Connecting TNF-alpha signaling pathways to iNOS expression in a mouse model of Alzheimer's disease: relevance for the behavioral and synaptic deficits induced by amyloid beta protein

Abstract

Increased brain deposition of amyloid beta protein (Abeta) and cognitive deficits are classical signals of Alzheimer's disease (AD) that have been highly associated with inflammatory alterations. The present work was designed to determine the correlation between tumor necrosis factor-alpha (TNF-alpha)-related signaling pathways and inducible nitric oxide synthase (iNOS) expression in a mouse model of AD, by means of both in vivo and in vitro approaches. The intracerebroventricular injection of Abeta(1-40) in mice resulted in marked deficits of learning and memory, according to assessment in the water maze paradigm. This cognition impairment seems to be related to synapse dysfunction and glial cell activation. The pharmacological blockage of either TNF-alpha or iNOS reduced the cognitive deficit evoked by Abeta(1-40) in mice. Similar results were obtained in TNF-alpha receptor 1 and iNOS knock-out mice. Abeta(1-40) administration induced an increase in TNF-alpha expression and oxidative alterations in prefrontal cortex and hippocampus. Likewise, Abeta(1-40) led to activation of both JNK (c-Jun-NH2-terminal kinase)/c-Jun and nuclear factor-kappaB, resulting in iNOS upregulation in both brain structures. The anti-TNF-alpha antibody reduced all of the molecular and biochemical alterations promoted by Abeta(1-40). These results provide new insights in mouse models of AD, revealing TNF-alpha and iNOS as central mediators of Abeta action. These pathways might be targeted for AD drug development.

Figure 1.

Intracerebroventricular injection of Aβ_1–40 disrupts learning and memory in mice. The spatial reference memory version of the Morris water maze test in Swiss mice (n = 9–10 mice per group) was used as a measure of cognition. A, Training trials were performed on day 7 after a single intracerebroventricular administration of aggregated Aβ_1–40 (400 pmol per mouse), inverse peptide Aβ_40–1 (400 pmol per mouse) (used as a negative control), or vehicle (PBS). Treatment with Aβ_1–40 significantly increased the escape latencies to find the submerged platform during the training trials (F_(2,26) = 104.32; p < 0.0001). B, The probe test session was performed 24 h after the training trials. Treatment with Aβ_1–40 reduced the frequency of time spent in the correct quadrant (F_(2,26) = 19.91; p < 0.0001). The values represent the mean ± SEM. *p < 0.05 compared with the vehicle-treated group (Newman–Keuls test).

Figure 2.

Involvement of TNF-α and iNOS in the cognitive deficits induced by Aβ_1–40 in mice. The spatial reference memory version of the Morris water maze test in mice was used as a measure of cognition. Training trials were performed on day 7 after a single intracerebroventricular administration of aggregated Aβ_1–40 (400 pmol per mouse) or vehicle (PBS). Data are presented as means ± SEM latency (seconds) for escape to a submerged platform (n = 8–10 mice per group). The probe test session was performed 24 h after training trials. Data are presented as means ± SEM of the frequency of time spent in the correct quadrant. A, B, E, F, Pretreatment with the specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) (A, B) or the preferential iNOS inhibitor AG (100 mg/kg, i.p., once per day) (E, F) improved the cognitive deficits induced by Aβ_1–40 in Swiss mice during training trials (AbTNF-α: F_(1,27) = 21.89, p < 0.0001; AG: F_(1,33) = 82.78, p < 0.0001) and test sessions (AbTNF-α: F_(1,27) = 5.26, p < 0.05; AG: F_(1,33) = 29.14, p < 0.0001) of the Morris water maze. C, D, G, H, TNFR1 knock-out (TNFR1^−/−; C, D) and iNOS knock-out (iNOS^−/−; G, H) mice were significantly more resistant than wild-type C57BL/6 mice to the deleterious effect of Aβ_1–40 in the spatial learning (TNFR1^−/−: F_(1,30) = 29.58, p < 0.0001; iNOS^−/−: F_(1,26) = 81.46, p < 0.0001) and spatial retrieval (TNFR1^−/−: F_(1,30) = 15.45, p < 0.001; iNOS^−/−: F_(1,26) = 28.46, p < 0.0001). *p < 0.05 compared with the vehicle/PBS-treated group; ^#p < 0.05 compared with the Aβ_1–40/PBS-treated group (Newman–Keuls test).

Figure 3.

TNF-α and iNOS participate in Aβ_1–40-induced synaptic disruption. Immunohistochemistry analysis for the presynaptic protein synaptophysin was performed on day 8 after aggregated Aβ_1–40 (400 pmol per mouse) intracerebroventricular injection. A–H, Relative optical densitometry of synaptophysin immunostaining in the CA1 (A, B), CA2 (C, D), and CA3 (E, F) subregions of the hippocampus and parietal cortex (G, H). Synaptophysin immunoreactivity was used as a measure of synaptic density. A, C, E, G, Pretreatment with the specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) or with the preferential iNOS inhibitor AG (100 mg/kg, i.p., once per day) prevented the Aβ_1–40-induced synaptic disruption in Swiss mice in the CA1 (F_(3,8) = 9.38; p < 0.01; A) and CA2 (F_(3,8) = 9.44; p < 0.01; C), but not in the CA3 (F_(3,8) = 4.02; p < 0.05; E) and parietal cortex (F_(3,8) = 3.26; p = 0.08; G). B, D, F, H, TNFR1 knock-out (TNFR1^−/−) and iNOS knock-out (iNOS^−/−) mice were significantly more resistant than wild-type C57BL/6 mice to the Aβ_1–40-induced synaptic disruption in the CA2 (F_(3,8) = 4.79; p < 0.05; D) and CA3 (F_(3,8) = 4.63; p < 0.05; F), but not in the CA1 (F_(3,8) = 3.17; p = 0.08; B) and parietal cortex (F_(3,8) = 4.00; p < 0.05; H). The values represent the mean ± SEM. *p < 0.05, **p < 0.01 compared with the control group (naive); ^#p < 0.05, ^##p < 0.01 compared with the Aβ_1–40/PBS-treated group (Newman–Keuls test).

Figure 4.

Requirement of TNF-α for Aβ_1–40-induced iNOS upregulation. Swiss mice were treated with aggregated Aβ_1–40 (400 pmol, i.c.v., per mouse) [except naive, untreated mice (N)], and the hippocampus and prefrontal cortex were isolated at the time points indicated. A, B, Total RNA was isolated from the hippocampus (A) and prefrontal cortex (B) for evaluation of TNF-α and iNOS expression by RT-PCR. β-Actin mRNA was assessed in all RNA samples as an internal control for the amount of RNA in each sample. TNF-α and iNOS mRNA increased in a time-dependent manner after Aβ_1–40 intracerebroventricular treatment, whereas β-actin levels remained constant. C, D, Aβ_1–40 induced a time-dependent and prolonged iNOS protein expression in the hippocampus (C) and prefrontal cortex (D). Western blot analysis revealed that pretreatment with the specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) prevented Aβ_1–40-induced iNOS expression in the hippocampus (F_(1,36) = 118.26; p < 0.0001) and prefrontal cortex (F_(1,36) = 87.69; p < 0.0001). Immunoblot for α-actin was used as a cytosolic loading control. Treatment with the reverse peptide Aβ_40–1 (6 h) had no effect on the iNOS protein expression. As a positive control, LPS (2.5 μg, i.c.v., 12 h) treatment induced a marked iNOS protein expression. E, F, Graph illustrating the temporal profile of the TNF-α and iNOS mRNA synthesis and iNOS protein expression in the hippocampus (E) and prefrontal cortex (F). Results were normalized by arbitrarily setting the densitometry from the maximal responsive group and are expressed as the mean of three to four independent experiments.

Figure 5.

Blockade of TNF-α prevents alterations in GSH-dependent antioxidant parameters induced by Aβ_1–40. Swiss mice were treated with a specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) 15 min before the administration of the aggregated Aβ_1–40 (400 pmol per mouse) or vehicle (PBS). A–F, Total GSH levels (GSH total; A, B) and antioxidant enzymes, GR (C, D), and GPx (E, F) were measured in the hippocampus and prefrontal cortex of Swiss mice at 24 h after Aβ_1–40 administration. The values represent the mean ± SEM of GSH total levels (micromoles per gram of wet tissue) and GR and GPx activities (milliunits per milligram of protein) (n = 5–7 mice per group). A, B, Treatment with AbTNF-α prevented the reduction in GSH total levels in the hippocampus (F_(3,16) = 18.04; p < 0.0001; A) and prefrontal cortex (B) of Aβ_1–40-treated mice. C–F, AbTNF-α treatment prevented the increase in GR and GPx activities in the hippocampus (C, E) (F_(3,16) = 2.05, p < 0.05 and F_(3,16) = 5.13, p < 0.001, respectively) and (D, F) prefrontal cortex (F_(3,16) = 5.68, p < 0.01 and F_(3,16) = 4.17, p < 0.05, respectively) induced by Aβ_1–40 injection. The values represent the mean ± SEM. *p < 0.05 compared with the vehicle/PBS-treated group; ^#p < 0.05 compared with the Aβ_1–40/PBS-treated group (Newman–Keuls test).

Figure 6.

Aβ_1–40-induced JNK/c-Jun pathway activation. A, B, Activation of cytoplasmic JNK in response to aggregate Aβ_1–40 intracerebroventricular injection was detected in the hippocampus (A) and prefrontal cortex (B) of Swiss mice, at the time points indicated, by phosphospecific antibody for diphosphorylated JNK (p-JNK; at Thr¹⁸³ and Tyr¹⁸⁵). Blots were reprobed with anti-JNK to verify equal loading. C, D, Kinetics of c-Jun protein translocation from cytoplasm into the nucleus in the hippocampus (C) and prefrontal cortex (D) after Aβ_1–40 administration. Activation of the JNK pathway is dependent on TNF-α. Pretreatment with the specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) prevented Aβ_1–40-induced JNK activation and subsequent c-Jun translocation in the hippocampus and prefrontal cortex. Immunoblot for lamin A/C was used as a nuclear loading control. N, Naïve, untreated mice. Treatment with the reverse peptide Aβ_40–1 (6 h) had no significant effect on the JNK/c-Jun activation. As a positive control, LPS (2.5 μg, i.c.v., 1 h) treatment induced activation in the JNK/c-Jun pathway. E, F, Pretreatment with JNK inhibitor SP600125 (25 mg/kg, i.p., 1 h before Aβ), but not with NF-κB inhibitor PDTC (100 mg/kg, i.p., 1 h before Aβ), prevented Aβ_1–40-induced c-Jun translocation in the hippocampal (E) and cortical (F) homogenates. Brain samples were collected 6 h after Aβ_1–40 intracerebroventricular injection. These data indicate an association between JNK and c-Jun proteins.(C), Cytoplasmatic; (N), nuclear.

Figure 7.

Aβ_1–40-induced NF-κB pathway activation. A, B, Cytoplasmic [(C)] and nuclear [(N)] samples of hippocampus (A) and prefrontal cortex (B) from Swiss mice were probed with a p65 NF-κB antibody at the time points indicated. Data indicate that Aβ_1–40 induced a translocation of p65 NF-κB from the cytoplasm to the nucleus. Pretreatment with the specific antibody against mouse TNF-α (AbTNF-α; 10 ηg, i.c.v., per mouse) prevented Aβ_1–40 induction of p65 NF-κB translocation in hippocampus and prefrontal cortex. Immunoblot for lamin A/C was used as a nuclear loading control. Treatment with the reverse peptide Aβ_40–1 (6 h) had no significant effect on the p65 NF-κB translocation. As a positive control, LPS (2.5 μg, i.c.v., 1 h) treatment resulted in p65 NF-κB migration into the nucleus. C, D, Aβ_1–40-induced p65 NF-κB translocation was prevented by pretreatment with NF-κB inhibitor PDTC (100 mg/kg, i.p.), but not by pretreatment with JNK inhibitor SP600125 (25 mg/kg, i.p.), in hippocampus (C) and prefrontal cortex (D). These data indicate that JNK/c-Jun and NF-κB pathways are independently activated by Aβ_1–40 intracerebroventricular injection. N, Naïve, untreated mice; C, control.

Figure 8.

Aβ_1–40-induced expression of iNOS is dependent on JNK/c-Jun and NF-κB activation. Involvement of JNK/c-Jun and NF-κB signaling pathways on iNOS expression was verified in the hippocampus (A) and prefrontal cortex (B) 24 h after Aβ_1–40 intracerebroventricular administration. Top, Pretreatment with the selective inhibitor of JNK, SP600126 (50 mg/kg, i.p.), or NF-κB blocker PDTC (100 mg/kg, i.p.) 1 h before Aβ_1–40 treatment reduced iNOS expression in the hippocampus (F_(3,8) = 40.00; p < 0.0001) and prefrontal cortex (F_(3,8) = 12.68; p < 0.01). Bottom, Graphs showing quantification of iNOS protein normalized by α-actin protein (loading control). The values represent the mean ± SEM. **p < 0.01 compared with the vehicle-treated group; ^#p < 0.05, ^##p < 0.01 compared with the Aβ_1–40-treated group (Newman–Keuls test).