Biomarkers of Tumor Heterogeneity in Glioblastoma Multiforme Cohort of TCGA.

Tumor Whole Slide Images (WSI) are often heterogeneous, which hinders the discovery of biomarkers in the presence of confounding clinical factors. In this study, we present a pipeline for identifying biomarkers from the Glioblastoma Multiforme (GBM) cohort of WSIs from TCGA archive. The GBM cohort endures many technical artifacts while the discovery of GBM biomarkers is challenged because "age" is the single most confounding factor for predicting outcomes. The proposed approach relies on interpretable features (e.g., nuclear morphometric indices), effective similarity metrics for heterogeneity analysis, and robust statistics for identifying biomarkers. The pipeline first removes artifacts (e.g., pen marks) and partitions each WSI into patches for nuclear segmentation via an extended U-Net for subsequent quantitative representation. Given the variations in fixation and staining that can artificially modulate hematoxylin optical density (HOD), we extended Navab's Lab method to normalize images and reduce the impact of batch effects. The heterogeneity of each WSI is then represented either as probability density functions (PDF) per patient or as the composition of a dictionary predicted from the entire cohort of WSIs. For PDF- or dictionary-based methods, morphometric subtypes are constructed based on distances computed from optimal transport and linkage analysis or consensus clustering with Euclidean distances, respectively. For each inferred subtype, Kaplan-Meier and/or the Cox regression model are used to regress the survival time. Since age is the single most important confounder for predicting survival in GBM and there is an observed violation of the proportionality assumption in the Cox model, we use both age and age-squared coupled with the Likelihood ratio test and forest plots for evaluating competing statistics. Next, the PDF- and dictionary-based methods are combined to identify biomarkers that are predictive of survival. The combined model has the advantage of integrating global (e.g., cohort scale) and local (e.g., patient scale) attributes of morphometric heterogeneity, coupled with robust statistics, to reveal stable biomarkers. The results indicate that, after normalization of the GBM cohort, mean HOD, eccentricity, and cellularity are predictive of survival. Finally, we also stratified the GBM cohort as a function of EGFR expression and published genomic subtypes to reveal genomic-dependent morphometric biomarkers.

Influence of Simulated Microgravity on Mammary Epithelial Cells Grown as 2D and 3D Cultures.

During space travel, astronauts will experience a unique environment that includes continuous exposure to microgravity and stressful living conditions. Physiological adaptation to this is a challenge and the effect of microgravity on organ development, architecture, and function is not well understood. How microgravity may impact the growth and development of an organ is an important issue, especially as space flight becomes more commonplace. In this work, we sought to address fundamental questions regarding microgravity using mouse mammary epithelial cells in 2D and 3D tissue cultures exposed to simulated microgravity. Mouse mammary HC11 cells contain a higher proportion of stem cells and were also used to investigate how simulated microgravity may impact mammary stem cell populations. In these studies, we exposed mouse mammary epithelial cells to simulated microgravity in 2D and then assayed for changes in cellular characteristics and damage levels. The microgravity treated cells were also cultured in 3D to form acini structures to define if simulated microgravity affects the cells' ability to organize correctly, a quality that is of key importance for mammary organ development. These studies identify changes occurring during exposure to microgravity that impact cellular characteristics such as cell size, cell cycle profiles, and levels of DNA damage. In addition, changes in the percentage of cells revealing various stem cell profiles were observed following simulated microgravity exposure. In summary, this work suggests microgravity may cause aberrant changes in mammary epithelial cells that lead to an increase in cancer risk.

CD36+ Fibroblasts Secrete Protein Ligands That Growth-Suppress Triple-Negative Breast Cancer Cells While Elevating Adipogenic Markers for a Model of Cancer-Associated Fibroblast.

Tumor and stroma coevolve to facilitate tumor growth. Hence, effective tumor therapeutics would not only induce growth suppression of tumor cells but also revert pro-tumor stroma into anti-tumoral type. Previously, we showed that coculturing triple-negative or luminal A breast cancer cells with CD36+ fibroblasts (FBs) in a three-dimensional extracellular matrix induced their growth suppression or phenotypic reversion, respectively. Then, we identified SLIT3, FBLN-1, and PENK as active protein ligands secreted from CD36+ FBs that induced growth suppression of MDA-MB-231 breast cancer cells and determined their minimum effective concentrations. Here, we have expanded our analyses to include additional triple-negative cancer cell lines, BT549 and Hs578T, as well as HCC1937 carrying a BRCA1 mutation. We show that the ectopic addition of each of the three ligands to cancer-associated fibroblasts (CAFs) elevates the expression of CD36, as well as the adipogenic marker FABP4. Lastly, we show that an agonist antibody for one of the PENK receptors induces growth suppression of all cancer cell lines tested but not for non-transformed MCF10A cells. These results clearly suggest that proteins secreted from CD36+ FBs induce not only growth suppression of tumor cells through binding the cognate receptors but also increasing adipogenic markers of CAFs to reprogram tumor stroma.

Protein Ligands in the Secretome of CD36+ Fibroblasts Induce Growth Suppression in a Subset of Breast Cancer Cell Lines.

Reprogramming the tumor stroma is an emerging approach to circumventing the challenges of conventional cancer therapies. This strategy, however, is hampered by the lack of a specific molecular target. We previously reported that stromal fibroblasts (FBs) with high expression of CD36 could be utilized for this purpose. These studies are now expanded to identify the secreted factors responsible for tumor suppression. Methodologies included 3D colonies, fluorescent microscopy coupled with quantitative techniques, proteomics profiling, and bioinformatics analysis. The results indicated that the conditioned medium (CM) of the CD36+ FBs caused growth suppression via apoptosis in the triple-negative cell lines of MDA-MB-231, BT549, and Hs578T, but not in the ERBB2+ SKBR3. Following the proteomics and bioinformatic analysis of the CM of CD36+ versus CD36- FBs, we determined KLF10 as one of the transcription factors responsible for growth suppression. We also identified FBLN1, SLIT3, and PENK as active ligands, where their minimum effective concentrations were determined. Finally, in MDA-MB-231, we showed that a mixture of FBLN1, SLIT3, and PENK could induce an amount of growth suppression similar to the CM of CD36+ FBs. In conclusion, our findings suggest that these ligands, secreted by CD36+ FBs, can be targeted for breast cancer treatment.

An enhanced loss function simplifies the deep learning model for characterizing the 3D organoid models.

MOTIVATION: Organization of the organoid models, imaged in 3D with a confocal microscope, is an essential morphometric index to assess responses to stress or therapeutic targets. In fact, differentiating malignant and normal cells is often difficult in monolayer cultures. But in 3D culture, colony organization can provide a clear set of indices for differentiating malignant and normal cells. The limiting factors are delineating each cell in a 3D colony in the presence of perceptual boundaries between adjacent cells and heterogeneity associated with cells being at different cell cycles. RESULTS: In a previous paper, we defined a potential field for delineating adjacent nuclei, with perceptual boundaries, in 2D histology images by coupling three deep networks. This concept is now extended to 3D and simplified by an enhanced cost function that replaces three deep networks with one. Validation includes four cell lines with diverse mutations, and a comparative analysis with the UNet models of microscopy indicates an improved performance with the F1-score of 0.83. AVAILABILITY AND IMPLEMENTATION: All software and annotated images are available through GitHub and Bioinformatics online. The software includes the proposed method, UNet for microscopy that was extended to 3D and report generation for profiling colony organization. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Overexpression of CD36 in mammary fibroblasts suppresses colony growth in breast cancer cell lines.

Human breast tumors are not fully autonomous. They are dependent on nutrients and growth-promoting signals provided by the supporting stromal cells. Within the tumor microenvironment, one of the secreted macromolecules by tumor cells is activin A, where we show to downregulate CD36 in fibroblasts. Downregulation of CD36 in fibroblasts also increases the secretion of activin A by fibroblasts. We hypothesize that overexpression of CD36 in fibroblasts inhibits the formation of solid tumors in subtypes of breast cancer models. For the first time, we show that co-culturing organoid models of breast cancer cell lines of MDA-MB-231 (e.g., a triple-negative line) or MCF7 (e.g., a luminal-A line) with CD36+ fibroblasts inhibit the growth and normalizes basal and lateral polarities, respectively. In the long-term anchorage-independent growth assay, the rate of colony formation is also reduced for MDA-MB-231. These observations are consistent with the mechanism of tumor suppression involving the downregulation of pSMAD2/3 and YY1 expression levels. Our integrated analytical methods leverage and extend quantitative assays at cell- and colony-scales in both short- and long-term cultures using brightfield or immunofluorescent microscopy and robust image analysis. Conditioned media are profiled with the ELISA assay.

Deep fusion of contextual and object-based representations for delineation of multiple nuclear phenotypes.

MOTIVATION: Nuclear delineation and phenotypic profiling are important steps in the automated analysis of histology sections. However, these are challenging problems due to (i) technical variations (e.g. fixation, staining) that originate as a result of sample preparation; (ii) biological heterogeneity (e.g. vesicular versus high chromatin phenotypes, nuclear atypia) and (iii) overlapping nuclei. This Application-Note couples contextual information about the cellular organization with the individual signature of nuclei to improve performance. As a result, routine delineation of nuclei in H&E stained histology sections is enabled for either computer-aided pathology or integration with genome-wide molecular data. RESULTS: The method has been evaluated on two independent datasets. One dataset originates from our lab and includes H&E stained sections of brain and breast samples. The second dataset is publicly available through IEEE with a focus on gland-based tissue architecture. We report an approximate AJI of 0.592 and an F1-score 0.93 on both datasets. AVAILABILITY AND IMPLEMENTATION: The code-base, modified dataset and results are publicly available. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Fusion of encoder-decoder deep networks improves delineation of multiple nuclear phenotypes.

BACKGROUND: Nuclear segmentation is an important step for profiling aberrant regions of histology sections. If nuclear segmentation can be resolved, then new biomarkers of nuclear phenotypes and their organization can be predicted for the application of precision medicine. However, segmentation is a complex problem as a result of variations in nuclear geometry (e.g., size, shape), nuclear type (e.g., epithelial, fibroblast), nuclear phenotypes (e.g., vesicular, aneuploidy), and overlapping nuclei. The problem is further complicated as a result of variations in sample preparation (e.g., fixation, staining). Our hypothesis is that (i) deep learning techniques can learn complex phenotypic signatures that rise in tumor sections, and (ii) fusion of different representations (e.g., regions, boundaries) contributes to improved nuclear segmentation. RESULTS: We have demonstrated that training of deep encoder-decoder convolutional networks overcomes complexities associated with multiple nuclear phenotypes, where we evaluate alternative architecture of deep learning for an improved performance against the simplicity of the design. In addition, improved nuclear segmentation is achieved by color decomposition and combining region- and boundary-based features through a fusion network. The trained models have been evaluated against approximately 19,000 manually annotated nuclei, and object-level Precision, Recall, F1-score and Standard Error are reported with the best F1-score being 0.91. Raw training images, annotated images, processed images, and source codes are released as a part of the Additional file 1. CONCLUSIONS: There are two intrinsic barriers in nuclear segmentation to H&E stained images, which correspond to the diversity of nuclear phenotypes and perceptual boundaries between adjacent cells. We demonstrate that (i) the encoder-decoder architecture can learn complex phenotypes that include the vesicular type; (ii) delineation of overlapping nuclei is enhanced by fusion of region- and edge-based networks; (iii) fusion of ENets produces an improved result over the fusion of UNets; and (iv) fusion of networks is better than multitask learning. We suggest that our protocol enables processing a large cohort of whole slide images for applications in precision medicine.