Researchers at Rockefeller University in New York City have provided further evidence toward solving the mystery of where in the cell the HIV-1 virus assembles. Using fluorescent labeling techniques and total internal reflection microscopy, the group watched individual HIV-1 virions as they assembled on the surface of cells. The work enhances scientific understanding of the process by which the virus replicates and someday could lead to new treatments.Researchers at Rockefeller University have observed what they believe to be the formation of individual HIV virus particles. Using GFP-labeled Gag proteins and total internal reflection microscopy, the investigators studied the assembly of virus particles. In one of their experiments, they used a GFP variant that was quenched by acid pH to determine whether the viruslike particle was separated from the cell’s cytosol. Then they acidified the cytosol in the cell. Section (a) shows the cells (top panel) and viruslike particles (bottom panel) after a 30-s acidification pulse occurring at 5 min. A red arrow points out a punctum that was sensitive to the acidification, a blue arrow highlights a punctum that was not sensitive, and a black arrow shows one cell-free viruslike particle. Section (b) shows the drop in normalized fluorescence at approximately 5 min 30 s. Section (c) shows the loss of fluorescence immediately after acidification for diffuse Gag (gray, n = 107), for cell-free viruslike particles (black, n = 107) and for Gag puncta at t = 0 min (red, n = 101) and at t = 30 min (blue, n = 136). Section (d) shows loss of fluorescence after acidification of diffuse Gag (gray, n = 728), of cell-free viruslike particles (black, n = 334), of wild-type Gag (red, n = 308) and of Gag with a late-domain mutation (blue, n = 366). VLP = viruslike particles; WT = wild type; LD = late domain. Reprinted with permission from Nature. According to Sanford M. Simon, who, along with Paul D. Bieniasz and Nolwenn Jouvenet recently published the research in the May 25, 2008, online edition of Nature, scientists initially thought that HIV-1 assembled on cell surfaces. However, as the group began noticing the virus inside cells, the conventional wisdom began to shift to a theory that HIV-1 assembles in multivesicular bodies. These small organelles are involved in the degradation pathway of the cell, Simon explained.“For us, there were two valid possibilities: The virus was assembling in these internal organelles, which then secreted them from the cells, [or] the virus was assembling on the surface, but when a lot of virus was being made in the cell, the cell was internalizing the virus and targeting it for these degradation compartments.” For more than a decade, his group has been developing techniques capable of watching internal compartments of the cell move to the surface and secrete cargo — techniques that are suited to addressing the question of where the HIV-1 virus assembles. The goal was to use optical methods to watch HIV assemble in situ. “Our approach is sensitive enough that we can clearly resolve between single viruses assembling and releasing from the surface and a large number of viruses being delivered together,” Simon explained. The scientists used green fluorescent protein to label Gag, a structural protein of HIV-1, and then engineered HeLa cells to co-express both the tagged and untagged proteins. They chose this route to help reduce the possibility that tagging Gag with GFP would disrupt the labeled protein’s function in the virus. “It was previously reported by another group that if you only expressed labeled Gag, you can detect some defects in viral assembly, but with a mixture of Gag that is labeled and not labeled, it assembled just fine,” Simon explained. “Since we wanted to observe as normal a process as possible, we used the mixture of labeled and unlabeled.” Using total internal reflection microscopy enabled the group to limit the excitation to within about 70 nm (approximately one-seventh the wavelength of light) of the coverslip-medium interface. That dramatically reduced the background fluorescence emitted by labeled Gag protein in the cytoplasm. The investigators recorded two types of fluorescence events at this stage of the experiments: slowly appearing fluorescence points, or puncta, and puncta that rapidly appeared and then rapidly disappeared. Using additional endosome-specific labeling, they determined that the second class of fluorescence events might be associated with endosomes, so they focused further effort on studying the slowly appearing puncta.“Jouvenet [lead author] has now imaged over 20,000 viruses leaving the cell and has only observed them assembling and leaving one at a time,” Simon explained. “She never observed a large number leave at once.” The researchers now want to test to “be sure the spots we were observing on the surface were really assembling virus.” For this, they developed a way to use Förster resonance energy transfer (FRET).Because the Gag molecules are structural proteins within the virus, they become close enough to each other during virus assembly for FRET to take place. The group chose GFP and mCherry because the pair has a relatively weak FRET pathway. The emission and absorption spectra overlap is relatively narrow, so a much higher density of molecules is required for a bright FRET signal. The investigators reported that this particular combination also offers better signal-to-noise ratio because of a much lower background.They found that, when fluorescence first becomes visible, the FRET coefficient is similar to that in areas with diffuse Gag. However, it rapidly increased to a stable level that was 2.5 times that of its initial fluorescence. The investigators then photobleached the mCherry and observed an increase in GFP fluorescence, confirming that FRET was taking place. To ensure that they were not observing a cycling of virus particles within the cytoplasm but actually individual virus particles assembling in the cell, they photobleached the GFP to see whether the fluorescence would recover. Recovery would indicate that new Gag molecules were cycling in. The Gag fluorescence would recover only while the virus was assembling. However, once the fluorescence had reached its steady state, indicating that the nascent virus had completed assembly, the fluorescence would no longer recover. To test whether the virus had separated from a cell, they substituted a variant of GFP that is not fluorescent when the pH is acid. Then they increased the CO2 pressure in the growth medium. An enzyme in the cytosol, carbonic anhydrase, accelerates the acidification reaction and rapidly quenches fluorescence in any GFP exposed to the cytosol. The increase in CO2 pressure was a short pulse. The higher acidity quenched the fluorescence in some of the puncta. However, some of the puncta were much more resistant to quenching. The researchers used an inverted Olympus IX-70 microscope with a 60×, 1.45-NA total internal reflection objective. For excitation, they used a HeNe laser operating at 543 nm or an argon-ion laser operating at 488 nm, both from Melles Griot. They reflected the excitation light off a 488/543 polychroic mirror from Chroma Technologies. They set up the optical train for dual-color imaging with a dual-emission splitter from Optical Insights containing a bandpass filter (515/30 nm) and a long-wave pass filter (580 nm). For image collection, they used a 12-bit, cooled CCD camera from Hamamatsu. They also used the microscope in epifluores?cence mode. For this they used a xenon arc lamp from Ushio in an Opti Quip housing.Simon explained that, to his knowledge, this is the first report of imaging a virus particle as it is released from a cell membrane. The methodology may be useful in a wide variety of applications that study viruses. The group’s research, he explained, is going in two different directions: continuing to study in more detail how HIV-1 assembles and applying the technique to other viruses.“Initially, we will focus on viruses that assemble on the cell surface, but soon we will be trying to apply it to other viruses,” he added.