Review Article

Modeling to Optimize Terminal Stem Cell Differentiation

Table 4


Preliminary step: Western Blots, RT-PCR, IFResearch goals and progress

MarkersAll (Tables 1 and 3) Confirm that all markers are expressed and differentiation of all cell types is occurring; replace any antibodies that do not work; revise differentiation protocol if necessary
Time courseEach day, d0–d10 Determine the peak day for expression of each marker
HypothesisNanog decreases by d2. Endoderm peaks at d3-d4. Mesoderm and ectoderm peak between d6-d7 Determine a day to be used as + and – control for flow cytometry
Possible problems and solutionsMarkers are expressed in undifferentiated cells— they cannot use ES cells as negative control Determine whether markers for each cell type are expressed at similar times or have varying temporal expression

Step 1: Flow cytometry 1Research goals

MarkersDifferentiated versus undifferentiated
(Table 2)
Using + and – controls from preliminary step, confirm that all antibodies work for flow cytometry; replace any antibodies that do not work
Time courseEach day, d0–d10, Determine the proportion of all cells that actually differentiate on any given day of the time course
HypothesisEarly on, the curve will resemble an exponential growth curve with a plateau between d8-d9 and a decrease on d10 Fit data to a curve to model differentiation
Possible problems and solutionsThe curve may be defined by a complex equation or there may not be a peak on the curve—Step 1 is not useful, proceed to Step 2 Use the curve and data from the preliminary step to estimate the peak day and standard deviation (SD) for differentiation of each cell fate

Step 2: Flow cytometry 2Research goals

Markers usedAll (Tables 1 and 3) Determine the peak day for differentiation of each cell type by cycling through iterations of data collection
Time coursePeak day ± SD as determined in step 1; 24 hour intervals for iteration 1; 12 hour intervals for iteration 2After iteration 1, the time between intervals and the time course being analyzed will be narrowed, and the number of replicates conducted will be increased
HypothesisPeak day for each cell type will be different; endo = d3; ecto = d7; meso = d7 Generate a comprehensive probabilistic model of ES cell differentiation over the course of time
Possible problems and solutionsDifferentiation of germ layers can’t be modeled—the experiment will be conducted with late markers of differentiated cell types

Step 3: Flow cytometry and STAT3 functionResearch goals

Markers UsedAll (Tables 1 and 3); GFP ES cell line Use the model generated in Step 2 as a tool to analyze the effect of STAT3 on differentiation
Time coursePeak day ± SD from the model generated in Step 2 Determine the peak day for differentiation of each cell type in cells that express dominant negative STAT3
HypothesisThe absence of STAT3 will prevent differentiation of mesoderm and will delay or decrease ectoderm differentiation Use statistical analysis (repeated measure ANOVA) to determine whether the loss of STAT3 leads to a change in the progression of ES cell differentiation over time
Possible problems and solutionsSTAT3 has no effect on differentiation—a new target for modification of differentiation will be chosen Confirm our results using GFP ES cell line and survival analysis