Virion stiffness regulates immature HIV-1 entry
© Pang et al.; licensee BioMed Central Ltd. 2013
Received: 21 June 2012
Accepted: 29 December 2012
Published: 10 January 2013
Skip to main content
© Pang et al.; licensee BioMed Central Ltd. 2013
Received: 21 June 2012
Accepted: 29 December 2012
Published: 10 January 2013
Human immunodeficiency virus type 1 (HIV-1) undergoes a protease-mediated maturation process that is required for its infectivity. Little is known about how the physical properties of viral particles change during maturation and how these changes affect the viral lifecycle. Using Atomic Force Microscopy (AFM), we previously discovered that HIV undergoes a “stiffness switch”, a dramatic reduction in particle stiffness during maturation that is mediated by the viral Envelope (Env) protein.
In this study, we show that transmembrane-anchored Env cytoplasmic tail (CT) domain is sufficient to regulate the particle stiffness of immature HIV-1. Using this construct expressed in trans with viral Env lacking the CT domain, we show that increasing particle stiffness reduces viral entry activity in immature virions. A similar effect was also observed for immature HIV-1 pseudovirions containing Env from vesicular stomatitis virus.
This linkage between particle stiffness and viral entry activity illustrates a novel level of regulation for viral replication, providing the first evidence for a biological role of virion physical properties and suggesting a new inhibitory strategy.
The main structural component of an HIV-1 particle is the viral Gag polyprotein, which polymerizes to form a protein shell surrounded by a lipid membrane. Expression of Gag alone is necessary and sufficient for viral particle assembly and budding . The viral surface contains the envelope protein (Env), which is synthesized as a precursor (gp160) that is cleaved by a cellular protease into receptor-binding (gp120) and transmembrane (gp41) subunits. gp120 and gp41 form a noncovalent complex that mediates viral entry [2, 3]. The Env transmembrane subunit of HIV-1 and other lentiviruses has an unusually long (~150 amino acids) cytoplasmic tail (CT) domain compared to other retroviruses (~20-30 amino acids). Gag interacts with Env via CT, which aids Env localization to viral budding sites and efficient incorporation into virions [4–6].
During the viral lifecycle, a virion needs to meet several distinct demands—efficient membrane fusion during entry, particle disassembly to release genetic material, assembly during budding, and stability in the extracellular environment before entry into the next cell. HIV-1 virions initially emerge from infected cells as immature particles. These particles then undergo a maturation process induced by HIV-1 protease cleavage of Gag into several products including three structural proteins: MA, CA and NC . Electron Microscopy (EM) shows that HIV-1 particles undergo a dramatic morphological change from a roughly spherically symmetric immature particle with a thick protein shell to a mature particle with a prominent conical core (capsid) formed by CA . In mature virions, only MA remains associated with the viral membrane, creating a thin protein shell. Because of this striking morphological change and the requirement to address diverse needs throughout their lifecycle, we hypothesized that the physical properties of viral particles would also change during maturation. Atomic force microscopy (AFM) has proven to be uniquely informative for measuring the mechanical properties of viral particles under native conditions. We and others have used AFM to measure the physical properties of several viruses [9–12].
Using AFM, we determined that HIV-1 immature particles are ~14-fold stiffer than mature ones, reflecting a dramatic biophysical change during maturation that we call the “stiffness switch” . Further studies determined that Env CT is required for the stiffness switch, as its deletion softens immature HIV-1 particles almost to the mature level . Immature HIV-1 virions are not infectious due to both entry and post-entry (e.g., integration) defects [14, 15]. Deletion of CT restores the entry activity of immature HIV-1 to the mature level [13–15]. These results suggest a strong inverse correlation between viral particle stiffness and entry activity. However, deletion of CT has been reported to induce conformational changes in gp120, which could affect viral entry activity . Therefore, a cause and effect relationship between particle stiffness and viral entry has not yet been demonstrated, and the link between virion physical properties and biological function remains an open question.
As we previously reported , HIV-1 Env, specifically the CT domain, is necessary to produce stiff immature particles. To discern whether CT alone can stiffen viral particles, we first deleted the Env ectodomain (all of gp120 and most of gp41), leaving the gp41 transmembrane (TM) and CT domains. This construct was poorly incorporated into virions, and introduction of a soluble trimeric coiled-coil ectodomain (to mimic trimeric Env) did not improve incorporation (data not shown). Next, we reasoned that replacing gp120 with a soluble globular protein like GFP might improve incorporation. Using GFP to replace the ectodomain, but leaving TM and CT intact (GFP-TM1) resulted in efficient incorporation (data not shown).
To investigate whether GFP-TM1 can stiffen viral particles similarly to WT Env, we generated immature particles with escalating amounts of GFP-TM1. As with Env, GFP-TM1 incorporation levels were calculated as a CT:Gag ratio. Increasing GFP-TM1 incorporation also gradually stiffens immature particles to beyond the WT immature level (Figure 2B). These data show that CT alone is sufficient to stiffen immature viral particles in a similar manner to WT Env.
Both viral stiffness and entry are controlled by Env. To investigate whether particle stiffness directly regulates viral entry, we separated Env’s stiffness-mediating and entry-inducing domains. GFP-TM1 stiffens viral particles without introducing entry activity (due to its missing ectodomain). ΔCT Env efficiently mediates entry, but contributes little to particle stiffness . These constructs allow us to modulate virion stiffness as an independent variable to determine its effect on viral entry activity.
We produced immature virions with varying amounts of GFP-TM1 (to modulate stiffness) and a fixed amount of ΔCT Env (to provide entry activity). GFP-TM1 and ΔCT Env incorporation levels were measured by WB as CT:Gag and gp120:Gag ratios, respectively. The specific entry activities of viruses with different GFP-TM1 incorporation levels were measured by a viral fusion (BlaM) assay and normalized to corresponding ΔCT viruses with no GFP-TM1.
HIV-1 is generally thought to enter target cells at the plasma membrane, though a recent report suggests endocytosis as the primary entry route . VSV, as a member of the Rhabdoviridae family, is unrelated to HIV-1 and enters target cells via endocytosis. VSV entry is mediated by its Env, VSVg [17–19]. Compared to HIV-1 Env, VSVg incorporates at much higher levels, requires no protease cleavage to mediate entry, and is unlikely to interact with GFP-TM1. Investigating the effect of GFP-TM1 incorporation on entry activity of HIV-1 pseudovirions coated with VSVg therefore provides an independent test of the relationship between particle stiffness and viral entry.
An important assumption of our study is that GFP-TM1 incorporation does not affect viral entry mediated by ΔCT Env or VSVg except by changing particle stiffness. As discussed earlier, GFP-TM1 cannot mediate viral entry itself, and GFP-TM1 is missing the critical self-associating residues of the gp41 ectodomain required for Env trimerization [20, 21]. Nevertheless, there is still the possibility that GFP-TM1 interacts with ΔCT Env or VSVg.
HIV-1 gp41 undergoes significant conformational changes to bring the viral and cellular membranes into close proximity to achieve membrane fusion. A likely mechanism by which particle stiffness affects viral entry activity is that stiffer particles may present a higher energy barrier to membrane deformation during fusion. Determining which step of entry is regulated by particle stiffness may help to further characterize this regulatory mechanism. We also showed that particle stiffness regulates viral entry mediated by the unrelated VSVg. Therefore, targeting particle stiffness (e.g., by preventing the stiffness switch during viral maturation), may represent a novel and general viral inhibitory strategy.
Gag is the major HIV structural protein, and the stiffness switch during HIV-1 maturation likely originates from changes in the Gag shell organization, which are visible by EM as a dramatic decrease in the thickness of the membrane-associated protein layer (thought to only consist of MA). MA relies on its N-terminal myristoyl group for membrane binding, and separation of Gag components from MA was shown to partially bury its myristoyl group, which destabilizes membrane binding and MA’s oligomeric structure [23, 24].
Current estimates of the number of HIV-1 Env spikes present on the viral surface are very low (~10 trimers per virion), compared to up to 2,000 Gag molecules in the immature virion protein shell [25–28]. An interesting question is how so few CT domains can dramatically alter a global property like particle stiffness in ~100 nm particles. We propose two possibilities: 1) CT functions during assembly to seed a specific packing arrangement of the Gag lattice (analogous to seeding of different crystal forms), 2) CT functions via its interaction with Gag to stabilize the whole Gag shell. Future studies may distinguish between these two models by establishing the timing of CT’s regulation of particle stiffness (e.g., during budding vs. in released immature particles).
Interestingly, the fusion activity of mature HIV-1 is also blocked by a cholesterol-binding compound, AME . Virions with CT truncations are AME-resistant . HIV-1 virions are enriched in cholesterol as a consequence of their budding from lipid rafts [31, 32], which raises the question whether lipid membrane components, especially cholesterol, are also important for particle stiffness. It is therefore tempting to speculate that AME may inhibit viral entry by stiffening mature particles, and AME resistance emerges by severing of the link between Env CT and Gag (via CT truncation).
Ultimately, mapping of the CT determinants involved in mediating stiffness and structural studies of the Gag/CT interaction will be required to reveal how CT affects the overall Gag shell organization. Future studies will also examine why HIV regulates entry activity via particle stiffness. Possibilities include prevention of viral entry until particle maturation is complete (to avoid non-productive entry, as suggested by ) or use of the immature virion as a more durable particle to facilitate distant spread of infection.
Overall, these results demonstrate the first direct linkage between a viral physical property and its biological activity. Future mechanistic and functional studies of viral particle stiffness may enable the design of novel entry inhibitors that exploit this linkage.
Plasmids used in this study
ΔEnv HIV-1 genome vector
Viral Env expression vectors
β-lactamase expression vector
GFP-TM1 expression vector
PLAP expression vector
To construct GFP-TM1, the GFP (green fluorescent protein) gene was obtained from pET9a-GFP-C37  by PCR using a 5′ KpnI-containing primer (5′-tctgggtacctagctctggcatggtgagcaagggcgagg) and a 3′ SacI-containing primer (5′-ctcgaggagctcttgtacag). The gp41 TM + CT fragment was obtained from pEBB-HXB2 by PCR using 5′ SacI-containing primer (5′-caatgagctctggcggttggaattggtttaacataacaaattgg) and 3′ BamHI-containing primer (5’-gtcccagataagtgctaaggatc). These two PCR products were annealed at the SacI digestion site. The generated GFP-TM1 fusion fragment was ligated into pEBB-HXB2 to replace the corresponding KpnI-BamHI fragment, which includes Env residues from V44 to L681 (HXB2 numbering).
Pseudovirion particles were produced by cotransfection of 293T cells with ΔEnv Int- HIV-1 genome vector, an Env-expressing vector (WT or ΔCT), pMM310, and pEBB-GFP-TM1. To generate control immature or mature WT virus, 2.5 μg of total DNA (1.23 μg genome vector, 0.819 μg Env expressing vector, and 0.45 μg pMM310) was transfected into ~106 cells using 10 μg polyethylenimine (PEI, Sigma). In most experiments, the amount of only one functional plasmid (Env-expressing vector, pMM310 or pEBB-GFP-TM1) is titrated, and an inactive balancer plasmid is added to keep total plasmid concentration constant in a titration series. The amount of each plasmid used in the control transfection (described above) is defined as 100%. 100% GFP-TM1, VSVg or PLAP input is the same as that of WT or ΔCT Env (0.819 μg). Unless otherwise indicated, the same amount of genome vector (1.23 μg) and pMM310 vector (0.45 μg) was used for all transfections. For Env/GFP-TM1 cotransfections, the Env input was kept constant at 100%, while GFP-TM1 levels were titrated. Media was changed 6 h after transfection to avoid PEI toxicity. Supernatants containing secreted viral particles were collected 30 h post-transfection and filtered through 0.2 μm Acrodisc syringe filters (Pall). Each series of viruses prepared on the same day is defined as one “batch”.
For western blot (WB) analysis of viral concentration and Env incorporation level, virus supernatants were purified by centrifugation through a sucrose cushion (20% sucrose in TNE buffer: 0.1 M NaCl, 1 mM EDTA, 10 mM Tris, pH 7.6) at 20,000 X g for 90 min at 4°C. A caveat of this method is that pelleted samples may also include vesicle-associated Env, which is difficult to distinguish from virion-associated Env. Since intact virions are needed to incorporate BlaM-Vpr, this caveat does not influence entry activity measurements, but may affect the accuracy of measured Env incorporation levels. The pellet was resuspended in SDS-PAGE reducing sample buffer and resolved by SDS-PAGE. WB was developed using rabbit polyclonal anti-CA and anti-VSVg (provided by W. Sundquist), mouse monoclonal anti-gp41 CT antibody (obtained from Chessie 8 hybridoma contributed by G. Lewis, NIH AIDS Research and Reference Reagent Program (ARRRP)), and sheep polyclonal anti-gp120 (contributed by M. Phelan, ARRRP). Secondary antibodies are goat anti-rabbit (IRDye 680, Li-Cor), donkey anti-mouse (IRDye 800, Li-Cor) and Rabbit anti-sheep (IRDye 800, Rockland). Blots were quantified using a Li-Cor Odyssey infrared scanner. GFP-TM1 and WT Env incorporation levels per virion were calculated as a CT:Gag ratio. ΔCT Env and VSVg incorporation levels were calculated as gp120:Gag and VSVg:Gag ratios, respectively. PLAP was detected by rabbit monoclonal anti-PLAP antibody (Abcam, ab16695).
For AFM measurements, virus-containing supernatant was first concentrated as previously described . Briefly, the filtered viral supernatant was pelleted onto a 5 ml cushion of OPTI-PREP (Iodixanol, Sigma) in a SW-28 rotor (21,000 rpm, 90 min, 4°C). ~90% of the upper layer supernatant was then aspirated, and the lower layer supernatant was collected with a syringe, combined with 3–5 volumes of TNE buffer, and concentrated by centrifugal ultrafiltration with a MWCO of 100,000 (Vivaspin 20, Sartorius) three times at 3,000 X g.
For AFM imaging and force measurements, virus particles were attached to HMDS-coated microscope glass slides using previously described methods [13, 40]. All AFM experiments were carried out using a Bioscope with a Nanoscope IV controller (Veeco) equipped with a dimension XY closed loop scanner mounted on an inverted optical microscope (Axiovert 200 M, Carl Zeiss AG). Images of virus particles were acquired in AFM tapping mode in a fluid environment (TNE buffer). Pyramidal silicone nitride probes either mlct (K = 1 N/m, Veeco) or NSC35 (K = 3-7 N/m, Micromasch) were used. The spring constants of the DNP probes were determined experimentally by measuring thermal fluctuations . Since the amplitude of the NSC35 thermal fluctuations was too low, we used the method of Sader et al.  to determine their spring constants.
Virus stiffness was determined based on indentation type experiments, as previously described [9, 13]. Briefly, for each indentation measurement, ~100 force-distance (FD) curves were performed at a scan rate of 0.5 Hz. Viral stiffness was derived mathematically from the slope of the FD curve. The stiffness of the virus was computed according to Hooke’s law on the assumption that our experimental system can be modeled as two springs (the virus and the cantilever) arranged in series. In this study, virus particles had a relatively wide range of point stiffnesses, which required using several types of cantilevers having different spring constants. In order to reduce error in the virus calculated point stiffness, we prefer that the measured point stiffness is not larger than 70% of the cantilever spring constant. If the majority of virus measured point stiffness values are larger than this threshold we use a stiffer cantilever to measure these particles. Particles are selected for measurement based on their appearance in AFM imaging in tapping mode (roughly round particles with 80–120 nm height and little background debris near them). Particles are excluded if particle stiffness rises or drops dramatically during measurement (reflecting poor virion attachment to the surface or irreversible “breaking”, respectively).
The viral entry assay was performed as described . Briefly, viruses mixed with DEAE-Dextran (4 μg/ml) were added onto HOS-CD4-CXCR4 cells (provided by B.Chen), followed by centrifugation at 1800 X g for 30 min at 4°C and then incubated at 37°C for 2 h. After aspiration of unbound viruses, 1 μM ß-lactamase substrate solution (CCF2-AM, Invitrogen) was incubated with cells at 13°C for 17 h. Uncleaved and cleaved CCF2-AM have emission peaks of 520 nm (green) and 447 nm (blue), respectively, under 409 nm excitation. Fluorescent signals from both channels were detected using an Olympus MVX10 fluorescent microscope and quantified using ImageJ software. The ratio of blue vs. green fluorescent signal was calculated as the BlaM assay signal. Each assay was performed in triplicate and their BlaM assay signals were averaged as the entry signal. In this assay, entry signal increases proportionally to the amount of virus added (Figure 10A). gp120 represents functional HIV-1 Env that is cleaved by cellular proteases and capable of mediating viral entry. To investigate the relationship between gp120 concentration and BlaM assay signal, immature viruses bearing different amounts of HIV-1 ΔCT Env were generated and measured for their entry activities. Entry signal per viral particle also increases proportionally with the amount of processed Env (gp120) per viral particle, showing that the BlaM assay signal depends on the processed Env concentration (Figure 10B).
Therefore, viral entry activity was normalized to gp120 levels for each GFP-TM1 titration series. The specific entry activity we report here represents the viral entry activity per processed Env. Relative viral entry activity was calculated by normalization of the entry activity of each virus to that of virus without GFP-TM1. For all measurements, the mean entry activities from at least two independent batches assayed separately are reported.
The BlaM assay signal also increases proportionally to VSVg incorporation level (Figure 10C), so VSV-specific entry activity was also normalized to VSVg levels. Similarly, a varying amount of pMM310 was used during transfection to create a series of viruses bearing titrating amounts of BlaM (detected by rabbit polyclonal anti-BlaM (Chemicon/Millpore)). 2-fold variation of BlaM incorporation level only slightly changed the BlaM assay signal (Figure 10D). Since anti-BlaM Western blots showed less than 2-fold variation in BlaM incorporation levels for all our GFP-TM1 titrating viruses (data not shown), BlaM incorporation level was not considered during normalization.
To investigate whether we can compare the entry activity of viruses from different batches, normalized entry activities of immature VSVg pseudovirions without GFP-TM1 were measured and found to be very similar between different batches (data not shown). Therefore, the entry activity of viruses is reported as a relative value to that of virus with no GFP-TM1 within each series.
TX100 treatment of HIV-1 particles was performed as described previously . Briefly, virus-containing supernatant was first concentrated using the sucrose cushion method described above. The pellet was then resuspended into 0.5% TX100 in TNE buffer and incubated at 4°C for 30 min before centrifugation in a Beckman TLA-55 rotor at 45,000 rpm for 30 min. After centrifugation, the pellet was resuspended in SDS-PAGE reducing sample buffer and analyzed by Western blot. WB was developed using primary antibodies (rabbit anti-CA, mouse anti-gp41 CT and sheep anti-gp120) and secondary antibodies (goat anti-rabbit, donkey anti-mouse and rabbit anti-sheep) described above and quantified using a Li-Cor Odyssey infrared scanner.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License ( http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.