Therapeutic potential of CPI-613 for targeting tumorous mitochondrial energy metabolism and inhibiting autophagy in clear cell sarcoma

Clear cell sarcoma (CCS) is an aggressive type of soft tissue tumor that is associated with high rates of metastasis. In the present study, we found that CPI-613, which targets tumorous mitochondrial energy metabolism, induced autophagosome formation followed by lysosome fusion in HS-MM CCS cells in vitro. Interestingly, CPI-613 along with chloroquine, which inhibits the fusion of autophagosomes with lysosomes, significantly induced necrosis of HS-MM CCS cell growth in vitro. Subsequently, we established a murine orthotropic metastatic model of CCS and evaluated the putative suppressive effect of a combination of CPI-613 and chloroquine on CCS progression. Injection of HS-MM into the aponeuroses of the thigh, the most frequently affected site in CCS, resulted in massive metastasis in SCID-beige mice. By contrast, intraperitoneal administration of CPI-613 (25 mg/kg) and chloroquine (50 mg/kg), two days a week for two weeks, significantly decreased tumor growth at the injection site and abolished metastasis. The present results imply the inhibitory effects of a combination of CPI-613 and chloroquine on the progression of CCS.


Introduction
Clear cell sarcoma (CCS) affects the deep soft tissues of young adults and is known to have high rates of metastasis [1,2]. Lymphatic metastasis is rare in other malignant soft tissue tumors but is commonly detected in CCS [3,4]. Radical surgical resection is the first line of treatment of CCS. However, the rate of local recurrence can reach as high as 84% and the rate of late metastases can be as high as 63%, and these are associated with the 5-20-year survival rate of 67-10% [5]. Considering that CCS is relatively resistant to conventional soft tissue sarcoma chemotherapy regimens, there is an urgent need to develop therapies that also control metastasis.
CPI-613 is a first-in-class agent that is an analog of α-lipoic acid coenzyme and targets tumor mitochondrial energy metabolism [6]. CPI-613 can also induce a burst of mitochondrial reactive oxygen species in cancer cells [7]. In line with this concept, CPI-613 specifically kills tumor cells without exhibiting toxicity to primary normal cell counterparts [7]. Serial clinical PLOS ONE | https://doi.org/10.1371/journal.pone.0198940 June 7, 2018 1 / 10 a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 trials on CPI-613 have been completed for hematological malignancies and pancreatic cancer, and the drug was well tolerated in patients [8,9]. In this report, we describe the putative suppressive efficacy of CPI-613 on the local and metastatic growth of CCS using a novel murine model with high rates of metastasis.
To detect autolysosomes, we employed the DALGreen agent (Dojindo Co., Kumamoto, Japan) according the manufacturer's protocol. Briefly, DALGreen, which is a small hydrophobic molecule, passes the cell surface membrane of live cells and is incorporated in the autophagosome. After a lysosome fuses with the autophagosome, the incorporated DALGreen begins to fluoresce as the acidity increases [13], and this was visualized under a confocal fluorescence microscope (Leica TCS SP8; Leica Corporation, Germany) and analyzed with a Guava Easy-Cyte cell analyzer (Hayward, CA, USA).
Cells were also stained with a fluorescein isothiocyanate (FITC)-conjugated Annexin V and propidium iodide (PI) (PromoCell GmbH, Heidelberg, Germany) followed by analysis with a confocal fluorescence microscope and cell analyzer.

Xenoplantation and CPI-613 treatment
The experimental protocol was approved by the Animal Care Committee of Gifu Graduate School of Gifu, Japan (approval No. 27-80). SCID-beige (CB17.Cg-PrkdcscidLystbg-J/CrlCrlj) mice were purchased from Charles River Laboratories Japan (Sizuoka, Japan) and were housed in the Animal facility of the Gifu Graduate School of Gifu, Gifu, Japan. Mice were monitored for signs of distress and were euthanized humanely according to the guidelines for housing mice in animal facility (100% CO 2 inhalation followed by cervical dislocation). We did not observe any signs of distress, including weight loss, appetite loss, or tumor ulceration, in any groups of mice in this study. No animal died during the experiment except when sacrificed at the experimental end point.
Mice were intraperitoneally injected with CPI-613 (25 mg/kg) and chloroquine (50 mg/kg), two days a week for two weeks, when tumor volumes reached approximately 2mm 3 . One week after the last CPI-613 and chloroquine injection, mice were sacrificed to examine their metastatic status. Euthanasia was performed under anesthesia and every effort was made to minimize suffering.

Histological analysis, antibodies, and immunohistochemical staining
The tissue sections were fixed in 10% neutral buffered formalin, embedded in paraffin, and cut into 4 μm thick sections. After deparaffinization and rehydration, the slices were stained with hematoxylin and eosin (H&E).
Immunohistochemical staining was performed according to a previously described procedure [15]. Briefly, antigen retrieval from deparaffinized sections was performed by autoclaving samples for 15 min in 10 mM citrate buffer (pH 6.0). Slides were then incubated for 30 min in 10% normal goat serum and subsequently incubated with antibodies overnight at 4˚C. Immunohistochemical staining was performed using an ImmPRESS Polymerized Reporter Enzyme Staining System (Vector Laboratories Inc., Burlingame, CA, USA).

CPI-613 along with chloroquine induced cell death in HS-MM cells
Cancer cells prevent the toxic buildup of cellular waste products by autophagy [15,16]. Subsequently, we asked whether CPI-613 along with chloroquine, which blocks lysosome function and the degradation of autophagy cargo, impaired the growth of HS-MM cells. Surprisingly, CPI-613, along with chloroquine, induced cell death in HS-MM in a CPI-613 dose-dependent manner (Fig 2C, 2D and 2E). Since a necroptosis inhibitor, necrostatin-1 did not affect cell death, noted as Annexin V negative and PI positive (Fig 2H and 2I), we assume that CPI-613 along with chloroquine induced necrotic cell death.

Orthotropic xenoplanted model of CCS in SCID-beige mice
Notably, lymph node metastasis and distant metastasis were detected at eight weeks in the liver, lungs, heart, and/or peritoneum of all inoculated SCID-beige mice. Peritoneal dissemination lesions were resected, minced, cultured, and transplanted into SCID-beige mice. After four repeated transplantations, accelerated cell growth was observed in the soft tissues of the thighs. Lymphatic metastasis and distant metastasis were detected within four weeks ( Fig 3A). Both histopathological and molecular examination revealed that tumors exhibited a phenotype that was consistent with human CCS comprising polygonal cells containing abundant, clear cytoplasm bordered by thin fibrous septa, which are often observed in human CCS cases (Fig 3B-3D). Tumor cells showed positive staining with anti-S-100, -HMB-45, and -MITF antibodies, consistent with typical CCS observations (Fig 3G-3I).
We also examined the expression of the EWSR1-ATF1 fusion gene in metastatic HS-MM cells. As demonstrated in Fig 3J, RT-PCR demonstrated EWSR1-ATF1 fusion gene expression in both primary injected and metastatic HS-MM cells.
We concluded that the present orthotropic metastatic tumor model was useful in evaluating new and advanced therapies for CCS.

Tumor suppressor effect of CPI-613 in conjunction with chloroquine in vivo
We next attempted to determine whether a combination of CPI-613 and chloroquine had an inhibitory effect on tumor progression in the present murine model. Notably, intraperitoneal injection of CPI-613 and chloroquine not only decreased tumor growth at the primary injection site but also abrogated metastasis (Fig 4). No significant tumor suppressor function was found with CPI-613 or chloroquine alone (Fig 4). Notably, tumor suppression was observed in all five treated mice in this study, in both female (Fig 4) and male mice (S3 Fig).

Discussion
CPI-613 is an analog of lipoic acid, which strongly disrupts mitochondrial metabolism through inhibiting the TCA cycle [6]. Despite the early dogma that cancer cells bypass the TCA cycle and solely use anaerobic glycolysis, recent advances demonstrate that certain cancer cells still utilize the TCA cycle for energy production and macromolecule synthesis [17]. Furthermore, CPI-613 disrupts the redox balance in cancer cell mitochondria [7].
In the present study, we first noted that CPI-613 markedly increased autolysosome formation (Fig 1B, 1C and 1F), but with a small amount of cell death (Fig 2B). Notably, chloroquine impaired autolysosome formation, or the fusion of autophagosomes with lysosomes and induced cell death in CPI-613-treated HS-MM CCS cells (Fig 2C-2E). These findings indicate that insufficient removal and degradation of CPI-613-induced autophagy cargo resulted in cell death in HS-MM CCS cells. Since Annexin V-positive apoptosis was not observed following CPI-613 and chloroquine treatment, the tumor cell death observed may be due to necrosis. In line with this idea, extensive studies to detect apoptosis (i.e., through apoptotic DNA fragmentation) have failed in CPI-613 and chloroquine-treated HS-MM cells (data not shown). It is generally accepted that cells with defects in apoptosis and autophagy fail to tolerate metabolic stress, undergo metabolic catastrophe, and die by necrosis [18,19].
CPI-613 is well-tolerated by recent phase I clinical trials [8,9]. In addition, chloroquine has long been used in the treatment or prevention of malaria, and has recently been applied to Therapeutic potential of CPI-613 in clear cell sarcoma patients with autoimmune disorders, such as rheumatoid arthritis and lupus erythematosus [20]. Clearly, tolerance of both CPI-613 and chloroquine is beneficial for their clinical application. Therefore, we attempted to examine the effect of CPI-613 along with chloroquine on CCS progression in vivo. However, to the best of our knowledge, there are no animal models with high metastatic rates (i.e., animal models that mimic human CCS progression). Indeed, our own previous study demonstrated that BALB/c nude mice xenoplanted with cultured CCS cells showed no metastasis [11].
Next, we developed a murine model, the first animal model for CCS with a high rate of metastasis using SCID-beige mice. SCID-beige is a double-mutant mouse strain with impaired lymphoid development and weak NK (Natural Killer) cell activity [21]. Notably, SCID-beige mice allow for tumor cell growth and the metastasis of cell lines derived from various human cancers, including breast cancer [22], pancreatic cancer [23], and Ewing sarcoma [24]. Although further study is needed, we speculate that impaired NK cell function is linked to the metastatic microenvironment in SCID-beige mice.
As demonstrated in Fig 3, xenoplanted tumor cells exhibited the typical histopathological features of CCS and immunohistochemical profiles (i.e., positive for HMB45, S-100, and MITF). Moreover, tumor cells appeared to harbor the EWS-ATF1 fusion gene, which is a canonical feature of CCS [25], at the primary injection site, ascites, and distant metastatic sites (Fig 3J). The similarity to the pathobiological phenotype and genetic background with similar progression pattern to human CCS may verify the present tumor model as an orthotropic metastatic CCS model.
Using a newly generated orthotropic metastatic model, we examined the efficacy of utilizing CPI-613 against CCS. As demonstrated in Fig 4, CPI-613 along with chloroquine significantly suppressed tumor growth and the metastasis of CCS cells in xeoplanted male SCID-Beige mice.
In conclusion, a combination of CPI-613 and chloroquine may have therapeutic potential for the treatment of patients with CCS, especially before or after surgical excision of the tumor (s) to control the local progression, recurrence, and metastasis.