The present work points to a role for bacterial translocation and subsequent TLR-4 pathway stimulation in the neuroinflammation induced by an experimental model of depression. To our knowledge, our results demonstrate for the first time that the TLR-4 signaling pathway becomes activated in brain cortex of rats exposed to an animal model of depression. This activation occurs with increased levels of the pro-inflammatory cytokine IL-1β and of one of the main enzymatic sources of inflammatory and oxidative mediators, COX-2 and its product PGE2. Interestingly, after 21 days of CMS, the COX-derived anti-inflammatory mediator 15d-PGJ2 appears decreased. As a consequence of this misbalance and the resulting enhancement of inflammation and oxidation in brain cortex after CMS exposure, an increment in lipid peroxidation takes place.
In the search for a mechanistic explanation for the observed TLR-4 activation, experiments using antibiotic intestinal decontamination suggest a pivotal role for anaerobic Gram-negative bacteria translocation on TLR-4-signaling pathway activation after stress exposure in brain cortex of rats.
In accordance with other studies carried out in different models of stress exposure, including CMS, our data show that there is inflammatory and oxidative/nitrosative damage in the brain after CMS [5, 38–40]. The increase of IL-1β mRNA levels detected in brain cortex also correlates with results obtained in previous studies [41–43]. This can be considered particularly significant, bearing in mind that this cytokine plays a central role in the sickness behavior detected in animals after LPS injection (LPS induces its release) and has been proposed as a possible actor involved in the pathophysiology of depression [6, 44]. Moreover, the actions of IL-1β in the CNS include increases in the production of other pro-inflammatory cytokines which can stimulate enzymatic sources of oxidative and nitrosative mediators .
Apart from cytokines, other mediators such as bacterial endotoxin (i.e. LPS, which we are showing here also increased after CMS) rapidly induce COX-2 and PGE2 production [46, 47]. The induction of COX-2 in the CNS by stress and the increase in the PGE2 levels in the brain cortex are well documented phenomena [48, 49] of significant importance in experimental models of depression and in depressive disorders , bearing in mind that PGE2, in turn, stimulates production of pro-inflammatory cytokines, expression of COX-2 and, as a co-factor, activity of indoleamine 2,3-dioxygenase (IDO), which reduces levels of 5-HT, a hallmark of depression.
On the other hand, it has been previously shown that, during the production of prostaglandins, reactive oxygen species (ROS) are generated, which are a main cause of oxidative/nitrosative damage as has been shown to occur after CMS, leading to an increase in lipid peroxidation markers (increase in the amount of MDA) . Although previous studies have revealed an increase in inducible nitric oxide synthase (iNOS) levels in the brain after acute and subacute stress protocols , after chronic exposure to a series of stressors of mild intensity (as occurs in CMS) the main isoform implicated is the constitutive, neuronal NOS (nNOS) isoform . Thus, the increase in lipid peroxidation observed in the specific experimental setting used in the present study should be attributed mainly to cyclooxygenase-derived products.
Activation of the transcription factor nuclear factor kappa B (NF-κB) controls the transcription of many acute-phase proteins and inflammatory genes both in humans and rodents, and is one of the earliest events in the stress-inflammation response in the brain [53, 54]. This transcription factor resides silent in the cytoplasm bound by an inhibitory protein, I kappa B alpha (IκBα). When a specific cellular pathway is stimulated, it produces phosphorylation and subsequent degradation of IκBα, activating NF-κB which translocates to cell nucleus where it recognizes specific DNA sequences in the promoter of target genes, among which are those that code for proteins involved in inflammation. Interestingly, no clear stimulation of NF-κB occurs in the brain cortex after CMS when its p65 subunit is analyzed. However, our results show that IκBα mRNA levels are increased after CMS. As it has been described to occur in other experimental settings, the increase in IκBα mRNA is an autoregulatory pathway switched on by NF-κB after prolonged stimulation as may be the case in CMS, thus restricting NF-κB action when chronically stimulated [55, 56].
Having described some components of the inflammatory response in the brain cortex to CMS exposure, we focused on a search for possible external stressors stimulating this response, as recently reviewed by Kubera et al. . All of the inflammatory parameters described up to this point can be induced by the Toll-like receptors (TLRs) pathway stimulation. TLRs, being the first line of defense against invading microorganisms, constitute the main agents of the innate immune response. Stimulation of TLRs causes an immediate defensive response, including the production of an array of antimicrobial peptides and inflammatory/oxidative mediators . During the last several years numerous studies have appeared regarding the role of TLRs in the pathophysiology of diverse CNS diseases such as multiple sclerosis, Alzheimer's disease and brain ischemia [16, 57, 58]. Now, our results show for the first time increases in expression of and mRNA levels for Toll-like receptor 4 (TLR-4) in the brain cortex in an experimental model of depression in rodents. Additionally, we have also found that CMS induces protein expression and synthesis of MD-2, which is the molecule that confers lipopolysaccharide responsiveness to TLR-4 .
Taken as a whole, the results presented here suggest that TLR-4 could be an important regulatory factor in the consequences of chronic stress in the brain, and also support a possibility for pharmacological or genetic manipulations of this pathway - although to date the selective inhibition of TLR-4 has proved to be a difficult challenge  - in order to minimize oxidative and inflammatory damage in the CNS after stress and in stress-related psycho- and neuro-pathologies such as depression.
There are several studies exploring endogenous ligands that activate TLR-4 after brain damage (e.g. protein S100 or nuclear protein high-mobility group box 1 after cerebral ischemia, pro-inflammatory cytokines after brain trauma) . However, knowledge about mechanisms that regulate TLR-4 activation in the brain in models of neuropsychiatric pathologies comes from previous studies based on stress exposure, which have shown increased intestinal permeability and a resultant bacterial translocation to the systemic circulation after stress exposure [21, 22]. As a result, there are circulating Gram-negative enterobacteria, which are a major source of LPS and can activate brain TLR-4 inducing a neuroinflammatory response. In order to clarify the origin of stress-induced activation of the TLR-4 pathway in CMS, we studied LPS and its binding protein (LBP; which serves as a lipid transfer protein that facilitates the transportation of LPS to the recognition protein CD14 and to TLR-4) levels in plasma. Our results show that CMS exposure produces increases in both LPS and LBP plasma levels. Thus, it is possible that CMS is causing an intestinal dysfunction followed by bacterial translocation, as occurs in different stress models in rodents , with LPS (from those Gram-negative bacteria) being the reason for the TLR-4 activation. This proposed mechanism, known as "leaky gut", also takes place in depressed patients, and has been related to the inflammatory pathophysiology of major depressive disorder .
To assess, in our experimental setting, whether the source of LPS and the consequent TLR-4 pathway stimulation, are bacteria translocated from the gut, we examined the effects of intestinal decontamination on the stress-induced inflammatory and oxidative/nitrosative changes revealed above. We used a standard stringent protocol (streptomycin and penicillin G) for only intestinal decontamination. This protocol has been used because it has demonstrated to lack any neuroprotective or anti-inflammatory effects on the CNS when used in other related protocols [24, 61]. By using this protocol, we can separate possible effects on the brain of the antibiotic used (i.e. the anti-neuroinflammatory effect of minocycline) from the effects caused by intestinal decontamination.
Our data show that animals subjected to CMS plus intestinal decontamination present a return to basal levels (control group values) for pro-inflammatory and oxidative/nitrosative parameters previously analyzed, including LPS and LBP plasma concentrations and TLR-4 and MD-2 expression and mRNA levels.
In this vein, of special relevance is the finding that antibiotic intestinal decontamination promotes decreases in IL-1β and COX-2/PGE2 in brain cortex. This result supports the notion that LPS from translocated bacteria stimulates TLR-4, and in that way produces the increases in IL-1β and COX-2/PGE2 levels in the CNS previously detected. More interestingly, intestinal decontamination is able to restore the disbalance between COX-derived inflammatory (PGE2) and anti-inflammatory (15d-PGJ2) components in the brain.
Our results also indicate that plasma corticosterone levels are increased after 21 days of CMS when compared with the control group, showing that even after this chronic stress exposure the hypothalamic-pituitary-adrenal (HPA) axis of these animals remains functioning. Additionally, it has been previously demonstrated that LPS stimulates the HPA axis  and thus, it is conceivable that the increase in the corticosterone levels after CMS could be caused, at least in part, by the increase in LPS levels detected here and not only by the stressors themselves. Supporting this idea, the intestinal decontamination that decreases LPS after CMS, also decreases plasma corticosterone levels, again supporting the role of intestinal bacteria as a source for the LPS detected in our study.
The effects of intestinal decontamination on depressive-like behavior were analyzed using a modified forced swimming test based on the method described by Porsolt , measuring behavioral despair. In spite of its anti-inflammatory effects after decreasing LPS levels, antibiotic decontamination failed to reverse the depressive-like behavior induced by CMS, which indicates a role for LPS-induced neuroinflammation after CMS without (at this level) behavioral consequences. Nonetheless, the fact that CMS-induced neuroinflammation is reversed by antibiotic intestinal decontamination is particularly relevant because neuroinflammation is considered an important biological event that might increase the risk of major depressive episodes much like more traditional psychosocial factors . Further studies using mixed protocols of experimental depression plus infective or inflammatory agents would aid in explaining the role of comorbid depression in inflammatory or immune-related pathologies.
The results presented here are in line with a hypothesis recently presented  according to which, external stressors to the brain, such as LPS, may up-regulate immune receptors such as TLR-4 that, in turn, may aggravate neuroinflammation due to locally produced internal stressors (prostanoids, some cytokines, transcription factors) thus causing a superinduction of (neuro)inflammatory responses.