Video Bioinformatics Analysis of Human Embryonic Stem Cell Colony Growth

Video bioinformatics is the automated processing, analysis, understanding, and data mining of biological spatio-temporal data extracted from microscopic videos. The purpose of this article is to demonstrate a method for measuring human embryonic stem cell colony growth using a video bioinformatics method.

This protocol demonstrates video bioinformatics method for measuring human embryonic stem cell colony growth by analyzing time-lapse videos collected in an icon bio station CT incubator, equipped with a camera for video imaging. Human embryonic stem cell or HESC colonies are cultured to 70%co fluency replated and incubated for 48 hours after which the HESC colonies are filmed for an additional 48 hours in the Biot ct. The rate of growth is determined using segmentation enhancement and measurement recipes developed with CL quant software.

The recipes are validated by comparison to Adobe Photoshop, an image analysis tool that enables precise manual mask fitting to each colony. The emerging field of video bioinformatics can be used to automate the processing, analysis and mining of biological data from microscopic videos. Hi, I am Serena Lin from Dr.Prou Talbot's lab in the Department of Cell Biology and Neuroscience here at the University of California Riverside.

Hi, I'm Sean Fino, also from the Talbot Lab. Hi, I am TI Satish from the Cell Molecular and Developmental Biology Graduate Program. Today we will be showing you how to measure human embryonic stem cell colony growth using video bioinformatics tools.

We use this procedure in our lab to measure the effects of environmental toxicants on human embryonic stem cell behavior. We'll be demonstrating this procedure in our new stem cell core facility. So let's get started.

To prepare HESC for video analysis, grow HESC On matrigel coated six well plates until 70%confluent aspirate medium from one well containing HESC and rinse two times with one milliliter of PBS. Add one milliliter of Accutane and incubate for one minute at 37 degrees Celsius in 5%carbon dioxide. Next, add 10 to 12 glass beads into the well and shake the plate gently until colonies completely detach.

Neutralize Accutane using one milliliter of human embryonic stem cell maintenance, medium or mouse embryonic fibroblasts conditioned medium. Collect detached cells without the beads into a 15 milliliter conical tube and centrifuge at 200 G for three minutes. Decant the SUP natant and break the pellet with 500 microliters of fresh human embryonic stem cell maintenance medium or M-T-E-S-R plate 100 microliters of the HESC suspension dropwise into multiple wells of a 12 well matrigel coated plate.

Rock the plate back and forth gently and observe cultures under a light microscope to make cell clumps are evenly distributed throughout the wells. Place the plate in an incubator at 37 degrees and 5%carbon dioxide for 48 hours to ensure cells are attached flat and large enough to easily visualize. After 48 hours, aspirate the medium and wash wells with 500 microliters of PBS to remove unattached cells, then add 500 microliters of M-T-E-S-R medium to each.

Well immediately place the plate into the Biot CT and begin collecting time-lapse images. Try to select fields with discreet single colonies not likely to grow into other colonies. In this protocol, cells are video recorded for an additional 48 hours with frames at seven minute intervals.

To create the segmentation recipe first, manually scan the entire video to verify that it contains a single colony that remains focused throughout the recording period. After this, open the segmentation wizard in the CL quants software and click the next button. Select the correct image channel and click the next button.

Pick soft matching and click Next. Select one or two want regions by circling the outer edge to the central area of the colony. This area should be as small as possible, but should be representative of the entire colony.

Do not pick too many regions as this may result in an inaccurate mask application. Click the next button. Select Don't want regions by circling regions that are not part of the colony and not part of the background.

These regions have patterns that should be suppressed, such as debris and dead cells. Click next. Select background by circling regions that are not colonies, debris, or dead cells.

The background should be uniform and likely to show up in every frame. Typically, we select the gray background around the colony. Click the next button.

A colored mask should be displayed over the region of interest if the mask does not accurately cover the region of interest. The segmentation threshold range at the bottom right hand corner of the screen may be increased or decreased if a change of mask is needed. Select updates soft matching regions.

This allows changing the want don't want or background areas. If the mask is satisfactory, select apply threshold and save my mask. Then click the next button.

The mask will then be displayed in a different color. Select finish. The recipe should be displayed on the right upper corner of the software as segmentation recipe.

Rename the recipe if desired by right clicking and inputting a new name. To apply the recipe. Right click and place the mouse over.

Apply recipe. A menu will appear allowing selection of the number of frames to use. Set the frame interval to every 20 frames to assure mask fidelity.

Randomly manually spot check 10%of the frames to which the mask was applied to create an enhancement recipe that eliminates unwanted regions in debris. Right click the enhancement recipe folder and then click new. A new enhancement recipe should be displayed from the original field of view.

Move the mouse over icons in the toolbar to identify the toggle enhancement module button located to the right of the image and click on it. Select labeling dropdown menu. Select labeling for connected select labeling for connected two.

A new task bar should be displayed. Grab initial mask from the segmentation recipe and drag it to the input box. The bar should now display input mask number.

Drag the icon with the input mask number to the mask zero icon. Find min size in task bar and adjust the number until only region of interest is displayed. Make sure to click on execute for each fitting trial.

Once the minimum size adjustment is satisfactory, click on the previous enhancement recipe at the bottom of the screen, select save to recipe. The software now prompts to overwrite selected recipe and select yes spot. Check the enhancement recipe using the same procedure as that of spot checking segmentation.

If satisfied with the segmentation and enhancement recipes, continue on to create the measurement template. To begin select, create measurement template on the right hand toolbar. In the measurement window, select move cursor over clip art cell figure.

Hold the control key and select the cell under the morphology section, uncheck default parameter and check area parameter. Then exit the template window. Select the icon.

Create measurement recipe. Rename the recipe. Move to and select template tab In the upper right corner, drag and drop the previously created measurement template onto the measurement window.

Click yes to continue in the measurement recipe window, select channel mappings. Then whole cell select mask mappings. Then select whole cell option For enhanced mask, select the measurement recipe under the recipe tab.

In the main window, select save icon within the measurement recipe window. Then close the window Next close and reopen the field of view. Right click the recipe list folder and then click new drag recipes in order of the necessary application into the recipe list, select lock recipe list, and run the recipes sequentially on the video data using Adobe Photoshop.

Every 20th frame of the same colonies was analyzed. Manually open a frame of a colony image in Adobe Photoshop. Click on the magic wand tool in the toolbar.

Click on the area around the colony so the entire field is covered except the colony region. Make sure the dotted line around the colony fits right around the periphery of the colony and not inside it. If the dotted line is not around the periphery, change the tolerance value in the upper toolbar accordingly.

Click edit in quick mask mode in the toolbar and click on the colony so the colony is selected. Go to the window, pull down menu and click histogram. The cache should be set to one.

If the cache is two, click the expression mark exclamation to change the cache to one. Record the pixel value into a spreadsheet. Repeat the above process for every 20th frame.

Data collected using Photoshop and CL quant software can then be plotted together. Our protocol for quantifying HESC colony growth demonstrates one application of video bioinformatics to a biological problem. Here we show a graph comparing the increase in colony size over 48 hours as determined using CL quant software and adobe Photoshop.

The graph shows five different colonies each analyzed by both software packages. Each colony is represented by a different color. The solid lines were derived using sail quant software while the dotted lines were collected using Photoshop.

The graph of the raw data shows that both methods of measurement are in good agreement to account for the fact that colonies have a different starting size. This figure shows the data re plotted as the percent increase in colony size. The growth rates are similar regardless of starting colony size and both measures of analysis give similar results.

The following graph displays the means of the previous percent change data on HESC colony growth. This graph clearly shows good agreement between the analysis done using video bioinformatics and Adobe Photoshop. We've just shown you how to measure human embryonic stem cell colony growth using video bioinformatics tools.

When doing this procedure using phase contrast microscopy, be sure to include the halo around the colony for precise colony selection by the software. It is also important to remember that other methods for culturing, incubating and video collecting can be used in place of the protocol we demonstrated. Once the recipes are developed and validated, they will perform analysis far more rapidly and with less experimental variation than manual analysis.

So that's it. Thanks for watching and good luck with your experiments.