Automatic Nuclei Segmentation in H&E Stained Breast Cancer Histopathology Images

The introduction of fast digital slide scanners that provide whole slide images has led to a revival of interest in image analysis applications in pathology. Segmentation of cells and nuclei is an important first step towards automatic analysis of digitized microscopy images. We therefore developed an automated nuclei segmentation method that works with hematoxylin and eosin (H&E) stained breast cancer histopathology images, which represent regions of whole digital slides. The procedure can be divided into four main steps: 1) pre-processing with color unmixing and morphological operators, 2) marker-controlled watershed segmentation at multiple scales and with different markers, 3) post-processing for rejection of false regions and 4) merging of the results from multiple scales. The procedure was developed on a set of 21 breast cancer cases (subset A) and tested on a separate validation set of 18 cases (subset B). The evaluation was done in terms of both detection accuracy (sensitivity and positive predictive value) and segmentation accuracy (Dice coefficient). The mean estimated sensitivity for subset A was 0.875 (±0.092) and for subset B 0.853 (±0.077). The mean estimated positive predictive value was 0.904 (±0.075) and 0.886 (±0.069) for subsets A and B, respectively. For both subsets, the distribution of the Dice coefficients had a high peak around 0.9, with the vast majority of segmentations having values larger than 0.8.     

Systems biology approach to developing S2RM-based “systems therapeutics” and naturally induced pluripotent stem cells

The degree to, and the mechanisms through, which stem cells are able to build, maintain, and heal the body have only recently begun to be understood. Much of the stem cell’s power resides in the release of a multitude of molecules, called stem cell released molecules (SRM). A fundamentally new type of therapeutic, namely “systems therapeutic”, can be realized by reverse engineering the mechanisms of the SRM processes. Recent data demonstrates that the composition of the SRM is different for each type of stem cell, as well as for different states of each cell type. Although systems biology has been successfully used to analyze multiple pathways, the approach is often used to develop a small molecule interacting at only one pathway in the system. A new model is emerging in biology where systems biology is used to develop a new technology acting at multiple pathways called “systems therapeutics”. A natural set of healing pathways in the human that uses SRM is instructive and of practical use in developing systems therapeutics. Endogenous SRM processes in the human body use a combination of SRM from two or more stem cell types, designated as S²RM, doing so under various state dependent conditions for each cell type. Here we describe our approach in using state-dependent SRM from two or more stem cell types, S²RM technology, to develop a new class of therapeutics called “systems therapeutics.” Given the ubiquitous and powerful nature of innate S²RM-based healing in the human body, this “systems therapeutic” approach using S²RM technology will be important for the development of anti-cancer therapeutics, antimicrobials, wound care products and procedures, and a number of other therapeutics for many indications.     

Bortezomib Resistance Can Be Reversed by Induced Expression of Plasma Cell Maturation Markers in a Mouse In Vitro Model of Multiple Myeloma

Multiple myeloma (MM), the second most common hematopoietic malignancy, remains an incurable plasma cell (PC) neoplasm. While the proteasome inhibitor, bortezomib (Bz) has increased patient survival, resistance represents a major treatment obstacle as most patients ultimately relapse becoming refractory to additional Bz therapy. Current tests fail to detect emerging resistance; by the time patients acquire resistance, their prognosis is often poor. To establish immunophenotypic signatures that predict Bz sensitivity, we utilized Bz-sensitive and -resistant cell lines derived from tumors of the Bcl-X_L/Myc mouse model of PC malignancy. We identified significantly reduced expression of two markers (CD93, CD69) in “acquired” (Bz-selected) resistant cells. Using this phenotypic signature, we isolated a subpopulation of cells from a drug-naïve, Bz-sensitive culture that displayed “innate” resistance to Bz. Although these genes were identified as biomarkers, they may indicate a mechanism for Bz-resistance through the loss of PC maturation which may be induced and/or selected by Bz. Significantly, induction of PC maturation in both “acquired” and “innate” resistant cells restored Bz sensitivity suggesting a novel therapeutic approach for reversing Bz resistance in refractory MM.     

Enabling user-guided segmentation and tracking of surface-labeled cells in time-lapse image sets of living tissues

To study the process of morphogenesis, one often needs to collect and segment time-lapse images of living tissues to accurately track changing cellular morphology. This task typically involves segmenting and tracking tens to hundreds of individual cells over hundreds of image frames, a scale that would certainly benefit from automated routines; however, any automated routine would need to reliably handle a large number of sporadic, and yet typical problems (e.g., illumination inconsistency, photobleaching, rapid cell motions, and drift of focus or of cells moving through the imaging plane). Here, we present a segmentation and cell tracking approach based on the premise that users know their data best-interpreting and using image features that are not accounted for in any a priori algorithm design. We have developed a program, SeedWater Segmenter, that combines a parameter-less and fast automated watershed algorithm with a suite of manual intervention tools that enables users with little to no specialized knowledge of image processing to efficiently segment images with near-perfect accuracy based on simple user interactions.     

Benefits and limitations of a new genome‐based PCR‐RFLP genotyping assay (GB‐RFLP): A SNP‐based detection method for identification of species in extremely young adaptive radiations

High‐throughput DNA sequencing technologies make it possible now to sequence entire genomes relatively easily. Complete genomic information obtained by whole‐genome resequencing (WGS) can aid in identifying and delineating species even if they are extremely young, cryptic, or morphologically difficult to discern and closely related. Yet, for taxonomic or conservation biology purposes, WGS can remain cost‐prohibitive, too time‐consuming, and often constitute a “data overkill.” Rapid and reliable identification of species (and populations) that is also cost‐effective is made possible by species‐specific markers that can be discovered by WGS. Based on WGS data, we designed a PCR restriction fragment length polymorphism (PCR‐RFLP) assay for 19 Neotropical Midas cichlid populations (Amphilophus cf. citrinellus), that includes all 13 described species of this species complex. Our work illustrates that identification of species and populations (i.e., fish from different lakes) can be greatly improved by designing genetic markers using available “high resolution” genomic information. Yet, our work also shows that even in the best‐case scenario, when whole‐genome resequencing information is available, unequivocal assignments remain challenging when species or populations diverged very recently, or gene flow persists. In summary, we provide a comprehensive workflow on how to design RFPL markers based on genome resequencing data, how to test and evaluate their reliability, and discuss the benefits and pitfalls of our approach.     

A novel three-dimensional variant of the watershed transform for segmentation of electron density maps

Electron density maps at moderate resolution are often difficult to interpret due to the lack of recognizable features. This is especially true for electron tomograms that suffer in addition to the resolution limitation from low signal-to-noise ratios. Reliable segmentation of such maps into smaller, manageable units can greatly facilitate interpretation. Here, we present a segmentation approach targeting three-dimensional electron density maps derived by electron microscopy. The approach consists of a novel three-dimensional variant of the immersion-based watershed algorithm. We tested the algorithm on calculated data and applied it to a wide variety of electron density maps ranging from reconstructions of single macromolecules to tomograms of subcellular structures. The results indicate that the algorithm is reliable, efficient, accurate, and applicable to a wide variety of biological problems.     

A Context-Aware Delayed Agglomeration Framework for Electron Microscopy Segmentation

Electron Microscopy (EM) image (or volume) segmentation has become significantly important in recent years as an instrument for connectomics. This paper proposes a novel agglomerative framework for EM segmentation. In particular, given an over-segmented image or volume, we propose a novel framework for accurately clustering regions of the same neuron. Unlike existing agglomerative methods, the proposed context-aware algorithm divides superpixels (over-segmented regions) of different biological entities into different subsets and agglomerates them separately. In addition, this paper describes a “delayed” scheme for agglomerative clustering that postpones some of the merge decisions, pertaining to newly formed bodies, in order to generate a more confident boundary prediction. We report significant improvements attained by the proposed approach in segmentation accuracy over existing standard methods on 2D and 3D datasets.     

In situ histone landscape of nephrogenesis

In the developing kidney, self-renewing progenitors respond to inductive signaling from the adjacent branching ureteric bud by undergoing mesenchyme-to-epithelium transition. Nascent nephrons subsequently undergo elongation, segmentation, and differentiation into a mature renal epithelium with diverse functions. Epigenetic mechanisms have been implicated in impacting cell fate decisions during nephrogenesis; however, the chromatin landscape of nephron progenitors and daughter differentiating cells are largely unknown. Here, we examined the spatiotemporal expression patterns of histone H3 methylation and histone methyltransferases in E15.5 mouse kidneys. Kidney sections were probed with antibodies against histone modifications, enzymes, and markers of progenitors and differentiation. The results revealed that: (1) nephron progenitor cells exhibit a broad histone methylation signature that comprises both “active” and “repressive” marks (H3K4me3/K9me3/K27me3/R2me2/R17me2); (2) nascent nephrons retain high H3K4me3 but show downregulation of H3K9/K27me3 and; (3) maturing epithelial tubules acquire high levels of H3K79me2/3. Consistent with respective histone marks, the H3K4 methyltransferase, Ash2l, is expressed in progenitors and nascent nephrons, whereas the H3K9/K27 methyltransferases, G9a/Ezh2, are more enriched in progenitors than nascent nephrons. We conclude that combinatorial histone signatures correlate with cell fate decisions during nephrogenesis.     

A watershed approach for improving medical image segmentation

In this paper, a novel watershed approach based on seed region growing and image entropy is presented which could improve the medical image segmentation. The proposed algorithm enables the prior information of seed region growing and image entropy in its calculation. The algorithm starts by partitioning the image into several levels of intensity using watershed multi-degree immersion process. The levels of intensity are the input to a computationally efficient seed region segmentation process which produces the initial partitioning of the image regions. These regions are fed to entropy procedure to carry out a suitable merging which produces the final segmentation. The latter process uses a region-based similarity representation of the image regions to decide whether regions can be merged. The region is isolated from the level and the residual pixels are uploaded to the next level and so on, we recall this process as multi-level process and the watershed is called multi-level watershed. The proposed algorithm is applied to challenging applications: grey matter–white matter segmentation in magnetic resonance images (MRIs). The established methods and the proposed approach are experimented by these applications to a variety of simulating immersion, multi-degree, multi-level seed region growing and multi-level seed region growing with entropy. It is shown that the proposed method achieves more accurate results for medical image oversegmentation.     

Engineering Antigen-Specific T Cells from Genetically Modified Human Hematopoietic Stem Cells in Immunodeficient Mice

There is a desperate need for effective therapies to fight chronic viral infections. The immune response is normally fastidious at controlling the majority of viral infections and a therapeutic strategy aimed at reestablishing immune control represents a potentially powerful approach towards treating persistent viral infections. We examined the potential of genetically programming human hematopoietic stem cells to generate mature CD8+ cytotoxic T lymphocytes that express a molecularly cloned, “transgenic” human anti-HIV T cell receptor (TCR). Anti-HIV TCR transduction of human hematopoietic stem cells directed the maturation of a large population of polyfunctional, HIV-specific CD8+ cells capable of recognizing and killing viral antigen-presenting cells. Thus, through this proof-of-concept we propose that genetic engineering of human hematopoietic stem cells will allow the tailoring of effector T cell responses to fight HIV infection or other diseases that are characterized by the loss of immune control.     

“Velcro” Engineering of High Affinity CD47 Ectodomain as Signal Regulatory Protein α (SIRPα) Antagonists That Enhance Antibody-dependent Cellular Phagocytosis

CD47 is a cell surface protein that transmits an anti-phagocytic signal, known as the “don''t-eat-me” signal, to macrophages upon engaging its receptor signal regulatory protein α (SIRPα). Molecules that antagonize the CD47-SIRPα interaction by binding to CD47, such as anti-CD47 antibodies and the engineered SIRPα variant CV1, have been shown to facilitate macrophage-mediated anti-tumor responses. However, these strategies targeting CD47 are handicapped by large antigen sinks in vivo and indiscriminate cell binding due to ubiquitous expression of CD47. These factors reduce bioavailability and increase the risk of toxicity. Here, we present an alternative strategy to antagonize the CD47-SIRPα pathway by engineering high affinity CD47 variants that target SIRPα, which has restricted tissue expression. CD47 proved to be refractive to conventional affinity maturation techniques targeting its binding interface with SIRPα. Therefore, we developed a novel engineering approach, whereby we augmented the existing contact interface via N-terminal peptide extension, coined “Velcro” engineering. The high affinity variant (Velcro-CD47) bound to the two most prominent human SIRPα alleles with greatly increased affinity relative to wild-type CD47 and potently antagonized CD47 binding to SIRPα on human macrophages. Velcro-CD47 synergizes with tumor-specific monoclonal antibodies to enhance macrophage phagocytosis of tumor cells in vitro, with similar potency as CV1. Finally, Velcro-CD47 interacts specifically with a subset of myeloid-derived cells in human blood, whereas CV1 binds all myeloid, lymphoid, and erythroid populations interrogated. This is consistent with the restricted expression of SIRPα compared with CD47. Herein, we have demonstrated that “Velcro” engineering is a powerful protein-engineering tool with potential applications to other systems and that Velcro-CD47 could be an alternative adjuvant to CD47-targeting agents for cancer immunotherapy.     

A Systemically-Administered Small Molecule Antagonist of CCR9 Acts as a Tissue-Selective Inhibitor of Lymphocyte Trafficking

A goal for developers of immunomodulatory drugs has long been a systemically administered small molecule that can selectively inhibit inflammation in specific tissues. The chemokine receptor CCR9 is an attractive target for this approach, as entry of T cells into the small intestine from blood requires interaction between CCR9 and its ligand CCL25. We have tested the ability of a small molecule CCR9 antagonist, CCX8037, to inhibit antigen-mediated T cell accumulation in the intestine. This compound prevented accumulation of gut-imprinted antigen-specific CD8 T cells within epithelium of the small intestine. Interestingly, the antagonist did not affect the robust generation of gut-imprinted CD8 T cells within mesenteric lymph nodes. To distinguish “gut-selective” from “general” T cell inhibition, we tested the drug’s ability to influence accumulation of T cells within skin, a tissue in which CCR9 plays no known role, and we found no appreciable effect. This study demonstrates the feasibility of creating systemically-administered pharmaceuticals capable of tissue-selective immune modulation. This proof of concept is of utmost importance for designing effective treatments against various autoimmune disorders localized to a specific tissue.     

Zooming in on the cell biology of disease

Today’s cell biology could be considered a fusion of disciplines that blends advanced genetics, molecular biology, biochemistry, and engineering to answer fundamental as well as medically relevant scientific questions. Accordingly, our understanding of diseases is greatly aided by an existing vast knowledge base of fundamental cell biology. Gunter Blobel captured this concept when he said, “the tremendous acquisition of basic knowledge will allow a much more rational treatment of cancer, viral infection, degenerative disease and mental disease.” In other words, without cell biology can we truly understand, prevent, or effectively treat a disease?

R. M. Perera     

Transcriptome analysis of cattle muscle identifies potential markers for skeletal muscle growth rate and major cell types

Background

This study aimed to identify markers for muscle growth rate and the different cellular contributors to cattle muscle and to link the muscle growth rate markers to specific cell types.

Results

The expression of two groups of genes in the longissimus muscle (LM) of 48 Brahman steers of similar age, significantly enriched for “cell cycle” and “ECM (extracellular matrix) organization” Gene Ontology (GO) terms was correlated with average daily gain/kg liveweight (ADG/kg) of the animals. However, expression of the same genes was only partly related to growth rate across a time course of postnatal LM development in two cattle genotypes, Piedmontese x Hereford (high muscling) and Wagyu x Hereford (high marbling). The deposition of intramuscular fat (IMF) altered the relationship between the expression of these genes and growth rate. K-means clustering across the development time course with a large set of genes (5,596) with similar expression profiles to the ECM genes was undertaken. The locations in the clusters of published markers of different cell types in muscle were identified and used to link clusters of genes to the cell type most likely to be expressing them. Overall correspondence between published cell type expression of markers and predicted major cell types of expression in cattle LM was high. However, some exceptions were identified: expression of SOX8 previously attributed to muscle satellite cells was correlated with angiogenesis. Analysis of the clusters and cell types suggested that the “cell cycle” and “ECM” signals were from the fibro/adipogenic lineage. Significant contributions to these signals from the muscle satellite cells, angiogenic cells and adipocytes themselves were not as strongly supported. Based on the clusters and cell type markers, sets of five genes predicted to be representative of fibro/adipogenic precursors (FAPs) and endothelial cells, and/or ECM remodelling and angiogenesis were identified.

Conclusions

Gene sets and gene markers for the analysis of many of the major processes/cell populations contributing to muscle composition and growth have been proposed, enabling a consistent interpretation of gene expression datasets from cattle LM. The same gene sets are likely to be applicable in other cattle muscles and in other species.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1403-x) contains supplementary material, which is available to authorized users.     

Mitochondrial metabolism in cancer metastasis: Visualizing tumor cell mitochondria and the “reverse Warburg effect” in positive lymph node tissue

We have recently proposed a new two-compartment model for understanding the Warburg effect in tumor metabolism. In this model, glycolytic stromal cells produce mitochondrial fuels (L-lactate and ketone bodies) that are then transferred to oxidative epithelial cancer cells, driving OXPHOS and mitochondrial metabolism. Thus, stromal catabolism fuels anabolic tumor growth via energy transfer. We have termed this new cancer paradigm the “reverse Warburg effect,” because stromal cells undergo aerobic glycolysis, rather than tumor cells. To assess whether this mechanism also applies during cancer cell metastasis, we analyzed the bioenergetic status of breast cancer lymph node metastases, by employing a series of metabolic protein markers. For this purpose, we used MCT4 to identify glycolytic cells. Similarly, we used TOMM20 and COX staining as markers of mitochondrial mass and OXPHOS activity, respectively. Consistent with the “reverse Warburg effect,” our results indicate that metastatic breast cancer cells amplify oxidative mitochondrial metabolism (OXPHOS) and that adjacent stromal cells are glycolytic and lack detectable mitochondria. Glycolytic stromal cells included cancer-associated fibroblasts, adipocytes and inflammatory cells. Double labeling experiments with glycolytic (MCT4) and oxidative (TOMM20 or COX) markers directly shows that at least two different metabolic compartments co-exist, side-by-side, within primary tumors and their metastases. Since cancer-associated immune cells appeared glycolytic, this observation may also explain how inflammation literally “fuels” tumor progression and metastatic dissemination, by “feeding” mitochondrial metabolism in cancer cells. Finally, MCT4(+) and TOMM20(-) “glycolytic” cancer cells were rarely observed, indicating that the conventional “Warburg effect” does not frequently occur in cancer-positive lymph node metastases.Key words: caveolin-1, oxidative stress, MCT4, metabolic coupling, tumor stroma, SLC16A3, monocarboxylic acid transporter, two-compartment tumor metabolism, metastasis, TOMM20, complex IV, OXPHOS, mitochondria, inflammation     

An Improved Artificial Bee Colony-Based Approach for Zoning Protected Ecological Areas

China is facing ecological and environmental challenges as its urban growth rate continues to rise, and zoning protected ecological areas is recognized as an effective response measure. Zoning inherently involves both site attributes and aggregation attributes, and the combination of mathematical models and heuristic algorithms have proven advantageous. In this article, an improved artificial bee colony (IABC)-based approach is proposed for zoning protected ecological areas at a regional scale. Three main improvements were made: the first is the use of multiple strategies to generate the initial bee population of a specific quality and diversity, the second is an exploitation search procedure to generate neighbor solutions combining “replace” and “alter” operations, and the third is a “swap” strategy to enable a local search for the iterative optimal solution. The IABC algorithm was verified using simulated data. Then it was applied to define an optimum scheme of protected ecological areas of Sanya (in the Hainan province of China), and a reasonable solution was obtained. Finally, a comparison experiment with other methods (agent-based land allocation model, ant colony optimization, and density slicing) was conducted and demonstrated that the IABC algorithm was more effective and efficient than the other methods. Through this study, we aimed to provide a scientifically sound, practical approach for zoning procedures.     

Some personal and historical notes on the utility of “deep-etch” electron microscopy for making cell structure/function correlations

This brief essay talks up the advantages of metal replicas for electron microscopy and explains why they are still the best way to image frozen cells in the electron microscope. Then it explains our approach to freezing, namely the Van Harreveld trick of “slamming” living cells onto a supercold block of metal sprayed with liquid helium at −269ºC, and further talks up this slamming over the alternative of high-pressure freezing, which is much trickier but enjoys greater favor at the moment. This leads me to bemoan the fact that there are not more young investigators today who want to get their hands on electron microscopes and use our approach to get the most “true to life” views of cells out of them with a minimum of hassle. Finally, it ends with a few perspectives on my own career and concludes that, personally, I''m permanently stuck with the view of the “founding fathers” that cell ultrastructure will ultimately display and explain all of cell function, or as Palade said in his Nobel lecture,electron micrographs are “irresistible and half transparent … their meaning buried under only a few years of work,” and “reasonable working hypotheses are already suggested by the ultrastructural organization itself.”