Giant cell arteritis: immune and vascular aging as disease risk factors

Susceptibility for giant cell arteritis increases with chronological age, in parallel with age-related restructuring of the immune system and age-induced remodeling of the vascular wall. Immunosenescence results in shrinkage of the naïve T-cell pool, contraction of T-cell diversity, and impairment of innate immunity. Aging of immunocompetent cells forces the host to take alternative routes for protective immunity and confers risk for pathogenic immunity that causes chronic inflammatory tissue damage. Dwindling immunocompetence is particularly relevant as the aging host is forced to cope with an ever growing infectious load. Immunosenescence coincides with vascular aging during which the arterial wall undergoes dramatic structural changes and medium and large arteries lose their pliability and elasticity. On the molecular level, elastic fibers deteriorate and matrix proteins accumulate biochemical modifications. Thus, the aging process impacts the two major biologic systems that liaise to promote giant cell arteritis; the immune system and the vessel wall niche.


Introduction
Giant cell arteritis (GCA) is a granulomatous disease that displays tissue tropism for large and medium arteries manifesting as aortitis and vasculitis of the second to fi fth branches of the aorta [1]. Granulomatous lesions are typically localized in the wall layers of the aff ected arteries; extra-vascular GCA is rare, and likely represents a distinct entity. Th e arteritis is almost always combined with intense systemic infl ammation and a powerful acute phase response. Similar to other infl ammatory syn dromes, GCA is a complex disorder with multiple pathogenic factors. An instigator initiating the infl am matory process has not been identifi ed; however, over whelming evidence has accumulated that abnormalities in innate and adaptive immunity play a critical role in the initiation and the perpetuation of the vasculitis.
Several unique factors of GCA have been informative in dissecting its immunopathogenesis. Th e disease is characterized by a stringent tissue tropism; meaning that granulomatous wall infi ltrates typically appear in arteries of selected vascular beds. Th is pathogenic feature strongly suggests that vessel-wall-specifi c factors drive GCA. Dendritic cells (DCs), similar to skin-residing Langerhans cells, have been implicated in providing initial signals that break the immune protection of arterial walls [2,3]. Vascular DCs are an endogenous cell population in the arteries. Th ey defi ne the immunological identity of blood vessels by expressing a vessel-specifi c profi le of pattern recognition receptors [4]. Activated vascular DCs recruit adaptive immune responses, sustain them in the vascular wall and shape their architecture and functional direction. Th e nature of the adaptive immune response underlying the granulomatous reaction is now well understood [5]. In essence, a selected population of T lymphocytes travels to the arterial lesion, undergoes local proliferation and activation and produces pro-infl ammatory cytokines regulating the functions of macrophages, vascular smooth muscle cells and endothelial cells. T cells that accumulate in the granulo matous lesions are non-random; identical T cells have been isolated from the right and left temporal artery of the same patient [6]. Th e selectivity of the T cells involved in the disease process is highly suggestive for antigen as the ultimate driver of the infl ammation. Vasculitis T cells are functionally selected as well, and belong to either the Thelper (Th ) 1 or the Th 17 lineage [7]. In GCA patients, the two T-cell lineages are diff erentially responsive to cortico steroid therapy and appear to be responsible for distinct aspects of the vasculitis process [8]. Th e multiplicity of T-cell abnormalities suggests complexity in disease-driving signals, emphasizes the intricacy of the immune-mediated damage to the blood vessel and encourages multifaceted therapeutic approaches.
While substantial progress has been made in dissecting the T-cell biology of GCA, one critical disease factor has been less well examined. Th e single strongest risk factor is age, with incidence steadily rising in the seventh to eighth decades of life [9,10]. Th e disease essentially does not exist in individuals younger than 50 years of age. Progressive age could pose a risk to develop GCA through two major trajectories: age-related remodeling of the immune system, both the innate as well as the adaptive arm; and aging-imposed restructuring of the blood vessel wall. Takayasu's arteritis, a granulomatous vasculitis of the aorta and its primary branches, aff ects individuals prior to age 40 years and thus unfolds on the background of a distinct immune system and a distinct vascular microenvironment. Understanding the impact of the aging process on the immune system and the vascular system are necessary components of a more comprehensive concept of GCA, its risk factors and its manageable pathogenic elements.

Age and the immune system
Th e human immune system is designed to protect the host from infections, destroy malignant cells, regulate tissue healing and prevent immune responses to self. Over a lifetime, the immune system is challenged with a broad spectrum of infectious pathogens and responds with massive expansion of the cells required to destroy the invader and memorize the encounter for future protection. Th e constant need to protect the host thus imposes enormous proliferative stress upon the bonemarrow-derived cells that diff erentiate into immunecompetent cell types. With progressive age, production of new immune cells deteriorates. Th is is particularly relevant for T lymphocytes as involution of the thymus essentially abrogates regeneration of T cells in middle age. Th e question arises whether age-imposed changes in the pool of T lymphocytes increases the risk for GCA.
Th e term immunosenescence is now used for a series of changes that are typically encountered in the immune system of the elderly (Table 1) [11]. First and foremost, the production of naïve T cells declines. In the thymus, bone-marrow-derived precursor cells develop into mature T cells and are checked for autoreactivity, at which time auto-reactive T cells are eliminated. T cells with low-affi nity for self antigens are released into the periphery where they form the pool of naïve T cells. Once T cells encounter their specifi c antigen, they enter a diff erentiation process and turn into memory and eff ector T cells. During middle age, the process of thymic T-cell production declines markedly and the naïve T-cell pool shrinks. Within the T-cell compartment, enddiff erentiated memory and eff ector T cells replace naïve T cells. Chronic persistent infections place a particular strain on the immune system. Viral persistence, such as in cytomegalovirus infection, can drive chronic immune stimulation, which gradually exhausts the reserve of naïve T cells [12].
A prototypic change in end-diff erentiated T cells is the loss of the CD28 receptor from the cell surface. CD4 + CD28 -T cells were fi rst observed in patients with rheumatoid arthritis (RA) and their accumulation and clonality has given rise to the concept that RA patients have premature immune aging [13]. Other abnormalities typically encountered in the old are shifts in the distribution of CD4 and CD8 cells, a contraction in T-cell receptor diversity and loss of T-cell telomeres [14]. From studies testing vaccine responses in the elderly, it is known that only a fraction of over-65-year olds are capable of mounting protective immune responses and that the frequency of senescent CD8 + CD28 -T cells is an excellent predictor of non-responsiveness [15].
It is easy to understand that progressive age and contracting numbers of naïve T cells go hand in hand with declining immunity. It is more diffi cult to understand how the derangements of the T-cell system aff ect the ability to tolerate self. It is now clear that selftolerance is an active mechanism that depends on a wellregulated immune system. Th e strongest evidence that the aging immune system loses the capability to fi ne-tune immune responses and suppress those that would lead to self damage comes from the observation that the very old often have a state of chronic smoldering infl ammation [16]. Th is chronic, low-grade infl ammation is sometimes called infl amm-aging. Cytokines, such as IL-6, TNF-α, and IL-1β, are suspected to have an important role in sustaining chronic infl ammation ( Figure 1). Such cytokines mainly derive from the innate immune system, suggesting that immunosenescence aff ects both the innate and adaptive immune systems. Continuous immune stimulation, as well as release of pro-infl ammatory cytokines from other senescent cells, such as fi broblasts, have also been implicated as causative factors. Th us, ageinduced abnormalities in immune and non-immune cells may join forces to promote infl amm-aging ( Figure 1). Parallel changes in the tissue microenvironment, leaving tissue-residing cells with a primed phenotype, could easily generate tissues that are more susceptible to the action of systemically produced pro-infl ammatory cytokines. Th e end result would be that enhanced synthesis of pro-infl ammatory mediators, such as IL-6, TNF-α, and IL-1β, exacerbates relatively benign tissue infl ammation and switches it into a destructive event.
In summary, immunosenescence describes a series of abnormalities in the innate and adaptive immune system ( Table 1) that create a state of chronic infl ammation, impair anti-pathogen immune responses, increase susceptibility to malignancy and autoimmunity, decrease anti-vaccine responses and undermine wound healing. It is conceivable that such abnormalities also render individuals susceptible to vessel wall infl ammation.

Aging of T cells and dendritic cells
Th e central cellular players in GCA are vessel-embedded DCs that sense danger signals and T cells that mount an adaptive immune response at the tissue site [17]. Both cell types are subject to aging, directly connecting immunosenescence and GCA susceptibility. T-cell aging is mostly determined by the inability of the thymus to replenish new T cells while maintaining memory T-cell responses and keeping check of chronic persistent infections [18]. Interestingly, aging of CD4 and CD8 T cells proceeds at diff erent speeds [14]. CD8 T cells age faster, accumulate more CD28cells and display a higher degree of clonality. Th is may be a consequence of persistent viral infections, such as cytomegalovirus, which impose considerable stress upon CD8 T cells and induce restructuring of the immune system with progressive age. Little is known about the role of CD8 T cells in GCA, other than the fact that they account for only a minor part of the vasculitic infi ltrate. Reports in the 1990s drew attention to the blood CD8 compartment in GCA [19,20].
A study performed in Denmark with 227 active, untreated GCA patients and 227 age-and gendermatched controls revealed a signifi cant reduction in the CD8 population compared to controls (12% CD8 cells in patients versus 20% in age-matched controls). In several subsequent studies, decreased CD8 T-cell counts could not be confi rmed as a universal marker of GCA. Rather, analysis of T-cell clonality revealed that the CD4/CD8 ratio in both patients and controls correlated closely with the number of clonal T-cell populations [21,22].
CD4 T cells, the major component of the GCA wall lesions, age slower, but essentially undergo similar changes as CD8 T cells with progressive age [14]. Detailed studies have shown that the aging process in CD4 cells is significantly accelerated in RA, an autoimmune syndrome with chronic infl ammation of the synovial membrane and more rapid atherosclerosis. RA has emerged as a model system to study the interrelationship between chronic infl ammation and immune aging. Molecular defects described in RA patients may provide helpful insights into how the immune system ages overall and which particular pathways are relevant in the setting of chronic infl ammation. In RA patients, CD4 T-cell immunosenescence has been molecularly linked to a defect in repairing damaged DNA. Th is defect involves the telomeric ends as well as non-telomeric genomic DNA. T cells from RA patients fail to properly upregulate telomerase and thus are much less effi cient in protecting and repairing telomeres [23]. Also, T cells from RA patients have impaired upregulation of ataxia telangiectasia mutated (ATM), a kinase critically relevant in recognizing double-strand breaks of non-telomeric DNA [16,24]. ATM has been implicated in participating in maintaining genomic integrity of the telomeric ends as well. In T cells from RA patients, damaged DNA accumulates and DNA repair mechanisms are blunted. Forced overexpression of ATM restores the ability to repair fragmented DNA and extends the life span of RA T cells. RA T cells sense that they fail to fi x DNA damage and chronically activate cell internal stress pathways [24,25]. It is currently unknown how these molecular networks are regulated in patients who have GCA. However, systemic infl ammation may undermine DNA damage recognition and repair, and therefore accelerate aging of the T-cell compartment. Recognizing the role of the DNA repair machinery in determining immunosenescence provides access to possible interventions to counteract the aging process in the immune system ( Table 2). T cells orchestrate immune responses, both protective as well as harmful, by releasing cytokines. Recently, two cytokine-committed T-cell lineages have been implicated in GCA [7]. Th 1 cells produce IFN-γ, target macrophages and provide strong pro-infl ammatory function. In earlier studies, tissue IFN-γ levels have been correlated with the production of vascular endothelial growth factor and platelet-derived growth factor, molecules implicated in the neointimal response that leads to luminal stenosis [26,27]. Th 17 cells secrete IL-17, a cytokine that is known to regulate recruitment of neutrophils and provide protective immunity during early immune reactions, particularly indolent infections.
Both Th 1 and Th 17 cells participate in GCA lesions. Th 17 cells can be relatively easily suppressed with corticosteroids, whereas Th 1 cells are steroid-resistant and persist into the chronic phase of disease. Recent studies have refi ned the characterization of vasculitic T cells and have demonstrated that expression of the NOTCH1 receptor on T cells has an immediate role in the infl ammatory process [28].
Th 17 cells have also been implicated in atherosclerosis, a prototypic age-related syndrome in which mechanisms of immune and vascular aging interplay. Cells producing IFN-γ and IL-17 have been found in atherosclerotic plaques and are thought to create a pro-infl ammatory environment that causes abnormal function of vascular smooth muscle cells that eventually leads to plaque growth and instability [29]. It is possible that the aged immune system is biased towards the generation of Th 1 cells, instead of relying more on Th 17 cells.
Whether DCs age at their own pace or in tandem with T cells, their major partner in the immune system, is yet unresolved. DC functions amenable to the aging process include their production, their homing to peripheral tissues, their responsiveness to 'danger signals' , their migration from the periphery to central lymphoid organs and their direct interaction with T cells in the priming process. Studies of DC populations in GCA lesions have revealed that they are the critical antigen-presenting cells during early and late stages of disease [2]. Th ey display an unusual phenotype in that they express the CCR7 chemokine receptor. CCR7 expression on DCs is a sign of maturation and enables the DCs to follow a gradient of CCL19 and CCL21, two chemokines that direct the cells through lymph vessels into lymph nodes. In GCA lesions, DCs themselves produce high amounts of CCL19 and CCL21, triggering local trapping of the cells and preventing them from leaving the infl ammatory tissue niche [30].
Two studies in the human system and several studies in murine systems have approached the question of whether DC function changes with aging and how this aff ects immune-competence. In an infl uenza vaccine response study, Toll-like receptor (TLR) function of human DCs was assessed among 104 subjects. Older individuals were found to have DCs with lower TNF-α and IFN-α pro duction in response to TLR stimulation compared to younger subjects. Older individuals also had higher intracellular cytokine levels in the absence of TLR triggering compared to young [31]. In the murine model, young and old marrow-derived DCs were compared, and it has been shown that in vivo DC migration and CCR7 signal transduction are defective in aged DCs compared to their younger counterparts. Th ough the migratory defect could be overcome by increasing the numbers of DCs, the defect in CD8 T-cell priming persisted [32]. Th e fi ndings of these studies support the idea that DC aging impairs functional activity and that they may contribute to the observed infl amm-aging seen in the elderly. Other studies in both human and mice have shown that DCs, particularly plasmacytoid DCs, are susceptible to the aging process and this occurs both via decrease in population size and TLR function [33][34][35][36]. A molecular profi le of the age-related changes in T-cell and DC function that renders individuals susceptible to GCA or is caused by the infl ammatory activity in GCA has yet to be defi ned.

Vascular aging as a risk factor for giant cell arteritis
GCA displays a stringent tissue tropism for certain vascular territories. Such tissue tropism may result from selective availability of disease-relevant antigens. Alternatively, vessel wall cell populations may contribute to disease initiation and the composition of the vessel wall may not be homogeneous through the vascular tree. Studies performed over the past 5 years have lent strong support for the second alternative. Specifi cally, a population of immune-sensing cells, vascular DCs, have been identifi ed in the walls of medium and large arteries [17]. A population of vascular DCs is positioned at the mediaadventitia border of the human aorta and carotids, as well as subclavian, mesenteric, iliac and temporal arteries. Such vascular DCs provide immunosurveillance functions by sensing pathogen-derived molecular patterns (PAMPs). Th ey utilize a spectrum of TLRs for this sensing function [4]. Remarkably, studies of normal human arteries from distinct vascular beds have shown that the TLR profi le in each arterial territory is vesselspecifi c and unique [4]. Th us, human arteries contribute to monitoring immune danger and they do so in a very selective way. Granulomatous lesions contain highly activated DCs, which have been implicated in promoting activation of vasculitic T cells [2]. Atherosclerotic plaques, another type of vascular lesion, are also populated by DCs, which shape the infl ammatory milieu and activate T cells [37]. Whether such lesional DCs derive directly from the wall-residing DC population or whether they are recruited from the blood stream is currently not known. Vascular DCs have been implicated in initiating and sustaining mural infl ammation. Signals received by vascular DCs determine the microarchitecture of the evolving vascular infl ammation [38] and DCs are indispensable for the recruitment and the guidance of T cells that enter the vessel wall [3]. Studies in human artery severe combined immunodefi ciency (SCID) mouse chimeras have demonstrated that each artery displays a selective sensing function and that each artery interacts with T cells in a vessel-specifi c pattern. In essence, at least part of the tissue tropism for GCA results from vessel-specifi c immune functions. Immune-sensing functions of blood vessels may well be subject to age-related changes. It is not yet fully understood how the number, distribution, mobility and responsiveness to danger signals changes in vascular DCs as the host ages. However, substantial changes in the wall structure make it almost inevitable that vascular DCs in an 80-year old function distinctly from those in a 20-year old.
Functionality of the TLR-based sensing machinery in human dendritic cells is defi nitely susceptible to agerelated changes. Both myeloid and plasmacytoid DCs have been shown to be less responsive to TLR stimulation in older individuals [32]. While a decline in TLR responsiveness may suggest less pro-infl ammatory functions deriving from 'old' DCs, such 'old' DCs would also loose anti-infl ammatory functions. Tolerogenicity of DCs is equally dependent on proper sensing functions. Deteriora tion of tolerogenecity with advancing age would thus turn tolerogenic DCs into immune stimulator DCs (Figure 2). Th e relative abundance of systemically distributed pro-infl ammatory cytokines (for example, IL-6, TNF-α, IL-1) may indeed alter DC behavior and bias DCs towards a diff erent maturation program. It is well established that older individuals fail to generate protective immune responses with similar effi ciency as younger counterparts [15,39]. While declining T-cell immunity has a central role in immunosenescence, DCs remain the major partners for T cells and are equally important in initiating and sustaining immune responses. Functional consequences for vascular DCs when chronically exposed to a milieu of infl ammation need to be explored.
Age-induced changes in blood vessel structure and function are not limited to wall-embedded immunecompetent DCs. From macroscopic and microscopic studies, it has long been known that the medial layer is prone to age-related degeneration [40,41] (Figure 3). Vascular smooth muscle cells decrease in number and function. Th e media becomes thinner and deposition of calcium is not unusual. Overall, the arterial wall can thicken because the intimal layer expands from one to several layers. Elasticity of the arteries, mediated by the elastic fi bers and membranes integrated into the media, deteriorates, clinically recognized as increased stiff ness [42]. Th e underlying mechanism relates to the fracture of elastic membranes and the fi zzing of elastic fi bers [43]. As life progresses, matrix proteins responsible for the facilitation of cell-cell interactions and the transmission of forces throughout the wall structure are biochemically modifi ed; changing structural features and transmural traffi cking patterns for cells and molecules. Matrix proteins also serve as reservoirs for molecules released from wall-residing cells that promote communication between diff erent layers. In essence, the tissue niche of the vessel wall is profoundly altered by the aging process. Biochemical changes associated with vascular aging may actually be accelerated by infl ammatory damage of wall structures. In particular, vasculitis typically leads to fragmentation of elastic membranes and destruction of cell-protective matrix layers.
One of the molecular pathways implicated in molecular aging relates to oxidative stress, a process incriminated in aging overall [44]. As humans age, protective mechanisms detoxifying oxygen radicals are impaired. As a consequence, reactive oxygen species are more plentiful in the aged host. Such highly reactive oxygen species, derivatives of metabolic processes in the cell, are able to induce multiple biochemical modifi cations of proteins as well as membrane-integrated lipids. Aging of the aorta is associated with accumulation of advanced glycation endproducts and Alcian-blue positive mucin within the medial layer [41]. Glycation end-products have so far not been reported in medium-sized arteries, such as the temporal artery. Finally, endothelial cells are subject to aging. As prototypic vascular components, they not only line the macrolumen of medium and large arteries, but also form the microvascular vasa vasorum tree. Many of the immune recognition processes involved in GCA are located in the adventitia and the adventitia-adjacent media [45]. Th is is the neighborhood dominated by the  vasa vasorum network. Disorganization of the vasa vasorum tree has been identifi ed as an early abnormality in atherosclerotic disease [46]. Vasa vasora provide the entrance for infl ammatory cells in GCA lesions. Whether they have additional immunoregulatory functions in the early and late steps of the vasculitis is currently unknown. Insuffi cient regeneration of endothelial layers by endothelial precursor cells has been suspected as an important disease mechanism in multiple vascular patho lo gies. Organ-specifi c precursor cell populations decline with advancing age. Here, the aging process would have profound consequences for vessel wall integrity and function.

Conclusion
Progressive age, the strongest of all risk factors for GCA, impacts the two major elements that interact to cause vascular infl ammation; the innate and adaptive arms of the immune system and the unique microenvironment of the vessel wall. Clinically, immune aging is characterized by the loss of protective immunity and the occurrence of host-injurious infl ammation. Failure to protect the host results from the decline of immune cell regeneration. As a consequence, the elderly accumulate an increasing load of pathogens. Cells that survive in the aging immune system are less well controlled, spontaneously release cytokines and have altered threshold settings in response to immune stimuli. GCA is a disease of uncontrolled DC and T-cell activation in a tissue site that is ordinarily inaccessible for immune responses. As the host ages, the tissue site itself undergoes numerous structural changes, summarized as vascular aging. Th is process brings with it changes in elastic fi bers, matrix proteins and remodeling of the medial and intimal layer, aff ecting cellular traffi c, turnover and survival. An attractive hypothesis is that the infectious load of the aging host triggers activation of vascular DCs that can no longer protect the structurally altered vascular wall, giving rise to tissue-injurious infl ammation. Studies are needed to precisely defi ne how the aging process aff ects GCA-susceptible blood vessels and how vascular DCs change functionally and molecularly as the host ages. The 'old' artery provides a distinct microenvironment that potentially increases the risk for the homing and the persistence of pro-infl ammatory immune cells and supplies a novel spectrum of neoantigens. Recent work has identifi ed at least two lineages of proinfl ammatory T cells contributing to GCA; Th 1 and Th 17 cells. It is currently unknown whether the process of immunosenescence aff ects Th 1 and Th 17 cells in a lineage-specifi c way or whether both T-cell types age in parallel. Th e molecular details of T-cell aging in GCA remain to be elucidated. In GCA patients, Th 1 and Th 17 cells diff er in therapeutic responsiveness and participation in early and chronic disease. Evidence suggests that the two lineages respond to distinct instigators. A combination of pathogen-derived and tissue-indigenous danger signals may result in the chronic granulomatous infl ammation that characterizes GCA. Th erapeutic approaches for GCA patients need to be tailored to the special needs of the aging host. Immunosuppression needs to be restricted to avoid further weakening of immunocompetence. DCs originate from the bone marrow and the development of DC-target therapies will require a much better understanding of their life cycle. Enhancing immune cell regeneration could help in clearing chronic infections. Prevention of vascular aging is a long-drawn process that should begin during middle age. Given similarities in the infl ammatory process that typifi es atherosclerosis and GCA, it is possible that managing standard vascular risk factors could function as adjuvant therapy in the treatment of large vessel vasculitis.