Increased progenitor proliferation and apoptotic cell death in the sensory lineage of mice overexpressing N-myc
- First Online:
- Cite this article as:
- Kobayashi, M., Hjerling-Leffler, J. & Ernfors, P. Cell Tissue Res (2006) 323: 81. doi:10.1007/s00441-005-0011-5
- 86 Views
N-myc, a member of the myc family of bHLH transcription factors, is expressed mainly in the nervous system, including derivatives of neural crest cells in the periphery during development, such as the sensory dorsal root ganglion (DRG). Previous studies suggest that N-myc is involved in the proliferation of progenitor cells in the sensory lineage. To address the role of N-myc in the development of peripheral sensory neurons, we have overexpressed N-myc in sensory progenitor cells. The overexpression of N-myc did not significantly affect the number of multipotent neural crest cells or glial differentiation but caused a brief and marked increase of both proliferation and apoptosis in the DRG at embryonic day 11 (E11), thus coinciding with the stage of cell-cycle exit. At E17, the total number of cells in the lumbar DRG of mice with forced expression of N-myc was significantly reduced compared with that in wild-type mice. Among the different DRG subpopulations examined, the number of parvalbumin-positive neurons representing large-diameter proprioceptive neurons increased significantly. Our results indicate that forced expression of N-myc in the sensory lineage leads to unscheduled cell-cycle re-entry and excessive apoptosis and show that N-myc can affect the composition of different functional subtypes of sensory neurons in the DRG.
KeywordsApoptosisCell proliferationDorsal root ganglionN-mycSensory neuronsDevelopmentMouse, transgenic
N-myc is a neuronal member of the myc family of genes and encodes a bHLH/leucine zipper transcription factor, which regulates cell proliferation, differentiation, transformation, and apoptosis. N-myc is expressed in several tissues including the brain and peripheral nervous system from embryonic to postnatal stages, whereas in the adult mouse, most of the differentiated tissues express a low level or undetectable levels of N-myc (Zimmerman et al. 1986; Semsei et al. 1989). In newborn mouse brain, post-mitotic neuroblasts express high levels of N-myc mRNA (Mugrauer et al. 1988). N-myc-deficient mice die by embryonic day 11 (E11) and show abnormalities in the organs in which N-myc is normally expressed (Stanton et al. 1992; Charron et al. 1992; Sawai et al. 1993). Loss of N-myc selectively in the nervous system is compatible with life and results in microencephaly, with a two-fold absolute reduction in brain mass relative to body weight (Knoepfler et al. 2002). The reduction in brain mass is correlated with a reduction of cell numbers, which in turn coincides with the reduced proliferation of progenitor cells during early embryonic development. These findings have been taken to show that N-myc is essential for the expansion of the progenitor cell populations during neurogenesis in the brain (Knoepfler et al. 2002).
The dorsal root ganglion is derived from the neural crest, which condenses into the ganglia in the lumbar region at around E10. During the course of migration and differentiation, the neural crest cell is believed to become progressively restricted by committing, first, to the sensory lineage and, then, to different subtypes of sensory neurons. The various subtypes of somatosensory neurons transmit different modalities of sensation. These subtypes are distinguished by cell-soma size and the expression of different types of ion channels and neuropeptides and depend on different neurotrophic factors for their survival during development (for reviews, see Snider and Wright 1996; Bibel and Barde 2000).
A number of different transcription factors have been shown to be important for sensory neuron development. Neurogenins (NGN) 1 and 2 of the bHLH family of transcription factors participate in the decision between sensory and autonomic lineages by increasing the probability of adopting a sensory fate (Ma et al. 1999; for a review, see Anderson 1999). Brn-3a, a member of the POU domain transcription factors, is required for the expression of neurotrophin receptors, and in its absence, sensory neurons die as a consequence of the loss of neurotrophic factor receptor signaling (McEvilly et al. 1996; Huang et al. 1999). Runx1 and Runx3, members of the Runt domain transcription factors, are expressed in a different subpopulation of dorsal root ganglion neurons (Levanon et al. 2001; for a review, see Levanon and Groner 2004). Runx1 is expressed in small-diameter nociceptive neurons, whereas Runx3 is expressed in large-diameter proprioceptive neurons. Runx3-deficient mice display a loss of functional TrkC-expressing proprioceptive neurons and a failure of the establishment of the central and peripheral axonal projections of these neurons (Levanon et al. 2002; Inoue et al. 2002).
In addition to the above transcription factors, N-myc has been reported to be expressed in the migrating neural crest and the dorsal root ganglia. N-myc protein is translocated from the nucleus to the cytoplasm prior to cell-cycle exit between E10 and E11 in the mouse (Kato et al. 1991; Wakamatsu et al. 1993). Although relatively little is known of the in vivo role of N-myc in the developing dorsal root ganglion, it is clearly of pivotal importance. N-myc-deficient mice show a marked reduction in size of the dorsal root ganglion (Stanton et al. 1992; Charron et al. 1992; Sawai et al. 1993), presumably as a consequence of reduced proliferation (Knoepfler et al. 2002). Studies in the chicken show that expression of N-myc in the migrating neural crest leads to more neurons in the ventral pathway in vivo, and as a result, more neural crest cells adopt a sympathetic neuron phenotype (Wakamatsu et al. 1997).
To clarify the function of N-myc in the developing dorsal root ganglion further, we have generated transgenic mice in which N-myc is overexpressed under the nestin promoter. Nestin is an intermediate filament protein that is expressed in most proliferating neuronal progenitor cells in the central and peripheral nervous systems (Dahlstrand et al. 1995). In the peripheral nervous system, endogenous nestin expression is observed in migrating neural crest cells, in the dorsal root ganglion (Dahlstrand et al. 1995), and in neural crest stem cells (Stemple and Anderson 1992). We have used the human nestin promoter, which drives expression as early as E7.5 in the neural plate, in progenitor cells of the central nervous system, and in early migrating neural crest cells at later stages (Lothian and Lendahl 1997). Using this promoter, we have analyzed the impact of N-myc overexpression on cell proliferation, apoptosis, and phenotypic differentiation in the developing sensory neuron lineage of the mouse.
Materials and methods
Construction of nes1852tk/N-myc transgenic mice
Generation of transgenic mice
The 9.5-kb HindIII-PacI fragment (Fig. 1a) was purified and injected into fertilized mouse eggs (CBA×C57Bl6). The injected eggs were subsequently transferred into pseudopregnant female mice. Transgenic embryos were identified by detecting amplicon corresponding to human N-myc DNA by using a real-time PCR system (GeneAmp 5700, ABI). Briefly, small pieces of tail or hind limb were excised from each embryo, and genomic DNA was extracted. The primer pair for human N-myc transgene was designed within the intronic sequence to avoid amplification from endogenous mouse genomic DNA (forward: 5′-TCT GGG CTG AGG AGA GAT CTA AA-3′, reverse: 5′-GCT TAT CTG TCT CTT CCG GCA G-3′).
Histochemistry and immunohistochemistry
Embryos were fixed in 4% paraformaldehyde (PFA) for 4 h to overnight (according to the size of the embryo), immersed in 20% sucrose, and frozen. Transverse sections (14 μm thick) were cut on a cryostat and used for all immunohistochemical analyses.
Sections were postfixed with ice-cold Webster (0.4% PFA and 2.5% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4) for 10 min and rinsed in ice-cold phosphate-buffered saline (PBS; 3×10 min). β-Galactosidase activity was visualized by incubating slides in 3.1 mM potassium ferricyanide, 3.1 mM potassium ferrocyanide, 1 mM MgCl2, and 0.4 mg/ml 5-bromo-4-chloro-3-indole-β-D-galactoside (X-gal) in 0.1 M phosphate buffer overnight at 37°C.
BrdU labeling was performed by intraperitoneal injection of 50 mg/kg BrdU (Sigma) into pregnant mice 4 h prior to collection of embryos. Sections were postfixed with 4% PFA for 15 min, rinsed three times in PBS, and incubated in 2 M HCl in 70% EtOH at -20°C for 15 min. Endogenous peroxidase activity was blocked in 2% hydrogen peroxide in PBS for 10 min at room temperature. The sections were rinsed twice in PBS, deproteinized in ice-cold 0.1 N HCl for 20 min, denatured in 2 N HCl in PBS at 37°C for 30 min, neutralized in 0.1 M borate buffer (pH 8.5) for 10 min at room temperature, and rinsed three times in PBS. They were incubated in TRIS-buffered saline (TBS)-0.1% Tween20 (TBST) containing 5% horse serum for 2 h at room temperature and incubated with mouse monoclonal anti BrdU antibody (1:1,000, Sigma) in TBST containing 5% horse serum overnight at 4°C. Immunostaining was visualized with the ABC immunoperoxidase kit (Vector Laboratories) and diaminobenzidine as substrate.
Apoptosis was detected by TUNEL method with the ApopTag Apoptosis Fluorescein In Situ Apoptosis Detection Kit (Serologicals, USA) according to the manufacturer's instructions.
Sections were post-fixed with 4% PFA for 20 min and rinsed three times in PBS. Following incubation with 4% bovine serum albumin (BSA) in PBS containing 10% donkey serum for 2 h at room temperature, sections were incubated with rabbit anti-parvalbumin antibody (1:500, Swant), guinea pig anti-calcitonin gene-related peptide (CGRP) antibody (1:500, Peninsula Laboratories), rabbit anti-glial fibrillary protein (GFAP) (1:400, Dako), mouse anti-βIII-tubulin (1:250, Promega), anti-Sox10 (Michael Wegner), or rabbit anti-N-Myc antibody (1:400, Santa Cruz Biotechnologies) in 2% BSA in PBS containing 0.03% Tween20 overnight at 4°C. After being rinsed five times in PBS, sections were incubated with species-specific and isotype-specific fluorescent antibodies (donkey Cy2-, Cy3-, or Cy5-conjugated anti-rabbit, anti-mouse, or anti-guinea pig IgG, diluted 1:100–1:400; Jackson) for 2 h at room temperature. Staining was visualized and photographed by using a Zeiss Axioplan 2 with either an LSM 510 confocal detector (Zeiss, Germany) or a color charge-coupled device camera (Jena Optik, Germany).
Series of every third section were stained as indicated above or with cresyl violet. Neuronal cell numbers were established by counting cells with a neuronal morphology, as identified by the Nissl staining, and with a clear nucleus with nucleoli. Neurons were also identified by the staining of βIII-tubulin. The total number of neurons in each ganglion was established by multiplying the sum of cell numbers from a series of sections by three. Data are presented as the mean±SEM of three or more animals, unless otherwise indicated.
Generation of N-myc overexpressing mice
We used the 1,852-bp human nestin second intron, which has extensively been studied previously and which possess transcriptional activity in cells of both the central and peripheral nervous systems (Lothian and Lendahl 1997). The human N-myc gene consisting of exon 2 and 3, a region containing all of the protein coding sequence, was inserted between the promoter and an IRES-βgeo cassette (Mountford et al. 1994; Fig. 1a). Transgenic mice were evaluated by measuring the copy number of the integrated human N-myc transgenes into the mouse genome by real-time PCR. Embryos that retained a high copy number (>20 copies/ng DNA) of human N-myc DNA were chosen for the study (nes1852tk/N-myc mice). Littermate animals carrying undetectable levels of human N-myc DNA were used as wild-type controls.
To confirm nestin promoter activity, performed X-gal staining on transverse cryosections of E14.5 mice. Wild-type mice showed no β-galactosidase activity (Fig. 1b), whereas nes1852tk/N-myc mice displayed a strong staining in the spinal cord, especially around the ventricular zone and in the outer edge of the marginal zone, and in some cells of the dorsal root ganglion (Fig. 1c). The pattern of X-gal staining in the nes1852tk/N-myc was similar to the nestin immunoreactivity previously described in the E15.5 mouse (Dahlstrand et al. 1995). Because N-myc can be localized to both the nucleus and the cytoplasm, we next immunohistochemically stained E11, E12, E14, and E17 dorsal root ganglia for N-myc. The cellular localization of N-myc in wild-type mice was consistent with previous studies with an almost exclusive cytoplasmic localization at E11; a similar localization was also seen at later stages. In contrast, overexpression of N-myc in the nes1852tk/N-myc mice led to a pronounced nuclear localization of N-myc at E11 (Fig. 1d, e), but this was largely lost between E11 and E12 (Fig. 1f, g). Because N-myc translocates from the nucleus to the cytosol between E10 and E11 during normal development (Wakamatsu et al. 1993), the overexpression thus results in forced nuclear localization for at least one extra day.
Increased cell proliferation in the early dorsal root ganglion of nes1852tk/N-myc mice
N-myc overexpression leads to increased apoptosis
Reduction of neuronal numbers by N-myc overexpression
E17 embryos were collected, and neuronal numbers were measured in order to determine the overall consequence of the increased proliferation and apoptosis on total neuronal cell number in the dorsal root ganglion. The number of neurons in the dorsal root ganglion was counted from lumbar level 1 to sacral level 1 in one embryo of each genotype. A marked reduction of neurons was found in the nes1852tk/N-myc mouse at all levels (Fig. 4a). Counts of neuronal numbers in the L4 dorsal root ganglion of several animals/genotype showed a significant reduction of 33% compared with wild-type mice (Fig. 4b).
Increased number of limb proprioceptive neurons in nes1852tk/N-myc mice
N-myc and the control of neuronal birth numbers in the sensory lineage
In the present study, we show that forced N-myc expression enhances cell proliferation in the developing dorsal root ganglion. N-myc is believed to play similar functions as c-myc, but in neuronal cells. c-Myc appears to control the decision to divide or not to divide and therefore functions as a crucial mediator of signals that determine organ and body size (Trumpp et al. 2001). N-myc is expressed at moderate levels in all neural crest cells before and during their migration. After migration, when the neural crest condenses into dorsal root ganglia at around E10, N-myc persists in the nuclear compartment of the cells (Kato et al. 1991; Wakamatsu et al. 1993). The sub-cellular localization of N-myc protein changes between E10 and E11. Although it is abundant in the nucleus of most cells at E10, few, if any, nuclear N-myc can be detected at E11 (Elshamy et al. 1998; Wakamatsu et al. 1993). In the cerebellum and the rest of the brain, N-myc is essential for the expansion of progenitor cells during neurogenesis, and in its absence, there is a severely reduced proliferation (Knoepfler et al. 2002; Kenney et al. 2003; Oliver et al. 2003). The overexpression of N-myc leads consistently to forced S-phase entry in cultured post-mitotic sympathetic neurons but not in cortical neurons, as shown by BrdU incorporation (Wartiovaara et al. 2002). Our results showing that overexpression of N-myc leads to increased proliferation of sensory neuron progenitor cells in the dorsal root ganglion are thus in agreement with the previously reported role of N-myc on cell proliferation.
Mice carrying a null mutation in the N-myc gene show dorsal root ganglia that are significantly smaller and contain fewer neurons than those of wild-type mice (Stanton et al. 1992; Charron et al. 1992; Sawai et al. 1993). Our results concerning the role of N-myc on progenitor proliferation in the sensory lineage suggest that the reduced number of neurons in the null mutant mice can at least partly be ascribed to a failure of proliferation and, thereby, of expansion of the sensory neuron progenitor pool.
N-myc and apoptosis of sensory progenitor cells
Differentiation and growth arrest of neuroblastoma cells are counteracted by N-myc (Peverali et al. 1996). Because cell-cycle exit is initiated at approximately E11 in the mouse (Lawson and Biscoe 1979), a prerequisite of the elimination of nuclear N-myc activity for neuronal differentiation of the sensory progenitor cells is consistent with its translocation from the nucleus to the cytoplasm between E10 and E11. Paradoxically, N-myc overexpression leads not only to an increased tumorigenic phenotype of neuroblastoma cells but also to an increased sensitivity to drug-induced apoptosis (Fulda et al. 1999). Furthermore, N-myc overexpression causes increased serum-withdrawal-induced apoptosis in human neuroblastoma cells (van Golen et al. 2003). Overexpression of nuclear N-myc in cells that are not scheduled for cell division may lead to apoptotic death (for reviews, see Evan and Vousden 2001; Hogarty 2003). During normal development of the dorsal root ganglion, little apoptotic cell death takes place at E11. Naturally occurring cell death starts around E12, and nuclear N-myc, under physiological conditions, is unlikely to participate in this developmental process. No increased apoptosis has consistently been seen in the dorsal root ganglion of N-myc null mutant mice at E13 (Knoepfler et al. 2002). However, in neurotrophin-3 (NT-3) null mutant mice, we have previously found a marked increase of apoptosis of progenitor cells at E11 as a consequence of the loss of trophic-factor survival signaling (ElShamy and Ernfors 1996). The trophic-factor-deprived apoptotic cells express high levels of nuclear N-myc and of cell-cycle-related proteins that normally are down-regulated at this stage (ElShamy et al. 1998). The finding that the neuronal deficit of the NT-3 null mutant mice is caused by increased apoptosis (ElShamy and Ernfors 1996) in dividing cells was initially disputed (Farinas et al. 1996). Farinas et al. (1996) claimed the reduction of neuronal numbers to be a mechanism independent of cell death, i.e., to be caused by a premature cell-cycle exit and differentiation. However, in a recent study, the crossing of NT-3 null mutant mice with mice carrying a targeted deletion of the proapoptotic gene Bax has been shown to restore neuronal numbers in the dorsal root ganglion (Patel et al. 2003), unequivocally establishing that the neuronal deficits seen in NT-3 null mutant mice are caused by apoptosis. The present study is consistent with this conclusion, because we have found that N-myc overexpression is sufficient to induce apoptotic cell death of sensory progenitors in vivo. It is therefore conceivable that the trophic-factor deprivation-induced death of sensory progenitor cells in the NT-3 deficient mice is caused by the increase of N-myc in these cells.
Early effects of the transgene on progenitor cells
The effect of N-myc overexpression on the enhancement of cell proliferation and apoptosis has been observed in dorsal root ganglion at E11 but not at later stages in our nes1852tk/N-myc mice. Nestin marks stem and progenitor cells that have the capacity to form both neurons and glia (Dahlstrand et al. 1995; Alvarez-Buylla et al. 2001). Consistent with this notion, nestin is expressed in migrating neural crest cells and progenitors in the dorsal root ganglion at E10.5, but not at E12.5 (Dahlstrand et al. 1995; Lothian and Lendahl 1997). As a consequence, we can define the effects of N-myc overexpression in the nes1852tk/N-myc mice as taking place in sensory progenitor cells. Of interest, the increase of TUNEL-positive cells in this study occurs in the pool of sensory progenitor cells but takes place at a stage when the cells are scheduled to exit the cell cycle (Lawson and Biscoe 1979) and results in the loss of around one third of the normal number of neurons at E17. The increased proliferation and apoptotic cell death are probably causally linked, as has previously been shown in NT-3 null mutant mice (ElShamy and Ernfors 1996; ElShamy et al. 1998). Thus, the forced expression of N-myc in sensory progenitor cells at the stage of cell-cycle exit may lead to unscheduled cell-cycle re-entry and, as a consequence, excessive apoptosis.
N-myc induced increase of proprioceptive neurons in the dorsal root ganglion
N-myc overexpression leads to an increase in the number of parvalbumin-positive neurons in lumbar dorsal root ganglia. Parvalbumin is a marker for the large-diameter proprioceptive neurons, whereas CGRP labels nociceptive neurons (Ernfors et al. 1994). The neuronal birth days in lumbar dorsal root ganglia are from E10.5 to E11.5 for large-diameter proprioceptive/mechanoreceptive neurons and from E11.5 to E13.5 for small-diameter nociceptive neurons (Lawson and Biscoe 1979). It is intriguing that we have found a selective increase of parvalbumin-positive neurons in nes1852tk/N-myc mice. Essentially, the increase in this subpopulation of sensory neurons could come about either by affecting cell fate or by a selective expansion of a predetermined proprioceptive progenitor subtype pool. By measuring the number of cells falling into the nociceptive and proprioceptive classes, one should be able to distinguish between these possibilities, since an effect on cell fate can be predicted to take place at the expense of another population of cells, whereas expansion of a selective progenitor subtype should not affect other subtypes. Our data in this respect is unfortunately inconclusive. Although we have found a consistent reduction of CGRP-positive cells in the L4, L5, and L6 dorsal root ganglia corresponding to the number of increased parvalbumin-positive neurons in the equivalent ganglia, the reduction does not reach statistical significance. However, if we assume that proliferation and apoptotic cell death are causally linked, then the increase of proprioceptive neurons by N-myc must be an effect independent of its role in cell proliferation and cell death; this also supports the idea that overexpression of N-myc affects cell fate. In N-myc−/− mice, not only do the number of neurons in the dorsal root ganglion decrease, but also the neurons are “small” (Sawai et al. 1993). This has been interpreted as showing that the neurons fail to mature. However, in the same study, the neurons have also been described as differentiated, as revealed by the expression of the relatively late neuronal marker, neurofilament (Sawai et al. 1993). In retrospect, the preferential appearance of “small” neurons may be caused by an absence of the large proprioceptive neurons. This agrees with our results and argues that N-myc directly affects the composition of the different functional subtypes of sensory neurons in the dorsal root ganglion.