Altered activation state of circulating neutrophils in patients with neovascular age-related macular degeneration

Age-related macular degeneration (AMD) is a chronic progressive disease of the aged macula [1]. In the early stages, the disease is clinically characterized by drusen, which are yellow deposits between Bruch’s membrane and the retinal pigment epithelium [1]. The late stages of AMD are characterized by localized atrophy of the retina or by choroidal neovascularizations (CNV) [1]. The latter instance is described as neovascular AMD due to its key feature where newly formed vessels of the choroid penetrate through Bruch’s membrane into the subretinal space [1]. Consequently, fluid and blood leak into the retina, irreversibly impairing vision and visual function [2]. The treatment is only able to keep vision at a stable level for some years, and neovascular AMD still remains the most common reason for irreversible vision loss in the developed world [2‐4].

The pathogenesis of neovascular AMD remains incompletely understood, but ageing and dysfunction of the immune system are believed to play a key role for the disease to develop in an aged macula [5‐7]. A current developing area of interest is how neutrophils play a role in disease development. This is particularly interesting since aging is the highest risk factor of developing neovascular AMD and aged neutrophils are characterized by changed surface expression and activity [5‐8]. Studies have found a higher neutrophil/lymphocyte ratio in patients with neovascular AMD [9‐11], and studies of donor eyes have shown infiltrating lipocalin-2-positive neutrophils at significantly higher levels in retina and the choroid in both early and late stages of AMD [12]. Lipocalin-2 is a protein expressed in neutrophils, and levels of intravitreal lipocalin-2 are significantly elevated in eyes with neovascular AMD [13].

Neutrophils are the most prominent granulocytes and are part of the innate immune system practicing granulocyte release and phagocytosis [14]. Following stimuli, the activated neutrophils adhere to the endothelial cells in the area of inflammation and migrate their way to the site of injury and infection. Neutrophilic action is mediated through an effective combination of cytotoxic granules, antimicrobial peptides, and neutrophil extracellular traps [15]. Circulating neutrophils contains myeloperoxidase (MPO), which can form a hypochlourous acid that is an efficient killer of pathogens. In Parkinson’s disease and Alzheimer’s disease, MPO is redistributed into the extracellular space where it mediates tissue damage [16, 17], which is a mechanism suspected of contributing to the pathogenesis of AMD. Accumulated MPO is a two-edged sword: it may be beneficial by clearing toxic retinal lipofuscin deposits, but may be harmful by causing lysosomal stress that results in cell death [18].

Several chemokines are suggested to play a role in recruiting monocytes/macrophages and neutrophils resulting in the formation of CNVs [19]. Zhou et al. studied laser-induced CNV on mice and found that neutrophils infiltrate the retinal tissue from the first day after stimulation and that neutrophil-depleted mice had significantly smaller CNV-response. [20] Lavalette et al. found that neutrophils infiltrated the choroid 10 h after laser stimulation expressing the proangionetic interleukin 1β [21]. Taken together, these findings support the theory that neutrophils may play an important part in the early CNV-response. In humans, we previously described that systemic levels of neutrophils correlate with CNV-lesion size in patients with neovascular AMD [22], which further supports a role for neutrophils in CNV-development in AMD.

Based on these findings, we hypothesized that systemic neutrophil expression and properties could be altered in patients with wet AMD. To investigate this further, we selected markers of interest representing key steps of neutrophil activity: activation, adhesion, and inflammation [23]. Neutrophil migration is a process, which involves activation following interaction between adhesion molecules on the neutrophils and their ligands on the endothelial cells. Cluster of differentiation (CD) 63 and CD66b are activation-molecules expressed on the surface of neutrophils after appropriate stimulation [24, 25]. CD162 (P-selectin glycoprotein ligand-1) and CD62L (P-selectin) are mediators of the first step of rolling, after which integrin molecules CD11a (lymphocyte function-associated antigen-1) and CD11b (macrophage associated antigen-1) participate [23, 24, 26]. The process of transmigration through the endothelial layer is mediated by CD54 and CD31 (platelet endothelial cell adhesion molecule-1) [27]. The degree of inflammatory activity of neutrophils can be studied by measuring the expression level of several markers. One such marker is the interleukin-1-receptor-2 (IL1-R2), which is the receptor of the highly proinflammatory cytokine interleukin-1 [28]. C-X-C chemokine receptor 2 (CXCR2) is also important in acute and chronic inflammation and is mainly regulated by interleukin-8 [29]. C-C chemokine receptor 5 (CCR5) is involved in resolution of inflammation [30]. The Duffy antigen receptor for chemokines (DARC) is suggested to play a role in inflammation since DARC seems to bind a large number of chemokines whereby it might have a protective role in preventing chemokine activation of neutrophils and inflammation [31].

In this study, we wished to study alterations of the innate immune system in neovascular AMD. We sampled blood from patients with neovascular AMD and compared them to that of aged-matched healthy control individuals. We did not include any participants who were actively smoking. Tobacco triggers acute inflammation mediated via Toll-like-receptors [32] and modulates the expression of pro-inflammatory cytokines and chemokines [33]. Also, tobacco smoking increases the risk of AMD significantly. The increase in risk is most pronounced in current smokers, but is also markedly higher in former smokers [34].

We recruited a total of 61 individuals, of which 58 provided blood sample for our neutrophil study. One of the healthy control individuals had a plasma CRP >15 mg/L (22 mg/L) and was excluded from analyses. In total, 28 healthy control individuals and 29 patients with neovascular AMD were included for analyses. Mean age was 77.2 (SD: 6.6) years and 79.4 (SD: 6.1) years respectively for patients and controls (P = 0.192; independent samples t-test). Slightly more patients (n = 20, 69%) were female when compared to controls (n = 12, 43%) (P = 0.047, χ²-test). The patients and the control groups were generally similar in their co-morbidities and lifestyle characteristics. White blood cell count, and leukocyte population percentages and counts also did not differ significantly between the groups. Increased plasma CRP was more likely in patients with neovascular AMD (odds ratio 3.7, P = 0.033) (Table 1).

We find that among non-smokers, patients with neovascular AMD had increased expression of the activity marker CD66b, decreased expression of the adhesion marker CD162, and lower expression of inflammation marker CXCR2. Our results suggest that neutrophils may have specific components that play a role in neovascular AMD.

Complement activation plays a key role in the pathogenesis of AMD. Histopathological studies have found complement components in eyes with drusen [41], which are also reflected in altered levels of systemic complement markers in patients with AMD [42, 43]. Interestingly, complement activation induces CD66b overexpression and drastically reduces the neutrophils’ phagocytic capacity [25]. In light of these findings, we hypothesize that CD66b overexpression on systemic neutrophils in patients with neovascular AMD may reflect an increased systemic level of complement components that in turn inhibit the phagocytic capacity of the neutrophils. Consequently, we speculate that the drusenoid macula may lack an appropriate neutrophil response to the increased inflammatory and angiogenic drive whereby CNV formation can proceed unhampered.

Several studies suggest that AMD is associated with immunosenescence and systemic presence of low-grade inflammation [36, 44, 45]. We confirm this association in CRP, which were higher in patients with neovascular AMD. On neutrophils, such inflammatory environments influence the expression of CD162 [46, 47]. CD162 is a type 1 membrane protein that is constitutively expressed on human neutrophils and able to interact with all 3 types of selectins: P-selectin on activated platelets and endothelial cells, E-selectin on endothelial cells, and L-selectin on leukocytes [26]. This interaction between CD162 and its ligands is the first adhesion step and leads to neutrophils rolling on the endothelium prior to adhesion. Severe systemic inflammation in humans causes rapid downregulation of CD162 on neutrophils [46]. Inflammation of lesser degree have similar impact on CD162, exemplified by one study of surgical stress after cardiopulmonary bypass surgery [47]. Based on these considerations, we speculate that low-grade inflammation in patients with AMD may cause CD162 downregulation on neutrophils that in turn are less adherent.

CXCR2 is a chemokine receptor that regulates neutrophil recruitment in inflammatory contexts. Its function in inflammation and tumor-related inflammatory activities is vital and CXCR2 modulation has been suggested for treatment [48‐50]. One study of CXCR2 deficient mice demonstrated that CXCR2 plays an important role for macrophage-dependent inflammatory response. In CXCR2 deficient mice, inflammatory responses were more excessive, more macrophages were recruited to sites of inflammation, and levels of anti- and pro-inflammatory cytokines were shifted towards relatively more pro-inflammatory levels [51]. Hence, CXCR2 controls the magnitude of macrophage response in inflammation. We and other groups have previously found that monocytes and macrophages may play a key role in AMD and particularly for CNV formation [22, 52‐55]. Experimental laser-induced lesions on mice retinae show that macrophages are important for the CNV formation and systemic depletion of macrophages lead to significantly lower lesion size [55]. In patients with neovascular AMD, monocyte levels are increased in the first 30 days of new CNV diagnosis [22]. In light of these findings, we hypothesize that a lower expression of CXCR2 and the aged and drusenoid macula may be an ill-matched couple that orchestrates a more excessive macrophage activity and pro-inflammatory environment where CNV formation can occur. Interestingly, rheumatoid arthritis which is associated with CXCR2 deficiency is also associated with developing AMD later in life [56, 57].

Limitations of this study should be noted when interpreting its results. Importantly, this was an observational, case-control study, which can only associate but not infer on causality. Thus, we cannot exclude that these findings may also reflect a post-CNV state in the blood. We can only speculate on causality. However, findings of previous studies suggesting that complement dysfunction, low-grade inflammation, and CXCR2 deficiency all comes prior to onset of AMD gives reasons to expect causality [45, 57, 58]. The exploratory approach in the study and the limited group sizes did not permit meaningful stratifications based on single nucleotide polymorphisms that could be interesting, e.g. in the complement system [58, 59] in light of our findings on CD66b. Our study design cannot determine whether the findings reflect immunological dysfunction that are seen in a broad range of retinal diseases or mechanisms that specifically lead to AMD. Future studies need to investigate other retinal and ophthalmological diseases to clarify such aspects.

In summary, we find that in non-smokers, patients with neovascular AMD present with neutrophils that have increased expression of the activity marker CD66b, decreased expression of the adhesion marker CD162, and lower expression of the resolution of inflammation marker CXCR2. Circulating neutrophils may play a role for neovascular AMD and experimental studies are warranted to fully clarify how and when these neutrophil dysfunctions contribute to disease development.