RPE phagocytic function declines in age-related macular degeneration and is rescued by human umbilical tissue derived cells

Abstract

Background: Age-related macular degeneration (AMD) is a leading cause of blindness among the elderly characterized by retinal pigment epithelium (RPE) degeneration with accumulation of abnormal intracellular deposits (lipofuscin) and photoreceptor death. RPE is vital for the retina and integrity of photoreceptors through its phagocytic function which is closely linked to formation of lipofuscin through daily phagocytosis of discarded photoreceptor outer segments (POS). Although phagocytosis has been implicated in AMD, it has not been directly shown to be altered in AMD. RPE phagocytic defect was previously shown to be rescued by subretinal injection of human umbilical tissue derived cells (hUTC) in a rodent model of retinal degeneration (RCS rat) through receptor tyrosine kinase (RTK) ligands and bridge molecules. Here, we examined RPE phagocytic function directly in the RPE from AMD patients and the ability and mechanisms of hUTC to affect phagocytosis in the human RPE.

Methods: Human RPE was isolated from the post-mortem eyes of normal and AMD-affected subjects and cultured. RPE phagocytic function was measured in vitro using isolated POS. The effects of hUTC conditioned media, recombinant RTK ligands brain-derived neurotrophic factor (BDNF), hepatocyte growth factor (HGF), and glial cell-derived neurotrophic factor (GDNF), as well as bridge molecules milk-fat-globule-EGF-factor 8 (MFG-E8), thrombospondin (TSP)-1, and TSP-2 on phagocytosis were also examined in phagocytosis assays using isolated POS. RNA was isolated from normal and AMD RPE treated with hUTC conditioned media and subjected to transcriptome profiling by RNA-Seq and computational analyses.

Results: RPE phagocytosis, while showing a moderate decline with age, was significantly reduced in AMD RPE, more than expected for age. hUTC conditioned media stimulated phagocytosis in the normal human RPE and significantly rescued the phagocytic dysfunction in the AMD RPE. RTK ligands and bridge molecules duplicated the rescue effect. Moreover, multiple molecular pathways involving phagocytosis, apoptosis, oxidative stress, inflammation, immune activation, and cholesterol transport were affected by hUTC in the RPE.

Conclusions: We demonstrated for the first time RPE phagocytic dysfunction in AMD, highlighting its likely importance in AMD, and the ability of hUTC to correct this dysfunction, providing insights into the therapeutic potential of hUTC for AMD.

Keywords: Age-related macular degeneration; Bridge molecules; Cell therapy; Human umbilical tissue derived cells; Phagocytosis; Receptor tyrosine kinase; Retinal pigment epithelium.

Fig. 1

Phagocytosis level in RPE from human eyes of donors of different ages. RPE cells were isolated from eyes of 6 normal donors (no significant ocular medical history) with age 31, 39, 59, 61, 71 and 79, and 4 AMD donors with age 65, 84, 86, and 88 and cultured as described in “Methods”. For phagocytosis assay, 5 × 10⁴ human RPE cells were plated in a 24-well plate, maintained in MEM containing 20% (v/v) FBS for at least 6 days, then in MEM with 5% FBS (v/v) before the assay. Circle shown for normal age 31 represents 10 counted numbers of internalized OS from 10 representative fields from one experiment. The circles for all the other age groups represent counted numbers of internalized OS from 10 representative fields per experiment from three independent experiments for each age group. A moderate negative (a) and weak negative (b) correlation between phagocytosis level and age was shown in the normal and AMD RPE, respectively

Fig. 2

Phagocytosis level in RPE from eyes of AMD donors. RPE cells were isolated from eyes of 4 AMD donors with age 65, 84, 86, and 88, and cultured as described in “Methods”. For best possible age matching RPE isolated from 3 normal donors without ocular diseases of age 61, 71 and 79 were used as normal control. 5 × 10⁴ human RPE cells are plated in a 24-well plate in MEM containing 20% (v/v) FBS for a week followed by medium change to MEM with 5% (v/v) FBS and subjected to phagocytosis assay. Normal RPE cells from eyes of donors without ocular diseases, AMD RPE cells from eyes of donors with AMD. Data represent the mean ± SEM (n = 16). ****P < 0.0001. Representative images of RPE containing fluorescent ingested POS are shown (scale bar = 10 µm)

Fig. 3

Effect of hUTC conditioned medium on phagocytosis in RPE from aged normal and AMD donors. hUTC conditioned medium (CM) was prepared as described in “Methods”. a, b RPE cells were isolated from eyes of 3 normal donors without ocular diseases at age 61, 71, and 79, and b from 4 AMD donors at age 65, 84, 86, and 88, respectively, and cultured before experiments. 5 × 10⁴ human RPE cells were plated in a 24-well plate in MEM containing 20% (v/v) FBS for a week. For untreated cells, medium was changed to MEM containing 5% (v/v) FBS. The cells to be treated were incubated with hUTC CM for 24 h and then subjected to phagocytosis assay in the presence of hUTC CM. N RPE cells from eyes of donors without ocular diseases, AMD RPE cells from eyes of donors with AMD, CM conditioned medium. Data represent the mean ± SEM (n = 26). ****P < 0.0001. Representative images of RPE containing fluorescent ingested POS are shown (scale bar = 10 µm)

Fig. 4

Effect of RTK ligands on phagocytosis in human RPE from donor eyes with AMD. The human RPE cells isolated from eyes of 4 AMD donors with age 65, 84, 86, and 88 were incubated with recombinant human BDNF (a), HGF (b), or GDNF (c) for 24 h, and then subjected to phagocytosis assay. The human RPE incubated with the hUTC CM was used as a positive control for the assay. Normal RPE cells from eyes of donors without ocular diseases, AMD RPE cells from eyes of donors with AMD, CM conditioned medium. Data represent the mean ± SEM (n = 4). ****P < 0.0001. Representative images of RPE containing fluorescent ingested POS are shown (scale bar = 10 µm)

Fig. 5

Effect of bridge molecules on phagocytosis in human RPE from donor eyes with AMD. The human RPE cells isolated from eyes of 4 AMD donors with age 65, 84, 86, and 88 were cultured as described in “Methods”. The photoreceptor OS were incubated with recombinant human MFG-E8 (a), TSP-1 (b), or TSP-2 (c) for 24 h and then fed to the RPE cells for phagocytosis assay in the absence of the hUTC CM. The POS pre-incubated with the hUTC CM was used as a positive control for the assay. Normal RPE cells from eyes of donors without ocular diseases, AMD RPE cells from eyes of donors with AMD, CM conditioned medium. Data represent the mean ± SEM (n = 4). ****P < 0.0001, ***P < 0.001, **P < 0.01, ns not significant. Representative images of RPE containing fluorescent ingested POS are shown (scale bar = 10 µm)

Fig. 6

Effect of hUTC conditioned medium on gene expression in human RPE cells. a Gene expression changes after 24-h hUTC CM treatment in human normal (X axis) and AMD (Y axis) RPE cells. The changes in gene expression between the two cell types are significantly correlated (P value = 0). b Enrichment of genes modulated by hUTC CM treatment in molecular and cellular functions (adjusted P value < 0.05). 1811 genes significantly upregulated or downregulated (fold change > 2 and adjusted P value < 0.05) in response to hUTC CM treatment were enriched in 19 molecular and cellular functions