Good nutrition may not be the first thing on our minds in the candy-heavy week after Halloween, but papers in the current issue of the Journal of Neuroscience will help to refocus attention on two vitamins and their relation to Alzheimer disease. The first report proposes a mechanism by which folate deficiency, a risk factor for AD, can elevate tau phosphorylation and harm neurons. The results, from Estelle Sontag and colleagues at the University of Texas Southwestern Medical Center in Dallas, indicate that defective protein methylation in folate-starved cells destabilizes the major tau protein phosphatase (PP2A), leading to enhanced tau phosphorylation. Overexpression of a PP2A methyltransferase restores phosphatase activity and reverses both tau hyperphosphorylation and cell toxicity. Taken together with similar in vivo data, the results suggest that PP2A is a prime target of folate deficiency in neurons, and its methylation could be a site of intervention in folate-related disorders and tauopathies.

The second paper focuses on the vitamin A metabolite retinoic acid, a well-known inducer of cell differentiation. Guo-Huang Fan and colleagues at the Virginia Commonwealth University School of Medicine in Richmond treated AD mice with the compound, and found that it could reduce Aβ deposition and improve cognitive function. “People have written reviews and suggested that retinoic acid might be useful in treatment of AD, but there was no evidence—we are the first to report data in an animal model,” Fan told ARF. The compound Fan and colleagues tested, all-trans retinoic acid, is already used to treat cancer, and the researchers hope they can quickly move the compound into trials for the treatment or prevention of AD.

Messed-up Methylation
Folic acid and its companion vitamin B12 are necessary for proper “one carbon” metabolism through pathways that shuttle methyl groups on and off DNA, lipids, and proteins. Low folate and low B12 are risk factors for AD and are linked to high homocysteine, itself a risk factor for cardiovascular disease and AD (see ARF related news story). In spite of much evidence in favor of the idea that folate and vitamin B12 protect against AD, a recent trial showed no effect of 18 months of supplements on people with mild to moderate AD (see ARF related news story).

Those results underscore the idea that the interactions among folate, B12, homocysteine, diet, and cognition are exceedingly complex, and not clear even now, after 150 years of study (for a thorough review see McCaddon, 2006). The new data from Sontag and colleagues helps to explain how changes in methylation of a specific protein lead to one of the hallmarks of folate deprivation, enhanced tau phosphorylation.

Using the N2a neuronal cell line, first author Jean-Marie Sontag and colleagues began by looking at methylation of the major tau phosphatase, PP2A. The catalytic subunit of PP2A requires methylation by the leucine carboxyl methyltransferase-1 (LCMT-1) enzyme before it can assemble into an active holoenzyme complex of A, Bα and C subunits. In folate-deficient cells, the investigators find, LCMT-1 is downregulated, demethylated PP2A C accumulates, the unassembled Bα subunit is destabilized, and tau phosphorylation increases. Overexpression of LCMT-1 or the Bα subunit reverses tau phosphorylation and renders the cells more resistant to cell death induced by folate deprivation. Knocking down LCMT-1, on the other hand, made cells more sensitive to folate deprivation. The de novo methylation of PP2A by LCMT-1 seemed to play a unique role in the cells, because its effects were not replicated when demethylation was prevented by knocking down the PP2A methylesterase PME-1.

To ask if similar changes in PP2A regulation occurred in vivo, the researchers examined mice that had been folate-deprived for eight weeks to lower serum folate and raise homocysteine. The mice had region-specific decreases in the S-adenosylmethionine/S-adenosylhomocysteine ratios in the brain, indicative of cellular folate deprivation. In the cortex and cerebellum, PP2A methylation was reduced by 75-80 percent, and both LCMT-1 and Bα proteins were downregulated. In regions where Bα was downregulated, significant increases in tau phosphorylation at the PHF-1 and Ser422 epitopes were observed.

The authors conclude that LCMT-1 appears to be a critical intermediate in the role of folate in the CNS. “Our cellular and mouse data reinforce the strong connection between LCMT-1-dependent ABαC holoenzyme formation/stabilization and tau regulation,” they write. They further propose that offsetting the neuronal loss of LCMT-1 or Bα could be a valuable therapeutic approach for tauopathies and folate-dependent CNS disorders.

Differentiating Treatment
As a cancer researcher, Guo-Huang Fan was struck by reports of cell cycle reactivation in neurons in the AD brain (e.g., see ARF related news story). The idea that postmitotic neurons try to proliferate and instead die by apoptosis made him ask whether drugs that promote differentiation and prevent proliferation might stop or slow the death of neurons in AD. Retinoic acid, the active metabolite of vitamin A, is a classical differentiation-inducing factor used to treat cancers. It occurs as cis and trans isomers, and Fan says they chose all-trans retinoic acid (ATRA) because it had a better clinical effect as a cancer treatment.

When first author Yun Ding and colleagues treated five-month-old APP/PS1 double transgenic mice (B6C3-Tg) for eight weeks with ATRA, they found significantly reduced Aβ deposition in the cortex and the hippocampus. The reduction in Aβ appeared to be due to a decrease in amyloid precursor protein (APP) processing, with no change in expression. ATRA treatment caused a reduction in APP phosphorylation, which could have affected processing. ATRA also decreased tau phosphorylation, and the compound appeared to decrease the activity of the APP and tau kinase, cyclin-dependent kinase 5 (Cdk5), though there was little effect on another tau kinase, GSK3β. Treated mice showed less inflammatory activation of microglia and astrocytes, and staining for neuronal markers suggested that treatment preserved synapses and neuronal size in the transgenic mice. Finally, the researchers showed the treated mice performed better in the Morris water maze test of learning and memory.

Fan and colleagues are keen to translate their animal results into human trials as soon as possible, although no experiments are planned yet. In addition, Fan noted, almost all differentiation-inducing factors used in cancer treatment prevent the activation of Cdk5 and GSK3β, suggesting that more of these drugs may be candidates for treating AD. His lab is now testing other such compounds, and he says they will have more results coming out soon.—Pat McCaffrey.

Sontag J-M, Nunbhakdi-Craig C, Montgomery L, Arning E, Bottiglieri T, Sontag E. Folate deficiency induces in vitro and mouse brain region-specific downregulation of leucine carboxyl methyltransferase-1 and protein phosphatase 2A Balpha subunit expression that correlate with enhanced tau phosphorylation. J. Neurosci. 2008 Nov 5;28(45):11477-11487. Abstract

Ding Y, Qiao A, Wang Z, Goodwin JS, Lee E-S, Block ML, Allsbrook M, McDonald MP, Fan G-H. Retinoic acid attenuates beta-amyloid deposition and rescues memory deficits in an Alzheimer’s disease transgenic mouse model. J. Neurosci. 2008 Nov 5;28(45):11622-11634. Abstract


Make a Comment

To make a comment you must login or register.

Comments on News and Primary Papers

  1. To test if a retinoic acid can reduce amyloid deposition in vivo, the authors used the APPswe/PS mutant transgenic mouse model. It is unclear to me which mouse model was used in the current study. Although it was stated in the text that line85 APPswe/PS1delta9 mice were used, articles describing several other APP transgenic mice models, including APPswe/PS1M146L transgenic mice, were cited. The authors stated that the mouse model used in this study begin to develop amyloid plaques as early as 2.5 months of age. In our experience, line85 APPswe/PS1delta9 mice do not have any amyloid plaque at such an early age. It appears to me that authors used APPswe/PS1M146L mice, rather than line85 APPswe/PS1delta9 mice.

    View all comments by Jungsu Kim
  2. Methylation and Tau
    The wealth of reports in the last decade confirming an association between homocysteine and Alzheimer disease hint that disturbed methylation might somehow relate to AD pathology (Smith, 2008; McCaddon and Hudson, 2007). A link between impaired methylation and neurofibrillary tangle formation was first proposed by Scott and Vafai in 2002 (Vafai and Stock, 2002). In support of this elegant hypothesis Obeid et al. found an association between phospho-tau and the ratio of the methyl donor S-adenosylmethionine (SAM) and its demethylated product S-adenosylhomocysteine (SAH) in the CSF of 182 patients with various neurological disorders (Obeid et al., 2007).

    SAH is a potent inhibitor of methyltransferase reactions, and last year Sontag et al. found that exposing neuroblastoma cells to SAH led to reduced methylation of PP2A (Sontag et al., 2007). Sontag’s group now show that folate deprivation downregulates PP2A carboxymethyltransferase expression in these cells, ultimately resulting in cell death. Protection is afforded by overexpressing either the methyltransferase or the Balpha regulatory subunit of PP2A, whereas knockdown of either protein accelerates the toxicity of folate deprivation. Reduced SAM and elevated SAH concentrations in folate deficient mice are also associated with enhanced tau phosphorylation in susceptible brain regions.

    Importantly, this work suggests an association between methylation status and one of the key pathological features of Alzheimer disease. Impaired methylation is likely to have other widespread effects in the nervous system. For example, alterations in the SAM/SAH ratio are also associated with PS1 and BACE upregulation and β amyloid deposition (Fuso et al., 2008). ”Tau-ists” and ”Bap-tists”’ should both be aware that the ”Meth-odists” may well be on to something!


    . B-vitamin deprivation induces hyperhomocysteinemia and brain S-adenosylhomocysteine, depletes brain S-adenosylmethionine, and enhances PS1 and BACE expression and amyloid-beta deposition in mice. Mol Cell Neurosci. 2008 Apr;37(4):731-46. PubMed.

    . Alzheimer's disease, oxidative stress and B-vitamin depletion. Future Neurology. 2007;2:537-47.

    . Folate and methylation status in relation to phosphorylated tau protein(181P) and beta-amyloid(1-42) in cerebrospinal fluid. Clin Chem. 2007 Jun;53(6):1129-36. PubMed.

    . The worldwide challenge of the dementias: a role for B vitamins and homocysteine?. Food Nutr Bull. 2008 Jun;29(2 Suppl):S143-72. PubMed.

    . Protein phosphatase 2A methyltransferase links homocysteine metabolism with tau and amyloid precursor protein regulation. J Neurosci. 2007 Mar 14;27(11):2751-9. PubMed.

    . Folate deficiency induces in vitro and mouse brain region-specific downregulation of leucine carboxyl methyltransferase-1 and protein phosphatase 2A B(alpha) subunit expression that correlate with enhanced tau phosphorylation. J Neurosci. 2008 Nov 5;28(45):11477-87. PubMed.

    . Protein phosphatase 2A methylation: a link between elevated plasma homocysteine and Alzheimer's Disease. FEBS Lett. 2002 May 8;518(1-3):1-4. PubMed.

  3. This is a well-designed and very exhaustive investigation of the effect of the administration of retinoic acid in a good murine model of Alzheimer disease. The beneficial effects of retinoic acid administration are plausible, consistent, and substantial. It is heartening to me to read this paper since previous publications by my colleagues and me (Lerer et al., 2008; Goodman et al., 2008; Goodman, 2006; Palha and Goodman, 2006; Goodman and Pardee, 2003) have suggested that increasing levels of retinoic acid in the brain might have therapeutic effects in late-onset Alzheimer disease through targeting of the retinol binding proteins, transthyretin, the retinoic acid receptors, or the retinoid receptor inhibitors (CYP26s). Administering retinoic acid analogs might also have benefit.

    To my knowledge there are no clinical trials in humans addressing the use of retinoids in Alzheimer disease. However, recently a preliminary drug trial has been published (Lerer et al., 2008) testing the effect of administering very low doses of the retinoid x receptor ligand, bexarotene—LGD1069 (Targretin)—to patients with chronic schizophrenia. The hippocampus and other brain regions affected in schizophrenia are similar to those implicated in Alzheimer disease. The drug was well tolerated, with the exception of an increase in serum cholesterol levels, and the trial demonstrated significant reduction of chronic symptoms. The unique insight of this trial was the extreme dilution of the retinoid analog. The usual clinical dosage of 300 mg/day was reduced to 75 mg/day, still achieving significant symptom reduction. Whether such a paradigm could prove successful in the treatment of Alzheimer disease remains to be demonstrated.

    See also:

    Goodman AB, McCaffery P, Palha JA, Simons, C, Pardee AB. Enhancement of brain retinoic acid levels, Chapt. 8 in H. John Smith, Claire Simons, & Robert D. E. Sewell, “Protein Misfolding in Neurodegenerative Diseases” p. 337-378, CRC Press Taylor & Francis Group New York 2008.


    . Bexarotene as add-on to antipsychotic treatment in schizophrenia patients: a pilot open-label trial. Clin Neuropharmacol. 2008 Jan-Feb;31(1):25-33. PubMed.

    . Retinoid receptors, transporters, and metabolizers as therapeutic targets in late onset Alzheimer disease. J Cell Physiol. 2006 Dec;209(3):598-603. PubMed.

    . Thyroid hormones and retinoids: a possible link between genes and environment in schizophrenia. Brain Res Rev. 2006 Jun;51(1):61-71. PubMed.

    . Evidence for defective retinoid transport and function in late onset Alzheimer's disease. Proc Natl Acad Sci U S A. 2003 Mar 4;100(5):2901-5. PubMed.

  4. Tauopathies are a group of diseases characterized by accumulation of tau protein. Tau protein has a novel physiological function in the brain—stabilizing the neurons. Alterations in the amount or the structure of tau protein might destabilize the microtubules, thus causing changes in subcellular structures like the lysosomes (1) or the mitochondria (2). Tau can be structurally modified by phosphorylation, glycosylation, oxidation, and crosslinking. These pathological forms of tau tend to form self-aggregates and thus forming the neurofibrilary tangles (NFTs). NFTs are typical findings in all tauopathies containing paired PHF comprising hyperphosphorylated tau (3).

    Alzheimer disease (AD) is the best known tauopathy that is characterized by accumulation of NFTs in the brain. In an animal model of neurodegenerative diseases, mice developed progressive accumulation of NFTs, neuronal loss, and memory decline (4). Suppressing the transgenic tau caused improvement in memory function, and neuron numbers stabilized. Unexpectedly, NFTs continued to accumulate. The authors concluded that tau accumulation rather than NFTs can cause cognitive decline or neuronal death in this model of tauopathy (4). Although, these results can not be extrapolated to humans without further research, they might suggest that lowering tau protein can improve cognitive function in humans as well (4).

    A very tempting hypothesis that would have implications for prevention and treatment of tauopathies is that lowering the modified forms of tau protein might protect the brain. The degree of tau phosphorylation and the phosphorylated residues regulate binding of tau to microtubules, thus affecting its function. There is a state of balance between kinase-mediated phosphorylation and phosphatase-mediated dephosphorylation. Tau can be phosphorylated by several kinases (proline and non-proline directed; glycogen synthase kinase 3, Cdk5, MAP kinase, JNK, PKC, calmodulin kinase II). Protein phosphatase 2A (PP2A) is the most important phosphatase in the brain acting on most phosphorylated sites of tau. PP2A is composed of three subunits: A, B, and C. The subunit Bα is involved in substrate recognition and is considered the regulatory subunit. The binary enzyme AC must be methylated on Leu-309 to be able to join the B subunit and form the functional enzyme. The methylation is mediated by leucine carboxy methyltransferase-1 (LCMT-1), an S-adenosyl methionine (SAM) dependent enzyme.

    In the recently published study by Sontag et al., the authors conducted several elegant experiments aiming at testing the effect of folate deprivation on the expression of hyperphosphorylated tau protein, PP2A enzyme and the methyltransferase required for its activation (5). The authors found that folate deficiency caused disturbed methylation potential in brain tissues. Furthermore, enhanced tau phosphorylation and cell death were related to downregulation of LCMT-1 and subsequent loss of ABαC complex. Folate deficiency caused an enhanced expression of the non-methylated form of PP2A enzyme. This condition did not change the existing PP2A, suggesting that binding of the ABαC complex is stable against demethylation by PME-1. In a previous study by Sontag et al., the authors found that mice fed a high methionine/low folate diet had higher brain S-adenosylhomocysteine (SAH) and lower expression of LCMT-1 (also called PP2A methyl transferase, or PPMT), thus leading to severe decrease of the steady state of PP2A (6). Another recent study found that SH-SY5Y neuroblastoma cells grown in a folate-deficient medium showed a decrease in the phosphatase activity, and this effect was reversible by adding SAM to the folate-deficient medium (7). Another interesting mechanism explaining the effect of folate deficiency on phosphorylated tau protein is related to enhancing activity of one or more of the kinases responsible for tau phosphorylation in the brain (7). This effect on the kinases is thought to be related to activating NMDA channels and thereby Ca++-dependent kinase pathways by homocysteine (Hcy). Therefore, folate deficiency is causally related to accumulation of tau protein; this effect is at least partly mediated by disturbed methylation status.

    The study by Sontag et al. opens new perspectives for future research dealing with the pathophysiology of dementia or studies aiming at developing protective or therapeutic measures (5). Methyl group metabolism is regulated by micronutrients such as folate, vitamin B12, and vitamin B6. Several methyl donors in human diet have been identified, such as methionine and choline. The role of folate ingested with the diet is to convert Hcy into methionine in the presence of methyl cobalamin. Methionine is activated in the presence of ATP and further converted into SAM. SAM is a methyl donor that participates in numerous biological reactions including DNA methylation, and metabolism of neurotransmitters and phospholipids. Betaine, a product of the nutrient choline, is an alternative methyl donor in the conversion of Hcy into methionine via homocysteine betaine methyltransferase. However, this pathway has probably no or a limited role in Hcy remethylation in the brain. Furthermore, all factors that cause accumulation of Hcy (renal insufficiency, vitamin B deficiency, and alcoholism and liver disorders) might cause disturbed methylation potential. This metabolic condition implies increased S-adenosyl homocysteine (SAH) that acts as a potent inhibitor of methyltransferases by preventing SAM binding.

    Deficiency of micronutrients is common in elderly people. Among other environmental factors, both hyperhomocysteinemia and B-vitamin deficiencies have been linked to increased risk for dementia and other age-related neurodegenerative diseases. Therefore, enhancing B vitamin status has a great potential to prevent metabolic conditions, associated memory disorders, and dementia. Future studies should test in-vivo the role of micronutrients in preventing or reversing phospho-tau accumulation in the brain. It is equally important in such studies to control for all sources of methyl donors in the diet.


    . Lysosomes are associated with microtubules and not with intermediate filaments in cultured fibroblasts. Proc Natl Acad Sci U S A. 1984 Feb;81(3):788-92. PubMed.

    . KIF1B, a novel microtubule plus end-directed monomeric motor protein for transport of mitochondria. Cell. 1994 Dec 30;79(7):1209-20. PubMed.

    . Microtubule-associated protein tau. A component of Alzheimer paired helical filaments. J Biol Chem. 1986 May 5;261(13):6084-9. PubMed.

    . Tau suppression in a neurodegenerative mouse model improves memory function. Science. 2005 Jul 15;309(5733):476-81. PubMed.

    . Folate deficiency induces in vitro and mouse brain region-specific downregulation of leucine carboxyl methyltransferase-1 and protein phosphatase 2A B(alpha) subunit expression that correlate with enhanced tau phosphorylation. J Neurosci. 2008 Nov 5;28(45):11477-87. PubMed.

    . Protein phosphatase 2A methyltransferase links homocysteine metabolism with tau and amyloid precursor protein regulation. J Neurosci. 2007 Mar 14;27(11):2751-9. PubMed.

    . Folate deprivation increases tau phosphorylation by homocysteine-induced calcium influx and by inhibition of phosphatase activity: Alleviation by S-adenosyl methionine. Brain Res. 2008 Mar 14;1199:133-7. PubMed.


News Citations

  1. Want to Keep Your DNA in Good Repair? Then Eat Your Spinach!
  2. The B Side—Vitamins Won’t Sharpen AD Brains, Trial Suggests
  3. Presenilins—Grinding Gears in the Cell Cycle?

Paper Citations

  1. . Homocysteine and cognition--a historical perspective. J Alzheimers Dis. 2006 Aug;9(4):361-80. PubMed.
  2. . Folate deficiency induces in vitro and mouse brain region-specific downregulation of leucine carboxyl methyltransferase-1 and protein phosphatase 2A B(alpha) subunit expression that correlate with enhanced tau phosphorylation. J Neurosci. 2008 Nov 5;28(45):11477-87. PubMed.
  3. . Retinoic acid attenuates beta-amyloid deposition and rescues memory deficits in an Alzheimer's disease transgenic mouse model. J Neurosci. 2008 Nov 5;28(45):11622-34. PubMed.

External Citations

  1. B6C3-Tg

Further Reading

Primary Papers

  1. . Folate deficiency induces in vitro and mouse brain region-specific downregulation of leucine carboxyl methyltransferase-1 and protein phosphatase 2A B(alpha) subunit expression that correlate with enhanced tau phosphorylation. J Neurosci. 2008 Nov 5;28(45):11477-87. PubMed.
  2. . Retinoic acid attenuates beta-amyloid deposition and rescues memory deficits in an Alzheimer's disease transgenic mouse model. J Neurosci. 2008 Nov 5;28(45):11622-34. PubMed.