"We have not studied the sporadic form of Alzheimer's disease sufficiently enough, and yet this makes up the major population of the patients," said Knuesel. "Now we have a model that would allow us to study, in a morphological context, the processing changes and upstream factors involved in initiation of plaque and tangle pathology—almost impossible to do in a transgenic mouse."

Knuesel and colleagues' work suggests that inflammation plays a prominent role in the sporadic form of the disease, perhaps even initiating it. They first found that in four-month-old triple-transgenic mice, a single systemic immune reaction—induced by injecting the viral mimic polyinosinic:polycytidylic acid (poly[I:C])—caused a dramatic boost in the plaque and tau pathology observed 11 months later. "I've never seen that much plaque deposition in triple-transgenic mice at that age," said Cindy Lemere, Brigham and Women's Hospital, Boston. "Giving that injection early in life led to very strong acceleration of AD-like pathology, showing that early exposure to an immune stimulus can fast-forward pathology."

Could an early immune challenge elicit AD-like pathology in wild-type mice as well? Knowing that some neurodevelopmental disorders might have their roots in gestation, the researchers wondered if the same might be true of Alzheimer's. The team found a window in late gestation when injecting mothers with poly(I:C), permanently altered their pups' immune systems. Circulating levels of several inflammatory cytokines remained elevated throughout the offspring's lives. The animals had more APP, and produced more Aβ42 and somatodendritic, hyperphosphorylated tau than controls, in which neuronal tau was predominantly axonal. A Y-maze test also demonstrated profound working memory deficits in these offspring at 22 months of age.

Knuesel and colleagues gave those same mice a second injection of poly(I:C) at 12 months of age, and looked at their brains three months later. Amyloid plaques—detected by an antibody against rodent Aβ40/42—were pronounced in the entorhinal (ERC) and piriform cortices. Relative to controls, there was also more Aβ immunoreactivity within the hippocampus, particularly in regions receiving projections from the ERC. The pattern suggests that the deposits started in the cortices and expanded to the hippocampus, much like they do in the early human form of Alzheimer's disease. The plaques resembled the diffuse ones found in humans, Knuesel said. Hyperphosphorylated tau also accumulated in the mouse neurons. While these aggregates were not similar to human neurofibrillary tangles (they were Gallyas silver stain-negative), Knuesel thought NFTs might form as the mice age. The mice also showed elevated microglial activation and some evidence of microglia degeneration. The researchers are conducting behavioral tests on the 12- to 15-month-old mice to test if they have cognitive deficits. The results suggest that an immune challenge early in development, followed by a second in adulthood, puts the brain at risk for AD-like pathology.

"The very early immune challenge seems to have long-term consequences for aging in the brain," said Lemere. "It appears that the brain is then set up so another immune challenge later in life makes the brain more susceptible to neurodegeneration."

Knuesel believes that the cytokines and chemokines produced by the mother as a result of infection cross the placental barrier, enter the fetus, and make their way across the underdeveloped blood-brain barrier to the central nervous system of the offspring. In late gestation, the fetus also contributes to the cytokine production. Knuesel thinks the elevated cytokines alter microglia and genes involved in early brain development and immune functions. Cytokines may also hamper division of microglial precursors so that there are fewer microglia in old age, making them less able to phagocytose protein deposits.

"I think it's a moderately low elevation of inflammatory cytokines that may damage the brain chronically," said Knuesel. Genetic factors and repeated infections in old age may also infer risk. "Then a systemic infection can be the last little kick that sends the system all the way downhill," Knuesel added.

Inflammation has long been thought to be involved in AD pathology, but it is not clear whether immune responses are a cause or a result of the disease (see ARF related news story). Microglia have been found to phagocytose plaques and clear them from the brain (see ARF related news story on Simard et al., 2006 and ARF related news story on El Khoury et al., 2007). But evidence that systemic inflammation in people with Alzheimer's disease speeds up cognitive decline suggests that some immune responses exacerbate AD (see Holmes et al., 2009). In addition, genomewide association studies reported that genes related to the innate immune system confer risk for AD (see ARF related news story on Harold et al., 2009 and Lambert et al., 2009).

There have been mixed reactions to Knuesel’s findings. At several conferences where she presented this work, scientists pointed out that rodent Aβ is not known to aggregate, leading them to question the nature of the amyloid deposits in these wild-type mice. "It would be helpful to clarify that they are extracellular versus intracellular, as well as the exact composition of β amyloid in these deposits, " said Lemere. Knuesel says she has done the biochemistry but is keeping the results under wraps until they are published, which she expects to be in the coming months.

This isn't the first time that Alzheimer's-like plaques have been induced in wild-type mice by infection. Chlamydia pneumoniae has been reported to induce amyloid plaques, for instance (see Little et al., 2004). Herpes simplex virus infection also causes Aβ42 to deposit in wild-type mice (see Wozniak et al., 2007 and recent ARF Webinar on herpes simplex virus as a possible trigger of AD). There is still much work to do before a link can be made between infection and the human form of Alzheimer's disease, including testing AD patients for immune markers and closely reviewing epidemiological data, Knuesel said.

"I don't think there's any epidemiology out there that has looked at that carefully," said Bruce Lamb of the Cleveland Clinic in Ohio. "But I think maybe it is time to do that kind of study." He added that the mechanism of how poly(I:C) works in these mice also needs further exploration.

"Knuesel’s two-hit strategy to develop a wild-type AD model has a sound scientific background because there is now genetic evidence for the role of innate immunity [in AD], and there is epidemiological and clinical evidence that systemic inflammatory mediators could contribute to the development of clinical Alzheimer’s," said Piet Eikelenboom of the Vrije Universiteit in Amsterdam, The Netherlands. But he is not yet convinced that the plaques seen in the rodents are comparable to the ones found in humans. He said he will reserve judgment until the new data are published.

Michael Chumley of the Texas Christian University in Fort Worth and his team are also looking into the effects of simulated bacterial infections on adult wild-type mice. In a poster presentation, graduate student Marielle Kahn reported that peripheral injections of lipopolysaccharide (LPS)—a bacterial coat component—over seven days in four- to six-month-old C57BL/6J mice (a common wild-type lab strain) led to immediate cognitive deficits. Animals had trouble with contextual fear learning and spent less time in the platform zone in the Morris water maze test. The mice were no longer sick at the time of testing, nor did they have elevated levels of some common pro-inflammatory cytokines left in their systems. However, their Aβ42 levels in the hippocampus were significantly elevated.

Taken together, these findings and Knuesel's support the idea that infection may trigger Alzheimer’s. "The implication could be that systemic inflammation is an instigating factor for Alzheimer's disease," said Chumley.

Chumley and colleagues do not yet know if mice fully recover cognitive function after the LPS injections, or if they continue to decline. The group will check for cognitive deficits a few weeks after injection, and will test the effects of repeated simulated infections on the mice. Recurring infections often occur in the older population, Chumley said, and the team wants to know if chronic infections and inflammation could drive Alzheimer's-like deficits and pathology.—Gwyneth Dickey Zakaib.