Systemic Factors Mediate Reversible Age-Associated Brain Dysfunction

Introduction

Aging-associated loss of function in mammals seems to involve a set of defined, potentially reversible changes in many tissues and organs, including the brain. Cognitive function/memory and neurogenesis in the subventricular zone (SVZ) of the lateral vesicles and subgranular zones (SGZ) of the hippocampus diminish with increasing age. So-called “heterochronic parabiosis,” in which the circulatory systems of old and young mice are interconnected, have been used to demonstrate several key factors whose altered expression leads to diminished cognitive function with age.

The most important identified brain factors so far include C-C motif chemokine 11 (CCL11) and growth/differentiation factor 11 (GDF11). CCL11 increases in the plasma and cerebrospinal fluid (CSF) of aging mice and humans and causes impaired learning and memory, at least in young mice—likely a conserved effect. ¹ The plasma concentration of GDF11 decreases with age, and treatment of old mice with GDF11, known to promote endothelial cell function and vasculature integrity, ² increases cognition and neurogenesis in old mice. So far these two key effectors appear to exert their effects on the brain by affecting supportive or auxiliary tissue. CCL11 is an immune/inflammatory regulator, and GDF11 maintains the cerebral vasculature. Interestingly, GDF11 has also been shown to play a potential rejuvenative role in heart ^3,4 and skeletal muscle. ⁵ Other unknown plasma-borne factors in young blood lead to phosphorylation and activation of the cAMP response element-binding protein (CREB), which contributes to partial rejuvenation of cognitive activity in old mice. ⁶ An emerging theme emerges from this pioneering work—supporting tissue, such as blood vessels and the immune system, plays a key role in the aging of diverse organs.

Type I Interferon Genes Are Induced in the Choroid Plexus of Old Mice and People, Adversely Impacting Cognitive Function

In an important paper that extends understanding of the connection between the immune and central nervous systems, Baruch et al. show that abnormal induction of type I interferon (IFN-I) signaling in the choroid plexus (CP) is linked to diminished cognition and reduced neurogenesis. ⁷ IFN-I signaling is mediated by IFNs-α and –βbeta. The CP, a monolayer of epithelial cells, produces the CSF and forms the brain–CSF barrier (B-CSF-B). The B-CSF-B integrates signals from the brain and circulation by controlling translocation of macromolecules between the circulation and the brain.

Baruch et al. used RNA sequencing to investigate differences between organs and tissues from young (3 months old) and old (22 months old) mice. All tissues examined showed aging-related changes, but only the CP showed an increase in genes corresponding to IFN-I signaling (irf7, ifn-beta, ifit1). ⁷ One interpretation of this result might be that the old mice have had chronic viral brain infections, because IFN-I is known to be induced by viral infection. However, the authors hypothesize that the increased IFN-I gene expression levels are associated with aging. To support this idea, they showed that different groups of mice in different animal centers all show the increased IFN-I response. At the same time, further investigation revealed a decrease in expression of genes involved in the type II interferon (IFN-II) response (icamI, cxcl10, ccl17). These data argue against a localized viral infection inducing IFN-I just in the investigators' mice, but by themselves do not rule out the possibility that older mice could have a greater propensity for viral infections in the brain and CP. To strengthen their argument, Baruch et al. also stained brain sections from humans, who died without apparent brain disease, for expression of IFN-I proteins. They observed an increase in an IFN-I–associated protein expression pattern in the CP of old people. ⁷

To test whether factors in the circulation were involved in the altered expression of IFN-I and IFN-II, heterochronic parabiosis experiments, in which the circulatory systems of old and young mice were surgically joined, were performed. Such experiments have shown previously that the blood of young mice increases cognition and neurogenesis of old partners, and that the blood of old mice reduces neurogenesis and cognition in young parabionts. ^1,2,6 Here, the blood of old mice did not increase IFN-I signaling in the CP of young mice, suggesting that factors that increase IFN-I signaling are not present in the circulation. However, the blood from the old mice did diminish expression of IFN-II genes (mediated primarily by IFN-γ), which resulted in increased expression of CCL11, a chemokine whose expression has been reported to impair neuronal plasticity. Consistent with the possibility that the blood of young mice has factors to maintain IFN-II levels is that blood from young parabionts restored IFN-II levels in old partners with one exception. Young blood did not reduce increased levels of IFN-II–associated CCL11 expression in old parabionts. These data suggest that there is more to increased CCL11 expression in old animals than the simple proposed model ⁸ wherein CCL11 is induced by IL-4 in the absence of significant levels of IFN-γ.

Young blood did not reduce IFN-I expression to normal levels. Moreover, induction of IFN-I is unlikely to be mediated by blood-borne factors. To test the possibility that IFN-I induction is dependent on factors in the CSF, primary cultures of CP cells were exposed to CSF from old and young mice. Only CSF from old mice induced IFN-I–dependent genes, suggesting that CSF-borne factors were responsible. ⁷

The authors propose the simple idea that in old mice, increased IFN-I levels, especially IFN-β, induced by CSF-borne factors, causes reduced IFN-II expression levels, resulting in increased CCL11 expression and subsequent reduced neurogenesis and cognition (Fig. 1). They promote this model because it is known that chronic IFN-I signaling can interfere with IFN-II–mediated inflammation resolution in infection. ^{9 –11} However, the parabiosis experiments in which IFN-II is regulated independently of IFN-I in aged animals suggests that this model is too simple. Nevertheless, Baruch et al. demonstrate the importance of IFN-II expression by showing that mutant mice with deficient IFN-II signaling, both via a IFN-γ receptor knockout or Tbx21 null mice (which are deficient in CD4⁺ cell–derived IFN-γ) have memory and spatial learning defects as well as reduced neurogenesis. IFN-γ has been reported to be a trophic factor for the CP, ¹² and mice deficient in IFN-II signaling also had reduced CP function, including reduced leukocyte trafficking.

OPEN IN VIEWER

FIG. 1.

Young blood and anti-interferon receptor antibody partially restore cognition and neurogenesis in old mice. (A) In young mice, factors in young blood and stromal interferon-γ (IFN-γ) help maintain choroid plexus (CP) function. (B) During aging, type I interferon (IFN-I) gene expression increases, type II interferon (IFN-II) decreases, and the CP produces C-C motif chemokine 11 (CCL11), which results in reduced leukocyte trafficking, affecting brain function. Both neurogenesis and cognitive abilities decline. (C) In heterochronic parabiosis, young blood (red) from the young parabiont partially restores cognition and neurogenesis. IFN-II signaling is also restored. However, CCL11 and IFN-I expression remain elevated. (D) In old mice, injection with an antibody to interferon-α/β receptor (IFNAR) in the cerebrospinal fluid (CSF), blocked IFN-I signaling, reduced CCL11 levels to normal, and partially restored CP function and brain plasticity, resulting in increased neurogenesis and cognitive ability. Red, young blood; brown, old blood; dark red, CCL11; magenta, IFN-I signal regulators; blue, IFN-γ; CP, choroid plexus; CSF, cerebrospinal fluid; EL, ependymal cell layer; IFNAR, interferonα/β receptor; α-IFNAR, ; IGN-1/β,. Color images available online at www.liebertpub.com/rej

Regardless of their model for IFN-I/II interaction in the CP of old animals, it is possible that IFN-I itself may have detrimental effects on brain function and neurogenesis. Indeed, exposure of CP epithelial cells in culture decreased expression of the neurotrophic factors insulin-like growth factor 1 (IGF-1) and bone-derived neurotrophic factor (BDNF). Furthermore, when an inhibitory anti-interferon receptor antibody (interferon-α/β receptor [α-IFNAR]) is injected into the CSF of mice, levels of IGF-I and BDNF are increased and CCL11 levels are decreased to normal levels.

Baruch et al. then investigated effects on cognition and neurogenesis. However, they report experiments in which they were forced to select cognitively impaired mice. Apparently many mice were actually cognitively normal when tested with a novel location recognition task, wherein an object is moved and the amount of time spent with the moved object is evaluated 24 hr later (increased time indicates better memory and cognition). As hypothesized the anti-α-IFNAR antibody raised scores on the task and increased neurogenesis. As might be expected increased expression of interleukin-10 (IL-10), an anti-inflammatory cytokine, was observed as well as reduced inflammation/damage of astrocytes and microglia. ⁷ These results are consistent with a detrimental role for chronic IFN-I signaling on CP function and on brain function. It would have been interesting to compare the IFN-I and IFN-II expression levels in the cognitively normal old mice with the cognitively challenged mice to help prove a possible association. Also, Baruch et al. did not show that inhibiting IFN-I signaling restored normal IFN-II signaling, with the exception of CCL11 levels, which given the data in Baurch et al. are clearly not dependent on IFN-II alone.

It is important to recognize that the source of inflammatory IFN-I signaling in the CSF has not been identified. Discovery of the source of chronic signaling should be a priority. Moreover, there is a formal possibility that the increased IFN-I in old mice and people results from old mammals being more prone to chronic low-level viral infection in the brain/CSF than younger mammals. Perhaps the heterogeneity in cognitive ability in the old genetically identical mouse population is due to infection status rather than aging per se. The role of endogenous retroviruses should also be studied given recent work connecting these agent to autoimmune processes. ¹³ Of related interest is that retrotransposon activation, which might be expected to activate IFN-I response, has been reported to be associated with aging in several systems, including mice. ^{14 –18}

Medical Implications

Although Baruch et al. did not completely succeed in completely linking reciprocal IFN-I and IFN-II dysregulation to each other during aging, they make a strong case that both IFN-II and IFN-I signaling plays a role in diminished CP function and cognitive decline in aging. That each effect is reversible increases the possibility that successful rejuvenation may be achieved by simultaneously addressing each defect. That IFN-I plays a significant role in brain dysfunction might be predicted from genetic neurological diseases associated with elevated IFN-α expression, such as Aicardi-Goutières syndrome (AGS). ¹⁹

The work of Baruch et al. needs to verified and extended and more carefully combined with studies that have identified other non-immune factors that affect the aging brain, such as GDF11, which effects brain vasculature. Perhaps there are analogous factors to GDF11 that affect the structural integrity of the ventricular system and the subarachnoid space.

IFN-I dysfunction and CCL11 overexpression are not affected by the blood of young parabionts. An interesting implication is that factors from young blood will be not enough to achieve full rejuvenation of the brain, even after identification and purification, because other factors from other reservoirs must also be modulated. However, at least one biotech startup is pursuing anti-aging therapeutics based on injection of factors from the blood of young humans. Even if IFN-I dysregulation is shown to be a major source of age-associated brain dysfunction, it is unlikely that anti-α-IFNAR will be investigated or approved as an anti-aging treatment, given the problematic regulatory and industry environment for anti-aging therapies.

Conclusion

Increased IFN-I and decreased IFN-II signaling at the CP decrease function of the B-CSF-B, which in turn negatively affects brain function. Such dysregulation is exemplary of a class of aging-associated bio-molecular changes that are potentially reversible. Dysregulation in the form of over- or under-expression of cytokines, growth factors, hormones, and metabolites in the blood and fluids, as well as altered expression of other signaling molecules on cell surfaces critical for cell–cell interactions, may be responsible for many detrimental changes associated with aging. Aging-associated changes in cells that play a supporting role in tissue/organ function seem to be especially prominent. Each dysregulated biomolecule presents a potential target for rejuvenation.