IGF-I gene delivery promotes corticospinal neuronal survival but not regeneration after adult CNS injury

Abstract

An unmet challenge of spinal cord injury research is the identification of mechanisms that promote regeneration of corticospinal motor axons. Recently it was reported that IGF-I promotes corticospinal axon growth during nervous system development. We therefore investigated whether IGF-I also promotes regeneration or survival of adult lesioned corticospinal neurons. Adult Fischer 344 rats underwent C3 dorsal column transections followed by grafts of IGF-I-secreting marrow stromal cell grafts into the lesion cavity. IGF-I secreting cell grafts promoted growth of raphespinal and cerulospinal axons, but not corticospinal axons, into the lesion/graft site. We then examined whether IGF-I-secreting cell grafts promote corticospinal motor neuron survival or axon growth in a subcortical axotomy model. IGF-I expression coupled with infusion of the IGF binding protein inhibitor NBI-31772 significantly prevented corticospinal motor neuron death (93% cell survival compared to 49% in controls, P<0.05), but did not promote corticospinal axon regeneration. Coincident with observed effects of IGF-I on corticospinal survival but not growth, expression of IGF-I receptors was restricted to the somal compartment and not the axon of adult corticospinal motor neurons. Thus, whereas IGF-I influences corticospinal axonal growth during development, its application to sites of adult spinal cord injury or subcortical axotomy fails to promote corticospinal axonal regeneration under conditions that are sufficient to prevent corticospinal cell death and promote the growth of other supraspinal axons. We conclude that developmental patterns of growth factor responsiveness are not simply recapitulated after adult injury, potentially due to post-natal shifts in patterns of IGF-I receptor expression.

Figure 1

Production and bioactivity of IGF-I. (A) IGF-I levels secreted into conditioned media by lentiviral-transduced MSCs, by ELISA (Mann Whitney *P<0.05). (B) Cerebellar granule neuron survival after 24hrs in conditioned medium from IGF-I-secreting cells, compared to GFP-transduced cells with known quantities of IGF-I protein added. Survival in presence of IGF-I-secreting cell medium is equal to that of the highest quantity of IGF-I protein (ANOVA P<0.005, post-hoc Fisher’s *P<0.05, **P<0.001). (C) MSC grafts transduced to express IGF-I exhibit significantly greater amounts of IGF-I protein after 4 weeks in vivo compared to naïve MSC grafts (Student’s t-test, *P=0.01). While naïve MSCs do not produce detectable levels of IGF-I in vitro prior to grafting, IGF-I is detectable in naïve grafts after 4 weeks in vivo, possibly due to migration of host cells into graft that express IGF-I.

Figure 2

IGF-I promotes growth of raphaespinal and rubrospinal axons into IGF-I-secreting cell grafts in sites of spinal cord injury. (A) Few 5HT-labeled serotenorgic axons penetrate a GFP-expressing cell graft in the C3 lesion site, four weeks after placement of spinal cord lesion. (B) In contrast, 5HT-labeled axons extensively penetrate an IGF-I-secreting cell graft. (C) Quantification reveals a 5-fold increase in seroteonergic axon penetration into IGF-I-secreting cell grafts (two-tailed t-test *P<0.02). (D) Similarly, tyrosine hydroxylase-labeled cerulospinal axons modestly penetrate GFP-producing cell grafts in the lesion site, and (E) more extensively penetrate IGF-I-secreting cell grafts. (F) Quantification reveals an 8-fold increase in the axon penetration of IGF-I-secreting cell grafts (two-tailed t-test **P<0.01). Scale bar = 25μm.

Figure 3

Corticospinal axons do not regenerate into IGF-I-secreting cell grafts. (A) Light-level label demonstrates corticospinal tract, adjacent to lesion site. Sagittal section, left rostral, right caudal. No axonal penetration of IGF-I-secreting cell graft is evident. (B) GFAP labeling delineates host/graft interface. (C) Merge with color rendering demonstrates absence of corticospinal axon penetration of IGF-I graft. (D) Quantification reveals rare corticospinal axonal penetration of either graft type (IGF-I: 0.67±0.49, GFP: 0.17±0.17; P=0.4). Thus, while IGF-I-secreting cell grafts elicit regeneration of other supraspinal axonal populations that modulate motor function, they do not elicit growth of corticospinal axons under these conditions. Scale bar = 100μm.

Figure 4

IGF-I prevents corticospinal motor neuron death after subcortical axotomy. (A) Nissl-stain section to illustrate lesion underlying motor cortex and graft in lesion site. Light brown staining present in layer V consists of corticospinal motor neurons retrogradely labeled with CTB. (B) Illustration of intact layer V motor cortex containing CTB-labeled corticospinal neurons. (C) Reductions in numbers of CTB-labeled motor neurons two weeks after subcortical axotomy, in subject that received GFP-producing cell graft with infusion of the IGF binding protein inhibitor NBI-31772. (D) Neurons remain largely labeled with CTB two weeks after subcortical axotomy in subjects that receive IGF-I-secreting cell grafts and infusions of NBI-31772. (E) Quantification indicates that IGF-I-secreting grafts combined with NBI-31772 infusion prevent axotomy-induced loss of CTB-labeled corticospinal motor neurons after subcortical axotomy (ANOVA P<0.005; * indicates significant differences on post-hoc Fisher’s). Scale bars, 500μm (A), 25μm (B–D).

Figure 5

Adult corticospinal motor neurons express the IGF-I receptor in the somal compartment. (A, B) IGF-I-receptor β subunit staining of large-diameter pyramidal neurons within layer V motor cortex. (C, D) IGF-I-receptor α subunit immunoloabeling in identified corticospinal motor neurons retrogradely labeled with CTB (arrowheads indicate co-labeled cells). In contrast, IGF-I receptor labeling is not detectable on lesioned or intact corticospinal motor axons in the cortex or spinal cord (not shown). (E) Heavy-chain neurofilament-immunoreactive axons do not penetrate subcortical grafts expressing IGF-I with NBI-31772 infusion (host-graft interface demarcated with dashed line), nor (F) control, GFP-producing subcortical grafts with NBI-31772 infusion. Scale bars, 100μm (A); 50μm (B–F).