Supplementary Materials for High-throughput screening of tyrosine kinase inhibitor cardiotoxicity with human induced pluripotent stem cells Arun Sharma, Paul W. Burridge, Wesley L. McKeithan, Ricardo Serrano, Praveen Shukla, Nazish Sayed, Jared M. Churko, Tomoya Kitani, Haodi Wu, Alexandra Holmström, Elena Matsa, Yuan Zhang, Anusha Kumar, Alice C. Fan, Juan C. del Álamo, Sean M. Wu, Javid J. Moslehi, Mark Mercola, Joseph C. Wu* *Corresponding author. Email: [email protected]Published 15 February 2017, Sci. Transl. Med. 9, eaaf2584 (2017) DOI: 10.1126/scitranslmed.aaf2584 The PDF file includes: Materials and Methods Fig. S1. hiPSCs exhibit characteristic morphologies and markers of pluripotent stem cells. Fig. S2. Quantitative and qualitative cell viability assays illustrate sorafenib, regorafenib, and ponatinib cytotoxicity in hiPSC-CMs. Fig. S3. Quantitative cell viability assays on additional hiPSC-CM lines demonstrate VEGFR2/PDGFR-inhibiting TKI toxicity. Fig. S4. Quantitative cell viability assays in hiPSC-CMs and hiPSC-ECs derived from patients receiving TKI treatment. Fig. S5. Commercially available, healthy control hiPSC-CMs exhibit alterations in cellular contractility after a 72-hour TKI treatment. Fig. S6. Heat maps of high-throughput contractility analysis on commercially available, healthy control hiPSC-CMs treated with TKIs. Fig. S7. Extended calculations for TKI safety index after a 72-hour TKI treatment on commercially available, healthy control hiPSC-CMs. Fig. S8. hiPSC-CMs exhibit alterations in cellular contractility after a 72-hour treatment with known QT interval–prolonging TKIs. Fig. S9. hiPSC-ECs exhibit EC characteristics and demonstrate cytotoxicity in response to TKI treatment. Fig. S10. hiPSC-CFs exhibit properties of adult cardiac fibroblasts and demonstrate cytotoxicity in response to TKI treatment. www.sciencetranslationalmedicine.org/cgi/content/full/9/377/eaaf2584/DC1
26
Embed
Supplementary Materials for · 2017-02-21 · Supplementary Materials for . High-throughput screening of tyrosine kinase inhibitor cardiotoxicity with human induced pluripotent stem
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Supplementary Materials for
High-throughput screening of tyrosine kinase inhibitor cardiotoxicity with human induced pluripotent stem cells
Arun Sharma, Paul W. Burridge, Wesley L. McKeithan, Ricardo Serrano, Praveen Shukla, Nazish Sayed, Jared M. Churko, Tomoya Kitani, Haodi Wu,
Alexandra Holmström, Elena Matsa, Yuan Zhang, Anusha Kumar, Alice C. Fan, Juan C. del Álamo, Sean M. Wu, Javid J. Moslehi, Mark Mercola, Joseph C. Wu*
Published 15 February 2017, Sci. Transl. Med. 9, eaaf2584 (2017) DOI: 10.1126/scitranslmed.aaf2584
The PDF file includes:
Materials and Methods Fig. S1. hiPSCs exhibit characteristic morphologies and markers of pluripotent stem cells. Fig. S2. Quantitative and qualitative cell viability assays illustrate sorafenib, regorafenib, and ponatinib cytotoxicity in hiPSC-CMs. Fig. S3. Quantitative cell viability assays on additional hiPSC-CM lines demonstrate VEGFR2/PDGFR-inhibiting TKI toxicity. Fig. S4. Quantitative cell viability assays in hiPSC-CMs and hiPSC-ECs derived from patients receiving TKI treatment. Fig. S5. Commercially available, healthy control hiPSC-CMs exhibit alterations in cellular contractility after a 72-hour TKI treatment. Fig. S6. Heat maps of high-throughput contractility analysis on commercially available, healthy control hiPSC-CMs treated with TKIs. Fig. S7. Extended calculations for TKI safety index after a 72-hour TKI treatment on commercially available, healthy control hiPSC-CMs. Fig. S8. hiPSC-CMs exhibit alterations in cellular contractility after a 72-hour treatment with known QT interval–prolonging TKIs. Fig. S9. hiPSC-ECs exhibit EC characteristics and demonstrate cytotoxicity in response to TKI treatment. Fig. S10. hiPSC-CFs exhibit properties of adult cardiac fibroblasts and demonstrate cytotoxicity in response to TKI treatment.
Fig. S11. hiPSCs demonstrate a TKI cytotoxicity profile that is unique from those of hiPSC-CMs, hiPSC-ECs, and hiPSC-CFs. Fig. S12. VEGFR2/PDGFR-inhibiting TKI treatment in hiPSC-CMs results in activation of compensatory insulin/IGF1 signaling. Fig. S13. IGF1 and insulin treatment activates cardioprotective Akt signaling in hiPSC-CMs. Fig. S14. IGF1 and insulin treatment rescues doxorubicin toxicity in hiPSC-CMs. Fig. S15. IGF1 and insulin treatment rescues ponatinib toxicity at early time points in hiPSC-CMs. Fig. S16. RNA-seq of hiPSC-CMs treated with the VEGFR2/PDGFR-inhibiting TKI sorafenib illustrates compensatory hyperactivation of VEGF signaling. Table S1. Small-molecule TKIs selected for high-throughput cardiotoxicity screen. Table S2. Adverse cardiac events associated with small-molecule TKIs selected for high-throughput cardiotoxicity screen. Legend for movie S1 References (32–50)
Other Supplementary Material for this manuscript includes the following: (available at www.sciencetranslationalmedicine.org/cgi/content/full/9/377/eaaf2584/DC1)
Movie S1 (.mp4 format). hiPSC-CMs before purification via glucose deprivation.
Supplementary Materials
Methods
Derivation of human induced pluripotent stem cells (hiPSCs). All the protocols for this study were approved
by the Stanford University Institutional Review Board. Briefly, peripheral blood was obtained via standard
blood draw from 8 healthy control patients and 2 patients receiving TKI as part of their cancer treatment
regimen. Approximately 10 mL of blood was collected in Vacutainer tubes (BD Biosciences). Using a Ficoll-
Paque PLUS gradient (GE Healthcare), we isolated peripheral blood mononuclear cells (PBMCs) and
subsequently cultured them at 1 million cells per mL in a humanized blood cell culture medium (Life
Technologies). PBMCs were reprogrammed using a Sendai virus vector expressing OCT4, KLF4, SOX2, and
MYC (OKSM) (Life Technologies) following manufacturer’s instructions. Three additional patient-specific,
healthy control lines were derived from human skin fibroblasts using a lentivirus reprogramming vector
expressing OKSM. To obtain skin fibroblasts, skin punch biopsies were digested with Collagenase IV and were
retained in GlutaMAX-containing DMEM cell culture medium (Life Technologies) supplemented with 10%
fetal bovine serum (FBS, Gibco) in sterile conditions at standard 37 °C and 5% CO2 cell culture incubator
conditions. Subsequently, hiPSC clones were picked and grown on growth factor-reduced Matrigel (Corning)-
coated 6-well tissue culture dishes (Greiner) in E8 pluripotent stem cell culture medium (Life Technologies).
Immunofluorescence and laser confocal microscopy. Beating hiPSC-CM sheets were incubated in TrypLE
for 2 min at 37 °C followed by mechanical dissociation for the next 6 min using a P1000 pipettor, centrifuged at
200 x g for 4 min, and plated on 0.1-0.2% gelatin-coated glass coverslips following TrypLE deactivation.
Immunostaining was performed per standard protocols using 4% paraformaldehyde for fixation, 0.2% Triton X-
100 for membrane permeabilization, and 4% bovine serum albumin for blocking. Cardiac troponin T (Abcam
AB45932 1:200) and alpha-actinin (Sigma A7811 1:200) antibodies were used for qualitative imaging. Imaging
was performed using a DMIL-LED inverted tissue culture microscope (Leica Microsystems), BioTek Cytation
5 (BioTek), or a Zeiss LSM 510Meta confocal microscope (Carl Zeiss AG) using Zen imaging software.
hiPSC-CM culture for contractility analysis. Day 30-40 post-differentiation patient-specific, healthy control
hiPSC-CMs or, where indicated, commercially available, healthy control iCell hiPSC-CMs were used for
lot 1031999) or patient-specific healthy control hiPSC-CMs were transferred to liquid nitrogen storage. The
hiPSC-CMs were thawed and plated at a density of 5,000 cell/well into 0.1% gelatin (Stem Cell Technologies,
07903)-coated 384 well plates (Greiner Bio-One, 781091). The cardiomyocytes were maintained for 48 hours in
iCell Cardiomyocyte plating media at 37 °C and 5% CO2. After 48 hours, the media was changed to iCell
Cardiomyocyte maintenance media (Cellular Dynamics International, CMM-100-120-001) supplemented with 5
mM D-(+)-glucose (Sigma-Aldrich, G7021). The cardiomyocytes were maintained at 37 °C and 5% CO2 for 10
days with media changes every other day prior to addition of the library of tyrosine kinase inhibitors (TKIs) and
subsequent imaging.
Quantification of cell motion for contractility analysis. Deformation vector maps were obtained from
analysis of image time series using a particle image velocimetry (PIV) routine (32). To quantify the
deformation, a two-step approach was implemented. The aim of the first step was to select a reference frame
that represented the diastolic relaxation of the cells. This was achieved by running PIV using the previous frame
as reference, resulting in a measurement of the instantaneous contraction/relaxation velocity vector map. The
chosen interrogation window size was 128x128 pixels with 64-pixel overlap. The spatially averaged value of
the magnitude of the velocity fields measured the overall motion of the cell culture at each instant of time. The
resulting signal presented two periodic peaks that corresponded to a cardiac cycle (contraction and relaxation).
In the time elapsed between consecutive contraction cycles, the cells are mostly at rest. These periods of
inactivity were automatically identified to choose several reference frames. Ensemble-PIV was run again using
the selected reference frame to obtain the deformation fields and a smaller interrogation window (64x64 pixel
with 32-pixel overlap) to obtain finer spatial resolution and reduced noise levels.
Extraction of contractility signals. Using the cell deformation vector maps obtained from PIV, we applied
Gauss’ divergence theorem to automatically quantify cell contractility as the relative change in area of beating
cells. Contractility signals were obtained as the spatial average of the magnitude of relative change in area,
resulting in periodic signals with one peak per cycle (Fig. S5) whose magnitude is proportional to the overall
axial tension of the cell culture.
Retrieval of contractility parameters. The contractility signals presented periodic peaks related to each
contraction/relaxation cycle. Several parameters were extracted from these signals to perform quantitative and
statistical comparisons. To improve the robustness of the parameter retrieval, all the peaks present in one
contractility signal were collapsed into a most-representative peak by performing time-dependent conditional
averaging (Fig. S6A). As a byproduct of conditional averaging, the mean time between contractility peaks
(Tpeak) was obtained. The rest of the parameters were measured from the single-cycle contractility profile
obtained after conditional averaging (Fig. S6B).
Differentiation and characterization of hiPSC-ECs. To derive hiPSC-ECs, hiPSCs were maintained in E8
medium until the start of differentiation, at which time they were treated with 6 µM GSK3β inhibitor
CHIR99021 in CDM3 medium on day 0 and 2 µM CHIR99021 on day 2. Medium was changed on day 4 and
day 6. On day 7 post-differentiation, cells were treated with 20 µg/mL FGF2 and with 25 µg/mL VEGFA
(Peprotech). Medium was changed every other day. At day 12 post-differentiation, cells were sorted using a
magnetic-activated cell sorter (Miltenyi Biotech) for the endothelial cell surface markers CD31 and CD144.
Cells were then plated on 384-well plates (Greiner) in EGM-2 endothelial medium (Lonza) supplemented with
50 ng/mL VEGFA (Peprotech). To further functionally characterize hiPSC-ECs, cells were subjected to a tube
formation assay. Briefly, 100,000 hiPSC-ECs were plated into a 24-well culture plate coated with Matrigel
(Corning). After 24 hours, spontaneous tube formation was examined.
Differentiation and characterization of hiPSC-CFs. To derive hiPSC-CFs, hiPSCs were maintained in E8
medium until the start of differentiation, at which time they were treated first with GSK3β inhibitor CHIR99021
on day 0 and then with Wnt signaling inhibitor Wnt-C59 on day 2, both in CDM3 medium. At day 4 post-
differentiation, cells were switched to FibroLife medium (Lifeline Cell Technology) for human fibroblast
expansion. This medium was changed every other day. At day 8, cells were re-passaged at low confluency to
selectively facilitate hiPSC-CF adhesion over hiPSC-CM. Cells were also negatively-sorted for CD31 and
CD144 using a magnetic-activated cell sorter to eliminate endothelial cells. Cells were cultured for the next 8
days, and FibroLife was changed every 2 days to maintain and expand the hiPSC-CF population. Primary
human CFs for comparison to hiPSC-CFs were obtained from ATCC.
RNA-seq data analysis. RNA was isolated using miRNeasy kit (Qiagen) following the manufacturer’s protocol
and DNase treatment was performed using RNase-free DNase kit (Qiagen). 100 ng of total RNA was converted
to cDNA and amplified using NuGen V2 RNA-seq kit (NuGen, San Carlos, CA). cDNA was fragmented to an
average of 300 bps using the Covaris S2, and Illumina sequencing adapters were ligated to 500 ng of cDNA
using NEBNext® mRNA Library Prep Reagent Set (New England Biolabs, Ipswich, MA). PCR was performed
on the adapter-ligated cDNA using the following conditions (denaturation at 98 °C for 30 seconds, following 12
cycles of denaturation at 98 °C for 10 seconds, annealing at 65 °C for 30 seconds, and extension at 72 °C for 30
seconds, ending with an extension at 72 °C for 5 min). Libraries were submitted to the Stanford Stem Cell
Institute Genome Center for sequencing using Illumina’s HiSeq2000 platform using paired-in reads at an
average length of 100 bps (2x100). Reads were mapped using Tophat 2.0.8b with the hg19 reference annotation.
Cuffcompare and Cuffdiff were then used to determine which gene levels were significantly different (p<0.05).
RNA-seq expression data was annotated by the Affymetrix Expression Console software (Affymetrix). The
Pearson Correlation Coefficient was calculated for each pair of samples using the expression level of transcripts
that showed standard deviation greater than 0.2 among all samples. Heat maps were generated with TM4 suite
in MeV software (Multiple Experiment Viewer, MeV team).
SUPPLEMENTARY FIGURES
Fig. S1. hiPSCs exhibit characteristic morphologies and markers of pluripotent stem cells. (A) In this
study, eight patient-specific healthy control hiPSC lines were derived from human peripheral blood
mononuclear cells (PBMCs) using Sendai virus reprogramming vectors expressing OCT4, KLF4, SOX2, and
MYC (OKSM). An additional three patient-specific healthy control lines were derived from human skin
fibroblasts using a lentivirus vector expressing OKSM. Two hiPSC lines from patients treated with TKIs
sunitinib and axitinib were also derived using PBMCs/Sendai virus reprogramming. (B) hiPSCs from all
individuals exhibited standard human induced pluripotent stem cell colony morphology and expressed typical
markers for pluripotency such as NANOG and TRA-1-81.
Fig. S2. Quantitative and qualitative cell viability assays illustrate sorafenib, regorafenib, and ponatinib
cytotoxicity in hiPSC-CMs. (A) Dose response curves quantifying TKI toxicity in hiPSC-CMs from healthy
control patients using the CCK-8 cell viability assay, which measures intracellular dehydrogenase activity in
viable cells. Purified day 30 hiPSC-CMs were treated with TKIs for 72 hours. n = 5 biological replicates
conducted per line. (B) Dose response curves representing quantification of TKI toxicity in hiPSC-CMs from
multiple patients using CellTiter-Glo luminescence-based cell viability assay measuring ATP output. Day 30
hiPSC-CMs were treated with TKI for 72 hours. n = 5 biological replicate experiments were conducted for each
line. (C) High-throughput fluorescence imaging of purified day 30 hiPSC-CMs, stained with cardiomyocyte specific markers TNNT2 and α-actinin (ACTN2), treated for 72 hours from 0 to 100 µM, and doxorubicin as a
positive control for toxicity. All data are expressed as means ± SEM.
cytotoxicity in response to TKI treatment. (A) Workflow for hiPSC-EC production. F indicates FGF2 and V
indicates VEGF. (B) hiPSC-ECs exhibited characteristic “cobblestone” morphology. They also absorbed low
density lipoprotein and formed characteristic tube-like structures when replated onto Matrigel basement membrane. (C) hiPSC-ECs stained positive for the endothelial-specific cell surface markers CD31 (PECAM1)
and CD144 (VE-Cadherin). (D) hiPSC-ECs stained positive for von Willebrand Factor (vWF), another marker
for endothelial cells. (E) Dose response curves representing quantification of TKI toxicity in hiPSC-ECs from
multiple patients using the PrestoBlue viability assay. n = 5 biological replicate experiments conducted for each
line. All data expressed as means ± SEM.
Fig. S10. hiPSC-derived cardiac fibroblasts exhibit properties of adult cardiac fibroblasts and
demonstrate cytotoxicity in response to TKI treatment. (A) Workflow for hiPSC-CF production. (B) hiPSC-
CFs expressed the mesodermal and mesenchymal marker vimentin. (C) A subset of hiPSC-CFs expressed
vimentin (VIM) and alpha smooth muscle actin (ACTA2), markers of myofibroblast differentiation. (D) hiPSC-
CFs exhibited a similar proliferative rate to primary CFs. n = 3 biological replicate experiments. (E) hiPSC-CFs
exhibited whorl-like morphologies, like primary CFs. (F) Dose response curves representing quantification of
TKI toxicity in hiPSC-CFs from multiple patients using the PrestoBlue viability assay. n = 5 biological replicate
experiments conducted per line. Data expressed as means ± SEM.
Fig. S11. hiPSCs demonstrate a TKI cytotoxicity profile that is unique from those of hiPSC-CMs, hiPSC-
ECs, and hiPSC-CFs. Dose response curves representing quantification of TKI toxicity in undifferentiated
hiPSCs from multiple patients using CellTiter-Glo luminescence-based cell viability assay which measures ATP
output. Initially, hiPSCs were seeded at 1,000 cells per well in a 384-well plate and allowed to grow for 3 days.
Then, hiPSCs were treated with TKI for 72 hours. n = 5 biological replicate experiments were conducted for
each line. All data are expressed as means ± SEM.
Fig. S12. VEGFR2/PDGFR-inhibiting TKI treatment in hiPSC-CMs results in activation of
compensatory insulin/IGF1 signaling. Representative kinase assays on day 30 post-differentiation hiPSC-
CMs can elucidate changes in receptor tyrosine kinase phosphorylation in response to treatment with
VEGFR2/PDGFR-inhibiting TKIs sorafenib, axitinib, regorafenib, cabozantinib, ponatinib, and sunitinib.
Specifically, this kinase assay examines phosphorylation changes in EGFR, ErbB2, PDGFRα, ErbB4,
VEGFR2, INSR, IGF1R, and Axl. (A) Sorafenib treatment resulted in dose dependent inhibition of
VEGFR2/PDGFR phosphorylation, but no significant compensatory increase in INSR/IGF1R signaling. (B)
Axitinib treatment resulted in strong dose dependent inhibition of VEGFR2/PDGFRA phosphorylation and a
strong compensatory increase in INSR/IGF1R signaling. (C) Regorafenib treatment resulted in dose dependent
inhibition of VEGFR2/PDGFR phosphorylation and a minor compensatory increase in INSR/IGF1R signaling.
(D) Cabozantinib treatment resulted in dose dependent inhibition of VEGFR2/PDGFR phosphorylation, but no
significant compensatory increase in INSR/IGF1R signaling. (E) Ponatinib treatment resulted in dose dependent
inhibition of VEGFR2/PDGFRA phosphorylation and a compensatory increase in INSR/IGF1R signaling. (F)
Sunitinib treatment resulted in dose dependent inhibition of VEGFR2/PDGFR phosphorylation and a minor
compensatory increase in INSR/IGF1R signaling.
Fig. S13. IGF1 and insulin treatment activates cardioprotective Akt signaling in hiPSC-CMs. (A)
Representative phosphorylated kinase array assessing for phosphorylation of a broad range of human kinases
(listed) after 12 hour treatment with IGF1 at 200 ng/mL. Treatment with IGF1 led to phosphorylation of Akt
signaling network members such as Akt, WNK1, and PRAS40. (B) Representative phosphorylated kinase
arrays assessing for phosphorylation of a broad range of human kinases (listed) after 12 hour treatment with
insulin at 20 µg/mL. Treatment with insulin led to increased phosphorylation of Akt signaling network members
such as Akt, WNK1, and PRAS40. Significant differences (*) defined by P < 0.05 per Student’s t-test. n = 3
biological replicate experiments, as shown in Figure 6.
Fig. S14. IGF1 and insulin treatment rescue doxorubicin toxicity in hiPSC-CMs. (A) Dose response curves
quantifying cytotoxicity in hiPSC-CMs following 72 hour treatment with 3 µM doxorubicin at increasing doses
of IGF1 (12 hour pretreatment). PrestoBlue viability assay was used, and n = 5 biological replicates were
conducted for each line. (B) Dose response curves quantifying cytotoxicity in hiPSC-CMs following 72 hour
treatment with 0-100 µM doxorubicin and 12 hour pretreatment with either 20 µg/mL insulin or 200 ng/mL
IGF1. PrestoBlue viability assay was used, and n = 5 biological replicates were conducted. All data are
expressed as means ± SEM.
Fig. S15. IGF1 and insulin co-treatment rescues ponatinib toxicity at early timepoints in hiPSC-CMs.
Dose response curves representing quantification of ponatinib toxicity in hiPSC-CMs with and without insulin
or IGF1 co-treatment using the CellTiter-Glo viability assay. A significant increase in cell viability was
observed in response to insulin and IGF1 co-treatment as early as 12 hours post-ponatinib treatment. n = 3
biological replicate experiments were conducted. All data are expressed as means ± SEM.
Fig. S16. RNA-sequencing of hiPSC-CMs treated with VEGFR2/PDGFR-inhibiting TKI sorafenib
illustrates compensatory hyperactivation of VEGF signaling. (A) Heatmap illustrating alterations in gene
expression for five patient-specific, healthy control hiPSC-CM lines treated with 1 µM VEGFR2/PDGFR-
inhibiting TKI sorafenib for 72 hours. Red indicates increase in gene expression, blue indicates decrease. (B)
Heatmap illustrating alterations in VEGF pathway genes following sorafenib treatment. (C) Diagram
illustrating alterations in expression of VEGF pathway network genes following 1 µM sorafenib treatment. Red
indicates increase in gene expression, blue indicates decrease.
SUPPLEMENTARY TABLES
Table S1. Small molecule TKIs selected for high-throughput cardiotoxicity screen