17 September 2010. New data reported in this month’s PLoS ONE may inspire some Alzheimer disease researchers to consider expanding their postmortem analyses to include tissue from a curious source—the liver. Analyzing autopsy specimens from AD patients and age-matched elderly, Daniele Piomelli, University of California, Irvine, with collaborators in the U.S. and Italy, traced a shortage of the lipid docosahexaenoic acid (DHA) in AD brains to defects in a liver enzyme needed for biosynthesis of the neuroprotective fatty acid. The study “broadens the pathology of AD beyond the brain,” Piomelli told ARF. “Individuals who have lower levels of this liver enzyme seem to make less DHA, causing a decrease in brain DHA, which could interact with other pathological hallmarks, such as Aβ and tau, to eventually lead to the full-fledged disorder.” While the authors believe the data could help refine selection of at-risk participants for future clinical trials, others found the findings provocative yet somewhat hard to interpret.
People get much of their DHA from fatty fish and other foods. Moreover, dietary supplementation drives up blood levels of DHA (Stark et al., 2004), which correlates with reduced dementia risk in some (see, e.g., Schaefer et al., 2006), but not all (e.g., Laurin et al., 2003), epidemiological studies. However, average daily DHA intake varies considerably, and on the global spectrum, typical Western American fare contains scant supplies of DHA-rich foods (Cordain et al., 2005). “When there is little DHA in the diet, the liver kicks in and starts using shorter-chain fatty acids, so-called omega-3s, to make DHA,” Piomelli said. Studies suggest the liver is equipped to supply adequate amounts of DHA (Rapoport and Igarashi, 2009), and that brain metabolism of the phospholipid depends on both the diet and the liver (Rapoport et al., 2007).
To probe the link between DHA and Alzheimer disease, first author Giuseppe Astarita and coauthors measured levels of free (i.e., non-esterified) DHA in brain samples from 37 AD patients and 17 controls matched for age and postmortem intervals. Consistent with past reports, amounts of free brain DHA were lower in AD patients for all regions examined—temporal cortex, frontal cortex, and, surprisingly, the cerebellum, an area typically less vulnerable to AD pathology. This led the authors to suspect that defects in systemic DHA biosynthesis, rather than simply brain atrophy, could be contributing to decreased brain levels of the phospholipid.
The researchers analyzed the brain specimens for phosphatidylethanolamine, a DHA-containing membrane phospholipid, and found a dearth in the AD group in all sampled areas. Probing the brain tissue for four omega-3 fatty acids that serve as DHA precursors, they discovered only one, α-linolenic acid, that differed between the two groups, with AD patients having modestly higher levels.
Since α-linolenic acid gets converted to DHA primarily in the liver, the scientists took a closer look at this DHA synthesis pathway in liver samples from a separate cohort of 14 AD and nine control subjects with no signs of liver disease. Levels of free DHA and phosphatidylethanolamine were down in the AD liver specimens, mirroring the trend in brain tissue. Furthermore, reduction in liver DHA content correlated with cognitive decline, as captured by recent scores on the Mini-Mental State Examination and a global deterioration scale.
Whereas DHA levels were reduced in the AD liver samples, the amount of DHA’s immediate precursor, tetracosahexaenoic acid (which is also made from α-linolenic acid), was elevated. This suggested that shortage of dietary lipids was not the root cause of brain DHA deficiency, and gave further reason to blame perturbed biosynthesis.
To explore that possibility, Astarita and colleagues looked at expression of 10 genes known to be involved in liver DHA metabolism. AD liver contained lower mRNA levels of D-bifunctional protein (DBP), which catalyzes the final step of DHA biosynthesis from tetracosahexaenoic acid. The other nine other genes did not differ between AD and control groups. The findings “indicate that a deficit in DBP activity impairs DHA biosynthesis in liver of AD patients, lessening the flux of this neuroprotective fatty acid to the brain,” the authors write.
“I think it’s a solid paper. The data look good,” said Joseph Quinn, Oregon Health and Sciences University, Portland. “However, I think interpreting the data is a little tricky.”
He shared two concerns, one dealing with causality. “They show a relationship between reduced DHA synthesis in the liver and dementia severity. One interpretation, as put forth in the paper, is that change in DHA synthesis contributes to disease pathogenesis,” Quinn said. “However, it is equally plausible that people with AD are not as healthy as people without AD, and may have worse synthetic liver activity.” Quinn said the demonstrated correlation between liver DHA levels and cognitive performance “actually argues that a change in DHA synthesis is an effect, rather than a cause, of the disease. If the liver change were causative, you’d probably see the change at the beginning of the disease (i.e., when MMSE scores are around 25). You wouldn’t have to wait for the scores to drop to 10.”
As a second consideration, Quinn mentioned that the contribution of liver DHA synthesis to the link between DHA levels and dementia remains rather modest. “Dietary intake of DHA is a much greater contributor than endogenous synthesis in the liver,” he said.
Quinn led the Alzheimer’s Disease Cooperative Study’s (ADCS) recent 18-month trial of DHA, which failed to improve cognition in mild to moderate AD patients, relative to the placebo group (see ARF related news story). However, post-hoc analysis did suggest a hint of benefit in ApoE4-negative AD patients, and a smaller six-month trial sponsored by Martek Biosciences Corporation, Columbia, Maryland, a major supplier of dietary DHA, found that the phospholipid modestly improved learning and memory recall in healthy elderly with mild memory complaints. The Martek trial did not obtain participants’ ApoE genotypes.
Based on those findings, experts considered a possible DHA trial in less impaired subjects, i.e., people with mild cognitive impairment (MCI). “I proposed this to the ADCS, but they declined to proceed,” Quinn told ARF, noting, in addition, that the consortium’s confidence in the Martek data had not been quite strong enough to support the launch of a large prevention trial of DHA. A Martek representative told ARF the company is “developing a comprehensive strategy” based on its own research and on DHA’s possible benefit in E4 non-carriers in the ADCS trial, but said its “development plans are confidential.”
Meanwhile, other studies are addressing whether DHA-containing formulations can boost cognition in earlier-stage populations. One trial in France, led by Bruno Vellas and Sandrine Andrieu of the University Hospital, Toulouse,
is testing omega-3 fatty acid supplementation, with or without physical and cognitive exercise, in frail elderly. Other trials, sponsored by the European LipiDiDiet project, are offering SouvenaidTM, a milkshake containing DHA and other omega-3 fatty acids, to people with prodromal AD (see ARF related news story).
Given the recent findings, though, “it’s reasonable to hypothesize that when you give a mixture of omega-3 fatty acids to people with faulty DHA production, you’re not going to speed production. You’re going to end up clogging the system,” Piomelli said. “These individuals process other omega-3 fatty acids normally but cannot convert them to the final product, DHA.”
Piomelli said his group would like to explore the possibility of developing a peripheral biomarker for AD vulnerability. “It could be that people who have DBP deficiency—which could result from oxidative stress in the liver or from a DBP gene mutation—might be more prone to develop AD,” he said. The beauty of this biomarker would be that, unlike Aβ and tau pathologies, “you might actually be able to do something about it,” Piomelli said. “You could go in with a little bit more DHA.” Coauthor and UC Irvine colleague Carl Cotman added that, based on the new data, future DHA trials should exclude people with liver dysfunction.—Esther Landhuis.
Astarita G, Jung K-M, Berchtold NC, Nguyen VQ, Gillen DL, et al. (2010) Deficient Liver Biosynthesis of Docosahexaenoic Acid Correlates with Cognitive Impairment in Alzheimer's Disease. PLoS ONE 5(9): e12538. Abstract