Comparison of RAW264.7, human whole blood and PBMC assays to screen for immunomodulators

The RAW264.7 mouse macrophage cell line is used extensively to carry out in vitro screens for immunomodulators. Compounds that are effective at reducing the expression of pro-inflammatory cytokines or nitric oxide (NO) from lipopolysaccharide (LPS)-stimulated RAW264.7 cells are often considered candidate anti-inflammatory agents for humans. There is, however, very little data on the reliability of this screen to identify bona fide human immunomodulators. We compared the efficacy of 37 purported immunomodulators to modulate LPS or E. coli-induced inflammatory responses in RAW264.7 cell, whole human blood and human peripheral blood mononuclear cell (PBMC) assays. Interestingly, there was no significant correlation (R = 0.315) between the responses obtained with RAW264.7 cells and the whole blood assay (WBA), suggesting that compounds demonstrating efficacy in RAW264.7 cells may be ineffective in humans, and, more importantly, compounds that are effective in humans may be missed with a RAW264.7 screen. Interestingly, there was also no significant correlation between the WBA and human PBMCs when the latter were cultured with 10% FCS, but a moderate correlation was seen when the PBMCs were cultured with 25% autologous plasma. The presence of plasma thus contributes to the overall inflammatory response observed in the WBA. We then asked if RAW264.7 cells, given that they are mouse macrophage-like cells, respond in a manner similar to primary murine derived macrophages. Intriguingly, there was no significant correlation (R = 0.012) with the 37 putative immunomodulators, pointing to distinct inflammatory response mechanisms in the two model systems. We conclude that the use of a WBA to confirm potential immunomodulators obtained from high throughput screening with RAW264.7 cells is advisable and that future screens be carried out using a WBA.