Induced Differentiation of Human Myeloid Leukemia Cells into M2 Macrophages by Combined Treatment with Retinoic Acid and 1α,25-Dihydroxyvitamin D3

Retinoids and 1α,25-dihydroxyvitamin D3 (1,25(OH)2D3) induce differentiation of myeloid leukemia cells into granulocyte and macrophage lineages, respectively. All-trans retinoic acid (ATRA), which is effective in the treatment of acute promyelocytic leukemia, can induce differentiation of other types of myeloid leukemia cells, and combined treatment with retinoid and 1,25(OH)2D3 effectively enhances the differentiation of leukemia cells into macrophage-like cells. Recent work has classified macrophages into M1 and M2 types. In this study, we investigated the effect of combined treatment with retinoid and 1,25(OH)2D3 on differentiation of myeloid leukemia THP-1 and HL60 cells. 9-cis Retinoic acid (9cRA) plus 1,25(OH)2D3 inhibited proliferation of THP-1 and HL60 cells and increased myeloid differentiation markers including nitroblue tetrazolium reducing activity and expression of CD14 and CD11b. ATRA and the synthetic retinoic acid receptor agonist Am80 exhibited similar effects in combination with 1,25(OH)2D3 but less effectively than 9cRA, while the retinoid X receptor agonist HX630 was not effective. 9cRA plus 1,25(OH)2D3 effectively increased expression of M2 macrophage marker genes, such as CD163, ARG1 and IL10, increased surface CD163 expression, and induced interleukin-10 secretion in myeloid leukemia cells, while 9cRA alone had weaker effects on these phenotypes and 1,25(OH)2D3 was not effective. Taken together, our results demonstrate selective induction of M2 macrophage markers in human myeloid leukemia cells by combined treatment with 9cRA and 1,25(OH)2D3.


Introduction
Retinoids play roles in numerous biological functions, such as cellular proliferation and differentiation, embryogenesis, immunity and metabolism [1]. An active natural retinoid, all-trans retinoic acid (ATRA), is effective in differentiation therapy for acute promyelocytic leukemia (APL) [2]. APL is a subtype of acute myeloid leukemia, which is characterized by a specific chromosomal abnormality t (15,17) associated with a genetic rearrangement between retinoic acid receptor a (RARa) (gene symbol, RARA) and the promyelocytic leukemia gene PML [2]. RARa plays a role in granulocytic differentiation of hematopoietic cells and the abnormal chimeric receptor PML-RARa has been implicated in APL pathogenesis by blocking the myeloid differentiation program and enhancing self-renewal of leukemic cells [3,4]. Pharmacological doses of ATRA induce differentiation of APL cells into granulocytes through degradation of PML-RARa and recovery of physiological RARa signaling [3,4].
Retinoids, including ATRA, 9-cis retinoic acid (9cRA) and synthetic RAR ligands, exhibit anti-tumor effects not only on APL but also on other malignancies, such as breast cancer, lung cancer, and head and neck cancer [5]. With regard to leukemia, ATRA was first reported to induce the differentiation of human myeloid leukemia HL60 cells towards the granulocytic lineage [6,7]. Importantly, HL60 cells are derived from non-APL leukemia without t (15,17) [8], and ATRA can also induce differentiation of leukemia cells from non-APL myeloid leukemia patients [9]. Retinoids in combination with other differentiation inducers, such as 1a,25-dihydroxyvitamin D 3 (1,25(OH) 2 D 3 ) and dibutyryl cAMP, synergistically induce differentiation of leukemia cells [10][11][12][13]. However, the underlying mechanisms of retinoidinduced differentiation of leukemia cells remain poorly understood and retinoids have not been utilized in the treatment of myeloid leukemia other than APL.
The active form of vitamin D 3 , 1,25(OH) 2 D 3 , regulates calcium and bone homeostasis, immunity, and cellular growth and differentiation through direct binding to the vitamin D receptor (VDR), and has been demonstrated to inhibit the proliferation and to induce the differentiation of various types of malignant cells, including breast, prostate and colon cancers as well as myeloid leukemia cells [14,15]. The administration of 1,25(OH) 2 D 3 and its analog has therapeutic effects in a mouse model of myeloid leukemia [16]. While ATRA induces granulocytic differentiation [6,7], 1,25(OH) 2 D 3 induces the differentiation of HL60 cells and other myeloid leukemia cells towards the monocyte and macrophage lineage [17,18]. Interestingly, ATRA induces monocytic differentiation of monoblastic leukemia U937 and THP-1 cells [13,19]. Combined treatment with 1,25(OH) 2 D 3 and retinoids induces the differentiation of HL60 cells and human monoblastic leukemia cells, such as THP-1 cells, to monocyte/macrophagelineage cells more effectively than 1,25(OH) 2 D 3 alone [10,12,20]. Although 1,25(OH) 2 D 3 has been shown to exert its biological effects on cellular proliferation and differentiation by genomic and/or non-genomic pathways [21], the detailed mechanisms remain unclear. Macrophages have been classified into two cell types, classically activated M1 macrophages and alternatively activated M2 macrophages [22,23]. While M1 macrophages produce proinflammatory cytokines and enhance microbicidal and tumoricidal immunity, M2 macrophages are involved in wound healing and immune regulation. Although retinoids and 1,25(OH) 2 D 3 play functional roles in monocytes and macrophages [4,24], the macrophage cell type resulting from 1,25(OH) 2 D 3 and/ or retinoid differentiation of myeloid leukemia cells has not been further characterized. In this study, we examined the effects of 1,25(OH) 2 D 3 in combination with retinoids on differentiation of myeloid leukemia cells and found that 1,25(OH) 2 D 3 in combination with 9cRA and ATRA induce the differentiation of myeloid leukemia cells to macrophages with M2-like phenotype.
Cell culture, cell growth, nitroblue tetrazolium (NBT) reduction, and interleukin-10 (IL-10) production Human myeloid leukemia HL60 and THP-1 cells (RIKEN Cell Bank, Tsukuba, Japan) were maintained in RPMI1640 medium containing 10% fetal bovine serum, 100 unit/ml penicillin, and 100 mg/ml streptomycin in a humidified atmosphere containing 5% CO 2 . Suspensions of cells (10 5 cells/ml) were cultured with or without test compounds at pharmacological concentrations (302 100 nM) according to our preliminary experiments and the previous reports [10][11][12]27]. Cell numbers were counted in a Z1S Coulter Counter (Beckman Coulter, Fullerton, CA). Cell morphology was examined in cell smears stained with May-Grünwald-Giemsa. NBT reduction was assayed colorimetrically and NBTreducing activity data were normalized to the cell numbers [28]. IL-10 levels in culture media were determined with the Human IL-10 ELISA MAX Standard kit (BioLegend, San Diego, CA).

Flow cytometry
Expression of cell surface antigens, CD14, CD11b and CD163, were determined with immunofluorescence staining and flow cytometry [29]. FITC mouse anti-human CD14, PE mouse anti-human CD11b, PE mouse anti-human CD163, and isotype control antibodies were purchased from Becton, Dickinson and Company (Franklin Lakes, NJ). The stained cells were assayed with a flow cytometer (BD FACSCalibur; Becton, Dickinson and Company) and analyzed with the BD CellQuest software (Becton, Dickinson and Company).

Reverse transcription and real-time quantitative polymerase chain reaction
Total RNAs from samples were prepared by the acid guanidine thiocyanate-phenol/chloroform method [30]. cDNAs were synthesized using the ImProm-II Reverse Transcription system (Promega Corporation, Madison, WI). Intron-spanning primers were as follows:

Statistical analysis
All values are shown as mean 6 S.D. We performed one-way ANOVA followed by Tukey's multiple comparisons or two-way ANOVA to assess significant differences using Prism 6 (Graphpad Software, La Jolla, CA).

Induction of differentiation of human myeloid leukemia cells by retinoids plus 1,25(OH) 2 D 3
We examined the effects of 9cRA and ATRA in the absence or presence of 1,25(OH) 2 D 3 on NBT-reducing activity, a marker of myeloid differentiation, in monoblastic leukemia THP-1 cells and promyelocytic leukemia HL60 cells. 9cRA (100 nM) increased NBT-reducing activity in THP-1 and HL60 cells, while ATRA (100 nM) was not effective (Fig. 1A), consistent with the previous reports showing that 9cRA is more potent than ATRA in inducing differentiation of leukemia cells [20,31,32]. In combination with 1,25(OH) 2 D 3 , 9cRA and ATRA effectively increased NBTreducing activity in these cells (Fig. 1A). Am80 (also called tamibarotene) is a potent synthetic RAR agonist that is used in the treatment of recurrent APL in Japan [33,34]. HX630 is an RXR selective agonist derived from LE135, an RAR antagonist [26]. HX630 does not exhibit RAR antagonistic activity but enhances the differentiation-inducing activity of Am80 in HL60 cells [33]. We also examined the effects of Am80 and HX630 in the absence or presence of 1,25(OH) 2 D 3 . Although Am80 and HX630 at 100 nM were not effective, the combination of 1,25(OH) 2 D 3 with Am80 but not HX630 significantly increased NBT-reducing activity in THP-1 and HL60 cells (Fig. 1A).
Induction of M2 macrophage markers in human myeloid leukemia cells by retinoids plus 1,25(OH) 2
The combination of 9cRA and 1,25(OH)2D3 induces IL-10 protein secretion in THP-1 and HL60 cells Finally, we examined IL-10 protein levels in conditioned media of THP-1 and HL60 cells treated with 9cRA and/or 1,25(OH) 2 D 3 . IL-10 protein levels from THP-1 cells treated with 9cRA and 1,25(OH) 2 D 3 alone were below detection limits, but the combination of these compounds effectively induced IL-10 protein secretion from these cells (Fig. 5). While IL-10 protein was detected in the culture media of untreated HL60 cells, 9cRA increased and 1,25(OH) 2 D 3 decreased the protein level. 9cRA plus 1,25(OH) 2 D 3 effectively increased the IL-10 protein levels in

Discussion
In this study, we found that combined treatment with retinoid and 1,25(OH) 2 D 3 induces the differentiation of human myeloid leukemia THP-1 and HL60 cells into the monocytic lineage with a M2 macrophage phenotype. ATRA induces granulocytic differentiation of promyelocytic leukemia HL60 cells but monocytic differentiation of monoblastic U937 and THP-1 cells [6,7,13,19]. Combination of ATRA or 9cRA with 1,25(OH) 2 D 3 effectively induces monocyte/macrophage phenotypes, such as phagocytic activity, monocyte-specific esterase, lysozyme secretion, and CSF1R expression, in HL60 and U937 cells [10,32,37]. We observed a monocytic morphology and increased CD14 expression in HL60 and THP-1 cells treated with 9cRA plus 1,25(OH) 2 D 3 (Figs. 1 and 2). ATRA plus 1,25(OH) 2 D 3 also induces differentiation of promyelocytic AML-193 cells into cells that display both a typical neutrophilic morphology and monocyte-specific properties, such as CD14 expression and monocytespecific esterase, a hybrid granulomonocytic phenotype [35]. RAR signaling plays an important role in hematopoiesis and RARa is involved in neutrophil development [4]. RAR and RXR signaling pathways have been reported to regulate monocyte/macrophage function [4]. However, it remains to be determined how retinoid signaling enhances monocytic differentiation induced by 1,25(OH) 2 D 3 in myeloid leukemia cells.
Among retinoids, 9cRA, ATRA and Am80, but not HX630, in combination with 1,25(OH) 2 D 3 exhibit effective differentiationinducing activity in these cells (Figs. 1 and 2). 1,25(OH) 2 D 3 acts as a ligand for the nuclear receptor VDR, which forms a heterodimer with RXR [38], and the VDR2RXR heterodimer is not permissive to RXR ligand activation [39]. RAR selective ligands exhibit stronger synergistic effects with 1,25(OH) 2 D 3 than RXR selective ligands in inhibiting proliferation and inducing differentiation of monoblastic U937 cells [27]. Combined effects of retinoid and 1,25(OH) 2 D 3 on differentiation of myelomonocytic leukemia cells are likely mediated by VDR and RAR activation. RXR also forms heterodimers with RAR and other nuclear receptors, including peroxisome proliferator-activated receptor (PPAR) and liver X receptor (LXR) [38]. The RAR2RXR heterodimer is activated by RXR ligand only in the presence of RAR ligand, a feature known as conditional permissivity [39]. 9cRA exhibits differentiation-inducing activity more effectively than ATRA in the absence or presence of 1,25(OH) 2 D 3 (Figs. 1 and 2), in agreement with previous reports [20,31,32]. Since 9cRA acts as a ligand for both RAR and RXR [40], synergistic activation may be due to binding to both RAR and RXR in the RAR2RXR heterodimer. In addition, RXR ligands can activate permissive heterodimers, such as PPAR2RXR and LXR2RXR [38]. PPARc ligand and LXR ligand have been reported to induce differentiation of myeloid leukemia cells [41,42]. RXR ligand activation of these permissive heterodimers may also contribute to the effect of 9cRA. However, the pure RXR ligand HX630 alone and in combination with 1,25(OH) 2 D 3 was not effective in inducing differentiation of THP-1 and HL60 cells, while the combination of 1,25(OH) 2 D 3 with the RAR selective agonist Am80 induced the differentiation of these cells (Figs. 1 and 2). These findings suggest that cooperation between VDR signaling and RAR signaling, not RXR signaling, plays a role in the differentiation of myeloid leukemia cells. VDR activation changes expression of many genes, including those involved in cellular proliferation, differentiation and apoptosis [21]. 1,25(OH) 2 D 3 treatment can modulate intracellular kinase pathways via a nongenomic mechanism, and it remains unknown whether the nongenomic actions are mediated through VDR or other proteins [21]. Although both genomic and non-genomic effects of 1,25(OH) 2 D 3 have been shown to play roles in differentiation induction of leukemia cells, the detailed mechanisms remain to be elucidated. RAR signaling may modulate the vitamin D signaling pathway or regulate other differentiation mechanisms. Further studies are needed to elucidate molecular mechanisms involving VDR, RAR and RXR signaling pathways in the induced differentiation of leukemia cells.
Combined treatment with 9cRA and 1,25(OH) 2 D 3 increased mRNA expression of CD163, ARG1, IL10, and TGFB1 genes (Fig. 3), surface expression of CD163 proteins (Fig. 4) and IL-10 secretion in THP-1 cells (Fig. 5). This combination also increased CD163, ARG1 and IL10 mRNA levels (Fig. 3) and IL-10 secretion in HL60 cells (Fig. 5). This phenotype has been characterized as M2 macrophages, although the classification of human macrophages remains controversial [23]. 9cRA plus 1,25(OH) 2 D 3 did not increase expression of the M1 macrophage gene IL12B in THP-1 and HL60 cells (Fig. 3). Although TNF mRNA levels were not increased, IL6 mRNA expression was effectively induced by the combination of 9cRA and 1,25(OH) 2 D 3 in THP-1 cells. ATRA plus 1,25(OH) 2 D 3 has been reported to induce mRNA and protein levels of tumor necrosis factor (TNF) and IL-6 in U937 cells [43] and to increase expression of inducible nitric oxide synthase (encoded by NOS2) and nitric acid production in U937 cells [44], while we observed no significant change in NOS2 mRNA expression in THP-1 and HL60 cells after treatment with ATRA or 9cRA in combination with 1,25(OH) 2 D 3 (Fig. 3, and data not shown). Tumor necrosis factor (TNF) and IL-6 are cytokines produced from M1 macrophages and NOS2 expression is a M1 macrophage marker [22,23,45]. Macrophages with a mixed phenotype expressing both M1 and M2 markers have been identified [46]. Thus, differentiated leukemia cells by 9cRA plus 1,25(OH) 2 D 3 are macrophage-like cells expressing primarily M2 markers with some M1 markers. Further analysis is required to reveal their functional characteristics.
The RAR signaling pathways play an important role in hematopoiesis and granulocytic differentiation [4,47]. With regard to macrophages, ATRA inhibits TNF production in mouse peritoneal macrophages activated by lipopolysaccharide and interferon c [48]. ATRA also reduces the synthesis of IL-12 and TNF and enhances IL-10 production in lipopolysaccharidestimulated human macrophages [49]. Although VDR is dispensable for normal myelopoiesis [50], the vitamin D signaling pathway is involved in the regulation of macrophage/monocyte function [4]. 1,25(OH) 2 D 3 suppresses activation of mouse macrophages by interferon c [51], and enhances the immunoglobulin-and complement-dependent phagocytosis activity of human blood monocytes [52]. Thus, both ATRA and 1,25(OH) 2 D 3 induce the macrophage/monocyte function common to M2 macrophages. Combined effects of retinoid and 1,25(OH) 2 D 3 on physiological monocyte/macrophage function remain to be elucidated.
In conclusion, our results indicate that combined treatment with retinoid and 1,25(OH) 2 D 3 induces differentiation of human myeloid leukemia THP-1 and HL60 cells into macrophage-like cells expressing M2 markers. Further study of human leukemia cell differentiation has the potential to extend differentiation-inducing therapy to the treatment of non-APL myeloid leukemia and to expand the understanding of human macrophage function.