Parentage and heritage of dendritic cells

Dendritic cells (DCs), despite their common functions of antigen-processing and T-lymphocyte activation, are diverse in surface markers, migratory patterns, and cytokine output. These differences can determine the fate of the T cells they activate. But tracing the ancestry of these specialized DC subtypes has proved to be a dubious enterprise. The original concept that all DCs were of myeloid origin was questioned when Ardavin et al (Nature. 1993;362:761-763) found a mouse thymus T precursor population, which although lacking a capacity to form macrophages or granulocytes, produced thymic DCs. Since these DCs also expressed some lymphoid markers, including CD8α, they were termed “lymphoid DCs.” It was assumed that murine DC lacking CD8α and expressing certain myeloid markers would be the product of myeloid precursors; so these were termed “myeloid DCs.” But our assumption that the early hemopoietic precursor determined the mature DC phenotype was upset when Traver et al (Science. 2000;290:2152-2154) found that bone-marrow myeloid-restricted precursors produced both CD8α⁺ and CD8α⁻ DCs. Manz et al (page3333) now complete the picture, demonstrating that lymphoid-restricted precursors are indeed efficient producers of DCs but that both lymphoid-restricted and myeloid-restricted precursors are able to generate both CD8α⁺ and CD8α⁻ DCs. These results from the Stanford laboratory are in close agreement with our own current findings (Wu and D'Amico, in preparation). Manz et al argue that, since myeloid precursors outnumber lymphoid precursors, most DCs will be of myeloid origin; however, this balance will depend on the relative efficiency of the 2 precursors at generating DCs, on the kinetics of development, and on local environmental influences.

Does this early developmental flexibility mean the ancestry of DCs is irrelevant and that different DC subtypes simply represent different activation states? Not quite. The mature DC subtypes are not readily interconvertible. Recent kinetic evidence (Kamath et al, J Immunol. 2000;165:6762-6770) indicates that the DC subtypes of spleen are the products of separate developmental streams, at least as far back as their immediate dividing precursors. We must now determine the points downstream of the conventional myeloid and lymphoid precursors where the DC sublineages branch off, and determine the cytokines and environmental factors that induce their specialized functions.