavity of an E13.5 embryo in coordination with the peak of declining neuron numbers.

RESULTS

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

NT-3 Rescues Select Populations of MNs from EPCD but not NPCD

To determine whether MNs in the brainstem and SC respond to NT-3, during NPCD and EPCD, we examined MN cell numbers in the lumbar and thoracic regions of the SC lateral motor column (LMC) and medial motor column (MMC), respectively, as well as in the facial motor nucleus (FMN) of the brainstem of E17.5 mouse embryos subjected to prior injection of either a sham (saline) or neurotrophin (NT-3) treatment at E13.5 (as described in the Experimental Procedures section). As discovered in previous studies, DM sham-injected MN counts were significantly reduced when compared with WT sham-injected numbers (Table 1; Kablar and Rudnicki,1999; Geddes et al.,2006).

Table 1. Cell Counts of Motor and Sensory Neurons in WT and DM Embryos Treated With Saline (Control) or NT-3
 WT salineWT NT-3DM salineDM NT-3
Mean% ControlMean% ControlMean% ControlMean% Control
  • Note: Mean values represent mean neuron cell number (μ) ± standard deviation. The % Control values represent μtreatmentsaline × 100% ± [(SD of μ) / μsaline] × 100%. Asterisks indicate a statistically significant difference from control group (**P < 0.001; *P < 0.05, multiple analysis of variance with Tukey's honest significance difference and least significant difference post hoc tests); n = 3 except where indicated with either a (where n = 2) or b (where n = 4). The numbers that round to zero are shown with two decimals. WT, wild-type; DM, double mutant; NT-3, neurotrophin-3; MMC, median motor column; LMC, lateral motor column; FMN, facial motor nucleus; TDRG, thoracic dorsal root ganglion; LDRG, lumbar dorsal root ganglion; Me5, mesencephalic nucleus.

  • ++

    Although the MMC and TDRG counts show no significant difference between the saline-treated and NT-3–treated populations, there is no significant difference to be found when comparing the WT saline- with the DM NT-3–treated embryos. We interpret this finding, particularly as it pertains to the TDRG population, to indicate that, in the diseased or excessive cell death (ECD) state, treatment of NT-3 during development serves to bring these types of cells back to a normal baseline count (as represented by WT saline; see Fig. 1).

  • ∧∧

    The results observed in the WT population indicate that NT-3 may in fact have a neurotoxic effect on the neuron cells in the Me5; these counts represent proprioceptive neuron cell bodies located in the brainstem. Of interest, the TDRG population of proprioceptive neurons, also a neural crest derivative, showed the strongest response to NT-3 during both excessive cell death and naturally occurring cell death (see Fig. 1).

LMC14 ± 1100 ± 517 ± 3119 ± 210.12 ± 0.10100 ± 890.92 ± 0.23**786 ± 198
MMC6 ± 2100 ± 329 ± 2137 ± 300.03 ± 0.06100 ± 1730.40 ± 0.26++1200 ± 794
FMN2526 ± 691100 ± 273934 ± 1002b156 ± 40308 ± 114100 ± 371022 ± 187*a332 ± 61
LDRG463 ± 87100 ± 19442 ± 3096 ± 6179 ± 88100 ± 49408 ± 452b228 ± 253
TDRG176 ± 51100 ± 29323 ± 11184 ± 6153 ± 26100 ± 17200 ± 104++b131 ± 68
Me5261 ± 61100 ± 23171 ± 14∧∧*66 ± 531 ± 5a100 ± 1572 ± 8*a230 ± 272
thumbnail image

Figure 1. Percentages of neuron increase and neuron rescue from naturally occurring (NPCD) and excessive (or diseased-like [EPCD]) programmed cell death (EPCD), respectively, after a single injection of neurotrophin-3 (NT-3) by means of amniotic sac puncture at embryonic day (E) 13.5. A: Percentage of muscle-associated neuron increase, where percentage increase is equal to (NT-3–injected WT/saline-injected WT) × 100%, of motor neurons (squares) and sensory neurons (circles) during NCPD indicates a cranial to caudal effect in motor neurons and a strong effect on the thoracic proprioceptive neuron population. B: Percentage of muscle-associated neuron rescue, where percentage rescue is equal to (NT-3–injected DM/saline-injected WT) × 100%, from EPCD of motor neurons (squares) and sensory neurons (circles), indicates a functional sensitivity to NT-3 that appears to have a stronger effect on sensory vs. motor neurons in the spinal cord, whereas the reverse is observed in the brainstem.

Download figure to PowerPoint

Setting the saline-injected embryos as 100% we were able to infer that both the LMC and FMN MNs respond to NT-3 during EPCD showing a significant difference in MN profiles per section, an increase of 686% and 232%, respectively, when compared with MN numbers in the control group at E17.5 (P ≤ 0.001 and P ≤ 0.05; Table 1). However, neurons in the MMC, although demonstrating the highest numeric increase when expressed in percentage (1,100%), did not prove to be significantly different from MN numbers in the appropriate (i.e., DM) control group (Table 1). On the other hand, although MMC counts showed no significant difference between the saline-treated and NT-3–treated neuron numbers in either group (WT and/or DM), there was also no significant difference found when comparing the DM NT-3–treated counts (0.4 ± 0.26) with the WT saline-treatment counts (6 ± 2; P > 0.05; Fig. 1B).

Although showing a tendency to be rescued from NPCD by increasing amounts moving in a caudal to cranial direction from LMC (19%) to MMC (37%) and finally FMN (56%), analysis of WT MN populations indicated no significant difference between neurotrophin- and sham-injected counts (Table 1).

NT-3 Rescues a Discriminative Population of Proprioceptive Sensory Neurons From EPCD While Exacerbating NPCD

To determine the extent to which proprioceptive sensory neurons are responsive to NT-3 during both NPCD and EPCD, two SC populations (lumbar DRG [LDRG] and thoracic DRG [TDRG]), as well as one brainstem population (the mesencephalic nucleus [Me5]), were exposed by means of serial sectioning through the appropriate regions at E17.5 (as described in Experimental Procedures section). Similar to the observations made regarding MN populations, DM, sham-injected, proprioceptive neuron counts were found to be significantly lower than WT sham-injected numbers as described in previous studies on untreated DMs (Kablar and Rudnicki,1999).

Accurate comparisons of proprioceptive neuron counts were obtained by again setting saline counts as 100%. Analysis of the three aforementioned cell populations revealed that, despite a tendency toward an increase in neuron numbers in WT TDRG (84%), DM LDRG (128%), and DM TDRG (31%), no significant difference was found between saline-treated and NT-3–treated neuron numbers in these locations. On the other hand, it is important to note that there was also no significant difference between the neuron counts in the T-DRG when comparing WT saline (176 ± 51) and DM NT-3 (200 ± 104) -injected embryos (P > 0.05; Table 1; Fig. 1B). In fact, the TDRG location proved to be unique in that the mean counts in the later group (DM NT-3) were found to be higher than the associated WT control.

Analysis of the Me5, however, indicated a significant increase in neuron numbers in the DM population when comparing the sham- and neurotrophin-injected groups (130%; P ≤ 0.05; Table 1). Alternatively, the Me5 counts in the WT population indicate a significant decrease in neuronal numbers when comparing the sham vs. the neurotrophin treatment (−34% of control; P ≤ 0.05; Table 1). The numbers obtained for LDRG (−4% of control) in the WT population, although not significant, indicate the same trend.

Neurons in the Spinal Cord and Brainstem Show Regional and Functional Specificity of Response to NT-3

As was the case in a recently published investigation in which we evaluated the effects of BDNF (Geddes et al.,2006), this study demonstrates novel findings in that no previous study has looked at the effects of NT-3 on all MANs in a single embryo. This approach has allowed us to determine whether MANs in distinct subpopulations respond differently to NT-3 for survival.

In all counts of SC MANs, when compared with WT-saline injected counts, the neurons in the thoracic region appear to respond to a greater degree than do the neurons in the lumbar region to the application of exogenous neurotrophin (WT LMC [18.71%] vs. WT MMC [37.21%], WT LDRG [−4.5%) vs. WT TDRG [83.9%], and DM LDRG [88.2%] vs. DM TDRG [113.2%]) with the exception of the DM MNs in which case the effect, shown as percentage rescue (PR), was found to be only slightly higher in the lumbar rather than thoracic region (DM LMC [6.3%] vs. DM MMC [6.2%]; Fig. 1A,B). However, when one considers significance at the P ≤ 0.05 level (where the DM MMC numbers are not significantly different from the WT MMC numbers), the reverse of this latter case is true. The same effect was observed when exogenous BDNF was applied in the same manner in a previous study by Geddes et al. (2006) in which there was a consistent greater effect on the numbers of neurons in the thoracic vs. lumbar region of the SC.

With regard to the DM counts, the response of both the MMC and TDRG neurons to NT-3 were found to be particularly interesting. For these two populations of neurons, no significant difference (P > 0.05) is observed when comparing DM NT-3–injected to WT control, indicating that these neuron subpopulations are restored to “normal” NPCD numbers in the presence of exogenous neurotrophin (Fig. 1B).

With respect to the head region, the MNs in the brainstem responded to a greater degree to NT-3 injection than did the sensory neurons in both WT (FMN [55.7%] vs. Me5 [−34.3%]) and, to a lesser degree, DM (FMN [40.5%] vs. Me5 [27.6%]) embryos. The response in numeric value of MN populations in both WT and DM to exogenous NT-3 relative to WT saline demonstrates a cranial-to-caudal gradient of response in which the FMN was found to show the highest level of percentage increase (PI) and PR (Fig. 1A,B). This cranial-to-caudal response of MNs to exogenous application of neurotrophin was also observed in response to BDNF administration (Geddes et al.,2006), where the greatest effect was found in the FMN and the least effect in the LMC in both the WT and DM populations.

Conversely, relative to WT saline counts, the proprioceptive neurons of the Me5 were found to show the lowest percentage value of response to the application of exogenous NT-3 in both WT (−34.3%) and DM (27.6%) counts. The sensory neurons showed the greatest response to NT-3 in the thoracic region, followed by the neurons in the lumbar region (as did the MN populations in DMs).

The ability of NT-3 to increase neuron numbers appears to reveal a regional sensitivity of MN response to exogenous NT-3 (decreasing in a cranial to caudal manner) along with variable responses of sensory neurons to NT-3 (two decreases and an increase in neuron numbers relative to control, with thoracic counts demonstrating the only increase). Therefore, the ability of NT-3 to increase neuronal numbers during NPCD can be summarized as follows: TDRG (83.9%) > FMN (55.7%) > MMC (37.2%) > LMC (18.7%) > LDRG (−4.5%) > Me5 (−34.3%).

The ability of NT-3 to rescue neuron numbers relative to WT saline reveals a functional sensitivity to NT-3 that appears to have a stronger effect on sensory than on MNs in the SC, whereas the reverse is observed in the brainstem. Therefore, the ability of NT-3 to rescue neuronal numbers during EPCD can be summarized as follows: TDRG (113.2%*) > MMC (6.2%*) > LDRG (88.2%) > FMN (40.5%) > Me5 (27.6%) > LMC (6.3%) (where * = P ≤ 0.05).

DISCUSSION

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

We designed an experiment to determine the relationship between MAN survival and the presence of exogenous neurotrophin in both WT and amyogenic mouse embryos. A single injection of NT-3 was administered, by means of amniotic sac puncture, at E13.5, and tissues were harvested, processed, and analyzed at E17.5. The PR and PI for EPCD and NPCD, respectively, were assessed according to appropriate counting criteria and statistical analysis (described in Experimental Procedures section).

The observed PR and PI of neurons in the TDRG, 113.2% and 83.9%, respectively, and the PR of neurons in the LDRG, 88.2%, are similar to results obtained in previous mouse knockout and NT-3–deprivation experiments. A sensory neuron loss of 55% to 78% has been observed in NT-3−/− mice (Ernfors et al.,1994a; Farinas et al.,1994,1996; Snider,1994; Tessarollo et al.,1994; Snider and Silos-Santiago,1996; reviewed in Farinas,1999; Liu and Jaenisch,2000; Huang and Reichardt,2001) and mice of postnatal day (P0) who lack TrkC, the primary receptor for NT-3, demonstrate a 20% to 30% loss of total DRG sensory neurons (reviewed by Farinas,1999) with a 100% loss of Ia muscle afferents and their associated sense organ muscle spindles and Golgi tendon organs (Barbacid,1994; Ernfors et al.,1994a). It should be noted that large diameter proprioceptive neurons constitute ∼20% of total DRG sensory neurons and that a vast majority of these express TrkC (Oakley et al.,2000). Along with this, it has been noted that, although mice lacking NT-3 can survive to term, after they are born, most of them display reduced activity, severe movement defects of the limbs, abnormal limb postures, do not appear to ingest food, and die within 24 hr (Ernfors et al.,1994a; Farinas et al.,1994). The proprioceptive defects observed in NT-3 null mice are also noted to a slightly lesser degree in TrkC−/− mice (heterozygote mutants generally die by P21) and are known to present in humans as peripheral neuropathies that affect large diameter axons and, at times, are associated with profound ataxia due to a position sense loss along with abnormal movements of extremities (Barbacid,1994). As one might expect, based on the timing of expression, in the different regions, during development, the defects observed in TrkC null mice are exhibited in the early prenatal (E.13) period of development, whereas the proprioceptive defects are observed in NT-3 null mice display themselves at a later stage of prenatal and postnatal development.

The results obtained as the mean PR value for the neurons in the Me5 (27.6%) are similar to a study published by Oakley et al. (1995) in which the investigators removed NT-3 from the peripheral tissues of chicks during the period of NPCD with injection of anti–NT-3 on the contralateral sides revealing a DRG neuron loss of 28.9%. Injection of anti–NT-3 on the ipsilateral side revealed a sensory neuron loss of 55.6%. The authors of this study concluded that large-diameter muscle afferents require NT-3 from the peripheral tissues to survive NPCD in vivo. On the other hand, our observations regarding the Me5 do not concur with studies in which larger neuronal losses of approximately 57% were observed in the sensory population of the Me5 ganglia (Farinas et al.,1994; Fan et al.,2000) likely due to further functional redundancy in neurotrophin requirements.

However, the observation that the PI and PR of Me5 (and LDRG) neurons show differing trends is in agreement with a study in which the investigators treated chicks by the addition of NT-3 onto the chorio-allantoic membrane (CAM) and found a rescue of −1.25% during NPCD (indicating a deleterious effect on neuron survival during this sensitive period), whereas chicks subjected to deafferentation responded favorably to exogenous application of NT-3, showing a 55.7% rescue of proprioceptive L3 DRG neurons at E16 (Qin-Wei et al.,1994). On the other hand, a separate earlier study indicated a rescue of 19.6% in L3 DRG neuron numbers in CAM-treated chicks during NPCD between E6 and E9 (Oppenheim et al.,1993). The knowledge that NT-3−/− mice show a 60% reduction of sensory neurons in the DRG compared with a greater loss of 92% in NT-3/BDNF/NT-4/5 knockouts helps to explain these results. During an already sensitive period in which neurotrophins and their receptors are regulated both spatially and temporally, our injection of exogenous NT-3 proved to have deleterious effects on these populations of MANs during NPCD when observed in mice at E17.5. It is possible that an overflow of the system with exogenous NT-3, at this time, is sufficient to down-regulate phosphorylation of the respective pathways. It would, therefore, be interesting to investigate, or quantify, the expression levels of both Trk and p75NTR, and the phosphorylation of specific pathways (i.e., by means of Western blot of phosphorylated vs. nonphosphorylated molecules) to determine whether a correlation exists between receptor levels, phosphorylation, and neuronal survival in both neurotrophin-treated and control embryos.

As mentioned in the introduction, previous studies have demonstrated by means of TUNEL labeling that an increase in apoptosis occurs in DMs as a result of the absence of muscle targets (Kablar and Rudnicki,1999). An ideal follow-up study for the current experiment would be to compare apoptosis levels in the areas of interest in sham- vs. neurotrophin-injected embryos. This could be accomplished by using either TUNEL staining or caspase-3 antibody for immunohistochemistry.

Our recent findings and the findings of in vitro and knockout mouse experiments (Ernfors et al.,1994b; Jones et al.,1994) do not agree with the suggestion that limb proprioceptors depend solely on NT-3 or that other neurotrophins cannot prevent cell death of NT-3–deprived limb proprioceptive neurons (Fan et al.,2000). Our findings do agree with the conclusions that developmental regulation by neurotrophins may differ between proprioceptive neurons located in the brainstem (CNS) and dorsal root ganglia (PNS), even though neurons of both locations are of neural crest origin and mediate the same sensory modality (Fan et al.,2000). Along with the fact that there is a known differential in the dependence on neurotrophins in different anatomic regions containing the same cell types (i.e., FMN vs. MN of the SC and Me5 vs. DRG sensory neurons), another study showed that both overexpression of NT-3 and intramuscular, but not intraperitoneal, injections of NT-3 served to significantly reduce the percentage of L5 DRG neurons lost after nerve crush compared with wild-type (Wright et al.,2002).

The PI of MNs in the LMC (18.71%), MMC (37.21%), and FMN (55.6%) as well as the PR of the MNs in the FMN (40.5%) closely relate to some of the findings on the influence of NT-3 and its associated receptor on MN survival. Neurotrophin-3 knockout experiments have noted no loss of MNs in the LMC and MMC (Yan et al.,1993; Snider,1994; Huang and Reichardt,2001; Patel et al.,2003) and facial motor nucleus neuron loss is shown to be nonsignificant, i.e., no obvious deficits were detected, in the brainstem of newborn NT-3−/− mice (Farinas et al.,1994). Whereas certain in vitro studies have demonstrated inadequate promotion of chick motoneuron survival through the use of NT-3 and BDNF, the application of these same neurotrophic factors on cultured fetal rat motoneurons demonstrated a survival-promoting effect for these agents (reviewed in Li et al.,1994).

In the first injection studies of this type, where P12 mice that had been axotomized at P5 and treated at the site of axotomy with gel foam (soaked in 10 μl of NT-3), the investigators found that 102% of L4 MNs were counted relative to the unoperated contralateral side and a larger 183% relative to the saline-treated ipsilateral SC MN numbers (Houenou et al.,1994; Li et al.,1994). Dorsal root ganglion neuron numbers were not investigated for NT-3 treatment; however, NGF was found to increase neuron numbers by 40% in the L3 DRG of embryonic chicks and by 65% in fetal mice injected on E14 and examined on E18 (Houenou et al.,1994).

Mouse knockout experiments show that the elimination of multiple endogenous neurotrophins (e.g., NT-3/BDNF or NT-3/BDNF/NT-4/5), and of respective receptors (e.g., TrkC/TrkB), is necessary to see a near complete absence of Trigeminal Me5 (95%) and DRG (92%) sensory neurons (Fan et al.,2000; Liu and Jaenish,2000; Huang and Reichardt,2001). This effect is observed to a lesser degree in the case of facial (22%) and SC (20%) MNs (Liu and Jaenish,2000), indicating the requirement for additional factors, proteins, or trophic agents, other than the listed neurotrophins, for MN survival to occur.

As others have previously suggested, our findings indicate that some populations of neurons require several different neurotrophins at a particular stage of their development whereas others have been shown to switch their trophic dependence, and/or Trk expression, during development in vitro (Enokido et al.,1999) and in vivo (Farinas,1999). For example, the muscles of mastication, but not of limbs, contained spindles in newborn and adolescent NT-3 and TrkC null mutant mice, suggesting that, at least in the muscles of mastication, the muscle secreted NT-3, or associated TrkC, may serve in functions other than those related to spindle assembly and that proprioceptive neurons innervating the jaw muscles, as suggested above, are dependent on factors other than NT-3 for maintenance and survival (Kucera et al.,1998).

Our current findings support the idea that there may be subpopulations of MANs that are not target-dependent, or that some MANs require two or more trophic agents, or finally that some of the trophic agents required for survival might not be secreted by the target tissues at all and may in fact be derived from other supporting tissues. Further investigation is required to determine whether NT-3, or any of the other neurotrophins, is of unique importance in preventing the death of developing motor and sensory neurons and at what developmental stage this dependence occurs. Our laboratory is currently following-up our series of single neurotrophin-injection experiments (BDNF, NT-3, and GDNF) by injecting developing embryos with combinations of neurotrophins (e.g., GDNF/NT-3 and GDNF/NT-3/BDNF).

EXPERIMENTAL PROCEDURES

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

Interbreeding of Embryos

Mouse embryos lacking both Myf5 and MyoD (Myf5−/−:MyoD−/−, double mutant, or amyogenic) were derived by a two-generation breeding scheme (Rudnicki et al.,1993). In brief, Myf5+/−:MyoD+/− mice were generated by breeding MyoD−/− knockout mice to Myf5+/− knockout mice. Myf5+/−:MyoD+/− were then interbred, resulting in nine different possible genotypes. The DM embryos occur at a probability of 1 in 16. When interbreeding mice (overnight) resulted in a vaginal plug, the morning of plug observation was assigned E0.5.

Injecting and Harvesting Embryos

In accordance with approved animal care protocols and institutional guidelines, embryos were injected in utero at E13.5 with 10μl of NT-3 (Regeneron) in saline (1 μg/ml). Pregnant females were anesthetized using isoflurane gas in oxygen. Under aseptic conditions, a partial laparotomy was used to expose the horns of the uterus. Saline was used ad lib to prevent the drying of tissues. A modified Hamilton syringe was used to inject 10μl of fluid into the amniotic fluid, embryos were replaced, and the abdomen was sutured to allow pregnant females to recover. Hamilton syringe tips were modified to an outer diameter of 0.15mm, allowing for a higher survival rate (60 to 70%) than usual when injected at E13.5 (summarized in Geddes et al.,2006). A survival rate of 6% made injections at E12.5 an impossibility, considering the already low probability of obtaining a DM embryo.

At E17.5, pregnant females were euthanized by cervical dislocation and embryos were collected by means of cesarean section. DM embryos were identified by physical characteristics: the absence of ribs, the absence of reflexive withdrawal of limb when extended with forceps, and the inability to have head tilted backward (due to fused cervical vertebrae; Rot-Nikcevic et al.,2006). Polymerase chain reaction was used to verify genotype of observed phenotype using isolated DNA from embryo tails.

Histological Processing

Harvested embryos were fixed in 4% paraformaldehyde overnight, rinsed in triphosphate buffered saline, and dehydrated using a series of ethanol dilutions increasing in strength. After further dehydration, followed by xylene-to-paraffin infiltration, embryos were embedded in blocks of paraffin wax.

Tissue Staining and Immunohistochemistry

Transverse and coronal serial sections were cut through the lumbar and thoracic body wall and the region of the cranium containing the brainstem, respectively, using a rotary microtome. Sections in the lumbar and thoracic regions were cut at 4 μ for immunohistochemistry staining of the lumbar (LMC) and thoracic (MMC) SC motor column neurons and DRG of the lumbar (L3–L4) and thoracic (T4–T6) level (previously outlined in Geddes et al.,2006). The collections of the sections were made several segments earlier than necessary to enable accurate selection of the appropriate region for counting. These sections were treated with mouse monoclonal anti–Islet-1/2 antibody 39.4D5 (1:4, Developmental Studies Hybridoma Bank) or rabbit polyclonal anti–tyrosine kinase C TrkC 8800 antibody (1:50, Santa Cruz). Hematoxylin was used for counterstaining. Sections of the brainstem were cut serially along the coronal plane at 40μl for cresyl violet staining (Nissl).

Cell Counting

Morphometric analysis of SC MNs and DRG proprioceptive neurons were performed as described before (Kablar and Rudnicki,1999) on transverse sections through the lumbar and thoracic regions. Motor neurons were counted on every ninth and tenth section at the lumbar level, where the LMC is bifurcated into a dorsal and ventral column (Lin and Carpenter,2003), and at the level of the thoracic SC, where only the MMC neurons and not LMC neurons are found. An average of the counts on the two sections was calculated. Mature and immature MNs could be identified and counted with the staining technique we used (Fig. 2D; reviewed by Geddes et al.,2006). Motor neuron profiles were identified by the characteristic MN morphology (nucleus with diameter >10 μm, large soma) and/or positive against Islet-1/2 antibody (Fig. 2D). The total numbers of MN profiles per motor column, per section, were counted for MMC and LMC.

thumbnail image

Figure 2. Cell count criteria. A: In the mesencephalic nucleus, only neurons with a visibly large soma with darker cresyl violet stain were counted (asterisk in Ai). B: In the facial motor nucleus, only cresyl violet-stained motor neurons with dark somas and large, pale nuclei with an intact nuclear membrane and visible nucleolus were counted (asterisk in inset). C: In the dorsal root ganglion, only the large proprioceptive neuron profiles with large soma (area > 190 μm2) with/without tyrosine kinase C (TrkC) receptor labeling and clear hematoxylin-stained nucleus were counted. D: In the spinal cord, motor neuron profiles with a clear nucleus (diameter > 10 μm) and large soma and/or positive against Islet-1/2 antibody were counted. Note asterisks in Di. Mature motor neuron negative for Islet-1/2 in Dii. Mature motor neuron positive for Islet-1/2 in Diii. Immature motor neuron positive for Islet-1/2. Original magnification, ×100. Scale bar = 50 μm in A–D.

Download figure to PowerPoint

DRG neuron counts were made on associated slides. Therefore, DRG tissue was collected at the same time as motor column sections cut at 4 μm. Immunohistochemistry, with TrkC as primary antibody, was used to label large neurons in the DRG. Large proprioceptive neurons were counted using the following criteria: large soma (area > 190 μm2), evident nucleus with intact membrane, and clear nucleolus and/or positive signal against TrkC, allowing reliable estimates of cell numbers (Fig. 2C; reviewed in Geddes et al.,2006). An estimate of the number of proprioceptive cells in each LDRG and TDRG was obtained by multiplying the average number of profiles per section by the length of the DRG as previously described (Geddes et al.,2006). The total numbers of proprioceptive neuron profiles per DRG were counted for TDRG and LDRG.

An unbiased stereological technique, the optical fractionator, was used to count the number of MNs in the FMN. Neurons were identified by their large, pale nuclei, with intact nuclear membrane and visible nucleolus (reviewed in Geddes et al.,2006; Fig. 2B). A systematic random sampling of serial sections, spanning the entire length of the nucleus, was used to obtain an estimate of the total number of MNs.

As the boundaries of Me5 are less easily defined, the optical dissector was used to count the number of large proprioceptive neurons in Me5. Only neurons showing a large, dark cresyl violet soma and located within the boundaries of the nucleus were counted in serial sections spanning the entire length of both the right and the left Me5, as previously reviewed (Fig. 2A; Geddes et al.,2006). The total number of neurons per brainstem Me5 nuclei was counted.

All stereological analysis were completed using an Olympus BMax microscope, using a ×100 magnification lens (Olympus, PlanAPO, 1.4 numerical aperture), a Microfire camera (Optronics), and a Prior (Optiscan) mechanical stage. Optical fractionation was done using Stereo Investigator 6 (MicroBrighField, Inc.) 6.55.1 (2005) Image Library Version 5.65. Cell measurements were made and images acquired using AxioVision version 4.4.1.0 (Zeiss).

Statistical Analysis

The data of neuron counts are expressed as mean ± standard deviation. The PI during NPCD (NT-3–injected WT/saline-injected WT × 100%) and the PR from EPCD (NT-3–injected DM/saline-injected WT × 100%), due to treatment, were calculated for each population. Comparisons between treated groups within and between WT and DM populations were made on transformed data (to logarithmic scale) using multiple analysis of variance followed by Tukey's honest significance difference and least significant difference post hoc to determine location of significance. Statistical significance was established at P ≤ 0.05. All calculations were performed using SPSS version 14.0 for Windows.

Acknowledgements

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

We thank Allison Geddes for her hard work, surgical assistance, and pioneering of staining and counting techniques in the region of the brainstem. We thank Dr. Paul Neumann for his help in statistical analysis. We also thank Anne C. Belliveau for technical assistance. B.K. received a Canadian Institutes of Health Research (CIHR) research grant and a contract for NT-3 supply from Regeneron Pharmaceuticals.

REFERENCES

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