Vascular and neuronal protection induced by the ocular administration of nerve growth factor in diabetic-induced rat encephalopathy

Abstract

Background: Based on our previous findings on the efficacy of ocular applied nerve growth factor as eye drops (oNGF) to act in brain and counteract neuronal damage, we hypothesized that oNGF treatment might revert neuronal atrophy occurring in diabetic brain also by controlling neurotrophin system changes. The major NGF brain target areas, such as the septum and the hippocampus, were used as an experimental paradigma to test this hypothesis.

Methods: Bilateral oNGF treatment was performed twice a day for 2 weeks in full-blown streptozotocin-treated adult male rats. The forebrain distribution of cholinergic and endothelial cell markers and NGF receptors were studied by confocal microscopy. The septo-hippocampal content of NGF mature and precursor form and NGF receptors expression were also analyzed by Elisa and Western blot.

Results: oNGF treatment recovers the morphological alterations and the neuronal atrophy in septum and normalized the expression of mature and pro-NGF, as well as NGF receptors in the septum and hippocampus of diabetic rats. In addition, oNGF stimulated brain vascularization and up-regulated the TRKA receptor in vessel endothelium.

Conclusions: Our findings confirm that reduced availability of mature NGF and NGF signaling impairment favors vascular and neuronal alterations in diabetic septo-hippocampal areas and corroborate the ability of oNGF to act as a neuroprotective agent in brain.

Figure 1

(A–H) Cell atrophy/death and vascularization in the septum. (A–C) shows the distribution of ChAT immunopositive cells in the medial septum – indicated as gray in A – of CTR (B), STZ (C), and STZ+NGF (D). Bar = 100 μm. The visible reduction in stained cells in STZ and increased immunoreactivity in STZ+NGF are confirmed by the cell count reported in graph (E). The levels of DNA fragmentation reported in graph F (expressed as arbitrary OD units) indicate the effects of diabetes and oNGF treatment on cell mortality in the septum. The pictures at the bottom show the distribution of IB4‐positive vessels in the septum of STZ (G) and STZ+NGF (H) rats (Bar = 50 μm). The graph on the right side (I) shows the difference in area fraction occupied by vessels in the three experimental rat groups. *P < 0.05; **P < 0.01 statistically significant changes.

Figure 2

(A–D) NGF receptor localization in the septum. The TRKA and p75NTR single stain and co‐stain in the CTR, STZ and STZ+NGF septum are showed in A: note the reduction in cells expressing the NGF receptors in STZ and the recovering effects of oNGF on TRKA/p75NTR cell number and morphology. Magnification shows the different localization of NGF receptors on blood vessels and surrounding area. Specifically, single p75NTR and p75NTR/TRKA co‐stain were found in cells inside and around the blood vessel (B), while only TRKA stains the vasal endothelium and co‐localizes with CD34 (C and D). Bar = 20 μm.

Figure 3

Effects of diabetes and oNGF on the hippocampal histology and vasculature. Tolouidine blue stain reveals changes in the pyramidal layer of CA2/CA3 (A) of STZ rats when compared with STZ+NGF or healthy control rats (not showed). Many picnotic and ghost cells were localized in the DG area of STZ, while no apparent differences between CTR and STZ+NGF were observed (B). Representative pictures of the GS‐IB4 (red stain) labeled large and small blood vessels in the CA2/3 and DG of CTR, STZ and STZ+NGF rats. A reduced vessel density and a relative augment of small vessels are observable in both the hippocampal areas of STZ rats compared with control rats and diabetic rats receiving NGF, see C. The graphs in D, reporting the computer analysis of the vessel density (expressed as fraction area occupied by vessels, see Materials and Methods and Results section for details) confirm the microscopy observations and show differences between the experimental groups. **P < 0.01 statistically significant changes.

Figure 4

(A–D) Distribution of TRKA and p75NTR in the CA2/3 area of the hippocampus. The effects of diabetes and oNGF on the NGF receptor distribution in CA layers are showed in A. Compared with CTR, TRKA expression is reduced in stratum pyramidalis (SP), and p75NTR immunoreactivity is shifted from the pyramidalis to the stratum oriens (SO) in diabetic rats, while in STZ+NGF rats, an increase in TRKA immunofluorescence is observable in stratum oriens (SO), pyramidale (SP), and lucidum (SL). Cells co‐localizing TRKA and p75NTR are identifiable. TRKA‐positive blood vessels were scarcely found in STZ CA2/3 area but largely increased in STZ+NGF with respect to CTR. The magnification in B and C shows the TRKA‐positive vessels and the p75NTR expression on vessels surrounding cells or intralumen cells. The co‐expression of TRKA and CD34 is showed in Figure 5D. Bar = 20 μm.

Figure 5

(A–D) NGF receptors in Dentate gyrus of hippocampus. In A, the effects of diabetes and oNGF on TRKA and p75 expression in DG are shown. Note the increased expression of TRKA on vessels and the TRKA/P75‐positive cells in the hilus of STZ+NGF rats. The white frames indicate the area of magnifications at the bottom panel in which it is possible to observe the different distribution of NGF receptors in the granular layer of the three rat groups. Particulars of p75NTR immunoreactive cells localized closely to TRKA‐stained vessels in STZ+NGF DG are shown in Figure 6B and C. Co‐expression of TRKA and CD34 endothelial marker in vessels is demonstrated in Figure 6D. Bar = 20 μm.

Figure 6

Expression of TRKA and p75NTR in the septum. The figure on the left shows the representative Western blots for the two NGF receptors. Two bands at 140 and 110 kDa identifying the mature and immature form of TRKA, respectively, were recognized, while a single band was found for p75NTR. Band intensities were normalized to β‐actin, and the receptor expression was quantified from three or four independent experiments, and average levels in septum of CTR, STZ, and STZ+NGF rats are showed in the graphs. *P < 0.0%; **P < 0.01 statistically significant changes.

Figure 7

Expression of TRKA and p75NTR in the hippocampus. The effects of diabetes and oNGF on the expression levels of TRKA (110 and 140 kDa), and p75 receptors are visible in the representative Western blot (figure on upper left side), and graphs. *P < 0.05; **P < 0.01 statistically significant changes. Quantifications of three or four independent experiments were performed, and β‐actin expression was used to normalize the sample variability.

Figure 8

NGF content in the septum and hippocampus. The effects of diabetes and oNGF on the NGF and proNGF content in the septum and hippocampus were analyzed by ELISA and Western blot, respectively. *P < 0.05; **P < 0.01 statistically significant changes.