A deep hybrid learning pipeline for accurate diagnosis of ovarian cancer based on nuclear morphology

Nuclear morphological features are potent determining factors for clinical diagnostic approaches adopted by pathologists to analyze the malignant potential of cancer cells. Considering the structural alteration of the nucleus in cancer cells, various groups have developed machine learning techniques based on variation in nuclear morphometric information like nuclear shape, size, nucleus-cytoplasm ratio and various non-parametric methods like deep learning have also been tested for analyzing immunohistochemistry images of tissue samples for diagnosing various cancers. We aim to correlate the morphometric features of the nucleus along with the distribution of nuclear lamin proteins with classical machine learning to differentiate between normal and ovarian cancer tissues. It has already been elucidated that in ovarian cancer, the extent of alteration in nuclear shape and morphology can modulate genetic changes and thus can be utilized to predict the outcome of low to a high form of serous carcinoma. In this work, we have performed exhaustive imaging of ovarian cancer versus normal tissue and developed a dual pipeline architecture that combines the matrices of morphometric parameters with deep learning techniques of auto feature extraction from pre-processed images. This novel Deep Hybrid Learning model, though derived from classical machine learning algorithms and standard CNN, showed a training and validation AUC score of 0.99 whereas the test AUC score turned out to be 1.00. The improved feature engineering enabled us to differentiate between cancerous and non-cancerous samples successfully from this pilot study.

A microfluidic device for characterizing nuclear deformations

Cell nuclei experience and respond to a wide range of forces, both in vivo and in vitro. In order to characterize the nuclear response to physical stress, we developed a microfluidic chip and used it to apply mechanical stress to live cells and measure their nuclear deformability. The device design is optimized for the detection of both nucleus and cytoplasm, which can then be conveniently quantified using a custom-written Matlab program. We measured nuclear sizes and strains of embryonic stem cells, for which we observed negative Poisson ratios in the nuclei. In addition, we were able to detect changes in the nuclear response after treatment with actin depolymerizing and chromatin decondensing agents. Finally, we showed that the device can be used for biologically relevant high-resolution confocal imaging of cells under compression. Thus, the device presented here allows for accurate physical phenotyping at high throughput and has the potential to be applied to a range of cell types.

Nucleus Morphometry in Cultured Epithelial Cells Correlates with Phenotype

Phenotype of cultured ocular epithelial transplants has been shown to affect clinical success rates following transplantation to the cornea. The purpose of this study was to evaluate the relationship between cell nucleus morphometry and phenotype in three types of cultured epithelial cells. This study provides knowledge for the development of a non-invasive method of determining the phenotype of cultured epithelium before transplantation. Cultured human conjunctival epithelial cells (HCjE), human epidermal keratinocytes (HEK), and human retinal pigment epithelial cells (HRPE) were analyzed by quantitative immunofluorescence. Assessments of nucleus morphometry and nucleus-to-cytoplasm ratio (N/C ratio) were performed using ImageJ. Spearman's correlation coefficient was employed for statistical analysis. Levels of the proliferation marker PCNA in HCjE, HEK, and HRPE correlated positively with nuclear area. Nuclear area correlated significantly with levels of the undifferentiated cell marker ABCG2 in HCjE. Bmi1 levels, but not p63α levels, correlated significantly with nuclear area in HEK. The N/C ratio did not correlate significantly with any of the immunomarkers in HCjE (ABCG2, CK7, and PCNA) and HRPE (PCNA). In HEK, however, the N/C ratio was negatively correlated with levels of the undifferentiated cell marker CK14 and positively correlated with Bmi1 expression. The size of the nuclear area correlated positively with proliferation markers in all three epithelia. Morphometric indicators of phenotype in cultured epithelia can be identified using ImageJ. Conversely, the N/C ratio did not show a uniform relationship with phenotype in HCjE, HEK, or HRPE. N/C ratio therefore, may not be a useful morphometric marker for in vitro assessment of phenotype in these three epithelia.

Nucleus deformation of SaOs-2 cells on rhombic µ-pillars

It has been previously shown that osteosarcoma (SaOs-2) cells respond to micropillared surfaces consisting of poly-L-lactic acid with strong deformation of the cell body and nucleus. Until now, cell nucleus deformation of SaOs-2 cells was only studied by exposing them to square shaped micropillars in an isotropic pattern. Here we report on experiments of the cell nucleus response of such cells to rhombic structures of different topographies generated from a rubbery polymer, namely poly(n-butyacrylate). It is observed that cells orientate themselves perpendicular to the long axis of the rhombi. While their spreading on the surface is not influenced by the opening angle of the structures, rhombic structures with sharper angles induce stronger deformation of the cells and accordingly more elongated nuclei.

Epithelial-mesenchymal transition in prostate cancer is associated with quantifiable changes in nuclear structure

BACKGROUND

Prostate cancer progression is concomitant with quantifiable nuclear structure and texture changes as compared to non-cancer tissue. Malignant progression is associated with an epithelial-mesenchymal transition (EMT) program whereby epithelial cancer cells take on a mesenchymal phenotype and dissociate from a tumor mass, invade, and disseminate to distant metastatic sites. The objective of this study was to determine if epithelial and mesenchymal prostate cancer cells have different nuclear morphology.

METHODS

Murine tibia injections of epithelial PC3 (PC3-Epi) and mesenchymal PC3 (PC3-EMT) prostate cancer cells were processed and stained with H&E. Cancer cell nuclear image data was obtained using commercially available image-processing software. Univariate and multivariate statistical analysis were used to compare the two phenotypes. Several non-parametric classifiers were constructed and permutation-tested at various training set fractions to ensure robustness of classification between PC3-Epi and PC3-EMT cells in vivo.

RESULTS

PC3-Epi and PC3-EMT prostate cancer cells were separable at the single cell level in murine tibia injections on the basis of nuclear structure and texture remodeling associated with an EMT. Support vector machine and multinomial logistic regression models based on nuclear architecture features yielded AUC-ROC curves of 0.95 and 0.96, respectively, in separating PC3-Epi and PC3-EMT prostate cancer cells in vivo.

CONCLUSIONS

Prostate cancer cells that have undergone an EMT demonstrated an altered nuclear structure. The association of nuclear changes and a mesenchymal phenotype demonstrates quantitative morphometric image analysis may be used to detect cancer cells that have undergone EMT. This morphometric measurement could provide valuable prognostic information in patients regarding the likelihood of [future] metastatic disease.

Protective effect of melatonin on methamphetamine-induced apoptosis in glioma cell line

Methamphetamine (METH) is a highly addictive drug causing neurodegenerative diseases. METH has been known to be neurotoxic by inducing oxidative stress, free radical, and pro-inflammatory cytokines. Previous studies have shown that METH could induce neuron and glial cell death, especially inducing glial cell-mediated neurotoxicity that plays a critical role in stress-induced central nervous system damage. Therefore, the aim of the present study is to explore the mechanisms of METH-induced cell death in the glial cell. METH-induced glial cells death is mediated via mitochondrial damage pathway. METH activates the upregulation of the Bax, cytochrome c, cleavage caspase 9 and 3 proteins, and downregulation of Bcl-XL protein in cascade. Pretreatment with melatonin, a neurohormone secreted by the pineal gland, effectively reduced glial cell death. Moreover, melatonin increased the Bcl-XL/Bax ratio but reduced the level of cytochrome c, cleavage caspase 9 and 3 proteins. Therefore, these results demonstrated that melatonin could reduce the cytotoxic effect of METH by decreasing the mitochondrial death pathway activation in glial cells. This outcome suggests that melatonin might be beneficial as the neuroprotection in neurodegenerative diseases caused by METH or other pathogens.

Role of actin dependent nuclear deformation in regulating early gene expression

The nucleus of a living cell is constantly undergoing changes in shape and size as a result of various mechanical forces in physiology. These changes correlate with alterations in gene expression, however it is unclear whether nuclear deformation alone is sufficient to elicit these alterations. We used T-cell activation as a model system to test the coupling between nuclear deformation (elongation) and gene expression. Naïve T-cell activation with surrogate antigens resulted in actin dependent nuclear elongation. This was accompanied with Erk and NF-κB signaling to the nucleus to induce CD69 expression. Importantly, inhibiting actin polymerization abolished both nuclear elongation and CD69 expression, while inhibiting Erk, NF-κB or microtubule depolymerization only abolished expression but not elongation. Immobilization of antigen-coated beads, under conditions where actin polymerization was inhibited, rescued both nuclear elongation and CD69 expression. In addition, fibroblast cells plated on fibronectin micropatterns of different sizes showed correlation between nuclear shape index and tenascin C expression. Upon inhibiting the signaling intermediate Erk, tenascin C expression was down regulated although the nuclear shape index remained unaltered. Our results highlight the importance of specific signaling intermediates accompanied with nuclear deformation in the modulation of cellular genomic programs.

[Postnatal development of the cerebellar Purkinje cell shape in guinea pig ontogenesis]

In sagittal cerebellum sections, morphometrical study of cerebellum of mature-born animals - guinea pigs - was performed using Nissl's procedure. A change of shape and volume of Purkinje cells and their nuclei in the course of the guinea pig postnatal ontogenesis was studied. It has been shown that both the growth process itself and the rate of formation of the definite form of Purkinje cells and of their nuclei in the course of ontogenesis proceeds non-uniformly. The most intensive growth of vertical and horizontal diameters of Purkinje cells and of their nuclei is observed during the 1st and 4th weeks of postnatal life. Especially rapid is an increase of horizontal diameters of Purkinje cells and of their nuclei, which impairs the ovoid-bear-like shape to the cerebellar Purkinje cells of adult guinea pigs.

Proper detection of the nuclear shape: ways and significance

Shape and size of the nucleus, coupled with changes in chromatin amount and distribution, still remain the basic microscopic criteria for cytological diagnoses. Diagnostic recognition of the nuclear shape in pathological histology and cytology has been always based on the assumption that it is the content in nucleic acids, which determines the nuclear shape. The present review challenges this opinion, focuses on the structure, and functions of the nuclear envelope and on how these features can be exploited in diagnostic pathology. In particular, we will present the contribution of thee-dimensional modeling to the understanding of nuclear irregularities in breast cancer and papillary thyroid carcinomas. Specifically, it will be shown how tagging the nuclear membrane with anti-Emerin antibodies can represent an additional and valuable tool in the differential diagnosis of thyroid lesions. Finally, the prognostic importance of detecting irregularities of the nuclear shape in breast carcinomas by immunofluorescence staining for nuclear proteins will be discussed.

A novel function of plant histone H1: microtubule nucleation and continuous plus end association

In higher plant cells, various microtubular arrays can be seen despite of their lack of structurally defined microtubule-organizing centers (MTOCs) like centrosomes in animal cells. Little is known about the molecular properties of the microtubule-organizing centers in higher plant cells. The nuclear surface contains one of these microtubule-organizing centers and generates microtubules radially toward the cell periphery (radial microtubules). Previously, we reported that histone H1 possessed the microtubule-organizing activity, and it was suggested that histone H1 localized on the nuclear surfaces in Tobacco BY-2 cells (Nakayama, T., Ishii, T., Hotta, T., and Mizuno, K. J. Biol. Chem. (submitted)). Here we show that histone H1 forms ring-shaped complexes with tubulin, and these complexes nucleated and elongated the radial microtubules continuously (processively) associating with their proximal ends where the incorporation of tubulin occurred. Furthermore, the polarity of radial microtubules was determined to be proximal end plus. Immunofluorescence microscopy of the isolated nuclei revealed that histone H1 localized on the nuclear surfaces, distinct from that in the chromatin. These results indicate that radial microtubules are organized by a novel MTOC that is totally different from MTOCs previously found in either plant or animal cells.

Application of Mie theory to determine the structure of spheroidal scatterers in biological materials

We present here the results of a numerical study on light scattering from nonspherical particles with relevance to detecting precancerous states in epithelial tissues. In previous studies of epithelial cell nuclei, the experimental light scattering data have been analyzed by comparison with Mie theory. However, given the spheroidal shape of many cell nuclei, the validity of this assumption demands a thorough investigation. We investigate this assumption by using the T-matrix method to model light scattered from spheroids with parameters relevant to epithelial cell nuclei. In our previous studies, we have developed a data analysis procedure that extracts the oscillatory component of the angular-scattering distribution for an ensemble of epithelial cell nuclei for comparison with Mie theory. We demonstrate that application of our analysis procedure to the predictions of the T-matrix method for spheroids, oriented such that their axis of symmetry is aligned with the incident light propagation direction, generally yields the spheroid dimension that is transverse to the incident light propagation direction with subwavelength accuracy.