HTLV-1 Viral Factor HBZ Induces CCR4 to Promote T-cell Migration and Proliferation

Human T-cell leukemia virus type 1 (HTLV-1) is a retrovirus that mainly infects CD4⁺ T cells. HTLV-1 causes adult T-cell leukemia (ATL) and several chronic inflammatory diseases, including HTLV-1–associated myelopathy/tropic spastic paraparesis (HAM/TSP), alveolitis, and uveitis, in a small population of infected individuals (1). C-C chemokine receptor type-4 (CCR4) is known as a marker of type-2 T helper (Th2) cells, regulatory T (Treg) cells, and skin-homing T cells (2). C-C chemokine ligand (CCL)-17 and CCL-22 are known as CCR4 ligands (3). Treg cells and skin-homing T cells migrate to peripheral tissues using CCR4 and in a ligand dependent manner (4, 5).

ATL cells have been reported to express CCR4 (6). Furthermore, HTLV-1–infected CD4⁺ and CD8⁺ T cells also express CCR4, indicating a close relationship between HTLV-1 infection and CCR4 expression (7, 8). As a mechanism of upregulated CCR4 expression, it has been reported that increased Fra2 and JunD expression activates the promoter of CCR4 (9). Recently, humanized anti-CCR4 antibody (mogamulizumab) has been approved for the treatment of ATL and cutaneous T-cell lymphoma (10–12). Recent studies have reported that gain-of-function mutations of the CCR4 gene are frequently found in ATL cases, and these mutations promote cell proliferation (13, 14). However, enhanced proliferation of cells with wild-type CCR4 was not observed in that study. ATL cells tend to infiltrate into skin and form skin lesions in which they proliferate (15). It is possible that CCR4 on ATL cells is implicated in the migration and proliferation of these cells.

HTLV-1 encodes several regulatory and accessory genes in its genome (1). Among them, the tax and HTLV-1 bZIP factor (HBZ) genes play important roles in viral replication and the proliferation of infected cells (16). The HBZ gene is consistently expressed in all HTLV-1–infected individuals (17). In HBZ transgenic (HBZ-Tg) mice that express HBZ only in CD4⁺ T cells, the populations of effector/memory T cells and Foxp3⁺ T cells increased (18). We reported that HBZ enhances transcription of the Foxp3 gene by activating the TGF-β/Smad pathway (19). Foxp3 expression is unstable in HBZ-Tg mice, resulting in increased IFNγ production in CD4⁺ T cells (20). Because loss of IFNγ remarkably retards development of dermatitis, IFNγ is implicated in inflammation caused by HBZ (21). Thus, HBZ is closely linked with the immunophenotype of HTLV-1–infected cells and with HTLV-1 pathogenesis.

In this study, we found that HBZ induces CCR4 expression in T cells by enhancing GATA3 expression. Furthermore, proliferation of HBZ-expressing CD4⁺ T cells was promoted in the presence of CCR4 ligands and an integrin signal via E-cadherin. These results show that the HBZ–CCR4 axis plays important roles in the infiltration and proliferation of HTLV-1–infected cells and ATL cells.

Experiments using clinical samples from ATL patients were conducted according to the principles expressed in the Declaration of Helsinki, and approved by the Institutional Review Board of Kyoto University (the approval number G310). ATL patients provided written informed consent for the collection of samples and subsequent analysis. All experiments using live animals and clinical samples were conducted in accordance with guidelines and regulations of Kyoto University. Mouse experiments were permitted by the Institutional Animal Research Committee of Kyoto University (the approval numbers are D13-02, D14-02, and D15-02).

C57BL/6 mice were purchased from CLEA Japan (Tokyo). C57BL/6 transgenic mice expressing the spliced form of the HBZ gene under the control of the CD4 promoter/enhancer/silencer have been described previously (22). All mice (6–14 weeks of age) used in this study were maintained in an SPF facility for breeding of mice. The animals were handled according to protocols approved by Kyoto University.

Suspension cell line and primary mouse primary cells were cultured in RPMI 1640 supplemented with 10% FBS and antibiotics. IL2 and 2-Mercaptoethanol were supplemented for culture of Kit225 and mouse cells, respectively. 293T and Plate-E were cultured in DMEM supplemented with 10% FBS and antibiotics. Kit225 was obtained from Dr. Toshiyuki Hori (Ritsumeikan University, Kyoto, Japan; ref. 23), and ED was provided by Dr. Michiyuki Maeda (Kyoto University, Kyoto, Japan; ref. 24). MT-1 and MT-2 cells were gifts from Dr. Isao Miyoshi (Kochi University, Kochi, Japan). HEK293T cells were provided by the ATCC. All cell lines were obtained from 1999 to 2015 and passaged for less than 6 months after receipt/resuscitation. Peripheral blood mononuclear cells (PBMC) were separated from whole blood of human using Ficoll (GE Healthcare). Flow cytometric assay was carried out using FACS Verse with Suite Software (BD Pharmingen). Data were analyzed by FlowJo software (Treestar). HBZ transfectants (EL4 cells) were selected in presence of G418 (200 μg/mL).

This experiment was conducted according to the principles expressed in the Declaration of Helsinki and approved by the Institutional Review Board of Kyoto University (Approval number: G310). PBMCs of ATL patients were separated from whole blood using Ficoll (GE Healthcare) according to the manufacturer's protocol. Skin samples from ATL patients were fixed in 10% formalin in phosphate buffer and embedded in paraffin. Hematoxylin and eosin staining was performed according to standard protocols. For immunohistochemistry, cell surface and intracellular antigens were detected using anti-human CD103 (Abcam: EPR4166(2)), anti-human CCR4 (BD bioscience: 1G1), and anti-human Ki67 (DAKO: MIB-1) as primary antibodies and ChemMate EnVision (DAKO) for visualization. Images were captured using a Provis AX80 microscope (Olympus) equipped with an OLYMPUS DP70 digital camera and detected using a DP manager system (Olympus).

Air pouches were raised in the dorsal skin by two serial injections of air. Seven days after the first injection, 10 μg of lipopolysaccharide (LPS) or homogenized spleen supernatant or a mixture of CCR4 ligands (0.5 μg of CCL17 and 0.5 μg of CCL22) in 1 mL D-PBS (-) (Nakarai) were inoculated into the pouch. After 24 hours, migrated cells were collected with 1 mL PBS with 5mM EDTA. Viable cells were counted by the trypan blue exclusion method and stained with fluorescent-conjugated monoclonal antibodies for cell surface markers and intracellular antigens. To prepare homogenized spleen supernatant, the whole spleen of a wild-type mouse (6–8 weeks of age) was frozen in 2 mL of D-PBS (-) and homogenized. After centrifugation, the supernatant was diluted 5-fold with D-PBS (-), and 1 mL of the diluted supernatant was inoculated into the air pouch. The CCR4 antagonist, C 021 dihydrochloride (C021), was purchased from Tocris Bioscience (25). Note that 0.5 mg of C021 was suspended with 0.2% Tween20-PBS and inoculated twice subcutaneously (not into the air pouch).

Splenocytes were obtained from HBZ-Tg mice (14 weeks old). A total of 5 × 10⁵ splenocytes were treated with CCR4 antagonist (1 μg/mL) for 20 hours. Ki-67 and phosphorylated Akt in CD4⁺ T cells were measured using flow cytometry.

C021 was dissolved in sterile PBS. Mice were anesthetized with 2,2,2-Tribromoethanol (Sigma-Aldrich) and inoculated subcutaneously with 0.5 mg of C021 every 2 days for 4 weeks. Inoculation sites were changed every time.

E-cadherin was diluted to 5 μg/mL in 1 mmol/L CaCl/PBS and used to coat a 48-well plate for 24 hours at 4°C. A total of 5 × 10⁵–10⁶ splenocytes were treated with Phytohaemagglutinin (PHA; 5 μg/mL) for 6 hours in the coated plate. CCL17 and CCL22 were added at 0.2 μg/mL each. Cells were stained using the BrdU Flow Kit (BD PharMingen), and cell proliferation was measured by flow cytometry. Dead cells were excluded by flow cytometry using forward and side scatter and the LIVE/DEAD Fixable Dead Cell Stain Kit (Invitrogen).

Nucleotides +993 to −119 of the CCR4 promoter region were amplified by PCR using human genomic DNA as a template, and cloned into pGL4.22 (Promega). A total of 1 to 2 × 10⁵ cells of an ATL cell line (TL-Om1), an HTLV-1–transformed cell line (MT-2), or an HTLV-1–unifected cell line (Kit225) were cotransfected using Lipofectamine LTX or the Neon system (Invitrogen) with a reporter plasmid, a Renilla luciferase control vector (pRL-TK), and the HBZ expression plasmid sHBZ (pcDNA3.1(-)HBZ myc-His; ref. 26). Cells were lysed 48 hours after transfection, and Firefly and Renilla luciferase activities were then measured using the Pickagene Kit (Toyo Inki). Relative luciferase activity was calculated as the ratio of Firefly to Renilla luciferase activities. CCR4 promoter mutants were generated by the deletion of all three GATA3 binding sites in the relevant region (−434/−425, −406/−397 and −285/−276).

Statistical analyses were performed using the Student unpaired t test.

We have reported that HBZ-Tg mice that specifically express the HBZ gene in CD4⁺ T cells develop T-cell lymphoma and inflammatory diseases with infiltration of T cells into the skin, lung, and intestine (18, 20). As shown in Fig. 1A, CD4⁺ T cells of HBZ-Tg mice express high levels of CCR4. In particular, its expression was higher on Foxp3⁺CD4⁺ T cells of HBZ-Tg mice compared with Foxp3⁻CD4⁺ T cells. To analyze whether HBZ induces CCR4 expression, we transduced HBZ-expressing retrovirus or lentivirus into primary mouse T cells or primary human T cells, respectively. HBZ increased CCR4 expression on both mouse and human CD4 T cells (Fig. 1B). We previously reported that HBZ induced Foxp3 expression through activation of the TGF-β/Smad pathway (19). Because human effector Treg cells are known to express CCR4 (27), we surmised that Foxp3 induction by HBZ might be the cause of CCR4 expression. Therefore, we analyzed the CCR4 expression of the Foxp3⁺ and Foxp3⁻ populations in the HBZ-transduced CD4 T cells, and found that HBZ induced CCR4 expression regardless of Foxp3 expression (Fig. 1C), suggesting that increased CCR4 expression is not an indirect effect of Foxp3 induction.

We have reported that not only HBZ protein but also HBZ RNA modulates the transcription of cellular genes (28), and both HBZ protein and RNA upregulate the expression of T-cell immunoreceptor with Immunoglobulin and ITIM Domain (TIGIT) on T cells (29). To study whether HBZ RNA and protein induce CCR4 expression on CD4⁺ T cells, we transduced recombinant retrovirus expressing GFP and either wild-type or mutant HBZ into mouse-activated CD4⁺ T cells (Fig. 1D). In the TTG mutant, the start codon (ATG) is replaced by TTG, so the TTG mutant does not generate a protein. The SM mutant contains silent mutations in the entire coding region of HBZ. Therefore, SM mutant encodes the same protein while its RNA sequence and secondary structure are altered. We found that both the TTG and SM mutants increased CCR4 expression on T cells (Fig. 1D), indicating that both HBZ RNA and protein are responsible for CCR4 expression.

To identify the region responsible for enhanced CCR4 expression, we transduced recombinant retroviruses expressing deletion mutants of the HBZ protein. Deletion of the activation domain severely impaired enhancement of CCR4 expression (Fig. 1E). For HBZ RNA, we used three mutants that contained silent mutations in different regions. The mutant with silent mutations in the first one third (1–207) of coding region could not increase CCR4 expression (Fig. 1F). This result is consistent with reports that this region of RNA is responsible for various activities of HBZ RNA (22, 28).

To further analyze whether HBZ indeed induces CCR4 expression in ATL cells, we suppressed HBZ expression using shRNA. GATA3 is known to be associated with CCR4 expression in Th2 cells. Furthermore, ectopic expression of GATA3 in Th1 cells induces CCR4 on their surfaces (30). We found that HBZ knockdown decreased CCR4 and GATA3 transcripts in the ED and MT-1 cell lines (Fig. 2A and B; Supplementary Materials and Methods), indicating a link between HBZ, GATA3, and CCR4. Next, we transduced the vector expressing HBZ into a mouse T-cell line, EL4. Expression of CCR4 and GATA3 was upregulated in HBZ-expressing EL4 transfectants (Fig. 2C). Furthermore, GATA3 transcripts were also increased in HBZ-transduced mouse primary CD4⁺ T cells (Fig. 2D) and ATL cells (Fig. 2E).

To investigate whether HBZ-GATA3–mediated CCR4 expression is observed in ATL cells, we analyzed CCR4, GATA3, and HBZ transcripts in ATL samples. We quantified spliced HBZ (sHBZ) and unspliced HBZ (usHBZ) mRNA in 14 ATL cases (Supplementary Fig. S1; Supplementary Materials and Methods). We also quantified CCR4 and GATA3 transcripts using quantitative PCR and found that there were no significant correlations between HBZ and CCR4 or between HBZ and GATA3 (Supplementary Fig. S2A and S2B). This finding suggests that other factors besides HBZ also influence the expression of CCR4 and GATA3.

Because both HBZ RNA and protein increase CCR4 expression, we studied the effect of HBZ RNA and protein on GATA3 transcription. As shown in Fig. 2F, both HBZ RNA and protein enhanced GATA3 transcription. These findings suggest that increased CCR4 expression is partially due to HBZ RNA and protein-mediated GATA3 expression.

To investigate the mechanism of HBZ-induced CCR4 expression in more detail, we cloned the human CCR4 promoter region (−993/+119) and inserted it into a luciferase reporter plasmid (Fig. 3A). We first checked whether transcription from the CCR4 promoter is activated in ATL cell lines. When an ATL cell line (TL-Om1) and an HTLV-1–transformed cell line (MT-2) were transfected with the CCR4 promoter plasmid, the promoter activity was activated in TL-Om1 and MT-2 more than in a human T-cell line, Kit225 (Fig. 3B). When we transduced the HBZ expression vector and reporter plasmid DNA into Kit225 cells, HBZ activated transcription from the CCR4 promoter (Fig. 3C). Next, we determined which region of the CCR4 promoter is critical for HBZ-induced transcription. Analysis of deletion mutants shows that the −458/−175 region is required for activation of transcription from the CCR4 promoter (Fig. 3D). This region contains three GATA3 binding sites (Fig. 3A). We next generated a reporter plasmid that lacked all three GATA binding sites. Deletion of these three GATA3 binding sites strongly inhibited the enhancing activity of HBZ on transcription from the CCR4 promoter (Fig. 3E). These data indicate that HBZ increases transcription from the CCR4 promoter via GATA3.

We have reported that CD4⁺ T cells of HBZ-Tg mice infiltrate into skin, lung, and intestines, and induce inflammation (20). To analyze the infiltrative potency of HBZ-expressing T cells, we utilized a mouse air pouch model (31). An air pouch was formed in the dorsal skin of HBZ-Tg mice, and cell migration was induced by inoculation of LPS (to simulate inflammation due to pathogens) or homogenized spleen supernatant (to simulate inflammation due to necrotic cells) into the pouch (Fig. 4A). CD4⁺ T cells of HBZ-Tg mice preferentially migrated into the pouch compared with those of non-Tg mice when either type of inflammation was induced (Fig. 4B). Next, we inoculated CCR4 ligands (CCL17 and CCL22) into the pouch to analyze CCR4-dependent migration (Fig. 4C). When CCL17 and CCL22 were present, the percentage of CD4⁺ T cells in the pouch of HBZ-Tg mice was strikingly increased compared with that of non-Tg mice (Fig. 4D). However, CD8⁺ and CD11b⁺ cells did not increase in the air pouch, suggesting the involvement of CCR4 specifically in CD4⁺ T-cell migration.

A recent study reported that proliferation of ATL cells with a gain-of-function mutation of CCR4 is promoted by stimulation with its ligand (13). However, it remains unknown whether wild-type CCR4 promotes the proliferation of expressing T cells. To explore the role of CCR4 in the proliferation of CD4⁺ T cells, we stained Ki67 as a proliferation marker and found that Ki67⁺CD4⁺ T cells increased in HBZ-Tg mice (Fig. 4E). The CCR4⁺ expression level was quite high in splenic CD4⁺ Ki67⁺ T cells of HBZ-Tg mice compared with either the Ki67⁻ population or non-Tg mice (Fig. 4F), suggesting that CCR4 is involved in cell proliferation.

It has been reported that skin tropic Treg cells are CD103⁺ and CCR4⁺ (32). Furthermore, we have also reported that CD4⁺ T cells in HBZ-Tg mice express CD103 (18). As well as CCR4, CD103 is highly expressed in splenic Ki67⁺ CD4⁺ T cells of HBZ-Tg mice compared with Ki67⁻CD4⁺ T cells (Fig. 4G). CD103, also called integrin αE, forms a heterodimer with integrin β7, a main ligand of which is E-cadherin. E-cadherin is an adhesion molecule found on epithelial cells. Previous studies have reported that integrin on the cell surface positively regulates cell proliferation via binding of extracellular matrix (ECM) molecules as well as cell adhesion (33–35). The finding that expression of both CD103 and CCR4 is increased on Ki67⁺CD4⁺ T cells of HBZ-Tg mice suggests that these molecules may be associated with cell proliferation. To analyze the effects of CD103 and CCR4 on cell proliferation, we cultured splenocytes of HBZ-Tg mice in the presence of CCR4 ligands (CCL17 and 22) and/or CD103 ligand (E-cadherin) and then measured BrdUrd incorporation in CD4⁺ T cells. Although either CCR4 ligands or CD103 ligand stimulation had no effect on BrdUrd incorporation, costimulation by CCL17/22 and E-cadherin augmented the proliferation of PHA-stimulated CD4⁺ T cells of HBZ-Tg mice (Fig. 4H). In addition, the Ki67 expression level in CD4⁺ T cells was also increased by the CCL17/22 and E-cadherin stimulation (Fig. 4I). These results indicate that CCR4 and CD103 are associated with the proliferation of expressing T cells. CD103 expression is increased on CD4⁺ T cells of HBZ-Tg mice. However, because HBZ does not directly induce CD103 expression in mouse primary CD4⁺ T cells (Supplementary Fig. S3), CD103 induction appears to be an indirect effect of HBZ.

In the air pouch model, the migration of CD4⁺ T cells in HBZ-Tg mice was enhanced (Fig. 4B). To further clarify the role of CCR4-dependent migration, we utilized a CCR4 antagonist, C 021 dihydrochloride (C021; ref. 25). C021 was administrated twice subcutaneously to HBZ-Tg mice undergoing the air pouch experiment (Fig. 5A), and then cells in the air pouch were analyzed. The numbers of total infiltrated cells and CD4⁺ T cells were suppressed compared with nontreated HBZ-Tg mice (Fig. 5B and C). In Fig. 4H, the proliferation of CD4⁺ T cells was augmented via chemokine and integrin signals. In accordance with this finding, activation (CD69, CD25, and CD154) and proliferation (Ki67 and PCNA) of CD4⁺ T cells that migrated into the air pouch were also suppressed by the C021 treatment (Fig. 5D and E), indicating that the CCR4-mediated signal enhances migration, cell activation, and proliferation of HBZ-Tg CD4⁺ T cells in vivo.

To further investigate CCR4-dependent proliferation, we repeatedly administered the CCR4 antagonist to HBZ-Tg mice. After HBZ-Tg mice had been treated with the antagonist for 4 weeks, we measured Ki67 expression in HBZ-Tg splenocytes (Fig. 6A). The CCR4 antagonist treatment suppressed the expression level of Ki67 in CD4⁺, but not CD8⁺, splenocytes of HBZ-Tg mice (Fig. 6A). Among the effector molecules of chemokine receptor signaling (36), the antagonist treatment inhibited only Akt phosphorylation in CD4⁺ T cells of HBZ-Tg mice (Fig. 6B). Previous studies reported that Akt signaling is involved in growth, apoptosis, and metastasis in cancer cells (37, 38). We found that the inhibition of Akt phosphorylation led to upregulation of p53 and p21 and downregulation of NFκB1 and NFκB2 (Fig. 6C). When CD4⁺ T cells were treated with the CCR4 antagonist in vitro for 20 hours, phosphorylated Akt and Ki67 were decreased (Fig. 6D), confirming that CCR4-medaited signaling is associated with T-cell activation. These results suggest that the Akt pathway activated by CCR4 is linked with proliferation.

This study suggests that CCR4 and CD103 expression is associated with the migration and proliferation of ATL cells. Therefore, we analyzed CCR4 and CD103 expression in ATL cases. Transcription of the CD103 gene was upregulated in some ATL cases (Supplementary Fig. S4). Using four ATL samples (ATL #3, #6, #15, and #16 in Supplementary Fig. S4) in which CD103 transcripts were increased, we confirmed CD103 and CCR4 expression on CD4⁺ T cells by flow cytometry (Fig. 7A). Because skin lesions are frequently observed in ATL patients, it is possible that ATL cells with CD103 and CCR4 expression preferentially infiltrate into skin and proliferate. Among skin lesions of 5 ATL patients, ATL cells frequently expressed CD103 and CCR4 (Supplementary Table S1; Supplementary Fig. S5). Percentages of Ki67-positive cells were also high in these skin lesions. It is noteworthy that CD103- and Ki67-positive cells accumulated in Pautrier's microabscess, suggesting a close linkage between proliferation and CD103 expression (Fig. 7B).

In most ATL cases, ATL cells express CCR4 (6). Furthermore, HTLV-1–infected T cells are CCR4⁺ in both the CD4⁺ and CD8⁺ subpopulations (7, 8). These findings suggest that CCR4 expression is caused by HTLV-1 infection, but not by leukemogenesis. Among the viral genes encoded by HTLV-1, only HBZ is constantly expressed in all ATL cells and HTLV-1–infected cells, suggesting a link between CCR4 expression and HBZ (1). Indeed, this study shows that HBZ RNA and protein both increase CCR4 expression through upregulated GATA3 expression. HBZ-induced CCR4 expression is implicated in the migration of T cells. Importantly, CCR4 and CD103 expression is also linked with the proliferation of T cells. Pathologic analyses of skin lesions in ATL patients showed that a high percentage of cells were positive for CD103 and Ki67. In sum, CCR4 and CD103 expression on ATL cells appears to be involved in their migration and proliferation. Because skin lesions are observed in approximately 70% of ATL patients (39), CCR4 and CD103 are important in pathogenesis of ATL.

A previous study did not find any direct effects of HBZ on the activity of the CCR4 promoter (9). In that study, a relatively short promoter region (-151/+25 in Fig. 3A: 176 bp), which contains only one GATA binding site, was used for the reporter assay (Fig. 3A). The reporter plasmid used for our study contains a longer promoter region (1,112 bp) that includes additional GATA binding sites. Indeed, deletion of three GATA3 biding sites severely impaired the enhancement of promoter activity by HBZ (Fig. 3E), indicating that HBZ-mediated GATA3 upregulation is critical to induce the CCR4 gene transcription through these GATA3 binding sites.

CCR4 plays crucial roles in skin-homing T cells (3). Its ligands, CCL17 and CCL22, are produced by endothelial cells, keratinocytes and dermal dendritic cells. IFNγ induces transcription of CCL17 and CCL22 genes in keratinocytes—a phenomenon that is implicated in atopic dermatitis (40). It has been reported that IFNγ production is increased in the HTLV-1–infected cells of HAM/TSP patients (7). We have also reported that IFNγ production is enhanced in HBZ-Tg mice (20). Collectively, these reports suggest that IFNγ produced by HTLV-1–infected cells can augment the infiltration of T cells through an IFNγ-CCL17/22-CCR4 cascade. In addition to CCR4, CD103 is also implicated in Treg cell migration into nonlymphoid tissues including skin, lung, intestine, and liver (32). Because CD103 expression is also enhanced in HBZ-Tg mice and ATL cases, it may further potentiate the migration of T cells into skin and other tissues.

Nonsense or frameshift mutations of the CCR4 gene, which result in truncation of the carboxy terminal of CCR4 protein, are frequently detected in ATL cases (13, 14). Although these mutations were found to increase migration toward CCL17 and CCL22 and promote proliferation, wild-type CCR4 did not enhance the proliferation of T cells promoted by CCL22 (13). In this study, we also observed that stimulation by CCL17/22 alone did not induce cell proliferation in CD4⁺ T cells of either HBZ-Tg or non-Tg mice (Fig. 4H). However, this study shows that stimulation by CCL17/22 and E-cadherin could promote proliferation of T cells expressing wild-type CCR4. Pathologic analyses suggest that ATL cells in skin lesions proliferate in response to the interaction of CCR4 and CD103 with these ligands. Treatment of leukemias/lymphomas using monoclonal antibodies depletes circulating cells by antibody-mediated cellular cytotoxicity in vivo (41). Skin lesions are more resistant to monoclonal antibodies due to suppressed penetration of antibody into skin (42). Therefore, skin-tropism and proliferation of ATL cells conferred by CCR4 and CD103 likely dampen efficacy of antibody therapy for ATL.

Previous studies reported that several chemokines promote not only migration but also cell proliferation (43–45). In this study, we observed enhanced proliferation of T cells when CCR4 ligands were combined with E-cadherin stimulation. Integrin signaling by various ECMs (by E-cadherin in this study) promotes metastasis and proliferation of cancer cells (33). When CCR4-expressing HTLV-1–infected cells and ATL cells respond to CCL17 and CCL22 by migrating into the skin, they encounter keratinocytes and endothelial cells that express E-cadherin. E-cadherin and CCL17/22 may then promote the proliferation of infiltrated infected cells and ATL cells as shown in this study. Consistent with this hypothesis, ATL cells in skin lesions are frequently positive for CCR4, CD103, and Ki67, which supports the idea that signaling via CCR4 and CD103 potentiates both the migration and proliferation of ATL cells.

CD103 expression is critical for resident memory T cells (46). HBZ increases CD103 expression on CD4⁺ T cells, and thus likely promotes the retention of tissue-resident memory T cells. Furthermore, this study reveals that CD103 also promotes the proliferation of ATL cells in skin lesions. Unlike CCR4 expression, CD103 expression is not directly induced by HBZ in transduced mouse CD4⁺ T cells (Supplementary Fig. S3), although CD103 expression is increased in the CD4⁺ T cells of HBZ-Tg mice. It remains to be determined how CD103 expression is enhanced in ATL cells and CD4⁺ T cells of HBZ-Tg mice.

HBZ-Tg mice develop inflammatory diseases and CD4⁺ T cells infiltrate into the skin, and intestines (20). We have reported that increased IFNγ production is implicated in these inflammations (20), and loss of IFNγ retards the onset of inflammatory diseases in these mice (21). However, HBZ-Tg/IFNγ KO mice still develop inflammatory diseases (later than HBZ-Tg mice), suggesting that factor(s) other than IFNγ contribute to the migration of CD4⁺ T cells and inflammation. CD4⁺ T cells of HBZ-Tg/IFNγ KO mice express CCR4 on their surfaces (21). Thus, it is possible that HBZ-induced CCR4 is an alternate mechanism for the infiltration of CD4⁺ T cells into skin and other tissues. Considering that HTLV-1 is transmitted only by cell-to-cell contact and transmission routes are breast-feeding, sexual intercourse, or transfer of blood, infected cells must enter into the breast milk or semen. It is possible that HTLV-1 utilizes CCR4 to cause infected cells to migrate into such places.

In this study, we demonstrate that HBZ induces CCR4 transcription through enhanced GATA3 expression. This CCR4 expression is implicated in the migration and proliferation of T cells. Furthermore, the ligands of CCR4 and CD103 promote the proliferation of T cells. Thus, HBZ possesses functions that enhance the migration and proliferation of T cells—phenomena that are important in pathogenesis by HTLV-1.

No potential conflicts of interest were disclosed.

Conception and design: K. Sugata, M. Matsuoka

Development of methodology: K. Sugata, K. Nakashima, K. Ohshima

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): K. Sugata, J.-i. Yasunaga, H. Kinosada, Y. Mitobe, R. Furuta, K. Nakashima

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): K. Sugata, J.-i. Yasunaga, R. Furuta, M. Mahgoub, M. Matsuoka

Writing, review, and/or revision of the manuscript: K. Sugata, J.-i. Yasunaga, M. Matsuoka

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Onishi, M. Matsuoka

Study supervision: J.-i. Yasunaga, M. Matsuoka

We thank T. Kitamura for the pMXs-Ig vector and Plat-E cells and H. Miyoshi for the pCS2-EF-GFP vector; Y. Satou, K. Yasuma, A. Kawatsuki, and P. Miyazato for valuable advice on experiments; and L. Kingsbury for proofreading.

This work was supported by JSPS KAKENHI Grant Number 25293219 (M. Matsuoka) and 13J05302 (K. Sugata), a grant from Mitsubishi Foundation (M. Matsuoka), and grants from the Ichiro Kanehara Foundation and SENSHIN Medical Research Foundation (J.-i. Yasunaga), in part by the Research Program on Emerging and Re-emerging Infectious Diseases from Japan Agency for Medical Research and Development, AMED. This study is also partially supported by the Joint Usage/Research Center program of the Institute for Virus Research, Kyoto University.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.