In This Issue

(See article on pages R9–R14.)

The weak response of the cellular immune system to soluble antigens limits the use of vaccination to control tumor growth and suppress viral infections. Dhodapkar et al. have previously argued that even apparently exceptional antigens, such as the keyhole limpet hemocyanin (KLH), owe their robust T cell-priming activity to contamination with bacterial lipopolysaccharide, an activator of dendritic cells (DCs). DCs, antigen presenting cells that stimulate CD4⁺ and CD8⁺ T cell activity directly, can now be maintained in culture and exposed to specific antigenic peptides, a treatment that induces their maturation and allows them to present these peptides to T cells in vivo or in vitro. Dhodapkar and colleagues recently showed that reintroducing even a single dose of such antigen-pulsed autologous DCs into a human host induces T cell responses, including antigen-dependent induction of IFN-γ, and lysis of antigen-bearing target cells. In their initial study, the authors had pulsed DCs with a viral epitope, a peptide from the matrix protein (MP) of influenza virus, in the presence or absence of KLH or another antigen. They froze away lymphocytes taken from 9 subjects at different times after vaccination and set these samples aside for later analysis. Now they have injected some of the individuals from this earlier study with DCs pulsed with MP alone, and they have compared the responses of T cells isolated before and after boosting. The initial CD8⁺ response to KLH peaked after 30–90 days but remained well above baseline after 1 year for several subjects. After boosting, responses in each of 3 individuals were quicker and more dramatic than had been seen before: Killing of cells displaying the MP antigen occurred more efficiently and anti-KLH antibodies increased to higher levels, as did the fraction of T-cells that induced IFN-γ in response to MP. Dhodapkar et al. also note that boosting permitted T-cells to respond to MP presented in more dilute form, suggesting that human T cell receptors, like immunoglobulins, are subject to selection for higher avidity interactions following the initial exposure to antigens.

(See article on pages 823–832.)

Here, Okada and co-workers explore a well-established but poorly understood effect of the prostaglandin PGE₂ on the development of bone resorbing osteoclasts. Osteoblast precursors can complete their differentiation in organ culture if they are co-cultured with and allowed to contact osteoblasts. Prostaglandins promote this differentiation, and inhibitors of prostaglandin synthesis, such as the non-steroidal anti-inflammatory drugs, diminish osteoclast differentiation in culture. Based on both the biochemical specificity of these inhibitors and genetic studies using mice lacking one of the two prostaglandin G/H synthase genes, PGHS-1 or -2, it appears that only PSGH-2 and its product PGE₂ are required in this process. Indeed, Okada et al. determine that PGE₂ can be limiting for osteoclast formation in this system. When treated to induce osteoclasts, bone marrow cells cultured from PSGH-2^+/– heterozygous animals produce approximately half as many osteoclast-like cells as do PSGH-2^+/+ cultures, and PGE₂ levels in the medium are reduced by a corresponding amount. Osteoclast formation in PSGH-2^–/– cultures, in which PGE₂ levels are nearly undetectable, is rarer still, but exogenous PGE₂ restores osteoclast formation to maximal efficiency in cultures of any PSGH-2 genotype. By co-culturing osteoclast precursors and osteoblasts from different genotypes, the authors show that prostaglandin synthesis is required in the latter cells but not the former. The effects of PGE₂ in this system may be mediated by expression of the RANK ligand, a known inducer of osteoclast formation whose mRNA is reduced 2-fold in PSGH-2^–/– osteoblasts cells.

(See article on pages 813–821.)

In the wake of a systemic viral infection, T and B lymphocytes specific for the virus undergo apoptosis and are cleared from peripheral organs. The importance of this activation-induced cell death (AICD) is evident from the phenotype of mice lacking Fas or Fas ligand (FasL), crucial mediators of immune cell apoptosis. Such mice mount an efficient reaction to murine cytomegalovirus (MCMV), but they show persistent inflammation and autoimmune reactivity in their lungs and livers long after the virus is cleared. Here, Zhang and colleagues follow up on evidence that antigen presenting cells (APCs) expressing FasL can induce tolerance to specific antigens, showing FasL⁺ APCs suffice to direct AICD, even in animals that otherwise lack FasL. When these APCs are pulsed with MCMV in culture and then introduced intravenously into FasL-deficient animals, they present the viral epitopes, promote AICD, and suppress inflammation more vigorously than other FasL⁺ macrophages. Although their anti-inflammatory effect is seen in the peripheral organs, these reintroduced lymphocytes home directly to the marginal zone of the spleen. Evidently, ongoing inflammation requires a constant supply of T-cells, and FasL expression by APCs allows the killing of newly emerging splenic T cells.