Future studies to quantify in more detail the diluting effect of differential expression in non-target cells on detectable gene expression of an individual cell type within a mixture, could be undertaken by means of titration experiments. A series of artificially mixed samples could be used, with, for example, decreasing proportions of MonoD added to Mono+ samples. We chose to study the interaction between monocytes and LPS because these inflammatory cells comprise a significant minority of PBMC and their interaction with LPS in initiating the inflammatory cascade in Gram negative bacterial infection is well described. It is likely that our results apply to the detection of gene expression in other situations. The proportion of genes in a particular cell type that can be detected in a cell mixture might be lower for target cells comprising a smaller proportion of the mixture or in which the cell type of interest is not the principal target or primary effector cell. However, it is probable that detection of the most differentially expressed genes would remain less susceptible to the influence of other cells within the cell mixture. To ensure gene expression induced by LPS stimulation was comparable in the monocytes analysed alone and in those analysed within PBMC, monocytes were separated after stimulation of PBMC with LPS, thus preserving the interactions between cells. The repeatability of our results was confirmed by the high degree of correlation between the two PBMC samples analysed from the same individual. This control validated the relevance of the differences in gene expression detected between PBMC and the monocyte samples. We included a number of other control samples to exclude the possibility that our results were confounded by artefactual or technical variability. Monocytes were isolated by different methods to TWS119 enable us to control for the possibility that gene expression might be induced by binding of CD14 antibodies during positive selection of monocytes. The similarity in differential gene expression between the Mono+ and Mono2 samples showed that changes in cell surface expression of CD14 or LPS-binding of CD14 did not have a major effect. The greater correlation between the PBMC and Mono+ samples than between the PBMC and Mono2 samples reflects the higher proportion of monocytes selected using positive isolation. Negative isolation resulted in up to 50% of non-monocytes cells remaining in the Mono2 sample. Possible explanations include the relatively lower efficacy of negative selection and the absence of antibodies to every nonmonocyte cell type present in PBMC. There is increasing evidence that the immune system and the central nervous system interact to modulate each other via autonomic pathways. Immune mediators and cytokines released by the innate immune system rapidly activate neuronal responses, resulting in amplified local immune responses designed to clear pathogens and triggering regional neural and systemic neuroendocrine responses.