In 2018, researchers created a stir with a multi-network analysis of postmortem tissue that suggested viruses—particularly human herpes—might disrupt molecular pathways in the brains of Alzheimer’s disease patients. Now, researchers led by Steven Jacobson, a virologist at the National Institutes of Health in Bethesda, Maryland, have taken a direct approach using quantitative RNA and DNA analysis to search for the little bugs. They found no more viruses in AD than control brain. While this study does not support a role for viruses in AD, it does not rule it out either, the authors say. Future work should try to figure out how these viruses might act early on in disease, without themselves being detectable anymore at life’s end.
- Postmortem brains screened for viral RNA/DNA.
- No differences found between AD and controls.
- Viral role in AD not ruled out.
The idea that pathogens might play a role in AD has been around, and often dismissed, for decades. That changed when researchers led by Joel Dudley at the Icahn School of Medicine at Mount Sinai, New York, published their paper in Neuron in 2018.
First author Benjamin Readhead and colleagues had examined a huge dataset, combining sophisticated bioinformatic approaches. They reported that viral load, especially of HHV-6A, correlated with changes in expression of human genes linked to AD, including genes in the amyloid pathway (Jul 2018 news). Scientists in the field applauded the approach, but also struggled to comprehend the complexity of the data and the analysis. A recent letter to Neuron even challenged the findings. Hyun-Hwan Jeong and Zhandong Liu at Baylor College of Medicine, Houston, Texas, reanalyzed the data and concluded that the statistical analysis was too weak to prove a link between viral load and AD (Jeong and Liu, 2019). Readhead and colleagues countered that Jeong and Liu misinterpreted some of the data, but acknowledged that different methods of analysis in comparative transcriptomics can yield dramatically different results (Readhead et al., 2019).
Still, the paper energized leaders in the AD fields—and indeed in virology. It prompted an invitation for Alzheimer-minded neurologists and neuroscientists to attend the 11th International Conference on HHV-6 and HHV-7 in Toronto last July (Jul conference news). There, the virologists brought their expertise to bear, raising fundamental questions about infection and AD, such as, what fraction of AD patients in Readhead et al. tested positive for HHV?
Low Load in LOAD. In MSBB and ROSMAP AD brain samples, viral load in late-onset AD is no higher than in controls. [Courtesy of Steven Jacobson and Neuron.]
Jacobson is a virologist who for many years has studied the role of herpes viruses in multiple sclerosis, epilepsy, glioma, and encephalitis (e.g., Leibovitch and Jacobson, 2018; Lin et al., 2016; Yao et al., 2009). He addressed this basic question using some of the same tissue samples Readhead had used, from the Mount Sinai Brain Bank (MSBB), New York, and from the Religious Orders Study/Memory and Aging Project at Rush University (ROSMAP), Chicago. The virologists also analyzed samples from the Baltimore Longitudinal Study of Aging, Johns Hopkins University, for a total of more than 1,200 AD and control brains.
First author Mary Alice Allnutt sifted for viruses among tissues taken from up to four different areas of each brain. She used PathSeq, an RNA-Seq-based tool developed by the Broad Institute, Cambridge, Massachusetts, to screen tissue for the presence of more than 25,000 microbes, including 118 human viruses. She backed up that data with digital droplet PCR, a highly sensitive third-generation quantitative amplification method, to find traces of viral DNA.
In short, neither method linked virus load to AD. PathSeq found HHV6 in four of 301 brains from MSBB, and two of 600 brains from ROSMAP. In MSBB, three of the four were from people diagnosed with definite, one from a person diagnosed with probable AD. The two positives in ROSMAP were from one control and one AD case.
PathSeq reflects the total of all transcripts found for an individual microbe. In this study, the scores suggested very low viral load in the positive samples, with the highest scoring 35. By contrast, postmortem analysis of people who had been infected with dengue or John Cunningham virus (JCV) scored 5,000 or higher for those viruses. There was no correlation between HHV PathSeq scores and AD severity.
PathSeq identified other viruses, including JCV, Epstein-Barr, and cytomegalovirus, but again scores were low and none significantly correlated with AD (see image above).
Digital droplet PCR told a similar story. Testing DNA from 364 ROSMAP samples and 344 from the Johns Hopkins dataset, Allnutt found no HHV6A in any of the AD samples and in only one control from ROSMAP. HHV6B turned up in a total of 20 out of 510 AD samples and five of 166 controls. Those frequencies were not statistically different and the viral loads were low.
What does this data mean? And how to reconcile it with Readhead et al.? Jacobson said his new data does not rule out a role for viruses in AD. “Our data just doesn’t tell me viruses are the cause of AD,” he told Alzforum. The techniques used in this paper would miss a decades-old infection that might have set in motion a slow pathogenic process toward the disease. Jacobson added that those associations are extremely hard to prove for ubiquitous viruses, such as HHVs. Liu said Jacobson’s work complements his reanalysis of the Readhead et al. data. “Taking both analyses together, we can conclude that HHVs do not contribute significantly to the etiology of AD,” he wrote to Alzforum. “The only remaining factor is whether the ROSMAP and MSSB datasets represent the diverse AD population,” he added.
For now, researchers may be left looking for the smoke rather than the gun. On that note, Readhead questioned the relevance of the dengue and JCV infection comparisons. “While these examples are clinically striking, they are characterized by productive, fulminant viral infection … and thus importantly different from the modes of activity that species such as HHV-6A/B demonstrate in a non-encephalitic brain, where the virus is likely to be in a latent or quiescent state and present at very low abundance,” Readhead told Alzforum.
Lavinia Alberi Auber, Swiss Integrative Center for Human Health, Fribourg, emphasized the stark differences between the two approaches. “This is salient and raises the need for methodological harmonization and specific guidelines for how pathogen-host interactions can be investigated in the setting of the more immuno-privileged brain,” she wrote (comment below). Alberi Auber said bioinformatics and wet-lab approaches should be run in parallel. Readhead agreed. “We expect that mere detection of a particular virus in a single dimension, e.g., RNA or DNA sequences, will convey only limited understanding of how viruses might contribute to the complex pathobiology of Alzheimer’s disease,” he wrote (comment below).
Readhead also noted that the prevalence of HHV6 in this new study was lower than has been reported in the literature. “Without a clearer understanding of how to reconcile the unusually low frequencies of detected HHV-6A/B with a diverse scientific literature that supports a much higher prevalence, it is difficult to properly evaluate potential associations with Alzheimer’s disease,” he wrote. Jacobson considers historical data inappropriate for comparison because there is no consensus in the literature, and different groups use different samples, extraction methods, assays, primers, etc. “Levels of sensitivity vary, and many times data are reported as positive or negative without absolute viral load,” he wrote.
Can scientists settle once and for all if viruses play a role in AD? Jacobson believes the only sure way is to run a prevention trial using antivirals. That would take decades and seems unlikely to happen. Meanwhile, bioinformatics will become more sophisticated. “We expect that an expansion of viral detection assays to include searches for viral proteins and non-coding RNAs will contribute valuable additional context to these investigations,” wrote Readhead.—Tom Fagan
- Aberrant Networks in Alzheimer’s Tied to Herpes Viruses
- Going Viral: Alzheimer’s Research at Herpes Conference
- Jeong HH, Liu Z. Are HHV-6A and HHV-7 Really More Abundant in Alzheimer's Disease?. Neuron. 2019 Dec 18;104(6):1034-1035. PubMed.
- Readhead B, Haure-Mirande JV, Ehrlich ME, Gandy S, Dudley JT. Clarifying the Potential Role of Microbes in Alzheimer's Disease. Neuron. 2019 Dec 18;104(6):1036-1037. PubMed.
- Leibovitch EC, Jacobson S. Viruses in chronic progressive neurologic disease. Mult Scler. 2018 Jan;24(1):48-52. PubMed.
- Lin CT, Leibovitch EC, Almira-Suarez MI, Jacobson S. Human herpesvirus multiplex ddPCR detection in brain tissue from low- and high-grade astrocytoma cases and controls. Infect Agent Cancer. 2016;11:32. Epub 2016 Jul 26 PubMed.
- Yao K, Honarmand S, Espinosa A, Akhyani N, Glaser C, Jacobson S. Detection of human herpesvirus-6 in cerebrospinal fluid of patients with encephalitis. Ann Neurol. 2009 Mar;65(3):257-67. PubMed.
- Chorlton SD. Reanalysis of Alzheimer's brain sequencing data reveals absence of purported HHV6A and HHV7. J Bioinform Comput Biol. 2020 Feb;18(1):2050012. PubMed.
- Allnutt MA, Johnson K, Bennett DA, Connor SM, Troncoso JC, Pletnikova O, Albert MS, Resnick SM, Scholz SW, De Jager PL, Jacobson S. Human Herpesvirus 6 Detection in Alzheimer's Disease Cases and Controls across Multiple Cohorts. Neuron. 2020 Mar 18;105(6):1027-1035.e2. Epub 2020 Jan 23 PubMed.