Robust single-shot 3D fluorescence imaging in scattering media with a simulator-trained neural network.

Imaging through scattering is a pervasive and difficult problem in many biological applications. The high background and the exponentially attenuated target signals due to scattering fundamentally limits the imaging depth of fluorescence microscopy. Light-field systems are favorable for high-speed volumetric imaging, but the 2D-to-3D reconstruction is fundamentally ill-posed, and scattering exacerbates the condition of the inverse problem. Here, we develop a scattering simulator that models low-contrast target signals buried in heterogeneous strong background. We then train a deep neural network solely on synthetic data to descatter and reconstruct a 3D volume from a single-shot light-field measurement with low signal-to-background ratio (SBR). We apply this network to our previously developed computational miniature mesoscope and demonstrate the robustness of our deep learning algorithm on scattering phantoms with different scattering conditions. The network can robustly reconstruct emitters in 3D with a 2D measurement of SBR as low as 1.05 and as deep as a scattering length. We analyze fundamental tradeoffs based on network design factors and out-of-distribution data that affect the deep learning model's generalizability to real experimental data. Broadly, we believe that our simulator-based deep learning approach can be applied to a wide range of imaging through scattering techniques where experimental paired training data is lacking.

Displacement-agnostic coherent imaging through scatter with an interpretable deep neural network.

Coherent imaging through scatter is a challenging task. Both model-based and data-driven approaches have been explored to solve the inverse scattering problem. In our previous work, we have shown that a deep learning approach can make high-quality and highly generalizable predictions through unseen diffusers. Here, we propose a new deep neural network model that is agnostic to a broader class of perturbations including scatterer change, displacements, and system defocus up to 10× depth of field. In addition, we develop a new analysis framework for interpreting the mechanism of our deep learning model and visualizing its generalizability based on an unsupervised dimension reduction technique. We show that our model can unmix the scattering-specific information and extract the object-specific information and achieve generalization under different scattering conditions. Our work paves the way to a robust and interpretable deep learning approach to imaging through scattering media.

Deep learning approach for Fourier ptychography microscopy.

Convolutional neural networks (CNNs) have gained tremendous success in solving complex inverse problems. The aim of this work is to develop a novel CNN framework to reconstruct video sequences of dynamic live cells captured using a computational microscopy technique, Fourier ptychographic microscopy (FPM). The unique feature of the FPM is its capability to reconstruct images with both wide field-of-view (FOV) and high resolution, i.e. a large space-bandwidth-product (SBP), by taking a series of low resolution intensity images. For live cell imaging, a single FPM frame contains thousands of cell samples with different morphological features. Our idea is to fully exploit the statistical information provided by these large spatial ensembles so as to make predictions in a sequential measurement, without using any additional temporal dataset. Specifically, we show that it is possible to reconstruct high-SBP dynamic cell videos by a CNN trained only on the first FPM dataset captured at the beginning of a time-series experiment. Our CNN approach reconstructs a 12800×10800 pixel phase image using only ∼25 seconds, a 50× speedup compared to the model-based FPM algorithm. In addition, the CNN further reduces the required number of images in each time frame by ∼ 6×. Overall, this significantly improves the imaging throughput by reducing both the acquisition and computational times. The proposed CNN is based on the conditional generative adversarial network (cGAN) framework. We further propose a mixed loss function that combines the standard image domain loss and a weighted Fourier domain loss, which leads to improved reconstruction of the high frequency information. Additionally, we also exploit transfer learning so that our pre-trained CNN can be further optimized to image other cell types. Our technique demonstrates a promising deep learning approach to continuously monitor large live-cell populations over an extended time and gather useful spatial and temporal information with sub-cellular resolution.

In vitro predatory activity of Arthrobotrys oligospora and after passing through gastrointestinal tract of small ruminants on infective larvae of trichostrongylides.

In vitro predatory activity of 157 native isolates of Arthrobotrys oligospora from China on larvae of trichostrongylides (Trichostrongylus colubriformis and Haemonchus contortus) in feces of sheep were assessed. The results showed that 135 of tested isolates of A. oligospora reduced the development of trichostrongylide larvae in feces by 90-99.99%, 11 isolates by 80-89.46% and 11 isolates by 14.58-78.82%. To understand their capacity of passing through gastrointestinal tract of sheep, 50 native isolates of A. oligospora were selected and assessed in sheep. Among these isolates, 16 isolates significantly reduced the number of larvae developing in the feces (P < 0.05); their percentage reduction of L3 ranged from 42.87% to 99.51% and the isolates tested were harvested in 5 g sub-samples of from sheep in each treatment group, indicating that these isolates had the capacity of preying larvae of trichostrongylides after the passage through gastrointestinal tract of sheep. The remaining isolates of A. oligospora were not able to survive after passage through gastrointestinal tract of sheep. In the following, the 16 isolates that presented more or less viability after sheep gastrointestinal passage were selected and assessed in goats. The results showed that the 11 isolates out of them could be able to pass through the digestive tract of goats without loss of ability of preying larvae of trichostrongylides in feces and their efficacies ranged from 53.88% to 94.28%, and that the isolates tested were harvested in 5 g sub-samples of feces from goats in each treatment group. In the current study, these isolates which demonstrated outstanding properties in vitro and could survive in the passage through the alimentary tract of sheep and goat should be potential candidates as a possible feed additive.

In vitro and in vivo studies of the native isolates of nematophagous fungi from China against the larvae of trichostrongylides.

To screen potential nematophagous fungi candidates for the biological control of parasitic nematodes in livestock, in vitro and in vivo studies of the native isolates of nematophagous fungi against the larvae of trichostrongylides were conducted. The in vitro predatory activity of 16 native nematophagous fungal isolates on the larvae of trichostrongylides in sheep feces was assessed. In the ten isolates of Duddingtonia flagrans, the reduction percentage for the infective larvae (L3) of Trichostrongylus colubriformis ranged from 57.21 to 99.83%, and that of Haemonchus contortus ranged from 62.12 to 99.88%. The analysis of the same assay on five isolates of Arthrobotrys superba and one isolate of A. cookedickinson (Monacrosporium cystosporum) showed comparable results with those for D. flagrans. To determine the excretion time of fungal isolates in feces after oral administration, D. flagrans (SDH035) were studied in vivo in sheep and rabbits. Results showed that the tested fungal isolates existed in sheep feces from 12 to 72 h after fungal treatment, and the fungal excretion in rabbit feces occurred at 4 h, reached a peak at 10 h, and declined gradually 18 h after oral administration. All the native fungal isolates were assessed after passing through the gastrointestinal tract of sheep. Treatment with isolates of D. flagrans significantly reduced the number of developing larvae in the feces, and the efficacies ranged from 55.15 to 98.82%. One out of the five isolates of A. superba and A. cookedickinson (BS002) survived after passing through the gastrointestinal tract, and the L3 reduction rates were 83.79 and 81.33%, respectively. Results of the present study provide information about the in vitro predatory activity of nematophagous fungi from China on the L3 of trichostrongylides and their ability to pass through the gastrointestinal tract before administering them for biocontrol.

The zinc-finger transcription factor KLF6 regulates cardiac fibrosis.

AIMS: Heart failure (HF) is one of the most devastating consequences of cardiovascular diseases. Regardless of etiology, cardiac fibrosis is present and promotes the loss of heart function in HF patients. Cardiac resident fibroblasts, in response to a host of pro-fibrogenic stimuli, trans-differentiate into myofibroblasts to mediate cardiac fibrosis, the underlying mechanism of which remains incompletely understood. METHODS: Fibroblast-myofibroblast transition was induced in vitro by exposure to transforming growth factor (TGF-ß). Cardiac fibrosis was induced in mice by either transverse aortic constriction (TAC) or by chronic infusion with angiotensin II (Ang II). RESULTS: Through bioinformatic screening, we identified Kruppel-like factor 6 (KLF6) as a transcription factor preferentially up-regulated in cardiac fibroblasts from individuals with non-ischemic cardiomyopathy (NICM) compared to the healthy donors. Further analysis showed that nuclear factor kappa B (NF-κB) bound to the KLF6 promoter and mediated KLF6 trans-activation by pro-fibrogenic stimuli. KLF6 knockdown attenuated whereas KLF6 over-expression enhanced TGF-ß induced fibroblast-myofibroblast transition in vitro. More importantly, myofibroblast-specific KLF6 depletion ameliorated cardiac fibrosis and rescued heart function in mice subjected to the TAC procedure or chronic Ang II infusion. SIGNIFICANCE: In conclusion, our data support a role for KLF6 in cardiac fibrosis.

The chromatin remodeling protein BRG1 contributes to liver ischemia-reperfusion injury by regulating NOXA expression.

AIMS: Hepatic ischemia-reperfusion injury (IRI) is a common complication secondary to liver transplantation. Extensive death of hepatocytes, typically in the form of apoptosis, is observed in and contributes to IRI. In the present study we investigated the role of BRG1 (encoded by Smarca4), a chromatin remodeling protein, in the pathogenesis of liver IRI focusing on the transcriptional mechanism and translational potential. METHODS: Smarca4f/f mice were crossed to Alb-Cre mice to generate hepatocytes-specific BRG1 knockout mice (CKO). Alterations in cellular transcriptome were evaluated by RNA-seq. RESULTS: BRG1 expression was up-regulated in liver tissues of mice subjected to I/R and in hepatocytes exposed to hypoxia-reoxygenation (H/R). Compared to wild type (WT) littermates, the BRG1 CKO mice displayed significant amelioration of liver injury following ischemia-reperfusion as evidenced by decreased ALT/AST levels and cell apoptosis. Primary hepatocytes isolated from the CKO mice were protected from H/R-induced apoptosis compared to those from the WT mice. RNA-seq analysis revealed phorbol-12-myristate-13-acetate-induced protein 1 (PMAIP1, also known as NOXA) as a novel target for BRG1. Consistently, NOXA knockdown attenuated liver IRI in mice. More importantly, administration of a small-molecule BRG1 inhibitor (PFI-3) protected the mice from liver IRI. CONCLUSIONS: Our data uncover a pivotal role for BRG1 in liver IRI and suggest that targeting BRG1 with small-molecule inhibitors can be considered as a reasonable therapeutic strategy.

Pupil engineering for extended depth-of-field imaging in a fluorescence miniscope.

Significance: Fluorescence head-mounted microscopes, i.e., miniscopes, have emerged as powerful tools to analyze in-vivo neural populations but exhibit a limited depth-of-field (DoF) due to the use of high numerical aperture (NA) gradient refractive index (GRIN) objective lenses. Aim: We present extended depth-of-field (EDoF) miniscope, which integrates an optimized thin and lightweight binary diffractive optical element (DOE) onto the GRIN lens of a miniscope to extend the DoF by 2.8× between twin foci in fixed scattering samples. Approach: We use a genetic algorithm that considers the GRIN lens' aberration and intensity loss from scattering in a Fourier optics-forward model to optimize a DOE and manufacture the DOE through single-step photolithography. We integrate the DOE into EDoF-Miniscope with a lateral accuracy of 70 µm to produce high-contrast signals without compromising the speed, spatial resolution, size, or weight. Results: We characterize the performance of EDoF-Miniscope across 5- and 10-µm fluorescent beads embedded in scattering phantoms and demonstrate that EDoF-Miniscope facilitates deeper interrogations of neuronal populations in a 100-µm-thick mouse brain sample and vessels in a whole mouse brain sample. Conclusions: Built from off-the-shelf components and augmented by a customizable DOE, we expect that this low-cost EDoF-Miniscope may find utility in a wide range of neural recording applications.

Targetable Brg1-CXCL14 axis contributes to alcoholic liver injury by driving neutrophil trafficking.

Alcoholic liver disease (ALD) accounts for a large fraction of patients with cirrhosis and hepatocellular carcinoma. In the present study we investigated the involvement of Brahma-related gene 1 (Brg1) in ALD pathogenesis and implication in ALD intervention. We report that Brg1 expression was elevated in mouse models of ALD, in hepatocyte exposed to alcohol, and in human ALD specimens. Manipulation of Brg1 expression in hepatocytes influenced the development of ALD in mice. Flow cytometry showed that Brg1 deficiency specifically attenuated hepatic infiltration of Ly6G+ neutrophils in the ALD mice. RNA-seq identified C-X-C motif chemokine ligand 14 (CXCL14) as a potential target for Brg1. CXCL14 knockdown alleviated whereas CXCL14 over-expression enhanced ALD pathogenesis in mice. Importantly, pharmaceutical inhibition of Brg1 with a small-molecule compound PFI-3 or administration of an antagonist to the CXCL14 receptor ameliorated ALD pathogenesis in mice. Finally, a positive correlation between Brg1 expression, CXCL14 expression, and neutrophil infiltration was detected in ALD patients. In conclusion, our data provide proof-of-concept for targeting the Brg1-CXCL14 axis in ALD intervention.

3D Chemical Imaging by Fluorescence-detected Mid-Infrared Photothermal Fourier Light Field Microscopy.

Three-dimensional molecular imaging of living organisms and cells plays a significant role in modern biology. Yet, current volumetric imaging modalities are largely fluorescence-based and thus lack chemical content information. Mid-infrared photothermal microscopy as a chemical imaging technology provides infrared spectroscopic information at submicrometer spatial resolution. Here, by harnessing thermosensitive fluorescent dyes to sense the mid-infrared photothermal effect, we demonstrate 3D fluorescence-detected mid-infrared photothermal Fourier light field (FMIP-FLF) microscopy at the speed of 8 volumes per second and submicron spatial resolution. Protein contents in bacteria and lipid droplets in living pancreatic cancer cells are visualized. Altered lipid metabolism in drug-resistant pancreatic cancer cells is observed with the FMIP-FLF microscope.

Robust single-shot 3D fluorescence imaging in scattering media with a simulator-trained neural network.

Imaging through scattering is a pervasive and difficult problem in many biological applications. The high background and the exponentially attenuated target signals due to scattering fundamentally limits the imaging depth of fluorescence microscopy. Light-field systems are favorable for high-speed volumetric imaging, but the 2D-to-3D reconstruction is fundamentally ill-posed, and scattering exacerbates the condition of the inverse problem. Here, we develop a scattering simulator that models low-contrast target signals buried in heterogeneous strong background. We then train a deep neural network solely on synthetic data to descatter and reconstruct a 3D volume from a single-shot light-field measurement with low signal-to-background ratio (SBR). We apply this network to our previously developed Computational Miniature Mesoscope and demonstrate the robustness of our deep learning algorithm on scattering phantoms with different scattering conditions. The network can robustly reconstruct emitters in 3D with a 2D measurement of SBR as low as 1.05 and as deep as a scattering length. We analyze fundamental tradeoffs based on network design factors and out-of-distribution data that affect the deep learning model's generalizability to real experimental data. Broadly, we believe that our simulator-based deep learning approach can be applied to a wide range of imaging through scattering techniques where experimental paired training data is lacking.

Deep-learning-augmented computational miniature mesoscope.

Fluorescence microscopy is essential to study biological structures and dynamics. However, existing systems suffer from a trade-off between field of view (FOV), resolution, and system complexity, and thus cannot fulfill the emerging need for miniaturized platforms providing micron-scale resolution across centimeter-scale FOVs. To overcome this challenge, we developed a computational miniature mesoscope (CM2) that exploits a computational imaging strategy to enable single-shot, 3D high-resolution imaging across a wide FOV in a miniaturized platform. Here, we present CM2 V2, which significantly advances both the hardware and computation. We complement the 3 × 3 microlens array with a hybrid emission filter that improves the imaging contrast by 5×, and design a 3D-printed free-form collimator for the LED illuminator that improves the excitation efficiency by 3×. To enable high-resolution reconstruction across a large volume, we develop an accurate and efficient 3D linear shift-variant (LSV) model to characterize spatially varying aberrations. We then train a multimodule deep learning model called CM2Net, using only the 3D-LSV simulator. We quantify the detection performance and localization accuracy of CM2Net to reconstruct fluorescent emitters under different conditions in simulation. We then show that CM2Net generalizes well to experiments and achieves accurate 3D reconstruction across a ~7-mm FOV and 800-µm depth, and provides ~6-µm lateral and ~25-µm axial resolution. This provides an ~8× better axial resolution and ~1400× faster speed compared to the previous model-based algorithm. We anticipate this simple, low-cost computational miniature imaging system will be useful for many large-scale 3D fluorescence imaging applications.

Upregulation of Neogenin-1 by a CREB1-BAF47 Complex in Vascular Endothelial Cells is Implicated in Atherogenesis.

Atherosclerosis is generally considered a human pathology of chronic inflammation, to which endothelial dysfunction plays an important role. Here we investigated the role of neogenin 1 (Neo-1) in oxidized low-density lipoprotein (oxLDL) induced endothelial dysfunction focusing on its transcriptional regulation. We report that Neo-1 expression was upregulated by oxLDL in both immortalized vascular endothelial cells and primary aortic endothelial cells. Neo-1 knockdown attenuated whereas Neo-1 over-expression enhanced oxLDL-induced leukocyte adhesion to endothelial cells. Neo-1 regulated endothelial-leukocyte interaction by modulating nuclear factor kappa B (NF-κB) activity to alter the expression of adhesion molecules. Neo-1 blockade with a blocking antibody ameliorated atherogenesis in Apoe -/- mice fed a Western diet. Ingenuity pathway analysis combined with validation assays confirmed that cAMP response element binding protein 1 (CREB1) and Brg1-associated factor 47 (BAF47) mediated oxLDL induced Neo-1 upregulation. CREB1 interacted with BAF47 and recruited BAF47 to the proximal Neo-1 promoter leading to Neo-1 trans-activation. In conclusion, our data delineate a novel transcriptional mechanism underlying Neo-1 activation in vascular endothelial cells that might contribute to endothelial dysfunction and atherosclerosis.

Development of a method for the isolation and culture of astrocytes from the canine cerebral cortex.

BACKGROUND: Astrocytes are considered key players in neuroimmunopathological processes, and they play a certain role in neuroinflammation. Rodent primary astrocyte cultures are commonly used in the study of human neuroinflammation. However, gene sequence homologies are closer between humans and dogs than between humans and rodents. NEW METHOD: We established protocols to isolate astrocytes from the canine forebrain. Cerebral hemispheres of 3-4-week-old dogs were used. The isolation procedure included the use of the Neural Tissue Dissociation Kit P, demyelination by the magnetic bead method, and separation and preparation by differential adhesion. RESULTS: We found a 96% astrocyte purification rate after isolation by differential adhesion. Purified canine astrocytes increased the secretion of interleukin-1ß, interleukin-6, and tumor necrosis factor-alpha, and increased the expression of glial fibrillary acidic protein after lipopolysaccharide stimulation. We sequenced the transcriptome of the purified canine astrocytes and analyzed the differentially expressed genes among the rodent, human, and canine astrocytes. Transcriptome profiling and gene ontology analysis of the genes co-expressed in humans and canines indicate that human and canine astrocytes may be different from their rodent counterparts in terms of mediated interactions with metals. COMPARED WITH THE EXISTING METHODS: The cells prepared by our method allow for the rapid separation of astrocytes with a relatively small resource scheme. The method also retains the cell phenotype and has an in vitro culture lifetime of approximately 2-3 months. CONCLUSION: We established a method for preparing canine astrocytes with high purity, which can be used to study the biological function of astrocytes in vitro.

Epigenetic Repression of Chloride Channel Accessory 2 Transcription in Cardiac Fibroblast: Implication in Cardiac Fibrosis.

Cardiac fibrosis is a key pathophysiological process that contributes to heart failure. Cardiac resident fibroblasts, exposed to various stimuli, are able to trans-differentiate into myofibroblasts and mediate the pro-fibrogenic response in the heart. The present study aims to investigate the mechanism whereby transcription of chloride channel accessory 2 (Clca2) is regulated in cardiac fibroblast and its potential implication in fibroblast-myofibroblast transition (FMyT). We report that Clca2 expression was down-regulated in activated cardiac fibroblasts (myofibroblasts) compared to quiescent cardiac fibroblasts in two different animal models of cardiac fibrosis. Clca2 expression was also down-regulated by TGF-ß, a potent inducer of FMyT. TGF-ß repressed Clca2 expression at the transcriptional level likely via the E-box element between -516 and -224 of the Clca2 promoter. Further analysis revealed that Twist1 bound directly to the E-box element whereas Twist1 depletion abrogated TGF-ß induced Clca2 trans-repression. Twist1-mediated Clca2 repression was accompanied by erasure of histone H3/H4 acetylation from the Clca2 promoter. Mechanistically Twist1 interacted with HDAC1 and recruited HDAC1 to the Clca2 promoter to repress Clca2 transcription. Finally, it was observed that Clca2 over-expression attenuated whereas Clca2 knockdown enhanced FMyT. In conclusion, our data demonstrate that a Twist1-HDAC1 complex represses Clca2 transcription in cardiac fibroblasts, which may contribute to FMyT and cardiac fibrosis.

Dual Regulation of Tank Binding Kinase 1 by BRG1 in Hepatocytes Contributes to Reactive Oxygen Species Production.

Excessive accumulation of reactive oxygen species (ROS) is considered a major culprit for the pathogenesis of non-alcoholic fatty liver disease (NAFLD). We have previously shown that deletion of Brahma related gene 1 (BRG1) mitigated NAFLD in mice in part by attenuating ROS production in hepatocyte. Here we report that BRG1 deletion led to simultaneous down-regulation in expression and phosphorylation of tank binding kinase 1 (TBK1) in vivo and in vitro. On the one hand, BRG1 interacted with AP-1 to bind to the TBK1 promoter and directly activated TBK1 transcription in hepatocytes. On the other hand, BRG1 interacted with Sp1 to activate the transcription of c-SRC, a tyrosine kinase essential for TBK1 phosphorylation. Over-expression of c-SRC and TBK1 corrected the deficiency in ROS production in BRG1-null hepatocytes whereas depletion of TBK1 or c-SRC attenuated ROS production. In conclusion, our data suggest that dual regulation of TBK1 activity, at the transcription level and the post-transcriptional level, by BRG1 may constitute an important mechanism underlying excessive ROS production in hepatocytes.

Single-cell cytometry via multiplexed fluorescence prediction by label-free reflectance microscopy.

Traditional imaging cytometry uses fluorescence markers to identify specific structures but is limited in throughput by the labeling process. We develop a label-free technique that alleviates the physical staining and provides multiplexed readouts via a deep learning-augmented digital labeling method. We leverage the rich structural information and superior sensitivity in reflectance microscopy and show that digital labeling predicts accurate subcellular features after training on immunofluorescence images. We demonstrate up to three times improvement in the prediction accuracy over the state of the art. Beyond fluorescence prediction, we demonstrate that single cell-level structural phenotypes of cell cycles are correctly reproduced by the digital multiplexed images, including Golgi twins, Golgi haze during mitosis, and DNA synthesis. We further show that the multiplexed readouts enable accurate multiparametric single-cell profiling across a large cell population. Our method can markedly improve the throughput for imaging cytometry toward applications for phenotyping, pathology, and high-content screening.