Mitotane Inhibits Sterol-O-Acyl Transferase 1 Triggering Lipid-Mediated Endoplasmic Reticulum Stress and Apoptosis in Adrenocortical Carcinoma Cells
Silviu Sbiera, Ellen Leich, Gerhard Liebisch, Iuliu Sbiera, Andreas Schirbel, Laura Wiemer, Silke Matysik, Carolin Eckhardt, Felix Gardill, Annemarie Gehl, Sabine Kendl, Isabel Weigand, Margarita Bala, Cristina L. Ronchi, Timo Deutschbein, Gerd Schmitz, Andreas Rosenwald, Bruno Allolio, Martin Fassnacht, and Matthias Kroiss
Department of Internal Medicine I, Endocrinology and Diabetes Unit (S.S., I.S., E.C., F.G., A.G., I.W., M.B., C.L.R., T.D., B.A., M.F.), University Hospital Würzburg, 97080 Würzburg, Germany; Comprehensive Cancer Center Mainfranken (S.S., A.R., M.F., M.K.), 97080 Würzburg, Germany; Institute of Pathology (E.L., A.R.), University of Würzburg, 97080 Würzburg, Germany; Institute of Clinical Chemistry and Laboratory Medicine (S.M., G.L., G.S.), University Hospital Regensburg, 93053 Regensburg, Germany; Department of Nuclear Medicine (A.S.), University Hospital Würzburg, 97080 Würzburg, Germany; and Clinical Chemistry and Laboratory Medicine (S.K., M.F.), University Hospital Würzburg, 97080 Würzburg, Germany
Adrenocortical carcinoma (ACC) is a rare malignancy that harbors a dismal prognosis in advanced stages. Mitotane is approved as an orphan drug for treatment of ACC and counteracts tumor growth and steroid hormone production. Despite serious adverse effects, mitotane has been clinically used for decades. Elucidation of its unknown molecular mechanism of action seems essential to develop better ACC therapies. Here, we set out to identify the molecular target of mitotane and altered downstream mechanisms by combining expression genomics and mass spectrometry technology in the NCI-H295 ACC model cell line. Pathway analyses of expression genomics data demonstrated activation of en- doplasmic reticulum (ER) stress and profound alteration of lipid-related genes caused by mitotane treatment. ER stress marker CHOP was strongly induced and the two upstream ER stress signalling events XBP1-mRNA splicing and eukaryotic initiation factor 2 A (elF2a) phosphorylation were activated by mitotane in NCI-H295 cells but to a much lesser extent in four nonsteroidogenic cell lines. Lipid mass spectrometry revealed mitotane-induced increase of free cholesterol, oxysterols, and fatty acids specifically in NCI-H295 cells as cause of ER stress. We demonstrate that mitotane is an inhibitor of sterol-O-acyl- transferase 1 (SOAT1) leading to accumulation of these toxic lipids. In ACC tissue samples we show variable SOAT1 expression correlating with the response to mitotane treatment. In conclusion, mitotane confers adrenal-specific cytotoxicity and down-regulates steroidogenesis by inhibition of SOAT1 leading to lipid- induced ER stress. Targeting of cancer-specific lipid metabolism opens new avenues for treatment of ACC and potentially other types of cancer. (Endocrinology 156: 3895-3908, 2015)
A drenocortical carcinoma (ACC) is an orphan malig- nant disease. Even after complete resection ACC har- bors a high risk of recurrence and metastatic spread. Ad- vanced ACC has a dismal prognosis with a median overall
survival of 12-15 months and limited therapeutic options (1, 2). Mitotane (1,1 dichloro-2(o-chlorophenyl)-2-(p- chloro-phenyl) ethane; o,p’-DDD), a compound with both antihormonal and antitumoral activity, is the mainstay of
ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in USA
Copyright @ 2015 by the Endocrine Society
Received April 24, 2015. Accepted August 20, 2015.
First Published Online August 25, 2015
Abbreviations: ACC, adrenocortical carcinoma; ACTB, beta-Actin; CI, confidence interval; CoA, coenzyme A; elF2a, eukaryotic initiation factor 2x; ER, endoplasmic reticulum; FCS, fetal calf serum; HR, hazard ratio; SOAT1, Sterol-O-acyl-transferase 1.
treatment both in an adjuvant setting and in metastatic disease and is the only drug approved for the therapy of ACC (3).
Its introduction to treatment of ACC dates back more than 50 years (4). In advanced disease, mitotane mono- therapy induces objective tumor response in approxi- mately 25% of patients with nonresectable ACC and is a cornerstone of most combination therapies (5). In addi- tion, a multicenter study suggested its efficacy also in an adjuvant setting (6) but adjuvant use remains controver- sial (7). There is evidence for a close correlation of mito- tane plasma levels with both efficacy and adverse effects. Plasma concentrations between 14 and 20 mg/L are aimed at (8) but adverse effects such as adrenal insufficiency, nau- sea, diarrhea, and dizziness frequently prevent achievement of these blood levels. Mitotane strongly induces hepatic cytochrome P450 enzyme 3A4 (CYP3A4) leading to in- creased metabolic clearance of glucocorticoids and mul- tiple drugs (9, 10). The many adverse effects, the narrow therapeutic window, and need for close monitoring ham- per the use of mitotane. Thus analogs with an improved benefit/risk ratio would represent a major advance, but have not been successfully developed. Detailed insights into the mechanism of action of mitotane should fuel these efforts significantly.
The mechanism of action of mitotane remains poorly understood. Earlier studies suggested metabolic activation of mitotane in adrenocortical cells and covalent binding to certain adrenal proteins, but their significance and identity remained elusive (11). Later, inhibition of steroid biosyn- thesis and cell death involving mitochondria has been re- ported with recent studies focusing on the influence of mitotane on mitochondrial respiratory chain activity (12) and mitochondrial morphofunctional changes (13). Anal- ysis of mRNA expression changes induced by mitotane at a genome-wide level by microarray (14) and a proteomics approach employing two-dimensional differential gel electrophoresis (15) could also not clarify the exact mo- lecular mode of action. In addition, the reason for the relatively specific effect of mitotane on cells of adrenocor- tical origin remains elusive. Taken together, mitotane treatment is unsatisfactory at multiple levels and improved treatment options are required but hampered by lack of insight into the molecular events triggered by this drug.
In this study we have used the ACC reference cell line NCI-H295 to perform genome-wide gene expression analyses and lipidomics mass spectrometry to identify mo- lecular mechanisms underlying mitotane action in an un- supervised fashion. Pathway information was used to guide functional assessment. We employed a panel of non- steroidogenic cell lines to elucidate the cell type specificity of molecular events triggered by mitotane and identified
the target of mitotane, SOAT1 in adrenocortical cells. We further investigated expression of SOAT1 in human ad- renocortical carcinoma tissues to analyze the clinical effect of our findings.
Materials and Methods
Cell culture
NCI-H295 and adherent variant NCI-H295R were cultured as described (16). In short, RPMI1640 medium supplemented with 10% fetal calf serum (FCS), 5.2 ng/ml sodium selenite, 100 µg/mL transferring, and 5 µg/mL insulin was used and the cells were cultured in flasks in a humid atmosphere at 37℃ and 5% CO2. Twenty percent of preconditioned medium were used for passaging. HeLa, HEK293, and IMR-32 cells were cultured in DMEM-AQ medium with 10% FCS, SW13 cells in L15 Lei- bowitz medium with 10% FCS and 1% L-glutamine and HepG2 cells in DMEM/nutrient mixture F12 Ham medium with 1% L-glutamine and 10% FCS. Short tandem repeat-profiling con- firmation was performed on all the cell lines in collaboration with Sabine Herterich from the Department of Clinical Chem- istry and Laboratory Medicine, University Hospital Würzburg. Mycoplasma contamination PCR tests have been performed reg- ularly and cells were maintained mycoplasma free.
Active substances
Mitotane was purchased from ISP Columbus, thapsigargin from Applichem and salubrinal, and Sandoz 58-035 and ator- vastatin from Sigma-Aldrich.
Apoptosis and viability tests
NCI-H295 cells (106/well)Nincubated with different concen- trations of mitotane or diluent were collected and stained with fluorescein-isothiocyanate-labeled Annexin V (BD Biosciences) for 15 minutes and for 2 minutes with propidium iodide (Sigma- Aldrich) at RT. After washing, fluorescence was measured by flow cytometry using a FACSCalibur instrument and data ana- lyzed with Cell Quest Pro (BD Biosciences). Viability testing us- ing WST1 reagent was performed according to manufacturer’s protocol (Roche).
Hormone measurements
NCI-H295 cells (106/well) were incubated with medium con- taining increasing concentrations of active substances or with the equivalent amount of solvent as negative control for several time periods. Cortisol, androstenedione, and dehydroepiandros- terone sulfate were measured in the supernatant with a Immu- lite2000 analyzer (Siemens Healthcare Diagnostics).
Gene expression profiling
RNA from 106 NCI-H295 cells treated with mitotane or sol- vent was extracted with NucleoSpin RNA II Kit (Macherey- Nagel) after homogenization using QIAshredder (QIAGEN). RNA quality was assessed using Agilent 2100 Bioanalyzer (Agi- lent) for high-quality RNA (RNA integrity number > 8.5). Ex- periments were performed in biological triplicates. Biotin-la- beled cRNA was synthesized, labeled, and fragmented using the
GeneChip 3’ IVT Express Kit (Affymetrix) and hybridized over- night at 45°℃ to a GeneChip Human Genome U133 Plus 2.0 microarray (Affymetrix). Fluorescence intensities were mea- sured and MAS5 normalized using the GeneChip 3000 7G Scan- ner and the Expression Console Software from Affymetrix.
Quantitative real-time PCR
Reverse transcription of 500 ng of total RNA was performed using the QuantiTect Reverse Transcription Kit (QIAGEN) and 40 ng cDNA each were employed for duplicate PCR on a CFX96 Real-time PCR system (Bio-Rad) with the following amplifica- tion settings: 95° for 3 minutes followed by 35 cycles of 95° 30 seconds, 60° 30 seconds, and 72° 30 seconds. Relative quanti- fication was performed using the ACT method and gene expres- sion levels normalized to beta-Actin (ACTB).
The following Taqman gene expression assays (Life Technol- ogies) were used for qRT-PCR: GDF15 (Hs00171132_m1), DUSP4 (Hs01027785_m1), TRIB3 (Hs01082394_m1), CHOP (Hs01090850_m1),ABCG1(Hs00245154_m1),SREBF1(Hs010 88691_m1), SCD (Hs01682761_m1), LDLR (Hs00181192_m1), SQLE (Hs01123768_m1), ABCA1 (Hs01059118_m1), BAX (Hs00180269_m1), GAPDH (Hs99999905_m1), and ACTB (Hs99999903_m1).
XBP1 mRNA splicing was measured with the following cus- tom-made primers and probes (Life Technologies):
· XBP1 spliced sense primer: 5’-CTGAGTCCGCAGCAG GT-3’;
· XBP1 spliced antisense primer: 5’-TGTCAGAGTCCATG GGAAGA-3’;
· XBP1 spliced FAM-labeled probe 5’-GGCCCAGTTGTCA CCTCCCC-3’;
· XBP1 unspliced sense primer: 5’-CTGAGTCCGCAGCACT CAGA-3’;
· XBP1 unspliced antisense primer: 5’-TCAGAGTCCATGG GAAGATGTTC-3’;
· XBP1 unspliced FAM-labeled probe 5’-CTATGTGCACCT CTGC-3’.
4-[125I]iodomitotane uptake assay
NCI-H295H cells (2 × 105) suspended in RPMI1640 me- dium with 10% fetal calf serum were incubated with 4 kBq of 4’-[125I]iodomitotane (see Supplemental Methods for 4’-[125I]i- odomitotane synthesis) for different periods of time at 37°℃ in glass tubes. For subcellular fractionation the Subcellular Protein fractionation Kit for Cultured Cells (Thermo) and Qproteome Mitochondria Isolation Kit (QIAGEN) were used, respectively. Uptake in each fraction was measured separately and specific proteins analyzed by immunoblotting.
Immunoblot analyses
Immunoblot analysis was performed as described (16) and the following antibodies used: Early Endosome Antigen 1 (EEA1; 1:1000; clone C45B10; Cell Signaling; http://1 degreebio. org/reagents/product/808351/?qid=762985), Mitochondrial Marker (MTC02; 1:1000; ab3298; Abcam; http://1degreebio.org/reagents/ product/1159447/?qid=762961), Sterol Regulatory Element Bind- ing Transcription factor 1 (SREBF1; 1:200; Clone H-160; Santa Cruz), CCAAT-enhancer-binding protein-Homologous Protein (CHOP; 1:1000; NBP2-13172; Novus Biologicals), Eukaryotic Translation Initiation Factor 2A eIF2@ (1:1000; D7D3; Cell Signaling;
http:/1degreebio.org/reagents/product/862810/?qid=762970), phospho-eIF2a (Ser51) (1:500; Clone D9G8; Cell Signaling; http://1degreebio.org/reagents/product/808408/?qid=762976), beta actin (ACTB; 1:1000; Clone D6A8; Cell Signaling; http:// 1degreebio.org/reagents/product//1420068/?qid=762978), Sterol O-Acyltransferase 1 (SOAT1; 1:1000; ab39327; Abcam; http://1degreebio.org/reagents/product/338156/?qid=762981) and SOAT 2 (1:1000; 100027; Cayman Chemical). Horse rad- dish peroxidase-conjugated secondary antibodies from GE Healthcare were used (1:5000). Signal was developed with West- ernSure Premium chemiluminescence substrate (Li-Cor) and documented with a C-Digit Instrument (Li-Cor). For probing immunoblots with SOAT1- and SOAT2-specific antibodies, the protein extracts were reduced and alkylated by treatment with 50mM NuPage Reducing Agent containing dithiothreitol (Life Technologies) and 107mM iodoacetamide (Sigma).
Gas chromatography-mass spectrometry and electrospray ionization tandem mass spectrometry analysis of lipids
Gas chromatography-mass spectrometry analysis of lipids was performed as described for oxysterols, cholesterol precur- sors (17), and total fatty acids (18). Free cholesterol (FC) and cholesteryl esters (CEs) were quantified by direct-flow injection electrospray ionization tandem mass spectrometry as described (19). Total protein was quantified with BCA assay and absolute lipid quantities expressed as nmol/mg protein.
SOAT1 assay
SOAT1 activity was measured as described (20) at Eurofins Cerep Panlabs (assay ID, 105900). Wistar rat liver was homog- enized in preparation buffer (5mM K2HPO4, pH 7.4, 0.25M Sucrose, 1mM EDTA, 1 mM dithiothreitol, 7mM KCl) using a motor-driven teflon pestle homogenizer. The homogenate was centrifuged at 17 000 × g for 15 minutes and the supernatant saved and centrifuged at 145 000 × g for 60 minutes. The pellet containing microsomes was resuspended in preparation buffer and centrifuged at 145 000 × g for 60 minutes, the supernatant discarded, and the pellet resuspended in preparation buffer. To measure SOAT1 activity, mitotane and ethanol control, respec- tively were preincubated with 0.16 mg/mL Wistar rat liver mi- crosomes in 200mM KH2PO4, pH 7.4 for 15 minutes at 37°℃ in a total volume of 500uL. Esterification was initiated by ad- dition of 18uM [14C]Palmitoyl-coenzyme A (CoA) (60 µCi/ umol, PerkinElmer NEN) for 10 minutes. Esterification was stopped using ethanol and lipid extraction performed with pe- troleum ether. An aliquot of 1 mL of lipid phase was applied to a SepPak silica gel column (Waters) and eluted with 3 mL pe- troleum ether:diethyl ether (98:2v/v). Activity of [14C]choles- terol ester was quantified using scintillation counting (PerkinEl- mer 1450 Microbeta). Lovastatin was used as positive control.
22-(N-(7-Nitrobenz-2-oxa-1,3-diazol-4-yl)amino)- cholesterol incorporation in lipid droplets
NCI-H295 (4 × 105) cells were seeded on a chambered slide (Nunc) and grown for 24 hours as described. HEK293 cells (1 × 106) were grown on chamber slides and transiently transfected using Lipofectamine 2000 reagent (Life Technologies) according to the manufaturer’s instructions using plasmids encoding SOAT1 (Origene, RC205774), SOAT2 (Origene, RC221499),
or empty vector, respectively. 22-NBD-cholesterol (1 µg/mL) (Life Technologies) and mitotane or Sandoz 58-035 (or diluent as control), was added for 4 hours, cells were washed in PBS and immediately imaged using a ZEISS Axiovert 135 microscope equipped with standard filter sets and AxioCam MR digital cam- era to detect lipid droplet formation (21).
Immunohistochemistry
Immunohistochemistry was performed as described (22) us- ing antibodies against SOAT1 (1:1000; ab39327; Abcam) on tissue specimens from four normal adrenals, 16 adrenocortical adenomas (six cortisol-producing adenomas, four endocrine in- active adenomas, and six aldosterone-producing adenomas) and 116 adrenocortical carcinomas (89 primary, 16 local recur- rences, and 11 metastases) with four prostate carcinomas, five colorectal carcinomas, five pancreatic carcinomas, and four ovarian carcinomas as extra-adrenal controls. Staining was eval- uated by two independent experienced observers (S.S. and S.I.) and the interobserver agreement coefficient was very high (Spearman r = 0.95; 95% confidence interval [CI], 0.93-0.96). Semiquantitative evaluation was performed using an H-score as described (22). Clinical data of adrenal tumors were retrieved from the German ACC registry and European Network for the Study of Adrenal Tumors registry. All patients gave informed consent, and the study was approved by the ethics committee of the University of Würzburg (Approval No. 88/2011). Thirty- seven tissue samples were collected within 3 months prior to mitotane therapy. From these patients, 12 received mitotane therapy in an adjuvant setting and 25 in a palliative setting. Response to therapy was defined as stable disease or better after at least 6 months for palliative and absence of tumor for at least 12 months in adjuvant treatment. The adjuvant-treated patient group was heterogenous regarding resection status, a factor of superordinate importance in tumor recurrency. Therefore, only patients treated with mitotane in a palliative setting were eval- uated for response to mitotane defined as progression-free survival.
Statistical analyses
All experiments were performed in biological triplicates, ex- perimental and control cells were taken from the same batch, and treatment has been performed on equal amount of cells from every batch. Correlation between expression data resulted from microarray and real-time assays and protein-staining interob- server agreement were performed using the Spearman r corre- lation test. Statistical significance was assessed by two-way ANOVA followed by Bonferroni post test. Kolmogorov-Smir- nov and d’Agostino and Pearson omnibus normality tests were performed where possible to test whether values have the same distribution in every group. Progression-free survival during mi- totane treatment was evaluated using Kaplan-Maier followed by Mantel-Cox log-rank analysis. All statistical analyses were per- formed using GraphPad Prism 6.0 for Windows and Mac. All error bars show SD. Box plots for semiquantitative immunohis- tochemistry data indicate mean values and quartiles and the whiskers represent the minimal and maximal values.
Results
Cell type-specific mitotane effects
By comparing cell viability of mitotane-treated NCI- H295 ACC cells and nonsteroidogenic HeLa (cervix), HepG2 (liver), HEK293 (embryonic kidney) and IMR-32 (neuroblastoma) cell lines using WST1 testing, we found an EC50 of 18.1uM (5.8 mg/L, NCI-H295), 56.0MM (17.9 mg/L, HeLa), and >100uM (>32.0 mg/L, HepG2, HEK293, IMR-32; Figure 1A) at 24 hours. Annexin V- propidium iodide fluorescence activated cell sorting anal- ysis of NCI-H295 cells demonstrated mitotane-induced apoptosis with an EC50 of 23.8uM (7.6 mg/l, Figure 1B) at 72 hours. Mitotane inhibited cortisol secretion into cul- ture supernatant at a similar EC50 of 19.1uM (6.1 mg/L; Figure 1C). Given that these terminal effects of mitotane treatment were apparent within the first 24 hours, we pos- tulated that the underlying molecular events should hap- pen early after exposure to mitotane. Accordingly, pro- apoptotic Bcl-2-associated X protein (BAX) mRNA was 2-fold overexpressed already after 6 hours mito- tane treatment (Figure 1D). We therefore chose this time interval to perform gene expression analysis, lipid mass spectrometry was additionally performed after 0.5 and 2 hours’ mitotane treatment (Figure 1E).
Mitotane induces ER stress leading to impaired steroidogenesis and apoptosis
By applying a 1.5-fold change cutoff, treatment with 50uM (16 mg/L) and 100µM (32 mg/L) mitotane led to gene expression changes in 567 and 1976 genes, respec- tively at microarray analysis (Supplemental Table 1). In- vestigation of gene ontology pathways using MetaCore showed that among the 30 most down-regulated genes after mitotane exposure, eight were implicated in lipid metabolism and steroidogenesis such as squalene epoxi- dase (SQLE), low-density lipoprotein receptor (LDLR), stearoyl-CoA desaturase (SCD), sterol regulatory element binding transcription factor 1 (SREBF1), and ATP-bind- ing cassette subfamily G (ABCG1) (Supplemental Figure 1). Of the 30 most overexpressed genes, six were related to apoptosis-like growth differentiation factor 15 (GDF15) and DNA-damage-inducible transcripts 1, 3, and 4 (DDIT1/DDIT3(CHOP)/DDIT4). Unsupervised analy- ses revealed that the ER stress response was a highly ac- tivated pathway with 22 of 53 genes altered after mitotane exposure (Supplemental Figure 2). Validation of ten se- lected, strongly regulated genes using qRT-PCR con- firmed a close correlation with microarray data (Spear-
A
B
A
I
4
100-
uM mitotane
—
-
fold change apoptosis
0
25
cell viability (%)
80-
50
3
75
100
60-
NCI-H295
SW13
40-
₴ HeLa
2
**
A HepG2
**
20-
7 HEK 293
IMR32
1
T
0
6
24
48
72
0
1
5
10
25
50
100
Time / (hours)
C
mitotane concentration (uM)
fold change cortisol secretion
2.5
uM mitotane
fold change BAX expression
8
uM mitotane
0
0
25
100
50
2.0
75
6
100
**
4
1.5
2
**
1.0
0
12
24
36
48
60
72
0
4
6
8
12
24
Time / (hours)
Time / (hours)
E Expression Genomics
Lipidomics
ΜΙΤΟ 50 μΜ ΜΙΤΟ 100 μΜ
control
MIΤΟ 50 μΜ ΜΙΤΟ 100 μΜ
control
0.5h
2h
6h
6h
total RNA
rev. transcr.
total lipids
microarray hybridization
derivatization
read-out + transformation
LC-MS GC-MS
pathway analysis
lipid identification
man r = 0.84; Figure 2A). Time course experiments showed that some genes, in particular those involved in lipid metabolism, changed in a biphasic manner with an early up-regulation followed by a steep decrease and late
recovery (Figure 2B). Immunoblot- ting confirmed the time-dependent increase of ER stress activation marker DDIT3/CHOP at protein level and a decrease of active 68 kDa SREBF1 (Figure 2C). We also detected changes in upstream signaling pathways of the ER stress response. As a marker of pro- tein kinase R-like kinase (PERK)- dependent ER stress, we found strong phosphorylation of eukaryotic initia- tion factor eIF2@ at serine 51 with total eIF2a protein unaltered (Figure 2C). The same protein changes were induced by treatment with ER stress inducer thapsigargin with lesser effect on SREBF1 protein expression compared with mitotane. As an indication of inos- itol-requiring kinase 1 (IRE1)-depen- dent ER stress, we demonstrated a dose- dependent strong increase of X-box binding protein 1 (XBP1) mRNA splic- ing in mitotane-treated NCI-H295 cells (Figure 2D; 100µM: 23.27-fold; P < .001).
Mitotane-induced ER stress is partially reversible with salubrinal and enhanced by thapsigargin
We next coincubated NCI-H295 cells with mitotane and the specific ER stress inhibitor salubrinal and found a significant decrease of mi- totane-induced CHOP mRNA ex- pression (Figure 3A) that was re- flected also in a decrease of proapoptotic BAX mRNA expres- sion (Figure 3B). Salubrinal treat- ment alone did not significantly ef- fect CHOP and BAX expression. Conversely, treatment with the ER stress inducer thapsigargin to- gether with mitotane increased CHOP and BAX expression (Fig- ure 3C). Thapsigargin alone not only increased CHOP, albeit to a lesser extent than mitotane, but also decreased cortisol secretion by almost 50% (18.7 ± 2.9 µg/dL vs 33.8 ± 5.5 µg/dL, filled circle, Figure 3D). In contrast, salubrinal had no significant effect on cortisol production either by itself (Figure 3E, filled circles) or in combination with mito- tane (Figure 3E, filled squares).
A
10-
r=0.84
B
10000
UM mitotane
1000
uM mitotane
fold change expression
o + 50
50
microarray (fold change)
fold change expression
1.
1000
€ 100
+ 100
0
·2
100
5-
100
**
0
0
10 ·· 6
10
10
GDF15
0
2
2
CHOP
1
Z
증
3
1
5
I
0
0
0
6
12
18
24
0
6
12
18
24
0.25 0.5
1
2
4
8
16
32
64
Time / (hours)
Time / (hours)
7.
5
3
&
4
100 - HM mitotane
0 0
100-
uM mitotane
fold change expression
00
+
50
fold change expression
+ 50
-5.
10
৳ 100
SREBF1
100
10
**
ABCG1
0
$
·9
1
I
1
5
T
5
#
-10-
0.1
B
0.1
Real-time PCR (normalized expression)
-
0.01
0
6
12
18
24
0.01
0
6
12
18
24
Time / (hours)
Time / (hours)
C
D
50μM mitotane
50uM thapsigargin
0
0.3
1
6
time [h]
0
0.3
1
6
time [h]
CHOP
CHOP
SREBF1
SREBF1
EIF2A
EIF2A
P-EIF2A
P-EIF2A
ACTB
ACTB
E
30-
ratio XBP1 spliced/unspliced
4
1
20-
10-
0
0
50
100
mitotane concentration (uM)
Accumulation of toxic lipids is a key feature of mitotane action and characteristic for ACC cells
Increasing evidence suggests a predominant role of lipids in activating ER stress in various cellular contexts (23, 24). We therefore used mass spectrometry to directly assess lipids after mitotane treatment of NCI-H295 and,
for comparison, hepatic HepG2 cells. Strikingly, a characteristic and prominent pattern of lipid altera- tions developed in a time- and dose- dependent manner in NCI-H295 cells but not in HepG2 cells (Figure 4A). NCI-H295 cells exhibited a 1.3-2.5 fold increase of free choles- terol after 2-6 hours (P < . 05; Sup- plemental Table 2). In addition, we also found a dose-dependent in- crease at 6 hours of nonenzymatically generated 7ß-hydroxycholesterol and 7-ketocholesterol (2.3- and 2.8-fold, respectively), whereas enzymatically formed oxysterols such as 24- and 25- hydroxycholesterol were undetect- able. Similarly, all saturated and monounsaturated fatty acids mea- sured were strongly increased after treatment with mitotane in NCI- H295 but not HepG2 cells (Figure 4A). The same effect was observed for plasmalogens (Supplemental Table 2). In turn, we found saturated and un- saturated cholesterol C16 and C18 cholesteryl esters to be decreased at early time points with a later increase observed at 100M. Of note, also the cholesterol precursors lathosterol, lanosterol, and 7-dehydrocholesterol were increased strongly at 6 hours with a 4.3-, 7.8-, and 5.4-fold change, respectively after treatment with 100µM mitotane. In contrast, in HepG2 cells none of the sterols exam- ined was significantly altered (Figure 4A and Supplemental Table 2).
We extended these lipidomics exper- iments to the panel of nonsteroidogenic cell lines employed previously (Figure 4B) that were exposed to 100uM mito- tane for 6 hours. Similar to HepG2 cells, FC, oxysterols, fatty acids, and choles- terol precursors did not exhibit any sig- nificant change in SW13, Hek293, HeLa, and IMR-32. In SW13 cells a strong decrease of CEs was observed whereas in HEK293 cells we found a mitotane induced an increase of CEs. Fatty acid composition of all these cell lines showed no substantial changes upon mitotane treatment (Supplemental Table 2). We next ex- amined whether differences in lipid profiles among cell lines
A
B
fold change CHOP expression
*
*
25-
1
fold change BAX expression
2.0-
1
20-
1.5-
15-
1.0-
10-
5-
0.5-
0
0.0
50nM salubrinal
-
+
-
+
50nM salubrinal
-
+
-
+
100uM mitotane
-
-
+
+
100uM mitotane
-
-
+
+
C
D
fold change CHOP expression
80
20
5uM thapsigargin+100uM mitotane
5uM thapsigargin+100uM mitotane
5uM thapsigargin
fold change cortisol
5uM thapsigargin
100μM mitotane
100μM mitotane
control
60
15
control
40
10
*
20
5
0
12
18
?
0
0
6
24
0
6
Time / (hours)
Time / (hours)
E
8
50nM salubrinal+100uM mitotane
100μM mitotane
fold change cortisol
50nM salubrinal
₹
6
control
4
2
0
0
6
12
18
24
Time / (hours)
are reflected by differences in their propensity to undergo ER stress using qRT-PCR of marker gene transcripts. Indeed, after 100uM mitotane treatment, fold change increase of CHOP and XBP1 spliced/unspliced ratio was increased 23.9- and 23.3-fold (P <. 001), respectively in NCI-H295 cells, but much less in SW13 (6.4/6.0, HepG2: 2.6/1.4-fold), HEK293 (2.5/3.7), HeLa (3.5/6.4), and IMR-32 cells (1.5/ 2.1) (Figure 4C).
To further clarify the role of FC in mediating ER stress, we combined mi- totane treatment with 3-hydroxy-3- methyl-glutaryl-CoA reductase inhib- itor atorvastatin. For the cells treated with 25 µM mitotane, cell viability in- creased with increasing concentra- tions of atorvastatin (37.4 ± 4.6 at OnM, 70.5±8.2at0.1nM, 85.6±7.8 at 1nM, and 83.3 ± 8.5nM). How- ever, atorvastatin in this concentration range failed to decrease the cytotoxic effects of 50uM mitotane treatment (Supplemental Figure 3).
SOAT inhibitor Sandoz 58-035 mimics mitotane effects in NCI- H295 cells
* Early increase of FC and de- crease of CEs measured by mass
spectrometry in NCI-H295 cells
exposed to mitotane led us to com- 12 18 24 pare these changes to those induced by the SOAT-inhibitor Sandoz 58- 035. We found significant decrease of CEs at both 50uM and 100uM (Supplemental Figure 4A) and a concomitant increase of FC at 100uM. Notably, also cholesterol precursors increased. Similar to mito- tane, SOAT inhibition by Sandoz 58- 035 significantly induced CHOP mRNA expression (2.6 ± 0.1- and 2.6 ± 0.2-fold change for 50 and 100µM, respectively) and activated XBP1 mRNA splicing (2.0 ± 0.4- and 1.8 ± 0.1-fold change for 50 and 100uM, respectively) in NCI-H295 cells (Supplemental Figure 4B). San- doz 58-035 also reduced cell via- bility (33.9 ± 3.2% decrease in vi- ability at 100uM, Supplemental Figure 4C) and led to a decrease of cortisol synthesis after 72 hours by almost half compared with control (from 6.4 + 0.4-fold in untreated cells to 3.7 + 0.2-fold at 100µM; Supple- mental Figure 4D).
Mitotane inhibits SOAT1 and SOAT 1 expression correlates with in vitro mitotane sensitivity
Immunoblotting of the two cholesterol acyl trans- ferases sterol-O-acyl transferase 1 and 2 in all cell lines
A
0.5
2
6
0.5
2
6
B
6
Treatment time (hours)
50
100
50
100
0
50
100
50
100
50
100
50
100
100
100
100
100
uM mitotane
free cholesterol
7-B-hydroxycholesterol
7-ketocholesterol
16:0 CE
16:1 CE
18:0 CE
18:1 CE
lathosterol
desmosterol
7-dehydrocholesterol
C12:0 FA
C14:0 FA
C15:0 FA
C16:0 FA
C16:1-C9 FA
C17:0 FA
C18:0 FA
C18:1-C9 FA
C18:1-C11 FA
C20:0 FA
C16 plasmalogens
C18 plasmalogens
NCI-H295R
HepG2
SW-13
HEK-293
HeLa
IMR32
Cell Lines
log2 fold-change
C
30-
NCI-H295
HepG2
SW13
HEK293
HeLa cervix
IMR32
adrenal
kidney
relative expression
neuro- blastoma
20-
☐ CHOP
☐ XBP1
spliced/unspliced
10-
T
T
**
**
a
0
M
0
50
100
0
50
100
O
50
100
0
50
100
0
100
0
O
100
0
100
O
100
0
100
0
100
0
100
uM mitotane
showed very prominent expression of SOAT1 in NCI- H295 cells but minimal expression at protein level in all other cell lines tested (Figure 5A). Inversely, SOAT2 was expressed at very low levels in NCI-H295 cells (Figure 5B).
We next used a microsomal SOAT1 assay to directly test the influence of mitotane on SOAT1 activity and found mitotane to inhibit SOAT1 activity with an IC50 of 21µM (7.0 mg/L). This corresponds to half of the con-
A
B
kDa
1
2
3
4
5
6
kDa
7
8
9
10
11
12
55
SOAT1
55
SOAT2
45
ß-Actin
45
ß-Actin
NCI-H295
HepG2
HeLa
HEK293
IMR32
SW13
NCI-H295
HepG2
HeLa
HEK293
IMR32
SW13
C
1.0-
D
40-
SOAT activity
.
% uptake
30-
0.5-
20-
10-
0
0.0
C50
21µM
MTC02
1
10
100
EEA1
mitotane concentration (UM)
Cytosol
Endo- membranes
mitochondria
E
F
1
G
H
2
1
A
2
3
J
K
L
M
N
O
P
Q
1.0
SOAT1 low (n=5)
0.8-
SOAT1 high (n=20)
0.6
0.4.
p=0.001
0.2-
0.0
0
24
48
72
time (months)
centration found to be clinically ef- fective in patients with ACC (Figure 5C). Given that SOAT1 is located within the ER membrane, we inves- tigated the subcellular distribution of mitotane using 4-[125I]iodomito- tane. Unsurprisingly, due to the lipo- philic nature of mitotane, the tracer was enriched in both endomembranes and mitochondrial membranes (Fig- ure 5D).
In addition, we observed strongly reduced incorporation of fluores- cently labeled 22-NBD-cholesterol into CE-rich lipid droplets in NCI- H295 and HEK293 cells overex- pressing SOAT1 treated with mito- tane similar to the SOAT inhibitor Sandoz 58-035 (Supplemental Fig- ure 5). SOAT2 transfection into HEK293 cells led to marginal forma- tion of lipid droplets compared with SOAT1.
SOAT1 is differentially expressed in different tumor entities and associates with response to mitotane in ACC
Using FFPE tissue samples, we in- vestigated SOAT1 protein expres- sion in the healthy adrenal gland, adrenocortical neoplasms, and ex- tra-adrenal tumors. Strong expres- sion was consistently detected throughout the normal adrenal cor- tex (H-score mean = 2.6 ± 0.5; n = 4; Figure 5, E and F) and adrenocor- tical adenomas (2.7 ± 0.5; n = 16; Figure 5, G and I) but expression was more variable in adrenocortical car- cinoma (2.0 ± 0.9; n = 116; Figure 5, J and M). Of note, SOAT1 showed a trend toward lower expression in metastases compared with primary tumors (1.7 ± 1.2; n = 11 vs 2.0 ± 1.0; n = 89; Supplemental Figure 6A). Among extra-adrenal tumors, prostate cancer samples showed re- markably high SOAT1 expression (2.8 ± 0.5; n = 4) but low SOAT1 was seen in the other tumors exam- ined (colon: 1.0 ± 0.7, n = 5; pan-
creatic: 0.8 ± 0.8, n = 5; and ovarian: 0.25 ± 0.5, n = 4; Figure 5, N-P; Supplemental Figure 6A).
We next examined whether SOAT1 expression in ad- renocortical carcinomas translates into response to mito- tane treatment. To this end, we analyzed the subset of tumors surgically removed no more than 2 months before initiation of palliative mitotane treatment (n = 25). Of note, time to progression in tumors with low SOAT1 ex- pression (H-score <2) was significantly shorter compared with those with high SOAT1 expression (H-score ≥ 2; median time to progression 3 mo vs 9 mo; hazard ratio [HR], 28; 95% CI, 4.3-186.0; P = . 001; Figure 5Q). Sim- ilarly, we found significantly higher SOAT1 expression in responders compared with nonresponders when patients with adjuvant mitotane treatment were included (n = 37; Supplemental Figure 6B). Considering the hormonal sta- tus of the tumor, SOAT1 was variably expressed in both nonsecreting and hormonally active tumors, however, in average significantly higher in secreting tumors (2.3 ± 0.8, n = 45 to 1.7 ± 0.9, n = 16; P = . 02; Supplemental Figure 6C). In contrast with mitotane-treated patients, high SOAT1 expression had no significant effect on the overall survival of all patients with ACC analyzed (n = 87; median overall survival, 86 vs 45 mo for low vs high SOAT1; HR, 0.79; 95% CI, 0.44-1.42; P = . 3; Sup- plemental Figure 3C).
Discussion
Toxic effects of mitotane on the adrenal cortex were first observed in dogs and subsequently exploited for treatment of adrenocortical carcinoma in animals and humans (25). Mitotane became the cornerstone of ACC treatment both in an adjuvant setting and in metastatic disease. Despite steady research including genome and proteome-wide ex- pression studies (12-14), the molecular mechanism by which mitotane acts on both steroidogenesis and cell sur- vival has remained elusive.
In this study we used pathway analyses of genome-wide gene expression data to decipher the effect of mitotane on the ACC model cell line NCI-H295 at the molecular level. Endoplasmic reticulum stress was discovered as a key mo- lecular pathway activated by mitotane. Using lipidomics analyses we demonstrated profound alterations of choles- terol metabolism characteristic for adrenocortical carci- noma cells to underlie ER stress leading to discovery of SOAT1 as the key molecular target of mitotane.
Gene expression data revealed profound disturbances in cell signaling after mitotane treatment with the ER stress response pathway and the “SCAP/SREBF transcriptional control of cholesterol and fatty acids biosynthesis” being
the two most strongly affected pathways. ER stress is con- sidered a major adaptive response of cells to unmet needs of biosynthesis and excess macromolecules such as un- folded proteins (26). In the event of sustained ER stress exceeding the cell’s adaptive mechanisms, ER stress ulti- mately leads to apoptosis. Confirmatory time course experiments quantitatively analyzing gene expression showed ER stress sensitive, proapoptotic genes GDF15, DUSP4, TRIB3, and CHOP to undergo a continuous in- crease in expression over time indicating early and steady activation of apoptosis pathways during mitotane treat- ment, as confirmed by Annexin V/propidium iodide flow cytometry analyses and increased BAX gene expression. There are three key proteins transducing ER stress up- stream of CHOP: PERK, IRE1, and activating transcrip- tion factor 6 (27). Using eIF2a phosphorylation as a read- out, we demonstrated strong mitotane-induced PERK pathway activation while robust increase in XBP1 mRNA splicing provided evidence for IRE1-pathway activation by mitotane treatment. Triggering ER stress through two pathways may explain the unexpectedly broad extent of transcriptional effects we observed after mitotane. Ac- cordingly, inhibition of ER stress by salubrinal attenuated CHOP and BAX expression in response to mitotane. In our experimental setting salubrinal seemed to have no sig- nificant effect on steroidogenesis, but this could be also due to the low concentration used or the short treatment time. In turn, ER stress experimentally triggered by thap- sigargin alone was sufficient to induce apoptosis and, no- tably, to repress steroidogenesis. The same pattern of pro- tein changes as observed with mitotane in the NCI-H295 cells appeared after thapsigargin treatment. Whereas ER stress is known to initiate apoptosis, there are only very few instances where an inhibitory effect on steroid hor- mone synthesis has been demonstrated (28). qRT-PCR revealed a biphasic behavior of genes involved in lipid metabolism with an early increase of expression and strong decrease at later time points with a trend toward baseline expression after prolonged exposure to mitotane. These changes may reflect early moderate activation of fatty acid synthesis via SREBF1 after mitotane treatment as a compensatory mechanism of increased FC, with sup- pression and cell death at later time points.
Lipid mass spectrometry confirmed profound altera- tions of lipid metabolism after mitotane treatment in ad- renocortical cells. Although in clinical practice mitotane consistently induces hepatic enzymes and liver toxicity, the pattern of profound changes in lipid homeostasis ob- served in NCI-H295 was virtually absent in the hepatic HepG2 cells and also four other nonsteroidogenic cell lines, including the adrenocortical cell line, SW13. In- crease of FC and fatty acids in NCI-H295 cells but early
decrease of CEs pointed to a specific metabolic event in- terfering with lipid metabolism. Of note, we found con- comitantly increased oxysterols 7-dehydrocholesterol and 7-ketocholesterol, indicative of oxidative stress. It is well known that oxysterols are key signaling molecules that trigger transcriptional down-regulation of de novo cho- lesterol biosynthesis (29). Increased cholesterol precursors provide direct evidence for a mitotane-induced block in cholesterol synthesis. Blocked cholesterol synthesis and depletion of cholesteryl esters will ultimately result in the observed impairment of steroidogenesis, which is a hall- mark of mitotane treatment. Furthermore, the importance of cholesterol for the efficacy of mitotane was furthermore proven by the counteracting effect of atorvastatin on mi- totane-induced cell toxicity.
Given that all these changes at the lipid level were ab- sent in other cell lines tested, mitotane effects seem to de- pend on different activities of lipid-metabolizing enzymes according to the respective cellular context. The strong decrease of CE, albeit with unchanged FC in the SW13 cells, points toward SOAT1 inhibition, which is expressed at higher levels in this cell line compared with the other nonsteroidogenic cells. We cannot at this point completely clarify the observed concomitant CE and FC increase in HEK293 cells. There is probably an off-target effect of mitotane on a yet-unidentified protein specific in this cell line as HeLa and IMR-32 cells, do not show any significant lipid pattern change. As downstream events, strong in- crease of CHOP expression and XBP1-mRNA splicing in NCI-H295 cells but much less in all other cell lines tested provided strong support for a model in which lipotoxicity is causative of mitotane-induced ER stress specifically in NCI-H295 cells. Lipotoxic ER stress has previously been demonstrated in other pathophysiological con- texts (30-32). ER stress has been shown to be a key pathway mediating stress-responsive regulation of mi- tochondrial function (33) with mitochondria-associ- ated ER membranes constituting the interface between these organelles (34). SOAT enzymes (Enzyme Com- mission 2.3.1.26) are located within these mitochon- dria-associated ER membranes (35) and catalyze cho- lesterol esterification (36, 37). This SOAT subcellular location complies with our observation that radioactive 4-[125I]iodomitotane can be experimentally retrieved from endomembranes and mitochondria. We therefore further investigated SOAT as putative molecular target of mitotane in NCI-H295 cells.
By directly measuring SOAT activity, we discovered that mitotane inhibits this enzyme with an IC50 very sim- ilar to the EC50 observed for cortisol secretion, apoptosis, and comparable with clinically effective plasma concen- trations. Inhibition of SOAT1 most likely is independent
of gene expression changes given that gene expression mi- croarray did not reveal significant changes of SOAT1 and SOAT2 expression. Furthermore, by treating the NCI- H295 cells with the SOAT inhibitor Sandoz 58-035 we observed the same effects on cell viability and changes in lipid profile, key ER stress gene expression, and cortisol secretion as with mitotane. Although these changes were more attenuated, probably due to efficacy restrictions, it further provides evidence that mitotane is a potent SOAT inhibitor. Of note, SOAT inhibitors developed for treat- ment of atherosclerosis frequently failed during clinical development because of adrenal toxicity (38, 39), suggest- ing that adrenocortical cells are extraordinarily sensitive to SOAT inhibition.
High SOAT1 expression only in the mitotane suscep- tible NCI-H295 ACC cells and normal adrenal tissue sam- ples but low SOAT2 expression confirmed previous data demonstrating SOAT1 as the relevant isoenzyme catalyz- ing cholesterol esterification in the active adrenal (40). By visualizing lipid droplets using a fluorescent cholesterol analog we were able to show that NCI-H295 cells highly expressing SOAT1 are rich in CE-containing lipid drop- lets. Upon treatment with either mitotane or Sandoz 58- 035 the number of lipid droplets decreased strongly, cor- roborating decrease of CE observed by mass spectrometry.
ABCG1
CM
FA
CHOL
CHOL
CHOL
ER
+
FA
CHOL
CHOL
APOPTOSIS
FA
MITO
SOAT1
ER-STRESS
CE
SREBF
IRE1 P
PERK
-
STEROIDOGENESIS
elF2a
XBP1_
XBP13
P
CHOP
Bax
SREBF
XBP1
Bcl-2
N
sterol responsive genes
UPR responsive genes
M
In line with this, wild-type HEK293 cells expressing low levels of SOAT1 did not show lipid droplets. After trans- fection with SOAT1, lipid droplets were observed in vir- tually all cells. At variance, SOAT2 transfection caused lipid droplet formation in only a small percentage of cells confirming the prevalent role of SOAT1 in cholesterol es- terification. Indeed, the adrenal is among the organs with highest SOAT1 activity and cholesterol esters are consid- ered to constitute a storage pool required for steroidogen- esis (36, 41).
We can reconcile our findings in a model (Figure 6) in which mitotane induces ER stress by increasing FC and fatty acids through inhibition of SOAT1 and indi- rectly through lipid-induced down-regulation of SREBF and SREBF-dependent genes (42). Disturbed lipid ho- meostasis perpetuates and exacerbates ER stress, acti- vating lipid-signaling pathways, leading to blocked cholesterol synthesis and steroidogenesis. With sus- tained ER stress, cells will undergo apoptosis by CHOP- dependent factors. This mechanism of action seems to be adrenocortical cell-specific as the nonsteroidogenic cells are not or much less affected at the same concen- trations of mitotane.
Expression of SOAT1 at high levels in benign adrenal tumors but the more variable expression in ACC high- lights the relevance of SOAT1 as a marker of intact sterol metabolism and hence, absent or low expression in approximately one third of ACC seems to reflect a less-well-differentiated phenotype without necessarily correlating with other aspects of tumor biology. Thus, we did not find a correlation of SOAT1 expression with overall survival. The observation that most other tu- mors investigated exhibit lower SOAT1 expression un- derscores the notion of a relative tissue-specific SOAT1 expression. As a remarkable exception, prostate cancer, which has previously been shown to exhibit strongly activated sterol metabolism (43), showed high SOAT1 expression.
Importantly, time to progression during palliative mitotane differed sharply between tumors with high and low SOAT1 expression. We consider this strong support for the concept that SOAT1 expression is a prerequisite for mitotane efficacy. It is, however, also not surprising that mitotane treatment failed in some patients despite high tumor SOAT1 expression. The pharmacokinetic profile of mitotane is unfavorable and timely effective mitotane plasma concentrations (44, 45) are missed in a significant proportion of patients (46). A shortcoming of our study is limited information on maximal mitotane plasma levels in this cohort of patients. However, our clinical data support the view that SOAT1 is necessary but not sufficient for clinical
response to mitotane treatment. Given that we found SOAT1 expression to be significantly higher in hormon- ally active tumors and given the high susceptibility of NCI-H295 cells to mitotane it is tempting to speculate that mitotane would have a better effect on hormonally active ACC. However, SOAT1 expression alone had a significant effect on mitotane response. With respect to the observed high expression of SOAT1 in prostate can- cer it is possible that SOAT inhibitors may favorably influence the clinical course of this malignancy as well.
Based on these results, future pharmacological ap- proaches for ACC may 1) exploit SOAT1 inhibition with more specific compounds, or 2) combine mitotane with inducers of ER stress acting by complementary modes of action. This could reduce mitotane doses, decrease adverse effects, and improve quality of life. Lipid-induced ER stress may become a novel approach for the treatment of certain types of cancer by exploiting tissue-specific lipid metabolism and SOAT1 enzymes may be attractive novel drug targets.
Acknowledgments
We thank Sabine Herterich for the STR cell-line identification, careful reading of the manuscript, and helpful comments.
Address all correspondence and requests for reprints to: Matthias Kroiss, MD, PHD, Comprehensive Cancer Center Mainfranken, University of Würzburg, Oberdürrbacher Straße 6, 97080 Würzburg, Germany. E-mail: Kroiss_M@ukw.de.
Authors contributions: Conception and design: S.S., M.F., B.A., and M.K .; Development, methodology, and acquisition: S.S., E.L., G.L., I.S., L.W.,S.M.,F.G.,A.G.,C.E.,S.K.,I.W., and M.K .; New reagents/analytic tools: A.S .; Analysis and interpre- tation of data: S.S., M.F., E.L., M.B., C.L.R., T.D., G.S., A.R., B.A., and M.K; Writing of the manuscript: S.S., M.F., B.A., and M.K.
This study was supported by the following grants: Deutsche Forschungsgemeinschaft (Grant FA 466/4-1 to M.F. and KR 4371/1-1 to M.K.), a fellowship of the Comprehensive Cancer Center Mainfranken to M.K., IZKF Würzburg (Grant B-281 to M.F.), and the ERA-NET “E-Rare” (Grant 01GM1407B to M.F.).
Microarray data are available in the ArrayExpress database (www.ebi.ac.uk/arrayexpress) under accession number E-MTAB- 3241.
Disclosure Summary: M.F. received a lecture fee from HRA Pharma. S.B., E.L., G.L., I.S., A.S., L.W., S.M., C.E., F.G., A.G., S.K., I.W., M.B., C.L.R., T.D., G.S., A.R., B.A., and M.K. have nothing to disclose.
References
1. Fassnacht M, Kroiss M, Allolio B. Update in adrenocortical carci- noma. J Clin Endocrinol Metab. 2013;98:4551-4564.
2. Else T, Kim AC, Sabolch A, et al. Adrenocortical carcinoma. Endocr Rev. 2014;35:282-326.
3. Terzolo M, Ardito A, Zaggia B, et al. Management of adjuvant mitotane therapy following resection of adrenal cancer. Endocrine. 2012;42:521-525.
4. Bergenstal D, Lipsett M, Moy R, Hertz R. Regression of adrenal cancer and suppression of adrenal function in men by o,p-DDD. Trans Am Physicians. 1959;72:341.
5. Fassnacht M, Terzolo M, Allolio B, et al. Combination chemother- apy in advanced adrenocortical carcinoma. N Engl J Med. 2012; 366:2189-2197.
6. Terzolo M, Angeli A, Fassnacht M, et al. Adjuvant mitotane treat- ment for adrenocortical carcinoma. N Engl J Med. 2007;356:2372- 2380.
7. Huang H, Fojo T. Adjuvant mitotane for adrenocortical cancer- A recurring controversy. J Clin Endocrinol Metab. 2008;93:3730- 3732.
8. Hermsen IG, Fassnacht M, Terzolo M, et al. Plasma concentrations of o,p’DDD, o,p’DDA, and o,p’DDE as predictors of tumor re- sponse to mitotane in adrenocortical carcinoma: Results of a retro- spective ENS@T multicenter study. J Clin Endocrinol Metab. 2011; 96:1844-1851.
9. Kroiss M, Quinkler M, Lutz WK, Allolio B, Fassnacht M. Drug interactions with mitotane by induction of CYP3A4 metabolism in the clinical management of adrenocortical carcinoma. Clin Endo- crinol (Oxf). 2011;75:585-591.
10. Chortis V, Taylor AE, Schneider P, et al. Mitotane therapy in ad- renocortical cancer induces CYP3A4 and inhibits 5a-reductase, ex- plaining the need for personalized glucocorticoid and androgen re- placement. J Clin Endocrinol Metab. 2013;98:161-171.
11. Cai W, Counsell RE, Djanegara T, Schteingart DE, Sinsheimer JE, Wotring LL. Metabolic activation and binding of mitotane in ad- renal cortex homogenates. J Pharm Sci. 1995;84:134-138.
12. Hescot S, Slama A, Lombès A, et al. Mitotane alters mitochondrial respiratory chain activity by inducing cytochrome c oxidase defect in human adrenocortical cells. Endocr Relat Cancer. 2013;20:371- 381.
13. Poli G, Guasti D, Rapizzi E, et al. Morpho-functional effects of mitotane on mitochondria in human adrenocortical cancer cells. Endocr Relat Cancer. 2013;20:537-550.
14. Zsippai A, Szabó DR, Tömböl Z, et al. Effects of mitotane on gene expression in the adrenocortical cell line NCI-H295R: A microarray study. Pharmacogenomics. 2012;13:1351-1361.
15. Stigliano A, Cerquetti L, Borro M, et al. Modulation of proteomic profile in H295R adrenocortical cell line induced by mitotane. En- docr Relat Cancer. 2008;15:1-10.
16. Kroiss M, Reuss M, Kühner D, et al. Sunitinib inhibits cell prolif- eration and alters steroidogenesis by down-regulation of HSD3B2 in adrenocortical carcinoma cells. Front Endocrinol (Lausanne). 2011;2:27.
17. Matysik S, Klünemann HH, Schmitz G. Gas chromatography-tan- dem mass spectrometry method for the simultaneous determination of oxysterols, plant sterols, and cholesterol precursors. Clin Chem. 2012;58:1557-1564.
18. Ecker J, Scherer M, Schmitz G, Liebisch G. A rapid GC-MS method for quantification of positional and geometric isomers of fatty acid methyl esters. J Chromatogr B Analyt Technol Biomed Life Sci. 2012;897:98-104.
19. Liebisch G, Binder M, Schifferer R, Langmann T, Schulz B, Schmitz G. High throughput quantification of cholesterol and cholesteryl ester by electrospray ionization tandem mass spectrometry (ESI- MS/MS). Biochim Biophys Acta. 2006;1761:121-128.
20. Largis EE, Wang CH, DeVries VG, Schaffer SA. CL 277,082: A
novel inhibitor of ACAT-catalyzed cholesterol esterification and cholesterol absorption. J Lipid Res. 1989;30:681-690.
21. Lada AT, Davis M, Kent C, et al. Identification of ACAT1- and ACAT2-specific inhibitors using a novel, cell-based fluorescence assay: Individual ACAT uniqueness. J Lipid Res. 2004;45:378- 386.
22. Sbiera S, Schmull S, Assie G, et al. High diagnostic and prognostic value of steroidogenic factor-1 expression in adrenal tumors. J Clin Endocrinol Metab. 2010;95:E161en]E171.
23. XueJ, Wei J, Dong X, Zhu C, Li Y, Song A, Liu Z. ABCG1 deficiency promotes endothelial apoptosis by endoplasmic reticulum stress- dependent pathway. J Physiol Sci. 2013;63:435-444.
24. Hou NS, Gutschmidt A, Choi DY, et al. Activation of the endoplas- mic reticulum unfolded protein response by lipid disequilibrium without disturbed proteostasis in vivo. Proc Natl Acad Sci U S A. 2014;111:E2271-E2280.
25. Hahner S, Fassnacht M. Mitotane for adrenocortical carcinoma treatment. Curr Opin Investig Drugs. 2005;6:386-394.
26. Ron D, Walter P. Signal integration in the endoplasmic reticulum unfolded protein response. Nat Rev Mol Cell Biol. 2007;8:519- 529.
27. Brewer JW. Regulatory crosstalk within the mammalian unfolded protein response. Cell Mol Life Sci. 2014;71:1067-1079.
28. Park SJ, Kim TS, Park CK, et al. hCG-induced endoplasmic retic- ulum stress triggers apoptosis and reduces steroidogenic enzyme expression through activating transcription factor 6 in Leydig cells of the testis. J Mol Endocrinol. 2013;50:151-166.
29. Sato R. Sterol metabolism and SREBP activation. Arch Biochem Biophys. 2010;501:177-181.
30. Fu S, Yang L, Li P, et al. Aberrant lipid metabolism disrupts calcium homeostasis causing liver endoplasmic reticulum stress in obesity. Nature. 2011;473:528-531.
31. Feng B, Yao PM, Li Y, et al. The endoplasmic reticulum is the site of cholesterol-induced cytotoxicity in macrophages. Nat Cell Biol. 2003;5:781-792.
32. Tabas I. The role of endoplasmic reticulum stress in the progression of atherosclerosis. Circ Res. 2010;107:839-850.
33. Rainbolt TK, Saunders JM, Wiseman RL. Stress-responsive regu- lation of mitochondria through the ER unfolded protein response. Trends Endocrinol Metab. 2014;25:528-537.
34. Raturi A, Simmen T. Where the endoplasmic reticulum and the mitochondrion tie the knot: The mitochondria-associated mem- brane (MAM). Biochim Biophys Acta. 2013;1833:213-224.
35. Rusiñol AE, Cui Z, Chen MH, Vance JE. A unique mitochondria- associated membrane fraction from rat liver has a high capacity for lipid synthesis and contains pre-Golgi secretory proteins in- cluding nascent lipoproteins. J Biol Chem. 1994;269:27494- 27502.
36. Spector AA, Mathur SN, Kaduce TL. Role of acylcoenzyme A: Cho- lesterol o-acyltransferase in cholesterol metabolism. Prog Lipid Res. 1979;18:31-53.
37. Chang TY, Li BL, Chang CC, Urano Y. Acyl-coenzyme A:choles- terol acyltransferases. Am J Physiol Endocrinol Metab. 2009;297: E1-E9.
38. Ogino M, Fukui S, Nakada Y, et al. Discovery of a potent and orally available acyl-CoA: Cholesterol acyltransferase inhibitor as an anti- atherosclerotic agent: (4-phenylcoumarin)acetanilide derivatives. Chem Pharm Bull (Tokyo). 2011;59:1268-1273.
39. Floettmann JE, Buckett LK, Turnbull AV, et al. ACAT-selective and nonselective DGAT1 inhibition: Adrenocortical effects-A cross- species comparison. Toxicol Pathol. 2013;41:941-950.
40. Lee HT, Roark WH, Picard JA, et al. Inhibitors of acyl-CoA:cho- lesterol O-acyltransferase (ACAT) as hypocholesterolemic agents: Synthesis and structure-activity relationships of novel series of sulfonamides, acylphosphonamides and acylphosphoramidates. Bioorg Med Chem Lett. 1998;8:289-294.
41. Ferraz-de-Souza B, Hudson-Davies RE, Lin L, et al. Sterol O-acyl-
transferase 1 (SOAT1, ACAT) is a novel target of steroidogenic factor-1 (SF-1, NR5A1, Ad4BP) in the human adrenal. J Clin En- docrinol Metab. 2011;96:E663-E668.
42. Griffiths B, Lewis CA, Bensaad K, et al. Sterol regulatory element binding protein-dependent regulation of lipid synthesis supports cell survival and tumor growth. Cancer Metab. 2013;1:3.
43. Yue S, Li J, Lee SY, et al. Cholesteryl ester accumulation induced by PTEN loss and PI3K/AKT activation underlies human prostate can- cer aggressiveness. Cell Metab. 2014;19:393-406.
44. van Slooten H, Moolenaar AJ, van Seters AP, Smeenk D. The treat-
ment of adrenocortical carcinoma with o,p’-DDD: Prognostic im- plications of serum level monitoring. Eur J Cancer Clin Oncol. 1984;20:47-53.
45. Terzolo M, Baudin AE, Ardito A, et al. Mitotane levels predict the outcome of patients with adrenocortical carcinoma treated adju- vantly following radical resection. Eur J Endocrinol. 2013;169: 263-270.
46. Kerkhofs T, Baudin E, Terzolo M, et al. Comparison of two mito- tane starting dose regimens in patients with advanced adrenocortical carcinoma. J Clin Endocrinol Metab. 2013;98:4759-4767.