Methods for determining if a subject has neurodegeneration or is suspected of neurodegeneration

Abstract

This disclosure provides a method of sample processing, the method comprising (a) obtaining a sample from a subject having or suspected of having neurodegeneration, (b) determining an amount of one or more taxa in the sample, and (c) comparing the amount of the one or more taxa to a control amount. Also provided are methods of treating neurodegeneration.

Description

Atty. Dkt. No.650053.01137 METHODS FOR DETERMINING IF A SUBJECT HAS NEURODEGENERATION OR IS SUSPECTED OF NEURODEGENERATION CROSS REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Provisional Application No. 63 / 561,628, filed on March 5, 2024, the contents of which are incorporated by reference in their entireties. STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with government support under grant number AG075501 awarded by the National Institutes of Health. The government has certain rights in this invention. FIELD

[0003] This invention relates to the detection and treatment of neurodegeneration in a subject. In particular, this invention relates to the detection of gut and oral microbiome indicators associated with neurodegeneration. BACKGROUND

[0004] Neurodegenerative disorders, such as Alzheimer’s disease, are debilitating conditions involving nervous system damage that place tremendous burden on patients, caregivers, and healthcare systems worldwide. Despite this, our understanding of the etiologies for many of these diseases is limited, as are available treatment options. Accordingly, a need exists for new and more effective methods of detecting and treating neurodegenerative disorders. SUMMARY

[0005] In an aspect, provided herein is a method of sample processing, the method comprising (a) obtaining a sample from a subject having or suspected of having neurodegeneration, (b) determining an amount of one or more taxa in the sample, and (c) comparing the amount of the one or more taxa to a control amount. In embodiments, the one or more taxa are selected from the group consisting of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis.

[0006] In some embodiments, determining the amount of the one or more taxa comprises isolating bacterial nucleic acids and sequencing the isolated bacterial nucleic acids. The amount of the one or more taxa may be indicative of neurodegeneration when the amount of Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control amount and / or the amount Turicibacter sanguinis is decreased in the sample, relative to the control amount. In some embodiments, theAtty. Dkt. No.650053.01137 amount of the one or more taxa is indicative of neurodegeneration when a ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount, is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1.

[0007] In a second aspect, provide herein is a method of treatment comprising (a) determining an amount of one or more taxa selected from the group consisting of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis in a sample obtained from a subject having or suspected of having neurodegeneration, (b) comparing the amount of the one or more taxa in the sample to a control amount, and (c) administering a treatment for neurodegeneration. In embodiments, the treatment comprises administering an agent to modulate the one or more taxa in the subject. The treatment may be administered when the amount of Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control amount and / or the amount Turicibacter sanguinis is decreased in the sample, relative to the control amount. In some embodiments, the treatment is administered when a ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount, is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] FIG. 1 is a Linear Discriminate Analysis Effect Size (LEfSe) plot for stool samples collected from subjects with mild cognitive impairment (MCI; green; N=14) or controls (red; N=10). Taxa highlighted in blue are bile-acid tolerant or capable of modifying bile acids. Data shown are statistically significant and represent enrichment of taxa within each cohort.

[0009] FIG. 2 is an Anvi’o plot of Turicibacter sangunins metagenome identified in stool samples collected from subjects with MCI (green) or CU controls (red). Individual patient samples are represented as rings spanning the T. sanguinis genome. Assembled contigs are represented by each radius. Color intensity is proportional to relative abundance for each contig within a given sample. N=14 for MCI; N=10 for CU. Overall, the relative abundance per person for T. sanguinis is 4.72x greater in controls relative to MCI.

[0010] FIG. 3 is an Anvi’o plot of Bilophila wadsworthia metagenome identified in stool samples collected from subjects with MCI (green) or CU controls (red). Individual patient samplesAtty. Dkt. No.650053.01137 are represented as rings spanning the B. wadsworthia genome. Assembled contigs are represented by each radius. Color intensity is proportional to relative abundance for each contig within a given sample. N=14 for MCI; N=10 for CU. Overall, the relative abundance per person for B. wadsworthia is 3.18x greater in MCI relative to controls.

[0011] FIGS. 4A-4C depict the impact of human cytomegalovirus (HCMV) infection on neurons from Alzheimer’s disease (AD) patients and iPSC-derived control neurons. (A) HCMV infection in both control and AD neurons increases insoluble pTau. (B) Immunofluorescence shows an increase in pTau expression in HCMV infected iPSC-derived neurons. (C) By 20dpi (days post-infection), HCMV infection leads to a near complete decreases in action potential generation (both spontaneous and evoked) across all tested lines. Analyses are restricted to MEA wells that demonstrate activity above threshold (wMFR > 0.15) (n = 5-15 wells per condition). Statistics in FIGS.4A and C were completed using two-way ANOVA. †, p<0.1; *, p < 0.05; **, p < 0.01.

[0012] FIGS. 5A-5C illustrate phage purification from stool samples. (A) Phage banding stained with EtBr to enable extraction. The 1.55 and 1.35 g / ml fractions are extracted for cleanup and sequencing. (B) Fluorescence microscopy of phage samples following purification and visualized with SYBR gold to verify a lack of contaminating microbes. The typical yield from one gram of stool is ~10e8particles / ml. (C) An electron micrograph of a prepared stool sample, depicting a phage likely in the syphoviridae family of dsDNA phage.

[0013] FIG. 6 illustrates the use of machine learning to identify dissimilatory features using Random Forest with Boruta algorithm. Metagenomic (MAGs) features (e.g., bacterial taxa, viral taxa, etc.) are associated with outcomes (e.g., MCI, Aβ+, sex, etc.). Shadow features are generated by rearranging the original features across observations to generate classifiers, destroying the relationship with the outcome. Random Forest competes features with randomized versions of themselves (i.e., training) subsets before computing a “threshold” to compete against original features. Features with larger importance relative to threshold or shadow variables are assigned a Gini score. Top features are used for further analyses.

[0014] FIG. 7 Cerebral vascular reactivity (CVR), cerebral blood flow (CBF), arterial transit time (ATT) t-test results of group comparisons with age, biological sex and grey matter density as covariates. Voxelwise group t-test results showing control- aMCI group for cluster-size corrected p < 0.01 (^^ < 0.05). CBF and CVR were higher and ATT was lower for the control group compared to aMCI group.

[0015] FIGS. 8A-8D. Discriminatory taxa from the 16S Random Forest and MaAsLin2 analyses and correlations with neurovascular and cognitivevascular data. (A) Discriminatory taxaAtty. Dkt. No.650053.01137 identified via MaAsLin2 using a p-value cutoff of 0.05. (B) Discriminatory taxa identified via Random Forest using a Gini score cutoff of 0.09. Discriminatory taxa identified by Random Forest and MaAsLin2 underwent Spearman Rho correlational analysis to identify taxa that were significantly correlated with neurovascular metrics, including cerebrovascular reactivity (CVR), cerebral blood flow (CBF), and arterial transit time (ATT). Correlations were calculated separately for all 24 participants (All), 14 aMCI participants (MCI), and 10 controls (Con). Discriminatory taxa also underwent Spearman Rho correlational analysis for cognitive analyses, including delayed recall (DR), category fluency (CF), and Trail Making Test B (TMTB). Correlations for cognitive tests were calculated for the total sample (All). (C) Shows Spearman correlations for cerebrovascular analyses, and (D) shows Spearman correlations for cognitive tests. Note. *p<0.05, **p<0.01, ***p< 0.001

[0016] FIGS.9A-9D. Discriminatory bacteria from metagenomics sequencing Random Forest and MaAsLin2 analyses and correlations with neurovascular and cognitive data. (A) Discriminatory taxa identified via MaAsLin2 using a p-value cutoff of 0.05. (B) Discriminatory taxa identified via Random Forest using a Gini score cutoff of 0.09. Discriminatory taxa underwent Spearman Rho correlational analysis to identify taxa that were significantly correlated with neurovascular metrics, including cerebrovascular reactivity (CVR), cerebral blood flow (CBF), and arterial transit time (ATT). Correlations were calculated separately for all 24 participants (All), 14 aMCI participants (MCI), and 10 controls (Con). Discriminatory taxa also underwent Spearman Rho correlational analysis for cognitive analyses, including delayed recall

[0017] (DR), category fluency (CF), and Trail Making Test B (TMTB). Correlations for cognitive tests were calculated for the total sample (All). (C) Shows Spearman correlations for cerebrovascular analyses, and (D) shows Spearman correlations for cognitive tests. Note. *p<0.05, **p<0.01, ***p< 0.001

[0018] FIGS. 10A-10C. Turicibacter is depleted and Bilophila wadsworthia is enriched in aMCI. (A) Ratio of Turicibacter to B. wadsworthia for individual participants. Red boxes indicate participants whose concentration of Bilophila is greater than that of Turicibacter. (B) B. wadsworthia is significantly enriched in aMCI compared to controls (p<0.01), and the ratio of Bilophila to Turicibacter is significantly higher in aMCI (p<0.001) than in controls (p>0.05). Turicibacter is enriched in controls, though not significantly so (p=0.093 via the 2-way ANOVA shown above and p=0.057 via Mann-Whitney U test comparing only control vs aMCI). Red circles and squares represent a subset of samples that underwent LefSe analysis due to similar abundance levels (p>0.9999). (C) Anvio plot showing the nucleotide variability in the B. wadsworthia MAG. Color intensity represents number of single nucleotide variants (SNVs) per kb pair.Atty. Dkt. No.650053.01137

[0019] FIGS.11A-11D. aMCI phageomes are distinct from controls. (A) Viral bins identified with control, aMCI, or both after use of a min3 filter. The vast majority of the viral contigs identified were bacteriophages (B) Examples of phage bins identified in (A). The bin on the right is only found in aMCI (n=7 / 14), and the bin on the left is only found in the controls (n=4 / 10). (C) Differential metabolic pathways encoded by phage contigs in aMCI vs control. Statistically significant comparisons are denoted by the following: p<0.05: *; p<0.01: **. (D) Assessment of phage lifestyle using the viral to bacterial ratio (VBR). Values to the right indicate an increase in lytic activity while values to the left indicate lysogeny. Controls (red) are more likely to have a more lytic VBR, whereas aMCI participants (green) have a more lysogenic gut virome (p<0.05).

[0020] FIGS. 12A-12D. Discriminatory viruses from metagenomics sequencing Random Forest and MaAsLin2 analyses and correlations with neurovascular and cognitive data. (A) Discriminatory viral contigs identified via MaAsLin2 using a p-value cutoff of 0.05. (B) Discriminatory viral contigs identified via Random Forest using a Gini score cutoff of 0.09. Discriminatory viral contigs underwent Spearman Rho correlational analysis to identify contigs that were significantly correlated with neurovascular metrics, including cerebrovascular reactivity (CVR), cerebral blood flow (CBF), and arterial transit time (ATT). Correlations were calculated separately for all 24 participants (All), 14 aMCI participants (MCI), and 10 controls (Con). Discriminatory viral contigs also underwent Spearman Rho correlational analysis for cognitive tests, including delayed recall (DR), category fluency (CF), and Trail Making Test B (TMTB). Correlations for cognitive tests were calculated for the total sample (All). (C) Shows Spearman correlations for cerebrovascular analyses, and (D) shows Spearman correlations for cognitive tests. Note. *p<0.05, **p<0.01, ***p< 0.001.

[0021] FIG. 13 is a graph illustrating increased taurochenodeoxycholic acid (TCDCA) concentrations in subjects with mild cognitive impairment (MCI), as compared to cognitively unimpaired controls. DETAILED DESCRIPTION

[0022] Neurodegenerative disorders remain a global health concern. For example, Alzheimer’s disease (AD), the most prevalent neurodegenerative disorder worldwide, is a devastating disorder characterized by the loss of basal forebrain and hippocampal neurons leading to cognitive decline, memory impairment, and general behavioral changes. Current reports estimate that there are 55 million affected individuals worldwide with 6.2 million cases in the United States1. With the high prevalence and progressive nature of the disease, this is a significant clinical problem as these numbers are predicted to double by 20501. The economic impact is an added burden for families, caregivers, and the health care system overall with healthcare costs totaling $355 billion in 20211.Atty. Dkt. No.650053.01137 Despite the wide prevalence of AD, the etiology remains unknown. Genetic risk factors and familial forms of AD have been identified, but ~95% of cases are sporadic in nature, suggesting that environmental factors likely play a role and has spawned investigations into putative AD contributions from viral pathogens and gut microbes. Accordingly, a need exists for new and more effective methods of detecting and treating neurodegenerative disorders.

[0023] The Examples demonstrate differences in the microbiome of subjects with neurodegeneration, as compared to healthy controls. In Example 1, stool samples were analyzed from subjects with mild cognitive impairment (MCI) and cognitively unimpaired (CU) control subjects. As compared to CU controls, the samples from MCI subjects showed increased abundance of the following bacterial taxa: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales. In addition, the samples from MCI subjects showed decreased abundance of the following bacterial taxa: Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis. The data also demonstrated a strong correlation for depletion of Turicibacter sanguinis and enrichment of Bilophila wadsworthia, with a Bilophila: Turicibacter of 15:1 in MCI relative to CU control samples. These results demonstrate a clear association between microbiome composition and neurodegeneration, and suggest that the abundance of certain bacterial taxa may be a marker of neurodegeneration and / or a pathological condition predisposed to neurodegeneration.

[0024] Thus, this disclosure presents methods of processing a sample from a subject having or suspected of having neurodegeneration and methods of treatment for neurodegeneration.

[0025] Methods of Sample Processing

[0026] In an aspect, the present disclosure provides a method of sample processing, the method comprising obtaining a sample from a subject having or suspected of having neurodegeneration. The term “biological sample” or “sample” as used herein includes, but is not limited to, a sample containing tissues, cells, and / or biological fluids isolated from a subject. Examples of biological samples include, but are not limited to, tissues, cells, biopsies, blood, plasma, lymph, serum, plasma, urine, saliva, mucus, stool, cerebrospinal fluid (CSF), and tears. A biological sample may be obtained directly from a subject (e.g., by blood, stool, or CSF sampling) or from a third party (e.g., received from an intermediary, such as a healthcare provider or lab technician). In some embodiments, the biological sample may be selected from the group consisting of a tissue sample, a cell sample, a biopsy sample, a blood sample, a CSF sample, a lymph sample, a serum sample,Atty. Dkt. No.650053.01137 a plasma sample, a urine sample, a saliva sample, a mucus sample, and a tear sample. In some embodiments, the biological sample is a blood sample, a stool sample, a saliva sample, or a CSF sample.

[0027] As used herein, a “subject” may be interchangeable with a “patient” or “individual” and means an animal, which may be a human or non-human animal, that may be in need of treatment. In embodiments, the subject is a human. In the methods, the sample is obtained from a subject having or suspected of having neurodegeneration. “Neurodegeneration,” which may be used interchangeably with “neurodegenerative diseases” and “neurodegenerative disorders,” refers broadly to any condition associated with cognitive impairment. Examples of neurodegeneration include, without limitation, mild cognitive impairment (MCI), Alzheimer’s disease (AD), Parkinson’s disease (PD), Lewy Body’s disease (LBD), frontotemporal dementia (FTD), and prion diseases.

[0028] In some embodiments, the subject has or is suspected of having mild cognitive impairment. As used herein, “mild cognitive impairment” or “MCI” refers to a condition in which a subject has cognitive decline relative to non-cognitively impaired people of the same or similar age. For instance, a subject exhibiting mild cognitive impairment may exhibit more problems with memory or thinking as compared to people of the same or similar age that are not cognitively impaired. Example memory or thinking problems can include, but are not limited to, having difficulty coming up with words, frequently losing things, and / or frequently forgetting about attending important events.

[0029] In some embodiments, the subject has or is suspected of having Alzheimer’s disease (AD). Subjects with AD may exhibit symptoms related to cognitive decline that are more severely exhibited than those associated with mild cognitive impairment and may also include more severe changes in thinking, memory, reasoning, and / or behavioral changes. In some embodiments, the subject has or is suspected of having AD, and the method further comprises determining if the subject is amyloid-β positive (Aβ+) or Amyloid-β negative (Aβ-).

[0030] In an aspect, the method further comprises determining an amount of one or more taxa in the sample from the subject. As used herein, “taxon” or “taxa” refers to a grouping of organisms within a taxonomic hierarchy based on common characteristics. In embodiments, the one or more taxa are bacterial taxa selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales,Atty. Dkt. No.650053.01137 Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis.

[0031] The amount of the one or more taxa in the sample may be determined using any suitable method known in the art, including, without limitation, quantitative PCR, droplet digital PCR, fluorescence in situ hybridization, flow cytometry, enzyme-linked immunosorbent assay, or nucleic acid sequencing. The terms “nucleic acid” and “oligonucleotide,” as used herein, refer to polydeoxyribonucleotides (containing 2-deoxy-D-ribose), polyribonucleotides (containing D- ribose), and to any other type of polynucleotide that is an N glycoside of a purine or pyrimidine base. There is no intended distinction in length between the terms “nucleic acid”, “oligonucleotide” and “polynucleotide”, and these terms will be used interchangeably. These terms refer only to the primary structure of the molecule. Thus, these terms include double- and single-stranded DNA, as well as double- and single-stranded RNA. For use in the present methods, an oligonucleotide also can comprise nucleotide analogs in which the base, sugar, or phosphate backbone is modified as well as non-purine or non-pyrimidine nucleotide analogs.

[0032] In some embodiments, determining the amount of the one or more taxa in the sample comprises isolating bacterial nucleic acids and sequencing the isolated bacterial nucleic acids. As used herein, “isolating” or “extracting” refer interchangeably to a process in which a nucleic acid is released from a cell, and separated from cell membranes, proteins, and other cellular components using physical and / or chemical methods. Nucleic acids can be extracted using various methods that are well known in the art, including, without limitation, those that rely on organic extraction, ethanol precipitation, silica-binding chemistry, cellulose-binding chemistry, and ion exchange chemistry. Many reagents and kits for nucleic acid isolation are commercially available.

[0033] Prior to sequencing, the isolated bacterial nucleic acids may be amplified by performing an amplification reaction. The term “amplification reaction” refers to any chemical reaction, including an enzymatic reaction, which results in increased copies of a template nucleic acid sequence or results in transcription of a template nucleic acid. Amplification reactions include reverse transcription, the polymerase chain reaction (PCR), including Real Time PCR (see U.S. Pat. Nos. 4,683,195 and 4,683,202; PCR Protocols: A Guide to Methods and Applications (Innis et al., eds, 1990)), and the ligase chain reaction (LCR) (see Barany et al., U.S. Pat. No.5,494,810). Exemplary “amplification reactions conditions” or “amplification conditions” typically comprise either two or three step cycles. Two-step cycles have a high temperature denaturation step followed by a hybridization / elongation (or ligation) step. Three step cycles comprise a denaturation step followed by a hybridization step followed by a separate elongation step.Atty. Dkt. No.650053.01137

[0034] In some embodiments, the bacterial nucleic acids are amplified using PCR. PCR is an in vitro method used to selectively amplify a specific DNA target sequence in a sample. PCR employs two main reagents: primers and a DNA polymerase. The term “primer,” as used herein, refers to an oligonucleotide capable of acting as a point of initiation of DNA synthesis under suitable conditions. Such conditions include those in which synthesis of a primer extension product complementary to a nucleic acid strand is induced in the presence of four different nucleoside triphosphates and an agent for extension (for example, a DNA polymerase or reverse transcriptase) in an appropriate buffer and at a suitable temperature. In PCR, a repeated series of reaction steps (i.e., template denaturation, primer annealing, and extension of the annealed primers by DNA polymerase) results in exponential amplification of the target sequence. See Saiki et al., 1985, Science 230:1350 for a detailed description of PCR.

[0035] A primer is preferably a single-stranded DNA. The appropriate length of a primer depends on the intended use of the primer but typically ranges from about 6 to about 225 nucleotides, including intermediate ranges, such as from 15 to 35 nucleotides, from 18 to 75 nucleotides and from 25 to 150 nucleotides. Short primer molecules generally require cooler temperatures to form sufficiently stable hybrid complexes with the template. A primer need not reflect the exact sequence of the template nucleic acid, but must be sufficiently complementary to hybridize with the template. The design of suitable primers for the amplification of a given target sequence is well known in the art and described in the literature cited herein

[0036] In some embodiments, the method comprises contacting the isolated bacterial nucleic acids with one or more sets of primers, each set specific to a target sequence associated with one taxon of the one or more taxa. As used herein, a primer is “specific,” for a target sequence if, when used in an amplification reaction under sufficiently stringent conditions, the primer hybridizes primarily to the target nucleic acid. The terms “target sequence”, “target region”, and “target nucleic acid,” as used herein, are synonymous and refer to a region or sequence of a nucleic acid which is to be amplified, sequenced, or detected. Typically, a primer is specific for a target sequence if the primer-target duplex stability is greater than the stability of a duplex formed between the primer and any other sequence found in the sample. One of skill in the art will recognize that various factors, such as salt conditions as well as base composition of the primer and the location of the mismatches, will affect the specificity of the primer, and that routine experimental confirmation of the primer specificity will be needed in many cases. Hybridization conditions can be chosen under which the primer can form stable duplexes only with a target sequence. Thus, the use of target-specific primers under suitably stringent amplification conditionsAtty. Dkt. No.650053.01137 enables the selective amplification of those target sequences that contain the target primer binding sites.

[0037] The term “hybridization,” as used herein, refers to the formation of a duplex structure by two single-stranded nucleic acids due to complementary base pairing. Hybridization can occur between fully complementary nucleic acid strands or between “substantially complementary” nucleic acid strands that contain minor regions of mismatch. Conditions under which hybridization of fully complementary nucleic acid strands is strongly preferred are referred to as “stringent hybridization conditions” or “sequence-specific hybridization conditions”. Stable duplexes of substantially complementary sequences can be achieved under less stringent hybridization conditions; the degree of mismatch tolerated can be controlled by suitable adjustment of the hybridization conditions. Those skilled in the art of nucleic acid technology can determine duplex stability empirically considering a number of variables including, for example, the length and base pair composition of the oligonucleotides, ionic strength, and incidence of mismatched base pairs, following the guidance provided by the art (see, e.g., Sambrook et al., 1989, Molecular Cloning-A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York; Wetmur, 1991, Critical Review in Biochem. and Mol. Biol. 26(¾):227-259; and Owczarzy et al., 2008, Biochemistry, 47: 5336-5353, which are incorporated herein by reference).

[0038] Primers can incorporate additional features which allow for the detection or immobilization of the primer but do not alter the basic property of the primer, that of acting as a point of initiation of DNA synthesis. For example, primers may contain an additional nucleic acid sequence at the 5′ end which does not hybridize to the target nucleic acid, but which facilitates cloning or detection of the amplified product, or which enables transcription of RNA (for example, by inclusion of a promoter) or translation of protein (for example, by inclusion of a 5′-UTR, such as an Internal Ribosome Entry Site (IRES) or a 3′-UTR element, such as a poly(A)n sequence, where n is in the range from about 20 to about 200). The region of the primer that is sufficiently complementary to the template to hybridize is referred to herein as the hybridizing region.

[0039] As used herein, a “polymerase” refers to an enzyme that catalyzes the polymerization of nucleotides. “DNA polymerase” catalyzes the polymerization of deoxyribonucleotides. Known DNA polymerases include, for example, Pyrococcus furiosus (Pfu) DNA polymerase, E. coli DNA polymerase I, T7 DNA polymerase and Thermus aquaticus (Taq) DNA polymerase, among others. “RNA polymerase” catalyzes the polymerization of ribonucleotides. The foregoing examples of DNA polymerases are also known as DNA-dependent DNA polymerases. RNA-dependent DNA polymerases also fall within the scope of DNA polymerases. Reverse transcriptase, which includesAtty. Dkt. No.650053.01137 viral polymerases encoded by retroviruses, is an example of an RNA-dependent DNA polymerase. Known examples of RNA polymerase (“RNAP”) include, for example, T3 RNA polymerase, T7 RNA polymerase, SP6 RNA polymerase and E. coli RNA polymerase, among others. The foregoing examples of RNA polymerases are also known as DNA-dependent RNA polymerase. The polymerase activity of any of the above enzymes can be determined by means well known in the art.

[0040] Nucleic acid sequencing is the process of determining the order of nucleotides in a nucleic acid molecule. Any suitable sequencing method may be used with the present invention. Suitable methods include, for example, RNA sequencing, 16S rRNA sequencing, 16S rDNA sequencing, single molecule real time (SMRT) sequencing, whole-genome shotgun sequencing (WGS), Illumina sequencing, Nanopore DNA sequencing, massively parallel signature sequencing (MPSS), Polony sequencing, 454 pyrosequencing, combinatorial probe anchor synthesis (cPAS), Ion Torrent semiconductor sequencing, DNA nanoball sequencing, and SOLiD sequencing. In some embodiments, the nucleic acid sequencing comprises 16S rDNA sequencing and / or whole genome shotgun sequencing.

[0041] For methods that utilize a high-throughput sequencing method, such as 16S rDNA sequencing, the amplicons must be converted into a sequencing library for sequencing. A “sequencing library” is a pool of DNA fragments that include adapters. Thus, in these embodiments, the methods may further comprise (a) fragmenting the amplicons, and / or (b) adding adapters to the amplicons. Methods of generating sequencing libraries are well known in the art. Adapters must be included in or added to the amplicons to allow them to interact with a high- throughput sequencing platform. In some embodiments, adapters are included in the primers such that the adapters are added to the amplicons during the DNA amplification step. In other embodiments, adapters are ligated to the amplicons following the DNA amplification step using a ligase enzyme.

[0042] DNA sequencing produces sequencing reads, i.e., sequences of the DNA fragments present in the sequencing library as determined by the sequencer. To analyze the sequencing reads, they are first cleaned up (e.g., trimmed, filtered for quality, de-noised to limit the impact of sequencing errors, de-replicated to reduce file size). Without further analysis, the resulting sequences lack genomic context. Thus, to determine the source of a read (i.e., organism from which the sequenced DNA fragment was derived), it must be mapped to a reference database. Methods for mapping sequencing reads to reference databases are available in the form of free tools, including mothur, QIIME, and various R packages.Atty. Dkt. No.650053.01137

[0043] 16S rDNA sequencing allows for the identification of genus and / or species level taxonomic assignments as amplicon sequence variants (ASVs). Further, 16S amplicons are generated quantitatively allowing for characterization of abundance of ASVs within samples. Thus, in some embodiments, determining the amount of the one or more taxa comprises quantitatively generating 16S amplicons and quantifying the quantitatively generated 16S amplicons.

[0044] In some embodiments, determining the amount of the one or more taxa comprises detecting and / or quantifying one or more taxonomic marker. As used herein, a “taxonomic marker” refers to a molecule that is produced by and / or associated with a taxon. In the methods, detecting an increase (or decrease) in a taxonomic marker may be indicative of an increase (or decrease) in the taxon with which the taxonomic marker is associated. In some embodiments, the one or more taxonomic marker comprises a protein and / or a metabolite. The protein may be an enzyme. As demonstrated in the Examples, the inventors discovered differential abundance of certain metabolic functions between subjects with neurodegeneration and control subjects. Thus, in some embodiments, the protein / enzyme is associated with methanogenesis, the PTS system (cellobiose-specific II component), the gamma-hexachorocyclohexane transport system, highly selective metalloendopeptidase, the Mce transport system, siroheme biosynthesis, propanoyl-CoA metabolism, and / or adenine ribonucleotide biosynthesis. As demonstrated in the Examples, the inventors discovered an increased abundance of bile salt-tolerant taxa. Thus, the taxonomic marker may be a molecule associated with bile acid modifications. In some embodiments, the taxonomic marker comprises a bile salt hydrolase. In some embodiments, the taxonomic marker comprises a modified bile acid metabolite. Exemplary modified bile acid metabolites included, without limitation, chenodeoxycholic acid (CDCA), deoxycholic acid (DCA), glycocholic acid (GCA), taurocholic acid (TCA), taurochenodeoxycholic acid (TCDCA), glycochenodeoxycholic acid (GCDCA), and tauroursodeoxycholic acid (TUDCA). In some embodiments, the taxonomic marker comprises butyrate, H2S, and / or taurine. The taxonomic marker may be detected using any suitable method known in the art. For example, and without limiting, when the taxonomic marker is a protein / enzyme, suitable methods for detecting and / or quantifying include functional assays (e.g., fluorometric assays, chromogenic assays), mass spectrometry, ELISA, and assays to measure protein / enzyme mRNA levels (e.g., qPCR). When the taxonomic marker is a metabolite, suitable methods for detecting and / or quantifying include spectrophotometric assays, fluorometric assays, chromatography methods (e.g., HPLC, LC-MS / MS, gas chromatography), and quantitative NMR.Atty. Dkt. No.650053.01137

[0045] In an aspect, the method comprises comparing the amount of the one or more taxa in the sample to a control amount. In some embodiments, the control amount can be the amount in a control sample from another subject that is known to not exhibit neurodegeneration and / or cognitive impairment. In the same or alternative embodiments, the control amount can be a quantified value of a biomarker, e.g., 16S rDNA for a specific taxon and / or a metabolite for a specific taxon, in one or more subjects known to not exhibit neurodegeneration and / or cognitive impairment. For instance, the control amount can be an average amount of 16S rDNA for a specific taxon in a cohort of subjects known to not exhibit neurodegeneration and / or cognitive impairment. In some embodiments, the control amount is determined using a control sample.

[0046] In some embodiments, the amount of the one or one or more taxa in the sample is indicative of neurodegeneration when the amount of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcalesis increased in the sample, relative to the control amount. In some embodiments, the amount of the one or one or more taxa in the sample is indicative of neurodegeneration when the amount Bilophila wadsworthia, relative to the control amount, is about 1.1 times greater, about 1.5 times greater, about 2 times greater, about 2.5 times greater, about 3 times greater, about 3.5 times greater, about 4 times greater, about 4.5 times greater, about 5 times greater, about 10 times greater, or about 20 times greater, or is within a range bounded by any of the foregoing. In some embodiments, the amount of Bilophila wadsworthia in the sample is about 2 times greater, or about 3 times greater, relative to the control amount.

[0047] In some embodiments, the amount of the one or one or more taxa in the sample is indicative of neurodegeneration when the amount of Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and / or Turicibacter sanguinis is decreased in the sample, relative to the control amount. In some embodiments, the amount of the one or one or more taxa in the sample is indicative of neurodegeneration when the amount of Turicibacter sanguinis, relative to the control amount, is about 1.1 times less, about 1.5 times less, about 2 times less, about 3 times less, about 3.5 times less, about 4 times less, about 4.5 times less, about 5 times less, about 10 times less, or about 20 times less, or is within a range bounded by any of the foregoing. In some embodiments, the amount of Turicibacter sanguinis in the sample is about 3 times less, or about for times less, relative to the control amount.

[0048] The method may further comprise determining a ratio of the amount of Bilophila wadsworthia to the amount of Turicibacter sanguinis in the sample, relative to the control amountAtty. Dkt. No.650053.01137 of each taxa. As an example, this ratio may may be calculated by (1) dividing the amount of Bilophila wadsworthia in the sample by the control amount of Bilophila wadsworthia to calculate a relative Bilophila wasdsworthia value, (2) dividing the amount of Turicibacter sanguinis in the sample by the control amount of Turicibacter sanguinis to calculate a relative Turicibacter sanguinis value, and (3) dividing the relative Bilophila wasdsworthia value by the relative Turicibacter sanguinis value. In some embodiments, the amount of the one or one or more taxa in the sample is indicative of neurodegeneration when the ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount Bilophila wadsworthia, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount of Turicibacter sanguinis, is 2:1, 5:1, 10:1, 15:1, 20:1, 25:1, 30:1, or 50:1, or is within a range bounded by any of the foregoing. In particular embodiments, the ratio is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1.

[0049] In the Examples, a differential abundances of certain phages between subjects with neurodegeneration and control subjects is shown. Thus, in some embodiments, the method may comprise determining an amount of one or more phage associated with the one or more taxa. In some embodiments, the phage is a lysogenic phage.

[0050] Methods of Treatment

[0051] In a second aspect, the present disclosure provides a method of treating neurodegeneration, the method comprising determining an amount of one or more taxa in a sample obtained from a subject having or suspected of having neurodegeneration, wherein the one or more taxa are selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis. The amount of the one or more taxa in the sample may suitably be determined using any method described herein.

[0052] The method further comprises (i) comparing the amount of the one or more taxa to a control amount and (ii) administering a treatment for neurodegeneration based on the comparing in (i). In some embodiments, the treatment is administered when the amount of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, and / or Micrococcales is increased in the sample, relative to the control amount.Atty. Dkt. No.650053.01137 In some embodiments, the treatment is administered when the amount Bilophila wadsworthia in the sample, relative to the control amount, is about 1.1 times greater, about 1.5 times greater, about 2 times greater, about 2.5 times greater, about 3 times greater, about 3.5 times greater, about 4 times greater, about 4.5 times greater, about 5 times greater, about 10 times greater, or about 20 times greater, or is within a range bounded by any of the foregoing. In some embodiments, the treatment is administered when the amount of Bilophila wadsworthia in the sample is about 2 times greater, or about 3 times greater, relative to the control amount.

[0053] In some embodiments, the treatment is administered when the amount of Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and / or Turicibacter sanguinis is decreased in the sample, relative to the control amount. In some embodiments, the treatment is administered when the amount of Turicibacter sanguinis in the sample, relative to the control amount, is about 1.1 times less, about 1.5 times less, about 2 times less, about 3 times less, about 3.5 times less, about 4 times less, about 4.5 times less, about 5 times less, about 10 times less, or about 20 times less, or is within a range bounded by any of the foregoing. In some embodiments, the treatment is administered when the amount of Turicibacter sanguinis in the sample is about 3 times less, or about for times less, relative to the control amount.

[0054] In some embodiments, the treatment is administered when the ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount Bilophila wadsworthia, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount of Turicibacter sanguinis, is 2:1, 5:1, 10:1, 15:1, 20:1, 25:1, 30:1, or 50:1, or is within a range bounded by any of the foregoing. In particular embodiments, the treatment is administered when the ratio is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1.

[0055] As used herein, “treatment” or “treating” neurodegeneration may include disease onset, slowing disease progression, and / or reducing the frequency / severity of one or more symptoms of neurodegeneration. Treatment may be assessed using any suitable method for assessing neurodegeneration, including, without limitation, assessments of verbal memory (Rey Auditory Verbal Learning Test, RAVLT), assessments psychomotor processing speed and mental flexibility (Trail Making Test, Parts A and B), assessments language (letter and category fluency), and the Quick Dementia Rating System (QDRS). Any suitable treatment for neurodegeneration may be administered. In some embodiments, the treatment comprises administering a cholinesterase inhibitor, an NMDA receptor antagonist, and / or an anti-amyloid agent. In some embodiments, the treatment comprises administering an agent to modulate the amount of the one or more taxa in the subject. For example, in some embodiments, the treatment comprises administering an agent to increase the amount of Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia,Atty. Dkt. No.650053.01137 Ruminococcus, and / or Turicibacter sanguinis in the subject. In some embodiments, the treatment comprises administering an agent to decrease the amount of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales in the subject. Suitable agents for modulating abundance of bacterial taxa are known in the art and include, without limitation, probiotic agents, prebiotic agents, and antibiotic agents. In some embodiments, the treatment comprises administering an agent to increase serotonin concentrations in the gut. The compositions exemplified above may be used individually (singly), in combination, sequentially, and / or any combination thereof.

[0056] By way of example, and without limiting, in some embodiments, the treatment may comprise obtaining a sample from a patient presenting with symptoms of neurodegeneration and analyzing the microbiome in the sample to determine an imbalance in one or more taxa consistent with the neurodegenerative symptoms (e.g., a decrease, relative to control, in Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and / or Turicibacter sanguinis, and / or an increase in of Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales). To alleviate and / or slow the progression of the neurodegenerative symptoms, an agent or combination of agents may be administered to increase (e.g., probiotics, prebiotics) one or more taxa that show decreased abundance in the sample and / or to decrease (e.g., antibiotics) one or more taxa that show increased abundance in the sample. The treatment may comprise sequential administration of an agent (e.g., an antibiotic) to produce an overall decrease in bacterial abundance, followed by administration of an agent (e.g., a prebiotic, a probiotic), to selectively increase one or more taxa that show decreased abundance in the sample. A probiotic to be administered may comprise a plurality of organisms (e.g., a plurality of taxa that showed decreased abundance in the sample) or a single organism that acts as a keystone organism to modulate the abundance of other taxa in the microbiome. A prebiotic to be administered may comprise a bacterial metabolite (e.g., a metabolite made by Turicibacter) to condition the microbiota, thereby modulating the abundance of other taxa in the microbiome.

[0057] The present disclosure is not limited to the specific details of construction, arrangement of components, or method steps set forth herein. The compositions and methods disclosed herein are capable of being made, practiced, used, carried out and / or formed in various ways that will beAtty. Dkt. No.650053.01137 apparent to one of skill in the art in light of the disclosure that follows. The phraseology and terminology used herein is for the purpose of description only and should not be regarded as limiting to the scope of the claims. Ordinal indicators, such as first, second, and third, as used in the description and the claims to refer to various structures or method steps, are not meant to be construed to indicate any specific structures or steps, or any particular order or configuration to such structures or steps. All methods described herein can be performed in any suitable order unless otherwise indicated herein or otherwise clearly contradicted by context. The use of any and all examples, or exemplary language (e.g., "such as") provided herein, is intended merely to facilitate the disclosure and does not imply any limitation on the scope of the disclosure unless otherwise claimed. No language in the specification, and no structures shown in the drawings, should be construed as indicating that any non-claimed element is essential to the practice of the disclosed subject matter. The use herein of the terms “including,” “comprising,” or “having,” and variations thereof, is meant to encompass the elements listed thereafter and equivalents thereof, as well as additional elements. Embodiments recited as “including,” “comprising,” or “having” certain elements are also contemplated as “consisting essentially of” and “consisting of” those certain elements.

[0058] As used in this specification and the claims, the singular forms “a,” “an,” and “the” include plural forms unless the context clearly dictates otherwise. For example, the term “a therapeutic” or “an antibody” should be interpreted to mean “one or more therapeutics” and “one or more antibodies,” respectively, unless the context clearly dictates otherwise. As used herein, the term “plurality” means “two or more.”

[0059] Furthermore, in those instances where a convention analogous to “at least one of A, B and C, etc.” is used, in general such a construction is intended in the sense of one having ordinary skill in the art would understand the convention (e.g., “a system having at least one of A, B and C” would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and / or A, B, and C together.). It will be further understood by those within the art that virtually any disjunctive word and / or phrase presenting two or more alternative terms, whether in the description or figures, should be understood to contemplate the possibilities of including one of the terms, either of the terms, or both terms. For example, the phrase “A or B” will be understood to include the possibilities of “A” or "B" or “A and B.”

[0060] All language such as “up to,” “at least,” “greater than,” “less than,” and the like, include the number recited and refer to ranges which can subsequently be broken down into ranges and subranges. Recitation of ranges of values herein are merely intended to serve as a shorthandAtty. Dkt. No.650053.01137 method of referring individually to each separate value falling within the range, unless otherwise indicated herein, and each separate value is incorporated into the specification as if it were individually recited herein. For example, if a concentration range is stated as 1% to 50%, it is intended that values such as 2% to 40%, 10% to 30%, or 1% to 3%, etc., are expressly enumerated in this specification. These are only examples of what is specifically intended, and all possible combinations of numerical values between and including the lowest value and the highest value enumerated are to be considered to be expressly stated in this disclosure. Use of the word “about” to describe a particular recited amount or range of amounts is meant to indicate that values very near to the recited amount are included in that amount, such as values that could or naturally would be accounted for due to manufacturing tolerances, instrument and human error in forming measurements, and the like. All percentages referring to amounts are by weight unless indicated otherwise.

[0061] No admission is made that any reference, including any non-patent or patent document cited in this specification, constitutes prior art. In particular, it will be understood that, unless otherwise stated, reference to any document herein does not constitute an admission that any of these documents forms part of the common general knowledge in the art in the United States or in any other country. Any discussion of the references states what their authors assert, and the applicant reserves the right to challenge the accuracy and pertinence of any of the documents cited herein. All references cited herein are fully incorporated by reference, unless explicitly indicated otherwise. The present disclosure shall control in the event there are any disparities between any definitions and / or description found in the cited references.

[0062] The following examples are meant only to be illustrative and are not meant as limitations on the scope of the invention or of the appended claims. EXEMPLARY EMBODIMENTS

[0063] Embodiment 1. A method of sample processing, comprising: (a) obtaining a sample from a subject having or suspected of having neurodegeneration; (b) determining an amount of one or more taxa in the sample from the subject, wherein the one or more taxa are selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella,Atty. Dkt. No.650053.01137 Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis; and (c) comparing the amount determined in (b) to a control amount.

[0064] Embodiment 2. The method of embodiment 1, wherein the determining comprises isolating bacterial nucleic acids from the sample and sequencing the isolated bacterial nucleic acids.

[0065] Embodiment 3. The method of embodiment 2, wherein the sequencing comprises 16S rDNA bacterial gene sequencing and / or wherein the sequencing comprises: (i) contacting the isolated bacterial nucleic acids with one or more sets of primers, each set specific to a target sequence associated with one taxon of the one or more taxa; (ii) amplifying the isolated bacterial nucleic acids to generate amplicons; and (ii) sequencing the amplicons to generate sequencing reads.

[0066] Embodiment 4. The method of embodiment 2 or 3, wherein the determining further comprises quantitatively generating 16S amplicons.

[0067] Embodiment 5. The method of any one of embodiments 2-4, the sequencing comprises whole genome shotgun sequencing of bacterial DNA.

[0068] Embodiment 6. The method of any one of embodiments 1-5, wherein the determining comprises detecting one or more taxonomic marker.

[0069] Embodiment 7. The method of embodiment 6, wherein the one or more taxonomic marker comprises an enzyme and / or a metabolite.

[0070] Embodiment 8. The method of any one of embodiments 1-7, wherein the determining comprises determining an amount of one or more taxa selected from the group consisting of: Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, NLAE zl H528, and Turicibacter sanguinis.

[0071] Embodiment 9. The method of embodiment 8, wherein the amount determined in (b) is indicative of neurodegeneration when the amount of Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control amount.

[0072] Embodiment 10. The method of embodiment 9, wherein the amount of Bilophila wadsworthia in the sample is about 2 times greater, or about 3 times greater, relative to the control amount.

[0073] Embodiment 11. The method of embodiment 8, wherein the amount determined in (b) is indicative of neurodegeneration when the amount of Turicibacter sanguinis is decreased in the sample, relative to the control amount.Atty. Dkt. No.650053.01137

[0074] Embodiment 12. The method of embodiment 11, wherein the amount of Turicibacter sanguinis in the sample is about 3 times less, or about four times less, relative to the control amount.

[0075] Embodiment 13. The method of embodiment 8, wherein a ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount of Bilophila wadsworthia, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount of Turicibacter sanguinis, is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1.

[0076] Embodiment 14. The method of any one of embodiments 1-13, wherein the sample is a stool sample, a saliva sample, a blood sample, and / or a cerebrospinal fluid (CSF) sample.

[0077] Embodiment 15. The method of embodiment 14, wherein the sample is a stool sample and / or a saliva sample.

[0078] Embodiment 16. The method of any one of embodiments 1-15, further comprising determining that the subject has mild cognitive impairment based on the comparing in (c).

[0079] Embodiment 17. The method of any one of embodiments 1-15, further comprising determining that the subject has Alzheimer’s disease (AD) based on the comparing in (c).

[0080] Embodiment 18. The method of embodiment 17, further comprising determining if the subject is amyloid-β positive (Aβ+) or Amyloid-β negative (Aβ-).

[0081] Embodiment 19. The method of any one of embodiments 1-18, wherein the control amount is determined using a control sample obtained from (i) a control subject not having neurodegeneration, or (ii) the subject at an earlier timepoint.

[0082] Embodiment 20. A method of treatment, comprising: (a) determining an amount of one or more taxa in a sample obtained from a subject having or suspected of having neurodegeneration, wherein the one or more taxa are selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis; (b) comparing the amount determined in (a) to a control amount; and (c) administering a treatment for neurodegeneration, based on the comparing in (b), optionally wherein the treatment comprises administering an agent to modulate the amount of the one or more taxa in the subject.

[0083] Embodiment 21. The method of embodiment 20, wherein the treatment is administered when the amount of Tannerella, Tanneralla forsythia,Atty. Dkt. No.650053.01137 Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control.

[0084] Embodiment 22. The method of embodiment 21, wherein the treatment is administered when the amount of Bilophila wadsworthia is in the sample about 2 times greater, or about 3 times greater, relative to the control.

[0085] Embodiment 23. The method of any one of embodiments 20-22, wherein the treatment is administered when the amount of Turicibacter sanguinis is decreased in the sample, relative to the control.

[0086] Embodiment 24. The method of embodiment 23, wherein the treatment is administered when the amount of Turicibacter sanguinis in the sample is about 3 times less, or about four times less, relative to the control.

[0087] Embodiment 25. The method of any one of embodiments 20-24, wherein the treatment is administered when a ratio of the amount of Bilophila wadsworthia to the amount of Turicibacter sanguinis in the sample is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1, relative to the control.

[0088] Embodiment 26. The method of any one of embodiments 20-25, wherein the control amount is determined using a control sample obtained from (i) a control subject not having or suspected of having neurodegeneration, or (ii) the subject at an earlier timepoint.

[0089] Embodiment 27. The method of any one of embodiments 20-26, wherein the determining comprises (i) isolating bacterial DNA from the sample and sequencing the isolated bacterial DNA.

[0090] Embodiment 28. The method of any one of embodiments 20-27, wherein the sample is a stool sample, a saliva sample, a blood sample, and / or a cerebrospinal fluid (CSF) sample. EXAMPLES

[0091] Example 1

[0092] Infectious viral agents are correlated with neurological tissue damage. Among putative viral influences, human herpesviruses, specifically human cytomegalovirus (HCMV) and herpes simplex virus 1 (HSV-1) are highly prevalent in adult population, and AD individuals exhibit the highest rate of seropositivity2-4. Recent evidence suggests that the alpha-herpesvirus HSV-1 does play a role in AD pathology4, 5. For example, studies have found that human stem cell-derived neural progenitor cells and organoids infected with HSV-1 exhibit increased amyloid beta (Aβ) expression, gliosis, and neuroinflammation6, 7. Peripheral blood mononuclear cells (PBMCs) from HCMV seropositive AD patients showed a greater pro-inflammatory response to anti-CD3 / CD28 stimulation compared to non-HCMV seropositive AD patients and healthy controls suggestingAtty. Dkt. No.650053.01137 that HCMV infection could enhance the pro-inflammatory environment contributing to AD pathology8. Another study found that HCMV antibody levels in the cerebral spinal fluid were positively correlated with neurofibrillary tangles9. The beta-herpesvirus HCMV is species specific, and the current dogma in the field is that HCMV easily infects NPCs but has limited tropism for differentiated neurons10, 11.

[0093] Porphyromonas gingivalis is strongly associated with Alzheimer’s Disease. One of the strongest correlations between bacteria and dementia and AD is the presence of Porphyromonas gingivalis and periodontal disease12. P. gingivalis DNA has been identified in the brains of patients with AD and in CSF of living subjects likely to have AD13. In addition, P. gingivalis outer membrane vesicles (OMVs) which carry bacterial lipids, proteins (e.g., gingipains), and metabolites have been shown to cross the blood brain barrier (BBB)13, 14and could deliver enzymes or metabolites capable of driving tauopathies and AD-like symptoms13.

[0094] Turicibacter depletion in MCI samples and bile acid pools. A strong correlation between gut bacteria and brain function has been demonstrated in mice by the Hsiao lab15. They show that Turicibacter sanguinis consumes serotonin to affect behavior and neurological symptoms in C57Bl / 6J animals16. Specifically, T. sanguinis possesses a 5-hydroxytryptamine (5- HT) transporter that functions like human serotonin transporter (SERT) and is similarly inhibited by selective-serotonin reuptake inhibitors (SSRI) including fluoxetine. Further, T. sanguinis has enzymatic capacity (e.g., bile salt hydrolases) to affect bile acid pools, steroids, and lipid composition profiles in a strain-dependent manner17. Turicibacter strains alter levels of chenodeoxycholic acid (CDCA), deoxycholic acid (DCA), glycocholic acid (GCA), taurocholic acid (TCA), taurochenodeoxycholic acid (TCDCA) and glycochenodeoxycholic acid (GCDCA) both in vitro and in mice colonized by those strains17, 18. Importantly, at least one bile acid, tauroursodeoxycholic acid (TUDCA) appears to protect neural tissue by inhibiting apoptosis, lowering oxidative stress and neuroinflammation19. As demonstrated in FIG. 13, the inventors discovered that subjects with MCI demonstrated increased levels of TCDCA. Without wishing to be bound by a particular theory, because lipid profiles are also affected by the ApoE^4 allele20which is strongly associated with AD21, 22, we expect that Turicibacter strains differentially affect host lipid pools ultimately promoting or preventing neurodegeneration.

[0095] Therefore, to investigate the impact of microbiome on neurodegeneration, stool and saliva samples were analyzed from subjects with mild cognitive impairment (MCI) and cognitively unimpaired (CU) controls. The MCI group included subjects with or without elevated amyloid beta (Aβ) expression (Aβ+; Aβ-).Atty. Dkt. No.650053.01137

[0096] Consistent with the dysbiosis model of neurodegeneration and other studies24, the analysis identified several taxa that were discriminatory for MCI relative to CU controls (FIG 1.). One prominent organism found in CU controls is Turicibacter sanguinis (FIGS. 1-2) which is known to affect bile acid pools. Turicibacter is part of the family Erysipelotrichaceae which are bile-acid tolerant organisms with hydrolases capable of modifying conjugated bile acids. Further, Bilophila wadsworthia respires taurine, which is an important component of conjugated bile acids, some of which are thought to prevent apoptosis and neurodegeneration (TUDCA)25.

[0097] It is worth noting that the data set showed no statistically significant changes in either alpha diversity (richness and evenness within samples) or beta diversity (sample similarity) when comparing MCI to CU controls, as has been reported elsewhere24, 26. This may reflect a lack of power for our preliminary study (total participants = 24), but also suggests that the discriminatory features found here may potentially serve to drive neurodegeneration without having arisen due to changes in gut composition or dysbiosis.

[0098] One of the strongest correlations between bacteria and dementia and AD is the presence of Porphyromonas gingivalis and periodontal disease13. Consistent with their results, we have identified a different oral pathogen, Tannerella forsythia, is differentially abundant in patients with mild cognitive impairment (MCI) relative to CU controls (FIG. 1). Both P. gingivalis and T. forsythia produce outer membrane vesicles with the potential to deliver toxins to host tissue.

[0099] Without wishing to be bound by a particular theory, the prevalence of bile-acid modifying organisms found in this analysis, including Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, NLAE zl H528, and Turicibacter sanguinis (FIG. 1, highlighted in blue), leads us to hypothesize that shifts in gut bacterial physiology affecting bile acid pools is a key feature to identify microbial drivers of neurodegeneration. Gut dysbiosis and corresponding bacterial physiology is widely regarded to affect obesity27, 28and cardiovascular disease29. A comprehensive view of bile acid profiles affecting neurodegeneration in only now beginning to emerge.

[0100] Bacterial strain Erysipelotrichaceae_UCG_003 may differentially alter the availability of bile acids which then supports the expansion of bile-acid tolerant NLAE zl H528 and Bilophila wadsworthia. B. wadsworthia would subsequently drive depletion of taurine-conjugated bile acids due to its ability to respire taurine and produce volatile hydrogen sulfide which is thought to affect barrier function18. As such, this constellation of organisms could present a physiological feed- forward loop to deplete critical bile acid components that would prevent Turicibacter from acting as a protective commensal in the human gut. In the absence of T. sanguinis, bile-acid driven dysbiosis (e.g., depletion of TUDCA) could then promote downstream neuroinflammation via apoptosis and degradation of host neural tissues. The data generated by this analysis demonstrateAtty. Dkt. No.650053.01137 a strong correlation for depletion of Turicibacter (FIG.2) and enrichment of Bilophila (FIG. 3), where the resulting Bilophila: Turicibacter ratio is 15:1 in MCI relative to CU control samples. Thus, we predict that unique bile acid profiles and identification of key metabolites (bile acids, taurine, H2S) will affect neural tissue in vitro.

[0101] Recent preliminary and published data demonstrates that HCMV is not restricted to NPCs as they found that terminally differentiated iPSC-derived neurons can be readily infected, produce new virus, and exhibit significant cellular and functional consequences30, which has important implications for neurodegenerative disease associated with aging. They showed that infection of neurons with the clinical strain HCMV strain expressing GFP induced an increase in insoluble S262-phosphorylated tau protein in AD patient and control neurons compared to mock conditions (FIG 4A), which is consistent with an increase of phosphorylated tau observed via immunofluorescence (FIG 4B). Moreover, infection essentially eliminated electrophysiological function in both AD patient and control iPSC-derived neurons (FIG 4C) further demonstrating the functional consequences that neurotrophic viruses can have on human neurons.

[0102] Future Studies (Phase I)

[0103] Despite our increasing understanding of the potential role for microbes as drivers of neurodegeneration, we lack sufficient animal models to capture the sporadic nature of the rise in AD in humans23. Similarly, animal microbiota is distinct from human gut microbes, especially at the sub-species or strain level, and thus we lack a clear understanding of the mechanistic components of AD pathology. or strain level, and thus we lack a clear understanding of the mechanistic components of AD pathology. Without wishing to be bound by a particular theory, we hypothesize that infectious agents gain access to central nervous system tissue, thereby acting as drivers of neurodegeneration, and that gut and oral dysbiosis may permit infectious disease agents to establish the cycle of neuroinflammation and neurodegeneration to promote dementia and AD. Identification of bacterial and / or viral signatures for neurodegeneration may allow for development of enhanced diagnostics and effective therapeutics in the battle against AD. To test this, we will characterize the gut and oral microbiota, associated bacterial metabolites, and viral factors correlated with neurological decline.

[0104] The overall objectives are to conduct metagenomic analyses of patient biospecimens to identify potential microbial drivers of neurodegeneration. We will examine microbial communities in blood, CSF, stool, and saliva samples from participants at baseline and in blood, stool, and saliva after a one year follow up-visit. We expect to identify discriminatory eukaryotic viruses, bacteria, and phage features (including enzymes and metabolites) distinguishing MCI,Atty. Dkt. No.650053.01137 Aβ+ or Aβ-, and CU cohorts. Top discriminatory features can be used to determine their impact on iPSC-derived neurons and organoids.

[0105] We will conduct metagenomic analyses of patient biospecimens to identify potential microbial drivers of neurodegeneration. We will examine microbial communities in blood, CSF, stool, and saliva samples from participants at baseline and in blood, stool, and saliva after a one year follow up-visit. We expect to identify discriminatory eukaryotic viruses, bacteria, and phage features (including enzymes and metabolites) distinguishing MCI, Aβ+ or Aβ-, and CU cohorts.

[0106] Define metagenomics of microbial communities in patient biospecimens. We will conduct 16S rDNA and whole-genome shotgun (WGS) sequencing for metagenomics on blood, cerebral spinal fluid (CSF), stool, and saliva samples from participants in all cohorts. We will analyze 640 baseline samples and 480 follow-up samples.

[0107] Study Design: Participants will be recruited from the MCW Disease Modifying Alzheimer’s Therapy (DMAT) and memory disorders clinics. Subjects will be screened for Aβ and neuropsychological (NP) testing for MCI during clinic visits. We will enroll three cohorts: (1) Aβ+ (n=80), (2) Aβ- (n=40), and (3) demographically matched cognitively unimpaired (CU) controls (n=40). Participants will be tracked longitudinally with a minimum of 1 follow-up visit at least one year after the first visit. Baseline stool, blood, CSF, and saliva will be collected. Stool, blood, and saliva will be collected during the follow-up visit. Samples will be analyzed for microbial factors and metabolic / proteomic changes associated with neurological disease. Samples will be processed and analyzed as described in Methods. Cellular and molecular impact on neural tissues will be investigated for a subset of samples for stem-cell modeling. The metabolomes and proteomes from a subset of samples will also be defined.

[0108] We will first conduct 16S rDNA sequencing to identify statistically significant shifts in gut microbial composition to identify samples that should be pursued for WGS metagenomic sequencing. Likewise, we will first assay stool and saliva samples as these are known to contain specific examples of bacteria and viruses described above. Further, the isolation and metagenomics pipelines are more robust for stool and saliva samples. Once we have identified samples for specific cohorts that display statistically significant differences, we will further assess CSF and blood for additional features. In addition, we will use the same datasets to identify phages and eukaryotic viruses as discriminatory features for samples distinguishing MCI and CU controls. We have extensive experience using next-generation sequencing for feature discovery and expect to identify robust signals within the data acquired here31-37.

[0109] Whole genome shotgun (WGS) sequencing of bacterial DNA is required because sub- species or strain-level sequence variation occurs widely for bacteria. WGS sequencing andAtty. Dkt. No.650053.01137 metagenome assembly will allow us to identify metabolic features enriched in Aβ+ samples relative to both Aβ- and CU controls. Furthermore, biosynthetic gene clusters encoding unique enzymatic functions are frequently mobilized and encoded within phages (viruses of bacteria) which do not possess 16S genes. Thus, WGS metagenomics allows us to discover unique genomic features associated with unique metabolite profiles with the potential to identify microbial discriminators of neurodegeneration. A subset of samples where phage-associated features are identified will be processed further for phage isolation and phageome determination (see Methods). It is worth noting that phage signatures are present in CSF38; thus, samples from any source may be assayed for phage nucleic acids.

[0110] Longitudinal Analysis: Our data from stool samples for MCI and CU controls (FIGS. 1-3) was produced from a cross-sectional study with cohorts of limited size. While we expect to see similar features, we also expect substantially increased resolution for discriminators of Aβ+, Aβ-, and CU controls. Importantly, we will collect blood, stool, and saliva during a second clinic visit. Metagenomic and metabolomic analyses of those samples, in comparison to baseline, will allow us to determine a subset of taxa, metagenomic features, and metabolites that may be responsible for neurocognitive decline within cohorts. A refined set of metabolites from these analyses will be used to further assess neurodegeneration in tissues.

[0111] Methods

[0112] Sample Collection: Blood and CSF samples: These samples will be collected during clinic visits. During clinic visits, subjects will be given additional kits to collect stool and saliva which will then be returned by mail to for downstream sequencing analyses and metagenomics analyses. For stool and saliva samples, we will utilize the DNA Genotek System kits, OmniGene Gut RNA / DNA kit (OMR-205) and OmniGene Saliva RNA / DNA kit (OMR-610), respectively. Multiple collection tubes will used for samples that may need to be processed separately (e.g., phage isolation). Bacterial and viral DNA and RNA will be isolated and processed for downstream 16S and whole-genome shotgun (WGS) metagenomic sequencing. For stool-based metabolomics we will utilize the OmniMet Gut kit for metabolites (ME-200). Samples will be analyzed to identify discriminatory metabolites including bile acids.

[0113] Batch effects: Multiple sample tubes will be collected for stool and saliva from each subject, split for analysis to avoid batch effects, and can be pooled to generate sufficient material for some downstream processes such as phage isolation. Stool samples are obtained via the OMNIgene-Gut DNA and RNA kit (OMR-205). Saliva samples will be obtained via the OMNIgene-Saliva DNA and RNA kit (OMR-610). Sample tubes will be split and treated separately for DNA or RNA isolation.Atty. Dkt. No.650053.01137

[0114] Isolation of DNA: We will utilize the Qiagen QIAamp DNeasy PowerFecal Pro Kit (cat#51804) for stool or DNeasy PowerBiofilm Kit (cat#24000) for saliva, blood, and CSF to extract DNA. We will use Zymobiomics DNA kit (cat#D4301) for additional cleanup and concentration (all DNA samples).

[0115] Isolation of RNA: We will utilize the Qiagen RNeasy PowerFecal Pro Kit (cat#78404) for stool or RNeasy PowerMicrobiome Kit (cat#26000) for saliva, blood, and CSF to extract RNA. We will use Zymobiomics RNA kit (cat#R2001) for additional cleanup and concentration (all RNA samples). Quality of both DNA and RNA is checked using the Nanodrop spectrophotometer and Qubit and Agilent Bioanalyzer. Samples are stored at -80°C until sequenced.

[0116] Isolation of phages from stool: Cesium chloride (CsCl) gradient centrifugation is used to prepare phage stocks from stool. Wet stool (~1 g from kit OMR-205) is resuspended in PBS, spun at 4000 x g for 15 min, and supernatant containing phage is decanted and stored on ice. The phage containing supernatant is then filtered through 0.45 µm filter 2x and then through a 0.2 µm filter 2x to remove bacteria or other debris. The filtered phage fraction is then subjected to ultracentrifugation (100,000 x g for 3h (for crude preparation) at 4°C) within a CsCl gradient spanning 1.75 g / ml to 1.35 g / ml in Beckman Ultra-Clear ultracentrifuge tubes (cat#344058). The 1.35 and 1.55 g / ml fractions are collected using an 18-gauge needle piercing the tube. The concentrated phage fraction is then dialyzed for several hours in PBS to gradually remove CsCl using a 10k molecular weight cut-off (MWCO) dialysis cassette. Finally, the cassette fraction is transferred to a centrifugal filter unit (10k MWCO) and spun at 4000 x g for 30 minutes to concentrate as desired. Aliquots can be combined and frozen at -80C for long term storage. To facilitate removal of fine contaminants such as LPS or peptidoglycan, fractions are centrifuged at 100,000 x g for 72 h and handled as described. Phage fractions are imaged on a 0.02 µm anodisc filter with SYBR-gold staining (FIGS.5A-C).

[0117] 16S rDNA bacterial gene sequencing: 16S rDNA libraries are generated using the V3- V4 region spanning primers 341F to 806R using the Illumina TruSeq DNA library prep system. PCR products are sequenced on the Illumina MiSeq platform using the 2x300-bp protocol. Amplicons are generated using a dual-indexing amplification strategy and optimized for sequencing.

[0118] 16S rDNA amplicon analysis: Paired-end reads are analyzed using Quantitative Insights into Microbial Ecology (QIIME2) which is updated regularly39. The current workflow is: Raw sequences are imported and summarized to check DNA quality. Sequences are denoised, filtered, trimmed, and chosen, following removal of chimeras, using the DADA2 plugin with a PHRED cutoff of 2540. Amplicon sequence variants (ASVs) are aligned with mafft41, 42and used toAtty. Dkt. No.650053.01137 construct a phylogeny with fasttree243. Alpha-diversity (richness and evenness of taxa within a population), beta-diversity (overlap for taxa shared between populations) and Principal Coordinates Analysis (PCoA) are estimated using the q2-diversity plugin after samples are rarefied. Taxonomy is assigned to ASVs using the q2-feature-classifier against SILVA reference databases. Changes in abundance of individual taxa are also analyzed using traditional univariate statistical methods. Linear discriminate analysis effect size (LEfSe)44, 45is used to determine discriminant ASVs (bacterial taxa; see FIG.1).

[0119] 16S rDNA sequence data suffices for identification of genus and / or species level taxonomic assignments as amplicon sequence variants (ASVs). Further, 16S amplicons are generated quantitatively allowing for characterization of abundance of ASVs within samples. In addition, due to the small size of the 16S amplicon, many samples can be sequenced and analyzed in a cost-effective manner. In short, 16S sequencing and statistical analyses allow us to ascribe which taxa and how much of each ASV is present within a given sample from a given experiment. However, 16S does not allow for resolution of strains (sub-species) nor for the accurate prediction of metabolic features that may be responsible for the phenotypes observed in our experiments. Therefore, 16S information is used to determine which samples show clear variation; those samples will then be sequenced via shotgun metagenomics, used to build MAGs to identify potential strains with critical functional pathways, thereby allowing us to identify discriminatory features of biological consequence within samples.

[0120] Whole metagenome sequencing for human samples: Sequences are generated on the Illumina NovaSeq platform producing 2x150 paired-end reads on an S4-flow cell. Depth of coverage is 10x greater than that used for mouse samples, resulting in ~100M reads per human sample.

[0121] Generation of Metagenomic Assembled Genomes (MAGs): Raw sequences are quality filtered using KneadData and Trimmomatic, assembled using MEGAHIT44to generate contigs >1000 bp. Anvi’o7.1 is used to simplify the assembled contigs. Bowtie2 is used to map short reads to the assembled contigs, followed by the generation of BAM files using samtools46, 47. Anvi’o is used to create a contigs database, which stores position of open reading frames, k-mer frequencies, and taxonomic and functional annotations. We run Hidden Markov Models to identify single-copy core genes and predict total genomes present. Taxonomic annotations are added using KAIJU48, functional annotations are imported using NCBI’s Cluster of Orthologous Groups and Pfam databases. Profile databases are generated for each sample (minimum length cut-off 2500 bp) using Anvi’o which provides sample-specific contig information such as mean coverage, standard deviation, and single nucleotide variant information. The individual profiles are merged into oneAtty. Dkt. No.650053.01137 profile with hierarchical clustering using Anvi’o and CONCOCT49, binning into preliminary MAGs. Downstream visualization, manual binning and analyses are performed using Anvi’o. MAGs are manually curated to generate genomes based on a high-quality threshold of 50% genome completion or genome size greater than 2Mb and less than 10% redundancy based on Campbell et al.50bacterial single-copy-core gene collection.

[0122] Analysis of MAGs: An abundance table of MAGs (FIG. 6) is exported from Anvi’o, where RStudio runs a random forest (rf) package to identify the top discriminatory features (MAGs) between samples (FIGS. 2-3). For smaller sam-ple sets, we implement a Boruta algorithm, which compares the importance of a real predictor variable with shadow variables using statistical testing and several permutations of rf. Top MAGs based on the Boruta algo-rithm are run through the rf-classification model to calculate their discriminatory capacity as a mean decrease Gini score. LEfSe45 is also used to determine discriminant MAGs.

[0123] Functional Analysis of Metabolic Pathways and KOs: A coverage table of metabolic pathways is exported from Anvi’o and analyzed by LEfSe to determine discriminant pathways between treatment groups. Top discriminatory pathways identified by LEfSe are used to generate network analysis. Using python modules (pandas, NumPy51, Matplotlib, and NetworkX), KO coverage will be imported, converted to a Pearson’s correlation matrix, and exported as a network. The network is reformatted and imported into Cytoscape52to visualize and annotate. To analyze the networks topology, the Prefuse Force Directed Layout based on correlation score (r-value) is implemented. The LDA score for each pathway is represented as node size. Gene coverage table of KOs is analyzed similarly and evaluated using Boruta algorithm via random forest to determine Gini scores. The log10 coverage of top KOs is visualized as a heatmap using gplots in Rstudio. These data are used to identify candidate microbial strains for inclusion in our synthetic microbial consortium.

[0124] In contrast to 16S gene sequencing, metagenomes provide information on the overall predicted metabolic and biochemical properties (e.g., biosynthetic gene clusters, toxin production, presence of phages, antibiotic resistance, and single nucleotide variants) within samples. For example, enrichment or depletion of a given organism or gene cluster in post-op HCD vs LCD would allow us to hypothesize a functional link between the microbiota and the overall resulting AnMet as measured for the host.

[0125] Determining the Phageome: Bacteriophages, or phages, are an important component of the microbiome due to their ability to modulate the gut ecosystem54. To analyze the taxonomic and functional state of phages in the gut, phages can be enriched from stool using CsCl density gradient ultracentrifugation. Following concentration and extraction of the phage fraction, DNAAtty. Dkt. No.650053.01137 is isolated using a virus-specific DNA isolation kit (Invitrogen PureLink™ Viral RNA / DNA Mini Kit cat #12280050). The DNA from these samples undergo whole metagenome sequencing at the as described above. Phage metagenomes are generated with Anvi’o as described above. However, rather than use Hidden Markov Models and KAIJU to predict the number of genomes present and annotate taxonomy (as for bacterial metagenomes), Virsorter2 and CheckV are run alongside Anvi’o using the Anvi’o-generated contigs database. Virsorter255is used to identify putative viral sequences and CheckV56is used to assess the quality and completeness of the Virsorter2-predicted genomes. DRAMv57is run to annotate the quality controlled Virsorter2 genomes, including the annotation of auxiliary metabolic genes present in the viral genomes. Viral sequences and annotations are uploaded into Anvi’o and the Anvi’o sequences are profiled and merged into a single profile. Anvi’o is used for downstream visualization of annotated genomes, and DRAMv annotations can be further analyzed using the Viral sequence identification SOP with VirSorter2 V.3 generated by the Sullivan lab55. Pairwise phage-host predictions are generated through DRAMv and a web platform that uses FASTA files to generate a probability score for each phage- host combination. Updated workflow will be incorporated as needed58, 59.

[0126] Determining the Eukaryotic virome: Metagenomes are generated with Anvi’o as described above following whole genome sequencing. During bacterial metagenome assembly, eukaryotic genes are systematically re-moved. To determine the eukaryotic virome, we will include those reads and follow the same processing as that for phageome determination. We will use Virsorter2 within the Anvi’o-generated contigs database. Virsorter255 identifies putative viral sequences and DRAMv57 is run to annotate Virsorter2 genomes. Viral sequences and annotations are uploaded into Anvi’o and the Anvi’o sequences are profiled and merged into a single profile. Anvi’o is used for downstream visualization of annotated genomes, and DRAMv annotations can be further analyzed using the Viral sequence identification SOP with VirSorter2 V.3 generated by the Sullivan lab55.

[0127] Future Studies (Phase II)

[0128] Information gathered from Phase I regarding 16S rDNA analyses and WGS metagenomics will allow us to identify bacterial strains (and their associated phages) for growth under anaerobic conditions. We have experience identifying specialized metabolites which affect microbial ecosystems62-65and expect to be able to do so in the context of this proposal. Discriminatory elements (metabolites, phages, bacterial enzymatic byproducts, eukaryotic viruses) can be used to assess impact on iPSC-derived neurons and organoids. Top known and unknown bacterial or phage features displaying statistically significant differences between Aβ+ samples and CU controls will be examined further.Atty. Dkt. No.650053.01137

[0129] Bacterial strains with known growth conditions can be obtained from ATCC (American Type Culture Collection) or the DSMZ (Deutsche Sammlung von Mikroorganismen und Zellkulturen) and grown in vitro in our anaerobic workstation. Based on the data described above, we predict that specific microbes (e.g., Porphyromonas gingivalis, Tannerella forsythia, Turicibacter sanguinis, Bilophila wadsworthia) (FIGS. 1-3) or their byproducts (e.g., outer membrane vesicles (OMVs), modified bile acids) will display statistically significant differences between Aβ+ samples and CU controls. Presence of microbial discriminators for Aβ+ samples will, therefore, be associated with neurodegeneration and selected for use.

[0130] Growth of anaerobic bacteria: All anaerobic strains currently in our possession are obtained from ATCC or DSMZ and handled in our Whitely A35 anaerobic workstation. We receive glass vials containing freeze-dried pellets under anaerobic gases (5% CO2 / 5% H2 / 90% N2). Pellets are typically rehydrated in brain-heart-infusion medium (BHI), subsequently transferred to BHI agar plates supplemented with sheep blood, grown for 3-5 days to produce colonies. We then use individual colonies for outgrowth in desired liquid media (per ATCC recommendations) specific for to each strain. Freezer stocks are generated in 20% glycerol. From freezer stocks, we streak onto appropriate plates for single colonies and then transfer into 2-3 ml of appropriate medium without shaking at 37oC in the anaerobic chamber and scale up volume as needed. Closed Falcon tubes can be transported to larger centrifuges or ultracentrifuges and returned to the anaerobic workstation as needed. We have successfully grown more than 20 gut related anaerobes for use in microbiome experiments using this workflow. Specifically, Porphyromonas gingivalis is grown in supplemented Tryptic soy broth with hemin and vitamin K1. Tannerella forsythia is grown in Tryptic soy broth plus 5% defibrinated sheep blood or Tryptic soy broth with hemin and N-acetyl muramic acid.

[0131] Isolation of outer membrane vesicles (OMVs): We have experience isolating OMVs66. Cultures are grown in appropriate media to an optical density corresponding to stationary phase which can vary for each organism. For OMV production, cultures are grown in multiple tubes or larger volumes. After growth, cultures are vortexed for 30 seconds to shear OMVs from cell surfaces, pelleted at 5000 x g for 20 min to remove whole cells, and supernatants are serially filter sterilized through 0.45 µm and 0.22 µm syringe filters to remove remaining debris. Cell-free supernatants are then centrifuged at 125,000 x g at for 3 h. Lipid pellets (containing OMVs) are suspended in PBS or appropriate buffer for downstream applications. Vesicles can be quantified by staining with FM 4-64 (Thermo-Scientific). Further processing of OMVs for use in metabolomics to determine composition or for direct application to astrocyte tissue culture orAtty. Dkt. No.650053.01137 organoids will be done as directed for those purposes. Growth media and other processing buffers would be used as a negative control for metabolomics determination of contents for OMVs.

[0132] Phage Characterization: Phages depend on bacterial hosts for propagation, either integrated in host genomes as prophages or as free particles following lytic induction. We expect to see prophage genomes within bacterial metagenomic datasets. Our data supports this where we see differentially abundant phages in MCI relative to CU controls (See, Example 2). However, because phages can exist independently following induction and are observed in CSF38, we may choose to isolate phages (using e.g., the CsCl centrifugation method; FIGS.5A-C) from samples independent from bacterial hosts. Phages typically carry remnant chromosomal DNA signatures from the site of integration in their bacterial host and frequently carry metabolic features that enable host cell physiology. Indeed, phage mobilization of traits such as antibiotic resistance genes, biosynthetic gene clusters, and other metabolism genes are readily observable in stool samples67-70. Importantly, phage associated sulfur metabolism genes have been implicated in taurine metabolism in mouse gut and may affect neurodegeneration18, 71. Thus, we will isolate phages from stool, saliva, and CSF samples and then sequence phageomes to identify potential traits of interest (e.g., bile acid metabolism) and to identify potential host sources. For example, CSF phages might derive from P. gingivalis, or T. forsythia or other unknown pathogenic organisms associated with MCI relative to CU controls. Features identified in this way will guide our use of specific metabolites for assessment on NPCs. It is worth noting that phageome sequences may identify known host-phage associations where both the host and phage are obtainable via the ATCC or DSMZ and can be used for further tool development in our laboratory.

[0133] Meta-transcriptomics on saliva samples: Based on the results from the 16S analysis and metagenomics and corresponding MRI and NP evaluations, we will select a subset of samples (N=10) from each of the 4 cohorts for meta-transcriptomic analyses from saliva samples. Multiple studies have identified P. gingivalis OMV contents that contribute to oral pathogenesis and could be drivers of neurodegeneration14, 72-74. We expect to identify gingipains13, 14, citrulline-modified proteins73, and other enzymes14capable of affecting host tissue using metabolomics. We will conduct meta-transcriptomics to identify differential gene expression regulating these features in P. gingivalis. Similar features may be present for Tannerella forsythia, although no reports have been made for T. forsythia genetic regulation of OMV production to our knowledge. Associated gene products and putative metabolites may be used to assess effects on tissue.

[0134] Based on the data described above (FIGS. 1-3), we expect to find discriminatory bacterial and phage features consistent with bile acid modifications, either enhanced or depleted, in stool samples from patients with MCI relative to controls. Specifically, we expect to findAtty. Dkt. No.650053.01137 differentially abundant genes encoding bile salt hydrolases from Turicibacter sanguinis or related organisms in the family, Erysipelotrichaceae, or other organisms known to be bile-salt tolerant or to utilize components of bile salts such as Bilophila wadsworthia which respires taurine to produce hydrogen sulfide18. The differential abundance of these organisms should correspond to differential abundances of modified bile acid metabolites and include chenodeoxycholic acid (CDCA), deoxycholic acid (DCA), glycocholic acid (GCA), taurocholic acid (TCA), taurochenodeoxycholic acid (TCDCA), glycochenodeoxycholic acid (GCDCA), and tauroursodeoxycholic acid (TUDCA). Additional metabolites (e.g., butyrate, H2S, taurine) are expected to be identified. Some genes encoding bile-acid modifications may be identified in phages associated with the above microbes.

[0135] Additional potential outcomes include discriminatory feature abundance from Aβ- subjects relative to other groups. Feature abundance from those samples may fall between Aβ+ samples and CU controls. Presence of microbial discriminators for Aβ+ samples will, therefore, be associated with neurodegeneration and selected for use.

[0136] Future Studies (Phase III)

[0137] Infection has been implicated as a contributing factor in the development of AD, and this includes both acute and chronic eukaryotic animal viral infections. For example, the neurotropic human herpesvirus, herpes sim-plex virus 1 (HSV-1) replicates in mucosal epithelia while entering innervating neurons to establish a latent life-long state with periods of reactivation (Reviewed in75). HSV-1 is often detected in brains of AD patients and, infection in vitro induces Aβ42 expression and tau phosphorylation in neuronal cells and tissues6, 76, 77. Analysis of large patient cohorts including diverse biomedical data types has implicated HSV-1 along with human her-pesviruses 6 and cytomegalovirus in contributing to AD neuropathology78. Experimental and computational analyses also implicated acute viral infections such as influenza A and SARS-CoV- 2 in AD79, including a recent study associating SARS-CoV-2 infection with herpesvirus reactivation in AD80. However, despite these associations, no specific animal virus has been identified as a causative agent of AD, and this supports the hypothesis that multiple drivers exist likely occurring in various combinations. The fact that many of these viruses are ubiquitous in the human populations further complicates defining potential drivers. Therefore, we will profile sali- va, serum, stool, and CSF viromes of Aβ+, Aβ-, and control cognitively unimpaired (CU) cohorts with the goal of defining positive and negative correlations of animal viruses with AD. These analyses will be completed in parallel to Phase I and provided sufficient data sets to define correlations.Atty. Dkt. No.650053.01137

[0138] Determine positive and negative correlations with AD using experimental and published data sets: Nucleic acid isolation and metagenomics analysis completed in Phase I will define bacteria, phage and animal virus families and genera in serum, CSF, stool, and saliva from MCI, Aβ+ or Aβ-, and CU cohorts. Animal viruses will be identified in the same analysis pipeline used in phage identification. In contrast to human phage viromes, limited studies exist defining human animal viromes. In these studies, we will follow the approach used by Ye et al81recently published, defining oral and gut viromes in hypertension81, 82. Animal viruses will be analyzed separate from phage defining alpha and beta diversity, family and genus composition per patient, and anatomical compartment (i.e., CSF, stool, saliva, serum), and differential composition between patient and between disease states. Linear discriminate analysis effect size (LEfSe)39, 44,45will be used to determine discriminant ASVs as described above. Currently, 273 NCBI RefSeq human-associated viral genomes are available at NCBI Virus82, 83and, in addition to defining correlations with AD, our studies will increase the robustness of available human virome datasets which are currently underdeveloped. Any contig that fails to align will be analyzed following the strategy by Abbas et al.84in the recent discovery of redondovirus genomes.

[0139] Confirmation of animal viruses with positive and negative correlations to AD: Studies on CSF, saliva and stool have uncovered animal viruses (reviewed in85). For example, lesser- known viruses such as anelloviruses, redondoviruses, and cressdnaviruses along with common human herpesviruses and papillomavirus subtypes can be detected in saliva86, 87. Interestingly, redondovirus has been associated with periodontal disease and elevated in the respiratory tract of critical ill patients84. Animal viruses have also been detected in CSF of otherwise healthy individuals including herpesviruses38. However, of the few studies published, bacteriophages are the dominant virus found in CSF38, 85, exceptions being animal viruses in cases of encephalitis and meningitis88. In our studies, the presence of animal viruses identified by metagenomics and determined to correlate with AD will be subsequently validated using TaqMan-based PCR on the given patient DNA or RNA sample (saliva, serum, CSF or stool) and samples lacking the identified virus serving as negative controls. Upon confirmation, subsequent PCRs will be done on all samples from the given individual to assess the other anatomical compartments. Finally, for common human viral pathogens for which diagnostic antibodies exist, we will complete serology as an additional confirmatory test.

[0140] Test determinants that contribute to AD pathologies: Infection by animal viruses potentially causes cellular damage, stimulate and / or suppress the immune system, and, in some cases, establish lifelong persistent infections challenging the immune system for the life of the host. Further, some viruses have established in vitro systems to study pathogenesis while otherAtty. Dkt. No.650053.01137 newly discovered animal virus such as redondovirus lack culture systems. Viruses are obligate intracellular parasites, and unlike bacteria, they do not have metabolic pathways that can be used for discriminatory features. Therefore, we will define discriminatory features for each virus identified based on current literature to include: (1) family and genus, (2) acute, persistent, or latent infection, (3) cell / tissue tropism including in the CNS, (4) benign (e.g., anelloviruses) or pathogenic (e.g., herpesviruses) infections, and (5) epidemiology. Further, we will analyze published NCBI metagenome data sets of viromes from similar tissue to assess frequency of detection. If this knowledge is not available and an in in vitro culture system is developed, we will obtain the virus from ATCC or an individual lab investigating the pathogen. We will work with the top ranked virus, based on linear discriminate analysis and validation as noted above, within our BSL2-designated lab after receiving institutional approval. We will evaluate viral replication kinetics, susceptibility to human cell lines (e.g., fibroblast, epithelial, endothelial) including neural progenitor cells (NPC). If we can culture the virus, materials including titered viral stocks and resulting conditioned media from infections will be provided to investigate the impact of infection and / or infected cell secretome on iPSC-derived neurons and choroid plexus organoids. If the virus has no available culture system, we will use a virus in the same family or use the next ranked virus to complete the proposed experiments. Regardless of our ability to culture the top candidate(s), upon completion, we will have a ranked list of known and likely previously unknown animal viruses from different samples sites (i.e., serum, stool, CSF, saliva) that differentially associate with AD compared to CU controls.

[0141] Unlike efforts to define bacterial and phage human microbiomes, defining eukaryotic animal viruses co-existing with the human population is in its infancy. In these studies, we will apply similar approaches developed for phage to uncover the animal virome in diverse patient cohorts and from diverse sample sites (see Determining the Eukaryotic Virome in Methods). Samples to be analyzed for eukaryotic viruses are from a subset of samples with highly discriminatory features determined by 16S rDNA sequencing in Phase I. The results from these studies may identify animal viruses associated with AD, as well as significantly advance our knowledge of the human virome by repurposing pipelines developed for defining phageomes.

[0142] References 1. Association As. 2022 Alzheimer's disease facts and figures. Alzheimers Dement. 2022;18(4):700-89. Epub 20220314. doi: 10.1002 / alz.12638. 2. Catterall WA. Cooperative activation of action potential Na+ ionophore by neurotoxins. Proc Natl Acad Sci U S A.1975;72(5):1782-6. doi: 10.1073 / pnas.72.5.1782.Atty. Dkt. No.650053.01137 3. Carbone I, et al. Herpes virus in Alzheimer's disease: relation to progression of the disease. Neurobiol Aging.2014;35(1):122-9. Epub 20130803. doi: 10.1016 / j.neurobiolaging.2013.06.024. 4. Harris SA, Harris EA. Herpes Simplex Virus Type 1 and Other Pathogens are Key Causative Factors in Sporadic Alzheimer's Disease. J Alzheimers Dis. 2015;48(2):319-53. doi: 10.3233 / JAD-142853. 5. Sait A, et al. Viral Involvement in Alzheimer's Disease. ACS Chem Neurosci. 2021;12(7):1049-60. Epub 20210309. doi: 10.1021 / acschemneuro.0c00719. 6. Abrahamson EE, et al. Modeling Abeta42 Accumulation in Response to Herpes Simplex Virus 1 Infection: 2D or 3D? J Virol.2021;95(5). Epub 20201202. doi: 10.1128 / JVI.02219-20. 7. Qiao H, et al. Cerebral Organoids for Modeling of HSV-1-Induced-Amyloid beta Associated Neuropathology and Phenotypic Rescue. Int J Mol Sci.2022;23(11). Epub 20220526. doi: 10.3390 / ijms23115981. 8. Westman G, et al. Increased inflammatory response in cytomegalovirus seropositive patients with Alzheimer's disease. PLoS One. 2014;9(5):e96779. Epub 20140507. doi: 10.1371 / journal.pone.0096779. 9. Lurain NS, et al. Virological and immunological characteristics of human cytomegalovirus infection associated with Alzheimer disease. J Infect Dis. 2013;208(4):564-72. Epub 20130509. doi: 10.1093 / infdis / jit210. 10. Teissier N, et al. Cytomegalovirus-induced brain malformations in fetuses. J Neuropathol Exp Neurol.2014;73(2):143-58. doi: 10.1097 / NEN.0000000000000038. 11. Krstanovic F, et al. Cytomegalovirus Infection and Inflammation in Developing Brain. Viruses.2021;13(6). Epub 20210604. doi: 10.3390 / v13061078. 12. Elwishahy A, et al. Gingivalis as a Risk Factor to Alzheimer's Disease: A Systematic Review. J Alzheimers Dis Rep.2021;5(1):721-32. Epub 20210913. doi: 10.3233 / ADR-200237. 13. Dominy SS, et al. Porphyromonas gingivalis in Alzheimer's disease brains: Evidence for disease causation and treatment with small-molecule inhibitors. Sci Adv. 2019;5(1):eaau3333. Epub 20190123. doi: 10.1126 / sciadv.aau3333. 14. Vermilyea DM, et al. PPAD Activity Promotes Outer Membrane Vesicle Biogenesis and Surface Translocation by Porphyromonas gingivalis. J Bacteriol. 2021;203(4). Epub 20210125. doi: 10.1128 / JB.00343-20. 15. Jameson KG, et al. Toward Understanding Microbiome-Neuronal Signaling. Mol Cell. 2020;78(4):577-83. Epub 20200409. doi: 10.1016 / j.molcel.2020.03.006.Atty. Dkt. No.650053.01137 16. Fung TC, et al. Intestinal serotonin and fluoxetine exposure modulate bacterial colonization in the gut. Nat Microbiol. 2019;4(12):2064-73. Epub 20190902. doi: 10.1038 / s41564-019-0540-4. 17. Lynch JB, et al. Gut microbiota Turicibacter strains differentially modify bile acids and host lipids. Nat Commun.2023;14(1):3669. Epub 20230620. doi: 10.1038 / s41467-023-39403-7. 18. Ye H, et al. Ecophysiology and interactions of a taurine-respiring bacterium in the mouse gut. Nat Commun.2023;14(1):5533. Epub 20230918. doi: 10.1038 / s41467-023-41008-z. 19. Khalaf K, et al. Tauroursodeoxycholic acid: a potential therapeutic tool in neurodegenerative diseases. Transl Neurodegener. 2022;11(1):33. Epub 20220604. doi: 10.1186 / s40035-022-00307-z. 20. Bantugan MA, et al. Associations of ApoE4 status and DHA supplementation on plasma and CSF lipid profiles and entorhinal cortex thickness. J Lipid Res. 2023;64(6):100354. Epub 20230322. doi: 10.1016 / j.jlr.2023.100354. 21. Palmqvist S, et al. An accurate fully automated panel of plasma biomarkers for Alzheimer's disease. Alzheimers Dement.2023;19(4):1204-15. Epub 20220811. doi: 10.1002 / alz.12751. 22. Chatterjee P, et al. Diagnostic and prognostic plasma biomarkers for preclinical Alzheimer's disease. Alzheimers Dement. 2022;18(6):1141-54. Epub 20210908. doi: 10.1002 / alz.12447. 23. Jankowsky JL, Zheng H. Practical considerations for choosing a mouse model of Alzheimer's disease. Mol Neurodegener. 2017;12(1):89. Epub 20171222. doi: 10.1186 / s13024- 017-0231-7. 24. Rey KM, et al. Dysbiosis of the Female Murine Gut Microbiome Exacerbates Neutrophil- mediated Vascular Allograft Injury by Affecting Immunoregulation by Acetate. Transplantation. 2022;106(11):2155-65. Epub 20221021. doi: 10.1097 / TP.0000000000004161. 25. Zangerolamo et al. The bile acid TUDCA and neurodegenerative disorders: An overview. Life Sci.2021;272:119252. Epub 20210223. doi: 10.1016 / j.lfs.2021.119252. 26. Vogt NM, et al. Gut microbiome alterations in Alzheimer's disease. Sci Rep. 2017;7(1):13537. Epub 20171019. doi: 10.1038 / s41598-017-13601-y. 27. Abenavoli L, et al. Gut Microbiota and Obesity: A Role for Probiotics. Nutrients. 2019;11(11). Epub 20191107. doi: 10.3390 / nu11112690. 28. Riedl RA, et al. Gut Microbiota Represent a Major Thermogenic Biomass. Function (Oxf). 2021;2(3):zqab019. Epub 20210415. doi: 10.1093 / function / zqab019. 29. Witkowski M, et al. Gut Microbiota and Cardiovascular Disease. Circ Res. 2020;127(4):553-70. Epub 20200730. doi: 10.1161 / CIRCRESAHA.120.316242.Atty. Dkt. No.650053.01137 30. Adelman JW, et al. Human cytomegalovirus induces significant structural and functional chsnges in terminally differentiated human cortical neurons. mBio. 2023;14(6):e0225123. Epub 20231115. doi: 10.1128 / mbio.02251-23. 31. Ghose C, et al. The Virome of Cerebrospinal Fluid: Viruses Where We Once Thought There Were None. Front Microbiol. 2019;10:2061. Epub 20190906. doi: 10.3389 / fmicb.2019.02061. 32. Bolyen E, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol.2019;37(8):852-7. doi: 10.1038 / s41587-019-0209-9. 33. Callahan BJ, et al. Bioconductor Workflow for Microbiome Data Analysis: from raw reads to community analyses. F1000Res. 2016;5:1492. Epub 20160624. doi: 10.12688 / f1000research.8986.2. 34. Katoh K, et al. MAFFT online service: multiple sequence alignment, interactive sequence choice and visualization. Brief Bioinform.2019;20(4):1160-6. doi: 10.1093 / bib / bbx108. 35. Katoh K, et al. MAFFT: a novel method for rapid multiple sequence alignment based on fast Fourier transform. Nucleic Acids Res.2002;30(14):3059-66. doi: 10.1093 / nar / gkf436. 36. Price MN, et al. FastTree 2--approximately maximum-likelihood trees for large alignments. PLoS One.2010;5(3):e9490. Epub 20100310. doi: 10.1371 / journal.pone.0009490. 37. Segata N, et al. Metagenomic biomarker discovery and explanation. Genome Biol. 2011;12(6):R60. Epub 20110624. doi: 10.1186 / gb-2011-12-6-r60. 38. Segata N, Huttenhower C. Toward an efficient method of identifying core genes for evolutionary and functional microbial phylogenies. PLoS One. 2011;6(9):e24704. Epub 20110912. doi: 10.1371 / journal.pone.0024704. 39. Danecek P, et al. Twelve years of SAMtools and BCFtools. Gigascience.2021;10(2). doi: 10.1093 / gigascience / giab008. 40. Li H, et al. Genome Project Data Processing S. The Sequence Alignment / Map format and SAMtools. Bioinformatics. 2009;25(16):2078-9. Epub 20090608. doi: 10.1093 / bioinformatics / btp352. 41. Menzel P, et al. Fast and sensitive taxonomic classification for metagenomics with Kaiju. Nat Commun.2016;7:11257. Epub 20160413. doi: 10.1038 / ncomms11257. 42. Alneberg J, et al. Binning metagenomic contigs by coverage and composition. Nat Methods.2014;11(11):1144-6. Epub 20140914. doi: 10.1038 / nmeth.3103. 43. Campbell JH, et al. UGA is an additional glycine codon in uncultured SR1 bacteria from the human microbiota. Proc Natl Acad Sci U S A. 2013;110(14):5540-5. Epub 20130318. doi: 10.1073 / pnas.1303090110.Atty. Dkt. No.650053.01137 44. Harris CR, et al. Array programming with NumPy. Nature.2020;585(7825):357-62. Epub 20200916. doi: 10.1038 / s41586-020-2649-2. PubMed PMID: 32939066; PMCID: PMC7759461. 45. Shannon P, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res.2003;13(11):2498-504. doi: 10.1101 / gr.1239303. 46. Frank JA, et al. Improved metagenome assemblies and taxonomic binning using long-read circular consensus sequence data. Sci Rep. 2016;6:25373. Epub 20160509. doi: 10.1038 / srep25373. 47. Zuppi M, et al. Phages in the Gut Ecosystem. Front Cell Infect Microbiol.2021;11:822562. Epub 20220104. doi: 10.3389 / fcimb.2021.822562. 48. Guo J, et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome. 2021;9(1):37. Epub 20210201. doi: 10.1186 / s40168-020-00990- y. 49. Nayfach S, et al. CheckV assesses the quality and completeness of metagenome-assembled viral genomes. Nat Biotechnol. 2021;39(5):578-85. Epub 20201221. doi: 10.1038 / s41587-020- 00774-7. 50. Shaffer M, et al. DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res.2020;48(16):8883-900. doi: 10.1093 / nar / gkaa621. 51. Johansen J, et al. Centenarians have a diverse gut virome with the potential to modulate metabolism and promote healthy lifespan. Nat Microbiol. 2023;8(6):1064-78. Epub 20230515. doi: 10.1038 / s41564-023-01370-6. 52. Shalon D, et al. Profiling the human intestinal environment under physiological conditions. Nature. 2023;617(7961):581-91. Epub 20230510. doi: 10.1038 / s41586-023-05989-7. PubMed PMID: 37165188; PMCID: PMC10191855. 53. Ferdous T, et al. The rise to power of the microbiome: power and sample size calculation for microbiome studies. Mucosal Immunol. 2022;15(6):1060-70. Epub 20220722. doi: 10.1038 / s41385-022-00548-1. 54. Kelly BJ, et al. Power and sample-size estimation for microbiome studies using pairwise distances and PERMANOVA. Bioinformatics. 2015;31(15):2461-8. Epub 20150329. doi: 10.1093 / bioinformatics / btv183. 55. Muller S, et al. Thiocillin contributes to the ecological fitness of Bacillus cereus ATCC 14579 during interspecies interactions with Myxococcus xanthus. Front Microbiol. 2023;14:1295262. Epub 20231124. doi: 10.3389 / fmicb.2023.1295262.Atty. Dkt. No.650053.01137 56. Muller S, et al. Bacillaene and sporulation protect Bacillus subtilis from predation by Myxococcus xanthus. Appl Environ Microbiol. 2014;80(18):5603-10. Epub 20140707. doi: 10.1128 / AEM.01621-14. 57. Muller S, et al. Predation by Myxococcus xanthus induces Bacillus subtilis to form spore- filled megastructures. Appl Environ Microbiol. 2015;81(1):203-10. Epub 20141017. doi: 10.1128 / AEM.02448-14. 58. Muller S, et al. Identification of Functions Affecting Predator-Prey Interactions between Myxococcus xanthus and Bacillus subtilis. J Bacteriol. 2016;198(24):3335-44. Epub 20161118. doi: 10.1128 / JB.00575-16. 59. Bretl DJ, et al. Type IV-pili dependent motility is co-regulated by PilSR and PilS2R2 two- component systems via distinct pathways in Myxococcus xanthus. Mol Microbiol. 2016;102(1):37-53. Epub 20160823. doi: 10.1111 / mmi.13445. 60. Sun M, et al. Underexplored viral auxiliary metabolic genes in soil: Diversity and eco- evolutionary significance. Environ Microbiol. 2023;25(4):800-10. Epub 20230105. doi: 10.1111 / 1462-2920.16329. 61. Shuwen H, Kefeng D. Intestinal phages interact with bacteria and are involved in human diseases. Gut Microbes.2022;14(1):2113717. doi: 10.1080 / 19490976.2022.2113717. 62. Kim M, et al. Gain and loss of antibiotic resistant genes in multidrug resistant bacteria: One Health perspective. J Microbiol. 2021;59(6):535-45. Epub 20210420. doi: 10.1007 / s12275- 021-1085-9. 63. Nadeem SF, et al. Antimicrobial resistance: more than 70 years of war between humans and bacteria. Crit Rev Microbiol. 2020;46(5):578-99. Epub 20200919. doi: 10.1080 / 1040841X.2020.1813687. 64. Rosshart SP, et al. Laboratory mice born to wild mice have natural microbiota and model human immune responses. Science. 2019;365(6452). Epub 20190801. doi: 10.1126 / science.aaw4361. 65. Choi JW, et al. Secretable Small RNAs via Outer Membrane Vesicles in Periodontal Pathogens. J Dent Res.2017;96(4):458-66. Epub 20170109. doi: 10.1177 / 0022034516685071. 66. Larsen DN, et al. Citrullinome of Porphyromonas gingivalis Outer Membrane Vesicles: Confident Identification of Citrullinated Peptides. Mol Cell Proteomics.2020;19(1):167-80. Epub 20191121. doi: 10.1074 / mcp.RA119.001700. 67. Waller T et al. Porphyromonas gingivalis Outer Membrane Vesicles Induce Selective Tumor Necrosis Factor Tolerance in a Toll-Like Receptor 4- and mTOR-Dependent Manner. Infect Immun.2016;84(4):1194-204. Epub 20160324. doi: 10.1128 / IAI.01390-15.Atty. Dkt. No.650053.01137 68. Feng S, et al. Mechanistic insights into the role of herpes simplex virus 1 in Alzheimer's disease. Front Aging Neurosci. 2023;15:1245904. Epub 20230907. doi: 10.3389 / fnagi.2023.1245904. 69. Eimer WA, et al. Alzheimer's Disease-Associated beta-Amyloid Is Rapidly Seeded by Herpesviridae to Protect against Brain Infection. Neuron. 2018;99(1):56-63 e3. doi: 10.1016 / j.neuron.2018.06.030. 70. Cairns DM, et al. A 3D human brain-like tissue model of herpes-induced Alzheimer's disease. Sci Adv.2020;6(19):eaay8828. Epub 20200506. doi: 10.1126 / sciadv.aay8828. 71. Readhead B, et al. Multiscale Analysis of Independent Alzheimer's Cohorts Finds Disruption of Molecular, Genetic, and Clinical Networks by Human Herpesvirus. Neuron. 2018;99(1):64-82 e7. Epub 20180621. doi: 10.1016 / j.neuron.2018.05.023.. 72. Levine KS, et al. Virus exposure and neurodegenerative disease risk across national biobanks. Neuron.2023;111(7):1086-93 e2. Epub 20230119. doi: 10.1016 / j.neuron.2022.12.029. 73. Itzhaki RF. COVID-19 and Alzheimer's Disease: What Is the Connection? J Alzheimers Dis.2023;91(4):1273-6. doi: 10.3233 / JAD-220955. 74. Ye HL, et al. Alterations of oral and gut viromes in hypertension and / or periodontitis. mSystems.2023:e0116923. Epub 20231218. doi: 10.1128 / msystems.01169-23. 75. Hatcher EL, et al. Virus Variation Resource - improved response to emergent viral outbreaks. Nucleic Acids Res. 2017;45(D1):D482-D90. Epub 20161128. doi: 10.1093 / nar / gkw1065. 76. Simmonds P, et al. Consensus statement: Virus taxonomy in the age of metagenomics. Nat Rev Microbiol.2017;15(3):161-8. Epub 20170103. doi: 10.1038 / nrmicro.2016.177. 77. Abbas AA, et al. Redondoviridae, a Family of Small, Circular DNA Viruses of the Human Oro-Respiratory Tract Associated with Periodontitis and Critical Illness. Cell Host Microbe. 2019;25(5):719-29 e4. doi: 10.1016 / j.chom.2019.04.001.. 78. Liang G, Bushman FD. The human virome: assembly, composition and host interactions. Nat Rev Microbiol.2021;19(8):514-27. Epub 20210330. doi: 10.1038 / s41579-021-00536-5. 79. Paietta EN, et al. Characterization of Diverse Anelloviruses, Cressdnaviruses, and Bacteriophages in the Human Oral DNA Virome from North Carolina (USA). Viruses.2023;15(9). Epub 20230826. doi: 10.3390 / v15091821. 80. Briese T, et al. A multicenter virome analysis of blood, feces, and saliva in myalgic encephalomyelitis / chronic fatigue syndrome. J Med Virol. 2023;95(8):e28993. doi: 10.1002 / jmv.28993.Atty. Dkt. No.650053.01137 81. Benninger F, Steiner I. CSF in acute and chronic infectious diseases. Handb Clin Neurol. 2017;146:187-206. doi: 10.1016 / B978-0-12-804279-3.00012-5.

[0143] Example 2: Gut Microbiota and Neurovascular Patterns in Amnestic Mild Cognitive Impairment

[0144] The interplay between the gut microbiome (GMB) and neurovascular function in Alzheimer’s disease (AD) is unclear. The goal of this study is to identify relationships between GMB, neurovascular functioning, and cognition in amnestic mild cognitive impairment (aMCI). Participants (N = 24) provided fecal samples for GMB sequencing (16S and shotgun metagenomics), underwent an MRI, and completed cognitive testing. Cerebral Vascular Reactivity (CVR), Cerebral Blood Flow (CBF) and Arterial Transit Times (ATT) were assessed. Statistical analyses evaluated relationships between discriminatory taxa, cerebrovascular metrics, and cognition. Sequencing revealed differentially abundant bacterial and viral taxa distinguishing aMCI from controls. Spearman correlations revealed that bacteria known to induce inflammation were negatively associated with CVR, CBF, and cognition, and positively associated with ATT. A reciprocal pattern emerged with taxa associated with gut health. Our results suggest that cerebrovascular changes and GMB may hold untapped potential as AD biomarkers and treatment targets.

[0145] Introduction

[0146] Alzheimer’s disease (AD) is the most prevalent neurodegenerative disease in older adults and a growing epidemic worldwide. After decades of research, there remains no effective treatment for AD. Treatment failures can be attributed, at least in part, to an incomplete understanding of the pathogenic mechanisms of AD, which underscores the urgent need to examine alternative pathogenic mechanisms and therapeutic targets for AD.

[0147] Within the last decade, the gut microbiome (GMB) has been increasingly implicated in the progression of AD pathogenesis, even at the preclinical and prodromal stages of the disease1,2. Pathways linking the gut and brain in AD have been proposed3. GMB studies (mostly using 16S rRNA sequencing) reveal bacterial taxa capable of distinguishing AD from MCI and controls, along with associations with cerebral spinal fluid (CSF) biomarkers of AD1,4,5and brain MRI in MCI and AD6,7. Of the few studies that have examined the relationship between the GMB and brain MRI in AD8, there is only one known study examining associations between the GMB and functional MRI in AD6, and one study examining diffusion MRI in AD7. Additionally, only one other study has investigated gut virome differences in prodromal AD9. To our knowledge, no study has examined the interplay between bacterial and viral taxa and neurovascular changes in AD.Atty. Dkt. No.650053.01137

[0148] Considering the substantial evidence connecting early vascular contributions to AD pathophysiology and dementia, cerebrovascular dysfunction has emerged as a major contributor to cognitive decline and disease progression in AD10. The blood-brain barrier (BBB), a neurovascular structure responsible for regulating the passage of cells and molecules to and from the central nervous system (CNS), has been identified as a key player in this process11. Disruption of the BBB allows the influx of neurotoxic blood-derived molecules, cells, and microbial pathogens into the brain, triggering inflammatory and immune responses that can initiate various pathways of neurodegeneration12,13. Growing evidence suggests that dysfunction and / or breakdown of BBB, as well as reductions and / or dysregulation of cerebral blood flow (CBF as measured by MRI), may occur in sporadic AD and experimental models of the disease prior to cognitive decline, Aβ deposition, and brain atrophy13,14. The cerebral vasculature is the locus where multiple pathogenic processes converge and contribute to cognitive impairment15and has given rise to speculation regarding the existence of a common pathological triad model that consists of vascular damage, neurodegeneration, and neuroinflammation16, expanding upon the traditional understanding of AD. Recent neuropathological research has provided evidence to support this notion, demonstrating that systemic infection can alter brain cytokine levels and exacerbate cerebral hypoperfusion and BBB leakiness associated with AD independent of the level of insoluble Aβ14. However, the complex crosstalk between neuroinflammation, immune responses, and vascular changes in the context of neurodegeneration remains poorly understood. Thus, it is essential to acquire a deeper understanding of the underlying pathophysiology. Mounting evidence indicates involvement of inflammation-inducing gut bacterial taxa in the onset and progression of AD1,2and provides insights into possible microbial mechanisms involved in the onset and progression of AD.

[0149] The purpose of the current study is to understand the interplay between gut microbes, neurovascular dysfunction, and cognition in patients clinically diagnosed with amnestic mild cognitive impairment (aMCI) compared to cognitively unimpaired older adult controls. We examined 16S and shotgun metagenomics sequencing, multimodal neuroimaging metrics to assess neurovascular dysfunction, and performance on neuropsychological measures of memory, language, processing speed, and executive functioning.

[0150] Our findings link bacteria known to induce inflammation to neurovascular and cognitive dysfunction in aMCI, possibly suggesting that GMB markers may predict cognitive changes in prodromal AD through neurovascular pathways. These results link gut microbial dysbiosis in aMCI to neurovascular and cognitive dysfunction and expands on current literature by suggesting that bacteriophages may play a role in the development of AD. Future research willAtty. Dkt. No.650053.01137 investigate the causality of microbial dysbiosis on the development of neurovascular dysfunction and cognitive decline.

[0151] Materials and Methods

[0152] Participants: All subjects provided written informed consent prior to participation in this study, which was approved by the local Institutional Review Board and conducted in accordance with the Declaration of Helsinki and its later amendments or comparable ethical standards. Participants (63 to 84 years old; 58% female, 96% White) were recruited from our ongoing NIA R21 study, which is a cross-sectional, non-interventional study that examines associations between bacterial taxa (stool), neurovascular functioning (brain MRI), and cognition. We selected these 24 participants for the current study because both stool and neuroimaging data were processed and analyzed. Participants (n = 14) were diagnosed with aMCI by a clinical neuropsychologist or memory disorders neurologist as part of routine clinical care at Froedtert and the Medical College of Wisconsin. The MCI diagnosis was documented in a recent clinical visit with a neurologist or neuropsychologist within 6 months or less of the first study visit. Ten participants were cognitive unimpaired control subjects as supported by a brief neuropsychological research evaluation. Exclusion criteria for all participants included a history of moderate to severe TBI, brain tumor, symptomatic stroke, severe psychiatric illness (i.e., schizophrenia and bipolar disorders), intellectual disability, dementia / major neurocognitive disorder, end stage renal or liver disease, drug addiction, alcoholism, inflammatory bowel disease, recent history of viral infections, antibiotic use within two months before stool sample collection, and age younger than 60 years. Participants with a history of hypertension (aMCI = 25%; controls = 40%), diabetes (aMCI = 14%; controls = 0%), or high cholesterol (aMCI = 43%; controls = 20%) were able to participate in the current study as these are common health conditions in older adult populations and AD risk factors. We acknowledge that our participants are not diverse in race but are in terms of other sample characteristics such as medical history and gender. We primarily recruited from a convenience sample and approached all potentially eligible individuals based on our inclusion and exclusion criteria noted above.

[0153] In this cross-sectional study, each participant completed two study visits. The first visit included collection of demographics, medical and family history information, and blood samples. Participants were sent home with a fecal collection kit, provided by our institution’s Center for Microbiome Research (CMR). Patients underwent an MRI at the second study visit. Sample characteristics including demographics (age, education, sex), BMI, and parental history of dementia were also collected. MCI and control groups did not statistically significantly differ on these variables.Atty. Dkt. No.650053.01137

[0154] Data Collection (Stool): Participants were provided with in-home fecal collection kits with instructions during their first study visit. They completed a 7-day diet log prior to providing and shipping a stool sample according to standard instructions provided by the CMR. Samples were frozen at -80°C until processed by the CMR. Genomic DNA was extracted using the Qiagen PowerLyzer PowerSoil Kit. Purified genomic DNA was then submitted to the University of Wisconsin-Madison Biotechnology Center for 16S and shotgun metagenomics sequencing.

[0155] 16S rDNA bacterial gene sequencing: 16S rDNA libraries were generated using the V3-V4 region spanning primers 341F to 806R at the University of Wisconsin Madison Biotechnology Center using the Illumina TruSeq DNA library prep system. PCR products were sequenced on the Illumina MiSeq platform using the 2x300-bp protocol. Amplicons were generated using a dual-indexing amplification strategy and optimized for sequencing.

[0156] Whole metagenome sequencing for human samples: Sequences were generated on the Illumina NovaSeqX Plus platform producing 2x150 paired-end reads on an S4-flow cell at the University of Wisconsin-Madison Biotechnology Center. Depth of coverage was ~50M reads per human sample.

[0157] Magnetic Resonance Imaging (MRI): MRI were conducted on the MCW research- dedicated GE Signa Premier 3T scanner using the Nova Medical 32-channel phased-array head coil. Details on the MRI protocol are available in our previously published work 17. In brief, a 3D T1-weighted MPRAGE anatomical image was acquired with TR / echo time (TE) = 2200 / 2.8 ms, field of view (FOV) = 24 cm, matrix size = 512 × 512 × 256, slice thickness= 0.5 mm, voxel size = 0.47 × 0.47 × 0.5 mm, and flip angle (FA) = 8°. Participants completed a breath-holding (BH) fMRI scan, an advanced MBME sequence, and these parameters: TR = 1000 ms, TE = 11,30,49 ms (three echoes), FOV = 24x24 cm, matrix size = 80 × 80, slice thickness = 3 mm (3 × 3 × 3 mm voxel size), 11 slices with multiband factor = 4 (44 total slices), FA = 60°, BW = 250 kHz, echo spacing = 0.51 ms, partial Fourier factor = 0.85, and in-plane acceleration with R = 2. The functional MBME scan lasted 320 s for a total of 320 volumes. The TEs for the MBME scan were set to the minimum possible. Our BH protocol has been described previously by our group (e.g., refs.17,18)— started with 72 sec of paced breathing, then four cycles of 16 sec of BH on expiration, 16 sec of self-paced recovery breathing, 24 sec of paced breathing, and ended with another 24 sec of paced breathing.

[0158] A Hadamard-encoded, multiple delay 3dPCASL was also collected with seven PLDs (1.0, 1.4, 1.7, 2.1, 2.6, 3.1, 3.7s). and a 3.5s labeling block19. A segmented stack of spirals readout with 4 arms of 640 points was used. A raw magnetization (M0) image was collected on the last repetition. Additional imaging parameters included: TR / TE = 7480 / 61 ms, slice thickness = 3Atty. Dkt. No.650053.01137 mm, 48 slices, FOV = 220 mm, and FA = 110°. Individual ATT was computed using the signal- weighted delay method (equation [1])20where ^^i is the ithPLD, ^^ is the transit delay and ∆^^(^^, ^^i) is the difference signal.

[0159] Equation 1: ^^^^^^^^ ൌ ^∑^ ^ୀ^ ^^^∆^^^^^,^^^^^ / ^∑^ ^ୀ^ ∆^^^^^,^^^^^

[0160] An ATT-corrected CBF map was estimated using the equation (equation [2]) where PLD represents the shortest post-labeling delay (1.0s), LD is the total time for labeling (3.5s), T1ais the arterial blood’s longitudinal relaxation (1.6s), T1t is the gray matter’s longitudinal relaxation (1.2s), ^^ is the combined efficiency of labeling and background suppression (0.64), ^^ is thecalculated ATT, Sr is equal to ^^ ^1 െ ^^ିమ భ.మ^ where ^^ is 0.9, Δ^^ is the average of the 7 PLD perfusion-weighted intensity of the reference image21.

[0161] Equation 2: ^^^^^^ ൌ^^^^^^ഃ / ^భೌ^ ௌ^∙∆ெ ଶఌ భ்ೌ^^ష^^౮ ^ು^ವషഃ,బ^ / ^భ^ି^ష^ೌ^ ^^ವశು^ವషഃ,బ^ / ^భ^^ெబ

[0162] CBF, excluded if datadid not display gray and / or no or few negative values. In addition, respiratory traces were collected during the BH scans. All subjects included in analyses had 4 clear BH periods. One MCI subject did not have a BH scan and one MCI subject was excluded due to poor quality of the BH scan. One MCI subject was excluded due to a poor quality CBF scan.

[0163] Clinical Assessments

[0164] Controls: Control participants completed a medical history questionnaire and a brief cognitive battery: verbal memory (Rey Auditory Verbal Learning Test, RAVLT)22, psychomotor processing speed and mental flexibility (Trail Making Test, Parts A and B)23, and language (letter and category fluency)23. Additionally, to further evidence that controls did not experience behavioral changes concerning for dementia, a study partner completed the Quick Dementia Rating System (QDRS), a 10-item questionnaire (scores range from 0 to 30 with higher scores representing greater cognitive and functional impairment)24.

[0165] aMCI Group: Clinical neuropsychological data on participants diagnosed with aMCI were obtained via retrospective electronic medical record review. These data include demographic variables, MCI diagnosis, medical history and medications, psychiatric history, and neuropsychological test scores. Common cognitive testing data elements were identified from clinical neuropsychological exams and entered into the study database. Common data elements were the Trail Making Test (TMT, Parts A and B)23, and letter and category fluency tasks (composite score was created from one of three different versions23,25,26), and verbal memory spontaneous delayed recall (composite score was created from Hopkins Verbal Learning Test-Atty. Dkt. No.650053.01137Revised27or RAVLT22). Two patients had a recent cognitive screening assessment as part of theirvisit with a memory disorders neurologist but did not have recent comprehensive neuropsychological testing. One patient had neuropsychological testing but no TMT results. Therefore, sample size for common cognitive testing data elements for the aMCI group ranged from 11 to 12.

[0166] Data Analysis

[0167] 16S sequencing and analysis: 16S V3-V4 amplicon libraries are generated at the University of Wisconsin Madison Biotechnology Center. PCR products are sequenced on the Illumina MiSeq platform using the 2x300-bp protocol. Paired-end reads are analyzed usingQuantitative Insights into Microbial Ecology (QIIME2) which is updated regularly28andprocessed by us using standard workflows, as described previously29. Taxonomy is assigned to ASVs against the SILVA 138 reference database. Changes in abundance of individual taxa are also analyzed using traditional univariate statistical methods. Linear discriminate analysis effectsize (LEfSe)30is used to determine discriminant ASVs (bacterial taxa) and confirmed via randomforest machine learning analyses using RStudio and MaAsLin231,32. Correlation statistical analyses were performed using “corr.test” function of the psych R package to calculate Spearman’s rank correlation coefficients and p- values33. The ggcorrplot R package, using ggplot2, was utilized to visualize correlation matrices as circles of neurovascular patterns against differentially abundant taxa34,35. Plots were arranged together using the patchwork R package36.

[0168] Whole genome shotgun sequencing and analysis: Sequences are generated on the Illumina NovaSeq platform producing 2x150 paired-end reads on an S4-flow cell at the University of Wisconsin-Madison Biotechnology Center. Raw sequences are quality filtered, assembled to generate contigs, and processed to create metagenomes (MAGs). We utilize Anvi’o for all metagenomics processing and visualization as described previously37. MAGs are manually curated based on a threshold of 50% completion (relative to a reference) or genome size greater than 2Mb and less than 10% redundancy based on bacterial single-copy-core gene collection38. RStudio runs random forest (rf) or MaAsLin231,32packages to identify the top discriminatory MAGs between samples. LEfSe30is also used to determine discriminant MAGs. Spearman rho correlational analysis, as described above for 16S analyses, was conducted to confirm associations between bacterial taxa and clinical data. Comparisons between Turicibacter and Bilophila were conducted using GraphPad Prism v10. Equal abundance subset analysis of Bilophila wadsworthia was conducted by choosing samples whose abundances were approximately the same value (p>.999 via Mann-Whitney U test comparing control (n=5) and aMCI (n=7)). The gene coverages from this subset of samples were run through LEfSe to identifyAtty. Dkt. No.650053.01137 differentially abundant genes between Bilophila associated with controls vs Bilophila associated with aMCI.

[0169] Whole genome shotgun virome analysis: The contigs file generated after quality controlling and assembling the raw sequencing reads, as described above, was used for downstream virome analyses. Viral contigs were identified using Virsorter239and were quality controlled using CheckV40. Viral genes, including auxiliary metabolic genes (AMGs) were functionally annotated using DRAMv41. Viral contigs were manually curated as previously described42to achieve greater confidence in viral identification. As with the bacterial MAGs, high quality viral contigs that were present in at least three samples underwent discriminant analysis using LEfSe, Random Forest, and MaAsLin2. MaAsLin2 was also used to conduct a correlational analysis with clinical variables. Viral contigs underwent Spearman rho correlational analysis as described above to further identify and confirm associations with the cognitive and neurovascular data. Virus hosts were identified using VirHostMatcher-Net43or via the NCBI blastn suite if a contig was unable to be identified using VirHostMatcher-Net. The lytic vs lysogenic state of the viral contigs was determined by running the BAM files (described above) through VIBRANT and running PropagAte45on the VIBRANT-predicted viral contigs to determine viral to bacterial ratio (VBR). As with the 16S and bacterial MAGs, Spearman rho correlational analysis was run comparing the viral contigs with clinical and neurovascular measures of cognitive impairment. Visualization of these correlations was generated the same as is described for 16S and bacterial MAGs.

[0170] Neuroimaging: MRI data were analyzed using AFNI46, FSL47, and Advanced Normalization Tools (ANTS) programs48. Our fMRI preprocessing procedures followed those of Cohen et al.18using the Human Connectome Project preprocessing pipeline49, adapted given the multiple echo data. Our imaging processing procedures and analytic methods were previously published in great detail17. In brief, the anatomical MPRAGE images was coregistered to MNI space using a linear followed by nonlinear registration. After discarding the first eight volumes, first-echo data was volume registered to the first volume using mcflirt in FSL. Echoes 2 and 3 were registered using the transformation matrices from the first echo. Then, multi-echo independent component analysis (ME-ICA50,51) was implemented using tedana version 23.0.1. Finally, the data was registered to the MPRAGE image and then to MNI space and smoothed using a 6mm FWHM Gaussian kernel.

[0171] The BH response was evaluated using a general linear model using phys2cvr, a python-based tool which generates time-shifted regressors to estimate CVR maps and their lag. BH regressor was generated by convolving a square wave, with ones during BH periods andAtty. Dkt. No.650053.01137 zeros otherwise, with the respiration response function (RRF)52. This was then used as the input regressor in phys2cvr. The BH regressor was shifted from −9 to 9s in steps of 1s, and for each voxel, the regressor that resulted in the highest positive t-score was chosen.

[0172] For CBF and ATT, first the M0 image was registered to the MPRAGE image using epi_reg in FSL47. The CBF and ATT maps were then registered to the MPRAGE image using the M0 to MPRAGE transformation matrix. CBF and ATT maps were normalized to MNI standard space using the MPRAGE to MNI transformation matrix computed above. For CBF, ATT, and CVR, a voxelwise t-test was used to compare between aMCI and control participants using 3dttest++ in AFNI with age and sex as covariates. Resulting t-score maps were thresholded at p< .01 and then cluster-corrected using 3dClustSim53in AFNI at ^^ < 0.05. Individual mean CBF,ATT, and CVR values were extracted from regions showing a significant difference between the control and aMCI groups at cluster-corrected p < 0.01 and compared between groups using a Mann-Whitney test. Spearman correlational analyses were conducted between CBF, ATT, and CVR values and cognitive test scores including tests of memory, executive functioning, processing speed, and language, as well as bacterial taxa. A p-value of .05 was considered statistically significant.

[0173] Results

[0174] Cognitive Measures and Neurovascular Data for aMCI and Control Cohorts. Means, SDs, and parametric (independent samples t-tests) and non-parametric (Mann-Whitney test) group comparison (control vs aMCI) results were calculated for cognitive measures, as well as CVR, CBF, and ATT values. The aMCI group had lower mean scores on tests of episodic memory (delayed recall), psychomotor processing speed (Trails A), executive functioning (Trails B), and language (category fluency, not letter fluency) (all p ≤ .01). Subsequent analyses with cognitive measures excluded letter fluency to focus on clinical measures that differentiated aMCI from control groups. Illustrated in FIG.7 are the results of the independent-samples t- tests for control- aMCI comparisons with age, biological sex, and GM density as covariates for CVR, CBF, and ATT. The control group had higher CVR and CBF and lower ATT values compared to aMCI participants at the cluster-wise corrected p-value threshold of .01. This was also evident with Mann Whitney non-parametric group comparisons—CVR (U = 17.00, z = -2.84, p = 0.003) and CBF (U = 24.00, z = -2.54, p = 0.010) values were higher in controls, whereas ATT values were lower in controls (U = 105.00, z = 2.48, p = 0.012). Significant and trend-level correlations emerged between cognitive measures and neurovascular metrics. Most notably, higher executive functioning (TMT-B) scores were positively correlated with CVR and CBF (rs = .527, p =0.020 and rs = .417, p = 0.067, respectively) and negatively associated with ATT (rs = -.357; p = 0.122).Atty. Dkt. No.650053.01137 Similar patterns emerged with other cognitive measures, mainly delayed recall and category fluency scores.

[0175] 16S Sequencing Analysis of aMCI and Control Participants. DNA from stool samples from both aMCI and control cohorts was assessed for changes in GMB.16S sequencing yielded 269 ASVs across both cohorts. None of the tested measures of alpha and beta diversity were significantly different between the two groups (p > 0.05). However, we did identify several differentially abundant bacterial taxa in each cohort (FIGS.8A-8B).

[0176] Several ASVs associated with aMCI participants are members of bacterial species known to be pro-inflammatory or have previously been associated with cognitive impairment, such as Bilophila (identified via all methods of differential abundance) and Faecalibacterium (identified via Random Forest analysis). Bacteria associated with a healthy GMB, including Akkermansia (identified via MaAsLin2 analysis) and Turicibacter (identified via Random Forest analysis) were more abundant in controls (FIGS.8A-8B).

[0177] Relationship between Bacterial Taxa, Neurovascular Patterns, and Cognition. Across all participants with both GMB and neurovascular data, we found a pattern of positive and negative Spearman correlations (significant and non-significant or trend-level). Pro- inflammatory bacterial taxa that were more abundant in aMCI, such as Bilophila, were generally negatively correlated with CVR and CBF values and positively correlated with ATT. Conversely, bacterial taxa associated with a healthy gut, such as Akkermansia, Turicibacter, and Subdoligranulum, were more abundant in controls and were typically positively correlated with CVR and CBF and negatively correlated with ATT. Statistically significant and trend-level findings between 16S and neurovascular metrics for the total sample, controls, and MCI are displayed in FIGS.8C-8D.

[0178] Similarly, we also found a pattern of positive and negative Spearman correlations across all participants with both GMB and cognitive data (FIGS. 8C-8D). Pro-inflammatory bacteria, such as Bilophila, Erysipelotrichaceae UCG-003, and Anaerostignum were associated with worse cognitive functioning, whereas bacteria associated with a healthy gut, including two Christensenellaceae R-7 group bacteria, Subdoligranulum, and Eisenbergiella, were associated with better cognitive functioning. Significant and trend-level correlational findings emerged more frequently with tests of delayed recall (memory) and category fluency (language). For 16S results, 9 of the 17 bacterial taxa that were associated with cognitive functioning (significant to trend-level) also showed significant or trend-level significant correlations with neurovascularmetrics. For instance,Subdoligranulum g. was positively correlated with TMT-B (rs= 0.476, p= 0.029) and category fluency (rs= 0.443, p = 0.039), as well as CVR (rs= 0.483, p = 0.023) andAtty. Dkt. No.650053.01137 CBF (rs= 0.408, p = 0.053), and was slightly negatively correlated with ATT (rs= -0.197, p = 0.367). Eubacterium g. was positively correlated with delayed recall (rs= .401, p = 0.065), category fluency (r = 0.519, p = 0.013), and CBF (rs= 0.447, p = 0.033) and was negatively correlated with ATT (rs= -0.524, p = 0.010). Tyzzerella sp., which positively correlated with all cognitive measures (trend level to significant correlational values) also positively correlated with CVR (rs= 0.451, p = 0.035) and CBF (rs= 0.409, p = 0.053), and negatively correlated with ATT(rs= -0.248, p = 0.254).Human gut-associated Parabacteroides was negatively correlated withall cognitive measures (ps < .05) as wellas CVR (rs= -0.562, p = 0.006). Blautia glucerasea wassignificantly negatively correlated with all cognitive measures (ps < .05) along with CVR (rs= - 0.451, p = 0.035) and CBF (rs= -0.321, p = 0.135), and positively correlated with ATT (rs=0.227, p = 0.299).

[0179] Shotgun Metagenomics Analysis of aMCI and Control Participants. 16S rDNA profiling accounts for taxonomic identification but cannot accurately identify functionality which is widely variable between strains of the same species. Thus, we conducted shotgun metagenomics sequencing to identify discriminatory bacterial functions associated with each cohort. After quality controlling the contigs binned with metaBAT2, we ended up with 243 MAGs with an average N50 of 42,341 base pairs (range: 5 kb – 389 kb). Many of the MAGs appeared to be the species-level resolution of taxa identified in the 16S dataset (FIGS.8A-8D). We used both Random Forest (with Boruta) and MaAsLin2 (see Methods) to identify discriminatory MAGs, which included Parabacteroides diastonis, Bilophila wadsworthia, Turicibacter sanguinis, Christensenellales CAG-74 UMGS1633, and CAG-349 (another member of the Christensenellales order). We also observed an enrichment for Alistipes indistinctus, Faecalibacterium prausnitzii, and Anaerostipes in aMCI and an enrichment for Oscillospiraceae,CAG-1031 (a potential probiotic member of the Muribaculaceae family54and Intestimonasmassiliensis in controls. A total of 22 differentially abundant bacterial MAGs were identified (FIGS. 9A-9B), including several with significant and trend-level correlations with the neurovascular and cognitive tests, as detailed below and in FIGS.9C-9D. Bilophila wadsworthia, which has previously been shown to be pro- inflammatory55, was significantly increased in aMCI (p<0.05), whereas Turicibacter sanguinis, which has been shown to be negatively correlated with the inflammatory markers IL-1β and IL-656, was enriched in the controls (FIGS. 10A-10B). Another species of Turicibacter - Turicibacter sp001543345 - was enriched in the controls as well. It is worth noting that one control participant was taking fluoxetine at the time of samplecollection; fluoxetine is known to deplete Turicibacter sanguinis57leading us to exclude thatsample during analysis of Turicibacter abundance. The subsequent analysis of 9 controls and 14Atty. Dkt. No.650053.01137 aMCI subjects produced statistically significant differences for Turicibacter abundance (Turicibacter sanguinis and Turicibacter sp001543345 combined) between the two groups (p < 0.05), consistent with the 16S analysis (FIGS.8A-8B). All DNA sequencing data generated from these studies have been made available in the National Library of Medicine, National Center for Biotechnology Information, Sequence Read Archive as BioProject ID PRJNA1161622.

[0180] Identification of Two Strains of Bilophila wadsworthia. MAGs attributed to B. wadsworthia were identified in both the aMCI cohort and controls but appeared to comprise two strains with extensive single nucleotide variants (SNVs) across the genome (FIG. 10C). To confirm that extensive SNVs correspond to unique gene composition and functionality, Linear Discriminant Analysis Effect Size (LEfSe) was conducted on a subset of participants with equal abundances of B. wadsworthia (n=5 controls, n=7 aMCI, p>0.999). This analysis yielded 77 unique genes differentially abundant in aMCI and 58 unique genes found in controls. These data support the conclusion that two unique strains of B. wadsworthia with unique metabolic properties are present in each cohort.

[0181] Relationship between select Bacterial Metagenomes, Neurovascular Patterns, and Cognition. Trend- level and significant correlations emerged between bacterial species and neurovascular and cognitive data (FIGS.9C-9D). Most notably, Alistipes indistinctus was moreabundant in aMCI and was negatively correlatedwith CVR (rs= -0.437, p = 0.042) and CBF(rs= -0.546, p = 0.007). In addition, A. indistinctus was significantly correlated with TMT-B (rs= -0.587, p = 0.005) and category fluency (rs= -0.422, p = 0.050). B. wadsworthia sp. was negatively associated (trend-level) with CVR (rs= -0.389, p = 0.074) and CBF (rs= - 0.402, p = 0.057), and statistically significantly correlated with cognition, including TMT-A (rs= -0.444, p = 0.044), Trails B (rs= -0.499, p = 0.022), and category fluency (rs= -0.503, p = 0.017). Anaerostipes was found to be more abundant in aMCI and was negatively correlated with CBF (rs= -0.499, p = 0.015) but not with other MRI or cognitive metrics. Turicibacter sp. wassignificantly positively correlated with CBF (rs= 0.423, p= 0.050), but no other significantcorrelations emerged with neurovascular or cognitive measures. Analyses with Turicibacter species excluded the participant who was on fluoxetine at the time of sample collection.

[0182] Viral Metagenome Analysis of aMCI and Control Cohorts. The human gut harborsbetween 108and 1010viral-like particles per gram of feces58. Thus, the shotgun metagenomicssequencing data was also analyzed using standard tools (see Methods) for differentially abundant viruses. We identified 2387 high quality viral contigs, the vast majority of which were identified to be bacteriophages LefSe analysis revealed that 36 phage contigs were differentially abundant between the cohorts, while MaAsLin2 revealed 74 differentially abundant phage contigs. 30Atty. Dkt. No.650053.01137 phage contigs were identified by both methods. An additional 9 contigs that were identified via LefSe had trend-level significance (p<0.1) in MaAsLin2 to differentiate the cohorts. Furthermore, there were 8 phage contigs only present in the control cohort and 28 phage contigs only present in aMCI participants (FIG. 11A). The data were used to predict phage-encoded metabolic pathways, identified using DRAMv.

[0183] Multiple metabolic functions were found to be differentially abundant between the two cohorts, including ATP and ADP synthesis from inosine monophosphate, propanoyl-CoA metabolism, siroheme biosynthesis, and methanogenesis (FIG. 11C). Lastly, because phages display specific lifestyles including lysis and lysogeny, we assessed the viral to bacterial ratio (VBR) to predict viral lifestyle, as described previously (see methods)45. We found that viruses associated with aMCI were significantly more likely to be lysogenic than lytic compared to the controls (FIG.11D, p<0.05).

[0184] Relationship between Select Viral Contigs, Neurovascular Patterns, and Cognition. Finally, viral contig abundance was significantly correlated with the neurovascular and cognitive data. We found that several phage contigs were significantly associated with clinical and / or neurovascular measures, including phages that are associated with Bacteroides ovatus, Roseburia intestinalis, and E. coli (FIGS. 12A-12D). Significant positive and negative correlations were revealed between several phage contigs, CVR, CBF, ATT, TMT A and B, delayed recall, and category fluency. It is worth noting that one Bilophila wadsworthia phagewas identified as enrichedin aMCI and was significantly associated with delayed recall (rs= -0.589, p = 0.004) and TMT-B (rs= -0.491, p= 0.033). This phage also was significantly associated with CVR (rs= -0.474, p=0.026) and had a trend level association with ATT (rs=0.352, p = 0.099) and contained an acyl-CoA synthetase capable of influencingcentralmetabolism (12A-12D). Additionally, a phage associated with Acinetobacter baumanii ATCC17978 thatwas enriched in controls was significantly associated with nearly all tested metrics,including CVR (rs=0.543, p=0.009), CBF (rs= 0.599, p = 0.0025), ATT (rs= -0.424, p = 0.044),delayed recall (rs= 0.427 , p= 0.048), and TMT-B (rs= 0.467, p= 0.033).

[0185] Discussion

[0186] After decades of research, Alzheimer’s disease remains a devastating and costly neurodegenerative condition without an effective treatment. Increasingly, there is promising evidence that the GMB may change our understanding of AD and improve early diagnosis and treatment. Examination of the gut-brain axis in AD is critically needed to understand the direct and indirect pathways that link gut microbial species to the central nervous system. Neurovascular changes can occur decades before the onset of AD symptomatology and disruption ofAtty. Dkt. No.650053.01137 neurovascular structures allows the influx of pathogens into the brain12-14. However, there is no known study that has yet examined connections between the GMB and neurovascular changes in prodromal AD. Therefore, we evaluated the associations between the GMB, neurovascular functioning as measured by MRI, and cognition in aMCI and control cohorts. To the best of our knowledge, this is the first study that examines the relationship between the GMB, including both bacteria and viruses, and neurovascular changes in aMCI.

[0187] Consistent with previous studies examining neurovascular changes in AD (e.g., ref.17), we found that aMCI participants had lower CVR and CBF and longer ATT compared to the control group (FIG. 7) and that CVR and CBF were associated with cognition. Similar to previously published results59, 16S sequencing analysis revealed an increase in proinflammatory bacterial taxa in the aMCI cohort, such as Bilophila and Erysipelotrichaceae UCG-003, and an enrichment for probiotic ASVs in the controls, including Akkermansia, Subdoligranulum, and Turicibacter. We found that several of the enriched ASVs for each cohort had many significant (p<0.05) and trend-level (p<0.1) correlations with the cognitive and neurovascular data. More specifically, ASVs enriched in aMCI had negative correlations with CVR and CBF and positive correlations with ATT. The opposite trend was true for ASVs enriched in controls, in which abundance was positively correlated with CVR and CBF and negatively correlated with ATT. Further, inflammation has been implicated in neurovascular changes associated with AD60, suggesting a possible biological relationship between bacterial presence and neurovascular function.

[0188] We utilized shotgun metagenomics sequencing to identify specific bacterial species and their associated metabolic pathways that may be enriched in our two cohorts. Similar to the 16S data, we identified several pro- inflammatory bacterial metagenomes enriched in aMCI and several health-promoting MAGs enriched in the controls. Notably, we saw an enrichment for Alistipes indistinctus in aMCI as well as significant correlations between this bacterium and measures of both neurovascular and cognitive function. Interestingly, A. indistinctus has been shown to deplete intestinal urate levels61, which is a risk factor for AD62. Bilophila wadsworthia was also enriched in aMCI and significantly correlated with multiple cognitive metrics. B. wadsworthia has been shown to deplete host urate levels via the breakdown of taurine (from bile acids) into hydrogen sulfide, which interferes with the production of urate63. In contrast, Turicibacter, which was enriched in controls and is associated with an increase in intestinal taurine levels, has a positive correlation with intestinal urate levels63.

[0189] Further, both Bilophila and Turicibacter are directly implicated in bile acid metabolism and altered bile acid profiles have been associated with neurodegeneration, including in MCI andAtty. Dkt. No.650053.01137AD64. For instance, Nho et al.64 demonstrated bile acid signatures that were associated with CSFAβ1-42, p-tau 181, t-tau, glucose metabolism,and atrophy by combining neuroimagingtechniques with targeted metabolomics. We did not see statistically significant correlations between neurovascular patterns and Bilophila and Turicibacter levels using our metagenomics data despite there being a significant association between Bilophila and CVR and between Turicibacter and CBF in our 16S data. The reason for this discrepancy could be due to loss of lower quality Bilophila and Turicibacter sequences, resulting in the formation of MAGs that do not fully represent those bacterial taxa. Both Turicibacter MAGs were less than 70% complete; thus, if those MAGs were complete, there may have been less of a divergence between the 16S and shotgun sequencing data. However, trend-level associations between Bilophila and CVR and Turicibacter and CBF were still identified, indicating that significance may be reached with a larger sample size than we had available for this study. Additionally, Bilophila was significantly associated with TMT A and B, as well as category fluency, further demonstrating its relationship to cognitive functioning. Several of the other differentially abundant bacterial MAGs that were enriched in aMCI seem to have conflicting findings within the current literature. For example, Vogt et al.1saw similar trends to us in which Blautia and Alistipes were associated with cognitiveimpairment. Li et al.65also saw an increase in Blautia in MCI, but unlike us, they observed adecrease and Alistipes and Parabacteroides. Further, Liu et al6observed a decrease in Blautia inaMCI patients. There are several possible reasons for discrepancies between all these studies. For one, there are cultural and environmental differences in patient populations (e.g., participants from China vs. United States), leading to potential differences in GMB results.

[0190] Another potential explanation is differences in identified bacterial strain. There are several species within each genus of bacteria and several strains within each species; thus, while one Blautia species may be pro- inflammatory, another may be anti-inflammatory or neutral, depending on the differences in metabolism between the species or strains. Another example of this is Parabacteroides diastonis, which was increased in our aMCI cohort. In one study, P.diastonis was positively associated with post-operative delirium66and in another, it had a negativeassociation with neurocognitive development67. However, it seems that P. diastonis has an ambivalent association with inflammation, as it has been shown to be both probiotic and anti- inflammatory68, depending on the context.

[0191] Finally, we assessed differences in the gut virome between controls and aMCI. It has been demonstrated that previous viral infections with known pathogens, such as HSV-1, influenza, SARS-CoV-2, and viral encephalitis, increase the risk of developing AD69-71. Another study examined non-pathogenic gut viruses (primarily bacteriophages) as possible contributors toAtty. Dkt. No.650053.01137 AD cognitive decline as assessed with a cognitive screening instrument9, but no studies have examined the relationship between gut viruses, sensitive neuropsychological tests, and neuroimaging. We identified differentially abundant phages, which may have the ability to metabolically reprogram their host bacterial physiology, thereby shifting the overall metabolic profile in the gut72. Phage-associated metabolic pathways are known to contribute to secondary metabolite production and bile acid metabolism73-74, ultimately changing the circulating metabolite profile within the host. Phages associated with aMCI were also more likely to be capable of lysogeny (integrating into the bacterial genome; FIGS.11A-11D), allowing for stable transmission within bacterial cells over time. Bacteria-phage stability could affect both protective or pathogenic features in either control or MCI cohorts. However, Johansen et al.75demonstrated that the gut virome (phageome) trends towards a more lytic state as people age. Lytic phages lead to a decrease in bacterial populations overall, conferring an element of passive immunity. Thus, our results suggest that increased lysogenic phages in the aMCI cohort represents a disruption in the normal aging GMB and may represent a loss of passive immunity to specific pathogenic bacteria capable of influencing neurodegeneration. Overall, our results suggest that disruption of the gut microbiota, including phage dynamics, alters the metabolic functionality of the gut ecosystem in aMCI relative to controls. Metabolic features of the microbiota may affect pro- inflammatory metabolites and circulating bile acids that, when combined with age- related BBB permeability changes, lead to accelerated neurovascular changes and neurodegeneration.

[0192] However, it is important to note limitations in this study. For one, due to cohort composition, we were unable to determine a potential role of race and ethnicity, medical history, and environmental factors on the aMCI-associated GMB and associated neurovascular dysfunction. We also acknowledge the small sample size we currently have for this study which may also contribute to the number of trend-level, rather than significant, correlations identified between the gut microbiota, neurovascular dysfunction, and cognition. We are repeating this study with additional participants to further validate the results shown here, along with the addition of other critical data elements (e.g., bacterial metabolites). It will also be important to conduct a similar study but in a longitudinal manner to assess changes to the GMB and neurovascular function over time. Further, we plan to assess blood (plasma) for AD biomarkers, including Aβ and p-tau and determine whether significant associations exist between these biomarkers, neurovascular changes, and dysbiosis. Lastly, we plan to obtain metabolomic data on production for SCFAs and bile acids for aMCI and control cohorts. Through active collaborations with community and academic institutional partners to conduct community engaged participatory research, we will work to improve research equity in our future gut-brain axis scientificAtty. Dkt. No.650053.01137 endeavors. Future studies will likely include a rodent model to test whether specific bacteria, phages, or metabolites can drive neurodegeneration in vivo. This study provides a connection between gut dysbiosis, cognitive functioning, and neurovascular changes associated with aMCI, paving the way for future studies to identify novel biomarker testing and therapeutic targets for neurodegeneration, cognitive decline, and AD.

[0193] References

[0194] 1. Vogt NM, et al. Gut microbiome alterations in Alzheimer's disease. Sci Rep. Oct 19 2017;7(1):13537. doi:10.1038 / s41598-017-13601-y

[0195] 2. Ferreiro AL, et al. Gut microbiome composition may be an indicator of preclinical Alzheimer's disease. Sci Transl Med. Jun 14 2023;15(700):eabo2984. doi:10.1126 / scitranslmed.abo2984

[0196] 3. Chandra S, et al. The gut microbiome in Alzheimer's disease: what we know and what remains to be explored. Mol Neurodegener. Feb 12023;18(1):9. doi:10.1186 / s13024-023-00595- 7

[0197] 4. Liu P, et al. Altered microbiomes distinguish Alzheimer's disease from amnestic mild cognitive impairment and health in a Chinese cohort. Brain Behav Immun. Aug 2019;80:633-643. doi:10.1016 / j.bbi.2019.05.008

[0198] 5. Nagpal R, et al. Modified Mediterranean-ketogenic diet modulates gut microbiome and short-chain fatty acids in association with Alzheimer's disease markers in subjects with mild cognitive impairment. EBioMedicine. Sep 2019;47:529-542. doi:10.1016 / j.ebiom.2019.08.032

[0199] 6. Liu P, et al. Gut microbiota interacts with intrinsic brain activity of patients with amnestic mild cognitive impairment. CNS Neurosci Ther. Feb 2021;27(2):163-173. doi:10.1111 / cns.13451

[0200] 7. Yamashiro K, et al. Free water in gray matter linked to gut microbiota changes with decreased butyrate producers in Alzheimer's disease and mild cognitive impairment. Neurobiol Dis. Apr 2024;193:106464. doi:10.1016 / j.nbd.2024.106464

[0201] 8. Montoro RA, et al. Structural and functional neuroimaging of the effects of the gut microbiome. Eur Radiol. Jun 2022;32(6):3683-3692. doi:10.1007 / s00330-021-08486-5

[0202] 9. Chaudhari DS, et al. Unique trans-kingdom microbiome structural and functional signatures predict cognitive decline in older adults. Geroscience. Oct 2023;45(5):2819-2834. doi:10.1007 / s11357-023-00799-1

[0203] 10. Sweeney MD, et al. Vascular dysfunction-The disregarded partner of Alzheimer's disease. Alzheimers Dement. Jan 2019;15(1):158-167. doi:10.1016 / j.jalz.2018.07.222

[0204] 11. Sweeney MD, et al. Blood-brain barrier breakdown in Alzheimer disease and otherAtty. Dkt. No.650053.01137 neurodegenerative disorders. Nature reviews Neurology. Mar 2018;14(3):133-150. doi:10.1038 / nrneurol.2017.188

[0205] 12. Bowman GL, et al. Blood-brain barrier breakdown, neuroinflammation, and cognitive decline in older adults. Alzheimers Dement. Dec 2018;14(12):1640-1650. doi:10.1016 / j.jalz.2018.06.2857

[0206] 13. Lee RL, Funk KE. Imaging blood-brain barrier disruption in neuroinflammation and Alzheimer's disease. Front Aging Neurosci.2023;15:1144036. doi:10.3389 / fnagi.2023.1144036

[0207] 14. Asby D, et al. Systemic infection exacerbates cerebrovascular dysfunction in Alzheimer's disease. Brain. Jul 282021;144(6):1869-1883. doi:10.1093 / brain / awab094

[0208] 15. Klohs J. An Integrated View on Vascular Dysfunction in Alzheimer's Disease. Neuro- degenerative diseases.2019;19(3-4):109-127. doi:10.1159 / 000505625

[0209] 16. Zlokovic BV. Neurovascular pathways to neurodegeneration in Alzheimer's disease and other disorders. Nature reviews Neuroscience. Nov 3 2011;12(12):723-38. doi:10.1038 / nrn3114

[0210] 17. Glass Umfleet L, et al. Decreased Cerebrovascular Reactivity in Mild Cognitive Impairment Phenotypes. J Alzheimers Dis.2023;94(4):1503-1513. doi:10.3233 / JAD-221156

[0211] 18. Cohen AD, et al. Improving the Breath-Holding CVR Measurement Using the Multiband Multi-Echo EPI Sequence. Front Physiol. 2021;12:619714. doi:10.3389 / fphys.2021.619714

[0212] 19. Cohen AD, et al. Longitudinal Reproducibility of MR Perfusion Using 3D Pseudocontinuous Arterial Spin Labeling With Hadamard-Encoded Multiple Postlabeling Delays. J Magn Reson Imaging. Jun 2020;51(6):1846-1853. doi:10.1002 / jmri.27007

[0213] 20. Dai W, et al. Reduced resolution transit delay prescan for quantitative continuous arterial spin labeling perfusion imaging. Magn Reson Med. May 2012;67(5):1252-65. doi:10.1002 / mrm.23103

[0214] 21. van der Thiel M, et al. Brain Perfusion Measurements Using Multidelay Arterial Spin-Labeling Are Systematically Biased by the Number of Delays. AJNR Am J Neuroradiol. Aug 2018;39(8):1432-1438. doi:10.3174 / ajnr.A5717

[0215] 22. Geffen G, et al. Performance measures of 16–to 86-year-old males and females on the auditory verbal learning test. Clinical Neuropsychologist.1990;4(1):45-63.

[0216] 23. Reitan RM, Wolfson D. The Halstead-Reitan neuropsychological test battery: Theory and clinical interpretation. vol 4. Reitan Neuropsychology; 1985.

[0217] 24. Galvin JE. The Quick Dementia Rating System (Qdrs): A Rapid Dementia Staging Tool. Alzheimers Dement (Amst). Jun 12015;1(2):249-259. doi:10.1016 / j.dadm.2015.03.003Atty. Dkt. No.650053.01137

[0218] 25. Sue Baron I. Delis-Kaplan executive function system. Child Neuropsychology. 2004;10(2):147-152.

[0219] 26. Ivnik RJ, et al. Neuropsychological tests' norms above age 55: COWAT, BNT, MAE token, WRAT-R reading, AMNART, STROOP, TMT, and JLO. The Clinical Neuropsychologist. 1996;10(3):262-278.

[0220] 27. Brandt J, Benedict RH. Hopkins verbal learning test--revised: professional manual. Psychological Assessment Resources; 2001.

[0221] 28. Bolyen E, et al. Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2. Nat Biotechnol. Aug 2019;37(8):852-857. doi:10.1038 / s41587-019- 0209-9

[0222] 29. Fernando DG, et al. Correction: A single, peri- operative antibiotic can persistently alter the post-operative gut microbiome after Roux-en-Y gastric bypass. Surg Endosc. Feb 2023;37(2):1614. doi:10.1007 / s00464-022-09604-0

[0223] 30. Segata N, Huttenhower C. Toward an efficient method of identifying core genes for evolutionary and functional microbial phylogenies. PLoS One. 2011;6(9):e24704. doi:10.1371 / journal.pone.0024704

[0224] 31. Liaw A, & Wiener, M. Classification and regression by random forest. R News. 2002;2:18-22.

[0225] 32. Mallick H, et al. Multivariable association discovery in population-scale meta- omics studies. PLoS Comput Biol. Nov 2021;17(11):e1009442. doi:10.1371 / journal.pcbi.1009442

[0226] 33. Revelle W. psych: Procedures for Psychological, Psychometric, and Personality Research. R package version 2.4.6. Northwestern University.

[0227] 34. Kassambara A. ggcorrplot: Visualization of a Correlation Matrix using 'ggplot2'. R package version 0.1.4.999.

[0228] 35. Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag.

[0229] 36. Pedersen T. Patchwork: The Composer of Plots. R package version 1.2.0.9000.

[0230] 37. Eren AM, et al. Community-led, integrated, reproducible multi-omics with anvi’o. Nature Microbiology.2021 / 01 / 012021;6(1):3-6. doi:10.1038 / s41564-020-00834-3

[0231] 38. Campbell JH, et al. UGA is an additional glycine codon in uncultured SR1 bacteria from the human microbiota. Proc Natl Acad Sci U S A. Apr 2 2013;110(14):5540-5. doi:10.1073 / pnas.1303090110

[0232] 39. Guo J, et al. VirSorter2: a multi-classifier, expert-guided approach to detect diverse DNA and RNA viruses. Microbiome. Feb 12021;9(1):37. doi:10.1186 / s40168-020-00990-y

[0233] 40. Nayfach S, et al. CheckV assesses the quality and completeness of metagenome-Atty. Dkt. No.650053.01137 assembled viral genomes. Nat Biotechnol. May 2021;39(5):578-585. doi:10.1038 / s41587-020- 00774-7

[0234] 41. Shaffer M, et al. DRAM for distilling microbial metabolism to automate the curation of microbiome function. Nucleic Acids Res. Sep 18 2020;48(16):8883-8900. doi:10.1093 / nar / gkaa621

[0235] 42. Viral sequence identification SOP with VirSorter2 V.3 protocols.io; 2021.

[0236] 43. Wang W, et al. A network-based integrated framework for predicting virus-prokaryote interactions. NAR Genom Bioinform. Jun 2020;2(2):lqaa044. doi:10.1093 / nargab / lqaa044

[0237] 44. Kieft K, et al. VIBRANT: automated recovery, annotation and curation of microbial viruses, and evaluation of viral community function from genomic sequences. Microbiome. Jun 102020;8(1):90. doi:10.1186 / s40168-020-00867-0

[0238] 45. Kieft K, Anantharaman K. Deciphering Active Prophages from Metagenomes. mSystems. Apr 262022;7(2):e0008422. doi:10.1128 / msystems.00084-22

[0239] 46. Cox RW. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput Biomed Res. Jun 1996;29(3):162-73. doi:10.1006 / cbmr.1996.0014

[0240] 47. Jenkinson M, et al. Fsl. Neuroimage. Aug 15 2012;62(2):782-90. doi:10.1016 / j.neuroimage.2011.09.015

[0241] 48. Avants BB, et al. Advanced normalization tools (ANTS). Insight j.2009;2(365):1-35.

[0242] 49. Glasser MF, et al. The minimal preprocessing pipelines for the Human Connectome Project. NeuroImage.10 / 15 / 2013;80(0):105-124. doi:10.1016 / j.neuroimage.2013.04.127

[0243] 50. DuPre E, et al. TE-dependent analysis of multi-echo fMRI with tedana. Journal of Open Source Software.2021;6(66)doi:10.21105 / joss.03669

[0244] 51. Kundu P, et al. Integrated strategy for improving functional connectivity mapping using multiecho fMRI. Proc Natl Acad Sci U S A. Oct 1 2013;110(40):16187-92. doi:10.1073 / pnas.1301725110

[0245] 52. Birn RM, et al. The respiration response function: the temporal dynamics of fMRI signal fluctuations related to changes in respiration. Neuroimage. Apr 1 2008;40(2):644-654. doi:10.1016 / j.neuroimage.2007.11.059

[0246] 53. Cox RW, et al. fMRI clustering and false-positive rates. Proc Natl Acad Sci U S A. Apr 252017;114(17):E3370-E3371. doi:10.1073 / pnas.1614961114

[0247] 54. Du J, et al. Spinal Cord Injury Changes the Structure and Functional Potential of Gut Bacterial and Viral Communities. mSystems. May 112021;6(3)doi:10.1128 / mSystems.01356-20

[0248] 55. Feng Z, et al. A human stool-derived Bilophila wadsworthia strain caused systemicAtty. Dkt. No.650053.01137 inflammation in specific-pathogen-free mice. Gut Pathog. 2017;9:59. doi:10.1186 / s13099-017- 0208-7

[0249] 56. Liu P, et al. Gut Microbiome Composition Linked to Inflammatory Factors and Cognitive Functions in First-Episode, Drug-Naive Major Depressive Disorder Patients. Front Neurosci.2021;15:800764. doi:10.3389 / fnins.2021.800764

[0250] 57. Fung TC, et al. Intestinal serotonin and fluoxetine exposure modulate bacterial colonization in the gut. Nat Microbiol. Dec 2019;4(12):2064-2073. doi:10.1038 / s41564-019- 0540-4

[0251] 58. Zuppi M, et al. Phages in the Gut Ecosystem. Front Cell Infect Microbiol. 2021;11:822562. doi:10.3389 / fcimb.2021.822562

[0252] 59. Catumbela CSG, et al. Clinical evidence of human pathogens implicated in Alzheimer's disease pathology and the therapeutic efficacy of antimicrobials: an overview. Transl Neurodegener. Jul 262023;12(1):37. doi:10.1186 / s40035-023-00369-7

[0253] 60. Meszaros A, et al. Neurovascular Inflammaging in Health and Disease. Cells. Jul 4 2020;9(7)doi:10.3390 / cells9071614

[0254] 61. Xu YX, et al. Alistipes indistinctus-derived hippuric acid promotes intestinal urate excretion to alleviate hyperuricemia. Cell Host Microbe. Mar 13 2024;32(3):366-381 e9. doi:10.1016 / j.chom.2024.02.001

[0255] 62. Qiao M, et al. The Influence of Serum Uric Acid Level on Alzheimer's Disease: A Narrative Review. Biomed Res Int.2021;2021:5525710. doi:10.1155 / 2021 / 5525710

[0256] 63. Murros KE. Hydrogen Sulfide Produced by Gut Bacteria May Induce Parkinson's Disease. Cells. Mar 122022;11(6)doi:10.3390 / cells11060978

[0257] 64. Nho K, et al. Altered bile acid profile in mild cognitive impairment and Alzheimer's disease: Relationship to neuroimaging and CSF biomarkers. Alzheimers Dement. Feb 2019;15(2):232-244. doi:10.1016 / j.jalz.2018.08.012

[0258] 65. Li B, et al. Mild cognitive impairment has similar alterations as Alzheimer's disease in gut microbiota. Alzheimers Dement. Oct 2019;15(10):1357-1366. doi:10.1016 / j.jalz.2019.07.002

[0259] 66. Zhang Y, et al. The association between gut microbiota and postoperative delirium in patients. Transl Psychiatry. May 92023;13(1):156. doi:10.1038 / s41398-023-02450-1

[0260] 67. Noble EE, et al. Gut microbial taxa elevated by dietary sugar disrupt memory function. Transl Psychiatry. Mar 312021;11(1):194. doi:10.1038 / s41398-021-01309-7

[0261] 68. Ezeji JC, et al. Parabacteroides distasonis: intriguing aerotolerant gut anaerobe with emerging antimicrobial resistance and pathogenic and probiotic roles in human health. GutAtty. Dkt. No.650053.01137 Microbes. Jan-Dec 2021;13(1):1922241. doi:10.1080 / 19490976.2021.1922241

[0262] 69. Levine KS, et al. Virus exposure and neurodegenerative disease risk across national biobanks. Neuron. Apr 52023;111(7):1086-1093 e2. doi:10.1016 / j.neuron.2022.12.029

[0263] 70. Wozniak MA, et al. Herpes simplex virus infection causes cellular beta- amyloid accumulation and secretase upregulation. Neurosci Lett. Dec 18 2007;429(2-3):95-100. doi:10.1016 / j.neulet.2007.09.077

[0264] 71. Liu N, et al. The viral hypothesis in Alzheimer's disease: SARS-CoV-2 on the cusp. Front Aging Neurosci.2023;15:1129640. doi:10.3389 / fnagi.2023.1129640

[0265] 72. Alkhalil SS. The role of bacteriophages in shaping bacterial composition and diversity in the human gut. Front Microbiol.2023;14:1232413. doi:10.3389 / fmicb.2023.1232413

[0266] 73. Campbell DE, et al. Infection with Bacteroides Phage BV01 Alters the Host Transcriptome and Bile Acid Metabolism in a Common Human Gut Microbe. Cell Rep. Sep 15 2020;32(11):108142. doi:10.1016 / j.celrep.2020.108142

[0267] 74. Dragos A, et al. Phages carry interbacterial weapons encoded by biosynthetic gene clusters. Curr Biol. Aug 232021;31(16):3479-3489 e5. doi:10.1016 / j.cub.2021.05.046

[0268] 75. Johansen J, et al. Centenarians have a diverse gut virome with the potential to modulate metabolism and promote healthy lifespan. Nat Microbiol. Jun 2023;8(6):1064-1078. doi:10.1038 / s41564-023-01370-6

Claims

Atty. Dkt. No.650053.01137 CLAIMS What is claimed:

1. A method of sample processing, comprising: (a) obtaining a sample from a subject having or suspected of having neurodegeneration; (b) determining an amount of one or more taxa in the sample from the subject, wherein the one or more taxa are selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis; and (c) comparing the amount determined in (b) to a control amount.

2. The method of claim 1, wherein the determining comprises isolating bacterial nucleic acids from the sample and sequencing the isolated bacterial nucleic acids.

3. The method of claim 2, wherein the sequencing comprises 16S rDNA bacterial gene sequencing and / or wherein the sequencing comprises: (i) contacting the isolated bacterial nucleic acids with one or more sets of primers, each set specific to a target sequence associated with one taxon of the one or more taxa; (ii) amplifying the isolated bacterial nucleic acids to generate amplicons; and (iii) sequencing the amplicons to generate sequencing reads.

4. The method of claim 3, wherein the determining further comprises quantitatively generating 16S amplicons.

5. The method of claim 2, the sequencing comprises whole genome shotgun sequencing of bacterial DNA.

6. The method of claim 1, wherein the determining comprises detecting one or more taxonomic marker.

7. The method of claim 6, wherein the one or more taxonomic marker comprises an enzyme and / or a metabolite.Atty. Dkt. No.650053.01137 8. The method of claim 1, wherein the determining comprises determining an amount of one or more taxa selected from the group consisting of: Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, NLAE zl H528, and Turicibacter sanguinis.

9. The method of claim 8, wherein the amount determined in (b) is indicative of neurodegeneration when the amount of Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control amount.

10. The method of claim 9, wherein the amount of Bilophila wadsworthia in the sample is about 2 times greater, or about 3 times greater, relative to the control amount.

11. The method of claim 8, wherein the amount determined in (b) is indicative of neurodegeneration when the amount of Turicibacter sanguinis is decreased in the sample, relative to the control amount.

12. The method of claim 11, wherein the amount of Turicibacter sanguinis in the sample is about 3 times less, or about four times less, relative to the control amount.

13. The method of claim 8, wherein a ratio of (i) the amount of Bilophila wadsworthia in the sample, relative to the control amount of Bilophila wadsworthia, to (ii) the amount of Turicibacter sanguinis in the sample, relative to the control amount of Turicibacter sanguinis, is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:

1.

14. The method of claim 1, wherein the sample is a stool sample, a saliva sample, a blood sample, and / or a cerebrospinal fluid (CSF) sample.

15. The method of claim 14, wherein the sample is a stool sample and / or a saliva sample.

16. The method of claim 1, further comprising determining that the subject has mild cognitive impairment based on the comparing in (c).

17. The method of any one of claim 1, further comprising determining that the subject has Alzheimer’s disease (AD) based on the comparing in (c).

18. The method of claim 17, further comprising determining if the subject is amyloid-β positive (Aβ+) or Amyloid-β negative (Aβ-).Atty. Dkt. No.650053.01137 19. The method of claim 1, wherein the control amount is determined using a control sample obtained from (i) a control subject not having neurodegeneration, or (ii) the subject at an earlier timepoint.

20. A method of treatment, comprising: (a) determining an amount of one or more taxa in a sample obtained from a subject having or suspected of having neurodegeneration, wherein the one or more taxa are selected from the group consisting of: Erysipelotrichaceae_UCG_003, Burkholderiales, Sutterellaceae, Eubacterium hallii, Bacteroides caccae, Sutterella, Tannerella, Tanneralla forsythia, Bilophila wadsworthia, NLAE zl H528, Anaerotignum lactatifermentans, Blautia glucasera, Anaerostignum, Ruminococcaceae, Micrococcaceae, Micrococcales, Oscillospirales, Tyzzerella, Collidextribacter massiliensis, Akkermansia, Ruminococcus, and Turicibacter sanguinis; (b) comparing the amount determined in (a) to a control amount; and (c) administering a treatment for neurodegeneration, based on the comparing in (b), optionally wherein the treatment comprises administering an agent to modulate the amount of the one or more taxa in the subject.

21. The method of claim 20, wherein the treatment is administered when the amount of Tannerella, Tanneralla forsythia, Erysipelotrichaceae_UCG_003, Bilophila wadsworthia, and / or NLAE zl H528 is increased in the sample, relative to the control.

22. The method of claim 21, wherein the treatment is administered when the amount of Bilophila wadsworthia is in the sample about 2 times greater, or about 3 times greater, relative to the control.

23. The method of claim 22, wherein the treatment is administered when the amount of Turicibacter sanguinis is decreased in the sample, relative to the control.

24. The method of claim 23, wherein the treatment is administered when the amount of Turicibacter sanguinis in the sample is about 3 times less, or about four times less, relative to the control.

25. The method of claim 20, wherein the treatment is administered when a ratio of the amount of Bilophila wadsworthia to the amount of Turicibacter sanguinis in the sample is about 5:1 to about 30:1, about 10:1 to about 20:1, or about 15:1, relative to the control.Atty. Dkt. No.650053.01137 26. The method of claim 20, wherein the control amount is determined using a control sample obtained from (i) a control subject not having or suspected of having neurodegeneration, or (ii) the subject at an earlier timepoint.

27. The method of claim 20, wherein the determining comprises (i) isolating bacterial DNA from the sample and sequencing the isolated bacterial DNA.

Images