Sirs

We are grateful for the comments of Subbarao Bondada and Ralph Chelvarajan on our recent review of neonatal immunity1. Their observation that neonatal splenic macrophages overproduce interleukin-10 (IL-10)2, which downregulates the production of pro-inflammatory cytokines, does seem to provide support for our suggestion that the dampening of inflammatory responses in neonates could reflect the need to avoid damage to developing tissues. It also draws attention to functional differences between neonatal and adult macrophages that, mainly for reasons of space, were not addressed in our review1. We note that their findings are similar to those of Genovese et al.3, who showed that neonatal macrophages produced more IL-10 than adult macrophages when stimulated with killed Listeria monocytogenes bacteria or lipopolysaccharide and that blockade of IL-10 abrogated the increased susceptibility of neonates to listeriosis, presumably by allowing increased pro-inflammatory responses. S.M.C. has previously proposed a role for IL-10 as a potential modulator of neonatal responses4, based on observations that CD5+ B cells (the most abundant B-cell subset in the neonatal spleen) and neonatal T cells secrete high levels of this cytokine5,6. Indeed, overproduction of IL-10 begins during pregnancy, in which again it has been suggested to have a role in dampening pro-inflammatory responses that might damage the developing fetus7.

Our proposal regarding the importance of marginal-zone B cells in response to T-cell-independent type 2 (TI-2) polysaccharide antigens was written with bacterial polysaccharide antigens in mind. We were aware of the reported effects of IL-1 and IL-6 on responses to trinitrophenyl (TNP)-Ficoll, which are cited as evidence by Bondada and Chelvarajan that macrophages, rather than marginal-zone B cells, dictate responsiveness to polysaccharides. However, although TNP-Ficoll has been widely used as a model for this group of antigens, its behaviour is rather anomalous. Responses to TNP-Ficoll can be detected early in ontogeny, long before responses to classical bacterial polysaccharides8, and neonatal exposure to TNP-Ficoll markedly primes animals for subsequent recall responses — an effect that is not seen following administration of bacterial polysaccharides9. Furthermore, TNP-Ficoll has been shown to promote the differentiation of developing B cells in the bone marrow and even to induce cytokine production by T cells10,11. Therefore, some caution needs to exercised when interpreting data that depend on this antigen. Nonetheless, mice in which marginal-zone macrophages were eliminated mounted normal antibody responses to TNP-Ficoll, indicating that B cells are more important than macrophages12. The well-reported effects of CpG-containing oligodeoxynucleotides on neonatal responses, which are also cited by Bondada and Chelvarajan, do not provide evidence that these molecules function through stimulation of macrophages. These immunomodulators are ligands for Toll-like receptor 9, which is expressed by various cell types, including B cells, macrophages and dendritic cells. We and Bondada's group have shown that CpG-containing oligodeoxynucleotides activate responses by highly purified neonatal B cells13,14. Therefore it is perfectly plausible that they might increase responses to TI-2 antigens by directly activating a subset of B cells that is normally unresponsive to this type of antigen.

All of these data indicate that there is an urgent need to examine the effects of pro-inflammatory cytokines on neonatal responses to classical bacterial polysaccharides; until these experiments are carried out, the jury is still out on the relative contribution of marginal-zone B cells and macrophages to adult-like responses to this class of antigen. We fully agree with Bondada and Chelvarajan that these experiments will have an important role in guiding the development of vaccines against infection with Streptococcus pneumoniae and other important bacterial pathogens that cause high mortality and morbidity in early life.