Brain Inflammaging
Inflammatory processes can be initiated in the aging brain by a number of pathophysiologically relevant changes or events, which are assumed the drive CNS senescence. In a nondiseased subject, the progression of inflammaging may remain moderate for quite a number of years. However, as soon as additional proinflammatory alterations take place, both brain inflammation and resulting neurodegeneration can be substantially accelerated and aggravated. This concerns especially the formation of clinically relevant amounts of Ap peptides and tau hyperphosphorylation with its numerous consequences for peripheral mitochondrial function, energy supply and neuronal connectivity.
Under subclinical conditions of solely age-related low-grade brain inflammation, the following processes deserve particular attention. The inevitable age-associated remodeling of the immune system, which is a consequence of progressive thymic involution, life-long repeated exposure to foreign antigens and exhaustion of several subtypes of leukocytes [36, 37, 45-51] can lead to a proinflammatory phenotype that makes the brain more susceptible to inflammation initiation. In the case of such an immune risk profile (IRP), elevated levels of proinflammatory cytokines and other inflammatory mediators are typically observed.
Low-grade brain inflammation can be enhanced by various mechanisms. The inflammatory state may still remain in a subclinical range, but contribute to the progression of aging. Moreover, the same processes can be involved in the development of pathological changes of clinical relevance. One of the inflammation- promoting mechanisms is SASP, which has turned out to be a sustained source of inflammatory signals and elevated formation of free radicals [52-55]. SASP represents the potentially problematic side of an otherwise favorable mechanism that serves the mitotic arrest of DNA-damaged cells, a way of keeping these cells alive and metabolically active but preventing them from entering a neoplastic development. However, these arrested cells which display the so-called DNA damage response (DDR) steadily release proinflammatory cytokines. Importantly, SASP is a feature of many nonimmune cells, which, however, stimulate and attract immune cells, thereby contributing to an increased formation of inflammation-induced reactive oxygen and nitrogen species. With regard to the CNS, SASP has been shown to occur in aging astrocytes [56]. Moreover, SASP was demonstrated in endothelial cells from vessels outside the brain [57, 58], but the existence of this mechanism in CNS circulation system appears highly likely.
A process central to brain inflammation is microglia activation. Microglia- associated inflammation can occur at different degrees of severity and is relevant already at low-grade. This variability should be seen on the background of basal microglia activities. Contrary to earlier belief, these cells are not generally inactive nor do they behave as a uniformly responding entity. Even ramified microglia is known to be continuously active in terms of movement and safeguarding the CNS microenvironment [59]. Moreover, microglia activation can lead to different phenotypes, which may be either neurodestructive and phagocytically active or, alternately, primarily neuroprotective and also growth promoting [59]. Various mechanisms and signals can lead to the stimulation of microglia and are based on a complex network of, sometimes mutual, interactions with astrocytes and neurons. For instance, glutamate excitotoxicity can cause microglia activation [60-63], whereas, on the other hand, primary immune responses that activate microglia may initiate excitotoxicity [64-66]. A further complexity results from the frequently occurring coactivation of microglial cells and astrocytes. The involvement of astrocytes may not only further stimulate the microglia, e.g., by nitroxidative stress or inflammatory signals such as NO or SASP-associated cytokines, but also promote neuronal excitation, by impaired glutamate uptake and enhanced NO release with consequences to Ca2+ uptake and mitochondrial function [56, 65, 67, 68]. From a certain level on, proinflammatory processes in different cell types can lead to vicious cycles based on positive feedback loops between neurons, astrocytes and microglia. This may expand the grade and area of inflammation and become further aggravated by recruitment of other immune cells. A particular role can be attributed to the assembly of inflammasomes formed as different subtypes in neurons (NLRP1 and AIM2), astrocytes (NLRP2) and microglia (NLRP3), which are known to cause the release of proinflammatory cytokines such as IL-ip and IL-18 and to induce apop- totic or pyroptotic cell death [69]. As soon as cells are dying, they liberate histone H1, which acts as an additional pro-inflammatory signal to microglial cells and also as a chemoattractant [70].
Brain inflammation and neuronal overexcitation are connected in several ways, among which the enhanced formation of free radicals seems to be of particular importance. The role of oxidative stress has been demonstrated in this context in neurological diseases [71], but may be likewise applicable to respective subclinical changes. There are mainly three sources for enhanced production of reactive oxygen species, (i) leukocytes including microglial cells that are activated in the course of inflammation and primarily form superoxide by NADPH oxidase (Nox) and hypochlorite by myeloperoxidase, (ii) other Nox subforms expressed in astrocytes, neurons and endothelial cells, and (iii) mitochondria. In addition to leukocytes, neurons have also been shown to express myeloperoxidase and, although their contribution to oxidant formation under basal and subclinical inflammatory conditions is uncertain, neurons were reported to upregulate this enzyme in AD [72]. Mitochondria seem to play an additional crucial part in brain inflammation. Notably, several theories of aging have focused on these organelles, with regard to oxidant formation, but also under various additional aspects such as apoptotic cell death, mitophagy with its consequences to peripheral mitochondrial depletion, interconnections with metabolic sensing and the role of aging suppressors, as summarized elsewhere [37, 73, 74]. These considerations have been also specifically discussed in the context of the aging brain [36, 75]. Instead of repeating all these details in full length, only findings that are critical to brain inflammaging and neurodegeneration shall be addressed. Free radicals are generated in mitochondria by electron dissipation from the electron transport chain (ETC), a process that is reinforced by damage to the respira- somes as it particularly occurs under conditions of enhanced NO formation. Details on actions of the NO radical, •NO, on peroxynitrite-derived free radicals OOH, CO3", ^NO2) and nitrosating metabolites such as N2O3, other NO congeners or nitro- sothiols have been summarized elsewhere [37, 73, 76-78]. These changes can lead to a vicious cycle of free radical formation and may end up in apoptosis or mitoph- agy. Notably, some frequently discussed consequences may not be as relevant to damage and aging as formerly believed. First, a breakdown of the mitochondrial membrane potential (A?mt), as it occurs during a superoxide flash, does not necessarily result in an immediate initiation of apoptosis or mitophagy, but only does this after prolonged duration [36, 79]. Second, the crucial step of cardiolipin peroxidation is not mainly caused by free radicals directly, but is typically catalyzed by the peroxidase activity of the cytochrome c/cardiolipin complex [80-83]. However, decreases in mitochondrial levels of reduced glutathione (GSH), which can be caused by excess of free radicals, favor this process and have been shown to be counteracted by overexpression of the mitochondrial subform of glutathione peroxidase, GPx4 [84]. Therefore, free radicals may act upstream of cardiolipin peroxidation. Third, the damage of mitochondrial DNA (mtDNA) by free radicals has been overrated. Apart from the fact that mitochondrial chromosomes are not naked, but rather densely associated with proteins different from histones (cf. ref. [37]), mtDNA mutator mice showed an age-related accumulation of mitochondrial mutations, but no substantial increase in free radical formation [85]. Subforms of NADPH oxidase as another relevant source of free radicals are associated with processes of normal and pathological types of aging, with neuroinflammation and are subject to activation by various proinflammatory signals and factors known to induce neurodegeneration [86-92]. Accumulating evidence speaks for a crucial role of Nox isoenzymes in oxidative damage as a consequence of microglia activation und also directs attention to the multiple links between microglia, astrocytes and neurons in aging and neurodegeneration.
