Subscribe to eTOC

Does a ‘Brain in a Dish’ Plus Herpes Equal Alzheimer's Disease?

Article In Brief

Exposure to herpes simplex virus type 1 among healthy human neurons in a 3-D bioengineered brain model led to pathologic dysfunction often found in Alzheimer's disease (AD). These results support the controversial idea that sporadic AD is often due to the long-term effects of early herpes virus infection.


Fluorescent immunostaining in which the red shows amyloid-beta fibrils. The green indicates HSV-1-infected cells, and the blue indicates dead cells stained with DAPI.

When healthy human neurons growing in a three-dimensional bioengineered brain model are exposed to herpes simplex virus type 1 (HSV-1), they develop pathology and electrophysiologic dysfunction reflective of Alzheimer's disease (AD), according to a new study published May 6 in the journal Science Advances.

The finding in a novel model for AD supports the controversial hypothesis that much of sporadic AD may be a long-term consequence of early herpesvirus infection.

HSV-1 is a widespread virus in human populations, and some epidemiological work has been done to try to correlate infection and treatment with subsequent risk of AD, most recently a paper from Taiwan of over 33,000 subjects, of whom over 8,000 had HSV infection. That study showed that the 20-year risk of developing some form of dementia was 2.5 times as likely among those who had been infected, and that antiviral treatment at the time of infection reduced the later risk of dementia.

While cautioning that the current report was as in vitro study, “these results support and amplify work from other labs, suggesting that viral infections like this could predispose toward certain kinds of dementia such as Alzheimer's disease,” said the principal investigator David Kaplan, PhD, professor and chair in the department of biomedical engineering at Tufts University in Medford, MA.


“This gives you something that doesnt exist today—the ability to do meaningful structural and functional screens in normal tissue, to determine important aspects of the disease process and potential treatments.”—DR. DAVID KAPLAN

But some experts not involved with the study are skeptical that the pathology is closely related to that seen in human AD.

Brain in a Dish

The current study is the result of two intersecting activities in the lab at Tufts University. One was the development of the 3D brain model. One was the development of the 3D brain, which over the past seven years his team has developed into a robust model for exploring questions of development and connectivity in healthy and diseased tissue.

The model brain is a squat cylinder, about six millimeters across and two millimeters high. The cylinder is a sponge formed from purified moth-silk, with a hollow center filled with a gel, usually collagen.

“We call the system the doughnut model,” Dr. Kaplan said. Neurons and astrocytes derived from human neural stem cells are allowed to grow in the sponge, and reach into the center to form synapses. “The system gives us nice control over the tissue, and we don't see significant necrosis,” even after many months, Dr. Kaplan added.

Most recently, Dr. Kaplan used the model to show that over a two-year period, cells derived from AD patients displayed phenotypic and electrophysiologic differences compared with those from healthy individuals.

In addition, the lab was able to demonstrate that Zika virus readily infected neurons derived from neural stem cells in two-dimensional culture, providing a valuable model for studying the disease and its treatment. First author on the new paper, Dana Cairns, PhD, a postdoctoral fellow in Dr. Kaplan's lab, brought the two approaches together to ask what might be the effects of HSV-1 infection in the 3D brain model.

“Our feeling was that exploring the effect of herpesvirus was a logical choice,” Dr. Kaplan said, “since it is very well known to infect and harbor in neurons, and it is widespread, and it can be handled safely in the lab.”

Study Design

The team began by infecting human neural stem cells with HSV-1, at virion:human cell ratios ranging from 1:1 to 1:100,000. Ratios of 1:100 quickly caused substantial cell death, and so a ratio of 1:10,000 was used for subsequent studies. In two-dimensional cell culture, infection led to the development of multicellular “plaque-like formations,” which contained amyloid-beta fibrils, and an average of 19 cell nuclei per formation. Phosphorylated tau aggregates were also observed. Cells bearing astrocyte markers displayed some phenotypic changes seen in the same cells in AD brain, and inflammatory mediators such as TNF-alpha and IL-6 were upregulated. The antiviral drug valocyclovir decreased formation of the plaque-like formations and reduced the levels of inflammatory mediators.


“This is a very interesting and novel paper that provides good support for herpes as an important cause of Alzheimers disease, rather than simply being associated with the disease, as some have suggested.”—DR. RUTH ITZHAKI

Next the team infused the doughnut model with human neural stem cells, followed by infection with HSV-1. Again they found amyloid-containing, multicellular, plaque-like formations in response to infection, with phosphorylated tau aggregates. Local field potential, a measure of the total electric current in a region, was reduced compared to healthy tissue, which Dr. Kaplan noted was reminiscent of the impaired brain functionality seen in AD patients.

The plaque-like formations seen here differ from the dense extracellular amyloid seen in the AD brain, Dr. Kaplan noted, possibly because the model tissue itself is not as dense as that of the brain.

The key point of the results, he stressed, is that unlike the large majority of AD models, but like the large majority of sporadic AD patients, the cells in this model did not carry any of the known genetic risk factors for AD. In this respect, the development of AD-like pathology may be more representative of the disease process in human AD than some other models.

“The real value of the model is that it provides a brain-like tissue in order to do rapid screening,” whether of the effects of other pathogens or of antivirals or other drugs to test for efficacy. While the current system would need to be dramatically scaled up, Dr. Kaplan suggested that would be quite practical with robotics. “This gives you something that doesn't exist today—the ability to do meaningful structural and functional screens in normal tissue, to determine important aspects of the disease process and potential treatments.”

Expert Commentary

“This is a very interesting and novel paper that provides good support for herpes as an important cause of Alzheimer's disease, rather than simply being associated with the disease, as some have suggested,” commented Ruth Itzhaki, PhD, emeritus professor in neuroscience and experimental psychology at the University of Manchester, UK. Dr. Itzhaki has studied the potential role of herpesvirus in AD for many years.

The role of apolipoprotein E genotype combined with herpesvirus infection has not been well studied, Dr. Itzhaki noted. “I hope there will be many more epidemiological studies done to further explore this issue,” said Dr. Itzhaki, who noted that she is involved in one such study.

Samuel E. Gandy, MD, PhD, professor of neurology and psychiatry at the Icahn School of Medicine at Mount Sinai, is less convinced that the model truly represents an Alzheimer's like disease process.


“Alzheimers disease pathology requires amyloid plaques that meet standard specifications.”—DR. SAMUEL E. GANDY

“I am generally supportive but the authors and readers should keep in mind that Alzheimer's disease pathology requires amyloid plaques that meet standard specifications,” Dr. Gandy said. In the current work, “I did not see anything that I recognized as a standard amyloid deposit.”

While Alzheimer's disease is a multicellular (and multi-molecular and multi-network) pathology, he added, “those are still areas of inquiry that researchers study. With Alzheimer's brain, one still must require structural proteinopathy in the form of tangles and plaques, and the densities of those structures are widely used to assess the severity of disease.” As such, he said, the results do not represent an advance on the model reported in 2014 in Nature by See Hoon Choi and Rudy Tanzi.

“An important, even irreplaceable role exists for three-dimensional induced pluripotent stem cell-derived, patient-specific models for brain disease is to provide access to living neuropathology that is otherwise inaccessible,” he said, rather than in humans or mice where the skull obstructs the view.

“The most important role I see is for such models is validating or even revealing events that only occur with native human molecules inside host human cells of various types. Still I cannot imagine modeling AD pathology without having in parallel access to a mouse model, so that I can go from genes to proteins to cells to tissues to electrophysiology to behavior.”


Dr. Itzhaki had no disclosures. Dr. Gandy has a grant from Constellation Wines for investigator initiated research using Constellation grapeseed extract to treat dementia. He is one of several founders of Recuerdo Inc., a company intended to offer concurrent access to multiple anti-amyloid and anti-cytokine interventions.

Link Up for More Information

• Dana M. Cairns, Nicolas Rouleau, Rachael N. Parker, et al. A 3D human brain–like tissue model of herpes-induced Alzheimer's disease Sci Adv 2020; 6(19):eeay8828.
    • Tzeng NS, Chung CH, Lin FH, et al. Anti-herpetic medications and reduced risk of dementia in patients with herpes simplex virus infections-A nationwide, population-based cohort study in Taiwan Neurotherapeutics 2018;15(2):417–429.
    • Choi SH, Kim YH, Hebisch M, et al. A three-dimensional human neural cell culture model of Alzheimer's disease Nature 2014;515:274–278.