Background
It has been shown that intraperitoneal injection of lipopolysaccharide (LPS), a cell-wall component of gram-negative bacteria, induces neuroinflammation, hippocampal cell loss, cognitive impairment, learning deficits and even β-amyloid plaque generation in the hippocampus [
1,
2], constituting a valid experimental model to study the physiological, behavioral, and emotional aspects of sickness behavior [
3]. A number of studies have shown that upon exposure to LPS, microglia become activated and produce proinflammatory mediators such as cytokines, chemokines, prostanoids, and reactive oxygen species. Current evidence indicates that these products are key mediators of the neuroinflammatory process and contribute to LPS-induced neuronal damage and subsequent cognitive loss [
1,
4-
8]. In line with this view, it has been shown that LPS administration impairs contextual fear conditioning [
8-
10] and spatial memory [
2,
11-
14] and avoidance learning [
15-
19] in rats and mice. In addition, LPS decreases the preference for a novel object in a novel object recognition test in mice [
20]. The comprehension of the mechanisms by which such a deficit occurs may unveil additional (or ratify already-known) regulatory mechanisms of memory consolidation and retrieval in this pathologic condition, thereby guiding the search for novel therapeutic strategies or compounds to mitigate these deficits.
The N-methyl-D-aspartate (NMDA) receptors seem to be particularly susceptible to the neuroinflammatory challenge, since inflammation decreases the expression of GluN1 [
21], GluN2A and GluN2B [
22] subunits, and NMDA-dependent long-term potentiation [
23] in the hippocampus. This finding is in agreement with previous reports that neuroinflammation decreases total NMDA receptors (NR1 immunoreactivity) in the hippocampus and entorhinal cortex [
24] and that LPS-treated animals present decreased MK801 binding [
25]. Accordingly, the partial NMDA receptor agonist D-cycloserine prevents the deleterious effects of LPS [
26] and closed head injury [
27] on memory consolidation. Therefore, it sounds possible that other positive allosteric modulators of the NMDA receptor attenuate LPS-induced cognitive deficits.
Polyamines, such as putrescine, spermidine, and spermine, allosterically activate NMDA receptors by binding at the lower lobe of the N-terminal domain of GluN1 and GluN2B dimer interface [
28]. Functionally, polyamines are involved in growth and differentiation, but they also regulate a broad array of cellular functions in both neurons and inflammatory cells [
29-
31]. Numerous reports have indicated that polyamines improve memory in several tasks and attenuate memory deficits induced by different amnesic agents [
32-
41]. In fact, agonists and antagonists of the polyamine binding site at the NMDA receptor respectively facilitate and impair memory in various tasks [
33,
34,
36,
39,
40,
42,
43], and the sequential activation of protein kinase C (PKC) and PKA/CREB (protein kinase A/cAMP response element-binding protein) pathways in the hippocampus has been implicated in the promnesic effect of polyamines [
44,
45]. However, it has recently been described that interruption of the NMDA receptor modulation by polyamines reverses Aβ
25–35-induced memory impairment in mice in a novel object recognition task [
46], suggesting that the role of polyamines in memory may vary in physiological and pathological conditions.
The assumption that the effects of polyamines on memory depend on physiological or pathological conditions is in agreement with clinical and experimental data that support a beneficial role for NMDA antagonists in Alzheimer’s disease [
46,
47]. However, one might ask whether this applies to every pathological or neuroinflammatory condition. Therefore, we investigated the effect of spermine on LPS-induced impairment of memory in a novel object recognition test. In addition, we investigated whether spermine alters LPS-induced increase of cortical and hippocampal cytokine levels in mice, since both anti- and proinflammatory activity have also been reported for these aliphatic amines [
29,
48-
51].
Discussion
This study showed that a polyaminergic agonist, spermine, reverses post-training bacterial endotoxin-induced memory impairment in a novel object recognition task. In addition, it showed that spermine, at doses higher than those capable of reversing the deleterious effect of LPS, increases the recognition index in the novel object recognition task. It also revealed that ifenprodil decreases the recognition index in the novel object recognition task, and a non-effective dose of ifenprodil reverses the memory-improving effect of spermine in LPS-treated animals, providing pharmacological evidence that the promnesic effect of spermine involves the polyamine binding site at the NMDA receptor. However, spermine failed to reverse the LPS-induced increase of cytokine levels in the cerebral cortex or hippocampus.
The currently described ability of spermine to reverse LPS-induced memory deficits is in agreement with a number of studies that have separately shown that while LPS disrupts [
1,
8,
13,
15,
53], polyamines improve memory [
32-
39,
54].
Several mechanisms have been implicated in LPS-induced alterations of neural functions. It has been suggested that memory impairment is triggered by direct stimulation of toll-like receptor 4 (TLR
4) by LPS. This TLR
4 activation recruits myeloid differentiation adaptor protein (MyD88) and ultimately activates the nuclear factor κB signaling pathway, increasing the production of proinflammatory cytokines by macrophages [
55], microglia [
56-
58] and astrocytes [
56,
59]. The LPS-induced activation of glial cells results in the release of neurotoxic substances, including nitric oxide, glutamate, cytotoxic cytokines, and superoxide radicals [
8,
56,
60,
61], the suppression of neurotrophic factor secretion, and impaired neuroplasticity and, consequently, learning and memory [
8,
9,
11,
17,
62,
63]. In addition, TLR
4 activates Src family kinases, which phosphorylate GluN2B subunit or NMDA receptor and enhance GluN2B-dependent Ca
2+ influx [
64,
65], promoting excitotoxicity. Indeed, LPS induces progressive and cumulative neuronal loss over time [
66-
68].
Conversely, there is a large body of evidence indicating that the promnesic effects of polyamines involve the activation of NMDA receptors, which have been critically implicated in learning and memory processes [
36,
40,
44,
45,
69-
71]. In the current study, ifenprodil, a noncompetitive GluN2B-containing NMDA receptor antagonist, not only impaired the consolidation of the memory of novel object recognition task, but also reversed (at a dose that did not alter memory
per se in our study) the improving effect of spermine on the memory of LPS-treated animals, supporting a role for NMDA receptors in this effect of spermine [
72-
74]. Interestingly, these results are in agreement with the study by Kranjac and colleagues [
26], who have shown that partial NMDA receptor agonist D-cycloserine rescues memory consolidation following systemic bacterial endotoxin exposure. Furthermore, Velloso and colleagues [
41] have found that post-training intrastriatal administration of spermine reverses the recognition memory deficits in the novel object recognition task induced by quinolinic acid, a model of Huntington’s disease. It is worth noting that spermine improved memory
per se in the novel object recognition task. To our knowledge, this is the first study showing that spermine improves memory. The finding that ifenprodil prevents the promnesic effect of spermine tempts us to propose that it may involve the same molecular targets proposed for spermidine [
33,
34,
36,
38]. However, further studies are necessary to clarify this point.
We have found that while LPS increased IL1-β, IL-6, TNF-α, and IFN-γ levels, it decreased IL-10 levels in the cerebral cortex and hippocampus. This is in agreement with previous studies that have shown that systemic administration of this endotoxin causes neuroinflammation [
56-
58]. Moreover, spermine increased IL1-β and IL-6 levels in the cerebral cortex and hippocampus
per se, indicating that it facilitates inflammation, but to a much lesser degree than LPS. Interestingly, spermine also decreased the anti-inflammatory cytokine IL-10 in both cerebral structures, further supporting a proinflammatory role for this polyamine in our experimental conditions. These results are in line with other studies [
51,
75] which show that the spermine facilitates inflammation in
vivo. Accordingly, it has been shown that polyamines promote macrophage influx into the murine central nervous system following pathogenic insult, in an
in vivo model of secondary central nervous system inflammation [
51]. Soulet and Rivest [
75] have shown that systemic LPS increases ornithine decarboxylase expression throughout the central nervous system and that this precedes an increase in the expression of proinflammatory molecules TLR
2 and TNF-α, in mice. Moreover, treatment with a difluoromethylornithine, an irreversible inhibitor of ornithine decarboxylase, decreases LPS-induced TNF-α and TLR
2 expression, proving evidence that the LPS-induced increase of proinflammatory molecules is polyamine-dependent. The intracerebral injection of spermine prior to LPS also increases the number of cells expressing both TNF-α and TLR
2 in the central nervous system [
75]. Interestingly, LPS exposure stimulates IL-1β release from astrocytes (
in vitro) through a mechanism that requires NMDA receptor stimulation [
76].
We hypothesized that polyamines could decrease LPS-induced cognitive impairment by interfering in cytokine levels because: (1) LPS, at doses that cause cognitive impairment, sequentially increases brain IL-6, IL-1β, and TNF-α mRNA levels in the hippocampus [
77,
78] and frontal cortex [
79]; (2) LPS-induced upregulation of IL-1β and TNF-α mRNA in hippocampal tissue of IL-6
(+/+) mice is absent in IL-6
(−/−) mice, which are also refractory to the LPS-induced impairment in working memory [
77]. In addition, overexpression of TNF-α in neurons or glial cells impairs passive avoidance memory [
80]; (3) Pharmacological treatments that decrease the levels of TNF-α, TNF-α receptor 1, and NF-κB p65 phosphorylation also decrease LPS-induced memory deficits [
81,
82], though conventional TNF
(−/−) knockout mice present cognitive dysfunction [
83,
84]. A recent review summarized the effect of TNF-α, IL-6, and IL-1 on learning and memory [
85], where the authors suggest that TNF-α and its receptors might mediate the disrupting effect of LPS on learning and memory. Therefore, considering the existing evidence supporting a role for cytokines, particularly TNF-α, on LPS-induced cognitive impairment, we thought that the memory-improving effects of spermine could involve a decrease of cytokine levels in the hippocampus.
It has recently been described that LPS reduces the number of excitatory synapses in the hippocampus and cerebral cortex, leading to synaptic deficits [
86] that may underlie LPS-induced cognitive deficits. In fact, LPS induces neurotoxic substance release and suppression of neurotrophic factors secretion, which are known to increase neuroplasticity and, consequently, learning and memory. Since polyamines increase excitatory activity [
28], it is possible that spermine actions involve a compensatory increase of excitatory transmission in LPS-treated animals. However, specific studies have to be performed, to clarify this point.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.
The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
Competing interests
The authors declare that they have no competing interests.
Authors’ contributions
PKSF, RPI, LT, and TD conceived and designed the experiments. PKSF and LT performed the behavioral experiments. TD performed the cytokine quantification assay. PKSF, RPI, CFM, and MAR analyzed the data. CFM, MAR, and PKSF wrote the paper. All authors read and approved the final manuscript.