. 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.


Please login to recommend the paper.


  1. Microbes and Alzheimer's Disease: Paradigm Change

    In this article Readhead et al. provide fresh impetus to the link between infection and Alzheimer’s disease (AD). A causal link has long been suspected (Itzhaki et al., 2016), underlined by the finding that the signature of AD, Aβ, is now established as an antimicrobial defense peptide (Soscia et al., 2010; Kumar et al., 2016). 

    Because the majority of the adult population already harbors subclinical infections with diverse viruses, notably including herpes simplex viruses types 1 and 2 (HSV-1/2), as well as human herpes viruses 6A, 6B, and 7 (HHV-6A/6B/7), and others such as Epstein–Barr virus (EBV), a declining immune system with age might permit reactivation of erstwhile silent viruses, exacerbating pathology from another cause, but playing no causal role. In the alternative, such viruses might be essential triggers.

    Herpesviruses: Evidence for Causation
    To place this in context, evidence has recently emerged directly and causally linking infection with HSV-1, and possibly by HSV-2, to AD development. Briefly, population data from Taiwan argue that 10-year AD development could be prevented, in 90 percent of treated patients, with aggressive antiviral medication at the time of acute herpes infection (Tzeng et al., 2018; Itzhaki and Lathe, 2018). This work establishes beyond reasonable doubt that herpesviruses can cause AD, as pioneered by Itzhaki (Itzhaki et al., 1997), but raises two key questions: (i) which herpesvirus? and (ii) how many cases? The new paper by Readhead et al. tackles both questions head-on.

    The New Results: Overabundance of HHV in AD
    To date most studies have focused on HSV-1. HSV-1 DNA is widely found in AD brain, as well as in control tissue, and association with Aβ plaques in AD brain has been demonstrated (Wozniak et al., 2009). By contrast, other herpesviruses such as HHV-6 have been predominantly implicated in multiple sclerosis (Leibovitch and Jacobson, 2014), and not so far in AD. The striking new results from Readhead et al. cut across this compartmentalization. In a long and often complex paper, they first report that transcripts of multiple herpesviruses are increased in AD brain—predominantly HHV-6A, HHV-6B, and HHV-7, although there was also evidence of over-representation of HSV-1 and HSV-2 transcripts. Notably, the overabundance was not restricted to a few odd cases, but from the data at hand appears to be a more general feature of AD. Unfortunately, other potential AD-linked pathogens including spirochetes (Miklossy, 2015) do not appear to have been systematically surveyed. Second, DNA variants involved in innate immunity and antiviral responses (that have previously been highlighted as risk factors for AD) were found to correlate with viral abundance and degree of AD pathology. Third, they show that there was also increased abundance of HHV-6A DNA, pointing to active viral replication in AD brain. Further work in this exciting compendium of results, including network and multivariate analysis, highlighted HHV-6A and HHV-7 as central associates of AD development. The authors rightly conclude that their data support an important role of these viruses, particularly HHV-6A and HHV-7, in AD development.

    Like HSV-1/2, both virus types have a very high prevalence of more than 80–90 percent in adults worldwide (Tanaka-Taya et al., 1996). Moreover, similarly to HSV-1/2, HHV-6 and -7 are well-known causes of viral encephalitis, in particular in immunocompromised individuals, and have also been shown to be associated with demyelinating brain diseases (Pietiläinen-Nicklén et al., 2014). In contrast to the eight other known human herpesviruses, HHV-6A and 6B integrate into the host genome with a very high frequency in both somatic and germline cells and can thus be inherited. The detection of HHV-6A/B or HHV-7 DNA is therefore not necessarily a sign of active infection, only the expression of lytic HHV-6A/B or HHV-7 transcripts is indicative of this, as in the study by Readhead and colleagues.

    In a very recent addition to the debate, Moir and colleagues (Eimer et al., 2018), upcoming also in Neuron, confirm that the AD signature protein, Aβ, binds to HSV-1 and HHV-6 surface glycoproteins, and mediates agglutination and protection against virus challenge, further reinforcing the link between herpesviruses, Aβ, and AD.

    Drivers or Passengers? Differential Tropism Could Point the Way Forward
    Readhead et al. highlight the challenge ahead: “Distinguishing the earliest drivers of disease from the ‘opportunistic passengers’ of a multi-decade neurodegenerative process is especially formidable ...” In other words, herpesvirus infection of affected brain areas might either be a causal component/cofactor for AD or a consequence of the tissue alterations in the affected AD areas. An important component that has to be considered is the frequently observed invasion of inflammatory immune cells in these areas. We suggest that the differential tropism of the viruses might offer an insight.

    Whereas HSV-1 and HSV-2 are considered to be “neurotropic,” in that they have a predilection to infect and replicate in neurons, the HHV-6A/B and HHV-7 as a group have generally been dubbed “lymphotropic” in that they principally target immune cells, including T cells and macrophages, and also are reported to infect glial cells e.g., oligodendrocytes. Work has been done on identifying receptors for HSV-1 and mapping them in human brain (e.g., Lathe and Haas, 2017), but much less is known about receptors and co-receptors for HHVs. None of the candidates identified so far—CD4, CD43, and CD134/TNFRSF4—alone permits infection by HHVs, noting that herpesviruses are unusual in that they require a cluster of host receptors for effective infection; a single receptor is insufficient. Therefore our conclusions are tentative. Indeed, the terms neuro- and lymphotropic are inaccurate. However, they serve to illustrate the potential for reciprocal interactions among HHV-6A/B, HHV-7, and HSV-1/2, in vivo, via several different pathways.

    AD is accompanied by a major CNS influx of proinflammatory immune cells, including macrophages (e.g., Fiala  et al., 2002), to boost pathogen elimination. Invading cells harboring episomal or integrated HHV genomes might have skewed the AD versus non-AD ratio of viral transcripts detected by Readhead et al. It would thus be very interesting to study the distribution of HHV genomes in neuronal versus non-neuronal cells.

    In addition, active herpesvirus infections can foster reactivation of other latent herpesviruses: for example, human cytomegalovirus infection is accompanied by reactivation of latent HSV-1 (Stowe et al., 2012). In this regard a 2016 paper by Chapenko et al. (Chapenko et al., 2016) is notable. Briefly, in this paper some 30–80 percent of all human brain samples studied (controls and “unspecified encephalopathy,” UCP, no precise diagnosis provided; mean age in both cases 58–59 years) were positive, as expected, for both HHV-6 and HHV-7 (HHV-6A and HHV-6B were not separately subtyped), and there was no significant difference in positivity between control and UCP. By contrast, in UCP frontal and temporal lobe there was a 100-fold, and highly significant, increase in HHV-6 genome content (Chapenko et al., 2016). Despite diagnostic caveats, it is clear that some event or events in these individuals switched on HHV replication—perhaps by genome reactivation induced by another triggering factor/agent, or by an influx of susceptible cells, or both.

    Conclusions: Rethinking Alzheimer’s Disease from Scratch
    Immense effort has been expended on targeting Aβ, without any success, and it now turns out that it is a defense molecule. The Readhead paper, combined with Tzeng et al. (Tzeng et al., 2018) and Eimer et al. (Eimer et al., 2018), is driving a paradigm change. Viruses are steadily moving to the fore as vital contributors to AD development, but which virus? Or all? Or are they all opportunist infections of a degenerating brain?

    Given that there is causal evidence for an involvement of HSV-1 in AD (Tzeng et al., 2018; Itzhaki and Lathe, 2018), and a recent paper from the team at Genentech reveals that a receptor gene for HSV-1 (PILRA, encoding a receptor that is probably not targeted by HHVs) is a significant risk factor for AD (Rathore et al., 2018), it seems that HSV-1 is likely to play a direct role. However, it could well be that HSV-1 brings in HHV by immune cell recruitment and/or reactivation, leading to “double pathology.” The Readhead paper is a convincing demonstration that HHVs are also likely to play a role.

    But what is it that first triggers AD? It does not seem to be infection per se, because these viruses are everywhere. Key “AD genes” encoding immune system modulators such as APOE are clearly important (and APOE alleles modulate susceptibility to diverse pathogens including herpes viruses, HIV, Chlamydia, and malaria), and lifestyle factors such as stress may play a role. Squinting ahead, it could be that a combination of infection, genes, age, and environment might explain a majority of AD cases.

    However, lest we spend too much time peering into the mist, we should focus on what we know. As it stands, the field has established that viruses are somewhere central in the causal chain. The intellectual property of many antivirals, such as aciclovir as the first-line drug against HSV-1/2 and another herpes virus, varicella-zoster virus (VZV), is no longer protected, and interest from pharmaceutical companies will probably be limited. But is that a good reason not to follow-up? To date, hundreds of drugs against AD have failed in clinical Phase 1-3 studies. Although these studies cost billions of dollars, not a single drug has been successful. We believe that the increasing evidence over the past few years—including the paper by Readhead and colleagues—that chronic infections and inflammatory processes are central to AD, clearly warrants revisiting aciclovir and other antiviral drugs (as well as vaccination) as potential routes to combating AD.

    Moreover, we should not limit ourselves to herpesviruses (both spirochetes and fungi have been associated with AD, indeed Readhead et al. detected traces of diverse pathogens such as human adenovirus, Ippy arenavirus, Torque teno virus, and Kaposi sarcoma-associated herpesvirus in AD brain). Nor should we overlook other disorders, Parkinson’s disease for example, where an infectious etiology has long been speculated.


    . Microbes and Alzheimer's Disease. J Alzheimers Dis. 2016;51(4):979-84. PubMed.

    . The Alzheimer's disease-associated amyloid beta-protein is an antimicrobial peptide. PLoS One. 2010 Mar 3;5(3):e9505. PubMed.

    . Amyloid-β peptide protects against microbial infection in mouse and worm models of Alzheimer's disease. Sci Transl Med. 2016 May 25;8(340):340ra72. PubMed.

    . 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.

    . Herpes Viruses and Senile Dementia: First Population Evidence for a Causal Link. J Alzheimers Dis. 2018;64(2):363-366. PubMed.

    . Herpes simplex virus type 1 in brain and risk of Alzheimer's disease. Lancet. 1997 Jan 25;349(9047):241-4. PubMed.

    . Herpes simplex virus type 1 DNA is located within Alzheimer's disease amyloid plaques. J Pathol. 2009 Jan;217(1):131-8. PubMed.

    . Evidence linking HHV-6 with multiple sclerosis: an update. Curr Opin Virol. 2014 Dec;9:127-33. Epub 2014 Oct 17 PubMed.

    . Historic evidence to support a causal relationship between spirochetal infections and Alzheimer's disease. Front Aging Neurosci. 2015;7:46. Epub 2015 Apr 16 PubMed.

    . Seroepidemiological study of human herpesvirus-6 and -7 in children of different ages and detection of these two viruses in throat swabs by polymerase chain reaction. J Med Virol. 1996 Jan;48(1):88-94. PubMed.

    . HHV-6-positivity in diseases with demyelination. J Clin Virol. 2014 Oct;61(2):216-9. Epub 2014 Jul 21 PubMed.

    . Distribution of cellular HSV-1 receptor expression in human brain. J Neurovirol. 2017 Jun;23(3):376-384. Epub 2016 Dec 15 PubMed.

    . Cyclooxygenase-2-positive macrophages infiltrate the Alzheimer's disease brain and damage the blood-brain barrier. Eur J Clin Invest. 2002 May;32(5):360-71. PubMed.

    . Reactivation of herpes simplex virus type 1 is associated with cytomegalovirus and age. J Med Virol. 2012 Nov;84(11):1797-802. PubMed.

    . Detection frequency of human herpesviruses-6A, -6B, and -7 genomic sequences in central nervous system DNA samples from post-mortem individuals with unspecified encephalopathy. J Neurovirol. 2016 Aug;22(4):488-97. Epub 2016 Jan 4 PubMed.

    . Paired Immunoglobulin-like Type 2 Receptor Alpha G78R variant alters ligand binding and confers protection to Alzheimer's disease. PLoS Genet. 2018 Nov;14(11):e1007427. Epub 2018 Nov 2 PubMed.

    View all comments by Jurgen Haas
  2. This work by Readhead and colleagues is an important contribution to the field. The authors have performed a really impressive, in-depth analysis of postmortem Alzheimer’s disease brain tissue samples using modern bioinformatics techniques, and found multiple lines of evidence linking viral infections, in particular human herpes virus 6A and 7, to Alzheimer’s disease.

    This is in line with the now-rapidly increasing evidence of viral infections as key drivers in the development of Alzheimer’s disease pathology. In recent years, it has been established that the Aβ peptide accumulating in Alzheimer’s disease amyloid plaques is a potent antimicrobial peptide and a key constituent of the brain’s innate immune defense against multiple pathogens. An emerging understanding of these findings could be that persistent viral infections within the brain could trigger this immune response over prolonged periods of time, resulting in pathological Aβ accumulation and eventually progressive cell death. The linking of known risk genes of Alzheimer’s disease to the immune defense and viral infections, as in the current work, provides further substance to this understanding of the drivers of the disease process.

    HHV6A is a ubiquitous pathogen, with nearly 100 percent of the population being infected; however, this primarily T-cell-associated virus seldom causes symptoms in immunocompetent adults. A limitation of studies of postmortem brain material is that the finding of pathogen associations might represent late superinfection of the diseased brain rather than true early inducers of the pathological processes.

    Until today, only herpes simplex has been confirmed to increase the risk of later Alzheimer’s disease development in large prospective population-based materials. With that said, I think the current and previous results indicate a role of HHV6A, at least in late Alzheimer’s disease, and I think future clinical antiviral drug trials to target Alzheimer’s disease development should seek to cover both HSV1 and HHV6A/HHV7.

    View all comments by Hugo Lövheim
  3. This is an interesting paper comparing postmortem brains of individuals with either AD neuropathology but no cognitive change (“preclinical AD”) or clinical AD, to brains of normal individuals. Subjects with clinical AD had higher levels of human herpesvirus-6A and -7. Viral abundance was associated with various modulators of amyloid precursor protein metabolism.

    However, no epidemiological data was presented on individuals who took part in these postmortem brain studies. As Readhead et al. note, we cannot know what the cognitive health trajectories of the “preclinical AD” patients would have been, and therefore disentangling molecules involved in disease progression from those responsible for maintenance of brain function is complex. Robust longitudinal population data sets are needed to assess the clinical relevance and generalizability of such findings, and how they relate to disease trajectory. In future, 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.

    View all comments by Charlotte Warren-Gash
  4. My concern about this work is that I find it difficult to square with the occurrence of Alzheimer’s disease in all Down's syndrome and in all carriers of some APP and PSEN mutations. The authors in their introduction bring up the idea of “slow virus” diseases. In fact the notion of slow virus diseases was fatally damaged by the identification of prion gene mutations causing these diseases in a hereditary fashion. This meant that, to explain these cases, one had to postulate a universal virus for which mutant prion genes were a universal receptor. In this case, we have to suppose that Down’s cases and the mutation cases either have a totally different disease mechanism or that for some reason, they are uniquely susceptible to these infections.

    View all comments by John Hardy
  5. This is an extremely exciting study that extends our understanding of the role of the human “virome” to neuropathology. The presence of viruses in human brains has long been recognized, but viral detection has been challenging and measurements of serum responses do not accurately reflect the viral load in the brain. In addition, highly neurotropic viruses like those found by the authors often display a stage of latency with periodic and not-well-understood cycles of reactivation and are thus difficult to quantify.

    This study tries to tackle these challenges by conducting a large-scale, comprehensive analysis of the association of viral constitutes in Alzheimer disease. The study confirms a previously suggested association of HSV-1 and AD but, unlike previous data, the authors now identify HHV-6A as the most notable virus associated with AD. This is intriguing as HHV-6A has already made a mark as a virus associated with demyelination diseases such as multiple sclerosis (MS). It is tempting to speculate that HHV-6A disrupts fundamentally the same processes in MS and AD, but their consequences are different and might depend on the interplay with the host genome.

    In addition, HHV-6A is well known for its ability to generate latent infections. What is the difference between the effects of latent and active infection of cells? We showed recently that expression of just a single latency gene product in oligodendrocyte precursor cells is enough to disrupt their migration, a function essential in the repair of myelin damage. Thus, latent infections may also contribute to adverse effects of these common viruses on the progression of AD.  

    Understanding the potential viral constitutes in the context of neurodegenerative diseases is vital for establishing more “human relevant” animals model. It should not be surprising that animal works often “do not work”—after all, we are ignoring a large amount of biologically active components that co-evolved with humans and that are not simply silent bystanders.

    The present study sets the stage for important future work to unlock the biological functions of the human virome.

    View all comments by Margot Mayer-Proschel
  6. When I first met Frank LaFerla more than 20 years ago at a Keystone meeting, he told me about a herpes simplex virus glycoprotein B fragment with homology to the Alzheimer's Aβ peptide. Frank and I spent some time working on the topic and published a paper in Biochemistry (Cribbs et al., 2000). We are now wondering whether the human herpesvirus 6A (HHV-6A) and human herpesvirus 7 (HHV-7) have homology to Aβ, as well?


    . Fibril formation and neurotoxicity by a herpes simplex virus glycoprotein B fragment with homology to the Alzheimer's A beta peptide. Biochemistry. 2000 May 23;39(20):5988-94. PubMed.

    View all comments by David Cribbs
  7. Shingles, induced by activated residual herpes virus, which had previously been silent, in terminal tissues, causes acute symptoms and pain within a few days. If the virus is activated in the CNS, the effect will also be seriously acute. This contradicts the chronic nature of familial and sporadic AD, whose development requires approximately 25 years since initial deposition of Aβ.

    In addition, I do not quite trust the pathway analysis. It depends on past publications, many of which have utilized IP-Western experiments and are irreproducible. I thus tend to agree with Charlotte Warren-Gash and John Hardy.

    View all comments by Takaomi Saido

Make a Comment

To make a comment you must login or register.

This paper appears in the following:


  1. Aberrant Networks in Alzheimer’s Tied to Herpes Viruses