Carfilzomib Treatment Causes Molecular and Functional Alterations of Human Induced Pluripotent Stem Cell–Derived Cardiomyocytes

Abstract

Background
Anticancer therapies have significantly improved patient outcomes; however, cardiac side effects from cancer therapies remain a significant challenge. Cardiotoxicity following treatment with proteasome inhibitors such as carfilzomib is known in clinical settings, but the underlying mechanisms have not been fully elucidated.

Methods and Results
Using human induced pluripotent stem cell‐derived cardiomyocytes (hiPSC‐CMs) as a cell model for drug‐induced cytotoxicity in combination with traction force microscopy, functional assessments, high‐throughput imaging, and comprehensive omic analyses, we examined the molecular mechanisms involved in structural and functional alterations induced by carfilzomib in hiPSC‐CMs. Following the treatment of hiPSC‐CMs with carfilzomib at 0.01 to 10 µmol/L, we observed a concentration‐dependent increase in carfilzomib‐induced toxicity and corresponding morphological, structural, and functional changes. Carfilzomib treatment reduced mitochondrial membrane potential, ATP production, and mitochondrial oxidative respiration and increased mitochondrial oxidative stress. In addition, carfilzomib treatment affected contractility of hiPSC‐CMs in 3‐dimensional microtissues. At a single cell level, carfilzomib treatment impaired Ca ²⁺ transients and reduced integrin‐mediated traction forces as detected by piconewton tension sensors. Transcriptomic and proteomic analyses revealed that carfilzomib treatment downregulated the expression of genes involved in extracellular matrices, integrin complex, and cardiac contraction, and upregulated stress responsive proteins including heat shock proteins.

Conclusions
Carfilzomib treatment causes deleterious changes in cellular and functional characteristics of hiPSC‐CMs. Insights into these changes could be gained from the changes in the expression of genes and proteins identified from our omic analyses.

Full text

Journal of the American Heart Association
J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 1
ORIGINAL RESEARCH
Carfilzomib Treatment Causes Molecular
and Functional Alterations of Human
Induced Pluripotent Stem Cell– Derived
Cardiomyocytes
Parvin Forghani , PhD; Aysha Rashid, BS; Fangxu Sun, PhD; Rui Liu , BS; Dong Li , PhD; Megan R. Lee, BS;
Hyun Hwang , BS; Joshua T. Maxwell , PhD; Anant Mandawat, MD; Ronghu Wu, PhD; Khalid Salaita, PhD;
Chunhui Xu , PhD
BACKGROUND: Anticancer therapies have significantly improved patient outcomes; however, cardiac side effects from cancer
therapies remain a significant challenge. Cardiotoxicity following treatment with proteasome inhibitors such as carfilzomib is
known in clinical settings, but the underlying mechanisms have not been fully elucidated.
METHODS AND RESULTS: Using human induced pluripotent stem cell- derived cardiomyocytes (hiPSC- CMs) as a cell model for
drug- induced cytotoxicity in combination with traction force microscopy, functional assessments, high- throughput imaging,
and comprehensive omic analyses, we examined the molecular mechanisms involved in structural and functional alterations
induced by carfilzomib in hiPSC- CMs. Following the treatment of hiPSC- CMs with carfilzomib at 0.01 to 10µmol/L, we ob-
served a concentration- dependent increase in carfilzomib- induced toxicity and corresponding morphological, structural, and
functional changes. Carfilzomib treatment reduced mitochondrial membrane potential, ATP production, and mitochondrial
oxidative respiration and increased mitochondrial oxidative stress. In addition, carfilzomib treatment affected contractility of
hiPSC- CMs in 3- dimensional microtissues. At a single cell level, carfilzomib treatment impaired Ca2+ transients and reduced
integrin- mediated traction forces as detected by piconewton tension sensors. Transcriptomic and proteomic analyses re-
vealed that carfilzomib treatment downregulated the expression of genes involved in extracellular matrices, integrin complex,
and cardiac contraction, and upregulated stress responsive proteins including heat shock proteins.
CONCLUSIONS: Carfilzomib treatment causes deleterious changes in cellular and functional characteristics of hiPSC- CMs.
Insights into these changes could be gained from the changes in the expression of genes and proteins identified from our
omic analyses.
Key Words: cardiomyocyte ■ cardiotoxicity ■ drug research ■ gene expression ■ stem cell
Anticancer therapies have significantly improved
the outcomes of patients with cancer over the
past decade. However, several common chemo-
therapeutic agents, including proteasome inhibitors,
are associated with an increased risk of arrhythmias,
conduction abnormalities, and other cardiac ad-
verse events. Cardiac toxicities have been reported
with Food and Drug Administration– approved pro-
teasome inhibitors in clinical trials.1 Carfilzomib, a
second- generation proteasome inhibitor for the treat-
ment of relapsed or refractory multiple myeloma, can
cause cardiotoxicity.1– 4 Clinical trials with carfilzomib
have indicated cardiotoxicity including heart failure and
cardiac arrhythmias.5– 7 A meta- analysis of 29 clinical
trials including 4164 patients who received carfilzomib
reported 8.6% and 4.9% incidence of all- grade and
high- grade cardiotoxicity, respectively.8 Another meta-
analysis of 24 clinical trials including 2594 patients
Correspondence to: Chunhui Xu, PhD, Emory University School of Medicine, 2015 Uppergate Drive, Atlanta, GA 30322. E- mail: chunhui.xu@emory.edu
For Sources of Funding and Disclosures, see page 26.
© 2021 The Authors and Emory University school of Medicine. Published on behalf of the American Heart Association, Inc., by Wiley. This is an open access
article under the terms of the Creative Commons Attribution- NonCommercial License, which permits use, distribution and reproduction in any medium, provided
the original work is properly cited and is not used for commercial purposes.
JAHA is available at: www.ahajournals.org/journal/jaha
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 2
Forghani et al Carfilzomib- Induced Cardiotoxicity
who received carfilzomib showed 18.1% and 8.2%
incidence of all- grade and high- grade cardiotoxicity,
respectively.9 Although carfilzomib is used frequently
in the treatment of multiple myeloma, the incidence
of cardiotoxicity caused by carfilzomib appears to be
higher than other proteasome inhibitors.8
Carfilzomib has been well- characterized for its potent
activity to irreversibly bind to and inhibit the chymotrypsin-
like site of the proteasome,10 blocking the ability of the
ubiquitin/proteasome system to degrade and recycle
misfolded or damaged proteins.11 Carfilzomib can effec-
tively induce apoptosis and kill multiple types of human
cancer cells with IC50 ranging from 50 to 300 nmo-
l/L after 24 hours exposure of cell cultures to carfilzo-
mib.12 Cardiomyocytes are also sensitive to proteasome
inhibition possibly because of high protein turnover of
contractile proteins.11 For example, exposure of primary
neonatal rat cardiomyocytes to submicromolar concen-
trations of carfilzomib induced apoptosis and myocyte
damage.13 A preclinical pharmacokinetics study identi-
fied chymotrypsin- like proteasomal activity of carfilzo-
mib that can potentially damage rat cardiomyocytes at
clinically relevant concentrations.
14 However, cellular and
molecular mechanisms underlying carfilzomib- induced
cardiotoxicity remain to be fully elucidated.
Because cardiovascular side effects of cancer ther-
apies are increasing, the development of a human
cell model is needed to facilitate the understanding of
cardiotoxicity- related mechanisms. Progress in hiP-
SC- CM research has provided a new platform for the
studies of drug- induced side effects and disease mod-
eling.15,16 hiPSC- CMs have translational potential to im-
prove current models by providing more precise and
clinically relevant characteristics on responses to drug
treatment.17 They can also overcome the differences
between human and animal cardiac physiology and
challenges in long- term maintenance of primary human
cardiomyocytes and can be engineered for scalable
manufacture. hiPSC- CMs have provided novel insights
for the study of genetic heart diseases and drug re-
sponses.18– 2 0 Patient- specific hiPSC- CMs have also
been used for pharmacogenetic studies to facilitate the
identification of cancer survivors with increased risk of
chemotherapy- related cardiomyopathy.21
To advance our understanding of the underlying
mechanisms contributing to the carfilzomib- induced
cardiotoxicity, here we provide a molecular and func-
tional view of hiPSC- CMs after carfilzomib treatment.
We found that carfilzomib induced dose- dependent
cytotoxicity and targeted mitochondria at physiologi-
cally relevant doses, leading to the disruption of cellu-
lar energy and contractility. Additionally, we examined
cellular function at the single cell level through traction
force measurements using a nucleic acid- based ten-
sion sensor along with Ca2+ transient imaging. Our
findings on reduction in traction forces, abnormal Ca2+
transients, mitochondrial dysfunction, and contractility
impairment, in combination with comprehensive tran-
scriptome and proteome analyses, illustrate the possi-
ble molecular mechanisms in cardiomyocyte functional
alteration after carfilzomib treatment. Our study also
provides a unique resource for the discovery of bio-
markers associated with cardiomyocyte dysfunction
and arrhythmias following carfilzomib therapy.
METHODS
Data Availability
Global gene expression profiling of RNA sequencing
(RNA- seq) data are available at the National Center for
CLINICAL PERSPECTIVE
What Is New?
• Treatment of human stem cell- derived cardio-
myocytes with carfilzomib resulted in oxida-
tive stress, mitochondrial dysfunction, and cell
death.
• Carfilzomib treatment negatively affected contrac-
tility, Ca2+ handling, and integrin- mediated traction
forces in human stem cell- derived cardiomyocytes.
• Carfilzomib treatment downregulated the ex-
pression of genes involved in extracellular matri-
ces, integrin complex, and cardiac contraction
and upregulated stress responsive proteins.
What Are the Clinical Implications?
• Improving mitochondrial function, Ca2+ han-
dling, cardiac contraction, and integrin-
mediated traction forces has the potential to
mitigate carfilzomib- induced cardiotoxicity.
• The molecules such as heat shock proteins
upregulated by carfilzomib treatment are po-
tential biomarkers for carfilzomib- induced
cardiotoxicity.
Nonstandard Abbreviations and Acronyms
3D 3- dimensional
DMSO dimethyl sulfoxide
ECM extracellular matrix
GO gene ontology
hiPSC- CMs human induced pluripotent stem
cell- derived cardiomyocytes
HSPs heat shock proteins
RNA- seq RNA sequencing
ROS reactive oxygen species
TGT tension gauge tether
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 3
Forghani et al Carfilzomib- Induced Cardiotoxicity
Biotechnology Information Gene Expression Omnibus
database with the accession number GSE16 3102.
The proteomics data were deposited in the public
MassIVE database with the identifier MSV000087350.
Institutional review board approval was not required,
because this study did not involve the use of animals
or human subjects.
Cardiomyocyte Differentiation and
Spheroid Formation
Human induced pluripotent stem cell lines SCVI- 273
(Stanford Cardiovascular Institute) and IMR- 90 (WiCell
Research Institute) were differentiated toward car-
diomyocytes using small molecules or growth fac-
tors based on previously published differentiation
protocols2 2,23 (Figure1A). Human induced pluripotent
stem cell monolayers were dissociated using Versene
(Thermo Fisher Scientific) and seeded onto 12- well
plates coated with 1:60 Matrigel (Thermo Fisher
Scientific). Cardiomyocyte differentiation was per-
formed using 6µmol/L CHIR 99021 (Selleck Chemicals)
in RPMI/B27 insulin- free medium (days 0– 1) followed
by 5µmol/L IWR- 1 (Sigma- Aldrich) for 2days on days
3 to 5. Medium was changed on day 5, and cells were
maintained in RPMI/B27 with insulin for the remain-
ing of days. In the second protocol, cells were treated
with 100 ng/mL recombinant human activin A (R&D
Systems) on day 0 and replaced with 10 ng/mL re-
combinant human bone morphogenic protein- 4 (R&D
Systems) in RPMI/B27 insulin- free medium from days
1 to 4. Differentiated cells were maintained in RPMI/
B27 with insulin for 5 weeks with medium change
every 2 to 3 days. Spheroids were generated using
AggreWell400 plates (1800cells per microwell for each
spheroid) from differentiated cultures on differentiation
day 6 and maintained in RPMI/B27 with insulin.
SCVI- 273 hiPSC- CMs were used for oxidative stress,
caspase activity, mitochondrial membrane potential,
ATP measurement, Ca2+ transients, measurements
of sarcomere length and cell structure, and RNA- seq
analysis. IMR- 90 hiPSC- CMs were used for quantita-
tive reverse transcription– polymerase chain reaction,
traction force, contractility, and proteomics analysis.
Both hiPSC lines SCVI- 273- and IMR90- derived CMs
were used for cell viability and mitochondrial function
with similar results, and data from SCVI- 273 hiPSC-
CMs were presented.
Assay to Determine Mitochondrial
Membrane Potential
To analyze changes in mitochondrial membrane po-
tential in hiPSC- CMs, we used tetramethyl rhoda-
mine methyl ester, a dye probe that accumulates in
mitochondria. Medium was removed following 1- day
treatment with carfilzomib, and cells were labeled with
100 nmol/L tetramethyl rhodamine methyl ester for
30minutes at 37 °C. Cells were then counter stained
with Hoechst (Thermo Fisher Scientific; H3570) and
imaged immediately using ArrayScan XTI Live High
Content Platform (Life Technologies).24 We quantified
the nuclear spots within the ring in channel 2 by using
intensity as a readout.
Drug Preparation and Treatment
The stock solution of 10 mmol/L carfilzomib (Selleck
Chemicals; PR- 171 and S2853) was prepared by dis-
solving 10mg carfilzomib in 1.389mL dimethyl sulfox-
ide (DMSO; Sigma- Aldrich) and stored at −80 °C. The
peak plasma concentration based on pharmacokinetic
characteristics of carfilzomib is 5.88 µmol/L. A dose
range from 0.01 to 10µmol/L was selected following
initial testing of carfilzomib at 1, 2, 10, and 20µmol/L.
To make ×2 final concentration of carfilzomib in RPMI/
B27 with insulin, drug dilution was performed on the
day of each experiment and kept on ice in the dark.
DMSO at 0.2% (v/v), a concentration corresponding to
the highest drug concentration, was used as vehicle
control. Doxorubicin (Adriamycin; Selleck Chemicals)
at 10μmol/L was used as a positive control for viability
testing and 1μmol/L for contractility and Ca2+ transient
analysis. Stock solution of doxorubicin (10 mmol/L)
was made in the same manner as carfilzomib. The
peak plasma concentration of doxorubicin is 1.8 to
11μmol/L.
For the rescue experiment, hiPSC- CMs were pre-
treated with ascorbic acid (Sigma- Aldrich) at 25 µg/
mL for 2 hours before the carfilzomib treatment and
then treated with carfilzomib and ascorbic acid for
24hours before the analysis of cell viability by CellTiter-
Blue assay (Promega). All treatments were adjusted to
equivalent concentrations of DMSO (solvent).
In Vitro Cytotoxicity Assays
For monolayer culture of hiPSC- CMs, cell viability
was measured using CellTiter- Blue assay (Promega).
Monolayer cultures of hiPSC- CMs (IMR- 90 and SCVI-
273) were plated in Matrigel (1/60)- coated 96- well plates
with clear bottom and black wall. Cells were allowed
to attach at 37 °C in RPMI- B27 with Rock inhibitor
(10 µmol/L) for 24hours before treatment with carfil-
zomib. Serially diluted carfilzomib was added to cells.
After 24 and 48 hours of carfilzomib treatment, cells
were incubated with 20µL of the CellTiter- Blue reagent
in 100µL of RPMI/B27 medium solution for 2hours, and
the reduction of resazurin to resorufin in live cells was
measured by fluorescence excitation at 530 nm and
emission at 590 nm using a BioTek micro- plate reader
and Gen5 3.03 software. For 3- dimensional (3D) hiPSC-
CMs, cell viability was measured by Live/Dead staining
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 4
Forghani et al Carfilzomib- Induced Cardiotoxicity
(Thermo Fisher Scientific). A master mix of 1µmol/L eth-
idium homodimer and 0.25µmol/L of calcein in RPMI/
B27 medium was added following a wash with PBS.
Cells were incubated for 25minutes at 37 °C, washed
twice with PBS, and suspended in 5mL of RPMI/B27
medium without phenol red. Live/dead- stained cells
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 5
Forghani et al Carfilzomib- Induced Cardiotoxicity
were observed at 645nm for ethidium homodimer and
530nm for calcein- acetoxymethyl ester.
Immunofluorescence Staining
hiPSC- CMs were dissociated with 0.05% Trypsin-
EDTA and reseeded in Matrigel- coated 96- well
culture plates at a density of 5×104 cells per well.
Retrieved cells were fixed in 4% paraformaldehyde
for 15minutes following gentle PBS wash and per-
meabilized using 90% cold methanol for 2 minutes
at room temperature. The cells were then blocked
with 10% normal goat serum in PBS at room tem-
perature for 1hour and incubated overnight at 4 °C
with the primary antibodies against NK X2.5 (Cell
Signaling; 1:1600) and α- actinin (Sigma- Aldrich;
1:800) diluted in 3% normal goat serum for the purity
as say.23 After the incubation with the primary anti-
bodies, the cells were washed twice with PBS and
incubated with secondary antibodies, Alexa Fluor
488- conjugated goat anti- mouse immunoglobulin
G1 (for α- actinin staining, Life Technologies) and
Alexa Fluor 594- conjugated goat anti- rabbit immu-
noglobulin G (for NKX2.5 staining; Life Technologies)
diluted at 1:1000 in PBS with 0.25% BSA. The nuclei
were counterstained with 7 µmol/L Hoechst33342
(Thermo Fisher Scientific) for 15 minutes at room
temperature and preimaged using an inverted micro-
scope (Axio Vert.A1). Images of immunocytochem-
istry were quantitatively analyzed using ArrayScan
XTI Live High Content Platform. The Cellomics Scan
Software (Thermo Fisher Scientific) was used to cap-
ture images, and data analysis was performed using
Cellomics View Software (Thermo Fisher Scientific).
Twenty fields per well were imaged using a ×10 ob-
jective. Spot threshold was set to 10 units, and de-
tection limit was set at 25 units. The percentage of
α- actinin– positive cells and the average intensity per
well were used as a readout.
Caspase 3/7 Detection
Fresh Caspase- GloR 3/7 reagent (Promega) was re-
constituted and added to cells as an indicator of apop-
tosis. Background readings were measured from wells
containing culture medium without cells. Illuminometer
readings were taken 1hour after adding the Caspase-
GloR 3/7 reagent.
Assay to Measure ATP Content
CellTiter- Glo 3D Cell Viability kit (Promega) was used
to detect alterations in the cellular ATP content. 3D
hiPSC- CMs were dissociated into single cells using
0.25% Trypsin- EDTA and replated into a 96- well plate
at a density of 4.5×104cells per well. Medium was re-
moved, and RPMI without phenol red was added at
100µL per well. The kit was thawed at 4 °C a day be-
fore and the reagent was added at 100µL per well (1:1
ratio) with 2minutes shaking. Measurement was per-
formed at Top Count NXT Microplate Luminescence
Counter (PerkinElmer) with integration time of 1sec-
ond per well after 20 minutes incubation in room
temperature.
Quantitative Reverse Transcription–
Polymerase Chain Reaction
Total RNA was extracted using Aurum total RNA mini
kit (Bio- Rad) according to manufacturer’s instructions.
One microgram total RNA was used for complemen-
tary DNA synthesis using the Superscript VILO comple-
mentary DNA synthesis kit (Thermo Fisher Scientific),
and reaction mixture was incubated using a C1000
touch thermal cycler (Bio- Rad) as follows: 25 °C for
10minutes, 37 °C for 2hours and 85 °C for 5minutes.
The reaction mixture was further diluted to 300 and
2µL complementary DNA as the template was sub-
jected to quantitative reverse transcription– polymerase
chain reaction, which was performed in triplicate for
each gene using a SYBR Green reaction master mix
(Bio- Rad). Real- time polymerase chain reaction con-
ditions included initial denaturation step at 95 °C for
10minutes, 40 cycles of 2- steps with 15seconds of
denaturation at 95 °C, followed by 1minutes of anneal-
ing at 60 °C using Applied Biosystems 7500 real- time
polymerase chain reaction systems. The messenger
RNA levels of the genes examined were normalized to
GAPDH messenger RNA levels. The primers used for
the genes are listed in Table1.
Figur e 1. Cfz treatment- induced dose- dependent cytotoxicity in human induced pluripotent stem cell- derived cardiomyocytes
(hiPSC- CMs).
A, Overall experimental design. B, Representative images of immunofluorescence staining for examining cardiac purity using NKX2.5
antibodies and cardiomyocyte purity (percent NKX 2.5- positive cells) of 3- dimensional (3D) hiPSC- CMs following the treatment
with Cfz analyzed by ArrayScan. C, Relative cell viabilit y of hiPSC- CMs 24 and 48hours after Cfz treatment measured by CellTiter-
Blue fluorescence assay (n=4 cultures). D, Cell viability of 3D hiPSC- CMs 24hours after Cfz treatment using Live/Dead staining (red
ethidium- stained cells were dead cells; green calcein- stained cells were live cells). Scale bar=100µm. E, Relative caspase3/7 activity
of hiPSC- CMs 48hours after Cfz treatment (n=4– 5 cultures). F, Representative images of MitoSOX staining and summary of MitoSOX
MFI in hiPSC- CMs 24 and 48hours after Cfz treatment (n=3– 4 cultures). Nuclei were counterstained with Hoechst. Scale bar=50µm.
G, Relative cell viability of hiPSC- CMs 24hours following Cfz treatment with and without ascorbic acid (n=3 cultures). hiPSC- CMs were
pretreated with ascorbate acid for 2hours, followed by cotreatment with Cfz for 24hours. Cfz indicates carfilzomib; D, day; DMSO,
dimethyl sulfoxide; h, hour; and MFI, mean fluorescence intensity. *P<0.05; **P<0.01; ***P<0.001; ****P<0.0001.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 6
Forghani et al Carfilzomib- Induced Cardiotoxicity
Detection of Mitochondrial Reactive
Oxygen Species
To analyze changes in mitochondrial reactive oxy-
gen species (ROS), we used MitoSOX Red (Thermo
Fisher Scientific) staining. The cells were washed with
PBS and incubated with 5 µmol/L MitoSOX Red for
15minutes at 37 °C and protected from light. Cells
were counter- stained with Hoechst (Thermo Fisher
Scientific) and imaged using ArrayScan XTI Live High
Content Platform (Life Technologies).
Seahorse Extracellular Flux Analysis of
Mitochondrial Respiration
Seahorse plates were coated with Matrigel at 1:50
dilution 1 day before cell seeding. hiPSC- CMs were
seeded at 2.5×105cells per well in 300µL of the me-
dium and were allowed to adhere for 1day in a 37 °C
humidified incubator with 5% CO2. The Seahorse XF
Sensor Cartridge was hydrated the day before by filling
each well of the XF Utility plate with 1mL of Seahorse
XF Calibrant Solution and kept in a non- CO2 37 °C in-
cubator for 24hours to remove CO2 from the media
to prevent interference with pH- sensitive measure-
ments. To pre- equilibrate, hiPSC- CMs were washed
once with nonbuffered RPMI supplemented with
10mmol/L glucose, 2 mmol/L sodium pyruvate, and
2mmol/L glutamine. Cells were maintained in 525µL
of XF Assay medium at 37 °C in a non- CO2 incuba-
tor for 1hour. Agilent Seahorse XF24 Analyzer (Agilent
Seahorse Bioscience) was used to analyze the mito-
chondrial function of the cells by sequential injections
of modulators. A mixture of oligomycin (2µmol/L), car-
bonyl cyanide- 4- (trifluoromethoxy) phenylhydrazone
(1µmol/L), and rotenone (0.5µmol/L) were suspended
in a prewarmed XF Assay medium and loaded into the
injection ports (75µL) of the hydrated sensor cartridge
corresponding to the order of injection. Each measure-
ment cycle consisted of 3 minutes of mixing, 2 min-
utes of waiting, and 3 minutes of measurements of
oxygen consumption respiration. Measurement cycles
were performed after each addition of the given com-
pounds. The data were analyzed using Wave 2.6 and
Report Generator Version: 4.0.
Ca2+ Transient Imaging
Live cell imaging of intracellular Ca2+ transients was per-
formed with dye Fluo- 4 AM (Thermo Fisher Scientific).
Cells were incubated in Tyrode solution25 containing
Table 1. Primers for Quantitative Reverse Transcription– Polymerase Chain Reaction
Gene Descr iption of fu ll name Accession code Primer
ATP 2A 2 ATPase, Ca2+ transporting, cardiac
muscle, slow twitch 2
NM _1706 65 Fo rward: TCAG CAGGAACTTTGTCACC
Reverse: GGGCAAAGTGTATCGACAGG
CASQ2 Calsequestrin 2 NM _0 012 32 Forward: TTATGTTCAAGGACCTGGGC
Reverse: GCCTCTACTACCATGAGCCG
GAPDH Glyceraldehyde- 3- phosphate
dehydrogenase
NM_ 001256799 For ward: CTGGGCTACACTGAGCACC
Reve rse: A AGTGGTCGT TGAGGGCAATG
MYH6 Myosin, h eavy chai n 6, cardiac muscle,
alpha (α- MHC)
NM _0 02471 Forward: CTTCTCCACCTTAGCCCTGG
Rever se: GCTGGCC CTTCAACTACAGA
MYL2 Myosin, light chain 2, regulatory,
cardiac, s low (MLC- 2V)
NM_000432 Forward: CGTTCTTGTCA ATGAAGC CA
Reve rse: CA ACGTGTTC TCCATGTTCG
RYR2 Ryanodine rece ptor 2, cardiac NM _0 01035 F orwa rd: CAAATCCTTC TGCTGCCA AG
Rever se: CGA AGACGAGATCCAGT TCC
SLC8A1 Solute carrier family 8 (sodium/calcium
exchanger), member 1
NM _021097 For ward: CTGGA ATTCGAGCTCTCCAC
Reve rse: ACATCTGG AGCTCGAG GAA A
OP A1 OPA1 mitochondrial dynamin like
GTPase
NM _015 56 0.3 Forward: TGAAAGCATCAAGTTTTTCTTG
Reverse: TGCTGAAGATGGTGAGAAGAAG
NDUFB5 NADH ubiquinone oxoreductase
subunit B5
NM _00 2492. 4 Forward: ATGGTCTCCAC TGTGTCGA A
Reve rse: GGTG GCAGCTCTGTCTG G
MFN2 Mitofusin 2 N M _0 14 8 74. 4 Forward: TTGCATCGAGAGAAG AGCAG
Reverse: GTCTTTTGGACTTCAGCCAT
MFN1 Mitofusin 1 NM_033540.3 Forwa rd: GTT TTCAC TGCTGACTGCGA
Reverse: GTGGCACTTGCTGAAGGATT
CO Q10A Coenzyme Q10 homolog A NM_14 45 76 Forward: CTTACCTTCGAGCCGTTCCTT Reverse:
CC ATGATTCTAC GC TCC GA GTA
UCP3 Uncoupling protein 3 NM_003356.4 Forward: AACGCAAAAAGGAGGGTGTA
Reverse:
CTCCAGGCCAGTACTTCAGC
Primers were retrieved from open access websites (http://prime rdepot.nci.nih.gov/ or http://pga.mgh.harva rd.edu/prime rbank/).
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 7
Forghani et al Carfilzomib- Induced Cardiotoxicity
Fluo- 4 AM at a final concentration of 10μmol/L in the
dark at 37 °C for 20minutes followed by a gentle wash
at room temperature in prewarm Tyrode solution.
Fluorescence images were acquired using the Image
Xpress Micro XLS System (Molecular Devices) with
excitation/emission at 488/515 to 600 nm at a rate of
5frames per second and ×10 magnification for 4 fields
per well.
Video- Based Contractility Measurement
Contractility of spontaneously beating hiPSC- CMs was
recorded using a phase- contrast inverted microscope
(Axio Vert.A1) equipped with Zeiss Axio Cam digital
camera system.
Videos were recorded for 30seconds (5frames/s)
under ×10 magnification and were processed and ex-
ported using Zeiss AxioVision LE imaging software.
Videos were converted to frame by frame image se-
quence using ImageJ (National Institutes of Health).
Video- based analysis of contractility was performed
using MATLAB and motion vector software (R2016b;
MathWorks).26
Probe Preparation and Traction Force
Measurement
Turn on tension gauge tether (TGT) probes were used
to measure molecular traction forces. DNA duplexes
were conjugated to the fibronectin mimic cyclic- Arg-
Gly- Asp- Phe- Lys (Phe is D isomer) (cRGDfK), fluoro-
phore (Cy3B), and quencher (BHQ) using previously
published protocols.27, 2 8 The duplex was tethered to
a surface using biotin- streptavidin binding. When in-
tegrin receptors apply sufficient tension, the duplex
will mechanically denature and specifically when the
applied force exceeds the tension tolerance of the
probe. The shearing TGT with a tension tolerance
of 56 pN was used as described previously.29 Glass
surfaces were activated and functionalized with
streptavidin. Next, biotinylated DNA tension probes
were added. 3D hiPSC- CMs were treated with carfil-
zomib for 1day, dissociated, and then reseeded on
the DNA- modified glass surfaces. Microscopy im-
aging of spontaneously contracting cells was per-
formed using a Nikon TIRF microscope with ×100
objectives.
Measurements of Sarcomere Length and
Cell Structure
3D- derived hiPSC- CMs were treated with carfilzomib
for 1day, and then dissociated and reseeded on the
glass- bottom microplates. Cells were stained with
antibodies against α- actinin, and microscopy imag-
ing was performed using a Nikon TIRF microscope in
reflection interference contrast microscopy and TRITC
channels with ×100 objectives. ImageJ software was
used to quantify cell morphology, cell spread area, cir-
cularity, and aspect ratio. A program was written to
obtain automated outlining of cells. To measure the z-
lines, individual z- lines were selected, and lengths were
measured per cell. For each cell, the average length of
≈20 to 30 z- lines was plotted.
RNA Sequencing
Total RNA was isolated from day 30 hiPSC- CMs in bio-
logical triplic ates following 1day of carfilzomib treatment
using Aurum total RNA mini kit (Bio- Rad) according to
the manufacturer’s instructions. RNA concentrations
were measured using Nanodrop Spectrophotometer
(Thermo Fisher Scientific). Library preparation and
RNA- seq was conducted by Novogene with 20 M
reads per sample, PE150 Mapped Homo sapiens
(GRCh38/hg38) to the genome using STAR (v2.6.1d)
with ensemble annotation. Fastp was used for length
limitation of adapter trimming (https://github.com/
novog ene- europ e/fastp). The differentially expressed
genes were used for analyses of gene ontology (GO)
terms and Kyoto Encyclopedia of Genes and Genomes
pathways, which were considered significantly en-
riched if the adjusted P<0.05.
Proteomic Analysis
Cells and culture media were collected from the trip-
licates of 3D hiPSC- CM cultures following 1 day of
carfilzomib treatment. Proteins were extracted from
3 to 4×106 hiPSC- CMs per sample by suspending
the cells in the lysis buffer as described previously.30
Proteins were purified through methanol- chloroform
precipitation. For the secretome analysis, the media
was passed through a filter (0.45µm) and then con-
centrated by centrifugation (molecular weight 3kDa
cutoff). Proteins were digested with trypsin over-
night. Then, purified peptides were labeled with the
6- plex tandem mass tag reagents for protein quan-
titation. The tandem mass tag– labeled samples (6
for the cell lysates and 6 for the secretomes) were
mixed and fractionated. Each fraction was analyzed
by liquid chromatography- tandem mass spectrom-
etry. The data analyses were conducted as reported
previously.30
Statistical Analysis
Statistical analyses were done using GraphPad Prism
version 8.00. Global differences were evaluated by
Dunnett and Turkey test (1- way ANOVA). P<0.05 was
taken as statistically significant. Results are presented as
mean±SD in all experiments. Sample sizes are indicated
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 8
Forghani et al Carfilzomib- Induced Cardiotoxicity
in figure legends. For group comparison of gene ex-
pression in RNA- seq analysis, Benjamin- Hochberg
correction was used to control false- discovery rate.
We considered genes to be significantly differentially
expressed between the two groups if adjusted P<0.05
and absolute value of log2 (fold change) >1.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 9
Forghani et al Carfilzomib- Induced Cardiotoxicity
RESULTS
Carfilzomib Treatment Induced a Dose-
and Duration- Dependent Cardiotoxicity
To examine whether carfilzomib treatment causes
cytotoxicity in hiPSC- CMs, we generated enriched
hiPSC- CMs (>90% NKX2.5- positive cells), treated
the cells with various doses of carfilzomib, and char-
acterized the cells 24 and 48 hours after treatment
(Figure 1A and 1B). We used a common chemo-
therapeutic drug, doxorubicin, at 10µmol/L (a peak
plasma concentration) as a control.31 Following carfil-
zomib treatment, we observed dose- and duration-
dependent increase of cell loss in carfilzomib- treated
cultures compared with DMSO- treated cultures
(Figure1C). As detected by CellTiter Blue assay, cul-
tures treated with carfilzomib at 10μmol/L and doxo-
rubicin for 24hours had lower cell viabilities compared
with cultures treated with DMSO. The cell viability in
cultures treated with carfilzomib for 48hours began
to decrease at doses as low as 0.1μmol/L, which
is significantly lower than the peak concentration of
the carfilzomib observed in patients’ plasma after
intravenous administration.32,33 Given that hiPSC-
CM spheroids (3D cultures) provide a more physi-
ologically relevant context for drug toxicity,34 we
examined whether carfilzomib could also induce cy-
totoxicity in 3D cultures. We generated hiPSC- CM
spheroids using microscale tissue engineering23 and
treated them with carfilzomib for 24 hours. Using
calcein and ethidium bromide as indicators for live
and dead cells, respectively, we found that cultures
treated with carfilzomib at 10 μmol/L and doxoru-
bicin had increased dead cells (ethidium bromide–
positive cells), whereas moderate toxicity was also
observed in cultures treated with carfilzomib at 0.1
and 1μmol/L (Figure1D). These results suggest that
carfilzomib induced dose- dependent cytotoxicity in
both 2- dimensional and 3D cultures.
To determine the mechanism of cell death, we ex-
amined caspase3/7 activity using Caspase- Glo 3/7
Luminescent assay. Increased caspase3/7 activation
was detected after carfilzomib treatment for 24hours,
indicating that carfilzomib- induced cytotoxicity might
be the result of apoptosis (Figure1E).
Carfilzomib Treatment Increased
Mitochondrial Superoxide and Reduced
Mitochondrial Function of hiPSC- CMs
Oxidative stress through generation of mitochondrial
superoxide and mitochondrial dysfunction plays im-
portant roles in cellular cytotoxicity,35,36 which could
affect cardiac function.37– 39 We therefore examined
the effect of carfilzomib treatment on mitochondrial
oxidative stress and mitochondrial function. Based
on fluorescence intensity of MitoSOX, a mitochondrial
superoxide indicator, the relative levels of mitochon-
drial superoxide were higher in cultures treated with
carfilzomib at 1 and 10μmol/L compared with DMSO-
treated cultures (Figure1F).
To examine if carfilzomib induced cytotoxicity
through oxidative stress, we evaluated if ascorbic acid, a
commonly used antioxidant, could rescue carfilzomib-
induced cell loss. Compared with carfilzomib- treated
cultures without ascorbic acid, the carfilzomib- treated
cultures with ascorbic acid had significantly attenu-
ated cell loss when cells were treated with carfilzomib
at 0.01, 0.1, and 1μmol/L (Figure1G), suggesting that
carfilzomib- induced cytotoxicity is in part mediated by
oxidative stress.
We also examined the effect of carfilzomib treat-
ment on mitochondria membrane potential by staining
the cells with tetramethylrhodamine methyl ester, a cell-
permeant fluorescent dye that is sequestered by active
mitochondria. hiPSC- CMs treated with carfilzomib at
0.1, 1, and 10μmol/L had >10- fold reduced levels of
tetramethylrhodamine methyl fluorescence intensity
(Figure2A and 2B). These results indicate a substan-
tial decrease in mitochondrial membrane potential and
increase in mitochondrial superoxide following carfilzo-
mib treatment, suggesting that oxidative stress could
play an important role in carfilzomib- mediated cardiac
cytotoxicity.
Figure 2. Cfz treatment reduced mitochondrial function of human induced pluripotent stem cell- derived cardiomyocytes
(hiPSC- CMs).
A, Measurement of TMRM fluorescence 24hours after Cfz treatment. Nuclei were counterstained with Hoechst. Scale bar=100µm.
B, Summar y of TMRM MFI analyzed by ArrayScan (n=5 cultures). C, Effects of Cfz treatment on the expression of genes related to
mitochondrial function in hiPSC- CMs (n=3 cultures). Gene expression is normalized to the housekeeping gene GAPDH and shown
as relative levels to the control (DMSO treated) group. D, Representative traces of oxygen consumption rate recording in hiPSC- CMs
upon sequential treatments with oligomycin, FCCP, and a mixture of rotenone and antimycin A. E, Quantification of ATP production,
basal respiration, and maximal respiration (n=4 cultures). The results were normalized to 1×106cells. F, Relative cellular ATP content in
3- dimensional hiPSC- CMs 24hours after Cfz treatment as measured by an ATP- based luminescence assay (n=5 cultures). *P<0.05;
**P<0.01; ***P<0.001; ****P<0.0001. Cfz indicates carfilzomib; COQ10 A, coenzyme Q10A; DMSO, dimethyl sulfoxide; FCCP, carbonyl
cyanide- 4- (trifluoromethoxy) phenylhydrazone; MFI, mean fluorescence intensity; MF N1, mitofusin 1; MFN2, mitofusin 2; NDUFB8,
NADH:ubiquinone oxidoreductase subunit B8; OCR, oxygen consumption rate; OPA1, OPA1 mitochondrial dynamin like GTPase;
TMRM, tetramethyl rhodamine methyl; and UCP3, uncoupling protein 3.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 10
Forghani et al Carfilzomib- Induced Cardiotoxicity
To understand the effects of carfilzomib on mito-
chondrial function, we examined the expression of
genes associated with mitochondrial function and per-
formed a Seahorse XF Cell Mito stress test 1day after
carfilzomib treatment. Carfilzomib treatment caused
a dose- dependent decrease in the expression of
genes associated with mitochondrial function, includ-
ing COQ10A (coenzyme Q10A), MFN1 (mitofusin 1),
MFN2 (mitofusin 2), and NDUFB5 (NADH:ubiquinone
oxidoreductase subunit B8), but not OPA1 (OPA1
mitochondrial dynamin like GTPase) and UCP3 (un-
coupling protein 3) (Figure 2C). The basal and maxi-
mal respiratory capacity and ATP production (serving
as an indicator of mitochondrial function) were mea-
sured by monitoring oxygen consumption respira-
tion after sequential injection of oligomycin, carbonyl
cyanide- 4- (trifluoromethoxy) phenylhydrazone, and
rotenone according to the manufacture’s instruction.
As shown in Figure2D and 2E, a significant decrease
in ATP production was detected following the treat-
ment of carfilzomib at 0.1, 1, and 10μmol/L in SCVI-
273 hiPSC- CMs. The reduced ATP levels were also
observed in 3D cultures treated with carfilzomib as
detected by CellTiter- Glo viability assay (Figure 2F).
Higher concentrations of carfilzomib (10μmol/L) also
decreased the basal and maximal mitochondrial res-
piration in SCVI- 273 hiPSC- CMs (Figure 2D and 2E).
Similarly, carfilzomib treatment reduced ATP produc-
tion and basal mitochondrial respiration in IMR90
hiPSC- CMs (data not shown). These results suggest
that carfilzomib induces mitochondrial damage and
consequently the cardiotoxicity.
Figure 3. Cfz increased abnormal intercellular Ca2+ transients of human induced pluripotent stem cell- derived
cardiomyocytes (hiPSC- CMs) and decreased the expression of genes associated with Ca2+ handling.
A, Representative Ca2+ transient traces from each group and summary of cells with normal and abnormal Ca2+ transients (n=9– 58
cells). hiPSC- CMs were treated with Cfz for 24hours and measured for Ca2+ transients. Numbers shown on the stack bars represents
percentages of cells with normal and abnormal Ca2+ transients. B, Relative expression levels of genes associated with Ca2+ handling
in hiPSC- CMs treated with Cfz for 24hours (n=3 cultures). *P<0.05; * *P<0. 01. ATP2A2 indicates ATPase sarcoplasmic/endoplasmic
reticulum Ca2+ transporting 2; CASQ2, calsequestrin 2; Cfz, carfilzomib; DMSO, dimethyl sulfoxide; RYR2, ryanodine receptor 2; and
SLC8A1, solute carrier family 8 member A1.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 11
Forghani et al Carfilzomib- Induced Cardiotoxicity
Carfilzomib Treatment Induced Abnormal
Ca2+ Transients and Dysfunctional
Contractility
There is increasing evidence that alteration in mito-
chondrial function results in increased ROS generation
and abnormal Ca2+ transients.37, 4 0 Considering the link
of oxidative stress and Ca2+ transients, we analyzed
the Ca2+ transient profile 24 hours after carfilzomib
treatment. Carfilzomib treatment of SCVI- 273 hiPSC-
CMs increased the proportion of cells with abnormal
Ca2+ transients (Figure 3A). Consistently, carfilzomib
treatment decreased the expression of genes as-
sociated with Ca2+ handling (SLC8 A1 [solute carrier
family 8 member A1], RYR2 [ryanodine receptor 2],
CASQ2 [calsequestrin 2], and ATP 2A2 [ATPase sar-
coplasmic/endoplasmic reticulum Ca2+ transporting
2]) (Figure3B). The downregulation of these genes is
consistent with the abnormal Ca2+ transients following
carfilzomib treatment. These results indicate that carfil-
zomib treatment impairs Ca2+ transients at the single
cell level, which may affect the contractility function of
hiPSC- CMs.
To address if mitochondrial dysfunction along with
abnormal Ca2+ transients was associated with con-
tractile dysfunction in carfilzomib- treated hiPSC- CMs,
we assessed the contractility of 3D hiPSC- CM static
spheroids after treatment with carfilzomib for 24 and
48hours using video microscopy with motion vector
analysis. Cessations in contraction occurred in cul-
tures treated with all concentrations (0.01, 0.1, 1, and
10µmol/L) of carfilzomib for 48hours. At 24hours after
carfilzomib treatment, a portion of hiPSC- CM spheroids
treated with carfilzomib at 0.01µmol/L or doxorubicin
and all spheroids treated with 0.1, 1, and 10µmol/L
stopped beating (Figure 4A). Similar results of ces-
sations in contraction were observed in hiPSC- CMs
derived from both SCVI- 273 and IMR90 lines treated
with carfilzomib at high doses. The average maximum
contraction, maximum relaxation velocity, and the beat
rate remained unchanged in the remaining beating cells
from cultures treated with carfilzomib at 0.01µmol/L or
doxorubicin compared with those from DMSO- treated
cultures (Figure 4B and 4C). Additionally, carfilzomib
reduced the expression of genes associated with
contractility (MYH6 [myosin heavy chain 6] and MYL2
[myosin light chain 2]) (Figure4D). These data suggest
that carfilzomib increases contractile dysfunction after
treatment.
Carfilzomib Treatment Reduced
Traction Forces and Caused Structural
Disorganization
To further examine the carfilzomib- induced alteration
in contraction, we used a DNA duplex TGT probe29
to measure integrin- mediated traction forces in single
cardiomyocytes. The TGT probe was modified with a
fibronectin mimetic ligand (cyclic- RGD [the tripeptide
consists of arginine, glycine, and aspartate]), fluoro-
phore (Cy3B), and quencher. When cell integrins bind
to the RGD ligand and transmit a threshold magnitude
of tension greater than the probe’s tension tolerance
(56 pN), the duplex mechanically denatures, and the
fluorophore separates from the quencher. Because
the biotin- anchored nucleic acid is also fluorescently
tagged, shearing of the top strand leads to ≈20- fold
enhancement in fluorescence. Moreover, the fluores-
cence signal is directly proportional to the number of
probes that experiences a threshold force exceeding 56
pN; therefore, the fluorescence signal provides a quan-
titative readout of integrin traction forces (Figure 5A).
3D hiPSC- CMs were treated with carfilzomib for 1day,
dissociated, and then reseeded onto the glass surface
with the TGT probe. We then quantified the fluores-
cence signal of single cardiomyocytes upon plating to
monitor the traction forces of spontaneously contract-
ing cardiomyocytes. As shown in Figure 5B and 5C,
the traction forces of hiPSC- CMs decreased in cultures
treated with carfilzomib at 0.01, 0.1, 1, and 10µmol/L or
doxorubicin compared with DMSO- treated cells. Thus,
carfilzomib reduced integrin- mediated traction forces.
To evaluate if structural changes were accompanied
with the alteration in contraction induced by carfil-
zomib treatment, we measured the cell structure of
hiPSC- CMs after carfilzomib treatment by immuno-
cytochemistry of α- actinin, a protein expressed in z-
lines of cardiomyocytes. Cells treated with carfilzomib
at 1 and 10µmol/L lacked clear z- lines, whereas the
DMSO- treated cells and cells treated with carfilzomib
at 0.01 and 0.1µmol/L had clear z- lines (Figure6A). We
also quantified cell size and shape along with sarco-
mere length to identify the link between the shape and
contraction. Cells treated with higher concentrations of
carfilzomib (1 and 10µmol/L) showed decreased cell
area compared with DMSO- treated cells; these cells
also had irregular peripheral borders resulting in an in-
creased circularity (Figure6B). In addition, cells treated
with carfilzomib at 0.01 and 0.1µmol/L or doxorubi-
cin had shorter z- line length and sarcomere length
compared with DMSO- treated cells (Figure6B). These
results indicate that carfilzomib treatment can induce
significant structural alteration parallel with contractility
dysfunction.
Transcriptomic and Proteomic Analyses
Revealed That Carfilzomib Dysregulated
Genes Related to Stress Response,
Extracellular Matrix, and Contractility
To further elucidate the mechanism of carfilzomib-
induced alteration in contractility, we compared global
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 12
Forghani et al Carfilzomib- Induced Cardiotoxicity
Figure 4. Cfz tr eatment induced contractility dysfunction in 3- dimensional (3D) human induced
pluripotent stem cell- derived cardiomyocytes (hiPSC- CMs).
A, Pie chart representing proportions of spheroids with or without beating arrest 24hours after Cfz treatment
(n=16– 60 cardiac spheroids). B, Contractilit y of static 3D hiPSC- CMs was video recorded and analyzed using
MATLAB. Representative heat maps and graphs of averaged magnitude of beating speed over time in Cfz- treated
hiPSC- CMs. Red cycles and blue triangles represent contraction and relaxation, respectively. Note: hiPSC- CMs
stopped beating following 24hours of treatment of Cfz at 0.1, 1, and 10µmol/L. C, Quantification of contraction,
relaxation, and beating rate among groups (n=14– 17 cardiac spheroids). D, Expression of genes encoding
contractile proteins MYH6 (myosin heavy chain 6) and MYL2 (myosin light chain 2) detected by quantitative reverse
transcription– polymerase chain reaction (n=3 cultures). Cfz indicates carfilzomib; and DMSO, dimethyl sulfoxide.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 13
Forghani et al Carfilzomib- Induced Cardiotoxicity
gene expression profile of hiPSC- CMs treated with
carfilzomib at 1 µmol/L for 24 hours with DMSO-
treated cells. RNA- seq analysis identified 5027 genes
that were differentially expressed based on absolute
log2 (fold change) >1 and adjusted P<0.05. Compared
with DMSO- treated cells, 1913 genes were upregu-
lated, whereas 3114 genes were downregulated in
carfilzomib- treated cells (Figure 7A). For example,
carfilzomib induced downregulation of genes involved
in cardiac muscle contraction (eg, ACTA 2 [actin alpha-
2, smooth muscle]) and integrin complex (eg, I T G A11
[integrin subunit alpha 11]). In addition, genes related to
heat shock stress were upregulated, including HSPA1B
(heat shock protein family A member 1B), HSPA6 (heat
shock protein family A member 60), HSPH1 (heat
shock protein family H member 1), and BAG3 (BAG
cochaperone 3), which participate in cellular response
to stress, cell death, and apoptosis.
We also performed analysis of GO terms using the
differentially expressed genes. GO terms related to
oxidative stress, heat shock proteins (HSPs) and
proteasomal protein catabolic process were upreg-
ulated, and GO terms related to extracellular matrix
(ECM) and cardiac contraction were downregulated
(Figure7B, Table2). In line with carfilzomib- mediated
mitochondrial oxidative stress, carfilzomib induced
upregulation of GO terms including response to ox-
idative stress, ROS metabolic process, response to
Figure 5. C fz treatment reduced t he traction forces of human induced pluripotent stem cell- derived cardiomyocytes
(hiPSC- CMs).
For quantification of molecular traction forces of spontaneously contracting cardiomyocytes, hiPSC- CMs were treated for 1day with
Cfz and reseeded on the glass bottom microplates coated with the DNA probes through biotin streptavidin conjugation. A, Illustrative
model of traction force measurement principle. The probes were decorated with peptide mimic (cRGDfk) of fibronectin, a fluorophore
(Cy3B), and quencher (BHQ - 2). Fluorescence intensity of the probes on the surface increases upon rupturing of the probes when cells
contract and apply a force to the probes greater than the force tolerance of around 56 pN. B, Summary of traction force measurement
in hiPSC- CMs treated with Cfz vs DMSO (n=20– 30 cells). ** **P<0.0001. C, Representative traction force microscopy of hiPSC-
CMs after Cfz treatment. Scale bar=12µm. Cfz indicates carfilzomib; DMSO, dimethyl sulfoxide; pN, pico newton; RICM, reflection
interference contrast microscopy; and TGT, tension gauge tether.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 14
Forghani et al Carfilzomib- Induced Cardiotoxicity
Figure 6. Cfz treatment led to structural alterations in human induced pluripotent stem cell- derived cardiomyocytes (hiPSC-
CMs).
hiPSC- CMs were treated with Cfz for 24 hours, fixed, and stained with antibodies against α- actinin. Cells were imaged using
fluorescence microscopy and quantitatively analyzed. A, Representative RICM and fluorescence images of hiPSC- CMs treated with
Cfz. Scale bar=12µm. B, Summar y of structural parameters of hiPSC- CMs after Cfz treatment (n=20 – 30 cells). **P<0.01; ****P<0.0001.
Note that Cfz treatment significantly decreased the spread area, z- line length, and sarcomere length and increased circularity. Cells
treated with Cfz at 1 and 10µmol/L did not show clear striation and z lines. Cfz indicates carfilzomib; DMSO, dimethyl sulfoxide; and
RICM, reflection interference contrast microscopy.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 15
Forghani et al Carfilzomib- Induced Cardiotoxicity
temperature stimulus, and cellular response to heat.
In contrast, carfilzomib induced downregulation of
GO terms of extracellular structure organization, ex-
tracellular matrix, actin binding, muscle system pro-
cess, actin cytoskeleton, transmembrane receptor
protein serine/threonine kinase signaling pathway,
ECM structural constituent, contractile fiber, re-
sponse to calcium ion, cardiac muscle contraction,
and integrin binding and pathway- restricted Smad
protein phosphorylation.
The top downregulated genes based on fold change
included those involved in integrin and ECM (I TG A 11,
MEGF6 [multiple EGF like domains 6], FJX1 [four-
jointed box kinase 1], MFAP4 [microfibril associated
protein 4], CCDC80 [coiled- coil domain containing 80],
and FNDC10 [fibronectin type III domain containing
10]), mitochondria (SLIT3 [slit guance ligand 3], DUT
[deoxyurine triphosphatase], PCK2 [phosphoenolpyru-
vate carboxykinase 2, mitochondrial], and ATP 6V0 E 2
[ATPase H+ transporting V0 subunit E2]) and mus-
cle contraction/tight junction (ACTA2 [actin alpha 2,
smooth muscle] and SYNPO [Synaptopodin] ) (Table3).
The top upregulated genes included genes involved
in response to oxidative stress, heat stress, and ROS
metabolic process, including HS PA1B, H S PA1A (heat
shock protein family H member 1), DNAJA1 (DnaJ
heat shock protein family member A1), BAG3, and
HSP90AB1 (heat shock protein 90 alpha family class
B member 1) (Table 4). Additionally, carfilzomib up-
regulated genes involved in mitogen- activated pro-
tein kinase- mediated signaling cascade (MAP2K3)
and downregulated genes involved in Smad pathway
(BM P10 and BMP7 ) (Figure 7B). Carfilzomib also al-
tered the expression of genes associated with ECM-
receptor interaction, cell cycle, protein digestion and
absorption, dilated cardiomyopathy, and hypertrophic
cardiomyopathy (Figure7C).
We further analyzed the differentially expressed
genes based on Z scores. As shown in the heatmap
(Figure7D), the expression of genes related to cardiac
muscle contraction was dramatically downregulated
in carfilzomib- treated cells compared with DMSO-
treated cells. These genes included response to cal-
cium ion (TNNT2 [troponin T2, cardiac type], EEF2K
[eukaryotic elongation factor 2 kinase], CARF [calcium
responsive transcription factor], MYL3 [myosin light
chain 3], MAP2K6 [mitogen- activated protein kinase
kinase 6], MYB [MYB proto- oncogene, transcription
factor], KCNJ5 [potassium inwardly rectifying channel
subfamily J member 5], KCNQ1 [potassium voltage-
gated channel subfamily Q member 1], TNNT2, MYH7
[myosin heavy chain 7], RYR2, and CASQ2), calcium
responsive proteins (RYR2, SLC 25A12 [solute carrier
family 25 member 12], CASQ2, and KCNMB1), and
contractile proteins (SYNPO [synaptopodin], ACTA 2,
FLNC [filamin C], BMF [Bcl2 modifying factor], MYL3,
MYO1D [myosin ID], ZNF185 [zinc finger protein 185
with LIM domain], TN NI1 [troponin I1, slow skeletal
type], TNNT3 [troponin T3, fast skeletal type], MYL2,
MYL5 [myosin light chain 5], MYLK [myosin light chain
kinase], MYO5C [myosin VC], MYO15B [myosin XVB],
MYH6, and MYL7 [myosin light chain 7]). In addition,
carfilzomib- dysregulated genes were involved in ECM,
integrin complex, and actinin cytoskeleton. These
genes included I TG A 11, SYNPO, SLC6A4 [solute car-
rier family 6 member 4], CASQ2, TNNI1, and MYL3
(Figure7D).
To examine the effect of carfilzomib at the protein
level, we performed quantitative proteomic analysis on
both the cell supernatant and cell lysate of hiPSC- CMs
treated with carfilzomib at 1µmol/L versus DMSO for
24hours. Out of the 4060 proteins quantified in the cell
lysate, 183 proteins were upregulated, and 39 proteins
were downregulated (Figure 8A). Out of the 298 pro-
teins detected in the cell supernatant, 6 proteins were
upregulated, and 18 proteins were downregulated after
carfilzomib treatment (P<0.05, absolute fold change
>1.3) (Figure 8B). The downregulated proteins in the
cell lysate included ANXA6 (annexin A6, a calcium-
dependent membrane and phospholipid binding pro-
tein), SPTN1 (spectrin alpha, non- erythrocytic 1, a
filamentous cytoskeletal protein highly expressed in
cardiac muscle at z- disc), and TPM1 (tropomyosin 1, a
protein that forms a complex with troponin T and regu-
lates actin- myosin interaction in response to intracellu-
lar Ca2+ concentration).
As shown in Figure 8A, carfilzomib treatment re-
duced the expression of proteins in the cells associ-
ated with metabolic process, including pyruvate kinase
M2 (PKM2), which is involved in glycolysis and reg-
ulates cardiomyocyte cell cycle, and protein kinase
cAMP- activated catalytic subunit alpha (PRKACA),
which is the catalytic subunit α of protein kinase A that
contributes to the control of glucose metabolism and
cell division. The downregulated proteins were asso-
ciated with GO terms of metabolic process (PRKACA
[protein kinase cAMP- activated catalytic subunit
alpha], SDHAF2 [succinate dehydrogenase complex
assembly factor 2], PGAM2 [phosphoglycerate mu-
tase 2], GOT1 [glutamic- oxaloacetic transaminase 1],
IMPA1 [inositol monophosphatase 1], CLEC16A [C-
type lectin domain containing 16A], and ATM [ATM
serine/threonine kinase]), mitochondrion organization
(SDHFA2, GGCT [gamma- glutamylcyclotransferase],
and MICOS13 [mitochondrial contact site and cristae
organizing system subunit 13]), striated muscle con-
traction (PRKACA, GSTM2 [glutathione S- transferase
mu 2], PGAM2, TPM1, and RCSD1 [RCSD domain
containing 1]), oxidation reduction (FDX1 [ferredoxin
1], HADH [hydroxyacyl- CoA dehydrogenase], IDH3G
[isocitrate dehydrogenase (NAD(+) 3 non- catalytic sub-
unit gamma], PGAM2, PKM [pyruvate kinase M1/2],
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 16
Forghani et al Carfilzomib- Induced Cardiotoxicity
TPI1 [triosephosphate isomerase 1], and SDHAF2), and
muscle system process (ANXA [annexin A], GSTM2
[glutathione S- transferase mu 2], PGAM2, PRKACA,
TPM1, and RCSD1) (Figure9A, Table5).
The enriched GO terms of upregulated proteins in
the cells included cellular response to stress, ubiquitin-
dependent protein catabolic process, protein folding,
regulation of cellular response to heat, cell death, and
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 17
Forghani et al Carfilzomib- Induced Cardiotoxicity
stress- activated mitogen- activated protein kinase cas-
cade (Figure9B). In addition, proteins associated with
oxidative stress, autophagy, apoptosis and cell cycle
were upregulated in the cells, including MAFG (MAF
bZIP transcription factor G, a transcription factor that
is induced following oxidative stress), ATG101 (autoph-
agy related 101, an essential protein for the initiation of
autophagy), RND3 (Rho family GTPase 3, a member
of the small Rho GTPase family that regulates apopto-
sis), and CDKN1A (cyclin dependent kinase inhibitor 1,
Figure8A, Table6).
The downregulated proteins in the secretome in-
cluded POSTN (periostin), C1S (complement C1s),
THBS1 (thrombospondin 1), and COL3A1 (collagen
type III alpha 1 chain) (Figure8B, Table7). THBS1 is
an adhesive glycoprotein that mediates cell- to- cell
and cell- to- ECM interactions, COL3A1 provides in-
structions for making type III collagen that strength-
ens and supports cardiac tissue, and POSTN is a
ligand for integrins providing the support for cell ad-
hesion. The downregulated proteins in secretome
are associated with GO terms including extracellular
Figure 7. Cfz treatment altered global gene expression in human induced pluripotent stem cell- derived cardiomyocytes
(hiPSC- CMs).
Gene expression profiling of hiPSC- CMs after treatment with 1μmol/L Cfz for 24hours was analyzed by RNA sequencing (n=3 cultures).
DEGs in cells after Cfz were compared with those in the cells treated with DMSO. Volcano plot representing 3114 downregulated
genes and 1913 upregulated genes are depicted in the red and blue dots, respectively. DEGs between the 2 groups were defined
based on adjusted P<0.05 and the absolute value of log2 (fold change) ≥1. B, Bubble plots representing enrichment analysis of DEGs
using GO enrichment analysis. C, Chord diagram of relationship between selected GO terms and relevant DEGs. Each GO term
is shown on the right, and genes contributing to these enrichments are presented on the left. Colored squares next to each gene
indicate log2 (fold change) from the highest to the lowest level. D, Heatmap of DEGs associated with cardiac muscle contraction (lef t)
and ECM, integrin complex, and actinin cytoskeleton (right). Red color indicated relatively high expression and blue color indicated
relatively low expression based on Z score. ACTA2 indicates actin alpha 2, smooth muscle; AKR1C3, aldo- keto reductase family
1 member C3; ANXA5, annexin A5; ARC, activit y regulated cy toskeleton associated protein; BMP10, bone morphogenetic protein
10; CACNB2, calcium voltage- gated channel auxiliary subunit beta 2; CACNG6, calcium voltage- gated channel auxiliary subunit
gamma 6; CAPN3, calpain 3; CARF, calcium responsive transcription factor; CASQ2, calsequestrin 2; CCDC80, coiled- coil domain
containing 80; CD248, CD248 molecule; Cfz, carfilzomib; CGN, cingulin; CHGA, chromogranin A; CLU, clusterin; COL1A1, collagen
type I alpha 1 chain; COL1A2, collagen type I alpha 2 chain; COL3A1, collagen type III alpha 1 chain; COL4A1, collagen type IV alpha
1 chain; COL4A3, collagen type IV alpha 3 chain; COL4A4, collagen type IV alpha 4 chain; COL4A5, collagen type IV alpha 5 chain;
COL4A6, collagen type IV alpha 6 chain; COL6A3, collagen type VI alpha 3 chain; COL9A1, collagen type IX alpha 1 chain; CORIN,
corin, serine peptidase; CPNE4, copine 4; CPNE5, copine 5; CRYAB, crystallin alpha B; CSRP3, cysteine and glycine rich protein
3; DCN, decorin; DEGs, differentially expressed genes; DMSO, dimethyl sulfoxide; DPEP1, dipeptidase 1; DUSP1, dual specificity
phosphatase 1; ECM, extracellular matrix; EGFL6, EGF like domain multiple 6; EGFR, epidermal growth factor receptor; EMILIN1,
elastin microfibril interfacer 1; ENDOG, endonuclease G; FERMT3, FERM domain containing kindlin 3; FLNC, filamin C; FN1, fibronectin
1; GAA, alpha glucosidase; GAS2, growth arrest specific 2; GO, gene ontology; GPX3, glutathione peroxidase 3; HAPLN1, hyaluronan
and proteoglycan link protein 1; HMGCR, 3- hydroxy- 3- methylglutaryl- CoA reductase; HSP90AA1, heat shock protein 90 alpha family
class A member 1; HSPA1A, heat shock protein family A; HSPA1B, heat shock protein family A; HSPA5, heat shock protein family A;
HSPA6, heat shock protein family A; ITGA1, integrin subunit alpha 1; ITGA11, integrin subunit alpha 11; ITGA3, integrin subunit alpha
3; ITGA8, integrin subunit alpha 8; ITGAX, integrin subunit alpha X; ITGB3, integrin subunit beta 3; ITGB4, integrin subunit beta 4;
ITGB6, integrin subunit beta 6; JUP, junction plakoglobin; KCNA5, potassium voltage- gated channel subfamily A member 5; KCNE5,
potassium voltage- gated channel subfamily E regulatory subunit 5; KCNIP2, potassium voltage- gated channel interacting protein
2; KCNJ5, potassium inwardly rectifying channel subfamily J member 5; KCNJ8, potassium inwardly rectifying channel subfamily
J member 8; KCNMB1, potassium calcium- activated channel subfamily M regulatory beta subunit 1; KCNQ1, potassium voltage-
gated channel subfamily Q member 1; LRRC15, leucine rich repeat containing 15; MAFG, MAF bZIP transcription factor G; MAP2K3,
mitogen- activated protein kinase kinase 3; MAP2K6, mitogen- activated protein kinase kinase 6; MCM2, minichromosome maintenance
complex component 2; MCM4, minichromosome maintenance complex component 4; MCM5, minichromosome maintenance complex
component 5; MCM6, minichromosome maintenance complex component 6; MFAP4, microfibril associated protein 4; MMP1, matrix
metallopeptidase 1; MMP10, matrix metallopeptidase 10; MMP12, matrix metallopeptidase 12; MMP24, matrix metallopeptidase 24;
MMP3, matrix metallopeptidase 3; MPP4, membrane palmitoylated protein 4; MSN, moesin; MUC4, mucin 4, cell surface associated;
MYBPC3, myosin binding protein C3; MYBPH, myosin binding protein H; MYH14, myosin heavy chain 14; MYH15, myosin heavy chain
15; MYH6, myosin heavy chain 6; MYH7, myosin heavy chain 7; MYL2, myosin light chain 2; MYL3, myosin light chain 3; MYO16, myosin
XVI; MYO7A, myosin VIIA; NEB, nebulin; NLGN1, neuroligin 1; NTN1, netrin 1; PCNA, proliferating cell nuclear antigen; PHOSPHO1,
phosphoethanolamine/phosphocholine phosphatase 1; PHPT1, phosphohistidine phosphatase 1; PKM, pyruvate kinase M1/2; POSTN,
periostin; PPIF, peptidylprolyl isomerase F; RASAL1, RAS protein activator like 1; RELL2, RELT like 2; RGS2, regulator of G protein
signaling 2; RND3, Rho family GTPase 3; RNF207, ring finger protein 207; RYR1, ryanodine receptor 1; RYR2, ryanodine receptor 2;
SCN2B, sodium voltage- gated channel beta subunit 2; SCN5A, sodium voltage- gated channel alpha subunit 5; SERPINE1, serpin
family E member 1; SGCD, sarcoglycan delta; SHH, sonic hedgehog signaling molecule; SLC8A1, solute carrier family 8 member A1;
SLC9A1, solute carrier family 9 member A1; SOD1, superoxide dismutase 1; SSC5D, scavenger receptor cysteine rich family member
with 5 domains; ST13, ST13 Hsp70 interacting protein; SUCNR1, succinate receptor 1; SYNPO, synaptopodin; TAFAZZIN, tafazzin,
phospholipid- lysophospholipid transacylase; TCAP, titin- cap; TIMP4, TIMP metallopeptidase inhibitor 4; TNNC1, troponin C1, slow
skeletal and cardiac type; TNNI1, troponin I1, slow skeletal type; TNNT2, troponin T2, cardiac type; TNNT3, troponin T3, fast skeletal
type; TSPAN32, tetraspanin 32; TXNIP, thioredoxin interacting protein; TXNRD1, thioredoxin reductase 1; VCP, valosin containing
protein; VEGFB, vascular endothelial growth factor B; VIL1, villin 1; WAS, WASP actin nucleation promoting factor; WNT5A, Wnt family
member 5A; XIRP2, xin actin binding repeat containing 2; and ZC3H12A, zinc finger CCCH- type containing 12A.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 18
Forghani et al Carfilzomib- Induced Cardiotoxicity
matrix organization, cell- matrix adhesion, and cell
motility (Figure 9C) and Kyoto Encyclopedia of Genes
and Genomes pathways including tight junction and
DNA replication (Table8). In addition, upregulation of
HSPA1A was observed in both the cells and the secre-
tome (Figure8A and 8B, Table7).
Comparison of the proteomic and transcriptomic
data revealed a set of overlapping genes and proteins
that were differentially expressed in response to car-
filzomib treatment (Figure 8C and 8D). We performed
additional GO- term analysis using these overlapping
genes and proteins (Figure 8E, Table9). The enriched
GO terms included response to heat, HSP binding, and
ATPase regulator activity (Figure8E). Several pathways
were upregulated, including apoptotic signaling (ATF3
[activating transcription factor 3], CDKN1A, DNAJA1,
HSPA1A, SERPINE1 [serpin family E member 1], BAG3,
and USP47 [ubiquitin specific peptidase 47]), stress-
activated mitogen- activated protein kinase cascade
(CRYAB, HMGCR [3- hydroxy- 3- methylglutaryl- CoA re-
ductase], DNAJA1, SKP1 [S- Phase Kinase Associated
Protein 1], and UBB [ubiquitin B]) and ATPase activ-
ity (DNAJA1, DNAJB1 [DnaJ heat shock protein fam-
ily (Hsp40) member B1], BAG3, HSPH1, and DNAJB4
[DnaJ heat shock protein family (Hsp40) member B4])
(Table9). We note that RNA- seq was analyzed using
hiPSC- CMs derived from the SCVI- 273 line, and pro-
teomics analysis was performed using hiPSC- CMs
derived from IMR- 90 hiPSCs. The consistency in the
alterations in the expression of these overlapping
genes and proteins in 2 cell lines suggests that the
observed alterations induced by carfilzomib treatment
are independent of the cell lines used.
In addition, carfilzomib treatment upregulated sev-
eral ubiquitin- related proteins. As detected by RNA-
seq, ubiquitin C was among the top 20 upregulated
proteins (Table 4). Proteomic analysis of cell lysates
also revealed that several ubiquitin- related proteins
were upregulated, including negative regulator of ubiq-
uitin like proteins 1, praja ring finger ubiquitin ligase
2, ubiquitin specific peptidase 33, and ubiquitin B
(Table6). GO- term analysis of differentially expressed
proteins identified ubiquitin protein ligase binding
(n=44) and ubiquitin- dependent ERAD (endoplasmic
reticulum- associated degradation) pathway (n=16).
These changes may indicate the direct effect of carfil-
zomib on the ubiquitin- proteasome pathway.
Taken together, the transcriptomic and proteomic
analyses of hiPSC- CMs indicate that downregulation
of contractile- related genes/proteins, and ECM and
integrin- related genes/proteins together with upreg-
ulation of HSPs and stress- activated pathways were
associated with cardiac- toxic effects of carfilzomib
treatment.
DISCUSSION
Using molecular tension sensors to quantify cellular
traction forces in combination with high- throughput
imaging, functional assessments, and transcriptomic
and proteomic analyses, we investigated mechanisms
of earlier cellular and molecular events associated
Table 2. Top Downregulated GO Terms Based on DEGs
GO term Category N o. of genes GO term ID Adjuste d P value Enrichment
Extracellular matrix Biological process 106 GO:0 031012 1.2 4E - 17 3.2
Extracellular structure organization Cellular component 108 GO:0043062 2.02E- 12 3
Extracellular matrix structural constituent Morphological
function
54 GO:0005201 1.3 7E - 10 3.6
Contractile fiber Biological process 49 GO:0043292 0.006207 2.3
Transmembrane receptor protein serine/threonine
kinase signaling pathway
Cellular component 68 GO:0007178 0.0 07959 2.1
Integrin complex Morphological
function
12 GO:0008305 0.010804 3.4
Actin binding Morphological
function
79 GO:0003779 0.012514 2
Integrin binding Biological process 29 GO:0005178 0.027535 2.5
Pathway- restricted Smad protein phosphorylation Biological process 19 GO:0060389 0.028713 3 .1
Cardiac muscle contraction Biological process 31 GO:0060048 0.028 713 2.7
Cardiac muscle tissue morphogenesis Biological process 20 GO:0055008 0.02 8713 3.2
Response to calcium ion Biological process 34 GO: 00 5159 2 0.03831 2.4
Muscle system process Biological process 77 GO:0003012 0.044283 1.8
Sarcomere organization Cellular component 14 GO :0 04 5214 0 .0 46 017 3.4
Actin cytoskeleton Cellular component 83 GO:001562 9 0.048669 1.7
DEGs indi cates differentiall y expressed genes; and GO, ge ne ontology.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 19
Forghani et al Carfilzomib- Induced Cardiotoxicity
with alterations of contraction after carfilzomib treat-
ment in hiPSC- CMs. Our results suggest a possible
role of ECM and integrin- related genes in contractility
and traction force defects as an early response to
carfilzomib treatment. We also identified increased
mitochondrial oxidative stress, reduced mitochondrial
Table 3. Top Downregulated DEGs
Gene ID Gene symbol Descr iption of fu ll name Adjuste d P value Log2, fold change
ENSG00000177943 MAMDC4 MAM doma in containing 4 4.94E- 56 − 9.1 9 9 08 7 57 2
ENSG00000171992 SYNPO Synaptopodin 8.8 4E - 51 −6.60093912
ENSG00000104738 MCM4 Minichromosome maintenance complex component 4 1.11E- 5 0 −6.261530975
ENSG00000137809 ITG A 11 Integrin subunit alpha 11 2.58E- 48 −6.196870403
ENSG00000184347 SLIT3 Slit guance ligand 3 5.1 3E - 46 −6 .15 7 7 8116 3
ENSG00000156427 FG F18 Fibroblast growth factor 18 6.1 4E - 43 −6 .1 50 8 72 2 0 9
ENSG00000185567 AHNAK2 AHNAK nucleoprotein 2 3.69E- 39 −6.0835 52672
ENSG00000128951 DUT Deoxyurine triphosphatase 7.18E - 3 9 − 6. 0 12 0 2411 6
ENSG00000225138 None None 1.01E- 38 −5.979760829
ENSG00000166482 MFAP4 Microfibril associated protein 4 2.35E - 38 − 5.9 732 36912
ENSG00000107796 ACTA 2 Actin alpha 2, smooth muscle 4.87E- 34 −5.900 336258
ENSG00000162591 MEGF6 Multiple EGF like domains 6 5.24 E- 33 −5.883525697
ENSG00000179431 FJX1 Four- jointed box kinas e 1 1. 35 E- 31 −5.691046545
ENSG00000157637 SLC 38 A10 Solu te carrier f amily 38 mem ber 10 2.21E- 31 − 5.6 69 74 82 57
ENSG00000171130 ATP6V0E2 ATPase H+ transporting V0 sub unit e2 8.55 E- 31 − 5. 62 73 08177
ENSG00000129103 SUMF2 Sulfatase modifying factor 2 9.96 E- 31 − 5.615074065
ENSG00000100889 PCK2 Phosphoenolpyruvate carboxykinase 2, mitochondrial 1.01E- 30 −5.60 670 1278
ENSG00000228594 FNDC10 Fibronectin ty pe III domain containing 10 1. 94E - 3 0 −5.571960443
ENSG00000091986 CCDC80 Coiled- coil domain containing 80 6.85E- 30 −5.561040949
ENSG00000100297 MCM5 Minichromosome maintenance complex component 5 2.30E- 29 −5.54716236 4
DEGs indicates differentially expressed genes.
Table 4. Top Upregulated DEGs
Gene ID Gene symbol Description of full name P value Log2, fold change
ENSG00000198431 TXNR D1 (Thioredoxin reductase 1) 4.82E- 178 4.006 453
ENSG00000204388 HSP A1B Heat shock protein fami ly A (Hsp70) member 1B 3.04E- 168 6.043998
ENSG00000173110 HSPA6 Heat shoc k protein family A (Hsp70) member 6 1.07E- 140 10.5290 2
ENSG00000178381 ZFAND2A Zinc finger AN1- type containing 2A 3.06E- 140 4.963423
ENSG00000080824 HSP 9 0A A1 Heat shock protein 90 al pha family class A member 1 4. 45 E- 12 0 3.5678 91
ENSG00000204389 HSP A1A Heat shock protein fami ly H (Hsp110) member 1 1.3 5E- 10 5 5.71313 8
ENSG00000120694 HSPH1 Heat shock protein family H (Hsp110) member 1 4.87E- 94 3.5 9318 8
ENSG00000151929 BAG3 BAG cochaperone 3 2.72E- 90 3.825662
ENSG00000187134 AKR1C1 Aldo- keto reductase fam ily 1 member C1 7.03E- 90 6.43 5566
ENSG00000211445 GPX3 Glutathione peroxidase 3 2.7 6E- 73 3.7 92141
ENSG00000248713 C4orf54 Chromosome 4 open reading frame 54 7.4 1E - 7 3 7. 2 6 7 9 6 3
ENSG00000096384 HSP 90AB1 Heat shock prote in 90 alpha family class B member 1 1.1 8E - 67 2.4725 03
ENSG00000132002 DNAJB1 DnaJ heat shock protein f amily (Hsp 40) member B1 4.49E- 66 4.255973
ENSG00000013275 PSMC4 Proteasome 26S subunit, ATPase 4 3.71E- 64 2.34 3216
ENSG00000150991 UBC Ubiquitin C 1.8 3E- 61 3. 29 4192
ENSG00000272899 ATP6V1FNB (ATP6V1F neighbor) 6.66E- 61 5 .12 39 17
ENSG00000023909 GCLM Glutamate- cysteine ligase modifier subunit 2.35E- 60 3.747229
ENSG00000116161 CACYBP Calcyclin binding protein 7. 5 4 E- 5 6 2. 4192 54
ENSG00000197170 PSM D12 Proteasome 26S subunit, non- ATPase 12 1.96 E- 54 2 .62 7123
DEGs indicates differentially expressed genes.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 20
Forghani et al Carfilzomib- Induced Cardiotoxicity
function, abnormal Ca2+ transients, and changes in the
sarcomere structure and cell morphology that might
impact the contractile potency and traction forces of
hiPSC- CMs after carfilzomib treatment. Transcriptomic
and proteomic analyses highlighted an important role of
the genes involved in ECM signaling including integrin
Figure 8. Profiling of proteins after Cfz treatment by proteomics.
DEPs were identified with abundance change by >1.3- fold (absolute log2 [fold change] >0.38) compared with the DMSO- treated group (n=3
cultures). A and B, Volcano plot illustrating proteins in the cell lysate (right) and secreted proteins (left) with statistically significant abundance
differences. Significantly upregulated proteins are marked in blue, and significantly downregulated proteins are in red. C and D, Venn diagram
showing overlapping of DEGs and DEPs after Cfz treatment in the cell lysate and the secretome. E, Bubble plot represents selected GO terms
based on overlapped DEGs and DEPs with significant upre gulation. ANX6 indicates annexin 6; ATG101, autophagy related 101; CDKN1A, cyclin
dependent kinase inhibitor 1A; Cfz, carfilzomib; C1S complement C1s; COL1A2, collagen type I alpha 2 chain; COL3A1, collagen type III alpha 1
chain; COL3A1, collagen type III alpha 1 chain; DEGs, differentially expressed genes; DEPs, differentially expressed proteins; DMSO, dimethyl
sulfoxide; FLNC, filamin C; FN1, fibronectin 1; GO, gene ontology; HMGCR, 3- hydroxy- 3- methylglutaryl- CoA reductase; HSPA1A, heat shock
protein family A, Hsp70 member 1A; HSPA1A, heat shock protein family A, Hsp70 member 1A; HSPA1A, heat shock protein family A, Hsp70
member 1A; KCTD10, potassium channel tetramerization domain containing 10; MAFG, MAF bZIP transcription factor G; MFAP4, microfibril
associated protein 4; MSN, moesin; Not Found; PHPT1, phosphohistidine phosphatase 1; PKM, pyruvate kinase M1/2; POSTN, periostin;
RNA- seq, RNA sequencing; RND3, Rho family GTPase 3; SERPINE1, serpin family E member 1; ST13, ST13 Hsp70 interacting protein; TPM,
tropomyosin; VCP, valosin containing protein; and VCP, valosin containing protein.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 21
Forghani et al Carfilzomib- Induced Cardiotoxicity
and actin filaments. In parallel, our study revealed an
increase in genes associated with stress response (eg,
HSPs). These results suggest that cardiotoxicity can
result from the cumulative response to both off- target
and on- target effects of carfilzomib treatment.
Since the Food and Drug Administration’s ap-
proval of carfilzomib in 2012, there has been increasing
evidence surrounding carfilzomib- associated adverse
cardiovascular events including cardiac arrest and car-
diac arrhythmias.1,3,4,6 Carfilzomib is also considered
as an approved drug with repurposing potential for
mechano- based therapeutic interventions,41 although
integrin- mediated contraction alteration of cardiomy-
ocytes after carfilzomib treatment has not yet been
Figure 9. Profiling of protein changes in response to Cfz treatment in human induced pluripotent stem cell- derived
cardiomyocytes.
A and B, GO- term enrichment of the downregulated and upregulated proteins in the cell lysate. C, GO- term enrichment of the
downregulated secreted proteins. D, Pearson correlation of biological triplicate experiments of control group (C1– 3) and Cfz- treated
group (T1– 3). Pearson correlation coefficient (r) values are depicted for each group. Cfz indicates carfilzomib; and GO, gene ontology.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 22
Forghani et al Carfilzomib- Induced Cardiotoxicity
characterized. This is the first study that applies a pi-
conewton tension sensor to examine cytotoxic effects
of carfilzomib on hiPSC- CMs. Using the TGT probes,
we found that carfilzomib decreased integrin- mediated
traction forces of hiPSC- CMs as measured by changes
in fluorescence intensity on the probes because of con-
traction of individual cells. Our study provides a new av-
enue of mechano- pharmacology platform to study the
impact of carfilzomib at molecular and cellular levels.
Recent evidence illustrates off- target effects of
carfilzomib treatment in animal models; however,
mechanisms underlying contraction alteration after
carfilzomib treatment in cardiomyocytes have not been
fully characterized. Our findings indicate the possibil-
ity of on- target and off- target effects of carfilzomib on
human cardiac cells. Using RNA- seq, we identified
early transcriptomic signatures of carfilzomib- induced
cardiotoxicity and important genes/pathways that
might mediate carfilzomib- induced cell death and
contraction defects. Following a 24- hour treatment
of hiPSC- CMs with carfilzomib, there was significant
upregulation of genes and related proteins involved
Table 5. Top Downr egulated DEPs in Cell Lysate
Protein I D Protein symbol Desc ription of f ull name P value Log2 (fold change)
P28161 GSTM2 Glutathione S- transferase mu 2 0.000199 −0.37869
Q5XKP0 QIL1 (MICOS13) Mitochondrial contact site and cristae organizing system
subunit 13
0.000256 −0.38072
P14 618 PKM Pyruvate kinas e M1/2 0.000615 −0.3854
O150 67 PFA S Phosphoribosylformylglycinamidine synthase 0. 0 0 114 4 −0.39249
Q142 54 FLOT2 Flotillin 2 0. 0 0 116 5 −0 .39 616
Q9H0R4 HDHD2 Haloacid dehalogenase like hydrolase domain containing 2) 0.0 0178 − 0.39726
P29218 I MPA1 Inositol monophosphatase 1 0.0 0218 7 −0.39969
P3 00 41 PRDX6 Peroxiredoxin 6 0.00259 8 −0.40337
P08243 ASNS Asparagine synthetase (glutamine- hydrolyzing) 0.0 0319 8 −0.40509
P6 0 174 TPI1 Triosephosphate isomerase 1 0. 00 3513 −0.40568
Q3MHD2 LS M12 LSM12 homolog 0.004002 −0.40938
P485 09 C D151 CD151 molecule (Raph blood group) 0.004477 − 0.41 01
P15259 PGAM2 Phosphoglycerate mutase 2 0.005065 − 0.41576
Q168 36 HADH Hydroxyacyl- CoA dehydrogenase 0.00 5345 − 0.41756
Q9 95 41 PLIN2 (Perilipin 2) 0.0 06138 − 0.4 2144
Q9Y235 APOBEC2 Apolipoprotein B mRNA editing enzyme catalytic subunit 2 0.0066 33 −0. 42492
Q13813- 2 S PTA N1 Spectrin alpha, non- erythrocytic 1 0.0068 85 −0.43392
O60669 SLC 16A7 Solute carrier family 16 membe r 7 0.0 08852 −0.43623
Q96A X9 MIB2 Mindbomb E3 ubiquitin protein ligase 2 0 .0 09 051 −0.44424
P515 53 IDH3G Isocitrate dehydrogenase (NAD (+) 3 non- cataly tic subunit
gamma)
0.009335 −0.44433
P0 8133 ANX A6 Annexi n A6 0.0096 85 −0.466 95
P17174 G OT1 Glutamic- oxaloacetic transaminase 1 0.011001 −0. 47138
P1010 9 FDX1 (Ferredoxin 1) 0.0143 75 −0 .4754 3
P430 07 SL C1A4 Solute carrier family 1 member 4 0.0148 45 − 0.48642
P30046 DDT D- dopachrome tautomerase 0.017 05 8 −0.5028 3
Q133 15 ATM ATM serine/threonine kinase 0. 0171 −0.50678
P34949 MPI Mannose phosphate isomerase 0.0183 24 −0.508 84
Q8IYM0 FAM186B Family with se quence similarity 186 member B 0.01849 6 −0 .515 45
Q9NX18 SDHAF2 Succinate dehydrogenase complex assembly factor 2 0. 021519 −0.51609
P80723 BA SP1 Brain abundant membrane attached signal protein 1 0.022129 −0.52675
O75223 GGCT Gamma- glutamylcyclotransferase) 0.022945 −0.53503
P17612 PRK ACA Protein kinase cAMP- activated catalytic subunit alpha 0. 02751 − 0.55372
P0 0 3 74 DHFR Dihydrofolate reductas 0.0275 65 −0.5787
Q8N 111 CEND1 Ce ll cycle exits a nd neurona l differe ntiation 1 0.033353 −0 . 61 374
B7Z596 TPM1 Tropomyosin 1 0.033428 − 0.72 819
DEPs indicates differentially expressed proteins.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 23
Forghani et al Carfilzomib- Induced Cardiotoxicity
in cellular stress. In particular, RNA- seq revealed that
carfilzomib treatment induced the overexpression of
genes encoding HSPs including HSPA1B , HS PA1A ,
and HSPA6. Consistent with our RNA- seq data, the
proteomic analysis also confirmed upregulation of
HSPs in both the cell lysate and the secretome, includ-
ing HSPAB1, and HSPA1A as well as molecules that
act as cochaperones including DNAJA&B, HSPH1,
Table 6. Top Upregulated DEPs in Cell Lysate
Protein I D Protein symbol Description of full nam e P value Log2, fold change
P1706 6 HSPA6 Heat shock protein family A (Hs p70) member 6 5.85E- 06 2. 920 214
Q8N6M9 ZFAN D2A Zinc finger AN1- type containing 2A 8.86E- 06 2.792291
P188 47 ATF 3 Activating transcription factor 3 2.03 E- 05 2.6 55 0 11
Q153 27 ANK RD1 Ankyrin repeat domain 1 2.93E- 05 2.301373
O76080 ZFAND5 Zinc finger AN1- type containing 5 5.43E- 0 5 2.210395
P0 8107 HSPA1A Heat shock protein family A (Hsp70) member 1A 6.26E- 0 5 2 .16 678 3
P38936 CD KN1A Cyclin dependent kinase inhibitor 1 7.03E- 05 2.13 516 6
Q9H0R8 GABARAPL1 GABA type A rece ptor associa ted protein like 1 8. 41E- 05 1. 98 5212
P25685 DNAJB1 DnaJ heat shock p rotein family (Hsp40) mem ber B1 0.000163 1.945932
P04035 HMGCR 3- hydroxy- 3- methylglutaryl- CoA reductase 0.000211 1.85 5423
P615 87 RND3 Rho fa mily GTPase 3 0.000249 1.6 49824
Q01581 HMGCS1 3- hydroxy- 3- methylglutaryl- CoA synthase 1 0.000271 1.626212
O155 25 MAFG MAF bZI P transcription factor G 0.000284 1.601714
Q9H3F6 KCTD10 Potassium channel tetramerization domain
containing 10
0.00037 1.507235
Q9UDY4 DNAJB4 DnaJ heat sh ock protein family (Hsp40) member B4 0.000379 1.484945
P54868 HMGCS2 3- hydroxy- 3- methylglutaryl- CoA synthase 2 0.000395 1.463084
P02792 FTL Ferritin light chain 0.000444 1.432189
Q9UBU8 MORF4L1 Mortality factor 4 like 1 0.000498 1.424 01
Q9BYN0 SRX N1 Sulfiredoxin 1 0.000514 1.416 93 7
Q92963 RI T1 Ras like without CAA X 1 0.000551 1.3 41017
P02 511 CR YAB (Crystallin alpha B) 0.000627 1.330892
Q9UH92 MLX MAX dimerization protein MLX) 0.000713 1.30305
Q92598 HSPH1 (Hea t shock protein family H (Hsp110) member 1) 0.00 072 1.28 85 95
Q99608 NDN Necdin, M AGE family me mber 0.000767 1.265368
Q9BY42 RTFDC1 Replication termination factor 2 0.000826 1.221335
Q135 01 SQS TM1 Sequestosome 1 0.000845 1.207748
P78362 SRPK2 SRSF protein kinase 2 0.000887 1.198661
Q9Y5A7 NUB1 Negative regulator of ubiquitin like proteins 1 0.000906 1.19 413 1
P6 8133 A CTA1 Actin alpha 1, skeletal muscle 0.000913 1.178388
O60333- 3 K IF1B Kinesin family member 1B 0.000962 1.153 8 91
P62699 YPEL5 Y ippee like 5 0.0 0105 8 1.12 52 07
O75794 C DC12 3 Cell div ision cycle 12 0. 0010 82 1.0 97219
O9 58 17 BAG3 BAG cochaperone 3 0.001229 1.091382
O4 3164 PJA 2 Praja ring finger ubiquitin ligase 2 0.0 01327 1.0 58091
Q07352 ZFP36L1 ZFP36 ring f inger protein like 1 0.0 0139 1. 04 69 27
O14950 M Y L12B Myosin light ch ain 12B 0.0 0160 2 1.007005
Q9UGL1 KDM5B Lysine demethylase 5B 0.0016 48 0.977882
P248 44 MYL9 Myosin light cha in 9 0.0 01717 0 .966 276
Q9UK73 FEM1B Fe m- 1 homo log B 0. 0 0174 0.953693
Q8TEY7 USP33 Ubiquitin specific peptidase 33 0 .0 0175 2 0.948084
P0 CG47 UBB Ubiquitin B 0.001801 0.928962
P46976 GYG1 Glycogenin 1 0.001869 0.92 6016
DEPs indicates differentially expressed proteins.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 24
Forghani et al Carfilzomib- Induced Cardiotoxicity
ZFAND2A, and CRYAB. It has been reported that pro-
teasome inhibitors such as bortezomib can upregulate
HSPs,42 and targeting HSPs has been considered to
overcome resistance to chemotherapeutic agents such
as imatinib and cisplatin.43 HSP90 has been found to
assist in degrading toxic metabolites by promoting
Table 7. Top Dysregulated DEPs in Secretome
Protein I D Protein symbol Description of full name P value Log2, fold change
P010 09 S ER PI NA1 Serpin family A member 1 0.0 0 2177 −1. 018 51
P0 9871 C1S Complement C1s 0.0 05218 −0 .71786
Q9UBP4 DKK3 Dickkop WNT signaling pathway inhibitor 3 0.00526 0 .74 3 39
P0 8107 HSPA1A Heat shock protein family A (Hsp70) member 1A 0.005936 1.11 59 0 6
Q150 63 POST N Periostin 0.006501 −0.55999
O6 08 14 HIS T1H2BK H2B clustered 0.006 567 −0 .27213
P55083 M FAP4 Microfibril associated protein 4 0.007602 − 0. 6915 8
Q86UP2 KT N1 Kinectin 1 0 .0 112 08 0.15 2319
Q128 41 FS T L1 Follistatin like 1 0 .012 976 −0. 5254
P08238 HSP90AB1 Heat shock protein 90 alpha family class B mem ber 1 0. 013136 − 0 .13 5 6 6
P02461 COL3 A1 Collagen type III alpha 1 chain 0. 01317 9 −0.5 5331
P26038 MSN Moesin 0.013885 0.42732
O95373 I PO7 Importin 7 0. 014671 −0.63705
P11142 HSPA8 Heat shock protein famil y A (Hsp70) member 8 0.016105 0. 36 4134
P31946 YWHAB Tyrosine 3- monooxygenase/tryptophan
5- monooxygenase activation protein beta)
0.016 69 6 0.35428
P0 8123 C O L1A2 Collagen type I alpha 2 chain 0. 017711 − 0.45736
Q14 315 FLNC Filamin C 0.0 214 56 0.693258
P0 2751 FN1 Fibronectin 1 0.0 2149 3 −0. 5477 9
P139 29 ENO3 Enolase 3 0.024 012 −0.33332
Q7Z7M1 G PR 144 ADGRD2- adhesion G protein- coupled receptor D2 0.024362 3.30262
P55072 VCP Valosin containing protein 0.025906 0.830314
P0 5121 SE RPI NE1 Serpin family E member 1 0.027676 0. 43316 7
P26038 MSN Moesin 0.013885 0.42732
P50502 ST13 ST13 Hsp70 interacting protein 0.044477 0.4126 65
DEPs indicates differentially expressed proteins.
Table 8. Downregulated DEPs in Secretome and Their Association With KEGG Pathways
Gene ID Protein symbol Descr iption of full name P value
Log2, fold
change KEGG pathway
ENSG00000177943 MAMDC4 MAM doma in containing 4 4.94E- 56 − 9 .19 9 0 9 None
ENSG00000171992 SYNPO Synaptopodin 8.8 4E - 51 −6.60094 N/A
ENSG00000104738 MCM4 Minichromosome
maintenance complex
component 4
1.11E - 50 −6.26153 (ko04530) Tight junction; (hsa04530) Tight junction
ENSG00000137809 ITG A11 Integrin subunit alpha 11 2.58E - 48 − 6 .19 6 8 7 (ko03030) DNA replication; (hsa03030) DNA
replication; (ko04110) Cell cycle
ENSG00000184347 SLIT3 Slit guidance ligand 3 5.1 3E - 46 −6 .15 7 78 (ko05412) Arrhythmogenic right ventricular
cardiomyopathy (ARVC); (hsa05412) Arrhythmogenic
right ventricular cardiomyopathy; (ko04512) ECM-
receptor interaction
ENSG00000156427 FGF18 Fibroblast growth factor 18 6.1 4E - 43 −6 .15 0 8 7 (ko04360) Axon guidance; (hsa04360) Axon guidance
ENSG00000185567 AHNAK2 AHNAK nucleoprotein 2 3.69 E- 39 −6.0 8355 (ko05218) Melanoma; (hsa05218) Melanoma;
(ko05224) Breast c ancer
ENSG00000128951 DUT Deoxyuridine triphosphatase 7.1 8 E- 3 9 − 6.01202 N /A
ENSG00000225138 SLC9A3- AS1 None 1.01E- 38 −5.97976 (ko00240) Pyrimidine metabolism; (hsa00240)
Pyrimidine metabolism; (hsanan01) drug metabolism
DEPs indicates diffe rentially expressed p roteins; and KEGG, Kyoto Encyclopedi a of Genes and G enomes.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 25
Forghani et al Carfilzomib- Induced Cardiotoxicity
Table 9. GO Terms and KEGG Pathways Associated With Overlapped Genes/Proteins Identif ied by DEPs From Cell Lysate
and DEGs by RNA- seq
Description Category Count GO/KEGG term ID Adj usted P value Enrichment
Response to heat GO term biological processes 15 GO:0009408 1E- 15 40
Heat shock protein binding GO term molecular functions 10 GO:0 031072 7.9E - 1 0 35
ATPase regulator activity GO term molecular functions 6GO:0060590 5E- 07 65
ATPase activator activity GO term molecular functions 4G O:0 001671 0.00016 71
Ferroptosis KEGG pathway 4ko04216 0.00079 44
Necroptosis KEGG pathway 6hsa04217 0.00079 16
Tau protein binding GO term molecular functions 4GO: 00 4815 6 0.001 39
Myeloid c ell activati on involved in
immune response
GO term biological processes 9GO:0002275 0.00 126 7. 3
Chaperone- mediated autophagy GO term biological processes 3GO: 00 6168 4 0.0 0158 83
Positive regulation of ATPase activity GO term biological processes 4GO:0 032781 0.0 02 32
Apoptotic signaling pathway GO term biological processes 9GO :009719 0 0.0 02 51 6.5
ATPase activity GO term molecular functions 7GO:0 0168 87 0.0 0501 8.3
Regulation of ATPase activity GO term biological processes 4GO:0043462 0.00 631 22
Cellular response to interleukin 1 GO term biological processes 5GO :00 71347 0.01 12
Interleukin 1– mediated signaling
pathway
GO term biological processes 4GO:0 070498 0 .012 59 17
Stress- activated MAPK cascade GO term biological processes 5G O: 00 5140 3 0.0 50 12 7.7
Regulation of extrinsic apoptotic
signaling pathway
GO term biological processes 4GO:2001236 0.0 5012 11
Intrinsic apoptotic signaling pathway GO term biological processes 5GO :00 97193 0 .05012 7. 7
Regulation of intrinsic apoptotic
signaling
GO term biological processes 4GO:20 01242 0.06 31 11
Respon se to oxidative str ess GO Biological Processes 6GO:0006979 0.06 31 5.8
Extrinsic apoptoti c signalin g pathway
via death domain receptors
GO term biological processes 3GO:0008625 0.1 15
Regulation of reactive oxygen
species metabolic process
GO term biological processes 4GO:2000377 0.1 8.9
Autophagy GO term biological processes 6GO:0006914 0.1174 9 4.9
Positive re gulation of cell death GO term biological processes 7GO :0010 942 0 .12 58 9 4.2
Pathways in ca ncer KEGG pathway 6hsa05200 0.1513 6 4.6
Mitogen- activated protein kinase
(MAPK) signaling pathway
KEGG pathway 4k o04 010 0.19 95 3 7
Apoptotic mitochondrial changes GO term biological processes 3GO:0008637 0.2138 11
ATPase activity, coupled GO term molecular functions 4GO:0042623 0. 2138 6.8
Macro autophagy GO term biological processes 4GO:0016236 0.35481 5.6
Cellular senescence KEGG pathway 3hsa04218 0.42658 7.7
Oxidoreductase activity GO molecular functions 6GO :00 16491 0.43652 3.5
PI3K- Akt signaling pathway KEGG pathway 4ko0 4151 0.4466 5.2
Regulation of Ras protein signal
transduction
GO term biological processes 3GO:0046578 0.56234 6.8
Regulation of MAPK cas cade GO term biological processes 5GO:0043408 0. 02 5119 2.9
Wnt signaling pathway GO term biological processes 4GO :00 1605 5 0.0 31623 3.4
Cell- cell signaling by wnt GO term biological processes 4GO:019 8738 0 .0316 23 3.4
Retina- expressed kinase (ERK )1 and
ERK2 cascade
GO term biological processes 3GO :0070371 0 .0 3 9811 4
Regulation of mitogen- activated
protein kinase activity
GO term biological processes 3GO:0043405 0. 0 39 811 3.8
DEGs indi cates differentiall y expressed genes; DEPs, di fferenti ally expressed proteins; GO, gene ontology; KEGG, Kyoto Encyclopedia of Gene s and
Genomes; and RNA- seq, RNA sequencing.
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 26
Forghani et al Carfilzomib- Induced Cardiotoxicity
ubiquitination and proteasome lysis.44 Overexpression
of HSPs is also linked to many pathological con-
ditions,45,46 and increased HSPs have a protective
function by blocking apoptosis, which allows cells to
survive in otherwise lethal conditions.47, 4 8 It is likely that
increased HSPs may overcome the toxic effect of the
inhibition of proteasomes after carfilzomib treatment
in hiPSC- CMs. Given that increased HSPs are detect-
able in the secretome following the carfilzomib treat-
ment, these HSPs may serve as biomarkers for early
detection of carfilzomib- induced cardiotoxicity.
Our results highlight the possible role of the ECM and
integrins in contractility defects as an early response to
carfilzomib treatment. Alterations in ECM can impact
cardiomyocyte function.49,50 ECM and secreted pro-
teins are important components of the stroma, which
could play important roles in the cell- cell communica-
tion and regulation of cell process including contrac-
tion.51 Integrins are sensors that transmit mechanical
signals to ion channels.52 The results of the GO term
analysis indicate that genes belong to the ECM, inte-
grin complex, and actinin cytoskeleton proteins were
downregulated in carfilzomib- treated cells. Carfilzomib
treatment also downregulated genes associated with
cardiac muscle cell contraction, including contractile
proteins (eg, MYL2 and MYH6) and Ca2+ handing pro-
teins. In line with RNA- seq data, our proteome data
in both cell lysate and secretome also indicate signifi-
cant reduction of proteins related to both contraction
and ECM including POSTN, COL3A1, THBS1, C1S,
ANXA6, and TPM.
The current results demonstrate that carfilzomib
interferes with mitochondrial function as indicated by
reduced ATP production and respiration, increased
mitochondrial ROS, and decreased mitochondrial
membrane potentials. The reduced mitochondrial
function and cellular energy impairment may contrib-
ute to cardiomyocyte dysfunction, which is consistent
with previously published in vivo off- target effects of
carfilzomib.53 Increased mitochondrial ROS is known
to be associated with cytotoxicity.35,38,54 We also found
similar results of increased mitochondrial ROS by treat-
ing hiPSC- CMs with doxorubicin, a cancer therapy
drug with well- known cardiovascular toxicity effect be-
cause of the generation of ROS,55,56 reduced contrac-
tile capacity,57 and adverse effect on Ca2+ transport.31
The increased mitochondrial ROS and decreased
mitochondrial function in carfilzomib- treated hiPSC-
CMs were also associated with the downregulation of
genes associated with mitochondrial function such as
nicotinamide adenine dinucleotide (NADH) ubiquinone
oxidoreductase subunit AB1, which is a crucial regu-
lator of mitochondrial energy and ROS metabolism
through coordinating the assembly of respiratory com-
plexes.58 In addition, our proteomic analysis indicated
downregulation of SDHFA2, a member of succinate
dehydrogenase complex, which plays essential roles
in both the mitochondrial electron transport chain and
the tricarboxylic acid cycle.
Consistent with the findings on increased mitochon-
drial oxidative stress, defects in mitochondrial function,
and increased gene expression and proteins related
to response to oxidative stress in carfilzomib- treated
hiPSC- CMs, we found that cell death in carfilzomib-
treated hiPSC- CMs could be rescued by targeting
oxidative stress with ascorbic acid. These results sug-
gest that oxidative stress plays an important role in
carfilzomib- induced cytotoxicity and that antioxidants
have the potential for cardioprotective therapy.
The carfilzomib- induced mitochondrial changes
may lead to other functional alterations including ab-
normal Ca2+ signaling. On the other hand, abnormal
Ca2+ signaling can induce alterations in respiratory
chain complexes, which lead to increased mitochon-
drial oxidative stress and reduced mitochondrial
membrane potential.35,37 Our results showed a sig-
nificant increase in Ca2+ transient abnormalities after
carfilzomib treatment, which is consistent with clini-
cal observations showing arrhythmias after chemo-
therapeutic drugs as a common phenomenon.40,59 In
pathological conditions, Ca2+ overload is accompa-
nied by alterations in mitochondrial function and mi-
tochondrial membrane potential,37,38,6 0 and abnormal
Ca2+ transients can be an indicator for proarrhythmic
behavior of cardiomyocytes.61 In cardiomyocytes, the
interplay between Ca2+ signaling and mitochondrial
function is necessary to maintain normal cardiac func-
tion.22,37 Carfilzomib treatment of hiPSC- CMs signifi-
cantly downregulated several genes encoding Ca2+
handling proteins, including RYR2 and CASQ2, which
play a critical role in excitation– contraction coupling in
cardiomyocytes. Consistently, carfilzomib treatment
increased abnormal Ca2+ transients in hiPSC- CMs.
Our findings support that possible mechanisms of
carfilzomib- induced cardiotoxicity are associated with
abnormal Ca2+ signaling and mitochondrial oxidative
stress, contributing to the development of arrhythmias
after carfilzomib treatment.
We also found that carfilzomib treatment reduced
the expression of proteins associated with metabolic
process. For example, carfilzomib reduced the expres-
sion of glucose metabolism- related enzymes including
PKM2 and PRKACA. Loss of function studies show
that PKM2 deletion in cardiomyocytes resulted in sig-
nificantly reduced cell cycle.62 PRKACA is a key regu-
latory enzyme in humans and contributes to the control
of cellular processes including glucose metabolism
and cell division. Defective regulation of protein kinase
A activity is also linked to the progression of cardiovas-
cular disease, and reduced protein kinase A activity is
associated with reduction of Ca2+ signaling in embry-
onic hearts.63 In addition, carfilzomib dysregulated the
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 27
Forghani et al Carfilzomib- Induced Cardiotoxicity
expression of proteins involved in autophagy, apop-
tosis, and cell cycle. For example, carfilzomib down-
regulated CLEC16A, a regulator of autophagy through
mTOR (the mammalian target of rapamycin) activity,64
and upregulated ATG101 (an autophagy factor required
for autophagosome formation), CDKN1A (a regulator of
cell cycle in response to stress stimuli), and RND3 (a
member of the small Rho GTPase family that regulates
apoptosis through the Rho kinase- dependent signal-
ing pathway).
Considering the complexity of cardiotoxicity, several
possible mechanisms can synergistically induce cyto-
toxic effects after carfilzomib treatment. As expected
from its on- target effect, proteasome inhibition by car-
filzomib treatment can promote cell death. Our study
also reveals potential novel mechanisms induced by
carfilzomib- treatment, including (1) targeting mitochon-
dria that may lead to oxidative stress, disruptions in
cellular energy, and contractility defect; and (2) trigger-
ing downregulation of ECM and integrin- related genes
that may result in reduced integrin- mediated traction
forces and alterations of cell structure and morphology.
We note that a replication study in human patients
is an important next step to examine if the alterations
of genes and proteins observed in hiPSC- CMs are also
observed in human patients. However, given the simi-
lar end points observed in hiPSC- CMs and human pa-
tients (eg, cytotoxicity and abnormal contractility), the
findings of hiPSC- CMs from this study may have im-
portant implications toward discovery of new therapies
to overcome clinical side effects after chemothera-
peutic drug treatment. In addition, molecular changes,
such as upregulation of heat shock- related genes and
other proteins, may provide a new avenue toward bio-
marker development for early detection of carfilzomib-
induced cardiotoxicity.
ARTICLE INFORMATION
Received April 27, 2021; accepted Oc tober 1, 2021.
Affiliations
Divisi on of Pediatri c Cardiolog y, Department of Pediatric s, Emory University
School of M edicine and Children’s Healthcare of Atlanta, Atlanta, G A (P.F.,
R.L., D.L., M.R.L., H.H., J.T.M., C.X.); Biomolecular Che mistry, Depa rtment
of Chemis try, Emory University, Atlanta, GA (A.R., K.S.); Scho ol of Chemistry
and Biochemistry and the Petit Institute for Bioengineering and Bioscience,
Georgia Institute of Technology, Atlanta, GA (F.S., R.W.); Department of
Medici ne & Winship Cancer Institute, Emory Universit y School of Medicine,
Atlanta, G A (A.M.); and Wallace H. Coulter Dep artment of B iomedical
Engineering, Georgia Institute of Technology and Emory University, Atlanta,
GA (K.S., C.X.).
Acknowledgments
Drs Forgh ani, Mandawat, and Xu concei ved and desi gned resea rch. Dr
Forghani, A. Rashid, Dr Sun, M. R. Lee, and D r Li performed resea rch and
acquire d the data. Dr Forghani, A. Rashid, Dr Sun, R. Liu, M.R. Lee, Dr Li, H.
Hwang, and D r Maxwell analyze d and interpre ted the data. Dr Fo rghani, Dr
Sun, and R. L iu perfor med statisti cal analysis. Dr Mandawat provided clinical
guidance. Drs Wu and Salaita contr ibuted new rea gents or analytic tools,
interpreted the data, and m ade critic al revision of the articl e for important
intellectual content. Drs Forghani and Xu wrote the ar ticle and made critica l
revisio ns of the article for impor tant intell ectual con tent. Dr Xu handl ed fund-
ing and supervision.
Sources of Funding
This stud y was suppor ted by the Children’s Heart Re search and O utcomes
Center at Em ory Unive rsity and Children’s Healthcare of Atlanta; the
Center for Pediatric Technology and Bio locity at Em ory Unive rsity and
Georgia Institute of Technology; Imagin e, Innovate and Impact (I3) Funds
from the Emory Schoo l of Medicine and through th e Georgia Cl inical and
Translational Science Awa rds Program (the National Institutes of He alth
award) (UL1- TR002378); the National Institute s of Health (R21AA025723,
R01HL136345, and R01AA028 527); and the Nationa l Science Fo undation
and the Center for the Advancement of Sc ience in Spa ce (CBET 1926387).
Disclosures
None.
REFERENCES
1. Cole DC, Frishma n WH. Cardiovas cular comp lications of p roteasome
inhibitors used in multiple myeloma. Cardiol Rev. 2018;26:122– 129. doi:
10.1097/CRD.00000 00000 000183
2. Kuhn DJ, Chen Q, Voorhees PM, S trader JS, Shenk KD, Sun CM, Demo
SD, Bennet t MK, van Leeu wen FWB, Chanan- K han AA, et a l. Potent
activity of carfilzomib, a novel, irreversible inhibitor of the ubiquitin-
proteasome pathway, against preclinical models of multiple myeloma.
Blood. 2007;110:3281– 3290. doi: 10.1182/blood - 2007- 01- 065888
3. Dimopoulos MA, Roussou M, Gavriatopoulou M, Psimenou E, Ziogas
D, Eleutherakis- Papaiakovo u E, Fotiou D, Migkou M, Ka nellias N,
Panagiotidis I, et al. Cardiac and renal c omplicati ons of carf ilzomib
in patients with multiple myeloma. Blood Adv. 2017;1:449– 454. doi:
10.1182/blood advan ces.20160 03269
4. Mangla A, Payd ary K, Liu J, Mbachi C, Yim B, Lad TE. Carfil zomib-
associ ated cardiova scular adve rse events in a non- cauca sian cohor t
of patients w ith multipl e myeloma: a real- wor ld experie nce. Hematol
Oncol. 2018;36:715– 717. do i: 10.1002/hon.2535
5. Morea u P, Richardson PG, Cavo M, Or lowski RZ, San Mig uel JF, Palumbo
A, Harousseau JL. Proteasome inhibitors in multiple myeloma: 10 years
la te r. Blood. 2012;120:947– 959. doi: 10.1182/blood - 2012- 04- 403733
6. Jai n T, Narayana samy H, Mikhael J, Reeder C B, Bergsage l PL, Mayo A,
Stewart A K, Mookada m F, Fonseca R. Systolic dysfunction associated
with carfilzomib use in patients with multiple myeloma. Blood Cancer J.
20 17;7:6 42. doi: 10 .103 8/ s414 0 8- 017- 0026 - 7
7. Bruno G, Bringhe n S, Maffei I, Iannaccon e A, Crea T, Ravera A, Astarita
A, Vallelo nga F, Salvini M, Gay F, et al. Cardiova scular organ damage
and blood pressure levels predict adverse events in multiple myeloma
patients u ndergoing carfilzomib therapy. Cancers (Basel). 2019;11:622.
doi: 10. 339 0/canc e rs110 5 06 22
8 . Shah C, Bish noi R, Jain A, Bej janki H, Xiong S, Wa ng Y, Zou F, Moreb JS.
Cardiotoxi city asso ciated with carfilzomib: systematic review and meta-
analys is. Leuk Lympho ma. 2018;59:2557– 2569. doi: 10.1080/10428
194.2018.1437269
9. Waxm an AJ, Clasen S, Hwa ng W- T, Garfa ll A, Vogl DT, Carver J, O’Quinn
R, Cohen A D, Stadtmauer EA, Ky B, et al. Carfilzomib- as sociated ca r-
diovascu lar advers e events: a systematic review and meta- ana lysis.
JAMA Oncol. 2018;4:e174519. doi: 10.1001/ja mao n col.2017.4519
10. Kortuem KM, Stewart AK. C arfilzomib. Blood. 2013;121:893– 897. doi:
10.1182/blood - 2012- 10- 459883
11. Hahn VS, Zhang K W, Sun L, Narayan V, Leniha n DJ, Ky B. Heart failure
with targ eted cancer therapies: m echanism s and cardio protection. Ci rc
Res. 2021;128:1576– 1593. do i: 10.1161/CIRCR ESAHA .121.318223
12. Han B, Yao W, Oh YT, Tong JS, Li S, Deng J, Yue P, Khuri FR, Sun SY.
The novel proteasome inhibitor carfilzomib activates and enhances ex-
trinsic a poptosis involving stabilization of death receptor 5. O ncotarget.
2015;6:17532– 17542. do i: 10.18632/ oncot arget.3947
13. Hasinoff BB, Patel D, Wu X. Molecul ar mechanisms of the card iotox-
icity of the proteasomal- targeted drugs bortezomib and carfilzomib.
Cardiovas c Toxicol. 2017;17:237– 250. doi: 10.1007/s1201 2- 016- 9378- 7
14. Efentakis P, Kremastiotis G, Varela A, Nikolaou P- E, Papanagnou E-
D, Davos CH, Tsoumani M, Agrogiannis G, Konstantinidou A, Kas tritis
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 28
Forghani et al Carfilzomib- Induced Cardiotoxicity
E, et al. Molecular mech anisms of carfilzomib- i nduced cardiotoxicity
in mice and the emergin g cardioprotective role of metformin. Blood.
2019;133:710– 723. doi: 10.1182/blo od - 2018- 06- 858415
15. Musun uru K, Sheik h F, Gupta RM, Ho user SR, Mah er KO, Milan DJ,
Terzic A, Wu JC; American Heart Association Council on Functional
G, Translation al B. Induced pluripotent stem cells for c ardiovascu-
lar disease modeling and precision medicine: a scientific statement
from the Am erican He art Asso ciation. Circ Genom Prec is Med.
2018;11:e000043. doi: 10.1161/HCG.00000 00000 000043
16. Mordwinki n NM, Burrid ge PW, Wu JC. A review of human pluripote nt
stem cell- derived cardiomyocytes for high- throughput drug discov-
ery, cardiotoxicity screening, and publication standards. J Cardiovasc
Transl Res. 2013;6:22– 30. doi: 10.1007/s1226 5- 012- 9423- 2
17. Hnatiuk AP, Briganti F, Staudt DW, Merco la M. Human iPSC modeling of
heart disease for d rug development. Cell Chem Biol. 2021;28:271– 282.
doi: 10.1016/j.chemb iol.2021.02.016
18. Bellin M, Mummer y CL. Inheri ted heart d isease— what can we ex-
pect from the second decade of huma n iPS cell research? FEBS Lett.
2016;590: 2482– 2493 .
19. Mercola M, C olas A, Willems E. Induced pluripote nt stem cells in c ar-
diovascular drug discovery. Circ Res. 20 13 ;112 :5 3 4– 548 . d o i: 10 .116 1/
CI R C R E S AH A .111. 25 026 6
20. Nair P, Prado M, Perea- Gil I, Ka rakikes I. Co ncise review: precision
matchmak ing: induce d pluripotent stem cells me et cardio- onco logy.
Stem Cells Transl Med. 2019;8:758– 767. doi: 10.1002/sctm.18- 0279
21. Singh P, Wang X, Hageman L, Chen Y, Magdy T, Landier W, Ginsbe rg
JP, Neglia JP, Sklar CA, Castel lino SM, et al. As sociation of G STM1
null variant with anthracycline- related cardiomyopathy after child-
hood can cer- a Child ren’s Oncology G roup ALTE03N1 repor t. Cancer.
2020;126:4051– 4058. doi: 10.1002/cncr.32948
22. Lan F, Lee A, Li ang P, Sanchez- Freire V, Nguyen P, Wang LI, Han L, Yen
M, Wang Y, Sun N, et al. Abnormal calcium handlin g properties under-
lie familial hyper trophic cardiomyopathy pathology in patient- specif ic
induce d pluripotent stem cells. Ce ll Stem Cell. 2013;12:10 1– 113. do i:
10.1016/ j.st em. 2012.10.010
23. Jha R, Wu Q, Singh M, Preining er MK, Han P, Ding G, Cho HC, Jo H,
Maher KO, Wagne r MB, et al. Simul ated microgravity and 3D c ulture
enhance induction, viability, proliferation and differentiation of cardiac
progenitors from human pluripotent stem cells. Sci Rep. 2016;6:30956.
doi: 10.1038/srep3 0956
24. Billis P, Will Y, Nadanaciva S. High- content imaging assays for identi-
fying compounds that generate superoxide and impair mitochondrial
membra ne potential in adherent eu karyotic cells. Curr Protoc Toxicol.
20 14;5 9: 25 .1.1 – 2 5.1.14. d oi : 1 0.100 2 /0 47 11 4 08 5 6. t x2 5 0 1s 5 9
25. Rampoldi A , Crooke SN, Preininger MK, Jha R, Max well J, Ding L,
Spearman P, Finn MG, Xu C. Targeted e limination of tumorige nic human
pluripotent stem cells using suicide- inducing virus- like particles. ACS
Chem Biol. 2018;13:2329– 2338. d oi: 10.1021/acsch emb io.8b0049 0
26. Hu ebsch N, Losk ill P, Mandegar MA, M arks NC, Shee han AS, Ma Z,
Mathur A, N guyen TN, Yoo JC, Judge LM, et al. Automated video-
based an alysis of contractilit y and calciu m flux in human- induced plu-
ripotent s tem cell- d erived cardiomyocytes cul tured over diffe rent spatial
scales. Tissue Eng Par t C Methods. 2015;21:467– 479. doi: 10.1089/ten.
te c. 2014 .02 83
27. Gao W, Shi P, Chen X, Zh ang L, Liu J, Fan X, Luo X. Clathri n- mediated
integrin alphaIIbbeta3 trafficking controls platelet spreading. Platelets.
2018;29:610 – 621.
28. Liu Y, Blanchfield L, Ma VP, Andargachew R, G alior K, Liu Z, Evavold
B, Salait a K. DNA- base d nanopar ticle tensi on sensors reveal that T-
cell rec eptors transmit define d pN forces to their antigens fo r enhanced
fi de lit y. Proc Natl Acad Sci USA. 2016;113:5610– 5 615. doi: 10.1073/
pna s .16 00 1 6 311 3
29. Ma R, Kell ner AV, Ma VP, Su H, Deal BR, Brockman JM, S alaita K. DN A
probes th at store mecha nical infor mation revea l transient piconewton
forces ap plied by T cells. Proc Natl Acad S ci USA. 20 19;116:169 49 –
16954. doi: 10.1073/pnas.19040 34116
30. Liu R, Sun F, Forghani P, Armand LC, Ramp oldi A, Li D, Wu R, Xu C.
Proteomic p rofiling reve als roles of stress response, Ca(2+) transient
dysregulation, and novel signaling pathways in alcohol- induced cardio-
toxicity. Alcohol Clin Exp Re s. 2020;44:2187– 2199.
31. Hanna AD, Lam A, Tham S, Du lhunty AF, Beard NA . Adverse ef fects of
doxorubicin and its metabolic product on cardiac RyR2 and SERCA2A.
Mol Pharmacol. 2014 ;8 6: 4 38 – 4 49. do i: 10 .1124 /mo l.114 .0 9384 9
32. Takemura G, Fujiwara H. Doxorubicin- induced cardiomyopathy from
the cardi otoxic mechanisms to manag ement. Prog Ca rdiovasc Dis.
2007;49:3 30 – 3 52. d oi: 10.1016/j.pcad.2 00 6.10.002
3 3. Ma illet A, Tan K, Chai X , Sadanand a SN, Mehta A, Ooi J, Hayden MR,
Pouladi MA, Ghosh S, Shim W, et al. Modeling d oxorubicin- induced
cardiotoxicity in human pluripotent stem cell derived- cardiomyocytes.
Sci Rep. 2016;6:25333. doi: 10.1038/srep2 5333
34. Polonchuk L, Cha bria M, Badi L, H oflack JC, Fig tree G, Davies MJ,
Gentile C. Cardiac sphe roids as prom ising in vitro models to stud y the
human heart microenvironment. Sci Rep. 2017;7:7005. doi: 10.1038/
s4159 8- 017- 06385 - 8
35. Starkov AA. The role of mitoc hondria in reacti ve oxygen specie s metab-
olism and signaling. Ann N Y Acad Sci. 20 0 8; 1147: 37 – 5 2. doi : 1 0.11 96 /
annals.1427.015
36. Bernardi P, Di Lisa F, Fogolar i F, Lippe G. From ATP to PTP and
back: a dua l function fo r the mitochon drial ATP synthas e. Circ Res.
20 15 ;116 :18 5 0 – 18 6 2. doi : 1 0.11 61/ C IR C R E S AH A .115 .30 6 55 7
37. Brooke s PS, Yoon Y, Robotham JL, Anders MW, Sheu SS. Calcium,
ATP, and ROS: a mitochondrial love - hate triangle. A m J Physiol Cell
Physiol. 2004;287:C817– C833. doi: 10.1152/ajpce ll.00139.2004
38. Indo HP, Davidson M, Yen HC, Suenaga S, Tomita K, Nishii T, Higuch i
M, Koga Y, Ozawa T, Majima HJ. Evide nce of ros gene ration by mi-
tochondria in cells with impaired electron tra nsport chain and mito-
chondr ial DNA damage. Mitochondrion. 200 7;7:106 – 118. doi: 10 .1016/j.
m it o . 20 0 6 .11 .0 2 6
39. Sabharwal SS, Schumacker PT. Mitochondrial ROS in cancer: initiators,
amplif iers or an Achilles’ heel? Nat Rev Cance r. 2014;14:709– 721. doi:
10.1038/nrc3803
40. Buza V, Rajagopalan B, Curtis AB. Ca ncer treatm ent- in duced arr hyth-
mias: focus on chemothe rapy and targ eted therapi es. Circ Arr hythm
Electrophysiol. 2 017; 10 :e 00 54 43 . d oi: 10.116 1/ CI RC EP.117. 005 44 3
41. Lampi MC, Reinhart- King CA. Targeti ng extracellular matrix stiffn ess
to attenuate d isease: from m olecular mechanis ms to clinical trials. Sci
Transl Med. 2018;10:eaao0475. doi: 10.1126/scitr anslm ed.aao0475
42. Shah SP, Nooka AK, Jaye DL, Bahlis NJ, Lonial S, Boise LH.
Bortezomib- induced heat shock response protects multiple myeloma
cells an d is activated by he at shock fac tor 1 serine 326 ph osphoryla-
tion. Oncota rget. 2016;7:59727– 59741. doi: 10.18632/ oncot arget.10847
43. Lu X Y, Xiao L, Wang L, R uden DM. Hsp90 inhibitors and drug re sis-
tance in c ancer: the potential benefits of combi nation therapies of
Hsp90 inhibitors and other anti- cancer drugs. Biochem Pharmacol.
20 12 ;8 3: 99 5– 10 04. do i: 10 .1016/j .b cp .2 011 .11.0 11
44. Pratt WB, M orishima Y, Peng HM, Os awa Y. Proposal for a role of the
Hsp90/Hsp70- based chaperone machinery in making triage decisions
when prote ins undergo oxidative and toxic damage. E xp Biol Med
(Maywood). 2010;235:278– 289. doi: 10.1258/ebm.2009.009250
45. Kamal A, Bo ehm MF, Burrows FJ. Therapeutic and diagnostic im pli-
cations of H sp90 activation. Trends Mol Med. 200 4;10:283– 290. doi:
10.1016/j.molmed.2004.04.006
46. Hoter A, El- Sabban ME, Naim HY. The HSP9 0 family: structure, regu-
lation, function, and implications in health and disease. Int J Mol Sci.
2018;19:2560. doi: 10.3390/ijms1 9092560
47. Crea gh EM, Sheehan D, Cotter TG. Heat shock proteins— mo dulators of
apoptosis in tumour cells. Leukemia. 20 00;14: 1161– 117 3. d oi : 10.1 038 /
sj.leu.2401841
48. Lanneau D, Br unet M, Frisan E, Solary E, Fon tenay M, Garrido C. Heat
shock proteins: essential proteins for apoptosis regulation. J Cell Mol
Med. 2 008 ;12 :74 3– 7 61. d oi : 1 0.1111 /j .158 2- 4 93 4.2 00 8. 00 27 3. x
49. Boothe SD, Myers JD, Pok S, S un J, Xi Y, Nieto RM, Cheng J, Jac ot
JG. The ef fect of substrate sti ffness on card iomyocyte action potentials.
Cell Biochem Biophys. 2016;74:527– 535.
50. Tallawi M, R ai R, Boccac cini AR, Aifantis KE. Effect of substrate me-
chanic s on cardiomyocyte maturation and grow th. Tissue Eng Part B
Rev. 2015;21:157– 165. doi: 10.1089/ten.teb.2014.0383
51. Schelbert EB, Fonarow GC, Bonow RO, Butler J, Gheorghiade M.
Therapeutic targets in heart failure: refocusing on the myocardial in-
terstitium. J Am Coll Cardio l. 2014;63:2188– 2198. doi: 10.1016/j.
jacc .2014.01.068
52. Be cchetti A , Petroni G, Arca ngeli A. Ion channel co nformations regulate
integrin- dependent signaling. Trends Cell Biol. 2019;29:298– 307. doi:
10 .10 16 / j . tc b . 2 01 8.1 2. 0 0 5
53. More b JS. Off- target effects of ca rfilzomib that cause c ardiotoxicity.
Blood. 2019;133:626– 628. doi: 10.1182/blood - 2018- 12- 889758
Downloaded from http://ahajournals.org by on December 7, 2021

J Am Heart Assoc. 2021;10:e022247. DOI: 10.1161/JAHA.121.022247 29
Forghani et al Carfilzomib- Induced Cardiotoxicity
54. Raturi A, Simmen T. Where th e endoplas mic reticulum and the mitochon-
drion tie the knot: the mitochondria- associated membrane (MAM). Biochim
Biophys Acta. 2013;1833:213– 224. doi: 10.1016/j.bbamcr.2012.04.013
55. Arai M, Yoguc hi A, Takizawa T, Yokoyama T, Kanda T, Kurabayashi M,
Nagai R. Mechanism of doxorubicin- induced inhibition of sarcoplasmic
reticulum Ca(2+)- ATPase gene transcription. Circ Res. 20 00;86:8 – 14.
56. Ber thiaume JM, Wallace KB. Adriamycin- induced oxidative mitochon-
drial cardiotoxicity. Cell Biol Toxicol. 2007;23:15– 25. doi: 10.1007/s1056
5- 006- 0140- y
57. Dimitrakis P, Romay- Oga ndo MI, Timolati F, Suter TM, Zuppinger C.
Effects of doxorubicin cancer therapy on autophagy and the ubiquitin-
proteasome system in long- term cultured adult rat cardiomyocytes. Cell
Tissue Res. 2012;350:361– 372. d oi: 10.1007/s0044 1- 012- 1475- 8
58. Hou T, Zhang R, Jian C, Din g W, Wang Y, Ling S, Ma Q, Hu X, Cheng H,
Wang X. NDUFAB1 confers cardio- protection by enhancing mitochondrial
bioenergetics through coordination of respiratory complex and supercom-
plex assembly. Cell Res. 2019;29:754 – 766. doi: 10.1038/s4142 2- 019- 0208 - x
59. Gu glin M, Aljaye h M, Saiyad S, Ali R, Curtis AB. Introducing a new en-
tity: chemotherapy- induced arrhythmia. Europace. 20 09 ;11:1579– 1586 .
doi: 10.1093/europ ace/eup300
60. Sauer H, Wartenberg M, Hesc heler J. Reactive oxygen s pecies as intra-
cellular messengers durin g cell growth and diffe rentiation. Cell Physiol
Biochem. 2001;11:173– 186. doi: 10.1159/00004 7804
61. Fil ice D, Dhahri W, Solan JL, Lampe PD, Steele E, Milani N, Van Biber B,
Zhu W- Z, Valdma n TS, Romagnuolo R, et al. Optic al mapping of human
embryonic stem cel l- derived cardiomyoc yte graft e lectrical activit y in
injured hearts. Stem Cell Res Ther. 2020;11:417. doi: 10.1186/s1328 7-
020 - 0 1919 - w
62. Mag adum A, Sing h N, Kurian A A, Munir I, Me hmood T, Brown K,
Sharka r MTK, Chepurko E, Sassi Y, Oh JG, et al. Pkm2 regulates car-
diomyocy te cell cycle and promotes cardiac rege neration. Circulation.
20 2 0; 141:1 24 9 – 12 65 . d o i: 10 .116 1/ CI R CU L AT I O N A H A.11 9. 043 0 67
63. Sui X, Kong N, Ye L, Han W, Zhou J, Zhang Q, He C, Pan H. p38 and
JNK MAPK pathways control th e balance of apoptosis and au tophagy
in response to chemothe rapeutic agents. Cance r Lett. 2014;344:174–
179 . do i: 10.1 01 6/ j.c an le t.20 13.11.0 19
64. Tam RC, Li MW, Gao YP, Pang YT, Yan S, Ge W, Lau CS, Chan
VS. Human CLEC16A regulates autophagy through modulating
mTOR activity. Exp Cell Res. 2017;352:304– 312. doi: 10.1016/j.
ye xc r.2 017.0 2. 017
Downloaded from http://ahajournals.org by on December 7, 2021