Brain type II calcium and calmodulin-dependent protein kinase: characterization of a brain-region specific isozyme and regulation by autophosphorylation

A variety of molecular approaches have been used to investigate the structural and enzymatic properties of rat brain type ll Ca^(2+) and calmodulin-dependent protein kinase (type ll CaM kinase). This thesis describes the isolation and biochemical characterization of a brain-region specific isozyme of the kinase and also the regulation the kinase activity by autophosphorylation.

The cerebellar isozyme of the type ll CaM kinase was purified and its biochemical properties were compared to the forebrain isozyme. The cerebellar isozyme is a large (500-kDa) multimeric enzyme composed of multiple copies of 50-kDa α subunits and 60/58-kDa β/β’ subunits. The holoenzyme contains approximately 2 α subunits and 8 β subunits. This contrasts to the forebrain isozyme, which is also composed of and β/β'subunits, but they are assembled into a holoenzyme of approximately 9 α subunits and 3 β/β ' subunits. The biochemical and enzymatic properties of the two isozymes are similar. The two isozymes differ in their association with subcellular structures. Approximately 85% of the cerebellar isozyme, but only 50% of the forebrain isozyme, remains associated with the particulate fraction after homogenization under standard conditions. Postsynaptic densities purified from forebrain contain the forebrain isozyme, and the kinase subunits make up about 16% of their total protein. Postsynaptic densities purified from cerebellum contain the cerebellar isozyme, but the kinase subunits make up only 1-2% of their total protein.

The enzymatic activity of both isozymes of the type II CaM kinase is regulated by autophosphorylation in a complex manner. The kinase is initially completely dependent on Ca^(2+)/calmodulin for phosphorylation of exogenous substrates as well as for autophosphorylation. Kinase activity becomes partially Ca^(2+) independent after autophosphorylation in the presence of Ca^(2+)/calmodulin. Phosphorylation of only a few subunits in the dodecameric holoenzyme is sufficient to cause this change, suggesting an allosteric interaction between subunits. At the same time, autophosphorylation itself becomes independent of Ca^(2+) These observations suggest that the kinase may be able to exist in at least two stable states, which differ in their requirements for Ca^(2+)/calmodulin.

The autophosphorylation sites that are involved in the regulation of kinase activity have been identified within the primary structure of the α and β subunits. We used the method of reverse phase-HPLC tryptic phosphopeptide mapping to isolate individual phosphorylation sites. The phosphopeptides were then sequenced by gas phase microsequencing. Phosphorylation of a single homologous threonine residue in the α and β subunits is correlated with the production of the Ca^(2+) -independent activity state of the kinase. In addition we have identified several sites that are phosphorylated only during autophosphorylation in the absence of Ca^(2+)/ calmodulin.

Latent variable models for neural data analysis

The brain is perhaps the most complex system to have ever been subjected to rigorous scientific investigation. The scale is staggering: over 10^11 neurons, each making an average of 10^3 synapses, with computation occurring on scales ranging from a single dendritic spine, to an entire cortical area. Slowly, we are beginning to acquire experimental tools that can gather the massive amounts of data needed to characterize this system. However, to understand and interpret these data will also require substantial strides in inferential and statistical techniques. This dissertation attempts to meet this need, extending and applying the modern tools of latent variable modeling to problems in neural data analysis.

It is divided into two parts. The first begins with an exposition of the general techniques of latent variable modeling. A new, extremely general, optimization algorithm is proposed - called Relaxation Expectation Maximization (REM) - that may be used to learn the optimal parameter values of arbitrary latent variable models. This algorithm appears to alleviate the common problem of convergence to local, sub-optimal, likelihood maxima. REM leads to a natural framework for model size selection; in combination with standard model selection techniques the quality of fits may be further improved, while the appropriate model size is automatically and efficiently determined. Next, a new latent variable model, the mixture of sparse hidden Markov models, is introduced, and approximate inference and learning algorithms are derived for it. This model is applied in the second part of the thesis.

The second part brings the technology of part I to bear on two important problems in experimental neuroscience. The first is known as spike sorting; this is the problem of separating the spikes from different neurons embedded within an extracellular recording. The dissertation offers the first thorough statistical analysis of this problem, which then yields the first powerful probabilistic solution. The second problem addressed is that of characterizing the distribution of spike trains recorded from the same neuron under identical experimental conditions. A latent variable model is proposed. Inference and learning in this model leads to new principled algorithms for smoothing and clustering of spike data.

The Role of Fucose in Learning and Memory: The Identification of a Fucosylprotein, Synapsin I, in the Rat Brain and the Characterization of its Fucosylation

Previous studies have shown that the glycoproteins containing the fucose moiety are involved in neuronal communication phenomena such as long-term potentiation and memory formation. These results imply that fucose containing glycoproteins might play an important role in learning and memory. To understand the role of fucose in neuronal communication, and the mechanisms by which fucose may be involved in information storage, the identification of fucosylproteins is essential. This report describes the identification and characterization of fucosylproteins in the brain, which will provide new insights into the role of the fucose involved molecular interactions.

Mathematical study of complex networks : brain, internet, and power grid

The dissertation is concerned with the mathematical study of various network problems. First, three real-world networks are considered: (i) the human brain network (ii) communication networks, (iii) electric power networks. Although these networks perform very different tasks, they share similar mathematical foundations. The high-level goal is to analyze and/or synthesis each of these systems from a “control and optimization” point of view. After studying these three real-world networks, two abstract network problems are also explored, which are motivated by power systems. The first one is “flow optimization over a flow network” and the second one is “nonlinear optimization over a generalized weighted graph”. The results derived in this dissertation are summarized below.

Brain Networks: Neuroimaging data reveals the coordinated activity of spatially distinct brain regions, which may be represented mathematically as a network of nodes (brain regions) and links (interdependencies). To obtain the brain connectivity network, the graphs associated with the correlation matrix and the inverse covariance matrix—describing marginal and conditional dependencies between brain regions—have been proposed in the literature. A question arises as to whether any of these graphs provides useful information about the brain connectivity. Due to the electrical properties of the brain, this problem will be investigated in the context of electrical circuits. First, we consider an electric circuit model and show that the inverse covariance matrix of the node voltages reveals the topology of the circuit. Second, we study the problem of finding the topology of the circuit based on only measurement. In this case, by assuming that the circuit is hidden inside a black box and only the nodal signals are available for measurement, the aim is to find the topology of the circuit when a limited number of samples are available. For this purpose, we deploy the graphical lasso technique to estimate a sparse inverse covariance matrix. It is shown that the graphical lasso may find most of the circuit topology if the exact covariance matrix is well-conditioned. However, it may fail to work well when this matrix is ill-conditioned. To deal with ill-conditioned matrices, we propose a small modification to the graphical lasso algorithm and demonstrate its performance. Finally, the technique developed in this work will be applied to the resting-state fMRI data of a number of healthy subjects.

Communication Networks: Congestion control techniques aim to adjust the transmission rates of competing users in the Internet in such a way that the network resources are shared efficiently. Despite the progress in the analysis and synthesis of the Internet congestion control, almost all existing fluid models of congestion control assume that every link in the path of a flow observes the original source rate. To address this issue, a more accurate model is derived in this work for the behavior of the network under an arbitrary congestion controller, which takes into account of the effect of buffering (queueing) on data flows. Using this model, it is proved that the well-known Internet congestion control algorithms may no longer be stable for the common pricing schemes, unless a sufficient condition is satisfied. It is also shown that these algorithms are guaranteed to be stable if a new pricing mechanism is used.

Electrical Power Networks: Optimal power flow (OPF) has been one of the most studied problems for power systems since its introduction by Carpentier in 1962. This problem is concerned with finding an optimal operating point of a power network minimizing the total power generation cost subject to network and physical constraints. It is well known that OPF is computationally hard to solve due to the nonlinear interrelation among the optimization variables. The objective is to identify a large class of networks over which every OPF problem can be solved in polynomial time. To this end, a convex relaxation is proposed, which solves the OPF problem exactly for every radial network and every meshed network with a sufficient number of phase shifters, provided power over-delivery is allowed. The concept of “power over-delivery” is equivalent to relaxing the power balance equations to inequality constraints.

Flow Networks: In this part of the dissertation, the minimum-cost flow problem over an arbitrary flow network is considered. In this problem, each node is associated with some possibly unknown injection, each line has two unknown flows at its ends related to each other via a nonlinear function, and all injections and flows need to satisfy certain box constraints. This problem, named generalized network flow (GNF), is highly non-convex due to its nonlinear equality constraints. Under the assumption of monotonicity and convexity of the flow and cost functions, a convex relaxation is proposed, which always finds the optimal injections. A primary application of this work is in the OPF problem. The results of this work on GNF prove that the relaxation on power balance equations (i.e., load over-delivery) is not needed in practice under a very mild angle assumption.

Generalized Weighted Graphs: Motivated by power optimizations, this part aims to find a global optimization technique for a nonlinear optimization defined over a generalized weighted graph. Every edge of this type of graph is associated with a weight set corresponding to the known parameters of the optimization (e.g., the coefficients). The motivation behind this problem is to investigate how the (hidden) structure of a given real/complex valued optimization makes the problem easy to solve, and indeed the generalized weighted graph is introduced to capture the structure of an optimization. Various sufficient conditions are derived, which relate the polynomial-time solvability of different classes of optimization problems to weak properties of the generalized weighted graph such as its topology and the sign definiteness of its weight sets. As an application, it is proved that a broad class of real and complex optimizations over power networks are polynomial-time solvable due to the passivity of transmission lines and transformers.

Identification of fucose-α(1-2)-galactose binding proteins in the mammalian brain

Fucose-α(1-2)-galactose (Fucα(1-2)Gal) carbohydrates have been implicated in cognitive functions. However, the underlying molecular mechanisms that govern these processes are not well understood. While significant progress has been made toward identifying glycoconjugates bearing this carbohydrate epitope, a major challenge remains the discovery of interactions mediated by these sugars. Here, we employ the use of multivalent glycopolymers to enable the proteomic identification of weak affinity, low abundant Fucα(1-2)Gal-binding proteins (i.e. lectins) from the brain. End-biotinylated glycopolymers containing photoactivatable crosslinkers were used to capture and enrich potential Fucα(1-2)Gal-specific lectins from rat brain lysates. Candidate lectins were tested for their ability to bind Fucα(1-2)Gal, and the functional significance of the interaction was investigated for one such candidate, SV2a, using a knock-out mouse system. Our results suggest an important role for this glycan-lectin interaction in facilitating synaptic changes necessary for neuronal communication. This study highlights the use of glycopolymer mimetics to discover novel lectins and identify functional interactions between fucosyl carbohydrates and lectins in the brain.

Effects of particle size ratio on single particle motion in colloidal dispersions

The motion of a single Brownian particle of arbitrary size through a dilute colloidal dispersion of neutrally buoyant bath spheres of another characteristic size in a Newtonian solvent is examined in two contexts. First, the particle in question, the probe particle, is subject to a constant applied external force drawing it through the suspension as a simple model for active and nonlinear microrheology. The strength of the applied external force, normalized by the restoring forces of Brownian motion, is the Péclet number, Pe. This dimensionless quantity describes how strongly the probe is upsetting the equilibrium distribution of the bath particles. The mean motion and fluctuations in the probe position are related to interpreted quantities of an effective viscosity of the suspension. These interpreted quantities are calculated to first order in the volume fraction of bath particles and are intimately tied to the spatial distribution, or microstructure, of bath particles relative to the probe. For weak Pe, the disturbance to the equilibrium microstructure is dipolar in nature, with accumulation and depletion regions on the front and rear faces of the probe, respectively. With increasing applied force, the accumulation region compresses to form a thin boundary layer whose thickness scales with the inverse of Pe. The depletion region lengthens to form a trailing wake. The magnitude of the microstructural disturbance is found to grow with increasing bath particle size -- small bath particles in the solvent resemble a continuum with effective microviscosity given by Einstein's viscosity correction for a dilute dispersion of spheres. Large bath particles readily advect toward the minimum approach distance possible between the probe and bath particle, and the probe and bath particle pair rotating as a doublet is the primary mechanism by which the probe particle is able to move past; this is a process that slows the motion of the probe by a factor of the size ratio. The intrinsic microviscosity is found to force thin at low Péclet number due to decreasing contributions from Brownian motion, and force thicken at high Péclet number due to the increasing influence of the configuration-averaged reduction in the probe's hydrodynamic self mobility. Nonmonotonicity at finite sizes is evident in the limiting high-Pe intrinsic microviscosity plateau as a function of bath-to-probe particle size ratio. The intrinsic microviscosity is found to grow with the size ratio for very small probes even at large-but-finite Péclet numbers. However, even a small repulsive interparticle potential, that excludes lubrication interactions, can reduce this intrinsic microviscosity back to an order one quantity. The results of this active microrheology study are compared to previous theoretical studies of falling-ball and towed-ball rheometry and sedimentation and diffusion in polydisperse suspensions, and the singular limit of full hydrodynamic interactions is noted.

Second, the probe particle in question is no longer subject to a constant applied external force. Rather, the particle is considered to be a catalytically-active motor, consuming the bath reactant particles on its reactive face while passively colliding with reactant particles on its inert face. By creating an asymmetric distribution of reactant about its surface, the motor is able to diffusiophoretically propel itself with some mean velocity. The effects of finite size of the solute are examined on the leading order diffusive microstructure of reactant about the motor. Brownian and interparticle contributions to the motor velocity are computed for several interparticle interaction potential lengths and finite reactant-to-motor particle size ratios, with the dimensionless motor velocity increasing with decreasing motor size. A discussion on Brownian rotation frames the context in which these results could be applicable, and future directions are proposed which properly incorporate reactant advection at high motor velocities.

Complex behavior and perception in Drosophila emerges from iterative feedback-regulated reflexes

For a hungry fruit fly, locating and landing on a fermenting fruit where it can feed, find mates, and lay eggs, is an essential and difficult task requiring the integration of both olfactory and visual cues. Understanding how flies accomplish this will help provide a comprehensive ethological context for the expanding knowledge of their neural circuits involved in processing olfaction and vision, as well as inspire novel engineering solutions for control and estimation in computationally limited robotic applications. In this thesis, I use novel high throughput methods to develop a detailed overview of how flies track odor plumes, land, and regulate flight speed. Finally, I provide an example of how these insights can be applied to robotic applications to simplify complicated estimation problems. To localize an odor source, flies exhibit three iterative, reflex-driven behaviors. Upon encountering an attractive plume, flies increase their flight speed and turn upwind using visual cues. After losing the plume, flies begin zigzagging crosswind, again using visual cues to control their heading. After sensing an attractive odor, flies become more attracted to small visual features, which increases their chances of finding the plume source. Their changes in heading are largely controlled by open-loop maneuvers called saccades, which they direct towards and away from visual features. If a fly decides to land on an object, it begins to decelerate so as to maintain a stereotypical ratio of expansion to retinal size. Once they reach a stereotypical distance from the target, flies extend their legs in preparation for touchdown. Although it is unclear what cues they use to trigger this behavior, previous studies have indicated that it is likely under visual control. In Chapter 3, I use a nonlinear control theoretic analysis and robotic testbed to propose a novel and putative mechanism for how a fly might visually estimate distance by actively decelerating according to a visual control law. Throughout these behaviors, a common theme is the visual control of flight speed. Using genetic tools I show that the neuromodulator octopamine plays an important role in regulating flight speed, and propose a neural circuit for how this controller might be implemented in the flies brain. Two general biological and engineering principles are evident across my experiments: (1) complex behaviors, such as foraging, can emerge from the interactions of simple independent sensory-motor modules; (2) flies control their behavior in such a way that simplifies complex estimation problems.

Binding site size limitations of imidazole-pyrrole polyamides for recognition in the minor groove of DNA

The discovery that the three ring polyamide Im-Py-Py-Dp containing imidazole (Im) and pyrrole (Py) carboxamides binds the DNA sequence 5'-(A,T)G(A,T)C(A,T)-3' as an antiparallel dimer offers a new model for the design of ligands for specific recognition of sequences in the minor groove containing both G,C and A,T base pairs. In Chapter 2, experiments are described in which the sequential addition of five N- methylpyrrolecarboxamides to the imidazole-pyrrole polyamide Im-Py-Py-Dp affords a series of six homologous polyamides, Im-(Py)_2-7-Dp, that differ in the size of their binding site, apparent first order binding affinity, and sequence specificity. These results demonstrate that DNA sequences up to nine base pairs in length can be specifically recognized by imidazole-pyrrole polyamides containing three to seven rings by 2:1 polyamide-DNA complex formation in the minor groove. Recognition of a nine base pair site defines the new lower limit of the binding site size that can be recognized by polyamides containing exclusively imidazole and pyrrolecarboxamides. The results of this study should provide useful guidelines for the design of new polyamides that bind longer DNA sites with enhanced affinity and specificity.

In Chapter 3 the design and synthesis of the hairpin polyamide Im-Py-Im-Py-γ-Im- Py-Im-Py-Dp is described. Quantitative DNase I footprint titration experiments reveal that Im-Py-Im-Py-γ-Im-Py-Im-Py-Dp binds six base pair 5'-(A,T)GCGC(A,T)-3' sequences with 30-fold higher affinity than the unlinked polyamide Im-Py-Im-Py-Dp. The hairpin polyamide does not discriminate between A•T and T•A at the first and sixth positions of the binding site as three sites 5'-TGCGCT-3', 5'-TGCGCA-3', and 5 'AGCGCT- 3' are bound with similar affinity. However, Im-Py-Im-Py-γ-Im-Py-Im-PyDp is specific for and discriminates between G•C and C•G base pairs in the 5'-GCGC-3' core as evidenced by lower affinities for the mismatched sites 5'-AACGCA-3', 5'- TGCGTT-3', 5'-TGCGGT-3', and 5'-ACCGCT-3'.

In Chapter 4, experiments are described in which a kinetically stable hexa-aza Schiff base La³⁺ complex is covalently attached to a Tat(49-72) peptide which has been shown to bind the HIV-1 TAR RNA sequence. Although these metallo-peptides cleave TAR site-specifically in the hexanucleotide loop to afford products consistent with hydrolysis, a series of control experiments suggests that the observed cleavage is not caused by a sequence-specifically bound Tat(49-72)-La(L)³⁺ peptide.

Social saliency : visual psychophysics and single-neuron recordings in humans

My thesis studies how people pay attention to other people and the environment. How does the brain figure out what is important and what are the neural mechanisms underlying attention? What is special about salient social cues compared to salient non-social cues? In Chapter I, I review social cues that attract attention, with an emphasis on the neurobiology of these social cues. I also review neurological and psychiatric links: the relationship between saliency, the amygdala and autism. The first empirical chapter then begins by noting that people constantly move in the environment. In Chapter II, I study the spatial cues that attract attention during locomotion using a cued speeded discrimination task. I found that when the motion was expansive, attention was attracted towards the singular point of the optic flow (the focus of expansion, FOE) in a sustained fashion. The more ecologically valid the motion features became (e.g., temporal expansion of each object, spatial depth structure implied by distribution of the size of the objects), the stronger the attentional effects. However, compared to inanimate objects and cues, people preferentially attend to animals and faces, a process in which the amygdala is thought to play an important role. To directly compare social cues and non-social cues in the same experiment and investigate the neural structures processing social cues, in Chapter III, I employ a change detection task and test four rare patients with bilateral amygdala lesions. All four amygdala patients showed a normal pattern of reliably faster and more accurate detection of animate stimuli, suggesting that advantageous processing of social cues can be preserved even without the amygdala, a key structure of the “social brain”. People not only attend to faces, but also pay attention to others’ facial emotions and analyze faces in great detail. Humans have a dedicated system for processing faces and the amygdala has long been associated with a key role in recognizing facial emotions. In Chapter IV, I study the neural mechanisms of emotion perception and find that single neurons in the human amygdala are selective for subjective judgment of others’ emotions. Lastly, people typically pay special attention to faces and people, but people with autism spectrum disorders (ASD) might not. To further study social attention and explore possible deficits of social attention in autism, in Chapter V, I employ a visual search task and show that people with ASD have reduced attention, especially social attention, to target-congruent objects in the search array. This deficit cannot be explained by low-level visual properties of the stimuli and is independent of the amygdala, but it is dependent on task demands. Overall, through visual psychophysics with concurrent eye-tracking, my thesis found and analyzed socially salient cues and compared social vs. non-social cues and healthy vs. clinical populations. Neural mechanisms underlying social saliency were elucidated through electrophysiology and lesion studies. I finally propose further research questions based on the findings in my thesis and introduce my follow-up studies and preliminary results beyond the scope of this thesis in the very last section, Future Directions.

I. Stokes flow past a thin screen. II. Viscous flows past porous bodies of finite size

Part I

The slow, viscous flow past a thin screen is analyzed based on Stokes equations. The problem is reduced to an associated electric potential problem as introduced by Roscoe. Alternatively, the problem is formulated in terms of a Stokeslet distribution, which turns out to be equivalent to the first approach.

Special interest is directed towards the solution of the Stokes flow past a circular annulus. A "Stokeslet" formulation is used in this analysis. The problem is finally reduced to solving a Fredholm integral equation of the second kind. Numerical data for the drag coefficient and the mean velocity through the hole of the annulus are obtained.

Stokes flow past a circular screen with numerous holes is also attempted by assuming a set of approximate boundary conditions. An "electric potential" formulation is used, and the problem is also reduced to solving a Fredholm integral equation of the second kind. Drag coefficient and mean velocity through the screen are computed.

Part II

The purpose of this investigation is to formulate correctly a set of boundary conditions to be prescribed at the interface between a viscous flow region and a porous medium so that the problem of a viscous flow past a porous body can be solved.

General macroscopic equations of motion for flow through porous media are first derived by averaging Stokes equations over a volume element of the medium. These equations, including viscous stresses for the description, are more general than Darcy's law. They reduce to Darcy's law when the Darcy number becomes extremely small.

The interface boundary conditions of the first kind are then formulated with respect to the general macroscopic equations applied within the porous region. An application of such equations and boundary conditions to a Poiseuille shear flow problem demonstrates that there usually exists a thin interface layer immediately inside the porous medium in which the tangential velocity varies exponentially and Darcy's law does not apply.

With Darcy's law assumed within the porous region, interface boundary conditions of the second kind are established which relate the flow variables across the interface layer. The primary feature is a jump condition on the tangential velocity, which is found to be directly proportional to the normal gradient of the tangential velocity immediately outside the porous medium. This is in agreement with the experimental results of Beavers, et al.

The derived boundary conditions are applied in the solutions of two other problems: (1) Viscous flow between a rotating solid cylinder and a stationary porous cylinder, and (2) Stokes flow past a porous sphere.

Advanced Monte Carlo Simulation and Machine Learning for Frequency Domain Optical Coherence Tomography

Optical Coherence Tomography(OCT) is a popular, rapidly growing imaging technique with an increasing number of bio-medical applications due to its noninvasive nature. However, there are three major challenges in understanding and improving an OCT system: (1) Obtaining an OCT image is not easy. It either takes a real medical experiment or requires days of computer simulation. Without much data, it is difficult to study the physical processes underlying OCT imaging of different objects simply because there aren't many imaged objects. (2) Interpretation of an OCT image is also hard. This challenge is more profound than it appears. For instance, it would require a trained expert to tell from an OCT image of human skin whether there is a lesion or not. This is expensive in its own right, but even the expert cannot be sure about the exact size of the lesion or the width of the various skin layers. The take-away message is that analyzing an OCT image even from a high level would usually require a trained expert, and pixel-level interpretation is simply unrealistic. The reason is simple: we have OCT images but not their underlying ground-truth structure, so there is nothing to learn from. (3) The imaging depth of OCT is very limited (millimeter or sub-millimeter on human tissues). While OCT utilizes infrared light for illumination to stay noninvasive, the downside of this is that photons at such long wavelengths can only penetrate a limited depth into the tissue before getting back-scattered. To image a particular region of a tissue, photons first need to reach that region. As a result, OCT signals from deeper regions of the tissue are both weak (since few photons reached there) and distorted (due to multiple scatterings of the contributing photons). This fact alone makes OCT images very hard to interpret.

This thesis addresses the above challenges by successfully developing an advanced Monte Carlo simulation platform which is 10000 times faster than the state-of-the-art simulator in the literature, bringing down the simulation time from 360 hours to a single minute. This powerful simulation tool not only enables us to efficiently generate as many OCT images of objects with arbitrary structure and shape as we want on a common desktop computer, but it also provides us the underlying ground-truth of the simulated images at the same time because we dictate them at the beginning of the simulation. This is one of the key contributions of this thesis. What allows us to build such a powerful simulation tool includes a thorough understanding of the signal formation process, clever implementation of the importance sampling/photon splitting procedure, efficient use of a voxel-based mesh system in determining photon-mesh interception, and a parallel computation of different A-scans that consist a full OCT image, among other programming and mathematical tricks, which will be explained in detail later in the thesis.

Next we aim at the inverse problem: given an OCT image, predict/reconstruct its ground-truth structure on a pixel level. By solving this problem we would be able to interpret an OCT image completely and precisely without the help from a trained expert. It turns out that we can do much better. For simple structures we are able to reconstruct the ground-truth of an OCT image more than 98% correctly, and for more complicated structures (e.g., a multi-layered brain structure) we are looking at 93%. We achieved this through extensive uses of Machine Learning. The success of the Monte Carlo simulation already puts us in a great position by providing us with a great deal of data (effectively unlimited), in the form of (image, truth) pairs. Through a transformation of the high-dimensional response variable, we convert the learning task into a multi-output multi-class classification problem and a multi-output regression problem. We then build a hierarchy architecture of machine learning models (committee of experts) and train different parts of the architecture with specifically designed data sets. In prediction, an unseen OCT image first goes through a classification model to determine its structure (e.g., the number and the types of layers present in the image); then the image is handed to a regression model that is trained specifically for that particular structure to predict the length of the different layers and by doing so reconstruct the ground-truth of the image. We also demonstrate that ideas from Deep Learning can be useful to further improve the performance.

It is worth pointing out that solving the inverse problem automatically improves the imaging depth, since previously the lower half of an OCT image (i.e., greater depth) can be hardly seen but now becomes fully resolved. Interestingly, although OCT signals consisting the lower half of the image are weak, messy, and uninterpretable to human eyes, they still carry enough information which when fed into a well-trained machine learning model spits out precisely the true structure of the object being imaged. This is just another case where Artificial Intelligence (AI) outperforms human. To the best knowledge of the author, this thesis is not only a success but also the first attempt to reconstruct an OCT image at a pixel level. To even give a try on this kind of task, it would require fully annotated OCT images and a lot of them (hundreds or even thousands). This is clearly impossible without a powerful simulation tool like the one developed in this thesis.

Delivery of Targeted Nanoparticles Across the Blood-Brain Barrier Using a Detachable Targeting Ligand

Chronic diseases of the central nervous system are poorly treated due to the inability of most therapeutics to cross the blood-brain barrier. The blood-brain barrier is an anatomical and physiological barrier that severely restricts solute influx, including most drugs, from the blood to the brain. One promising method to overcome this obstacle is to use endogenous solute influx systems at the blood-brain barrier to transport drugs. Therapeutics designed to enter the brain through transcytosis by binding the transferrin receptor, however, are restricted within endothelial cells. The focus of this work was to develop a method to increase uptake of transferrin-containing nanoparticles into the brain by overcoming these restrictive processes.

To accomplish this goal, nanoparticles were prepared with surface transferrin molecules bound through various liable chemical bonds. These nanoparticles were designed to shed the targeting molecule during transcytosis to allow increased accumulation of nanoparticles within the brain.

Transferrin was added to the surface of nanoparticles through either redox or pH sensitive chemistry. First, nanoparticles with transferrin bound through disulfide bonds were prepared. These nanoparticles showed decreased avidity for the transferrin receptor after exposure to reducing agents and increased ability to enter the brain in vivo compared to those lacking the disulfide link.

Next, transferrin was attached through a chemical bond that cleaves at mildly acidic pH. Nanoparticles containing a cleavable link between transferrin and gold nanoparticle cores were found to both cross an in vitro model of the blood-brain barrier and accumulate within the brain in significantly higher numbers than similar nanoparticles lacking the cleavable bond. Also, this increased accumulation was not seen when using this same strategy with an antibody to transferrin receptor, indicating that behavior of nanoparticles at the blood-brain barrier varies depending on what type of targeting ligand is used.

Finally, polymeric nanoparticles loaded with dopamine and utilizing a superior acid-cleavable targeting chemistry were investigated as a potential treatment for Parkinson’s disease. These nanoparticles were capable of increasing dopamine quantities in the brains of healthy mice, highlighting the therapeutic potential of this design. Overall, this work describes a novel method to increase targeted nanoparticle accumulation in the brain.