What is a cell transition

Imaging of HIV-1 Envelope-Induced Virological Synapse and Signaling on Synthetic Lipid Bilayers

Summary

This article describes a method for the formation of an HIV-1-induced envelope of virological synapses on glass supported planar bilayers to be visualized by Total Internal Reflection Fluorescence (TIRF) microscopy. The method can also be combined with immunofluorescent staining to detect the activation and redistribution of signaling molecules that occur during HIV-1-induced envelope virological synapse formation.

Abstract

Human immunodeficiency virus type 1 (HIV-1) infection occurs most efficiently via cell to cell transmission 2,10,11. This cell to cell between CD4 + T cells is the formation of a virological synapse (VS), which is an F-actin-dependent cell-cell transition formed on the engagement of HIV-1 envelope gp120 on the infected cell with CD4 and the chemokine receptor (CKR) CCR5 or CXCR4 on the target cell 8. In addition to gp120 at the receptors, other membrane proteins, especially the adhesion molecule LFA-1 and its ligands, the ICAM family play an important role in the formation and transmission of the virus as seen on the surface of virus-infected donor cells and target cells, as VS. also on the envelope of HIV-1 virions 1,4,5,6,7,13. VS formation is accompanied by intracellular signaling pathways that are transduced as a result of gp120 encroachment on its receptors. In fact, we recently show that CD4 + T cell interaction with gp120 recruitment and phosphorylation of signaling molecules with the TCR signalosome including Lck, CD3ζ, ZAP70, LAT, SLP-76 Itk and PLCγ 15 assigned induced.

In this article, we present a method to visualize supramolecular arrangement and membrane-proximal signaling events in the course of VS formation. We use the glass-assisted planar bi-layer system according to a reductionist model to map the surface of HIV-infected cells, which represent the viral envelope gp120 and the cellular adhesion molecule ICAM-1. The protocol describes general procedures for monitoring HIV-1 gp120-induced VS and signal activation events, which include i) bi-layer preparation and assembly in a flow cell, ii) injection of cells and immunofluorescent staining to detect intracellular signaling molecules interacting on cells with HIV-1 gp120 and ICAM-1 on bi-layers, iii) image acquisition by TIRF microscopy, and iv) analysis of the data. This system generates high resolution images of VS interface in addition achieved with the conventional cell-cell system as it allows depending on capturing different clusters of individual molecular components of VS along with signaling molecules recruited to these sub-domains.

Protocol

1. Label GP120

Fluorescent dyes used in this protocol are effective for binding to proteins, are sufficiently stable for photo-imaging microscopy, and correspond to the excitation wavelengths of the laser with our microscope. These three criteria must be met when selecting fluorescent molecules for tagging proteins.

  1. His stock exchange gp120 DH12 (a gift from Dr. Michael Cho, of Iowa State University) protein buffer to sterile PBS with a centrifugal filter unit (30 kDa MWCO) 6 day. Make sure gp120 concentration is not more than 1 mg / ml. Note: His 6-marked gp120 proteins from other X4 or R5 tropic strains were also tested and are now commercially available from Immune Technologies.
  2. Add sodium bicarbonate pH 9.0 in a protein solution to obtain 50 mM final concentration of sodium bicarbonate at ~ pH 8.5.
  3. Add amine-reactive Alexa Fluor 488 (AF488) fluorescent dye that is dissolved in water at a 10-fold molar excess. Incubate this mixture for 1-2 hours at room temperature in the dark.
  4. To remove the excess color, use the centrifugal filter unit as before and with fresh sterile PBS on the column. Repeat this step until all free dye is removed (3-4 washes with 4 ml PBS). Turn down until gp120 is concentrated to about 1 mg / ml. Note: The removal of free dye by dialysis or centrifugal filter unit only works if the dye is water-soluble, or if gel filtration is used.
  5. To measure the fluorescence intensity per molecule of protein (F / P) of the labeled protein, the concentrations (in µM) of the fluorescent dye at the appropriate wavelength (e.g. at 488 nm for AlexaFluor 488) and protein (we use NanoDrop a spectrophotometer using the 'Proteins and Labels' setting) and then dividing these two numbers to give you the F / P ratio. The same procedure is used for other proteins and dyes, such as ICAM-1 and Cy5. Note: Others Types of spectrophotometers can get the protein concentration and fluorescent dye concentration in order to get the F / P ratio when a NanoDrop is not available.

2. Flow module and bilayer preparation

The general procedure for flow cell and bilayer fabrication in detail above 14 has been described. Here we specifically outline the protocol on bilayers with His 6-marked Prepare gp120 DH12 on a Bioptech flow cell. For studies with infectious viruses or infected cells and when small amounts are required due to the limited reagents, a disposable flow chamber ibidi (sticky foil I 0,2 Luer) can be used and followed with the same two-layer preparation procedure of 2.4-2.9 steps. Information for ibidi flow chambers can be obtained from the company's website, http://www.ibidi.com/service/display_material/CA_8p_EN_150dpi.pdf

  1. Prepare Pirhana solution (45 ml sulfuric acid (trace metal grade 98%) and 15 ml 30% hydrogen peroxide) 14. Using plastic clamps, dip the cover in the Pirhana solution for 15 minutes.
  2. Transfer cover slips to a beaker filled with pica-purified water and wash under running water for 2 minutes (1 minute on each side). Place on the shelf to dry.
  3. Assemble Bioptech's FCSII chambers as before 14 described.
  4. A liposome mixture containing 12.5% ​​Ni 2+ chelate lipids is for the capture and presentation of His 6-gp120 double layer on the surface 16 used. Add 1 μl drop of the mixture to the liposome-microaqueduct slide without touching the tip to the glass. To prepare the maximum of 5 double layers per measuring cell, repeat up to 4 times more so that there are five 1μl points as shown before 14.
  5. Place coverslip on top of lipids. Cover the white locking ring with a stainless steel clamp base unit, turn the whole apparatus on and seal. Mark location of double layers with a marker. Incubate for 10 minutes at room temperature.
  6. Attach tubing with two-way stopcock to the left perfusion tube as shown in the video. A positive meniscus at the end of the tubing with the 3-way stopcock on it (tubing was previously filled with HEPES-buffered saline (HBS) with human serum albumin (HSA) (HBS / HSA: 50 ml 10x HBS [10x HEPES-buffered Saline: 200mM HEPES, pH 7.2, 1.37M NaCl, 50mM KCl, 7mM Na 2 HPO 4, 60 mM D-glucose] 20 ml HSA 25x, 500 µl 1 M CaCl 2, 1 ml of 1 M MgCl 2 and dH 2 0 up to a total volume of 500ml and filters with 0.22μ m filter) and is free of bubbles. Connect the tube to the right perfusion tube and push the buffer through the measuring cell.
  7. Block bilayer with ~ 300 μl of casein containing 100 μM NiCl 2, by adding 1 ml syringe and attaching it to the open part of the 3-way stopcock. GenTLY push buffers through and close both taps before disconnecting the syringe. Incubate for 30 min at room temperature.
  8. In His 6-marked AF488-labeled gp120 (the concentration is determined as previously in 16 described). We use ~ 250 molecules of gp120 / μm 2, the local density of env clusters on HIV-1 virion area 17 to imitate. We also use ~ 250 molecules / µ m 2 of ICAM-1 on the bilayer as previously reported 14. Place in flow cell with casein as done. Incubate for 30 min in the dark (cover with foil) at room temperature. The same procedure is used to make other proteins like His 12-marked ICAM-1 on the bilayer Cy5-labeled take over.
  9. Wash bilayers with 5 ml of buffer.

3. Quality control of double layers via FRAP

Proteins that are in bilayers must be laterally diffusible. Before adding cells to bilayers, it is important to insure the mobility of proteins in the prepared bilayer. Here we describe a fluorescence recovery after photobleaching (FRAP) method 3, the manual can be done on most objectively illuminated TIRF, wide field epifluorescence, or laser scanning confocal microscopes. The goal is to image completely uniform fields of fluorescence in the lipid bilayer, spot bleach, and to monitor the return of the quenched molecules of the bleached. A 50% recovery within 2 minutes may be acceptable. All major microscope manufacturers (Leica, Nikon, Olympus, Zeiss) sell objectively illuminated TIRF systems that work well for this application. While every business offers variations of TIRF lighting and other operational features, the image quality is essentially the same.

  1. Use a low stimulation intensity to get a fading. Use a high numerical aperture goal such as an NA of 1.3 to 1.45, 40x, 60x, or 100x. When using a standard fluorescence or TIRF microscope to reduce the light intensITY, put gray filters in the excitation light path. Most laser systems can be set for low illumination by setting an acousto-optic filter (AOTF) or acousto-optic beam splitter (AOBS) with low voltage.
  2. Take a "pre-bleach" picture. When using a cooled CCD camera that can be set to binning 2x2 or 4x4 to compensate for poor lighting conditions, the gain can be set to high or long exposure times (roughly on the order of seconds) can be used. When using a confocal, the pinhole can be opened; the gain is set high and slow scanning or averaging is used.
  3. Bleach a place. If using a standard fluorescence or TIRF microscope, remove any ND filters to allow for maximum illumination intensity, close the field diaphragm, and expose the sample for a few seconds. When using a confocal laser scanning, set the maximum laser power and zoom in about 4x to 8x to bleach a place. A warning for confocal use: do not zoom in too high, because the bilayer can be damaged by START-UP TIME only concentrated photons.
  4. At 10 to 15 second intervals, take pictures with the same lighting and recording settings as in 3.1 and 3.2. Within two to three minutes, the spot should recover intensity approaching that of the "pre-bleach" image. Rapid recovery is the sign of a successful mobile double shift. If the spot remains dark, and especially if the edge retains high contrast, the fluorescent proteins are immobile in the bilayer and should not be used for the experiment.

Note: In our current microscope we can deliver 0.9 mW of 641 nm light to a circular spot with an area of ​​240 μm 2, the fluorescent ICAM bleaches in less than 4 seconds at typical concentrations. Readers should find that critical alignment of an Hg or Xe lamp with bleach times less than 30 seconds is more than sufficient to perceive this test in a repeatable manner. We would also refer you to that 3 referring for additional det is missing.

4. Injection of cells and immunofluorescent stains

  1. Warm up flow cell to 37 ° C (this can be done in a 37 ° C incubator without CO 2 carried out for about 30 minutes).
  2. In activated human CD4 + T cells (2-5x10 6 / Flow cell) in HBS / HSA ~ 400 μl from 1 ml syringe. Allow cells to attach bilayer for ~ 45 min in the 37 ° C incubator. When the ibidi chambers are deployed, additional buffers are delivered every 15-20 min.
  3. Attach the cells by injection of 2% paraformaldehyde and incubate for 10 minutes at 37 ° C
  4. Wash 3x with 1 ml PBS buffer room temperature. Permeabilize cells by injecting 0.1% Triton X-100 for 5 min at room temperature.
  5. Wash with 1 ml room temperature PBS buffer (if probing for phosphorylated proteins, you can add sodium vanadate at 1mm final concentration or another phosphatase inhibitor in PBS to prevent phosphate removal during processing) 3x. Lock using casein with 5% goat serum for 25 minutes at room temperature. The type of animal serum depends on the secondary antibody.
  6. Wash 3x with 1 ml PBS buffer room temperature. In primary antibody in ~ 300 μl buffer / flow cell for 30 min - 1 h at room temperature. To capture the initial membrane-proximal activation signal, we use specific antibodies against phosphorylated Lck (PLCK), total Lck, or Fyn.
  7. Wash 3x with 1 ml PBS buffer room temperature. Carried out in suitable fluorescently labeled secondary antibodies in buffer as for primary antibodies. For our experiments we use Alexa Fluor 568-labeled goat anti-rabbit secondary antibody. Incubate for 20 min at room temperature. Wash 3x with 1 ml PBS buffer.
  8. NOTE: It is important to have a control bilayer that is treated with secondary antibodies only (not a primary antibody). Follow steps 4.1-4.5 and skip 4.6, then go to step 4.7. This determines the amount of non-specific binding of the secondary antibody. Alternatively, one can use labeled primary antibodies directly.

5. Image acquisition by TIRF microscopy

TIRF microscopy enables excitation of fluorescence signals that are only at 250 nm or thinner level at the glass substrate and cell interface. This guarantees image acquisition only signals from the lipid bilayer and from the directly attached cell membranes. Therefore only molecules at the synapse are mapped. Since TIRF only enter the membrane-proximal area at the bottom of the cell, proteins that are internalized or distributed in the dorsal membrane will not be recognized. It may then be important to additionally acquire images from standard wide-angle or confocal microscopy to determine accumulation or distribution of proteins in the synaptic area, as compared to the other volumes of the cell.

All major microscope manufacturers (Leica, Nikon, Olympus, Zeiss) sell objectively illuminated TIRF systems that work well for this application. While every business offers variations of TIRF lighting and other operational features, the image quality is essentially the same. Each TIRF microscope has its own specifications for operation, but the following general rules apply to obtaining high-resolution imaging.

  • Lighting should be low to avoid bleaching.
  • The camera should have a linear response.
  • The camera should be able to image a wide dynamic range with no saturation.
  • All settings should be constant for quantitative analysis.

6. Data analysis

To get the intracellular signals through CD4 + To examine T cell interaction with gp120 activated in the VS, quantitative analyzes of TIRF images will be carried out to determine whether intracellular signaling molecules such as Lck and Fyn are activated and adjusted to the synapse 15 to determine. Here we describe a simple method for most image analysis application packages such as ImageJ and Metamorph. We have ImageJ, which runs on Windows, Macintosh, and Linux, for our method here 12 used.

in the Analyze> Set Tab Measurements, the Check box, as well as the mean gray value. If you make a region of interest on the picture that a polygon or a freehand shape is traced closed and no> measure, the software will put the data from this measurement in a table of results. The area is expressed in number of pixels or in μm 2 be. The mean is the average intensity in arbitrary units. It has to be subtracted from it via background intensity. Multiplying the mean minus the background area gives the integrated intensity of the protein in arbitrary units.

  1. Make the measurements understand and environment.
  2. Open pictures for an experimental state.
  3. For each cell, track and measure the region of interest (mean) and track and measure an area outside the area of ​​interest (in the background).
  4. If made with an experimental condition, either copy and paste measurements into Excel or other spreadsheet programs, or save the measurements.
  5. Go back to 6.2 for each experimental condition.
  6. When the measurements are complete, calculate the results. The formulas are:
    Average intensity = average - background
    Integrated Intensity = Medium Intensity * Range

7. Representative results

In order to measure the activation and adjustment of the first membrane-proximal signaling molecules Lck and Fyn as a result of the interaction with gp120 in the VS, primary human CD4 were used + T cells on bilayers carry gp120 and ICAM-1 15 introduced. Cells introduced to bilayers with only ICAM-1 served as a control to define the basal levels of signaling. Specific fluorescence intensity measured within the gp120 contact area for cells on gp120 and ICAM-1 with bilayers and within the total contact area for interaction with cells of the ICAM-1 bilayer. An increase in the average intensity of the fluorescence is a sign of augmented recruitment and activation of the signaling molecule at the VS. This will be demonstrated by a comparable increase in the integrated fluorescence intensity. However, if there is no change in the integrated fluorescence intensity, this indicates a redistribution of the signaling molecule.

After the cells with bilayers interacted with gp120 and ICAM-1, a total of Lck and PLCK (Y394), the VS interface was recruited and colocalized with gp120 (Fig. 1 & 2). The average total intensity of Lck (Fig. 1) was higher on the bilayer containing both gp120 and ICAM-1 than with ICAM-1 alone, but the integrated intensities (Fig. 1) were similar, suggesting that Lck was in A central cluster is redistributed on CD4 + T cell binding to gp120. However, quantification by PLCK (Y394) (Fig. 2) showed that the mean intensity on bilayers containing gp120 and ICAM-1 was higher than that on the ICAM-1 bilayers alone, and the integrated intensity was higher on bilayers with both gp120 and ICAM- 1 than bilayers on ICAM-1 alone. This shows that while the levels of total Lck at the interface in cells are similar to adherent bilayers on gp120 and ICAM-1 and ICAM-1 alone, gp120 binding increased phosphorylation of the Y394 residue at the Lck activation loop. In contrast, Fyn was not set on the VS (Fig. 3) as Fyn was present in the contact area of ​​cells on ICAM-1 bilayers alone than on bilayers containing both gp120 and ICAM-1. Hence we conclude that Lck, not Fyn, is the active kinase in the HIV-1 gp120-induced VS.

Facts: Membrane-proximal signaling in HIV-1 gp120-induced VS. Images representative cells on the GP120 + ICAM-1 bilayer (upper images) and the ICAM-1 bilayer (lower images) are shown. Fluorescence intensities of the individual cells were quantified in hand-drawn areas of the cell traces represented by the area marked with the yellow line in FIG. Quantification of the mean and integrated intensities detected by TIRF microscopy are presented in the left and right graphs. A total of 30 to 350 cells were quantified for each condition. Bars = 5 um. Data from one of the three replicate experiments are shown.


1. CD4 + T cells were introduced on bilayers with gp120 and ICAM-1 or ICAM-1 alone for 45 min and then fixed and stained total Lck.


2. CD4 + T cells were bilayered introduced with gp120 and ICAM-1 or ICAM-1 alone for 45 min and then fixed and stained for PLCK (Y394).


Figure 3. CD4 + T cells were introduced on bilayers with gp120 and ICAM-1 or ICAM-1 alone for 45 minutes and then fixed and stained for a total of Funen.

Discussion

Previous studies have visualized VS in the cell-cell conjugate system, but these studies do not provide high-resolution images enough to visualize the supramolecular structures at the synapse. In our laboratory we used the glass-assisted planar bilayer system to visualize the surface of the infected cells, the virus envelope gp120 and the cellular adhesion molecule ICAM-1. In conjunction with TIRF microscopy, which detects fluorescence signals within 100-200 nm from the bilayer surface with a high signal-to-noise ratio, we were able to detect supramolecular separation of gp120 from ICAM-1 in the VS. In addition, the standard immunostaining method can be applied to the bilayer system and can also be recognized and quantified here, the specific setting of active Lck, but not Fyn, the gp120 contact area at VS 15. Therefore, the planar bilayer offers an experimental system for high-resolution imaging of the synapse interface in a 2D plane by TIRF microscopy as well as wide-angle or confocal illumination methods. However, the system also has limitations, adjusting ligand mobility, out-of-plane, and fluctuations of biological membranes are represented by planar bilayers. In addition, this is a in vitro system and therefore has other limitations, such as the lack of other membrane molecules that would be present on an infected cell and the cytoskeletal machines that regulate molecular motility and cellular mobility. The dynamics and distribution of the molecules, such as trimers versus monomers of gp120, cannot physiologically be represented on the bilayer core. However, even with these limitations, this system is still very valuable for studying virus-cell or cell-cell interactions, and these methods can serve as a useful guide for researchers trying to detect high-resolution images that are not supramolecular organization can be seen in the conventional cell-cell conjugated system.

Disclosures

The authors declare no conflict of interest.

Acknowledgments

This work was supported by NIH Grants AI071815 (CEH) and the Roadmap Nanomedicine Development Center Award PN2EY016586 (MLD).

Materials

SurnameCompanyCatalog NumberComments
His tagged HIV gp120Gift from Dr. ChoDH12
His tagged HIV gp120Immune TechnologyVarious X4 or R5 tropic
HSAWilliams Medical Company521302Human Serum Albumin 25%
Amicon Ultra Centrifugal FiltersEMD MilliporeUFC803024Ultracel-30K
Lck Rabbit mAbCell Signaling Technology2787
p-Lck Rabbit pAbSanta Cruz Biotechnology, Inc.Sc-101728Tyr 394
Fyn Rabbit mAbEMD Millipore04-353
Alexa Fluor 568 2 ° AbMolecular Probes, Life TechnologiesA11034Goat anti-Rabbit IgG (H + L)
Alexa Fluor 488Molecular Probes, Life TechnologiesA-20000carboxylic acid succinimidyl ester
FCS2 ChamberBioptechs060319-2-03
Microaqueduct SlideBioptechs130119-5
30mm round w / holesBioptechs1907-08-7500.75mm thick
Rectangle gasketBioptechs1907-1422-25014x24 0.25mm thick
Tygon tubingBioptechs202022751/16 "(25ft)
Three-way stopcockBio wheel7328103
Two-way stopcockBio wheel7328102
Sticky Slide I 0.2 LuerIbidi80168
Cover glassesIbidi10812
1ml syringeBD Biosciences309659
DOGS-NTAAvanti Polar Lipid, Inc790404C
DOPCAvanti Polar Lipid, Inc850375C
Glass coverslipBioptechs40-1313-031940mm
TIRF microscopeNikon Instruments
ImageJNational Institutes of Healthhttp://rsbweb.nih.gov/ij/

DOWNLOAD MATERIALS LIST

References

  1. Bastiani, L., Laal, S., Kim, M., Zolla-Pazner, S. Host cell-dependent alterations in envelope components of human immunodeficiency virus type 1 virions.J. Virol. 71, 3444-3450 (1997).
  2. Dimitrov, D.S., Willey, R.L., Sato, H., Chang, L.J., Blumenthal, R., Martin, M.A. Quantitation of human immunodeficiency virus type 1 infection kinetics.J. Virol. 67, 2182-2190 (1993).
  3. Dustin, M. L. Adhesive Bond Dynamics in Contacts between T Lymphocytes and Glass-supported Planar Bilayers Reconstituted with the Immunoglobulin-related Adhesion Molecule CD58.J. Biol. Chem. 272, 15782-15788 (1997).
  4. Fortin, J. F., Cantin, R., Lamontagne, G., Tremblay, M. Host-derived ICAM-1 glycoproteins incorporated on human immunodeficiency virus type 1 are biologically active and enhance viral infectivity.J. Virol. 71, 3588-3596 (1997).
  5. Frank, I., Stoiber, H., Godar, S., Stockinger, H., Steindl, F., Katinger, H. W., Dierich, M. P. Acquisition of host cell-surface-derived molecules by HIV-1.AIDS. 10, 1611-1620 (1996).
  6. Hioe, C. E., Bastiani, L., Hildreth, J. E., Zolla-Pazner, S. Role of cellular adhesion molecules in HIV type 1 infection and their impact on virus neutralization.AIDS Res. Hum. Retroviruses. 14, Suppl 3. S247-S254 (1998).
  7. Hioe, C. E., Chien, P. C., Lu, C., Springer, T. A., Wang, X. H., Bandres, J., Tuen, M. LFA-1 expression on target cells promotes human immunodeficiency virus type 1 infection and transmission.J. Virol. 75, 1077-1082 (2001).
  8. Jolly, C., Kashefi, K., Hollinshead, M., Sattentau, Q. J. HIV-1 cell to cell transfer across an Env-induced, actin-dependent synapse.J. Exp. Med. 199, 283-293 (2004).
  9. Jolly, C., Mitar, I., Sattentau, Q. J. Requirement for an intact T-cell actin and tubulin cytoskeleton for efficient assembly and spread of human immunodeficiency virus type 1.J. Virol. 81, 5547-5560 (2007).
  10. McDonald, D., Wu, L., Bohks, S.M., KewalRamani, V.N., Unutmaz, D., Hope, T.J. Recruitment of HIV and its receptors to dendritic cell-T cell junctions.Science. 300, 1295-1297 (2003).
  11. Pearce-Pratt, R., Malamud, D., Phillips, D.M. Role of the cytoskeleton in cell-to-cell transmission of human immunodeficiency virus.J. Virol. 682898-682905 (1994).
  12. Rasband, W. S. ImageJ. National Institutes of Health. Bethesda, Maryland, USA. (1997).
  13. Rizzuto, C.D., Sodroski, J.G. Contribution of virion ICAM-1 to human immunodeficiency virus infectivity and sensitivity to neutralization.J. Virol. 71, 4847-4851 (1997).
  14. Vardhana, S., Dustin, M. Supported Planar Bilayers for the Formation of Study of Immunological Synapses and Kinapse.J. Vis. Exp. (19), e947-e947 (2008).
  15. Vasiliver-Shamis, G., Cho, M. W., Hioe, C. E., Dustin, M. L. Human immunodeficiency virus type 1 envelope gp120-induced partial T-cell receptor signaling creates an F-actin-depleted zone in the virological synapse.J. Virol. 83, 11341-11355 (2009).
  16. Vasiliver-Shamis, G., Tuen, M., Wu, T., Starr, T., Cameron, T., Thomson, R., Kaur, G., Liu, J., Visciano, M., Li, H ., Kumar, R., Ansari, R., Han, D., Cho, M., Dustin, ML, Hioe, CE Human immunodeficiency virus type 1 envelope gp120 induces a stop signal and virological synapse formation in noninfected CD4 + T cells.J. Virol. 82, 9445-9457 (2008).
  17. Zhu, P., Liu, J., Bess, J., Chertova, E., Lifson, J., Grise, H., Ofek, G., Taylor, K., Roux, K. Distribution and three-dimensional structure of AIDS virus envelope spikes.Nature. 441, 847-852 (2006).