The antipsychotic drug pimozide is a novel chemotherapeutic agent for adult T cell leukemia


 Patients with adult T cell leukemia (ATL), caused by the human T cell leukemia virus type 1 (HTLV-1), exhibit poor prognosis owing to drug resistance. Pimozide, a dopamine D2 receptor subfamily antagonist and antipsychotic drug, has been shown to exhibit anticancer activity. Herein, we investigated whether pimozide exerts anti-ATL effects and explored the mechanisms underlying these effects. While pimozide inhibited cell growth and survival in HTLV-1-infected T cells, it exerted limited effects on uninfected T cells. The dopamine D2 receptor subfamily mRNA expression levels in HTLV-1-infected T cells were high. Pimozide induced G1 cell cycle arrest concomitant with the upregulation of p21/p27/p53 and the suppression of cyclin D2/E, CDK2/4/6 and c-Myc expression and pRb phosphorylation. Pimozide also induced apoptosis via the activation of caspases and upregulation of pro-apoptotic proteins and downregulation of anti-apoptotic proteins. Additionally, it promoted ROS generation and increased the expression of the ER stress marker ATF4 and the DNA damage-inducible protein GADD45a and the phosphorylation of the DNA damage marker H2AX. Furthermore, pimozide-induced cytotoxicity was partially inhibited by a ROS scavenger and pan-caspase and necroptosis inhibitors, indicating the involvement of caspase-dependent and -independent lethal pathways. The activities of the NF-κB, Akt, STAT3/5 and AP-1 signaling pathways were inhibited via the dephosphorylation of IκBα, IKKα/β, Akt and STAT3/5, in addition to reduced JunB and JunD expression in HTLV-1-infected T cells in response to pimozide treatment. Pimozide also exhibited potent anti-ATL activity in the xenograft mouse model. These findings demonstrated the efficacy of pimozide as a potential therapeutic agent for ATL.


Introduction
Adult T cell leukemia (ATL) is an intractable peripheral T cell neoplasm caused by infection with the retrovirus human T cell leukemia virus type 1 (HTLV-1) [1]. The lifetime risk of developing ATL in HTLV-1 carriers is estimated to be 6 − 7% for men and 2 − 3% for women in Japan [1]. ATL exhibits a diversity in its clinical features such as leukocytosis with increased abnormal lymphocytes, lymphadenopathy, hepatosplenomegaly, skin lesions, hypercalcemia and frequent complication of opportunistic infections

Apoptosis analysis
The cells were treated with pimozide for up to 48 h, followed by permeabilization by incubation with digitonin, following which the cells were labeled with a phycoerythrin-conjugated APO2.7 antibody (1:10; Beckman Coulter, Inc., Marseille, France), as described previously [7]. The percentage of apoptotic cells was quanti ed immediately after staining using an Epics XL Flow Cytometer. In addition, to evaluate the morphological features of the nuclei, the cells were stained with Hoechst 33342 (Dojindo Molecular Technologies, Inc., Kumamoto, Japan) and observed under a DMI6000 microscope (Leica Microsystems, Wetzlar, Germany).

Caspase activity assay
The activity levels of caspase-3, caspase-8 and caspase-9 were quanti ed using Colorimetric Caspase Assay kits (Medical & Biological Laboratories, Co., Nagoya, Japan) in accordance with the manufacturer's instructions. In brief, the cells were lysed in the lysis buffer provided with the kit, and the cell lysates were incubated with the respective caspase-speci c labeled substrates. The chromophore ρ-nitroanilide released upon cleavage from the substrates was quanti ed using a Wallac 1420 Multilabel Counter.
Caspase activity was measured as the ratio between the colorimetric output in the treated sample and that in the control, with the value of the latter set to 1.
Quanti cation of reactive oxygen species (ROS) The generation of intracellular ROS was detected using CellROX, a uorescence probe. After exposure to pimozide at different concentrations for 24 h, the cells were incubated with the CellROX Green reagent (Thermo Fisher Scienti c, Waltham, MA, USA) for 60 min in the dark and washed with phosphatebuffered saline. The changes in CellROX Green-induced uorescence were analyzed using an SH800 Flow Cytometer (Sony Biotechnology Inc., Tokyo, Japan).

Xenograft tumor model
HUT-102 cell suspensions (1 ´ 10 7 /0.2 ml of RPMI-1640 medium) were injected subcutaneously into 5week-old female C.B-17/Icr-severe combined immunode cient (SCID) mice (Japan SLC, Inc., Hamamatsu, Japan) on day 0. The mice were randomly divided into control and treatment groups (n = 9, each). Either pimozide (25 mg/kg) or a solvent (0.5% methylcellulose, Wako Pure Chemical Industries) was administered via oral gavage ve times per week between days 1 and 28. The tumor size was measured weekly using shifting calipers to calculate the tumor volume using the following formula: p/6 ´ h ´ l ´ w [8]. The body weight of each mice was also measured weekly. At the time of sacri ce on day 28, the xenograft tumors and blood samples were collected rapidly. The weight of each tumor was measured, and the sera were stored at −80 °C until they were assayed for human soluble interleukin-2 receptor a (sIL-2Ra). The studies were approved by the Animal Care and Use Committee at the University of the Ryukyus (A2019149).

Measurement of serum sIL-2Ra levels
The serum sIL-2Ra levels were measured in mice treated or untreated with pimozide using enzyme-linked immunosorbent assay (ELISA) for human sIL-2Ra (Diaclone SAS, Besançon, France) in accordance with the manufacturer's instructions.

Statistical analysis
Data are presented as the mean ± standard deviation (SD). Statistical analysis was performed using a Student's t-test or ANOVA along with the Tukey-Kramer test. Differences were considered signi cant at P < 0.05.

HTLV-1-infected T cells were sensitive to pimozide
The treatment of HTLV-1-infected T cells with pimozide at various concentrations for up to 72 h inhibited cell proliferation and survival in a dose-and time-dependent manner, as indicated by the reduction in WST-8 activity (Fig. 1a). As a negative control for ATL, uninfected T cells were also treated with pimozide. There were limited effects exerted on the proliferation and survival of uninfected T cells (Fig. 1b). Since pimozide is an antagonist of the dopamine D2 receptor subfamily (D2, D3 and D4) [6], we analyzed the expression of dopamine receptors on T cells. RT-PCR analysis revealed that the D2, D3 or D4 receptors were not expressed in uninfected T cells, whereas they were expressed in all HTLV-1-infected T cells (Fig.  1c). There ndings indicated that pimozide, which is a potent dopamine D2 receptor subfamily antagonist, inhibits the proliferation and survival of HTLV-1 infected T cells, and the expression of the dopamine D2 receptor subfamily could be associated with the inhibitory effect to an extent.

Effects of pimozide on the cell cycle in HTLV-1-infected T cells
To determine whether pimozide inhibits cell cycle progression to suppress cell proliferation, the effect of pimozide on cell cycle distribution was analyzed using PI staining. After MT-2 and HUT-102 cells were treated with pimozide at the indicated concentrations for 24 h, the number of cells in the G1 phase increased, whereas the number of cells in the S phase decreased compared to the number of untreated cells (Fig. 2a). The western blotting results showed the reduced expression of the cell cycle markers cyclin D2, cyclin E, CDK2, CDK4 and CDK6, along with increased expression of p21, p27 and p53 (Fig. 2b), which is consistent with the G1 arrest observed in the ow cytometric analysis. Pimozide also downregulated c-Myc expression and induced pRb dephosphorylation (Fig. 2b). These results indicated that pimozide hindered cell proliferation in association with the inhibition of G 1 -S progression.

Induction of apoptosis and necroptosis in HTLV-1-infected T cells treated with pimozide
Next, we investigated whether pimozide treatment induced apoptosis in HTLV-1-infected T cells. The microscopic examination of pimozide-treated MT-2 and HUT-102 cell nuclei after Hoechst 33342 staining revealed striking morphological changes, with the cells exhibiting apoptotic characteristics, along with nuclear fragmentation and chromatin condensation (Fig. 3a). The apoptotic effects induced by pimozide on HTLV-1-infected T cells were also analyzed by APO2.7 staining using ow cytometry [7]. Pimozide was observed to induce apoptosis in MT-2, HUT-102 and MT-4 cells (Fig. 3b). Pimozide induced the cleavage of a known caspase-3 substrate, PARP, as well as of caspase-3, caspase-8 and caspase-9 in a dose-dependent manner (Fig. 3c), thus unveiling the role of caspase in pimozide-induced apoptosis.
Furthermore, caspase-3, -8 and -9 activity levels were measured in cells treated with pimozide (Fig. 3d). This experiment also demonstrated the activation of the three caspases (Fig. 3d). To evaluate the role of caspase-dependent cell death, the broad-range caspase inhibitor z-VAD-FMK was used to selectively inhibit the apoptotic pathway. As shown in Fig 4a, HUT-102 and MT-4 cells were pretreated with z-VAD-FMK for 2 h, followed by incubation with pimozide for 24 h. z-VAD-FMK reduced the pimozide-induced inhibition of cell viability in the WST-8 assay partially but signi cantly. These results demonstrated the prevalence of apoptosis while also indicating the involvement of caspase-independent cell death.
Besides apoptosis, another mode of programmed cell death, necroptosis, has also been identi ed [9]. We further evaluated whether necroptosis is required for the induction of cell death by pimozide. Pimozide-induced cell death was reduced in the presence of the necroptosis inhibitor necrostatin-1 (Fig. 4b). These results indicated that apoptosis and necroptosis are involved in pimozidetriggered cell death.
Pimozide induced ROS accumulation ROS accumulation could partially contribute to cell apoptosis and necroptosis [10,11]. We evaluated the effect of pimozide on the ROS levels in MT-2 and HUT-102 cells. Flow cytometry experiments revealed the elevation in ROS production in cells after pimozide treatment (Fig. 5a). To determine whether the generation of ROS induced by pimozide was related to its ability to induce the apoptosis of HTLV-1infected T cells, we analyzed the apoptotic effects induced by pimozide in the presence of the ROS scavenger NAC. As shown in Fig. 5b, NAC impeded the pimozide-mediated induction of apoptosis.

Modulation of apoptotic regulatory protein expression in pimozide-induced apoptosis
The balance between pro-apoptotic and anti-apoptotic proteins eventually determines whether cells will undergo apoptosis or survive. We evaluated whether pimozide induces cell death by modulating the expression of Bcl-2 and IAP family members, which eventually determine the cell response to apoptotic stimuli. As shown in Fig. 5c, pimozide treatment caused the downregulation of Bcl-xL, Mcl-1, c-IAP2, XIAP and survivin; conversely, it induced the expression of the pro-apoptotic protein Bax. These results indicated that Bcl-2 and IAP family proteins may be involved in pimozide-induced apoptosis.
When produced in excess, ROS can induce extensive damage in DNA, proteins and lipids [10]. To evaluate whether the cytotoxic response induced by pimozide is mediated via DNA damage and/or whether the increased ROS levels caused oxidative damage to DNA, the levels of phosphorylated H2AX, which is a DNA damage marker, and the expression of the DNA damage-inducible protein GADD45a were assessed. As postulated, the expression of phosphorylated H2AX and GADD45a was upregulated in cells treated with pimozide (Fig. 5c).
The proper functioning of the endoplasmic reticulum (ER) is essential for most cellular activities as well as for survival. ER stress leads to mitochondrial dysfunction and apoptosis [12]. ROS is known to induce ER stress-dependent apoptosis [10]. The involvement of ER stress in pimozide-induced cell apoptosis was investigated. The levels of ATF4, an ER stress marker, increased in response to pimozide treatment (Fig. 5c). The results indicated that ER stress signaling is one of the potential pathways involved in the induction of apoptosis.
Pimozide suppresses the activities of STAT3/5, NF-κB, Akt and AP-1 STAT3/5 act as signaling mediators involved in cell growth and survival. Both are constitutively activated in ATL, and STAT activation is associated with cell cycle progression [13]. Çyclin D2, c-Myc, Bcl-xL and Mcl-1 are regulated by STAT3/5 [14]. Therefore, agents that target STAT3/5 could be useful for the treatment of ATL. Since pimozide has been reported to inhibit STAT3/5 activation [15,16], we assessed the effect exerted by pimozide on STAT3/5 activity in HTLV-1 infected T cells, in which both proteins are activated. Pimozide was observed to inhibit STAT3/5 phosphorylation (Fig. 6a). The NF-κB, Akt and AP-1 pathways, which are common cell survival pathways, are also constitutively activated in HTLV-1-infected T cells [17,18]. As shown in Fig. 6a, pimozide suppressed the phosphorylation of IkBa and its upstream kinases IKKa/b, whereas it enhanced the expression of IkBa. In addition, pimozide treatment decreased the levels of Akt phosphorylation (Fig. 6a). AP-1 is a dimeric transcription factor composed of proteins belonging to the Jun, Fos and ATF protein families [18]. JunB and JunD were expressed at high levels and mediated AP-1 DNA-binding activity in HTLV-1-infected T cells [18−20]. JunB and JunD expression was also reduced upon treatment with pimozide (Fig. 6a). Therefore, in addition to STAT3/5, the NF-κB, Akt and AP-1 pathways were suppressed via the dephosphorylation of IKKα/β, IκBα and Akt as well as via the inhibition of JunB and JunD expression, in response to pimozide treatment in HTLV-1-infected T cells.

Pimozide inhibited the growth of ATL xenografts in SCID mice
Lastly, we evaluated the antitumor effects of pimozide on the growth of HTLV-1-infected T cell HUT-102 xenografted tumors in vivo. The tumor volumes in mice from the untreated control group increased, whereas the tumor volumes reduced by 32%, a signi cant reduction (P < 0.05), in mice from the treated group (Fig. 6b). The tumor weight decreased in mice treated with pimozide (Fig. 6d). Notably, pimozide treatment was well-tolerated, and no signi cant effects were exerted on body weight (Fig. 6c).
Furthermore, the serum biomarker levels were assessed using ELISA. Pimozide treatment reduced the levels of sIL-2Ra; however, the differences were not statistically signi cant (Fig. 6e). These results implied that pimozide exhibits anti-ATL activities in vivo without generating signi cant side effects.

Discussion
Several epidemiological studies and clinical data have shown a lower rate of cancer incidence among patients with schizophrenia, and medication used to treat schizophrenia may be the possible mediators of this effect [4]. While pimozide is a neuroleptic drug with manageable side effects in clinical use, it exhibits therapeutic effects in certain types of cancer [6]. Pimozide has also shown e cacy in the treatment of metastatic melanoma and has been shown to be well-tolerated in clinical trials [21,22]. In addition, no hematological toxicity ascribable to pimozide has been reported [16]. In this study, we investigated the cytotoxic effects of pimozide and showed that pimozide could be used as a novel anti-ATL drug, based on its ability to suppress cell proliferation and increase cell death in selectively HTLV-1infected T cells in vitro and to inhibit tumor growth in xenografts in vivo.
Pimozide inhibited cell proliferation of HTLV-1-infected T cells as a result of G1 cell cycle arrest and downregulated expression of cyclin D2, cyclin E, CDK2, CDK4, CDK6 and c-Myc along with increased expression of p21, p27 and p53. Cyclin D2-CDK4/6 and cyclin E-CDK2 induce pRb phosphorylation, which initiates DNA synthesis. p21 and p27 inhibit cyclin-CDK complexes. c-Myc induces the expression of cyclin D2, cyclin E, CDK2, CDK4 and CDK6. Moreover, c-Myc represses p21 and p27 expression [23]. The pimozide-induced reduction in cyclin and CDK levels and increase in p21 and p27 levels may be partially attributed to the downregulation of c-Myc, which results in pRb dephosphorylation.
Pimozide promoted the apoptosis of HTLV-1-infected T cells and increased ROS generation. In several studies, ROS generation has been associated with apoptosis and necroptosis induction [10,11]. NAC, a ROS scavenger, suppressed pimozide-induced apoptosis, indicating that the anti-ATL effect of pimozide was linked to ROS generation. Moreover, pimozide cytotoxicity was impacted by pan-caspase or necroptosis inhibition. These results indicated that apoptosis and necroptosis are involved in the pimozide-induced elimination of HTLV-1-infected T cells.
At elevated levels, ROS can cause extensive DNA damage [10]. Pimozide also exerts genotoxic effects on DNA by promoting DNA double-strand breaks, which corresponds to an increase in the phosphorylation of histone H2AX [24]. ROS also activate p53, which regulates multiple signaling pathways triggered by cellular stress, inducing either cell cycle arrest or apoptosis [10,25,26]. Pimozide was shown to induce p53 expression. Upon induction, p53 in uences downstream genes, including p21 and GADD45α, to promote cell cycle arrest and DNA damage. p53 also regulates apoptosis by downregulating pro-survival proteins, such as members of the Bcl-2 family and the IAP family, and by upregulating pro-apoptotic proteins [25,26]. Previous studies have shown that ROS plays an essential role in ER stress induced by various antitumor agents [27]. In addition, p53 also triggers ER stress [28]. Pimozide treatment induced the expression of the ER stress marker ATF4. These results implied that pro-apoptotic ER stress also contributes to cell death induced by pimozide. Pimozide induces cell cycle arrest and cell death via a ROS-mediated p53-regulated stress pathway.
Multiple signaling pathways, including STATs, NF-κB, Akt and AP-1 signaling, are essential for the pathogenesis of ATL, as these pathways directly or indirectly regulate the expression of genes that induce or maintain the proliferation and survival of ATL cells [13,17,18]. Therefore, these signaling pathways are attractive targets for ATL therapy. Previous studies have shown that pimozide exerts anti-neoplastic effects by suppressing STAT3/5 activity [15,16]. We con rmed that the anti-ATL effects of pimozide were related to STAT3/5 suppression, as reported previously [15,16]. Pimozide was also shown to act as a potent anti-ATL agent by inhibiting the NF-κB, Akt and AP-1 signaling pathways. Pimozide impacted the NF-κB and Akt signaling pathways by inhibiting IκBα, IKKα/β and Akt phosphorylation. Furthermore, pimozide suppressed AP-1 signaling by inhibiting JunB and JunD expression. We also observed the effects exerted by pimozide on NF-κB, AP-1, Akt and STAT3/5 signaling by measuring protein expression from the target genes. Pimozide treatment of HTLV-1-infected T cells downregulated the expression of their target proteins, including cyclin D2/E, CDK2/4/6, c-Myc, Bcl-xL, Mcl-1, c-IAP2, XIAP and survivin, which indicates the inactivation of these pathways [14, 29 − 36]. Therefore, we propose that pimozide suppresses the proliferation of HTLV-1-infected T cells by inhibiting the NF-κB, AP-1, Akt and STAT3/5 pathways. The results of the present study indicated that the concurrent blockade of these pathways may be effective in suppressing cell proliferation.
Notably, the cytotoxic effects of pimozide appeared to be speci c to HTLV-1-infected T cells, as the proliferation and survival of uninfected T cells remained unaffected by pimozide administration at concentrations below 5 µM. Since pimozide is an antagonist of the dopamine D2 receptor subfamily, we evaluated the expression of these receptors in HTLV-1-infected T cells and found that the levels of the dopamine D2 receptor subfamily mRNA were elevated compared to the corresponding levels in uninfected T cells. In future, further studies should be conducted to evaluate the association between the expression of these receptors and the anti-ATL effects exerted by pimozide. Although studies using fresh leukemic cells obtained from ATL patients are required to evaluate pimozide as an anti-ATL agent, pimozide treatment was found to be well-tolerated in a murine ATL model, with no signi cant effects exerted on body weight. In conclusion, the present study provides compelling evidence that pimozide demonstrates anti-ATL activity via multiple molecular mechanisms, and could therefore be considered a novel candidate for ATL treatment.

Consent for publication
All authors consent to the publication pf this study.

Availability of data and materials
The data and materials used in this study are available from the corresponding author on reasonable request.

Competing interests
No potential competing interest is reported by the authors.

Figure 5
ROS triggers apoptosis by altering the expression of apoptosis-related proteins in pimozide-treated cells.
a. ROS production in pimozide-treated cells. Cells were treated without (red) or with 5 μM (yellow) or 10 μM (blue) pimozide for 24 h. ROS levels were quanti ed using CellROX staining and ow cytometry.
Representative images from one of three experiments are shown (top panels). Analysis of ROS production intensity (bottom panels). b. Cells were pretreated with NAC at the indicated concentrations for 2 h. After pretreatment, pimozide was added at the indicated concentrations and time intervals. Cells were stained with APO2.7 and analyzed using ow cytometry. Data are presented as the mean ± SD from triplicate cultures. *P < 0.001 compared to group treated only with pimozide. c. Expression of apoptosisrelated proteins in pimozide-treated MT-2 cells. Cells were treated with pimozide at the indicated concentrations for 24 h and subjected to immunoblotting analysis. Actin was used as a loading control.

Figure 6
Pimozide affected STAT3/5, NF-κB, Akt and AP-1 activation and suppressed tumor growth in the murine xenograft model. a. Pimozide affected STAT3/5, NF-κB, Akt and AP-1 activation. MT-2 cells were treated with pimozide at the indicated concentrations for 24 h. Cell extracts were prepared to evaluate STAT3/5, NF-kB, Akt and AP-1 activation using western blot analysis. Actin was used as a loading control. b. Effect of pimozide on tumor growth in severe combined immunode cient mice. The xenograft mice were treated with or without pimozide. Tumor size was measured weekly to calculate the tumor volume. c. Changes in the body weight of mice during the experimental period. Tumor weight (d) and serum concentrations of sIL-2Rα (e) in mice. Data are presented as the mean ± SD (n = 9). *P < 0.05 compared to controls.