Research

Methods

Our pipeline is built on a tight integration of approaches as cells are gene edited, assayed, imaged and modeled.

Stem Cells and Gene Editing

undefined

We have begun our work with human induced pluripotent stem (hiPS) cells because they are diploid, well characterized, propagate in vitro, have a stable karyotype and can differentiate into many other cell types. This unique combination allows us to study multiple cell types as well as the transitions to these differentiated cell types. Working with hiPS cells also allows us to study cells derived from people with different genetic backgrounds. We chose human cells to exploit the enormous genetic information on humans that is now emerging, particularly for disease, in the wake of the Human Genome Project.

Our end goal of imaging live cells and capturing details of their structure, organization, and activities has led us to pursue a genome editing strategy that specifically tags intracellular structures of interest. We use Crispr-Cas9 technology to tag a collection of genes with a fluorescent protein at endogenous loci of interest. We introduce the tag into living cells through electroporation using a ribonuclear protein approach, which enables us to work with better efficiency and lower off-target effects.

These new gene editing technologies introduce reliable markers for the molecular assemblies that comprise the cell and sensors for the major cellular activities. We use these gene edited lines to make valuable observations about cell organization and activities.

We conduct quality control steps during the gene-edited cell line generation process, including genotyping, stem cell marker analysis, karyotype testing, deep sequencing and image-based assays. We also create clonal lines for each gene to minimize variability and ensure that the cell line is consistent for long term experiments. Generating each clonal line takes five to six months.

This year, we hope to generate up to 25 different cell lines for imaging a variety of intracellular structures. These lines will be available to the community for study.

Protocols

We have created a number of protocols to describe our process in creating our cell lines and generating our data.

Assay Development

Imaging living cells is critical since cell organization is dynamic, reflecting the particular state and behavior of the cell under study. Our first project is to understand the physical organization of the cell. We are asking questions such as: Where are all of the cell's major structures? Are there key structures that serve as organizational linchpins? What are the organizational principles of the cell, and what is the hierarchy of the players? How does that system change as the cell goes through its cycle or differentiates, and what are the mechanisms?

To address these questions, we are developing quantitative, image-based assays of cellular organization and activities, both routine and specialized. In doing this, we are leveraging and synergize with the cell biology community that has developed these kinds of assays.

The goal is to interrogate the cell as an integrated system in a variety of settings. We will observe the location and activities of these structures in the living cell both as it goes about its routine activities—dividing, migrating, producing energy, making proteins, etc.—and in its specialized life, for example as a beating heart muscle cell or barrier creating epithelial cell. We will also observe how the cell organization responds, as a system, to genetic perturbation or drugs in changing environments and in disease models.

Determining the cell's coordinate system, thus defining its polarity, is our first step. We are identifying key reference structures, determining the position of secondary structures relative to the reference set, and computationally integrating these structures into a modeled cell. We are using chemical dyes and gene edited cells that illuminate the nucleus and cell periphery as our first set of reference structures and will also incorporate other key reference structures such as the microtubule organizing center, Golgi, or cell junctions. We develop and validate 3D image processing methods to extract key features of each cellular structure at the appropriate level of detail for data analysis and modeling. The resulting computational model will serve as our first digital assay in which we can probe questions about cellular behavior and organization.

Automated Imaging

undefined

We use several methods to capture high resolution images of whole, living cells. Spinning disk microscopes are fast, reliable and are our workhorses to collect our largest data sets. We use the Yokogawa spinning disk CSU-X1 with two sCMOS cameras. We have three of these microscopes in our pipeline and two similar systems that are dedicated to assay development and research and development.

For resolution beyond the diffraction limit of conventional light microscopes, we also employ laser scanning microscopes with point-spread-function sampling, which provides super resolution images of cellular structures with much greater signal to noise ratios. These more flexible microscopes represent the cutting edge of cellular microscopy. We have chosen to use the Zeiss LSM 880 Airyscan with fast acquisition mode. We use two of these microscopes in the imaging pipeline and two similar systems for assay development and research and development.

In our pipeline, a robot prepares 96-well plates of cells and scans them to identify and quantify colonies. The spinning disk microscopes find these colonies and images individual cells in higher resolution, in 3D and over time. This automation is crucial to the pipeline, enabling us to image up to 16,000 cells per day.

We analyze the images in a cluster based image processing pipeline. The output of the imaging and image processing pipeline consists of images of segmented cells, quantitative features of these cells and meta data describing the sample preparation and imaging steps. The quantitative features include the geometry of individual cells, the intensity of signal, texture and information about the cellular neighborhood. Metadata includes the colony history and sample preparation, microscope settings, quality control standards and information about the local environment in which the images were taken.

When possible, we use open-source software, like CellProfiler image analysis software to process the data, and OMERO to organize and store our images. Our coding is primarily done in Python.

Integrated Models

Fluorescence microscopy is the primary means by which we interrogate the spatial organization of subcellular structures. Since only a small number of these structures can be imaged simultaneously, we use computational modeling to integrate these images into a unified understanding of how cells are organized. Automation is also critical, so that we can gather hundreds or thousands of images and build these models with little human intervention.

The computational models we produce will:

  • Represent the physical organization of undifferentiated and differentiated cell states
  • Describe the spatial organization and shape variation of cells, tissues and subcellular structures
  • Predict cell type and state from subcellular structures
  • Provide models for transitions between cellular states
  • Replace cartoons with data-driven representations
  • Serve as a platform to unite modeling across multiple scales

As a first step to building our computational models, we will leverage two software packages developed by team members, CellOrganizer and cellPACK, to establish statistical models for spatial distribution of subcellular components and to construct high resolution, multi-scale representations of cells from multiple image and structural biology sources. Subsequent modeling efforts will address mechanistic questions about organization and transitions using structural models as a platform for whole cell multiscale modeling.

Allen Cell Explorer

Our product, the Allen Cell Explorer (ACE), is deigned to become a multi-scale and dynamic cell biology reference resource. ACE will provide quick access to data, methods, and tools for researchers. Primarily, ACE will highlight results via an interactive image database with online interfaces designed to simplify and enhance both visual analysis and learning of cellular and molecular biology. We intend to create a resource akin to a map that captures cells at multiple spatial levels in all their complexity, and presents them in a quantitative format ripe for analysis, hypothesis generation, and deeper study. To satisfy the demands of researcher while inspiring the next generation of researchers with the wonderment of cell biology, the Allen Cell Explorer comprises two key features: an interactive research portal and educational tools for exploring cell structure and behavior.