Spinal motoneuron synaptic plasticity after axotomy in the absence of inducible nitric oxide synthase

Abstract

Background: Astrocytes play a major role in preserving and restoring structural and physiological integrity following injury to the nervous system. After peripheral axotomy, reactive gliosis propagates within adjacent spinal segments, influenced by the local synthesis of nitric oxide (NO). The present work investigated the importance of inducible nitric oxide synthase (iNOS) activity in acute and late glial responses after injury and in major histocompatibility complex class I (MHC I) expression and synaptic plasticity of inputs to lesioned alpha motoneurons.

Methods: In vivo analyses were carried out using C57BL/6J-iNOS knockout (iNOS(-/-)) and C57BL/6J mice. Glial response after axotomy, glial MHC I expression, and the effects of axotomy on synaptic contacts were measured using immunohistochemistry and transmission electron microscopy. For this purpose, 2-month-old animals were sacrificed and fixed one or two weeks after unilateral sciatic nerve transection, and spinal cord sections were incubated with antibodies against classical MHC I, GFAP (glial fibrillary acidic protein - an astroglial marker), Iba-1 (an ionized calcium binding adaptor protein and a microglial marker) or synaptophysin (a presynaptic terminal marker). Western blotting analysis of MHC I and nNOS expression one week after lesion were also performed. The data were analyzed using a two-tailed Student's t test for parametric data or a two-tailed Mann-Whitney U test for nonparametric data.

Results: A statistical difference was shown with respect to astrogliosis between strains at the different time points studied. Also, MHC I expression by iNOS(-/-) microglial cells did not increase at one or two weeks after unilateral axotomy. There was a difference in synaptophysin expression reflecting synaptic elimination, in which iNOS(-/-) mice displayed a decreased number of the inputs to alpha motoneurons, in comparison to that of C57BL/6J.

Conclusion: The findings herein indicate that iNOS isoform activity influences MHC I expression by microglial cells one and two weeks after axotomy. This finding was associated with differences in astrogliosis, number of presynaptic terminals and synaptic covering of alpha motoneurons after lesioning in the mutant mice.

Figure 1

Iba-1 and NeuN double-labeling in C57BL6/J and iNOS^-/- mice, one and two weeks after axotomy. Observe the strong upregulation of Iba-1 labeling in C57BL6/J (WT, F) and iNOS^-/- (N) one week after lesion, as compared to the unlesioned side (B and J). Two weeks after axotomy, the microglial reaction decreased (H and P). The microglia presented in close apposition to axotomized motoneurons (G and O). (A and I) NeuN labeling on the contralateral side of C57BL6/J and iNOS^-/- mice, respectively. (E and M) NeuN labeling on the ipsilateral side of C57BL6/J and iNOS^-/- mice, respectively. (C and K) Iba-1 and NeuN double-labeling in the unlesioned side of C57BL6/J and iNOS^-/- mice. (D and L) Iba-1 labeling on the unlesioned side of C57BL6/J and iNOS^-/- mice, two weeks after axotomy. (Q) Graph representing quantification of the integrated density of pixels (×10³) for Iba-1 immunohistochemistry in the neuropil of the unlesioned and lesioned sides, one (1 w) and two (2 w) weeks after axotomy. * = p < 0.05 (Mann-Whitney U test). Scale bar = 50 μm. (R) Schematic representation of the sciatic nerve pool in the ventral horn of spinal cord. One motoneuron is shown in detail with apposed pre-synaptic terminals. The dashed circles represent areas where measurements of the integrated density of pixels were performed.

Figure 2

Classical MHC I and Iba-1 double-staining in C57BL6/J and iNOS^-/- mice, one and two weeks after axotomy. Observe the significantly stronger upregulation of MHC I (ER-Hr52) labeling in C57BL6/J (WT) mice one week after lesion (B), which is co-localized with microglial Iba-1 positive cells (C). The lesioned side (F) of the spinal cord of iNOS^-/- mice shows almost no labeling, (A and E) Iba-1 labeling on the lesioned side of C57BL6/J and iNOS^-/- mice, respectively. Two weeks after lesioning, ER-HR52 immunolabeling is even stronger than in the acute phase (D). (H) No MHC I labeling was seen in the ventral horn of iNOS^-/- mice, two weeks after sciatic transection. (I). Western blotting analysis of MHC I expression in spinal cord ventral horn ipsi- and contralateral to axotomy. Note the upregulation of MHC I in wild type mice after lesion, which is not observed in iNOS^-/- animals. β-Actin was used as sample loading control. IL = ipsilateral; CL = contralateral. * = p < 0.05; ** = p < 0.01 (Mann-Whitney U test). (J) Graph representing quantification of the integrated density of pixels for MHC I immunolabeling in the neuropil adjacent to large motoneurons, one (1 w) and two (2 w) weeks after axotomy. * = p < 0.05, Student t test. Scale bar = 50 μm.

Figure 3

GFAP and NeuN double-labeling in C57BL6/J and iNOS^-/- mice, one and two weeks after axotomy. Observe the strong upregulation of GFAP labeling in C57BL6/J (WT, F) and iNOS^-/- (N) one week after lesion, as compared to the unlesioned side (B and J). The iNOS^-/- unlesioned side (J) shows milder basal astroglial labeling as compared to control mice. After lesioning the astroglial reaction presented a diffuse distribution surrounding the axotomized pool of motoneurons in WT mice (G), and was more concentrated around the motoneurons in iNOS-deficient mice (O). (A and I) NeuN labeling on the contralateral side of C57BL6/J and iNOS^-/- mice, respectively. (E and M) NeuN labeling on the ipsilateral side of C57BL6/J and iNOS^-/- mice, respectively. (C and K) GFAP and NeuN double-labeling on the unlesioned side of C57BL6/J and iNOS^-/- mice. (H and P) GFAP labeling in C57BL6/J and iNOS^-/- mice, two weeks after axotomy, lesioned side. (D and L) GFAP labeling in C57BL6/J and iNOS^-/- mice, two weeks after axotomy, unlesioned side. (Q) Graph representing quantification of the integrated density of pixels for GFAP in the neuropil of the unlesioned and lesioned sides, one (1 w) and two weeks (2 w) after axotomy. * = p < 0.05, ** = p < 0.01 and *** = p < 0.001 (Student t test). Scale bar = 50 μm.

Figure 4

Classical MHC I and GFAP double-staining in C57BL6/J and iNOS^-/- mice, one week after axotomy. Observe that GFAP labeling in C57BL6/J (WT) mice is not co-localized with MHC I- (ER-Hr52) positive cells (C). Again, axotomized iNOS^-/- mice did not express classical MHC I (E). (A and D) GFAP labeling on the lesioned side of C57BL6/J and iNOS^-/- mice, respectively. Scale bar = 50 μm.

Figure 5

Synaptophysin immunostaining in C57BL6/J and iNOS^-/- mice, one week after unilateral axotomy. Note that one week after lesioning, there was a stronger decrease in labeling especially in the areas surrounding the motoneurons. This decrease was more intense in iNOS^-/- mice (D) than in C57BL6/J mice (B). (A and C) Unlesioned side of C57BL6/J and iNOS^-/- mice, respectively. The dashed circle indicates the motor nucleus containing the alpha motoneurons. (E) Graph representing quantification of the integrated density of pixels in the neuropil adjacent to large motoneurons. * = p < 0.05 (Student t test). Scale bar = 50 μm.

Figure 6

iNOS and nNOS labeling in C57BL6/J and iNOS^-/- mice, one and two weeks after unilateral axotomy. (A) positive iNOS labeling within the motor nucleus in C57BL6/J (WT) one week after axotomy (1 w). The lesioned side (B) of the spinal cord of WT mice shows no labeling two weeks after axotomy (2 w). (C and D) nNOS immunolabeling in C57BL6/J and iNOS^-/- ventral horn one week after lesioning, respectively. (E) Western blot analysis of nNOS expression in iNOS knockout mice and WT mice. Observe the absence of statistically significant nNOS upregulation after peripheral axotomy as well as the similar expression of such protein in both strains. β-Actin was used as sample loading control. IL = ipsilateral; CL = contralateral. Scale bar = 50 μm.

Figure 7

Synaptic covering obtained by a detailed analysis of inputs in apposition to the surface of sciatic motoneurons. (A) Percentage of absolute retraction of nerve terminals from apposition to the postsynaptic membrane in the different mice strains. One week after axotomy, a significant loss of covering can be seen in iNOS^-/- mice as compared to C57BL6/J mice. (B - D) Percentage of synaptic covering by F-, S- and C-terminals on the unlesioned and lesioned sides. Note that after injury the iNOS^-/- mice presented a greater loss of F- type terminals than the WT mice. * = p < 0.05 and ** = p < 0.01 (Mann-Whitney U test).

Figure 8

Quantitative ultrastructural analyses of the number of synaptic terminals apposed to the surface of motoneurons after axotomy. (A and E) Representative micrographs of the surface of unlesioned motoneurons from C57BL6/J and iNOS^-/- mice strains, respectively. One week after peripheral lesion, C57BL6/J mice show a few synaptic terminals detached from the surface of lesioned motoneurons (B). Such detachment occurred to a greater degree in the ventral horn of the iNOS knockout strain (F). Observe the comparatively lower degree of synaptic covering in iNOS^-/- mice as compared to C57BL6/J mice. DT = detached terminal, T = apposed terminal, Mt = motoneuron. The black arrows indicate the location of motoneuron membrane surfaces from which synaptic terminals were detached. (C) Graph of the total number of synaptic terminals per 100 μm of motoneuron membrane in lesioned and unlesioned neurons. Observe that unlesioned iNOS^-/- mice show greater synaptic elimination as compared to WT. This difference is even higher after axotomy. (F, G and H) Graphs of F-, S- and C-terminal numbers per 100 μm of motoneuron membranes, respectively. A greater loss of F-type terminals can be seen one week after axotomy in iNOS^-/- mice. * = p < 0.05 and ** = p < 0.01 (Mann-Whitney U test). Scale bar = 1 μm.

Figure 9

Graphs showing frequency distributions of gaps between terminals along the motoneuron membrane. The general gap distribution was retained in both strains one week after lesion indicating that, apart from differences regarding number and covering of terminals, synaptic elimination occurred in such a way that clusters of boutons were preserved. (A and B) Graph of the number of gaps between apposed terminals per length of membrane (μm) of C57BL6/J (WT) mice on the unlesioned and lesioned sides, respectively. (C and D) Graph of the number of gaps between apposed terminals per length of membrane (μm) of iNOS^-/- mice on the unlesioned and lesioned sides, respectively.