An Exploration of Cellular Signatures of Ageing, and the Effects of Age-Related Dysregulated Myeloid Circadian Rhythm in Retinal Disease.

Abstract

Age-related macular degeneration (AMD) is the most common cause of vision loss and blindness in the elderly population across the developed world. This disease is characterised by a gradual degeneration and death of photoreceptors leading to central vision loss. There is currently no effective pharmacological cure available to patients. This is due in part to the complicated nature of the disease and the lack of accurate in vivo and in vitro models of AMD. Ageing is the single greatest risk of AMD development; therefore, this study explores the link between ageing, development of cellular senescence and AMD, with a specific focus on the role of peripheral myeloid cells.

Although AMD is a multifactorial disease, ageing is the single greatest risk factor associated with the development of the disease. Ageing is driven by the accumulation of senescent cells, which are non-dividing but exhibit high metabolic activity. The accumulation of senescent cells has been implicated in the development of a number of age-related diseases including Alzheimers disease and cardiovascular disease. The exact role of senescence in the development of AMD remains unknown. However, the increased production of pro-inflammatory cytokines such as IL6 through the production of the senescence associated secretory phenotype are hypothesised to disrupt the homeostatic ocular environment, leading to an increased level of inflammation and oxidative stress in the retina. The increased inflammatory retinal environment is thought to trigger the death of photoreceptor and retinal pigment epithelial (RPE) cells, eventually leading to retinal degeneration.

At present, relatively little is known about the process of healthy ageing in the retina and the exact molecular changes that occur that tip the balance from a healthy aged retina to a retinal degeneration phenotype. Therefore, in the first part of this study, we investigated the link between ageing and the expression of several immune, stress response and senescent/cell cycle markers in both retinal and peripheral tissue using a model of naturally aged mice to establish a baseline for typical age-related changes in the retina and peripheral tissue.

We detected increased retinal and peripheral immune senescence in our naturally aged C57 (22-24 months) mice compared to the young (8-10 weeks) cohort of mice. As no one gold standard marker of senescence is available, this study employed use of a range of senescent markers that detect distinct hallmarks of senescence. We observed a significant increase in the levels of diverse senescent markers including Beta-galactosidase, the lipofuscin marker Sudan black B, cell cycle proteins 16, p19 and p21, DNA damage marker gamma-H2AX and oxidised nucleic acid marker in the retina and RPE of aged mice. Interestingly, expression of cell cycle markers appeared to be perivascular in the inner retina of aged mice but were not found to co-localise with endothelial cells. Further investigations into the expression of senescence in specific retina cell types, demonstrated that aged retinal pericytes and neurons express DNA damage markers, but retinal microglial cells were protected from the process of ageing by this measure. Thus, indicating that the impact of natural ageing differs between retinal cell types. The distinct ageing signature between retinal cell types was found to also occur between retinal tissues. Use of a large-scale qPCR array focusing on distinct pathways of ageing, oxidative stress, and inflammation, demonstrated that the aged neural retina exhibits a distinct gene signature compared to the RPE/Choroid. The gene signature of the aged RPE/Choroid aligned with the aged peripheral liver tissue more closely than the aged neural retina. We detected very few changes in the gene expression of the retina during the ageing process, indicating that the neural retina has a number of internal mechanisms in place to protect against potential damaging systemic or mesenchymal factors that may occur as part of the ageing process.

Having characterised and confirmed increased levels of senescence in a healthy aged retina, we next investigated the impact of healthy ageing on immunosenescence and effector function in peripheral immune cells. Use of the nuclear proliferation marker, Ki67 and the DNA damage marker gamma-H2AX demonstrated that the aged bone marrow and splenic lymphocytes and monocytes had an enhanced immunosenescent phenotype compared to young mice. Like previous reports, we observed that the natural ageing process impacted the effector function of peripheral immune cells. Stimulation with several pattern recognition receptor ligands resulted in an aberrant production of pro-inflammatory cytokines and chemokines in aged bone marrow derived macrophages (BMDMs) compared to young BMDMs. Finally, we observed that aged BMDMs showed a dysregulated expression and rhythm of key circadian rhythm genes.

The age-related association with senescence and AMD combined with numerous reports of the detection of peripheral monocytes in the retina of AMD patients lead us to hypothesise that infiltrating peripheral monocytes may have adapted a senescent phenotype. Previously, we had detected increased immunosenescence as part of the natural ageing process, therefore we next investigated if AMD disease phenotype would alter the levels of immunosenescence in peripheral immune cells. We detected no change in the levels of senescence in T cells between donor groups. However, peripheral natural killer (NK) cells and monocytes isolated from patients with AMD were found to have a heightened immunosenescent phenotype compared to age-matched, healthy donors. Enhanced immunosenescence has previously been reported to be linked with circadian rhythm dysfunction. Therefore, we next assessed the expression and rhythm of key circadian rhythm cells in peripheral monocyte derived macrophages (MDMs) isolated from patients with AMD and healthy, age matched and young donors. Our investigations demonstrated that MDMs isolated from patients with AMD showed increased circadian rhythm disruption compared to age-matched and young donors. Finally, recent studies have indicated that loss of circadian rhythm is associated with an increase frailty phenotype. Therefore, frailty index data gathered by the Irish Longitudinal Study on Ageing (TILDA) in Wave 1 in 2009 and Wave 5 in 2019 was analysed to assess the rate and development of frailty within the cohort over the 10 years. Patients with a diagnosis of AMD were identified to have an accelerated rate of frailty over the 10-year period compared to healthy, age-matched participants, and this was independent of confounding factors such as sex, age, smoking status and importantly vision. Therefore, this provides evidence to indicate that the development of AMD is systemic.

In the final portion of the study, we developed and utilised a novel mouse model and confirmed that the combination of the natural ageing process and heightened systemic inflammation was sufficient to induce a retinal degeneration phenotype. We detected significantly higher numbers of mononuclear phagocytes (MPs) at the RPE and in the sub-retinal space of aged Bmal1-/-LysMCre mice compared to age-matched control mice. The increased retinal inflammation phenotype detected in aged Bmal1-/-LysMCre mice was enhanced following the addition of a HCD for 8 weeks. Supplementation with the HCD in aged Bmal1-/-LysMCre mice was observed to induce an RPE atrophy-like phenotype, with RPE cells noted to have increased disorganisation and areas of RPE-loss.

In conclusion, the evidence in this thesis supports the hypothesis that the dysregulation of peripheral myeloid cells in combination with environmental stressors (HCD) are implicated in the development of AMD. Our data indicates that rather than being a disease local to the retina, AMD likely has a strong systemic component driving pathology.