In 2018, tuberculosis (TB) was the leading cause of death by a single infectious disease, causing 10 million new cases and 1.5 million deaths. Upon Mycobacterium tuberculosis (Mtb) infection most people develop non-transmissible latent TB, and some develop active TB disease. Healthy hosts have a 5-15% lifetime risk of progressing from latent to active TB, and prior studies have demonstrated that a type I interferon-stimulated gene (ISG) signature can predict progression to active disease. Type I interferon (IFN) and ISG expression is induced by DNA-sensing pathways and has anti-viral functions. However, the innate immune mechanisms and genetics controlling type I IFN responses following Mtb infection are not well understood. We hypothesize that genetically regulated higher type 1 IFN responses are associated with lower anti-microbial responses and higher Mtb replication in macrophages, as well as increased risk of TB disease. To test this hypothesis, blood was collected from 40 healthy donors. Donors were genotyped using Illumina MEGA^EX SNP Array. Monocytes were isolated, differentiated into monocyte-derived macrophages (MDMs), and stimulated with 3 ligands (supercoiled plasmid DNA, cyclic guanosine monophosphate–adenosine monophosphate, and sheared calf thymus DNA) to activate DNA-sensing pathways and induce a type I IFN response. RNA was isolated at 4 and 24 hours. Interferon-beta (IFN-β) and interleukin-6 (IL-6) gene expression were quantified using Real-Time PCR. Donors had highly variable IFN-β induction upon ligand stimulation. Donors’ genotypes will be linked to these in vitro phenotypes to identify expression quantitative trait loci (eQTLs) that regulate IFN-β expression. We will assess if these functional polymorphisms in genes of interest are associated with TB disease using patient samples from a Brazilian cohort. The results of this investigation will identify novel pathways that control TB progression that can inform vaccine development and host-directed therapeutic approaches.

Millions of people have vision diseases that are not yet treatable, leading to blindness. Mouse models exist for some inherited retinal diseases, and thus have helped develop vision loss therapies. However, other common retinal diseases like glaucoma lack accurate mouse models. Human pluripotent stem cells (hPSCs) are a promising technology that provide a new way to model human retinal diseases. hPSCs can be induced to become layered, 3D mini-retinas called retinal organoids. Retinal organoids mirror early neurogenesis of the human retina, and thus can be used for modeling developmental disorders. However, we and other research groups have shown that as retinal organoids mature, they lose many features of the normal human retina. In particular, the neurons that are damaged during glaucoma, called retinal ganglion cells (RGCs), are not well preserved in organoids. As the organoid matures, the RGC layer becomes disorganized and RGCs migrate through the retina. RGCs are the projection neurons of the human retina, so their axons extend through the optic nerve and carry visual information to the brain. Unlike in human development, retinal organoids are grown in isolation from their brain targets. We wondered whether we could preserve RGC organization by providing organoid RGCs with synaptic targets. We investigated this hypothesis by combining the retinal organoids with their natural targets in the brain, the lateral geniculate nucleus and the superior colliculus. It is not yet known how to make these brain regions from hPSCs, so we used newborn mouse brain and combined the retinal organoids with these brain regions into structures called “assembloids.” We are now testing whether these assembloids preserve the RGC survival and laminar organization that retinal organoids lack. Promoting RGC survival and organization will allow organoids to become better in vitro models for glaucoma and improve the outcome for patients with vision loss.

Since Computed Tomography (CT) scans expose the patients to high x-ray radiation dose which may potentially induce lifetime risk of cancers, researchers have been finding ways to reduce the radiation dose while maintaining the high quality of reconstructed images. One approach to lower the total X-ray radiation dose is to reduce the number of projections acquired, which generated sparse-view CT image. However, when the number of view angles is too less to satisfy the Shannon/Nyquist sampling theorem, serious streaking artifacts will appear on the reconstructed images. In this work, we present two iterative reconstruction algorithms, which implement convolutional neural networks (CNN) in each iterative step, to help eliminate these defects. The first algorithm is LEARN, which uses a CNN in place of a regularization function while solving a least squares problem. The second algorithm is based on SART and the superiorization methodology (an iterative method for constrained optimization), where the CNN is used to perturb the solution between SART iterations. We use Tensorflow and the Pyro-NN library in Python to train on data obtained from The Cancer Imaging Archive’s QIN LUNG CT dataset, and compare the performance of these two frameworks from the perspectives of PSNR value (often used to exam the quality of an image), training loss, penalty term, and learning rate. Besides testing on different sparse-view imaging datasets, we also demonstrate the performance of the proposed networks in limited angle CT image, where some view angles are missing due to geometric constraints.

Retinal diseases tend to affect specific neuron subtypes, ranging from age-related macular degeneration, which is caused by the deterioration of photoreceptors near the central portion of the retina (macula), to glaucoma, in which abnormally high intraocular pressure leads to ganglion cell death. Unfortunately, adult mammals are not able to regenerate retinal neurons. However, zebrafish and other amphibians can completely regenerate their retinal neurons in many different models of damage, and restore retinal structure and visual function. The source of regeneration stems from the resident Müller glia cells, which normally provide neuronal support and span all three retinal layers. A critical gene for the initiation of transforming Müller glia into neurons was found to be Ascl1. This led our lab to hypothesize that the introduction and upregulation of Ascl1 in mammalian Müller glia might stimulate them to become retinal neurons after damage, as occurs in these other regenerating species. Indeed, after introducing Ascl1 into the Müller glia of mice, we found newly regenerated retinal interneurons (bipolar cells) that successfully integrated into the retinal circuitry and functionally responded to light stimulus. In addition to Ascl1, we have identified two other transcription factors, that when introduced in combination with Ascl1, stimulate the generation of two different retinal neurons (ganglion cells and amacrine cells). We are currently developing a model of glaucoma, damaging the ganglion cells with a neurotoxin, and then testing the visual acuity using Optomotry to determine whether regenerated ganglion cells will mediate a functional improvement. Ectopic expression of a proneural transcription factor to stimulate retinal regeneration provides a potential therapeutic intervention for treating blinding diseases, that even now, have few modest treatment options.

Marine heatwaves are the phenomena of abnormally warm ocean surface temperatures that last for an extended period of time. The most severe marine heatwave of recent times occurred from 2013 to 2016 in the Northeastern Pacific. This event, nicknamed ‘The Blob’, was scientifically fascinating because the ocean-atmosphere system maintained itself for so long in an anomalous state. In mid-2019, a marine heatwave with a likeness to ‘The Blob’ began forming. This research project focuses on analyzing the anomalous patterns in sea surface temperature, clouds, radiative fluxes, and turbulent fluxes that arise during the formation and duration of this event. We set out to understand if the more recent 2019 marine heatwave evolves in a similar way to that of ‘The Blob,’ and how it differs. This project uses NOAA Climate Forecast System Reanalysis (CFSR) data, which assimilates measurements using complex models to create the best estimates of atmospheric and oceanic variables with complete global spatial coverage. With this project, we aim to understand the atmospheric response to marine heatwaves using geospatial plots of mean temperature, fluxes, and cloud cover. We expect to see differences in the atmosphere response with regards to the net flux that caused the quick dissipation of the recent marine heatwave.