Background
Toll-like receptors (TLRs) are a family of innate immune system receptors involved in sensing and response to pathogen-associated molecular patterns (PAMPs) and endogenous ligands known as damage-associated molecular patterns (DAMPs) that are released upon cell damage and necrosis [
1]. Toll proteins were discovered in
Drosophila melanogaster [
2]. Later, a mammalian homologue for Toll was found, and TLR4, the receptor for lipopolysaccharide (LPS) present in Gram-negative bacteria, was identified [
3,
4]. TLRs are widely expressed in a diversity of mammalian immune and non-immune cells, and they are present in the brain, where their expression is not restricted to microglia [
5] but expands to astrocytes [
6], oligodendrocytes [
7], and neurons [
8]. The functional implication of TLR expression in neurons is not well understood yet. It has been proposed that TLR4 may contribute to neural plasticity and development in neurons [
9]. In addition, recent studies indicate that TLR4 expression is upregulated with normal aging [
10], suggesting an altered regulation of the innate immune response in aging that may be relevant in different neurodegenerative disorders such as Alzheimer’s disease (AD). AD, the most common form of dementia, is strongly associated to aging and is characterized by the gradual loss of memory and cognitive function. The causes for AD are still not well understood. However, it is well known that AD is associated to formation of amyloid plaques made up of amyloid β peptide (Aβ), mainly Aβ
1–42, derived from the altered metabolism of the amyloid precursor protein after being processed by β- and γ-secretases. AD is also linked to intracellular neurofibrillary tangles, composed of abnormally hyper phosphorylated tau protein [
11].
Interestingly, McGeer and McGeer proposed in the late 1980s that innate immunity had an important role in neurodegenerative diseases (revised in [
11,
12]). Although this theory was not well accepted initially, consensus is growing about the involvement of an inflammatory component in AD. Microglia is responsible for immunity in the brain and becomes activated by signals released by surrounding cells. In the AD brain, the sites of neuroinflammation are surrounding senile plaques, which show increased levels of pro-inflammatory factors, such as pro-inflammatory cytokines, complement components, and proteases [
12,
13]. Recently, the
tlr4 gene has emerged as a candidate susceptibility gene for AD. For example, a genetic study proposed that a polymorphism in TLR4 (Asp299Gly) may decrease the risk of AD independently of a polymorphism in apolipoprotein E, suggesting the involvement of the innate immunity in neurodegeneration in general, and of TLR4 in AD, in particular [
14,
15]. Consistently, AD brains show increased expression of TLR4 [
15]. Furthermore, this receptor plays an important role in microglial neurotoxicity, since LPS binding induces its activation, thus releasing toxic substances to neurons [
16]. Consequently, instead of counteracting the damage caused by pathogens, TLR4 activation may lead to increased damage due to the release of toxic factors such as nitric oxide and oxygen free radicals [
17]. In this manner, it appears that Aβ may sensitize microglia to stimulation by some TLR ligands like LPS [
18] since co-administration of Aβ and LPS increases activation of TLR4, leading to increased release of nitric oxide and tumor necrosis factor α [
19]. However, the possible interactions of LPS and Aβ on hippocampal neurons have not been assessed yet. In this work, we aimed at investigating the interplay between neuroinflammation and AD in the context of aging.
To accomplish the above goal, we have employed here aged cultures of rat hippocampal neurons that are considered a model of aging and/or senescence, since some of the changes occurring in the elderly in vivo are mimicked in neurons aged in vitro [
20,
21]. Our results show that rat hippocampal neurons express TLR4 and expression increases with time in culture consistently with in vivo aging. We also found that LPS increases cytosolic [Ca
2+] and promotes neuron cell death only in aged cultures. Treatment with AD-related oligomers of the amyloid β peptide (Aβo) further enhanced TLR4 expression, Ca
2+ responses induced by LPS and neuron cell death, suggesting the interplay between TLR4 and Aβo in neuron cell death associated to aging and AD.
Discussion
The present study reveals aging and/or senescence as a critical factor required for a marked LPS- and Aβ oligomer-induced neuron cell death as well as for TLR4 expression increase in rat hippocampal neurons. In addition, it discloses an exacerbated interplay between TLR4 and Aβ in the context of aging and/or senescence with consequences in neuronal damage. Our results show that the endotoxin associated to Gram-negative bacteria LPS, by acting on TLR4, increases [Ca2+]cyt and promotes cell death in aged cultures of rat hippocampal neurons, but not in young cultures of hippocampal neurons. Several evidences support this conclusion. First, CAY16014, an specific inhibitor of TLR4 activation, inhibits LPS-induced [Ca2+]cyt increase and LPS-induced neuron cell death. Second, LPS promotes neuron cell death only in aged cultures of hippocampal neurons that show Ca2+ responses to LPS and increased expression of TLR4. In fact, in young cultured neurons expressing low levels of TLR4, LPS did not increase [Ca2+] and neither promoted neuronal apoptosis. Thus, aging and/or senescence correlates with enhanced TLR4 expression underlying LPS-mediated, hippocampal neuron death.
We must stress that our results have been obtained in an in vitro model of aging and/or senescence that may not reflect entirely in vivo aging. Evidently, long-term cultured neurons do not undergo the complex processes and interactions involved in in vivo aging. For example, actual young and aged neurons are very different from the metabolic point of view and only limited evidence suggests that these differences remain between short- and long-term cultured hippocampal neurons. However, long-term cultured neurons are amenable for single-cell studies such as the ones performed in this study. In addition, aged cultures of rat hippocampal neurons display many of the hallmarks of aging in vivo, including accumulation of reactive oxygen species, increased oxidative damage of cell proteins, protein carbonyl formation lipofuscin granules, heterochromatic foci, activation of the Jun N-terminal protein kinase and p53/p21 pathways, gradual loss of cholesterol, and changes in Ca
2+ channel density and NMDA receptor subunits similar to those found in in vivo aging [
20,
21]. In addition, Letiembre et al. [
10] reported that several TLR mRNAs are upregulated in the mouse-aged brain including TLR4 and its co-receptor CD14, as well as TLR1/2/5/7, while in contrast, transcripts of TLR3/6/8 did not change or even decreased as in the case of
tlr9. Therefore, long-term cultures of rat hippocampal neurons may provide a reasonable as well as amenable model to investigating aging and/or senescence-related changes in Ca
2+ signaling in individual neurons. In addition, our results are consistent with increased expression of TLR4 in aged brain reported in vivo [
10].
Our results pose the question on whether age-related changes in TLR4 gene expression contribute to susceptibility to neurodegeneration in the elderly. Consistent with this possibility, it has been reported that activation of the central innate immune system may lead to exacerbated neuroinflammation and prolonged sickness behavior in response to LPS in aged mice compared to adult mice [
32]. In addition, we have recently reported that increased susceptibility to excitotoxicity and brain damage in aging are also reflected in in vitro aged cultures of rat hippocampal neurons that are much more sensitive to glutamate-induced neuron cell death than young neurons [
25]. This effect is induced by dramatic increases in the rise in [Ca
2+]
cyt induced by the glutamate receptor-agonist N-methyl
d-aspartate associated to changes in receptor subunits [
25] similar to those reported in in vivo aging [
20]. Therefore, the aged culture of rat hippocampal neurons may be a good model to assess aging and/or senescence-related changes in the susceptibility to brain damage induced by excitotoxicity and neuroinflammation induced by Ca
2+ signals that may contribute largely to age-related neurodegeneration. In this regard, it is relevant the mechanism by which LPS increases [Ca
2+]
cyt in hippocampal neurons. In macrophages, it is well established that intracellular Ca
2+ participates as a second messenger in TLR4-dependent signaling that increases [Ca
2+]
cyt by activating a vanilloid member of the transient receptor potential superfamily of cation channels (TRP channels). In fact, TRPV2 is involved in the LPS-induced Ca
2+ mobilization from intracellular Ca
2+ store and extracellular Ca
2+ [
33]. In sensory neurons, however, it has been recently reported that LPS exerts fast, membrane delimited, excitatory actions via TRPA1 [
34], another TRP cation channel that is critical for transducing environmental irritant stimuli into nociceptor activity. Surprisingly, the effects of LPS on nociceptors were independent of TLR4 activation [
34]. In hippocampal neurons, both LPS-induced [Ca
2+]
cyt rises and neuron cell death were prevented by TLR4 antagonist and the effects correlated with changes in expression of TLR4 suggesting that LPS-induced effects in aged hippocampal neurons depend on TLR4.
We show here that glial cells also show Ca2+ responses induced by LPS, sometimes in the form of [Ca2+]cyt oscillations preceding Ca2+ responses in neurons. These results invite speculation on whether neuronal Ca2+ responses to LPS may be secondary to activated glia. For example, TLR4 activation in astrocytes might promote release of glutamate or other molecules that indirectly affect Ca2+ currents in neurons. Further research is required to fully exploring this and other possibilities. In any case, direct responses to LPS are definitely involved as most often neuronal Ca2+ responses are observed in the absence of glial responses. In addition, TLR4 expression is observed in morphologically and immunologically identified neurons, LPS-induced Ca2+ responses in neurons correlated with increased TLR4 expression in aged neurons and are prevented by inhibitors of TLR4 activation.
It could be claimed that LPS effects in aging cultures may be mediated by excitotoxicity events related to changes in NMDA receptor expression. However, several evidences are against this possibility. First, LPS has no or minor effects on expression of NMDA receptor subunits. Second, Ca2+ responses to NMDA and LPS are dissociated. Thus, Ca2+ responses to LPS are missing in 9–10 DIV neurons that display large Ca2+ responses to NMDA. Finally, MK801, a NMDA receptor antagonist, did not inhibit Ca2+ responses to LPS at concentrations that inhibit largely Ca2+ responses to NMDA.
Neuroinflammation has been associated to neurodegeneration and AD pathology, particularly in the elderly. For example, healthy aged individuals are more likely to suffer profound memory impairments following a challenging life event such as a severe bacterial infection than younger counterparts [
35]. Since TLR4 can sense not only molecular patterns derived from bacteria but also from necrotic damage, these data support the hypothesis of a decisive role of aging in exacerbating TLR4-mediated effects. Our data discloses an exacerbated interplay between Aβo and TLR4 in context of aging with consequences in neuronal damage. Specifically, we show here that treatment of rat hippocampal neurons with Aβo increased expression of TLR4, enhanced [Ca
2+] responses to LPS and increased LPS-induced neuron cell death. Most importantly, all these effects were observed only in aged cultures of rat hippocampal neurons but not in young cultures, indicating that aging and/or senescence is the critical factor required for Aβo-mediated enhancement of LPS-induced damage. It may be argued that coincidence of the neuron damaging effect by LPS and Aβo is very unlikely. However, a non-infectious tissue injury can release TLR4 endogenous ligands called DAMPs that trigger sterile inflammation in the nervous system. Therefore, in the brain, the simultaneous accumulation of Aβ and DAMP-induced TLR4 activation is plausible upon stress and/or brain damage, particularly in the context of aging. In fact, human aging is associated to increased serum levels of pro-inflammatory cytokines, a chronic subclinical condition named as “inflammaging”, as well as increased NF-кB signaling, a transcription factor that is a master regulator of TLR-mediated inflammation. Recent evidences in rat models show that elevation in the levels of inflammatory cytokines induced by TLR4 activation may promote the accumulation of Aβ, which may in turn increase TLR4 levels, thus creating a positive feedback loop that may contribute largely to the progression of AD [
36]. This feedback loop may be also amplified further by the age-associated increased expression of NMDA receptors that could be targeted simultaneously by LPS (or DAMPs) and Aβo. Our data is consistent with the report and further demonstrates an exacerbated crosstalk between Aβo and TLR4 on [Ca
2+] responses and cell death in aged hippocampal neurons, which might be relevant to the pathogenesis of age-related neurodegenerative diseases. Based on this, it is tempting to speculate that aging is the critical contributing factor that enables the vicious cycle between TLR4 and Aβo to promote AD.
Acknowledgements
We thank Mr. David del Bosque for technical support.