Nature © Macmillan Publishers Ltd 1997 letters to nature NATURE | VOL 388 | 21 AUGUST 1997 787 13. Reid, P. A. & Watts, C. Cycling of cell-surface MHC glycoproteins through primaquine-sensitive intracellular compartments. Nature 346, 655–657 (1990). 14. Neefjes, J. J., Stollorz, V., Peters, P. J., Geuze, H. J. & Ploegh, H. L. The biosynthetic pathway of MHC class II but not class I molecules intersects the endocytic route. Cell 61, 171–183 (1990). 15. Roosnek, E., Demotz, S., Corradin, G. & Lanzavecchia, A. Kinetics of MHC–antigen complex formation on antigen-presenting cells. Immunol. 140, 4079–4082 (1988). 16. Lanzavecchia, A., Reid, P. A. & Watts, C. Irreversible association of peptides with class II MHC molecules in living cells. Nature 357, 249–252 (1992). 17. Nelson, C. A., Petzold, S. J. & Unanue, E. R. Peptides determine the lifespan of MHC class II molecules in the antigen-presenting cell. Nature 371, 250–252 (1994). 18. Amigorena, S., Drake, J. R., Webster, P. & Mellman, I. Transient accumulation of new class II MHC molecules in a novel endocytic compartment in B lymphocytes. Nature 369, 113–120 (1994). 19. Germain, R. N. & Hendrix, L. R. MHC class II structure, occupancy and surface expression determined by post-endoplasmic reticulum antigen binding. Nature 353, 134–139 (1991). 20. Pierre, P. et al. Developmental regulation of MHC class II transport in mouse dendritic cells. Nature this issue. EDS to complete on page Acknowledgements. We thank D. Scheidegger and M. Dessing for technical assistance, E. Long for discussion, and C. Watts, F. Sallusto, K. Karjalainen and M. Colonna for critically reading the manuscript. The Basel Institute for Immunology was founded and is supported by F. Hoffmann-La Roche, Basel, Switzerland. Correspondence and requests for material should be addressed to M.C. (e-mail: cella@bii.ch). Developmental regulation of MHC class II transport in mouse dendritic cells Philippe Pierre*, Shannon J. Turley*, Evelina Gatti, Michael Hull, Joseph Meltzer†, Asra Mirza†, Kayo Inaba†, Ralph M. Steinman† & Ira Mellman Department of Cell Biology, Yale University School of Medicine, 333 Cedar Street, PO Box 208002, New Haven, Connecticut 06520, USA † The Rockefeller University, 1230 York Avenue, New York, New York 10021, USA * These authors contributed equally to this work. ......................................................................................................................... Dendritic cells (DCs) have the unique capacity to initiate primary and secondary immune responses 1–3 . They acquire antigens in peripheral tissues and migrate to lymphoid organs where they present processed peptides to T cells. DCs must therefore exist in distinct functional states, an idea that is supported by observa- tions that they downregulate endocytosis and upregulate surface molecules of the class II major histocompatibility complex (MHC) upon maturation 4–7 . Here we investigate the features of DC maturation by reconstituting the terminal differentiation of mouse DCs in vitro and in situ. We find that early DCs, corre- sponding to those found in peripheral tissues, exhibit a phenotype in which most class II molecules are intracellular and localized to lysosomes. Upon maturation, these cells give rise to a new intermediate phenotype in which intracellular class II molecules are found in peripheral non-lysosomal vesicles, similar to the specialized CIIV population seen in B cells. The intermediate cells then differentiate into late DCs which express almost all of their class II molecules on the plasma membrane. These variations in class II compartmentalization are accompanied by dramatic alterations in the intracellular transport of the new class II molecules and in antigen presentation. We found that although early DCs could not present antigen immediately after uptake, efficient presentation of the previously internalized antigen occurred after maturation, 24–48 hours later. By regulating class II transport and compartmentalization, DCs are able to delay antigen display, a property crucial to their role in immune surveillance. Mouse bone marrow is a major source of DCs when cultivated with granulocyte–macrophage colony-stimulating factor (GM- CSF) 8 . Immunofluorescence microscopy of these cultures revealed three distinct developmental stages. Cells were identified as DCs by the expected repertoire of antigens and expression of MHC class II, cell shape, and adherence. As reported previously 8 , DCs were found by immunofluorescence or FACS to be negative or weakly positive for the granulocyte marker GR1, negative for the macrophage marker SER-4, but strongly positive for CD11c and MHC class II. Contaminating SER-4 or GR1-positive cells were negative for class II and judged not to be DCs. After 4–5 days, DCs were found in proliferating clusters loosely attached to adherent stromal cells 8 . By confocal microscopy, most of the cells present in or migrating out from the clusters showed little MHC class II on their surface, but contained abundant intracellular class II (Fig. 1a). The class II-positive vesicles represented lysosomes (MIICs) and late endosomes, being positive for lgp-B/lamp-2 and H2-M (Fig. 1a). Thus, they were characteristic of MIIC as defined in human lymphoblasts and human DCs 9–12 . As the MIIC-containing cells were present in proliferating clusters, we defined them as ‘early’ DCs. With increasing time in culture, two additional cell populations were detected. The first of these (‘intermediate’ DCs) was present transiently and comprised non-adherent cells that had little surface MHC class II (Fig. 1b). They were strikingly unlike the early cells, however, because most of their intracellular class II was in a vesicle population that was devoid of lysosomal markers (Fig. 1b, arrows), and thus reminiscent of non-lysosomal, class II-positive CIIV isolated from A20 B cells 13 . At later times in particular, the vesicles accumulated directly beneath the plasma membrane (Fig. 1b, right), whereas the lysosomes became concentrated in the perinuclear region. The third major DC population accumulated with time until by 8–10 days it represented almost all of the non-adherent class II- positive cells. These ‘late’ cells had a more classical DC phenotype, with long processes that stained for class II (green) (Fig. 1c). Little class II remained intracellularly, with most of the now largely class II-depleted H2-M/lamp-positive lysosomes visualized as poorly resolved clusters of red-staining vesicles in the perinuclear region. Further characterization indicated that markers such as DEC-205 and 2A1 were absent from early cells but expressed at moderate and high levels on intermediate and late cells, respectively 8,14 (results not shown). Early cells, but not late cells, were capable of efficient fluid endocytosis (W. Garrett and I.M., unpublished results), as found previously for human cells 4 . To determine whether early cells pass through the intermediate phenotype before reaching maturity, we produced highly purified populations of early cells by gently dislodging and isolating pro- liferating clusters on serum columns, followed by depletion of contaminating cells by fluorescent-activated cell sorting (FACS) 8 . This approach yielded 95% pure populations of early DCs. As quantified in Fig. 1d , after 5–8 h in culture, the early cells had nearly disappeared and 60% of the population of exhibited the inter- mediate phenotype (for example, see Fig. 1b); 20–30% exhibited the late or mature phenotype (as shown in Fig. 1c). After 24 h, 90% of the cells were found to be of the late phenotype. As the number of cells remained constant throughout, these results strongly suggest that there is a sequential relationship in the maturation pathway. The rapid maturation kinetics observed using purified cluster-derived cells probably reflect their greater developmental synchrony. To ensure that the developmental sequence was not peculiar to bone marrow cultures, we investigated whether tissue DCs had similar properties. We first examined epidermal Langerhans cells 15 . In epidermal explants, class II was present in these cells in a punctuate pattern which co-localized with lysosomal marker H2- M, reminiscent of early bone marrow DCs (Fig. 2, upper right panels). If explants were incubated in culture medium and the Langerhans cells allowed to mature in situ, within 4 h the degree of class II and H2-M co-localization decreased, with class II staining remaining punctuate but becoming progressively less coincident with H2-M, which became concentrated at the cell body (Fig. 2, right panels). This pattern was consistent with the intermediate