Recognition of bacterial or viral DNA by the vertebrate immune system is based on the presence of unmethylated CG dinucleotides in particular sequence contexts ("CpG motifs"). At the time when we first identified the optimal CpG motif used by the human immune system to detect microbial DNA (J Immunol 2000:164:944) it became clear that CpG DNA has distinct properties as compared to other microbial molecules such as LPS, making CpG DNA a promising candidate for immunotherapy. The explanation for the distinct properties of CpG DNA now comes with the recent advances in the understanding of a new family of molecules, the so-called Toll-like receptors (TLR). Ten members of this family (TLR1-10) represent a combinatorial repertoire to discriminate among a wide spectrum of pathogen-associated microbial molecules, with TLR9 being the receptor for CpG DNA. While studies on molecular structure and signalling of TLRs are usually being performed in cell lines, our hypothesis was that the key to the understanding of their specific immunological activity lies in the study of primary immune cell subsets taken from fresh blood. Here Veit Hornung and colleagues made the surprising finding that different immune cell subsets in peripheral blood show a distinct profile of mRNA expression for TLR1 through TLR10; and that for CpG DNA as one representative microbial molecule the mRNA expression profile of TLR9 correlates with its functional profile. It turned out that only two immune cell subsets in peripheral blood express considerable amounts of TLR9 mRNA, the B cell and the plasmacytoid dendritic cell, and that only these two cell subsets are sensitive to CpG DNA. This information provided clarity in a field shaken by controversy. Besides direct activation of B cells and plasmacytoid dendritic cells, we demonstrated that CpG DNA stimulates monocytes, NK cells and T cells via soluble factors released by plasmacytoid dendritic cells, and not directly. These findings put plasmacytoid dendritic cell, also called the principal type I producing cell (IPC), in the focus of CpG research. This has important implications for the further development of CpG DNA as therapeutic.

The potential of CpG ODN as a cancer therapeutic is demonstrated in two recent publications from our group (J Immunol 2002; 169:3892; Eur J Immunol 2002 32:3235). The appreciation of plasmacytoid dendritic cells as the key sensors of CpG motifs led us to the identification of different classes of CpG oligonucleotides. CpG-A (prototype ODN 2216, Eur J Immunol 2001; 31:2154) and CpG-B (prototype ODN 2006, J Immunol 2000: 164: 1617) differ in the regulation of type I interferon induced in plasmacytoid dendritic cells (J Immunol 2003 170:4465-74); as a result, CpG-A drives monocytes toward a dendritic cell-like phenotype (J Immunol 2003 170:3468). The newest class of CpG oligonucleotides, CpG-C, combines features of CpG-A and CpG-B: CpG-C (prototype M362) induces high amounts of type I interferon in plasmacytoid dendritic cells and strongly activates B cells (Eur J Immunol 2003 33: 1633-41).

Gunther Hartmann, M.D.
Division of Clinical Pharmacology
Medizinische Klinik Innenstadt, Klinikum der LMU
Munich, Germany