CCR 4 and its ligands : from bench to bedside

Chemokines and chemokine receptors orchestrate cell migration and homing in the body. Humans have at least 44 chemokines that are further classified into four subfamilies based on the N-terminal conserved cysteine motifs: CXC, CC, C and CX3C. All the known chemokine receptors are seven transmembrane-type receptors. Humans have 18 chemotactic and 5 atypical non-chemotactic (recycling or scavenging) receptors. CC chemokine receptor 4 (CCR4) is the receptor for two CC chemokine ligands (CCLs)—CCL17 (also called thymusand activation-regulated chemokine) and CCL22 (macrophage-derived chemokine). Among the various T-cell subsets, CCR4 is predominantly expressed by Th2 cells, cutaneous lymphocyte antigen-positive skin-homing T cells and Treg cells. Thus, CCR4 attracts much attention for its possible clinical applications in diseases involving these T-cell subsets. Furthermore, CCR4 is often highly expressed by mature T-cell neoplasms such as adult T-cell leukemia/lymphoma (ATL) and cutaneous T-cell lymphomas (CTCLs). This article is a brief overview of basic and clinical research on CCR4 and its ligands, which has eventually led to the development of a humanized defucosylated anti-CCR4 antibody ‘Mogamulizumab’ for treatment of relapsed/refractory ATL and CTCLs.


Introduction
Chemokines are a family of small, structurally related 'chemotactic cytokines' and two transmembrane proteins with an N-terminal chemokine domain perched on a mucin-like stalk (1)(2)(3).The family members are further classified into four subfamilies according to the N-terminal cysteine motifs: CXC, CC, C and CX 3 C.There are at least 44 chemokines in humans (Fig. 1).All known chemokine receptors are heptahelical seven-transmembrane receptors (1)(2)(3).Humans have 18 chemokine receptors that mediate chemotactic responses through coupling to the Gαi class of the heterotrimeric G proteins (Fig. 1).Furthermore, there are five atypical chemokine receptors (now termed ACKR1 to 5) that are not coupled with heterotrimeric G proteins and are primarily involved in scavenging chemokines through coupling to β-arrestin (Fig. 1) (3,4).
The primary task of the chemokine superfamily is to orchestrate cell migration and homing in health and disease.Thus, chemokines are involved in various biological processes such as homeostatic migration and homing of lymphocytes, inflammatory mobilization of leukocytes, cell migration and homing during development, angiogenesis, wound healing and cancer metastasis (1)(2)(3).Notably, chemokines and chemokine receptors often have highly promiscuous relationships: a single chemokine binds to several chemokine receptors, while a single chemokine receptor interacts with multiple ligands (Fig. 1) (1)(2)(3).
Such complex ligand-receptor relationships are most probably due to rapid expansions of chemokines and chemokine receptors through multiple gene duplication events during vertebrate evolution.The dramatic increases in ligands and receptors of the chemokine superfamily during vertebrate evolution are likely to be related to the development and sophistication of the vertebrate immune systems.The most primordial chemokine receptors are CXC chemokine receptor 4 (CXCR4) and CXCR7 that are present in jawless fish and both interact with CXC chemokine ligand 12 (CXCL12) (5).Indeed, the CXCR4/CXCR7-CXCL12 axis plays pivotal roles in the organogenesis of heart and central nervous system as well as in homing of various stem cells to particular places of the embryo during development (3).On the other hand, the evolutionary appearance of CC chemokine receptor 4 (CCR4) is relatively recent, definitively present in reptiles along with CXCR3, CCR2, CCR5 and CX 3 CR1, while CCR1 and CCR3 are only seen in mammals (5).
Adult T-cell leukemia/lymphoma (ATL) is a highly aggressive mature CD4 + CD25 + T-cell neoplasm etiologically associated with human T-lymphotropic tropic virus type 1 (HTLV-1), which is endemic in certain areas of the world including Southern Japan and infects 15-20 million people worldwide (6).The incidence of ATL in HTLV-1 carriers is relatively low, with a long latency period of several decades and a cumulative lifetime risk of ~5% in Japan.ATL is also a spectrum of T-cell proliferative diseases categorized into four clinical subtypes: acute, lymphoma, chronic and smoldering.Acute ATL is the most frequent and aggressive subtype, refractory to conventional chemotherapy, and characterized by flower cells in the blood and frequent lymphadenopathy, hepatosplenomegaly, skin lesions and hypercalcemia (7).On the other hand, cutaneous T-cell lymphomas (CTCLs) are a highly heterogeneous group of lymphoproliferative disorders caused by mature skin-invasive T cells such as mycosis fungoides, Sezary syndrome and anaplastic large cell lymphoma (ALCL) (8).CTCLs are not etiologically associated with HTLV-1.
In this article, we first describe the original discovery of CC chemokine ligands 17 and 22 (CCL17 and CCL22) and their shared receptor CCR4.We then highlight the association of CCR4 with Th 2 responses and diseases such as atopic dermatitis (AD) (for which CCL17 is now a useful biomarker) and mature T-cell neoplasms.We then focus on the development and clinical assessment of a humanized defucosylated anti-CCR4 antibody 'Mogamulizumab' for the treatment of relapsed/refractory ATL and CTCLs (9)(10)(11).
At that time, there existed a chemokine dogma; namely, CC chemokines (represented by MCP-1/CCL2) were monocyte chemoattractants and their genes were clustered on human chromosome 17, whereas CXC chemokines (represented by IL-8/CXCL8) were neutrophil mobilizers and their genes were clustered on chromosome 4 (13,14).Its strong expression in the thymus, however, suggested that TARC/CCL17 might be a chemoattractant for the cells of T-cell lineage.Indeed, radio-labeled TARC/CCL17 specifically bound to peripheral blood T cells but not to monocytes or granulocytes (12).TARC/CCL17 induced vigorous chemotactic responses in two human T-cell lines, HUT102 and HUT78, but not in monocytes or granulocytes (12).
Furthermore, the gene for TARC/CCL17 was mapped to chromosome 16 instead of 17, where the classical CC chemokine gene cluster exists (12).Thus, TARC/CCL17 had several features that were quite new at that time: the first CC chemokine chemotactic for lymphoid cells but not for monocytes; and the first CC chemokine not mapped to the major CC chemokine cluster on chromosome 17.These features were in fact shared by an emerging group of lymphocytespecific and dendritic cell-specific chemokines that were subsequently identified in rapid succession (15).The presence of chemokines that might control lymphocyte recirculation and homing had long been postulated but was still missing at that time (16).Thus, the discovery of TARC/CCL7 was indeed a herald for an important breakthrough in the field of chemokines and even immunology at large.Another notable point worth mentioning here is that the two human T-cell lines originally used to demonstrate the chemotactic activity of TARC/CCL17 were in fact derived from patients with ATL and CTCL, respectively (see below).
In an attempt to identify the receptor for TARC/CCL17, we next examined its binding to a panel of 'orphan' seventransmembrane type receptors and some known chemokine receptors (17).Much to our surprise, CCL17 was found to bind to CCR4, which had been previously reported as a new receptor for MCP-1/CCL2, macrophage inflammatory protein 1α (MIP-1α)/CCL3 and RANTES/CCL5 (18).Subsequently, we conclusively demonstrated that TARC/CCL17 bound to CCR4 and induced vigorous chemotactic and calcium flux responses in CCR4-transfected cells, whereas no other CC chemokines tested at that time, including MCP-1/CCL2, MIP-1α/CCL3 and RANTES/CCL5, showed any significant binding or signaling activity on CCR4 (17).Thus, CCR4 was re-assigned as a highly specific receptor for TARC/CCL17.These results were again rather contradictory to the prevalent chemokine dogma at that time; namely, chemokine receptors often interact with multiple ligands that are also shared by other receptors (13,14).
Separately, Godiska et al. isolated a novel CC chemokine from a human monocyte-derived macrophage library and termed it 'macrophage-derived chemokine' (MDC, now also called CCL22) (19).They demonstrated that MDC/CCL22 was also highly expressed in the thymus.However, rather inconsistent to its tissue expression, they reported that it was chemotactic for monocytes and IL-2-activated NK cells (19).Subsequently, we demonstrated that MDC/CCL22 was another high-affinity ligand of CCR4 (20).Although only 37% identical in amino acid sequence, CCL17 and CCL22 are both highly expressed in the thymus (suggesting their roles in T-lineage cell migration) and encoded by genes closely mapped on human chromosome 16q13 (20).Thus, TARC/ CCL17 and MDC/CCL22 are a pair of CC chemokines obviously generated from a common ancestor gene by gene duplication and thus share CCR4.Importantly, however, TARC/CCL17 and MDC/CCL22 are not equivalent; MDC/ CCL22 is dominant over TARC/CCL17 in cross-desensitization of CCR4 (20,21).The reason why MDC/CCL22 was originally reported to be chemotactic for cells not expressing CCR4 such as monocytes remained unclear (19).

CCR4 and T-cell subsets
In the thymus, CCR4 was shown to be mainly expressed by medullary CD4-single positive thymocytes expressing CD30, supporting its possible role in negative selection (22).However, no gross abnormality in T-cell development has been reported in CCR4-deficient mice so far.In the periphery, mature T cells are known to be composed of multiple functional subsets.It is now known that each T-cell subset has a dominant-type chemokine receptor (Table 1) (1-3).In this regard, Sallusto et al. were the first to open this paradigm.They reported that CCR3, which is mainly expressed by eosinophils and basophils (23,24), was also selectively expressed by T h 2 cells (25).Their conclusion provoked an exciting notion that CCR3 was the common migratory receptor for all types of cells involved in allergic responses.
Meanwhile, Kyowa Hakko Kogyo in collaboration with us generated a murine IgG1 mAb against human CCR4 (KW2160) (Fig. 2).Therefore, we examined the expression of CCR4 in PBMCs from healthy donors using KW2160 (26).CCR4 was found to be expressed on a fraction of effector/ memory T cells, which produced IL-4 and IL-5 but not IFN-γ upon in vitro culture (26).Conversely, when naive T cells were polarized in vitro into T h 1 and T h 2, CCR4 was predominantly expressed by T h 2-polarized cells (26).On the other hand, we did not observe any significant CCR3 expression in either T h 1 or T h 2 cells.Collectively, we concluded that CCR4 but not CCR3 was the chemokine receptor preferentially expressed by T h 2 cells (26) Therapeutic use of anti-CCR4 13 expression of CCR4 in T h 2 cells derived from in vitro polarization of cord blood cells (27,28).
To further define the skewed expression of chemokine receptors by T h 1 and T h 2 cells, we performed double staining for surface chemokine receptors and intracellular cytokines in PBMCs from healthy donors and patients with AD (29).We found that T h 1 and T h 2 cells selectively expressed CXCR3 and CCR4, respectively, whereas neither subset expressed CCR3 (29).Consistently, there were increases in peripheral memory CD4 + T cells expressing CCR4 and decreases in those expressing CXCR3 in AD patients (29).

CCR4 and allergy
In accordance with the T h 2-predominant expression of CCR4, the importance of CCR4 and its ligands in allergic diseases has been amply demonstrated in mouse models of asthma and AD (54-59) and human clinical samples from asthma (60)(61)(62)(63)(64)(65), AD (66-71), allergic contact dermatitis (72) and eosinophilic pneumonia (73,74).Furthermore, we and others demonstrated that the blood content of TARC/CCL17 and MDC/CCL22 is highly elevated in allergic diseases such as AD (75)(76)(77)(78), asthma (79) and allergic rhinitis (80).We have further demonstrated that platelets contain CCL17, the amount of which is increased in AD patients, thus resulting in further increases in the serum content of CCL17 in AD patients (77).Accordingly, a serum TARC/CCL17 ELISA kit was approved in 2008 by the Japanese Pharmaceutical and Medical Devices Agency (PMDA) as a blood test for AD (Alaport® TARC).TARC/CCL17 is now regarded as a highly valuable biomarker to objectively monitor disease activity of AD during therapy (81).

CCR4 and skin-homing T cells
The crucial role of CCR4 in skin-homing T cells has also been amply demonstrated (31,82,83).Furthermore, CCL17 and CCL22 are not just redundant CCR4 ligands for skin-homing of T cells.It was originally noted that CCL22 was functionally dominant over CCL17 in ligand-mediated desensitization of CCR4 (20).Mariani et al. further demonstrated that CCL22 internalized and desensitized CCR4 more efficiently than CCL17 did (21).It was also shown that CCL17 and CCL22 were differentially expressed in inflamed skin mainly by endothelial cells and dermal dendritic cells, respectively (30,78,84).Accordingly, these two chemokines may provide sequential guidance of CCR4-expressing T cells into inflamed skin; CCL17 promotes vascular recognition and emigration of CCR4-expressing T cells at the endothelial surface, while CCL22 takes over the subsequent steps and guides migration of CCR4-expressing T cells within skin tissue (21).

CCR4 and T reg cells
The important role of the CCR4 axis in immune suppression by T reg cells has also been well documented in animal models and human samples (85)(86)(87)(88)(89)(90)(91)(92)(93)(94).For example, in an allogeneic cardiac transplantation model, tolerance induction by anti-CD154 plus donor-specific transfusion, which was accompanied by intragraft accumulation of Foxp3 + T cells, was not achieved in Ccr4 −/− recipients (86).Ccr4 −/− T reg cells failed to accumulate in the mesenteric lymph nodes to suppress colitis in a mouse adoptive transfer model of inflammatory bowel disease (88).Accumulation of T reg cells in the skin and lung airways was impaired in Ccr4 −/− mice, resulting in severe inflammatory diseases (94).

CCR4 and T-cell neoplasms
Given that various T-cell subsets differentially express chemokine receptors (Table 1), we postulated that the chemokine receptor expression pattern might be useful to determine cellular origin and tissue tropism of T-cell neoplasms.To test this hypothesis, we examined expression of various chemokine receptors including CCR4 in a panel of established human T-cell lines (95).It was quickly apparent that T-cell lines derived from patients with ATL as well as those established by infection with HTLV-1 consistently expressed CCR4 at high levels (95).We therefore examined blood samples from a total of 24 acute ATL cases and clearly demonstrated that primary leukemic cells from the majority of cases highly expressed CCR4 at both mRNA and protein levels and strongly responded to its ligands TARC/CCL17 and MDC/ CCL22 in chemotaxis assays (95).These findings suggested that ATL is predominantly derived from CCR4 + T-cell subsets such as T h 2 cells and T reg cells, and CCR4 may also account for the frequent skin involvement of ATL (95).Subsequently, Ishida et al. also confirmed the frequent expression of CCR4 in ATL by immunohistochemical staining of a total of 103 formalin-fixed and paraffin-embedded ATL tissues and further provided evidence that CCR4 + ATL patients had a poor prognosis compared to CCR4 − ones (96).Furthermore, FOXP3, the master transcription factor of T reg cells, was also found to be frequently expressed in ATL (97).This may support the T reg origin of ATL, although a novel HTLV-1-encoded oncogene HBZ was also shown to up-regulate FOXP3 in ATL (98).CCR4 was also shown to be frequently expressed by CTCLs (99,100) and ALCL (101)(102)(103)(104) and by a fraction of peripheral T-cell lymphoma (PTCL) (101,104).CCR4 expression by these T-cell neoplasms may also relate to their cellular origins and frequent skin involvement.
We also noted that primary ATL cells often expressed CCR4 at much higher levels and vigorously migrated to TARC/ CCL17 and MDC/CCL22 with an optimal concentration about 10-to 100-fold less than normal CD4 + CD25 + T cells (95,105).Such highly elevated expression of CCR4 in primary ATL cells might be due to a transcription factor(s) that is highly active in ATL and thus could be involved in ATL oncogenic processes, making CCR4 a kind of ATL tumor marker.To test this hypothesis, we examined the transcriptional regulation of CCR4 expression in ATL ( 106).An AP-1 site and a GATA-3 site were found as the major regulatory elements in the CCR4 promoter.The involvement of GATA3 might account for the T h 2-selective expression of CCR4 (106).Therefore, we focused on the AP-1 site and examined the AP-1 family transcription factors involved in CCR4 expression in ATL.The results show that ATL aberrantly expresses FRA-2, a FOS group member of the AP-1 family, which, in association with JUND, a JUN group member of the AP-1 family, enhances CCR4 expression as well as cell proliferation of ATL (106).Similar results were obtained from CCR4-expressing CTCLs (107).We have further demonstrated that FRA-2/JUND induces various protooncogenes such as c-MYB, MDM2 and SOX4 in ATL and CTCLs (106,107).In turn, SOX4 induces the gene encoding histone deacetylase 8 (HDAC8) in ATL and CTCLs (108), a highly promising chemotherapeutic target of ATL and CTCLs (109).Thus, we have uncovered the presence of a hitherto unknown oncogenic cascade involving FRA-2/JUND in ATL and CTCLs, which may eventually provide useful diagnostic and therapeutic molecules for these mature T-cell neoplasms (Fig. 3).
We also observed that HTLV-1-immortalized T-cell lines commonly expressed CCR4 (95).In this regard, the HTLV-1encoded potent transcriptional activator Tax was not responsible for CCR4 up-regulation (95).Therefore, HTLV-1 might be preferentially transmitted to CCR4 + T cells.To test this possibility, we examined the expression of various chemokines in HTLV-1-infected T cells (110).We found that HTLV-1-infected T cells strongly expressed CCL22 through potent induction by HTLV-1-encoded Tax.Accordingly, HTLV-1-infected T cells secrete a large amount of CCL22 and attract CCR4 + T cells, thus leading to preferential transmission of HTLV-1 to CCR4 + T cells through close cell-to-cell contacts (110).Attraction of CCR4 + T reg cells via CCL22 may also help HTLV-1-infected T cells evade host immune attacks in vivo (111).We have further shown that c-MAF, one of the critical transcription factors for T h 2 differentiation, suppresses the transcriptional activity of HTLV-1 Tax, thus potentially promoting HTLV-1 latency in T h 2 cells (112).This may also help immune evasion of HTLV-1-infected T cells in vivo.Collectively, there may be multiple reasons for the preferential transmission of HTLV-1 to CCR4 + T cells.

Drug development
The dominant expression of CCR4 in T h 2 cells makes it a potential therapeutic target for allergic diseases such as asthma and AD (113).Furthermore, because of the expression of CCR4 on T reg cells, CCR4 blockade may also be useful to enhance the efficacy of cancer vaccines (114).Accordingly, a number of small-molecule CCR4 antagonists have been generated and evaluated in animal models of allergic diseases, some giving highly promising results (113,115).However, most compounds are still preclinical.GlaxoSmithKline recently disclosed a series of indazole sulfonamides (116) and identified a clinical development compound GSK2239633, which entered phase I clinical trials.However, its poor bioavailability precluded further development.It thus remains to be seen whether other compounds in the pipeline will successfully go into human trials (113).
The frequent expression of CCR4 in ATL and CTCLs also makes it a possible therapeutic target for these T-cell neoplasms.Therefore, Kyowa Hakko Kogyo generated a chimeric antibody KM2760-from the mouse anti-CCR4 mAb KM2160 and a human IgG1-that was also defucosylated by Potelligent® technology to enhance its binding to FcγRIIIa, the dominant FcγR of NK cells (Fig. 2).KM2760 showed Therapeutic use of anti-CCR4 15 highly enhanced antibody-dependent cellular cytotoxicity (ADCC) activity in vitro and efficient antitumor activity against CCR4-expressing leukemic cells in vivo (117).Further studies demonstrated that KM2760 maintained high ADCC levels even at a 10-fold lower CCR4 density on target cells (118) and mediated ADCC by allogeneic and autologous effector cells against leukemic cells from ATL and CTCLs in vitro (119,120) as well as in vivo using humanized NOG (NOD/Shi-scid/ IL-2Rγ null ) mice (121,122).
Separately, a group from Harvard University also generated a humanized monoclonal anti-CCR4 antibody mAb2-3 and demonstrated a potent antitumor effect against a CTCL cell line in vitro and in vivo as well as inhibition of T reg cell chemotaxis and T reg -mediated suppression of effector T-cell proliferation (123).
Kyowa Hakko Kogyo then fully humanized KM2760 to generate KW0761 (Fig. 2), which was proven to be as efficient as KM2760 in mediating ADCC against primary ATL cells in preclinical studies (124).In 2006, a phase I clinical trial of KW0761 in patients with relapsed CCR4 + ATL and PTCL was started.Intravenous infusion of KW0761 at escalating doses of 0.01, 0.1, 0.5 and 1.0 mg/kg was given once a week for 4 weeks.KW0761 was well tolerated at all doses and achieved objective responses in 5 out of 15 evaluable patients: two complete (one at 0.1 mg/kg, one at 1.0 mg/ kg) and three partial (one at 0.01 mg/kg, two at 1.0 mg/kg) responses (125).
The subsequent phase II study was started in 2010 for relapsed CCR4 + ATL patients with an infusion of 1.0 mg/kg once a week for 8 weeks (126).Objective responses were obtained in 13 out of 26 evaluable patients with an overall response rate of 50% (126).A common adverse event was skin rash, which was also of higher grades in patients with favorable responses but was manageable (126).Accordingly, the Japanese PMDA approved KW0761 (Mogamulizumab) in March of 2012 as a therapeutic agent for relapsed/refractory ATL.Subsequently, a phase II study of Mogamulizumab in patients with relapsed PTCLs and CTCLs was also conducted at a dose of 1.0 mg/kg once for 8 weeks with objective responses in 13 out of 37 patients including 5 complete responses (127).Mogamulizumab may also be applicable to allergic diseases such as asthma and atopic diseases by targeting T h 2 cells (9) and to cancer adjuvant therapy by targeting T reg cells (128).
It should be noted, however, that the interim report for the post-marketing surveillance from May 29 to September 28, 2012, revealed nine skin-related severe adverse effects including four cases of Steven-Johnson syndrome (SJS) with one fatal case (129).The skin lesions of an SJS case demonstrated a massive infiltration of CD4 + and CD8 + T cells coupled with almost a total absence of FOXP3 + T reg cells, suggesting that T reg depletion was partly responsible for the severe skin lesions (129).Thus, the effect of Mogamulizumab on CCR4 + T reg cells could be double-edged, not only enhancing antitumor immunity but also increasing the risk of severe immune-mediated adverse effects.This may raise concern on the application of Mogamulizumab to less life-threatening diseases such as asthma and AD.The outcome of an ongoing phase I clinical trial of Mogamulizumab for asthma in the USA remains to be seen (113).

Concluding remarks
Despite tremendous efforts by the academic sector and pharmaceutical companies, only two chemokine receptor antagonists have been licensed so far for clinical use: Pfizer's CCR5 inhibitor Maraviroc and Anormed's CXCR4 inhibitor Plerixafor (113).Both Maraviroc and Plerixafor were developed as inhibitors of the entry of HIV-1 R5 and X4 strains, respectively.However, Plerixafor was approved for the application of bone marrow stem cell mobilization for autologous transplantation (113).Meanwhile, Mogamulizumab (KW0761) is a fully humanized monoclonal anti-CCR4 and the first biologic agent targeting a chemokine receptor.It was licensed for clinical use against relapsed/refractory ATL in March 2012 in Japan.It is also the first product of Potelligent Technology that aims to potentiate ADCC activity through defucosylation at the Fc region of antibody to enhance its binding to the Fc receptor on effector cells.Mogamulizumab has been proven to be quite effective in the treatment of relapsed/refractory ATL and CTCLs (126,127).However, the frequent and sometimes grave side effects including skin rash have also been noted (129).Thus, Mogamulizumab has both bright and dark sides.In fact, Mogamulizumab is still at its early stages of clinical use and may require much more experience to optimize its applications, treatment protocols and combinations with other drugs to maximize its benefits and to minimize its adverse effects.Mogamulizumab may also be useful for T reg -depleting cancer adjuvant therapy and therapy for T h 2mediated allergic diseases.If all goes well, Mogamulizumab may also provide a much needed proof of principle for the development of small-molecule CCR4 antagonists.

Table 1 .
. Although only at the mRNA level, Bonecchi et al. as well as Sallusto et al. also reported the dominant T-cell subsets and their dominant-type chemokine receptors