TNF-α protein synthesis inhibitor restores neuronal function and reverses cognitive deficits induced by chronic neuroinflammation

Abstract

Background: Chronic neuroinflammation is a hallmark of several neurological disorders associated with cognitive loss. Activated microglia and secreted factors such as tumor necrosis factor (TNF)-α are key mediators of neuroinflammation and may contribute to neuronal dysfunction. Our study was aimed to evaluate the therapeutic potential of a novel analog of thalidomide, 3,6'-dithiothalidomide (DT), an agent with anti-TNF-α activity, in a model of chronic neuroinflammation.

Methods: Lipopolysaccharide or artificial cerebrospinal fluid was infused into the fourth ventricle of three-month-old rats for 28 days. Starting on day 29, animals received daily intraperitoneal injections of DT (56 mg/kg/day) or vehicle for 14 days. Thereafter, cognitive function was assessed by novel object recognition, novel place recognition and Morris water maze, and animals were euthanized 25 min following water maze probe test evaluation.

Results: Chronic LPS-infusion was characterized by increased gene expression of the proinflammatory cytokines TNF-α and IL-1β in the hippocampus. Treatment with DT normalized TNF-α levels back to control levels but not IL-1β. Treatment with DT attenuated the expression of TLR2, TLR4, IRAK1 and Hmgb1, all genes involved in the TLR-mediated signaling pathway associated with classical microglia activation. However DT did not impact the numbers of MHC Class II immunoreactive cells. Chronic neuroinflammation impaired novel place recognition, spatial learning and memory function; but it did not impact novel object recognition. Importantly, treatment with DT restored cognitive function in LPS-infused animals and normalized the fraction of hippocampal neurons expressing the plasticity-related immediate-early gene Arc.

Conclusion: Our data demonstrate that the TNF-α synthesis inhibitor DT can significantly reverse hippocampus-dependent cognitive deficits induced by chronic neuroinflammation. These results suggest that TNF-α is a critical mediator of chronic neuroinflammation-induced neuronal dysfunction and cognitive impairment and targeting its synthesis could provide an effective therapeutic approach to several human neurodegenerative diseases.

Figure 1

Cytokine and chemokine gene expression. TNF-α mRNA levels in the hippocampus were significantly elevated in LPS-vehicle rats (*p < 0.05 vs. aCSF-vehicle rats) and reduced to control levels in LPS-DT rats (#p < 0.05 vs. aCSF-vehicle rats). TNFR2 mRNA levels were significantly elevated in LPS-vehicle rats (*p < 0.05 vs. aCSF-vehicle) and similarly lowered to control levels in LPS-DT rats (##p < 0.01 vs. LPS-vehicle; not significant vs. aCSF-vehicle). IL-1β mRNA levels were significantly increased compared to aCSF-V levels in both LPS-vehicle and LPS-DT groups (***p < 0.001 and *p < 0.05 vs. aCSF-vehicle rats). However, DT treatment did induce a significant reduction in the gene transcript level compared to the LPS-vehicle group (#p < 0.05 vs. LPS-vehicle rats). Treatment with LPS and/or DT induced no significant differences in gene transcript levels for IL-6, IL-4 and CX3CL1 and CD200.

Figure 2

OX-6 immunoreactive microglia staining. Staining from the LPS-vehicle and LPS-DT rats displayed elevated numbers of OX-6 immunoreactive microglia per square millimeter within the DG and CA3 regions of the hippocampus (**p < 0.01 vs. aCSF-veh). No significant difference was observed between LPS-vehicle and LPS-DT rats.

Figure 3

TLR-mediated signaling pathway related gene expression. Real time RT-PCR analysis from the hippocampus revealed that the gene expressions of TLR2 and TLR4 were increased in LPS-vehicle rats (***p < 0.001 and *p < 0.05 vs. aCSF-vehicle rats), while treatment with DT restored expression levels back to control (###p < 0.001 and #p < 0.05 vs. LPS-vehicle). Whereas LPS had no effect on Hmgb1 expression, treatment of LPS infused animals with DT decreased the expression of Hmgb1 (#p < 0.05 vs. LPS-vehicle). Treatment with DT reduced the expression of IRAK1 in both aCSF-DT and LPS-DT rats ($p > 0.05 vs. aCSF-vehicle and #p < 0.05 vs. LPS-vehicle). No significant differences in hippocampal transcript levels of MyD88 or NFκB p65 were observed between the treatment groups.

Figure 4

Novel place and novel object recognition test. (A). Novel place recognition familiarization phase. Analysis of the total amount of object exploration showed no significant difference across experimental groups. (B). 5 min delay novel place recognition. aCSF-vehicle, aCSF-DT and LPS-DT rats showed a preference for the novel place recognition over the familiar location (*p < 0.05 and **p < 0.01 vs. familiar place) whereas LPS-vehicle rats did not. (C). 24 h delay novel place recognition. aCSF-vehicle and aCSF-DT showed a preference for the novel place recognition over the familiar location (*p < 0.05 and ***p < 0.001 vs. familiar place) whereas LPS-vehicle and LPS-DT rats did not (D). Novel object recognition. aCSF-vehicle, aCSF-DT, LPS-vehicle and LPS-DT rats all showed a preference for the novel object over the familiar one (*p < 0.05; **p < 0.01; ***p < 0.001 vs. familiar object).

Figure 5

Morris water maze test. (A). Visible platform training. There was no significant group difference in average swim velocity or escape latency to the platform (B). Hidden platform training. aCSF-vehicle, aCSF-DT and LPS-DT groups improved their daily performance during the acquisition phase of the Morris water maze task, whereas LPS-vehicle rats showed an impaired learning profile. The average escape latency of LPS-vehicle rats was significantly higher than that of aCSF-vehicle animals on the second (*p < 0.05) and third day (***p < 0.001). The average escape latency of LPS-DT rats did not significantly differ from aCSF-vehicle rats, and was significantly lower than that of LPS-vehicle animals on the third day of training (^##p < 0.01). (C). Probe test. aCSF-vehicle, aCSF-DT and LPS-DT groups spent a significantly higher percentage of time in the target quadrant than the other quadrants (***p < 0.001 vs. other quadrants) whereas LPS-vehicle rats did not.

Figure 6

Proportion of neurons expressing Arc in response to behavior. The proportion of pyramidal neurons expressing behaviorally-induced Arc in the CA3 area was significantly increased in LPS-vehicle animals (*p < 0.05 vs. aCSF-vehicle). DT normalized this proportion back to control levels in LPS-DT rats (#p < 0.05 vs. LPS-vehicle). Similar trends were observed in the DG and the CA1 area, albeit statistical significance was not achieved.