Biology subjects

Kim, E.

Publications and source records attributed to Kim, E..

12 recordsLinked to original sources

bioRxiv2018

The impact of tumor stromal architecture on therapy response and clinical progression

-- Advances in molecular oncology research culminated in the development of targeted therapies that act on defined molecular targets either on tumor cells directly (such as inhibitors of oncogenic kinases), or indirectly by targeting the tumor microenvironment (such as anti-angiogenesis drugs). These therapies can induce strong clinical responses, when properly matched to patients. Unfortunately, most targeted therapies ultimately fail as tumors evolve resistance. Tumors consist not only of neoplastic cells, but also of stroma, whereby \"stroma\" is the umbrella term for non-tumor cells and extracellular matrix (ECM) within the tumor microenvironment, possibly excluding immune cells1. We know that tumor stroma is an important player in the development of resistance. We also know that stromal architecture is spatially complex, differs from patient to patient and changes with therapy. However, to this date we do not understand the link between spatial and temporal changes in stromal architecture and response of tumors to therapy, in space and time. In this project we sought to address this gap of knowledge using a combination of mathematical and statistical modeling, experimental in vivo studies, and analysis of clinical samples in therapies that target tumor cells directly (in lung and breast cancers) and indirectly (in kidney cancer). This knowledge will inform therapy choices and offer new angles for therapeutic interventions. Our main question is: how does spatial architecture of stroma impact the emergence or evolution of resistance to targeted therapies, and how can we use this knowledge clinically?

cancer biology↗

bioRxiv2018

Common ancestry of heterodimerizing TALE homeobox transcription factors across Metazoa and Archaeplastida

Homeobox transcription factors (TFs) in the TALE superclass are deeply embedded in the gene regulatory networks that orchestrate embryogenesis. Knotted-like homeobox (KNOX) TFs, homologous to animal MEIS, have been found to drive the haploid-to-diploid transition in both unicellular green algae and land plants via heterodimerization with other TALE superclass TFs, representing remarkable functional conservation of a developmental TF across lineages that diverged one billion years ago. To delineate the ancestry of TALE-TALE heterodimerization, we analyzed TALE endowment in the algal radiations of Archaeplastida, ancestral to land plants. Homeodomain phylogeny and bioinformatics analysis partitioned TALEs into two broad groups, KNOX and non-KNOX. Each group shares previously defined heterodimerization domains, plant KNOX-homology in the KNOX group and animal PBC-homology in the non-KNOX group, indicating their deep ancestry. Protein-protein interaction experiments showed that the TALEs in the two groups all participated in heterodimerization. These results indicate that the TF dyads consisting of KNOX/MEIS and PBC-containing TALEs must have evolved early in eukaryotic evolution, a likely function being to accurately execute the haploid-to-diploid transitions during sexual development.\n\nAuthor summaryComplex multicellularity requires elaborate developmental mechanisms, often based on the versatility of heterodimeric transcription factor (TF) interactions. Highly conserved TALE-superclass homeobox TF networks in major eukaryotic lineages suggest deep ancestry of developmental mechanisms. Our results support the hypothesis that in early eukaryotes, the TALE heterodimeric configuration provided transcription-on switches via dimerization-dependent subcellular localization, ensuring execution of the haploid-to-diploid transition only when the gamete fusion is correctly executed between appropriate partner gametes, a system that then diversified in the several lineages that engage in complex multicellular organization.

evolutionary biology↗

bioRxiv2018

Signal sequences in the genome of Mononegavirales regulate the generation of copy-back defective viral genomes

Defective viral genomes of the copy-back type (cbDVGs) are the primary initiators of the antiviral immune response during infection with respiratory syncytial virus (RSV) both in vitro and in vivo. However, the mechanism governing cbDVG generation remains unknown, thereby limiting our ability to manipulate cbDVG content in order to modulate the host response to infection. Here we report a specific genomic signal that mediates the generation of RSV cbDVGs. Using a customized bioinformatics tool, we identified regions in the RSV genome frequently used to generate cbDVGs during infection. We then created a minigenome system to validate the function of one of these sequences and to determine if specific nucleotides were essential for cbDVG generation at that position. Further, we created a recombinant virus that selectively produced a specific cbDVG based on variations introduced in this sequence. The identified sequence was also found as a common site for cbDVG generation during natural RSV infections, and common cbDVGs generated at this sequence were found among samples from various infected patients. These data demonstrate that sequences encoded in the viral genome are critical determinants of the location of cbDVG generation and, therefore, this is not a stochastic process. Most importantly, these findings open the possibility of genetically manipulating cbDVG formation to modulate infection outcome. Author summaryCopy-back defective viral genomes (cbDVGs) regulate infection and pathogenesis of Mononegavirales. cbDVG are believed to arise from random errors that occur during virus replication and the predominant hypothesis is that the viral polymerase is the main driver of cbDVG generation. Here we describe a specific genomic sequence in the RSV genome that is necessary for the generation of a large proportion of the cbDVG population present during infection. We identified specific nucleotides that when modified altered cbDVG generation at this position, and we created a recombinant virus that selectively produced cbDVGs based on mutations in this sequence. These data demonstrate that the generation of RSV cbDVGs is regulated by specific viral sequences and that these sequences can be manipulated to alter the content and quality of cbDVG generated during infection.

microbiology↗

bioRxiv2018

Hierarchical organization of the human cell from a cancer coessentiality network

Genetic interactions mediate the emergence of phenotype from genotype. Systematic survey of genetic interactions in yeast showed that genes operating in the same biological process have highly correlated genetic interaction profiles, and this observation has been exploited to infer gene function in model organisms. Systematic surveys of digenic perturbations in human cells are also highly informative, but are not scalable, even with CRISPR-mediated methods. As an alternative, we developed an indirect method of deriving functional interactions. We show that genes having correlated knockout fitness profiles across diverse, non-isogenic cell lines are analogous to genes having correlated genetic interaction profiles across isogenic query strains, and similarly implies shared biological function. We constructed a network of genes with correlated fitness profiles across 400 CRISPR knockout screens in cancer cell lines into a \"coessentiality network,\" with up to 500-fold enrichment for co-functional gene pairs, enabling strong inference of human gene function. Modules in the network are connected in a layered web that gives insight into the hierarchical organization of the cell.

systems biology↗

bioRxiv2018

Using neural networks to bridge scales in cancer: Mapping signaling pathways to phenotypes

Cancer is an evolving system subject to mutation and selection. Selection is driven by the microenvironment that the cancer cells are growing in and acts on the cell phenotype, which is in turn modulated by intracellular signaling pathways regulated by the cell genotype. Integrating all of these processes requires bridging different biological scales. We present a mathematical model that uses a neural network as a means to connecting these scales. In particular, we consider the mapping from intracellular pathway activity to phenotype under different microenvironmental conditions.

cancer biology↗

bioRxiv2018

Lrfn2-mutant mice display suppressed synaptic plasticity and inhibitory synapse development and abnormal social communication and startle response

SALM1, also known as LRFN2, is a PSD-95-interacting synaptic adhesion molecule implicated in the regulation of NMDA receptor (NMDAR) clustering largely based on in vitro data, although its in vivo functions remain unclear. Here, we found that mice lacking SALM1/LRFN2 (Lrfn2-/- mice) show a normal density of excitatory synapses but altered excitatory synaptic function, including enhanced NMDAR-dependent synaptic transmission but suppressed NMDAR-dependent synaptic plasticity in the hippocampal CA1 region. Unexpectedly, SALM1 expression is detected in both glutamatergic and GABAergic neurons, and Lrfn2-/- CA1 pyramidal neurons show decreases in the density of inhibitory synapses and frequency of spontaneous inhibitory synaptic transmission. Behaviorally, ultrasonic vocalization was suppressed in Lrfn2-/- pups separated from their mothers, and acoustic startle was enhanced, but locomotion, anxiety-like behavior, social interaction, repetitive behaviors, and learning and memory were largely normal in adult Lrfn2-/- mice. These results suggest that SALM1/LRFN2 regulates excitatory synapse function, inhibitory synapse development, and social communication and startle behaviors in mice.\n\nSignificance StatementSynaptic adhesion molecules regulate synapse development and function, which govern neural circuit and brain functions. The SALM/LRFN family of synaptic adhesion proteins consists of five known members whose in vivo functions are largely unknown. Here we characterized mice lacking SALM1/LRFN2 (SALM1 knockout) known to associate with NMDA receptors and found that these mice showed altered NMDA receptor-dependent synaptic transmission and plasticity, as expected, but unexpectedly also exhibited suppressed inhibitory synapse development and synaptic transmission. Behaviorally, SALM1 knockout pups showed suppressed ultrasonic vocalization upon separation from their mothers, and SALM1 knockout adults showed enhanced responses to loud acoustic stimuli. These results suggest that SALM1/LRFN2 regulates excitatory synapse function, inhibitory synapse development, social communication, and acoustic startle behavior.

neuroscience↗

bioRxiv2018

Galleria mellonella as an Insect Model for P. destructans, the Cause of White-Nose Syndrome in Bats

Pseudogymnoascus destructans is the fungal pathogen responsible for White-nose Syndrome (WNS), a disease that has killed millions of bats in North America over the last decade. A major obstacle to research on P. destructans has been the lack of a tractable infection model for monitoring virulence. Here, we establish a high-throughput model of infection using larvae of Galleria mellonella, an invertebrate used to study host-pathogen interactions for a wide range of microbial species. We demonstrate that P. destructans can kill G. mellonella larvae in an inoculum-dependent manner when infected larvae are housed at 13{degrees}C or 18{degrees}C. Larval killing is an active process, as heat-killed P. destructans spores caused significantly decreased levels of larval death compared to live spores. We also show that fungal spores that were germinated prior to inoculation were able to kill larvae 3-4 times faster than non-germinated spores. Lastly, we identified chemical inhibitors of P. destructans and used G. mellonella to evaluate these inhibitors for their ability to reduce virulence. We demonstrate that two chemicals, trifluoperazine and amphotericin B, can effectively block larval killing by P. destructans and thereby establish that this infection model can be used to screen biocontrol agents against this fungal pathogen.

microbiology↗

bioRxiv2017

Targeting the Untargetable: Predicting Pramlintide Resistance Using a Neural Network Based Cellular Automata

De novo resistance is a major issue for the use of targeted anticancer drugs in the clinic. By integrating experimental data we have created a hybrid neural network/agent-based model to simulate the evolution and spread of resistance to the drug Pramlintide in cutaneous squamous cell carcinoma. Our model can eventually be used to predict patient responses to the drug and thus enable clinicians to make decisions regarding personalized, precision treatment regimes for patients.

cancer biology↗

bioRxiv2017

Edge Detection of Cryptic Lamellipodia Assisted by Deep Learning

Cell protrusion plays important roles in cell migration by pushing plasma membrane forward. Cryptic lamellipodia induce the protrusion of submarginal cells in collective cell migration where cells are attached and move together. Although computational image analysis of cell protrusion has been done extensively, the study on protrusion activities of cryptic lamellipodia is limited due to difficulties in image segmentation. This study seeks to aid in the computational analysis of submarginal cell protrusion in collective cell migration by using deep learning to detect the cryptic lamellipodial edges from fluorescence time-lapse movies. Due to the noisy features within overlapping cells, the conventional image analysis algorithms such as Canny edge detector and intensity thresholding are limited. By combining Canny edge detector, Convolutional Neural Networks (CNNs), and local intensity thresholding, we were able to detect cryptic lamellipodial edges of submarginal cells with high accuracy from the fluorescence time-lapse movies of PtK1 cells stained with a plasma membrane marker. We used relatively small effort to prepare the training set to train the CNN to detect the cryptical lamellipodial edges in fluorescence time-lapse movies. This work demonstrates that deep learning can be combined with the conventional image analysis algorithms to facilitate the computational analysis of highly complex time-lapse movies of collective cell migration.

bioinformatics↗

bioRxiv2017

Expanding discovery from cancer genomes by integrating protein network analyses with in vivo tumorigenesis assays

Approaches that integrate molecular network information and tumor genome data could complement gene-based statistical tests to identify likely new cancer genes, but are challenging to validate at scale and their predictive value remains unclear. We developed a robust statistic (NetSig) that integrates protein interaction networks and data from 4,742 tumor exomes and used it to accurately classify known driver genes in 60% of tested tumor types and to predict 62 new candidates. We designed a quantitative experimental framework to compare the in vivo tumorigenic potential of NetSig candidates, known oncogenes and random genes in mice showing that NetSig candidates induce tumors at rates comparable to known oncogenes and 10-fold higher than random genes. By reanalyzing nine tumor-inducing NetSig candidates in 242 patients with oncogene-negative lung adenocarcinomas, we find that two (AKT2 and TFDP2) are significantly amplified. Overall, we illustrate a scalable integrated computational and experimental workflow to expand discovery from cancer genomes.

cancer biology↗

bioRxiv2016

Genetic and transgenic reagents for Drosophila simulans, D.mauritiana, D. yakuba, D. santomea and D. virilis

Species of the Drosophila melanogaster species subgroup, including the species D. simulans, D. mauritiana, D. yakuba, and D. santomea, have long served as model systems for studying evolution. Studies in these species have been limited, however, by a paucity of genetic and transgenic reagents. Here we describe a collection of transgenic and genetic strains generated to facilitate genetic studies within and between these species. We have generated many strains of each species containing mapped piggyBac transposons including an enhanced yellow fluorescent protein gene expressed in the eyes and a phiC31 attP site-specific integration site. We have tested a subset of these lines for integration efficiency and reporter gene expression levels. We have also generated a smaller collection of other lines expressing other genetically encoded fluorescent molecules in the eyes and a number of other transgenic reagents that will be useful for functional studies in these species. In addition, we have mapped the insertion locations of 58 transposable elements in D. virilis that will be useful for genetic mapping studies.

genetics↗

bioRxiv2016

Harnessing the lymphocyte meta-phenotype to optimize adoptive cell therapy

There is an urgent need for reliable effective therapy for patients with metastatic sarcoma. Approaches that manipulate the immune system have shown promise for patients with advanced, widely disseminated malignancies. One of these approaches is adoptive cell therapy (ACT), where tumor-infiltrating lymphocytes (TIL) are isolated from the tumor, expanded ex vivo, and then transferred back to the patient. This approach has shown great promise in melanoma, leading to an objective response in approximately half of treated patients [14]. Standard protocols involve characterization of TIL populations with respect to adaptive CD4+ and CD8+ T-lymphocytes, but neglect the possible role of the innate lymphoid repertoire. Due to toxicity and the high cost associated with ACT, the IFN-{gamma} release assay is currently used as a proxy to identify suitable TIL isolates for ACT. Efforts in TIL-ACT for sarcoma, which are pre-clinical and pioneered at Moffitt Cancer Center, have shown that only a minority of the TIL cultures show tumor specific activity in ex vivo IFN-{gamma} assays. Surprisingly, internal melanoma trial data reveal a lack of correlation between IFN-{gamma} assay and clinical outcomes, highlighting the need for a more reliable proxy. We hypothesize the existence of a predictable TIL meta-phenotype that leads to optimal tumor response. Here, we describe preliminary efforts to integrate prospective and existing patient data with mathematical models to optimize the TIL meta-phenotype prior to re-injection.

cancer biology↗