Nature Neuroscience
- 9, 1371 - 1381 (2006)
Published online: 22 October 2006; | doi:10.1038/nn1789
IGF-I specifically enhances axon outgrowth of corticospinal motor neuronsP Hande Özdinler1, 2 & Jeffrey D Macklis1, 21 MGH-HMS Center for Nervous System Repair, Departments of Neurosurgery and Neurology, and Program in Neuroscience, Harvard Medical School, Massachusetts General Hospital, Boston, Massachusetts 02114, USA. 2 Harvard Stem Cell Institute, Harvard University, Boston, Massachusetts 02114, USA.
Correspondence should be addressed to Jeffrey D Macklis jeffrey_macklis@hms.harvard.edu Corticospinal motor neurons (CSMN) are among the most complex CNS neurons; they control voluntary motor function and are prototypical projection neurons. In amyotrophic lateral sclerosis (ALS), both spinal motor neurons and CSMN degenerate; their damage contributes centrally to the loss of motor function in spinal cord injury. Direct investigation of CSMN is severely limited by inaccessibility in the heterogeneous cortex. Here, using new CSMN purification and culture approaches, and in vivo analyses, we report that insulin-like growth factor-1 (IGF-I) specifically enhances the extent and rate of murine CSMN axon outgrowth, mediated via the IGF-I receptor and downstream signaling pathways; this is distinct from IGF-I support of neuronal survival. In contrast, brain-derived neurotrophic factor (BDNF) enhances branching and arborization, but not axon outgrowth. These experiments define specific controls over directed differentiation of CSMN, indicate a distinct role of IGF-I in CSMN axon outgrowth during development, and might enable control over CSMN derived from neural precursors.Corticospinal motor neurons (CSMN) and cortico–brain stem neurons are the cerebral cortex component of motor neuron circuitry. Together with spinal motor neurons in the spinal cord, CSMN control the most precise aspects of voluntary motor function1. CSMN progressively degenerate in ALS, as well as in related motor neuron diseases such as hereditary spastic paraplegia and primary lateral sclerosis2,
3,
4. CSMN damage also contributes centrally to the loss of motor function in spinal cord injury5,
6. In contrast to spinal motor neurons, which have been studied in great detail, little is known about specific controls over the differentiation, survival or connectivity of CSMN. Understanding the cellular and molecular controls over their lineage-specific differentiation, survival and axon outgrowth will elucidate the basic developmental biology of CSMN, and might enable both a better understanding of diseases involving CSMN and, potentially, their cellular repair7,
8,
9.
Insulin-like growth factor-1 (IGF-I) is important in the development and maturation of the CNS (refs. 10,11). Transient IGF-I expression occurs during projection neuron development12, including that of Purkinje neurons13. Transgenic mice with increased IGF-I expression during postnatal development exhibit an increase in brain size14, whereas Igf1-null mice (Igf1-/-), and individuals with deletions in the IGF1 gene, show defects in brain development and in the formation of functional connections15,
16,
17. The IGF-I receptor (IGF-IR) is expressed in the developing cortex, cerebellum and hippocampus12, and heterozygous mice (Igf1r+/-) show growth retardation and reduced brain volume16. IGF-I enhances survival of spinal motor neurons in the spinal cord18,
19. Although the mechanisms are not yet clear, delivery of the Igf1 gene to spinal motor neurons markedly enhances the survival of transgenic mice carrying the human SOD1G93A mutation related to familial ALS (ref. 20). Because of its effects on spinal motor neurons and mutant mouse survival, IGF-I is a candidate for clinical trials in ALS.
Although CSMN are the 'upper motor neuron' half of the circuitry that degenerates in ALS, and although the observed in vivo mouse survival effects of IGF-I may be partly due to improved CSMN survival and/or function, the roles of IGF-I and other growth factors regarding CSMN have not yet been directly investigated. However, several indirect lines of evidence suggest a central role for IGF-I in CSMN growth and/or survival. IGF binding protein-4 (IGFBP4), one of six IGF-I–binding proteins, and one that specifically inhibits IGF-I–mediated signaling events21, is specifically and increasingly expressed in CSMN during development8. Igf1 knockout mice exhibit a marked loss of axon density in the corticospinal tract. The dorsal funiculus, the main location of the corticospinal tract in the spinal cord, is reduced in size, suggesting defects in the outgrowth of axons and/or survival of CSMN (ref. 16). Together, these results suggest the importance of IGF-I in CSMN development, maturation and/or axon outgrowth.
Peptide growth factors that control the differentiation, development of polarity and process extension of CSMN have not been previously addressed, though CSMN survival requirements have been investigated. These studies suggested that a variety of growth factors, including BDNF, NT-3 and CNTF, enhance CSMN survival22, though findings regarding GDNF have been ambiguous22. The limited specificity and consistency in these findings may stem in part from the lack of any surface markers with which to purify CSMN. Information about CSMN cell biology has been gathered either by means of in vivo analysis following intracortical manipulations23 or using mixed cortical cells in culture24.
Here, we used both in vitro and in vivo analyses, and found that IGF-I specifically enhances both the rate and the extent of CSMN axon outgrowth, that this effect is distinct from the role of IGF-I in neuronal survival, and that the enhancement of CSMN axon outgrowth by IGF-I is mediated via the IGF-I receptor and its downstream signaling pathways. In our in vitro experiments, we used homogeneous populations of developing CSMN purified by fluorescence-activated cell sorting (FACS) following retrograde labeling from their axonal terminals8,
25,
26. Initial receptor expression analysis of purified CSMN revealed high expression of IGF-IR, indicating that CSMN are competent to respond to IGF-I. Experiments assessing the effects of IGF-I and other peptides for which CSMN possess receptors revealed a specific effect of IGF-I on axon outgrowth and that CSMN axon outgrowth can be selectively and specifically disrupted in vivo by IGF-IR blockade. These results reflect the direct response of purified CSMN and indicate that IGF-I is a potent enhancer of CSMN axonal elongation, both during CNS development and for CSMN newly recruited from neural precursors, that might potentially contribute to cellular repair of corticospinal circuitry.
Results
FACS-purified CSMN retain their characteristics in vitro
CSMN were retrogradely labeled with green fluorescent microspheres, which were microinjected on postnatal day (P) 2 into the corticospinal tract at the C1 level of the spinal cord using ultrasound-guided microinjection (Fig. 1a–c). Two days after injection (P4), retrogradely labeled CSMN were visualized (Fig. 1d), and anterior-to-posterior sectioning of the cortex verified that labeling was restricted to layer V of the motor cortex (Fig. 1e). The motor cortex was microdissected using a fluorescence dissecting scope. Three gates were used to purify live and pure populations of CSMN by FACS: (i) a gate selected for high levels of green fluorescence (Fig. 1f); (ii) second gate selected for live CSMN using EthD-1 exclusion (Fig. 1g,h); and (iii) a gate selected for the appropriate size and surface characteristics of CSMN, as determined by forward and side scatter (Fig. 1i). Such FACS purification yields a virtually pure population of CSMN (Fig. 1j,k).
 | | |
FACS-purified CSMN retain the basic stereotypical morphologic characteristics of CSMN in vitro, where they possess a large pyramidal cell body, a primary dendrite and an axon (Fig. 2a). After two days in vitro (2 DIV), cultured CSMN became polarized and showed characteristics of mature neurons, as detected by neurofilament (NF) and microtubule-associated protein (MAP2) immunocytochemistry (Fig. 2b,c). Cultured CSMN express Er81 (ref. 27; 100%, n = 22 neurons, Fig. 2d) and Otx1 (ref. 28; 100%, n = 13 neurons, Fig. 2e), markers of layer V neurons. In addition, cultured CSMN expressed CTIP2 (100%, n = 18 neurons, Fig. 2f), a CSMN-specific transcription factor9,
10, but did not express LMO4 (0%, n = 19 neurons, Fig. 2g), a callosal projection neuron (CPN) marker, excluded from CSMN (ref. 9). These results indicate not only that FACS-purified CSMN can be effectively cultured as pure populations in vitro, but also that they retain the stereotypical cellular and molecular characteristics of mature CSMN in culture.
| | Figure 2. FACS-purified CSMN retain the cellular and molecular characteristics of mature CSMN in culture. | | |  | (a) Photomicrograph of FACS-purified P4 CSMN cultured for 2 DIV. Essentially all cells were CSMN labeled with green fluorescent microspheres and retained the basic stereotypical morphologic characteristics of CSMN in vitro. (b,c) Cultured CSMN became polarized and showed characteristics of mature neurons, as detected by NF and MAP2 (c, arrow indicates the longest dendrite, and asterisks indicate the bifurcated axon) immunocytochemistry. (d,e) After 2 DIV, FACS-purified P4 CSMN expressed the appropriate deep cortical layer markers ER81 (d; ref. 27) and Otx1 (e; ref. 28), and the CSMN-specific transcription factor CTIP-2 (f; refs. 8,9), but not LMO4 (g), which is specifically excluded from CSMN (ref. 8). DIC, differential interference contrast. Scale bars, 40  m.
 Full Figure and legend (71K) |
| |
Receptor expression of FACS-purified CSMN was investigated immediately after purification both by immunocytochemistry and by reverse transcriptase polymerase chain reaction (RT-PCR) (Fig. 3a,b). CSMN expressed mainly IGF-IR (Fig. 3a, 85%, n = 241 neurons) and TrkB receptors (83%, n = 267 neurons), but did not express TrkA (1%, n = 156 neurons), and showed lower TrkC receptor expression (59%, n = 213 neurons) and CNTFR receptor expression (59%, n = 294 neurons). RT-PCR analysis confirmed the expression of Igf1r, Ntrk2 (TrkB), Ntrk3 (TrkC) and Cntfr, but did not detect mRNA for Ntrk1 (TrkA), Pdgfr and Kdr (VEGFR2/flk-1) (Fig. 3b). In contrast, mixed dissociated cortical cells displayed mRNA expression for all receptors tested (Fig. 3b). CSMN, in P4 cortex, expressed both IGF-IR (Fig. 3c) and TrkB receptors in vivo29. In addition, after 2 DIV, FACS-purified CSMN expressed both IGF-IR (Fig. 3d,e) and TrkB (Fig. 3f,g), but did not express TrkA. These results indicate that CSMN as a pure population have a distinct receptor expression profile both in vitro and in vivo. Purified CSMN retain their stereotypical receptor expression profile in vitro, expressing both IGF-IR and TrkB receptors, suggesting that CSMN can respond to both IGF-I and BDNF.
| | Figure 3. FACS-purified CSMN express a distinct set of growth factor and neurotrophin receptors, both in vitro and in vivo. | | |  | (a) After FACS, P4 CSMN expressed IGF-IR (upper panels) and TrkB (middle panels) receptors, but not TrkA receptors (bottom panels), as demonstrated by immunocytochemistry. (b) Similarly, we confirmed this at the mRNA level by RT-PCR: after FACS, P4 CSMN expressed Igf1r , TrkB, TrkC and Cntfr receptors, but did not express TrkA, Pdgfr and Kdr receptors (upper panels). In contrast, dissociated heterogeneous cortical cells showed mRNA expression of all these receptors (bottom panels) by RT-PCR analysis. (c) In vivo in P4 cortex, almost all retrogradely labeled CSMN (arrows) expressed IGF-IR, whereas not all IGF-IR+ cells were CSMN (arrowhead). (d–g) After 2 DIV, P4 CSMN expressed IGF-IR (d,e) and TrkB (f,g) receptors in vitro. DIC, differential interference contrast. Scale bars, 20 m.
Full Figure and legend (43K) |
| |
IGF-I specifically enhances CSMN axon outgrowth
IGF-I and BDNF exhibited distinct effects on CSMN morphology: IGF-I enhanced CSMN axon outgrowth, whereas BDNF induced branching and arborization (Fig. 4). In control medium, CSMN displayed pyramidal morphology, extended an axon, and showed limited branching and arborization (Fig. 4a,b, n = 43 neurons). In contrast, in the presence of IGF-I, CSMN showed extensive axon outgrowth (Fig. 4c,d, n = 40 neurons). When grown in the presence of BDNF, however, CSMN showed elaborate branching and arborization, but not increased axon outgrowth (Fig. 4e,f, n = 40 neurons). The same results were observed with CSMN isolated from Bax-/- mice (n = 8 mice), which provide a means to differentiate controls over morphologic differentiation and survival (ref. 30 and Fig. 4b,d,f). Together, these results strongly indicate that the observed effects of IGF-I and BDNF on CSMN polarity, axonal outgrowth and process branching are distinct from their role in neuronal survival. Although IGF-I and BDNF were equivalently potent survival factors for CSMN (82  3% and 78 4%, respectively (mean s.e.m.; data not shown), their effects on CSMN differentiation are quite distinct.
| | |
The effects of IGF-I were specific to axon extension. After 2 DIV, CSMN matured and showed distinct expression of NF and MAP2 (Fig. 2b,c). In the presence of IGF-I and BDNF, the longest neurite always expressed NF, whereas dendrites expressed MAP2 (Supplementary Fig. 1 online). These results confirmed that the longest neurite of purified CSMN is indeed an axon and that quantification of axon length is possible and specific (Fig. 5a–c).
| | Figure 5. IGF-I induces a marked and specific increase in CSMN axon outgrowth. | | |  | (a–c) Camera lucida drawings of representative CSMN in control (a), IGF-I (b) or BDNF (c) conditions; quantitative analysis included assessment of axon length (dotted lines) and process complexity via Sholl analysis (concentric circles). (d) Bar graph of soma diameter of CSMN cultured in the presence of control medium, IGF-I or BDNF. (e) Bar graph representation of average CSMN axon length when cultured in the presence of control medium, IGF-I or BDNF. IGF-I induced substantially longer axons (***P < 0.0001), whereas BDNF resulted in modestly reduced axon length (*P < 0.001) compared to control. (f) Frequency histogram of the percentage of CSMN with axons of different length. In IGF-I, there was a marked shift to much longer axons, with many fewer short axons, compared to either control medium or BDNF. (g) Bar graph representation of the average number of branch points per CSMN in the presence of control medium, IGF-I or BDNF. BDNF markedly increased CSMN process branching (***P < 0.0001). (h) Sholl analysis of CSMN cultured in the presence of control medium, IGF-I or BDNF. Error bars indicate s.e.m. Scale bar, 20 m.
Full Figure and legend (55K) |
| |
Quantification of axon length indicated that IGF-I markedly increased axon length of purified CSMN. In the presence of IGF-I or BDNF, the size of CSMN soma did not change (Fig. 5d, P = 0.6, Mann-Whitney -test), further supporting the hypothesis that the observed effects were not due to changes in neuronal viability. In control cases, CSMN axons averaged 196 9 m (mean s.e.m.) in length (Fig. 5e, n = 43 neurons). In the presence of IGF-I, the average axon length increased more than twofold, to 396 27 m (Fig. 5e, n = 40 neurons, P < 0.0001, unpaired t-test with Welch correction, d.f. = 47). In contrast, in CSMN exposed to BDNF, average axon length was decreased, to 138 5 m (Fig. 5e, n = 40 neurons, P < 0.001, unpaired t-test with Welch correction, d.f. = 41). We observed similar results with CSMN isolated from Bax-/- mice (n = 8 P4 Bax-/- mice). In the presence of IGF-I, CSMN axon length reached 409 37 m; in contrast, in CSMN exposed to BDNF, average axon length was 144 18 m. In IGF-I, there was a notable increase in the percentage of CSMN with longer axons (Fig. 5f); IGF-I increased the maximum axon length 2.5- to 3-fold compared to that in the control and BDNF conditions, respectively. Notably, a subset of CSMN displayed extensive axon outgrowth specifically in the presence of IGF-I, with some axons extending further than 1 mm in isolated culture. Many CSMN treated with IGF-I displayed axons longer than 500 m, but none of the control or BDNF-treated CSMN had axons longer than 425 m and 325 m, respectively. Other growth factors such as NGF, NT-3, NT-4 and CNTF did not substantially affect CSMN morphology. When directly added into culture medium, they resulted in an axon outgrowth of only 170 18 m, 212 25 m, 230 35 m and 224 38 m, respectively, further reinforcing the specificity of IGF-I–mediated axon outgrowth in CSMN (Supplementary Fig. 2 online).
In the presence of IGF-I, the total number of branch points was not significantly reduced (Fig. 5g, P = 0.2, unpaired t-test, d.f. = 85). In contrast, the presence of BDNF markedly increased both the branch points per CSMN (Fig. 5g, P < 0.0001, unpaired t-test, d.f. = 86). BDNF, but not IGF-I, substantially increased the level of arborization in CSMN, as analyzed by Sholl analysis (Fig. 5h). Together, these results demonstrate that IGF-I specifically and markedly enhances axon extension of CSMN.
Local application of IGF-I increases CSMN axon outgrowth
Local application of IGF-I–coated beads near a location close to both CSMN cell bodies and axon hillocks had immediate and marked effects on the rate of CSMN axon outgrowth, both in wild-type (n = 26 neurons) and Bax-/- (n = 10 neurons) CSMN (Fig. 6a–f). In the presence of control beads, coated with bovine serum albumin (BSA) or phosphate-buffered saline (PBS) (n = 26 neurons), the rate of axon outgrowth was slow (0.06 mm min-1, Fig. 6a–c,g–i, and Supplementary Video 1 online) and did not differ from the rate in the absence of beads. IGF-I enhancement of axon outgrowth was evident as soon as 5 min after application (Fig. 6d–f,j–m and Supplementary Video 2 online). In contrast, single BDNF-coated beads induced immediate and local branching at the site of placement (Fig. 6n–p and Supplementary Video 3 online; n = 12 neurons). For quantitative analysis, cases in which single beads were located within 10 m of the axon hillock were assessed over time (n = 4 control; n = 3 IGF-I; n = 2 BDNF). The average rate of axon outgrowth increased almost 20-fold, to  1.1 m min-1, in the presence of a single IGF-I–coated bead versus control beads (Fig. 6m). Axons did not change their direction or turn toward either IGF-I– or BDNF-coated beads, indicating that neither of these growth factors act as a directional chemoattractant. Together, these experiments demonstrate potent, rapid and specific effects of IGF-I on the rate and extent of CSMN axon outgrowth; indeed, IGF-I markedly increases the rate of axonal elongation, to that observed during initial CNS development31,
32.
| | Figure 6. Local application of IGF-I results in immediate and marked increase in the rate of CSMN axon outgrowth. | | |  | (a–c) Photomicrographs obtained at t = 0 min (a), t = 15 min (b) and t = 30 min (c), of a single CSMN cultured in the presence of a single control bead. (d–f) Photomicrographs obtained at t = 0 min (d), t = 15 min (e) and t = 30 min (f), of a single CSMN cultured in the presence of a single IGF-I–coated bead. (g–i) Magnification of boxed areas in a. (j–l) Magnification of boxed area in d. These panels indicate the location of the growing tip of the axon at t = 0 min (red arrow; fixed in all images) and the location of the tip of the axon at each subsequent time (green star). In the presence of control beads, very little axon outgrowth was observed over such time scales (g–i), whereas axon growth in the presence of an IGF-I–coated bead was immediate and rapid (j–l). (m) Bar graph of the average rate of axon outgrowth of CSMN in the presence of a single control, IGF-I–coated or BDNF-coated bead. (n–p) Photomicrographs obtained at t = 0 min (n), t = 15 min (o) and t = 30 min (p), of a single CSMN cultured in the presence of a BDNF-coated bead. Branching was observed near the BDNF-coated bead (arrowhead) Scale bar, 20 m.
Full Figure and legend (77K) |
| |
Blockade of IGF-I signaling reduces axon outgrowth in vitro
In vitro blockade of IGF-I receptor–mediated signaling, by means of the addition of either a function-blocking antibody to IGF-IR (anti–IGF-IR, ref. 33; n = 19 neurons) or H-1356 (ref. 34), the IGF-I analog that acts as a competitive inhibitor at the IGF-I receptor, directly to the culture medium significantly reduced IGF-I–mediated axon outgrowth of CSMN (Fig. 7a–c, P < 0.001, Kruskal-Wallis Test, n = 16 neurons). The average length of the axon outgrowth in the presence of IGF-I (409 57 m, n = 14 neurons) was reduced markedly, to 177 23 m in the presence of H-1356 and to 147 10 m in the presence of anti–IGF-IR (Fig. 7c). This was comparable to axon outgrowth in the control (155 26 m, n = 13 neurons), confirming that the enhancement of axon outgrowth in the presence of IGF-I is mediated via the IGF-I receptor.
| | Figure 7. Blockade of IGF-I signaling via the IGF-I receptor and the PI3K and ERK/MAPK pathways results in axon outgrowth defects in vitro. | | |  | (a,b) Representative photomicrographs of CSMN cultured in the presence of IGF-I (a), and IGF-I and H-1356 (b). (c) Bar graph representation of average CSMN axon length when cultured in the presence of control medium, IGF-I, IGF-I and anti–IGF-IR antibody, IGF-I and H-1356, IGF-I and the PI3K inhibitor LY 294002, IGF-I and the ERK inhibitor PD 98059, and IGF-I and both LY 294002 and PD 98059. Addition of either IGF-I receptor blockers, or inhibitors of PI3K or ERK, directly to the culture medium significantly reduced CSMN axon length, to control levels (***P < 0.001). (d–g) CSMN cultured in the presence of control medium exhibited very low levels of p-AKT (d,e) and p-ERK (f,g) (arrows). (h,i) CSMN cultured in the presence of IGF-I exhibited high levels of p-AKT (h) and p-ERK (i), and both were specifically localized to the tips of growth cones of CSMN axons (arrows). (j–m) CSMN cultured in the presence of IGF-I with LY 294002 or PD 98059 showed markedly reduced amounts of p-AKT (j,k) and p-ERK (l,m), respectively. (n–q) CSMN cultured in the presence of BDNF exhibited low levels of p-AKT (n,o), which was not restricted to the tips of the growth cone (arrows); in the presence of BDNF, no p-ERK was detected (p,q; arrows). Error bars indicate s.e.m. Scale bars, 50 m.
Full Figure and legend (42K) |
| |
We investigated whether IGF-IR–mediated axon outgrowth of CSMN occurs via the PI3K and extracellular signal–regulated kinase/mitogen-activated protein kinase (ERK/MAPK) signaling pathways, known to operate in both the peripheral nervous system and in retinal ganglion cells35,
36. Inhibition of either PI3K (via LY 294002, 5 M, n = 17 neurons) or ERK (via PD 98059, 5 M, n = 18 neurons) significantly reduced the average length of axon outgrowth in the presence of IGF-I (to 153 15 m and 135 12 m, respectively), and the presence of both inhibitors resulted in an additive reduction in the average length of axons (99 14 m, n = 17 neurons), indicating that both the PI3K and ERK/MAPK signaling pathways mediate axon outgrowth (ref. 35 and Fig. 7c).
To investigate the specificity of PI3K and ERK/MAPK signaling pathways in mediating the effects of IGF-I on axon outgrowth, we investigated whether phospho-AKT and phospho-ERK were present in CSMN cultured in the presence of control (Fig. 7d–g), IGF-I (Fig. 7h–m) or BDNF (Fig. 7n–q). We found that in the presence of IGF-I, both phospho-AKT and phospho-ERK were increased (Fig. 7h,i) when compared to those in the control (Fig. 7d–g) and were specifically localized to the tips of growth cones of CSMN axons (Fig. 7h,i). In contrast, CSMN cultured in the presence of BDNF had a low amount of phospho-AKT, which was not localized to the tips of growth cones of CSMN axons (Fig. 7n,o). Similarly, in CSMN exposed to BDNF, no detectable amount of phospho-ERK was observed (Fig. 7p,q), suggesting distinct differences between IGF-I– and BDNF-mediated signaling. In addition, we found diminished levels of phospho-AKT (Fig. 7j,k) and phospho-ERK (Fig. 7l,m) when LY 294002 or PD 98059, respectively, were added to the culture medium containing IGF-I.
IGF-IR signaling blockade reduces CSMN outgrowth in vivo
We found high-level expression of IGF-IR on CSMN axons in the dorsal funiculus of the spinal cord (Supplementary Fig. 3 online), consistent with our findings that CSMN somata express IGF-IR in vivo (Fig. 3c).
Local application of anti–IGF-IR antibody into the C1 spinal cord at P1, when CSMN have not yet innervated their spinal targets and are thus not dependent on these targets for survival (Fig. 8a–c), resulted in a notable and specific interruption of corticospinal axon outgrowth as assayed at P3 (Fig. 8a–k, n = 7). In marked contrast, application of either antibody to TrkA (anti-TrkA, Fig. 8j, n = 4) or antibody to TrkB (anti-TrkB, Fig. 8k, n = 7; n = 4 "anti-TrkB-out" in ref. 37; n = 3 "anti-TrkB-ext" in ref. 38) did not result in any axon outgrowth defects in CSMN axons. In vehicle-injected controls (n = 6), we observed that CSMN labeled with 1'-dioctadecyl-3,3,3',3'-tetramethyl-indocarbocyanine perchlorate (DiI) passed through the pons, crossed the midline in the pyramidal decussation, entered the spinal cord, and extended long axons (Fig. 8c,d,f,g,l, 3.0 0.1 mm). In the presence of anti–IGF-IR, the length of CSMN axons (the corticospinal tract, CST) from the pyramidal decussation was substantially reduced (Fig. 8l, 1.5 0.3 mm), with associated disorganization, apparent defasciculation of axon outgrowth (Fig. 8h,i) and occasional misrouting of isolated axons. Injection of anti-TrkA or anti-TrkB did not reduce the length of CSMN axons (respectively, 2.9 0.8 mm and 2.7 0.4 mm from the pyramidal decussation, Fig. 8l). CSMN death did not occur after the application of either anti–IGF-IR or anti-TrkB, as assayed by means of both Hoechst staining for apoptotic nuclear fragmentation and cleaved caspase-3 (CC-3) immunocytochemistry on retrogradely labeled CSMN (Supplementary Fig. 3, n = 4 mice). In the presence of "anti-TrkB-out" (ref. 37) or "anti-TrkB-ext" (ref. 38), BDNF-mediated branching of CSMN was markedly decreased in culture: the total number of branch points per CSMN in the presence of BDNF, 13 1, dropped to 7 2 with "anti-TrkB-out" (n = 20 neurons) and to 5 1 with "anti-TrkB-ext" (n = 21 neurons) (control, 3 1; n = 12 neurons), indicating that these TrkB antibodies are both substantially function blocking. We verified the binding of both anti–IGF-IR and anti-TrkB to CSMN axons by double labeling the spinal cord from experimental mice with fluorescent secondary antibody and anterogradely transported DiI. We found that both IGF-IR and TrkB antibodies bind exclusively to the DiI-labeled CSMN axons in the dorsal funiculus (Supplementary Fig. 3). Taken together, these experiments demonstrate that blockade of the IGF-I receptor during initial CST development specifically and significantly reduces axon extension of CSMN (P < 0.0001, analysis of variance (ANOVA), Tukey-Kramer multiple comparison test, Fig. 8l), and this is not due to CSMN cell death (Supplementary Fig. 3). These results indicate the importance of IGF-IR–mediated IGF-I signaling for CSMN axon outgrowth in vivo.
| | Figure 8. Localized in vivo blockade of IGF-I signaling markedly reduces CSMN axon growth. | | |  | (a) At P1, antibodies or control vehicle were injected into the C1 level of the spinal cord, and corticospinal tract (CST) axons were traced anterogradely by injection of DiI into the motor cortex. (b,c) Schematic of CST axon outgrowth at P1 and P3. All injections were performed at P1, and all CSMN axon outgrowth investigations were performed at P3. (d,e) Representative photomicrograph of stacked images of DiI-labeled CST axons of a representative control mouse spinal cord injected with vehicle (d, n = 6) and a representative mouse spinal cord injected with anti–IGF-IR antibody (e, n = 7). (f) Schematic of DiI-labeled CST axons taken from a single 100-m-thick section. (g–i) Magnification of boxed area in f. (g) A representative CST injected with vehicle. (h,i) Representative corticospinal tracts injected with anti–IGF-IR antibody. These images show marked axon outgrowth defects, with apparent defasciculation of axon outgrowth and occasional misrouting of isolated axons (arrows). (j,k) Representative photomicrograph of stacked images of DiI-labeled CST axons of mouse spinal cords injected with anti-TrkA antibody (j, n = 4), and anti-TrkB antibody (k, n = 7). (l) Bar graph representation of average corticospinal tract length measured from the pyramidal decussation at P3. Injection of anti–IGF-IR antibody resulted in marked and specific reduction in average CST axon length (***P < 0.001). Error bars indicate s.e.m. PD, pyramidal decussation; SC, superior colliculus; IC, inferior colliculus; CST, corticospinal tract.
Full Figure and legend (69K) |
| |
Discussion
The experiments reported here demonstrate that IGF-I is a potent and specific activator and enhancer of axon outgrowth by CSMN, both in vitro and in vivo. These effects are mediated via the IGF-I receptor and its downstream PI3K and ERK/MAPK signaling pathways, and are distinct from the role of IGF-I in neuronal survival. The findings that even local IGF-I application rapidly enhances axon outgrowth of CSMN and that in vivo blockade of the IGF-I receptor, but not the TrkB receptor, causes axon outgrowth defects in the corticospinal tract demonstrate a distinct role of IGF-I for CSMN.
In vitro approaches have been very productively developed to investigate spinal motor neurons and retinal ganglion cells directly in culture, either via antibody-based methods and/or via immunopanning39,
40. Owing to the lack of any specific surface markers for CSMN and to their presence in the highly heterogeneous cerebral cortex, we retrogradely labeled CSMN from their axonal terminals with fluorescent microspheres and obtained homogeneous populations of CSMN purified by FACS (refs. 8,25,26). Our in vitro experiments used pure cultures of CSMN and thereby are the first to directly elucidate growth factor controls over CSMN-specific differentiation. The cultured CSMN maintain the stereotypic, morphological, molecular and receptor phenotype of developing CSMN in vivo, making these findings directly applicable to CSMN development and regeneration in vivo. Our in vivo experiments confirm that the central results of the culture experiments are directly applicable to in vivo development of CSMN. In addition, these results may allow us to enhance the extent and rate of axon outgrowth of CSMN, and may have implications for repairing diseased corticospinal circuitry and establishing long-term functional connectivity.
The interaction between neuronal and non-neuronal cells in the spinal cord of chimeric SOD1 G93A mice has demonstrated the influence of astroglia and microglia on spinal motor neuron survival in vivo41. Our study highlights the importance of developing in vitro approaches by which pure populations of the highly relevant CSMN are cultured in defined medium. This strategy eliminates confounding variables that might be present in in vivo experiments, in which observed effects on CSMN may be due to unknown responses of neighboring cells to introduced factors41, and thereby secondary effects on CSMN.
A developing body of evidence complementary to our results strongly supports the idea that IGF-I is important in CSMN development and especially in axon outgrowth, in vivo: (i) Igf1-/- mice show defects in the pyramidal tract, with highly reduced axonal number and density but only a limited reduction in neuron number17; (ii) Igf1-/- mice also show a reduction in the size of the dorsal funiculus17; (iii) IGF-I protein is ten times more concentrated in the spinal cord than in the cortex during the first two postnatal weeks of development42, just when corticospinal axons enter and elongate within the corticospinal tract; (iv) CSMN express high levels of IGF-IR both in vivo and in vitro; and (v) there is a specific and gradual increase in the level of the IGF-binding protein Igfbp4 expression in CSMN when corticospinal tract axons shift from elongation to their branching/arborization mode8. Because IGFBP4 (one of six IGF-I–binding proteins) specifically inhibits IGF-I effects21, it is tempting to speculate that CSMN might reduce IGF-I–mediated axon outgrowth at later stages of development, in part by specifically and gradually increasing Igfbp4 gene expression. These previously reported findings strongly support our direct results regarding the importance of IGF-I for CSMN development and maturation and especially for the rate and extent of axon outgrowth.
Here, we present direct evidence demonstrating the role of IGF-I as a specific and potent enhancer of CSMN axon outgrowth. IGF-I does not enhance axon outgrowth in pure populations of CPN (ref. 25), which are similar to CSMN in that they are glutamatergic cortical projection neurons, but project to the contralateral hemisphere of the cortex, even though callosal neurons express high levels of IGF-IR (ref. 25) and their survival is supported by IGF-I (refs. 25,26). Similarly, IGF-I does not induce axon outgrowth in pure populations of retinal ganglion cells, but BDNF and CNTF do35. A recent study indicates the importance of IGF-I receptor–mediated signaling for the establishment of neuronal polarity in dissociated hippocampal neurons in culture43. Our results, in combination with these previous results, strongly indicate that the powerful effect of IGF-I on axon outgrowth is specific to CSMN and is distinct from its role in neuronal survival. BDNF, another survival factor for CSMN, induces dendritic branching and arborization but not axon outgrowth, dissociating the effects of survival, branching and axon elongation. Further, findings in CSMN isolated from Bax-/- mice are essentially identical to those in CSMN isolated from wild-type mice in terms of their response to IGF-I, further dissociating the support of survival from the enhancement of axon elongation of CSMN.
Although a number of shared pathways that have general roles in axon outgrowth have been identified as also operating in corticospinal projections, no specific controls have previously been identified. Wnt proteins participate in anterior-posterior guidance of corticospinal tract (CST) axons via Ryk-mediated repulsion44. Similarly, growth-inhibitory molecules associated with myelin, such as Nogo, myelin-associated glycoprotein (MAG) and oligodendrocyte-myelin glycoprotein (OMgp) inhibit axon outgrowth in the adult CNS (ref. 45). In marked contrast, no single molecule and/or growth factor has yet been identified as specifically enhancing axon outgrowth of specific brain neurons. Here, we report that IGF-I, a growth factor known to enhance survival of a variety of neuronal populations, including spinal motor neurons20, specifically enhances both the rate and extent of axon outgrowth of CSMN via both the PI3K and ERK/MAPK signaling pathways. The differences between both the level and the location of IGF-I– and BDNF-mediated activation of p-AKT and p-ERK suggest that distinct elements of IGF-I– and BDNF-mediated signaling are involved.
Further, we found that IGF-I receptor signaling is both specific for and critical to axon outgrowth during initial development of the CST in vivo. Blockade of IGF-IR signaling, by local application of anti–IGF-IR antibody at the C1 region of the spinal cord at P1, resulted in both a significant reduction in the average axon length and also disorganization and defasciculation of the CST. These results indicate that IGF-I receptor signaling is required for normal CST axon outgrowth in vivo and that this requirement is specific for IGF-I receptors and CSMN. In contrast, although CSMN express TrkB receptors, blockade of TrkB signaling by blocking antibodies did not affect axon outgrowth in the developing corticospinal tract, strongly indicating the specificity of IGF-IR–mediated signaling for CSMN axon outgrowth.
Multiple growth factors have been considered for clinical trials in ALS. Trials with CNTF (ref. 46) and BDNF (ref. 47) have been unsuccessful, but IGF-I results are inconclusive48,
49. In these studies, the role of IGF-I on CSMN was not investigated. The inconsistency of IGF-I treatment trials may be due to ineffective delivery of IGF-I and/or the limited half-life of IGF-I in vivo. Recently, adeno-associated virus delivery of the Igf1 gene to spinal motor neurons markedly enhanced survival of both spinal motor neurons and transgenic mice carrying the human SOD1 G93A mutation related to familial ALS (ref. 20).
A major challenge to the repair of complex long-distance neocortical circuitry and neocortical output circuitry lies in the fact that molecules and molecular controls that control, direct and enhance progressive differentiation, axonal projection and connectivity of neural precursors and developing neurons along a specific neuronal lineage have not yet been established. Quite relevant to our investigations, endogenous neural precursors are capable of differentiating into a small number of CSMN and extending long axons into the spinal cord of adult mice7. However, in this set of previous experiments, when the neurogenesis of CSMN in adult mice was induced, most of the newborn neurons died during an extended period of axonal elongation into the spinal cord. These results motivate the idea that both support of CSMN survival and enhancement of axonal elongation to supportive targets might enable newly recruited CSMN to join CNS circuitry and effect functional cellular repair. We speculate that most of the newborn CSMN fail to extend an axon rapidly enough and thus fail to innervate supportive targets. Our current results indicate that IGF-I is an important positive control over the elongation of CSMN axons and might potentially be applied to enhance the outgrowth of CSMN in order to increase connectivity, functional integration and long-term survival.
Taken together, these experiments demonstrate that IGF-I is a potent and specific enhancer of CSMN axon elongation, that its effect is independent of its role in supporting CSMN survival and that these effects are mediated via IGF-I receptor signaling pathways. This role is highly relevant to understanding the basic developmental biology of this critical population of cortical neurons and to develop future treatments to promote the survival of vulnerable or diseased CSMN involved in ALS and other motor neuron degenerations. It is also relevant for developing strategies for regenerating injured spinal cord and for potential cellular replacement strategies aimed at manipulating neural precursors or stem cells. Further direct understanding of the factors that exert control over CSMN and other distinct neuronal populations may enable important and new insights into CNS development, circuit formation, regeneration and repair.
Methods
Mice.
CD1, C57/Bl6 (Charles River Laboratories) and Bax-/- mice (C57/Bl6 background, Jackson Laboratories) were used in these experiments. All mouse studies were approved by the Massachusetts General Hospital Institutional Animal Care and Use Committee and performed in accordance with institutional and federal guidelines.
CSMN labeling, dissociation and purification.
CSMN were retrogradely labeled with green fluorescent microspheres (Lumafluor), by means of ultrasound-guided microinjection (Vevo 660; VisualSonics) into the CST at the C1 level of the spinal cord at P2. At P4, retrogradely labeled motor cortex was microdissected using a fluorescence-equipped dissecting microscope (SMZ-1500; Nikon) in the presence of cold dissociation medium (20 mM glucose, 0.8 mM kynurenic acid, 0.05 mM D(-)-2-amino-5-phosphonovaleric acid (AP5), 50 U ml-1 penicillin, 0.05 mg ml-1 streptomycin, 0.9 M Na2SO4 and 0.014 M MgCl2, pH=7.35, and supplemented with B27) and enzymatically digested for 15–20 min (0.16 mg liter-1 L-cysteine HCl, 12 U ml-1 papain and 1U ml-1 DNAseI, pH = 7.35, prepared in dissociation medium) at 37 °C. Enzymatic digestion was blocked by dissociation medium containing 10 mg ml-1 ovomucoid (Sigma) and 10 mg ml-1 bovine serum albumin (BSA), and cells were mechanically dissociated in trituration buffer (OPTIMEM, supplemented with 20 mM glucose, 0.4 mM kynurenic acid, 0.025 mM AP5, B27 and BSA). The supernatant was collected in trituration buffer for FACS purification using a FACSVantage SE Diva flow cytometer (Becton Dickinson). Retrograde labeling, dissociation and FACS purification yielded approximately 1,000 live CSMN per P4 pup, approximately 15–20% of total CSMN (6,000 per hemisphere). Each experiment was repeated at least 4–5 times, and results were highly reproducible and comparable.
Reverse transcription-PCR.
RNA was prepared from FACS-purified CSMN immediately after sorting or from dissected P4 cortex (Microprep kit; Stratagene). cDNA was transcribed and RT-PCR was performed as previously described26, with minor modifications.
Culturing pure populations of CSMN.
FACS-purified CSMN were cultured on glass coverslips (2  103 to 3 103 cells per coverslip, 10 mm, Fisherbrand) coated with poly-L-lysine (10 g ml-1, Sigma), in 24-well culture plates (at 37 °C, in a humidified tissue culture incubator in the presence of 5% CO2, for 2 d). CSMN were plated (i) in serum-free medium (SFM; 0.034 mg liter-1 BSA, 1 mM L-glutamine, 25 U ml-1 penicillin, 0.025 mg ml-1 streptomycin, 35 mM glucose and 0.5% B27 in Neurobasal-A medium (Life Technologies)); (ii) in conditioned medium (CM; SFM conditioned overnight by P2 cortical cells); and (iii) in control medium (control; 1:1 CM:SFM). CSMN were cultured for 2 DIV. In some experiments, the PI3K inhibitor LY 294002 (LC Laboratories; 5 M), the ERK inhibitor PD 98059 (LC Laboratories; 5 M), the IGF-I analog that acts as a competitive inhibitor on the IGF-I receptor H-1356 (ref. 34; Bachem; 10 M), or function-blocking anti–IGF-IR antibody (ref. 33; Santa Cruz) were added directly to the culture medium.
CSMN morphology assays.
Control medium was used to investigate the role of specific growth factors on CSMN morphology, by means of both Sholl analysis and an analysis of soma diameter, axon length and total number of process branch points per CSMN. All assays were conducted blind. Growth factors of interest were added directly to control medium (bath applied, 25 ng ml-1). We used extremely rigorous inclusion criteria for our detailed analysis. We used only those blinded samples that, at 2 DIV, contained green microspheres, exhibited CSMN morphology in vitro with a large soma size, defined apical dendrite and a minimum axon length of 100 m, and lacked any cellular contact with any other CSMN in culture. Each sample that fit these criteria (usually 1 in 10–40 CSMN observed) was photographed and analyzed by an investigator blind to the culture conditions. Soma diameter, axon length and total number of branch points per CSMN were measured using OpenLab quantification software; the accuracy of measurements was verified with US National Institutes of Health (NIH) imaging software. For Sholl analysis, we used grids of concentric circles 10 m apart and counted the number of times the processes crossed at each concentric circle. All statistical analysis was performed using InStat software (version 3.0a, Graphpad) via parametric and/or nonparametric analyses, as appropriate, with a minimum significance level set at P < 0.01. These experiments were repeated in Bax-/- mice30.
Immunocytochemistry.
Cells were fixed with 2% paraformaldehyde (PFA) for 10 min at 19–22 °C, and washed with PBS three times. Immunocytochemistry was performed as previously described26. In this study, we used primary antibodies to MAP2 (1:500; Sigma), NF (neurofilament heavy chain (NF-H)), 1:500; Sigma), TrkB (1:1,000; gift from D. Kaplan, Hospital for Sick Children, Toronto), IGF-IR (1:250; Santa Cruz), ER81 (1:500; gift from T. Jessell, Columbia Univ., New York), Otx1 (1:2,000; Developmental Studies Hybridoma Bank), CTIP2 (1:300; gift from M. Leid, Oregon State Univ., Corvallis, Oregon), LMO4 (1:200; Santa Cruz), phospho-AKT (1:50; Cell Signaling Technology), phospho-ERK (1:50; Cell Signaling Technology) and cleaved caspase-3 (CC-3, 1:500; New England Biolabs). Appropriate secondary antibodies were obtained from the Molecular Probes Alexa series. Hoechst (1:10,000; Molecular Probes) was used to visualize nuclear staining. Coverslips were mounted on slides and analyzed using a Nikon E1000 fluorescence microscope equipped with an X-Cite 120 illuminator (EXF0), and images were acquired using a cooled CCD digital camera (Coolpix). Confocal images were collected using a BioRad Radiance 2100 Rainbow laser-scanning confocal microscope based on a Nikon E800 microscope.
Local application of growth factors.
Growth factors were passively adsorbed on latex beads (blue, 10 m diameter, Polysciences); these adsorb growth factors and release the peptides, and thus have immediate effects50. The amount of growth factor on beads (0.4 ng IGF-I or BDNF per bead) was determined using a Micro BCA Protein Assay kit (Pierce). Growth factor–coated beads were placed adjacent to the CSMN cell body and axon hillock using a micromanipulator on a heated stage of a Zeiss Axiovert 200 microscope equipped with Hoffman DIC optics and an X-cite 120 Fluorescence illuminator unit; 5% CO2 flowed directly into the chamber. Images (at 40 with 1.6 optical zoom) were acquired every 1 min with a cooled CCD camera (Photometrics Coolsnap) and collected with MetaView imaging analysis software (version 6.0). Image stacks were assembled using NIH ImageJ software (version 1.31), and time-lapse movies were compiled using Quicktime software (version 6.5).
In vivo antibody application and DiI tracing.
Anti-TrkA (gift from L. Reichardt, Univ. of California, San Francisco), "anti-TrkB-out" (raised against the extracellular domain of the receptor37; gift from D. Kaplan), "anti-TrkB-ext" (raised against the extracellular domain of the receptor38; gift from L. Reichardt), anti–IGF-IR (function-blocking33, Santa Cruz) antibodies, and control vehicle were injected into the C1 level of the spinal cord at P1, using an ultrasound-guided microinjection system. CSMN were traced anterogradely in vivo using DiI, by injection into the motor cortex at P1 (10 g ml-1 in dimethylformamide, Fig. 8a). At P3, pups were perfused, brains with spinal cord were fixed en bloc in 4% PFA overnight, and sagittal sections (100 m) were obtained along the anterior-posterior axis. Images of all sections that contained CSMN axons were digitally 'collapsed' to obtain a two-dimensional image of the DiI-labeled corticospinal tract. Average axon length in the CST was determined by measuring the longest axonal extension for each sample from the pyramidal decussation in the medulla as the rostral reference point, using OpenLab morphometric and quantification software. For some high magnification analysis, a single 100-m section was used to investigate axon outgrowth and fasiculation along the corticospinal tract. All statistical analysis was performed using InStat software (version 3.0a, Graphpad) via nonparametric analyses, as appropriate, with a minimum significance level set at P < 0.01. In some cases, following anti-receptor antibody application, CSMN were retrogradely labeled from the pons via ultrasound-guided microinjection of green fluorescent microspheres, and potential CSMN death was investigated via both Hoechst staining for apoptotic nuclear fragmentation and CC-3 immunocytochemistry in retrogradely labeled CSMN.
Note: Supplementary information is available on the Nature Neuroscience website.
Author contributions
This study was jointly designed by P.H.Ö. and J.D.M.; experiments were performed by P.H.Ö.; P.H.Ö. and J.D.M. jointly analyzed and interpreted data and jointly wrote the paper.
Received 21 August 2006; Accepted 25 September 2006; Published online: 22 October 2006.
REFERENCES
- Winhammar, J.M., Rowe, D.B., Henderson, R.D. & Kiernan, M.C. Assessment of disease progression in motor neuron disease. Lancet Neurol. 4, 229–238 (2005). | PubMed |
- Bruijn, L.I., Miller, T.M. & Cleveland, D.W. Unraveling the mechanisms involved in motor neuron degeneration in ALS. Annu. Rev. Neurosci. 27, 723–749 (2004). | Article | PubMed | ISI | ChemPort |
- Pasinelli, P. & Brown, R.H. Molecular biology of amyotrophic lateral sclerosis: insights from genetics. Nat. Rev. Neurosci. 7, 710–723 (2006). | Article | PubMed | ChemPort |
- Rosler, K.M., Truffert, A., Hess, C.W. & Magistris, M.R. Quantification of upper motor neuron loss in amyotrophic lateral sclerosis. Clin. Neurophysiol. 111, 2208–2218 (2000). | Article | PubMed | ChemPort |
- Hains, B.C., Black, J.A. & Waxman, S.G. Primary cortical motor neurons undergo apoptosis after axotomizing spinal cord injury. J. Comp. Neurol. 462, 328–341 (2003). | Article | PubMed |
- Schwab, M.E. Repairing the injured spinal cord. Science 295, 1029–1031 (2002). | Article | PubMed | ISI | ChemPort |
- Chen, J., Magavi, S.S. & Macklis, J.D. Neurogenesis of corticospinal motor neurons extending spinal projections in adult mice. Proc. Natl. Acad. Sci. USA 101, 16357–16362 (2004). | Article | PubMed | ChemPort |
- Arlotta, P. et al. Neuronal subtype-specific genes that control corticospinal motor neuron development in vivo. Neuron 45, 207–221 (2005). | Article | PubMed | ChemPort |
- Molyneaux, B.J., Arlotta, P., Hirata, T., Hibi, M. & Macklis, J.D. Fezl is required for the birth and specification of corticospinal motor neurons. Neuron 47, 817–831 (2005). | Article | PubMed | ISI | ChemPort |
- Popken, G.J. et al. In vivo effects of insulin-like growth factor-I (IGF-I) on prenatal and early postnatal development of the central nervous system. Eur. J. Neurosci. 19, 2056–2068 (2004). | Article | PubMed |
- Wilkins, A., Chandran, S. & Compston, A. A role for oligodendrocyte-derived IGF-1 in trophic support of cortical neurons. Glia 36, 48–57 (2001). | Article | PubMed | ChemPort |
- Bondy, C.A. & Lee, W.H. Patterns of insulin-like growth factor and IGF receptor gene expression in the brain. Functional implications. Ann. NY Acad. Sci. 692, 33–43 (1993). | PubMed | ChemPort |
- Bondy, C.A. Transient IGF-I gene expression during the maturation of functionally related central projection neurons. J. Neurosci. 11, 3442–3455 (1991). | PubMed | ChemPort |
- O'Kusky, J.R., Ye, P. & D'Ercole, A.J. Insulin-like growth factor-I promotes neurogenesis and synaptogenesis in the hippocampal dentate gyrus during postnatal development. J. Neurosci. 20, 8435–8442 (2000). | PubMed | ChemPort |
- Baker, J., Liu, J.P., Robertson, E.J. & Efstratiadis, A. Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75, 73–82 (1993). | Article | PubMed | ISI | ChemPort |
- Liu, J.P., Baker, J., Perkins, A.S., Robertson, E.J. & Efstratiadis, A. Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75, 59–72 (1993). | Article | PubMed | ISI | ChemPort |
- Beck, K.D., Powell-Braxton, L., Widmer, H.R., Valverde, J. & Hefti, F. Igf1 gene disruption results in reduced brain size, CNS hypomyelination, and loss of hippocampal granule and striatal parvalbumin-containing neurons. Neuron 14, 717–730 (1995). | Article | PubMed | ISI | ChemPort |
- Neff, N.T. et al. Insulin-like growth factors: putative muscle-derived trophic agents that promote motoneuron survival. J. Neurobiol. 24, 1578–1588 (1993). | Article | PubMed | ChemPort |
- Dobrowolny, G. et al. Muscle expression of a local Igf-1 isoform protects motor neurons in an ALS mouse model. J. Cell Biol. 168, 193–199 (2005). | Article | PubMed | ISI | ChemPort |
- Kaspar, B.K., Llado, J., Sherkat, N., Rothstein, J.D. & Gage, F.H. Retrograde viral delivery of IGF-1 prolongs survival in a mouse ALS model. Science 301, 839–842 (2003). | Article | PubMed | ISI | ChemPort |
- Clemmons, D.R. et al. Role of insulin-like growth factor binding proteins in the control of IGF actions. Prog. Growth Factor Res. 6, 357–366 (1995). | Article | PubMed | ChemPort |
- Giehl, K.M. Trophic dependencies of rodent corticospinal neurons. Rev. Neurosci. 12, 79–94 (2001). | PubMed | ISI | ChemPort |
- Harrington, A.W. et al. Secreted proNGF is a pathophysiological death-inducing ligand after adult CNS injury. Proc. Natl. Acad. Sci. USA 101, 6226–6230 (2004). | Article | PubMed | ChemPort |
- Junger, H. & Varon, S. Neurotrophin-4 (NT-4) and glial cell line-derived neurotrophic factor (GDNF) promote the survival of corticospinal motor neurons of neonatal rats in vitro. Brain Res. 762, 56–60 (1997). | Article | PubMed | ChemPort |
- Catapano, L.A., Arlotta, P., Cage, T.A. & Macklis, J.D. Stage-specific and opposing roles of BDNF, NT-3 and bFGF in differentiation of purified callosal projection neurons toward cellular repair of complex circuitry. Eur. J. Neurosci. 19, 2421–2434 (2004). | Article | PubMed |
- Catapano, L.A., Arnold, M.W., Perez, F.A. & Macklis, J.D. Specific neurotrophic factors support the survival of cortical projection neurons at distinct stages of development. J. Neurosci. 21, 8863–8872 (2001). | PubMed | ISI | ChemPort |
- Hevner, R.F. et al. Beyond laminar fate: toward a molecular classification of cortical projection/pyramidal neurons. Dev. Neurosci. 25, 139–151 (2003). | Article | PubMed | ISI | ChemPort |
- Frantz, G.D., Bohner, A.P., Akers, R.M. & McConnell, S.K. Regulation of the POU domain gene SCIP during cerebral cortical development. J. Neurosci. 14, 472–485 (1994). | PubMed | ISI | ChemPort |
- Giehl, K.M. et al. Endogenous brain-derived neurotrophic factor and neurotrophin-3 antagonistically regulate survival of axotomized corticospinal neurons in vivo. J. Neurosci. 21, 3492–3502 (2001). | PubMed | ISI | ChemPort |
- Deckwerth, T.L. et al. BAX is required for neuronal death after trophic factor deprivation and during development. Neuron 17, 401–411 (1996). | Article | PubMed | ISI | ChemPort |
- David, S. & Aguayo, A.J. Axonal elongation into peripheral nervous system "bridges" after central nervous system injury in adult rats. Science 214, 931–933 (1981). | PubMed | ISI | ChemPort |
- Goldberg, J.L. & Barres, B.A. Neuronal regeneration: extending axons from bench to brain. Curr. Biol. 8, R310–R312 (1998). | Article | PubMed | ChemPort |
- Li, S.L., Kato, J., Paz, I.B., Kasuya, J. & Fujita-Yamaguchi, Y. Two new monoclonal antibodies against the alpha subunit of the human insulin-like growth factor-I receptor. Biochem. Biophys. Res. Commun. 196, 92–98 (1993). | Article | PubMed | ChemPort |
- Pietrzkowski, Z., Wernicke, D., Porcu, P., Jameson, B.A. & Baserga, R. Inhibition of cellular proliferation by peptide analogues of insulin-like growth factor 1. Cancer Res. 52, 6447–6451 (1992). | PubMed | ISI | ChemPort |
- Goldberg, J.L. et al. Retinal ganglion cells do not extend axons by default: promotion by neurotrophic signaling and electrical activity. Neuron 33, 689–702 (2002). | Article | PubMed | ISI | ChemPort |
- Liu, R.Y. & Snider, W.D. Different signaling pathways mediate regenerative versus developmental sensory axon growth. J. Neurosci. 21, RC164 (2001). | PubMed | ChemPort |
- Fryer, R.H. et al. Developmental and mature expression of full-length and truncated TrkB receptors in the rat forebrain. J. Comp. Neurol. 374, 21–40 (1996). | Article | PubMed | ISI | ChemPort |
- Huang, E.J. et al. Expression of Trk receptors in the developing mouse trigeminal ganglion: in vivo evidence for NT-3 activation of TrkA and TrkB in addition to TrkC. Development 126, 2191–2203 (1999). | PubMed | ISI | ChemPort |
- Bloch-Gallego, E. et al. Survival in vitro of motoneurons identified or purified by novel antibody-based methods is selectively enhanced by muscle-derived factors. Development 111, 221–232 (1991). | PubMed | ChemPort |
- Meyer-Franke, A., Kaplan, M.R., Pfrieger, F.W. & Barres, B.A. Characterization of the signaling interactions that promote the survival and growth of developing retinal ganglion cells in culture. Neuron 15, 805–819 (1995). | Article | PubMed | ChemPort |
- Clement, A.M. et al. Wild-type nonneuronal cells extend survival of SOD1 mutant motor neurons in ALS mice. Science 302, 113–117 (2003). | Article | PubMed | ISI | ChemPort |
- Rotwein, P., Burgess, S.K., Milbrandt, J.D. & Krause, J.E. Differential expression of insulin-like growth factor genes in rat central nervous system. Proc. Natl. Acad. Sci. USA 85, 265–269 (1988). | Article | PubMed | ChemPort |
- Sosa, L. et al. IGF-1 receptor is essential for the establishment of hippocampal neuronal polarity. Nat. Neurosci. 9, 993–995 (2006). | Article | PubMed | ChemPort |
- Liu, Y. et al. Ryk-mediated Wnt repulsion regulates posterior-directed growth of corticospinal tract. Nat. Neurosci. 8, 1151–1159 (2005). | Article | PubMed | ISI | ChemPort |
- Schwab, M.E. Nogo and axon regeneration. Curr. Opin. Neurobiol. 14, 118–124 (2004). | Article | PubMed | ISI | ChemPort |
- A double-blind placebo-controlled clinical trial of subcutaneous recombinant human ciliary neurotrophic factor (rHCNTF) in amyotrophic lateral sclerosis. ALS CNTF Treatment Study Group. Neurology 46, 1244–1249 (1996).
- A controlled trial of recombinant methionyl human BDNF in ALS: the BDNF Study Group (Phase III) Neurology, 52, 1427–1433 (1999).
- Borasio, G.D. et al. A placebo-controlled trial of insulin-like growth factor-I in amyotrophic lateral sclerosis. European ALS/IGF-I Study Group. Neurology 51, 583–586 (1998). | PubMed | ISI | ChemPort |
- Lai, E.C. et al. Effect of recombinant human insulin-like growth factor-I on progression of ALS. A placebo-controlled study. The North America ALS/IGF-I Study Group. Neurology 49, 1621–1630 (1997). | PubMed | ISI | ChemPort |
- Horch, H.W. & Katz, L.C. BDNF release from single cells elicits local dendritic growth in nearby neurons. Nat. Neurosci. 5, 1177–1184 (2002). | Article | PubMed | ISI | ChemPort |
Acknowledgments
We thank B. Molyneaux for help and advice on neonatal surgery; L. Catapano and J. Chen for help with neuron isolation; D. Dombkowski for expert advice and optimization of FACS methods; D. Scadden for generous access to tissue culture facilities near the FACS facility; N. Kishi for help with Sholl analysis; J. Emsley for advice on statistical analysis; L. Reichardt, D. Kaplan, T. Jessell and M. Leid for gifts of antibodies; F. Briggs, A. Eswar and A. Palmer for technical assistance; A. Chandawarkar and J. Nagurney for help with blinded data analysis and immunocytochemistry; T. Jakobs for help with live imaging; and J. Emsley, P. Arlotta, S. Sohur, J. Menezes and other members of the Macklis lab for critical reading of the manuscript. This work was supported by grants from the US National Institutes of Health (NS49553, NS45523 and NS41590) and the ALS Association (to J.D.M.). P.H.O. was supported by a Harvard Center for Neurodegeneration and Repair Postdoctoral Fellowship.
Competing interests statement:
The authors declare that they have no competing financial interests. |
