In the first, JoAnne McLaurin, Peter St. George-Hyslop, and colleagues at the University of Toronto show that certain orally delivered cyclohexanehexol (aka inositol) stereoisomers can block the accumulation of soluble Aβ oligomers in the brain of transgenic mice. The compounds reverse memory deficits (as measured by performance in the Morris water maze), reduce plaque load, and reverse other signs of Aβ pathology. The results strengthen the case that high-molecular-weight oligomers of Aβ (like Aβ*, see ARF related news story) play a major role in producing memory deficits in mice, and pave the way for the testing of these inositols to prevent or reverse Alzheimer disease in people. A phase I trial of their most effective compound, scyllo-inositol, has just been launched under the name AZD-103.

The in vivo work with inositol stereoisomers follows on first author McLaurin’s previous observations that Aβ oligomerization is enhanced by phosphatidylinositol lipids. She discovered that several isomers of inositol itself compete for the lipid binding and prevent Aβ oligomer formation in vitro and toxicity in cell culture (McLaurin et al., 1998 and McLaurin et al., 2000).

It’s important to note that inositol comes in eight possible stereoisomeric forms, of which three are found in brain. McLaurin’s earlier work indicated that myo-inositol, the common stereoisomer that makes up inositol phospholipids and is widely available as a nutritional supplement, has no effect on Aβ aggregation. The current paper also reiterates that myo-inositol does not affect pathology in the mouse model, so forget loading up on nutraceutical myo-inositol—it will have no benefit.

It is the other isomers that appear more promising. In a prophylactic treatment regimen, TgCRND8 mice were fed epi- or scyllo-inositol ad libitum beginning at 6 weeks old until they were 4 or 6 months old. A treatment regimen was also done which consisted of dosing 4-month-old mice for 1 month. In both cases, the mice that got scyllo-cyclohexanehexol showed a significant reduction in brain Aβ levels, with decreases in soluble or insoluble Aβ40 or Aβ42 ranging from 20 to 60 percent. These reductions correlated with improved performance in the water maze compared to untreated animals. In fact, the scyllo-treated mice navigated the water maze just as well as normal nontransgenic controls after several days of testing. Measurements of synaptic loss and glial inflammation also showed improvement, and there was a reduction in mortality in the scyllo treated mice. Overall, the epi-stereoisomer showed lower activity.

To confirm if the inositol isomers were inhibiting Aβ aggregation, McLaurin and colleagues measured soluble oligomeric Aβ in brain, using the oligomer-specific antibody from Charles Glabe’s lab (see ARF related news story). This antibody recognizes soluble Aβ assemblies larger than 40 kDa, and dot blots of brain homogenates showed that scyllo treatment decreased amount of brain Aβ it picked up. Western blotting confirmed that brain homogenates from the scyllo-treated mice had less of the high-molecular-weight Aβ species and increased lower-order species, particularly trimers. Finally, dose response studies demonstrated that scyllo-inositol caused progressive decline in the immunoreactive oligomers, which was accompanied by significant behavioral improvement in the water maze and a decrease in the number of plaques.

While it’s not clear if the Aβ oligomer that McLaurin observes corresponds to the Aβ* recently described by Karen Ashe and coworkers, the current work fits right in with that study’s conclusion that accumulation of soluble aggregated forms of Aβ could be responsible for memory impairment in AD.

Transition Therapeutics in Toronto has just started a phase 1 trial of scyllo-cyclohexanehexol (AZD-103), according to company CEO Tony Cruz. They are expediting development and expect to be in a large phase 2 trial in AD patients by early next year if all goes well. AZD-103 follows another aggregation inhibitor, 3-amino-1-propane sulfonic acid (Alzhemed) which is currently in phase 3 trials, with results expected in early 2007.

Another way to rid the brain of toxic Aβ plaques and oligomers is to let the immune system do the work. But since the unexpected toxicity of an Aβ peptide immunogen halted the Elan vaccine trial, researchers have been looking for a safer stimulant. Now, a paper from the lab of Yoh Matsumoto at the Tokyo Metropolitan Institute for Neuroscience, along with Matthias Staufenbiel at Novartis in Basel, Switzerland, presents data on a DNA vaccine that might fit the bill. In their paper, to be published online in PNAS, first author Yoshio Okura and colleagues describe a nonviral Aβ DNA vaccine which reduced Aβ load in APP23 mice even when the mice were vaccinated later in life. Importantly, the researchers found no inflammation or T cell response in either the APP23 mice or in wild-type mice given the vaccine.

The vaccines tested consisted of plasmids that that express Aβ1-42 alone, or with a leader sequenced attached to increase secretion, with or without an immunoglobulin Fc sequence to maintain stability. In cells, all three constructs produced Aβ, but only the latter two supported secretion. In a prophylactic trial, vaccine was administered by intramuscular injection starting at 3-4 months of age, before amyloid appears in the APP23 mice. In this mode, all three vaccines dramatically reduced amyloid in the frontal cortex of mice at 7 months to 15-30 percent of the levels of untreated mice. By one year, the two secretion-competent vaccines produced reductions of 30-50 percent in brain Aβ load, while the Aβ peptide construct had no effect. Reduction was long-lived, since it was still apparent at 18 months.

In a therapeutic scheme, the mice that got the vaccines when they were 1 year old showed decreased amyloid deposition in the cortex and hippocampus at 18 months. In fact, there was no significant difference between therapeutic and prophylactic protocols. Both paradigms also decreased intracellular Aβ deposition in cortical pyramidal neurons, an early event in AD.

The vaccines raised anti-Aβ antibody titers in both APP and normal mice by about two- to fourfold, much lower than the increase (up to 10,000-fold) seen after immunization with Aβ peptides. Yet even at these low levels, reduction of Aβ burden was significantly correlated with antibody titers. The researchers could detect no Aβ-responsive T cells or histological signs of brain inflammation in either normal or transgenic mice after DNA vaccination. They suggest that the fact that DNA vaccines do not require adjuvants may be an advantage over peptide immunization by giving more control over the strength of the immune response. The steady, low level of protein expression driven by the plasmid vaccine generates a “gentle and quiet” immune reaction, they say, but one that is nonetheless sufficient to drive down amyloid burden. Behavioral studies will be necessary to determine if this kinder, gentler approach is really tough enough to improve mental functioning.

Any successful treatment for AD may have to deal with tau hyperphosphorylation in addition to Aβ deposition. Treatment of pure tauopathies like frontotemporal dementia-17 will also require taming the aberrant phosphorylation of mutant tau that leads to neurodegeneration. In another paper this week in PNAS online, Hanno Roder, from Sirenade Pharmaceuticals in Martinsried, Germany, along with Michael Hutton and colleagues at the Mayo Clinic in Jacksonville, Florida, show that they can tackle tau phosphorylation head on with a new kinase inhibitor. The compound, a brain-permeable derivative of the natural product K252a, shuts down mutant tau phosphorylation and prevents the onset of motor impairments in transgenic mice expressing the P301L mutant human tau. At the same time, they show the treatment does not reduce the number of neurofibrillary tangles in the mice. The results join other recent studies pointing to tau hyperphosphorylation, leading to the formation of soluble aggregates as the pathogenic lesion in these mice (see ARF related news story, ARF news story, and Spires et al., 2006) and suggest the new inhibitor or related compounds may be useful for treating tauopathies.

Starting with K252a, first author Sylvie Le Corre and the team optimized the kinase inhibitor for oral availability and brain penetration. In its modified form, the new compound, SRN-003-556, inhibited a range of kinases including Erk, Cdc2, GSK3β, PKA, and PKC with about equivalent potency, and prevented phosphorylation of tau at multiple residues in cells. To look at in vivo effects, the researchers used JNPL3 transgenic mice, which develop neurofibrillary tangles, neuronal loss, and motor deficits as they age due to the human tau P301L transgene. Treating the mice at the first sign of motor deficits delayed the development of severe deficits as measured by a wire hang and a beam balance test.

In the brain and spinal cord of these animals, they found a 64 KDa form of soluble aggregated phospho-tau, and showed it was decreased in the treated animals. In both treated and untreated groups, the level of 64 KDa tau, and not the amount of neurofibrillary tangles, correlated with the stage of phenotypic disease for each mouse. The results echo those of the Aβ aggregation inhibitors in implicating soluble aggregates as the species of interest for therapeutic attack.

Kinase inhibitors could be useful in treating Parkinson disease as well, according to a paper from Mark Cookson and colleagues at the National Institute on Aging in Bethesda, Maryland, and collaborators in London. In a report online in the Neurobiology of Disease, Cookson and colleagues show that the kinase activity of the PD gene LRRK/dardarin is required for the mutant protein to form inclusion bodies and cause cell death.

The LRRK/dardarin gene, which accounts for up to 6 percent of inherited PD, encodes a large protein with multiple motifs, including a GTPase domain and a kinase domain. PD-causing mutations are found throughout the protein, but it is not clear how they work to produce disease.

To find out, first author Elisa Greggio expressed a number of dardarin mutants in cells, and replicated a previous finding (Smith et al., 2005) that the mutants formed inclusions at a much higher rate than wild-type protein, which rarely showed such aggregates. She then made kinase-inactive mutants of the PD proteins by introducing a triple alanine substitution in the active site. When these proteins were introduced into cells, they formed inclusion bodies less frequently than wild-type proteins. Similar results were obtained when they tested the ability of the mutants to kill SH-SY5Y or primary rat neurons: kinase dead versions of several dardarin PD mutants lost their ability to kill the cells as measured by TUNEL staining and morphological changes.

Using immunostaining, the researchers showed that dardarin is in fact present in cell bodies and processes in mid-brain melanized neurons in both normal and PD brain. These results suggest that their neuronal model may be relevant to disease, and that inhibitors of the dardarin kinase activity should be considered potential therapeutic agents for patients with LRRK2 mutations, and possibly also sporadic PD.—Pat McCaffrey.

References:
McLaurin J, Kierstead ME, Brown ME, Hawakes CA, Lambermon MHL, Phinney AL, Darabie AA, Cousins JE, Rrench JE, Land MF, Chen F, Wong SSN, Mount HTJ, Fraser PE, Westaway D, St George-Hyslop P. Cyclohexanehexanol inhibitors of Aβ aggregation prevent and reverse Alzheimer phenotype in a mouse model. Nature Medicine. 2006 June 11; Advance Online Publication. Abstract

Okura Y, Miyakoshi A, Kohyama K, Park I, Staufenbiel M, Matsumoto Y. Nonviral Aβ DNA vaccine therapy against Alzheimer’s disease: Long-term effects and safety. PNAS. 2006 June 20; 103: 9619-9624. Abstract

Le Corre S, Klafki HW, Plesnila N, Hubinger G, Obermeier A, Sahagun H, Monse B, Seneci P, Lewis J, Eriksen J, Zehr C, Yue M, McGowan E, Dickson D, Hutton M, Roder HM. An inhibitor of tau hyperphosphorylation prevents severe motor impairments in tau transgenic mice. PNAS Early Edition. Proc Natl Acad Sci U S A. 2006 Jun 20;103(25):9673-8. Epub 2006 Jun 12. Abstract

Greggio E, Jain S, Kingsbury A, Bandopadhyay R, Lewis P, Kaganovich A, van der Brug MP, Beilina A, Blackinton J, Thomas KJ, Ahmad R, Miller DW, Kesavapany S, Singleton A, Lees A, Harvey RJ, Harvey K, Cookson MR. Kinase activity is required for the toxic effects of mutant LRRK2/dardarin. Neurobiol Dis. 2006 May 31; [Epub ahead of print]. Abstract