Probing PPARγ? Agonists: Could Diabetes Drugs Treat Alzheimer's Disease?
Elena Galea and Douglas Feinstein led this live discussion on 16 December 2002. Readers are invited to submit additional comments by using our Comments form at the bottom of the page.
See the Preamble (.pdf) to this discussion
Look up abstracts of the Society for Neuroscience Conference in Orlando at http://sfn.scholarone.com/itin2002/index.html
View Transcript of Live Discussion — Posted 28 August 2006View Comments By:
Siegfried Hoyer — Posted 11 December 2002
Peroxisome Proliferator Activated Receptor-γ (PPARγ) is a nuclear hormone receptor that upon activation induces de-novo gene transcription, which then leads to increased sensitivity to insulin in some cells. This has provided a rationale for the use of PPARγ agonists to treat type II diabetes. Two agonists, pioglitazone and rosiglitazone—both members of the thiazolidinedione (TZD) class of drugs—are currently approved for the treatment of diabetes. The discovery that PPARγ reduces inflammation (Combs et al., 2000; Heneka et al., 1999; Ricote et al., 1999; Combs et al., 2001; Jiang et al., 1998), together with the commercial availability of PPARγ agonists, has raised interest in TZD drugs for the treatment of neurological diseases with an inflammatory component. Studies show that TZD-PPARγ agonists are highly protective in the MPTP mouse model of Parkinson's disease (Breidert et al., 2002), and in the MOG-induced experimental allergic encephalomyelitis model of multiple sclerosis (Feinstein et al., 2002; Diab et al., 2002; Natarajan and Bright, 2002; Niino et al., 2001).
Int' Veld et al. showed in 2001 that nonsteroidal antiinflammatory drugs (NSAIDs) protect against AD. The mechanism by which they do so is now an area of active debate (see, for example, live chat). Perhaps it is less well-known in the AD community that NSAIDs can also stimulate PPARγ pathways. This raises the question of whether exerting a similar effect with different drugs, namely TZD-PPARγ could open up novel treatment strategies for AD. As with statins, TZD drugs have been taken safely for years, but unlike statins, epidemiological and mechanistic research on the relevance of TDZ drugs in AD is only now picking up steam. It is not known whether TZD-treated diabetics are at lower risk of developing AD, but there are two ongoing clinical trials investigating the effect of pioglitazone and rosiglitazone on patients with AD.
At the 32nd Annual Society for Neuroscience last month in Orlando, several presentations dealt with the effect of PPARγ in Alzheimer's. Topics included:
1. the efficacy of PPARγ agonists in patients or animal models, as compared with NSAIDs;
2. the mechanism of protection; candidates here include inhibition of inflammation, regulation of Ab production or clearance, or direct anti-poptotic effects; and
3. whether the effect of TZDs is really mediated by PPARγ or is receptor-independent.
1. Studies in animal models and humans
The notion that inflammation contributes to the development of Alzheimer's disease arose from epidemiological studies showing that long-term treatment with NSAIDs such as ibuprofen reduces the risk of developing the disease (Int't Veld et al., 2001). The observation that ibuprofen inhibits amyloid plaque burden and plaque-associated inflammation in APP-overexpressing mice provided experimental evidence that NSAIDs could reduce pathological markers associated with Alzheimer's disease (Jantzen et al., 2002; Lim et al., 2000). Eddie Koo's laboratory has shown that ibuprofen and structurally related molecules can directly reduce the production of amyloidb42 in vitro (Weggen et al., 2001). In Orlando, Morihara et al., Golde et al., and Eriksen et al. (Abstracts 722.5, 722.6, 722.7) provided in-vitro evidence indicating that the ibuprofen effect could be due to the COX-independent allosteric modification of γ-secretase. These studies support the idea that NSAIDs provide protection in Alzheimer's primarily by regulating the processing of amyloid, and that the decrease in glial inflammation is secondary to the reduction of plaque content, though this continues to be hotly debated.
Several studies sought to determine whether PPARγ activation would mimic the effect of NSAIDs. Heneka et al. reported that both pioglitazone and ibuprofen mixed into the chow of APP-overexpressing mice (London mutation) for one week decreased insoluble, i.e., plaque-associated, Ab42, but not Ab40, by 20 percent (Abstract 722.10). Glial inflammation was also reduced. Yan et al. studied the effect of pioglitazone and ibuprofen administered orally for four months in 11-month-old transgenic mice bearing the Swedish APP mutation (Abstract 483.5). Ibuprofen reduced both amyloid plaque burden and soluble amyloid Ab42 by 60 percent. By contrast, pioglitazone had no effect on plaque burden but reduced soluble Ab42 by 50 percent, and Ab40 by 20 percent. It could be argued that the drugs may independently regulate amyloid production and plaque formation—the latter being the result of deposition and removal—and that pioglitazone may primarily target production. Whether the difference in length of treatment or the mouse strain accounts for the differences between the two studies remains to be elucidated.
Feinstein et al. presented data showing that PPARγ agonists reduce the brain's inflammatory response to injection of aggregated Ab42 (Abstract 123.2). In this model, previous ablation of the locus coeruleus, a noradrenergic nucleus that is the major source of noradrenaline in brain, leads to a robust cortical inflammatory response to Ab42. Loss of locus coeruleus occurs in the majority of Alzheimer's patients, leading Feinstein to propose that reduced noradrenaline levels might exacerbate the inflammatory responses to Ab in Alzheimer's disease. Ab injection induced robust IL-1b and iNOS expression, and these responses were reduced by coinjection of ibuprofen or the TZD ciglitazone, as well as by oral pioglitazone, a more therapeutically relevant experiment.
A study by Galea et al. tested the effect of ibuprofen and pioglitazone in mice overexpressing transforming growth factor b (TGF-b (Abstract 685.11). These animals display astrocyte and microglia activation, vascular deposition of amyloid detected by thioflavin, and vascular dysfunction. Seven-week-old mice received oral ibuprofen or pioglitazone for two months. The drugs decreased the glia activation, but not the amyloid deposition. This result suggests that antiinflammatory drugs cannot interfere with TGF-induced amyloidosis, perhaps implying that they would not be appropriate for the treatment of cerebral amyloid angiopathies.
Taken together, these presentations provide evidence that orally administered PPARγ agonists can reduce brain glia inflammation. They demonstrate that some TZD drugs—at least pioglitazone—can cross the blood-brain barrier. However, while ibuprofen clearly decreases Ab42 content in APP overexpressing mice, evidence for a comparable effect of TZDs remains preliminary and conflicting, and needs to be replicated. A caveat of the studies is that they tested only one TZD dose, which was chosen based on previous studies evaluating its insulin-sensitizing effects. Hence, higher doses might be more effective on AD-related pathology. Eriksen et al. (Abstract 722.7) proposed at the meeting that higher NSAID doses may be necessary to regulate amyloid metabolism than are needed to inhibit inflammation; the same may hold for TZDs. This could create potential safety problems in future AD trials, (see related news story).
A key question regarding the use of TZDs in AD is whether they induce functional recovery, and if so, when they should be taken with respect to pathology development. TZD-treated AD mouse models have not yet been analyzed for changes in the standard learning paradigms. The available epidemiology on ibuprofen has not been yet been replicated with any of the TZD-PPARγ agonists, even though several million type II diabetes patients have been prescribed TZDs for up to five years. (Pioglitazone and rosiglitazone have been on the market since mid-1999; troglitazone was withdrawn in 1999 for liver toxicity, but patients had been taking it for three years.) This begs the question whether diabetics on TZDs are less prone to develop AD.
Despite the lack of epidemiological data and a dearth of animal studies, the safety record of the drugs has eased approval of two NIH-funded clinical trials. One is a pilot study led by Gary Landreth and David Geldmacher to examine the effects of pioglitazone on cognitive function. Patients will be treated for 18 months; results are expected in 2004. The pioglitazone dose, 45 mg./day, is the highest allowable antidiabetic dosage, so as to stay within the FDA-approved levels. The other trial, conducted by Suzanne Craft, tests the effect of rosiglitazone, and is in phase II to end in 2003. Craft et al. released some preliminary results in Orlando. People with AD appeared to have improved verbal memory after six months on rosiglitazone, thus providing a first hint for therapeutic benefit of TZD-PPARγ agonists in humans (Abstract 822.4, see related news story). The authors attribute the recovery to the insulin-sensitizing effects of rosiglitazone. In figuring out the mechanism, one must bear in mind that rosiglitazone, unlike pioglitazone, does not cross the blood-brain barrier.
2. Mechanisms of action
Could TZD-PPARγ agonists—like ibuprofen—directly regulate γ-secretase? Sagi et al. reported in-vitro data that activation of PPARs had no effect on the processing of amyloid b (Abstract 193.3). How else, then, might TZDs reduce Ab content in vivo? Sastre et al. (Abstract 483.3) showed that proinflammatory cytokines increase the production of total APP and Ab40 and 42 in neuronal cell lines N2a and SK-N-SH that transiently overexpress APP (Swedish mutation). Both ibuprofen and pioglitazone reversed this increase. Interestingly, the cytokines increased the expression and activity of BACE, but not γ-secretase. These findings indicate that inflammation and amyloid production are intertwined phenomena, the idea being that amyloid plaques or soluble Ab trigger glia activation and the release of cytokines, which, in turn, would stimulate Ab production further. Based on these findings, we speculate for the sake of discussion that TZDs could reduce the production of Ab by inhibiting the cytokine-induced expression and activation of BACE.
The findings reveal two major mechanisms by which treatment with anti-inflammatory agents could be protective in AD. One is the reduction of Ab42 production via inhibition of γ-secretase, as done by ibuprofen. For their part, TZDs and other NSAIDs might suppress the deleterious effects of inflammation, including BACE activation or the oxidative damage wrought by activated microglia. The preventive effect of ibuprofen in Alzheimer's disease could be thus due to both inflammation-dependent and independent mechanisms. By contrast, any potential therapeutic benefits of PPARγ agonists may be limited to their capacity to reduce inflammation. Note, however, one study at the meeting that showed TZDs can rescue neurons from apoptosis triggered by overexpressing FAD-APP mutations (McPhie et al., Abstract 123.1). This suggests a protective effect of TZDs independent of inflammation.
3. Is there PPARγ in brain?
While pioglitazone clearly reduces glia inflammation in vivo, the existence of PPARγ in brain remains uncertain. Wanderi et al. reported absence of expression in normal animals, and up-regulation only in neurons after ischemia (Abstract 694.13). Galea et al. did detect PPARγ expression in control mice, and down-regulation after chronic treatment with pioglitazone, but only after the antigen was concentrated by immunoprecipitation, suggesting low levels of PPARγ expression in brain (Abstract 685.11). There could be novel PPARs in brain, or TZD drugs could have PPARγ-independent effects. Richardson et al. proposed the latter in Orlando based on their finding that PPAR-antagonists failed to inhibit the antiinflammatory effect of TZDs in vitro (Abstract 304.12). Recent papers concerning the actions of PPAR agonists in diabetic models have also suggested that some TZD effects may be receptor-independent (Brunmair et al., 2001; Chawla et al., 2001; Lennon et al., 2002).
Clearly, much remains to be learned about these drugs. For this online discussion, let's focus on the two major issues of effectiveness of TZDs in AD and their mechanism(s) of action. We propose the following questions:
1. What additional animal data is needed to support the case of PPAR agonists as a treatment for AD?
2. For clinical trial planning, what are—or were—the considerations in choosing from among available TZDs? How to set the dose? NSAID trials wrestle with the problem that the dose needed to see the Ab-reducing effect exceeds what people have been taking for standard indications such as arthritis, raising safety questions. Would the doses that work in diabetes be sufficient for AD? Like ibuprofen, would TZDs work better in the prevention than in the treatment of fully developed AD?
3. What are the neuroprotective mechanisms of TZDs, and are they comparable in AD, Parkinson's disease, and multiple sclerosis? What is the contribution of inflammation-dependent and independent mechanisms?
4. Are there other, perhaps metabolic, connections between diabetes and Alzheimer's that could shed light on the potential usefulness and modus operandi of TZD drugs in both diseases?
5. Since the preferred target of NSAIDs and TZDs is Ab42, not Ab40, can these drugs do any good in cerebral amyloid angiopathies, which mostly exhibit vascular deposits of Ab40? How do we find out?
6. Are the antiinflammatory effects of ibuprofen mediated by PPARs? In APP models, is the dampening-down of glia secondary to ibuprofen's reduction of plaque deposition, or due directly to its antiinflammatory actions? In this case, is it mediated by PPARs? We believe so, since ibuprofen has been shown to have COX-independent antiinflammatory effects (Tegeder et al., 2001), and at higher doses becomes a PPARγ agonist.
7. Since activated microglia appear to carry out beneficial (phagocytosis of Ab) and deleterious actions (release of NO), is the general inhibition of microglia activation by TZDs and NSAIDs "good" or "bad"? Are there pathways exclusive to one or the other mechanism that could be exploited therapeutically? Jantzen et al. have described this year an ibuprofen analog that could specifically increase the phagocytosis of microglia.
Breidert T, Callebert J, Heneka MT, Landreth G, Launay JM, Hirsch EC. Protective action of the peroxisome proliferator-activated receptor-γ agonist pioglitazone in a mouse model of Parkinson's disease. J Neurochem 2002;(82):615-624. Abstract
Brunmair B, Gras F, Neschen S, Roden M, Wagner L, Waldhausl W, Furnsinn C. Direct thiazolidinedione action on isolated rat skeletal muscle fuel handling is independent of peroxisome proliferator-activated receptor-γ-mediated changes in gene expression. Diabetes 2001;(50):2309-2315. Abstract
Chawla A, Barak Y, Nagy L, Liao D, Tontonoz P, Evans RM. PPAR-γ dependent and independent effects on macrophage-gene expression in lipid metabolism and inflammation. Nat Med 2001;(7):48-52. Abstract
Combs CK, Bates P, Karlo JC, Landreth GE. Regulation of b-amyloid-stimulated proinflammatory responses by peroxisome proliferator-activated receptor a. Neurochem Int 2001;(39):449-457. Abstract
Combs CK, Johnson DE, Karlo JC, Cannady SB, Landreth GE. Inflammatory mechanisms in Alzheimer's disease: inhibition of b-amyloid-stimulated proinflammatory responses and neurotoxicity by PPARγ agonists. J Neurosci 2000;(20):558-567. Abstract
Diab A, Deng C, Smith JD, Hussain RZ, Phanavanh B, Lovett-Racke AE, Drew PD, Racke MK. Peroxisome proliferator-activated receptor-γ agonist 15-deoxy-d (12,14)-prostaglandin J(2) ameliorates experimental autoimmune encephalomyelitis. J Immunol 2002;(168):2508-2515. Abstract
Feinstein DL, Galea E, Gavrilyuk V, Brosnan CF, Whitacre CC, Dumitrescu-Ozimek L, Landreth GE, Pershadsingh HA, Weinberg G, Heneka MT. Peroxisome proliferator-activated receptor-γ agonists prevent experimental autoimmune encephalomyelitis. Ann Neurol 2002;(51):694-702. Abstract
Heneka MT, Feinstein DL, Galea E, Gleichmann M, Wullner U, Klockgether T.
Peroxisome proliferator-activated receptor gamma agonists protect cerebellar
granule cells from cytokine-induced apoptotic cell death by inhibition of
inducible nitric oxide synthase.J Neuroimmunol 1999 Dec; 22: 156-68.
Int' V, Ruitenberg A, Hofman A, Launer LJ, van Duijn CM, Stijnen T, Breteler MM, Stricker BH. Nonsteroidal antiinflammatory drugs and the risk of Alzheimer's disease. N Engl J Med 2001;(345):1515-1521. Abstract
Jantzen PT, Connor KE, DiCarlo G, Wenk GL, Wallace JL, Rojiani AM, Coppola D, Morgan D, Gordon MN. Microglial activation and b -amyloid deposit reduction caused by a nitric oxide-releasing nonsteroidal antiinflammatory drug in amyloid precursor protein plus presenilin-1 transgenic mice. J Neurosci 2002;(22):2246-2254. Abstract
Jiang C, Ting AT, Seed B. PPAR-γ agonists inhibit production of monocyte inflammatory cytokines. Nature 1998;(391):82-6. Abstract
Lennon AM, Ramauge M, Dessouroux A, Pierre M. MAP kinase cascades are activated in astrocytes and preadipocytes by dPGJ2 and the thiazolidinedione ciglitazone through PPAR γ independent mechanisms involving ROS. J Biol Chem 2002. Abstract
Lim GP, Yang F, Chu T, Chen P, Beech W, Teter B, Tran T, Ubeda O, Ashe KH, Frautschy SA, Cole GM. Ibuprofen suppresses plaque pathology and inflammation in a mouse model for Alzheimer's disease. J Neurosci 2000;(20):5709-5714. Abstract
Natarajan C, Bright JJ. Peroxisome proliferator-activated receptor-γ agonists inhibit experimental allergic encephalomyelitis by blocking IL-12 production, IL-12 signaling and Th1 differentiation. Genes Immun 2002;(3):59-70. Abstract
Niino M, Iwabuchi K, Kikuchi S, Ato M, Morohashi T, Ogata A, Tashiro K, Onoe K. Amelioration of experimental autoimmune encephalomyelitis in C57BL/6 mice by an agonist of peroxisome proliferator-activated receptor-γ. J Neuroimmunol 2001;(116):40-8. Abstract
Ricote M, Huang JT, Welch JS, Glass CK. The peroxisome proliferator-activated receptor(PPARγ) as a regulator of monocyte/macrophage function. J Leukoc Biol 1999;(66):733-739. Abstract
Tegeder I, Pfeilschifter J, Geisslinger G. Cyclooxygenase-independent action of cyclooxygenase inhibitors. FASEB J 2001;(15):2057-2072. Abstract
Weggen S, Eriksen JL, Das P, Sagi SA, Wang R, Pietrzik CU, Findlay KA, Smith TE, Murphy MP, Bulter T, Kang DE, Marquez-Sterling N, Golde TE, Koo EH. A subset of NSAIDs lower amyloidogenic Ab42 independently of cyclooxygenase activity. Nature 2001;(414):212-216. Abstract
Submit a Comment on this Live Discussion