Autophagy and Herpesvirus: A collaboration Contributing to Alzheimer's Disease

Since the 1960s viral pathogenesis researchers have considered herpesviruses as an underlying factor for Alzheimer's disease (AD). We reported molecular interactions between herpes simplex virus type 1 (HSV-1) and the amyloid precursor protein (APP), the parent of amyloid plaques pathognomonic for AD (Sapute-Krishnan et al., 2003 and 2006; Chen et al., 2011, Bearer and Wu, 2019). Furthermore, others report biochemical interactions between HSV-1 and autophagy. Using several brain banks for specimens of four brain regions in post-mortems of individuals with and without cognitive impairment prior to death. Readhead et al. 2018 found molecular-genetic evidence linking activity of 6 different human herpesviruses to AD, including HSV-1, HSV-2, HHN6, HHN7, VZV and CMV to AD. Of these, HHN6, a common virus causing a minor childhood illness thought to be a nuisance, emerged as most significant. Using a quantitative trait loci approach, a network of candidate AD-associated genes were found that correlated with viral load and activity. These ontology networks did not specifically consider autophagy genes. Our hypothesis is that viral replication and egress highjacks cellular membrane systems and thereby alters autophagic function. Those individuals carrying genetic variations that protect against this dynamic will be less vulnerable to cognitive impairment despite viral load, or viral load will be diminished. Here we first prepared lists of autophagy genes (ATG), including 180 we uniquely identified through machine learning, as well as lists from publications (Mitzushima, 2019) and websites (Autophagy Gene List, Tanpaku.org). We applied software developed by Readhead et al. 2018, available through Synapse.com, to expression and sequence data from post-mortem brains obtained from publications and public sites hosted by Alzheimer's Center brain banks. We first determined ATG expression levels correlating with either non-AD (<1 plaque per section, Braak<3, and no dementia, or pre-clinical AD, defined as Braak III-IV with no cognitive impairment. Virtually all ATG were down-regulated in pre-clinical compared to non-AD controls. Next we searched the list of quantitative trait loci (QTL) that correlated with increased viral load and activity for ATG genes from 300+ brains in the Nun's Study brain bank (ROSMAP) and the Mount Sinai Brain Bank (MSBB). Lastly, we correlated those ATG-associated QTL with expression levels of these genes in control and preclinical AD (Liang et al. 2007, 2008 and 2010). We found that decreased ATG expression due to single nucleotide polymorphisms correlate with viral load and AD. This study suggests autophagy is a novel mechanism linking herpesvirus to AD, which may aid in finding new diagnostic and therapeutic targets. Since HHV6 is a common infection of childhood, infecting nearly 100% of humans, identifying genetic vulnerabilities to persistence and progression will be critically important for prevention of adult AD.