The TNF-α receptor, CD120a, has recently been shown to be localized to both plasma membrane lipid rafts and to the trans Golgi complex. Through a combination of both confocal microscopy and sucrose density gradient ultracentrifugation, we show that amino acid sequences located within the death domain (DD) of CD120a are both necessary and sufficient to promote the appropriate localization of the receptor to lipid rafts. Deletion of the DD (CD120a.Δ321-425) prevented the receptor from being targeted to lipid rafts and resulted in a uniform plasma membrane localization. A similar loss of raft localization was also observed following pairwise deletion of the six α-helices that comprise the DD. In all situations, the loss of the ability of CD120a to become localized to lipid rafts following mutagenesis was paralleled by a failure of the receptor to initiate apoptosis. Furthermore, introduction of the lpr mutation into CD120a (CD120a.L351N) also resulted in both a loss in the ability of the receptor to signal apoptosis and to be appropriately localized to rafts. In contrast to CD120a, CD120b, which lacks a DD, is mainly expressed in the bulk plasma membrane and to a lesser extent in lipid rafts, but is absent from the Golgi complex. However, a chimeric receptor in which the DD of CD120a was fused to the cytoplasmic domain of CD120b was predominantly localized to lipid rafts. Collectively, these findings suggest that in addition to its role in CD120a signaling, an appropriately folded and functionally active DD is required for the localization of the receptor to lipid rafts.

The TNFR, TNF-R1, is localized to lipid raft and nonraft regions of the plasma membrane. Ligand binding sets in motion signaling cascades that promote the activation of p42mapk/erk2 and NF-κB. However, the role of receptor localization in the activation of downstream signaling events is poorly understood. In this study, we investigated the dynamics of TNF-R1 localization to lipid rafts and the consequences of raft localization on the activation of p42mapk/erk2 and NF-κB in primary cultures of mouse macrophages. Using sucrose density gradient ultracentrifugation and a sensitive ELISA to detect TNF-R1, we show that TNF-R1 is rapidly and transiently recruited to lipid rafts in response to TNF-α. Disruption of lipid rafts by cholesterol depletion prevented the TNF-α-dependent recruitment of TNF-R1 to lipid rafts and inhibited the activation of p42mapk/erk2, while the activation of NF-κB was unaffected. In addition, phosphorylated p42mapk/erk2, but not receptor interacting protein, I-κB kinase-γ, or I-κBα was detected in raft-containing fractions following TNF-α stimulation. These findings suggest that TNF-R1 is localized to both lipid raft and nonraft regions of the plasma membrane and that each compartment is capable of initiating different signaling responses. We propose that segregation of TNF-R1 to raft and nonraft regions of the plasma membrane contributes to the diversity of signaling responses initiated by TNF-R1.

We previously demonstrated that the macrophage cell lines RAW 264.7 and WEHI-3 exhibit distinct patterns of gene expression in response to IFN-γ. This difference is controlled at the transcriptional level and results from a specific inability of the less mature WEHI-3 cells to utilize either the IFN-stimulated response element or the γ-activated sequence DNA regulatory element in response to stimulation with IFN-γ, while other aspects of IFN-γ gene induction remain intact. In the work described here, we examined the components of the IFN-γ signal transduction pathway in RAW 264.7 and WEHI-3 cells to determine whether differences in pathway components or activity exist in WEHI-3 cells that could give rise to this difference in transcriptional response. Reverse transcriptase-PCR (RT-PCR) and flow cytometric analyses indicated that the levels of IFN-γ receptor mRNA accumulation and protein expression are comparable for RAW 264.7 and WEHI-3 cells. RT-PCR and immunoblot analyses revealed that the principal components of this signaling pathway, including JAK1, JAK2, and STAT1, are present in both RAW 264.7 and WEHI-3 cells. However, analysis of STAT1 DNA-binding activity by electrophoretic mobility shift assay and of STAT1 phosphorylation by immunoblot revealed that this DNA-binding factor is active in RAW 264.7, but not in WEHI-3, cells after IFN-γ stimulation. These results demonstrate that the components of the IFN-γ signal transduction pathway are intact in WEHI-3 cells, but stimulation of these cells by IFN-γ does not result in STAT1 activation.