IL-6 mediated degeneration of forebrain GABAergic interneurons and cognitive impairment in aged mice through activation of neuronal NADPH oxidase - PubMed (original) (raw)

IL-6 mediated degeneration of forebrain GABAergic interneurons and cognitive impairment in aged mice through activation of neuronal NADPH oxidase

Laura L Dugan et al. PLoS One. 2009.

Abstract

Background: Multiple studies have shown that plasma levels of the pro-inflammatory cytokine interleukin-6 (IL-6) are elevated in patients with important and prevalent adverse health conditions, including atherosclerosis, diabetes, obesity, obstructive sleep apnea, hypertension, and frailty. Higher plasma levels of IL-6, in turn, increase the risk of many conditions associated with aging including age-related cognitive decline. However, the mechanisms underlying this association between IL-6 and cognitive vulnerability remain unclear.

Methods and findings: We investigated the role of IL-6 in brain aging in young (4 mo) and aged (24 mo) wild-type C57BL6 and genetically-matched IL-6(-/-) mice, and determined that IL-6 was necessary and sufficient for increased neuronal expression of the superoxide-producing immune enzyme, NADPH-oxidase, and this was mediated by non-canonical NFkappaB signaling. Furthermore, superoxide production by NADPH-oxidase was directly responsible for age-related loss of parvalbumin (PV)-expressing GABAergic interneurons, neurons essential for normal information processing, encoding, and retrieval in hippocampus and cortex. Targeted deletion of IL-6 or elimination of superoxide by chronic treatment with a superoxide-dismutase mimetic prevented age-related loss of PV-interneurons and reversed age-related cognitive deficits on three standard tests of spatial learning and recall.

Conclusions: Present results indicate that IL-6 mediates age-related loss of critical PV-expressing GABAergic interneurons through increased neuronal NADPH-oxidase-derived superoxide production, and that rescue of these interneurons preserves cognitive performance in aging mice, suggesting that elevated peripheral IL-6 levels may be directly and mechanistically linked to long-lasting cognitive deficits in even normal older individuals. Further, because PV-interneurons are also selectively affected by commonly used anesthetic agents and drugs, our findings imply that IL-6 levels may predict adverse CNS effects in older patients exposed to these compounds through specific derangements in inhibitory interneurons, and that therapies directed at lowering IL-6 may have cognitive benefits clinically.

PubMed Disclaimer

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Aging increases plasma and brain levels of IL-6, and induces Nox2 protein expression which is blocked by targeted deletion of IL-6.

(a) Plasma IL-6 was measured by ELISA in young (4 month) and old (24 month) C57BL6 mice and was significantly increased in old mice; n = 4, mean±SEM (F(1,6) = 9.530, P = 0.021). (b) IL-6 mRNA expression was determined by real-time PCR in brain and was also increased significantly in old animals; n = 4, mean±SEM (F(1,6) = 10.288, P = 0.018). (c) Confocal imaging of brain slices prepared from young (4 months) and old (23 months) mice depicting Nox2 (green) and MAP2 (red) immunoreactivity in the CA1 hippocampal region. Nox2 expression was increased in old animals with respect to young, whereas this staining was absent in slices prepared from Nox2-deficient (gp91phox -/-, 12 months old) mice. Bar = 100 µm. For immunofluorescence detection AlexaFluor 488 and 568 conjugated secondary antibodies were used. Merged images show co-localization in yellow. (d, e) Western blot analysis of brain protein extracts obtained from young and old wild-type and IL-6-/- mice showed increased Nox2 expression with age which was significantly lower in old IL-6-/- mouse brain (n = 4, p<0.05, mean±SEM). Expression in young animals, either WT or IL-6-/-, was below detection levels. Lack of the Nox2 band in brain proteins extracted from an old gp91phox-/- mouse (far right lane) confirmed specificity of the anti-Nox2 antibody and location of Nox2- specific band. N = 4 animals per age per genotype (F(3,12) = 9.704, P = 0.002, *young versus old, P<0.05; # old versus old-IL6-/-, P<0.05 by Tukey's).

Figure 2. Imaging of in vivo superoxide production in brain of young and old animals: role of Nox2 and IL-6.

(a) Confocal imaging of in vivo DHE oxidation reveals that superoxide-mediated DHE oxidation is increased in neurons in old mice, and this increase is blocked by the NADPH oxidase inhibitor, apocynin, or by the SOD mimetic, C3. Young (4 mo) and old (24 mo) mice were treated orally for 7 days with the Nox inhibitor, apocynin, or for 21 days with the SOD mimetic, C3. (b) Quantitative analyses of fluorescence in hippocampal CA1 and cortex were performed as described in detail in the methods section. Images were obtained using a 40× WI objective. Values are mean±SEM, n = 3. Significance was determined by ANOVA: for CA1 (*old versus young, P = 0.01;**old versus old+apocynin (P = 0.02) or old+C3 (P = 0.01) and cortex (Ctx) (*old versus young, P = 0.01;**old versus old+apocynin (P = 0.01) or old+C3 (P = 0.006) by Tukey's. (c) Live animal imaging of DHE oxidation using a GE eXplore Optix™-MX2 whole animal scanner also demonstrates increased DHE oxidation in brain of old (24 mo) relative to young (4 mo) mice. Fluorescence images of brain through intact skull and skin are shown on the left , and fluorescence intensity from ox-DHE is mapped onto a linear pseudocolor scale. Images to the right are ex vivo scans of the same brains showing greater anatomical detail, including fluorescence from the cortical hemispheres and cerebellum. (d) Repeated injections of IL-6 on two consecutive days (5 µg/kg) in old mice induced Nox2 expression (top images, green), and produced a small increase in DHE oxidation (red). However, repeated injections of IL-6 every 12 h (5 µg/kg, times three) to produce sustained systemic levels resulted in a dramatic increased in ox-DHE fluorescence compared to saline-injected old controls (bottom images). Note that the laser excitation power was decreased from 260 mW in (c) to 100 mW (d) to avoid saturating the camera in the scanner.

Figure 3. Increased synaptosomal Nox-2 activation and superoxide production in old wild-type mice is absent in old IL-6-/- mice; determination by electrochemical oxygen consumption analysis and EPR spin-trapping spectroscopy.

(a) NADPH-induced oxygen consumption is significantly decreased by the Nox inhibitor apocynin (300 µM, black bars). Oxygen consumption by synaptosomal Nox from the cortex of young (3–4 M, n = 9) and old (23–24 M, n = 9) animals at 37°C was induced by the addition of 5 mM NADPH to samples containing 2–5 mg synaptosomal protein. The observed Nox specific activity was significantly higher in IL-6 injected young animals (white bars, p = 0.04, n = 5) and in old brains (light gray bar, p = 0.01, n = 9) and was significantly lower in IL-6-/- old mice (dark gray bar, p = 0.04, n = 4). IL-6 treatment: Two IL-6 i.p. injections (5 mg/kg), separated by 12 hrs were administered to young and old animals before the groups were sacrificed 12 hrs later for synaptosomal isolation. (b) Representative EPR spectra recorded 3 minutes after mixing 5 mM NADPH with 0.2–0.5 mg synaptosomal protein isolated from brain of old mice at 37°C with 70 mM DEPMPO in the absence or in the presence of 200 µM AEBSF or 300 µM apocynin. In each incubation, 10 mM DETC was included to inhibit the SOD enzyme, resulting in enhanced and quantifiable EPR signals. Using computer simulations we concluded that the observed signals are attributable to a mixture of adducts of DEPMPO with hydroxyl radical and lipid derived carbon centered radical and both radicals are derived from superoxide since the signals were completely abolished in the presence of SOD; not shown. (c) Quantification of the observed EPR signals accumulated over the first 3 minutes from NADPH addition, n = 4–9 per group. Superoxide yield in synaptosomes isolated from different groups follows essentially the same activity pattern seen by oxygraphy and shown in (a). The asterisks (*) designate statistical significance by ANOVA followed by Tukey test with respect to young control while (#) indicate statistical significance with respect to old control. EPR settings are described fully in methods. Data are mean ± SEM. (d) Nox-generated superoxide burst in synaptosomes isolated from IL-6-/- KO mice is significantly lower than that in their old control brains.

Figure 4. PV Interneurons are decreased with aging, but spared in IL-6 -/- mice or mice treated chronically with an SOD mimetic, C3.

(a) Images, depicting the regions analyzed for PV-interneuron immunostaining in the prelimbic region (between Bregma 2.2 and 1.9), and dorsal hippocampus (between Bregmas −1.2 and −1.7). Four consecutive slices were analyzed for each animal. All cells in the regions were counted, and counts per slice were calculated as described in Methods (b) PV-cell counts in the regions of CA1, CA3 and dentate gyrus (DG) of old and old-C3 treated animals are expressed as % of their control. Aging was accompanied by a statistically significant decrease in PV-interneuron number in all regions analyzed and treatment of animals from middle age with the SOD-mimetic C3 (Old-C3) prevented the reduction of PV-interneuron numbers in CA1 and CA3, but not in DG. (c) The loss of PV neurons in all of the studied regions was partially rescued in IL-6 KO mice. * indicates statistical significance with respect to young animals (P<0.05) and # indicates statistical significance with respect to old animals (P<0.05) as determined by ANOVA followed by Tukey's test. YM and OM: n = 9 animals per group; OM + C3: n = 7 animals.

Figure 5. IL-6-induction of Nox2 and subsequent decrease of parvalbumin and GAD67 expression in PV-positive interneurons is NFκB-dependent.

Nuclear extracts of brain samples obtained from wild-type or IL-6-deficient animals were analyzed for content of the NFkB p65 subunit by colorimetric quantification of p65 using a commercially available kit (TransAM). A high concentration of p65 was found in nuclear extracts obtained from old wild type mice (a), and this was greatly diminished in nuclear extracts obtained from similar-age IL-6 deficient mice. The specificity of the binding to wild type oligonucleotide was confirmed by competition with either sense (oligo) or mutated oligonucleotides (mutant oligo). ANOVA analysis yielded a significant effect of genotype (F(1,18) = 9.414, P<0.05). (b) Application of IL-6 to neurons in culture induced Nox2 expression (b, top panel) and DHE oxidation (b, bottom panel and c) that was blocked by the NFκB inhibitor SN50. Dissociated primary neuronal cultures were treated with IL-6 (10 ng/ml) for 24 h in the absence or presence of SN50 (5 µM), an inhibitor of NFκB nuclear translocation. Dihydroethidium (DHE, 1 µg/ml) was applied for the last hour as described. After fixation, cells were immunostained for Nox2. * = significant with respect to control. # = significant with respect to IL-6, F(2,346) = 12.826, P<0.05). (d. e) SN50 also prevented the decrease in expression of parvalbumin and GAD67 induced by IL-6 exposure in PV-interneurons (*F(2,409) = 234.7, P<0.05; #F(2,409) = 192.8, P<0.05).

Figure 6. Old IL-6-/- mice or old mice treated with the SOD mimetic, C3, exhibit preserved spatial working memory.

(a) Old IL-6-/- mice show improved learning performance on the Novel Object task. Old (22 month) C57BL6 and IL-6-/- mice (on a C57B6 background) were tested for habituation to each of three novel objects, and IL-6-/- mice showed significantly better performance than wild-type mice (statistics shown in Results). (b) Barne's maze escape latency (time to find the darkened chamber) showed a statistically significant decrease in latency in old IL-6-/- mice but wild-type mice (statistics included in results section). (c) Preservation of spatial learning and memory on the Morris water maze by chronic treatment with C3. Young (6 M), old (24 M), and old-_C3_-treated (24 M) male mice were tested on components of the Morris water maze as previously described. Graph shows performance on the probe trial component for the specific set of male mice used in the current study. Data from the probe component of the trial are shown for the immediate (1 h) recall, and delayed (24 h) recall. C3 significantly preserved performance at both time-points. Data are mean±SEM, and were analyzed by 2-Way ANOVA followed by Tukey's post-hoc test, n = 13 young, n = 10 old, n = 12 old-C3.

References

1. Krabbe KS, Pedersen M, Bruunsgaard H. Inflammatory mediators in the elderly. Exp Gerontol. 2004;39:687–699. -PubMed
1. Maggio M, Guralnik JM, Longo DL, Ferrucci L. Interleukin-6 in aging and chronic disease: a magnificent pathway. J Gerontol A Biol Sci Med Sci. 2006;61:575–584. -PMC -PubMed
1. Ershler WB, Keller ET. Age-associated increased interleukin-6 gene expression, late-life diseases, and frailty. Annu Rev Med. 2000;51:245–270. -PubMed
1. Giuliani N, Sansoni P, Girasole G, Vescovini R, Passeri G, et al. Serum interleukin-6, soluble interleukin-6 receptor and soluble gp130 exhibit different patterns of age- and menopause-related changes. Exp Gerontol. 2001;36:547–557. -PubMed
1. Walston J, Hadley EC, Ferrucci L, Guralnik JM, Newman AB, et al. Research agenda for frailty in older adults: toward a better understanding of physiology and etiology: summary from the American Geriatrics Society/National Institute on Aging Research Conference on Frailty in Older Adults. J Am Geriatr Soc. 2006;54:991–1001. -PubMed

IL-6 mediated degeneration of forebrain GABAergic interneurons and cognitive impairment in aged mice through activation of neuronal NADPH oxidase - PubMed (original) (raw)