Also, it is consistent with previous observations that genes upregulated in response to Ginsenoside-F4 interferon signaling are expressed at high levels in lupus patients. Many of these interferon responsive genes are present at increased levels in the activated versus resting lymphocytes observed here, suggesting that NK cells, T helper cells, and monocytes are the circulating cells principally responsive to type I interferon. Type I interferons are also known to promote maturation of monocytes into conventional dendritic cells, and this activity is consistent with our observation that activated dendritic cells appear in the blood of SLE patients. CDCs respond to bacterial infections and mediate the associated exacerbations in SLE patients, and they have been found to release IFN-alpha upon TLR4 ligation following priming with IFN-beta, consistent with their apparent involvement here. Conventional methods that use gene expression to quantify biological processes must use a single gene or a set of wellcorrelated genes, but solving a system of gene expression equations Sipeimine surmounts this limitation. For instance, RSAD2 would be a relatively poor cell type marker but is very valuable when quantifying cell types by the methods used here, since it is significantly differentially expressed between many different cell types. This gene is also important here because it is well-established to be induced by type 1 interferons. Among the activated cell types considered here, RSAD2 expression is high in those that are found to be more abundant in SLE patients. Therefore, this gene and the results here based on it support current views that the interferon signature observed in most SLE patients is due to the simultaneous activation of several classes of leukocytes by type 1 interferons and that this class of cytokines is central to the disease. The dynamics of activation appear to differ among cell types. For instance, monocytes constitute approximately the same fraction of blood in all samples examined, and the distribution of the proportion of resting to activated monocytes was fairly uniform across its observed range. In contrast, both NK and T helper lymphocytes were essentially fully activated or fully resting. Unlike FACS, the algorithm used here cannot differentiate between a population of a given cell type that are all homogeneously mildly activated and a population of that cell type that consists of a mixture of some resting cells and some strongly activated cells, since the parameters used to assess activation are merely the total quantity of each gene��s mRNA present in the sample. The monocytes observed here might possess the ability to adopt a mildly activated state, or perhaps they function in discrete activation states like lymphocytes appear to but differ in that the entire population of monocytes need not all be resting or activated in unison. In some cases the level of activation of a patient��s white blood cells was correlated among different cell types. For instance, dendritic cells�� and NK cells�� levels of activation are positively correlated. However, some types of lymphocytes exhibit more complex negative relationships. The maximum level of activated NK cells and T helper cells 
in SLE patients appears to be constant at approximately 30%. This is clearly seen in most SLE patients, but healthy individuals typically have substantially fewer activated lymphocytes. Although the results presented here offer no direct evidence about the mechanisms by which this pattern occurs, the data are consistent with the existence of a regulatory system.