ELSEVIER
Surgery
journal homepage: www.elsevier.com/locate/surg
SURGERY
MWEMBER 2018
$
=
Novel repurposing of sulfasalazine for the treatment of adrenocortical carcinomas, probably through the SLC7A11/xCT-hsa-miR-92a-3p- OIP5-AS1 network pathway
Check for updates
Chitra Subramanian, PhD, MBAª, Kelli McNamara, MSª, Seth W. Croslow, BSb, Yanqi Tan, BSb, Daniel Hess, BSª, Katja Kiseljak-Vassiliades, DOC, Margaret E. Wierman, MDC, Jonathan V. Sweedler, PhDb, Mark S. Cohen, MD, FSSOa,d,*
ª Department of Biomedical and Translational Sciences, Carle Illinois College of Medicine, University of Illinois Urbana-Champaign, Champaign, IL
b Department of Chemistry, University of Illinois Urbana-Champaign, Champaign, IL
” Division of Endocrinology, Metabolism and Diabetes, University of Colorado Anschutz School of Medicine, Aurora, CO
d Department of Surgery, Carle Illinois College of Medicine, University of Illinois Urbana-Champaign, Champaign, IL
ARTICLE INFO
Article history: Accepted 13 July 2024 Available online 18 October 2024
ABSTRACT
Background: Recent multigenomic analysis of adrenocortical carcinomas (ACCs) identified SLC7A11/xCT as a novel biomarker. The Food and Drug Administration-approved anti-inflammatory drug, sulfasala- zine (SAS), induces ferroptosis by blocking SLC7A11 expression. We hypothesize that SAS could be repurposed to target ACC cells.
Methods: Expression of SLC7A11 and its association with ACC survival was analyzed using Gene Expression Profiling Interactive Analysis (GEPIA). The validated ACC cell lines NCI-H295R, ACC1, and ACC2 were grown in 2D culture. In vitro studies included the CellTiter-Glo assay to calculate viability, Western blot (WB) analysis for apoptosis and other target protein changes, reverse transcriptase poly- merase chain reaction for steroidogenic enzyme changes, C11BODIPY for lipid peroxidation, and mass spectrometry for changes in lipids.
Results: The Cancer Genome Atlas Program database analysis in GEPIA showed that SLC7A11 and linked long noncoding RNA OAP5-AS1 are highly expressed in ACC tumors versus normal adrenals (n = 77 vs 128; P < . 05). This was associated with poor overall and disease-free survival with hazard ratios of 4.3 and 5.2 for SLC7A11 and 4.8 and 2.7 for OAP5-AS1, respectively. ACC cell line half-inhibitory maximum concentration values after 72-hour SAS treatment ranged from 412 nM (ACC1) to 799 nM (ACC2), and all showed cleavage of poly (ADP-ribose) polymerase, upregulation of p-Akt and p-ERK, and downregulation of GPX4 and SLC7A11 (P < . 05) by WB analysis. Sphere formation, migration, and invasion assay showed inhibition, and lipid peroxidation using C11BODIPY, increase in intracellular iron, induction of oxidative stress, and significant upregulation of oxidized polyunsaturated fatty acid phospholipids (P < . 05 each) by mass spectrometry suggests induction of ferroptosis.
Conclusion: SAS downregulates tSLC7A11 in ACCs, targets the Akt/ERK pathway and lipid metabolism, and induces cell death in vitro, warranting additional translational studies to define its therapeutic potential in ACC.
@ 2024 The Author(s). Published by Elsevier Inc. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Presented at the 44th Annual Meeting of the American Association of Endocrine Surgeons, Dallas, TX, April 20-22, 2024.
Chitra Subramanian and Mark S. Cohen contributed equally to this work as corresponding authors.
* Reprint requests: Mark S. Cohen, MD, FSSO, Department of Biomedical and Translational Sciences, Carle Illinois College of Medicine, University of Illinois, Urbana-Champaign, 506 S Mathews Ave, Urbana, Champaign, IL 61801.
E-mail address: meddean@illinois.edu (M.S. Cohen).
Introduction
Adrenocortical carcinoma (ACC) is a rare endocrine tumor with an incidence of 1-2 million cases per year.1 Unfortunately, most patients with ACC are diagnosed at an advanced stage, leading to a poor prognosis overall, with 5-year survival rate for patients with advanced ACC being less than 35%.1,2 Although complete surgical resection can cure early-stage localized ACC tumors, one-fourth of the patients with localized tumors
https://doi.org/10.1016/j.surg.2024.07.075
experience recurrence after complete resection. Standard of care treatment for ACC patients with unresectable or metastatic dis- ease often includes either mitotane alone3 or in combination with chemotherapeutics, such as etoposide (E), doxorubicin (D), and cisplatin (P), known as the Italian Protocol (mitotane + EDP), or other systemic regimens or trials.4 The response rate with first line of therapy is limited, with only 1.3% of patients achieving a complete response and 19.2% achieving a partial response. Moreover, the associated toxicities with these systemic drugs create a barrier to completing the therapy.1,2 Despite substantial advances in understanding the molecular mechanisms involved in ACC due to genomic studies, targeted therapies to date have not been effective in improving patient survival, and diagnosis of ACC remains a challenge. Therefore, there remains an imperative need for the identification of novel biomarkers that can help predict outcomes and better diagnose the disease at an earlier timepoint in its evolution.
Metabolic reprogramming is a hallmark of cancer progres- sion.5 As tumors progress from their initial stage to localized tumors and eventually to metastatic cancer, their metabolic properties and phenotypes change.6 Despite the known involvement of metabolic genes in tumor initiation and pro- gression, a comprehensive analysis of metabolic signatures in ACC has not yet been conducted. Thus, our goal was to identify novel therapeutic strategies targeting metabolism by evaluating metabolic signatures that could serve as prognostic and diag- nostic biomarkers of ACC survival. We analyzed metabolic genes and transporters that were differentially regulated in both pri- mary and metastatic ACC tumors from 4 primary (93 tumors) and 2 metastatic (100 tumors) ACC datasets (GSE12368, GSE10927, GSE14922, GSE19750, GSE90713, and GSE143383) compared with 35 normal adrenals. We identified 25 upregulated and 93 downregulated genes (P value <. 05; with >2-fold changes versus normal) in ACC out of the 3,507 metabolism-related genes and transporters identified from the MSigDB database (https://www. gsea-msigdb.org/gsea/msigdb/). Pathway analysis of the signifi- cantly differentially regulated genes in the Database for Anno- tation Visualization and Integrated Discovery (DAVID) identified the solute carrier transport mechanism and glutathione (GSH) metabolism as the top differentially regulated pathways, in addition to steroidogenesis, lipid metabolism, and other path- ways (https://david.ncifcrf.gov). Correlation analysis of the genes in the Gene Expression Profiling Interactive Analysis (GEPIA) database (http://gepia.cancer-pku.cn) showed that the expres- sion of the cysteine/glutamate antiporter, SLC7A11, was highly correlated with all of the top upregulated genes (P < . 0001) in the database. Because SLC7A11 plays a role in GSH metabolism and solute carrier transport in ACC, we evaluated the SLC7A11 inhibitor sulfasalazine (SAS), a Food and Drug Administration-approved drug for inflammatory bowel disease and arthritis,7 for its potential role as a novel therapy for ACC.
Methods
ACC database analysis and identification of the SLC7A11-miRNA- lncRNA network
The interactive web browser GEPIA (http://gepia.cancer-pku.cn) uses RNA sequencing data of tumor and normal samples from The Cancer Genome Atlas Program (TCGA) and the genotype-tissue expression datasets to profile the differential gene expression, correlation, and survival analysis of cancers including ACC. Expression levels, hazard ratios (HRs), and Kaplan-Meyer survival curves for SLC7A11 and associated long noncoding RNAs (lncRNAs) in ACC tumors were compared with normal adrenal samples and
analyzed using ACC TCGA datasets. The miRSystem database (http://mirsystem.cgm.ntu.edu.tw) integrates 7 well-known microRNA (miRNA) target gene prediction programs that include DIANA, miRanda, miRBridge, PicTar, PITA, rna22, and TargetScan. This was used to identify the miRNAs targeted by SLC7A11 in ACC. Differential regulation of miRNAs in ACC that are identified as SLC7A11 targets was then determined using the database of Differentially Expressed MiRNAs in human Cancers (dbDEMC; https://www.biosino.org/dbDEMC). These miRNAs were used as the input in miRnet (https://www.mirnet.ca), a miRNA-centric network visual analytics platform, to determine the lncRNAs associated with SLC7A11 and the target miRNAs in ACCs.
Adrenocortical carcinoma cell culture and viability assay
Genetically validated ACC cell lines ACC1, ACC2, and NCI-H295R were grown in 2D culture in appropriate growth medium. Approximately 1000 cells/well were plated in 96-well plates. Twenty-four hours after plating, the cells were treated with varying concentrations of SAS starting from 0.4 uM to 4 mM. Viability of the cells 72 hours after treatment was determined by luminescence on a BioTek Synergy plate reader (BioTek, Winooski, VT) after the addition of the CellTiter-Glo luminescent reagent (Promega, Fitchburg, WI). The percent viability was calculated as a percentage of solvent control-treated cell viability as 100%. The experiments were conducted in triplicate, and the values were presented as mean ± standard deviation. The half-inhibitory maximum con- centrations (IC50) for SAS were calculated using GraphPad Prism software (Boston, MA).
Immunoblot analysis after treatment with sulfasalazine
ACC cells NCI-H295R, ACC1, and ACC2 were treated with varying concentrations of SAS and solvent control for 24 hours. After treatment, the cells were lysed, and the proteins were quantified using the bicinchoninic acid (BCA) assay reagent (ThermoFisher Scientific, Waltham, MA). Equal amounts of proteins were sepa- rated using sodium dodecylsulfate polyacrylamide gel electropho- resis, and the proteins were transferred to nitrocellulose membrane (GE Healthcare Life Sciences, Piscataway, NJ). The membranes were blocked, incubated with primary and secondary antibodies, and then visualized using a chemiluminescent reagent (ThermoFisher Scientific). Actin was used as a loading control.
Evaluation of lipid peroxidation using C11BODIPY
The level of lipid peroxidation, an indicator for ferroptosis in- duction, was measured after treatment of ACC cells plated in glass bottom 60-mm plates using C11BODIPY (ThermoFisher Scientific). Once the cells were attached, they were treated for 24 hours with IC50 concentrations of SAS followed by treatment with 5 uM C11BODIPY for 15 minutes at 37℃. The cells were washed and fixed, and images were taken using a Keyence microscope (Keyence, Itasca, IL).
Glutathione assay
GSH was measured using a luminescent-based GSH/GSSG-Glo assay (Promega) as per the manufacturer’s protocol. Approximately 10,000 cells were plated on a clear white bottom 96-well plates, and the cells were left to adhere overnight. The next day, the media was removed, and then the cells were treated with varying con- centrations of SAS in Hank’s balance salt buffer for 4 hours. Then the luminescent intensity was measured in a BioTek Synergy plate reader (BioTek).
Analysis of intracellular iron concentration and induction of oxidative stress
ACC cells were treated with the IC50 values of SAS for 24 hours. Intracellular iron levels and induction of oxidative stress were measured using an iron assay kit (Abcam, Waltham, MA) and an oxidative detection kit (ThermoFisher Scientific) according to the manufacturer’s protocol.
Transfection of miRNA mimetics
The ACC cells were transfected with miRNA memetics or a control using Lipofectamine 3000 (ThermoFisher Scientific). Twenty-four hours after transfection, RNA was isolated, reverse transcribed, and SYBR Green quantitative reverse transcriptase polymerase chain reaction (RT-PCR) was performed using SLC7A11 and miRNA92a-3p specific primers. Relative gene levels were calculated using the delta-delta CT method.
Evaluation of migration, invasion, and sphere formation
Migration and invasion assays were conducted in triplicate us- ing chambers that were either noncoated or precoated with Matrigel (BD Biosciences). The inserts were placed into a 24-well plate containing culture medium with 10% fetal bovine serum, and approximately 50,000 ACC cells in serum-free medium were seeded onto the atypical side of the insert. Then, the plates were incubated for 24-48 hours after treatment with SAS. After the treatment, noninvaded cells were removed using a cotton swab, and the invaded cells on the lower side of the insert membrane were stained with crystal violet, air-dried, and imaged. For sphere formation, the cells were plated in ultralow attachment plates in triplicate and treated with varying concentrations of SAS. The number of formed spheres was then analyzed and imaged using a Zeiss microscope.
Mass spectrometry-based single-cell lipid metabolic profiling
Cultured ACC cells (control and 24-hour IC50 concentration of SAS treated) were deposited onto indium tin oxide-coated micro- scope slides, cleaned using 150 mM ammonium acetate, and then a matrix solution containing 45 mg/mL 2,5-dihydroxybenzoic acid dissolved in 70% methanol + 0.1% trifluoroacetic acid was deposited using an HTX M5-Sprayer (HTX Technologies). Matrix- assisted laser desorption/ionization mass spectrometry (MALDI MS) analysis was performed with a SolariX XR 7T FTICR mass spectrometer (mass spectrometer equipped with an APOLLO II dual MALDI/electrospray ionization [ESI] source; Bruker). Data were acquired in positive mode with a mass range of m/z 100-1,600 with 2M data points, yielding a transient length of 0.9787 seconds. micro-MS was used as previously described for targeted, image- guided acquisitions.8 Custom python scripts using DATSIGMA software analyzed the data, binning it with a 3-ppm continuous bin width across the m/z 400 to 1,000 lipid region. Features present in <5% of samples were removed, and the remaining features were mass matched to the LIPID MAPS Structure Database with a ±0.005 Da tolerance. Cells with <5 lipid matches were excluded from the dataset. For any m/z features with isobaric lipid species, the lowest PPM mass error assignment was chosen.
Structural characterization of lipid species via bulk LC-MS/MS
Lipids were extracted using the Bligh and Dyer method. Samples were supplemented with SPLASH LIPIDOMIX Mass Spec Standard as an internal standard and subjected to biphasic separation. The
upper layer was dried using a SpeedVac concentrator before redissolving in 200 uL of a mixture of isopropanol/acetonitile/water 65/30/5. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis was performed on a Waters Acquity UHPLC coupled with Synapt G2-Si Time of Flight Mass spectrometer (Agilent) with an ESI source. Positive and negative data were ac- quired separately, with full MS scans followed by data-dependent acquisition for fragmentation. Peak detection, alignment, and identification were performed on MS-DIAL (Ver 4.90) with a built- in in silico LC-MS/MS lipidomics database. Identification was based on MS2 match, with the score cutoff set at 90%.
Results
Amino acid transporter SLC7A11 upregulation is associated with poor survival
The expression levels of SLC7A11 RNA in ACC patient samples were studied using both the TCGA and the genotype-tissue expression datasets with 77 ACC tumors and 128 normal adrenal cortex samples and put into the GEPIA website for analysis using parameters of a log fold change >1 with a P value cutoff of <. 01. The analysis showed significant upregulation of SLC7A11 in ACC tumors compared with the normal adrenal cortex (Figure 1, A). Further- more, Kaplan-Meyer survival analysis with median cutoff for the ACC TCGA tumors indicated that high expression levels of SLC7A11 correlated with poor overall and disease-free survival. The log-rank P values for overall and disease-free survival were .00011 and 3.3e-5, respectively (Figure 1, B). The HRs for the overall and disease-free survival for SLC7A11 were calculated based on the Cox proportional hazards model. The results showed that higher expression levels of SLC7A11 can be a potential biomarker for ACC as the HRs were 5.2 (p(HR) = . 00046) for the overall survival and 4.3 (p(HR) = . 00012) for the disease-free survival.
SLC7A11 inhibitor sulfasalazine reduces the viability of ACC cells
The viability of 3 different ACC cell lines (NCI-H295R, ACC1, and ACC2) was determined after treatment with SAS for 24 and 72 hours (Figure 1, C) using the CellTiter-Glo reagent according to the manufacturer protocol (Promega). The results showed that SAS reduced the viability of all 3 cell lines in a dose-dependent manner. The IC50 values were 0.6031 ± 0.044, 0.4125 ± 0.055, and 0.7994 ± 0.09 mM after 72 hours and 1.913 ± 0.249, 1.772 ± 0.04, and 2.335 ± 0.347 after 24 hours for the NCI-H295R, ACC1, and ACC2 cell lines, respectively.
Treatment of ACC cells with SAS reduces glutathione activity, increases intracellular iron, and causes oxidative stress, leading to lipid peroxidation
SLC7A11 plays a role in regulating GSH activity by helping in cystine uptake, which affects redox homeostasis.9,10 Therefore, we investigated the impact of different concentrations of SAS on the GSH activity and reactive oxygen species (ROS) of ACC cells. Treatment of ACC1, ACC2, and NCI-H295R cells with SAS resulted in a decrease of over 80% in GSH levels, starting from 1 mM (Figure 2, A). We also observed a dose-dependent decrease in the ratio of reduced to oxidized GSH (GSH/GSSG) for all 3 cell lines starting from 1 mM SAS. The investigation of induction of oxidative stress using CM-H2DCFDA staining after SAS treatment resulted in a 7- to 8-fold increase in ROS (Figure 2, C). Furthermore, immune blot analysis of ACC cells showed a reduction in the expression levels of SLC7A11 and GPX4 with increasing doses of SAS (Figure 2, D). During ferroptosis, inhibition of GPX4 results in an increase in
A
B
Disease Free Survival
Overall Survival
1.0
Low SLC7A11 TPM
1.0
2.0
High SLC7A11 TPM
…
Low. SLC7A11 TPM
Logrank p=3.3e-05
High SLC7A11 TPM
HR(high)=4.3
Logrank p=0.00011
0.8
p(HR)=0.00012
0.8
HR(high)=5.2
p(HR)=0.00046
1.5
n(high)=38
n(low)=37
n(high)=38
Log2(TPM+1)
Percent survival
Percent survival
0.6
0.6
n(low)=37
1.0
0.4
0.4
0.5
0.2
8
8
ACC
Normal
0.0
8
0
50
100
150
0
50
100
150
C
Months
Months
100
75
% Viability
50
25
ACCI
ACC2
NC1-H295R
0
I
-1.5
-1.0
-0.5
0.0
0.5
1.0
Sulfasalazine (mM)
| IC50 (mM) | ACC1 | ACC2 | NCI-H295R |
|---|---|---|---|
| 72 h | 0.4125±0.055 | 0.7994±0.09 | 0.6031±0.044 |
| 24 h | 1.772±.0.04 | 2.335±0.347 | 1.913±0.249 |
A
D
1.2
2.5
NCI-H295R
ACC1
ACC2
Fold Change in GSH
1
ONCI-H295R
DNCI-H295R
DACC1
Fold Change in GSH/GSSG
BACC1
BACC2
2
BACC2
0
0.5
1
2 SAS (m.M)
0
0.5
1
2
SAS (m.M)
0
0.5
1
2
SAS (mM)
0.8
1.5
SLC7A11
SLC7A11
SLC7A11
0.6
**
1
0.008
0.008
0.009
1
0.061
0.012
0.011
1
0.539
0.088
0.019
0.4
1
GPX4
GPX4
GPX4
0.5
1
0.640
0.368
0.391
0.2
1
1.161
0.417
0.232
1
0.436
0.197
0.064
Actin
Actin
Actin
0
0
Control
1
2
4
Control
1
2
4
SAS (mM)
SAS (mM)
B
C
E
NCI-H295R
ACC1
ACC2
2.5
Relative iron concentration
OControl OSAS
10
DCFDA intensity (Fold change)
O Control OSAS
2
8
Control
**
1.5
6
1
4
0.5
2
SAS
0
0
NCI-H295R
ACC1
ACC2
NCI-H295R
ACC1
ACC2
C11 non-ox/C11ox/DAPI
Figure 2. Treatment of ACC cells with SAS reduces glutathione activity, increases intracellular iron, and causes oxidative stress, leading to lipid peroxidation. (A) The Promega GSH/ GSSG-Glo assay was used to measure GSH/GSSG activity in ACC cells treated with varying concentrations of SAS. (B) An intracellular iron assay after treatment of ACC cells with SAS. (C) ACC cells treated with SAS showed increased ROS, which was reversed by treatment with the oxidative stress inhibitor N-acetyl cysteine. (D) Immunoblot analysis of ACC cells after SAS treatment revealed a decrease in the glutathione pathway proteins GPX4 and SLC7A11. Actin was used as a loading control. (E) Induction of lipid peroxidation by SAS in ACC cells using C11BODIPY. The cells were treated with an IC50 concentration of SAS for 4 hours, followed by treatment with C11BODIPY and DAPI for 30 minutes. The cells were then fixed and imaged using a Keyence microscope. The induction of lipid peroxidation was indicated by a change in the color of C11BODIPY from red to green. In addition, the nuclei were stained using DAPI for better visualization (E). The statistical significance is denoted by * P < . 05, ** P < . 01, *** P < . 001. ACC, adrenocortical carcinoma; DAPI, 4’,6- diamidino-2-phenylindole; GSH, glutathione; IC50, half-inhibitory maximum concentration; ROS, reactive oxygen species; SAS, sulfasalazine.
intracellular iron levels to induce lipid peroxidation.10,11 The intracellular iron level measured showed a 1.6- to 2.2-fold increase in intracellular Fe2+ iron concentrations after SAS treatment compared with control cells (Figure 2, B). To evaluate the induction of lipid peroxidation after treatment of ACC cells with SAS, we monitored changes in the green fluorescence of cells that had been labeled with 5 AM BODIPY 581/591 dye. Figure 2, C, showed an increase in green fluorescence intensity by 2.8- to 4.0-fold and a decrease in red intensity by 40%-50% after treatment of ACC cells with SAS (Figure 2, E), indicating the induction of lipid peroxida- tion. Induction of ferroptosis was further confirmed by measuring the oxidized lipids by MS (Table II ). Several polyunsaturated fatty acid (PUFA) oxidized phospholipids showed an increase after SAS treatment.
SAS treatment of ACC cells induces cell death by modulating the Akt/ NFKB/ERK pathway
Lipid peroxidation is a crucial process involved in cell death mechanisms such as apoptosis, autophagy, and ferroptosis.12 In this study, we investigated the effects of different concentrations of SAS on ACC cells by analyzing the induction of apoptosis and
autophagy using immunoblot analysis. Our results showed a dose- dependent increase in poly (ADP-ribose) polymerase cleavage and LC3-II in ACC cells treated with SAS at IC50 concentrations (Figure 3, A). To understand the mechanism through which SAS induces cell death, we performed an immunoblot analysis of Akt/ MEK/NFKB pathway proteins. The results revealed that SAS treatment led to a dose-dependent decrease in Akt and p-P65 levels and an increase in p-ERK and p-Akt levels. However, the levels of total ERK and total P65 remained unchanged, supporting that the induction of cell death by SAS is in part through the Akt/ MEK/NFKB pathways in ACC cells (Figure 3, B).
Treatment of ACC cells with sulfasalazine reduces migration, invasion, and sphere formation
To examine whether treatment of ACC cells with SAS decreases invasive potential of ACC cells, we performed migration, invasion, and sphere assay after treatment of ACC cells with varying con- centrations of SAS. As seen from Figure 4, A-C, a dose-dependent decrease in migration, invasion, and sphere formation was observed. There is a significant reduction in migration and inva- sion (50%) at 2 mM SAS and a nearly total reduction (>90%) at
A
NCI-H295R
ACC1
ACC2
0
0 .. 5
1
2
SAS (mM)
0
0 .. 5
1
2
SAS (mM)
0
0 .. 5
1
2
SAS (mM)
PARP
PARP
PARP
Cleaved PARP
Cleaved PARP
Cleaved PARI
LC3-I
LC3-I
LC3-I
LC3-II
LC3-II
LC3-II
Actin
Actin
Actin
B NCI-H295R
ACC1
ACC2
0
0.5
1
2
SAS (mM)
0
0.5
1
2
SAS (mvM)
0
0.5
1
2
SAS (mvM)
p-Akt
p-Akt
p-Akt
1
3.4 0
4.03
3.98
1
1.30
1.50
1.66
1
1.67
2.17
4.65
Akt
Akt
Akt
1
0.75
0.27
0.27
1
0.70
0.66
0.61
1
1.10
0.85
0.13
p-ERK
p-ERK
p-ERK
1
1.35
1.27
6.89
1
1.02
2.31
2.52
1
1.5
5.96
5.90
ERK
ERK
ERK
1
0.95
0.94
0.90
1
1.08
1.05
1.01
1
1.20
1.09
0.98
p-p65
p-p65
p-p65
1
0.70
0.28
0.11
1
1.18
1.25
0.49
1
1.09
0.92
0.20
p65
p65
p65
1
0.95
0.91
1.06
1
1.06
1.05
1.07
1
1.24
1.14
1.04
Actin
Actin
Actin
A
Migration
B
Invasion
C
Control
SAS 2µM
SAS 4uM
Control
SAS 2µM
SAS 4µM
Control
SAS 0.5 mM
SAS 1mM
SAS 2.mM
Migration
Invasion
% Change in number of Spheres
NCI-H295R
ONCI-H295R
ONCI-H295R
100
ACC1
DACC1
100
DACC1
ACC2
WACC2
100
WACC2
75
% Migration
75
% Invasion
75
50
50
50
25
25
25
0
0
0
Control
2uM
4uM
Control
2uM
4uM
Control
0.5mM
1mM
2mM
4 mM SAS for all the ACC cells. Sphere formation showed a 25% decrease at 0.5 mM SAS and a greater than 80% decrease around 2 mM.
Construction of the SLC7A11-miRNA-lncRNA network
To understand the regulatory network of SLC7A11-miRNA- lncRNA, we investigated the negative modulation of gene transcription and translation by miRNAs, as well as the positive regulation of target gene translation by IncRNAs that act as endogenous miRNA sponges.13,14 Therefore, to construct the SLC7A11-miRNA-lncRNA network, we first scanned the miRNA signatures that regulate SLC7A11 using miRSystem (http:// mirsystem.cgm.ntu.edu.tw). Of the 11 miRNAs identified from 5 or more miRNA target gene prediction sites, only 6 miRNAs are involved in the SLC7A11 regulation in ACC (Table I). Next, using miRnet, we have identified the 9 lncRNAs regulated by the miRNAs (Table I). We observed that hsa-miR-92a-3p is downregulated in ACC tumors compared with normal tissues (Figure 5, A). To further understand the link between miR92a-3p and SLC7A11, we trans- fected the ACC cells with miRNA mimetics and control miRNA and then evaluated the changes in SLC7A11 and miR92a-3p levels by RT-PCR. Our results indicated that the upregulation of miRNA 92a-
| miRNA | lncRNA |
|---|---|
| hsa-miR-25-3p | LINC01128 |
| hsa-miR-27a-3p | PAX8-AS1 |
| hsa-miR-32-5p | LINC00963 |
| hsa-miR-363-3p | KCNQ1OT1 |
| hsa-miR-92a-3p | PWAR5 |
| hsa-miR-92b-3p | OIP5-AS1 |
| GABPB1-AS1 | |
| CCDC144NL-AS1 |
ACC, adrenocortical carcinoma; lncRNA, long noncoding RNA; miRNA, microRNA.
3p resulted in the downregulation of SL7A1, confirming that miRNA92a-3p regulates SLC7A11 (Figure 5, B). Furthermore, OIP5- AS1 was identified as the lncRNA associated with poor disease- free and overall survival in TCGA ACC patient tumors using the GEPIA website. The survival analysis showed an HR of 4.8 (p(dis- ease-free survival) = . 00011) and 2.7 (p(overall survival) = . 022), respectively (Figure 5, C). Based on this, it appears that the competing RNA (ceRNA) regulatory network OIP5-AS1-hsa-miR- 92a-3p-SLC7A11 (Figure 5, D) plays an important role in ACC.
Lipid metabolic profiling by single-cell mass spectrometry
As lipid metabolic pathways are involved in lipid peroxidation and ferroptosis,15 we performed single-cell lipid metabolic profiling of the ACC cell line NCI-H295R after treatment with SAS via MALDI MS and LC-MS/MS. The results showed that a total of 146 lipid features were upregulated and 42 were downregulated, with 70 and 23 of them being oxidized lipid features, respectively (Figure 6, A and B). Further analysis revealed that the majority of statistically significant features were upregulated in several lipid classes including phosphatidylethanolamine-ceramides, phos- phatidylcholine, sphingomyelins, and sterols, as shown in Figure 6, C. Putative lipid assignments for MALDI MS measure- ments, including their mass error, Q values, fold change, and LC- MS/MS structural characterization (where applicable), are shown in Table II.
Discussion
Despite advances in cancer therapy and the identification of molecular markers for ACC through genomic studies, ACC has an unfavorable prognosis.1 Although several research groups and companies are evaluating novel targeted or immune-modulating therapies for ACC, these drugs take 7-10 years or more to get from bench to bedside, very few if any of these will advance into the clinic, and to date none have shown significant improvements in survival in trials. An alternate and vastly more expedited approach
A
Hsa-mir-92a-3p
C
Disease Free Survival
Overall Survival
8.5
0
Low OIP5-AS1 TPM
High.OIP5-AS1.TPM
1.0
Logrank p=1.9e-05
Low OIP5-AS1 TPM
High OIP5-AS1. TPM
8
0.8
HR(high)=4.8
Logrank p=0.017
p(HR)=0.000+1
Percent survival
n(high)=38
0.8
HR(high)=2.7
7.5
p(HR)=0.022
n(low)=38
Percent survival
n(high)=38
TPM
0.6
7
0.6
n(low)=38
6.5
0.4
0.4
6
0.2
5.5
0.2
5
0.0
ACC
Normal
8
0
50
100
150
Months
0
50
100
150
Months
B
D
4
miRNA 92a-3p
1.200
SLC7A11
“Control
OmiRNA mimics
Fold Change in terms of Control
OControl ImiRNA mimetics
Fold Change in terms of Control
1.000
3
0.800
*
**
hsB-mir-278-3p
2
0.600
DIP5-AS1
0.400
han-mir-363-3p
1
had mor 32-5p
0.200
haut-mir-92b-3p
hsh mir-923.3p
fisa-mir-25-3p
0
0.000
NCI-H295R
ACC1
ACC2
NCI-H295R
ACC1
ACC2
involves repurposing of already FDA-approved drugs that target important tumorigenic mechanisms in ACC as a platform for developing novel therapies for this cancer. Metabolic rewiring plays a crucial role in cancer progression and therapy resistance. Intrinsic and extrinsic factors affect cancer metabolism, resulting in meta- bolic heterogeneity among various tumors, including ACC.6,16 The extrinsic factors that alter metabolism depend on the tumor microenvironment as well as the patient’s systemic metabolism, whereas the intrinsic factors responsible include alternations in cell signaling or gene expression, and genetic mutations, among others.6 Therefore, in analyzing the TCGA database to identify differentially expressed metabolic genes and transporters associ- ated with the survival of patients with ACC, we identified the cystine/glutamate antiporter SLC7A11 as one of the hub metabolic genes that is significantly altered in ACCs compared with normal adrenal tissue.
In this study, we investigated the role of SLC7A11 in cancer progression and its potential as a therapeutic target for ACC. Our analysis of TCGA data on the GEPIA platform revealed higher expression of SLC7A11 in ACC cells than in normal adrenal cells, and this overexpression was associated with both poor overall and disease-free survival. To evaluate the potential of SLC7A11 inhibi- tion as an anticancer strategy, we used SAS, an FDA-approved drug that inhibits SLC7A11. Our in vitro SAS treatment decreased the viability of ACC cells, suggesting its potential to be repurposed as an anticancer drug for ACC.
SLC7A11 is responsible for importing extracellular cystine and exporting intracellular glutamate for GSH biosynthesis.11 GSH depletion by SAS treatment activates various cell death mecha- nisms, including autophagy, apoptosis, and ferroptosis, depending on the cancer model.17-22 In our study, SAS treatment led to fer- roptosis in ACC cells, as evidenced by the downregulation of GPX4, increase in intracellular iron, ROS, lipid peroxidation, and oxidation of PUFA phospholipids. In addition, we observed cleavage of poly (ADP-ribose) polymerase and upregulation of LC3-II, indicating apoptotic and autophagic cell death due in part to the generation of lipid peroxidation during ferroptosis. Immunoblot analysis of Akt/ NFKB/MAPK pathway genes revealed that SAS treatment of ACC cells targeted these pathways at the same IC50 concentrations, leading to apoptosis and autophagic cell death. As lipid peroxida- tion and ferroptosis alter lipid metabolism, we conducted single- cell lipid metabolic profiling by MALDI MS, and similar to our previous findings,15 we noted that oxidized phospholipids were upregulated. In addition, we found upregulation of oxidized sphingomyelin and sterols. Further evaluation of migratory po- tential of ACC cells also showed downregulation of migration, in- vasion, and sphere formation, further supporting that SAS could be repurposed as a potential novel treatment for ACC.
Noncoding RNAs, such as miRNAs and lncRNAs, play a signifi- cant role in regulating gene expression. These biomarkers can be easily detected using liquid biopsy, which is a minimally invasive diagnostic and prognostic tool.23 Studies have shown that the
A
Downregulated
· Not Significant
Upregulated
B
Percent of Cells:
10%
50%
100%
Total Lipids
Oxidized Lipids
ACC Control
ACC Sulfa 2 mM
ACC Control
ACC Sulfa 2 mM
160
60
140
-Log, (Adj P Value)
50
120
-Log,,(Adj P Value)
100
40
80
30
60
20
40
20
10
0
1.11. 9= 0.05
0
0.05
-1.5
-1.0
Log2(Fold Change)
-0.5
0.0
0.5
1.0
-1.5
-1.0
Log2(Fold Change)
-0.5
0.0
0.5
1.0
C
CerPE
PC
SM
ST
2.00
ACC Control
ACC Sulfa 2 mM
Fold Change (Treated/Control)
1.75
1
1
1.50
**
L
T
L
I
1.25
I
I
I
**
I
L
I
1.00
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
0.75
0.50
T
I
I
0.25
0.00
34:3;03
36:1;02
38:1;02
38:1;021
40:1;02:
O-28:0
O-32:1
O-32:3
O-34:3
O-38:4
34:2,02
36:0;02
36:1;021
40:1;021
42:3;02
24:1;05;T
26:0;06
28:2;07
30:2;03
30:3;0
Lipid Species
ceRNA network is associated with the pathophysiology of cancer cells and the tumor microenvironment.24,25 Therefore, to construct a ceRNA network of mRNA-miRNA-lncRNA, we con- ducted an analysis to identify the miRNAs and lncRNAs that are associated with SLC7A11 in ACCs. Our research revealed that SLC7A11 is associated with the miRNA hsa-miR-92a-3p and the lncRNA OIP5-AS1 in ACC, indicating that the SLC7A11-hsa-miR- 92a-3p-OIP5-AS1 ceRNA network plays an important role in ACC cells.
Study limitations
A major limitation of this study is that gene analysis correla- tions with survival and in vitro studies may not translate into improved in vivo or clinical outcomes. Although similar concen- trations of SAS were used in vivo for FDA regulatory studies, several more in vivo validation studies will be needed to evaluate the potential role of repurposing SAS as a novel therapy in ACC. An additional limitation of our study is that we only evaluated the efficacy of SAS in SLC7A11-expressing ACC cells. Testing the effect of SAS in SLC7A11 knockout cells would further validate and support our findings. Overall, these initial studies are promising though and suggest that inhibiting SLC7A11 could be a new therapeutic approach for ACC and other cancers that rely on this pathway for redox homeostasis.
Funding/Support
This research was funded in part by the National Institutes of Health (R01 CA173292 and R01 CA216919; M.S.C. and B.S.J.B.) and
the Carle Illinois College of Medicine, University of Illinois Urbana- Champaign Cancer Center.
Conflict of Interest/Disclosure
The authors have nothing to disclose regarding this manuscript.
Acknowledgments
We would like to thank Dr Catherine Alicia Best-Popescu and Jeoge Maldonado De Jesus for help with the Keyence microscope.
CRediT authorship contribution statement
Chitra Subramanian: Writing - review & editing, Writing - original draft, Visualization, Validation, Supervision, Software, Re- sources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptuali- zation. Kelli McNamara: Visualization, Validation, Methodology, Data curation. Seth W. Croslow: Writing - review & editing, Visualization, Validation, Software, Resources, Methodology, Investigation, Formal analysis, Data curation. Yanqi Tan: Writing - review & editing, Validation, Software, Methodology, Formal analysis, Data curation. Daniel Hess: Visualization, Validation, Methodology, Data curation. Katja Kiseljak-Vassiliades: Resources. Margaret E. Wierman: Resources. Jonathan V. Sweedler: Writing - review & editing, Supervision, Software, Project administration, Methodology. Mark S. Cohen: Writing - review & editing, Visualization, Validation, Supervision, Software, Resources,
| Lipid species | Adduct | Mass error (PPM) | Q value | Fold change | LC-MS annotation |
|---|---|---|---|---|---|
| PC O-32:1 | [M+H]+ | 2.644124 | 1.19E-60 | 0.498756 | PC O-16:1_16:0, PC O-16:0_16:1 |
| PC O-34:3 | [M+H]+ | 0.538666 | 4.43E-56 | 1.829166 | PC O-18:1_16:2 |
| ST 24:1;O5;T | [M+Na]+ | 6.873735 | 1.81E-36 | 0.461391 | |
| CerPE 38:1;O2 | [M+Na]+ | 0.40564 | 2.34E-36 | 1.803135 | |
| PE 28:2;O2 | [M+H]+ | 0.903045 | 2.66E-35 | 0.425482 | |
| SM 34:2;O2 | [M+Na]+ | 2.625974 | 7.2E-26 | 0.352534 | SM 18:2;O2/16:0 |
| SM 42:3;O2 | [M+H]+ | 1.848045 | 3.51E-24 | 1.367788 | SM 18:1;O2/24:2 |
| FA 40:7;O3 | [M+K]+ | 3.306013 | 1.31E-21 | 0.286423 | |
| PC O-38:4 | [M+K]+ | 4.912667 | 7.69E-20 | 1.280641 | PC O-16:0_22:4, PC O-18:0_20:4, PC O- 18:1_20:3 |
| IPC 34:1;O2 | [M+H]+ | 4.484084 | 2.38E-18 | 1.238379 | |
| SM 36:0;O2 | [M+K]+ | 4.795369 | 6.19E-16 | 1.273483 | SM 20:0;O2/16:0 |
| PE 28:2;O2 | [M+H]+ | 6.923348 | 2.15E-15 | 1.414084 | |
| PC O-28:0 | [M+H]+ | 4.664968 | 1.41E-14 | 1.277998 | PC O-12:0_16:0, PC O-14:0_14:0 |
| CerP 26:1;O2 | [M+H]+ | 2.962316 | 1.19E-13 | 1.504632 | |
| SM 40:1;O2 | [M+Na]+ | 3.581792 | 4.71E-12 | 1.18543 | SM 18:1;O2/22:0 |
| PA O-38:0 | [M+K]+ | 6.0722 | 2.07E-11 | 1.398345 | |
| PS 24:3;O2 | [M+H]+ | 4.766811 | 6.29E-11 | 0.446751 | |
| PA O-34:5 | [M+H]+ | 2.907626 | 8.76E-10 | 1.269255 | |
| ST 30:3;O | [M+K]+ | 7.338124 | 2E-09 | 1.54066 | |
| PC O-32:3 | [M+H]+ | 4.618335 | 3.97E-09 | 1.419456 | |
| PC O-38:2 | [M+K]+ | 1.788677 | 1.54E-08 | 1.299967 | |
| CerP 36:1;O2 | [M+Na]+ | 4.188489 | 2.68E-08 | 0.481728 | |
| SM 36:1;O2 | [M+Na]+ | 0.663492 | 3.6E-08 | 1.451994 | |
| ST 30:2;O3 | [M+K]+ | 9.651363 | 6.67E-08 | 1.179246 | |
| SM 32:1;O2 | [M+Na]+ | 3.870826 | 1.04E-07 | 1.602628 | SM 16:1;O2/16:0, SM 18:1;O2/14:0 |
| LPS O-26:1;O | [M+Na]+ | 5.451068 | 1.33E-07 | 0.801872 | |
| SM 38:1;O2 | [M+H]+ | 3.291044 | 2.53E-07 | 1.625608 | |
| CerPE 38:1;O2 | [M+H]+ | 0.27871 | 3.13E-07 | 1.370171 | |
| PS O-36:1 | [M+H]+ | 0.901388 | 3.46E-07 | 1.425103 | |
| CAR 14:2;O2 | [M+K]+ | 3.194704 | 3.67E-07 | 1.405073 | |
| PC O-36:3 | [M+H]+ | 4.412113 | 5.78E-07 | 1.270034 | PC O-16:0_20:3, PC O-18:1_18:2 |
| SM 38:1;O2 | [M+H]+ | 2.632835 | 1.32E-06 | 1.11201 | |
| ST 24:1;O4;GlcA | [M+H]+ | 2.283378 | 1.45E-06 | 1.442313 | |
| PC O-38:7 | [M+Na]+ | 2.953641 | 1.58E-06 | 1.401014 | PC O-16:1_22:6 |
| PC O-30:0 | [M+H]+ | 2.310273 | 8.7E-06 | 1.524171 | PC O-16:0_14:0 |
| PC 36:5;O3 | [M+H]+ | 1.08625 | 1.74E-05 | 1.307198 | |
| PS O-34:1 | [M+H]+ | 3.60698 | 2.32E-05 | 1.804126 | |
| LPS O-28:1;O | [M+H]+ | 4.801414 | 6.54E-05 | 0.650831 | |
| CerPE 40:1;O2 | [M+K]+ | 5.61527 | 7.59E-05 | 1.850652 | |
| IPC 36:1;O2 | [M+H]+ | 4.575981 | 8.41E-05 | 1.260163 | |
| CerP 26:1;O2 | [M+Na]+ | 0.946356 | 8.78E-05 | 1.64145 | |
| Hex2Cer 32:1;O2 | [M+Na]+ | 2.334878 | 0.000157 | 1.328935 | |
| PC O-32:0 | [M+H]+ | 1.804077 | 0.000169 | 1.326458 | PC O-16:0_16:0, PC O-17:0_15:0 |
| PC 22:2;O | [M+Na]+ | 2.387172 | 0.000304 | 1.385097 |
CerPE, phosphatidylethanolamine-ceramides; FT-ICR MS, Fourier-transform ion cyclotron resonance mass spectrometry; LC-MS/MS, liquid chromatography-tandem mass spectrometry; PC, phosphatidylcholine; SM, sphingomyelins; ST, sterols.
Project administration, Investigation, Funding acquisition, Conceptualization.
References
1. Allolio B, Fassnacht M. Clinical review: adrenocortical carcinoma: clinical up- date. J Clin Endocrinol Metab. 2006;91:2027-2037.
2. Pommier RF, Brennan MF. An eleven-year experience with adrenocortical carcinoma. Surgery. 1992;112:963-970.
3. Lo Iacono M, Puglisi S, Perotti P, et al. Molecular mechanisms of mitotane action in adrenocortical cancer based on in vitro studies. Cancers. 2021;13:5255.
4. Else T, Kim AC, Sabolch A, et al. Adrenocortical carcinoma. Endocr Rev. 2014;35: 282-326.
5. Hanahan D. Hallmarks of cancer: new dimensions. Cancer Discov. 2022;12: 31-46.
6. Faubert B, Solmonson A, DeBerardinis RJ. Metabolic reprogramming and cancer progression. Science. 2020;368, eaaw5473.
7. Ma MZ, Chen G, Wang P, et al. Xc- inhibitor sulfasalazine sensitizes colorectal cancer to cisplatin by a GSH-dependent mechanism. Cancer Lett. 2015;368: 88-96.
8. Comi TJ, Neumann EK, Do TD, et al. MicroMS: a Python platform for image-guided mass spectrometry profiling. J Am Soc Mass Spectrom. 2017;28:1919-1928.
9. Xie YR, Chari VK, Castro DC, et al. Data-driven and machine learning-based framework for image-guided single-cell mass spectrometry. J Proteome Res. 2023;22:491-500.
10. Koppula P, Zhuang L, Gan B. Cystine transporter SLC7A11/xCT in cancer: fer- roptosis, nutrient dependency, and cancer therapy. Protein Cell. 2021;12: 599-620.
11. Latunde-Dada GO. Ferroptosis: role of lipid peroxidation, iron and ferriti- nophagy. Biochim Biophys Acta Gen Subj. 2017;1861:1893-1900.
12. Su LJ, Zhang JH, Gomez H, et al. Reactive oxygen species-induced lipid per- oxidation in apoptosis, autophagy, and ferroptosis. Oxid Med Cell Longev. 2019;2019:5080843.
13. Oliveto S, Mancino M, Manfrini N, et al. Role of microRNAs in translation regulation and cancer. World J Biol Chem. 2017;8:45-56.
14. Tay Y, Rinn J, Pandolfi PP. The multilayered complexity of ceRNA crosstalk and competition. Nature. 2014;505:344-352.
15. Lee JY, Kim WK, Bae KH, et al. Lipid metabolism and ferroptosis. Biology. 2021;10:184.
16. Wang Q, Sun N, Meixner R, et al. Metabolic heterogeneity in adrenocortical carcinoma impacts patient outcomes. JCI Insight. 2023;8, e167007.
17. Nagane M, Kanai E, Shibata Y, et al. Sulfasalazine, an inhibitor of the cystine- glutamate antiporter, reduces DNA damage repair and enhances radiosensi- tivity in murine B16F10 melanoma. PLoS One. 2018;13, e0195151.
18. Schnelldorfer T, Gansauge S, Gansauge F, et al. Glutathione depletion causes cell growth inhibition and enhanced apoptosis in pancreatic cancer cells. Cancer. 2000;89:1440-1447.
19. Han HY, Kim H, Jeong SH, et al. Sulfasalazine induces autophagic cell death in oral cancer cells via Akt and ERK pathways. Asian Pac J Cancer Prev. 2014;15: 6939-6944.
20. Kerkhove L, Geirnaert F, Rifi AL, et al. Repurposing sulfasalazine as a radio- sensitizer in hypoxic human colorectal cancer. Cancers. 2023; 15:2363.
21. Sendo K, Seino M, Ohta T, Nagase S. Impact of the glutathione synthesis pathway on sulfasalazine-treated endometrial cancer. Oncotarget. 2022;13: 224-236.
22. Liu J, Lou C, Zhen C, Wang Y, Shang P, Lv H. Iron plays a role in sulfasalazine- induced ferroptosis with autophagic flux blockage in K7M2 osteosarcoma cells. Metallomics. 2022;14, mfac027.
23. Subramanian C, McCallister R, Cohen MS. Multi-genomic analysis of 260 adrenocortical cancer patient tumors identifies novel network BIRC5-hsa-miR-
335-5p-PAX8-AS1 strongly associated with poor survival. Surgery. 2023;173: 43-51.
24. Gao L, Zhao Y, Ma X, Zhang L. Integrated analysis of lncRNA-miRNA-mRNA ceRNA network and the potential prognosis indicators in sarcomas. BMC Med Genomics. 2021;14:67.
25. Chen J, Song Y, Li M, et al. Comprehensive analysis of ceRNA networks reveals prognostic lncRNAs related to immune infiltration in colorectal cancer. BMC Cancer. 2021;21:255.
Discussion
Check for updates
Dr Priya Dedhia (Columbus, OH): Did you look at actual cortisol production or other hormonal production with treatment with the drug? Did you look at metabolomic pathways and specifically this gene?
Dr Mark S. Cohen (University of Illinois Urbana-Champaign): So the answer to both is not yet at this time. It will be important to understand where the synergy is between this and other drugs we typically use for standard treatment of adrenocortical carcinoma (ACC) such as cisplatin.
Dr Matthew Nehs (Boston, MA): Did you look at total cellular oxidative stress as a contributing mechanism to sulfasalazine, because the anti-inflammatory and oxidative stress is driving a lot of this malignancy? Did you look at that as a mechanism, in addi- tion to lipid peroxidation?
Dr Mark S. Cohen: We are performing those studies now and comparing several normal nonmalignant cell lines to look at the effect of reactive oxygen species or oxidative stress in this pathway, especially if ferroptosis is going to be a potential
mechanism of cell death. Based on our early studies, I can say it looks like there is a differential selectivity for ACC cells compared with normal cells and other types of cancers. So I think there is a real mechanism there. We still need to perform a few more studies to elucidate that.
Dr Michael Demeure (Newport Beach, CA): Just wondering if you have been looking at other cancers, notably colorectal cancer, where sulfasalazine has been studied and how your data compare in that context?
Dr Mark S. Cohen: That’s what we are looking at and want to understand. This is certainly not going to be a single-agent therapy for ACC. The real question is, Does the anti-inflammatory effect and some of the oxidative effect synergize with other treatments? Can we potentially lower toxicity dosing and create a combination ef- fect that’s going to improve outcomes? We are looking at a number of different cancers comparatively to see if there’s selectivity for ACC. Also the effect on steroidogenic enzymes is something that was not seen in colorectal tumors.