The effect of deafness duration on neurotrophin gene therapy for spiral ganglion neuron protection

Abstract

A cochlear implant can restore hearing function by electrically exciting spiral ganglion neurons (SGNs) in the deaf cochlea. However, following deafness SGNs undergo progressive degeneration ultimately leading to their death. One significant cause of SGN degeneration is the loss of neurotrophic support that is normally provided by cells within the organ of Corti (OC). The administration of exogenous neurotrophins (NTs) can protect SGNs from degeneration but the effects are short-lived once the source of NTs has been exhausted. NT gene therapy, whereby cells within the cochlea are transfected with genes enabling them to produce NTs, is one strategy for providing a cellular source of NTs that may provide long-term support for SGNs. As the SGNs normally innervate sensory cells within the OC, targeting residual OC cells for gene therapy in the deaf cochlea may provide a source of NTs for SGN protection and targeted regrowth of their peripheral fibers. However, the continual degeneration of the OC over extended periods of deafness may deplete the cellular targets for NT gene therapy and hence limit the effectiveness of this method in preventing SGN loss. This study examined the effects of deafness duration on the efficacy of NT gene therapy in preventing SGN loss in guinea pigs that were systemically deafened with aminoglycosides. Adenoviral vectors containing green fluorescent protein (GFP) with or without genes for Brain Derived Neurotrophic Factor (BDNF) and Neurotrophin-3 (NT3) were injected into the scala media (SM) compartment of cochleae that had been deafened for one, four or eight weeks prior to the viral injection. The results showed that viral transfection of cells within the SM was still possible even after severe degeneration of the OC. Supporting cells (pillar and Deiters' cells), cells within the stria vascularis, the spiral ligament, endosteal cells lining the scala compartments and interdental cells in the spiral limbus were transfected. However, the level of transfection was remarkably lower following longer durations of deafness. There was a significant increase in SGN survival in the entire basal turn for cochleae that received NT gene therapy compared to the untreated contralateral control cochleae for the one week deaf group. In the four week deaf group significant SGN survival was observed in the lower basal turn only. There was no increase in SGN survival for the eight week deaf group in any cochlear region. These findings indicated that the efficacy of NT gene therapy diminished with increasing durations of deafness leading to reduced benefits in terms of SGN protection. Clinically, there remains a window of opportunity in which NT gene therapy can provide ongoing trophic support for SGNs.

Figure 1. Composite images of GFP expression in GP cochleae

(A) Three weeks following injection of Ad-GFP or Ad-GFP-NT into the SM of GPs which had been deaf for one week, GFP expression was predominantly detected in the lower and upper basal turn. Limited expression was also observed in the lower and upper middle turns. (B) Three weeks after injection of Ad-GFP or Ad-GFP-NT in GPs which had been deaf for four weeks, GFP expression was predominantly detected in the lower and upper basal turn but was also observed in the middle and apical regions of one GP. GFP was located in the spiral limbus and OC in basal region and also in the endosteal cells lining the perilymphatic spaces in the middle and apical regions. (C) Following injections into cochleae that had been deaf for eight weeks, expression was detected in the OC in the basal region, spiral limbus, endosteal cells and along Reissner's membrane. Limited expression in middle and apical regions was observed in one case.

Figure 2. GFP expression in the normal OC

A) GFP (green) was observed in the outer hair cells (OHC), Deiters' cells (DC) and the outer pillar cell (OP) in this example. The inner pillar cell (IP – stained red with phalloidin) was not transfected. B) In this example GFP expression was observed in the inner hair cell (IHC), the outer hair cells, the inner pillar cells and the interdental cells (IC) on the spiral limbus. The outer pillar cell (red) was not transfected. C) A surface preparation of the lower basal turn of a normal OC exhibiting widespread GFP expression within the OC. The peripheral fibres of the SGNs are labelled with anti-neurofilament (red). Scale bar 50 μm.

Figure 3. GFP expression and peripheral fiber responses in deaf GP cochlea

(A-C) Cochlear mid-modiolar sections from one, four and eight week deafened groups were stained with anti-calretinin (red) and phalloidin (blue) to identify transfected cells (green) in the OC. In each case, the location of the habenula perforata (HP) is shown. (A) An example of the degenerating OC from the lower basal turn of one week deafened GP showing GFP expression in an inner pillar cell (IP) around a collapsed tunnel of Corti. Degenerating supporting cells (SC, most likely Deiters' cells) are stained red. (B) A section from lower basal turn of a four week deafened GP showing GFP expression within an outer pillar cell (OP) in the degenerated OC. In this example degenerated SCs (red) were not transfected and the tectorial membrane (TM) can be observed. (C) A section from lower basal turn of an eight week deafened GP that exhibited total loss of the OC and no GFP expression in this region. However, interdental cells (IC) of the spiral limbus were transfected with GFP. (D) Surface preparation of the lower basal turn from a one week deafened GP showing Ad-GFP-NT transfection in SCs in the degenerating OC and in ICs of the spiral limbus. Neurofilament-labelled resprouting peripheral fibers (* red) can be seen to project towards the Ad-GFP-NT expressing cells (green) and also on the inner sulcus towards the spiral limbus. DAPI staining shows cell nuclei (blue) (E) A surface preparation of the lower basal turn from a four week deafened GP showing Ad-GFP-NT transfection in SCs in the degenerating OC. Although neurofilament-labelled resprouting peripheral fibers were evident (*) there was no tendency for them to project towards the Ad-GFP-NT transfected cells. (F) In this example of a surface preparation from the lower basal turn from the eight week deafened GP there were no GFP transfected cells in the region of the basilar membrane. GFP expression was observed in the ICs on the spiral limbus. Neurofilament-labelled resprouting peripheral fibers (*) were also observed. Scale bar 50 μm.

Figure 4. Degeneration of the organ of Corti following aminoglycoside deafening

Example micrographs of the organ of Corti in the normal cochlea and residual Corti structures in the one, four and eight week group from the right control cochlea. There was a progressive collapse of the organ of Corti with longer durations of deafness (scale bar 50μm). The area of the organ of Corti was measured and the average area (±SEM) is plotted for each experimental group. Organ of Corti area was significantly greater in the normal hearing cochleae compared to the other groups. In the deafened cochleae there was no difference in area between the one and four week group, however in the eight week group the area was significantly smaller compared to the shorter deafness durations.

Figure 5. SGN survival or loss in the lower basal turn following viral injections

GPs were deafened for one week, four weeks or eight weeks prior to the injection with Ad-GFP (GFP) or Ad-GFP-NT (NT). Three weeks following the viral injections the density of SGNs was determined. Example confocal images of SGNs in the lower basal turn are presented for the one week, four week and eight week groups that received Ad-GFP-NT in the treated (left) cochlea. The right cochlea served as the untreated control. Cochleae were stained for NF200 (red) and DAPI (blue); scale bar 50μm. Data is presented as the average (±SEM) of the difference in the SGN density between the injected and the non-injected cochlea. A positive average difference indicates greater SGN survival in the injected cochlea. Injection of Ad-NT following one week and four weeks of deafness, but not eight weeks of deafness, resulted in significant SGN survival in the lower basal turn compared to the contralateral non-injected cochlea.