Dissecting the involvement of tropomyosin-related kinase A and p75 neurotrophin receptor signaling in NGF deficit-induced neurodegeneration

Abstract

NGF, the principal neurotrophic factor for basal forebrain cholinergic neurons (BFCNs), has been correlated to Alzheimer's disease (AD) because of the selective vulnerability of BFCNs in AD. These correlative links do not substantiate a comprehensive cause-effect mechanism connecting NGF deficit to overall AD neurodegeneration. A demonstration that neutralizing NGF activity could have consequences beyond a direct interference with the cholinergic system came from studies in the AD11 mouse model, in which the expression of a highly specific anti-NGF antibody determines a neurodegeneration that encompasses several features of human AD. Because the transgenic antibody binds to mature NGF much more strongly than to proNGF and prevents binding of mature NGF to the tropomyosin-related kinase A (TrkA) receptor and to p75 neurotrophin receptor (p75NTR), we postulated that neurodegeneration in AD11 mice is provoked by an imbalance of proNGF/NGF signaling and, consequently, of TrkA/p75NTR signaling. To test this hypothesis, in this study we characterize the phenotype of two lines of transgenic mice, one in which TrkA signaling is inhibited by neutralizing anti-TrkA antibodies and a second one in which anti-NGF mice were crossed to p75NTR(exonIII(-/-)) mice to abrogate p75NTR signaling. TrkA neutralization determines a strong cholinergic deficit and the appearance of beta-amyloid peptide (Abeta) but no tau-related pathology. In contrast, abrogating p75NTR signaling determines a full rescue of the cholinergic and Abeta phenotype of anti-NGF mice, but tau hyperphosphorylation is exacerbated. Thus, we demonstrate that inhibiting TrkA signaling activates Abeta accumulation and that different streams of AD neurodegeneration are related in complex ways to TrkA versus p75NTR signaling.

Fig. 1.

Cholinergic deficit in anti-TrkA TgMNAC13 transgenic mice. (A and B) ChAT immunostaining in the basal forebrain of (A) 2-mo-old WT mice, (B) 2-mo-old TgMNAC13 mice, and (C) 6-mo-old TgMNAC13 mice. (D) Quantification of ChAT-immunoreactive neurons in the basal forebrain of WT, AD10, and TgMNAC13 mice at 2, 6, and 15 mo of age. Bars represent mean ± SEM. *, P < 0.05 vs. WT mice; #, P < 0.05 vs. AD11 mice. (Scale bar: 200 μm.)

Fig. 2.

Aβ peptide staining and memory deficits in TgMNAC13 mice. (A–C) Immunoenzymatic reaction showing Aβ-immunoreactive accumulation (stained with anti-Aβ mAb 4G8) in the CA1 region of the hippocampus of 14-mo-old (A) AD11 and (B) TgMNAC13 mice. (C) Aβ-immunoreactive dystrophic dendrites were absent in WT mice. (D–I) Representative confocal images of neurons in the CA1 radial layer of AD11 (D–F) and TgMNAC13 (G–I) mice showing distribution of anti-Aβ– (red in D and G) and MAP-2–immunoreactive dendrites (green in E and H). The merge of the two figures shows that Aβ is found both intracellularly (yellow) and extracellularly (red) in both AD11 (F and J) and TgMNAC13 (I and K) mice. (L) Quantification of Aβ-immunoreactive clusters in WT, AD10, and TgMNAC13 mice at 2, 6, and 15 mo of age. (M) ORT in 2- and 6-mo-old TgMNAC13 mice. Results of the ORT test phase are represented with the discrimination index. WT mice of all ages and 2-mo-old AD11 mice discriminated between the unfamiliar object and the familiar one, whereas TgMNAC13 mice did not. At 6 mo of age, both AD11 and TgMNAC13 mice did not discriminate between the familiar and the unfamiliar object. In L and M: *P < 0.05 versus WT mice; #P < 0.05 versus AD11 mice. Bars represent mean ± SEM. (Scale bar: 20 μm in A–C, 25 μm in D–I, and 5 μm in J and K.)

Fig. 3.

Cholinergic and Aβ phenotype in anti-NGF AD10 × p75NTR^{exonIII(−/−)} (AD12) mice. (A–D) ChAT immunostaining in 2-mo-old (A) WT mice, (B) p75NTR^{exonIII(−/−)} mice, (C) AD10 mice, and (D) AD12 mice. (E) ChAT-immunoreactive cells in 6-mo-old AD12 mice. (F) Quantification of ChAT-immunoreactive neurons in the basal forebrain of WT, AD10, p75NTR^{exonIII(−/−)}, and AD12 mice at 2, 6, and 15 mo of age. (G–J) Immunostaining with mAb 4G8 for Aβ peptide in the hippocampus of 6-mo-old (G) WT, (H) p75NTR^{exonIII(−/−)}, (I) anti-NGF AD10, and (J) AD12 mice. (K) Quantification shows that Aβ-immunoreactive clusters are absent in AD12 mice at all ages. Bars represent mean ± SEM. (Scale bar: 200 μm in A–E and 20 μm in G–J.)

Fig. 4.

Hyperphosphorylated tau expression in 2-mo-old AD12 mice. Immunostaining for hyperphosphorylated tau with mAb AT8 in the retrosplenial cortex of (A) WT mice, (B) p75NTR^{exonIII(−/−)} mice, (C) AD10 mice, and (D) AD12 mice. In AD12 mice, staining also was observed in (E) parietal cortex, (F) hippocampus, (G) red nuclei, and (H) substantia nigra (pars reticulata). (I and J) Enlargements of D showing the distribution of phosphorylated tau in the cell body and dendrites of retrosplenial cortical neurons. (Scale bar: 20 μm in A–H and 50 μm in I.)

Fig. 5.

Tau phenotype in AD12 mice. (A) AT8 neuronal labeling in the cortex of 6-mo-old AD10 mice. (B) Microglial expression and faint neuronal labeling of phospho-tau (mAb AT8) in 6-mo-old AD12 mice. (C) mAb AT8-positive tufted astroglial cell in the hippocampal radial layer of a AD12 mouse. (D) Spheroidal swellings (arrows) and fragmentation of cellular processes in a subset AT8-labeled microglia. (E–G) Representative confocal images of brain sections of 6-mo-old AD12 mice after immunostaining with anti-phospho-tau (green in E) and an antibody against the microglial marker CD45 (red in F). The merge of the two figures (G) localizes phospho-tau in microglia cells (yellow labeling). (H–K) Representative merged confocal images of brain sections of 6-mo-old (H) WT, (I) AD10, (J) p75NTR^{exonIII(−/−)}, and (K) AD12 mice after immunostaining with mAb AT8 (anti-phospho-tau) (green) and antibodies against the astroglial marker GFAP (red). (K) In AD12, GFAP and tau colocalize. (L) Densitometric quantification of the normalized levels of phosphorylated tau from Western blot analysis. (M–P) Immunohistochemistry against the neuronal marker NeuN in the retrosplenial cortex of (M) WT, (N) p75NTR^exonIII (^−/−), (O) AD10, and (P) AD12 mice. (Q) Quantification of NeuN-immunoreactive neurons in 6-mo-old WT and AD12 mice. Bars represent mean ± SEM. (Scale bar: 16 μm in A–D and in 25 μm in E–K.)