Application of Artificial Intelligence/Machine Vision & Learning for the Development of a Live Single-cell Phenotypic Biomarker Test to Predict Prostate Cancer Tumor Aggressiveness.

To assess the usefulness and applications of machine vision (MV) and machine learning (ML) techniques that have been used to develop a single cell-based phenotypic (live and fixed biomarkers) platform that correlates with tumor biological aggressiveness and risk stratification, 100 fresh prostate samples were acquired, and areas of prostate cancer were determined by post-surgery pathology reports logged by an independent pathologist. The prostate samples were dissociated into single-cell suspensions in the presence of an extracellular matrix formulation. These samples were analyzed via live-cell microscopy. Dynamic and fixed phenotypic biomarkers per cell were quantified using objective MV software and ML algorithms. The predictive nature of the ML algorithms was developed in two stages. First, random forest (RF) algorithms were developed using 70% of the samples. The developed algorithms were then tested for their predictive performance using the blinded test dataset that contained 30% of the samples in the second stage. Based on the ROC (receiver operating characteristic) curve analysis, thresholds were set to maximize both sensitivity and specificity. We determined the sensitivity and specificity of the assay by comparing the algorithm-generated predictions with adverse pathologic features in the radical prostatectomy (RP) specimens. Using MV and ML algorithms, the biomarkers predictive of adverse pathology at RP were ranked and a prostate cancer patient risk stratification test was developed that distinguishes patients based on surgical adverse pathology features. The ability to identify and track large numbers of individual cells over the length of the microscopy experimental monitoring cycles, in an automated way, created a large biomarker dataset of primary biomarkers. This biomarker dataset was then interrogated with ML algorithms used to correlate with post-surgical adverse pathology findings. Algorithms were generated that predicted adverse pathology with >0.85 sensitivity and specificity and an AUC (area under the curve) of >0.85. Phenotypic biomarkers provide cellular and molecular details that are informative for predicting post-surgical adverse pathologies when considering tumor biopsy samples. Artificial intelligence ML-based approaches for cancer risk stratification are emerging as important and powerful tools to compliment current measures of risk stratification. These techniques have capabilities to address tumor heterogeneity and the molecular complexity of prostate cancer. Specifically, the phenotypic test is a novel example of leveraging biomarkers and advances in MV and ML for developing a powerful prognostic and risk-stratification tool for prostate cancer patients.

Probing Mechanical Properties of Brain in a Tuberous Sclerosis Model of Autism.

Causes of autism spectrum disorders (ASD) are understood poorly, making diagnosis and treatment challenging. While many studies have investigated the biochemical and genetic aspects of ASD, whether and how mechanical characteristics of the autistic brain can modulate neuronal connectivity and cognition in ASD are unknown. Previously, it has been shown that ASD brains are characterized by abnormal white matter and disorganized neuronal connectivity; we hypothesized that these significant cellular-level structural changes may translate to changes in the mechanical properties of the autistic brain or regions therein. Here, we focused on tuberous sclerosis complex (TSC), a genetic disorder with a high penetrance of ASD. We investigated mechanical differences between murine brains obtained from control and TSC cohorts at various deformation length- and time-scales. At the microscale, we conducted creep-compliance and stress relaxation experiments using atomic force microscope(AFM)-enabled indentation. At the mesoscale, we conducted impact indentation using a pendulum-based instrumented indenter to extract mechanical energy dissipation metrics. At the macroscale, we used oscillatory shear rheology to quantify the frequency-dependent shear moduli. Despite significant changes in the cellular organization of TSC brain tissue, we found no corresponding changes in the quantified mechanical properties at every length- and time-scale explored. This investigation of the mechanical characteristics of the brain has broadened our understanding of causes and markers of TSC/ASD, while raising questions about whether any mechanical differences can be detected in other animal models of ASD or other disease models that also feature abnormal brain structure.

Characterizing Multiscale Mechanical Properties of Brain Tissue Using Atomic Force Microscopy, Impact Indentation, and Rheometry.

To design and engineer materials inspired by the properties of the brain, whether for mechanical simulants or for tissue regeneration studies, the brain tissue itself must be well characterized at various length and time scales. Like many biological tissues, brain tissue exhibits a complex, hierarchical structure. However, in contrast to most other tissues, brain is of very low mechanical stiffness, with Young's elastic moduli E on the order of 100s of Pa. This low stiffness can present challenges to experimental characterization of key mechanical properties. Here, we demonstrate several mechanical characterization techniques that have been adapted to measure the elastic and viscoelastic properties of hydrated, compliant biological materials such as brain tissue, at different length scales and loading rates. At the microscale, we conduct creep-compliance and force relaxation experiments using atomic force microscope-enabled indentation. At the mesoscale, we perform impact indentation experiments using a pendulum-based instrumented indenter. At the macroscale, we conduct parallel plate rheometry to quantify the frequency dependent shear elastic moduli. We also discuss the challenges and limitations associated with each method. Together these techniques enable an in-depth mechanical characterization of brain tissue that can be used to better understand the structure of brain and to engineer bio-inspired materials.

Onset of heterogeneity in culture-expanded bone marrow stromal cells.

Inconsistencies among in vitro and in vivo experiments using adult mesenchymal stem cells (MSCs) confound development of therapeutic, regenerative medicine applications, and in vitro expansion is typically required to achieve sufficient cell numbers for basic research or clinical trials. Though heterogeneity in both morphology and differentiation capacity of culture-expanded cells is noted, sources and consequences are not well understood. Here, we endeavored to observe the onset of population heterogeneity by conducting long-term continuous in vitro observation of human adult bone marrow stromal cell (BMSC) populations, a subset of which has been shown to be stem cells (also known as bone marrow-derived MSCs). Semi-automated identification and tracking of cell division and migration enabled construction of cell lineage maps that incorporated cell morphology. We found that all BMSCs steadily grew larger over time; this growth was interrupted only when a cell divided, producing two equally sized, morphologically similar daughter cells. However, a finite probability existed that one or both of these daughters then continued to increase in size without dividing, apparently exiting the cell cycle. Thus, larger BMSCs are those cells that have exited the normal cell cycle. These results hold important implications for MSC in vitro culture expansion and biophysical sorting strategies.

Directional cell migration in an extracellular pH gradient: a model study with an engineered cell line and primary microvascular endothelial cells.

Extracellular pH (pH(e)) gradients are characteristic of tumor and wound environments. Cell migration in these environments is critical to tumor progression and wound healing. While it has been shown previously that cell migration can be modulated in conditions of spatially invariant acidic pH(e) due to acid-induced activation of cell surface integrin receptors, the effects of pH(e) gradients on cell migration remain unknown. Here, we investigate cell migration in an extracellular pH(e) gradient, using both model α(v)ß(3) CHO-B2 cells and primary microvascular endothelial cells. For both cell types, we find that the mean cell position shifts toward the acidic end of the gradient over time, and that cells preferentially polarize toward the acidic end of the gradient during migration. We further demonstrate that cell membrane protrusion stability and actin-integrin adhesion complex formation are increased in acidic pH(e), which could contribute to the preferential polarization toward acidic pH(e) that we observed for cells in pH(e) gradients. These results provide the first demonstration of preferential cell migration toward acid in a pH(e) gradient, with intriguing implications for directed cell migration in the tumor and wound healing environments.