Background
Stroke is a leading cause of death and the major cause of long-lasting disabilities in industrialized countries ([
1]). Indeed, patients surviving stroke will carry a major risk for development of vascular and/or Alzheimer’s style dementia later in the life [
2‐
4]. This risk is particularly elevated in elderly population as various cellular processes are altered in aging. Aging hampers the normal physiology of the cell, leading to a metabolic dysfunction, oxidative stress, inflammation, and/or DNA damage [
5‐
7]. Growing evidence suggests that aging in the brain is associated with a progressive loss of immune homeostasis (a chronic low-level inflammation) leading to an overall increase in the pro-inflammatory cytokines including IL-1β, TNF-α [
7,
8]. In keeping with previous evidence, we recently demonstrated that processes associated with aging significantly affect microglia activation patterns and innate immune signaling after stroke in both female and male mice [
9,
10]. In particular, we observed a marked deregulation of Toll-like receptor 2 (TLR2) induction patterns in activated microglia followed by alterations in the innate immune downstream signaling events and larger infarctions [
9,
10]. However, how aging affects immune signaling in neurons and/or microglia/neurons crosstalk in response to ischemic injury remains unclear.
In a search for proteins that may affect microglia/neuron immune crosstalk in aging brain, we focused our study on transactive response (TAR) DNA binding protein 43 (TDP-43). Generally localized in the nucleus, TDP-43 belongs to the family of heterogeneous nuclear ribonuclear proteins that are highly conserved in different species [
11]. TDP-43 regulates gene expression by controlling several processes such as pre-mRNA splicing [
11], mRNA stabilization [
12], mRNA transport, and translation [
13]. TDP-43 has been identified as a major constituent of ubiquitinated nuclear and cytoplasmic inclusions in frontotemporal lobar degeneration [
14], ALS [
15] and Alzheimer’s disease [
16,
17]. Normally, localized in the nucleus, under pathological conditions TDP-43 forms insoluble ubiquitinated inclusions in which it is abnormally phosphorylated and cleaved to generate a 35 and a 25 kDa C-terminal fragments lacking the N-terminus nuclear localization signal [
15,
18]. In addition to processes associated with chronic neurodegeneration, increasing evidence suggests that deregulation of TDP-43 neurons may occur following brain injuries including single and repetitive traumatic brain injury (TBI) [
19,
20], while Uchino and colleagues recently reported presence of TDP-43-positive inclusions in aging brains [
21]. To date, the molecular mechanisms by which TDP-43 may induce neurodegeneration and neuronal death remain elusive. However, our previous work suggests that TDP-43 may serve as a modulator of inflammation, acting as co-activator of p65 NF-κB [
22]. Here, we hypothesized that gradual age-related accumulation of cytoplasmic TDP-43 may trigger activation of NF-κB pathogenic pathways, leading to a deregulation of innate immune response and thus increasing susceptibility of neurons to ischemic injury.
The current study was designed to identify and characterize the age-related expression patterns of TDP-43 in neurons and microglia and to evaluate its role as modulator of inflammation following ischemic injury. We report here an age-related increase and long-lasting mislocalization of TDP-43 after stroke. The observed accumulation of cytoplasmic TDP-43 was associated with an increase in microglial activation and innate immune signaling seen by in vivo bioluminescence imaging and immunofluorescence analysis. The presence of ubiquitinated TDP-43 aggregates and its cleaved TDP-35 and TDP-25 fragments was markedly increased in older, 12-month-old mice, which showed larger infarctions alongside with an increase in neuronal death. We next showed that increase and/or overexpression of the cytoplasmic TDP-43 drives the NF-κB response and further increase levels of pro-inflammatory markers and ischemic injury after stroke. Overall, our results suggest that TDP-43 may act as an age-related modulator of inflammation after stroke. Based on our results, we propose that therapies targeting cytoplasmic TDP-43 may have a potential to modulate post-ischemic inflammation and to protect dying neurons in the ischemic microenvironment. Of note, the post-mortem analysis of the brains autopsied at different time points after human stroke suggests the presence of TDP-43 immunoreactive structures localized in the cytoplasm of the neurons in periphery and the core region of the ischemic lesion.
Discussion
The work presented here provides an important in vivo evidence for a pathogenic role of TDP-43 in stroke. Based on the results presented in this study, we propose here that age-related deregulation of TDP-43 exacerbates inflammation and ischemic injury and may contribute to post-stroke neurodegenerative processes. By investigating TDP-43 expression patterns after stroke in young and 12-month-old mice, we showed (i) a marked increase and accumulation of TDP-43 in cytoplasmic compartment in neurons and microglia (ii) levels of mislocalized/cytoplasmic TDP-43 and its cleaved pathogenic fragments TDP-35 and TDP-25 were more elevated in aged mice, (iii) observed deregulation of TDP-43 was associated with an increase in post-stroke inflammation and larger infarctions, (iv) overexpression of TDP-43 further exacerbated ischemic injury and markedly enhances inflammation via activation of NF-κB, and (v) mislocalization of TDP-43 into cytoplasmic compartment occurred also in human stroke.
Cerebral ischemia is characterized by a marked acute and chronic inflammatory response. We and others have shown that post-stroke inflammation may have a marked chronic component that may last several months following an initial ischemic event and may contribute to development of the chronic brain injury [
23,
36]. However, the molecular mechanisms driving the long-lasting post-stroke inflammation, and potentially leading to a neurodegeneration, remain elusive. Based on the results described in the current study, we hypothesized that the age-related accumulation of TDP-43 in the cytoplasm (see Fig.
2) may drive chronic inflammation after stroke and thus contribute to ischemic injury.
To date, little is known about the role of TDP-43 in the pathogenesis of stroke. The alterations in TDP-43 expression in response to ischemic injury have been recently described by Kanazawa and colleagues in an acute rat model of ischemic injury [
37]. However, the study has been limited to a 24 h after stroke time period [
37]. Our findings are generally in agreement with the initial report. However, we observed a long-lasting (up to 30 days post MCAO) cytoplasmic accumulation/deregulation of TDP-43 after stroke (Fig.
1). In addition, our study extends the initial report in several ways. First, in keeping with recent report of the TDP-43 contribution in aging, we investigated the expression patterns of the TDP-43 after stroke in two different age groups (3 and 12 month old). Importantly, our results revealed markedly increased accumulation of the cytoplasmic TDP-43 in older mice. The fact that in older mice the cytoplasmic TSP-43 was present even at the baseline levels may have significantly affected the initial brain response to ischemic injury. Indeed, the levels of pathogenic TDP-35 and TDP-25 fragments after stroke were significantly increased in 12-month-old mice when compared to young animals, thus further suggesting a strong component of aging in TDP-43 mediated pathology. As previously mentioned, TDP-43 expression is normally restricted to nucleus and has a role in the regulation of gene transcription, mRNA splicing, mRNA stability, and transport. However, in pathological conditions, TDP-43 becomes mislocalized to the cytoplasm of both neurons and glial cells and cleaved by the caspases into 35 kDa and 25 kDa fragments to form potentially pathogenic aggregates [
38].
Here, we presume that cytoplasmic accumulation of TDP-43 during ischemic stroke occurred because of deregulation of its nuclear import. In fact, the C-terminal fragments of TDP-43 that are formed during aging/stress lack functional nuclear localization signal [
39]. Cytoplasmic accumulation of TDP-43 protein was reported in different neurodegenerative diseases [
37,
40] [
17]. For example, development of many age-related pathological and biochemical changes like formation of C-terminal TDP-43 fragments, TDP-43/ubiquitin aggregates, and neuroinflammation have been reported in mouse model of ALS [
41]. Indeed, in TDP-43 proteinopathies, dying neurons display the presence of ubiquitin inclusions as the ubiquitin-dependent protein degradation pathways were hampered [
42]. Another important feature of TDP-43 proteinopathies is a presence of phosphoTDP-43 aggregates. Evidence suggests that casein kinase 1 phosphorylates TDP-43 to form insoluble phosphoTDP-43 aggregates in most of the TDP-43 proteinopathies [
43]. To investigate the formation and presence of insoluble phosphoTDP-43 aggregates after ischemic injury and in aging, we collected urea-SDS insoluble fraction and performed immunoblots. Surprisingly, we did not detect phosphorylated TDP-43 aggregates in any of tested age groups after stroke, suggesting that TDP-43 pathology has distinct molecular signature after stroke when compared to chronic neurodegenerative TDP-43 proteinopathies.
Another hallmark of the brain response to ischemic injury and neurodegeneration is activation of glial cells. Indeed, previous studies reported TDP-43 cytoplasmic inclusions glial cells in the spinal cords of the ALS patients [
15] while in vitro studies using microglia and astrocyte culture exhibits TDP-43 mislocalization in induced neuroinflammatory conditions [
41]. Furthermore, our previous studies have demonstrated that binding of its N-terminal and RRM1 domains to p65, TDP-43 acts as co-activator of p65 NF-kB thus leading to enhanced activation of the NF-κB pathways [
22]. Importantly, NF-κB may interact with specific proteins and DNA sequences to trigger inflammation and ischemic injury-induced neuronal apoptosis [
44]. It was reported that NF-κB translocate to the nucleus from the cytoplasm in cerebral ischemic injury [
45]. Indeed, in the present study, we demonstrated a TDP-43-mediated deregulation and activation of NF-κB pathway. We measured the nuclear phosphorylation levels of P65 subunit of NF-κB after stroke in both 3- and 12-month-old mice. We observed that after the ischemic injury, the phospho-P65 subunit travels to the nucleus from the cytoplasm in both groups. There was an increased amount of phospho-P65 subunit in the nucleus of 12-month-old mice compared to 3-month-old mice 72 h after MCAO. In an additional
proof-
of-
concept experiments using TDP-43 A315T transgenic mice that overexpress TDP-43 in the cytoplasm, at the similar levels as 12-month-old WT mice, we observed comparable levels of nuclear phosphorylation levels of P65 subunit of NF-κB after stroke.
An important question here is how deregulation of TDP-43 affects neuronal survival and microglia-neuron crosstalk after ischemic injury. Substantial loss of nuclear TDP-43 in ischemic neurons may lead to nuclear dysfunction. Indeed, a series of in vitro experiments showed that TDP-43 depletion/silencing may cause disturbance in cell cycle leading to cell death [
46]. Additional evidence suggests that TDP-25 fragments formed in the cytoplasm of neurons can gain toxic functions and thus cause tissue damage albeit the mechanism remains unknown [
47]. Importantly, both of these neuropathological features were detected in ischemic brain tissue and affected neurons. Therefore, we examined whether the observed age-related TDP-43 deregulation may lead ultimately lead to more neuronal damage. Indeed, we found a significantly larger infarctions in the 12-month-old than 3-month-old mice 72 h after MCAO as well as an increased in cleaved caspase-3 levels in aging brain after 72 h post ischemia, while the analyses of the ischemic lesion in the context of TDP-43 cytoplasmic overexpression also showed direct correlation between TDP-43 expression levels and the size of the ischemic lesion. The important question that has been raised here is whether the observed deregulation in TDP-43 expression patterns is present in human stroke? Importantly, analyses of the post-mortem and post-stroke brain tissues revealed the presence of the cytoplasmic TDP-43 immunoreactive structures in human stroke resembling those observed after single brain trauma [
20]. Together, our results suggest that deregulation of TDP-43 may represent a converging pathogenic pathway that drives neuroinflammation following acute brain injuries and in chronic neurodegeneration.