No NOS Promotes Tau Pathology in APP Transgenic Mice
A new mouse model has achieved a long-sought pathological prize—the production of hyperphosphorylated and aggregated tau proteins in a background of elevated Aβ. The trick to linking Aβ to tau’s tangles, it turns out, is to get rid of nitric oxide (NO). In a paper published August 14 in PNAS online, Hana Dawson and colleagues from Duke University, Durham, North Carolina, report that knocking out the gene for inducible nitric oxide synthase (NOS2) in an APP transgenic mouse leads to the hyperphosphorylation and aggregation of endogenous mouse tau protein and neuronal death.
The results, showing that the decrease in NOS exacerbates Aβ production and generates pathological tau species, run contrary to previous work finding NOS knockouts are protected against amyloid pathology (Nathan et al., 2005). In general, NO has been considered harmful to neurons (see ARF related news story and ARF news story), but the new results suggest that under some circumstances, NO may actually protect them.
The new mouse, generated by first author Carol Colton and coworkers, resulted from crossing the APPsw transgenic line Tg2576 with a NOS2 knockout. The progeny lacked the iNOS protein and displayed total NOS activity of about one-third that of APPsw mice. Hyperphosphorylated tau appeared in the soma and dendrites of cortical and hippocampal neurons in the APPsw/NOS-/- animals—phospho-tau was not seen in the APPsw or NOS-/- mice. Aggregated tau was detected by immunostaining or electron microscopy after filtration of brain lysates, and in intact tissue using thioflavin F staining.
Knockout of NOS2 also enhanced amyloid pathology in the mice. Total brain Aβ levels were five to six times higher in APPsw/NOS2-/- compared to APPsw littermates. The increase was mostly due to insoluble Aβ peptides. These results suggest that NO might influence Aβ generation or clearance by an unknown mechanism.
Finally, the mice showed neuron loss, which is not normally seen in APPsw mice. In three of four APPsw/NOS2-/- animals, cortical neurodegeneration was apparent after staining with Fluoro-Jade C, an anionic fluorescein derivative that specifically stains degenerating neurons (Schmued et al., 2005). Some neurons appeared to be undergoing apoptosis, as activated caspase-3 was detected in hippocampal neurons. Tau cleavage was also elevated in the same neurons compared to those in the APPsw mice.
The mechanism by which loss of NO promotes the accumulation of phospho-tau remains a mystery, but the authors propose two possibilities. Nitrosylation of tau inhibits tangle formation, so lack of NO might contribute to tau aggregation. Also, NO activates the Akt kinase, which inhibits tau phosphorylation via GSK3, and loss of NOS might release this inhibition.
In contrast to these results, NOS2 knockout in a double transgenic APPsw and mutant human presenilin background was previously reported to result in fewer plaques, longer lifespan, and less microglial activation. The authors of the current report speculate that the presenilin mutation, which has been implicated in NO-mediated toxicity (Hashimoto et al., 2004), could account for the different results.
As NO is generally accepted as a danger to neurons (see ARF related news story), one contribution of this work is to shade the view a bit to consider that the levels and timing of NO production may determine its impact. In light of the results that lowering NO over the life of a mouse causes the appearance of tau pathology, it could be that the induction of iNOS documented in Alzheimer disease (see Meda et al., 1995) might even represent a protective response, the authors speculate.—Pat McCaffrey
University of Illinois at Chicago
Nitric oxide signaling via the second messenger molecule cGMP is essential
for normal physiological brain function. NO itself is produced by the NO
synthase (NOS) family of enzymes, and NO bioactivity is also exerted by
metabolites of NO and by cGMP-independent pathways. In many brain regions,
activation of NOS and NO/cGMP signaling is a consequence of activation of
glutamatergic excitatory amino acid receptors and cholinergic muscarinic
receptor subtypes. The NO/sGC/cGMP signal transduction system is
considered to be important for modulating synaptic transmission and
plasticity in brain regions such as the hippocampus and cerebral cortex,
which are critical for learning and memory. We have long argued that NO
plays a decisive role in signal transduction cascades that are compromised
in AD, and therefore that drugs delivering NO bioactivity represent targets for
AD therapy. Acceptance of this argument has been impeded by a popular view
that NO is neurotoxic and causative in diseases such as AD.
NO, before realization of its essential role in human physiology, was best
known as a toxic atmospheric pollutant, and it is easy to demonstrate NO
toxicity toward brain cells in vitro. Excitotoxic neurodegeneration as
occurs in ischemic stroke has been linked with increased NO levels. It
has been proposed that disease states where chronic inflammation is a
causative factor would benefit from the use of inhibitors of iNOS, which is
induced under such conditions and can produce high levels of cellular NO.
Some researchers have espoused a simplistic paradigm of “bad” iNOS and “good” nNOS and eNOS. NO has been accepted as a causative factor
in brain diseases despite many reports demonstrating
anti-neurodegenerative properties and even neuroprotection in response to
insults such as amyloid-β (Aβ) neurotoxicity. There is also 130 years of
drug therapy with nitrates, drug sources of NO bioactivity, to support the
safety of NO and NO-based medications.
Recently, 1) nitrates have been reported to reverse cognition deficits
induced by cholinergic neurodegeneration and reduce amyloid load in
transgenic AD mouse models, and 2) direct links have been shown from
amyloid protein to neuronal dysfunction mediated by damage to NO signaling
The present work of Colton and coworkers serves to emphasize the
fallacy in the presumption that NO and/or iNOS causes
neurodegeneration. The work draws further links between loss of brain NO
bioactivity and both buildup of amyloid and hyperphosphorylated tau
deposits, the key pathological brain markers of AD. The Colton transgenic
mouse model superimposes an iNOS knockout on a standard AD transgenic
model leading to a pathophysiology that mimics aspects of human AD:
Interpretation of data from NOS knockout transgenics is especially difficult because knockout of one isozyme leads to compensatory changes in the remaining two isoforms. In this transgenic,
the researchers took care to note that eNOS protein levels increased and
nNOS levels fell, which may have also contributed to the observed
pathology. The observation of apoptotic cell death and raised caspase
activity may provide a pathway for formation of neurofibrillary tangles of
hyperphosphorylated tau initiated by loss of NO bioactivity. There remain
many questions to answer, in particular the mechanism of loss of iNOS
activity in the early stages of AD, but the work provides another strong
supporting plank for the argument that drugs that reinforce NO bioactivity
represent a valid and urgent approach to AD.
Colton and colleagues reported that crossing the APPsw transgenic line with the NOS2-/- mouse led to the novel Tg2576/NOS2-/- bigenic mouse which recapitulates the key pathological features of Alzheimer disease (AD), that is, β amyloid deposition, accumulation of hyperphosphorylated tau, and neuronal loss.
The significance of these results is twofold: they indicate that nitric oxide (NO) plays a role in the development of AD pathological hallmarks and they highlight the neuroprotective properties of NO. This view is supported by other findings obtained using NO-releasing derivatives of anti-inflammatory and antioxidant compounds (reviewed in Gasparini et al., 2004; 2005). In particular, HCT 1026 and NCX 2216, two NO-releasing derivatives of the nonsteroidal anti-inflammatory drug flurbiprofen, have been investigated in neuroinflammation and AD transgenic models. Besides showing improved anti-inflammatory activity (Prosperi et al., 2001; 2004), these compounds have additional properties of potential benefit for AD. For example, it has been shown that chronic administration of both HCT 1026 and NCX 2216 reduce β amyloid load in APPsw/presenilin 1 mutant double transgenic mice (Jantzen 2002; Van Groen and Kadish, 2005). Besides this, they also have peculiar activities that are mediated by NO. Specifically, these compounds act as PPARγ agonists on cultured rat microglia (Bernardo et al., 2005; 2006), inhibit Nf-kB activation (Fratelli et al., 2003), attenuate the loss of cholinergic cells following LPS-induced neuroinflammation, and reduce LPS-induced caspase-3, -8 and -9 activity in rat brain (Wenk et al., 2000).
It is worth noting that the kinetics of NO release and its concentration at the tissue level are critical to achieve such effects. The above-mentioned compounds release small amounts of NO with slow kinetics. Fast NO donors do not share the same effects (Gasparini, unpublished observations), indicating that the amount of NO and the timing are important to determine the neuroprotective effects of NO.
Altogether, these results show that approaches focusing on NO bioavailability and biogenesis could represent a valuable therapeutic strategy for AD treatment.
Bernardo A, Ajmone-Cat MA, Gasparini L, Ongini E, Minghetti L. Nuclear receptor peroxisome proliferator-activated receptor-gamma is activated in rat microglial cells by the anti-inflammatory drug HCT1026, a derivative of flurbiprofen. J Neurochem. 2005 Feb;92(4):895-903. PubMed.
Bernardo A, Gasparini L, Ongini E, Minghetti L. Dynamic regulation of microglial functions by the non-steroidal anti-inflammatory drug NCX 2216: implications for chronic treatments of neurodegenerative diseases. Neurobiol Dis. 2006 Apr;22(1):25-32. PubMed.
Fratelli M, Minto M, Crespi A, Erba E, Vandenabeele P, Del Soldato P, Ghezzi P. Inhibition of nuclear factor-kappaB by a nitro-derivative of flurbiprofen: a possible mechanism for antiinflammatory and antiproliferative effect. Antioxid Redox Signal. 2003 Apr;5(2):229-35. PubMed.
Gasparini L, Ongini E, Wenk G. Non-steroidal anti-inflammatory drugs (NSAIDs) in Alzheimer's disease: old and new mechanisms of action. J Neurochem. 2004 Nov;91(3):521-36. PubMed.
Gasparini L, Ongini E, Wilcock D, Morgan D. Activity of flurbiprofen and chemically related anti-inflammatory drugs in models of Alzheimer's disease. Brain Res Brain Res Rev. 2005 Apr;48(2):400-8. PubMed.
Jantzen PT, Connor KE, DiCarlo G, Wenk GL, Wallace JL, Rojiani AM, Coppola D, Morgan D, Gordon MN. Microglial activation and beta -amyloid deposit reduction caused by a nitric oxide-releasing nonsteroidal anti-inflammatory drug in amyloid precursor protein plus presenilin-1 transgenic mice. J Neurosci. 2002 Mar 15;22(6):2246-54. PubMed.
Prosperi C, Scali C, Barba M, Bellucci A, Giovannini MG, Pepeu G, Casamenti F. Comparison between flurbiprofen and its nitric oxide-releasing derivatives HCT-1026 and NCX-2216 on Abeta(1-42)-induced brain inflammation and neuronal damage in the rat. Int J Immunopathol Pharmacol. 2004 Sep-Dec;17(3):317-30. PubMed.
Prosperi C, Scali C, Pepeu G, Casamenti F. NO-flurbiprofen attenuates excitotoxin-induced brain inflammation, and releases nitric oxide in the brain. Jpn J Pharmacol. 2001 Jun;86(2):230-5. PubMed.
van Groen T, Kadish I. Transgenic AD model mice, effects of potential anti-AD treatments on inflammation and pathology. Brain Res Brain Res Rev. 2005 Apr;48(2):370-8. PubMed.
Wenk GL, McGann K, Mencarelli A, Hauss-Wegrzyniak B, Del Soldato P, Fiorucci S. Mechanisms to prevent the toxicity of chronic neuroinflammation on forebrain cholinergic neurons. Eur J Pharmacol. 2000 Aug 18;402(1-2):77-85. PubMed.