Through the Looking Glass: Time-lapse Microscopy and Longitudinal Tracking of Single Cells to Study Anti-cancer Therapeutics

Summary

Here, we describe a method of long-term time-lapse microscopy to longitudinally track single cells in response to anti-cancer therapeutics.

Abstract

The response of single cells to anti-cancer drugs contributes significantly in determining the population response, and therefore is a major contributing factor in the overall outcome. Immunoblotting, flow cytometry and fixed cell experiments are often used to study how cells respond to anti-cancer drugs. These methods are important, but they have several shortcomings. Variability in drug responses between cancer and normal cells, and between cells of different cancer origin, and transient and rare responses are difficult to understand using population averaging assays and without being able to directly track and analyze them longitudinally. The microscope is particularly well suited to image live cells. Advancements in technology enable us to routinely image cells at a resolution that enables not only cell tracking, but also the observation of a variety of cellular responses. We describe an approach in detail that allows for the continuous time-lapse imaging of cells during the drug response for essentially as long as desired, typically up to 96 hr. Using variations of the approach, cells can be monitored for weeks. With the employment of genetically encoded fluorescent biosensors numerous processes, pathways and responses can be followed. We show examples that include tracking and quantification of cell growth and cell cycle progression, chromosome dynamics, DNA damage, and cell death. We also discuss variations of the technique and its flexibility, and highlight some common pitfalls.

Introduction

Live-cell microscopy and longitudinal tracking of single cells is not a new technique. From the earliest microscopes, enthusiasts and scientists have observed and studied single cells and organisms, their behaviors, and development^1-3. A famous example from the late David Rogers at Vanderbilt University in the 1950s shows a human neutrophil in a blood smear chasing a Staphylococcus aureus bacterium and eventually the process of phagocytosis⁴. This live-cell movie is an excellent illustration of how multiple processes can be observed and correlated in a single experiment: sensing of a chemical gradient, mechanics and speed of cell motility, cell shapes dynamics, adhesion, and phagocytosis of a pathogen.

The advent of fully automated microscopes and highly sensitive digital cameras has resulted in an increasing numbers of investigators using microscopy to ask fundamental questions in cell biology ranging from how cells move^5,6 and divide^7,8 to organelle dynamics and membrane trafficking^9-11. Non-fluorescent, brightfield microscopy, including phase-contrast (PC), which garnered the Nobel Prize for Frits Zernike in 1953, and differential interference contrast (DIC) allow for the observation of cells and nuclei but also sub-cellular structures including microtubule bundles, chromosomes, nucleoli, organelle dynamics, and thick actin fibers¹². Genetically encoded fluorescent proteins and the development of fluorescent dyes against organelles have dramatically impacted time-lapse microscopy^13-15. While not the focus of this article, imaging in cell spheroids and in situ (intravital microscopy) using confocal and multiphoton microscopy represent another expansion of the approach, and there are outstanding articles that use and discuss these approaches^16-19.

The responses of cells to anti-cancer drugs or natural products are determined on the molecular and cellular scale. Understanding cell responses and fates following treatment often involves population averaging assays (e.g., immunoblotting, whole well measures), or fixed time-points with immunofluorescent detection and flow cytometry, which measure single cells. Heterogeneity in single cell responses to drugs within a population, in particular in tumors, may explain some of the variability in response seen across cell lines and tumors that are treated with the same drug at saturation. Long-term longitudinal approaches to follow a given single cell or a population of cells is a less common but very powerful approach that allows for the direct study of molecular response pathways, different phenotypes (e.g., cell death or cell division), observation of cell-to-cell variability within a population, and how these factors contribute to population response dynamics^20-22. Optimistically, being able to observe and quantify single cell responses will help improve our understanding of how drugs work, why they sometimes fail, and how to best use them.

The technique of long-term time-lapse microscopy, longitudinal tracking, and analysis of drug responses is available to many investigators and can be simple, using only transmitted light to observe phenotype responses^20,21. The main components of the approach include: appropriate preparation of the cells of interest, an automated microscope with environmental chamber, a camera integrated with a computer to acquire and store the images, and software to review the time-lapse and measure and analyze the cells and any fluorescent biosensors. We provide a detailed protocol with many tips for conducting time-lapse microscopy of cultured cells for as long as several days using brightfield and/or widefield epifluorescent microscopy. This protocol can be used for any cell line that can be grown in culture to study their responses to anti-cancer therapies. We provide examples of data acquired and analyzed using multiple different genetically-encoded fluorescent biosensors and an example of phase-contrast microscopy, briefly discuss different types of probes, the advantages and disadvantages of long-term time-lapse and longitudinal tracking, what can be learned for this approach that is difficult to understand from non-direct approaches, and some variations that we hope will be of interest and value to inexperienced researchers who have not considered using the approach, and to experienced researchers.

Access restricted. Please log in or start a trial to view this content.

Protocol

 The following protocol uses parameters defined by the experiments in Figures 4 and 6 regarding acquisition settings and experimental conditions. Many of these parameters can be modified to fit other experiments (i.e., exposure times, binning, fluorescent channels, etc.). All procedures must adhere to institutional guidelines and regulations and be approved by the institutional biosafety committee. Microscope manufacturer websites contain excellent information for live cell imaging.

1. Microscopes and Imaging Software

1. Perform live cell imaging on a wide variety of inverted microscopes. The most common microscopes are widefield epifluorescence and spinning-disc confocal. Here, use an automated and motorized widefield epifluorescence microscope with 200 watt metal halide light source.
2. Obtain a stage-top or microscope environmental chamber. Here, use a stage-top environmental chamber.
3. Use commercial software to operate microscopes with utilities to execute time-lapse microscopy.
4. Use commercially available software or ImageJ for image analysis. Many other analysis tools are available commercially, and there are custom-made programs many of which have been published^8,18.

2. Visualizing Cellular Processes and Phenotypic Responses

1. Visualize cells with brightfield microscopy. Differential interference contrast and phase contrast alone can be very informative to study responses to anti-cancer drug responses. These processes can include mitosis, cell motility and apoptosis.
2. Visualize cell structures, organelles and processes with fluorescent biosensors. Fluorescent probes are informative in tracking specific subcellular processes. These can include microtubule dynamics, mitochondrial and endoplasmic reticulum dynamics, protein accumulation and localization, and molecular signaling (e.g., phosphorylation, calcium).

3. Preparation of Samples

1. Grow cells in cell culture certified dishes in a standard, humidified cell culture incubator (e.g., 37 ^oC, 5% CO_2, 80% relative humidity).
   1. Grow HT1080 cells in MEM with EBSS. Supplement the media with 10% v/v FBS, 1% v/v Pen/Strep, 1% v/v sodium pyruvate and 1% v/v non-essential amino acids.
2. Two days prior to imaging, plate 50,000 HT1080 FUCCI cells into 3 wells of a 12-well #1.5 glass bottom dish in a certified sterile laminar-flow hood. Adjust the number of cells plated to achieve ~60% confluence for the start of the experiment. Depending on the cell line and nature of experiment the cell density may be less.
   Note: Cell density may have profound influence on the growth of the cell line and on the experimental results. Count cells to minimize experiment-to-experiment variability due to density.
   1. Depending on the environmental chamber and the necessary conditions for the experiment, plate cells in single well dishes (typically 35- or 60-mm), 4 well 35-mm dishes, 6-, 12- or 24-well cell culture plates, or coverslip slides in various formats. Use glass-bottom plates with #1.5 glass as most objective lenses are optimized for this thickness of glass, and imaging through cell culture plastic is very poor.
      Note: Some cell lines do not grow and survive on glass. In these cases, the glass can be coated in order to enhance cell adherence (e.g., poly-lysine or collagen). Some companies manufacture optical cell culture plastic, it must be determined empirically if this is a viable option.

4. Environmental Chamber Set-up

1. Prior to any experimentation, set-up the environmental chamber so that it operates at ~80% humidity and so the temperature at the sample position is 37 ^oC. Most cultured cells grow in 5% CO₂/balance air atmosphere due to sodium bicarbonate buffer in the medium. Depending on the set-up, either set the environmental controller to 5% CO₂ and it will mix 100% CO₂ with air, or use pre-mixed, certified 5% CO₂/balance air gas. Follow manufacturer directions for gas flow-rates.
   Note: Some cell lines grow in CO₂-independent medium in which case they can be maintained without CO₂. Specifically designed imaging media is also available that limits the addition of autofluorescent compounds. Growth in theses mediums for the desired cell type must be determined empirically beforehand.
2. Prior to imaging, ensure the water reservoir is filled (following manufacturer's directions) with sterile distilled water. Turn on the environmental chamber to the desired temperature setting and place it into the microscope stage inset. To save gas, do not start the flow at this point.
   Note: If there is any free space between the stage-top chamber and stage inset and/or any tension on connection cords to the chamber it may introduce motion artifacts into the experiment.
   1. Lay pieces of paraffin film over the edges of the stage inset opening prior to inserting the chamber into it to couple the chamber to the stage, and make sure that no connections are pulling on the chamber.
3. Allow the chamber to equilibrate to 37 ^oC. Maintain a stable temperature to prevent temperature fluctuations during imaging which can affect cell physiology and introduce drift. Reaching temperature equilibration typically requires 30 min to 1 hr, depending on the environment.
   1. Insert a 'dummy' dish with water in the wells into the chamber while warming. An imaging dish filled with water mimics the sample, helping ensure the chamber is sufficiently warmed and stabilized. Filling the chamber with pre-warmed sterile water reduces the time required to temperature stabilization and minimizes the temperature decrease upon addition of the experimental sample.

5. Microscope Set-up

1. Turn on the microscope, associated computer, and any required peripherals.
   1. Depending on the light source, wait to turn it on until needed (e.g., LED).
2. With the objective turret in a low position, select the 20X 0.7 NA objective to be used.
3. Position the sample over the objective - this will make it easier to find the cells when the sample is in the chamber.
4. Define imaging parameters at this time if they are known from previous experiments.

6. Transporting Cells to Microscope and into the Chamber

1. Transport the cells to be imaged from the incubator to the environmental chamber. Ensure that this is done quickly to limit the effects of a comparatively low CO₂ atmosphere and decreased temperature on the cells. Follow all institutional biosafety guidelines for sample transport and clean-up in case of a spill.
   1. Place the cells in a sealed Styrofoam container or insulated bag to avoid large environmental changes.
2. Insert the sample imaging dish into the chamber following manufacturer's directions.
3. After the cells have been secured within the environmental chamber, seal the chamber to maintain a stable environment and immediately turn on the source(s) of atmospheric gas.
   1. As some environmental chambers do not utilize a tight, sealed lid, to retard evaporation, layer sterile, embryo-certified mineral oil on top of the growth medium. Wrap gas-permeant paraffin film around the perimeter-edges of the sample dish being careful not to obstruct the glass imaging area. Localized, secondary humidification methods can be employed. Condensation on the lid of the sample dish can diminish performance of brightfield microscopy.
4. In order to minimize the effects of thermal drift early during the experiment, wait 30 min before starting the time-lapse. The required time varies based on imaging dish size and other factors. Determine empirically.

7. Setting up the Imaging

1. Select imaging conditions that will best represent the data. Take precautions to avoid phototoxicity by limiting exposure times, using lower intensity light and selecting probes that are excited by longer wavelengths of light.
2. Define x, y and z-plane and desired wavelength(s) for each position to be imaged. The time resolution is limited by the number of positions and wavelengths. For long-term time-lapse of most cellular processes, acquire one image every 10 - 20 min; higher time resolution (short intervals, e.g.,1 min) provides more data points and therefore more robust cell tracking, but also results in more integrated light exposure and larger data sets.
   1. As HT1080 FUCCI have dynamic green and red fluorescent proteins (see Representative Results), use the following exposure times: FITC/GFP - 50 msec, Texas Red/TRITC - 40 msec, Brightfield - 20 msec. Use a 2 x 2 bin.
3. Enable software controlled autofocusing using the default parameters under advanced settings. Define an autofocus range of 10 µm with the recommended step size. Be sure to do this before starting the time-lapse. Autofocus with brightfield images and never with fluorescence to reduce phototoxicity and photobleaching.
   1. Use software controlled autofocusing systems on any microscope with a motorized stage, and directly focus on the sample. However, they can limit the acquisition speed of the experiment. Hardware controlled continuous focusing systems work well for time-lapse microscopy and improve speed, allowing for more positions to be imaged or a higher rate of acquisition. However, they rely on detection of the air-glass interface and may lose focus of the sample with variations in glass thickness across the well.
4. Replace half of the media with media containing the desired drug concentration that has been warmed to 37 ^oC. Partial media replacement helps to reduce thermal drift. The conditions in this experiment are Vehicle (DMSO), 1 µM selinexor and 10 µM PD0332991.
   Note: This step can also be done after starting the acquisition by pausing and restarting the experiment. This allows for longitudinal tracking of pre- and post-drug responses of single cells. If drug stability or metabolism is a concern, pausing and replacing the media can be done with the same method. To test for drug degradation or metabolism, media from treated cells can also be placed on naïve cells to test drug action.
5. Start the time-lapse.
6. As the time-lapse runs, ensure that the imaged fields remain in focus and the chamber maintains a humid 37 ^oC.
   1. Adjust focus as needed by pausing the acquisition during the time gap between time points.
7. If necessary, add water to the chamber during longer experiments. Avoid cooling the chamber by adding pre-warmed, sterile distilled water at 37 ^oC.

8. Ending the Time-lapse

1. When the experiment has run to completion, stop the acquisition if it was not set to stop automatically.
2. Ensure that the time-lapse has been saved properly onto the hard drive, although most software packages automatically save during the acquisition if not finished properly data can be lost.
3. Remove the chamber from the microscope and dispose of the cells into biohazardous waste following approved institutional biosafety procedures.

9. Longitudinal Tracking and Analysis of Time-lapse Data

1. Choose the methodology for analysis that is appropriate for the biological processes of interest. Many plugins for ImageJ, programs using MatLab, and custom platforms have been produced for specific applications. The following methodology covers tracking of nuclei and analysis of nuclear fluorescent probes as demonstrated in Figures 4 and 5.
2. Open the .tif image stacks for the utilized channels in ImageJ. Alternatively, open native files from the acquisition program directly in ImageJ using the Bioformats plugin.
3. Draw a region of interest (ROI) within the nucleus of one cell using the brightfield image or the nuclear label channel (if used) and add it to the ROI manager. Note: If the imaged cells have a nuclear label (e.g., histone H2B-EGFP), then automatic cell tracking may be utilized to create regions of interest (ROIs) representing individual nuclei through time.
4. Proceed to the next time point and position the ROI within the same nucleus. Add the ROI to the ROI manager.
5. Continue to track and make ROIs for the individual cell until there is a cellular event (mitosis, apoptosis, etc.) or the cell can no longer be tracked (i.e., moves out of frame or the experiment ends).
6. When ROIs have been established for cells through the time they are in the field, superimpose them onto the fluorescent channels of interest. Measure the mean intensity of the fluorescence in each channel for every time point. Save the ROI list for later use.
   1. Various measurements can be made within the ROI for applications in different experiments. Click Analysis → Set Measurements… to bring up a menu to choose the analytical method (e.g., mean intensity within the ROI, integrated intensity, etc.).
7. Note cell fates for individual cells (e.g., apoptosis, division, survival) while tracking. This allows for creation of survival curves and further population analysis.
8. With the mean intensity for both channels, create a scatter plot to display cell cycle dynamics in response to therapeutics (Figure 5B, C).
   Note: Plots can be created for individual cells over time or averaged over a population of tracked cells. Quantitative imaging can be critical for establishing complex responses and cell relationships in response to cancer therapeutics. Long-term time-lapse microscopy, longitudinal tracking and data analysis is a multi-step process, with many options for the type of microscopy and analysis tools, and follows the general outline provided in Figure 1.

Access restricted. Please log in or start a trial to view this content.

Results

Long-term time-lapse microscopy and direct longitudinal tracking allows for the study of many anti-cancer effects during drug response. Following the general outline in Figure 1, multiple examples of cells are shown expressing validated fluorescent reporters that treated with anti-cancer drugs, tracked, and analyzed using different approaches.

Phase contrast microscopy alone is very informative and robustly repo...

Access restricted. Please log in or start a trial to view this content.

Discussion

Advantages of Time-lapse Microscopy and Longitudinal Tracking

The microscope is an ideal instrument for longitudinal studies of drug response as it allows investigators to track individual cells and their fates as well as the entire population. Variability in drug response within a population of cells is a major issue for anti-cancer therapeutic design. Longitudinal tracking of single cells allows investigators to observe this variability and begin to understand the underlying...

Access restricted. Please log in or start a trial to view this content.

Materials

Name	Company	Catalog Number	Comments
Taxol (paclitaxel)	Sigma	T7191	microtubule stabilizing drug
Etoposide	Selleckchem	S1225	topoisomerase II inhibitor
Selinexor	Karyopharm Therapeutics	na	XPO1/CRM1 inhibitor, gift
Kinesin-5 inhibitor	Merck Serono	na	gift, also available from American Custom Chemicals Corporation. CAS 858668-07-2
Cell growth medium	HyClone (Fisher) or Mediatech		many companies available
5% CO2/balance air, certified	Airgas	Z03NI7222004379
35 mm Dish, 20 mm glass bottom	Cellvis	D35-20-1.5-N	many companies available
35 mm 4-well Dish, 20 mm glass bottom	Cellvis	D35C4-20-1.5-N	many companies available
35 mm Dish, gridded glass bottom	MatTek	P35G-2-14-CGRD	many companies available
Multi-well, glass bottom	Cellvis	P12-1.5H-N	many companies available
Olympus IX81 inverted epifluorescence microscope	Olympus
Olympus IX2-UCB controller	Olympus
PRIOR LumenPro200	Prior Scientific	Lumen200PRO
PRIOR Proscan III motorized stage	Prio Scientific	H117
STEV chamber	InVivo Scientific	STEV.ECU.HC5 STAGE TOP
Environmental Controller Unit	InVivo Scientific	STEV.ECU.HC5 STAGE TOP
Hamamatsu ORCA R2 CCD with controller	Hamamatsu	C10600
Nikon Eclipse Ti	Nikon
Nikon laser launch	Nikon
SOLA light engine	lumencor
iXon Ultra 897 EM-CCD	ANDOR
TOKAI HIT inclubation chamber	TOKAI HIT	TIZSH