Lipopolysaccharides (LPSs) as Potent Neurotoxic Glycolipids in Alzheimer’s Disease (AD)

Lipopolysaccharides (LPSs) are microbiome-derived glycolipids that are among the most potent pro-inflammatory neurotoxins known. In Homo sapiens, the major sources of LPSs are gastrointestinal (GI)-tract-resident facultative anaerobic Gram-negative bacilli, including Bacteroides fragilis and Escherichia coli. LPSs have been abundantly detected in aged human brain by multiple independent research investigators, and an increased abundance of LPSs around and within Alzheimer’s disease (AD)-affected neurons has been found. Microbiome-generated LPSs and other endotoxins cross GI-tract biophysiological barriers into the systemic circulation and across the blood–brain barrier into the brain, a pathological process that increases during aging and in vascular disorders, including ‘leaky gut syndrome’. Further evidence indicates that LPSs up-regulate pro-inflammatory transcription factor complex NF-kB (p50/p65) and subsequently a set of NF-kB-sensitive microRNAs, including miRNA-30b, miRNA-34a, miRNA-146a and miRNA-155. These up-regulated miRNAs in turn down-regulate a family of neurodegeneration-associated messenger RNA (mRNA) targets, including the mRNA encoding the neuron-specific neurofilament light (NF-L) chain protein. While NF-L has been reported to be up-regulated in peripheral biofluids in AD and other progressive and lethal pro-inflammatory neurodegenerative disorders, NF-L is significantly down-regulated within neocortical neurons, and this may account for neuronal atrophy, loss of axonal caliber and alterations in neuronal cell shape, modified synaptic architecture and network deficits in neuronal signaling capacity. This paper reviews and reveals the most current findings on the neurotoxic aspects of LPSs and how these pro-inflammatory glycolipids contribute to the biological mechanism of progressive, age-related and ultimately lethal neurodegenerative disorders. This recently discovered gut-microbiota-derived LPS–NF-kB–miRNA-30b–NF-L pathological signaling network: (i) underscores a direct positive pathological link between the LPSs of GI-tract microbes and the inflammatory neuropathology, disordered cytoskeleton, and disrupted synaptic-signaling of the AD brain and stressed human brain cells in primary culture; and (ii) is the first example of a microbiome-derived neurotoxic glycolipid having significant detrimental miRNA-mediated actions on the expression of NF-L, an abundant filamentous protein known to be important in the maintenance of neuronal and synaptic homeostasis.


Introduction
The microbiome of the human gastrointestinal (GI)-tract consists of a dynamic and highly interactive internal prokaryotic ecosystem that possesses a staggering complexity and diversity. Composed of about~10 15 microorganisms, this microbial bionetwork constitutes a significant fraction of the 'human metaorganism', with considerable commensal and/or symbiotic benefit to the human host. In fact, through extracellular fluid; cerebrospinal fluid; lymphatic and glymphatic dissemination; endocrine, systemic and neurovascular circulation; and vesicular trafficking and/or central and peripheral nervous system (CNS, PNS) passage, this microbiome and its secreted components strongly impact the health, well-being, immunity and vitality, as well as digestive, nutritive and neurological health of the human host [1][2][3][4][5][6][7]. As a complex microbial repository, (i) this compartment includes a broad spectrum of microbes, which represent mostly facultative anaerobes or anaerobic bacteria, followed in abundance by archaebacteria, fungi, protozoa, viruses and other microorganisms [1][2][3]; (ii) represents the largest source and highest density of microbial species found anywhere in nature [4,5]; and (iii) comprises about 1-3% of the human body mass (2 to 6 pounds of microbes in a 200-pound adult) collectively constituting the largest 'diffuse organ system' in the human body, and at least as metabolically active as the liver [1][2][3][4][5][6][7]. Relatively recent data from US National Institutes of Health (NIH)-funded interdisciplinary Human Microbiome Project (HMP) initially classified over~200 thousand diverse, non-redundant prokaryotic genomes in the human GI-tract microbiome comprising about~5 thousand different GI-tract microbial species, which together encode about 200 million different protein sequences. Hence, genomic studies of the human GI-tract microbiome indicate that this microbial repository currently represents about~1 thousand times the genetic complexity of all protein-coding sequences found in the entire human genome [1][2][3][4][5][6][7]. (https://www.hmpdacc.org/hmp/overview/; last accessed on 10 October 2022).

The Human Gastrointestinal (GI)-Tract Microbiome
The human GI-tract is essentially a~3.5 cm in diameter and~7 m long flexible tube that varies in pH and oxygen availability along its entire length. Approximately~99.5% of all of the resident microbes of the human GI-tract microbiome consist of facultative and/or obligate anaerobic bacteria from just 2 of the 52 major bacterial divisions: Bacillota (Firmicutes) and Bacteroides [1][2][3][4][5]. These form the essential 'bacterial-microbial core' of the human GI-tract microbiome. The bacterial phylum Bacillota (Firmicutes) mostly possess a Gram-positive cell wall structure, while Bacteroides are an abundant class of obligate anaerobic, Gram-negative pleomorphic-to-rod-shaped bacteria [5][6][7]. As it might be expected, the resident microorganisms of the intestinal epithelium in the deeper and more anaerobic regions of the small intestine are the most enriched in anaerobic and obligate anaerobic microbial species, and these deeper GI-tract regions are heavily populated by the phylum Bacteriodetes [1][2][3][4][5][6][7][8][9][10][11]. In these deeper regions of the human GI-tract microbiome, B. fragilis can be about~100-fold more abundant than the phylum Proteobacteria and the genus-species Escherichia coli [1][2][3][8][9][10][11][12][13][14]. In fact, Bacteroides attain extremely high densities in the most anaerobic regions of the human gut at about 10 10 to 10 11 cells per cubic millimeter of intestinal content and represent the major single source of Gram-negativebacterium-derived lipopolysaccharides (LPSs) anywhere in the human body ( [7,11,13,14]; see below). While GI-tract-derived microorganisms are generally beneficial to human health, when stressed or in pathological states, enterotoxigenic forms of these same microbes have considerable potential to secrete extremely toxic elements, including multiple types of Gram-negative-microbe-derived glycolipids, including LPSs, extremely potent inducers of pro-inflammatory and altered immune-signaling specifically associated with infection and disease.

The Nature of Lipopolysaccharides (LPSs)
bacteria are highly immunogenic, amphipathic glycoconjugates known as LPSs [8][9][10][11][12]. Normally, LPSs have a major natural function to provide structural integrity and a permeability barrier to protect Gram-negative bacteria from microbial toxins and bile salts and from biophysical, biochemical and environmental stress [7][8][9]. Gram-negative bacteria are typically encapsulated by a thin, rigid peptidoglycan meshwork consisting of glycosaminoglycan chains interlinked with short peptides that constitute the bacterial cell wall, and this barrier is surrounded by an outer membrane (OM) embedded with species-specific varieties of LPSs [8][9][10][11][12]. The microbial OM is not a phospholipid bilayer but is instead a highly asymmetric capsule containing phospholipids in the inner leaflet and LPS molecules in the outer leaflet [8][9][10][11][12]. Non-covalently bound and ultimately shed from the OM matrix into surrounding biofluids, large LPS species of approximately 1000 kDa per subunit consist of a hydrophobic lipid-A domain (also known as the LPS endotoxin) attached to a core oligosaccharide (core-OS) and a distal O-antigen (OA), also known as an O-polysaccharide [8][9][10][11][12] (https://pubchem.ncbi.nlm.nih.gov/compound/Lipopolysaccharide; last accessed on 20 September 2022). Interestingly, OA structures, which represent the outermost and most immunologically exposed component of bacterial LPSs, are extremely diverse, and approx-imately~200 different serogroups have been identified for Escherichia coli alone [3][4][5]11,12]. LPS core-OSs often contain highly immunogenic non-carbohydrate components including amino acids, phosphate groups and/or ethanolamine substituents that are highly variable in composition amongst all Gram-negative bacterial species and even within strains of the same species [5][6][7][9][10][11][12]. Together as a group, (i) LPSs are highly varied in their basic structure, organization and immunogenicity; (ii) they are classified as pathogen-associated molecular pattern (PAMP) molecules containing microbial-conserved molecular motifs recognized by toll-like receptors (TLRs) and other pattern recognition receptors (PRRs); (iii) they activate cells of the innate-immune system, such as macrophages and neutrophils, which synthesize pro-inflammatory factors that include cytokines, chemokines (chemotactic cytokines) and adipokines such as interleukin-1β (IL-1β), tumor necrosis factor (TNF), free radicals and reactive oxygen species (ROS); (iv) they act as prototypical endotoxins that promote the efflux of nitric oxides and eicosanoids; (v) because of their amphipathic character, they relatively easily pass through GI-tract biophysical barriers into the systemic circulation, especially when these barriers are damaged, diseased or 'leaky', such as in 'leaky gut syndrome' [13][14][15][16][17][18][19][20]; (vi) they invariably lead to significant innate-immune and pro-inflammatory responses in the infected host tissue; (vii) they strongly induce proinflammatory transcription factor signaling, which includes an up-regulation of NF-kB (p50/p65)-DNA binding and the transactivation of pro-inflammatory gene expression, especially in neuroglia and other related brain cell types; (viii) they up-regulate the abundance of NF-kB (p50/p65)-sensitive microRNAs, such as miRNA-30b, miRNA-34a, miRNA-146a and miRNA-155; (ix) which is followed by a significant down-regulation of the expression of neuron-specific messenger RNA (mRNA) targets, including those that encode synaptic (SYN), neurofilament (NF) and other structural and cytoarchitectural components of the neuron [9,10,17-23].

LPS-Mediated Effects on Neurofilament Light (NF-L) Chain Gene Expression in Alzheimer's Disease (AD)
The most recent evidence of a neuropathological pathway involving GI-tract-microbiomederived LPSs and/or BF-LPS, beginning with the LS-mediated up-regulation of proinflammatory transcription factor NF-kB (p50/p65) and miRNA-30b signaling, and ending with the specific down-regulation of the neuron-specific cytoskeletal NF-L element is noteworthy [29,39]. This pathway: (i) for the first time, underscores a positive pathological link between the LPSs of GI-tract microbes and the disordered cytoskeleton and disrupted synaptic signaling of the AD brain and stressed human neuronal-glial (HNG) cells in primary culture; (ii) in part characterizes the up-regulation of NF-kB (p50/p65), a brain-abundant transcription factor known to support inflammatory neuropathology; (iii) is representative of a local and progressive pathological signaling system along the gut-brain axis with potential to operate over the life-span of Homo sapiens; and (iv) it is the first example of a microbiome-derived neurotoxic glycolipid having significant detrimental microRNA-mediated actions on the expression of NF-L, an abundant neuronspecific filamentous protein known to be important for the maintenance of neuronal cell structure and shape, and synaptic homeostasis [27,39,53]. MicroRNA miRNA-30b was previously implicated in targeting inflammasome function [36,39]. Importantly, the sensitivity of a small pro-inflammatory microRNA family consisting of miRNA-30b, miRNA-34a, miRNA-146a, miRNA-155 and others to NF-kB (p50/p65) signaling appears to be in part be due to the presence of one-to-several NF-kB-DNA-binding sites in the immediate promoters of the genes encoding these miRNAs [35,[38][39][40]49,50,[53][54][55][56]. An improved understanding of LPS molecular genetic interactions along the GI tract-CNS axis involving GI-tract-derived microbial neurotoxins, AD and other related disorders should further support the hypothesis of altered LPS-miRNA-mRNA coupled signaling networks, and these concepts are currently supported by recently described experimental-findings in the scientific literature [29,30,[34][35][36][38][39][40][41][42][43][44][45][46][47][48][49][50][53][54][55][56][57]. For example, targeting and/or modulating GI-tract-microbiome-LPS-mediated, miRNA-30b-regulated NF-L pathways and other miRNA-mediated gene expression circuitry could be valuable in the design of future therapeutic strategies that: (i) could support and maintain cytoarchitectural components essential for neuronal shape, axonal caliber, inter-neuronal signaling and synaptic plasticity; and (ii) may more effectively manage the many neurological diseases in which NF-L gene expression and abundance play a defining and/or determinant role [29,39,[55][56][57]. Recent emerging evidence on dietary-based modifications of microbial dysbiosis may also be an attractive means to alter the abundance, speciation and complexity of enterotoxigenic forms of AD-relevant microbes such as B. fragilis and their potential for their long-term pathological discharge of highly neurotoxic microbial-derived glycolipid/lipoglycan secretions such as LPS [51,52,[58][59][60] (Figure 2). and/or modulating GI-tract-microbiome-LPS-mediated, miRNA-30b-regulated NF-L pathways and other miRNA-mediated gene expression circuitry could be valuable in the design of future therapeutic strategies that: (i) could support and maintain cytoarchitectural components essential for neuronal shape, axonal caliber, inter-neuronal signaling and synaptic plasticity; and (ii) may more effectively manage the many neurological diseases in which NF-L gene expression and abundance play a defining and/or determinant role [29,39,[55][56][57]. Recent emerging evidence on dietary-based modifications of microbial dysbiosis may also be an attractive means to alter the abundance, speciation and complexity of enterotoxigenic forms of AD-relevant microbes such as B. fragilis and their potential for their long-term pathological discharge of highly neurotoxic microbial-derived glycolipid/lipoglycan secretions such as LPS [51,52,[58][59][60] (Figure 2).

Figure 2. Highly schematized flow chart of a pathological circuit in AD brain flowing along the gut-brain axis (see manuscript text).
The interactions among GI-tract-microbiome-derived LPSs, pro-inflammatory cytokines and ROS in the generation of the pro-inflammatory transcription factor NF-kB (p50/p65) complex are not completely understood; other miRNAs and mRNAs are expected to be involved in this pathological cascade; importantly, the GI-tract microbiome provides a longterm and life-long source of LPSs; specific components of this pathological pathway provide multiple therapeutic targets for the more effective clinical management of AD.

Summary
The GI-tract microbiome of Homo sapiens is a rich and dynamic source of microbes that possess a staggering degree of diversity and complexity [3][4][5][6][7]. Early studies suggested that the identification and characterization of unique microbial signatures in neuropsychiatric diseases could provide new possibilities in targeted anti-or pro-biotic treatments [31,32]. However, HMP and parallel GI-tract microbiome studies involving multiple human populations and animal models of AD provided evidence indicating that due to wide variation in the microbial and genetic composition of individual GI-tract microbiomes even during healthy aging, it would be difficult to associate the abundance, speciation and/or complexity of any single microbial genus, species or classification with any single human disease [1][2][3][4][5][6][7][8][9][10]45,47,55,59,[61][62][63][64][65][66][67][68][69]. This is especially relevant to highly complex and heterogeneous neurological syndromes, such as AD, prion disease and other agerelated neurodegenerative disorders, against their complex background of familial genetics, patient age, gender, drug history, age of onset and duration, inter-current disease and other critical environmental and lifestyle risk factors [61][62][63]70].
While commensal GI-tract microbes are beneficial and essential to human health, the enterotoxigenic forms of these same microorganisms have considerable potential to secrete highly neurotoxic biopolymers, including multiple varieties of Gram-negative bacterial-derived glycolipids such as LPSs, which are extremely potent inducers of pro-inflammatory and altered innate-immune and immunological signaling in infection, aging and diseases from AD to cancer [36,. One major characterized pathogenic role of Figure 2. Highly schematized flow chart of a pathological circuit in AD brain flowing along the gut-brain axis (see manuscript text). The interactions among GI-tract-microbiome-derived LPSs, pro-inflammatory cytokines and ROS in the generation of the pro-inflammatory transcription factor NF-kB (p50/p65) complex are not completely understood; other miRNAs and mRNAs are expected to be involved in this pathological cascade; importantly, the GI-tract microbiome provides a long-term and life-long source of LPSs; specific components of this pathological pathway provide multiple therapeutic targets for the more effective clinical management of AD.

Summary
The GI-tract microbiome of Homo sapiens is a rich and dynamic source of microbes that possess a staggering degree of diversity and complexity [3][4][5][6][7]. Early studies suggested that the identification and characterization of unique microbial signatures in neuropsychiatric diseases could provide new possibilities in targeted anti-or pro-biotic treatments [31,32]. However, HMP and parallel GI-tract microbiome studies involving multiple human populations and animal models of AD provided evidence indicating that due to wide variation in the microbial and genetic composition of individual GI-tract microbiomes even during healthy aging, it would be difficult to associate the abundance, speciation and/or complexity of any single microbial genus, species or classification with any single human disease [1][2][3][4][5][6][7][8][9][10]45,47,55,59,[61][62][63][64][65][66][67][68][69]. This is especially relevant to highly complex and heterogeneous neurological syndromes, such as AD, prion disease and other age-related neurodegenerative disorders, against their complex background of familial genetics, patient age, gender, drug history, age of onset and duration, inter-current disease and other critical environmental and lifestyle risk factors [61][62][63]70].
While commensal GI-tract microbes are beneficial and essential to human health, the enterotoxigenic forms of these same microorganisms have considerable potential to secrete highly neurotoxic biopolymers, including multiple varieties of Gram-negative bacterial-derived glycolipids such as LPSs, which are extremely potent inducers of proinflammatory and altered innate-immune and immunological signaling in infection, aging and diseases from AD to cancer [36,. One major characterized pathogenic role of LPSs appears to be the stimulation of cytokine-, chemokine-and/or ROS-mediated patho-logical signaling programs that drive the induction of pro-inflammatory transcription factor NF-kB (p50/p65), which subsequently promotes the transcriptional up-regulation of NF-kB-sensitive microRNAs. These, in turn, target AD-or related-disease-associated mRNAs, which ultimately down-regulate critically pathologically relevant gene expression programs, resulting in the initiation, development and/or propagation of human disease [21,29,34,49,50,56,[63][64][65][66][67][68][69][72][73][74][75]. For example, the LPS-mediated, ROS-and NF-kB-regulated up-regulation of microRNAs miRNA-30b, miRNA-146a and miRNA-155 in transgenic murine models of AD and in AD are now known to target and down-regulate the expression of important neuron-specific neurofilament and synaptic elements important in supporting and maintaining homeostasis in brain cells and neural signaling capabilities [51,52,[57][58][59][60]71] (Figure 2). An improved understanding of molecular-genetic signaling along the GI-tract-microbiome-CNS/PNS axis in healthy aging and in AD has significant potential for the development of new diagnostic, prognostic and therapeutic strategies and response-to-treatment efficacy monitoring for the more effective clinical management of AD and other types of progressive, age-related and ultimately lethal neurodegenerative disorders in which LPSs and other microbiome-derived neurotoxins appear to be involved.

Institutional Review Board Statement:
The scientific material in this review paper was obtained, compiled and synthesized from the references listed at the end of this paper; ethical review and approval were waived in this study, as this review papers is simply a compilation of scientific facts. Data Availability Statement: All data used in this review are openly available and freely accessible on MedLine (www.ncbi.nlm.nih.gov), where they are listed by the last names of the individual authors.

Acknowledgments:
The research in this article is due to be presented in part at Society for Neuroscience (SFN) Annual Meeting, San Diego, CA, USA, in November 2022. Sincere thanks are extended to L. Cong, C. Eicken, K. Navel, W. Poon, T. Bond, E. Head, Y. Zhao and the late P.N. Alexandrov, J.M. Hill and H.M. Wisniewski for helpful discussions in this research area and to D. Guillot for expert technical assistance, medical artwork and general organization. Research on microRNAs, ethnobiology, botanical neurotoxins, and pro-inflammatory and pathogenic signaling in the Lukiw laboratory involving the microbiome, the innate immune response, amyloidogenesis, synaptogenesis and neuro-inflammation in AD, prion disease and other human neurological and plant-viroid-based diseases was supported through an unrestricted grant to LSU Eye Center from Research to Prevent Blindness (RPB); The Brown Foundation; Joe and Dorothy Dorsett Innovation in Science Healthy Aging Award; Louisiana Biotechnology Research Network (LBRN); and NIH grants NEI EY006311, NIA AG18031 and NIA AG038834 (W.J.L.). The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute on Aging, the National Center for Research Resources or the National Institutes of Health.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.
Ethics Statement: The culture of human brain tissue cells, acquisition and handling procedures were carried out in strict accordance with the ethics review board policies at both the donor institutions and at LSU-HSC. The work in this study was approved by the Institutional Biosafety Committee/Institutional Review Board (IBC/IRB) of LSU Health Sciences Center, New Orleans, LA, 70112, USA.