Background:  The entry process of HIV infection requires the engagement of gp120 surface protein of HIV to host cell CD4 and CXCR4 or CCR5. Previous studies have shown that receptor density can influence the efficiency of the entry process. Studies in our lab and others have also shown receptor and co-receptors to be localized to ezrin-containing membrane structures. Published data in our lab have recently shown that the CD4 and CCR5 receptors are highly mobile and that mobility may be important for HIV infection. We now expand this line of investigation to include CXCR4.

Methods:  For this project we designed a CXCR4-YFP fusion protein then tested its ability to localize appropriately, internalize its natural ligand and to be used by HIV for entry using previously published methods. To further characterize CD4 and the co-receptors, we created photo-activatable GFP (PA-GFP) fusion proteins. PA-GFP differs from wild type GFP by a single mutation at the threonine 203 position (T203H). This mutation initially has barely detectable GFP fluorescence. However, once it is irradiated with 413-nm light, GFP fluorescence increases by as much as 100-fold. We transfected cells with each receptor-PA-GFP fusion and photo-activated select areas of the cell using a digital diaphragm. We tracked GFP movement by taking a snapshot of the cells every 5 seconds for a total of 100 frames with activation of the entire cell at frame 95. To quantify the GFP diffusion, we identified regions of interest both proximal and distal to the area of photo-activation. The resulting data are expressed as the signal of each region of interest at each time point divided by the signal of each region of interest after total activation (e.g., Sx/Sf).

Results:  We have found that CXCR4 works as a co-receptor and localizes in ezrin-containing structures with CCR5 and CD4. However, these proteins do not form complexes with each other because they are independently internalized. Using PA-GFP, we have found differences in the rate of diffusion of the 3 proteins specifically noting that CCR5 diffuses the most rapidly and completely.

Conclusions:  CD4, CXCR4, and CCR5 fluorescently labeled fusion proteins function and localize normally making them an effective way to study the dynamics of these proteins. In addition, preliminary studies with our PA-GFP constructs suggest differences in the rate and degree of diffusion between the 3 proteins and that CXCR4 may contain an immobile phase similar to that previously described for CD4.

Keywords: Fusion; CXCR4; Membrane