Explore the vast range of titles available.

Aberrant expression of the E26 transformation-specific (ETS) transcription factors characterizes numerous human malignancies. Many of these proteins, including EWS:FLI1 and EWS:ERG fusions in Ewing sarcoma (EwS) and TMPRSS2:ERG in prostate cancer (PCa), drive oncogenic programs via binding to GGAA repeats. We report here that both EWS:FLI1 and ERG bind and transcriptionally activate GGAA-rich pericentromeric heterochromatin. The respective pathogen-like HSAT2 and HSAT3 RNAs, together with LINE, SINE, ERV, and other repeat transcripts, are expressed in EwS and PCa tumors, secreted in extracellular vesicles (EVs), and are highly elevated in plasma of patients with EwS with metastatic disease. High human satellite 2 and 3 (HSAT2,3) levels in EWS:FLI1- or ERG-expressing cells and tumors were associated with induction of G2/M checkpoint, mitotic spindle, and DNA damage programs. These programs were also activated in EwS EV-treated fibroblasts, coincident with accumulation of HSAT2,3 RNAs, proinflammatory responses, mitotic defects, and senescence. Mechanistically, HSAT2,3-enriched cancer EVs induced cGAS-TBK1 innate immune signaling and formation of cytosolic granules positive for double-strand RNAs, RNA-DNA, and cGAS. Hence, aberrantly expressed ETS proteins derepress pericentromeric heterochromatin, yielding pathogenic RNAs that transmit genotoxic stress and inflammation to local and distant sites. Monitoring HSAT2,3 plasma levels and preventing their dissemination may thus improve therapeutic strategies and blood-based diagnostics.

The formation of hypoxic microenvironments within solid tumors is known to contribute to radiation resistance, chemotherapy resistance, immune suppression, increased metastasis, and an overall poor prognosis. It is therefore crucial to understand the spatial and molecular mechanisms that contribute to tumor hypoxia formation to improve the efficacy of radiation treatment, develop hypoxia-directed therapies, and increase patient survival. The objective of this study is to present a number of complementary novel methods for quantifying tumor hypoxia and proliferation in multiplexed immunofluorescence images, especially in relation to the location of perfused blood vessels. A standard marker analysis strategy is to take a positive pixel count approach, in which a threshold for positive stain is used to compute a positive area fraction for hypoxia. This work is a reassessment of that approach, utilizing not only cell segmentation but also distance to nearest blood vessel in order to incorporate spatial information into the analysis. We describe a reproducible pipeline for the visualization and quantitative analysis of hypoxia using a vessel distance analysis approach. This methodological pipeline can serve to further elucidate the relationship between vessel distance and microenvironment-linked markers such as hypoxia and proliferation, can help to quantify parameters relating to oxygen consumption and hypoxic tolerance in tissues, as well as potentially serve as a hypothesis generating tool for future studies testing hypoxia-linked markers.

Heterogeneity within solid tissue microenvironments has been well established as a contributor to patient outcome for a variety of different diseases and their etiology. With the advent of next-generational artificial intelligence algorithms for image segmentation and classification, current digital pathology workflows stand to gain considerable benefits by incorporating such tools. As pathological assessment is still widely regarded as the gold standard for numerous patient stratification and disease progression endpoints, it is important to incorporate artificial intelligence in a manner that works alongside pathologist’s classification, rather than aim to replace them. In this dissertation, I have investigated approaches for the development and application of computational methods for studying tissue microenvironments through the lens of spatial biology.First, we present an overview of hypoxia, one of the main determinants of microenvironmental heterogeneity, the means of measuring and targeting it, and the value of spatially-oriented analysis of hypoxia in high dimensionality histopathology datasets. Chapter 1 demonstrates the development of computational methods for the quantitative assessment of hypoxia and proliferation spatial gradients within histological tissue sections. In chapter 2, we apply these methods to two separate studies: evaluating the efficacy of the hypoxia activated prodrug evofosfamide in head and neck squamous cell carcinoma, and understanding the role and therapeutic potential of the autophagy activating kinase ULK1 in pancreatic cancer. In chapter 3, we build upon these methods to characterize spatial heterogeneity in glioblastoma microenvironments through the analysis of a high dimensionality dataset using the multiplexed immunohistochemistry imaging modality, Imaging Mass Cytometry (IMC). We explore the differential localization of specific cell subsets to regions of hypoxia and perform a comparison of the spatial distribution and localization of hypoxia biomarkers.To further test the robustness of the IMC analysis method developed in the previous chapter, we next apply this workflow in chapter 4 to characterize the renal localization of immune cells contributing to acute kidney injury (AKI) following treatment of cancers with immune checkpoint inhibitors (ICI). We identify specific cell subpopulations associated with AKI caused by ICI versus healthy tissue and corroborate our findings with supplementary patient information. Further building upon this pathologist-in-the-loop IMC analysis methodology, we use these tools to explore the single cell landscape of renal transplant rejection in chapter 5, incorporating further spatial characterizations such as the localization of immune cells to renal structures across disease types, to building a patient classifier capable of identifying spatially-derived features indicative of transplant rejection.In chapter 6, we delve back into the characterization of glioblastoma heterogeneity, this time using spatial transcriptomics. We develop a novel spatial transcriptomic signature of hypoxia in glioblastoma using registration of paired hypoxia immunohistochemistry images and piloting a spatial analysis method using open-source tools. This work highlights the critical role of spatial dimension in gene expression research, introducing a workflow that tackles batch variation and augments the spatial feature extraction of gene expression, thus unveiling novel perspectives on the transcriptional microenvironment of glioblastoma.

There is no neoadjuvant immunotherapy for early-stage oral cancer patients. We report a single-arm, open-label, pilot clinical trial assessing the efficacy and safety of topical toll-like receptor-7 (TLR-7) agonist, imiquimod, utilized in a neoadjuvant setting in early-stage oral squamous cell carcinoma (OSCC). The primary endpoint is reduction in tumor cell counts assessed by quantitative multiplex immunofluorescence and the immune-related pathologic response. The secondary endpoint is safety. 60% of patients experienced a 50% reduction or greater in tumor cell count post-treatment (95% CI = 32% to 84%). Similarly, 60% of patients had immune-related major pathologic response (irMPR) with two complete pathologic responses, and 40% had partial response (PR) with the percent residual viable tumor ranging from 25% to 65%. An increase in functional helper and cytotoxic T-cells significantly contributed to a reduction in tumor (R=0.54 and 0.55, respectively). The treatment was well tolerated with the application site mucositis being the most common adverse event (grades 1-3), and no grade 4 life-threatening event. The median follow-up time was 17 months (95% CI = 16 months - not reached), and one-year recurrence-free survival was 93% of evaluable patients. Neoadjuvant imiquimod immunotherapy could be safe and promising regimen for early-stage oral cancer. ClinicalTrials.gov, Identifier NCT04883645.

Malformations of cortical development such as tuberous sclerosis complex arise within a heterogeneous cellular landscape that conventional histopathology only partially resolves. Here, we combined a 19-marker multiplex immunofluorescence panel with a machine learning-driven image analysis pipeline to map and quantify over 365 000 cells from paediatric surgical cortex, defining the single-cell architecture of TSC lesions. Microtubers were objectively delineated by vimentin and detected in all TSC samples but absent from non-dysplastic controls. Within these structures, balloon cells exhibiting strong pS6 activation occupied lesion cores and were confined to microtuber boundaries, whereas dysmorphic neurons were more diffusely distributed into adjacent cortex. The microtuber niche was dominated by astroglial remodeling: immature and reactive vimentin-positive astrocytes, including Lamp5-positive subsets, accumulated at and around lesion rims, while mature GFAP-positive astrocytes showed only modest changes. Distance-based spatial analyses revealed neuronal exclusion from microtuber centres with gradual recovery in surrounding tissue, indicating local network disruption. Unsupervised clustering and niche modelling recapitulated these spatial gradients, identifying a glial-dominated ecosystem that concentrates balloon cells, increases inter-neuronal distances, and reduces cell-cell interactions. Together, these data support a model in which cortical tubers arise through the coalescence of microtubers orchestrated by balloon cells and reactive gliosis during corticogenesis. Beyond elucidating disease architecture, our automated framework enables reproducible lesion detection, quantitative cell-state mapping, and spatial readouts applicable across malformations of cortical development.

Medical image segmentation is a difficult but important task for many clinical operations such as cardiac bi-ventricular volume estimation. More recently, there has been a shift to utilizing deep learning and fully convolutional neural networks (CNNs) to perform image segmentation that has yielded state-of-the-art results in many public benchmark datasets. Despite the progress of deep learning in medical image segmentation, standard CNNs are still not fully adopted in clinical settings as they lack robustness and interpretability. Shapes are generally more meaningful features than solely textures of images, which are features regular CNNs learn, causing a lack of robustness. Likewise, previous works surrounding model interpretability have been focused on post hoc gradient-based saliency methods. However, gradient-based saliency methods typically require additional computations post hoc and have been shown to be unreliable for interpretability. Thus, we present a new architecture called Shape Attentive U-Net (SAUNet) which focuses on model interpretability and robustness. The proposed architecture attempts to address these limitations by the use of a secondary shape stream that captures rich shape-dependent information in parallel with the regular texture stream. Furthermore, we suggest multi-resolution saliency maps can be learned using our dual-attention decoder module which allows for multi-level interpretability and mitigates the need for additional computations post hoc. Our method also achieves state-of-the-art results on the two large public cardiac MRI image segmentation datasets of SUN09 and AC17.

The formation of hypoxic microenvironments within solid tumors is known to contribute to radiation resistance, chemotherapy resistance, immune suppression, increased metastasis, and an overall poor prognosis. It is therefore crucial to understand the spatial and molecular mechanisms that contribute to tumor hypoxia formation to improve the efficacy of radiation treatment, develop hypoxia-directed therapies, and increase patient survival. The objective of this study is to present a number of complementary novel methods for quantifying tumor hypoxia and proliferation, especially in relation to the location of perfused blood vessels. Multiplexed immunofluorescence staining can produce whole slide scanned image datasets that are amenable for computational pathology analysis. A standard marker analysis strategy is to take a positive pixel count approach, in which a threshold for positive stain is used to compute a positive area fraction for hypoxia. This work is a reassessment of that approach, utilizing not only cell segmentation but also distance to nearest blood vessel in order to incorporate spatial information into the analysis. We describe a reproducible pipeline for the visualization and quantitative analysis of hypoxia using a vessel distance analysis approach. This methodological pipeline can serve to further elucidate the relationship between vessel distance and microenvironment-linked markers such as hypoxia and proliferation, can help to quantify parameters relating to oxygen consumption and hypoxic tolerance in tissues, as well as potentially serve as a hypothesis generating tool for future studies testing hypoxia-linked markers. Footnotes * https://github.com/STTARR/Vessel-Distance-Analysis

Ewing sarcoma (EwS) is an aggressive childhood malignancy with a high propensity for metastasis. By analyzing cohorts of patients and age-matched healthy donors, we establish that EwS metastatic progression is accompanied by elevated plasma levels of multiple proinflammatory cytokines, interferons and extracellular vesicles (EVs). The latter were enriched with transcripts derived from LINE, SINE and ERV retroelements and from locus-specific pericentromeric regions, including HSAT2. We show that some of these RNAs, including HSAT2 and HERV-K, are selectively transmitted in EwS EVs and taken up by stromal fibroblasts and peripheral blood CD33+ myeloid cells and CD8+ T-cells, inducing immune exhaustion, immunosuppressive phenotypes and proinflammatory responses. Moreover, EwS EV-derived repeat RNAs were propagated and serially transmitted in recipient cell EVs, reminiscent of viral infection. As such, this study uncovers a novel mechanism driving cancer-associated inflammation, immunosuppression and metastatic progression.

The evolutionary reasons for variation in nose shape across human populations have been subject to continuing debate. An import function of the nose and nasal cavity is to condition inspired air before it reaches the lower respiratory tract. For this reason, it is thought the observed differences in nose shape among populations are not simply the result of genetic drift, but may be adaptations to climate. To address the question of whether local adaptation to climate is responsible for nose shape divergence across populations, we use Qst-Fst comparisons to show that nares width and alar base width are more differentiated across populations than expected under genetic drift alone. To test whether this differentiation is due to climate adaptation, we compared the spatial distribution of these variables with the global distribution of temperature, absolute humidity, and relative humidity. We find that width of the nares is correlated with temperature and absolute humidity, but not with relative humidity. We conclude that some aspects of nose shape may indeed have been driven by local adaptation to climate. However, we think that this is a simplified explanation of a very complex evolutionary history, which possibly also involved other non-neutral forces such as sexual selection.