CJD and Scrapie Require Agent-Associated Nucleic Acids for Infection
Sotirios Botsios and Laura Manuelidis*
Department of Surgery, Section of Neuropathology, Yale Medical School, New Haven 06510, Connecticut
Unlike Alzheimer's and most other neurodegenerative diseases, Transmissible Spongiform Encephalopathies (TSEs) are all caused by actively replicating infectious particles of viral size and density. Different strain-specific TSE agents cause CJD, kuru, scrapie and BSE, and all behave as latent viruses that evade adaptive immune responses and can persist for years in lymphoreticular tissues. A foreign viral structure with a nucleic acid genome best explains these TSE strains and their endemic and epidemic spread in susceptible species. Nevertheless, it is widely believed that host prion protein (PrP), without any genetic material, encodes all these strains. We developed rapid infectivity assays that allowed us to reproducibly isolate infectious particles where >85% of the starting titer separated from the majority of host components, including PrP. Remarkably, digestion of all forms of PrP did not reduce brain particle titers. To ask if TSE agents, as other viruses, require nucleic acids, we exposed high titer FU-CJD and 22L scrapie particles to potent nucleases. Both agent-strains were propagated in GT1 neuronal cells to avoid interference by complex degenerative brain changes that can impede nuclease digestions. After exposure to nucleases that are active in sarkosyl, infectivity of both agents was reproducibly reduced by 99%. No gold-stained host proteins or any form of PrP were visibly altered by these nucleases. In contrast, co-purifying protected mitochondrial DNA and circular SPHINX DNAs were destroyed. These findings demonstrate that TSE agents require protected genetic material to infect their hosts, and should reopen investigation of essential agent nucleic acids.
DISCUSSION We previously showed that isolated brain particles without detectable PrP maintain their high infectivity, a finding that excludes a prion protein agent [Kipkorir et al., 2015]. The above results demonstrate that particles from two different agent strains require protected nucleic acids for infection. PrP and its PrP-res component were not affected by nuclease digestion, yet after nucleases these proteins yielded no significant infectivity, a finding avenues for the discovery, prevention, and treatment of chronic diseases of unknown cause. In the realm of late-onset neurodegenerative diseases, it is possible that a subset of Alzheimer's disease is initially set in motion by unrecognized environmental pathogens of low virulence that are no longer actively replicating at later stages of neurodegeneration.
The above data demonstrates yet again that isolated TSE infectious particles, characterized by their homogeneous size, and density separation into 30% sucrose, contain many protected nucleic acids of viral size. Nucleases destroyed all visible nucleic acids along with infectivity. Improved particle isolation procedures that recovered 85–100% of the starting infectivity along with rapid and quantitative culture assays for infectivity, rather than PrP-res amyloid seeding, revealed unambiguous titer losses of >99% after straightforward nuclease digestion. These data provide compelling arguments for reinvigorating experiments to identify causal agent nucleic acids in TSE (PrP amyloid) diseases. In contrast to previous limited nuclease studies, the positive demonstrations of TSE agent destruction here is based, at least in part, on particle isolation procedures that minimized PrP and many extraneous host proteins, reduced background nucleic acids to very low levels (<0.01% of whole cells), used extended incubations with more powerful nucleases at 37°C, and included sarkosyl in buffer conditions that disaggregate PrP amyloid but do not affect infectivity.
PCR for mtDNA, an endogenous virus-like structure, also provided a superior quantitative assay for complete digestion of protein-protected nucleic acids. Nucleases in others’ previous studies failed to destroy obvious nucleic acids and hence no significant titer decrease was observed. Those nuclease experiments may also have been compromised by the typically small (0.5–5%) agent population investigated which selectively copurified with amyloid [Manuelidis, 2013]. PrP amyloid and associated plaque molecules in brain can trap the infectious agent with other molecules in an insoluble and poorly permeable aggregate [Manuelidis et al., 1997] that hinders effective nuclease digestions, and PrP-res accumulation is an innate host response that can help to clear infection [Miyazawa et al., 2012]. The tiny population of agent that remained viable in our nuclease studies could similarly reside in a resistant PrP aggregate, or in a keratinase resistant protein or biofilm [Miyazawa et al., 2011a; Botsios et al., 2015].
The requirement for agent nucleic acids was demonstrated with two very different TSE strains, and 22L scrapie agent particles were more susceptible to nuclease inactivation than FU-CJD. This difference parallels the greater sensitivity of 22L scrapie to chemical disruption, and emphasizes molecular or structural differences among agent strains that cannot be ascribed to host species or cell type. The invariant phenotype of individual TSE agent strains despite passage in different species and cell types suggests a DNA rather than a typical single stranded RNA viral genome. While we do not know if particle RNA, DNA, or both are required for infection, it is now possible to resolve these alternatives using advanced deep sequencing methods to probe nucleic acids in the more purified highly infectious ASB sol particles.
The concept that a TSE virus induces a pathological PrP-res amyloid change is not unlike recent observations made for other viral infections. TSE agents all replicate exponentially, and only later induce PrP-res amyloid in an arithmetic process that barely increases the total PrP [Manuelidis, 2013]. Other viruses also induce protein aggregates that are “prion-like” and antiviral [Hou et al., 2011]. Mammalian viruses can also initiate changes that lead to protein aggregation and pathology years after infection, as in postencephalitic Parkinson0 s disease. Subacute paramyxovirus infections also induce neurofibrillary tangles that are indistinguishable from those commonly found in standard non-infectious Alzheimer0 s disease (AD) brain, as well as in brains subjected to repeated trauma (dementia pugilistica). Many of these protein aggregates can transfer between cells in a non-infectious process. Hence, it is misleading to equate such pathological protein transfers with an active and real infectious process in mammals. Despite recent claims, AD amyloid is not infectious. It cannot be serially passaged or diluted, and it only accelerates old age neuropathologic changes when massive amounts are injected intracerebrally into aging primates [Ridley et al., 2006; Manuelidis, 2013]. While protein misfolding and amyloid seeding can occur in both artificial test tube and in vivo settings, and pathologic overlaps between TSEs and other neurodegenerative diseases such as AD have been highlighted for years, the term prion has become all-inclusive. It obscures differences among these diseases that have heterogeneous and largely unknown origins.
Interestingly, Herpes simplex virus can also initiate b-amyloid processes [Civitelli et al., 2015] and other environmental viruses may incite, contribute to and/or define different AD subsets. We identified two environmental viral metagenomic sequences hidden in TSE particle nucleic acids, and these were in the expected low amounts for TSE agent titers [Manuelidis, 2011]. Because these circular SPHINX DNAs (acronym for Slow Progressive Hidden INfections of variable X latency) can also be found in uninfected cells and brain, their DNA presence is insufficient to be the sole cause of TSE infection. It remains to be seen if their transcriptional activity, or antibody detected Rep A or other ORF proteins give them different attributes. SPHINX DNAs were not detectable in our reagents in numerous tests, indicating they were not lab contaminants, and the 1.76 element was identified in two human Multiple Sclerosis brains and sera in Germany [Whitley et al., 2014]. This further independently links them to progressive brain pathology. The circular structure of these elements is also consistent with a satellite type of virus, one that may need additional molecular components for infection. This circular DNA structure also led us to suggest they might have a role in neoplastic transformation [Manuelidis, 2011]. Interestingly, because of their presence in many cows, combined with the epidemiology of milk and meat consumption and breast cancer, zur Hausen  recently proposed they have a role in breast tumors. A more widespread source for these remnant phage DNAS of commensal Acinetobacter may also be considered because as previously shown, brains from laboratory hamsters and mice contained both SPHINX sequences even thought these rodents have no known exposure to cow's milk. The finding of these DNAs in TSE infectious particles surely should encourage a deeper investigation of viral nucleic acids in TSEs, and probably other neurodegenerative diseases. Because these small circular sequences were present in brain tissue without any accompanying Acinetobacter sequences, it suggests they can replicate symbiotically in mammalian cells. It remains to be seen if they have been incorporated during mammalian evolution or after birth. Regardless, they may represent the tip of the iceberg of other unsuspected bacterial viruses that can be incorporated by mammals and sometimes act as stealth pathogens.
We thank Joan Steitz for her encouragement and insightful suggestion on this manuscript. This work was supported by a grant from the Prusoff Foundation and contributions of CJD family members.
J. Cell. Biochem. 9999: 1–12, 2016. © 2016 Wiley Periodicals, Inc.
KEY WORDS: MICROBIOME; VIRAL STRAINS; GENOME; NEURODEGENERATION; SPHINX DNAs; PRION AMYLOID; ALZHEIMER’S DISEASE