Retinal tumorigenesis and neurodegeneration : strategies to promote tumor cell death and support retinal cell survival to preserve vision

Abstract

Our vision is central to our quality of life enabling us to interact with our environment and lead productive lives. The ability to see begins in the retina, one of the paramount features of the eye, as it takes on the task of converting light signals to nerve impulses that are relayed to the visual centers of the brain. Proper functioning of the retina is essential for our visual experience. In this thesis, two different aspects of retinal dysfunction are highlighted together with potential therapeutic interventions. One of these is the childhood cancer retinoblastoma, which is characterized by uncontrolled cell proliferation after the loss of the RB1 gene. And, the other is retinal neurodegeneration which is characterized by the dysfunction and death of the retinal cells. We targeted the pyrimidine ribonucleotide synthesis pathway as a possible strategy to prevent retinoblastoma cell growth and to promote cell death. In the case of retinal neurodegeneration, we explored the neuroprotective effect of human neural progenitor cell (hNPC) derived neurotrophic factors to support retinal cell survival.

The rationale to target pyrimidine ribonucleotide synthesis as a strategy to treat retinoblastoma cells was based on our findings in paper I. In this paper we discovered a novel dihydroorotate dehydrogenase (DHODH) inhibitor after screening a library of compounds for their ability to increase p53 transcription factor activity in two reporter cell lines, the ARN8 melanoma cell line and the T22 mouse fibroblasts. Amongst the hit compounds, we selected compound HZ00 because it was more active on the ARN8 cells than on the T22 fibroblasts and due to its favorable medicinal chemistry properties. Regarding the mechanism of action of HZ00 on the p53 pathway, we observed that it was able to induce p53 synthesis without affecting p53 mRNA levels. Moreover, HZ00 was able to kill cancer cells and also synergize with an inhibitor of p53 degradation both in vitro and in vivo. During our studies on the identification of the target for HZ00, we observed that short treatment times with HZ00 caused ARN8 cells to accumulate in S phase and long treatments led to an increase in the proportion of cells in SubG1. Moreover, the supplementation with an excess of uridine was able to prevent the cell death effect. Based on these and other facts, we eventually narrowed down the target for HZ00 to DHODH, and confirmed this finding with enzymatic assays using purified DHODH. A further search for more potent analogues of HZ00 led to the identification of HZ05. HZ05 was able to accumulate a number of different cancer cells in S phase, which was followed by increase in SubG1 levels. However, the U2OS cell line had a higher propensity to accumulate in S phase. Further analysis showed an increase in p53 expression in the S phase cells. When we pretreated the U2OS cells with HZ05 followed by nutlin-3a, cell death was observed. This indicated that DHODH inhibitors could sensitize cancer cells to p53 degradation inhibitors by accumulating them in S phase with high p53 levels.

Retinoblastoma is typically a TP53 wild-type tumor characterized by inactivation of the RB1 gene. This deficiency leads to the loss of the cell cycle’s major G1-S checkpoint protein Rb. In paper I, we described that DHODH inhibitors can activate p53 and also cause cancer cells to accumulate in S phase, and that this is eventually followed by death. Since retinoblastoma cells already have a dysfunctional G1-S checkpoint, the use of DHODH inhibitors seemed like a rational approach to promote retinoblastoma cell death. In paper II, we investigated the potential of the DHODH inhibitor brequinar to reduce retinoblastoma cell growth and promote cell death, both as a single agent and in combination with a nucleoside transport inhibitor, dipyridamole. Similar to the effects seen in paper I, we saw that brequinar as a single agent was able to accumulate retinoblastoma cells in S phase and to an extent, also cause an increase in cell death. However, the response was slow and required a treatment time of 6 days. When we treated the retinoblastoma cells with a combination of brequinar and dipyridamole, the cells responded with an S phase accumulation as early as 24 hours after treatment, with most of them driven towards cell death with increasing time. This synergistic effect was also seen with other DHODH and nucleoside transport inhibitors. Moreover, the combination treatment was effective in the presence of uridine at physiologic plasma concentrations. Further investigation showed activation of caspases 3 and 7 as well as an increased expression of cleaved PARP-1, indicating the onset of apoptosis. Additionally, the treatment of a p53 mutant retinoblastoma cell line also responded to brequinar and the combination treatment, suggesting that targeting pyrimidine ribonucleotide synthesis could be an attractive strategy to eliminate both p53 wild-type and p53 mutant retinoblastoma cells.

In paper III and IV, we investigated the neurodegenerative events in an in vivo and in vitro model of retinal neurodegeneration with an emphasis on photoreceptor degeneration, second order neuron remodeling and glia reactivity. Furthermore, in paper IV we assessed the neuroprotective potential of hNPC derived neurotrophic factors in porcine retinal explant cultures. In paper III, the in vivo pdgf-bret/ret mouse model showed severe vascular defects due to the detachment of pericytes from the vascular endothelium. Degenerative events were followed on postnatal day (P) 7, 10, 15 and 28. These events were quite evident at P15 and worsened by P28, and included photoreceptor cell death, shortening of the cone outer segments and synaptic disassembly in the outer plexiform layer (OPL). Rod bipolar cells underwent remodeling and the Müller cells showed increased expression of GFAP (glial fibrillary acidic protein). The microglia also changed to their reactive amoeboid-like phenotype. For the in vitro porcine retinal explant model in paper IV, photoreceptor death increased significantly by 3 days in vitro. This was associated to the loss of the cone outer segments, mislocalization of opsin and synaptic disassembly in the OPL. Furthermore, we observed a loss and remodeling of horizontal cells, as well as severe gliosis of the Müller cells. The hNPC cocultured explants were observed to maintain photoreceptor survival through preservation of the cone outer segments, better opsin trafficking and retaining synaptic integrity. However, Müller cell gliosis was only alleviated by a decreased density of GFAP immunoreactive Müller cells. Both the in vivo and in vitro model of neurodegeneration demonstrated the vulnerability of photoreceptors to different mechanisms of retinal injury. The hNPC derived neurotrophic factors had the potential to preserve photoreceptors in the porcine retinal explants, but were not able to completely eliminate Müller cell gliosis.