Sundari Chetty, Felicia Walton Pagliuca, Christian Honore, Anastasie Kweudjeu, Alireza Rezania, and Douglas A. Melton

Size: px
Start display at page:

Download "Sundari Chetty, Felicia Walton Pagliuca, Christian Honore, Anastasie Kweudjeu, Alireza Rezania, and Douglas A. Melton"

Transcription

1 A simple tool to improve pluripotent stem cell differentiation Sundari Chetty, Felicia Walton Pagliuca, Christian Honore, Anastasie Kweudjeu, Alireza Rezania, and Douglas A. Melton Supplementary Information List of Figures Supplementary Figure 1. DMSO treatment enhances differentiation of hescs that have a low propensity for differentiation. Supplementary Figure 2. Human ESCs which fail to differentiate into definitive endoderm retain characteristics of undifferentiated ESCs. Supplementary Figure 3. DMSO treatment of the H1 cell line enhances differentiation into derivatives of the endocrine lineage following directed differentiation. Supplementary Figure 4. DMSO treatment enhances terminal differentiation. Supplementary Figure 5. DMSO treatment promotes neural differentiation. Supplementary Figure 6. DMSO treatment promotes cardiomyocyte differentiation. Supplementary Figure 7. DMSO treatment promotes endocrine differentiation. Supplementary Figure 8. Improved differentiation potential following DMSO treatment. Supplementary Figure 9. Optimization of the DMSO treatment further enhances differentiation potential. Supplementary Figure 10. Impact of DMSO treatment on subsequent stages of stepwise differentiation. Supplementary Figure 11. Live-cell imaging during a 24 h DMSO treatment and subsequent 48 h release. Supplementary Figure 12. Prolonged cell cycle arrest inhibits directed differentiation. Supplementary Table 1. Efficiencies of differentiation for individual hesc and hipsc lines. Supplementary Table 2. Impacts of the DMSO treatment on the total yield of differentiated cells. Supplementary Note. Technical note for further optimization of differentiation following DMSO treatment.

2 Supplementary Figure 1: DMSO treatment enhances differentiation of hescs that have a low propensity for differentiation a b c d e f DMSO-treated HUES6 C-peptide HUES8 C-peptide (a) Treatment with Wnt3a and Activin A for 4 d induces hescs into definitive endoderm as assessed by immunostaining for Sox17. Initial plating densities are shown in millions (m) of cells. (b, c) The propensity to differentiate into Sox17+ cells in HUES8 increases from 25% to 80%, with enhanced differentiation potential at high starting densities. (d) In low density HUES8 cultures, DMSO treatment prior to the onset of directed differentiation doubles the proportion of cells differentiating into Sox17+ cells, from ~25% to 50%. (e) Prior to the onset of definitive endoderm induction, a 24 h 2% DMSO treatment significantly improves differentiation in the low-propensity HUES6 cell line, inducing up to 50% Sox17+ cells. (f) DMSO treatment prior to ~20 d directed differentiation protocol in the low propensity HUES6 cell line promotes terminal differentiation of c-peptide+ cells at levels similar to the high propensity HUES8 cell line. Error bars represent Nature Methods: SEM doi: /nmeth.2442 of 2-4 biological replicates; Scale bar, 200µM; **p 0.01.

3 Supplementary Figure 2: Human ESCs which fail to differentiate into definitive endoderm retain characteristics of undifferentiated ESCs a b c Following definitive endoderm induction, (a) HUES6 and (b) HUES8 cells which fail to differentiate into Sox17+ definitive endoderm cells remain Oct4+ in control and DMSO treated conditions. (c) Similarly, in the H1 cell line, cells which fail to differentiate into CD184+ definitive endoderm cells retain high levels of expression of the hesc surface marker CD9. Error bars represent SEM of 2-3 biological replicates; Scale bar, 200µM.

4 Supplementary Figure 3: DMSO treatment of the H1 cell line enhances differentiation into derivatives of the endocrine lineage following directed differentiation a Definitive Endoderm SOX17 Control 7.0±0.7% 53.4±3.8% 32.8±4.4% 6.8±0.6% 1% DMSO 12.8±2.0% 76.0±1.6% 3.7±0.4% 7.5±1.0% b Releative gene expression (fold change over undifferentiated H1 cells) c HNF-3B d Control 8.6% 0.2% 1% DMSO 25.3% 10.5% Endocrine lineage NKX % Control 3.8% 38.8% Synaptophysin 5.7% 1% DMSO 25.3% 31.3% Releative gene expression (fold change over undifferentiated H1 cells) Islet-1

5 In the H1 cell line (a cell line predicted to have a low propensity for endodermal differentiation), DMSO treatment promotes terminal endocrine differentiation expressing genes and proteins with functional implications. An initial 24 h DMSO treatment enhances the differentiation potential of the H1 cell line into definitive endoderm as assessed by the proportion of (a) Sox17/HNF-3B (FOXA2)+ cells. (b) In addition to Sox17, genes characteristic of endodermal cells (i.e. FGF 17 and CXCR4) are also upregulated in the definitive endoderm cells generated from DMSO-treated cultures. (c, d) Furthermore, the DMSO effects persist to improve differentiation potential in subsequent stages of differentiation, generating endocrine cells with upregulated expression of genes (i.e. Pdx1, Nkx6.1, Ngn3, NeuroD1, and Arx) characteristic of mature endocrine derivatives. Upregulated expression of (c) synaptophysin and Islet-1 as well as the (d) pancreatic hormones (INS, GCG, SST) in DMSO-derived H1 pancreatic cells is characteristic of terminally differentiated functional endocrine cells. Error bars represent SEM of 3 biological replicates.

6 Supplementary Figure 4: DMSO treatment enhances terminal differentiation a b c DMSO treatment improves terminal differentiation into the (a,c) ectodermal (as assessed by Nestin and Tuj1 expression) and (b,c) mesodermal (as assessed by Troponin C expression) lineages in the low propensity HUES6 cell line following two different 12 d differentiation protocols. Error bars represent SEM of two biological replicates; Scale bar, 200µM; *p 0.05.

7 Supplementary Figure 5: DMSO treatment promotes neural differentiation Neuronal morphology (as assessed by Nestin and Tuj1 immunostaining) of DMSO treated HUES6 cultures following 12 d of differentiation. Scale bar, 100µM.

8 Supplementary Figure 6: DMSO treatment promotes cardiomyocyte differentiation DMSO treated HUES6 cultures differentiate into functional cardiomyocytes expressing Nkx2.5 and Troponin T. Scale bar, 200µM (top panel), 100µM (bottom panel).

9 Supplementary Figure 7: DMSO treatment promotes endocrine differentiation An initial 24 h DMSO treatment rescues the capacity of HUES8 cells (differentiated under low density conditions) to terminally differentiate into c-peptide+ and glucagon+ cells following ~20 d of stepwise directed differentiation. Scale bar, 200µM.

10 Supplementary Figure 8: Improved differentiation potential following DMSO treatment a b c (a) The magnitude of the DMSO effect is maximal in the poorest performing cell lines. (b, c) A 24 h DMSO treatment prior to ectodermal (Sox1) induction raises the differentiation potential of HUES1 (a cell line predicted to have a low propensity for ectodermal differentiation) from 28% to 93%. Error bars represent SEM of four biological replicates; Scale bar, 200µM; ***p

11 Supplementary Figure 9: Optimization of the DMSO treatment further enhances differentiation potential a ** b (a) Extending the DMSO treatment from 24 h to 48 h substantially increases the differentiation potential across many hesc and hipsc lines representative of a broad array of derivation procedures (HUES4, H9, H14, BG01V/hOG, FA, FB, ips SevA, ips SevB). On average, control lines differentiate at an efficiency of approximately 15-20% following mesodermal (Brachyury) induction. DMSO increases these levels up to 30% and 60% following a 24 h and 48 h treatment, respectively. Each point corresponds to an individual well. (b) Extending the DMSO treatment from 24 h to 48 h further enhances the differentiation potential to ~80% towards all germ layers, indicating that the impacts of the DMSO treatment can be substantially greater when optimized. Error bars represent SEM of 2-4 biological replicates. *p 0.05, **p 0.01, ***p 0.001; **[p h DMSO relative to 24 h DMSO in (a).

12 Supplementary Figure 10: Impact of DMSO treatment on subsequent stages of stepwise differentiation a b c (a) Schematic of stepwise differentiation into pancreatic (Pdx1+) progenitor cells. (b) Across multiple human ESC and ipsc lines, an initial 24 h DMSO treatment prior to stepwise differentiation has lasting effects improving differentiation potential into Pdx1+ cells. (c) Average fold changes are reported for each cell line at the initial definitive endoderm (Sox17) stage and at the pancreatic progenitor (Pdx1) stage. With further differentiation, the impacts of the DMSO treatment are greater at the Pdx1 stage than at the definitive endoderm stage across many cell lines. On average, the DMSO effects persist to improve Pdx1 induction by nearly 15-fold in hescs and 5-fold in hipscs. DE, definitive endoderm, PP, pancreatic progenitor; Error bars Nature represent Methods: doi: /nmeth.2442 SEM of 2-4 biological replicates.

13 Supplementary Figure 11: Live-cell imaging during a 24 h DMSO treatment and subsequent 48 h release DMSO treatment promotes a reversible growth inhibition in low density HUES6 cells.

14 Supplementary Figure 12: Prolonged cell cycle arrest inhibits directed differentiation a b (a) DMSO treated HUES6 cells were induced with Wnt3a and AA for 24h in the presence of aphidicolin, a DNA polymerase inhibitor which arrests cells from progressing into S phase of the cell cycle. The DMSO enhanced differentiation potential of HUES6 is substantially inhibited with only 7% of cells differentiating into brachyury+ cells when cell cycle progression to S phase is blocked by aphidicolin treatment. (b) In some cell lines (i.e. DiPS line ), a high concentration of DMSO can become inhibitory and reduce differentiation potential. Error bars represent SEM of 2-3 biological replicates; Scale bar, 200µM.

15 Supplementary Table 1: Efficiencies of differentiation for individual hesc and hipsc lines Individual differentiation efficiencies (as determined by the fraction of cells that differentiated into each germ layer) ± SEM for each cell line are listed. N/A, not assessed.

16 Supplementary Table 2: Impacts of the DMSO treatment on the total yield of differentiated cells The average fold change of the total number of cells that differentiated into each germ layer (relative to control) is listed for each hesc or hipsc line. N/A, not assessed.

17 Supplementary Note: Technical note for further optimization of differentiation following DMSO treatment. While a h treatment with 1% or 2% DMSO was beneficial for most cell lines at 80-90% confluence, differentiation efficiencies can be further optimized by taking into account the doubling time of a cell line and its endogenous ability to activate Rb upon cell contact. In cell lines with a slower doubling time, extending the duration of DMSO treatment (i.e. 48 h) can further enhance differentiation efficiency. Since the average cycling times across many hes and hips cell lines commonly range from ~24-48 h 21,22, promoting differentiation following a 24 h or 48 h DMSO treatment will likely improve differentiation rates. For further optimization, population doubling times can be determined for a specific cell line 21. In few ips cell lines, a high concentration of DMSO can become inhibitory and reduce differentiation potential, as in the DiPS line (Supplementary Fig. 12b), and thus avoiding permanent cell cycle arrest is also important. In addition to the duration of DMSO treatment, the duration of induction of differentiation signals and the time at which differentiation potential is assessed can also be optimized for each cell line as these parameters are likely to be influenced by a cell line s cycling time. For instance (as shown in Figure 2 for the 4 d BMP4 mesodermal induction protocol), prolonged treatment of inductive signals is known to induce differentiation into alternative (non-brachyury+) fates, and thus duration of induction is an additional component to take into account to optimize differentiation potential for any given cell line. Similarly, the kinetics of transcription factor expression is also likely to vary from one cell line to another and thus the time at which differentiation potential is assessed can be optimized for each cell line depending on its cycling time. We also note that different media compositions (i.e. serum content) induce different degrees of cell death and differentially regulate cell-to-cell contact during a differentiation protocol. In differentiation protocols with a high degree of cell death and disruption of cell contact, extending the duration of DMSO treatment to 48 h (or increasing the concentration from 1% to 2%) may provide additional benefits. Likewise, shortening the DMSO treatment to 24 h (or using a low 1% concentration) can be beneficial when cell survival is not significantly affected by a differentiation protocol. In general, the DMSO effects are especially strong in serum-free cultures, making the method particularly useful for therapeutic applications. Finally, the DMSO effects persist across passages and in various media that promote the growth and expansion of human pluripotent stem cells (MEF-conditioned media, mtesr, and Essential 8 media). DMSO concentrations of 1-2% did not have any significant toxic effects on hes and hips cells.