We are studying retroviral envelope proteins (Env) and cellular receptor proteins, which mediate viral entry. Some studies involve mutating the receptor, or the Env, to see what portions of these proteins are involved in various steps in the fusion process. Other approaches involve the use of antibodies that bind to Env and either do, or do not, block fusion. In one set of experiments we engineered cells to express wild-type HIV or murine leukemia virus (MLV) envelope proteins along with a portion of these proteins that is known to form a homo-trimer (the N-heptad repeat region). The heptad repeat regions were linked to green fluorescent protein to allow visualization and antibody detection. The heptad repeat chimeric proteins formed mixed hetero-trimers with wild-type Env. Unexpectedly, we found that this very efficiently blocked transport of Env out of the endoplasmic reticulum (ER) and cleavage of the Env precursor protein into the mature form of Env required for membrane fusion. Since the Env proteins of many viruses in addition to retroviruses (e.g. influenza, SARS) go through similar steps of trimerization and cleavage, this work suggests a potentially general mechanism to inhibit formation of infectious virus via molecules that interact with Env during synthesis. In another set of experiments, we investigated the role of a cellular enzyme, protein disulfide isomerase (PDI), reported to catalyze disulfide bond rearrangements in HIV envelope protein during membrane fusion. We found that while antibodies to PDI inhibited virus infection, over-expression of wild-type or mutant PDI, intracellularly or on the cell surface, or inhibiting the synthesis of PDI via small inhibitory RNAs (siRNA), had little effect on cell fusion mediated by HIV Env. In contrast, a cell surface thiol-reactive reagent (DTNB) blocked HIV Env-mediated cell fusion extensively. These results suggest that other cell surface thiol-containing molecules act in parallel with PDI to promote Env-mediated fusion. To further evaluate our PDI-siRNA constructs, we collaborated with a group at Harvard Medical School that studies polyoma virus entry; using our constructs, they found that infection with polyoma virus was inhibited by down-regulating PDI, consistent with a role for PDI in uncoating polyoma virus capsid in the ER. In other studies, we investigated the mechanism of antibody neutralization of HIV. Antibodies that neutralize multiple strains of HIV are rare, so it is important to understand how these antibodies function. Three of five epitopes for broadly neutralizing antibodies are located in the membrane-proximal region of the HIV TM Env protein, which has led to the idea that epitope location may be crucially important. To investigate this, we transferred an epitope for one of these neutralizing antibodies (2F5) to the analogous position in the TM gene of a murine leukemia virus (MLV), or to a different location in the MLV SU protein. As controls, we inserted epitopes (HA and his6) for other antibodies at the same positions. The 2F5 antibody neutralized MLV carrying the 2F5 epitope in either location, while antibodies to HA and his6 did not block Env-mediated fusion, despite binding as well as 2F5. This indicates that the important feature is not epitope location, but rather a structural characteristic of the neutralizing antibody. If the structural characteristic can be identified, it might enable the engineering of monoclonal antibodies to make them neutralizing. In another line of experiments, we observed an unexpected property of quantum dots relevant to their use as fluorescent probes in vivo. In following up on studies of ability of cell surface polyanions to inhibit entry by certain strains of HIV, we noticed that polyanionic quantum dots could be concentrated at the cell surface by incubating them with polycations, or by attaching a polybasic peptide (related to the Tat protein of HIV) to their surface. After binding to the cell surface, the quantum dots were endocytosed and trafficked to lysosomes. Surprisingly, we noticed that once inside cells, the fluorescence intensity of the quantum dots increased several fold when they were illuminated. This is the opposite of the common phenomenon of photobleaching. Since quantum dots start out about as bright as they can be (quantum efficiency approaching 1), we surmise that something in the cell reduces their fluorescence efficiency in a photo-reversible way. Cellular modification of fluorescence intensity could affect the use of quantum dots as quantitative probes in vivo. Better understanding of this phenomenon might enable creation of quantum dots whose fluorescence is turned on in specific locations by focused laser light, which could be useful to follow the motion of molecules labeled with such dots.
Showing the most recent 10 out of 13 publications
