Langerhans Cells Dendritic Cells Is a Direct Precursor of Subset of Human Blood + /Cd11c + a Cd1a

Home , Glycophorin A, Langerhans cell, Langerin, Lineage markers

A CD1a+/CD11c+ Subset of Human Blood Dendritic Cells Is a Direct Precursor of Langerhans Cells

This information is current as Tomoki Ito, Muneo Inaba, Kayo Inaba, Junko Toki, Shinji of October 2, 2021. Sogo, Tomoko Iguchi, Yasushi Adachi, Kazuyuki Yamaguchi, Ryuichi Amakawa, Jenny Valladeau, Sem Saeland, Shirou Fukuhara and Susumu Ikehara J Immunol 1999; 163:1409-1419; ; http://www.jimmunol.org/content/163/3/1409 Downloaded from

References This article cites 40 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/163/3/1409.full#ref-list-1 http://www.jimmunol.org/

Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication by guest on October 2, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 1999 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. A CD1a؉/CD11c؉ Subset of Human Blood Dendritic Cells Is a Direct Precursor of Langerhans Cells1

Tomoki Ito,*† Muneo Inaba,† Kayo Inaba,‡ Junko Toki,† Shinji Sogo,§ Tomoko Iguchi,*† Yasushi Adachi,† Kazuyuki Yamaguchi,* Ryuichi Amakawa,* Jenny Valladeau,¶ Sem Saeland,¶ Shirou Fukuhara,* and Susumu Ikehara2*

Based on the relative expression of CD11c and CD1a, we have identiﬁed three fractions of dendritic cells (DCs) in human peripheral blood, including a direct precursor of Langerhans cells (LCs). The ﬁrst two fractions were CD11c؉ DCs, comprised of a major CD1a؉/CD11c؉ population (fraction 1), and a minor CD1a؊/CD11c؉ component (fraction 2). Both CD11c؉ fractions displayed a monocyte-like morphology, endocytosed FITC-dextran, expressed CD45RO and myeloid markers such as CD13 and CD33, and possessed the receptor for GM-CSF. The third fraction was comprised of CD1a؊/CD11c؊ DCs (fraction 3) and

resembled plasmacytoid T cells. These did not uptake FITC-dextran, were negative for myeloid markers (CD13/CD33), Downloaded from and expressed CD45RA and a high level of IL-3R␣, but not GM-CSF receptors. After culture with IL-3, fraction 3 acquired the characteristics of mature DCs; however, the expression of CD62L (lymph node-homing molecules) remained unchanged, indicating that fraction 3 can be a precursor pool for previously described plasmacytoid T cells in lymphoid organs. Strikingly, the CD1a؉/CD11c؉ DCs (fraction 1) quickly acquired LC characteristics when cultured in the presence of GM-CSF ؉ IL-4 ؉ TGF-␤1. Thus, E-cadherin, Langerin, and Lag Ag were expressed within 1 day of culture, and typical Birbeck granules were ؊ ؉ ؊ ؊ observed. In contrast, neither CD1a /CD11c (fraction 2) nor CD1a /CD11c (fraction 3) cells had the capacity to differ- http://www.jimmunol.org/ entiate into LCs. Furthermore, CD14؉ monocytes only expressed E-cadherin, but lacked the other LC markers after culture in these cytokines. Therefore, CD1a؉/CD11c؉ DCs are the direct precursors of LCs in peripheral blood. The Journal of Immunology, 1999, 163: 1409–1419.

endritic cells (DCs)3 are bone marrow-derived profes- of epidermal LCs. The other pathway includes a CD1aϪ/CD14ϩ sional APCs that initiate and regulate immune responses intermediate that can differentiate into monocyte-derived or dermal D (1–3). Recent studies have demonstrated that DCs in situ DCs that lack LC markers. GM-CSF and TNF-␣ are required in can exist in different maturational states and pathways of differ- both pathways, but TGF-␤1 is found to be important for the for- entiation. Subtypes of DCs with some differences in morphology, mation of Birbeck granules and Lag Ag of LCs (20–22). It has also by guest on October 2, 2021 phenotype, and function include interdigitating cells in lymphoid been found that peripheral blood monocytes cultured with a com- organs (4), peripheral blood DCs (5, 6), Langerhans cells (LCs) in bination of GM-CSF, IL-4, and TNF-␣ differentiate into mature the epidermis of the skin (7, 8), dermal DCs (9, 10), and thymic DCs (23), and that LCs may develop if TGF-␤1 is added (24). The DCs (11, 12). In the human tonsil, DCs in the germinal center (13) identification of DC intermediates in the peripheral blood is, there- and plasmacytoid T cells in the T cell zones (14) have recently fore, important, because it is thought that most blood DCs are been identified as new DC subsets. either en route from the bone marrow to peripheral tissues, or from Hemopoietic progenitors that differentiate into various types of nonlymphoid tissues to the regional lymph nodes and spleen. DCs are present in the cord blood (15, 16), bone marrow (17), and Previously, DCs were purified from the peripheral blood after a peripheral blood (18, 19). At least two separate developmental 1- to 2-day culture period. These DCs were morphologically, phe- ϩ pathways of DCs from CD34 progenitors have been shown (16, notypically, and functionally different from the DCs present in the ϩ Ϫ 19): one pathway involves CD1a /CD14 cells with the features circulation (25, 26). Recently, DCs have been isolated from the peripheral blood by immunoselection methods without culture. DCs thus prepared are not the morphologically typical DCs, and, † *First Department of Internal Medicine and First Department of Pathology, Kansai interestingly, there are two subsets with distinct phenotype (27– Medical University, Moriguchi, Osaka, Japan; ‡Department of Zoology, Kyoto Uni- versity, Sakyo, Kyoto, Japan; §Cellular Technology Institute, Otsuka Pharmaceutical 29). The nature, origin, and commitment of these DCs require Co., Tokushima, Japan; and ¶Schering-Plough, Laboratory for Immunological Re- further analysis. In the present study, we have isolated fractions of search, Dardilly, France blood DCs using magnetic beads and a multicolor sorting system. Received for publication January 11, 1999. Accepted for publication May 20, 1999. We have distinguished three distinct fractions by extensive surface The costs of publication of this article were defrayed in part by the payment of page marker analyses, and have examined their functions and develop- charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. mental capacities. During these investigations, we have identified 1 This work was supported by a grant from the Ministry of Health and Welfare’s a distinct subset of DCs that quickly differentiates into LCs. Study Group on Specific Diseases Research and by grants from the Ministry of Ed- Strunk et al. have reported that, in the peripheral blood, only ucation, Science, and Culture, and the Japan Private School Promotion Foundation. CD34ϩ progenitors expressing the skin-homing receptor differen- 2 Address correspondence and reprint requests to Dr. Susumu Ikehara, First Depart- tiate into LCs (after 10- to 18-day culture) (19). It has recently ment of Pathology, Kansai Medical University, 10-15 Fumizono-cho, Moriguchi City, ␤ Osaka 570-8506, Japan. E-mail address: [email protected] been reported that, in the presence of GM-CSF, IL-4, and TGF- 1 ϩ 3 Abbreviations used in this paper: DC, dendritic cell; EM, electron microscope; LC, for 6 days, human peripheral blood monocytes (CD14 cells) can Langerhans cell; PerCP, peridini-chlorophyll protein; PI, propidium iodide. differentiate into LCs characterized by the expression of CD1a,

Table I. Abs used for phenotypical analysisa

Clone Isotype Source Clone Isotype Source

CD1a BB-5 IgG1 TAG CD42b SZ2 IgG1 IM CD1a BL-6 IgG1 IM CD44 F10-44-2 IgG2a Serotec CD1c M241 IgG1 ANC CD45 HI30 IgG1 PM CD2 T11 IgG1 C CD45RA 2H4 IgG1 C CD3 Leu-4 IgG1 BD CD45RO UCHL1 IgG2a IM CD3 M2AB IgG1 EXA CD49d HP2/1 IgG1 IM CD3 UCHT1 IgG1 PM CD49e SAM1 IgG2b IM CD4 Leu-3a IgG1 BD CD50 HP2/19 IgG2a IM CD4 7E14 IgG1 EXA CD54 Leu-54 IgG2b BD CD5 T1 IgG2a C CD54 D3.6 IgG2b EXA CD5 M28623 IgG2a EXA CD56 Leu-19 IgG1 BD CD7 3A1 IgG2b C CD61 SZ21 IgG1 IM CD7 G34 IgG2a EXA CD62L TQ-1 IgG1 C CD8 T8 IgG1 C CD62P CLBThromb/6 IgG1 IM CD9 C3-3A2 IgG1 ANC CD64 10.1 IgG1 PM CD10 J5 IgG2a C CD68 KP-1 IgG1 Dako CD11a HI111 IgG1 PM CD70 BU69 IgG1 ANC CD11b ICRF(44) IgG1 PM CD71 LO1.1 IgG2a BD

CD11b Mo1 IgM C CD72 J4-117 IgG2a PM Downloaded from CD11c Leu-M5 IgG2b BD CD80 BB-1 IgM ANC CD11c 3.9 IgG1 Serotec CD83 HB15a IgG2b IM CD13 WM15 IgG1 PM CD86 IT2.2 IgG2b PM CD14 Leu-M3 IgG2b BD CD90 5E10 IgG1 PM CD14 FWKW-1 IgG2b EXA CD95(FAS) UB2 IgG1 MBL CD14 UCHM1 IgG2a ANC CD103 2G5 IgG2a IM CD15 80H5 IgM IM CD114(G-CSFR) LMM741 IgG1 PM CD15s 2H5 IgM PM CD116(GM-CSFR) M5D12 IgM PM http://www.jimmunol.org/ CD16 Leu-11a IgG1 BD CD117(C-KIT) 95C3 IgG1 IM CD16 J5511 IgG1 EXA CD121a(IL-1R 1) 6B5 IgG2a PM CD16 3G8 IgG1 PM CD122(IL-2R␤) TIC-1 IgG1 END CD18 L130 IgG1 BD CD123(IL-3R␣) 9F5 IgG1 PM CD19 B4 IgG1 C CD130 AM64 IgG1 PM CD19 1G9 IgG1 EXA CD135(Flt3) SF1.340 IgG1 IM CD19 HIB19 IgG1 PM CD152(CTLA-4) BNI3 IgG2a IM CD20 B1 IgG2a C CD154(CD40L) 24-31 IgG1 ANC CD21 BL13 IgG1 IM E-cadherin HECD-1 IgG1 Takara

CD23 9P25 IgG1 IM CLA HECA-452 RatIgM PM by guest on October 2, 2021 CD25(IL-2R␣) 2A3 IgG1 BD Lag IgG1 Dr. Imamura CD27 M-T271 IgG1 ANC HLA-DR L243 IgG2a BD CD28 KOLT-2 IgG1 Nichirei HLA-DR Tu36 IgG2b PM CD29 4B4 IgG1 C HLA-DQ TU169 IgG2a PM CD30 BerH8 IgG1 PM HLA-DQ 1a3 IgG2a LET CD32 AT10 IgG1 Serotec Glycophorin A KC16 IgG1 IM CD33 My9 IgG2b C TCR-␣␤ WT31 IgG1 BD CD33 Leu-M9 IgG1 BD TCR-␥␦ 11F2 IgG1 BD CD34 HPCA-2 IgG1 BD CD34 QBEnd10 IgG1 IM CD35 AR2 IgG1 EXA CD36 FA6.152 IgG1 IM CD38 Leu-17 IgG1 BD CD38 T16 IgG1 IM CD40 5C3 IgG1 PM CD41 SZ22 IgG1 IM

Abbreviations: TAG, BioSource International, Tago Products (Flynn Road, CA); IM, Immunotech (Marseille, France); ANC, Ancell (Bayport, MN); C, Coulter Immunology (Hialeah, FL); BD, Becton Dickinson; EXA, Exalpha (Boston, MA); PM, PharMingen; Serotec, Serotec (Kindlington, U.K.); Nichirei (Tokyo, Japan); Dako (Glostrup, Denmark); MBL, Medical & Biological Laboratories (Nagoya, Aichi, Japan); END, Endogen (Woburn, MA); Takara, Takara Biomedicals (Kusatsu, Shiga, Japan); LET, Leinco Tech- nologies (Manchester, MO).

E-cadherin, Lag Ag, and Birbeck granules (24). We show in this cording to Reddy et al. (30). Brieﬂy, blood mononuclear cells were layered study, however, that the most direct precursors of LCs in human onto human ␥-globulin-coated plates for 1 h, and nonadherent cells were blood are a distinctive subset of CD1aϩ/CD11cϩ/CD14Ϫ DCs. washed away. The dish-adherent cells were incubated in fresh medium for 24 h. The supernatant was harvested and frozen at Ϫ20°C until use. The following recombinant human cytokines were purchased from Boehringer Materials and Methods Mannheim (Indianapolis, IN): GM-CSF (used at a concentration of 100 ng/ml), TNF-␣ (2.5 ng/ml), TGF-␤1 (1 ng/ml), IL-3 (10 ng/ml), and IL-4 Media and reagents (50 ng/ml). The sources of mAb used in our studies are listed in Table I. The following culture medium was used throughout experiments: RPMI FITC-, PE-, PerCP-, PE Cy5-, or biotin-labeled isotype-matched controls 1640 supplemented with 2 mM L-glutamine, 100 U/ml penicillin, 100 were purchased from Becton Dickinson (Sunnyvale, CA) or PharMingen ␮g/ml streptomycin, and heat-inactivated 10% human AB serum (ICN Bio- (San Diego, CA). When unlabeled mAb or biotinylated mAbs were used as Ј medicals, Aurora, OH). Monocyte-conditioned medium was prepared ac- the primary reagents, PE-labeled goat anti-mouse IgG F(ab )2 (Jackson The Journal of Immunology 1411

FIGURE 1. Detection and puriﬁcation of DC fractions in the peripheral blood. DC fractions identiﬁed by four-color EPICS ELITE cell sorter (A and B). A DC-enriched populationwasstainedwithFITC-labeledanti- CD1a mAb, PE-labeled anti-CD11c mAb, PerCP-labeled anti-HLA-DR mAb, and biotinylated mAbs against lineage markers (CD3, CD7, CD14, CD16, CD19), followed by streptavidin-Red 613. Three fractions of DC (B), detected as CD1aϩ/CD11cϩ (fraction 1), CD1aϪ/CD11cϩ (fraction 2), and CD1aϪ/CD11cϪ (fraction 3), are present in the linϪ/HLA-DRϩ fraction (designated as R in A). These three fractions of DCs were also obtained by a two-color FACStar cell sorter (C and D) (details described in Materials and Methods). Fraction 1 was sorted as linϪ/ CD1aϩ cells after staining with PE (or FITC)-labeled mAbs against lineage markers

plus FITC (or PE)-labeled anti-CD1a mAb Downloaded from (R1 in C). Fractions 2 and 3 were puriﬁed as linϪ/CD1aϪ/CD11cϩ cells (R2 in D) and linϪ/CD1aϪ/CD11cϪ cells (R3 in D), respectively, after being stained with FITC-labeled mAbs against lineage markers/CD1a plus PE-labeled anti-CD11c mAb. Figures representative of more than 10 experiments http://www.jimmunol.org/ are shown.

ImmunoResearch, West Grove, PA), FITC-labeled goat anti-mouse IgG Enrichment and analyses of peripheral blood DCs Ј F(ab )2 (Becton Dickinson), FITC-labeled polyclonal anti-rat Ig (Phar- Mingen), or tri-color-labeled streptavidin (streptavidin-RED670; Life PBMC were isolated by Lymphoprep (Nycomed Pharma, Oslo, Norway) Technologies, Glasgow, U.K.) were employed as secondary reagents. The gradient centrifugation of heparinized blood (50–60 ml) obtained from mAbs against Langerin (DCGM4) (31) and Lag-1 (32) were used to de- each healthy volunteer. PBMC were incubated with anti-CD3 (HIT3a) and termine the differentiation into LCs. anti-CD14 (M5E2) mAbs (both from PharMingen) for 30 min on ice, and by guest on October 2, 2021 A cell permeabilization kit, FIX & PERM (Caltag, Burlingame, CA), cells binding these mAb were removed using sheep anti-mouse Ig-coated was used for intracellular staining. magnetic beads (M-450; Dynal, Oslo, Norway). CD3Ϫ/CD14Ϫ cells were

FIGURE 2. Morphology of DC fractions. Freshly prepared or cultured peripheral blood DC fractions were histologically examined by light or electron microscopy. Monocyte-conditioned medium was used for the cultures of fractions 1 and 2. IL-3-supplemented medium was used for fraction 3. A–C, Fraction 1; D–F, fraction 2; G–I, fraction 3. A, D, and G, Freshly prepared DCs stained by May-Giemsa solution (ϫ250). B, E, and H, Freshly prepared DCs examined by EM (ϫ3000). C, F, and I, DCs, cultured for 3 days, stained by May-Gi- emsa solution (ϫ250). No Birbeck granules were observed in any fractions under these conditions. Results are representative of eight experiments. 1412 LC PRECURSORS IN HUMAN PERIPHERAL BLOOD Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021 The Journal of Immunology 1413 further incubated with CD4-conjugated microbeads (Miltenyi Biotec., Ber- Epon. Ultrathin sections were stained with lead citrate and uranyl acetate, gisch Gladbach, Germany), and the CD4ϩ cells were then enriched through and studied using a Hitachi H-600a electron microscope (Hitachi, Ibaragi, Mini MACS magnetic separation column (Miltenyi Biotec.). Using this Japan). protocol, the percentage of DCs (Ͻ1% of total PBMC) increased to 20– 50%, depending on the individuals (n Ͼ 30). The resultant DC-enriched population (CD4ϩ/CD3Ϫ/CD14Ϫ cells) was stained with PE-labeled anti- Mixed leukocyte reaction CD11c (LeuM5), FITC-labeled anti-CD1a (BB-5), PerCP-labeled Freshly prepared blood DCs, cultured DCs (in the presence of GM-CSF HLA-DR (L243), and a mixture of biotinylated mAbs against lineage ϩ markers (CD3; M2AB, binding to a different determinant from that recog- and IL-3), or blood monocytes (purified by magnetic beads as CD14 cells) were ␥ irradiated at 15 Gy, and graded doses were then added to 2 ϫ nized by the previous anti-CD3 mAb, CD7; G34, CD14; UCHM1, binding 5 ϩ to a different determinant from that recognized by the previous anti-CD14 10 allogeneic T cells (collected by magnetic beads as CD3 cells) in mAb, CD16; 3G8 and CD19; HIB19), followed by RED613-streptavidin 96-well flat-bottom culture plates for 6 days. The cells were pulsed with 1 ␮ 3 (Life Technologies). The stained cells were analyzed (or sorted for Ci of [ H]TdR during the last 16 h of the culture period. cytologic assay, T cell proliferation assay, and analyses of endocytosis) by an EPICS ELITE flow cytometer (Coulter, Hialeah, FL). Consequently, three phenotypically distinct fractions of DCs were found: CD1aϩ/ Cell cycle and viability assays ϩ Ϫ ϩ Ϫ ϩ Ϫ ϩ CD11c /lin /DR (fraction 1), CD1a /CD11c /lin /DR (fraction 2), 5 Ϫ Ϫ Ϫ ϩ For cell cycle analyses, 10 cells were suspended in 70% ethanol at 4°C for and CD1a /CD11c /lin /DR cells (fraction 3). 1 h, followed by resuspension in 500 ␮l PBS, 250 ␮l RNase (1 mg/ml; Sigma, St. Louis, MO), and 250 ␮l propidium iodide (PI) (100 ␮g/ml; Purification and characterization of DC fractions Molecular Probes, Eugene, OR). Cell viability was determined using An- The DC fractions were obtained also by a two-color FACStar cell sorter nexin V-FITC Apoptosis Detection Kit (Genzyme, Cambridge, MA). The (Becton Dickinson), as follows. 1) Purification of CD1aϩ/CD11cϩ DCs stained cells were analyzed with a FACScan.

(fraction 1): after staining the DC-enriched population with PE (or FITC)- Downloaded from labeled mAbs against lineage markers plus FITC (or PE)-labeled anti- CD1a mAb, CD1aϩ/CD11cϩ DCs were sorted as CD1aϩ/linϪ cells (note ϩ Ϫ ϩ ϩ Ϫ Results that all CD1a /lin DCs are quite comparable with CD1a /CD11c /lin Identification of three fractions of DCs in peripheral blood DCs when analyzed after staining with anti-CD11c mAb). 2) Purification of CD1aϪ/CD11cϩ (fraction 2) and CD1aϪ/CD11cϪ (fraction 3) DCs: the DC-enriched populations prepared from the peripheral blood DC-enriched population was stained with FITC-labeled mAbs against lin- (CD4ϩ/CD3Ϫ/CD14Ϫ cells) were four-color analyzed after stain- eage markers and CD1a plus PE-labeled anti-CD11c (LeuM5), and the linϪ/CD1aϪ/CD11cϩ and linϪ/CD1aϪ/CD11cϪ fractions were then col- ing with PE-labeled anti-CD11c, FITC-labeled anti-CD1a, PerCP- http://www.jimmunol.org/ lected using a FACStar as fractions 2 and 3, respectively. These fractions labeled HLA-DR, and a mixture of biotinylated mAbs against lin- Ϫ ϩ collected using the FACStar were phenotyped. Purity of the sorted cells eage markers, followed by RED613-streptavidin. lin /HLA-DR was always greater than 96% by reanalysis using a FACScan (Becton cells (DCs) (Fig. 1A) were of medium cell size (between lympho- Dickinson). The sorted fractions, thus prepared, were then stained with ϩ cytes and monocytes) by light scatter profiles. Two distinct pop- PerCP-labeled HLA-DR. The cells bearing HLA-DR (DR ) were gated as Ϫ ϩ DCs, and further analysis was conducted using a FACScan after staining ulations of lin /DR were observed with respect to the expression with a panel of mAbs (listed in Table I) conjugated with FITC or PE. of CD11c (Fig. 1B). Furthermore, CD1a (detected by mAb clone Before staining, the cells were incubated with an excess amount of unla- BB-5)-positive cells were newly observed as a blood DC fraction beled polyclonal human or mouse Ig to block nonspecific binding of (Fig. 1B). Therefore, three fractions of DCs were identified with

labeled mAbs. ϩ ϩ Ϫ ϩ by guest on October 2, 2021 the phenotypes of CD1a /CD11c /lin /DR (fraction 1), Ϫ ϩ Ϫ ϩ Ϫ Ϫ Culture of DCs CD1a /CD11c /lin /DR (fraction 2), and CD1a /CD11c / linϪ/DRϩ cells (fraction 3). CD1a has been detected only on epi- The sorted DCs were cultured in 96-well flat-bottom tissue culture plates at ϫ 5 dermal LCs (8, 34) and dermal DCs (10) in vivo (or on DCs de- 1 10 cells/well in medium supplemented with monocyte-conditioned ϩ medium (final concentration 50% v/v) for 1–6 days. In the culture of rived from CD34 progenitors (15, 16) or monocyte-derived DCs CD1aϪ/CD11cϪ DCs (fraction 3), medium supplemented with IL-3 was (23) in the presence of cytokines in vitro), but not on freshly iso- used. In some experiments, a mixture of cytokines such as GM-CSF ϩ lated blood DCs (27, 28). CD1aϩ DCs were also observed when a TGF-␤1, GM-CSF ϩ IL-4, GM-CSF ϩ TNF-␣, GM-CSF ϩ TNF-␣ ϩ TGF-␤1, or GM-CSF ϩ IL-4 ϩ TGF-␤1 was used. Half of the medium different anti-CD1a mAb (BL-6) was used (data not shown). The was removed and replenished every 2 days. overall intensity of CD1a expression, however, was lower than that on CD34ϩ progenitor-derived DCs (data not shown) when deter- Analysis of endocytosis mined by either BB-5 or BL-6 Ab. ϩ ϩ To assess endocytosis of DC fractions, FITC-dextran (Polysciences, War- The percentage of CD1a /CD11c DCs was similar to that of rington, PA) was used according to the method described previously (33). CD1aϪ/CD11cϪ DCs, and that of CD1aϪ/CD11cϩ DCs was 1/3 to Briefly, the cells were incubated with 0.1 mg/ml FITC-dextran at 37°C for 1/10 that of CD1aϩ/CD11cϩ DCs. The frequency of DC fractions, 15 or 45 min. The results were displayed as mean fluorescence intensity ϩ ϩ Ϫ ϩ Ϫ CD1a /CD11c DCs, CD1a /CD11c DCs, and CD1a / after subtracting the background in which cells were incubated with FITC- Ϫ dextran at 4°C. CD11c DCs, ranged from 0.56 to 0.18%, 0.14 to 0.03%, and 0.50 to 0.13% of PBMCs, respectively, when more than 30 healthy Cytologic assays volunteers were examined. These three fractions of blood DCs Cells were cytocentrifuged onto a slide and stained with May-Giemsa so- were usually purified by a FACStar, as described in Materials and lution, or by mAb to CD68 or Lag, then visualized using the ABC method Methods. Region 1 (R1) in Fig. 1C shows the sorting gate of frac- using DAKO LSAB (labeled streptavidin biotin) kit plus hematoxylin and tion 1 as linϪ/CD1aϩ cells (note that all CD1aϩ cells are CD11cϩ, diaminobenzidine. as shown in Fig. 1B), and R2 and R3 in Fig. 1D represent the For electron-transmission microscopy, cells were fixed with 2.5% glu- Ϫ Ϫ ϩ Ϫ taraldehyde in 0.1 M cacodylate buffer (pH 7.4) and postfixed with 1% sorting gates of lin /CD1a /CD11c cells (fraction 2) and lin / Ϫ Ϫ OsO4. After dehydration with graded ethanol, they were embedded in CD1a /CD11c cells (fraction 3), respectively.

FIGURE 3. Expression of surface molecules on freshly isolated fractions of blood DCs. Freshly prepared blood DC fractions (puriﬁed by a FACStar) were stained with anti-HLA-DR mAb and the panel of mAbs. The DRϩ cells were gated and analyzed by a FACScan. Expression of surface molecules is shown as shaded histograms. Open histograms in the ﬁgures represent negative controls stained with isotype-matched control Abs. Results shown are representative of at least three experiments. 1414 LC PRECURSORS IN HUMAN PERIPHERAL BLOOD Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 4. Expression of surface molecules on cultured blood DC fractions. Cultured blood DC fractions were stained with anti-HLA-DR mAb and a panel of mAbs. The DRϩ cells were analyzed by a FACScan. The three fractions of DCs were cultured for 3 days with monocyte-conditioned medium for fractions 1 and 2, and IL-3-supplemented medium for fraction 3. Results shown are representative of at least three experiments.

Morphology of peripheral blood DC fractions bodies and vesicles, in contrast to typical DCs. After 3-day culture, Freshly prepared or cultured (for 3 days in monocyte-conditioned both fractions showed typical DC features such as long cell pro- medium or IL-3-supplemented medium) DC fractions were exam- cesses, large cell size (Fig. 2, C and F), and a prominent tubular- ined by light and electron microscopes (EM). Freshly sorted vesicular system (data not shown). Ϫ Ϫ CD1aϩ/CD11cϩ DCs (fraction 1) (Fig. 2, A and B) and CD1aϪ/ In contrast to fractions 1 and 2, freshly isolated CD1a /CD11c CD11cϩ DCs (fraction 2) (Fig. 2, D and E) were found to be DCs (fraction 3) displayed lymphoplasmacytoid morphology; monocyte like, characterized by short cell processes, indented nu- that is, a medium size with round nucleus, cytoplasm-contain- cleus, high nuclear-cytoplasmic ratio, and few cytoplasmic dense ing juxtanuclear Golgi apparatus, and parallel arrays of rough The Journal of Immunology 1415

FIGURE 6. Analysis of endocytosis. Fractions of DCs were incubated with 0.1 mg/ml FITC-dextran at 37°C for 15 or 45 min. After washing three times, uptake of FITC-dextran was analyzed by a FACScan. The results are shown as mean ﬂuorescence intensity, from which the background ﬂuorescence (incubated at 4°C) is subtracted. CD34ϩ progenitor- ϩ

FIGURE 5. Stimulatory activity of blood DC fractions. Graded doses of derived DCs were obtained by the culture of CD34 progenitor cells col- Downloaded from stimulator cells (freshly sorted DC fractions or cultured DC fractions in the lected from the human umbilical cord blood samples (according to presence of GM-CSF and IL-3 for 3 days) were irradiated and cocultured institutional guidelines) for 12 days in the presence of GM-CSF and TNF-␣ ϩ with 2 ϫ 105 allogeneic CD3 T cells for 6 days. Cells were then pulsed (16, 17). Fraction 1 (F), fraction 2 (f), fraction 3 (E), monocytes (Œ), and ϩ with 1 ␮Ci of [3H]TdR during the last 16 h of culture. CD14 monocytes CD34ϩ progenitor-derived DCs (Ⅺ). Results shown are representative of (purified using magnetic beads) were also prepared and used as stimulators. three experiments. Dashed lines represent freshly prepared DCs and straight lines show cultured DCs. Fraction 1 (F), fraction 2 (f), fraction 3 (E), and monocytes http://www.jimmunol.org/ Œ ϩ ϩ ( ). Vertical bars represent SDs of triplicate cultures. Representative data Fraction 1. CD1a /CD11c DCs expressed CD33 and CD13 of three experiments are shown. (myeloid-associated markers), CD2high, CD1c, CD49e, CD95 (Fas), CD45RO, CD32/CD64 (Fc␥ receptors), and cytokine recep- endoplasmic reticulum (Fig. 2, G and H). These cells correspond tors (GM-CSFR (CD116) and IL-3R␣ (moderately positive)). Fur- to plasmacytoid T cells within the T cell zone of lymphoid organs, thermore, they dimly expressed CD5 and CD11b. After culture, as recently characterized by Grouard et al. (14). After culture with CD11b was up-regulated, and CD32/64 was down-regulated. IL-3, they showed a typical DC morphology (Fig. 2I) with marked CD62L became undetectable during culture in contrast to freshly decreases in rough endoplasmic reticulum (data not shown), and prepared cells. On the basis of the expressions of CD2, CD9, by guest on October 2, 2021 this process was dependent on the presence of IL-3. No Birbeck CD11b, CD11c, CD13, CD32, CD33, CD64, and GM-CSFR, frac- granules were found in either of the three presently identified fraction 1 is thought to be closely associated with the monocyte lin- tions of DCs before or after culture with monocyte-conditioned eage. Actually, the phenotype of fraction 1 was quite similar to that medium or IL-3 (data not shown). of blood monocyte-derived DCs and that of CD34ϩ progenitor- derived DCs (but not LC type) (16). Phenotype of peripheral blood DC fractions Fraction 3. In contrast to fraction 1, CD1aϪ/CD11cϪ DCs ex- The fluorometrical characterization of the three fractions of DCs pressed dimly or little CD33, CD13, CD1c, CD2, CD49e, was next conducted. The surface phenotype of the three DC frac- CD45RO, CD32, CD64, but were significantly positive for tions before and after culture is highlighted in Figs. 3 and 4. The CD45RA. In comparison with the other two fractions, they were common features of these three fractions are as follows: none of highly positive for CD4, but moderately positive for HLA-DR/DQ. the DC fractions expressed lineage markers for T cells (CD8, TCR, They also possessed IL-3R␣ (brightly positive), but not GM- Thy-1/CD90), B cells (CD10, CD20), NK cells (CD56), mono- CSFR. Fraction 3 brightly expressed intracytoplasmic CD68, but cytes (CD14), granulocytes (CD15, CD35), or erythrocytes (gly- the other two fractions were only weakly positive (data not cophorin A) before or after culture. In addition, they possessed shown). Therefore, fraction 3 is closely related to previously re- neither CD27 and CD30 (TNF receptor families) nor CD70 (ligand ported plasmacytoid T cells, not only morphologically but also for CD27). IL-1R type I (CD121a), G-CSFR (CD114), IL-2R␤- according to phenotypical features. Furthermore, freshly prepared chain (CD122), and IL-6R␤ (CD130) were negative on all freshly fraction 3 had no CD95. However, CD95 was induced during prepared DC fractions (data not shown). All fractions clearly ex- 3-day culture with IL-3, whereas the expression of CD62L re- pressed MHC class II, CD4, CD9, CD36, CD38, CD45, and ad- mained unchanged. When compared with culture in medium alone, hesion molecules (CD11a, CD15s, CD18, CD29, CD44, CD49d, this fraction exhibited increased levels of CD83 and HLA-DR/DQ CD50, and CD54). Costimulatory molecules (CD80, CD86, and in the presence of IL-3 (data not shown). CD40) were dimly expressed on all fractions (fraction 3 little ex- Fraction 2. CD1aϪ/CD11cϩ DCs were a minor population, and pressed CD40), and were up-regulated after culture. Activation uniformly expressed CD11c and CD33, like fraction 1. However, molecules CD25 and the DC-associated marker CD83 were also they were found to be heterogeneous in the expression of some detected on all fractions after overnight culture, but not on freshly surface molecules such as CD2, CD13, CD45RA, CD45RO, IL- isolated cells. In addition, CLA (a skin-homing molecule (35), 3R␣, and GM-CSFR, although they were morphologically homo- which is constitutively expressed on LCs (36)), and L-selectin geneous. Fraction 2, in contrast to fraction 1, had no CD1a, CD1c, (CD62L, known to be a lymph node-homing molecule (37)) were CD11b, or CD64. However, they had CD11c, CD33, CD13, and present on all fresh DC fractions. The characteristic features of GM-CSFR, indicating that they were monocyte-lineage cells. each fraction are as follows. Based on the expression of the CD2 molecule, fraction 2 could be 1416 LC PRECURSORS IN HUMAN PERIPHERAL BLOOD Downloaded from http://www.jimmunol.org/ by guest on October 2, 2021

FIGURE 7. Differentiation of fraction 1 DCs to LCs in the presence of GM-CSF, IL-4, and TGF-␤1. A, Induction of LC-speciﬁc molecules on fraction 1 DCs. DCs (fraction 1) and CD14ϩ monocytes were cultured in the presence of GM-CSF, IL-4, and TGF-␤1 for 6 days, and cells harvested daily and stained with anti-E-cadherin, anti-Langerin (DCGM-4), or anti-CD1a (BB-5) and anti-HLA-DR mAbs. The DRϩ cells were gated and analyzed by a FACScan. The expression of E-cadherin, Langerin, or CD1a is shown as shaded histograms. Note that E-cadherin and Langerin were induced on fraction 1 DCs already after 1 day of culture. Open histograms represent negative controls stained with isotype-matched control Abs. B, Histocytochemical examination of Lag in fraction 1 DCs cultured in GM-CSF, IL-4, and TGF-␤1. Cultures were cytocentrifuged and stained with mAb against Lag. Original magniﬁcation: ϫ250. Results shown are representative of three experiments.

divided into two minor subfractions: CD2high and CD2low DCs. Functional characterization of peripheral blood DC fractions After purifying these two populations and staining them with a The stimulatory activity for allogeneic T cells was compared low panel of mAbs, we found that the former were CD13 / among the three DC fractions, as shown in Fig. 5. The order in the ϩ Ϫ ϩ ␣ϩ CD45RA /CD45RO /GM-CSFR (bright)/IL-3R (bright). On stimulatory activity of freshly prepared fractions was as follows: low high ϩ the other hand, CD2 DCs were CD13 /CD45RA (dim)/ fraction 1 ϭ fraction 2 Ͼ fraction 3, being dependent on the order ϩ ϩ ␣ϩ CD45RO /GM-CSFR (dim)/IL-3R (dim). Therefore, both of the expression of MHC class II molecules. The stimulatory ac- subpopulations in fraction 2 seem to be related (but not completely tivities became stronger after culture for 3 days with GM-CSF plus matched) to fractions 1 and 3 on the basis of the expression of IL-3 (note that fractions 1 and 2 are dependent on GM-CSF, and surface molecules. The physiological features of the heterogeneous fraction 3 is dependent on IL-3), indicating that they become fraction are unknown at present. mature DCs. Ϫ Finally, CD1a molecules were not induced on the CD1a cells The capacity of the DC fractions for endocytosis was examined even when both fraction 2 and fraction 3 were cultured with mono- by uptake of FITC-dextran. As shown in Fig. 6, freshly isolated cyte-conditioned medium or IL-3-supplemented medium, respec- fractions 1 and 2 had a potent ability to endocytose FITC-dextran tively (Fig. 4), or with any combinations of cytokines examined during a 45-min incubation compared with CD34ϩ progenitor- (GM-CSF, IL-3, TNF-␣, and TGF-␤1) (data not shown). derived mature DCs, although this capacity was less than The Journal of Immunology 1417

FIGURE 8. Morphology of fraction 1 DCs cultured for 6 days in the presence of GM-CSF, IL-4, and TGF- ␤1. Typical Birbeck granules (arrow and inset) are observed (ϳ20–25% of cultured DCs) by EM examina- tions. Original magnification: left, ϫ6,000; right, ϫ12,000; inset, ϫ20,000. Downloaded from monocytes. The endocytic activity of fractions 1 and 2 decreased cultured with GM-CSF ϩ IL-4 ϩ TGF-␤1 or other cytokines after culture for 3 days with monocyte-conditioned medium (data tested (GM-CSF, IL-3, TNF-␣, and TGF-␤1) (data not shown). not shown). Fraction 3, which had been freshly isolated or cultured Furthermore, induction of LC-specific molecules was not observed (with IL-3), showed little FITC-dextran uptake. in CD14ϩ monocytes during culture with the same cytokine com- ϩ ϩ ␤ ϩ ϩ bination (GM-CSF IL-4 TGF- 1). On day 6, E-cadherin and CD1a /CD11c peripheral blood DCs rapidly differentiate CD1a became positive, as has been reported by Geissmann et al. http://www.jimmunol.org/ into LCs (24); however, the more sensitive Lag and Langerin markers were It should be noted that no Birbeck granules were found in CD1aϩ/ not detected in our culture condition (Fig. 7). Furthermore, when CD11cϩ DCs (fraction 1) before or after culture with monocyte- we examined the frequency of dividing cells (evaluated by PI conditioned medium. To further characterize fraction 1, we com- staining) and viable cells (evaluated by PI and annexin V staining) pared the effects of combinations of cytokines (GM-CSF ϩ TGF- before and after the culture, the percentages of dividing cells were ␤1, GM-CSF ϩ IL-4, GM-CSF ϩ TNF-␣, GM-CSF ϩ IL-4 ϩ less than 1% (before and after 1-, 3-, and 6-day culture), and the TGF-␤1, and GM-CSF ϩ TNF-␣ ϩ TGF-␤1) on their maturation percentages of viable cells were more than 90% throughout the and differentiation. The expression of E-cadherin, the LC-re- culture period (data not shown). In addition, 15 Gy irradiation did by guest on October 2, 2021 stricted Langerin (detected by mAb DCGM-4), and Lag Ag was not alter LC differentiation from fraction 1 (data not shown). These kinetically examined. In the presence of GM-CSF ϩ IL-4 ϩ TGF- findings clearly demonstrate that 1) fraction 1 has the capacity to ␤1, E-cadherin and Langerin were detected (Fig. 7A) already after differentiate into LCs without cell proliferation, and 2) fraction 1 is 1 day of culture, and intracellular Lag was also slightly positive the direct/immediate precursor of LCs, compared with monocytes. both in FACS (data not shown) and cytochemical (Fig. 7B) exam- As shown in Fig. 9, 3 days after culture in the presence of inations. Expression of the LC-specific molecules increased at 3 GM-CSF, IL-4, and TGF-␤1, fraction 1 DCs expressed higher in- days, as shown in Fig. 7. Finally, fraction 1 showed DC morphol- tensities of cutaneous lymphocyte-associated Ag and CD1a, with ogy, and typical Birbeck granules were observed 6 days after cul- lower HLA-DR, CD54, CD40/CD80/CD86, and CD83 (mostly ture (Fig. 8). In addition, the intensity of CD1a expression in- negative, a few positive) than those cultured with monocyte-con- creased concomitantly (Fig. 7A). ditioned medium. Activation-associated markers, CD25 and It should be noted that, regardless of the time point examined, CD71, were not induced, and CD2 was down-regulated in contrast differentiation to LCs from fraction 1 was only observed with the to the cells before culture. Furthermore, CD23 (Fc⑀RII) was in- combination of GM-CSF ϩ IL-4 ϩ TGF-␤1, but not with any duced on fraction 1 under this condition (but not with monocyte- cytokine alone nor any combinations tested. Of importance, nei- conditioned medium), as reported by Schmitt et al. (38) on LCs. ther fraction 2 nor 3 had the capacity to differentiate into LCs when These findings suggest that the combination of IL-4, GM-CSF, and

FIGURE 9. Phenotype of fraction 1 DCs after culture with GM-CSF, IL-4, and TGF-␤1. Fraction 1 DCs were cultured with GM-CSF, IL-4, and TGF-␤1 for 3 days, and the cells were harvested and stained with the mAbs listed in the ﬁgure. The DRϩ cells were gated and analyzed by a FACScan. Shaded histograms represent expression of the surface molecules, and open histograms represent isotype controls. Representative results of three experiments are shown. 1418 LC PRECURSORS IN HUMAN PERIPHERAL BLOOD

TGF-␤1 is essential for differentiation of fraction 1 into LCs rather additional presence of TGF-␤1. We did not observe acquisition of than terminal maturation and/or activation. a complete LC phenotype from CD14ϩ monocytes in our culture Our results represent the first observation showing that LCs can condition (Fig. 7). Furthermore, the differentiation of LCs from originate from DCs circulating in peripheral blood, as their direct fraction 1 occurred during a very short-term culture. In addition, precursors. the possibility that a dividing or a distinct subpopulation in fraction 1 selectively differentiates into LCs was ruled out by the following: there was Ͻ1% of dividing cells, there was a low cell mortality Discussion rate, and a low rate of apoptosis throughout the short-term culture. We have identified three fractions of DCs in the peripheral blood Therefore, fraction 1 is a distinct direct/immediate precursor of using a panel of mAb. Two of these fractions, fraction 1 (CD1aϩ/ LCs in the peripheral blood, and might be an intermediate stage CD11cϩ DCs) and fraction 2 (CD1aϪ/CD11cϩ DCs), showed between monocytes and LCs. more potent APC activity and monocyte-like morphology. Frac- LCs are myeloid-lineage DCs that originate from the bone mar- tion 3 (CD1aϪ/CD11cϪ DCs) displayed poor endocytic activity row. However, little is known about their migration/differentiation and a lymphoplasmacytoid morphology. All fractions were con- pathway. Our results might help elucidate this question, as fraction firmed to exhibit dendritic morphology and acquired CD83 during 1 is the direct precursor of LCs (in response to GM-CSF ϩ IL-4 ϩ culture. TGF-␤1) and a bipotent cell that can be induced to differentiate As previously reported, less mature APCs in the blood, with into non-LCs such as dermal DCs (in response to monocyte-con- little expression of CD11c and CD33, have been speculated to be ditioned medium). This indicates that common precursors for the on the way from the bone marrow to nonlymphoid tissues such as LC lineage and the monocyte-associated DC lineage are present in the skin, and more mature potent APCs bearing CD11c and CD33 the peripheral blood. It can be speculated that part of this fraction Downloaded from are thought to be en route to the lymph nodes from the skin (27). migrates to the epidermis to differentiate into LCs, and that the When freshly isolated, these DCs display immature dendritic mor- other migrates to nonlymphoid tissues such as the dermis and dif- phology and express CD4 (5, 27), but not CD83 (26, 39), in agree- ferentiate into dermal DCs. This process might be dependent upon ment with our results (Figs. 2 and 3). Notably, DCs in the blood cytokines produced by regional microenvironments. have been found within a CD1a-negative fraction (26–28, 39). Fraction 3 (CD1aϪ/CD11cϪ DCs) is another homogeneous sub- However, we have identified one fraction of DCs as CD1a low- set in the peripheral blood, and distinct from fraction 1, based on http://www.jimmunol.org/ positive cells (fraction 1). The difference in detection of CD1a its morphology, the expression of lineage-associated Ags, lack of between our present results and previous studies could be due to ability to uptake FITC-dextran, and dependency on IL-3. From our the following: 1) in previous studies, T cells were depleted by detailed studies on blood DCs, the features of fraction 3 corre- SRBC-rosette formation, and this caused the depletion of some spond completely to those of plasmacytoid T cells in the T zone of DCs bearing CD1a because this DC population coexpresses CD2 the lymphoid organ (found to be close to or within the high en- (binding site of SRBC) (5, 26, 27); 2) part of the CD1aϩ DC dothelial venule) that could be derived from CD4ϩ/CD11cϪ blood fraction coexpressing CD11b might be depleted along with my- DCs (14). Fraction 3 possesses high levels of the lymph node- elomonocyte-lineage cells when anti-CD11b mAb was used in im- homing molecule; CD62L, and its expression, remains unchanged by guest on October 2, 2021 munoselection (29, 39, 40); 3) the level of CD1a expression on this during culture, although it is down-regulated on fraction 1 (Figs. 3 fraction is much lower than that on the CD34ϩ progenitor-derived and 4). These findings support the possibility that fraction 3 in the DCs or monocyte-derived DCs; therefore, CD1alowϩ DCs were blood may be a precursor pool for plasmacytoid T cells that di- classified as CD1a-negative cells; in this context, Brown et al. rectly migrate to the T zone in the lymphoid tissues via the high reported that CD1a was weakly expressed on a small amount of endothelial venule. Furthermore, fraction 3 requires IL-3, but not blood DCs (41); and 4) different anti-CD1a mAbs employed in GM-CSF, for cell maturation and development, whereas typical detection, this being considered to be a main reason. In our study, myeloid-lineage DCs are known to be dependent on GM-CSF, a difference in the intensity of CD1a was actually observed when indicating that fraction 3 is not in the monocyte-associated DC two anti-CD1a mAbs (clones BB-5 and BL-6) were compared. The lineage. Grouard et al. have pointed out that plasmacytoid T cells CD11cϩ/CD33ϩ DC subset in the blood that O’Doherty et al. (27) in the lymphoid organs are lymphoid-derived DCs for this reason and Thomas et al. (28) previously identified might contain fraction (14). However, fraction 3 also expresses several myelo-monocytic 2 (CD1aϪ/CD11cϩ DCs) and a part of fraction 1 (CD1aϩ/CD11cϩ markers (CD9; Fig. 3, CD36 and CD68; data not shown); there- DCs). In their studies, CD1alowϩ DCs might be classified as CD1a- fore, it is not evident at the moment whether they indeed are lym- negative cells. phoid-derived DCs. This issue will require further studies on these Fraction 1 is morphologically similar to monocytes and, after cells. Our results clearly indicate that fraction 3 is a different lin- culture, the cells resemble monocyte-derived DCs. Furthermore, eage from that previously examined in in vitro differentiation stud- fraction 1 has myelomonocyte-associated Ags, and has the capac- ies of CD34ϩ progenitors to LCs or dermal DCs. ity to uptake FITC-dextran. These findings indicate that fraction 1 Although examination of the minor fraction, fraction 2 (CD1aϪ/ is a homogeneous subset, and most likely a CD14-negative mono- CD11cϩ DCs), has been hampered by the limitation of cells col- cyte-associated or monocyte-derived DC. Strikingly, cells in frac- lected from the peripheral blood, this fraction has similar features tion 1, but not fraction 2 or 3, rapidly became cells with typical LC to fraction 1 in their morphology, allostimulatory activity, and the characteristics in the presence of GM-CSF, IL-4, and TGF-␤1. expression of several myelo-monocytic Ags, suggesting that cells E-cadherin (Fig. 7A) and Lag (Fig. 7B) were detected after short- in this fraction are monocyte-associated or monocyte-derived DCs, term culture, and the newly described LC-specific Ag, Langerin or that they are in an intermediate stage preceding fraction 1. Fur- (31), was concomitantly induced (Fig. 7A). Furthermore, the dif- thermore, fraction 2 is phenotypically similar to recently reported ferentiation of LCs from fraction 1 was confirmed by the appear- DCs located in the germinal center (GCDC) (13); therefore, the ance of Birbeck granules (Fig. 8). DCs bearing CD1a are induced possibility exists that fraction 2 becomes GCDCs, which have been from CD14ϩ blood monocytes after culture with GM-CSF and presumed to be derived from blood CD4ϩ/CD11cϩ DCs. IL-4 (23, 33). Geissmann et al. (24) have recently reported that In summary, we have characterized three fractions of DCs in the blood monocytes differentiate into LCs after a 6-day culture in the human peripheral blood using a panel of mAbs. It is noted that The Journal of Immunology 1419 fraction 1 has the capacity to become LCs. These results suggest 20. Strobl, H., E. Riedl, C. Scheinecker, C. Bello-Fernandez, W. F. Pickl, K. Rappersberger, O. Majdic, and W. Knapp. 1996. TGF-␤1 promotes in vitro that blood DC fractions (subsets) are different not only in their ϩ development of dendritic cells from CD34 hematopoietic progenitors. J. Im- maturational stages, but also in their lineage or differentiation path- munol. 157:1499. ways. Our studies of the DC fractions were conducted using 21. Borkowski, T. A., J. J. Letterio, A. G. Farr, and M.C. Udey. 1996. A role for ␤ healthy volunteers. Therefore, the findings in this study also offer endogenous transforming growth factor 1 in Langerhans cell biology: the skin of transforming growth factor ␤1 null mice is devoid of epidermal Langerhans basic data for the diagnosis of some immune-related diseases. We cells. J. Exp. Med. 184:2417. are now proceeding with experiments related to these points. 22. Strobl, H., C. Bello-Fernandez, E. Riedl, W. F. Pickl, O. Majdic, S. D. Lyman, and W. Knapp. 1997. Flt3 ligand in cooperation with TGF-␤1 potentiates in vitro development of Langerhans type dendritic cells in serum-free medium. Blood Acknowledgments 90:1425. 23. Zhou, L.-J., and T. F. Tedder. 1996. CD14ϩ blood monocytes can differentiate We thank Y. Tokuyama and M. Shinkawa for their skillful technical as- into functionally mature CD83ϩ dendritic cells. Proc. Natl. Acad. Sci. USA 93: sistance, F. Ishida (Central Research Center, Kansai Medical University) 2588. for sorting cells on a FACStar, and K. Ando for manuscript preparation. 24. Geissmann, F., C. Prost, J.-P. Monnet, M. Dy, N. Brousse, and O. Hermine. 1998. We thank Dr. S. Imamura and Dr. K. Yoneda (Faculty of Medicine, Kyoto Transforming growth factor ␤1, in the presence of granulocyte/macrophage col- University) for anti-Lag 1 Ab. ony-stimulating factor and interleukin 4, induces differentiation of human peripheral blood monocytes into dendritic Langerhans cells. J. Exp. Med. 187:961. 25. Freudenthal, P. S., and R. M. Steinman. 1990. The distinct surface of human References blood dendritic cells, as observed after an improved isolation method. Proc. Natl. Acad. Sci. USA 87:7698. 1. Steinman, R. M. 1991. The dendritic cell system and its role in immunogenicity. 26. Zhou, L.-J., and T. F. Tedder. 1995. Human blood dendritic cells selectively Annu. Rev. Immunol. 9:271. express CD83, a member of the immunoglobulin superfamily. J. Immunol. 154: 2. Banchereau, J., and R. M. Steinman. 1998. Dendritic cells and the control of 3821. immunity. Nature 392:245. Downloaded from 3. Inaba, K., J. P. Metlay, M. T. Crowley, and R. M. Steinman. 1990. Dendritic cells 27. O’Doherty, U., M. Peng, S. Gezelter, W. J. Swiggard, M. Betjes, N. Bhardwaj, pulsed with protein antigen in vitro can prime antigen-specific, MHC-restricted T and R. M. Steinman. 1994. Human blood contains two subsets of dendritic cells, cells in situ. J. Exp. Med. 172:631. one immunologically mature and the other immature. Immunology 82:487. 4. Bjo¨rck, P., L. F.-Romo, and Y.-J. Liu. 1997. Human interdigitating dendritic cells 28. Thomas, R., and P. E. Lipsky. 1994. Human peripheral blood dendritic cell sub- directly stimulate CD40-activated naive B cells. Eur. J. Immunol. 27:1266. sets: isolation and characterization of precursor and mature antigen-presenting 5. O’Doherty, U., R. M. Steinman, M. Peng, P. U. Cameron, S. Gezelter, cells. J. Immunol. 153:4016. I. Kopeloff, W. J. Swiggard, M. Pope, and N. Bhardwaj. 1993. Dendritic cells 29. Fanger, N. A., D. Voigtlaender, C, Liu, S. Swink, K. Wardwell, J. Fisher, freshly isolated from human blood express CD4 and mature into typical immu- R. F. Graziano, L. C. Pfefferkorn, and P. M. Guyre. 1997. Characterization of nostimulatory dendritic cells after culture in monocyte-conditioned medium. expression, cytokine regulation, and effector function of the high affinity IgG http://www.jimmunol.org/ J. Exp. Med. 178:1067. receptor Fc␥RI(CD64) expressed on human blood dendritic cells. J. Immunol. 6. Thomas, R., L. S. Davis, and P. E. Lipsky. 1993. Isolation and characterization 158:3090. of human peripheral blood dendritic cells. J. Immunol. 150:821. 30. Reddy, A., M. Sapp, M. Feldman, M. Subklewe, and N. Bhardwaj. 1997. A 7. Rowden, G., M. G. Lewis, and A. K. Sullivan. 1977. Ia antigen expression on monocyte conditioned medium is more effective than defined cytokines in me- human epidermal Langerhans cells. Nature 268:247. diating the terminal maturation of human dendritic cells. Blood 90:3640. 8. Romani, N., A. Lenz, H. Stossel, U. Stanzl, O. Majdic, P. Fritsch, and G. Schuler. 31. Valladeau, J., V. Duvert-Frances, C. Dezutter-Dambuyant, C. Vincent, J.-J. Pin, 1989. Cultured human Langerhans cells resemble lymphoid dendritic cells in C. Massacrier, J. Vincent J. Davoust, and S. Saeland. 1998. A monoclonal anti- phenotype and function. J. Invest. Dermatol. 93:600. body against Langerin, a protein specific of Langerhans cells, is internalized in 9. Cerio, R., C. E. M. Griffiths, K. D. Cooper, B. J. Nickoloff, and J. T. Headington. coated pits and Birbeck granules. J. Leukocyte Biol. Suppl. 2:25. 1989. Characterization of factor XIIIa positive dermal dendritic cells in normal 32. Kashihara, M., M. Ueda, Y. Horiguchi, F. Furukawa, M. Hanaoka, and and inflamed skin. Br. J. Dermatol. 121:421. S. Imamura. 1986. A monoclonal antibody specifically reactive to human Lang- by guest on October 2, 2021 10. Nestle, F. O., X.-G. Zheng, C. B. Thompson, L. A. Turka, and B. J. Nickoloff. erhans cells. J. Invest. Dermatol. 87:602. 1993. Characterization of dermal dendritic cells obtained from normal human skin reveals phenotypic and functionally distinctive subsets. J. Immunol. 151: 33. Sallusto, F., and A. Lanzavecchia. 1994. Efficient presentation of soluble antigen 6535. by cultured human dendritic cells is maintained by granulocyte/macrophage colony-stimulating factor plus interleukin 4 and down-regulated by tumor necrosis 11. Ardavin, C., L. Wu, C.-L. Li, and K. Shortman. 1993. Thymic dendritic cells and ␣ T cells develop simultaneously in the thymus from a common precursor popu- factor- . J. Exp. Med. 179:1109. lation. Nature 362:761. 34. Shibaki, A., L. Meunier, C. Ra, S. Shimada, A. Ohkawara, and K. D. Cooper. 12. Wu, L., L. Chung-Leung, and K. Shortman. 1996. Thymic dendritic cell precur- 1995. Differential responsiveness of Langerhans cell subsets of varying pheno- sors: relationship to the T lymphocyte lineage and phenotype of the dendritic cell typic states in normal human epidermis. J. Invest. Dermatol. 104:42. progeny. J. Exp. Med. 184:903. 35. Berg, E. L., T. Yoshino, L. S. Rott, M. K. Robinson, R. A. Warnock, 13. Grouard, G., I. Durand, L. Filgueira, J. Banchereau, and Y.-J. Liu. 1996. Den- T. K. Kishimoto, L. J. Picker, and E. C. Butcher. 1991. The cutaneous lympho- dritic cells capable of stimulating T cells in germinal centres. Nature 384:364. cyte antigen is skin lymphocyte homing receptor for the vascular lectin endothe- 14. Grouard, G., M.-C. Rissoan, L. Filgueira, I. Durand, J. Banchereau, and Y.-J. Liu. lial cell-leukocyte adhesion molecule 1. J. Exp. Med. 174:1461. 1997. The enigmatic plasmacytoid T cells develop into dendritic cells with in- 36. Koszik, F., D. Strunk, I. Simonitsch, L. J. Picker, G. Stingl, and E. Payer. 1994. terleukin (IL)-3 and CD40-ligand. J. Exp. Med. 185:1101. Expression of monoclonal antibody HECA-452-defined E-selectin ligands on 15. Caux, C., C. Dezutter-Dambuyant, D. Schmitt, and J. Banchereau. 1992. GM- Langerhans cells in normal and diseased skin. J. Invest. Dermatol. 102:773. ␣ CSF and TNF- cooperate in the generation of dendritic Langerhans cells. Nature 37. Nakache, M., E. L. Berg, P. R. Streeter, and E. C. Butcher. 1989. The mucosal 360:258. vascular addressin is a tissue-specific endothelial cell adhesion molecule for cir- 16. Caux, C., B. Vanbervliet, C. Massacrier, C. Dezutter-Dambuyant, culating lymphocytes. Nature 337:179. B. de Saint-Vis, C. Jacquet, K. Yoneda, S. Imamura, D. Schmitt, and ϩ 38. Schmitt, D. A., T. Bieber, J.-P. Cazenave, and D. Hanau. 1990. Fc receptors of J. Banchereau. 1996. CD34 hematopoietic progenitors from human cord blood human Langerhans cells. J. Invest. Dermatol. 94:15S. differentiate along two independent dendritic cell pathways in response to GM- CSF ϩ TNF␣. J. Exp. Med. 184:695. 39. Fearnley, D. B., A. D. McLellan, S. I. Mannering, B.D. Hock, and D. N. J. Hart. 17. Reid, C. D. L., A. Stackpoole, A. Meager, and J. Tikerpae. 1992. Interactions of 1997. Isolation of human blood dendritic cells using the CMRF-44 monoclonal tumor necrosis factor with granulocyte-macrophage colony-stimulating factor antibody: implications for studies on antigen-presenting cell function and immu- and other cytokines in the regulation of dendritic cell growth in vitro from early notherapy. Blood 89:3708. bipotent CD34ϩ progenitors in human bone marrow. J. Immunol. 149:2681. 40. McLellan. A. D., R. V. Sorg, L. A. Williams, and D. N. J. Hart. 1996. Human 18. Romani, N., S. Gruner, D. Brang, E. Kampgen, A. Lenz, B. Trockenbachter, dendritic cells activate T lymphocytes via a CD40:CD40 ligand-dependent path- G. Konwalinka, P. O. Fritsch, R. M. Steinman, and G. Schuler. 1994. Prolifer- way. Eur. J. Immunol. 26:1204. ating dendritic cell progenitors in human blood. J. Exp. Med. 180:83. 41. Brown, K. A., P. Bedford, M. Macey, D. A. McCarthy, F. Leroy, A. J. Vora, 19. Strunk, D., C. Egger, G. Leitner, D. Hanau, and G. Stingl. 1997. A skin homing A. J. Stagg, D. C. Dumonde, and S. C. Knight. 1997. Human blood dendritic molecule defines the Langerhans cell progenitor in human peripheral blood. cells: binding to vascular endothelium and expression of adhesion molecules. J. Exp. Med. 185:1131. Clin. Exp. Immunol. 107:601.