Discussion

Bloodstream form trypanosomes encode two structurally related GPI-anchored proteins, variant surface glycoprotein (VSG) and transferrin receptor (TfR), that are both critical for survival in the mammalian host. VSG is central to antigenic variation; TfR internalizes transferrin (Tf) for iron acquisition. Both are expressed by the active telomeric expression site, but ~20% of all TfR transcripts derive from ‘silent’ sites. VSG is a homodimer; TfR is a heterodimer of ESAG6 (E6, GPI-anchored) and ESAG7 (E7). We have shown that progression of GPI-anchored proteins correlates with GPI-valence: homodimeric VSG (GPI²) is a surface protein; heterodimeric TfR (GPI¹) is in the flagellar pocket; and truncated VSG (GPI⁰) is degraded in the lysosome. Overexpressed TfR accesses the plasma membrane suggesting a saturable mechanism for retention in the flagellar pocket. However, surface TfR is non-functional for Tf binding and we proposed that such TfR represents GPI² E6 homodimers that are unable to bind transferrin - mimicking native VSG [J. Cell Sci. (2005) 118:5499]. We now create RNAi cells for suppression of native TfR subunits. Silencing is lethal and is complemented by placing resistant (RNAi^R) E6/E7 genes into the active TfR locus. When expressed alone each subunit conforms to the valence model: GPI² E6 homodimers appear on the surface like native VSG; GPI⁰ E7 homodimers are delivered to the lysosome like GPI-minus VSG. Finally, we created an RNAi^R GPI² TfR by fusing the C-terminal domain of E6 to E7 (E7^GPI). When expressed together (E6:E7^GPI) these proteins form functional GPI² heterodimers that rescue growth, mediate Tf binding and uptake, and localize to the cell surface. Collectively our results with GPI-modified TfR validate the GPI valence hypothesis.