Pathology-modulating neuromodulation therapy design

Abstract

Methods, systems, and devices, including computer programs encoded on computer storage media, are provided for optimizing neurostimulation therapy for the treatment of neurological and neurodegenerative diseases. In particular, algorithms are used to provide predicted regional pathological density maps of neuropathology and predict the location of future spread. Neurostimulation therapy parameters, including location, intensity, and frequency of neurostimulation, can be adjusted accordingly to treat the neuropathology and reduce aggregation and spread.

Description

[Background technology]

[0001] An emerging hypothesis for the progression of neurodegenerative diseases is the spread of misfolded proteins throughout the central nervous system (Goedert et al., 2013; Jucker and Walker, 2013; Oliveira et al., 2021). Building on the long-standing Braak hypothesis (Braak et al., 2002; Braak et al., 2003), which correlates the progression of pathology with motor and cognitive symptoms, it has been suggested that small seeds of misfolded versions of disease proteins spread from cell to cell, subsequently causing pathology and neurodegeneration (Volpicelli-Daley et al., 2011). Misfolding and spreading of the protein α-synuclein (α-syn) is implicated in many neurodegenerative disorders, collectively referred to as synucleinopathies, including Parkinson's disease (PD), dementia with Lewy bodies (LBD), and multiple system atrophy (MSA) (Kordower et al., 2008; Lee and Trojanowski, 2006). In mouse models, even small injected preformed fibrils (PFF) seeds of α-syn can recruit endogenous α-syn protein and induce widespread pathology in a "prion-like" manner (Luk et al., 2012). Although the precise mechanisms and genetic pathways involved in the pathogenesis of whole-brain disease resulting from small PFF seeds remain largely unknown, many recent studies have shown that these fibrils spread along axonal pathways (Henderson et al., 2019; Henrich et al., 2020; Pandya et al., 2019). Summary of the Invention

[0002] Methods, systems, and devices, including computer programs encoded on computer storage media, are provided for optimizing neurostimulation therapy for the treatment of neurological and neurodegenerative diseases. In particular, algorithms are used to provide predicted regional pathological density maps of neuropathology and predict the location of future spread. Neurostimulation therapy parameters, including location, intensity, and frequency of neurostimulation, can be adjusted accordingly to treat the neuropathology and reduce aggregation and spread.

[0003] In one aspect, a computer-implemented method for predicting the location of occurrence of pathological protein aggregates in the brain of a subject having a neurological or neurodegenerative disease, the method comprising: a) receiving an image of the subject's brain; b) identifying pathological protein aggregates in the image using a machine learning algorithm; c) mapping the locations of the pathological protein aggregates to neuroanatomical regions; d) modeling a discretized distribution of the pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model; and e) using a regional gene density map of each neuroanatomical region to predict the locations of the pathological protein aggregates. f) calculating the distribution of gene effects on regional spread and disintegration in the body; f) predicting changes in regional density of pathological protein aggregates as a function of time by modeling the spread, aggregation, disintegration, and spatial gene expression of pathological protein aggregates across neuroanatomical regions, where spread is assumed to emerge retrogradely between anatomically interconnected neuroanatomical regions, and where spread is modeled as diffusion through a weighted directed graph connecting neuroanatomical regions of the subject's brain; and g) predicting past, current, and future locations of pathological protein aggregates based on the modeling.

[0004] In certain embodiments, the method further includes adjusting one or more programmed neural stimulation parameters based on the predicted change in regional density of pathological protein aggregates as a function of time. For example, the duration, amplitude, frequency, pulse width, and location of the neural stimulation, or any combination thereof, may be adjusted.

[0005] In certain embodiments, the method further includes instructing the neurostimulation device to apply neurostimulation to brain locations predicted to contain pathological protein aggregates to treat a neurological or neurodegenerative disease in the subject.

[0006] In certain embodiments, the method further includes instructing a neurostimulator device to apply electrical stimulation to brain locations where pathological protein aggregates are not yet present but are predicted to appear in the future based on predicting changes in regional density of pathological protein aggregates as a function of time.

[0007] In certain embodiments, the method includes performing image registration to a coordinate space comprising a plurality of voxels, each voxel being represented as a cubic volume element centered at a coordinate in the coordinate space; identifying a location in x, y, z coordinates of each pathological protein aggregate in the coordinate space; measuring a volume of each pathological protein aggregate from a total number of voxels occupied by each pathological protein aggregate; and calculating an aggregate density for each voxel, the aggregate density for each voxel being determined from a total number of pathological protein aggregates centered within the same voxel. and calculating a total aggregate size for each voxel, the total aggregate size being the total size of all pathological protein aggregates centered within the same voxel; calculating an average aggregate size for each voxel as the total aggregate size for each voxel divided by the aggregate density for each voxel; calculating a total signal intensity for each voxel from the total intensities of all pathological protein aggregates centered within the same voxel; and calculating an average signal intensity for each voxel as the total signal intensity for each voxel divided by the aggregate density.

[0008] In certain embodiments, modeling the discretized distribution of pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model includes:

[0009]

number

[0010] where c i,jwhere σ represents the total count of pathological protein aggregates in a discretized size bin indexed by I in the brain region indexed by j, the L matrix represents the Laplacian matrix of the weighted directed graph connecting the neuroanatomical regions of the brain, η is chosen as a hyperparameter that slows down the spread of large aggregates as an inverse power of their size, and λ is chosen as a hyperparameter that accelerates the collapse of pathological protein aggregates proportionally to a power of their size.

[0011] In certain embodiments, the initial values ​​of α and μ are adapted by sweeping through a two-dimensional grid and selecting the values ​​that result in the smallest mean square error between the predicted and actual counts of pathological protein aggregates.

[0012] In a particular embodiment, with initial values ​​μ=0, k=0, ξ=0, and c being a one-dimensional size vector, the set of differential equations simplifies to a standard network diffusion model.

[0013] In certain embodiments, the method further includes quantifying the sensitivity of the model to a particular neuroanatomical pathway in the brain, wherein a Jacobian matrix is ​​calculated by taking the partial derivative of the output of the model with respect to weights of anatomical coupling strengths between two neuroanatomical regions encoded in the model, and the elements of the Jacobian matrix represent the relative contribution of the anatomical coupling between the two neuroanatomical regions to the spread of pathological protein aggregates to a particular region of the brain.

[0014] In certain embodiments, the method further includes using the model to generate a ranking of candidate seed locations for a given pathological state c at t=T months post injection (MPI) by a method that includes: using each of the neuroanatomical regions simulated forward in time until t=T MPI as a separate seed location at t=0 in the model, where each of the simulation results for the different neuroanatomical regions is compared to the observed state c using a pairwise similarity metric, where the similarity metric is a correlation coefficient between total regional aggregate counts across the observed and simulated states; and using the similarity metric value to sort seed locations for the neuroanatomical regions as likely sites leading to the observed pathological state c, whereby a ranking of candidate seed locations for the given pathological state c at t=T MPI is generated.

[0015] In certain embodiments, the method further includes predicting the MPI for a time since seeding, t=T, for a given pathological state c by a method that includes: comparing the whole-brain distribution of aggregate sizes for state c to simulated distributions at various t using a pairwise similarity metric without considering seed locations, where the distribution of simulated aggregate sizes across the whole brain is assumed to be invariant with respect to which neuroanatomical region is used as the seed location at t=0; and calculating the mean squared error between the stimulus distribution and the observed distribution, where, when deciding among multiple candidate t values, the mean squared errors are inverted and normalized to sum to one to provide a predicted probability for each t being a correct estimate of T for the given pathological state c.

[0016] In certain embodiments, gene effects on regional spread and decay of pathological protein aggregates are determined by a method that includes: assuming that the α (spread) and μ (decay) parameters are region-dependent, such that spread from a particular neuroanatomical region is proportional to the gene density in that region; normalizing all genes to the same range so that only the regional distribution of gene expression is compared to the whole-brain expression of that gene, where α is a vector and the product of α and the Laplacian connectivity matrix L has the effect of correcting the regional connectivity encoded in the model; and normalizing each gene vector to have a mean of 1 and a standard deviation Σ empirically set to preserve the correlation between predicted and observed whole-brain aggregate counts, where the normalization is chosen so that the product has the effect of maintaining a trace of the original Laplacian connectivity matrix L. In some embodiments, a derivation of normalization to preserve the trace of the original Laplacian connectivity matrix L is to assume that the vector s is sampled from a multivariate normal distribution with mean 1 and standard deviation Σ, where s ∼N(1,Σ), and use the definition of matrix trace to represent s as a diagonal square matrix S, where the trace of the product of S and the Laplacian connectivity matrix L is

[0017]

number

[0018] where l represents the diagonal of L, the trace is equal to the dot product of s and l, and has expectation equal to the sum of the entries of l, s·l~N(1·l,lΣl) E[s·l]=Tr(L) The method involves recovering the definition of the trace of L according to the formula: After each gene is coded into the model, comparing the net effect on regional correlation between the simulated and actual data with a baseline correlation that does not include the gene, and providing an ordered list of genes ranked by the relevance of their spatial expression map in improving the model's regional predictions.

[0019] In a particular embodiment, the cubic volume element has a width in coordinate space of 100 μm.

[0020] In certain embodiments, one or more pathological protein aggregates are mapped to a single voxel.

[0021] In certain embodiments, the computer-implemented method further includes performing multi-dimensional Gaussian filtering to account for variations in image registration between different samples.

[0022] In certain embodiments, the computer-implemented method further includes segmenting the image to generate a plurality of image segments.

[0023] In certain embodiments, the locations of the pathological protein aggregates are mapped to neuroanatomical regions of the Allen Human Brain Reference Atlas. In some embodiments, the mapping comprises performing image registration to transform the locations of the pathological protein aggregates into the coordinate space of the Allen Human Brain Reference Atlas. In some embodiments, anatomically interconnected neuroanatomical regions are identified from the Allen Connectivity Atlas.

[0024] In certain embodiments, the neuroanatomical region is anterior amygdala region, anterior cingulate region, dorsal, anterior cingulate region, ventral, nucleus accumbens, anterior dorsal nucleus, anterior hypothalamic nucleus, agranular insular region, dorsal, agranular insular region, posterior, agranular insular region, ventral, nucleus ambiguus, anterior medial nucleus, dorsal, anterior medial nucleus, ventral, oblique lobule, accessory olfactory bulb, anterior olfactory nucleus, anterior pretectal nucleus, arcuate hypothalamic nucleus, dorsal auditory cortex, primary auditory cortex, ventral auditory cortex, anterior ventral nucleus of the thalamus, basolateral amygdala, basomedial amygdala, bed nucleus of the stria terminalis, CA1 area, CA2 area, CA3 area, central nucleus of the amygdala, central lobule, lateral central nucleus of the thalamus, claustrum, central linear raphe, medial central nucleus of the thalamus, cortical amygdala region, anterior, cortical Substantia amygdaloidea, posterior portion, caudate nucleus, superior central raphe, anterior cerebellum, cuneiform nucleus, dorsal cochlear nucleus, dentate gyrus, dorsomedial nucleus of the hypothalamus, dentate nucleus, dorsal peduncle region, dorsal raphe nucleus, rhinal region, entorhinal region, lateral portion, entorhinal region, medial portion, dorsal zone, caudate nucleus, dorsal portion, caudate nucleus, ventral portion, flocculus, dorsal cerebellar nucleus, anterior column, cerebral cortex, basal striatum, globus pallidus, external segment, globus pallidus, internal segment, nucleus reticularis magnocellularis, gustatory region, intercalated amygdaloid nucleus, inferior colliculus, central nucleus, inferior colliculus, dorsal nucleus, inferior colliculus, external nucleus, infralimbic region, intermediate dorsal nucleus of the thalamus, inferior olivary complex, nucleus intermedius, interpeduncular nucleus, intermediate reticular nucleus, lateral amygdaloid nucleus, lateral vestibular nucleus, lateral dorsal nucleus of the thalamus, lateral geniculate complex dorsal part of the lateral geniculate complex, ventral part of the lateral habenula, lateral area of ​​the hypothalamus, lateral postoptic nucleus of the thalamus, lateral preoptic area, lateral reticular nucleus, lateral septal nucleus, caudal part, lateral septal nucleus, rostral part, lateral septal nucleus, abdomen, magnocellular nucleus, magnocellular reticular nucleus, medial dorsal nucleus of the thalamus, medullary reticular nucleus, dorsal part, medullary reticular nucleus , abdomen, medial amygdala nucleus, median preoptic nucleus, medial geniculate complex, dorsal part, medial geniculate complex, medial part, medial geniculate complex, abdomen, medial habenula, medial mammillary nucleus, main olfactory bulb, primary motor area, secondary motor area, medial preoptic nucleus, medial preoptic area, medial pretectal area, mesencephalic reticular nucleus, medial septal nucleus, medial vestibular nucleus, diagonal zone nucleus, Nucleus indeterminate, nucleus of the lateral lemniscus, nucleus of the lateral olfactory tract, tubercle (X), nucleus of the optic tract, nucleus of the posterior commissure, nucleus of the solitary tract, orbital region, lateral part, orbital region, medial part, orbital region, ventrolateral part, olfactory tubercle, posterior amygdaloid nucleus, piriform amygdaloid region, periaqueductal gray matter, paraunculoid gyrus, parvocellular reticular nucleus, parabrachial nucleus, central pontine gray matter, perirhinal region, parafascicular nucleus, paraflocculus lobule, pontine gray matter, paracellular reticular nucleus, dorsal part, paracellular reticular nucleus, lateral part, posterior nucleus of the hypothalamus, piriform region, prelimbic area, dorsal premammillary nucleus, posterior thalamic complex, posterior bordering nucleus of the thalamus, posterior uncinate gyrus, peripeduncular nucleus, pedunculopontine nucleus, preuncleate gyrus, paramedian lobule, pontine reticular nucleus, caudal part, pontine reticular nucleus, precursor nucleus, principal sensory nucleus of the trigeminal nerve,Parathalamic nucleus, posterior parietal related area, paraventricular hypothalamic nucleus, paraventricular thalamic nucleus, periventricular hypothalamic nucleus, posterior, periventricular hypothalamic nucleus, preoptic area, lobule VIII, postchiasmatic area, nucleus of reuniens, rhomboid nucleus, nucleus raphe magnus, nucleus red, midbrain reticular nucleus, posterior ruber area, retrosplenial area, lateral agranular area, retrosplenial area, dorsal, retrosplenial area, ventral, thalamic reticular nucleus, Paraventricular region, superior colliculus, motor-related, superior colliculus, sensory-related, septal fimbria, substantia innominata, lobule simplex, medial subthalamic nucleus, substantia nigra, pars compacta, substantia nigra, pars reticulata, superior olivary complex, subparafascicular region, subparafascicular nucleus, magnocellular part, subparafascicular nucleus, parvocellular part, spinal vestibular nucleus, spinal nucleus of the trigeminal nerve, tail, spinal nucleus of the trigeminal nerve, interpolar part, spinal nucleus of the trigeminal nerve, oral part, primary somatic sensory Selected from the sensory cortex, barrel area, primary somatosensory cortex, lower limbs, primary somatosensory cortex, mouth, primary somatosensory cortex, nose, primary somatosensory cortex, trunk, primary somatosensory cortex, upper limbs, additional somatosensory areas, subthalamic nucleus, uncinate gyrus, supramammillary nucleus, supratrigeminal nucleus, superior vestibular nucleus, temporal association area, posterior piriform transition area, tegmental reticular nucleus, triangular septal nucleus, stria tecta, tubercular nucleus, motor nucleus of the trigeminal nerve, ventral anterolateral complex of the thalamus, ventral cochlear nucleus, facial motor nucleus, visceral area, anterolateral visual area, anteromedial visual area, lateral visual field, primary visual area, posterolateral visual area, posteromedial visual area, ventromedial nucleus of the thalamus, ventromedial nucleus of the hypothalamus, ventral posterolateral nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, parvocellular portion, ventral tegmental area, and hypoglossal nucleus.

[0025] In certain embodiments, the computer-implemented method further comprises predicting where the pathological protein aggregates occurred in the subject's brain.

[0026] In certain embodiments, the machine learning algorithm uses an artificial neural network.

[0027] In certain embodiments, the machine learning algorithm uses a deep learning algorithm.

[0028] In particular embodiments, the deep learning algorithm uses a convolutional neural network, a deep neural network, a recurrent neural network, a deep residual neural network, a long short-term memory network, a deep belief network, a multi-layer perceptron, or deep reinforcement learning.

[0029] In certain embodiments, the machine learning algorithm is supervised, semi-supervised, or unsupervised.

[0030] In certain embodiments, the subject is a human subject.

[0031] In certain embodiments, modeling the spreading, aggregation, and disintegration of pathological protein aggregates in the brain of a human subject uses simulations generated based on measuring the spreading, aggregation, and disintegration of pathological protein aggregates in a non-human animal model, and the experimentally measured pathology in the non-human animal model is used to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human subject.

[0032] In certain embodiments, the simulation further uses measured brain anatomy or brain function from a non-human animal to predict the past, current, and future locations of pathological protein aggregates in a human subject.

[0033] In certain embodiments, the non-human animal is a mammal. In some embodiments, the mammal is a rodent or a primate. In some embodiments, the rodent is a mouse.

[0034] In certain embodiments, the simulation further uses measured brain anatomy or brain function from other human subjects to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human subject.

[0035] In certain embodiments, the computer-implemented method further includes receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; and displaying a change in total aggregate size for each voxel, the volume of each pathological protein aggregate for each voxel, and the aggregate density for each voxel in the image taken after neurostimulation is applied to the subject's brain compared to the image taken before neurostimulation is applied to the subject's brain.

[0036] In certain embodiments, the computer-implemented method further includes receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; modulating one or more programmed neurostimulation parameters based on any changes in past, current, or future locations where pathological protein aggregates are predicted to appear; and instructing the neurostimulation device to apply the modulated neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease in the subject.

[0037] In certain embodiments, the computer-implemented method further includes storing a user profile for the subject, the user profile including information regarding programmed neurostimulation parameters to be used to apply neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease based on the location where the computer-implemented method predicts that pathological protein aggregates exist or will occur in the future.

[0038] In another aspect, a non-transitory computer-readable medium is provided that includes program instructions that, when executed by a processor in a computer, cause the processor to perform the methods described herein.

[0039] In another aspect, a kit is provided that includes a non-transitory computer-readable medium and instructions for treating a neurological or neurodegenerative disease in a subject using neurostimulation.

[0040] In another aspect, a method for treating a neurological or neurodegenerative disease in a subject is provided, the method comprising imaging pathological protein aggregates in the subject's brain; using a computer-implemented method described herein to predict where the pathological protein aggregates will occur based on locations of the pathological protein aggregates detected in the subject's brain by imaging; and applying neurostimulation to the locations in the brain where the pathological protein aggregates are detected in the subject's brain by imaging and where the computer-implemented method predicts that the pathological protein aggregates will occur.

[0041] In certain embodiments, imaging is performed using computed tomography (CT), single photon emission computed tomography (SPECT), magnetic resonance imaging, functional magnetic resonance imaging, optogenetic functional magnetic resonance imaging, or positron emission tomography (PET).

[0042] In certain embodiments, the method further includes adjusting the stimulation frequency and pulse width of the neurostimulation to target specific neuronal cell types or circuits in the brain at locations in the brain where the computer-implemented method predicts that pathological protein aggregates are present or will occur in the future.

[0043] In certain embodiments, the neurological or neurodegenerative disease is a synucleinopathy, including, but not limited to, Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy.

[0044] In certain embodiments, the neurological or neurodegenerative disease is Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia.

[0045] In certain embodiments, the pathological protein aggregates comprise alpha-synuclein aggregates.

[0046] In certain embodiments, applying the neural stimulation includes applying the neural stimulation using electrodes.

[0047] In certain embodiments, the electrodes are deep electrodes or surface electrodes.

[0048] In certain embodiments, the electrodes are non-penetrating surface electrode arrays or penetrating brain electrode arrays.

[0049] In certain embodiments, applying the neurostimulation includes applying deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation.

[0050] In certain embodiments, applying the neural stimulation comprises optogenetically applying the neural stimulation.

[0051] In certain embodiments, neural stimulation is applied optogenetically by a method comprising: introducing a recombinant polynucleotide encoding a light-responsive ion channel into a neuron at a location in the brain where a computer-implemented method predicts that pathological protein aggregates exist or will occur in the future, wherein the light-responsive ion channel is expressed in the neuron; and illuminating the light-responsive ion channel with light of a wavelength that activates the light-responsive ion channel, wherein conduction of ions by the light-responsive ion channel in response to absorption of light results in hyperpolarization or depolarization of the neuron.

[0052] In certain embodiments, the light-responsive ion channel is a light-responsive anion-conducting opsin or a light-responsive proton conductance modulator.

[0053] In certain embodiments, the light-responsive anion-conducting opsin is capable of activating chloride ions (Cl - ) is conducted.

[0054] In certain embodiments, the anion-conducting opsin is an anion-conducting channelrhodopsin or a halorhodopsin.

[0055] In certain embodiments, the halorhodopsin is Natronomonas pharaonis halorhodopsin (NpHR), enhanced NpHR (eNpHR) 1.0, eNpHR 2.0, or eNpHR 3.0.

[0056] In certain embodiments, the anion-conducting channelrhodopsin is iC1C2, SwiChR, SwiChR++, or iC++.

[0057] In certain embodiments, the light-responsive proton conductance modulator is bacteriorhodopsin or archaealhodopsin.

[0058] In certain embodiments, the light-responsive proton conductance modulator is Arch from Halorubrum sodomense, ArchT from Halorubrum sp., TP009 from Leptosphaeria maculans, or Mac from Leptosphaeria maculans.

[0059] In certain embodiments, the light-responsive ion channel is a light-responsive cation-conducting opsin.

[0060] In certain embodiments, the light-responsive cation-conducting opsin is a photoresponsive opsin that binds calcium cations (Ca 2+ ) is conducted.

[0061] In certain embodiments, the light-responsive cation-conducting opsin is a light-responsive cation-conducting channelrhodopsin.

[0062] In certain embodiments, the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin or Volvox carteri channelrhodopsin.

[0063] In certain embodiments, the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin-1 (ChR1), Chlamydomonas reinhardtii channelrhodopsin-2 (ChR2), Volvox carteri channelrhodopsin-1 (VChR1), or a chimeric ChR1-VChR1 channelrhodopsin.

[0064] In certain embodiments, the polynucleotide encoding the light-responsive ion channel is provided by a viral vector.

[0065] In certain embodiments, the viral vector is a lentiviral vector or an adeno-associated viral (AAV) vector.

[0066] In certain embodiments, viral vectors are stereotactically injected into the brain at locations where computer-implemented methods predict that pathological protein aggregates exist or will occur in the future.

[0067] In certain embodiments, the vector further comprises a neuron-specific promoter operably linked to the polynucleotide encoding the light-responsive ion channel.

[0068] In certain embodiments, expression of the light-responsive ion channel is inducible.

[0069] In certain embodiments, illuminating the light-responsive ion channels comprises delivering light from a light source to the light-responsive ion channels using a fiber optic-based optical neural interface.

[0070] In certain embodiments, the light source is a solid state diode laser.

[0071] In certain embodiments, applying the neurostimulation includes applying the neurostimulation to a motor cortical region or a subcortical region of the brain.

[0072] In certain embodiments, multiple cycles of neural stimulation are performed.

[0073] In certain embodiments, the method further includes evaluating the effectiveness of the treatment of the neurological or neurodegenerative disease in the subject. In some embodiments, evaluating includes imaging the subject's brain to measure the size and identify the location of pathological protein aggregates after the neural stimulation. In some embodiments, evaluating includes measuring brain function of the subject after the neural stimulation. For example, measuring brain function can be measured by performing electroencephalography (EEG), stereoelectroencephalography (sEEG), electrocorticography (ECoG), magnetoencephalography (MEG), single-photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), optogenetic functional magnetic resonance imaging, or positron emission tomography (PET).

[0074] In certain embodiments, the method further includes modulating one or more programmed neurostimulation parameters to improve brain function.

[0075] In another aspect, a system for treating a neurological or neurodegenerative disease in a subject is provided, the system comprising: a neurostimulation device; and a processor programmed to instruct the neurostimulation device to deliver neurostimulation to the subject's brain in a manner effective to treat the neurological or neurodegenerative disease in the subject according to the computer-implemented methods described herein, wherein the neurostimulation is applied to the brain at predicted current locations of pathological protein aggregates, at predicted future locations of pathological protein aggregates, or at predicted past locations of pathological protein aggregates, or combinations thereof.

[0076] In certain embodiments, the neurostimulator device comprises an electrode. In some embodiments, the electrode is a deep electrode or a surface electrode. In some embodiments, the electrode is a non-penetrating surface electrode array or a penetrating brain electrode array.

[0077] In certain embodiments, the neurostimulator device performs deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation.

[0078] In certain embodiments, the system further comprises a user interface comprising an input electronically coupled to the processor for instructing the neurostimulation device to apply neurostimulation to the brain of the subject to treat a neurological or neurodegenerative disease in the subject.

[0079] In certain embodiments, the user interface is password protected and operable by medical personnel.

[0080] In certain embodiments, the system further comprises a display. In certain embodiments, the display displays an image of the subject's brain showing predicted current, past, or future locations of pathological protein aggregates determined by the computer-implemented method. In some embodiments, the display displays information regarding the coordinates of each pathological protein aggregate, the volume of each pathological protein aggregate, the aggregate density of each voxel, the total aggregate size of each voxel, the average aggregate size of each voxel, the total signal intensity of each voxel, the average signal intensity of each voxel, or a mapping of the locations of pathological protein aggregates to neuroanatomical regions, or any combination thereof. In some embodiments, the display displays information regarding the distribution of gene effects on the regional spread and disintegration of pathological protein aggregates. In some embodiments, the display displays information regarding predicted changes in the regional density of pathological protein aggregates as a function of time determined by modeling the spread, aggregation, disintegration, and spatial gene expression of pathological protein aggregates across neuroanatomical regions. In some embodiments, the display displays information regarding the predicted past, current, and future locations of pathological protein aggregates based on the modeling.

[0081] In certain embodiments, the system is for use in the treatment of a synucleinopathy, including, but not limited to, Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy.

[0082] In certain embodiments, the system is for use in treating Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia. [Brief explanation of the drawings]

[0083] [Figure 1A] Tissue clearing and light-sheet fluorescence microscopy capture changes in whole-brain pathology at various time points after seeding. α-syn PFFs were unilaterally injected into the striatum of mice, and cohorts of mice were perfused at various time points ranging from 2 weeks to 18 months post-injection (MPI). Each extracted mouse brain was processed for fluorescent immunolabeling of α-syn pathology and whole-brain clearing using the iDISCO+ protocol. Brains were imaged three-dimensionally by light-sheet fluorescence microscopy to visualize both antibody fluorescence and tissue autofluorescence for anatomical mapping. [Figure 1B] Axial projection of autofluorescence from an imaged mouse brain (left) and α-syn pSer129 immunolabeled pathology (right). [Figure 1C] A quantitative pipeline aligns autofluorescence to an anatomical atlas, and a trained classifier segments α-syn pathology. [Figure 1D] A quantitative pipeline aligns autofluorescence to an anatomical atlas, and a trained classifier segments α-syn pathology. [Figure 1E] Both whole-brain spread and subsequent reduction in pathology are observed in glass-brain reconstructions of representative samples at each time point, with each aggregate color-coded by Allen Criterion atlas region. [Figure 1F] Total α-syn inclusion counts versus time post-injection quantifies this spreading and subsequent propensity for decay. [Figure 1G]Normalized density distributions of average aggregate size at each voxel across various time points show a general increase in aggregate volume during the initial, apparent prion-like spread, followed by a subsequent decrease as large aggregates disappear. Data are expressed as mean ± SD. Isoctx - isocortex, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. See also Figures 6, 8, and 9. [Figure 2A] Statistical analysis at both the regional and voxel levels demonstrates biphasic spread and decay between cortical and subcortical regions. The computational pipeline used to process each brain sample consists of alignment to a reference atlas, segmentation of three-dimensional aggregate volumes, and then using these two to map each aggregate to a neuroanatomical region or voxel within a shared coordinate space within the Allen Reference Atlas (ARA). This allows for longitudinal group statistical comparisons at both the regional and brain voxel levels. [Figure 2B] Voxel-level statistics using heat maps from pairs of time points facilitate the discovery of voxel clusters with statistically significant (p<0.05) vulnerability to early pathological spread and distinct accumulations of mean aggregate volumes. [Figure 2C] Grouping into ARA regions before statistical testing yields similar results: Isoctx - isocortex, OLF - olfactory region, HPF - hippocampal formation, CTX sp - cortical subplate, CNU - caudate nucleus, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. See also Figure 10 and Table S1. [Figure 3A] The computational model describes spreading, aggregation, and collapse. The computational model describes the spreading of aggregation across the nodes of a directed graph, which relies on anatomical connectivity estimates from the Allen connectivity atlas. Each node represents an atlas region, and each edge represents anatomical neuronal connectivity between two regions. Thicker lines represent higher anatomical connectivity. [Figure 3B]To model the interactions between aggregates of various sizes within the model, the volume of each aggregate is discretized into one of several size bins that are tracked as separate model variables within each region. Particles of discrete sizes within each region can accumulate in an additive combination of volumes. [Figure 3C] The fitted model accurately simulates both the longitudinal whole-brain counts for each discretized aggregate size (r=0.98) and the regional counts for each size (r=0.72). [Figure 3D] The raw time series output from the computational model demonstrates the model's ability to capture the dynamics of each discretized aggregate size. The black line represents the model prediction of total aggregates of a given size, while the gray line represents the actual observed counts. [Figure 3E] Jacobian calculations between adjacent time points quantify the sensitivity of the model to specific anatomical connections. The top 10% Jacobian matrix elements for each time point pair are displayed: Isoctx - isocortex, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. See also Figure 11. [Figure 4A] Seeding α-syn fibrils in different brain regions results in consistent volume distribution of aggregate formation but distinct spreading patterns across regions, both of which are predicted by a computational model fitted to the striatal dataset (Figure 4A). α-syn PFFs are injected into new seed locations in independent cohorts for the main olfactory bulb (MOB), substantia nigra (SN), and dentate gyrus (DG). Mice are perfused at 0.5, 2, and 4 MPI. [Figure 4B] The distribution of aggregate sizes for various seed locations shows consistency across time points. The 0.5 MPI sample consistently contained a larger number of small aggregates, but this distribution shifted toward larger aggregates at later time points. Data are presented as mean ± SD. However, voxel-level and regional statistics from 0.5 MPI to 4 MPI indicate that different seed locations result in distinct downstream spreading patterns. [Figure 4C]Voxel-level and regional statistics show that different seed locations result in different downstream spreading patterns. [Figure 4D] Voxel-level and regional statistics show that different seed locations result in different downstream spreading patterns. [Figure 4E] Correlation of each in vivo seed location with model output from in silico seeding of all 424 regions in the ARA. Both ipsilateral and contralateral results for the MOB, SN, and DG are labeled. In all cases, the model can accurately distinguish between unknown datasets with various distinct α-syn PFF seed locations. Side-by-side comparisons of actual and simulated histograms of aggregate volume for each additional seeding site, and confusion matrices between actual and simulated states, demonstrate that the model can predict progression since initial seeding (MPI) using discretized aggregate size histograms. See also Figures 12 and 13. [Figure 5A] Integrating spatial transcriptomics data into computational models reveals genes associated with spreading across seed locations. Encoding region-specific gene densities from the Allen ISH database into the models allows comparison of the association of each gene with spreading and decay parameters in improving predictive power. [Figure 5B] A joint heatmap of spreading and disruption gene rankings shows the clustering of genes associated with either spreading or disruption, further labeling genes involved in Parkinson's disease and synucleinopathies. [Figure 5C] Histogram of genes for each parameter that improved model performance. Genes are grouped by the cell type with the highest transcription level of that gene, taken from the Allen Atlas. [Figure 5D]Simulation results from all genes in the Allen ISH database tested separately for various PFF seed positions. Regional correlations between the simulated output of each gene and the entire time series for the seed position are reported, and the ordering of shared genes on the x-axis is determined by the ranked genes in striatum in ascending order. [Figure 5E] Highly ranked coding genes from the striatal dataset consistently improved the predictive power of the model for other seed regions, as measured by correlation coefficients, whereas lower percentile genes consistently decreased predictive power. Data are presented as mean ± SD. [Figure 6] Related to Figure 1, three-dimensional acquisition from whole-brain tissue clearing and light-sheet microscopy provides a readout of neuroanatomical contralateral and whole-brain α-syn pathology. Sagittal and axial projections of autofluorescence. Sagittal and axial projections of pSer129 labeling. [Figure 7A] Serial histological sectioning detects α-syn pathology in similar regions as whole-brain sections, but is inadequate for the longitudinal or regional comparisons associated with the STAR method. Conventional serial histology shows a layer-wide staining pattern in the striatum and multiple cortical regions. [Figure 7B] Distribution along the AP axis of histological coronal sections taken from animals at 2 and 6 MPI. [Figure 7C] Comparison of area aggregate counts between iDISCO acquisition (top; 2,6MPI) and extrapolated results from traditionally immunolabeled sections (bottom; 2,6MPI). [Figure 7D] The results indicate that serial histological estimation captures neither the absolute number of aggregates nor the relative numbers between regions. Data are expressed as mean ± SD. Isoctx - isocortex, OLF - olfactory region, HPF - hippocampal formation, CTX sp - cortical subplate, CNU - caudate nucleus, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. [Figure 8A]With reference to Figure 1, the alignment pipeline precisely aligns the 3D acquisitions into a shared ARA coordinate space. Alignment quality was optimized using a mutual information similarity metric that converges to a similar range of values ​​across samples. [Figure 8B] We calculated the linear unfolding of the cleared brain using the singular values ​​of the affine transformation matrix used in the registration process. [Figure 8C] The green outline of the atlas was overlaid onto the brain autofluorescence channel in both the imaging plane (sagittal) and the reslicing plane (coronal, axial) to demonstrate the registration quality. Data are presented as mean ± SD. [Figure 9A] With reference to Figure 1, the image segmentation pipeline performs accurate detection of pathological aggregates. Overview of a machine learning classifier (random forest) trained on 3D structural features and generating a probability of each voxel as foreground (pathology) or background. ROC and PRC curves quantify the segmentation performance on the test dataset. [Figure 9B] Raw extracts of pSer129 labeling along with the corresponding segmentation maps. [Figure 10] Western blot heatmap quantification for each time point after CP seeding shows spread and reduction. Montage of average voxel-level density maps for all acquisition time points after PFF seeding in the striatum. [Figure 11A] Computational model parameters were optimized and then evaluated across the entire dataset. An anatomical connectivity matrix (obtained from the Allen Connectivity Atlas) describing both ipsilateral and contralateral connectivity strengths between 212 regions. [Figure 11B] Sections from the ARA atlas used to group aggregates into neuroanatomical regions. [Figure 11C] Various versions of Laplacian matrices of directed weighted graphs representing regional connectivity were tested, including both forward and reverse anatomical connectivity matrices from the Allen connectivity atlas, and matrices weighted by the Euclidean distance between regions. [Figure 11D]Tree showing the hierarchy of brain regions in ARA, with regions used in the model highlighted in red. [Figure 11E] The number of size bins used in the size discretization was also swept. Increasing this hyperparameter improved model performance at a decreasing rate. A discretization of seven equally spaced sizes was chosen for all simulations to reduce the dimensionality of the model output and stay within computational limits. Similarly, increasing the number of regions used in the model only improved model performance. Isoctx - isocortex, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. [Figure 12] Heatmap quantification of additional seed locations reveals distinct initial spreading patterns. Montage of averaged voxel-level density maps for all acquisition time points after PFF seeding in the main olfactory bulb (MOB), substantia nigra (SN), and dentate gyrus (DG). [Figure 13A] The model generalizes to predict whole-brain and regional spreading patterns from alternative α-syn seeding sites. For different seeding sites, the model trained on the striatal dataset can accurately predict α-syn aggregate counts longitudinally across the whole brain, across various discretized sizes, and across neuroanatomical regions. [Figure 13B] Statistical tests across neuroanatomical regions detect distinct changes in aggregate count or average size for each seeding site: Isoctx - isocortex, OLF - olfactory region, HPF - hippocampal formation, CTX sp - cortical subplate, CNU - caudate nucleus, TH - thalamus, HY - hypothalamus, MB - midbrain, HB - hindbrain, CB - cerebellum. [Figure 14A] Whole-brain immunolabeling and quantification of alpha-synuclein pathology after clearing informs candidate regions to target for neuromodulation. After injection of alpha-synuclein into the striatum, clearing and imaging of whole brain tissue captures the spread of pathology from the striatum to many distant brain regions at 2WPI. [Figure 14B] The quantification pipeline performs detection of each aggregate while aligning it to the ARA. [Figure 14C] Regional grouping of alpha-synuclein spread and comparison across subjects provides a list of candidate target regions for stimulation, from which motor areas consistently show high pathological levels. [Figure 15A] Repetitive optogenetic stimulation of the motor cortex after striatal injection of alpha-synuclein PFFs alters whole-brain pathology. Schematic showing the experimental paradigm: striatal injection of alpha-synuclein PFFs and implantation of optical fibers into the secondary motor cortex (MOs), followed by daily optogenetic stimulation for two weeks. Each daily stimulation consisted of ten 1-minute stimulation periods with a 1-minute rest between each period. [Figure 15B] Maximum intensity projections (MIPs) of cleared and labeled brains, alongside enlarged cortical sections from the ipsilateral hemisphere, show a reduction in pathological aggregation from the control to the stimulated group. [Figure 16A] Whole-brain statistical analysis validates that optogenetic stimulation modulates α-synuclein inclusion counts at both the individual voxel and neuroanatomical region levels. Statistical comparisons at the voxel level between treatment and control groups reveal that pathological aggregate counts are significantly reduced in many ipsilateral subcortical clusters and near the stimulation site, while aggregate counts are significantly increased in many contralateral cortical clusters. [Figure 16B] Optogenetic stimulation effects quantified across neuroanatomical regions by counting the number of significant voxels per region and hemisphere, or by calculating total aggregate counts per region and performing statistical comparisons across treatment and control groups, show results consistent with voxel-level quantification. [Figure 17A] Whole-brain functional activity measured with optogenetic fMRI during optogenetic stimulation predicts downstream pathological changes. Whole-brain BOLD fMRI was measured during optogenetic stimulation of the secondary motor cortex (layer V). [Figure 17B]BOLD activation maps overlaid with statistical changes in pathology measured by iDISCO show high colocalization between positive BOLD and decreased aggregate counts, whereas negative BOLD colocalizes with increased aggregate counts. [Figure 17C] Statistical activation maps show positive functional activity at the stimulation site and in subcortical regions, and negative activity in the contralateral cortex. [Figure 18] Brain clearing and immunolabeling allows for the capture of pathological conditions throughout the brain. Maximum intensity projections of cleared brains imaged with light-sheet fluorescence microscopy demonstrate the spread of pathological α-synuclein throughout the brain after injection into the striatum. The bottom of the white line indicates the site of stimulation in the secondary motor cortex (layer V) of stimulated subjects. The tip of the inverted triangle indicates the site of injection of alpha-synuclein PFFs in the striatum for both control and stimulated subjects. [Figure 19] Optogenetic stimulation power was selected by finding the minimum optogenetic laser power required to produce consistent rotational behavior. On each stimulation day for a given subject, power was increased until consistent rotational behavior was observed. [Figure 20] Comparison of pathology between control and stimulated mice across two separate cohorts of subjects. Statistical maps comparing α-synuclein aggregate counts at the voxel level for two different cohorts of animals that underwent two separate batches of iDISCO treatment. Each cohort consisted of a separate control and stimulated group, all imaged 2 weeks post-injection (WPI). A consistent stimulation effect, consisting primarily of a decrease in ipsilateral aggregation and an increase in contralateral aggregation, is evident in both cohorts. [Figure 21] Wild-type mice with no ChR2 expression show no change in whole-brain pathology after 2 weeks of sham stimulation. Statistical comparison of whole-brain pathology between control and sham-stimulated mice at 2 weeks post-injection. Wild-type mice had no ChR2 expression but still received daily laser power. Voxel-based statistical maps between control and sham mice show little or no change in aggregate counts. [Figure 22]Individual ofMRI activation maps for each subject. Rows represent subjects used for optogenetic fMRI, and a single column in each row represents the fMRI activation map for a 6-minute acquisition. Each image is a threshold statistical map (p<0.01, corrected) showing both positive and negative activation based on z-scores from a generalized linear model fit to the voxel time series. The ofMRI maps are highly consistent within and across all subjects. [Figure 23A] Optogenetics functional magnetic resonance imaging bridges the scale. Optogenetics allows cell-type-specific stimulation, such as selective targeting of D1- or D2-MSNs in the striatum. [Figure 23B] Selective stimulation of D1- or D2-MSNs in the striatum caused mice to exhibit contralateral or ipsilateral rotations, respectively (D1-MSN stimulation: n = 12 animals, mean ± SEM, ***p < 0.001, two-tailed paired t-test; D2-MSN stimulation: n = 11 animals, mean ± SEM, *p < 0.05, **p < 0.005, two-tailed paired t-test). [Figure 23C] Optogenetic fMRI techniques combine optogenetic stimulation with fMRI readout. [Figure 23D] ofMRI with selective stimulation of D1- and D2-MSNs yields distinct whole-brain activity associated with distinct behaviors (n = 12 animals). Group-wise phase maps thresholded for only active voxels within the brain show heterogeneity in the temporal dynamics of the evoked responses. [Figure 23E] Time series of any region can be extracted from four-dimensional fMRI data. [Figure 23F] Electrophysiological recordings mirror the fMRI response in terms of the polarity of neural activity changes. This figure is based on Lee et al. (2). [Figure 24A] Computational modeling of MRI data reveals functional interaction dynamics across the brain. The cortico-basal ganglia-ganglion-thalamus network comprises numerous network nodes spanning the brain. [Figure 24B]The anatomical connectivity of the direct and indirect pathways involves many common anatomical regions with different cell types in the caudate-putamen (CPu). [Figure 24C] Anatomical connections were used as a priori generative network models. In addition to the direct and indirect pathways shown in Figure 24B, other established anatomical connections, such as the hyperdirect pathway, were also included. [Figure 24D] The fMRI time series generated by DCM closely matched the experimental ofMRI time series. [Figure 24E] DCM utilized ofMRI data to estimate causal influences (effective connectivity) between regions of interest during D1- and D2-MSN stimulation, respectively. [Figure 24F] Graph and matrix representation of the effective connectivity network for D1-MSN stimulation. [Figure 24G] Graphical and matrix representation of the effective connectivity network for D2-MSN stimulation. Significant and near-significant represent parameters with p<0.05 and 0.05≦p<0.10, respectively (one-sample t-test, multiple comparison correction across connections with FDR p<0.10). CPu: caudate putamen; GPe: globus pallidus external; GPi: globus pallidus internal; STN: subthalamic nucleus; SNr: substantia nigra; THL: thalamus; CTX: cortex. This figure is based on Bernal-Casas et al. (4). [Figure 25A] Brain circuit functional modeling at the single-cell spiking level can be enabled through a multiscale approach. We illustrate schematically the location of optogenetic stimulation and in vivo extracellular recording for the study of the cortico-basal ganglia-thalamic network. [Figure 25B] We demonstrate single-cell spiking-level modeling using ofMRI and single-unit recording data. The large-scale model constructed with ofMRI data is extended to a single-neuron-level biophysical model. Each ROI consists of many simulated single neurons. The model is validated by directly comparing the simulated spiking trains with experimental data. [Figure 25C]The model can allow experimental single-unit recording data and simulated data to be directly compared. [Figure 25D] The model should be designed so that the spike rates of all ROIs simulated by the single-cell spiking-level biophysical model statistically match the experimental data. ofMRI combined with biophysical modeling can enable successful reproduction of single-cell spiking-level dynamics induced by cell-type-specific optogenetic stimulation, such as D1- and D2-MSN stimulation. [Figure 26A] To understand patho-function interactions, whole-brain pathodynamics can be modeled in parallel with ofMRI. Alpha-synuclein PFFs injected into seed locations induce pathology at various time points post-injection, which can then be captured by iDISCO tissue clearing and light-sheet fluorescence microscopy (LSFM). Machine learning-based automated segmentation and registration techniques can streamline the quantification of each pathological marker within the Allen Reference Atlas (ARA). [Figure 26B] Comparison of averaged heat maps across cohorts can show whole-brain pathology changes over many months post-injection. [Figure 26C] Modeling longitudinal data based on whole-brain anatomical connectivity can capture regional differences in pathology. [Figure 26D] Reweighting the connectivity matrix can allow coding of genetic contributions within the model. [Figure 26E] Whole-brain colocalization analysis between optogenetic stimulation-induced alpha-synuclein pathology changes and optogenetic fMRI activity can reveal pathofunction relationships. In this example, positive activity colocalizes with decreased pathology, while negative activity highly colocalizes with increased pathology. [Figure 26F] Montages of modulated alpha-synuclein lesions and ofMRI brain activity maps show a high degree of colocalization with opposite polarity. [Figure 27]Schematic diagram showing the clinical use of the present technology. Pathology in a subject is imaged. The technology estimates the dynamics of the pathology, including simulating the pathological spread throughout the brain. The technology is used to optimize neuromodulation therapy to treat the pathology. [Figure 28] Schematic of the technical approach. DETAILED DESCRIPTION OF THE INVENTION

[0084] Methods, systems, and devices, including computer programs encoded on computer storage media, are provided for optimizing neurostimulation therapy for the treatment of neurological and neurodegenerative diseases. In particular, algorithms are used to provide predicted regional pathological density maps of neuropathology and predict the location of future spread. Neurostimulation therapy parameters, including location, intensity, and frequency of neurostimulation, can be adjusted accordingly to treat the neuropathology and reduce spread.

[0085] Before the present methods, systems, and devices, including a computer program encoded on a computer storage medium, are described, it is to be understood that the present invention is not limited to the particular methods, systems, devices, or computer programs encoded on a computer storage medium described, as such may, of course, vary. It is also to be understood that the terminology used herein is for the purpose of describing particular embodiments only, and is not intended to be limiting, since the scope of the present invention will be limited only by the appended claims.

[0086] Where a range of values ​​is provided, it is understood that each intervening value between the upper and lower limit of that range is also specifically disclosed, to the tenth of the unit of the lower limit, unless the context clearly dictates otherwise. Each smaller range between any stated or intervening value in a stated range and any other stated or intervening value in that stated range is encompassed within the invention. The upper and lower limits of these smaller ranges may independently be included or excluded in the range, and each range where either limit is included in the smaller range, neither limit is included in the smaller range, or both limits are included in the smaller range is also encompassed within the invention, subject to any specifically excluded limit in the stated range. Where a stated range includes one or both of the limits, ranges excluding either or both of those included limits are also included in the invention.

[0087] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Although any methods and materials similar or equivalent to those described herein can be used in the practice or testing of the present invention, some potential and preferred methods and materials are described herein. All publications mentioned herein are incorporated by reference to disclose and describe the methods and / or materials related to the cited publications. It is understood that the present disclosure supersedes any disclosure of an incorporated publication to the extent there is a conflict.

[0088] As will be apparent to those skilled in the art upon reading this disclosure, each of the individual embodiments described and illustrated herein has distinct components and features which may be readily separated from or combined with the features of any of the other embodiments without departing from the scope or spirit of the invention. Any described method can be carried out in the order of events described or in any other order which is logically possible.

[0089] It should be noted that as used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the context clearly dictates otherwise. Thus, for example, a reference to "a neuron" includes a plurality of such neurons, a reference to "the measurement" includes one or more measurements and equivalents thereof known to those skilled in the art, and so forth.

[0090] The publications discussed herein are provided solely for their disclosure prior to the filing date of the present application. Nothing herein should be construed as an admission that the present invention is not entitled to antedate such publication by virtue of prior invention. Further, the dates of publication provided may be different from the actual publication dates which may need to be independently confirmed.

[0091] definition The term "about" is meant to encompass a deviation of plus or minus 5 percent, particularly with respect to a given quantity.

[0092] The terms "individual," "subject," "host," and "patient" are used interchangeably herein and refer to any subject with a brain, including, but not limited to, mammals, including invertebrates and arthropods (e.g., insects, crustaceans, arachnids), cephalopods (e.g., octopus, squid), amphibians (e.g., frogs, salamanders, caesars), fish, reptiles (e.g., turtles, crocodiles, snakes, amphibians, lizards, tuatara), chimpanzees, other ape, and monkey species, including humans and non-human mammals such as non-primates; laboratory animals such as mice, rats, rabbits, hamsters, guinea pigs, and chinchillas; farm animals such as dogs and cats; domestic animals such as sheep, goats, pigs, and cows; and birds, including chickens, turkeys, and other galliformes, ducks, and geese, including domestic, wild, and game birds. In some cases, the methods of the present invention are used in laboratory animals, in veterinary applications, and in the development of animal models for disease, including, but not limited to, rodents, including mice, rats, and hamsters, primates, and transgenic animals.

[0093] As used herein, the term "user" refers to a person who interacts with the devices and / or systems disclosed herein to perform one or more steps of the methods disclosed herein. The user may be a patient receiving treatment. The user may be a medical professional, such as the patient's physician.

[0094] The term "synucleinopathy" includes any disease associated with alpha-synuclein aggregation. This term includes neurodegenerative diseases associated with the pathological accumulation of alpha-synuclein aggregates in neurons or glia. Synucleinopathies include, but are not limited to, Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophies such as infantile neuroaxonal dystrophy and Hallervorden-Spatz syndrome, Shy-Drager syndrome, striatonigral degeneration, and olivopontocerebellar atrophy. This term also includes neurodegenerative diseases in which alpha-synuclein pathology contributes to the pathological progression of the disease, but is not the primary protein component of the disease-related pathology, such as Alzheimer's disease, amyotrophic lateral sclerosis, and frontotemporal dementia. Certain mutations cause alpha-synuclein to form amyloid-like fibrils that contribute to the pathogenesis of the disease. For example, mutations in alpha-synuclein, A53T, A30P, E46K, H50Q, and G51D, are associated with Parkinson's disease.

[0095] The terms "treatment," "treating," "treat," and the like are used herein generally to refer to obtaining a desired pharmacological and / or physiological effect. The effect can be preventative, in that a disease or its symptoms are completely or partially prevented, and / or therapeutic, in that a disease and / or adverse effects resulting from the disease are partially or completely stabilized or cured. The term "treatment" encompasses any treatment of disease in a mammal, particularly a human, and includes (a) preventing the appearance of a disease and / or symptom in a subject who may be predisposed to the disease or symptom but has not yet been diagnosed as having the disease or symptom, (b) inhibiting the disease and / or symptom, i.e., arresting its development, or (c) alleviating disease symptoms, i.e., causing regression of the disease and / or symptom. Those in need of treatment include those already affected (e.g., those with a synucleinopathy) as well as those in whom prevention is desired (e.g., those susceptible to developing a synucleinopathy, those with a genetic predisposition to developing a synucleinopathy, those suspected of having a synucleinopathy, etc.).

[0096] A therapeutic treatment is one that the subject is administered prior to administration, and a prophylactic treatment is one that the subject is suffering from prior to administration. In some embodiments, the subject has an increased likelihood of suffering from or is suspected of suffering from the disease prior to treatment. In some embodiments, the subject is suspected of having an increased likelihood of suffering from the disease.

[0097] As used herein, "neuronal activity" can refer to the electrical activity of neurons (e.g., changes in the membrane potential of neurons), as well as indirect measures of the electrical activity of one or more neurons. Thus, neural activity can refer to changes in field potential, changes in intracellular ion concentrations (e.g., intracellular calcium concentrations), and changes in magnetic resonance induced by neuronal electrical activity, as measured, for example, by blood oxygen level-dependent (BOLD) signal in functional magnetic resonance imaging.

[0098] As used herein, the term "survival" means the time from the start of treatment to death.

[0099] The term "animal" is used herein to include all vertebrates except humans. This term also includes animals in all stages of development, including embryonic, fetal, neonatal, and adult stages. Animals may include any member of the subphylum Chordata, including, but not limited to, non-human primates such as chimpanzees and other ape and monkey species; livestock such as cows, sheep, pigs, goats, and horses; domesticated mammals such as dogs and cats; laboratory animals including rodents such as mice, rats, and guinea pigs; and birds, including chickens, turkeys, and other galliformes, domesticated birds such as ducks, geese, wild birds, and game birds.

[0100] Systems and computer-implemented methods The present disclosure provides systems and computer-implemented methods for use in practicing the subject methods. In some embodiments, the system may include a processor programmed to predict where pathological protein aggregates will occur in the brain of a subject with a neurological or neurodegenerative disease, and a display component for displaying information regarding the predicted location of the pathological protein aggregates in the subject's brain. The system may also include one or more graphics boards for processing and outputting graphical information to the display component. For example, the display may be used to display an image of the subject's brain showing the current location of the pathological protein aggregates and / or the predicted past, present, or future location of the pathological protein aggregates as determined by the computer-implemented method.

[0101] In some embodiments, a computer-implemented method is used to predict the location of occurrence of pathological protein aggregates in the brain of a subject with a neurological or neurodegenerative disease, wherein the processor a) receives images of the subject's brain, b) identifies pathological protein aggregates in the images using a machine learning algorithm, c) maps the locations of the pathological protein aggregates to neuroanatomical regions, d) models the discretized distribution of pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model, e) calculates the distribution of gene effects on the regional spread and decay of the pathological protein aggregates using the regional gene density map of each neuroanatomical region, and f) calculates the distribution of gene effects on the regional spread and decay of the pathological protein aggregates across the neuroanatomical regions. and g) predicting past, current, and future locations of pathological protein aggregates based on the modeling.

[0102] In certain embodiments, the computer-implemented method further includes adjusting one or more programmed neural stimulation parameters based on the predicted change in regional density of pathological protein aggregates as a function of time. For example, the duration, amplitude, frequency, pulse width, and location of the neural stimulation, or any combination thereof, may be adjusted.

[0103] In certain embodiments, the computer-implemented method further includes instructing the neurostimulation device to apply neurostimulation to brain locations predicted to have pathological protein aggregates present to treat a neurological or neurodegenerative disease in the subject.

[0104] In certain embodiments, the computer-implemented method further includes instructing the neurostimulator device to apply electrical stimulation to brain locations where pathological protein aggregates are not yet present but are predicted to appear in the future based on predicting changes in regional density of pathological protein aggregates as a function of time.

[0105] In certain embodiments, the method includes performing image registration to a coordinate space comprising a plurality of voxels, each voxel being represented as a cubic volume element centered at a coordinate in the coordinate space; identifying a location in x, y, z coordinates of each pathological protein aggregate in the coordinate space; measuring a volume of each pathological protein aggregate from a total number of voxels occupied by each pathological protein aggregate; and calculating an aggregate density for each voxel, the aggregate density for each voxel being determined from a total number of pathological protein aggregates centered within the same voxel. and calculating a total aggregate size for each voxel, the total aggregate size being the total size of all pathological protein aggregates centered within the same voxel; calculating an average aggregate size for each voxel as the total aggregate size for each voxel divided by the aggregate density for each voxel; calculating a total signal intensity for each voxel from the total intensities of all pathological protein aggregates centered within the same voxel; and calculating an average signal intensity for each voxel as the total signal intensity for each voxel divided by the aggregate density.

[0106] In certain embodiments, modeling the discretized distribution of pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model includes:

[0107]

number

[0108] where c i,j where σ represents the total count of pathological protein aggregates in a discretized size bin indexed by I in the brain region indexed by j, the L matrix represents the Laplacian matrix of the weighted directed graph connecting the neuroanatomical regions of the brain, η is chosen as a hyperparameter that slows down the spread of large aggregates as an inverse power of their size, and λ is chosen as a hyperparameter that accelerates the collapse of pathological protein aggregates proportionally to a power of their size.

[0109] In certain embodiments, the initial values ​​of α and μ are adapted by sweeping through a two-dimensional grid and selecting the values ​​that result in the smallest mean square error between the predicted and actual counts of pathological protein aggregates.

[0110] In a particular embodiment, the initial value of μ=0, the initial value of k=0, ξ=0, and c is a one-dimensional size vector, and the set of differential equations simplifies to a standard network diffusion model.

[0111] In certain embodiments, the computer-implemented method further includes quantifying the sensitivity of the model to a particular neuroanatomical pathway in the brain, wherein a Jacobian matrix is ​​calculated by taking the partial derivative of the output of the model with respect to weights of anatomical coupling strengths between two neuroanatomical regions encoded in the model, and the elements of the Jacobian matrix represent the relative contribution of the anatomical coupling between the two neuroanatomical regions to the spread of pathological protein aggregates to a particular region of the brain.

[0112] In certain embodiments, the computer-implemented method further includes using the model to generate a ranking of candidate seed locations for a given pathological state c at t=T months post injection (MPI) by a method that includes: using each of the neuroanatomical regions simulated forward in time until t=T MPI as a separate seed location at t=0 in the model, where each of the simulation results for the different neuroanatomical regions is compared to the observed state c using a pairwise similarity metric, where the similarity metric is a correlation coefficient between total regional aggregate counts across the observed and simulated states; and using the similarity metric value to sort seed locations for the neuroanatomical regions as likely sites leading to the observed pathological state c, whereby a ranking of candidate seed locations for the given pathological state c at t=T MPI is generated.

[0113] In certain embodiments, the computer-implemented method further includes predicting the MPI for a time t=T since seeding for a given pathological state c by a method that includes: comparing the whole-brain distribution of aggregate sizes for state c to simulated distributions at various t using a pairwise similarity metric without considering seed locations, where the distribution of simulated aggregate sizes across the whole brain is assumed to be invariant with respect to which neuroanatomical region is used as the seed location at t=0; and calculating the mean squared error between the stimulus distribution and the observed distribution, where, when deciding among multiple candidate t values, the mean squared errors are inverted and normalized to sum to one to provide a predicted probability for each t being a correct estimate of T for the given pathological state c.

[0114] In certain embodiments, gene effects on regional spread and decay of pathological protein aggregates are determined by a method that includes: assuming that the α (spread) and μ (decay) parameters are region-dependent, such that spread from a particular neuroanatomical region is proportional to the gene density in that region; normalizing all genes to the same range so that only the regional distribution of gene expression is compared to the whole-brain expression of that gene, where α is a vector and the product of α and the Laplacian connectivity matrix L has the effect of correcting the regional connectivity encoded in the model; and normalizing each gene vector to have a mean of 1 and a standard deviation Σ empirically set to preserve the correlation between predicted and observed whole-brain aggregate counts, where the normalization is chosen so that the product has the effect of maintaining a trace of the original Laplacian connectivity matrix L. In some embodiments, a derivation of normalization to preserve the trace of the original Laplacian connectivity matrix L is to assume that the vector s is sampled from a multivariate normal distribution with mean 1 and standard deviation Σ, where s ∼N(1,Σ), and use the definition of matrix trace to represent s as a diagonal square matrix S, where the trace of the product of S and the Laplacian connectivity matrix L is

[0115]

number

[0116] where l represents the diagonal of L, the trace is equal to the dot product of s and l, and has expectation equal to the sum of the entries of l, s·l~N(1·l,lΣl) E[s·l]=Tr(L) The method involves recovering the definition of the trace of L according to the formula: After each gene is coded into the model, comparing the net effect on regional correlation between the simulated and actual data with a baseline correlation that does not include the gene, and providing an ordered list of genes ranked by the relevance of their spatial expression map in improving the model's regional predictions.

[0117] In a particular embodiment, the cubic volume element has a width in coordinate space of 100 μm.

[0118] In certain embodiments, one or more pathological protein aggregates are mapped to a single voxel.

[0119] In certain embodiments, the computer-implemented method further includes performing multi-dimensional Gaussian filtering to account for variations in image registration between different samples.

[0120] In certain embodiments, the computer-implemented method further includes segmenting the image to generate a plurality of image segments.

[0121] In certain embodiments, the locations of the pathological protein aggregates are mapped to neuroanatomical regions of the Allen Human Brain Reference Atlas. In some embodiments, the mapping comprises performing image registration to transform the locations of the pathological protein aggregates into the coordinate space of the Allen Human Brain Reference Atlas. In some embodiments, anatomically interconnected neuroanatomical regions are identified from the Allen Connectivity Atlas.

[0122] In certain embodiments, the neuroanatomical region is anterior amygdala region, anterior cingulate region, dorsal, anterior cingulate region, ventral, nucleus accumbens, anterior dorsal nucleus, anterior hypothalamic nucleus, agranular insular region, dorsal, agranular insular region, posterior, agranular insular region, ventral, nucleus ambiguus, anterior medial nucleus, dorsal, anterior medial nucleus, ventral, oblique lobule, accessory olfactory bulb, anterior olfactory nucleus, anterior pretectal nucleus, arcuate hypothalamic nucleus, dorsal auditory cortex, primary auditory cortex, ventral auditory cortex, anterior ventral nucleus of the thalamus, basolateral amygdala, basomedial amygdala, bed nucleus of the stria terminalis, CA1 area, CA2 area, CA3 area, central nucleus of the amygdala, central lobule, lateral central nucleus of the thalamus, claustrum, central linear raphe, medial central nucleus of the thalamus, cortical amygdala region, anterior, cortical Substantia amygdaloidea, posterior portion, caudate nucleus, superior central raphe, anterior cerebellum, cuneiform nucleus, dorsal cochlear nucleus, dentate gyrus, dorsomedial nucleus of the hypothalamus, dentate nucleus, dorsal peduncle region, dorsal raphe nucleus, rhinal region, entorhinal region, lateral portion, entorhinal region, medial portion, dorsal zone, caudate nucleus, dorsal portion, caudate nucleus, ventral portion, flocculus, dorsal cerebellar nucleus, anterior column, cerebral cortex, basal striatum, globus pallidus, external segment, globus pallidus, internal segment, nucleus reticularis magnocellularis, gustatory region, intercalated amygdaloid nucleus, inferior colliculus, central nucleus, inferior colliculus, dorsal nucleus, inferior colliculus, external nucleus, infralimbic region, intermediate dorsal nucleus of the thalamus, inferior olivary complex, nucleus intermedius, interpeduncular nucleus, intermediate reticular nucleus, lateral amygdaloid nucleus, lateral vestibular nucleus, lateral dorsal nucleus of the thalamus, lateral geniculate complex dorsal part of the lateral geniculate complex, ventral part of the lateral habenula, lateral area of ​​the hypothalamus, lateral postoptic nucleus of the thalamus, lateral preoptic area, lateral reticular nucleus, lateral septal nucleus, caudal part, lateral septal nucleus, rostral part, lateral septal nucleus, abdomen, magnocellular nucleus, magnocellular reticular nucleus, medial dorsal nucleus of the thalamus, medullary reticular nucleus, dorsal part, medullary reticular nucleus , abdomen, medial amygdala nucleus, median preoptic nucleus, medial geniculate complex, dorsal part, medial geniculate complex, medial part, medial geniculate complex, abdomen, medial habenula, medial mammillary nucleus, main olfactory bulb, primary motor area, secondary motor area, medial preoptic nucleus, medial preoptic area, medial pretectal area, mesencephalic reticular nucleus, medial septal nucleus, medial vestibular nucleus, diagonal zone nucleus, Nucleus indeterminate, nucleus of the lateral lemniscus, nucleus of the lateral olfactory tract, tubercle (X), nucleus of the optic tract, nucleus of the posterior commissure, nucleus of the solitary tract, orbital region, lateral part, orbital region, medial part, orbital region, ventrolateral part, olfactory tubercle, posterior amygdaloid nucleus, piriform amygdaloid region, periaqueductal gray matter, paraunculoid gyrus, parvocellular reticular nucleus, parabrachial nucleus, central pontine gray matter, perirhinal region, parafascicular nucleus, paraflocculus lobule, pontine gray matter, paracellular reticular nucleus, dorsal part, paracellular reticular nucleus, lateral part, posterior nucleus of the hypothalamus, piriform region, prelimbic area, dorsal premammillary nucleus, posterior thalamic complex, posterior bordering nucleus of the thalamus, posterior uncinate gyrus, peripeduncular nucleus, pedunculopontine nucleus, preuncleate gyrus, paramedian lobule, pontine reticular nucleus, caudal part, pontine reticular nucleus, precursor nucleus, principal sensory nucleus of the trigeminal nerve,Parathalamic nucleus, posterior parietal related area, paraventricular hypothalamic nucleus, paraventricular thalamic nucleus, periventricular hypothalamic nucleus, posterior, periventricular hypothalamic nucleus, preoptic area, lobule VIII, postchiasmatic area, nucleus of reuniens, rhomboid nucleus, nucleus raphe magnus, nucleus red, midbrain reticular nucleus, posterior ruber area, retrosplenial area, lateral agranular area, retrosplenial area, dorsal, retrosplenial area, ventral, thalamic reticular nucleus, Paraventricular region, superior colliculus, motor-related, superior colliculus, sensory-related, septal fimbria, substantia innominata, lobule simplex, medial subthalamic nucleus, substantia nigra, pars compacta, substantia nigra, pars reticulata, superior olivary complex, subparafascicular region, subparafascicular nucleus, magnocellular part, subparafascicular nucleus, parvocellular part, spinal vestibular nucleus, spinal nucleus of the trigeminal nerve, tail, spinal nucleus of the trigeminal nerve, interpolar part, spinal nucleus of the trigeminal nerve, oral part, primary somatic sensory Selected from the sensory cortex, barrel area, primary somatosensory cortex, lower limbs, primary somatosensory cortex, mouth, primary somatosensory cortex, nose, primary somatosensory cortex, trunk, primary somatosensory cortex, upper limbs, additional somatosensory areas, subthalamic nucleus, uncinate gyrus, supramammillary nucleus, supratrigeminal nucleus, superior vestibular nucleus, temporal association area, posterior piriform transition area, tegmental reticular nucleus, triangular septal nucleus, stria tecta, tubercular nucleus, motor nucleus of the trigeminal nerve, ventral anterolateral complex of the thalamus, ventral cochlear nucleus, facial motor nucleus, visceral area, anterolateral visual area, anteromedial visual area, lateral visual field, primary visual area, posterolateral visual area, posteromedial visual area, ventromedial nucleus of the thalamus, ventromedial nucleus of the hypothalamus, ventral posterolateral nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, parvocellular portion, ventral tegmental area, and hypoglossal nucleus.

[0123] In certain embodiments, the computer-implemented method further comprises predicting where the pathological protein aggregates occurred in the subject's brain.

[0124] In certain embodiments, the subject is a human subject, and modeling the spread, aggregation, and disintegration of pathological protein aggregates in the brain of the human subject uses simulations generated based on measuring the spread, aggregation, and disintegration of pathological protein aggregates in a non-human animal model, and the experimentally measured pathology in the non-human animal model is used to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human subject. In certain embodiments, the simulation further uses brain anatomy or brain function measured from the non-human animal to predict the past, current, and future locations of pathological protein aggregates in the human subject.

[0125] In some embodiments, the non-human animals used to develop the simulation are non-human mammals, including, but not limited to, non-human primates (including chimpanzees and other ape and monkey species); laboratory animals (e.g., mice, rats, rabbits, hamsters, guinea pigs, and chinchillas); livestock (e.g., dogs and cats); farm animals (e.g., sheep, goats, pigs, horses, and cattle); and birds (e.g., poultry, wild birds, and game birds, including chickens, turkeys, and other galliformes, ducks, and geese).

[0126] In certain embodiments, the simulation further uses measured brain anatomy or brain function from other human subjects to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human patient. In some embodiments, pathological protein aggregates in the brains of other human subjects are monitored to provide experimental data regarding the spread, aggregation, and decay of pathological protein aggregates over time, which is used to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human patient. Pathological protein aggregates can be monitored using any suitable medical imaging technique, such as, but not limited to, imaging performed using computed tomography (CT), single-photon emission computed tomography (SPECT), magnetic resonance imaging, functional magnetic resonance imaging, optogenetic functional magnetic resonance imaging, and positron emission tomography (PET).

[0127] In certain embodiments, the computer-implemented method further includes receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; and displaying a change in total aggregate size for each voxel, the volume of each pathological protein aggregate for each voxel, and the aggregate density for each voxel in the image taken after neurostimulation is applied to the subject's brain compared to the image taken before neurostimulation is applied to the subject's brain.

[0128] In certain embodiments, the computer-implemented method further includes receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; modulating one or more programmed neurostimulation parameters based on any changes in past, current, or future locations where pathological protein aggregates are predicted to appear; and instructing the neurostimulation device to apply the modulated neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease in the subject.

[0129] In certain embodiments, the computer-implemented method further includes storing a user profile for the subject, the user profile including information regarding programmed neurostimulation parameters to be used to apply neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease based on the location where the computer-implemented method predicts that pathological protein aggregates exist or will occur in the future.

[0130] Analyzing and identifying pathological protein aggregates in brain images may involve the use of algorithms or classifiers. In certain embodiments, a machine learning algorithm is used to identify pathological protein aggregates in brain images. The machine learning algorithm may include a supervised learning algorithm. Examples of supervised learning algorithms may include average-order dependence estimators (AODEs), artificial neural networks (e.g., backpropagation), Bayesian statistics (e.g., Bayesian classifiers, Bayesian networks, Bayesian knowledge bases), case-based reasoning, decision trees, inductive logic programming, Gaussian process regression, group methods of data processing (GMDH), learning automata, learning vector quantization, minimum message length (decision trees, decision graphs, etc.), lazy learning, instance-based learning nearest neighbor algorithms, analogical modeling, probabilistic approximately correct learning (PAC) learning, ripple-down rules, knowledge acquisition methodologies, symbolic machine learning algorithms, sub-symbolic machine learning algorithms, support vector machines (SVMs), random forests, ensembles of classifiers, bootstrap aggregating (bagging), and boosting. Supervised learning may include ordinal classification such as regression analysis and informative fuzzy networks (IFNs). Alternatively, supervised learning methods may include statistical classification such as AODEs, linear classifiers (e.g., Fisher linear discriminant, logistic regression, naive Bayes classifiers, perceptrons, and support vector machines), quadratic classifiers, k-nearest neighbors, boosting, decision trees (e.g., C4.5, random forests), Bayesian networks, and hidden Markov models.

[0131] Machine learning algorithms may also include unsupervised learning algorithms. Examples of unsupervised learning algorithms may include artificial neural networks (recurrent or convolutional), data clustering, expectation-maximization algorithms, self-organizing maps, radial basis function networks, vector quantization, generative terrain maps, information bottleneck methods, and IBSEAD. Unsupervised learning may also include association rule learning algorithms such as the Apriori algorithm, the Eclat algorithm, and the FP-growing algorithm. Hierarchical clustering, such as single-linkage clustering and conceptual clustering, may also be used. Alternatively, unsupervised learning may include divisive clustering, such as the K-means algorithm and fuzzy clustering.

[0132] In some cases, the machine learning algorithm includes a reinforcement learning algorithm. Examples of reinforcement learning algorithms include, but are not limited to, temporal difference learning, Q-learning, and learning automata. Alternatively, the machine learning algorithm may include data preprocessing.

[0133] In some embodiments, the machine learning algorithm uses an artificial neural network. In some embodiments, the machine learning algorithm uses a deep learning algorithm, which may include the use of a convolutional neural network, a deep neural network, a recurrent neural network, a deep residual neural network, a long short-term memory network, a deep belief network, a multilayer perceptron, or deep reinforcement learning. For descriptions of deep learning algorithms, see, for example, Pedrycz et al. Deep Learning: Algorithms and Applications (Studies in Computational Intelligence Book 865, Springer, 2019), Goodfellow et al. Deep Learning (Adaptive Computation and Machine Learning series, The MIT Press, 2016), and Various Deep Learning Algorithms in Computational Intelligence (edited by Oscar Humberto Montiel Ross, Mdpi AG, 2023), which are incorporated herein by reference in their entireties.

[0134] In certain embodiments, the computer-implemented method further includes segmenting the image to generate a plurality of image segments. Any suitable method known in the art can be used for image segmentation to facilitate identification of pathological protein aggregates. Automatic or semi-automatic image analysis methods can be used for image segmentation. Various factors can complicate image analysis, including noise, autofluorescence, low resolution, blur, unstable brightness, overlapping targets, unclear boundaries, deformations, etc. In some cases, human intervention may be required to accurately identify pathological protein aggregates in the image. If a classifier is insufficient to accurately automatically identify a single pathological protein aggregate, a human can outline at least some of the pathological protein aggregates in the image to generate a set of pathological protein aggregates that can be used to train a machine learning algorithm. A variety of software programs are currently available for image segmentation, including, but not limited to, the Ilastik Toolkit, which uses a random forest classifier for cell segmentation; DeepCell, which uses a deep learning algorithm that utilizes deep convolutional neural networks for image segmentation; Open Segmentation Framework (OpSeF), which semi-automates image segmentation using a deep learning convolutional neural network for which the user manually provides some training data; CellSeg, which uses a masked region convolutional neural network (R-CNN) for image segmentation; CODEX image processing pipeline software, which uses a reference cell marker, a reference nuclear stain, and a reference membrane stain to assist in image segmentation; and CellProfiler, which uses conventional thresholding for image segmentation and classifies pixels as foreground if they are brighter than a certain "threshold" intensity value (cells appear as bright objects on a dark background in fluorescence microscopy images), illumination correction, declustering, and watershed segmentation.For descriptions of image segmentation techniques and software, see, e.g., Kreshuk et al. (2019) Methods Mol. Biol. 2040:449-463, Kreshuk et al. (2014) PLoS One 9(2):e87351, David A. Van Valen et al. (2016) PLoS Comput. Biol. 12(11):e1005177, Dobson et al. (2021) Curr. Protoc. 1(5):e89, Stirling et al. (2021) BMC Bioinformatics 22(1):433, Soliman (2015) Biol Proced Online 17:11, Schapiro et al. (2017) Nat. Methods 14:873-876, Ljosa et al. (2009) PLoS See Comput. Biol. 5(12):e1000603, and Lee et al. (2022) BMC Bioinformatics 23(1):46, which are incorporated herein by reference.

[0135] Additional relevant clinic metrics may also be stored within the system, including, but not limited to, the subject's age, height, body mass index (BMI), sex, weight, and / or diagnosis, or other metrics applicable to the present technology. In some cases, such metrics may be obtained from the subject's electronic medical record (EMR) or another applicable cloud-based storage technology, or in other cases, may be measured and then stored in the subject's electronic medical record or another applicable cloud-based storage technology.

[0136] The methods can be implemented in digital electronic circuitry, or in computer software, firmware, or hardware. The disclosed and other embodiments may be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer-readable medium for execution by, or to control the operation of, a data processing apparatus. The computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, a composition of matter providing a machine-readable propagated signal, or any combination thereof.

[0137] A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and can be deployed in any form, such as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program can be stored in part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program in question, or in multiple cooperating files (e.g., files storing one or more modules, subprograms, or portions of code). A computer program can be deployed to be executed on one computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communications network.

[0138] In further aspects, a system for performing the computer-implemented methods as described may include a processor, a storage component (i.e., memory), a display component, and other components typically present in a general-purpose computer. In some embodiments, the processor is provided by a computer or a handheld device (e.g., a cell phone or tablet). The storage component stores information accessible by the processor, including instructions that can be executed by the processor and data that can be retrieved, manipulated, or stored by the processor.

[0139] The storage component includes instructions. For example, the storage component includes instructions for determining predicted locations of occurrence of pathological protein aggregates in the brain of a subject having a neurological or neurodegenerative disease according to methods described herein. A computer processor is coupled to the storage component and configured to execute the instructions stored in the storage component to receive images of the subject's brain and analyze the images according to one or more algorithms as described herein to predict past, current, and future locations of pathological protein aggregates in the subject's brain.

[0140] The processor and / or memory may be operatively connected to a display device, for example, via a wired connection, such as a universal serial bus (USB) connection, or a wireless connection, such as a Bluetooth connection. Any convenient display device can be used, such as a liquid crystal display (LCD), a light-emitting diode (LED) display, a plasma (PDP) display, a quantum dot (QLED) display, or a cathode ray tube display device. The display component displays information regarding the location of pathological protein aggregates within the subject's brain. In certain embodiments, the display displays an image of the subject's brain showing the predicted current, past, or future locations of pathological protein aggregates determined by the computer-implemented method. In some embodiments, the display displays information regarding the coordinates of each pathological protein aggregate, the volume of each pathological protein aggregate, the aggregate density of each voxel, the total aggregate size of each voxel, the average aggregate size of each voxel, the total signal intensity of each voxel, the average signal intensity of each voxel, or a mapping of the locations of pathological protein aggregates to neuroanatomical regions, or any combination thereof. In some embodiments, the display displays information regarding the distribution of gene effects on the regional spread and disintegration of pathological protein aggregates. In some embodiments, the display displays information regarding the spread, aggregation, disintegration, and predicted changes in the regional density of pathological protein aggregates as a function of time determined by modeling spatial gene expression across neuroanatomical regions. In some embodiments, the display displays information regarding predicted past, current, and future locations of pathological protein aggregates based on the modeling.

[0141] The storage component may be of any type capable of storing information accessible by a processor, such as a hard drive, memory card, ROM, RAM, DVD, CD-ROM, USB flash drive, writable memory, and read-only memory. The processor may be a general-purpose processor, a graphics processor unit, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof designed to perform the functions described herein. A general-purpose processor may be a microprocessor, but alternatively, the processor may be a controller, microcontroller, or state machine, combinations thereof, or the like. A processor may also be implemented as a combination of computing devices, e.g., a combination of a DSP and a microprocessor, multiple microprocessors, one or more microprocessors in conjunction with a DSP core, or any other such configuration. While described herein primarily with reference to digital technology, a processor may also include primarily analog components. The computing environment may include any type of computer system, including, but not limited to, computer systems based on computational engines in microprocessors, graphics processor units, mainframe computers, digital signal processors, portable computing devices, personal organizers, device controllers, and appliances, to name a few.

[0142] The steps of a method, process, or algorithm described in connection with the embodiments disclosed herein may be embodied directly in hardware, in a software module executed by a processor, or a combination of the two. The software modules, engines, and associated databases may reside in memory resources such as RAM memory, FRAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, a hard disk, a removable disk, a CD-ROM, or any other form of non-transitory computer-readable storage medium, medium, or physical computer storage known in the art. An exemplary storage medium may be coupled to the processor such that the processor can read information from, and write information to, the storage medium. Alternatively, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC. The ASIC may reside in a user terminal. Alternatively, the processor and the storage medium may reside as discrete components in a user terminal.

[0143] Instructions may be any set of instructions to be executed by a processor directly (such as machine code) or indirectly (such as script). In that regard, the terms "instructions," "steps," and "program" may be used interchangeably herein. Instructions may be stored in object code form for direct processing by a processor, or in any other computer language, including a script or a collection of independent source code modules that are interpreted on demand or pre-compiled.

[0144] Data may be retrieved, stored, or modified by a processor in accordance with instructions. For example, the system is not limited by any particular data structure, but data may be stored in a computer register, in a relational database as a table with multiple different fields and records, in an XML document, or in a flat file. Data may also be formatted in any computer-readable format, such as, but not limited to, binary values, ASCII, or Unicode. Furthermore, data may include any information sufficient to identify related information, such as numbers, descriptive text, unique codes, pointers, references to data stored in other memory (including other network locations), or information used by a function to calculate related data.

[0145] In particular embodiments, a processor and storage component may comprise multiple processors and storage components, which may or may not be stored within the same physical enclosure. For example, some of the instructions and data may be stored on a removable CD-ROM, while others may be stored in a read-only computer chip. Some or all of the instructions and data may be stored in a location that is physically remote from the processor, but still accessible. Similarly, a processor may comprise a collection of processors that may or may not operate in parallel.

[0146] In some embodiments, the method may be performed using a cloud computing system, in which images of the subject's brain may be exported to a cloud computer, which runs the program and returns the output to the user.

[0147] Components of systems for carrying out the methods of the present disclosure are further described in the Examples below.

[0148] Neuromodulation Therapies for the Treatment of Neurological and Neurodegenerative Diseases In some embodiments, the subject methods are used to treat synucleinopathies, which may include any disease associated with alpha-synuclein aggregation. Pathological aggregates of alpha-synuclein may accumulate, for example, in neurons or glia. Synucleinopathies include, but are not limited to, Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophies such as infantile neuroaxonal dystrophy and Hallervorden-Spatz syndrome, Shy-Drager syndrome, striatonigral degeneration, and olivopontocerebellar atrophy. Furthermore, the subject methods can be used to treat neurodegenerative diseases, such as Alzheimer's disease, amyotrophic lateral sclerosis, and frontotemporal dementia, in which alpha-synuclein pathology contributes to the pathological progression of the disease but is not the primary protein component of the disease-associated pathology. Certain mutations cause alpha-synuclein to form amyloid-like fibrils that contribute to the pathogenesis of the disease. For example, mutations in alpha-synuclein, A53T, A30P, E46K, H50Q, and G51D, are associated with Parkinson's disease.

[0149] Neuromodulation can be achieved using electrical stimulation (e.g., from implanted clinically approved electrodes), transcranial magnetic stimulation, transcranial electrical stimulation, or focused ultrasound, among other techniques. For electrical stimulation, a deep brain stimulator can be used. Cortical layers or cell types can be specifically targeted using genetically encodable modulation techniques such as optogenetics. Deep brain stimulation (DBS), transcranial magnetic stimulation, transcranial electrical stimulation, or optogenetics can be used for neuromodulation of specific neuronal cell types or neural circuits in the brain at locations where pathological protein aggregates are currently, previously, or are predicted to appear in the future. For example, parameters such as the duration and location of neuromodulation can be tailored to the patient based on the patient's current pathological state and the predicted impact of the neuromodulation parameters on the pathology. Anticipated future pathology can also be taken into consideration when selecting neuromodulation parameters. One or more neurostimulation therapy parameters, including but not limited to the location, intensity, and frequency of neurostimulation, can be adjusted accordingly to treat neuropathology and reduce aggregation and spread. Methods of neuromodulation are described in further detail below.

[0150] Electrical stimulation In certain embodiments, electrical stimulation is applied to the subject's brain using electrodes. The method includes placing electrodes in regions of the subject's brain to deliver electrical stimulation to the brain to disrupt the aggregation and / or spread of pathological protein aggregates and prevent or slow disease progression. The electrodes can be non-penetrating surface electrodes, extracranial electrodes, e.g., subcranially or cranially mounted electrodes (in the case of a burr hole cap or skull-mounted neurostimulator), or brain-penetrating deep electrodes. Electrical stimulation can be applied to the brain using the electrodes in a manner effective to treat neurological or neurodegenerative diseases.

[0151] As used herein, the phrase "an electrode" or "the electrode" refers to a single electrode or multiple electrodes, such as an electrode array. As used herein, the term "contact," as used in the context of an electrode in contact with a region of the brain, refers to a physical association between the electrode and the region. The electrode is capable of conducting electricity to a specific target within the brain. Electrodes used in the methods disclosed herein can be unipolar (cathode or anode) or bipolar (e.g., having an anode and a cathode).

[0152] Positioning the electrode may be performed using standard surgical procedures for placing intracranial electrodes. In certain cases, placing the electrode may involve positioning the electrode on the surface of a designated region of the brain where pathological protein aggregates are known to appear (e.g., based on medical imaging) or are predicted to appear now, in the past, or in the future. The electrode may contact at least a portion of the brain surface in the designated region. In some embodiments, the electrode may contact substantially the entire surface area in the designated region. In some embodiments, the electrode may also contact a region adjacent to the designated region.

[0153] In some embodiments, an electrode array disposed on a flat support substrate can be used. The surface area of ​​the electrode array can be determined by the desired contact area between the electrode array and the brain. Electrodes for implantation on the brain surface, such as surface electrodes or surface electrode arrays, can be obtained from commercial sources. Commercially obtained electrodes / electrode arrays can be modified to achieve the desired contact area. In some cases, non-penetrating brain electrodes (also referred to as surface electrodes) that can be used in the methods disclosed herein can be electrocorticography (ECoG) electrodes, subcortical electrodes, or electroencephalography (EEG) electrodes. In certain embodiments, multiple electrodes are positioned in one or more designated brain regions.

[0154] In certain cases, placing an electrode in a target region or site may involve positioning a penetrating electrode (also referred to as a deep electrode) within a designated region of the brain. For example, an electrode may be placed in an area of ​​the brain where pathological protein aggregates are known to appear (e.g., based on medical imaging) or are predicted to appear now, in the past, or in the future. In some embodiments, the electrode may also contact an area adjacent to the particular region of the brain. In some embodiments, one or more electrodes or electrode arrays are used to target one or more regions of the brain where pathological protein aggregates are predicted to appear now, in the past, or in the future.

[0155] The depth to which the electrodes are inserted into the brain can be determined by the desired level of contact between the electrode array and the brain. Penetrating brain electrode arrays can be obtained from commercial suppliers. Commercially available electrode arrays can be modified to achieve a desired insertion depth into brain tissue.

[0156] Positioning of electrodes to deliver electrical stimulation to the brain can be performed using standard surgical procedures for placing electrodes for deep brain stimulation. For example, electrodes can be placed in target areas of the brain where pathological protein aggregates are known to appear (e.g., based on medical imaging) or are predicted to appear now, in the past, or in the future. Medical imaging, for example, using magnetic resonance imaging (MRI) or computed tomography (CT), can be used to provide guidance for DBS electrode placement and verify correct placement of the electrodes in the brain. In addition, a neurostimulator that generates electrical pulses is placed under the skin of the chest, typically under the collarbone or in the abdomen. In some embodiments, the neurostimulator is attached to the skull. The surgical procedure can include placing electrodes in the brain through a small hole in the skull. Electrode leads are tunneled under the skin down the neck and under the skin of the chest to connect to a neurostimulator implanted in the chest.

[0157] Electrical current is delivered to the electrodes by a neurostimulator. Parameters such as pulse width, shape, frequency, amplitude, pattern, and temporal distribution can be adjusted to modulate neural activity and neural circuits, reducing or preventing the aggregation and spread of pathological protein aggregates to treat neurological or neurodegenerative diseases. In some embodiments, a closed-loop system is used to automatically adjust DBS settings in response to changes in the predicted location of pathological protein aggregates. In other embodiments, an open-loop system is used in which DBS settings are adjusted by a user or physician based on the predicted location of pathological protein aggregates.

[0158] Electrical stimulation can be applied using a single electrode, an electrode pair, or an electrode array. In some embodiments, the number of electrodes used to deliver electrical stimulation to the brain ranges from 8 to 32, including any number within this range, for example, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, or 32 electrodes. In some embodiments, electrical stimulation is applied to two or more sites. The sites to which electrical stimulation is applied can be alternated or otherwise spatially or temporally patterned. Electrical stimulation can be applied to these sites simultaneously or sequentially. The sites selected for stimulation can vary for different subjects and depend on the locations where pathological protein aggregates are known to be present (e.g., by medical imaging) or are predicted to be present.

[0159] In some embodiments, an electrode array arranged on a flat support substrate can be used to electrically stimulate a region of the brain. The surface area of ​​the electrode array can be determined by the desired contact area between the electrode array and the brain. In some cases, cylindrical, paddle, or plate-shaped electrode arrays can be used in the methods disclosed herein for deep brain stimulation. Such electrode arrays for implantation in the brain can be obtained from commercial suppliers. Commercially obtained electrodes / electrode arrays can be modified to achieve the desired contact area.

[0160] The exact number of electrodes included in an electrode array (e.g., for electrical stimulation) can vary. In certain embodiments, an electrode array can include two or more electrodes, such as three or more, including four or more, e.g., about 3-6 electrodes, about 6-12 electrodes, about 12-18 electrodes, about 18-24 electrodes, about 24-30 electrodes, about 30-48 electrodes, about 48-72 electrodes, about 72-96 electrodes, or about 96 or more electrodes. The electrodes can be arranged in a regular, repeating pattern (e.g., a grid, such as a grid with about 1 cm spacing between electrodes), or there can be no pattern. Electrodes that conform to the target site for optimal delivery of electrical stimulation can be used. One such example is a single multi-contact electrode with eight contacts separated by 2.5 mm intervals. Each contact has a span of about 2 mm. Another example is an electrode with two 1 cm contacts with a 2 mm intervening gap. Yet another example of an electrode that can be used in the present method is a bifurcated or trifurcated electrode to cover the target site. Each of these trifurcated electrodes has four 1-2 mm contacts, spaced 2.5 mm apart, with a span of 1.5 mm.

[0161] The size of each electrode can also vary depending on factors such as the number of electrodes in the array, the location of the electrodes, the material, the age of the patient, and other factors. In certain embodiments, the electrode array has a size (e.g., diameter) of about 5 mm or less, such as about 4 mm or less, including 4 mm to 0.25 mm, 3 mm to 0.25 mm, 2 mm to 0.25 mm, 1 mm to 0.25 mm, or about 3 mm, about 2 mm, about 1 mm, about 0.5 mm, or about 0.25 mm.

[0162] In certain embodiments, the method further comprises mapping the subject's brain to optimize electrode positioning for applying electrical stimulation. The electrode positioning is optimized to maximize clinical response to electrical stimulation for treating a neurological or neurodegenerative disease, which may include a synucleinopathy. In some embodiments, DBS is optimized to achieve a neurophysiologically defined change, e.g., reduction of alpha-synuclein aggregation, spread of alpha-synuclein aggregates, and / or improvement of brain function.

[0163] Evaluating the effectiveness of electrical stimulation at a specific site to treat a neurological or neurodegenerative disease can be performed using any standard method. In some embodiments, the effectiveness of electrical stimulation is evaluated by imaging the subject's brain to measure the size and identify the location of pathological protein aggregates after neural stimulation. In some embodiments, the effectiveness of electrical stimulation is evaluated by measuring the subject's brain function after neural stimulation. For example, measuring brain function can be measured by performing electroencephalography (EEG), stereoelectroencephalography (sEEG), electrocorticography (ECoG), magnetoencephalography (MEG), single-photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), optogenetic functional magnetic resonance imaging, or positron emission tomography (PET).

[0164] In some cases, the severity of symptoms of a neurological or neurodegenerative disease may be further assessed using a visual analog scale or a verbal rating scale. In certain embodiments, the method further comprises assessing one or more motor and / or non-motor symptoms of the subject using the Movement Disorder Society-sponsored Unified Parkinson's Disease Rating Scale-Revised (MDS-UPDRS), the Hoehn and Yahr (HnY) scale, or the Montreal Cognitive Assessment (MoCA) scale.

[0165] As described herein, the subject methods involve applying electrical stimulation to regions of the brain where pathological protein aggregates are known to or predicted to appear, currently, in the past, or in the future. Parameters for applying electrical stimulation to the brain can be empirically determined during treatment or can be predefined, such as from a pilot study with subjects. For example, various stimulation settings, including baseline (stimulation off), optimal therapeutic stimulation, modified ineffective stimulation, and maximum tolerated stimulation, can be applied to determine optimal therapeutic stimulation parameters for treating a neurological or neurodegenerative disease in a region where pathological protein aggregates are known to be present (e.g., by medical imaging) or predicted to be evaluated according to the methods described herein.

[0166] Parameters of the electrical stimulation may include one or more of frequency, pulse width / duration, duty cycle, intensity / amplitude, pulse pattern, programmed duration, programmed frequency, etc. In certain embodiments, the parameters are adjusted to target specific neuronal cell types or neural circuits in the brain at locations where pathological protein aggregates are present or predicted to occur in the future.

[0167] Frequency refers to the pulses generated per second during stimulation and is described in units of Hertz (Hz, e.g., 60 Hz = 60 pulses per second). The frequency of electrical stimulation used in the methods of the invention can vary widely depending on numerous factors and can be empirically determined during treatment of a subject or can be predefined. In certain embodiments, the methods can involve applying electrical stimulation to the brain at a frequency between 2 Hz and 250 Hz, e.g., 25 Hz and 200 Hz, 50 Hz and 250 Hz, 50 Hz and 185 Hz, 50 Hz and 150 Hz, 75 Hz and 200 Hz, 100 Hz and 200 Hz, 100 Hz and 180 Hz, 100 Hz and 160 Hz, or 130 Hz and 150 Hz. In some embodiments, electrical stimulation to the brain is applied at a frequency of about 120 Hz to about 160 Hz, including any pulse frequency within this range, such as 120 Hz, 122 Hz, 124 Hz, 126 Hz, 128 Hz, 130 Hz, 132 Hz, 134 Hz, 136 Hz, 138 Hz, 140 Hz, 142 Hz, 144 Hz, 146 Hz, 148 Hz, 150 Hz, 152 Hz, 154 Hz, 156 Hz, 158 Hz, or 160 Hz. In some embodiments, non-integer pulse frequencies are used (e.g., 130.2 Hz, 130.4 Hz, etc.).

[0168] The electrical stimulation may be applied in pulses, such as monophasic or biphasic pulses. The time span of a single pulse is referred to as the pulse width or pulse duration. The pulse width used in the present methods may vary widely depending on many factors (e.g., disease severity, patient condition, etc.) and may be empirically determined or predefined. In certain embodiments, the present methods may involve applying the electrical stimulation at a pulse width of about 10 μs to 500 μs, e.g., 20 μs to 450 μs, 40 μs to 450 μs, 60 μs to 450 μs, 60 μs to 220 μs, 60 μs to 120 μs, or 60 μs to 90 μs. In some embodiments, the electrical stimulation to the brain is applied with a pulse width of about 60 μsec to about 210 μsec, including any pulse width within this range, e.g., 60 μsec, 65 μsec, 70 μsec, 75 μsec, 80 μsec, 85 μsec, 90 μsec, 95 μsec, 100 μsec, 105 μsec, 110 μsec, 115 μsec, 120 μsec, 125 μsec, 130 μsec, 135 μsec, 140 μsec, 145 μsec, 150 μsec, 155 μsec, 160 μsec, 165 μsec, 170 μsec, 175 μsec, 180 μsec, 185 μsec, 190 μsec, 195 μsec, 200 μsec, 205 μsec, 210 μsec, 215 μsec, or 220 μsec.

[0169] Electrical stimulation may be applied for stimulation periods of 0.1 seconds to 1 month, with rest periods (i.e., no electrical stimulation) between them. In certain cases, the duration of electrical stimulation may be 0.1 seconds to 1 week, 1 second to 1 day, 10 seconds to 12 hours, 1 minute to 6 hours, 10 minutes to 1 hour, etc. In certain cases, the duration of electrical stimulation may be 1 second to 1 minute, 1 second to 30 seconds, 1 second to 15 seconds, 1 second to 10 seconds, 1 second to 6 seconds, 1 second to 3 seconds, 1 second to 2 seconds, or 6 seconds to 10 seconds. The rest period between each stimulation period may be 60 seconds or less, 30 seconds or less, 20 seconds or less, or 10 seconds. In some embodiments, electrical stimulation may be applied for one year or more, two years or more, three years or more, five years or more, or ten years or more. In some embodiments, electrical stimulation may be continued indefinitely as part of a long-term DBS treatment plan.

[0170] The electrical stimulation may be applied at a current amplitude of 0.1 mA to 30 mA, e.g., 0.1 mA to 25 mA, e.g., 0.1 mA to 20 mA, 0.1 mA to 15 mA, 0.1 mA to 10 mA, 0.1 mA to 2 mA, 0.1 mA to 1 mA, 1 mA to 20 mA, 1 mA to 10 mA, 2 mA to 30 mA, 2 mA to 15 mA, 2 mA to 10 mA, or 1 mA to 3 mA. In some embodiments, the amplitude of the current is between 0.1 mA and 3.5 mA, or any amplitude of current within this range, for example, 0.1 mA, 0.2 mA, 0.3 mA, 0.4 mA, 0.5 mA, 0.6 mA, 0.7 mA, 0.8 mA, 0.9 mA, 1.0 mA, 1.1 mA, 1.2 mA, 1.3 mA, 1.4 mA, 1.5 mA, 1.6 mA, 1.7 mA, 1.8 mA, 1.9 mA, 2.0 mA, 2.1 mA, 2.2 mA, 2.3 mA, 2.4 mA, 2.5 mA, 2.6 mA, 2.7 mA, 2.8 mA, 2.9 mA, 3.0 mA, 3.1 mA, 3.2 mA, 3.3 mA, 3.4 mA, or 3.5 mA.

[0171] The electrical stimulus may be applied at a voltage amplitude of 0.1 V to 15 V, e.g., 0.1 V to 10 V, 0.1 V to 5 V, 1 V to 10 V, 1 V to 5 V, or 1 V to 3.5 V. In some embodiments, the voltage amplitude is 1 V to 3.5 V, or any voltage amplitude within this range, e.g., 1 V, 1.1 V, 1.2 V, 1.3 V, 1.4 V, 1.5 V, 1.6 V, 1.7 V, 1.8 V, 1.9 V, 2.0 V, 2.1 V, 2.2 V, 2.3 V, 2.4 V, 2.5 V, 2.6 V, 2.7 V, 2.8 V, 2.9 V, 3.0 V, 3.1 V, 3.2 V, 3.3 V, 3.4 V, or 3.5 V.

[0172] Electrical stimulation having the above-described parameters can be applied for a program duration of about one day or less, e.g., 18 hours, 6 hours, 3 hours, 2 hours, 1 hour, 45 minutes, 30 minutes, 20 minutes, 10 minutes, or 5 minutes, or less, e.g., 1-5 minutes, 2-10 minutes, 2-20 minutes, 2-30 minutes, 5-10 minutes, 5-30 minutes, or 5-15 minutes, 10-400 minutes, 25-300 minutes, 50-200 minutes, or 75-150 minutes, including pulse application and intervening rest periods. The program can be repeated at a desired program frequency to reduce or prevent the aggregation and spread of pathological protein aggregates and alleviate symptoms of a neurological or neurodegenerative disease in a subject. Thus, a treatment plan can include a program for electrical stimulation at a desired program frequency and program duration. In certain embodiments, the computer-implemented methods described herein are used to analyze brain images taken at different time points, and programmed neurostimulation parameters are adjusted based on any changes in the locations where pathological protein aggregates are predicted to appear or develop. In some embodiments, the treatment plan is controlled by a control unit that communicates with a pulse generator connected to one or more DBS electrodes in a closed-loop treatment plan.

[0173] Upon completion of the treatment plan, the patient may be evaluated for the effectiveness of the treatment, and if necessary, the treatment plan may be repeated. In certain cases, the treatment plan may be modified before being repeated. For example, one or more of the frequency, pulse width, current amplitude, period of electrical stimulation, programmed duration, programmed frequency, and / or placement of DBS or sensing electrodes may be modified before initiating a second treatment plan.

[0174] Application of the present methods may include a preliminary step of selecting patients for treatment based on need as determined by a clinical evaluation, which may include an assessment of the severity of the neurological or neurodegenerative disease (e.g., a neurological or neurodegenerative disease lasting at least three months), physical condition, medication regimen, cognitive assessment, anatomical assessment, behavioral assessment, and / or neurophysiological assessment. In certain cases, subjects may be further evaluated to determine whether neurostimulation completely or partially (e.g., at least 50%) alleviates the neurological or neurodegenerative disease. Such patients may undergo neurostimulation on a temporary, trial basis to determine whether neurostimulation reduces or prevents the aggregation and spread of pathological protein aggregates or decreases the severity of symptoms of the neurological or neurodegenerative disease experienced by the patient.

[0175] Optogenetics In some embodiments, optogenetics is used in an effective manner to reduce or prevent the aggregation and spread of pathological protein aggregates to treat neurological or neurodegenerative diseases. Optogenetics is used to enable optical control of activation (i.e., depolarization) or inhibition (i.e., hyperpolarization) of neurons that have been genetically modified to express light-responsive ion channels. In some embodiments, the light-responsive ion channels are natural or synthetic opsins that use retinal cofactors (e.g., all-trans retinal for microbial opsins) to respond to light. For example, light-responsive cation-conducting opsins (e.g., Ca 2+Opsins that conduct anion ions (e.g., channelrhodopsins that conduct chloride ions) can be used to activate or depolarize neurons. Light-responsive anion-conducting opsins (e.g., channelrhodopsins or halorhodopsins that conduct chloride ions) or light-responsive proton conductance modulators (e.g., bacteriorhodopsin or archaealhodopsin) can be used to inhibit or hyperpolarize neurons. The levels of retinoids present in mammalian brains are usually sufficient for expressed opsins to function without cofactor supplementation. For descriptions of optogenetics and its use in controlling neural activity, see, e.g., Aravanis et al. (2007) J Neural Eng 4:S143-S156, Arenkiel et al. (2007) Neuron 54:205-218, Boyden et al. (2005) Nat Neurosci 8:1263-1268, Chow et al. (2010) Nature 463:98-10, Gradinaru et al. (2007) J Neurosci 27:14231-14238, Gradinaru et al. (2008) Brain Cell Biol 36:129-139, Gradinaru et al. (2010) Cell 141:1-12, Li et al. (2005) Proc Natl Acad Sci 102:17816-17821,Lin et al.2009.Characterization of engineered channelrhodopsin variants with improved properties and kinetics.Biophys J 96:1803-1814,Yizhar et al.(2011)Microbial opsins:A family of single-component tools for optical control of neural activity.Cold Spring Harbor Protoc,Zhang et (2007) Nat Methods 4:139-141, Zhang et al. (2006) Nat Methods 3:785-792, Zhang et al.(2007)Nature 446:633-639,Zhang et al.(2008)Nat Neurosci 11:631-633; and US Patent No. 10,914,803; US Patent No. 10,589,123; US Patent No. 10,583,309; US Patent No. 10,568,51 No. 6; No. 10,568,307; No. 10,538,560; No. 10,478,499; No. 10,220,092; No. 10,196,43 Nos. 1, 10,087,223, 10,052,383, 9,969,783, 9,878,176, 9,855,442, 9,757,587, 9,458,208, and 8,834,546, all of which are incorporated herein by reference in their entireties.

[0176] In some embodiments, the target neuron is genetically modified to express a light-responsive ion channel that, when stimulated with an appropriate light stimulus, hyperpolarizes or depolarizes the stimulated target neuron. The term "genetic modification" refers to a permanent or temporary genetic change induced in a cell after the introduction of heterologous nucleic acid (i.e., nucleic acid exogenous to the cell) into the cell. The genetic change ("modification") can be achieved by integration of the heterologous nucleic acid into the genome of the host cell or by transient or stable maintenance of the heterologous nucleic acid as an extrachromosomal element. When the cell is a eukaryotic cell, a permanent genetic change can be achieved by introduction of the nucleic acid into the genome of the cell. Suitable methods of genetic modification include the use of viral infection, transfection, conjugation, protoplast fusion, electroporation, particle gun technology, calcium phosphate precipitation, direct microinjection, and the like.

[0177] In some cases, target cells expressing the light-responsive polypeptide can be activated or inhibited upon exposure to light of various wavelengths, hi some cases, the target cells expressing the light-responsive polypeptide are neuronal cells that express the light-responsive polypeptide, and exposure to light of various wavelengths results in depolarization or polarization of the neuron.

[0178] In some examples, the light-responsive polypeptide is a light-responsive ion channel polypeptide. The light-responsive ion channel polypeptide is adapted to allow one or more ions to pass through the plasma membrane of a target cell when the polypeptide is illuminated with light at an activating wavelength. The light-responsive protein may be characterized as an ion pump protein, which facilitates the passage of a small number of ions through the plasma membrane per photon of light, or as an ion channel protein, which allows ions to flow freely through the plasma membrane when the channel is open. In some embodiments, the light-responsive polypeptide depolarizes excitable cells when activated by light at an activating wavelength. In some embodiments, the light-responsive polypeptide hyperpolarizes excitable cells when activated by light at an activating wavelength.

[0179] In some cases, the light-responsive polypeptide mediates a hyperpolarizing current in a target cell in which it is expressed when the cell is illuminated with light. Non-limiting examples of light-responsive polypeptides that can mediate a hyperpolarizing current can be found, for example, in U.S. Patent Nos. 9,359,449 and 9,175,095. Non-limiting examples of hyperpolarizing light-responsive polypeptides include NpHr, eNpHr2.0, eNpHr3.0, eNpHr3.1, or GtR3. In some cases, the light-responsive polypeptide mediates a depolarizing current in a target cell in which it is expressed when the cell is illuminated with light. Non-limiting examples of depolarizing light-responsive polypeptides include "C1V1," ChR1, VChR1, and ChR2. Further information regarding other light-responsive cation channels, anion pumps, and proton pumps can be found in US Patent Application Publication No. 2009 / 0093403 and US Patent No. 9,359,449.

[0180] In some embodiments, the photoresponsive polypeptide can be activated by blue light (e.g., in the range of 490 nm to 450 nm). In one embodiment, the photoresponsive polypeptide can be activated by light having a wavelength of about 473 nm. In some embodiments, the photoresponsive polypeptide can be activated by yellow light (e.g., in the range of 590 nm to 560 nm). In another embodiment, the photoresponsive polypeptide can be activated by light having a wavelength of about 560 nm. In another embodiment, the photoresponsive polypeptide can be activated by red light (e.g., in the range of 700 nm to 635 nm). In another embodiment, the photoresponsive polypeptide can be activated by light having a wavelength of about 630 nm. In other embodiments, the photoresponsive polypeptide can be activated by violet light (e.g., in the range of 450 nm to 400 nm). In one embodiment, the photoresponsive polypeptide can be activated by light having a wavelength of about 405 nm. In other embodiments, the photoresponsive polypeptide can be activated by green light (e.g., in the range of 560 nm to 520 nm). In other embodiments, the photoresponsive polypeptide may be activated by cyan light (e.g., in the range of 520 nm to 490 nm). In other embodiments, the photoresponsive polypeptide may be activated by orange light (e.g., in the range of 635 nm to 590 nm). One skilled in the art will recognize that each photoresponsive polypeptide has its own range of activation wavelengths.

[0181] In some cases, a brain region having neurons containing light-responsive polypeptides is illuminated using one or more optical fibers. The optical fiber can be configured in any suitable manner to direct light emitted from a suitable light source, e.g., a laser or light-emitting diode (LED) light source, to the brain region. The optical fiber can be any suitable optical fiber. In some cases, the optical fiber is a multimode optical fiber. The optical fiber can include a core defining a core diameter, through which light from the light source passes. The optical fiber can have any suitable core diameter. In some cases, the core diameter of the optical fiber is 10 mm or more, e.g., 20 mm or more, 30 mm or more, 40 mm or more, 50 mm or more, 60 mm or more (including 80 mm or more), and 1,000 mm or less, e.g., 500 mm or less, 200 mm or less, or 100 mm or less (including 70 mm or less). In some embodiments, the core diameter of the optical fiber is within the range of 10 to 1,000 mm, e.g., 20 to 500 mm, 30 to 200 mm, including 40 to 100 mm.

[0182] The optical fiber end implanted in the target region of the brain can have any suitable configuration suitable for illuminating the region of the brain with light stimulation delivered through the optical fiber. In some cases, the optical fiber includes an attachment device at or near the distal end of the optical fiber, which corresponds to the end that is inserted into the subject. In some cases, the attachment device is configured to connect to the optical fiber and facilitate attachment of the optical fiber to the subject, such as the subject's skull. Any suitable attachment device can be used. In some cases, the attachment device includes a ferrule, such as a metal, ceramic, or plastic ferrule. The ferrule can have any dimensions suitable for holding and attaching the optical fiber.

[0183] In certain embodiments, the methods of the present disclosure can be implemented using any suitable electronic components to control and / or adjust various optical components used to illuminate a brain region. The optical components (e.g., light sources, optical fibers, lenses, objective lenses, mirrors, etc.) can be controlled by a controller, for example, to adjust the light source to illuminate the brain region with light pulses. The controller can include a driver for the light source that controls one or more parameters related to the light pulses, such as, but not limited to, the frequency, pulse width, duty cycle, wavelength, and intensity of the light pulses. The controller can communicate with components of the light source (e.g., collimators, shutters, filter wheels, movable mirrors, lenses, etc.).

[0184] In some embodiments, the photoresponsive polypeptide is activated by a light pulse that can have a duration of any of about 1 millisecond (ms), about 2 ms, about 3 ms, about 4 ms, about 5 ms, about 6 ms, about 7 ms, about 8 ms, about 9 ms, about 10 ms, about 15 ms, about 20 ms, about 25 ms, about 30 ms, about 35 ms, about 40 ms, about 45 ms, about 50 ms, about 60 ms, about 70 ms, about 80 ms, about 90 ms, about 100 ms, about 200 ms, about 300 ms, about 400 ms, about 500 ms, about 600 ms, about 700 ms, about 800 ms, about 900 ms, about 1 second, about 1.25 seconds, about 1.5 seconds, or about 2 seconds (including any time between these numbers). In some embodiments, the photoresponsive polypeptide has an optical power of about 0.05 mW / mm², including any value between these numbers. 2 , about 0.1mW / mm 2 , about 0.25mW / mm 2 , approximately 0.5 mW / mm 2 , approximately 0.75 mW / mm 2 , about 1mW / mm 2 , approximately 2 mW / mm 2 , about 3mW / mm 2 , approximately 4 mW / mm 2 , about 5 mW / mm 2 , approximately 6 mW / mm 2 , approximately 7 mW / mm 2 , approximately 8 mW / mm 2 , approximately 9 mW / mm 2, about 10mW / mm 2 , about 20mW / mm 2 , about 50mW / mm 2 , about 100mW / mm 2 , about 250mW / mm 2 , about 500mW / mm 2 , about 750mW / mm 2 , about 1000mW / mm 2 , about 1100mW / mm 2 , about 1200mW / mm 2 , about 1300mW / mm 2 , about 1400mW / mm 2 , about 1500mW / mm 2 , about 1600mW / mm 2 , about 1700mW / mm 2 , about 1800mW / mm 2 , about 1900mW / mm 2 , about 2000mW / mm 2 , about 2100mW / mm 2 , about 2200mW / mm 2 , about 2300mW / mm 2 , about 2400mW / mm 2 , about 2500mW / mm 2 , about 2600mW / mm 2 , about 2700mW / mm 2 , about 2800mW / mm 2 , about 2900mW / mm 2 , about 3000mW / mm 2 , about 3100mW / mm 2 , about 3100mW / mm 2 , about 3300mW / mm 2 , about 3400mW / mm 2 , or about 3500 mW / mm 2 The photodiode is activated by a light pulse that can have an optical power density of either

[0185] The light stimulus used to activate the light-responsive polypeptide may include light pulses characterized by, for example, frequency, pulse width, duty cycle, wavelength, intensity, etc. The light stimulus includes two or more different sets of light pulses, each set of light pulses characterized by a different temporal pattern of the light pulses. The temporal pattern may be characterized by any suitable parameter, including, but not limited to, frequency, period (i.e., total duration of the light stimulus), pulse width, duty cycle, etc.

[0186] The light pulses can have any suitable frequency. In some cases, the set of light pulses comprises a single pulse of light that is sustained for the entire duration of the light stimulus. In some cases, the set of light pulses has a frequency of 0.1 Hz or more, for example, 0.5 Hz or more, 1 Hz or more, 5 Hz or more, 10 Hz or more, 20 Hz or more, 30 Hz or more, 40 Hz or more, including 50 Hz or more, or 60 Hz or more, or 70 Hz or more, or 80 Hz or more, or 90 Hz or more, or 100 Hz or more, and has a frequency of 100,000 Hz or less, for example, 10,000 Hz or less, 1,000 Hz or less, 500 Hz or less, 400 Hz or less, 300 Hz or less, or 200 Hz or less, including 100 Hz or less. In some embodiments, the light pulses have a frequency in the range of 0.1 to 100,000 Hz, for example, 1 to 10,000 Hz, 1 to 1,000 Hz, including 5 to 500 Hz, or 10 to 100 Hz.

[0187] In some cases, the two sets of light pulses are characterized by having different parameter values, such as different pulse widths (e.g., short or long). The light pulses can have any suitable pulse width. In some cases, the pulse width is 0.1 ms or more, e.g., 0.5 ms or more, 1 ms or more, 3 ms or more, 5 ms or more, 7.5 ms or more, 10 ms or more (including 15 ms or more, or 20 ms or more, or 25 ms or more, or 30 ms or more, or 35 ms or more, or 40 ms or more, or 45 ms or more, or 50 ms or more), and 500 ms or less, e.g., 100 ms or less, 90 ms or less, 80 ms or less, 70 ms or less, 60 ms or less, 50 ms or less, 45 ms or less, 40 ms or less, 35 ms or less, 30 ms or less, 25 ms or less (including 20 ms or less). In some embodiments, the pulse width is in the range of 0.1 to 500 ms, for example, 0.5 to 100 ms, 1 to 80 ms, including 1 to 60 ms, or 1 to 50 ms, or 1 to 30 ms.

[0188] The average power of the light pulses measured at the tip of the optical fiber that delivers the light pulses to the brain region can be any suitable power. In some cases, the power is 0.1 mW or more, e.g., 0.5 mW or more, 1 mW or more, 1.5 mW or more (including 2 mW or more);

[0189] The power may be 2.5 mW or more, or 3 mW or more, or 3.5 mW or more, or 4 mW or more, or 4.5 mW or more, or 5 mW or more, and may be 1,000 mW or less, e.g., 500 mW or less, 250 mW or less, 100 mW or less, 50 mW or less, 40 mW or less, 30 mW or less, 20 mW or less, 15 mW or less (including 10 mW or less), or 5 mW or less. In some embodiments, the power is in the range of 0.1 to 1,000 mW, e.g., 0.5 to 100 mW, 0.5 to 50 mW, or 1 to 20 mW, including 1 to 10 mW, or 1 to 5 mW.

[0190] The wavelength and intensity of the light pulses can vary and can depend on the activation wavelength of the light-responsive polypeptide, the optical transparency of the region of the brain, the desired volume of the brain to be illuminated, and the like.

[0191] The volume of the brain region illuminated by the light pulse can be any suitable volume. In some cases, the illuminated volume is less than 0.001 mm. 3 More than, for example, 0.005 mm 3 More than 0.001mm 3 More than 0.005mm 3 More than 0.01mm 3 Over 0.05mm 3 More than (0.1mm 3 (including above) and 100mm 3 For example, 50 mm 3 Below, 20mm 3 Below, 10mm 3 Below, 5mm 3 Below, 1mm 3 Less than (0.1mm 3 In certain cases, the irradiated volume is between 0.001 and 100 mm 3 , e.g., 0.005 to 20 mm 3 , 0.01~10mm 3 , 0.01~5mm 3 (0.05 to 1 mm 3 (including)

[0192] In some embodiments, a light-responsive polypeptide expressed in a cell can be fused to one or more amino acid sequence motifs selected from the group consisting of a signal peptide, an endoplasmic reticulum (ER) export signal, a membrane trafficking signal, and / or an N-terminal Golgi export signal. The one or more amino acid sequence motifs that enhance light-responsive protein trafficking to the plasma membrane of a mammalian cell can be fused to the N-terminus, C-terminus, or both the N-terminus and C-terminus of the light-responsive polypeptide. In some cases, the one or more amino acid sequence motifs that enhance light-responsive polypeptide trafficking to the plasma membrane of a mammalian cell are fused internally within the light-responsive polypeptide. Optionally, the light-responsive polypeptide and the one or more amino acid sequence motifs can be separated by a linker. In some embodiments, the light-responsive polypeptide can be modified by the addition of a trafficking signal (ts) that enhances protein trafficking to the cell membrane. In some embodiments, the trafficking signal can be derived from the amino acid sequence of the human inward rectifier potassium channel Kir2.1. In some embodiments, the signal peptide sequence in the protein can be deleted or replaced with a signal peptide sequence from a different protein.

[0193] Exemplary light-responsive polypeptides and amino acid sequence motifs for use in the present systems and methods are disclosed, for example, in U.S. Pat. Nos. 10,538,560, 10,568,307, 9,284,353, 9,359,449, and 9,365,628, which are incorporated herein by reference.

[0194] Light-responsive polypeptides of interest include, for example, step function opsin (SFO) 6 proteins or stabilized step function opsin (SSFO) proteins, which may have specific amino acid substitutions at key positions in the retinal binding pocket of the protein. See, for example, WO 2010 / 056970, the disclosure of which is incorporated herein by reference in its entirety. The polypeptide may be a cation channel derived from Volvox carteri (VChR1), optionally containing one or more amino acid substitutions, such as C123A, C123S, D151A, etc. The light-responsive cation channel protein may be a C1V1 chimeric protein derived from the VChR1 protein of Volvox carteri and the ChR1 protein of Chlamydomonas reinhardti, wherein the protein comprises the amino acid sequence of VChR1, with at least the first and second transmembrane helices replaced by those of ChR1, and optionally with an amino acid substitution at amino acid residues E122 or E162. In other embodiments, the light-responsive cation channel protein is a C1C2 chimeric protein derived from the ChR1 and ChR2 proteins from Chlamydomonas reinhardtii, which is responsive to light and capable of mediating a depolarizing current in a cell when the cell is illuminated with light. In some embodiments, the depolarizing light-responsive polypeptide is a red-shifted variant of the depolarizing light-responsive polypeptide from Chlamydomonas reinhardtii, referred to as a "ReaChR polypeptide" or "ReaChR protein" or "ReaChR." In some embodiments, the depolarizing light-responsive polypeptide is an SdChR polypeptide from Scherffelia dubia, which is capable of transporting cations across the cell membrane when the cell is illuminated with light. In some embodiments, the depolarizing light-responsive polypeptide is a CnChR1 from Chlamydomonas noctigama, which is capable of transporting cations across the cell membrane when the cell is illuminated with light.In some embodiments, the light-responsive cation channel protein is a CsChrimson chimeric protein derived from the CsChR protein of Chloromonas subdivisa and the CnChR1 protein from Chlamydomonas noctigama, where the N-terminus of the protein comprises residues 1-73 of CsChR followed by residues 79-350 of CnChR1, and is capable of mediating a depolarizing current in cells when the cells are illuminated with light. In some embodiments, the depolarizing light-responsive polypeptide can be, for example, ShChR1 from Stigeoclonium helveticum, where the ShChR1 polypeptide is capable of transporting cations across the cell membrane when the cells are illuminated with light.

[0195] In some embodiments, the depolarizing light-responsive polypeptide is derived from Chlamydomonas reinhardtii (CHR1, and particularly CHR2), and this polypeptide is capable of transporting cations across the cell membrane when the cell is illuminated with light and mediating a depolarizing current within the cell when the cell is illuminated with light. In some embodiments, a CaMKIIa-driven humanized channelrhodopsin CHR2 H134R mutant fused to EYFP is used for optogenetic activation. The light used to activate the light-responsive cation channel protein from Chlamydomonas reinhardtii can have a wavelength of about 460 to about 495 nm, or can have a wavelength of about 480 nm. The light-responsive cation channel protein can additionally contain substitutions, deletions, and / or insertions introduced into the native amino acid sequence to increase or decrease its sensitivity to light, to increase or decrease its sensitivity to specific wavelengths of light, and / or to increase or decrease the ability of the light-responsive cation channel protein to modulate the polarization state of the cell's plasma membrane. Furthermore, the light-responsive cation channel protein can contain one or more conservative amino acid substitutions and / or one or more non-conservative amino acid substitutions. Light-responsive proton pump proteins containing substitutions, deletions, and / or insertions introduced into the native amino acid sequence preferably retain the ability to transport cations across cell membranes. The protein can contain one or more of various amino acid substitutions, such as H134R, T159C, L132C, E123A, etc. The protein can further contain a fluorescent protein, such as, but not limited to, yellow fluorescent protein, red fluorescent protein, green fluorescent protein, or cyan fluorescent protein.

[0196] Neurons can be selectively activated or inhibited optogenetically by engineering them to express one or more light-responsive polypeptides configured to hyperpolarize or depolarize the neuron. Suitable light-responsive polypeptides and methods for their use are further described below.

[0197] The light-responsive polypeptide for use in the present disclosure can be any suitable light-responsive polypeptide for selectively activating a subtype of neuron by irradiating the neuron with an activating light stimulus. In some examples, the light-responsive polypeptide is a light-responsive ion channel polypeptide. The light-responsive ion channel polypeptide is adapted to allow one or more ions to pass through the plasma membrane of a target cell when the polypeptide is irradiated with light at an activating wavelength. The light-responsive protein can be characterized as an ion pump protein, which facilitates the passage of a small number of ions through the plasma membrane per photon of light, or as an ion channel protein, which allows ions to flow freely through the plasma membrane when the channel is open. In some embodiments, the light-responsive polypeptide depolarizes the cell when activated by light at an activating wavelength. In some embodiments, the light-responsive polypeptide hyperpolarizes the cell when activated by light at an activating wavelength. Suitable hyperpolarizing and depolarizing polypeptides are known in the art and include, for example, channelrhodopsin (e.g., ChR2), mutants of ChR2 (e.g., C128S, D156A, C128S+D156A, E123A, E123T), iC1C2, C1C2, GtACR2, NpHR, eNpHR3.0, C1V1, VChR1, VChR2, SwiChR, Arch, ArchT, KR2, ReaChR, ChiEF, Chronos, ChRGR, CsChrimson, etc. In some cases, the light-responsive polypeptide comprises bReaCh-ES, as described, for example, in Rajasethupathy et al., Nature. 2015 Oct. 29;526(7575):653, which is incorporated by reference. Hyperpolarizing and depolarizing opsins are described in various publications; see, e.g., Berndt and Deisseroth (2015) Science 349:590; Berndt et al. (2014) Science 344:420; and Guru et al. (July 25, 2015) Intl. J. Neuropsychopharmacol. pp. 1-8 (PMID 26209858).

[0198] The light-responsive polypeptide can be introduced into neurons using any suitable method. In some cases, neurons of a desired subtype are genetically modified to express the light-responsive polypeptide. In some cases, neurons can be genetically modified using a viral vector, such as an adeno-associated viral vector, containing a nucleic acid having a nucleotide sequence encoding the light-responsive polypeptide. The viral vector can include any suitable control elements (e.g., promoters, enhancers, recombination sites, etc.) for controlling the expression of the light-responsive polypeptide according to the neuron subtype, timing, the presence of inducers, etc.

[0199] "Operably linked" refers to a juxtaposition wherein the components so described are in a relationship permitting them to function in their intended manner. For example, a promoter is operably linked to a nucleotide sequence (e.g., a protein-coding sequence, e.g., a sequence encoding an mRNA; a non-protein-coding sequence, e.g., a sequence encoding a photoreactive protein, etc.) if the promoter affects its transcription and / or expression.

[0200] Neuron-specific promoters and other control elements (e.g., enhancers) are known in the art. Suitable neuron-specific control sequences include the neuron-specific enolase (NSE) promoter (see, e.g., EMBL HSENO2, X51956; see also, e.g., U.S. Pat. No. 6,649,811, U.S. Pat. No. 5,387,742); the aromatic amino acid decarboxylase (AADC) promoter; the neurofilament promoter (see, e.g., GenBank HUMNF 1, L04147); the synapsin promoter (see, e.g., GenBank HUMSYNIB, M55301); the thy-1 promoter (see, e.g., Chen et al. (1987) Cell 51:7-19; and Llewellyn et al. (2010) Nat. Med. 16:1161); the serotonin receptor promoter (see, e.g., GenBank S62283); tyrosine hydroxylase promoter (TH) (see, e.g., Nucl. Acids. Res. 15:2363-2384 (1987) and Neuron 6:583-594 (1991)); GnRH promoter (see, e.g., Radovick et al., Proc. Natl. Acad. Sci. USA 88:3402-3406 (1991)); L7 promoter (see, e.g., Oberdick et al., Science 248:223-226 (1990)); DNMT promoter (see, e.g., Bartge et al., Proc. Natl. Acad. Sci. USA 85:3648-3652 (1988)); enkephalin promoter (see, e.g., Comb et al., Proc. Natl. Acad. Sci. USA 85:3648-3652 (1988)) al., EMBOJ. 17:3793-3805 (1988); myelin basic protein (MBP) promoter; CMV enhancer / platelet-derived growth factor-beta promoter (see, e.g., Liu et al. (2620) Gene Therapy 11:52-60); motor neuron-specific gene Hb9 promoter (see, e.g., U.S. Patent No. 7,632,679; and Lee et al. (2620) Development 131:3295-3306); and Ca 2+- the alpha subunit of the calmodulin-dependent protein kinase II (CaMKII) promoter (see, e.g., Mayford et al. (1996) Proc. Natl. Acad. Sci. USA 93:13250). Other suitable promoters include the elongation factor (EF) 1 and dopamine transporter (DAT) promoters.

[0201] In some cases, neuronal subtype-specific expression of light-responsive polypeptides can be achieved by using recombination systems, such as Cre-Lox recombination, Flp-FRT recombination, etc. Cell type-specific expression of genes using recombination is described, for example, in Fenno et al., Nat Methods, 2014 July;11(7):763; and Gompf et al., Front Behav Neurosci. 2015 July.2;9:152, which are incorporated herein by reference.

[0202] In some embodiments, the vector is a recombinant adeno-associated virus (AAV) vector. AAV vectors are relatively small DNA viruses that can integrate into the genome of the cells they infect in a stable and site-specific manner. They can infect a wide range of cells without inducing any effects on cell growth, morphology, or differentiation, and do not appear to be involved in human pathologies. The AAV genome has been cloned, sequenced, and characterized. It encompasses approximately 4,700 bases and contains inverted terminal repeat (ITR) regions of approximately 145 bases at each end, which serve as viral replication origins. The remainder of the genome is divided into two essential regions with encapsidation functions: the left part of the genome containing the rep gene, which is involved in viral replication and viral gene expression, and the right part of the genome containing the cap gene, which encodes the viral capsid protein.

[0203] The application of AAV as a vector for gene therapy has developed rapidly in recent years. Wild-type AAV can infect dividing or non-dividing cells or tissues of mammals, including humans, at relatively high titers and can also integrate into human cells at a specific site (on the long arm of chromosome 19) (see Kotin et al., Proc. Natl. Acad. Sci. USA, 1990, 87:2211-2215; Samulski et al., EMBO J., 1991, 10:3941-3950, the disclosures of which are incorporated herein by reference in their entirety). AAV vectors lacking the rep and cap genes lose the specificity of site-specific integration but can still mediate long-term stable expression of exogenous genes. AAV vectors exist in cells in two forms: one as an episome outside the chromosome, and the other as a chromosomal integration, the former being the predominant form. Furthermore, AAV has not been found to be associated with any human disease, nor has any change in biological characteristics resulting from integration been observed. There are 16 serotypes of AAV reported in the literature, designated AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10, AAV11, AAV12, AAV13, AAV14, AAV15, and AAV16, respectively. AAV5 was originally isolated from humans (Bantel-Schaal, and H. zur Hausen. Virology, 1984, 134:52-63), while AAV1-4 and AAV6 were all discovered in research on adenoviruses (Ursula Bantel-Schaal, Hajo Delius, and Harald zur Hausen. J. Virol., 1999, 73:939-947).

[0204] AAV vectors can be prepared using any convenient method. Adeno-associated viruses of any serotype are suitable (see, e.g., Blacklow, pp. 165-174 of "Parvoviruses and Human Disease," JR Pattison, ed. (1988); Rose, Comprehensive Virology 3:1, 1974; P. Tattersall, "The Evolution of Parvovirus Taxonomy," In Parvoviruses (J.R. Kerr, S.F. Cotmore, M.E. Bloom, R.M. Linden, C.R. Parrish, Eds.), pp. 5-14, Hudder Arnold, London, UK (2006); and D.E. Bowles, J.E. Rabinowitz, R.J. Samulski, "The Genus Dependovirus," (J.R. Kerr, S.F. Cotmore, M.E. Bloom, R.M. Linden, C.R. Parrish, Eds.), pp. 15-23, Hudder Arnold, London, UK (2006), the disclosures of which are incorporated herein by reference in their entireties). Methods for purifying vectors can be found, for example, in U.S. Patent Nos. 6,566,118, 6,989,264, and 6,995,006, and WO / 1999 / 011764, entitled "Methods for Generating High Titer Helper-free Preparation of Recombinant AAV Vectors," the disclosures of which are incorporated herein by reference in their entireties. The preparation of hybrid vectors is described, for example, in PCT Application No. PCT / US2005 / 027091, the disclosure of which is incorporated herein by reference in its entirety.The use of AAV-derived vectors to transfer genes in vitro and in vivo has been described (see, e.g., International Patent Application Publication Nos. WO 91 / 18088 and WO 93 / 09239; U.S. Pat. Nos. 4,797,368, 6,596,535, and 5,139,941; and European Patent No. 0 488 528, all of which are incorporated herein by reference in their entireties). These publications describe various AAV-derived constructs in which the rep and / or cap genes have been deleted and replaced with a gene of interest, and the use of these constructs to transfer the gene of interest in vitro (into cultured cells) or in vivo (directly into an organism). Replication-deficient recombinant AAV according to the present invention can be prepared by co-transfecting a plasmid containing a nucleic acid sequence of interest flanked by two AAV inverted terminal repeat (ITR) regions and a plasmid carrying the AAV encapsidation genes (rep and cap genes) into a cell line infected with a human helper virus (e.g., adenovirus). The resulting AAV recombinant is then purified by standard techniques.

[0205] In some embodiments, vectors for use in the methods of the invention are encapsidated into viral particles (e.g., AAV viral particles, including, but not limited to, AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAV10, AAV11, AAV12, AAV13, AAV14, AAV15, and AAV16). Thus, the invention includes recombinant viral particles (recombinant, because they contain a recombinant polynucleotide) comprising any of the vectors described herein. Methods for producing such particles are known in the art and are described in U.S. Patent No. 6,596,535.

[0206] It is understood that one or more vectors can be administered to a neuronal cell. When two or more vectors are used, they can be administered at the same time or at different times.

[0207] system The present disclosure also provides systems for use, for example, in practicing the subject methods. In certain embodiments, the system includes a neurostimulation device and a processor programmed to instruct the neurostimulation device to deliver neurostimulation to a subject's brain in a manner effective to treat the subject's neurological or neurodegenerative disease, according to the computer-implemented methods described herein, where the neurostimulation is applied to the brain at known locations of aggregated pathological proteins (e.g., based on medical imaging), or at predicted current locations of pathological protein aggregates, or at predicted future locations of pathological aggregates, or at predicted past locations of pathological protein aggregates, or combinations thereof. The system may be an open-loop or closed-loop system configured to perform the methods provided herein. In some embodiments, the system may include a DBS electrode adapted for positioning in a region of the brain where pathological protein aggregates are known to exist (e.g., by medical imaging) or are predicted to appear in the past, present, or future, to deliver electrical stimulation to that region of the brain. In a closed-loop system, the system may also include a computational means and a control unit programmed to instruct the DBS electrodes to apply electrical stimulation to areas of the brain where pathological protein aggregates are present or predicted to appear in a manner effective to treat a neurological or neurodegenerative disorder in the subject. In certain embodiments, the neurostimulation intervention can take the form of non-invasive stimulation, including transcranial electrical stimulation or transcranial magnetic stimulation. In some embodiments, one or more programmed stimulation parameters are modulated according to an algorithmic control law based on the location where pathological protein aggregates are present or predicted to appear, and the modulated electrical stimulation is delivered to the brain via the control unit, pulse generator, and DBS electrodes in a manner effective to treat a neurological or neurodegenerative disorder. The closed-loop system may include an on-body pulse generator connected to the implanted DBS electrodes, and thus can automatically apply electrical stimulation to the brain upon receiving communication from a control unit programmed according to the computer-implemented methods described herein.

[0208] The processor of the closed-loop system may execute programming as described herein to predict where pathological protein aggregates will occur in the subject's brain and / or evaluate the effectiveness of the treatment, and may modulate parameters of the treatment as needed without user intervention. Thus, the closed-loop system may not necessarily include a user interface through which a user can instruct DBS electrodes to apply electrical stimulation to the brain to treat the subject's neurological or neurodegenerative disease. However, in some embodiments, a user interface may be included within the closed-loop system that can be used to confirm, override, or modify the recommendations of the closed-loop system.

[0209] Components of systems for carrying out the methods of the present disclosure are further described in the Examples below.

[0210] Administration of pharmacological agents Embodiments of the methods and systems provided in the present disclosure may also include administration of an effective amount of at least one pharmacological agent. An "effective amount" refers to a dosage sufficient to treat a neurological or neurodegenerative disease in a subject as desired. The effective amount may vary somewhat from subject to subject and may depend on factors such as the subject's age and health, the type of neurological or neurodegenerative disease, the severity of the neurological or neurodegenerative disease being treated, the duration of treatment, the nature of any concurrent treatments, the form of the agent, the pharmaceutically acceptable carrier, if used, the route and method of delivery, and similar factors within the knowledge and expertise of those skilled in the art. Appropriate dosages may be determined according to routine pharmacological procedures known to those skilled in the art, as described in more detail below.

[0211] When pharmacological approaches are used in the treatment of neurological or neurodegenerative disorders, the specific nature of the agent and dosing schedule will vary depending on the specific nature of the disorder being treated. Exemplary pharmacological agents that may be used to treat Parkinson's disease include L-DOPA (l-3,4-dihydroxyphenylalanine, also known as levodopa), carbidopa (N-amino-α-methyl-3-hydroxy-L-tyrosine monohydrate), carbidopa-levodopa (Rytary, Sinemet, Duopa), dopamine agonists (pramipexole (Mirapex)), and the like. monoamine oxidase B (MAO-B) inhibitors (including, but not limited to, selegiline (Zelapar), rasagiline (Azilect), and safinamide (Xadago); catechol O-methyltransferase (COMT) inhibitors (including, but not limited to, entacapone (Comtan), opicapone (Ongentys), and tolcapone (Tasmar)); anticholinergics (including but not limited to benztropine (Cogentin) and trihexyphenidyl); adenosine receptor antagonists (including but not limited to A2A receptor antagonists, e.g., istradefylline (Nourianz)), or antipsychotics (including but not limited to nufazide (Pimavanserin)), or any combination thereof.

[0212] In certain aspects, administration of the pharmacological agent involves the use of a pharmacological delivery device, such as, but not limited to, a pump (implantable or external device), an epidural syringe, a syringe or other injection device, a catheter, and / or a reservoir operably associated with a catheter. For example, in certain embodiments, a delivery device used to deliver at least one pharmacological agent to a subject can be a pump, syringe, catheter, or reservoir operably associated with a connecting device, such as a catheter or tube. Containers suitable for delivery of at least one pharmacological agent to a pharmacological agent administration device include containment devices that can be used to deliver, place, attach, and / or insert at least one pharmacological agent into a delivery device for administration of the pharmacological agent to a subject, and include, but are not limited to, vials, ampoules, tubing, capsules, bottles, syringes, and bags. Administration of the pharmacological agent can be performed by a user or by a closed-loop system.

[0213] kit Also provided are kits that include software for implementing the computer-implemented methods described herein for predicting locations where pathological protein aggregates will occur in the brain of a subject with a neurological or neurodegenerative disease and / or for instructing a neurostimulation device to apply neurostimulation to brain locations where pathological protein aggregates are predicted to be present (e.g., past, present, and / or future) to treat the neurological or neurodegenerative disease in the subject. In some embodiments, the kit includes a non-transitory computer-readable medium and instructions for using the computer-implemented methods as described herein to treat a subject with a neurological or neurodegenerative disease based on locations where pathological protein aggregates are predicted to occur in the subject's brain. In some embodiments, the kit includes a system including a processor programmed according to the computer-implemented methods described herein and a display component for displaying information regarding locations where pathological protein aggregates are predicted to occur in the subject's brain.

[0214] In addition to the above components, the subject kits may further include (in certain embodiments) instructions for practicing the subject methods. These instructions may be present in the subject kits in a variety of forms, one or more of which may be present in the kit. One form in which these instructions may be present is as information printed on a suitable medium or substrate (e.g., one or more sheets of paper having the information printed thereon), in the kit packaging, in a package insert, etc. Another form in which these instructions may be present is as a computer-readable medium having the information recorded thereon, such as a diskette, compact disc (CD), flash drive, etc. Another form in which these instructions may be present is as a website address that can be used via the Internet to access the information at a remote location.

[0215] usefulness The disclosed methods and systems can be used to optimize neurostimulation therapy to alter pathology for the treatment of neurological or neurodegenerative diseases. Computer-implemented methods are provided for optimizing neurostimulation therapy parameters, including, but not limited to, neurostimulation location, intensity, and frequency. For example, the location of neurostimulation can be set based on where an algorithm predicts pathological protein aggregates will appear. Other parameters, such as stimulation frequency and pulse width, can then be set to target specific neuronal cell types or circuits within the brain.

[0216] In particular, the subject methods can be used to treat synucleinopathies, including any disease associated with alpha-synuclein aggregation. Synucleinopathies include neurodegenerative diseases associated with the pathological accumulation of alpha-synuclein aggregates in neurons or glia, including, but not limited to, Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophies such as infantile neuroaxonal dystrophy and Hallervorden-Spatz syndrome, Shy-Drager syndrome, striatonigral degeneration, and olivopontocerebellar atrophy. The subject methods can also be used to treat neurodegenerative diseases in which alpha-synuclein pathology contributes to the pathological progression of the disease but is not the primary protein component of the disease-associated pathology, such as Alzheimer's disease, amyotrophic lateral sclerosis, and Pick's disease.

[0217] Examples of Non-Limiting Aspects of the Disclosure Aspects, including embodiments of the present subject matter described above, may be beneficial alone or in combination with one or more other aspects or embodiments. Without limiting the foregoing, certain non-limiting aspects of the present disclosure, numbered 1 through 95, are provided below. As would be apparent to one of skill in the art upon reading this disclosure, each of the individually numbered aspects may be used or combined with any of the preceding or following individually numbered aspects. This is intended to provide support for all such combinations of aspects, and is not limited to the combinations of aspects explicitly provided below.

[0218] 1. A computer-implemented method for predicting the location of occurrence of pathological protein aggregates in the brain of a subject having a neurological or neurodegenerative disease, the method comprising: a) receiving an image of a subject's brain; b) using machine learning algorithms to identify pathological protein aggregates in the images; c) mapping the location of pathological protein aggregates to neuroanatomical regions; and d) modeling the discretized distribution of pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model; and e) calculating the distribution of gene effects on the regional spread and disruption of pathological protein aggregates using regional gene density maps for each neuroanatomical region; f) predicting changes in regional density of pathological protein aggregates as a function of time by modeling the spreading, aggregation, disintegration, and spatial gene expression of pathological protein aggregates across neuroanatomical regions, where spreading is assumed to occur retrogradely between anatomically interconnected neuroanatomical regions, and spreading is modeled as diffusion through a weighted directed graph connecting neuroanatomical regions of the subject's brain; g) predicting past, current, and future locations of pathological protein aggregates based on the modeling. 2. The computer-implemented method of embodiment 1, further comprising adjusting one or more programmed neurostimulation parameters based on predicting changes in the regional density of pathological protein aggregates as a function of time. 3. The computer-implemented method of aspect 2, wherein the one or more programmed neurostimulation parameters are selected from duration, amplitude, frequency, pulse width, and location of the neurostimulation. 4. The computer-implemented method of any one of aspects 1-3, further comprising instructing a neurostimulation device to apply neurostimulation to brain locations predicted to have pathological protein aggregates present, to treat a neurological or neurodegenerative disease in the subject. 5. The computer-implemented method of embodiment 4, further comprising instructing a neurostimulation device to apply electrical stimulation to brain locations where pathological protein aggregates are not yet present but are predicted to appear in the future based on predicting changes in regional density of pathological protein aggregates as a function of time.

[0219] 6. Performing image registration to a coordinate space containing a plurality of voxels, each voxel being represented as a cubic volume element centered at a coordinate in the coordinate space; identifying the x, y, z coordinate location of each pathological protein aggregate within coordinate space; determining the volume of each pathological protein aggregate from the total number of voxels occupied by each pathological protein aggregate; calculating an aggregate density for each voxel, the aggregate density for each voxel being determined from the total number of pathological protein aggregates centered within the same voxel; calculating a total aggregate size for each voxel, the total aggregate size being the total size of all pathological protein aggregates centered within the same voxel; calculating an average aggregate size for each voxel as the total aggregate size for each voxel divided by the aggregate density for each voxel; calculating the total signal intensity of each voxel from the total intensities of all pathological protein aggregates centered within the same voxel; 6. The computer-implemented method of any one of aspects 1-5, further comprising: calculating a mean signal intensity for each voxel as the total signal intensity for each voxel divided by the aggregate density. 7. Modeling the discretized distribution of pathological protein aggregates in each neuroanatomical region as a set of differential equations using the Smoluchowski network model is

[0220]

number

[0221] where c i,j7. The computer-implemented method of any one of aspects 1-6, wherein x represents the total count of pathological protein aggregates in a discretized size bin indexed by / in a brain region indexed by j, the L matrix represents the Laplacian matrix of a weighted directed graph connecting neuroanatomical regions of the brain, η is selected as a hyperparameter that slows down the spread of large aggregates as an inverse power of their size, and λ is selected as a hyperparameter that accelerates the collapse of pathological protein aggregates proportionally to a power of their size. 8. The computer-implemented method of aspect 7, wherein the initial values ​​of α and μ are adapted by sweeping a two-dimensional grid and selecting the values ​​that result in the smallest mean square error between the predicted and actual counts of pathological protein aggregates. 9. The computer-implemented method of embodiment 7, wherein the initial value of μ=0, the initial value of k=0, ξ=0, and c is a one-dimensional size vector, and the set of differential equations simplifies to a standard network diffusion model.

[0222] 10. The computer-implemented method of any one of aspects 7 to 9, further comprising quantifying the sensitivity of the model to a particular neuroanatomical pathway in the brain, wherein a Jacobian matrix is ​​calculated by taking partial derivatives of the output of the model with respect to weights of anatomical coupling strengths between two neuroanatomical regions encoded in the model, and elements of the Jacobian matrix represent the relative contributions of the anatomical coupling between the two neuroanatomical regions to the spread of pathological protein aggregates to a particular region of the brain. 11. Use the model to rank candidate seed locations for a given pathological condition c at t = T months post injection (MPI): using each of the neuroanatomical regions simulated forward in time until t=T MPI as a separate seed location at t=0 in the model, wherein each of the simulation results for the different neuroanatomical regions is compared to the observed state c using a pairwise similarity metric, the similarity metric being the correlation coefficient between total regional aggregate counts across the observed and simulated states; 11. The computer-implemented method of any one of aspects 7-10, further comprising: using the similarity metric value to sort seed locations for neuroanatomical regions as likely to lead to an observed pathological condition c, wherein a ranking of candidate seed locations for a given pathological condition c in the t=T MPI is generated. 12. Let the time since seeding t = T MPI for a given pathological condition c be comparing the whole-brain distribution of aggregate sizes for state c to simulated distributions at various t using a pairwise similarity metric without considering seed locations, where the distribution of simulated aggregate sizes across the whole brain is assumed to be invariant with respect to which neuroanatomical region is used as the seed location at t=0; 12. The computer-implemented method of claim 11, further comprising predicting by a method comprising: calculating a mean squared error between the stimulus distribution and the observation distribution, where, when determining from among multiple candidate t values, the mean squared errors are inverted and normalized to sum to one to provide a predicted probability for each t being a correct estimate of T for a given pathological condition c.

[0223] 13. Genetic effects on the regional spread and disintegration of pathological protein aggregates Assuming that the α (spread) and μ (decay) parameters are region-dependent, with the spread from a particular neuroanatomical region being proportional to the gene density in that region; normalizing all genes to the same range so that only the regional distribution of gene expression is compared to the whole-brain expression of that gene, where α is a vector and the product of α and the Laplacian connectivity matrix L has the effect of modifying the regional connectivity encoded in the model; 13. The computer-implemented method of any one of aspects 7-12, wherein the predicted and observed whole-brain aggregate counts are determined by a method comprising: normalizing each gene vector to have a mean of 1 and a standard deviation Σ empirically set to preserve the correlation between the predicted and observed whole-brain aggregate counts, wherein the normalization is selected to have the effect of maintaining the trace of the original Laplacian connectivity matrix L. 14. The derivation of normalization to preserve the trace of the original Laplacian connectivity matrix L is Assume that a vector s is sampled from a multivariate normal distribution with mean 1 and standard deviation Σ, where s ∼N(1,Σ); Using the definition of matrix trace, express s as a diagonal square matrix S, where the trace of the product of S and the Laplacian connectivity matrix L is

[0224]

number

[0225] where l represents the diagonal of L, the trace is equal to the dot product of s and l, and has expectation equal to the sum of the entries of l, s·l~N(1·l,lΣl) E[s·l]=Tr(L) After each gene is coded into the model, the definition of the trace L is restored according to the formula: Comparing the net effect on regional correlation between the simulated data and the actual data with a baseline correlation that does not include the gene; and providing an ordered list of genes ranked by their spatial expression map relevance in improving the model's regional predictions. 15. The computer-implemented method of any one of aspects 1-14, wherein the cubic volume element has a width of 100 μm in coordinate space.

[0226] 16. The computer-implemented method of any one of aspects 1-15, wherein the one or more pathological protein aggregates are mapped to a single voxel. 17. The computer-implemented method of any one of aspects 1-16, further comprising performing multi-dimensional Gaussian filtering to account for variations in image registration between different samples. 18. The computer-implemented method of any one of aspects 1-17, further comprising segmenting the image to generate a plurality of image segments. 19. The computer-implemented method of any one of aspects 1-18, wherein mapping comprises mapping locations of pathological protein aggregates to neuroanatomical regions of the Allen Human Brain Reference Atlas. 20. The computer-implemented method of embodiment 19, wherein the mapping comprises performing image registration to transform the locations of the pathological protein aggregates into the coordinate space of the Allen Brain Reference Atlas.

[0227] 21. The computer-implemented method of aspect 19 or 20, wherein the anatomically interconnected neuroanatomical regions are identified from the Allen Connectivity Atlas. 22. Neuroanatomical areas include anterior amygdala region, anterior cingulate region, dorsal, anterior cingulate region, ventral, nucleus accumbens, anterior dorsal nucleus, anterior hypothalamic nucleus, agranular insular region, dorsal, agranular insular region, posterior, agranular insular region, ventral, nucleus ambiguus, anterior medial nucleus, dorsal, anterior medial nucleus, ventral, oblique lobule, accessory olfactory bulb, anterior olfactory nucleus, anterior pretectal nucleus, arcuate hypothalamic nucleus, dorsal auditory cortex, primary auditory cortex, ventral auditory cortex, anterior ventral nucleus of the thalamus, basolateral amygdala, basomedial amygdala, bed nucleus of the stria terminalis, CA1 area, CA2 area, CA3 area, central nucleus of the amygdala, central lobule, lateral central nucleus of the thalamus, claustrum, central linear raphe, medial central nucleus of the thalamus, cortico-amygdala region, anterior, cortico-amygdala region, posterior region, caudate nucleus, superior central raphe, anterior lobe of the cerebellum, cuneiform nucleus, dorsal cochlear nucleus, dentate gyrus, dorsomedial nucleus of the hypothalamus, dentate nucleus, dorsal peduncular region, dorsal nucleus raphe, external rhinal region, entorhinal region, lateral region, entorhinal region, medial region, dorsal zone, caudate nucleus, dorsal, caudate nucleus, ventral, flocculus, dorsal column, cerebral cortex, basal striatum, globus pallidus, external segment, globus pallidus, internal segment, nucleus reticularis magnocellularis, gustatory region, intergeniculate nucleus of the amygdala, inferior colliculus, central nucleus, inferior colliculus, dorsal nucleus, inferior colliculus, external nucleus, inferior colliculus, infralimbic region, intermediate dorsal nucleus of the thalamus, inferior olivary complex, nucleus intermedius, interpeduncular nucleus, intermediate reticular nucleus, lateral amygdala, lateral vestibular nucleus, lateral dorsal nucleus of the thalamus, dorsal of the lateral geniculate complex, lateral geniculate nucleus Ventral, lateral habenula, lateral hypothalamic area, lateral posterior nucleus of the thalamus, lateral preoptic area, lateral reticular nucleus, lateral septal nucleus, caudal (caudodorsal) part, lateral septal nucleus, rostral (rostroventral) part, lateral septal nucleus, ventral, magnocellular nucleus, magnocellular reticular nucleus, medial dorsal nucleus of the thalamus, medullary reticular nucleus, dorsal part, medullary reticular nucleus, ventral, medial squamous peach nucleus, median preoptic nucleus, medial geniculate complex, dorsal part, medial geniculate complex, medial part, medial geniculate complex, abdomen, medial habenula, medial mammillary nucleus, main olfactory bulb, primary motor area, secondary motor area, medial preoptic nucleus, medial preoptic area, medial pretectal area, mesencephalic reticular nucleus, medial septal nucleus, medial vestibular nucleus, diagonal band nucleus, nuclear indeterminate, lateral Nuclei of the lateral lemniscus, nuclei of the lateral olfactory tract, tubercle (X), nuclei of the optic tract, nuclei of the posterior commissure, nucleus of the solitary tract, orbital region, lateral part, orbital region, medial part, orbital region, ventrolateral part, olfactory tubercle, posterior amygdaloid nucleus, piriform amygdaloid region, periaqueductal gray matter, paraunculoid gyrus, parvocellular reticular nucleus, parabrachial nucleus, central pontine gray matter, perirhinal region, parafascicular nucleus, paraflocculus lobule, pontine gray matter, paracellular reticular nucleus, dorsal part, paracellular reticular nucleus, lateral part, posterior nucleus of the hypothalamus, piriform region, prelimbic area, dorsal premammillary nucleus, posterior thalamic complex, posterior bordering nucleus of the thalamus, posterior uncinate gyrus, peripeduncular nucleus, pedunculopontine nucleus, preuncleate gyrus, paramedian lobule, pontine reticular nucleus, caudal part, pontine reticular nucleus, precursor nucleus, principal sensory nucleus of the trigeminal nerve, parathalamic nucleus,Posterior parietal-related area, paraventricular hypothalamic nucleus, paraventricular thalamic nucleus, periventricular hypothalamic nucleus, posterior, periventricular hypothalamic nucleus, preoptic tract, lobule VIII, postchiasmatic area, nucleus of reuniens, rhomboid nucleus, nucleus raphe magnus, nucleus red, midbrain reticular nucleus, posterior ruber area, retrosplenial area, lateral agranular area, retrosplenial area, dorsal, retrosplenial area, ventral, thalamic reticular nucleus, paraventricular area, superior colliculus, motor-related , superior colliculus, sensory related, septum fimbria, substantia innominata, lobule simplex, medial subthalamic nucleus, substantia nigra, pars compacta, substantia nigra, pars reticulata, superior olivary complex, subparafascicular area, subparafascicular nucleus, magnocellular part, subparafascicular nucleus, parvocellular part, spinal vestibular nucleus, spinal nucleus of the trigeminal nerve, tail, spinal nucleus of the trigeminal nerve, interpolar part, spinal nucleus of the trigeminal nerve, oral part, primary somatosensory cortex, barrel area, primary somatosensory cortex, lower limbs 22. The computer-implemented method of any one of aspects 1-21, wherein the visual field is selected from: primary somatosensory cortex, mouth, primary somatosensory cortex, nose, primary somatosensory cortex, trunk, primary somatosensory cortex, upper limb, additional somatosensory area, subthalamic nucleus, uncinate gyrus, supramammillary nucleus, supratrigeminal nucleus, superior vestibular nucleus, temporal association area, posterior piriform transition area, nucleus of the tegmental reticularis, nucleus septalis triangularis, stria tecta, nucleus tuberculosa, motor nucleus of the trigeminal nerve, ventral anterior thalamic complex, ventral cochlear nucleus, facial motor nucleus, visceral area, anterior lateral visual area, anterior medial visual area, lateral visual field, primary visual area, posterolateral visual area, posteromedial visual area, ventromedial nucleus of the thalamus, ventromedial nucleus of the hypothalamus, ventral posterolateral nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, parvocellular portion, ventral tegmental area, and hypoglossal nucleus.

[0228] 23. The computer-implemented method of any one of aspects 1-22, further comprising predicting where in the subject's brain the pathological protein aggregates have occurred. 24. The computer-implemented method of any one of aspects 1-23, wherein the machine learning algorithm uses an artificial neural network. 25. The computer-implemented method of any one of aspects 1-24, wherein the machine learning algorithm uses a deep learning algorithm. 26. The computer-implemented method of aspect 25, wherein the deep learning algorithm uses a convolutional neural network, a deep neural network, a recurrent neural network, a deep residual neural network, a long short-term memory network, a deep belief network, a multilayer perceptron, or deep reinforcement learning. 27. The computer-implemented method of any one of aspects 1-26, wherein the machine learning algorithm is supervised, semi-supervised, or unsupervised. 28. The computer-implemented method of any one of aspects 1-27, wherein the subject is a human subject.

[0229] 29. The computer-implemented method of aspect 28, wherein modeling the spread, aggregation, and disintegration of pathological protein aggregates in the brain of a human subject uses simulations generated based on measuring the spread, aggregation, and disintegration of pathological protein aggregates in a non-human animal model, and the experimentally measured pathology in the non-human animal model is used to design simulation parameters for predicting the past, current, and future locations of pathological protein aggregates in the human subject. 30. The computer-implemented method of aspect 29, wherein the simulation further uses measured brain anatomy or brain function from a non-human animal to predict the past, current, and future locations of pathological protein aggregates in a human subject. 31. The computer-implemented method of aspect 29 or 30, wherein the non-human animal is a mammal. 32. The computer-implemented method of aspect 31, wherein the mammal is a rodent or a primate. 33. The computer-implemented method of aspect 32, wherein the rodent is a mouse. 34. The computer-implemented method of any one of aspects 29 to 33, wherein the simulation further uses brain anatomy or brain function measured from other human subjects to design simulation parameters for predicting past, current, and future locations of pathological protein aggregates in the human subject.

[0230] 35. receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using a second image; and 35. The computer-implemented method of any one of aspects 1 to 34, further comprising: displaying a change in total aggregate size for each voxel, a volume of each pathological protein aggregate for each voxel, and an aggregate density for each voxel in images taken after the neurostimulation is applied to the brain of the subject compared to images taken before the neurostimulation is applied to the brain of the subject. 36. receiving a second image of the brain, the second image being taken after neurostimulation is applied to the brain; repeating steps (b) through (o) using a second image; and modulating one or more programmed neurostimulation parameters based on any changes in past, current, or future locations where pathological protein aggregates are predicted to appear; 35. The computer-implemented method of any one of aspects 1-34, further comprising: instructing a neurostimulation device to apply modulated neurostimulation to a brain of the subject to treat a neurological or neurodegenerative disease in the subject.

[0231] 37. The computer-implemented method of any one of aspects 1-36, wherein the computer-implemented method further comprises storing a user profile for the subject, the user profile including information regarding programmed neurostimulation parameters used to apply neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease based on the location where pathological protein aggregates are present or predicted to occur in the future.

[0232] 38. A non-transitory computer-readable medium comprising program instructions that, when executed by a processor in a computer, cause the processor to perform a method according to any one of aspects 1-37. 39. A kit comprising the non-transitory computer-readable medium of aspect 38 and instructions for treating a neurological or neurodegenerative disease in a subject using neurostimulation.

[0233] 40. A method for treating a neurological or neurodegenerative disease in a subject, the method comprising: imaging pathological protein aggregates in the brain of a subject; predicting where pathological protein aggregates will occur based on the locations of pathological protein aggregates detected in the subject's brain by imaging, using the computer-implemented method of any one of aspects 1-37; applying neurostimulation to locations in the brain where pathological protein aggregates are detected in the subject's brain by imaging and where the computer-implemented method predicts that the pathological protein aggregates will occur. 41. The method of aspect 40, wherein the imaging is performed using computed tomography (CT), single photon emission computed tomography (SPECT), magnetic resonance imaging, functional magnetic resonance imaging, optogenetic functional magnetic resonance imaging, or positron emission tomography (PET). 42. The method of aspect 40 or 41, further comprising adjusting the stimulation frequency and pulse width of the neurostimulation to target specific neuronal cell types or circuits in the brain at locations in the brain where the computer-implemented method predicts that pathological protein aggregates are present or will occur in the future. 43. The method of any one of aspects 40 to 42, wherein the neurological or neurodegenerative disease is a synucleinopathy. 44. The method of aspect 43, wherein the synucleinopathy is Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy.

[0234] 45. The method of any one of aspects 40-42, wherein the neurological or neurodegenerative disease is Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia. 46. ​​The method of any one of aspects 40-45, wherein the pathological protein aggregates comprise alpha-synuclein aggregates. 47. The method of any one of aspects 40-46, wherein applying the neurostimulation includes applying the neurostimulation using electrodes. 48. The method of aspect 47, wherein the electrode is a deep electrode or a surface electrode. 49. The method of aspect 47, wherein the electrode is a non-brain penetrating surface electrode array or a brain penetrating electrode array.

[0235] 50. The method of any one of aspects 40-49, wherein applying the neurostimulation comprises applying deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation. 51. The method of any one of aspects 40-46, wherein applying the neural stimulation comprises optogenetically applying the neural stimulation. 52. Nerve stimulation is introducing a recombinant polynucleotide encoding a light-responsive ion channel into a neuron at a location in the brain where a pathological protein aggregate is present or predicted to occur in the future by a computer-implemented method, wherein the light-responsive ion channel is expressed in the neuron; 52. The method of embodiment 51, wherein the neuron is optogenetically applied by a method comprising: illuminating the light-responsive ion channel with light of a wavelength that activates the light-responsive ion channel, wherein conduction of ions by the light-responsive ion channel in response to absorption of the light results in hyperpolarization or depolarization of the neuron. 53. The method of aspect 52, wherein the light-responsive ion channel is a light-responsive anion-conducting opsin or a light-responsive proton conductance modulator. 54. Light-responsive anion-conducting opsins are capable of transporting chloride ions (Cl - 54. The method of embodiment 53, wherein the

[0236] 55. The method of aspect 53 or 54, wherein the anion-conducting opsin is an anion-conducting channelrhodopsin or halorhodopsin. 56. The method of aspect 55, wherein the halorhodopsin is Natronomonas pharaonis halorhodopsin (NpHR), enhanced NpHR (eNpHR) 1.0, eNpHR 2.0, or eNpHR 3.0. 57. The method of aspect 55, wherein the anion-conducting channelrhodopsin is iC1C2, SwiChR, SwiChR++, or iC++. 58. The method of aspect 53, wherein the light-responsive proton conductance modulator is bacteriorhodopsin or archaealhodopsin. 59. The method of aspect 58, wherein the light-responsive proton conductance modulator is Arch from Halorubrum sodomense, ArchT from Halorubrum sp., TP009 from Leptosphaeria maculans, or Mac from Leptosphaeria maculans.

[0237] 60. The method of embodiment 52, wherein the light-responsive ion channel is a light-responsive cation-conducting opsin. 61. Light-responsive cation-conducting opsins are photoconductive opsins that react with calcium cations (Ca 2+ 61. The method of embodiment 60, wherein the 62. The method of aspect 60 or 61, wherein the light-responsive cation-conducting opsin is a light-responsive cation-conducting channelrhodopsin. 63. The method of aspect 62, wherein the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin or Volvox carteri channelrhodopsin. 64. The method of aspect 63, wherein the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin-1 (ChR1), Chlamydomonas reinhardtii channelrhodopsin-2 (ChR2), Volvox carteri channelrhodopsin-1 (VChR1), or a chimeric ChR1-VChR1 channelrhodopsin.

[0238] 65. The method of any one of aspects 52 to 64, wherein the polynucleotide encoding the light-responsive ion channel is provided by a viral vector. 66. The method of aspect 65, wherein the viral vector is a lentiviral vector or an adeno-associated viral (AAV) vector. 67. The method of aspect 65 or 66, wherein the viral vector is stereotactically injected into the brain at a location where the computer-implemented method predicts that pathological protein aggregates are present or will occur in the future. 68. The method of any one of aspects 65-67, wherein the vector further comprises a neuron-specific promoter operably linked to the polynucleotide encoding the light-responsive ion channel. 69. The method of any one of aspects 52 to 68, wherein expression of the light-responsive ion channel is inducible.

[0239] 70. The method of any one of aspects 52-69, wherein illuminating the light-responsive ion channel comprises delivering light from a light source to the light-responsive ion channel using an optical fiber-based optical neural interface. 71. The method of aspect 70, wherein the light source is a solid-state diode laser. 72. The method of any one of aspects 52-71, wherein applying the neurostimulation comprises applying the neurostimulation to a motor cortical region or a subcortical region of the brain. 73. The method of any one of aspects 40-72, wherein multiple cycles of neural stimulation are performed. 74. The method of any one of aspects 40-73, further comprising assessing the effectiveness of treating a neurological or neurodegenerative disease in the subject.

[0240] 75. The method of aspect 74, wherein the evaluating comprises imaging the subject's brain after neurostimulation to measure the size and identify the location of pathological protein aggregates. 76. The method of aspect 74 or 75, wherein evaluating comprises measuring the subject's brain function after neurostimulation. 77. The method of aspect 76, wherein measuring brain function comprises performing electroencephalography (EEG), stereoencephalography (sEEG), electrocorticography (ECoG), magnetoencephalography (MEG), single-photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), optogenetic functional magnetic resonance imaging, or positron emission tomography (PET). 78. The method of aspect 76 or 77, further comprising modulating one or more programmed neurostimulation parameters to improve brain function. 79. The method of any one of aspects 74-78, further comprising assessing the severity of symptoms of the neurological or neurodegenerative disease using a visual analog scale, a verbal rating scale, the Movement Disorder Society-sponsored Unified Parkinson's Disease Rating Scale-Revised (MDS-UPDRS), the Hoehn and Yahr (HnY) scale, or the Montreal Cognitive Assessment (MoCA) scale.

[0241] 80. A system for treating a neurological or neurodegenerative disease in a subject, the system comprising: A neurostimulation device; 38. A system comprising: a processor programmed according to the computer-implemented method of any one of aspects 1 to 37, wherein the processor is programmed to instruct a neurostimulation device to deliver neurostimulation to a brain of a subject in a manner effective to treat a neurological or neurodegenerative disease in the subject, wherein the neurostimulation is applied to the brain at a predicted current location of a pathological protein aggregate, at a predicted future location of a pathological protein aggregate, or at a predicted past location of a pathological protein aggregate, or a combination thereof. 81. The system of aspect 80, wherein the neurostimulation device comprises an electrode. 82. The system of aspect 81, wherein the electrode is a deep electrode or a surface electrode. 83. The system of aspect 82, wherein the electrode is a non-brain penetrating surface electrode array or a brain penetrating electrode array. 84. The system of any one of aspects 80 to 83, wherein the neurostimulation device performs deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation.

[0242] 85. The system of any one of aspects 80-84, further comprising a display. 86. The system of aspect 85, wherein the display displays an image of the subject's brain showing the predicted current location, past location, or future location of the pathological protein aggregates determined by the computer-implemented method. 87. The system of aspect 85 or 86, wherein the display displays information regarding the coordinates of each pathological protein aggregate, the volume of each pathological protein aggregate, the aggregate density of each voxel, the total aggregate size of each voxel, the average aggregate size of each voxel, the total signal intensity of each voxel, the average signal intensity of each voxel, or a mapping of the location of the pathological protein aggregate to a neuroanatomical region, or any combination thereof. 88. The system of any one of aspects 85 to 87, wherein the display displays information regarding the distribution of gene effects on the regional spread and disruption of pathological protein aggregates. 89. The system of any one of aspects 85 to 88, wherein the display displays information regarding the spread, aggregation, and disintegration of pathological protein aggregates across a neuroanatomical region, and predicted changes in the regional density of pathological protein aggregates as a function of time as determined by modeling spatial gene expression.

[0243] 90. A system described in any one of aspects 85 to 89, wherein the display displays information regarding predicted past locations, current locations, and future locations of pathological protein aggregates based on modeling. 91. The system of any one of aspects 85 to 90, further comprising a user interface having an input electronically coupled to the processor for instructing the neurostimulation device to apply neurostimulation to the subject's brain to treat a neurological or neurodegenerative disease in the subject. 92. The system of aspect 91, wherein the user interface is password protected and operable by a medical professional. 93. The system of any one of aspects 85 to 92, wherein the neurological or neurodegenerative disease is a synucleinopathy. 94. The system of aspect 93, wherein the synucleinopathy is Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy. 95. The system of any one of aspects 85 to 92, wherein the neurological or neurodegenerative disease is Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia.

[0244] It will be apparent to those skilled in the art that various changes and modifications can be made in the present invention without departing from the spirit or scope of the present invention.

[0245] experiment The following examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the present invention, and are not intended to limit the scope of what the inventors regard as the invention, nor are they intended to represent that the following experiments are all or the only experiments performed. Efforts have been made to ensure accuracy with respect to numbers used (e.g., amounts, temperatures, etc.), but some experimental error and deviation should be accounted for. Unless otherwise noted, parts are parts by weight, molecular weight is weight average molecular weight, temperature is in degrees Celsius, and pressure is at or near atmospheric.

[0246] All publications and patent applications cited herein are herein incorporated by reference to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference.

[0247] The present invention has been described in terms of particular embodiments discovered or proposed by the inventors to comprise preferred modes for practicing the invention. Those skilled in the art will, in light of this disclosure, appreciate that numerous modifications and changes can be made in the particular embodiments exemplified without departing from the intended scope of the invention. All such modifications are intended to be included within the scope of the appended claims.

[0248] Example 1 Mesoscale binding and gene expression enable whole-brain modeling of α-synuclein spreading, aggregation, and decay kinetics Parkinson's disease (PD) is the second most common neurodegenerative disorder. It is characterized by postural instability, tremor, rigidity, and bradykinesia (Goetz, 2011; Kalia and Lang, 2015). These clinical symptoms are primarily caused by the loss of dopaminergic neurons from the substantia nigra. A hallmark pathology of PD is the presence of Lewy bodies (LBs), cytoplasmic neuronal inclusions composed of misfolded aggregates of the protein α-synuclein (α-syn) (Dickson, 2012; Goedert et al., 2013; Kalia and Lang, 2015; Oliveira et al., 2021; Spillantini et al., 1997). A long-standing hypothesis for the pathogenesis of this debilitating disease is the Braak hypothesis, which postulates that pathological α-syn seeds form early in the disease, subsequently spread throughout the nervous system, and correlate with the progression of motor and cognitive symptoms (Beach et al., 2009; Braak et al., 2002; Braak et al., 2003). An exciting area in neurodegenerative disease research is the emerging phenomenon of prion-like spread of neurodegenerative disease proteins, including α-syn, in PD (Aguzzi and Rajendran, 2009; Angot et al., 2010; Cushman et al., 2010; Guo and Lee, 2014; Jucker and Walker, 2013). Prions are well established as the protein-based infectious agents underlying spongiform encephalopathies (e.g., bovine spongiform encephalopathy in cattle and Creutzfeldt-Jakob disease in humans). In these rare but devastating diseases, the prion protein PrP converts from a normal soluble form to an aggregated, self-templated, infectious form. This process initiates the relentless spread of pathology throughout the brain and occasional neurodegeneration (Aguzzi and Calella, 2009; Prusiner, 1998). But can this phenomenon be extended to more common neurodegenerative diseases like PD?

[0249] Early hints of this type of possibility came from postmortem analyses of individuals who received fetal nigral transplants as a treatment for PD and subsequently died several years later, leading to autopsy (Kordower et al., 2008; Li et al., 2008). In some of the fetal transplants (who were only 11–16 years old at the time of autopsy), Lewy bodies composed of α-syn fibrils were present. This finding was surprising, as it suggested that α-syn may have somehow propagated from the diseased host tissue into the new graft tissue. These findings were consistent with the long-standing Braak hypothesis that PD pathology appears to spread throughout the brain in an anatomically defined manner (Braak et al., 2003).

[0250] Although several studies have questioned this "prion-like" hypothesis due to its inability to explain the sparse, graded distribution in postmortem human PD brains (Surmeier et al., 2017), studies in primary neurons and animal models have demonstrated that α-syn pathology can and does spread in a cell-to-cell manner, causing excitability impairment and ultimately neuronal degeneration (Desplats et al., 2009; Hansen et al., 2011; Volpicelli-Daley et al., 2011). Injection of α-syn fibrils into α-syn-overexpressing transgenic mice causes pathology and degeneration (Luk et al., 2012b). Importantly, α-syn fibrils injected into wild-type mice cause the spread of pathology along anatomically interconnected brain regions, resulting in a decrease in tyrosine hydroxylase-positive dopaminergic neurons and motor deficits (Luk et al., 2012a). Contrary to arguments that these results may be secondary to signaling pathways induced by aggregated α-syn, injection of α-syn fibrils into α-syn knockout mice does not result in spreading pathology, degeneration, or motor impairment. Furthermore, injection of fibrils into α-syn heterozygous mice results in reduced pathology and motor impairment (Luk et al., 2012a), demonstrating that spreading is α-syn dependent. These observations have been extended to rats, nonhuman primates, and human neurons (Bieri et al., 2019; Gribaudo et al., 2019; Paumier et al., 2015; Prusiner et al., 2015; Recasens et al., 2014; Shimozawa et al., 2017). These results indicate that α-syn transneuronal spreading plays an important role in neurodegeneration in PD.

[0251] A picture emerges in which small amounts of α-syn fibril seeds (either spontaneously formed in the human brain or by direct injection into the mouse brain) can template the conversion of endogenous α-syn into an aggregated state and set in motion a flywheel that inexorably drives the nervous system toward disease. A major challenge now is defining how these aggregates spread from one brain region to the next. Are some regions selectively vulnerable? Are others resistant? These questions need to be answered across space and time. Several recent studies have applied computational network diffusion models to predict early stages of α-syn spreading patterns (Henderson et al., 2019a; Pandya et al., 2019), providing evidence that anatomical connectivity can accurately predict these patterns. Such spatiotemporal models are crucial for further understanding and ultimately treating progressive diseases like PD at variable points in their progression. Recent studies have also used transgenic animal models and cell-specific labeling techniques to explore the genetic determinants behind α-syn spreading (Henderson et al., 2020; Henrich et al., 2020), indicating that the level of expression of genes such as endogenous α-syn or GBA1 is a further important factor. However, previous studies did not quantify complete three-dimensional whole-brain pathology or report spreading patterns beyond 6 months post-injection. We hypothesized that both complete spatial representation and tracking of late-stage spreading would be important for characterizing the changes in pathology and degeneration known to emerge in such progressive neurodegenerative diseases.

[0252] Here, we used tissue clearing and light-sheet fluorescence microscopy to three-dimensionally image α-syn pathology in whole mouse brains and a computational pipeline to anatomically map each neuronal inclusion to the Allen Reference Atlas (ARA). Merging the data into the ARA coordinate system enabled comparison with previous studies that mapped mesoscale axonal projections between neuroanatomical regions (Oh et al., 2014) and spatial transcriptomics across many genes (Lein et al., 2007). Statistical comparison of brain maps at various stages of disease progression revealed biphasic spread and decay curves with distinct timing across regions. Furthermore, longitudinal tracking of the size of individual α-syn aggregates across regions revealed a pattern of steady increase and rapid decrease in average aggregate size per region, suggesting both prion-like aggregation and subsequent neurodegeneration. To capture these simultaneous effects, we developed a computational model incorporating pathology spread, aggregation, decay, and spatial gene expression throughout the brain. The development and validation of this model provides the basis for tracking both the origin and progression of this highly complex disease.

[0253] result Whole-brain quantification of α-synuclein pathology using tissue clearing and light-sheet microscopy Direct injection of α-syn preformed fibrils (PFFs) induces whole-brain pathology and neurodegeneration, serving as a robust model of Parkinson's disease (Henderson et al., 2019b; Luk et al., 2012a). To track pathology throughout disease progression, we injected α-syn PFFs ipsilaterally into the striatum or caudate-putamen (CP) of 8- to 10-week-old mice and processed the mouse brains for immunohistochemistry at various time points up to 18 months postinjection (Figure 1A). Building on recent advances in three-dimensional whole-brain immunolabeling and imaging, we optimized the iDISCO+ protocol to immunolabel α-syn aggregates (using an antibody to specifically detect aggregated endogenous α-syn phosphorylated on serine 129) (Renier et al., 2016; Renier et al., 2014) and imaged the samples using light-sheet microscopy (Figures 1B and 6). We validated virtual sections from these three-dimensional datasets against conventional serial histology (Figure 7) and found them consistent with previous studies using the same injection site (Bieri et al., 2019; Henderson et al., 2019a; Luk et al., 2012a). Pathology developed in both the ipsilateral striatum, intermediate layers of the cortex, and substantia nigra at both 2 and 6 months postinjection (Figure 7A). After imaging, we used a quantification pipeline built on multiple open-source software tools, including Ilastik (Berg et al., 2019) (a machine learning library) and ClearMap (Renier et al., 2016) (a registration library), to detect each α-syn aggregate and assign it to a voxel or anatomical region from the Allen Reference Atlas (Figures 1C, 1D, 8, and 9) (Oh et al., 2014). Because we were able to capture the three-dimensional volume of each α-syn neuronal inclusion (Figure 1D), we were able to visualize aggregate distribution and aggregate size across time points (Figure 1E). Plotting total α-syn aggregate counts in the whole brain against time points in months post-injection (MPI) revealed a biphasic curve (Figure 1F) beginning with pathological spread (between 0 and 6 months post-injection) followed by decay (between 8 and 18 months post-injection).We observed a relatively large spike of smaller aggregates at the earliest time points after injection, but at 4 and 8 months post-injection, the volume distribution shifted to twice the average aggregate volume per voxel. Strikingly, at the final 18-month time point, this distribution shifted back to a lower aggregate burden without the presence or reappearance of smaller-sized aggregates (Figure 1G).

[0254] Statistical analysis across longitudinal groups reveals region-dependent spread, accumulation, and decay. To define the spatiotemporal patterns of α-syn pathology, we performed statistical comparisons at the voxel and region levels between cohorts at various sacrifice time points after injection. Due to variability in spreading patterns across adjacent time points, we instead used time points separated by at least 3 months for these comparisons. Our quantitative pipeline captured both total aggregate counts and average aggregate volumes for each voxel (Figure 2A). Therefore, we performed statistical comparisons of each of these two metrics between each pair of selected time points. When comparing total aggregate counts at the voxel level (Figure 2B), we observed statistically significant clusters for different brain subregions, with widely varying rates of both spreading and decay. For example, while widespread pathology in the cortex was already consolidating and beginning to decay by 4 months postinjection, aggregates only began to appear in various subcortical clusters, including within the thalamus and contralateral hippocampus. Statistics comparing mean aggregate size at the voxel level show a similar initial increase from 0.5 to 4 MPI in both the cortex and subcortex, followed by a whole-brain decrease from 4 to 8 MPI (Figure 2B). Clusters of significant increases or decreases from voxel-level analysis are generally due to anatomical brain region boundaries, and we observed a similar biphasic trend when calculating both α-syn aggregate counts and mean size metrics across regions from the Allen Brain Atlas (Figure 2C). Thus, different brain regions exhibit distinct spatiotemporal patterns / kinetics of α-syn spreading, accumulation, and decay.

[0255] Computational model of spreadingPrevious applications of computational models (Henderson et al., 2019a) only accounted for α-syn spread, despite being able to explain much of the initial regional variation in spread. Therefore, they cannot generalize to later time points throughout the disease, particularly at 4 MPI, when pathology disruption begins to emerge (Figure 2B). To more accurately predict the progression of α-syn pathology, we developed a novel computational model by incorporating mechanistic insights into α-syn pathogenesis and trafficking from recent in vivo and in vitro studies. The key steps in this model consist of α-syn uptake into neurons, intracellular processing and interactions, and ultimately, the release of pathological α-syn. A set of differential equations models the discretized distribution of α-syn aggregate counts in each neuroanatomical region (Figure 3A). This model first assumes rapid uptake of injected α-syn fibrils into neurons within the target region. This is confirmed by studies showing that extracellular α-syn fibrils are incorporated into neurons by endocytosis (Brahic et al., 2016; Desplats et al., 2009; Henderson et al., 2019b; Konno et al., 2012). These injected fibrils are considered the smallest discrete pathological units that can exist in the brain. However, as misfolded α-syn is processed through endolysosomal and cytoplasmic compartments, it can both recruit endogenous α-syn to a pathogenic fibrillar state and fuse with existing fibrils to form larger aggregates. Building on previous work (Bieri et al., 2019), we hypothesize the retrograde spread of any α-syn aggregates through the brain connectome, which our model incorporates as diffusion through a directed weighted graph (Figures 3B and 11A). We derived this anatomical connectivity from the Allen Connectivity Atlas (Oh et al., 2014), which includes 424 regions across the brain (Figures 11A and 11B; Table S1).Our model does not incorporate fragmentation of α-syn aggregates inside the cytoplasm, because previous studies have shown that the fragmentation rate is undetectably low ( Gaspar et al., 2017 ).

[0256] While we set most model parameters a priori, we also fitted model parameters controlling the rate of spreading and decay to data from α-syn PFF injections into the striatum, ranging from 0.5 to 18 MPI. Despite only being fitted to maximize the model's output in predicting whole-brain α-syn aggregate counts (Figures 3C and 3D), the resulting model further captured regional variability in pathology with a high Pearson correlation coefficient of 0.72 (Figure 3C), providing evidence for the theory of a primarily retrograde neuronal spreading mechanism. We tested additional networks based on either inter-regional anterograde connectivity or Euclidean distance, but none were able to capture both whole-brain and regional variability (Figure 11C). Calculating model sensitivity through a Jacobian matrix allows for weighting of brain pathways that explain the most significant α-syn spread. This analysis highlights many retrograde pathways in the cortico-basal ganglia-thalamo-cortical loop. Examples of the highest Jacobian weights are early spread from the ipsilateral striatum to many cortical regions, such as the infralimbic area (ILA), main olfactory bulb (MOB), and gustatory cortex (GU), followed at later times by spread from the cortex to the thalamus and other subcortical regions (Figure 3E).

[0257] Predicting spread patterns for different injection sites After finding that our model, based primarily on anatomical connectivity, could accurately predict pathology resulting from α-syn PFFs injected into the striatum, we next tested the generalizability of this model to different seed locations (i.e., injection sites). We performed additional injections of α-syn PFFs across various brain regions and compared our model's predictions with the actual quantified pathological conditions. Because one of the general theories behind α-synucleinopathies is that a single seeding event can lead to spread throughout the nervous system, we chose a variety of different seed locations relevant to Parkinson's disease and other synucleinopathies: the substantia nigra pars compacta, the main olfactory bulb, and the dentate gyrus (Figure 4A). Using our iDISCO immunolabeling, imaging, and computational processing pipeline, we quantified voxel-level and regional aggregate density and average size for the brain at 0.5 MPI, 2 MPI, and 4 MPI for each seed location. These seed locations induced a remarkably consistent distribution of aggregate sizes (Figure 4B), with the earliest 0.5 MPI maps showing spikes of small aggregates, and this distribution trending toward larger aggregates over time. However, statistical testing of 0.5-4 MPI at both the voxel and neuroanatomical levels yielded clear spatial patterns of spreading and aggregation depending on seed location (Figures 4C and 13B).

[0258] We then applied our computational model to predict regional pathology density for each discretized size bin over time. This involved modifying the initial conditions of the simulated model and integrating forward in time, while keeping all parameter and hyperparameter values ​​fixed. After iterating through each region in silico and generating a complete time series of disease progression, we selected the region that best predicted pathology for a given dataset (Figure 4E). This method consistently predicted the correct ipsilateral hemisphere and neuroanatomical region for the initial seed from among all other seed locations (Figure 4E). Because all additional seed locations showed a consistent distribution of aggregate volume across the brain (Figure 4B), we hypothesized that through inversion of this model, we could also predict duration since injection, which we found to be the case (Figure 4F).

[0259] Coding spatial transcriptomics into region models Consistent with current models (Henderson et al., 2019a), anatomical connectivity appears to be a primary driver of α-syn spreading patterns, and our computational model, with edges simply weighted by anatomical retrograde connectivity, was able to predict highly correlated regional spreading patterns. However, Parkinson's disease is associated with a diverse set of genetic susceptibility factors. Do these converge and influence α-syn spreading? Snca, the gene encoding α-syn itself, directly influences propagation, as knockout of α-syn expression in mice is sufficient to prevent widespread pathology after injection of α-syn PFFs, almost certainly due to the absence of endogenous α-syn converting to aggregated forms (Luk et al., 2012a; Luna et al., 2018; Taguchi et al., 2014). Other PD genes have also been associated with α-syn spreading. Transgenic mice engineered to express a PD-causing mutation in the Lrrk2 gene (LRRK2:G2019S) showed increased α-syn aggregates upon PFF injection (Bieri et al., 2019; Henderson et al., 2019a), similar to human iPS neurons (Bieri et al., 2019).

[0260] We hypothesized that encoding regional gene data into the computational model would enable us to rank the effects of genes on the model's distinct spreading and decay steps, potentially improving the model's predictive power, as measured by the Pearson correlation coefficient of actual versus predicted regional variation. We encoded 19,893 regional gene maps from the Allen In Situ Hybridization (ISH) database (Lein et al., 2007) into our previously fitted computational model using the same 424 regions across the brain (Figure 5A). Resimulating α-syn progression with each encoded gene density map provided the distribution of their effects on spreading and decay parameters (Figure 5B). Interestingly, the Lrrk2 gene improved the model's regional predictions when incorporated into the spreading parameters, where it ranked very highly among all genes (94th percentile). This is consistent with the hypothesis that Lrrk2 is important in vesicular trafficking pathways ( Henderson et al., 2019b ) and recent evidence that reducing Lrrk2 levels reduces α-syn aggregates ( Bieri et al., 2019 ).

[0261] After associating each gene with its most likely cell type using the Allen Cell Type RNA-Seq Database (Tasic et al., 2018), we explored the relationship between a gene's cell type and its ranking. Because our model assumes a neuronal mechanism of transport, we assumed that genes from neuronal cell types would dominate the model's spread term, which is indeed the case (Figure 5C). However, for the decay term in this model, we unexpectedly found a cluster of genes expressed predominantly in oligodendrocytes, with myelin basic protein (Mbp) being the highest-ranked gene (Figure 5C). This finding is consistent with emerging data integrating genome-wide association studies with cell type atlases showing that oligodendrocytes play a key role in Parkinson's disease (Bryois et al., 2020). Furthermore, accumulating evidence suggests that another synucleinopathy, MSA (multiple system atrophy), an aggressive degenerative disease characterized by oligodendroglial cytoplasmic α-syn inclusions, behaves like a prion and that α-syn may indeed be a prion (Prusiner et al., 2015). Furthermore, recent studies suggest that the cytoplasmic environment of oligodendrocytes promotes the formation of particularly potent α-syn seeds that can spread to neurons (Peng et al., 2018). The enrichment of oligodendrocyte genes in our model (Figure 5C) further supports a role for oligodendrocytes in α-syn spreading.

[0262] Finally, we found that resimulating the model using other seed positions and assessing gene importance produced remarkably consistent rankings, regardless of the injection site-based dataset (Figure 5D). Applying the average of the top percentile of ranked genes improved this correlation across all injection sites, whereas conversely, averaging the lowest percentile of genes dramatically reduced the model's predictive power (Figure 5E). While future research is needed to define the functional impact of these top-ranked genes on α-syn spreading, this list provides a resource for testing hypotheses.

[0263] Consideration Here, we demonstrate the ability of a mesoscale computational model to predict both the origin and progression of neurodegenerative diseases by using this computational model in conjunction with quantitative high-resolution whole-brain imaging. As longitudinal statistics at the voxel level across neuroanatomical regions demonstrate, the spread and decay of pathology following α-syn PFF seeding are highly dynamic in nature, with many regions containing overlapping phases of spread and decay. Nevertheless, our computational model is based on known mechanisms of α-syn PFF pathogenesis and accurately reconstructs longitudinal counts of α-syn aggregates of various sizes across 424 brain regions. While retrograde anatomical connectivity can explain much of the regional variability in spread, by separately encoding 19,893 genes from a spatial transcriptomics database into the model, we were able to further clarify the relative contribution of genes in the model's spread and decay terms.

[0264] While incorporating regional gene information into the model only reveals correlations between genes and the spread and disruption of α-syn pathology, we provide evidence that the top-ranked genes in this analysis generalize to other PFF seed locations. With this list of genes in hand, future studies will aim to test their functional impact on α-syn spreading, and some may even represent therapeutic targets for slowing or halting spreading.

[0265] Idiopathic PD, which represents the majority of cases, can be seeded from various parts of both the nervous system and peripheral organs (Challis et al., 2020; Kim et al., 2019; Peelaerts et al., 2015; Sacino et al., 2014). Comparisons between in silico simulations of α-syn pathogenesis and data from various injection sites provide a testbed for this model and demonstrate its generalizability in predicting origins and future patterns for any seeding dataset. More broadly, this model promises to analyze human brain imaging data (e.g., once accurate α-syn PET ligands are developed) to rewind the clock and predict how and where α-syn pathology develops. The clock can also be advanced to predict future trajectories and tailor therapeutic interventions accordingly. We propose this clinical application of such a model as it relates to PD or any progressive protein-spreading neurodegenerative disorder, such as Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia (Goedert et al., 2010; Guo and Lee, 2014; Jucker and Walker, 2013). Because our model provides a metric for predicting in vivo seed location given an unknown set of pathological conditions, being able to predict seed location and progression given pathological conditions has high utility in clinical diagnostic and therapeutic applications for many of these neurodegenerative diseases.

[0266] We found that encoding gene transcription levels by neuroanatomical region into our model significantly affected the fit to the data, potentially indicating the influence of relative gene transcription levels on the spreading and decay process. This comprehensive modeling and gene encoding approach enabled this ability to rank genes. While future validation is needed to explore specific genes of interest for detailed implications, the main point is that we can begin to form hypotheses from these gene rankings. We also highlight a key finding in this study: this process resulted in similar rankings of genes across datasets from different pre-formed fibril seeding locations (Figures 5D and 5E), and these seeding locations result in very different patterns of pathology spreading (Figure 4C). This consistency supports the unique information provided by this approach, and we believe that future studies will enable further exploration of the role of these genes.

[0267] Future extensions to this model may include considering additional mechanisms, such as blood vessels, ventricles, or glial density. Many studies have also demonstrated seed spread from both the gut and peripheral nervous system. However, due to the current lack of quantitative atlases connecting the periphery-brain or gut-brain axes and the technical difficulties in imaging cleared whole-body mice using light-sheet microscopy, we focused solely on the dynamics of pathology across the central nervous system. However, our model can integrate these datasets as they become available. Furthermore, while gene encoding from spatial transcriptomics databases revealed the potential implications of these genes in either the spread or decay of pathology, this analysis does not consider protein expression levels or any effects caused by genetic variation. This computational model may help form hypotheses for further studies examining these effects.

[0268] Because our labeling approach using iDISCO utilizes polyclonal secondary antibodies, we expect that multiple polyclonal secondary antibodies will bind to each primary antibody, thus providing fluorescent signal amplification. An inherent limitation of this approach is that the size of alpha-synuclein aggregates, as measured by three-dimensional morphology in the fluorescence channel, does not directly indicate the actual aggregate size (it is nonlinear but monotonic). However, we expect this multiple binding to affect all samples equally. Given that we have performed extensive and rigorous statistical testing across many cohorts of mice, we expect our statistical tests to still reveal regions of interest that show significant increases or decreases in aggregate size.

[0269] Clinical detection of pathological α-syn and other proteinaceous seeds for neurodegenerative diseases is currently primarily performed via postmortem analysis. However, numerous advances in nuclear medicine, as well as PET and SPECT radiotracers for detecting amyloid deposits in vivo, appear to enable quantitative assessment of a patient's pathological state. Current tracers work well for amyloid-beta and tau, but are still under development for α-syn. Given the diverse etiologies observed clinically, being able to distinguish between synucleinopathies with different origins and pathways is undoubtedly essential. The generalizability and interpretability of the computational model we present here offers unique advantages, as it allows both inferring the progression of α-syn spreading patterns given the current pathological state, or, conversely, generating the likely seed locations and time since seeding that led to this state. All these applications will help enable more accurate disease classification and prediction of clinical phenotypes for a wide range of neurodegenerative diseases.

[0270] Limitations of this study The method presented in this study demonstrated that whole-brain imaging and computational modeling can accurately describe and predict the longitudinal dynamics of pathology in neurodegenerative diseases. While the model can accurately reconstruct the observed dynamics of whole-brain pathology change over time across 424 neuroanatomical regions, it made several key assumptions based on recent findings in the literature. Specifically, these include neuronal uptake of injected α-syn fibrils (Brahic et al., 2016; Desplats et al., 2009; Henderson et al., 2019b; Konno et al., 2012), synaptic spread of α-syn pathology (Bieri et al., 2019), prion-like aggregation of pathology into larger units, and eventual disintegration of this pathology (Luk et al., 2012a). Future studies can extend the model to incorporate parameters not actively considered and test the sensitivity of the model's predictions to these assumptions. For example, grouping aggregate counts into 424 neuroanatomical regions may not fully capture the complex pathodynamics we observed. We tested multiple sets of neuroanatomical regions across the atlas hierarchy of regions (Figure 13D), but this number was ultimately constrained by the Allen connectivity atlas (Oh et al., 2014). Future studies can fully utilize the high-resolution images obtained in our study to create more detailed models that take into account pathodynamics across cortical layers or more granular anatomical regions.

[0271] Experimental model and subject details animal Mouse care and procedures were performed in accordance with institutional guidelines and approved by the Stanford Administrative Panel on Animal Care (APLAC). Male C57Bl6J mice (Jackson Laboratory, catalog number 000664), aged 10–12 weeks, were used for stereotaxic injections. Mice were housed under specific pathogen-free conditions with a 12-h light / dark cycle and free access to food and water.

[0272] Learn more about how PFF preparation Expression and purification of mouse wild-type α-syn were performed as previously described (Ghee et al., 2005). α-syn fibril formation was induced by incubation in 50 mM Tris-HCl, pH 7.5, 150 mM KCl buffer at 37°C in an Eppendorf Thermomixer with continuous shaking at 600 rpm. α-syn fibrils were centrifuged twice at 15,000 g for 10 min and resuspended in PBS. All fibrils were fragmented by sonication for 20 min in 2 ml Eppendorf tubes in a Vial Tweeter driven by an ultrasonic processor UIS250v (250 W, 2.4 kHz; Hielscher Ultrasonic, Teltow, Germany) prior to in vivo use. 5 μg of fibrils per mouse were used for in vivo mouse experiments. The fibrils were endotoxin-free as assessed using the Pierce LAL Chromogenic Endotoxin Quantification Kit.

[0273] fibril injection Stereotaxic injections were performed on 10- to 12-week-old adult mice. Animals were placed in a stereotaxic frame and anesthetized with 2% isoflurane (2 L / min oxygen flow rate) delivered through an anesthesia nose cone. Ophthalmic ointment was applied to prevent corneal drying during surgery. The area around the incision was trimmed, cleaned, and disinfected. A small hole was drilled above the injection site. PFF or vehicle solution was injected unilaterally into the dorsal striatum, hippocampus, olfactory tract, or substantia nigra using the following coordinates (from bregma): Striatum - anterior (AP) = +0.4 mm, lateral (ML) = + / - 1.85 mm from midline, depth (DV) = -2.7 mm (from dura). Olfactory bulb: AP, +4.50 mm; ML -0.75 mm; DV -1 mm. Dentate gyrus: AP = -2 mm, ML = 1.5 mm, DV = -2.1 mm. Substantia nigra pars compacta: AP -3.1 mm, ML 1.2 mm, DV -3.75 mm. Mice were injected with sonicated PFF (5 μg / mouse) or PBS vehicle control. PFF was sonicated prior to injection. A 1 μl volume was injected at a rate of 100 nl / min using a 5 μl Hamilton syringe with a 32G needle. To limit backflow along the injection track, the needle was kept in situ for 5 minutes before being slowly withdrawn. The skin was closed with silk sutures. Each mouse was injected subcutaneously with an analgesic and monitored during recovery. Animals were sacrificed 2 weeks to 18 months after injection.

[0274] Tissue processing Mice were anesthetized with isoflurane and transcardially perfused with 0.9% saline followed by 25 ml of 4% PFA. Brains were dissected and post-fixed in 4% paraformaldehyde (PFA) pH 7.4 for 48 hours at 4°C. Brains for histology were stored in 30% sucrose in 1x PBS at 4°C. PFA-fixed brains were sectioned at 35 μm (coronal sections) using a cryomicrotome (Leica) and stored in cryoprotective medium (30% glycerol, 30% ethylene glycol) at -20°C. Brains for iDISCO tissue clearing and labeling were stored in PBS containing 0.05% sodium azide.

[0275] Tissue transparency Each sample was fully immunolabeled and cleared using the previously described iDISCO protocol (Renier et al., 2014), which describes sample pretreatment, blocking, immunolabeling, and clearing steps in more detail. A methanol pretreatment step was performed for all samples. For primary immunolabeling, anti-phospho-synuclein (pSer129) rabbit polyclonal antibody was used at a 1:1000 dilution for 7 days, while donkey anti-rabbit IgG (H+L) Alexa Fluor 647 nm antibody was used for secondary immunolabeling at a 1:1000 dilution for 7 days. All other clearing parameters were used as previously reported (Renier et al., 2014).

[0276] immunohistochemistry Tissue processing and immunohistochemistry were performed on free-floating sections according to standard published techniques. A 1:6 to 1:12 series of coronal sections was used for all histological experiments. Sections were rinsed three times in TBST, pretreated with 0.6% H2O2 and 0.1% Triton X-100, and blocked in 5% goat serum in TBST. Free-floating coronal sections were incubated overnight with mouse-α-syn pSer129 antibody (81A; 1:5000, Covance / BioLegend catalog number MMS-5091). After overnight incubation at 4°C, sections were rinsed three times in TBST. Primary antibody staining was revealed using a fluorescently labeled secondary antibody (Thermo Fisher Scientific catalog number A-21137). Sections were counterstained with DAPI, mounted on Superfrost Plus slides (Fisher Scientific), and coverslipped using Prolong Diamond antifade mountant (Thermo Fisher Scientific catalog number P36961). Images of pSer129 aggregates were acquired using a Leica DMI6000B inverted fluorescence microscope by investigators blinded to treatment groups.

[0277] Microscopy Each sample was imaged using a LaVision Biotec Ultramicroscope II within two days of completing iDISCO clearing. The following microscope settings were used for all acquisitions: full sheet width, 0.103 numerical aperture, 3.5 μm mechanical step size, and 7 μm light sheet thickness. A 488 nm excitation laser and a 460 / 40 nm emission filter (center wavelength / FWHM) were used for each autofluorescence acquisition. To detect fluorescent α-syn pathology, a 639 nm excitation laser and a 620 / 60 nm emission filter were used. The left and right hemispheres of each brain sample were imaged separately. Each acquisition was performed in the sagittal plane. Each acquired slice had an in-plane resolution of 4.0625 × 4.0625 μm, with a slice resolution of 3.5 μm.

[0278] Quantification and statistical analysis Alignment, segmentation, and quantification After using a machine learning model to binarize the raw iDISCO α-syn fluorescence channel into foreground (pathology) and background (autofluorescence) voxels, we use binary morphological operations to find bound components. Each bound component is considered a distinct α-syn aggregate, and the location (x, y, z), peak intensity (in the corresponding raw data), and volume (number of voxels) of each aggregate are preserved.

[0279] The nonlinear transformation resulting from the registration process is used to convert each aggregate into the Allen Reference Atlas (ARA) coordinate space. Each voxel in this coordinate space is a 100 μm-wide cube centered at that coordinate. Because the atlas has a lower spatial resolution (100 μm) than the raw data (4.0625 μm), multiple aggregates may map to a single voxel in the ARA space. For a given ARA voxel, density is defined as the total number of aggregates centered within that voxel. The total size of the voxel is also defined as the sum of the sizes of all aggregates centered within that voxel. The average aggregate size for each voxel is calculated as its total size divided by the density. Similar calculations are performed for the total intensity and average intensity at each voxel. Density, average aggregate size, and average aggregate intensity are considered separate metrics.

[0280] To account for variations in alignment quality between samples, a multidimensional Gaussian filter (σ = 15 voxels) was applied to the density, mean aggregate size, and mean aggregate intensity spatial maps to smooth values ​​across neighboring voxels. This filter size was empirically determined based on the alignment results. Average smoothed spatial maps for various time points are shown in Figures 10 and 12.

[0281] Sections from immunohistochemistry were also segmented for pathology and aligned to the ARA using a similar computational pipeline applied in two dimensions instead of three. For each brain section, the corresponding coronal ARA slice was first manually selected. The DAPI channel of each section was then aligned to this atlas slice. Aggregates from the pSer129 channel were also detected using a machine learning model. For a given brain sample, total aggregate counts for each neuroanatomical region across all imaged histological sections were calculated. Because histological sections capture only a sparse representation of brain volume, aggregate counts for each region were extrapolated by dividing them by the total volume observed for that region and then multiplying by the total volume of that region in the ARA.

[0282] statistical analysis Smoothed maps from the image processing pipeline were used for two-tailed T-tests at each voxel between samples at different time points. Due to the variability of spreading patterns between adjacent time points, statistical tests were only performed between time points with appropriate intervals: 0.5 MPI vs. 4 MPI, 4 MPI vs. 8 MPI, and 8 MPI vs. 18 MPI. Therefore, the 2, 6, and 12 MPI time points were omitted. To account for the large number of voxels at 25 mm resolution, multiple comparison correction was performed using the Benjamini-Hochberg method (Benjamini and Hochberg, 1995). Corrected p-values ​​were thresholded at 0.05 to determine significance. Similar analyses were performed for counts grouped by ARA anatomical region.

[0283] Computational Modeling This Smoluchowski network model has been extensively described in previous studies ( Fornari et al., 2020 ; Wattis, 2006 ) and is governed by the following set of differential equations:

[0284]

number

[0285] c i,jrepresents the total count of aggregates in the discretized size bin indexed by I in the brain region indexed by j. The L matrix represents the Laplacian matrix of the weighted directed graph connecting various neuroanatomical regions of the brain, obtained from the Allen Connectivity Atlas (Oh et al., 2014). Because this system of differential equations does not have a closed-form solution, numerical integration using the software package SciPy was used to solve for the state dynamics given initial conditions. η was selected as a hyperparameter that slows down the spreading of large aggregates as an inverse power of their size, and λ was selected as a hyperparameter that accelerates the collapse of aggregates proportional to a power of their size. Initial values ​​for α and μ, which control the rate of spreading between nodes and the collapse at a given node, respectively, were fitted by sweeping through a 2D grid (0.001–1000) and selecting the values ​​that yielded the smallest mean squared error (α = 0.037, μ = 0.139) between predicted and actual counts. For μ = 0, k = 0, ξ = 0, and a one-dimensional size vector c, the above system of differential equations simplifies to the standard network diffusion model used in previous studies ( Henderson et al., 2019a ; Pandya et al., 2019 ).

[0286] To quantify the sensitivity of the model to specific neuroanatomical pathways in the brain, we calculated a Jacobian matrix by taking the partial derivative of the model's output with respect to the weights of the anatomical connection strengths between two regions encoded in the model. The elements of this matrix represent the relative importance of that anatomical connection in the spread of aggregates to a particular region.

[0287] This model can be used to generate a ranking of candidate seed locations for a given pathological condition c at t=T MPI. To generate this ranking, each of the 424 neuroanatomical regions is used as a separate seed location in the model at t=0 and simulated forward in time until t=T MPI. Each of the 424 simulation results is then compared to the observed condition c using a pairwise similarity metric. In this case, the similarity metric was the correlation coefficient between the total regional aggregate counts across the observed and simulated conditions. The similarity metric values ​​can then be used to sort the 424 seed locations as likely sites leading to the observed pathological condition c.

[0288] Similarly, this model can be used to predict the time since seeding t = T MPI for a given pathological condition c. A property of this computational model is that the distribution of simulated aggregate sizes across the brain is invariant, regardless of which neuroanatomical region is used as the seed location at t = 0. Thus, the whole-brain distribution of aggregate sizes for condition c can be compared to simulated distributions at various t using a pairwise similarity metric, without taking seed location into account. The mean squared error was calculated between the stimulus and observed distributions. When deciding among multiple candidate t values ​​(0.5 MPI, 2 MPI, 4 MPI), the mean squared error was inverted and normalized to sum to 1, providing the predicted probability for each t that is a correct estimate of T for a given pathological condition c.

[0289] Additionally, to encode regional genetic differences and assess their differential effects on model performance, we reconsider the α (spread) and μ (decay) parameters as region-dependent. Spread from a particular region is proportional to the gene density in that region. All genes are normalized to the same range, so that we compare only the regional distribution of gene expression to the whole-brain expression of that gene. Here, since α is considered a vector, its product with the Laplacian connectivity matrix L has the effect of modifying the regional connectivity encoded in the model. To preserve the previous fit of the model to capture longitudinal spread across the brain, we normalize each gene vector to have a mean of 1 and a standard deviation Σ, which is empirically set to preserve the correlation between predicted and observed whole-brain counts. The normalization was chosen so that this product has the effect of preserving the trace of the original matrix L. The derivation of this normalization, which preserves the trace of the Laplacian matrix, is as follows: Assume that the vector s is sampled from a multivariate normal distribution with mean 1 and standard deviation Σ. s~N(1,Σ)

[0290] Using the definition of matrix trace, and expressing s as a diagonal square matrix S, the trace of the product of S and the Laplacian matrix L is:

[0291]

number

[0292] Thus, the trace is equivalent to the dot product of s and l, with expectation equal to the sum of the entries of l, recovering the definition of the trace of L. s·l~N(1·l,lΣl) E[s·l]=Tr(L)

[0293] After encoding each gene into the model, its net effect on regional correlation between simulated and real data is compared to a baseline correlation that does not include the gene. This provides an ordered list of genes ranked by their spatial expression map relevance in improving the model's regional predictions.

[0294] References Aguzzi, A., and Calella, AM (2009). Prions: Protein Aggregation and Infectious Diseases. Physiol Rev 89, 1105-1152.10.1152 / physrev.00006.2009.

[0295] Aguzzi, A., and Rajendran, L. (2009). The transcellular spread of cytosolic amyloids, prions, and prionoids. Neuron 64,783-790.10.1016 / j.neuron.2009.12.016.

[0296] Angot, E., Steiner, JA, Hansen, C., Li, JY, and Brundin, P. (2010). Are synucleinopathies prion-like disorders? Lancet Neurol 9,1128-1138.10.1016 / S1474-4422(10)70213-1.

[0297] Beach,T.G.,Adler,C.H.,Lue,L.,Sue,L.I.,Bachalakuri,J.,Henry-Watson,J.,Sasse,J.,Boyer,S.,Shirohi,S.,Brooks,R.,et al.(2009).Unified staging system for Lewy body disorders:correlation with nigrostriatal degeneration,cognitive impairment and motor dysfunction.Acta Neuropathol 117,613-634.10.1007 / s00401-009-0538-8.

[0298] Benjamini,Y.,and Hochberg,Y.(1995).Controlling the false discovery rate:a practical and powerful approach to multiple testing.Journal of the Royal statistical society:series B(Methodological)57,289-300.

[0299] Berg,S.,Kutra,D.,Kroeger,T.,Straehle,C.N.,Kausler,B.X.,Haubold,C.,Schiegg,M.,Ales,J.,Beier,T.,Rudy,M.,et al.(2019).ilastik:interactive machine learning for(bio)image analysis.Nature Methods 16,1226-1232.10.1038 / s41592-019-0582-9.

[0300] Bieri,G.,Brahic,M.,Bousset,L.,Couthouis,J.,Kramer,N.J.,Ma,R.,Nakayama,L.,Monbureau,M.,Defensor,E.,Schule,B.,et al.(2019).LRRK2 modifies alpha-syn pathology and spread in mouse models and human neurons.Acta Neuropathol 137,961-980.10.1007 / s00401-019-01995-0.

[0301] Braak,H.,Del Tredici,K.,Bratzke,H.,Hamm-Clement,J.,Sandmann-Keil,D.,and Rub,U.(2002).Staging of the intracerebral inclusion body pathology associated with idiopathic Parkinson’s disease(preclinical and clinical stages).J Neurol 249 Suppl 3,III / 1-5.10.1007 / s00415-002-1301-4.

[0302] Braak,H.,Del Tredici,K.,Rub,U.,de Vos,R.A.,Jansen Steur,E.N.,and Braak,E.(2003).Staging of brain pathology related to sporadic Parkinson’s disease.Neurobiol Aging 24,197-211.10.1016 / s0197-4580(02)00065-9.

[0303] Brahic,M.,Bousset,L.,Bieri,G.,Melki,R.,and Gitler,A.D.(2016).Axonal transport and secretion of fibrillar forms of alpha-synuclein,Abeta42 peptide and HTTExon 1.Acta Neuropathol 131,539-548.10.1007 / s00401-016-1538-0.

[0304] Bryois,J.,Skene,N.G.,Hansen,T.F.,Kogelman,L.J.A.,Watson,H.J.,Liu,Z.J.,Brueggeman,L.O.,Breen,E.O.M.,Bulik,A.,Arenas,E.N.,et al.(2020).Genetic identification of cell types underlying brain complex traits yields insights into the etiology of Parkinson’s disease.Nat Genet 52,482-+.10.1038 / s41588-020-0610-9.

[0305] Challis,C.,Hori,A.,Sampson,T.R.,Yoo,B.B.,Challis,R.C.,Hamilton,A.M.,Mazmanian,S.K.,Volpicelli-Daley,L.A.,and Gradinaru,V.(2020).Gut-seeded alpha-synuclein fibrils promote gut dysfunction and brain pathology specifically in aged mice.Nat Neurosci 23,327-336.10.1038 / s41593-020-0589-7.

[0306] Cushman,M.,Johnson,B.S.,King,O.D.,Gitler,A.D.,and Shorter,J.(2010).Prion-like disorders:blurring the divide between transmissibility and infectivity.J Cell Sci 123,1191-1201.10.1242 / jcs.051672.

[0307] Desplats,P.,Lee,H.J.,Bae,E.J.,Patrick,C.,Rockenstein,E.,Crews,L.,Spencer,B.,Masliah,E.,and Lee,S.J.(2009).Inclusion formation and neuronal cell death through neuron-to-neuron transmission of alpha-synuclein.P Natl Acad Sci USA 106,13010-13015.10.1073 / pnas.0903691106.

[0308] Dickson,D.W.(2012).Parkinson’s Disease and Parkinsonism:Neuropathology.Csh Perspect Med 2.ARTN a009258 10.1101 / cshperspect.a009258.

[0309] Fornari,S.,Schafer,A.,Kuhl,E.,and Goriely,A.(2020).Spatially-extended nucleation-aggregation-fragmentation models for the dynamics of prion-like neurodegenerative protein-spreading in the brain and its connectome.J Theor Biol 486.ARTN 110102 10.1016 / j.jtbi.2019.110102.

[0310] Gaspar,R.,Meisl,G.,Buell,A.K.,Young,L.,Kaminski,C.F.,Knowles,T.P.J.,Sparr,E.,and Linse,S.(2017).Secondary nucleation of monomers on fibril surface dominates alpha-synuclein aggregation and provides autocatalytic amyloid amplification.Q Rev Biophys 50,e6.10.1017 / S0033583516000172.

[0311] Ghee,M.,Melki,R.,Michot,N.,and Mallet,J.(2005).PA700,the regulatory complex of the 26S proteasome,interferes with alpha-synuclein assembly.FEBS J 272,4023-4033.10.1111 / j.1742-4658.2005.04776.x.

[0312] Goedert,M.,Clavaguera,F.,and Tolnay,M.(2010).The propagation of prion-like protein inclusions in neurodegenerative diseases.Trends Neurosci 33,317-325.10.1016 / j.tins.2010.04.003.

[0313] Goedert,M.,Spillantini,M.G.,Del Tredici,K.,and Braak,H.(2013).100 years of Lewy pathology.Nat Rev Neurol 9,13-24.10.1038 / nrneurol.2012.242.

[0314] Goetz,C.G.(2011).The history of Parkinson’s disease:early clinical descriptions and neurological therapies.Cold Spring Harb Perspect Med 1,a008862.10.1101 / cshperspect.a008862.

[0315] Gribaudo,S.,Tixador,P.,Bousset,L.,Fenyi,A.,Lino,P.,Melki,R.,Peyrin,J.M.,and Perrier,A.L.(2019).Propagation of alpha-Synuclein Strains within Human Reconstructed Neuronal Network.Stem Cell Rep 12,230-244.10.1016 / j.stemcr.2018.12.007.

[0316] Guo,J.L.,and Lee,V.M.Y.(2014).Cell-to-cell transmission of pathogenic proteins in neurodegenerative diseases.Nat Med 20,130-138.10.1038 / nm.3457.

[0317] Hansen,C.,Angot,E.,Bergstrom,A.L.,Steiner,J.A.,Pieri,L.,Paul,G.,Outeiro,T.F.,Melki,R.,Kallunki,P.,Fog,K.,et al.(2011).alpha-Synuclein propagates from mouse brain to grafted dopaminergic neurons and seeds aggregation in cultured human cells.J Clin Invest 121,715-725.10.1172 / Jci43366.

[0318] Henderson,M.X.,Cornblath,E.J.,Darwich,A.,Zhang,B.,Brown,H.,Gathagan,R.J.,Sandler,R.M.,Bassett,D.S.,Trojanowski,J.Q.,and Lee,V.M.Y.(2019a).Spread of alpha-synuclein pathology through the brain connectome is modulated by selective vulnerability and predicted by network analysis.Nat Neurosci 22,1248-+.10.1038 / s41593-019-0457-5.

[0319] Henderson,M.X.,Sedor,S.,McGeary,I.,Cornblath,E.J.,Peng,C.,Riddle,D.M.,Li,H.L.,Zhang,B.,Brown,H.J.,Olufemi,M.F.,et al.(2020).Glucocerebrosidase Activity Modulates Neuronal Susceptibility to Pathological alpha-Synuclein Insult.Neuron 105,822-836 e827.10.1016 / j.neuron.2019.12.004.

[0320] Henderson,M.X.,Trojanowski,J.Q.,and Lee,V.M.(2019b).alpha-Synuclein pathology in Parkinson’s disease and related alpha-synucleinopathies.Neurosci Lett 709,134316.10.1016 / j.neulet.2019.134316.

[0321] Henrich,M.T.,Geibl,F.F.,Lakshminarasimhan,H.,Stegmann,A.,Giasson,B.I.,Mao,X.,Dawson,V.L.,Dawson,T.M.,Oertel,W.H.,and Surmeier,D.J.(2020).Determinants of seeding and spreading of alpha-synuclein pathology in the brain.Sci Adv 6.10.1126 / sciadv.abc2487.

[0322] Jucker,M.,and Walker,L.C.(2013).Self-propagation of pathogenic protein aggregates in neurodegenerative diseases.Nature 501,45-51.10.1038 / nature12481.

[0323] Kalia,L.V.,and Lang,A.E.(2015).Parkinson’s disease.Lancet 386,896-912.10.1016 / S0140-6736(14)61393-3.

[0324] Kim,S.,Kwon,S.H.,Kam,T.I.,Panicker,N.,Karuppagounder,S.S.,Lee,S.,Lee,J.H.,Kim,W.R.,Kook,M.,Foss,C.A.,et al.(2019).Transneuronal Propagation of Pathologic alpha-Synuclein from the Gut to the Brain Models Parkinson’s Disease.Neuron 103,627-641 e627.10.1016 / j.neuron.2019.05.035.

[0325] Konno,M.,Hasegawa,T.,Baba,T.,Miura,E.,Sugeno,N.,Kikuchi,A.,Fiesel,F.C.,Sasaki,T.,Aoki,M.,Itoyama,Y.,and Takeda,A.(2012).Suppression of dynamin GTPase decreases alpha-synuclein uptake by neuronal and oligodendroglial cells:a potent therapeutic target for synucleinopathy.Mol Neurodegener 7,38.10.1186 / 1750-1326-7-38.

[0326] Kordower,J.H.,Chu,Y.,Hauser,R.A.,Freeman,T.B.,and Olanow,C.W.(2008).Lewy body-like pathology in long-term embryonic nigral transplants in Parkinson’s disease.Nat Med 14,504-506.10.1038 / nm1747.

[0327] Lein,E.S.,Hawrylycz,M.J.,Ao,N.,Ayres,M.,Bensinger,A.,Bernard,A.,Boe,A.F.,Boguski,M.S.,Brockway,K.S.,Byrnes,E.J.,et al.(2007).Genome-wide atlas of gene expression in the adult mouse brain.Nature 445,168-176.10.1038 / nature05453.

[0328] Li,J.Y.,Englund,E.,Holton,J.L.,Soulet,D.,Hagell,P.,Lees,A.J.,Lashley,T.,Quinn,N.P.,Rehncrona,S.,Bjorklund,A.,et al.(2008).Lewy bodies in grafted neurons in subjects with Parkinson’s disease suggest host-to-graft disease propagation.Nat Med 14,501-503.10.1038 / nm1746.

[0329] Luk,K.C.,Kehm,V.,Carroll,J.,Zhang,B.,O’Brien,P.,Trojanowski,J.Q.,and Lee,V.M.(2012a).Pathological alpha-synuclein transmission initiates Parkinson-like neurodegeneration in nontransgenic mice.Science 338,949-953.10.1126 / science.1227157.

[0330] Luk,K.C.,Kehm,V.M.,Zhang,B.,O’Brien,P.,Trojanowski,J.Q.,and Lee,V.M.(2012b).Intracerebral inoculation of pathological alpha-synuclein initiates a rapidly progressive neurodegenerative alpha-synucleinopathy in mice.J Exp Med 209,975-986.10.1084 / jem.20112457.

[0331] Luna,E.,Decker,S.C.,Riddle,D.M.,Caputo,A.,Zhang,B.,Cole,T.,Caswell,C.,Xie,S.X.,Lee,V.M.Y.,and Luk,K.C.(2018).Differential alpha-synuclein expression contributes to selective vulnerability of hippocampal neuron subpopulations to fibril-induced toxicity.Acta Neuropathol 135,855-875.10.1007 / s00401-018-1829-8.

[0332] Oh,S.W.,Harris,J.A.,Ng,L.,Winslow,B.,Cain,N.,Mihalas,S.,Wang,Q.,Lau,C.,Kuan,L.,Henry,A.M.,et al.(2014).A mesoscale connectome of the mouse brain.Nature 508,207-214.10.1038 / nature13186.

[0333] Oliveira,L.M.A.,Gasser,T.,Edwards,R.,Zweckstetter,M.,Melki,R.,Stefanis,L.,Lashuel,H.A.,Sulzer,D.,Vekrellis,K.,Halliday,G.M.,et al.(2021).Alpha-synuclein research:defining strategic moves in the battle against Parkinson’s disease.NPJ Parkinsons Dis 7,65.10.1038 / s41531-021-00203-9.

[0334] Pandya , S. , Zeighami , Y. , Freeze , B. , Dadar , M. , Collins , DL , Dagher , A. , and Raj , A. (2019). 192,178–194.10.1016 / j.neuroimage.2019.03.001.

[0335] Paumier, KL, Luk, KC, Manfredsson, FP, Canaan, NM, Lipton, JW, Collier, TJ, Steece-Collier, K, Kemp, CJ, Celano, S, Schulz, E, et al Alpha-synuclein pathology and bilateral nigrostriatal degeneration.

[0336] Peelaerts,W.,Bousset,L.,Van der Perren,A.,Moskalyuk,A.,Pulizzi,R.,Giugliano,M.,Van den Haute,C.,Melki,R.,and Baekelandt,V.(2015).alpha-Synuclein strains cause distinct synucleinopathies after local and systemic administration.Nature 522,340–344.10.1038 / nature14547.

[0337] Peng,C.,Gathagan,R.J.,Covell,D.J.,Medellin,C.,Stieber,A.,Robinson,J.L.,Zhang,B.,Pitkin,R.M.,Olufemi,M.F.,Luk,K.C.,et al.(2018).Cellular milieu imparts distinct pathological alpha-synuclein strains in alpha-synucleinopathies.Nature 557,558-563.10.1038 / s41586-018-0104-4.

[0338] Prusiner,S.B.(1998).Prions.Proc Natl Acad Sci U S A 95,13363-13383.10.1073 / pnas.95.23.13363.

[0339] Prusiner,S.B.,Woerman,A.L.,Mordes,D.A.,Watts,J.C.,Rampersaud,R.,Berry,D.B.,Patel,S.,Oehler,A.,Lowe,J.K.,Kravitz,S.N.,et al.(2015).Evidence for alpha-synuclein prions causing multiple system atrophy in humans with parkinsonism.Proc Natl Acad Sci U S A 112,E5308-5317.10.1073 / pnas.1514475112.

[0340] Recasens,A.,Dehay,B.,Bove,J.,Carballo-Carbajal,I.,Dovero,S.,Perez-Villalba,A.,Fernagut,P.O.,Blesa,J.,Parent,A.,Perier,C.,et al.(2014).Lewy body extracts from Parkinson disease brains trigger alpha-synuclein pathology and neurodegeneration in mice and monkeys.Ann Neurol 75,351-362.10.1002 / ana.24066.

[0341] Renier,N.,Adams,E.L.,Kirst,C.,Wu,Z.,Azevedo,R.,Kohl,J.,Autry,A.E.,Kadiri,L.,Umadevi Venkataraju,K.,Zhou,Y.,et al.(2016).Mapping of Brain Activity by Automated Volume Analysis of Immediate Early Genes.Cell 165,1789-1802.10.1016 / j.cell.2016.05.007.

[0342] Renier,N.,Wu,Z.,Simon,D.J.,Yang,J.,Ariel,P.,and Tessier-Lavigne,M.(2014).iDISCO:a simple,rapid method to immunolabel large tissue samples for volume imaging.Cell 159,896-910.10.1016 / j.cell.2014.10.010.

[0343] Sacino , AN , Brooks , M. , Thomas , MA , McKinney , AB , Lee , S. , Regenhardt , RW , McGarvey , NH , Ayers , JI , Notterpek , L. , Borchelt , DR , et al Proc Natl Acad Sci USA 111,10732–10737.10.1073 / pnas.1321785111 rapid-onset motor phenotype in transgenic mice.

[0344] Shimozawa , A. , Ono , M. , Takahara , D. , Tarutani , A. , Imura , S. , Masuda-Suzukake , M. , Higuchi , M. , Yanai , K. , Hisanaga , SI , and Hasegawa , M. (2017).Acta Neuropathol Commun 5.12.10.1186 / s40478-017-0413-0.

[0345] Spillantini , MG , Schmidt , ML , Lee , VM , Trojanowski , JQ , Jakes , R. , and Goedert , M. (1997).Alpha-synuclein in Lewy bodies.Nature 388,839-840.10.1038 / 42166.

[0346] Surmeier,D.J.,Obeso,J.A.,and Halliday,G.M.(2017).Parkinson’s Disease Is Not Simply a Prion Disorder.J Neurosci 37,9799-9807.10.1523 / JNEUROSCI.1787-16.2017.

[0347] Taguchi,K.,Watanabe,Y.,Tsujimura,A.,Tatebe,H.,Miyata,S.,Tokuda,T.,Mizuno,T.,and Tanaka,M.(2014).Differential expression of alpha-synuclein in hippocampal neurons.PLoS One 9,e89327.10.1371 / journal.pone.0089327.

[0348] Tasic,B.,Yao,Z.,Graybuck,L.T.,Smith,K.A.,Nguyen,T.N.,Bertagnolli,D.,Goldy,J.,Garren,E.,Economo,M.N.,Viswanathan,S.,et al.(2018).Shared and distinct transcriptomic cell types across neocortical areas.Nature 563,72-78.10.1038 / s41586-018-0654-5.

[0349] Volpicelli-Daley, L.A., Luk, K.C., Patel, T.P., Tanik, S.A., Riddle, D.M., Stieber, A., Meaney, D.F., Trojanowski, J.Q., and Lee, V.M. (2011). Exogenous alpha-synuclein fibrils induce Lewy body pathology leading to synaptic dysfunction and neuron death. Neuron 72, 57-71. 10.1016 / j.neuron.2011.08.033.

[0350] Wattis, J.A. (2006). An introduction to mathematical models of coagulation-fragmentation processes: a discrete deterministic mean-field approach. Physica D: Nonlinear Phenomena 222, 1-20.

[0351]

Table 1

[0352]

Table 2-1

[0353]

Table 2-2

[0354]

Table 2-3

[0355]

Table 2-4

[0356] [Table 2-5]

[0357] [Table 2-6]

[0358] Example 2 Neuromodulation alters α-synuclein spreading dynamics in vivo and is predicted by changes in whole-brain function Optogenetics is a powerful tool for selectively stimulating specific neuronal cell types with high spatial and temporal specificity (13). In Alzheimer's disease, recent studies have demonstrated that optogenetic stimulation at gamma frequencies can attenuate amyloid pathology (14), a discovery that has since led to noninvasive alternatives for both hearing and vision (14, 15). In synucleinopathies such as Parkinson's disease, optogenetics has been shown to rescue motor symptoms in α-synuclein-induced disease models (16, 17) and as a potential treatment after Parkinson's disease neurodegeneration (18-20). However, no studies have demonstrated that stimulation can be used in vivo to affect the spread and aggregation of whole-brain α-synuclein pathology. Existing treatments, such as deep brain stimulation, are effective in correcting circuit imbalances caused by neurodegeneration in Parkinson's disease (21). Interventions that can target susceptible neurons with spatial, temporal, and cell-type specificity and influence the formation of pathology would be an important complement for treating Parkinson's disease and other neurodegenerative disorders.

[0359] Here, we introduce a method for altering disease progression and modulating α-syn aggregation at the whole-brain level using cell-type-specific stimulation of neural circuits with optogenetics. Using whole-brain clearing with quantitative mapping of pathology (22, 23), we demonstrate that repetitive optogenetic stimulation affects the spreading dynamics of α-synuclein aggregates in a region-specific manner. To understand the mechanisms behind this intervention in pathological spreading, we further explore colocalization with functional activity. Because many neurodegenerative diseases are progressive in nature and have distinct neuroanatomical profiles, it is important to be able to design and evaluate interventions that can modify the underlying pathological disease state with spatial and temporal precision.

[0360] result Direct injection of α-syn PFFs into the dorsal striatum results in widespread pathology throughout the brain, selective neurodegeneration, and Parkinson's disease movement disorders (9). We used the tissue clearing and immunolabeling technique iDISCO to measure whole-brain pathology resulting from PFF injection (23). iDISCO allows intact tissue to be labeled, cleared, and imaged in three dimensions without sectioning, allowing for the preservation and eventual quantification of three-dimensional α-syn aggregates throughout the whole brain. After PFF injection into the striatum, the extracted whole brain was immunolabeled, cleared, and imaged by light-sheet fluorescence microscopy (LSFM) (Figure 14A, Figure 18). We used previously validated automated quantification pipelines, ClearMap and Ilastik (22, 24), for pathology segmentation and alignment to the Allen Reference Atlas (ARA) (Figure 14B) (25). For each subject, we automatically counted α-syn aggregates within both neuroanatomical regions and 25 mm voxels to determine the relative sensitivity of these regions of interest to pathology (Figure 14C). This approach captured widespread pathology across both ipsilateral and contralateral hemispheres, consistent with previous reports of α-syn aggregates after PFF injection into the striatum (9, 10). Across cohorts of mice 2 weeks post-injection (WPI), we compared total aggregate counts across neuroanatomical regions from ARA (Figure 14C). The motor cortex (MO) had abundant α-syn pathology. The canonical pathway within the basal ganglia circuit consists of cortical to striatal projections (26), and several recent studies have identified retrograde anatomical connectivity as a prime candidate for predicting downstream α-syn pathology (10, 27). Therefore, to stimulate motor cortical neurons involved in pathological synaptic spreading, we chose to optogenetically target layer 5 of the secondary motor cortex in the ARA (28), which projects strongly to the striatum.

[0361] After unilateral injection of α-syn PFFs into the striatum and a two-day rest period, we subjected a cohort of Thy1-ChR2 mice (N = 8) in the treatment group to 20 min of daily stimulation for two weeks (Figure 15A). We also injected a control group of Thy1-ChR2 mice (N = 8) with α-syn PFFs and implanted optical fibers, but they did not receive stimulation for two weeks. After the two-week period, we used iDISCO and LSFM to capture whole-brain pathology in both groups of animals. Maximum intensity projections (MIPs) of the raw data showed the effect of stimulation on whole-brain pathology (Figure 15B), indicating a significant reduction in pathology at the stimulation site. To quantify and localize these effects, we performed voxel-level statistical tests across the two groups in ARA space, which revealed significant clusters of both increased and decreased aggregate counts throughout the brain (Figure 16A). Most notably, optogenetic stimulation reduced α-syn aggregates at the stimulation site in the motor cortex and various subcortical regions, while a significant increase in α-syn aggregate counts was observed throughout the contralateral cortex. Statistical comparisons between neuroanatomical regions confirmed these findings (Figure 16B). To verify whether consistent results could be obtained across different iDISCO experimental batches, we performed these experiments on two separate cohorts of animals (N = 3 (control), 3 (stimulation); N = 5 (control), 5 (stimulation)). Statistical maps obtained from the two cohorts showed remarkably similar effects (Figure 20). Figure 16A shows the effect of the intervention on these two combined cohorts (N = 8 (control); N = 8 (stimulation)). To account for the potential effect of laser stimulation on pathologies such as heating or visual stimulation artifacts, we included an additional sham group (N = 3) of wild-type mice that underwent injection, implantation, and sham laser stimulation for 2 weeks. Whole brain pathology from this group showed no difference when compared to control wild-type mice (N=3) (FIG. 21).

[0362] Optogenetic functional magnetic resonance imaging (ofMRI) is a technique for examining spatiotemporal changes in brain-wide activity during cell-type-specific optogenetic stimulation (29, 30). We stimulated the same motor cortex using optogenetics during fMRI recordings to measure changes in brain function induced by neural stimulation (Figure 17A). Statistical analysis of these recordings yielded activation maps showing regions of significantly increased or decreased downstream neuronal activity (Figure 17C). These activation maps, when colocalized with changes in pathology (Figure 16A), showed striking similarity with opposite polarity (Figure 17B). Positive activity, representing active brain voxels significantly driven by stimulation, showed high colocalization with decreased pathology. However, negative activity, representing voxels significantly more active across stimulations, was highly localized to regions with increased pathology.

[0363] In this study, we introduce a neurostimulation method to manipulate the spread of misfolded proteins throughout the brain. The ability to capture whole-brain pathological conditions using tissue clearing allows for a whole-brain readout to determine changes in pathology induced by neurostimulation. Integration with MRI enabled the colocalization of activation maps with changes in pathology, demonstrating that both the polarity and localization of downstream activity during stimulation are highly indicative of the resulting changes in pathology. Further research is required to clarify the complete mechanism by which misfolded proteins (e.g., α-syn) spread throughout the connectome and the biochemical pathways linking neuronal activity to altered spreading kinetics. However, the ability to readout and reliably manipulate the spread of pathological proteins such as α-syn offers significant clinical utility in the development of future therapeutics targeting the progression of neurodegenerative diseases.

[0364] Materials and Methods Animals. Mouse care and procedures were performed in accordance with institutional guidelines and approved by the Stanford Administrative Panel on Animal Care (APLAC). Male Thy1-ChR2-YFP mice (Jackson Laboratory, catalog number 007612), aged 10–12 weeks, were used for stereotaxic injections. Mice were housed under specific pathogen-free conditions with a 12-h light / dark cycle and free access to food and water. Mice were excluded from the study if they experienced seizures after any of the daily stimulations. Therefore, a total of two mice were excluded based on this criterion and were not used for tissue clearing and immunolabeling. A total of 25 mice were used in this study for tissue clearing: 6 mice (3 in each control and stimulation group) for the initial cohort of 2 weeks of stimulation, 10 mice (5 in each control and stimulation group) for the follow-up cohort of 2 weeks of stimulation, and 6 mice (3 in each control and sham group) for the comparison between control and sham wild-type mice with no opsin expression. An additional 3 mice were used to measure brain activity using optogenetic fMRI.

[0365] Preparation of PFFs. Expression and purification of mouse wild-type α-syn were performed as previously described (Ghee et al., 2005). α-Syn fibril formation was induced by incubation in 50 mM Tris-HCl, pH 7.5, 150 mM KCl buffer at 37°C in an Eppendorf Thermomixer with continuous shaking at 600 rpm. α-Syn fibrils were centrifuged twice at 15,000 g for 10 min and then resuspended in PBS. All fibrils were fragmented by sonication for 20 min in 2 ml Eppendorf tubes in a Vial Tweeter driven by an ultrasonic processor UIS250v (250 W, 2.4 kHz; Hielscher Ultrasonic, Teltow, Germany) prior to in vivo use. 5 μg of fibrils were used for each in vivo mouse experiment.

[0366] Fibril Injection and Fiber Grafting. Stereotaxic injections were performed on 10-12 week-old adult mice. Animals were placed in a stereotaxic frame and anesthetized with 2% isoflurane (oxygen flow rate of 2 L / min) delivered through an anesthesia nose cone. Ophthalmic ointment was applied to prevent corneal drying during surgery. The area around the incision was trimmed, cleaned, and disinfected. A small hole was drilled above the injection site. PFF solution was injected unilaterally into the dorsal striatum using the following coordinates (from bregma): anterior = +0.4 mm, lateral = + / - 1.85 mm from the midline, depth = -2.7 mm (from the dura). Mice were injected with sonicated PFF (5 μg / mouse). A 1 μl volume was injected at a rate of 100 nl / min using a 5 μl Hamilton syringe with a 32G needle. To limit backflow along the injection track, the needle was kept in situ for 5 min and then slowly withdrawn. The skin was closed with silk sutures.

[0367] After fibril injection, a custom-designed fiber optic implant was then attached and fixed onto the skull using Metabond (Parkell Inc.), with the fiber optic extending from the base of the implant to the desired depth (approximately 0.2 mm above the stimulation site). The target stimulation site was in layer V of the secondary motor cortex (MOs) using the following coordinates: ML = -0.5 mm; AP = 1.7 mm; DV = 0.45 mm. After surgery, mice were administered buprenorphine (0.05 mg / kg, subcutaneously [sc]) twice daily for 2 days to minimize postoperative discomfort.

[0368] Optogenetic stimulation. Mice in the treatment group were subjected to daily optogenetic stimulation for 14 days, with one stimulation session per day. Each stimulation session consisted of 10 alternating 1-minute stimulation and 1-minute rest blocks, totaling 20 minutes. The laser for optogenetic stimulation was delivered at 10 Hz with a 30% duty cycle, resulting in a 30 ms pulse width. Stimulation parameters were selected based on physiological firing frequency, and the delivered laser power was minimized to induce only a steady-state rotational bias in the mice. On each day of stimulation for the subjects, the minimum laser power required to induce robust rotational behavior was determined (Figure 19).

[0369] Tissue clearing. Each sample was thoroughly immunolabeled and cleared using the iDISCO protocol described previously (Renier et al., 2014). Anti-phospho-synuclein (pSer129) rabbit polyclonal antibody was used as the primary antibody at a dilution of 1:1000, and donkey anti-rabbit IgG (H+L) Alexa Fluor 647 nm antibody was used as the secondary antibody at a dilution of 1:1000.

[0370] Microscopy. Each sample was imaged within 2 days of completing iDISCO clearing using a LaVision Biotec Ultramicroscope II. The following microscope settings were used for all acquisitions: full sheet width, 0.103 numerical aperture, 3.5 μm mechanical step size, and 7 μm light sheet thickness. 4xx / 488 nm excitation and emission filters were used for each autofluorescence acquisition, and a 6xx / 647 nm filter was used to detect fluorescent α-synuclein pathology. The left and right hemispheres of each brain sample were imaged separately. The light sheet, and therefore each acquired slice, was acquired in the sagittal plane. Each acquisition had an in-plane resolution of 4.0625 × 4.0625 μm, with a slice resolution of 3.5 μm. After light sheet imaging, each acquired hemisphere was segmented for pathology and aligned to a standardized atlas using ClearMap and Ilastik software (Renier et al., 2016; Berg et al., 2019).

[0371] Statistics. Smoothed maps of α-syn aggregate counts were used for two-tailed t-tests at each voxel between samples in the two comparison groups. To perform multiple comparison correction on statistical maps, a nonparametric, threshold-free, cluster-based technique (Ridgway et al., 2012) was used to localize clusters with significant changes (p<0.05) while accounting for the correlation structure within these maps. This correction method takes into account both the distribution of cluster sizes and the magnitude of test statistics within each cluster, which is necessary to capture both diffuse and highly localized treatment effects. For statistical tests between neuroanatomical regions in the Allen Criterion Atlas, multiple comparisons were corrected using the Benjamini-Hochberg method to set the false discovery rate (Benjamini et al., 1995).

[0372] Optogenetic fMRI Experiments and Data Analysis. Optogenetic fMRI scanning was performed using a 7 Tesla Bruker Biospec small animal MRI system. Mice were scanned under very light anesthesia (0.3%–0.7% isoflurane mixed with O2 and NO). Each of the MRI scans consisted of six 15-second pulse trains of light stimulation delivered once per minute over a 6-minute period. Most stimulation parameters during the scans were the same as those used for treatment (10 Hz, 30% duty cycle), except that the laser power was increased to 1.5 mW to account for the effects of anesthesia. fMRI voxels significantly modulated by optogenetic stimulation were identified using a generalized linear model in the FSL software package (Jenkinson et al., 2012). Individual activity maps showing significant positive and negative activity during stimulation are shown in Figure 22. Using the FSL alignment package, the aligned scans were then averaged across four acquisitions for each subject and then across all three subjects, resulting in the single activity map presented in Figure 17.

[0373] References 1.M.Jucker,L.C.Walker,Self-propagation of pathogenic protein aggregates in neurodegenerative diseases.Nature 501,45-51(2013).

[0374] 2.L.M.A.Oliveira et al.,Alpha-synuclein research:defining strategic moves in the battle against Parkinson’s disease.NPJ Parkinsons Dis 7,65(2021).

[0375] 3.M.Goedert,M.G.Spillantini,K.Del Tredici,H.Braak,100 years of Lewy pathology.Nat Rev Neurol 9,13-24(2013).

[0376] 4.H.Braak et al.,Staging of the intracerebral inclusion body pathology associated with idiopathic Parkinson’s disease(preclinical and clinical stages).J Neurol 249 Suppl 3,III / 1-5(2002).

[0377] 5.H.Braak et al.,Staging of brain pathology related to sporadic Parkinson’s disease.Neurobiol Aging 24,197-211(2003).

[0378] 6.L.A.Volpicelli-Daley et al.,Exogenous alpha-synuclein fibrils induce Lewy body pathology leading to synaptic dysfunction and neuron death.Neuron 72,57-71(2011).

[0379] 7.V.M.Lee,J.Q.Trojanowski,Mechanisms of Parkinson’s disease linked to pathological alpha-synuclein:new targets for drug discovery.Neuron 52,33-38(2006).

[0380] 8.J.H.Kordower,Y.Chu,R.A.Hauser,T.B.Freeman,C.W.Olanow,Lewy body-like pathology in long-term embryonic nigral transplants in Parkinson’s disease.Nat Med 14,504-506(2008).

[0381] 9.K.C.Luk et al.,Pathological alpha-synuclein transmission initiates Parkinson-like neurodegeneration in nontransgenic mice.Science 338,949-953(2012).

[0382] 10.M.X.Henderson et al.,Spread of alpha-synuclein pathology through the brain connectome is modulated by selective vulnerability and predicted by network analysis.Nat Neurosci 22,1248-1257(2019).

[0383] 11.S.Pandya et al.,Predictive model of spread of Parkinson’s pathology using network diffusion.Neuroimage 192,178-194(2019).

[0384] 12.M.T.Henrich et al.,Determinants of seeding and spreading of alpha-synuclein pathology in the brain.Sci Adv 6,(2020).

[0385] 13.O.Yizhar,L.E.Fenno,T.J.Davidson,M.Mogri,K.Deisseroth,Optogenetics in neural systems.Neuron 71,9-34(2011).

[0386] 14.H.F.Iaccarino et al.,Gamma frequency entrainment attenuates amyloid load and modifies microglia.Nature 540,230-235(2016).

[0387] 15.A.J.Martorell et al.,Multi-sensory Gamma Stimulation Ameliorates Alzheimer’s-Associated Pathology and Improves Cognition.Cell 177,256-271 e222(2019).

[0388] 16.C.Qi,S.Varga,S.J.Oh,C.J.Lee,D.Lee,Optogenetic Rescue of Locomotor Dysfunction and Dopaminergic Degeneration Caused by Alpha-Synuclein and EKO Genes.Exp Neurobiol 26,97-103(2017).

[0389] 17.H.C.Moon et al.,Effect of optogenetic modulation on entopeduncular input affects thalamic discharge and behavior in an AAV2-alpha-synuclein-induced hemiparkinson rat model.Neurosci Lett 662,129-135(2018).

[0390] 18.A.V.Kravitz et al.,Regulation of parkinsonian motor behaviours by optogenetic control of basal ganglia circuitry.Nature 466,622-U627(2010).

[0391] 19.V.Gradinaru,M.Mogri,K.R.Thompson,J.M.Henderson,K.Deisseroth,Optical Deconstruction of Parkinsonian Neural Circuitry.Science 324,354-359(2009).

[0392] 20.L.A.V.Magno et al.,Optogenetic Stimulation of the M2 Cortex Reverts Motor Dysfunction in a Mouse Model of Parkinson’s Disease.J Neurosci 39,3234-3248(2019).

[0393] 21.M.S.Okun,Deep-Brain Stimulation for Parkinson’s Disease.New England Journal of Medicine 367,1529-1538(2012).

[0394] 22.N.Renier et al.,Mapping of Brain Activity by Automated Volume Analysis of Immediate Early Genes.Cell 165,1789-1802(2016).

[0395] 23.N.Renier et al.,iDISCO:a simple,rapid method to immunolabel large tissue samples for volume imaging.Cell 159,896-910(2014).

[0396] 24.S.Berg et al.,ilastik:interactive machine learning for(bio)image analysis.Nature Methods 16,1226-1232(2019).

[0397] 25.S.M.Sunkin et al.,Allen Brain Atlas:an integrated spatio-temporal portal for exploring the central nervous system.Nucleic Acids Res 41,D996-D1008(2013).

[0398] 26.J.L.Lanciego,N.Luquin,J.A.Obeso,Functional Neuroanatomy of the Basal Ganglia.Cold Spring Harbor Perspectives in Medicine 2,(2012).

[0399] 27. C. Mezias, N. Rey, P. Brundin, A. Raj, Neural connectivity predicts spreading of alpha-synuclein pathology in fibril-injected mouse models: Involvement of retrograde and anterograde axonal propagation. Neurobiology of Disease 134, (2020).

[0400] 28.NNFoster et al.,The mouse cortico-basal ganglia-thalamic network.Nature 598,188-+(2021).

[0401] 29.JHLee et al., Global and local fMRI signals driven by neurons defined optogenetically by type and wiring.Nature 465,788-792(2010).

[0402] 30. HJLee et al., Activation of Direct and Indirect Pathway Medium Spiny Neurons Drives Distinct Brain-wide Responses. Neuron 91, 412-424 (2016).

[0403] Example 3 Resolving brain circuit function and dysfunction using computational modeling and optogenetic fMRI Can we systematically design treatments for brain disorders such as Parkinson's disease? To make this possible, we need to go beyond simple correlations and develop a complete algorithmic description of how specific cells or brain regions cooperatively contribute to behavior or symptoms in the context of brain-wide networks. What is needed to generate such algorithms? Cell types cannot be ignored, as even neurons in the same location in the brain can drive completely opposite functions and resulting behaviors (1, 2). Neurons often interact with larger brain-wide networks. Therefore, a limited field of view within the brain is insufficient to understand these algorithms. Therefore, to obtain the data necessary to reconstruct these algorithms of behavioral control, imaging systems capable of measuring cell-type-specific whole-brain function are required. Optogenetic functional magnetic resonance imaging (ofMRI) (3) has begun to achieve this goal. Using ofMRI, we can select cell-type-specific modulation targets while monitoring the results of such modulation across the entire brain in vivo with high spatiotemporal resolution. This has opened a new window into the study of brain function. We can see how modulating specific brain components leads to specific behaviors of interest and also directly observe the inner workings of the brain that lead to such behaviors. Through computational modeling of ofMRI signals measured across the entire brain (4), we can quantitatively describe cell-type-specific, large-scale brain function at the regional level. Once regional interaction maps are reconstructed, we envision combining biophysical modeling to enable cell-type-specific, large-scale modeling of brain function at the single-cell spiking level. In addition, while restoring brain function is the ultimate goal of treating neurological disorders, understanding how significant pathology is related to brain function is also crucial. In this review, we discuss approaches taken to date and approaches that could be used in the future to achieve these goals.

[0404] Cross-linking scales using optogenetic fMRI. ofMRI is a technology that combines optogenetic stimulation with fMRI readout. Optogenetics (5, 6) uses light to enable cell-type-specific, millisecond-scale activity modulation, while high-field fMRI measures the resulting hemodynamic responses in living subjects across the entire brain. In the first proof-of-concept study (3), motor cortex excitatory neurons evoked fMRI responses that could be measured across the entire brain with subsecond temporal resolution. To accelerate scientific discoveries through ofMRI, multiple technological innovations have been made. Real-time imaging with robustness to motion in living subjects has been developed, achieving data acquisition, reconstruction, motion correction (7), and 3D image analysis with high accuracy in approximately 12 ms. To elucidate cortical layer- and subnucleus-specific responses, novel compressed sensing (8-10) and machine learning-based fMRI techniques have been developed, which achieve significant reductions in voxel volume. MR-compatible optrodes and electrodes have also been developed for simultaneous electrophysiological recording to examine the neural basis of ofMRI hemodynamic signals (11, 12). They can achieve simultaneous acquisition of electrophysiological recordings during an fMRI session, providing information with higher temporal resolution in regions of interest identified by ofMRI.

[0405] Utilizing these advanced ofMRI techniques, the capabilities and accuracy of ofMRI have been extensively tested. It has been shown that the location, polarity, and temporal shape of neural activity can be accurately inferred from ofMRI signals (3, 13, 14), and neural activity can be measured by ofMRI across multiple synapses (11, 15). Cell type, location, frequency (14, 15), and intensity of stimulation (11) have been shown to dramatically alter the location and shape of activity throughout the brain. It has also been shown that whole-brain neurodynamics, as measured by ofMRI, can accurately predict distinct behaviors (2, 14, 15).

[0406] Many studies have used ofMRI to improve our understanding of fundamental circuits related to behavior, memory, and cognition. For example, ofMRI studies have identified that frequency-dependent thalamic activity drives distinct whole-brain functions in circuits related to arousal, attention, and somatosensory function (16-18). ofMRI studies have revealed distinct dorsal and ventral hippocampal control of brain-wide function (15, 19), circuits related to cell-type-specific targeting of the somatosensory cortex (20), and cerebellar cortical control across the forebrain and midbrain (21). Another study revealed brain-wide dynamics governing how the medial prefrontal cortex modulates reward-related behaviors through distal regions such as the striatum (22). Recent studies have explored fMRI with cell-type-specific activation of astrocytes (23). These studies have shown that the observed hemodynamic activity can be tightly coupled to neuronal activity using either simultaneous or tracking electrophysiology. Furthermore, although most ofMRI studies have been performed in rodents, they have also been applied to nonhuman primates, where both saccade latency and whole-brain activity were found to depend on specific neuronal targets in the motor cortex ( 24 ).

[0407] The ability to probe and read out whole-brain activity using ofMRI has also advanced our understanding of dysfunctional circuitry associated with neurological disorders. To study epilepsy, ofMRI offers the unique advantage of being able to optogenetically induce seizures at precise origins on demand while simultaneously measuring the resulting whole-brain activity with electrophysiological recordings. This has enabled studies that can generate models for predicting and classifying seizures using early activity markers of seizures (15, 25). Furthermore, the long-term effects of seizures on global brain function can be measured to understand how the disease progresses and how seizures develop and are maintained (26). These advances aid in understanding the circuit mechanisms underlying seizures and help design intervention parameters, such as stimulation location and frequency, to effectively suppress seizures. Optogenetic fMRI can also elucidate the mechanisms behind existing therapies, such as post-stroke recovery. Activation measured by ofMRI highly predicts the degree of recovery and has identified the sensory circuits involved in this process (27).

[0408] We can now begin to uncover detailed circuit mechanisms that have previously been difficult to understand. As an example, we consider a study revealing how medium spiny neurons (MSNs) expressing D1 and D2 receptors dynamically modulate global brain function and dysfunction (Figure 23). The cortico-basal ganglia-thalamic circuit is involved in many important brain functions, including motor control and reward mechanisms. Neurological disorders such as Parkinson's disease, Huntington's disease, addiction, and autism involve this network. The caudate-putamen houses D1-MSNs and D2-MSNs (Figure 23A), a key node separating the direct and indirect pathways (Figure 23B). To assess the brain-wide dynamics driven by inhibitory D1- and D2-MSNs, Lee et al. performed whole-brain fMRI during repeated 20-second periods of optogenetic D1- or D2-MSN stimulation (2) (Figure 23C). Active voxels were identified as significantly synchronized with the repeated stimulation (Figure 23D). Local signals at the site of stimulation were positive for both inhibitory D1- and D2-MSN stimulation (Figure 23E), shedding light on the widely debated question of whether inhibitory neuronal activity induces positive or negative fMRI signals. In most regions of the ipsilateral cortico-basal ganglia-thalamus network, evoked responses in a given region exhibited qualitatively different temporal profiles between D1- and D2-MSN stimulation (Figures 23D and 23E). Given the diversity of BOLD responses evoked by D1- and D2-MSN stimulation, we attempted to verify whether the responses reflected underlying neuronal activity using single-unit recordings. For example, fMRI time series in the thalamus exhibited robust and reliable increases and decreases in response to D1- and D2-MSN stimulation, respectively (Figures 23D and 23E). Indeed, single units showed increased firing rates during D1-MSN stimulation and decreased firing rates during D2-MSN stimulation (Fig. 23F). These results demonstrate that ofMRI can detect cell-type-specific changes in multisynaptic activity throughout the brain with high spatiotemporal precision, and that electrophysiological recordings in conjunction with ofMRI can support these findings.

[0409] As with any technology, ofMRI has caveats for future improvements, as well as potential pitfalls that need to be avoided. Channelrhodopsin (ChR2) is known to induce synchronized neuronal activity upon optical stimulation. Therefore, before initiating an ofMRI study, it is important to first investigate the behavioral effects of optogenetic stimulation as a means to ensure that the generated behavior is of interest in either normal physiological or pathological conditions. For example, in our ofMRI experiments with D1- and D2-MSN stimulation, we observed increased contralateral and ipsilateral rotations, respectively (2) (Figure 23B). This indicates that two separate stimuli result in opposing behaviors known to be associated with movement disorders. The fact that ChR2 induces synchronization upon optical stimulation also makes it suitable for studying pathological oscillations in several disease models / conditions. Excessive beta-band oscillations in Parkinson's disease have been extensively investigated using ChR2-induced oscillations (28, 29). Some newer opsins, such as stabilized step-function opsins (SSFOs) (30), modulate target neurons by increasing their excitability and amplifying existing spontaneous activity, extending the applicability of ofMRI to more physiological conditions. There are many aspects of ofMRI that can be further improved. For example, ofMRI would benefit greatly from higher spatiotemporal resolution, real-time feedback-based stimulus control, and imaging during behavior (7, 31).

[0410] It is also important to note that other techniques have been developed to study brain circuits, such as high-speed volumetric calcium imaging (32, 33), probes for high-density electrophysiological recording (34, 35), and wide-field calcium and voltage imaging (36). Compared to ofMRI, these new advances offer higher spatiotemporal resolution, although they are limited by recording depth and field of view. For example, wide-field "whole-cortex" calcium imaging allows simultaneous recording of cortical regions at up to 30 Hz, but its depth range is limited to the superficial layers of the cortex. Combining fMRI with wide-field calcium imaging has been shown to mitigate the limitations of both techniques (36). Therefore, future ofMRI studies can be integrated with other techniques to exploit their complementary strengths.

[0411] Cell-type-specific modeling of large-scale brain function. Large-scale neural network models (37-41) utilizing experimental data from PET (42), fMRI (43), and EEG / MEG (44) have contributed significantly to our understanding of brain function. However, these modeling efforts are based on data from carefully designed experiments that attempt to isolate specific brain functions, while multiple networks and pathways mediated by different cell types simultaneously participate in integrating brain function. Therefore, without the ability to elucidate the contributions from different cell types across the brain, models have been limited in their capabilities. The development of MRI techniques opens new opportunities for whole-brain computational modeling because they uniquely measure cell-typ...

Claims

1. 1. A computer-implemented method for predicting the location of occurrence of pathological protein aggregates in the brain of a subject having a neurological or neurodegenerative disease, the method comprising: a) receiving an image of the brain of the subject; b) using machine learning algorithms to identify pathological protein aggregates in the image; c) mapping the location of said pathological protein aggregates to neuroanatomical regions; d) modeling the discretized distribution of said pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model; e) calculating the distribution of gene effects on the regional spread and disruption of said pathological protein aggregates using regional gene density maps for each neuroanatomical region; f) predicting changes in regional density of the pathological protein aggregates as a function of time by modeling the spreading, aggregation, disintegration, and spatial gene expression of the pathological protein aggregates across the neuroanatomical regions, where spreading is assumed to occur retrogradely between anatomically interconnected neuroanatomical regions, and spreading is modeled as diffusion through a weighted directed graph connecting the neuroanatomical regions of the brain of the subject; g) predicting past, current, and future locations of said pathological protein aggregates based on said modeling.

2. 10. The computer-implemented method of claim 1, further comprising adjusting one or more programmed neurostimulation parameters based on the predicted change in the regional density of the pathological protein aggregates as a function of time.

3. The computer-implemented method of claim 2 , wherein the one or more programmed neurostimulation parameters are selected from duration, amplitude, frequency, pulse width, and location of neurostimulation.

4. 4. The computer-implemented method of claim 1, further comprising instructing a neurostimulation device to apply neurostimulation to the brain locations predicted to contain pathological protein aggregates to treat a neurological or neurodegenerative disease in the subject.

5. 5. The computer-implemented method of claim 4, further comprising instructing the neurostimulator device to apply electrical stimulation to the brain locations where pathological protein aggregates are not yet present but are predicted to appear in the future based on predicting the change in the regional density of the pathological protein aggregates as a function of time.

6. performing image registration to a coordinate space containing a plurality of voxels, each voxel represented as a cubic volume element centered at a coordinate in the coordinate space; identifying the x, y, z coordinate location of each pathological protein aggregate within said coordinate space; determining the volume of each pathological protein aggregate from the total number of voxels occupied by each pathological protein aggregate; calculating an aggregate density for each voxel, the aggregate density for each voxel being determined from a total number of pathological protein aggregates centered within the same voxel; calculating a total aggregate size for each voxel, the total aggregate size being the total size of all the pathological protein aggregates centered within the same voxel; calculating an average aggregate size for each voxel as the total aggregate size for each voxel divided by the aggregate density for each voxel; calculating the total signal intensity of each voxel from the total intensities of all the pathological protein aggregates centered within the same voxel; 6. The computer-implemented method of claim 1, further comprising: calculating a mean signal intensity for each voxel as the total signal intensity for each voxel divided by the aggregate density.

7. modeling the discretized distribution of the pathological protein aggregates in each neuroanatomical region as a set of differential equations using a Smoluchowski network model, [Equation 1] wherein c ｉ，ｊ 7. The computer-implemented method of claim 1, wherein η represents the total count of pathological protein aggregates in a discretized size bin indexed by / in a brain region indexed by j, the L matrix represents the Laplacian matrix of the weighted directed graph connecting the neuroanatomical regions of the brain, η is selected as a hyperparameter that slows down the spreading of large aggregates as an inverse power of their size, and λ is selected as a hyperparameter that accelerates the collapse of pathological protein aggregates proportionally to a power of their size.

8. 8. The computer-implemented method of claim 7, wherein the initial values ​​of α and μ are adapted by sweeping a two-dimensional grid and selecting the values ​​that result in the smallest mean square error between the predicted and actual counts of the pathological protein aggregates.

9. 8. The computer-implemented method of claim 7, wherein the initial value of μ=0, the initial value of k=0, ξ=0, and c is a one-dimensional size vector, and the set of differential equations simplifies to a standard network diffusion model.

10. 10. The computer-implemented method of claim 7, further comprising quantifying the sensitivity of the model to specific neuroanatomical pathways in the brain, wherein a Jacobian matrix is ​​calculated by taking partial derivatives of the output of the model with respect to weights of the anatomical coupling strengths between two neuroanatomical regions encoded in the model, and elements of the Jacobian matrix represent the relative contribution of the anatomical coupling between the two neuroanatomical regions to the spread of pathological protein aggregates to specific regions of the brain.

11. The model is used to rank candidate seed locations for a given pathological condition c at t=T months post injection (MPI) as: using each of the neuroanatomical regions simulated forward in time until t=T MPI as a separate seed location at t=0 in the model, wherein each of the simulation results for a different neuroanatomical region is compared to an observed state c using a pairwise similarity metric, the similarity metric being the correlation coefficient between total regional aggregate counts across the observed and simulated states; using the similarity metric value to sort the seed locations for the neuroanatomical regions as likely to lead to the observed pathological condition c, whereby the ranking of candidate seed locations for the given pathological condition c at t=T MPI is generated; The computer-implemented method of any one of claims 7 to 10, further comprising generating by a method comprising:

12. The time since seeding for a given pathological condition c, t=T MPI, is defined as: comparing the whole-brain distribution of aggregate sizes for state c to simulated distributions at various t using a pairwise similarity metric without considering the seed locations, where the distribution of simulated aggregate sizes across the whole brain is assumed to be invariant with respect to which neuroanatomical region is used as the seed location at t=0; calculating the mean squared error between the stimulus distribution and the observation distribution, where, when deciding among multiple candidate t values, the mean squared errors are inverted and normalized to sum to one to provide a predicted probability for each t being a correct estimate of T for the given pathological condition c; The computer-implemented method of claim 11 , further comprising predicting by a method comprising:

13. The genetic effect on the regional spread and disintegration of the pathological protein aggregates is Assuming that the α (spread) and μ (attenuation) parameters are region-dependent, with the spread from a particular neuroanatomical region being proportional to the gene density in that region; normalizing all genes to the same range so that only the regional distribution of gene expression is compared to the whole-brain expression of that gene, where α is a vector and the product of α and the Laplacian connectivity matrix L has the effect of modifying the regional connectivity encoded in the model; normalizing each gene vector to have a mean of 1 and a standard deviation Σ empirically set to preserve the correlation between predicted and observed whole-brain aggregate counts, the normalization being chosen to have the effect of the product preserving the trace of the original Laplacian connectivity matrix L; The computer-implemented method of any one of claims 7 to 12, wherein the determination is made by a method comprising:

14. The derivation of the normalization to preserve the trace of the original Laplacian connectivity matrix L is: Assume that the vector s is sampled from a multivariate normal distribution with mean 1 and standard deviation Σ, where s ∼ N(1,Σ); Using the definition of matrix trace, expressing s as a diagonal square matrix S, wherein the trace of the product of S and the Laplacian connectivity matrix L is [Equation 2] where l represents the diagonal of L, and said trace is equal to the dot product of s and l, with expectation equal to the sum of the entries of l; s・l ~ N (1・l, l Σl) E[s・l] = Tr(L) Recovering the definition of the trace of L according to the formula: After each gene is encoded in the model, Comparing the net effect on regional correlation between the simulated data and the actual data with a baseline correlation that does not include the gene; and and providing an ordered list of genes ranked by their spatial expression map relevance in improving the regional prediction of the model.

15. The computer-implemented method of any one of claims 1 to 14, wherein the cubic volume element has a width of 100 μm in the coordinate space.

16. The computer-implemented method of any one of claims 1 to 15, wherein one or more pathological protein aggregates are mapped to a single voxel.

17. The computer-implemented method of any one of claims 1 to 16, further comprising performing multi-dimensional Gaussian filtering to account for variations in image registration between different samples.

18. The computer-implemented method of any one of claims 1 to 17, further comprising segmenting the image to generate a plurality of image segments.

19. 19. The computer-implemented method of claim 1, wherein said mapping comprises mapping the locations of the pathological protein aggregates to neuroanatomical regions of the Allen Human Brain Reference Atlas.

20. 20. The computer-implemented method of claim 19, wherein said mapping comprises performing image registration to transform the locations of the pathological protein aggregates into the coordinate space of the Allen Brain Reference Atlas.

21. 21. The computer-implemented method of claim 19 or 20, wherein anatomically interconnected neuroanatomical regions are identified from the Allen Connectivity Atlas.

22. The neuroanatomical regions include anterior amygdala region, anterior cingulate region, dorsal, anterior cingulate region, ventral, nucleus accumbens, anterior dorsal nucleus, anterior hypothalamic nucleus, agranular insular region, dorsal, agranular insular region, posterior, agranular insular region, ventral, nucleus ambiguus, anterior medial nucleus, dorsal, anterior medial nucleus, ventral, oblique lobule, accessory olfactory bulb, anterior olfactory nucleus, anterior pretectal nucleus, arcuate hypothalamic nucleus, dorsal auditory cortex, primary auditory cortex, ventral auditory cortex, anterior ventral nucleus of the thalamus, basolateral amygdala nucleus, basomedial amygdala nucleus, bed nucleus of the stria terminalis, CA1 area, CA2 area, CA3 area, central nucleus of the amygdala, central lobule, lateral central nucleus of the thalamus, claustrum, central linear raphe, medial central nucleus of the thalamus, cortico-amygdala region, anterior, cortico-amygdala region, posterior. , caudate nucleus, superior central raphe, anterior cerebellum, cuneiform nucleus, dorsal cochlear nucleus, dentate gyrus, dorsomedial nucleus of the hypothalamus, dentate nucleus, dorsal peduncular region, dorsal nucleus raphe, external rhinal region, entorhinal region, lateral region, entorhinal region, medial region, dorsal zone, caudate nucleus, dorsal, caudate nucleus, ventral, flocculus, dorsal column, cerebral cortex, basal striatum, globus pallidus, external segment, globus pallidus, internal segment, nucleus reticularis magnocellularis, gustatory region, intergeniculate nucleus of the amygdala, inferior colliculus, central nucleus, inferior colliculus, dorsal nucleus, inferior colliculus, external nucleus, inferior colliculus, infralimbic region, intermediate dorsal nucleus of the thalamus, inferior olivary complex, nucleus intermedius, interpeduncular nucleus, intermediate reticular nucleus, lateral amygdala, lateral vestibular nucleus, lateral dorsal nucleus of the thalamus, dorsal of the lateral geniculate complex, lateral geniculate Ventral, lateral habenula, lateral hypothalamic area, lateral posterior thalamus, lateral preoptic area, lateral reticular nucleus, lateral septal nucleus, caudal (caudodorsal) part, lateral septal nucleus, rostral (rostral) part, lateral septal nucleus, ventral, magnocellular nucleus, magnocellular reticular nucleus, mediodorsal thalamus, medullary reticular nucleus, dorsal part, medullary reticular nucleus, abdomen, medial amygdala nucleus, median preoptic nucleus, medial geniculate complex, dorsal part, medial geniculate complex, medial part, medial geniculate complex, abdomen, medial habenula, medial mammillary nucleus, main olfactory bulb, primary motor area, secondary motor area, medial preoptic nucleus, medial preoptic area, medial pretectal area, mesencephalic reticular nucleus, medial septal nucleus, medial vestibular nucleus diagonal zone nucleus, nucleus indeterminate, lateral Nuclei of the lateral lemniscus, nuclei of the lateral olfactory tract, tubercle (X), nuclei of the optic tract, nuclei of the posterior commissure, nucleus of the solitary tract, orbital region, lateral part, orbital region, medial part, orbital region, ventrolateral part, olfactory tubercle, posterior amygdaloid nucleus, piriform amygdaloid region, periaqueductal gray matter, paraunculoid gyrus, parvocellular reticular nucleus, parabrachial nucleus, central pontine gray matter, perirhinal region, parafascicular nucleus, paraflocculus lobule, pontine gray matter, paracellular reticular nucleus, dorsal part, paracellular reticular nucleus, lateral part, posterior nucleus of the hypothalamus, piriform region, prelimbic area, dorsal premammillary nucleus, posterior thalamic complex, posterior bordering nucleus of the thalamus, posterior uncinate gyrus, peripeduncular nucleus, pedunculopontine nucleus, preuncleate gyrus, paramedian lobule, pontine reticular nucleus, caudal part, pontine reticular nucleus, precursor nucleus, principal sensory nucleus of the trigeminal nerve, parathalamic nucleus,Posterior parietal-related area, paraventricular hypothalamic nucleus, paraventricular thalamic nucleus, periventricular hypothalamic nucleus, posterior, periventricular hypothalamic nucleus, preoptic tract, lobule VIII, postchiasmatic area, nucleus of reuniens, rhomboid nucleus, nucleus raphe magnus, nucleus red, midbrain reticular nucleus, posterior ruber area, retrosplenial area, lateral agranular area, retrosplenial area, dorsal, retrosplenial area, ventral, thalamic reticular nucleus, paraventricular area, superior colliculus, motor-related , superior colliculus, sensory related, septum fimbria, substantia innominata, lobule simplex, medial subthalamic nucleus, substantia nigra, pars compacta, substantia nigra, pars reticulata, superior olivary complex, subparafascicular area, subparafascicular nucleus, magnocellular part, subparafascicular nucleus, parvocellular part, spinal vestibular nucleus, spinal nucleus of the trigeminal nerve, tail, spinal nucleus of the trigeminal nerve, interpolar part, spinal nucleus of the trigeminal nerve, oral part, primary somatosensory cortex, barrel area, primary somatosensory cortex, lower limbs 22. The computer-implemented method of claim 1, wherein the retinal region is selected from: primary somatosensory cortex, mouth, primary somatosensory cortex, nose, primary somatosensory cortex, trunk, primary somatosensory cortex, upper limb, additional somatosensory region, subthalamic nucleus, uncinate gyrus, supramammillary nucleus, supratrigeminal nucleus, superior vestibular nucleus, temporal association region, posterior piriform transition region, nucleus of the tegmental reticularis, nucleus septalis triangularis, stria tecta, nucleus nodosa, motor nucleus of the trigeminal nerve, ventral anterior thalamic complex, ventral cochlear nucleus, facial motor nucleus, visceral region, anterior lateral visual region, anterior medial visual region, lateral visual field, primary visual region, posterolateral visual region, posteromedial visual region, ventromedial nucleus of the thalamus, ventromedial nucleus of the hypothalamus, ventral posterolateral nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, ventral posteromedial nucleus of the thalamus, parvocellular portion, ventral tegmental region, and hypoglossal nucleus.

23. 23. The computer-implemented method of any one of claims 1 to 22, further comprising predicting where pathological protein aggregates have occurred in the brain of the subject.

24. The computer-implemented method of any one of claims 1 to 23, wherein the machine learning algorithm uses an artificial neural network.

25. The computer-implemented method of any one of claims 1 to 24, wherein the machine learning algorithm uses a deep learning algorithm.

26. 26. The computer-implemented method of claim 25, wherein the deep learning algorithm uses a convolutional neural network, a deep neural network, a recurrent neural network, a deep residual neural network, a long short-term memory network, a deep belief network, a multilayer perceptron, or deep reinforcement learning.

27. 27. The computer-implemented method of any one of claims 1 to 26, wherein the machine learning algorithm is supervised, semi-supervised, or unsupervised.

28. The computer-implemented method of any one of claims 1 to 27, wherein the subject is a human subject.

29. 29. The computer-implemented method of claim 28, wherein modeling the spreading, aggregation, and disintegration of the pathological protein aggregates in the brain of the human subject uses simulations generated based on measuring the spreading, aggregation, and disintegration of the pathological protein aggregates in a non-human animal model, and experimentally measured pathology in the non-human animal model is used to design simulation parameters for predicting the past location, the current location, and the future location of the pathological protein aggregates in the human subject.

30. 30. The computer-implemented method of claim 29, wherein the simulation further uses measured brain anatomy or brain function from the non-human animal to predict the past location, the current location, and the future location of the pathological protein aggregate in the human subject.

31. 31. The computer-implemented method of claim 29 or 30, wherein the non-human animal is a mammal.

32. 32. The computer-implemented method of claim 31, wherein the mammal is a rodent or a primate.

33. 33. The computer-implemented method of claim 32, wherein the rodent is a mouse.

34. 34. The computer-implemented method of any one of claims 29 to 33, wherein the simulation further uses measured brain anatomy or brain function from other human subjects to design simulation parameters for predicting the past locations, the current locations, and the future locations of the pathological protein aggregates in the human subjects.

35. receiving a second image of the brain, the second image being taken after the neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; 35. The computer-implemented method of claim 1, further comprising: displaying a change in the total aggregate size for each voxel, the volume of each pathological protein aggregate in each voxel, and the aggregate density for each voxel in the images taken after the neurostimulation is applied to the brain of the subject compared to the images taken before the neurostimulation is applied to the brain of the subject.

36. receiving a second image of the brain, the second image being taken after the neurostimulation is applied to the brain; repeating steps (b) through (o) using the second image; modulating one or more programmed neurostimulation parameters based on any changes in the past, current, or future locations where the pathological protein aggregates are predicted to appear; 35. The computer-implemented method of claim 1, further comprising: instructing a neurostimulation device to apply modulated neurostimulation to the brain of the subject to treat the neurological or neurodegenerative disease in the subject.

37. 37. The computer-implemented method of any one of claims 1 to 36, further comprising storing a user profile of the subject, the user profile including information regarding the programmed neurostimulation parameters used to apply neurostimulation to the brain of the subject to treat the neurological or neurodegenerative disease based on the location of pathological protein aggregates present or predicted to occur in the future.

38. A non-transitory computer readable medium comprising program instructions which, when executed by a processor in a computer, cause the processor to perform the method of any one of claims 1 to 37.

39. 40. A kit comprising the non-transitory computer-readable medium of claim 38 and instructions for treating a neurological or neurodegenerative disorder in a subject using neurostimulation.

40. 1. A method for treating a neurological or neurodegenerative disease in a subject, the method comprising: imaging pathological protein aggregates in the brain of the subject; predicting where pathological protein aggregates will occur based on the locations of the pathological protein aggregates detected in the brain of the subject by said imaging, using the computer-implemented method of any one of claims 1 to 37; applying neurostimulation to locations in the brain where the pathological protein aggregates are detected in the subject's brain by the imaging and where the computer-implemented method predicts that pathological protein aggregates will occur.

41. 41. The method of claim 40, wherein the imaging is performed using computed tomography (CT), single photon emission computed tomography (SPECT), magnetic resonance imaging, functional magnetic resonance imaging, optogenetic functional magnetic resonance imaging, or positron emission tomography (PET).

42. 42. The method of claim 40 or 41, further comprising adjusting a stimulation frequency and pulse width of the neurostimulation to target specific neuronal cell types or circuits in the brain at the location in the brain where the computer-implemented method predicts that the pathological protein aggregates are present or will occur in the future.

43. 43. The method of any one of claims 40 to 42, wherein the neurological or neurodegenerative disease is a synucleinopathy.

44. 44. The method of claim 43, wherein the synucleinopathy is Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy.

45. 43. The method of any one of claims 40 to 42, wherein the neurological or neurodegenerative disease is Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia.

46. The method of any one of claims 40 to 45, wherein the pathological protein aggregates comprise alpha-synuclein aggregates.

47. 47. The method of any one of claims 40 to 46, wherein applying the neurostimulation comprises applying the neurostimulation using electrodes.

48. 48. The method of claim 47, wherein the electrodes are deep electrodes or surface electrodes.

49. 48. The method of claim 47, wherein the electrodes are a non-penetrating surface electrode array or a penetrating brain electrode array.

50. 50. The method of any one of claims 40 to 49, wherein applying the neurostimulation comprises applying deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation.

51. 47. The method of any one of claims 40 to 46, wherein applying the neural stimulation comprises optogenetically applying the neural stimulation.

52. Nerve stimulation is introducing a recombinant polynucleotide encoding a light-responsive ion channel into a neuron at the location in the brain where the computer-implemented method predicts that pathological protein aggregates exist or will occur in the future, wherein the light-responsive ion channel is expressed in the neuron; illuminating the light-responsive ion channel with light at a wavelength that activates the light-responsive ion channel, wherein conduction of ions by the light-responsive ion channel in response to absorption of light results in hyperpolarization or depolarization of the neuron; 52. The method of claim 51 , wherein the optogenetic is applied by a method comprising:

53. 53. The method of claim 52, wherein the light-responsive ion channel is a light-responsive anion-conducting opsin or a light-responsive proton conductance modulator.

54. The light-responsive anion-conducting opsin is capable of reacting with chloride ions (Cl － 54. The method of claim 53, wherein the

55. 55. The method of claim 53 or 54, wherein the anion-conducting opsin is an anion-conducting channelrhodopsin or halorhodopsin.

56. 56. The method of claim 55, wherein the halorhodopsin is Natronomonas pharaonis halorhodopsin (NpHR), enhanced NpHR (eNpHR) 1.0, eNpHR 2.0, or eNpHR 3.

0.

57. 56. The method of claim 55, wherein the anion-conducting channelrhodopsin is iC1C2, SwiChR, SwiChR++, or iC++.

58. 54. The method of claim 53, wherein the light-responsive proton conductance modulator is bacteriorhodopsin or archaealhodopsin.

59. 59. The method of claim 58, wherein the light-responsive proton conductance modulator is Arch from Halorubrum sodomense, ArchT from Halorubrum sp., TP009 from Leptosphaeria maculans, or Mac from Leptosphaeria maculans.

60. 53. The method of claim 52, wherein the light-responsive ion channel is a light-responsive cation-conducting opsin.

61. The light-responsive cation-conducting opsin is a calcium cation (Ca ２＋ 61. The method of claim 60, wherein the

62. 62. The method of claim 60 or 61, wherein the light-responsive cation-conducting opsin is a light-responsive cation-conducting channelrhodopsin.

63. 63. The method of claim 62, wherein the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin or Volvox carteri channelrhodopsin.

64. 64. The method of claim 63, wherein the light-responsive cation-conducting channelrhodopsin is Chlamydomonas reinhardtii channelrhodopsin-1 (ChR1), Chlamydomonas reinhardtii channelrhodopsin-2 (ChR2), Volvox carteri channelrhodopsin-1 (VChR1), or a chimeric ChR1-VChR1 channelrhodopsin.

65. 65. The method of any one of claims 52 to 64, wherein the polynucleotide encoding the light-responsive ion channel is provided by a viral vector.

66. 66. The method of claim 65, wherein the viral vector is a lentiviral vector or an adeno-associated viral (AAV) vector.

67. 67. The method of claim 65 or 66, wherein the viral vector is stereotactically injected into the brain at the location where the computer-implemented method predicts that pathological protein aggregates exist or will occur in the future.

68. 68. The method of any one of claims 65 to 67, wherein the vector further comprises a neuron-specific promoter operably linked to the polynucleotide encoding the light-responsive ion channel.

69. 69. The method of any one of claims 52 to 68, wherein expression of the light-responsive ion channel is inducible.

70. 70. The method of any one of claims 52-69, wherein illuminating the light-responsive ion channel comprises delivering light from a light source to the light-responsive ion channel using a fiber optic-based optical neural interface.

71. 71. The method of claim 70, wherein the light source is a solid state diode laser.

72. 72. The method of any one of claims 52 to 71, wherein applying the neurostimulation comprises applying neurostimulation to a motor cortical region or a subcortical region of the brain.

73. 73. The method of any one of claims 40 to 72, wherein multiple cycles of the neural stimulation are performed.

74. 74. The method of any one of claims 40 to 73, further comprising assessing the effectiveness of said treatment of said neurological or neurodegenerative disease in said subject.

75. 75. The method of claim 74, wherein said evaluating comprises imaging the brain of said subject after said neurostimulation to measure the size and identify the location of said pathological protein aggregates.

76. 76. The method of claim 74 or 75, wherein said evaluating comprises measuring brain function of the subject after said neurostimulation.

77. 77. The method of claim 76, wherein measuring the brain function comprises performing electroencephalography (EEG), stereoencephalography (sEEG), electrocorticography (ECoG), magnetoencephalography (MEG), single photon emission computed tomography (SPECT), functional magnetic resonance imaging (fMRI), optogenetic functional magnetic resonance imaging, or positron emission tomography (PET).

78. 78. The method of claim 76 or 77, further comprising modulating one or more programmed neurostimulation parameters to improve the brain function.

79. 79. The method of any one of claims 74-78, further comprising assessing the severity of symptoms of the neurological or neurodegenerative disease using a visual analog scale, a verbal rating scale, the Movement Disorder Society-sponsored Unified Parkinson's Disease Rating Scale-Revised (MDS-UPDRS), the Hoehn and Yahr (HnY) scale, or the Montreal Cognitive Assessment (MoCA) scale.

80. 1. A system for treating a neurological or neurodegenerative disease in a subject, the system comprising: A neurostimulation device; 38. A system comprising: a processor programmed in accordance with the computer-implemented method of any one of claims 1 to 37, wherein the processor is programmed to instruct the neurostimulation device to deliver neurostimulation to the brain of the subject in a manner effective to treat the neurological or neurodegenerative disease in the subject, wherein the neurostimulation is applied to the brain at a predicted current location of the pathological protein aggregate, at a predicted future location of the pathological protein aggregate, or at a predicted past location of the pathological protein aggregate, or a combination thereof.

81. 81. The system of claim 80, wherein the neurostimulator device comprises an electrode.

82. 82. The system of claim 81, wherein the electrodes are deep electrodes or surface electrodes.

83. 83. The system of claim 82, wherein the electrodes are a non-penetrating surface electrode array or a penetrating brain electrode array.

84. 84. The system of any one of claims 80 to 83, wherein the neurostimulation device performs deep brain stimulation, transcranial magnetic stimulation, or transcranial electrical stimulation.

85. A system according to any one of claims 80 to 84, further comprising a display.

86. 86. The system of claim 85, wherein the display displays an image of the subject's brain showing the predicted current location, past location, or future location of the pathological protein aggregates determined by the computer-implemented method.

87. 87. The system of claim 85 or 86, wherein the display displays information regarding the coordinates of each pathological protein aggregate, the volume of each pathological protein aggregate, the aggregate density of each voxel, the total aggregate size of each voxel, the average aggregate size of each voxel, the total signal intensity of each voxel, the average signal intensity of each voxel, or the mapping of the locations of the pathological protein aggregates to neuroanatomical regions, or any combination thereof.

88. 88. The system of any one of claims 85 to 87, wherein the display displays information regarding the distribution of genetic effects on the regional spread and disruption of the pathological protein aggregates.

89. 89. The system of any one of claims 85-88, wherein the display displays information regarding the spreading, aggregation, and disintegration of the pathological protein aggregates throughout the neuroanatomical region and predicted changes in regional density of the pathological protein aggregates as a function of time as determined by modeling the spatial gene expression.

90. 90. The system of any one of claims 85 to 89, wherein the display displays information about the predicted past, current, and future locations of the pathological protein aggregates based on the modeling.

91. 91. The system of any one of claims 85 to 90, further comprising a user interface comprising an input electronically coupled to the processor for instructing the neurostimulation device to apply neurostimulation to the brain of the subject to treat the neurological or neurodegenerative disease in the subject.

92. 92. The system of claim 91, wherein the user interface is password protected and operable by a medical professional.

93. The system of any one of claims 85 to 92, wherein the neurological or neurodegenerative disease is a synucleinopathy.

94. 94. The system of claim 93, wherein the synucleinopathy is Parkinson's disease, dementia with Lewy bodies, multiple system atrophy, neuroaxonal dystrophy Shy-Drager syndrome, striatonigral degeneration, or olivopontocerebellar atrophy.

95. 93. The system of any one of claims 85 to 92, wherein the neurological or neurodegenerative disease is Alzheimer's disease, amyotrophic lateral sclerosis, or frontotemporal dementia.

Images