Compensatory Cross-Modal Plasticity Persists After Sight Restoration

Abstract

Sensory deprivation prompts extensive structural and functional reorganizations of the cortex resulting in the occupation of space for the lost sense by the intact sensory systems. This process, known as cross-modal plasticity, has been widely studied in individuals with vision or hearing loss. However, little is known on the neuroplastic changes in restoring the deprived sense. Some reports consider the cross-modal functionality maladaptive to the return of the original sense, and others view this as a critical process in maintaining the neurons of the deprived sense active and operational. These controversial views have been challenged in both auditory and vision restoration reports for decades. Recently with the approval of Luxturna as the first retinal gene therapy (GT) drug to reverse blindness, there is a renewed interest for the crucial role of cross-modal plasticity on sight restoration. Employing a battery of task and resting state functional magnetic resonance imaging (rsfMRI), in comparison to a group of sighted controls, we tracked the functional changes in response to auditory and visual stimuli and at rest, in a group of patients with biallelic mutations in the RPE65 gene ("RPE65 patients") before and 3 years after GT. While the sighted controls did not present any evidence for auditory cross-modal plasticity, robust responses to the auditory stimuli were found in occipital cortex of the RPE65 patients overlapping visual responses and significantly elevated 3 years after GT. The rsfMRI results showed significant connectivity between the auditory and visual areas for both groups albeit attenuated in patients at baseline but enhanced 3 years after GT. Taken together, these findings demonstrate that (1) RPE65 patients present with an auditory cross-modal component; (2) visual and non-visual responses of the visual cortex are considerably enhanced after vision restoration; and (3) auditory cross-modal functions did not adversely affect the success of vision restitution. We hypothesize that following GT, to meet the demand for the newly established retinal signals, remaining or dormant visual neurons are revived or unmasked for greater participation. These neurons or a subset of these neurons respond to both the visual and non-visual demands and further strengthen connectivity between the auditory and visual cortices.

Keywords: RPE65 gene; auditory; cross-modal plasticity; functional magnetic resonance imaging; low vision; resting state functional connectivity; sight restoration.

FIGURE 1

Schematic block diagram of the auditory paradigm. The auditory task consisted of four active sound blocks interspersed with three 15-s rest blocks. Each sound block was 30 s long containing 10, 2,000 ms long sound pairs with 1,000 ms between the two sound pairs to allow for subject response. There were a total of 40 randomly selected sound pairs, from a list of 72 wave forms, presented in each experiment.

FIGURE 2

Schematic block diagram of the visual stimuli. The checkerboard paradigm consisted of 15 s active blocks of contrast-reversing (8 Hz) checkerboards interleaved with 15 s presentation of a blank screen as rest blocks (R). Active blocks were comprised of 3 blocks for each contrast that were randomly presented for high (H), medium (M), and low (L) contrasts interspersed with 9 rest blocks (Ashtari et al., 2011).

FIGURE 3

Functional magnetic resonance imaging response to auditory stimulation. The results from the task-based auditory fMRI are presented for (A) sighted controls, (B) RPE65 patients before gene therapy, and (C) RPE65 patients 3 years after GT. The FX effect group analyses were processed at FDR corrected p < 0.02 (q < 0.05), with an extent threshold of >100 mm², and the RFX analyses were performed at an uncorrected p < 0.007 with an extent threshold >100 mm². Sighted controls showed the expected cortical activations in and around the bilateral Heschl’s gyrus, the inferior frontal gyrus (IFG), and the somatosensory cortex and deactivations in the visual cortex and within the default mode network. Compared to sighted controls, baseline RPE65 patients had similar patterns of auditory cortical activation, yet significant cortical activations within the bilateral visual cortices. Three years after retinal gene therapy, RPE65 patients showed the similar cortical activation patterns that were significantly enhanced.

FIGURE 4

Pre– and post–gene therapy comparison for auditory and visual stimulations. Direct group comparisons of pre– and 3 years post–retinal gene therapy were performed using the FX group-level analysis in response to (A) auditory and (B,C) visual stimulations. Group results are presented at an FDR-corrected p < 0.002 and p < 0.005 for auditory and vision, respectively (q < 0.05). In response to the same auditory stimulation, RPE65 patients showed increased bilateral auditory activation within the auditory cortex, as well as significant level of cross-modal activations along the bilateral calcarine sulci for their 3-year time point. The group comparison of the visual stimulation results between the baseline and 3 years post–gene therapy time points showed significantly more activations in and around the visual cortex for both the left and right eyes (left > right) 3 years after retinal gene therapy.

FIGURE 5

Functional magnetic resonance imaging response to visual stimulation. The visual stimulation tasks were presented either to the left or right eye of (A) normal-sighted controls, (B) RPE65 patients before GT, and (C) RPE65 patients 3 years after GT. All task-based cortical activations were processed using FX analyses at FDR corrected p < 0.007 (q < 0.05), with an extent threshold of >100 mm² and RFX analyses at an uncorrected p < 0.007 with an extent threshold > 100 mm². Sighted controls showed significant bilateral cortical activations in the visual cortex. At baseline, RPE65 patients expressed asymmetric and attenuated hemispheric activations in the left eye because the left eye was untreated in seven of eight patients. However, visual stimulation of the right eye in RPE65 patients displayed nearly symmetrical bilateral activation of the visual cortex. Three years post–FO clinical trial when RPE65 patients’ contralateral eye was treated, symmetrical bilateral visual cortex activations were observed for both left and right eyes.

FIGURE 6

Functional connectivity of the primary auditory areas. Anatomical locations for both the left and right Brodmann primary auditory areas (BA41) (A). The functional connectivity group results for FX and RFX analysis are presented for (B) sighted controls, (C) RPE65 patients at baseline, and (D) RPE65 patients 3 years post–gene therapy. All fixed-effects group analyses were performed at FDR corrected p < 0.02 (q < 0.05), with an extent threshold of >100 mm², and all RFX analyses were performed at an uncorrected p < 0.05 with an extent threshold >100 mm². Sighted controls showed significant bilateral functional connectivity between auditory, sensorimotor, frontal, and visual cortices. At baseline, compared to controls and post–gene therapy time point, RPE65 patients expressed similar, yet attenuated connectivity patterns particularly along the bilateral visual cortices. Three years after receiving their gene therapy, RPE65 patients’ connectivity of the primary auditory areas to the visual cortex significantly increased, yet remained attenuated as compared to the observed connectivity of the sighted controls.

FIGURE 7

Functional connectivity of the primary visual areas. (A) Left and right Brodmann primary visual areas (BA17) as anatomical seed points for functional connectivity analyses. Results for the FX and RFX group analyses are shown for (B) sighted controls, (C) RPE65 patients at baseline, and (D) RPE65 patients 3 years post–gene therapy. All FX group analyses were performed at FDR-corrected p < 0.02 (q < 0.05), with an extent threshold of >100 mm², and all RFX analyses were performed at an uncorrected p < 0.04 with an extent threshold >100 mm². Sighted controls presented significant extended functional connectivity between the left and right BA17 areas and other brain regions, particularly for the primary auditory areas (circled yellow). Connectivity pattern for the RPE65 patients at baseline showed significant functional connectivity across the left and right visual cortex, but considerably reduced levels of connectivity between the primary visual and primary auditory cortices (circled yellow). However, 3 years after gene therapy, functional connectivity between the primary visual to primary auditory cortices significantly increased compared to baseline, but did not reach the same magnitude as observed in sighted controls.