CONFERENCE COVERAGE SERIES
Quebec City, Canada
23 – 26 June 2019
In light of recent work implicating human herpesviruses in AD, virologists invited Alzheimer’s researchers to join them at the 11th International Conference on HHV-6 and HHV-7, held in June in Quebec City, Canada. The herpesvirus experts had more questions than AD researchers had answers, but together they started to plot a path toward addressing them.
This is Part 1 of a two-part story.
In a rare meeting of two worlds, experts on human herpesviruses and Alzheimer’s disease shared the stage at a special workshop at the 11th International Conference on HHV-6 and HHV-7, held June 23 to 26 in Quebec City, Canada. The virologists hosted the gathering because recent studies from within the AD field implicating these exact viruses in the disease had piqued their interest. The meeting was a mix of review and some new data on viruses in the AD brain and other tissues, and what they might be doing there. After a day of discussion, the virologists and the Alzheimerologists reached a consensus: Herpesviruses do not cause AD. And yet, because they seed, and speed, amyloid plaque deposition and inflame the immune system, they likely press on the gas pedal to accelerate the disease process. The scientists also agreed that much more work needs to be done on how viruses affect the brain. Most importantly: Could targeting viral infection slow the onslaught of dementia? This is an entirely open question.
This confluence of fields signals a growing interest in how microbes might contribute to AD. This was for decades a fringe topic in the AD field but is now being taken more seriously.
In the 1990s, Ruth Itzhaki, now emeritus at the University of Manchester, U.K., first identified brain infection with herpes simplex virus (HSV-1), best known for causing cold sores, as a risk factor for AD (Itzhaki et al., 1997). Subsequent evidence suggested that recurrent reactivation of a latent infection, or low-grade persistent viral activity in the brain, could over time inflame and damage neurons, leading to neurodegeneration and dementia. The idea is being tested in a Phase 2 trial. It started in 2018 at Columbia University, and evaluates whether 18 months of daily valacyclovir (brand name Valtrex) slows cognitive decline in 130 people with mild AD who tested positive for HSV-1 or 2 (Devanand, 2018). A Phase 2 trial of the antiviral drug pleconaril is ongoing in Poland and the Czech Republic (Nov 2018 conference news).
A Roseolovirus. Human Herpesvirus 6. [Courtesy of Bernard Kramarsky, National Cancer Institute.]
There were hints of a role for other herpesviruses in AD, too. PCR analysis of postmortem brain found a higher prevalence of HHV-6 RNA and DNA in AD tissue than healthy brain (Lin et al., 2002), but serological studies produced mixed results (Carbone et al., 2014; Agostini et al., 2015).
Then, in 2018 a trio of papers grabbed the attention of a wide group of virologists. A multi-omics analysis of AD data sets from Joel Dudley, Icahn School of Medicine at Mount Sinai, New York, reported higher expression of HHV-6A and HHV-7 RNAs in AD brain samples, and that HSV-1, HSV-2, and HHV-6A regulated the expression of AD risk genes (Jun 2018 news). Around the same time, Robert Moir and Rudolph Tanzi, Massachusetts General Hospital, linked viruses to amyloid pathology. They showed that HSV-1 and HHV-6 could initiate and accelerate deposition of Aβ plaques, which sequester the viruses and render them less infectious in mice (Jun 2018 news). Lastly, an epidemiological study indicated that old people taking antivirals for an HSV-1 outbreak were protected against developing dementia (Tzeng et al., 2018; for commentary, see Itzhaki and Lathe, 2018).
Network News: A multidimensional analysis of human brain tissue found HHV-6A presence correlated with changes in expression of multiple genes related to AD. [Courtesy of Readhead et al., Neuron, 2018.]
In Quebec, Philip Pellett, Wayne State University School of Medicine, Detroit, opened the meeting by welcoming the newcomers, i.e. the handful of neurobiologists in the crowd of virologists. Himself a herpesvirus expert, Pellett described what’s known about HHV-6A, -6B, and 7, collectively called roseoloviruses, and disease. The viruses primarily infect, and persist in, lymphoid cells, though they are common inhabitants of brain tissue and CSF. It’s established that HHV-6B causes roseola in infants and toddlers, marked by a transient fever and characteristic rash. But proving causality for other roseolovirus diseases has been more difficult because they occur in adults who are already seropositive from earlier infection, and who carry different numbers of infected cells whose viral activity waxes and wanes over time.
Several neurologic inflammatory diseases have been associated with HHV-6B. In immunocompromised patients, reactivation of the virus causes encephalitis. Multiple lines of evidence support a role for HHV-6A and -6B in some cases of multiple sclerosis. The association of HSV-1 with dementia meets some, but not all of the criteria for causation, Pellett said. To his mind, the epidemiological data tying antiviral treatment to a lower risk of dementia provides the most compelling argument for the link.
Remarking on Dudley’s bioinformatics study, Pellett stressed the need for independent confirmation, as well as analysis of viral activity in individual patients and in single cells. Pellet asked many questions of the nascent virus-in-Alzheimer’s evidence. Among them:
Finally, Pellet cautioned investigators in the AD field not to assume that what goes for one herpesvirus goes for all, because different viruses interact in unique ways with the host immune system and cellular machinery.
Taking the stage for his plenary, Dudley welcomed Pellett’s questions and agreed it was important to maintain a healthy skepticism about the results. Recapping his group’s bioinformatics analysis linking viral RNAs to AD-related gene expression, pathology, and cognition, Dudley emphasized that the results are entirely correlative. They say nothing about whether the viruses cause or accelerate AD pathogenesis, or opportunistically grow as a consequence of disease.
Trying to parse cause and effect, Benjamin Readhead, Arizona State University, Tempe, the first author on Dudley’s paper, is delving deeper into their data set. He presented new work on how ApoE might affect the virus-AD relationship. The ApoE4 allele is the strongest genetic risk for late-onset AD, and reduces the disease’s age of onset. ApoE4 has pleotropic effects on the immune system and increases susceptibility to viral, bacterial and parasitic infections. People with two copies of the E4 allele have 15 times the risk of AD compared to people without one, whereas ApoE2 protects against AD.
Readhead found two instances where ApoE alleles correlated with virus abundance. First, people with one copy of E4 had more of the HSV-1 latency-associated transcripts than those with none, and those with two had higher still. Second, E2 carriers had lower expression of the HHV-6B transactivator gene than non-carriers.
The number of E4 alleles also signaled higher neurofibrillary tangle density. In ongoing work, causal modeling was consistent with HSV-1 mediating at least part of ApoE4’s effect on tangles, meaning that ApoE4’s association with higher tangle abundance may be due to its influence on viral infection. That is consistent with Itzhaki’ s work showing that HSV-1 infection boosts risk of AD in ApoE4 carriers, and with Moir’s data on the amyloidogenic effects of virus particles. Readhead suggested that lack of HHV-6B virus in ApoE2 carriers might explain some part of the protective effects of that allele.
Echoing Pellet’s remarks, the virologists in the audience wanted to know a lot more about the infection. Do Dudley and Readhead know if the viral RNA comes from blood? Where was the organism in the brain? What cells are infected? Readhead said the virally regulated gene profiles gave some hints. HHV6A and 6B tended to be linked to neuronally expressed genes, as well as oligodendrocytic and microglial genes, while HSV 1 was associated with microglia gene expression. “That’s indirect evidence, and comes from tissue homogenates. We really need to go down to the single-cell level for the expression data. We also need more comprehensive immune profiling in the periphery," he said.
In the meantime, the group is studying viral infection in three-dimensional cerebral organoids, which enable them to directly measure how viruses change gene expression in single cells, and how that affects plaque and tangle formation. Readhead believes the problem of how viruses interact with each other also needs to be studied. Almost all people are infected with HHV-6B and approximately half have HSV-1, as well. “In isolation, one might not have an effect, but jointly, they could become a different beast,” he said.
The story could get even more complicated. “There are probably a dozen viruses implicated in some corner of the analysis,” Readhead told Alzforum. The existing analysis is limited in that the available RNA-expression data is skewed toward mammalian sequences; the microbiome profile is sparse. “The data was optimized to characterize human gene expression, so in looking for microbial sequences we are rifling through the junk drawer,” he said. “We need to expand the profile to look for viruses, bacteria, anything we can pick up.”
Steven Jacobson is chief of the Viral Immunology section at the National Institute of Neurological Disorders and Stroke, Bethesda, Maryland. He has studied HHV-6 extensively as a trigger for multiple sclerosis. At the NIH, Jacobson was tasked with exploring the possible connection of HHV-6 with AD, which he did by analyzing tissue from some 800 postmortem brains. He tapped the same cohorts Dudley had analyzed, but used a different computational technique to detect viral RNA. Jacobson applied the PathSeq algorithm, optimized to detect human pathogens, on the RNA-Seq data from the Mount Sinai Brain Bank and the Religious Orders Study/Memory and Aging Project (ROSMAP).
Jacobson detected HHV-6A and -6B sequences in 1.3 percent of the samples, and the frequency was the same in AD and healthy controls. Likewise, he detected no correlation between viral load and plaque pathology or cognition. In addition, Jacobson performed a sensitive PCR for HHV-6 DNA, developed to study MS, on 346 samples from the Johns Hopkins Brain Resource Center (Blauwendraat et al., 2019), and 87 saliva samples from a Swiss AD cohort. Again, his team did not find much. Overall, about 3 percent of the samples tested positive, with no difference between control and AD. For comparison, these scientists find the virus in about 10 percent of brains samples from people with MS. “We don’t see that overrepresentation in brains from people with AD,” he said.
These results contradict the Dudley study, in which Readhead found HHV-6 RNA in about 30 percent of people in both groups, and higher levels of viral transcripts in the AD group, Readhead told Alzforum. An earlier study reported HHV-6 in 70 percent of samples of AD brain, and 30 percent of unaffected (Lin et al., 2002). Other studies report baseline infection rates of 25 to 35 percent (Chan et al., 2001; Prusty et al., 2018). Readhead told Alzforum he found it difficult to explain the discrepancies, without looking more closely at the different methodologies used, but said, “I’m glad someone else is starting to look at this.”
Jacobson emphasized that his results do not rule out a role for the virus. “It’s fascinating that pathogens can nucleate amyloid plaques, but that doesn’t mean you have to find the virus in the brain at the time of disease,” Jacobson said. Also, HHV-6 is tricky to track, he added. Its copy number is low, and it forms focal infections that may be missed when sampling tissue. Besides, HHV-6 may be but one of many viruses that rile up the immune system as people age. “Maybe it’s one of multiple triggers for AD. It may be related, but it’s not likely to be the cause of disease,” Jacobson said.
Jacobson would like to see additional bioinformatics analyses of more samples, and mechanistic studies. “There is much more work to be done on Alzheimer’s. The collective knowledge of the virology field should be brought to bear on the problem,” he said.—Pat McCaffrey
This is Part 2 of a two-part story.
At a workshop during the 11th International Conference on HHV-6 and HHV-7, held June 23 to 26, 2019, in Quebec City, Canada, experts on human herpesviruses treated Alzheimer’s researchers to a lesson on the intricacies of viral biology, and what that might mean for the brain. Viruses disrupt mitochondria, autophagy, and even synapses, according to new data presented at the meeting. While some of the work was preliminary, it signaled the rapidly growing interest of virologists, immunologists, and neurobiologists alike in understanding the interaction of viruses and the innate immune system with the brain.
How do herpesviruses access the brain? One route goes through the nose. Olfactory neurons form a bridge from the outside environment into a person’s brain. They project from the nasal mucous membrane across two synapses into AD-susceptible areas of the entorhinal cortex and hippocampus, with no blood-brain barrier in the way. Olfactory neurons can serve as a conduit into the brain for pathogens, including herpesvirus and oral bacteria, that have been implicated in AD and Parkinson’s disease (Jan 2019 news; May 2017 news).
Lavinia Alberi Auber, University of Fribourg, Switzerland, probed the olfactory system for signs of HHV-6. Auber studies age-related decline in olfaction, a nonspecific harbinger of future AD, PD, or dementia with Lewy bodies (DLB). Auber has seen neuropathological changes that may underlie this gradual loss. Studying postmortem olfactory tissue from 38 people who had symptoms ranging from mild cognitive impairment to severe AD, Auber noticed a prominent and progressive tauopathy in olfactory neurons. She also detected a weak amyloid pathology in the synaptic terminal in the olfactory bulb that looked to her like intracellular amyloid deposits. These could be a response to nerve damage or infection, she believes.
Using immunohistochemistry, Auber also detected certain viral proteins, which indicate a new infection or viral reactivation, only in AD patients. In people with early AD, these antigens were limited to a few neurons. In more advanced cases, staining appeared as puncta scattered across the whole olfactory tract, and co-localized with phosphorylated tau. One synapse away, in the olfactory region of the entorhinal cortex, the viral antigens co-localized with amyloid plaques. In several autopsy cases, she found neurons and microglia that tested positive for HHV6A scattered throughout the entorhinal cortex.
The results suggest that HHV-6A, and -6B, may spread from the nose through the olfactory circuit into the brain. It remains to be seen if this contributes to, or is an effect of, AD progression in these patients. One complication of the study is that herpesviruses can reactivate after death, and Auber agreed that postmortem delay may have precipitated this in some samples. The work needs to be confirmed and replicated in more subjects, she said.
Conveniently, olfactory neurons can be biopsied in living people. Auber would like to do this to screen for HHV-6 infection. She is evaluating saliva as a less invasive source for detection of specific viruses, and for monitoring changes in the saliva microbiome over the course of Alzheimer’s disease.
Robert Moir of Massachusetts General Hospital in Boston reviewed his work with Rudy Tanzi on the anti-microbial function of amyloid peptides. Soluble Aβ oligomers bind to bacteria, fungi, viruses, then fibrillize, trapping invaders in a sticky amyloid net that stops their spread, they’ve shown (May 2016 conference news). In this way, viruses and other microbes instigate and accelerate the formation of amyloid plaques (Jun 2018 news; May 2019 news). Ruth Itzhaki’s group found HSV-1 DNA in amyloid plaques, but no one has looked for HHV-6 yet, Moir said (Wozniak et al., 2009). Fibrils not only sequester pathogens, they kill them by piercing their membranes and complexing with copper to create an oxidative burst, the researchers postulate. Moir said his group can watch antigens disappear over a few days in their in vitro systems, as plaque-produced oxidizers destroy the epitopes. That could make it harder to identify pathogens in autopsy tissue, he said.
“Amyloidosis and AD pathology in general are emerging from this interface between pathogens and the innate immune system,” Moir said. “Aβ doesn’t spontaneously aggregate; it needs a seed, and microbes provide that seed. It’s likely not one pathogen, but several that can seed.” The waning of adaptive immunity as people age could explain why AD arises late in life, even though many viral infections begin early. “The blood-brain barrier loses integrity with age, and we become susceptible to neuroinflammation and new infections. As we pass age 40, adaptive immunity starts to decline and innate immunity takes up the slack,” Moir believes.
Roberta Rizzo, an immunologist and microbiologist at the University of Ferrara, Italy, takes a similar view. “I don’t think a primary infection is causing dementia. If so, we should find viremia or antibody titers in AD, but this is not the case. We don’t find huge infections in the brain. Perhaps reactivation of a latent infection is enough to aggravate the immune system in a brain that is no longer able to control inflammation,” she told Alzforum. She recently reported that HHV-6 infection of peripheral, immune, natural-killer T cells induces expression of ApoE, a gene expressed in the brain mostly by astrocytes under normal conditions (Rizzo et al. 2019). She is now studying the response of microglia to HHV-6 infection in vitro.
For neurons, HHV-6 infection affects the structure of synapses, according to work presented by Margot Mayer-Proschel, a neuroscientist at University of Rochester, New York. Her lab also established that expression of the HHV-6A latency-associated transcript U94A in human oligodendrocyte precursor cells disables their migration to sites of axonal injury (Campbell et al., 2017). This could contribute to demyelination in multiple sclerosis. Mayer-Proschel traced the migration defect to U94A’s ability to induce cytoskeletal dysfunction.
The cytoskeleton also supports dendritic structures more generally. When Mayer-Proschel expressed U94A in purified cortical neurons, the cells’ dendritic arborization diminished. They did not lose material for dendrites; they expressed all the building blocks but did not assemble them. In these U94A-expressing neurons, otherwise sub-toxic doses of soluble Aβ assemblies induced beads to form in dendrites and reduced the complexity of dendritic branching. To Mayer-Proschel, this suggests HHV-6A infection could modify disease by destabilizing neuronal dendrites. This raises the possibility that individual differences in the brain virome could contribute to variation in disease progression in the same way a person’s genetic profile does, Mayer-Proschel speculated.
Herpesviruses hijack other cellular pathways implicated in AD, too. For example, many viruses antagonize autophagy though the cellular regulator Beclin 1 (Munz, 2011). Autophagy slows the buildup of amyloid but the process is impaired in mouse models of amyloidosis (Rocchi et al., 2017; Pickford et al., 2008). In Quebec, Xiaonan Dong, an immunologist at the University of Southwestern Texas Medical Center, Dallas, showed that HSV-1 also suppresses autophagy, and that this inhibition is required for the virus to survive and replicate.
Mitochondrial insufficiency is a common theme in neurodegeneration and again, viruses may play a role. In human T cells, HHV-6 infection degrades the mitochondrial membrane potential, weakens oxidative phosphorylation, and fragments the mitochondrial network. Those results, presented by Christine Birdwell, Texas A&M Health Science Center, College Station, suggest the virus damages mitochondria to ensure its own replication. Whether mitochondrial distress occurs, and might help drive AD pathogenesis, in neurons or microglia infected with HHV-6 remains to be seen.
Do animal models of HHV-6 offer insight into its potential role in AD? Because the virus does not infect mouse cells, researchers have taken two approaches to study the virus in vivo. In one, they added the human gene for the HHV-6 receptor, CD46, into the mice (Reynaud et al., 2014). Cynthia Liefer, an immunologist at the Cornell University College of Veterinary Medicine, Ithaca, New York, described those mice. After an intracranial injection of virus into the right hemisphere, the animals developed persistent infection and chromosomal integration of the virus, most prominently in Purkinje cells in the cerebellum and in serotonergic neurons of the Raphe nuclei in the brain stem. The mice develop behavior changes reminiscent of anxiety and depression.
The second option is to use the murine roseolovirus, a natural mouse pathogen genetically related to HHV-6. Tarin Bigley of Washington University, St. Louis, described neonatal infection of wild-type mice, which led to autoimmune disease in adults. Bigley told Alzforum that he is collaborating with David Holtzman’s lab at WashU on studies to infect 5XFAD mice with this virus.
Finally, marmoset monkeys can be infected with HHV-6 and they mount an immune response to it. Steven Jacobson, National Institute of Neurological Disorders and Stroke, Bethesda, Maryland, has used these primates to study the role HHV-6 in MS. His group found that monkeys infected with HHV-6A did not get MS, but did get severe experimental autoimmune encephalomyelitis (EAE). In HHV-6A infected monkeys, brain lesions developed faster, along with a burst of virus production in response to the EAE stimulus. The animals with HHV-6A also died sooner (Leibovitch et al., 2018). Jacobson considers this the strongest evidence that this virus can trigger MS. He told Alzforum he is starting a collaboration to study HHV-6 infection in a marmoset model of AD.
At the end of the day, the virologists urged AD researchers in the room to adopt an attitude of healthy skepticism about the role of HHV-6 in AD, and to continue to research the question. Those efforts are getting some backing, as the NIH recently designated infectious etiology of AD as a high-priority research topic. This makes some funds available to tackle the many outstanding questions raised at the conference, said Miroslaw Mackiewicz of the National Institute on Aging. Foreign institutions are eligible to apply.—Pat McCaffrey.
Comments
No Available Comments
Make a Comment
To make a comment you must login or register.