Assess the differentiation potential of human pluripotent stem cells

Alexander M. Tsankov; Broad Institute, Harvard University

(See a list of the products featured in this article.)

Human pluripotent stem cells (hPSCs) hold great promise for tissue engineering, regenerative medicine, and disease modeling [1]. The number of hPSC lines has dramatically increased in the past decade, which has created a need for hPSC quality standards that can ensure comparable results across laboratories [2]. We recently introduced a qPCR-based ScoreCard assay that uses gene expression signatures to quantify the differentiation potential of hPSC lines [3]. The improved ScoreCard method enables a rapid, more reproducible assessment of functional pluripotency than the teratoma assay and allows for a wider range of applications than previous genomic approaches, including screening of small molecules, quantifying perturbations of lineage regulators, and assessing different culture conditions.

Comparison of ScoreCard and teratoma assays

Formation of teratomas in mice is the most frequently used assay for quantifying the differentiation potential of hPSCs. However, teratoma generation is a very costly, time-consuming, and variable assay [2,4] and is not an efficient way to assess the quality of thousands of new cell lines (Figure 1A). To circumvent these issues, genomic approaches that instead use gene expression signatures to quantify pluripotency have emerged. PluriTest uses microarray measurements to assess with great accuracy the molecular signatures of pluripotency of a new cell line against a large database of hPSC lines [5]. Also, the original ScoreCard assay utilized the NanoString™ nCounter™ gene expression technology to evaluate both molecular and functional pluripotency [6], defined as differentiation into the three germ layers.

We have developed a qPCR-based ScoreCard assay (available commercially as the Applied Biosystems™ TaqMan® hPSC Scorecard™ Panel), which presents several advantages over previous approaches [3]. The assay is highly accessible to all labs with qPCR machines and more cost-effective than the previous ScoreCard assay. The gene panel leverages recent data on genome-wide expression of early germ layer differentiation [7] to improve the uniqueness of marker genes. We also improved the statistical analysis using the weighted Z-method, which combines information across multiple genes in a weighted, assay-dependent manner while also taking into account dependencies between genes [8]. The weighted Z-method thus enables a wider array of applications than the previous ScoreCard assay and provides a statistical measure of functional pluripotency.

We compared the predictive power of cell line differentiation potential as quantified by the teratoma assay and by the qPCR-based ScoreCard assay following embryoid body (EB) differentiation. We calculated the variance in assay scores between different cell lines and within replicates from the same cell lines. The between-group and within-group variances for the teratoma-predicted differentiation potential were very similar when quantified by an independent pathologist and when using gene expression signatures of our panel (Figure 1B, left). In contrast, we found that the EB differentiation potential scores had a significantly lower within-replicate variance than between–cell line variance (P < 0.0005, F-tests), even after culturing the replicates for more than 10 passages (Figure 1B, right). These results show that EB differentiation potential as quantified by the qPCR-based ScoreCard assay is a more quantitative and reproducible measure of a cell line’s germ layer propensity than the teratoma assay.

Figure 1. Comparison of teratoma and qPCR-based ScoreCard assays. (A) Schematic of the timelines for teratoma formation (top) and qPCR expression assay (bottom) for assessing hPSC utility (EC = ectoderm, ME = mesoderm, EN = endoderm, EB = embryoid body, and PL = pluripotency). (B) Ratio of between-group to within-group variance for germ layer differentiation potential as quantified by teratoma formation (left) and by the qPCR-based ScoreCard assay (right). Germ layer variance ratios are shown using different colored bars, and the asterisks above bars indicate a significantly lower variance between replicates than between cell lines (P < 0.0005, F-test). Reprinted by permission from Macmillan Publishers Ltd: Tsankov AM, Akopian V, Pop R et al. (2015) A qPCR ScoreCard quantifies the differentiation potential of human pluripotent stem cells. Nat Biotechnol 33:1182–1192.

New applications of the qPCR-based ScoreCard assay

The ScoreCard differentiation potential of cells lines based on the weighted Z-method correlates highly (R ≥ 0.83, P < 10–3, Pearson correlation) with established measures of directed differentiation efficiency (Figure 2A). This high proportionality between the ScoreCard assay’s measure of differentiation potential and germ layer efficiency enables several new applications of the qPCR-based ScoreCard assay, including assessing the effects of culture conditions, small molecules, and knockdown of key transcriptional regulators.

Using the qPCR-based ScoreCard assay, we observed a substantial difference in the gene expression signatures of 11 hPSC lines grown both on mouse embryonic fibroblast (MEF) feeder cells and in feeder-free culture conditions. Figure 2B shows that pluripotency markers were more highly expressed in hPSC lines grown in feeder-free conditions (P = 7 × 10–4, weighted Z-method), while markers of the three germ layers were more highly expressed in cell lines grown on feeders. This result suggests that a feeder culture introduces higher background differentiation, possibly due to differences in signaling [9]. We further observed that cell lines adapted on feeder-free culture for several passages had an even greater reduction of endoderm marker expression (Figure 2C).

The ScoreCard assay also allowed us to quantify the effect of different small molecules on endoderm differentiation. We found that replacing recombinant protein WNT3A with the less costly LiCl molecule did not affect the differentiation potential of cell line HUES64 (Figure 2D). However, compound IDE1 decreased LEFTY1 expression and endoderm differentiation potential [10].

In addition, we used the ScoreCard assay to quantify the effect of knocking down key lineage regulators. We knocked down transcription factor OTX2 in undifferentiated hPSCs using three distinct short hairpin RNAs (shRNAs) and observed lower overall activation of ectoderm marker genes following directed differentiation towards ectoderm (Figure 2E), supporting the hypothesis that OTX2 plays a key role in establishing early ectoderm cell fate [11,12]. We also observed a higher overall expression of mesoderm markers in the OTX2 knockdown ectoderm cells (Figure 2E, bottom), suggesting that OTX2 may act as a repressor of key mesoderm genes.

Figure 2. New applications of the qPCR-based ScoreCard assay. (A) Linear regression shows a high correlation between directed differentiation potential and traditional measures of ectoderm (dEC), mesoderm (dME), and endoderm (dEN) efficiency, using FACS quantification for established cell surface markers CD56 and CD184. (B) Box plot of the distribution of mean expression difference between feeder-free and feeder-cultured hPSC lines for all genes belonging to the four gene classes (EC = ectoderm, ME = mesoderm, EN = endoderm, and PL = pluripotency). (C) Box plots of the distribution of mean expression difference between adapted and unadapted hPSC lines for all genes belonging to the four gene classes. EN mean expression decreases after adaptation of lines for 1+ passages (left) and 6+ passages (right) in feeder-free culture. (D) Heatmaps showing gene expression level (left) and differentiation potential2D (right) of several different protocols for endoderm differentiation. WNT = WNT3A, AA = activin A, LiCl = lithium chloride, IDE1 = inducer of definitive endoderm-1, R&D = R&D Systems, TFS = Thermo Fisher Scientific. (E) Heatmaps showing gene expression level (left) and differentiation potential2D (right) for three shRNA knockdowns of OTX2 during ectoderm differentiation in HUES64. We observe decreased EC expression and differentiation potential2D (top) and increased ME expression (bottom) in the knockdowns compared to control experiments. Reprinted by permission from Macmillan Publishers Ltd: Tsankov AM, Akopian V, Pop R et al. (2015) A qPCR ScoreCard quantifies the differentiation potential of human pluripotent stem cells. Nat Biotechnol 33:1182–1192.

Future directions

We recently developed a qPCR-based ScoreCard assay [3] with an improved gene expression panel, statistical analysis, and utility for a wider range of applications (latest information on the TaqMan hPSC Scorecard Assay is available at thermofisher.com/scorecardbp74). The qPCR-based ScoreCard assay allows for more quantitative and reproducible assessment of differentiation potential than the teratoma assay and is highly accessible, 5 to 10 times faster, and more cost-effective (Figure 1A). An area of focus for future development is incorporation of the improved algorithm described in [3] into the analysis module of the TaqMan hPSC ScoreCard Assay so that it is available online for all users. Also, single-cell transcriptomics could further improve the gene selection process and the reduction of gene expression markers needed to maintain statistical power while further reducing the assay cost.

Acknowledgments: This article was contributed by Alexander M. Tsankov, who is a member of the Meissner laboratory at the Broad Institute of MIT and Harvard, the Harvard Stem Cell Institute, and the Department of Stem Cell and Regenerative Biology, Harvard University. Please address correspondence to: atsankov@broadinstitute.org.

References

  1. Daley GQ (2012) The promise and perils of stem cell therapeutics. Cell Stem Cell 10:740–749.
  2. Dolgin E (2010) Putting stem cells to the test. Nat Med 16:1354–1357.
  3. Tsankov AM, Akopian V, Pop R et al. (2015) A qPCR ScoreCard quantifies the differentiation potential of human pluripotent stem cells. Nat Biotechnol 33:1182–1192.
  4. Müller FJ, Goldmann J, Löser P et al. (2010) A call to standardize teratoma assays used to define human pluripotent cell lines. Cell Stem Cell 6:412–414.
  5. Müller FJ, Schuldt BM, Williams R et al. (2011) A bioinformatic assay for pluripotency in human cells. Nat Methods 8:315–317.
  6. Bock C, Kiskinis E, Verstappen G et al. (2011) Reference Maps of human ES and iPS cell variation enable high-throughput characterization of pluripotent cell lines. Cell 144:439–452.
  7. Gifford CA, Ziller MJ, Gu H et al. (2013) Transcriptional and epigenetic dynamics during specification of human embryonic stem cells. Cell 153:1149–1163.
  8. Lipták T (1958) On the combination of independent tests. Magyar Tud Akad Mat Kutato Int Közl 3:171–196.
  9. Teo AK, Arnold SJ, Trotter MW et al. (2011) Pluripotency factors regulate definitive endoderm specification through eomesodermin. Genes Dev 25:238–250.
  10. Borowiak M, Maehr R, Chen S et al. (2009) Small molecules efficiently direct endodermal differentiation of mouse and human embryonic stem cells. Cell Stem Cell 4:348–358..
  11. Tsankov AM, Gu H, Akopian V et al. (2015) Transcription factor binding dynamics during human ES cell differentiation. Nature 518: 344–349.
  12. Ziller MJ, Edri R, Yaffe Y et al. (2015) Dissecting neural differentiation regulatory networks through epigenetic footprinting. Nature 518:355–359.

Learn more about:

Article download

Download a hyperlink-enabled, printer-friendly version of this article.
Download now