More Like this

1. Using a Human Liver Tissue Equivalent (hLTE) Platform to Define the Functional Impact of Liver-Directed AAV Gene Therapy

   Ramamurthy, R ; Zheng, WT ; George, S ; Wan, M ; Zhou, Y ; Lu, B ; Bishop, C ; Atala, A ; Porada, C ; Almeida-Porada, G ( December 2021 , ASH)

   2938 Using a Human Liver Tissue Equivalent (hLTE) Platform to Define the Functional Impact of Liver-Directed AAV Gene Therapy 63rd ASH Annual Meeting and Exposition, December 11-14, 2021, Georgia World Congress Center, Atlanta, GA Program: Oral and Poster Abstracts Session: 801. Gene Therapies: Poster II Hematology Disease Topics & Pathways: Bleeding and Clotting, Biological, Translational Research, Hemophilia, Genetic Disorders, Clinically Relevant, Diseases, Gene Therapy, Therapies Sunday, December 12, 2021, 6:00 PM-8:00 PM Ritu M Ramamurthy1*, Wen Ting Zheng2*, Sunil George, PhD1*, Meimei Wan1*, Yu Zhou, PhD1*, Baisong Lu, PhD1*, Colin E Bishop, PhD1*, Anthony Atala, M.D.1*, Christopher D Porada, PhD1* and M. Graca Almeida-Porada, MD3 1Fetal Research and Therapy Program, Wake Forest Institute for Regenerative Medicine, Winston Salem, NC 2Massachusetts Institute of Technology, Cambridge, MA 3Fetal Research and Therapy Program, Wake Forest Institute For Regenerative Medicine, Winston-Salem, NC Clinical trials employing AAV vectors for hemophilia A have been hindered by unanticipated immunological and/or inflammatory responses in some of the patients. Also, these trials have often yielded lower levels of transgene expression than were expected based upon preclinical studies, highlighting the poor correlation between the transduction efficiency observed in traditional 2D cultures of primary cells in vitro, and that observed inmore »those same cell types in vivo. It has been also recognized that there are marked species-specific differences in AAV-vector tropism, raising the critical question of the accuracy with which various animal models will likely predict tropism/vector transduction efficiency, and eventual treatment success in humans. Human liver tissue equivalents (hLTEs) are comprised of major cell types in the liver in physiologically relevant frequencies and possess the ability to recapitulate the biology and function of native human liver. Here, we hypothesize that hLTEs can be used as a better model to predict the efficacy and safety of AAV gene therapy in humans. We fabricated hLTEs using 75% hepatocytes, 10% stellate cells, 10% Kupffer cells, and 5% liver sinusoid-derived endothelial cells in 96-well Elplasia plates with 79 microwells per well. hLTEs were transduced at an MOI of 105vg/cell, on the day of fabrication, with the clinically relevant serotypes AAV5 (hLTE-5) or AAV3b (hLTE-3b), both encoding a GFP reporter. After 4 days of self-aggregation, live/dead assay was performed to confirm viability. Non-transduced hLTEs served as negative controls (hLTE(-)), and hLTEs exposed to 20 mM acetaminophen were used as positive controls for liver inflammation/damage. Incucyte® Live-Cell Imaging system was used to track the aggregation and GFP expression of hLTEs. Over the course of the next 5 days, media was collected to determine hepatic functionality, RNA was isolated to assess dysregulation of genes involved in inflammation and fibrosis, DNA was isolated to determine whether AAV vectors integrate into the genome of human hepatocytes and, if so, to define the frequency at which this occurs and the genomic loci of integration, and hLTEs were fixed and processed at appropriate times for histological analyses and transmission electron microscopy (TEM). TEM analysis revealed that all groups exhibited microvilli and bile-canaliculus-like structures, demonstrating the formation of a rudimentary biliary system and, more importantly, proving that hLTEs resemble native liver structure. Incucyte® imaging showed that AAV5 and AAV3b transduction impaired formation of hLTEs (57.57 ± 2.42 and 24.57 ± 4.01 spheroids/well, respectively) in comparison with hLTE(-) (74.86 ± 3.8 spheroids/well). Quantification of GFP expression demonstrated that AAV5 yielded the most efficient transduction of hLTEs (fold change in GFP expression compared to control: 2.73 ± 0.09 and 1.19 ± 0.03 for hLTE-5 and hLTE-3b, respectively). Chromogenic assays showed decreased urea production in cell culture supernatants of AAV transduced groups compared to the non-transduced hLTEs on days 6 and 10 of culture, demonstrating decreased hepatocyte functionality. However, ALT and AST levels were similar in all groups. On day 10, hLTEs were either used for RNA isolation or fixed in 4% PFA and processed for histology. Masson’s Trichrome and Alcian Blue/Sirius Red staining was performed to detect fibrosis, which was then quantified using ImageJ. These analyses showed no significant increase in fibrosis in either hLTE-5 or hLTE-3b compared to hLTE(-). Nevertheless, RT2 PCR Array for Human Fibrosis detected dysregulation of several genes involved in fibrosis/inflammation in both hLTE-5 and hLTE-3b (16/84 and 26/84, respectively). In conclusion, data collected thus far show successful recapitulation of native liver biology and demonstrate that AAV5 transduces hLTEs more efficiently than AAV3b. However, impaired self-aggregation and decreased hepatocyte functionality was observed in both AAV-transduced groups. Studies to address the incidence and location(s) of AAV integration are ongoing. We have thus shown that the hLTE system can provide critical new knowledge regarding the efficacy and safety of AAV gene therapy in the human liver. Disclosures: No relevant conflicts of interest to declare.« less
2. The Temple University Digital Pathology Corpus: The Breast Tissue Subset

   https://doi.org/10.1109/SPMB52430.2021.9672275  

   Wevodau, Z. ; Doshna, B. ; Jhala, N. ; Akhtar, I. ; Obeid, I. ; Picone, J. ( December 2021 , Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB))

   Obeid, I. (Ed.)

   The Neural Engineering Data Consortium (NEDC) is developing the Temple University Digital Pathology Corpus (TUDP), an open source database of high-resolution images from scanned pathology samples [1], as part of its National Science Foundation-funded Major Research Instrumentation grant titled “MRI: High Performance Digital Pathology Using Big Data and Machine Learning” [2]. The long-term goal of this project is to release one million images. We have currently scanned over 100,000 images and are in the process of annotating breast tissue data for our first official corpus release, v1.0.0. This release contains 3,505 annotated images of breast tissue including 74 patients with cancerous diagnoses (out of a total of 296 patients). In this poster, we will present an analysis of this corpus and discuss the challenges we have faced in efficiently producing high quality annotations of breast tissue. It is well known that state of the art algorithms in machine learning require vast amounts of data. Fields such as speech recognition [3], image recognition [4] and text processing [5] are able to deliver impressive performance with complex deep learning models because they have developed large corpora to support training of extremely high-dimensional models (e.g., billions of parameters). Other fields that do notmore »have access to such data resources must rely on techniques in which existing models can be adapted to new datasets [6]. A preliminary version of this breast corpus release was tested in a pilot study using a baseline machine learning system, ResNet18 [7], that leverages several open-source Python tools. The pilot corpus was divided into three sets: train, development, and evaluation. Portions of these slides were manually annotated [1] using the nine labels in Table 1 [8] to identify five to ten examples of pathological features on each slide. Not every pathological feature is annotated, meaning excluded areas can include focuses particular to these labels that are not used for training. A summary of the number of patches within each label is given in Table 2. To maintain a balanced training set, 1,000 patches of each label were used to train the machine learning model. Throughout all sets, only annotated patches were involved in model development. The performance of this model in identifying all the patches in the evaluation set can be seen in the confusion matrix of classification accuracy in Table 3. The highest performing labels were background, 97% correct identification, and artifact, 76% correct identification. A correlation exists between labels with more than 6,000 development patches and accurate performance on the evaluation set. Additionally, these results indicated a need to further refine the annotation of invasive ductal carcinoma (“indc”), inflammation (“infl”), nonneoplastic features (“nneo”), normal (“norm”) and suspicious (“susp”). This pilot experiment motivated changes to the corpus that will be discussed in detail in this poster presentation. To increase the accuracy of the machine learning model, we modified how we addressed underperforming labels. One common source of error arose with how non-background labels were converted into patches. Large areas of background within other labels were isolated within a patch resulting in connective tissue misrepresenting a non-background label. In response, the annotation overlay margins were revised to exclude benign connective tissue in non-background labels. Corresponding patient reports and supporting immunohistochemical stains further guided annotation reviews. The microscopic diagnoses given by the primary pathologist in these reports detail the pathological findings within each tissue site, but not within each specific slide. The microscopic diagnoses informed revisions specifically targeting annotated regions classified as cancerous, ensuring that the labels “indc” and “dcis” were used only in situations where a micropathologist diagnosed it as such. Further differentiation of cancerous and precancerous labels, as well as the location of their focus on a slide, could be accomplished with supplemental immunohistochemically (IHC) stained slides. When distinguishing whether a focus is a nonneoplastic feature versus a cancerous growth, pathologists employ antigen targeting stains to the tissue in question to confirm the diagnosis. For example, a nonneoplastic feature of usual ductal hyperplasia will display diffuse staining for cytokeratin 5 (CK5) and no diffuse staining for estrogen receptor (ER), while a cancerous growth of ductal carcinoma in situ will have negative or focally positive staining for CK5 and diffuse staining for ER [9]. Many tissue samples contain cancerous and non-cancerous features with morphological overlaps that cause variability between annotators. The informative fields IHC slides provide could play an integral role in machine model pathology diagnostics. Following the revisions made on all the annotations, a second experiment was run using ResNet18. Compared to the pilot study, an increase of model prediction accuracy was seen for the labels indc, infl, nneo, norm, and null. This increase is correlated with an increase in annotated area and annotation accuracy. Model performance in identifying the suspicious label decreased by 25% due to the decrease of 57% in the total annotated area described by this label. A summary of the model performance is given in Table 4, which shows the new prediction accuracy and the absolute change in error rate compared to Table 3. The breast tissue subset we are developing includes 3,505 annotated breast pathology slides from 296 patients. The average size of a scanned SVS file is 363 MB. The annotations are stored in an XML format. A CSV version of the annotation file is also available which provides a flat, or simple, annotation that is easy for machine learning researchers to access and interface to their systems. Each patient is identified by an anonymized medical reference number. Within each patient’s directory, one or more sessions are identified, also anonymized to the first of the month in which the sample was taken. These sessions are broken into groupings of tissue taken on that date (in this case, breast tissue). A deidentified patient report stored as a flat text file is also available. Within these slides there are a total of 16,971 total annotated regions with an average of 4.84 annotations per slide. Among those annotations, 8,035 are non-cancerous (normal, background, null, and artifact,) 6,222 are carcinogenic signs (inflammation, nonneoplastic and suspicious,) and 2,714 are cancerous labels (ductal carcinoma in situ and invasive ductal carcinoma in situ.) The individual patients are split up into three sets: train, development, and evaluation. Of the 74 cancerous patients, 20 were allotted for both the development and evaluation sets, while the remain 34 were allotted for train. The remaining 222 patients were split up to preserve the overall distribution of labels within the corpus. This was done in hope of creating control sets for comparable studies. Overall, the development and evaluation sets each have 80 patients, while the training set has 136 patients. In a related component of this project, slides from the Fox Chase Cancer Center (FCCC) Biosample Repository (https://www.foxchase.org/research/facilities/genetic-research-facilities/biosample-repository -facility) are being digitized in addition to slides provided by Temple University Hospital. This data includes 18 different types of tissue including approximately 38.5% urinary tissue and 16.5% gynecological tissue. These slides and the metadata provided with them are already anonymized and include diagnoses in a spreadsheet with sample and patient ID. We plan to release over 13,000 unannotated slides from the FCCC Corpus simultaneously with v1.0.0 of TUDP. Details of this release will also be discussed in this poster. Few digitally annotated databases of pathology samples like TUDP exist due to the extensive data collection and processing required. The breast corpus subset should be released by November 2021. By December 2021 we should also release the unannotated FCCC data. We are currently annotating urinary tract data as well. We expect to release about 5,600 processed TUH slides in this subset. We have an additional 53,000 unprocessed TUH slides digitized. Corpora of this size will stimulate the development of a new generation of deep learning technology. In clinical settings where resources are limited, an assistive diagnoses model could support pathologists’ workload and even help prioritize suspected cancerous cases. ACKNOWLEDGMENTS This material is supported by the National Science Foundation under grants nos. CNS-1726188 and 1925494. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation. REFERENCES [1] N. Shawki et al., “The Temple University Digital Pathology Corpus,” in Signal Processing in Medicine and Biology: Emerging Trends in Research and Applications, 1st ed., I. Obeid, I. Selesnick, and J. Picone, Eds. New York City, New York, USA: Springer, 2020, pp. 67 104. https://www.springer.com/gp/book/9783030368432. [2] J. Picone, T. Farkas, I. Obeid, and Y. Persidsky, “MRI: High Performance Digital Pathology Using Big Data and Machine Learning.” Major Research Instrumentation (MRI), Division of Computer and Network Systems, Award No. 1726188, January 1, 2018 – December 31, 2021. https://www. isip.piconepress.com/projects/nsf_dpath/. [3] A. Gulati et al., “Conformer: Convolution-augmented Transformer for Speech Recognition,” in Proceedings of the Annual Conference of the International Speech Communication Association (INTERSPEECH), 2020, pp. 5036-5040. https://doi.org/10.21437/interspeech.2020-3015. [4] C.-J. Wu et al., “Machine Learning at Facebook: Understanding Inference at the Edge,” in Proceedings of the IEEE International Symposium on High Performance Computer Architecture (HPCA), 2019, pp. 331–344. https://ieeexplore.ieee.org/document/8675201. [5] I. Caswell and B. Liang, “Recent Advances in Google Translate,” Google AI Blog: The latest from Google Research, 2020. [Online]. Available: https://ai.googleblog.com/2020/06/recent-advances-in-google-translate.html. [Accessed: 01-Aug-2021]. [6] V. Khalkhali, N. Shawki, V. Shah, M. Golmohammadi, I. Obeid, and J. Picone, “Low Latency Real-Time Seizure Detection Using Transfer Deep Learning,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2021, pp. 1 7. https://www.isip. piconepress.com/publications/conference_proceedings/2021/ieee_spmb/eeg_transfer_learning/. [7] J. Picone, T. Farkas, I. Obeid, and Y. Persidsky, “MRI: High Performance Digital Pathology Using Big Data and Machine Learning,” Philadelphia, Pennsylvania, USA, 2020. https://www.isip.piconepress.com/publications/reports/2020/nsf/mri_dpath/. [8] I. Hunt, S. Husain, J. Simons, I. Obeid, and J. Picone, “Recent Advances in the Temple University Digital Pathology Corpus,” in Proceedings of the IEEE Signal Processing in Medicine and Biology Symposium (SPMB), 2019, pp. 1–4. https://ieeexplore.ieee.org/document/9037859. [9] A. P. Martinez, C. Cohen, K. Z. Hanley, and X. (Bill) Li, “Estrogen Receptor and Cytokeratin 5 Are Reliable Markers to Separate Usual Ductal Hyperplasia From Atypical Ductal Hyperplasia and Low-Grade Ductal Carcinoma In Situ,” Arch. Pathol. Lab. Med., vol. 140, no. 7, pp. 686–689, Apr. 2016. https://doi.org/10.5858/arpa.2015-0238-OA.« less
3. Microfluidic and Microscale Assays to Examine Regenerative Strategies in the Neuro Retina

   https://doi.org/10.3390/mi11121089  

   Vazquez, Maribel ( December 2020 , Micromachines)

   Bioengineering systems have transformed scientific knowledge of cellular behaviors in the nervous system (NS) and pioneered innovative, regenerative therapies to treat adult neural disorders. Microscale systems with characteristic lengths of single to hundreds of microns have examined the development and specialized behaviors of numerous neuromuscular and neurosensory components of the NS. The visual system is comprised of the eye sensory organ and its connecting pathways to the visual cortex. Significant vision loss arises from dysfunction in the retina, the photosensitive tissue at the eye posterior that achieves phototransduction of light to form images in the brain. Retinal regenerative medicine has embraced microfluidic technologies to manipulate stem-like cells for transplantation therapies, where de/differentiated cells are introduced within adult tissue to replace dysfunctional or damaged neurons. Microfluidic systems coupled with stem cell biology and biomaterials have produced exciting advances to restore vision. The current article reviews contemporary microfluidic technologies and microfluidics-enhanced bioassays, developed to interrogate cellular responses to adult retinal cues. The focus is on applications of microfluidics and microscale assays within mammalian sensory retina, or neuro retina, comprised of five types of retinal neurons (photoreceptors, horizontal, bipolar, amacrine, retinal ganglion) and one neuroglia (Müller), but excludes the non-sensory, retinal pigmented epithelium.
4. Gene-free methodology for cell fate dynamics during development

   https://doi.org/10.7554/eLife.30743  

   Corson, Francis ; Siggia, Eric D ( December 2017 , eLife)

   At first, embryos are made up of identical cells. Then, as the embryo develops, these cells specialize into different types, such as heart and brain cells. Chemical signals sent and received by the cells are key to forming the right type of cell at the right time and place. The cellular machinery that produces and interprets these signals is exceedingly complex and difficult to understand. In the 1950s, Conrad Waddington presented an alternative way of thinking about how an unspecialized cell progresses to one of many different fates. He suggested visualizing the developing cell as a ball rolling along a hilly landscape. As the ball travels, obstacles in its way guide it along particular paths. Eventually the ball comes to rest in a valley, with each valley in the landscape representing a different cell fate. Although this “landscape model” is an appealing metaphor for how signaling events guide cell specialization, it was not clear whether it could be put to productive use. The egg-laying organ in the worm species Caenorhabditis elegans is called the vulva, and is often studied by researchers who want to learn more about how organs develop. The vulva develops from a small number of identical cellsmore »that adopt one of three possible cell fates. Two chemical signals, called epidermal growth factor (EGF) and Notch, control this specialization process. Corson and Siggia have now constructed a simple landscape model that can reproduce the normal arrangement of cell types in the vulva. When adjusted to describe the effect of genetic mutations that affect either EGF or Notch, the model could predict the outcome of mutations that affect both signals at once. The twists and turns of cell paths in the landscape could also account for several non-intuitive cell fate outcomes that had been assumed to result from subtle regulation of EGF and Notch signals. Landscape models should be easy to apply to other developing tissues and organs. By providing an intuitive picture of how signals shape cellular decisions, the models could help researchers to learn how to control cell and tissue development. This could lead to new treatments to repair or replace failing organs, making regenerative medicine a reality.« less
5. Scan-IT: Domain segmentation of spatial transcriptomics images by graph neural network

   Cang, Zixuan ; Ning, Xinyi ; Nie, Annika ; Xu, Min ; Zhang, Jing ( November 2021 , British Machine Vision Conference)

   Complex biological tissues consist of numerous cells in a highly coordinated manner and carry out various biological functions. Therefore, segmenting a tissue into spatial and functional domains is critically important for understanding and controlling the biological functions. The emerging spatial transcriptomic technologies allow simultaneous measurements of thousands of genes with precise spatial information, providing an unprecedented opportunity for dissecting biological tissues. However, how to utilize such noisy, sparse, and high dimensional data for tissue segmentation remains a major challenge. Here, we develop a deep learning-based method, named SCAN-IT by transforming the spatial domain identification problem into an image segmentation problem, with cells mimicking pixels and expression values of genes within a cell representing the color channels. Specifically, SCAN-IT relies on geometric modeling, graph neural networks, and an informatics approach, DeepGraphInfomax. We demonstrate that SCAN-IT can handle datasets from a wide range of spatial transcriptomics techniques, including the ones with high spatial resolution but low gene coverage as well as those with low spatial resolution but high gene coverage. We show that SCAN-IT outperforms state-of-the-art methods using a benchmark dataset with ground truth domain annotations.