The study used 9 blood samples from 8 healthy donors (HDs; 4 men and 4 women; mean age, 39 years) and 25 blood samples from 21 patients with HAM (5 men and 16 women; mean age, 69 years). Patients with HAM were diagnosed according to World Health Organization guidelines (Osame, 1990). PBMCs were separated using Pancoll^® density gradient centrifugation (density: 1.077 g/mL; PANBiotech GmbH, Aidenbach, Germany). The separated PBMCs were frozen in cryopreserving fluid (Cell Banker 1; Mitsubishi Chemical Medience Corporation, Tokyo, Japan) and stored in liquid nitrogen. This study was approved by the Bioethics Committee of the St. Marianna University School of Medicine (Approval ID No. 1646). All participants gave their written informed consent.

PBMCs from HDs (n = 4) and patients with HAM (n = 4) were used as starting materials. CD4⁺ T cells were negatively isolated from the PBMCs using a human CD4⁺ T cell isolation kit (Miltenyi Biotec, Bergisch Gladbach, Germany). After removing the MACS column from the separator, the cells were flushed out and used as CD4⁺ T cell-depleted PBMCs. Subsequent manipulations were performed using a previously reported method (Araya et al., 2014). Briefly, total RNA was prepared using TRIZOL (Invitrogen, Carlsbad, CA, United States). Each RNA was amplified and labeled with cyanine 3, using the Agilent Quick Amp Labeling Kit, 1-color (Agilent Technologies, Santa Clara, CA, United States). Cyanine 3-labeled cRNA was fragmented and hybridized to an Agilent Human GE 4x44K Microarray (design ID 014850) loaded with a total of 41,000 probes, excluding control probes. After washing, the microarray was scanned using an Agilent DNA microarray scanner. Intensity values for each feature scanned were quantified using Agilent feature extraction software (version 9.5.3.1) with background subtraction. All data were analyzed using GeneSpring GX software (Agilent Technologies). The box plots were analyzed and visualized using R version 4.2.2. Clustering was performed using R package gplots (distance, Pearson correlation; linkage rule, Ward’s method). Gene Ontology analysis was performed by DAVID Bioinformatics Resources.¹

CD4⁺CCR4⁺ cells were isolated from the PBMCs of HDs (n = 5) and patients with HAM (n = 5) using a previously described method (Araya et al., 2014). Total RNA was extracted from isolated CD4⁺CCR4⁺ cells, and cDNA was generated using ReverTra Ace (TOYOBO, Osaka, Japan). Using the prepared cDNA as a template, the expression levels of EZH2 in CD4⁺CCR4⁺ cells of HDs and patients with HAM were analyzed via quantitative PCR. Primer sets and a probe used to detect EZH2 expression were EZH2#35-F (TGTGGATACTCCTCCAAGGAA), EZH2#35-R (GAGGAGCCGTCCTTTTTCA), and Universal ProbeLibrary #35 (Roche Diagnostics, Rotkreuz, Switzerland). Relative quantification of mRNA was determined with the comparative Ct method using GAPDH as an internal control (Yamano et al., 2009). The following equation was used to determine the relative expression level of the target gene: target gene expression = 2-(Ct[target] - Ct[GAPDH]).

The selective EZH2 inhibitors (GSK126 and tazemetostat) and EZH1/2 dual inhibitors (OR-S1 and valemetostat) were synthesized and provided by Daiichi Sankyo, Co., Ltd. (Tokyo, Japan). Prednisolone, which was used as a positive control, was purchased from LKT Laboratories, Inc. (St. Paul, MN, United States). Dimethylsulfoxide (DMSO) was purchased from Sigma-Aldrich (St. Louis, MO, United States).

For cell proliferation assay, human PBMCs in RPMI 1640 (FUJIFILM Wako Chemicals, Osaka, Japan) supplemented with 10% heat-inactivated fetal bovine serum and 1% penicillin and streptomycin antibiotic solution (FUJIFILM) were cultured at 37°C in 5% CO₂. For the cell viability assay, we used HTLV-1-infected cells lines (HCT-4 and HCT-5) established from cerebrospinal fluid cells of patients with HAM. HCT-4 cells were cultured in RPMI 1640 supplemented with 10% heat-inactivated fetal bovine serum, 1% L-glutamine (FUJIFILM), 1% penicillin and streptomycin antibiotic solution, and 100 U/mL IL-2 (Cell Science & Technology Institute, Inc., Sendai, Japan). HCT-5 cells were cultured in RPMI 1640 supplemented with 20% heat-inactivated fetal bovine serum, 1% L-glutamine, 1% penicillin and streptomycin antibiotic solution, and 200 U/mL IL-2.

To examine the effect of EZH1/2 inhibitors on the viability of HCT-4 and HCT-5 cells over time, the cells were cultured under the following conditions: 3 × 10⁶ HCT-4 or 2.5 × 10⁶ HCT-5 cells were seeded and cultured in 20 mL of culture medium with DMSO, 1 μM OR-S1, or 1 μM valemetostat for 21 days. Cell passages were performed as follows. First, the volume of culture medium containing 3 × 10⁶ HCT-4 or 2.5 × 10⁶ HCT-5 cells in the DMSO-treated groups was calculated. Next, in all groups, the calculated volume of the cell culture medium was transferred to a new flask, and the fresh medium was added up to 20 mL. Finally, DMSO, OR-S1, or valemetostat was added again. Cells were collected from the culture medium 7, 11, 14, and 21 days after the initiation of the culture and used for cell viability assay.

To examine the effect of various concentrations of EZH1/2 inhibitors on the viability of HCT-4 and HCT-5 cells, cell culture was also performed as follows: 4.5 × 10⁵ HCT-4 or 3.75 × 10⁵ HCT-5 cells were seeded in 6-well plates and cultured in 3 mL of culture medium with DMSO, OR-S1 dilution series, or valemetostat dilution series for 14 days, with repeated passages every 3 to 4 days. The passages were performed in the same manner as described above, except for the number of cells. The number of seeded cells was 4.5 × 10⁵ cells for HCT-4 and 3.75 × 10⁵ cells for HCT-5. Cells were collected from the culture medium 14 days after culture initiation for the cell viability assay.

PBMCs from patients with HAM (n = 8) were seeded at 1 × 10⁵ cells/well in 96-well round bottom plates. They were cultured for 7 days in the presence or absence of OR-S1, valemetostat, GSK126, tazemetostat, or prednisolone. During the last 16 h, 1 μCi of ³H-thymidine was added to each well, and then the cells were harvested and counted using a 2450 MicroBeta² Plate Counter (Perkin Elmer, Boston, MA). The assay was performed in triplicate. To verify the effect of the EZH1/2 inhibitors, PBMCs from another eight patients with HAM were also used. Using the average counts of ³H-thymidine incorporation in the DMSO-treated group as 100%, the average of the relative values in each drug-treated group was calculated and defined as the ³H-thymidine incorporation rate (%). The IC50 of each drug was calculated by substituting the data obtained from this experiment into the following equation: IC50 = 10ˆ[LOG(A/B)∗(50-C)/(D-C) + LOG(B)], where A is the higher concentration considering the two values that sandwich 50% of ³H-thymidine incorporation rate, B is the lower concentration considering the same two values, C is the ³H-thymidine incorporation rate (%) determined for B, and D is the ³H-thymidine incorporation rate (%) determined for A.

PBMCs from patients with HAM (n = 8) were seeded at 5 × 10⁵ cells/well in 48-well plates. They were cultured for 12 days in the presence or absence of OR-S1, valemetostat, or prednisolone. Supernatants were collected and stored at −80°C. The cells were then harvested for DNA extraction or flow cytometric analysis. The concentrations of IFN-γ, TNF-α, IL-6, and IL-10 in the culture supernatants were measured with a cytometric bead array kit (BD Biosciences, Franklin Lakes, NJ, United States), using a flow cytometer FACSCantoII (BD Biosciences) according to the manufacturer’s instructions. HTLV-1 proviral loads were measured with ABI Prism 7500 SDS (Applied Biosystems, Carlsbad, CA, United States) using a previously described method (Yamano et al., 2002). HTLV-1 (pX) per 100 cells = (copy number of pX) / ((copy number of β-actin)/2) × 100.

Zombie NIR™ Fixable Viability Kit (BioLegend, San Diego, CA, United States) was used to exclude dead cells. PBMCs were immunostained with a combination of the following fluorescence-conjugated antibodies to cell surface markers: CD3 (UCHT1), CD4 (OKT4), and CD8 (RPA-T8). Cells were fixed with 70% EtOH for 90 min at −20°C, then intracellularly stained with Ki67 antibodies (MOPC-21). The stained cells were analyzed using FACSCantoII (BD Biosciences). Data were processed using FlowJo software (BD biosciences).

HCT-4 and HCT-5 cells were cultured as described in Section 2.5. To measure cell viability, 100 μL of cell suspensions were seeded in 96-well plates, 10 μL of cell counting kit-8 (CCK-8) (Dojindo Laboratories, Kumamoto, Japan) was added to each well, and after 3 h of incubation at 37°C, the absorbance at 450 nm was measured using an iMark microplate reader (Bio-Rad Laboratories, Hercules, CA, United States). Cell viability was calculated using the following formula: cell viability = (experimental optical density value - blank optical density value) / (DMSO-treated control optical density value - blank optical density value) × 100%. The IC50 of each drug in the cell viability assay was calculated using the same method as for the cell proliferation assay.

HCT-4 or HCT-5 cells were cultured for 21 days with 1 μM OR-S1 or 1 μM valemetostat. The harvested cells were stained with PE annexin V and 7-aminoactinomycin D (7-AAD) according to the instructions provided with PE Annexin V Apoptosis Detection Kit I (BD Biosciences) and analyzed with a BD FACSCantoII (BD Biosciences). The flow cytometric data was processed using FlowJo software (BD Biosciences).

The unpaired t-test was used to compare EZH2 mRNA expression in CD4⁺CCR4⁺ cells between HDs and patients with HAM. The Friedman test for repeated measurements was used, followed by the Dunn multiple comparison test, to analyze differences among various drug concentrations. Statistical analysis was performed using GraphPad Prism 7 (GraphPad Software, San Diego, CA, United States). Statistical significance was set at a p-value <0.05.

Microarray analysis revealed common gene expression patterns in CD4⁺ T cells from patients with HAM (Figure 1A). The gene ontology analysis of differentially expressed genes showed that most upregulated genes belonged to the mitotic cell cycle, which is closely related to cell proliferation. In contrast, downregulated genes were involved not only in cytoplasmic translation but also in transcriptional regulation and T-cell differentiation (Figure 1B). Based on the microarray data, the median expression levels of the EZH1 and EZH2 genes in CD4⁺ T cells from patients with HAM were 1.3- and 2.4-fold higher, respectively, than in CD4⁺ T cells from HDs (Figure 1C). Next, EZH2 mRNA expression levels in CD4⁺CCR4⁺ T cells from other patients with HAM (n = 5) were examined using RT-qPCR and found to be 2.6 times higher than to those from HDs (p = 0.001, Figure 1D). Furthermore, we examined our previous microarray data (Accession No. GSE57259) and found that EZH1 and EZH2 in CD4⁺CD25⁺CCR4⁺ T cells from patients with HAM were 0.94- and 3.9-fold higher than those in HDs, respectively.

We evaluated the effects of EZH2 selective inhibitors and EZH1/2 dual inhibitors on the spontaneous proliferation of HAM patients’ PBMCs that could proliferate in the absence of mitogens or exogenous growth factors. Since this phenomenon is known to be inhibited by corticosteroids (Itoyama et al., 1988; Ijichi et al., 1989), prednisolone was used as a positive control. Experimental results demonstrated that EZH2 inhibitors (GSK126 and tazemetostat) inhibited spontaneous PBMC proliferation in patients with HAM (n = 8) in a concentration-dependent manner (Figure 2A). Both drugs significantly inhibited it at 1 μM (p < 0.01 for both). The IC50 values of these drugs were 724.3 nM and 214.2 nM, respectively. EZH1/2 dual inhibitors (OR-S1 and valemetostat) also showed a concentration-dependent inhibition of spontaneous proliferation of PBMCs from the same patients with HAM (Figure 2B). Both drugs significantly inhibited it at 0.1 μM (p < 0.05 for both).

Their IC50 values were 26.5 nM and 19.4 nM, respectively. The inhibitory effect of OR-S1 and valemetostat on the spontaneous proliferation of PBMCs from other patients with HAM (n = 8) was also observed, confirming the reproducibility of the inhibitory effect, and both drugs showed significant inhibition at 0.1 μM (p < 0.05 and p < 0.01, respectively) (Supplementary Figure S1).

EZH1/2 inhibitors and prednisolone reduced the percentage of Ki67⁺ cells in all viable cells on day 7 after culture (Figure 3 and Supplementary Figure S2). Further analysis revealed that all drugs could reduce the percentage of Ki67⁺ cells in CD4⁺ T cells and in CD8⁺ T cells. The pattern of proliferation inhibition of CD4⁺ T cells and CD8⁺ T cells by EZH1/2 inhibitors was different from that by prednisolone. Valemetostat had a higher inhibitory effect than OR-S1 in all three patients (Figure 3 and Supplementary Figure S2).

To evaluate the effect of OR-S1 and valemetostat on cytokine production, we measured the concentrations of the pro-inflammatory cytokines IFN-γ, TNF-α, and IL-6 and the anti-inflammatory cytokine IL-10 in the culture supernatants of PBMCs of patients with HAM (n = 8) on day 12 after culture, in the presence or absence of each drug. Prednisolone at 1 μg/mL, the positive control, showed a tendency to decrease IFN-γ and IL-6 levels, and it significantly decreased TNF-α level (p < 0.01) (Figures 4A–C). OR-S1 and valemetostat hardly changed IFN-γ and TNF-α production. There was a wide variation in the change in IL-6 production by OR-S1, but only 0.1 μM OR-S1 produced a significant change (p < 0.05). OR-S1 and valemetostat increased IL-10 levels in a concentration-dependent manner (Figure 4D). The increase was significant in response to 1 μg/mL prednisolone, 1 μM OR-S1, 100 nM valemetostat, and 1 μM valemetostat (p < 0.01, p < 0.0001, p < 0.05, and p < 0.01, respectively).

To evaluate the effect of OR-S1 and valemetostat on HTLV-1 proviral load, we measured the HTLV-1 proviral loads in PBMCs of patients with HAM (n = 8) on day 12 after culture, in the presence or absence of each drug. Prednisolone did not significantly reduce proviral loads, as previously reported (Figure 4E). OR-S1 and valemetostat showed a tendency to decrease HTLV-1 proviral load at 0.1 μM and significantly decreased it at 1 μM.

To investigate the potential of OR-S1 and valemetostat to kill HTLV-1-infected cells, we used HCT-4 and HCT-5 cells that were HTLV-1-infected cell lines established from cerebrospinal fluid cells from patients with HAM. The viability of HCT-4 and HCT-5 cells decreased over time in the presence of 1 μM OR-S1 or 1 μM valemetostat (Figure 5A) and also decreased in a concentration-dependent manner with OR-S1 or valemetostat (Figure 5B). The IC50 values of OR-S1 and valemetostat for HCT-4 cells were 7.63 nM and 5.92 nM, respectively, and those of HCT-5 cells were 185.4 nM and 90.6 nM, respectively.

To investigate the mechanism by which OR-S1 and valemetostat kill HTLV-1-infected cells, we performed apoptotic cell analysis using annexin V/7-AAD staining. The percentage of annexin V(+)7-AAD(−) HCT-4 cells were 6.1, 21.3, and 21.9% in the DMSO-, OR-S1-and valemetostat-treated groups, respectively (Figure 6A). Similarly, the percentages of annexin V(+)7-AAD(−) HCT-5 cells were 26.0, 42.6, and 60.6%, respectively (Figure 6B). Thus, an increase in early apoptotic cells was observed with EZH1/2 inhibitors.