An extensive, multi-cohort, multi-network analysis of what goes wrong in the Alzheimer’s brain and why, has turned up none other than human herpes and other viruses. The big data analyses, a four-year project, is detailed in the June 21 Neuron. Led by Joel Dudley at the Icahn School of Medicine at Mount Sinai, New York, the authors can’t be sure if the microbes are causing molecular and genetic networks to run amok or are merely taking advantage of the disruption to flex their minuscule genomes. Even so, other scientists say the findings are cause for a rethink.
- Viral load linked to activation of gene networks in preclinical AD.
- Viruses may be driving production of Aβ.
- Is viral reactivation a cause or consequence of AD pathology?
“This is in line with the now-rapidly increasing evidence of viral infections as key drivers in the development of Alzheimer’s disease pathology,” wrote Hugo Lövheim, Umeå University, Sweden. The new findings tie viral load to increased expression of components of the amyloid pathway in people with preclinical AD. This fits with the related hypothesis that Aβ normally functions as an endogenous antimicrobial defense, Lövheim notes (full comment below). On that note, an upcoming paper in the same journal reports that Aβ fibrils neutralize herpesviruses in cells and animals. Rudy Tanzi and Robert Moir at Massachusetts General Hospital, Charlestown, co-led this second study, currently available as a “sneak peek” on Cell Press. “The combined story might be the beginning of a paradigm shift, in which low-grade, subclinical infection, e.g., by herpes-type viruses and other microbes, could seed plaques,” wrote Tanzi.
Others were more cautious. Charlotte Warren-Gash, from the London School of Hygiene and Tropical Medicine, noted that robust longitudinal population data sets are needed to assess the clinical relevance and generalizability of such findings, and how they relate to the trajectory of Alzheimer’s disease. John Hardy of University College London noted that the work is difficult to square with the occurrence of Alzheimer’s disease in Down’s syndrome and in all carriers of some APP and PSEN mutations (full comment below).
Viruses and AD. Multi-omic evidence links viral load to the molecular, pathological, and clinical manifestations of dementia. [Courtesy of Ben Readhead.]
Alzheimer’s has long been linked to viruses. In 1997, Ruth Itzhaki and colleagues at the University of Manchester, U.K., reported that, when present in the brain, the cold sore virus herpes simplex 1 (HSV-1) was a strong risk factor for the disease, and later the scientists located HSV-1 in amyloid plaques (Itzhaki et al., 1997; Wozniak et al., 2009). They also found that AD brains had much more human herpesvirus 6 (HHV-6) than did controls (Lin et al., 2002).
In longitudinal studies, Lövheim and colleagues reported that people with anti-HSV-1 antibodies in their blood were at higher risk for incident AD (Lövheim et al., 2014; Lövheim et al. 2014). Others have linked Epstein-Barr, cytomegalovirus, and herpes viruses to AD, as well (e.g., Carbone et al., 2014).
Recently, nationwide population-based studies in Taiwan reported that varicella zoster virus and HSVs are risk factors for dementia, and that antivirals can reduce that risk—in the case of HSVs by 80 percent (Chen et al., 2018; Tzeng et al., 2018; for commentary, see Itzhaki and Lathe, 2018). These population studies are important, said Itzhaki. “Proof of a microbe being a cause of a noninfectious disease can be shown by either treating with an agent that specifically targets the microbe, or by preventing it with a specific vaccine. Despite shortcomings, the Taiwan studies are essential first steps,” Itzhaki wrote to Alzforum.
That these viruses are common makes potential connections to dementia troubling. By some estimates, 90 percent of the population carries HSV-1 by age 70, and almost everyone is infected with HHV-6 in the first three years of life. These viruses can lie dormant for decades.
Despite the apparently ubiquitous threat, the Alzheimer’s field thus far has largely dismissed links to viruses (Feb 2011 webinar). Lack of mechanistic explanations have seeded doubt. “Previous studies of viruses and Alzheimer’s have been very indirect and correlative,” noted Dudley, who splits his time between Mt. Sinai and the ASU-Banner Neurodegenerative Disease Research Center at Arizona State University in Tempe. The new work attempts to change that. “We performed a sophisticated computational analysis using multiple levels of genomic information measured directly from affected brain tissue, which allowed us to identify how the viruses directly interact with, or co-regulate, known Alzheimer’s genes.”
Upregulated Host Genes. Host genes upregulated by the HHV-6A virus include ones linked to AD, including PSEN1, CLU, BIN1, and PICALM. [Courtesy of Readhead et al., Neuron 2018.]
Joint first authors Ben Readhead, Jean-Vianney Haure-Mirande, and colleagues built on molecular network analysis pioneered by Bin Zhang and co-author Eric Schadt, both also from Mount Sinai (May 2013 webinar on Zhang et al., 2013). Readhead is now at ASU-Banner Neurodegenerative Disease Research Center at Arizona State University, Tempe.
The scientists started out by looking for differences among gene-regulatory networks in both cognitively and pathologically healthy controls and people with preclinical AD. For that they used a data set created by researchers at the Translational Genomics Institute, Phoenix. Freely shared raw data was the foundation for a lot of the work in this Neuron paper. “It’s a success story for the idea of open data access and sharing data in its most raw form,” Readhead emphasized.
The preclinical samples were from people who had moderate neurofibrillary tangle pathology—Braak Stage II to IV—and moderate plaques by CERAD score, but no evidence of cognitive impairment (Liang et al., 2010). Readhead and colleagues focused on gene expression in areas known to be subject to profound neuronal loss in AD, namely the entorhinal cortex and the hippocampus.
The researchers first built probabilistic causal networks. In other words, by comparing control and preclinical disease samples, they identified groups of related genes likely to be involved in the disease process at this stage. To get a handle on what regulates those groups, they identified key drivers, i.e. genes that likely control the activity of large subnetworks within a given network.
Perhaps unsurprisingly, genes previously linked to AD popped up among the drivers, including APP, PICALM, and ABCA1. But here, also, is where the first hints of viral involvement emerged. When the scientists looked at the sorts of biological mechanisms that could explain the differences between the key drivers in each network, Readhead and Haure-Mirande realized that they shared a set of DNA motifs, i.e., sequences that bind a specific set of transcription factors known as C2H2 zinc fingers. C2H2-TFs have been implicated in viral biology. For example, the C2H2-TF SP1 binds Epstein-Barr virus proteins, regulates transcription of HIV genes, and promotes replication of human cytomegalovirus. This pointer to viral involvement emerged from an unbiased analysis.
Could viruses be driving causal networks in AD? There were other hints. “We had a laundry list of reasons why we thought this might be the case and got to the point where we thought we should really look into this carefully,” Readhead told Alzforum. “Once we realized we had a signal, the essential goal became to look across as many perspectives as possible for how viruses interact with AD biology. For example, is it a specific virus, group of viruses, or pan virus effect, i.e., just the fact of having a viral infection? The paper builds up these perspectives.”
For that, the researchers capitalized on the Accelerated Medicines Partnership-AD project. Coordinated by the NIH, this public-private partnership funds multi-omic profiling of large data sets. Readhead and colleagues started with samples from the Mount Sinai Brain Bank, looking for correlations between AD and viral RNA and DNA. HHV-6A and HHV-7 tended to be more abundant in the anterior prefrontal cortices (213 total samples) and superior temporal gyri (96 samples) of people with AD, likely AD, and possible AD than in controls. In some cases and brain regions, other viruses negatively correlated. For example, human adenovirus B1 and HHV-6B were more abundant in the anterior prefrontal cortices of controls than of cases. Readhead thinks expression of some viruses might compete with expression of others.
The correlation extended to individual viral elements, such as repeats that flank viral genes, and to other cohorts. In 151 temporal cortex samples from the Mayo Clinic and 247 dorsolateral prefrontal cortex samples from the Religious Orders Study at Rush University, Chicago, more HHV-6A and HHV-7 were found in people with probable AD than controls.
The correlation didn’t always hold. No difference in viral load emerged between AD and control samples from the Memory and Aging Project at Rush. Nonetheless, Chris Gaiteri at Rush found the cohort replication compelling. “Keep in mind that these cohorts are selected in different ways and from different parts of the country. If you don’t see replication it does not necessarily mean that your original observation is wrong, but to see replication in three different cohorts is pretty impressive,” Gaiteri said. Furthermore, viral load correlated with neuropathological and clinical facets of disease. Expression of certain HHV-6A genes in the prefrontal cortex correlated with clinical dementia rating, amyloid plaque density, and Braak score.
What could these viruses be doing in the AD brain? To address this, the authors turned to viral quantitative trait loci. These are host genetic variants that associate with viral abundance. They found 1,672 vQTLs across four brain regions that associated with 16 viruses. Once again, HHV-6A stood out, having 103 vQTL markers. These included genes involved in mucosal immunity, innate immunity, and viral sensing. For each virus, the authors found multiple vQTLs that associated with AD genetic risk variants, CDR score, or AD neuropathology. “These results indicate a significant overlap between the genetic basis for AD traits and the abundance of specific viral species and viral genomic features,” write the authors.
The researchers took it a step further. By setting vQTLs as “causal anchors,” they constructed directional relationships between viral species and host genes. In this way they built a new set of networks. These now are networks of host genes that regulate, or are regulated by, each virus in any of four different brain regions: the anterior prefrontal cortex, superior temporal gyrus, parahippocampal gyrus, and inferior frontal gyrus. By this analysis, three viruses—HSV-1, HSV-2, and HHV-6A—controlled host genes in all four regions.
Genes most commonly regulated by the viruses included BACE1, Fyn kinase, and PPAR-γ, all of which have been implicated in AD. For HHV-6A specifically, the host genes purportedly regulated by the virus had considerable overlap with AD and included genes involved in APP processing and Aβ metabolism, such as presenilin 1, clusterin, Bin1, and PICALM (see image above). Others are known to mediate immune responses to viruses. Fyn promotes interferon-γ1 production during viral infection, while PPAR-γ is known to suppress viral replication. Links between other viruses and other AD genes also emerged. The authors found that HAdV-C-induced expression of complement receptor 1, and that HSV-2 inhibited TOMM40.
All told, the findings hint that viruses that have lain dormant in the brain for decades may be active in AD, or that there might be some form of chronic low-grade infection that goes undetected.
Is this a cause or a consequence of the disease? “That’s the most important question,” said Readhead. “Are these viruses opportunistic bystanders in a compromised host, or do they accelerate the disease once the brain becomes dysfunctional?” Readhead suspects the latter. “I think it’s plausible that they impact how quickly disease progresses once established, though we don’t know for certain,” he said. “The striking overlaps between the sets of genes that viruses perturb and AD risk genes and APP processing genes are too compelling to dismiss.”
There is no love lost between neurons and viruses. Even outside of age-related neurodegeneration, the tiny parasites cause nasty pathology, from shingles to temporary paralysis and even fatal encephalitis. Among the different cell types in the brain, HHV-6A might affect neurons directly. When Readhead and colleagues used quantitative RNA-Seq analysis to tease out relationships between viral load, AD trait, and the abundance of specific cell types, they found that viruses correlated with loss of neurons, particularly in the superior temporal gyrus. They traced this to specific networks of genes, including suppression of the gene encoding miR-155, a micro RNA known for its neuroprotective effects. When they crossed miR-155 knockouts with APP/PS1 mice and tested four-month-old offspring, they saw that plaques were larger and more numerous than in APP/PS1 controls. The work suggests that viruses could directly affect plaque growth.
That fits with Tanzi and Moir’s idea of Aβ being antimicrobial. Previously, they reported that the peptide erects a physical barrier against invading bacteria and fungi, cocooning them in amyloid fibrils (May 2016 news). Others have reported that Aβ40/42 can prevent HSV-1 entry into cells (Bourgade et al., 2015). Now, the MGH researchers report in their upcoming paper that Aβ can bind an agglutinate HSV-1 and HHV-6 in cells and in mice, preventing acute encephalitis. Whether host networks have evolved to mobilize in response to viral threats by upping production of Aβ remains unknown.
How about therapeutic implications? Don’t reach for the Valtrex yet. It remains to be seen how viral load correlates with disease progression. “As Readhead et al. note, we cannot know the cognitive health trajectories of the ‘preclinical AD’ patients, and disentangling molecules involved in disease progression from those responsible for maintenance of brain function is complex,” noted Warren-Gash. “Linking electronic health records and –omics data may provide insights into potentially tractable mechanisms of AD pathogenesis, but many questions remain about when, how, and in whom interventions would be indicated,” she added (full comment below). In years past, Itzhaki has applied for funding for anti-viral AD clinical trials but was turned down.—Tom Fagan
- Herpes Simplex and Alzheimer’s—Time to Think Again?
- Can Network Analysis Identify Pathological Pathways in Alzheimer’s
Research Models Citations
- Itzhaki RF, Lin WR, Shang D, Wilcock GK, Faragher B, Jamieson GA. Herpes simplex virus type 1 in brain and risk of Alzheimer's disease. Lancet. 1997 Jan 25;349(9047):241-4. PubMed.
- Wozniak MA, Mee AP, Itzhaki RF. Herpes simplex virus type 1 DNA is located within Alzheimer's disease amyloid plaques. J Pathol. 2009 Jan;217(1):131-8. PubMed.
- Lin WR, Wozniak MA, Cooper RJ, Wilcock GK, Itzhaki RF. Herpesviruses in brain and Alzheimer's disease. J Pathol. 2002 Jul;197(3):395-402. PubMed.
- Lövheim H, Gilthorpe J, Adolfsson R, Nilsson LG, Elgh F. Reactivated herpes simplex infection increases the risk of Alzheimer's disease. Alzheimers Dement. 2015 Jun;11(6):593-9. Epub 2014 Jul 17 PubMed.
- Lövheim H, Gilthorpe J, Johansson A, Eriksson S, Hallmans G, Elgh F. Herpes simplex infection and the risk of Alzheimer's disease-A nested case-control study. Alzheimers Dement. 2014 Oct 7; PubMed.
- Carbone I, Lazzarotto T, Ianni M, Porcellini E, Forti P, Masliah E, Gabrielli L, Licastro F. Herpes virus in Alzheimer's disease: relation to progression of the disease. Neurobiol Aging. 2014 Jan;35(1):122-9. PubMed.
- Chen VC, Wu SI, Huang KY, Yang YH, Kuo TY, Liang HY, Huang KL, Gossop M. Herpes Zoster and Dementia: A Nationwide Population-Based Cohort Study. J Clin Psychiatry. 2018 Jan/Feb;79(1) PubMed.
- Tzeng NS, Chung CH, Lin FH, Chiang CP, Yeh CB, Huang SY, Lu RB, Chang HA, Kao YC, Yeh HW, Chiang WS, Chou YC, Tsao CH, Wu YF, Chien WC. 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 Apr;15(2):417-429. PubMed.
- Itzhaki RF, Lathe R. Herpes Viruses and Senile Dementia: First Population Evidence for a Causal Link. J Alzheimers Dis. 2018;64(2):363-366. PubMed.
- Zhang B, Gaiteri C, Bodea LG, Wang Z, McElwee J, Podtelezhnikov AA, Zhang C, Xie T, Tran L, Dobrin R, Fluder E, Clurman B, Melquist S, Narayanan M, Suver C, Shah H, Mahajan M, Gillis T, Mysore J, MacDonald ME, Lamb JR, Bennett DA, Molony C, Stone DJ, Gudnason V, Myers AJ, Schadt EE, Neumann H, Zhu J, Emilsson V. Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer's disease. Cell. 2013 Apr 25;153(3):707-20. PubMed.
- Liang WS, Dunckley T, Beach TG, Grover A, Mastroeni D, Ramsey K, Caselli RJ, Kukull WA, McKeel D, Morris JC, Hulette CM, Schmechel D, Reiman EM, Rogers J, Stephan DA. Neuronal gene expression in non-demented individuals with intermediate Alzheimer's Disease neuropathology. Neurobiol Aging. 2010 Apr;31(4):549-66. PubMed.
- Bourgade K, Garneau H, Giroux G, Le Page AY, Bocti C, Dupuis G, Frost EH, Fülöp T Jr. β-Amyloid peptides display protective activity against the human Alzheimer's disease-associated herpes simplex virus-1. Biogerontology. 2015 Feb;16(1):85-98. Epub 2014 Nov 7 PubMed.
No Available Further Reading
- Readhead B, Haure-Mirande JV, Funk CC, Richards MA, Shannon P, Haroutunian V, Sano M, Liang WS, Beckmann ND, Price ND, Reiman EM, Schadt EE, Ehrlich ME, Gandy S, Dudley JT. Multiscale Analysis of Independent Alzheimer's Cohorts Finds Disruption of Molecular, Genetic, and Clinical Networks by Human Herpesvirus. Neuron. 2018 Jul 11;99(1):64-82.e7. Epub 2018 Jun 21 PubMed.