Effects of mesenchymal stromal cells and human recombinant Nerve Growth Factor delivered by bioengineered human corneal lenticule on an innovative model of diabetic retinopathy

Abstract

Introduction: Diabetic retinopathy (DR) is a microvascular complication of diabetes in which neurodegeneration has been recently identified as a driving force. In the last years, mesenchymal stromal cells (MSCs) and neurotrophins like Nerve Growth Factor (NGF), have garnered significant attention as innovative therapeutic approaches targeting DR-associated neurodegeneration. However, delivering neurotrophic factors directly in the eye remains a challenge. Hence, this study evaluated the effects of MSCs from human amniotic fluids (hAFSCs) and recombinant human NGF (rhNGF) delivered by human corneal lenticule (hCL) on a high glucose (HG) induced ex vivo model simulating the molecular mechanisms driving DR.

Methods: Porcine neuroretinal explants exposed to HG (25 mM for four days) were used to mimic DR ex vivo. hCLs collected from donors undergoing refractive surgery were decellularized using 0.1% sodium dodecyl sulfate and then bioengineered with hAFSCs, microparticles loaded with rhNGF (rhNGF-PLGA-MPs), or both simultaneously. Immunofluorescence (IF) and scanning electron microscopy (SEM) analyses were performed to confirm the hCLs bioengineering process. To assess the effects of hAFSCs and rhNGF, bioengineered hCLs were co-cultured with HG-treated neuroretinal explants and following four days RT-PCR and cytokine array experiments for inflammatory, oxidative, apoptotic, angiogenic and retinal cells markers were performed.

Results: Data revealed that HG-treated neuroretinal explants exhibit a characteristic DR-phenotype, including increased level of NF-kB, NOS2, NRF2 GFAP, VEGFA, Bax/Bcl2 ratio and decreased expression of TUBB3 and Rho. Then, the feasibility to bioengineer decellularized hCLs with hAFSCs and rhNGF was demonstrated. Interestingly, co-culturing hAFSCs- and rhNGF- bioengineered hCLs with HG-treated neuroretinal explants for four days significantly reduced the expression of inflammatory, oxidative, apoptotic, angiogenic and increased retinal markers.

Conclusion: Overall, we found for the first time that hAFSCs and rhNGF were able to modulate the molecular mechanisms involved in DR and that bioengineered hCLs represents a promising ocular drug delivery system of hAFSCs and rhNGF for eye diseases treatment. In addition, results demonstrated that porcine neuroretinal explants treated with HG is a useful model to reproduce ex vivo the DR pathophysiology.

Keywords: corneal lenticule; diabetic retinopathy; mesenchymal stromal cells; ocular delivery; rhNGF.

Figure 1

Molecular analyses of HG-treated porcine neuroretinal explants. mRNA levels of inflammatory/oxidative (NFκB, NOS2, NRF2, and GFAP), pro-angiogenic (VEGFA), apoptotic (Bax/Bcl-2), and retinal (TUBB3 and Rho) markers in porcine neuroretinal explants cultured for 4 days in the presence or absence of HG (25 mM). Results are shown as the mean ± error standard (SEM) (n≥5); *p<0.05; **p<0.01; ***p<0.001.

Figure 2

Immunofluorescence analyses of HG-treated porcine neuroretinal explants. Representative immunoflorescence (IF) images of CTRL and HG-treated neuroretinal explants and quantification of fluorescence intensity for GFAP (in red) and Rho (in green) markers. Nuclei were stained with DAPI (in blue) (n=3); *p<0.05.

Figure 3

Hystological analyses of HG-treated porcine neuroretinal explants. (A) Representative H&E images (10X and 20X) of CTRL and HG-treated neuroretinal explants and (B) quantification anlyses of cells density expressed as number of cells/mm2 of area for GCL and INL (n=3); *p<0.05; ***p<0.001.

Figure 4

Analyses of BioE_hCLs. Representative immunoflorescence (IF) and scanning electron microscope (SEM) images: hCLs (CTRL_hCL); decellularized hCLs with SDS 0.1% (Decell_hCL); Decell_hCL bioengineered with hAFSCs (BioE_hCL_A); Decell_hCL bioengineered with Fluo-PLGA-MPs or rhNGF-PLGA-MPs (BioE_hCL-B); Decell_hCL bioengineered with hAFSCs and Fluo-PLGA-MPs or rhNGF-PLGA-MPs (BioE_hCL_C).

Figure 5

Immunofluorescence analyses of NGF receptor and Fluo-MPs in porcine neuroretinal explants. Representative immunofluorescence (IF) images of (A) NGF receptors in porcine neuroretinal explants. Alexa Fluor 576 was used to stain NGFR p75 and Alexa Fluor 488 for TrkA, and (B) Fluo-PLGA-MPs were incorporated into the porcine neuroretina layers after release from BioE_hCL.

Figure 6

Molecular analyses of HG-treated porcine neuroretinal explants co-cultured with BioE_hCL. mRNA levels of inflammatory/oxidative (NF-κB, NOS2, NRF2, and GFAP), pro-angiogenic (VEGFA), apoptotic (Bax/Bcl-2), and retinal (TUBB3 and Rho) markers in HG-treated porcine neuroretinal explants co-cultured with each of the three BioE_hCL. Results are shown as the mean ± error standard (SEM) (n≥4); *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001.

Figure 7

Cytokine array analyses of HG-treated porcine neuroretinal explants co-cultured with BioE_hCL. Cytokine release (A) Heatmap and (B) scatter dot plot showing the release of pro-inflammatory (GM-CSF, IL-1β, and IL-1 ra), anti-inflammatory (IL-10), pro-angiogenic (PDGF-BB), and neuroprotective (INSULIN) cytokines in conditioned media of CTRL and HG-treated porcine neuroretinal explants co-cultured with each of the three BioE_hCL. (n ≥ 6 for pooled conditioned media).