Differential pH-dependent cellular uptake pathways among foamy viruses elucidated using dual-colored fluorescent particles
© Stirnnagel et al.; licensee BioMed Central Ltd. 2012
Received: 20 May 2012
Accepted: 13 August 2012
Published: 30 August 2012
It is thought that foamy viruses (FVs) enter host cells via endocytosis because all FV glycoproteins examined display pH-dependent fusion activities. Only the prototype FV (PFV) glycoprotein has also significant fusion activity at neutral pH, suggesting that its uptake mechanism may deviate from other FVs. To gain new insights into the uptake processes of FV in individual live host cells, we developed fluorescently labeled infectious FVs.
N-terminal tagging of the FV envelope leader peptide domain with a fluorescent protein resulted in efficient incorporation of the fluorescently labeled glycoprotein into secreted virions without interfering with their infectivity. Double-tagged viruses consisting of an eGFP-tagged PFV capsid (Gag-eGFP) and mCherry-tagged Env (Ch-Env) from either PFV or macaque simian FV (SFVmac) were observed during early stages of the infection pathway. PFV Env, but not SFVmac Env, containing particles induced strong syncytia formation on target cells. Both virus types showed trafficking of double-tagged virions towards the cell center. Upon fusion and subsequent capsid release into the cytosol, accumulation of naked capsid proteins was observed within four hours in the perinuclear region, presumably representing the centrosomes. Interestingly, virions harboring fusion-defective glycoproteins still promoted virus attachment and uptake, but failed to show syncytia formation and perinuclear capsid accumulation. Biochemical and initial imaging analysis indicated that productive fusion events occur predominantly within 4–6 h after virus attachment. Non-fused or non-fusogenic viruses are rapidly cleared from the cells by putative lysosomal degradation. Quantitative monitoring of the fraction of individual viruses containing both Env and capsid signals as a function of time demonstrated that PFV virions fused within the first few minutes, whereas fusion of SFVmac virions was less pronounced and observed over the entire 90 minutes measured.
The characterized double-labeled FVs described here provide new mechanistic insights into FV early entry steps, demonstrating that productive viral fusion occurs early after target cell attachment and uptake. The analysis highlights apparent differences in the uptake pathways of individual FV species. Furthermore, the infectious double-labeled FVs promise to provide important tools for future detailed analyses on individual FV fusion events in real time using advanced imaging techniques.
KeywordsRetrovirus Foamy virus Entry Disassembly Intracellular targeting Time-lapse microscopy Live-cell imaging
As virus replication is strictly dependent on infecting susceptible cells, viruses have evolved several strategies to enter their host. Within the host, their journey of amplification is initiated by adsorption to specific cellular receptor molecules on the target cell surface. Often, two kinds of binding events are involved. In a first attachment step, viruses are concentrated at the cell surface. This process is relatively nonspecific, frequently involving cell surface carbohydrate structures and surfing along cellular protrusions such as filopodia [1, 2]. Following this first step, a second, more specific interaction with the specific cellular receptor of proteinaceous, lipid or carbohydrate nature promotes viral entry. Depending on the virus species, different cellular uptake pathways are exploited (reviewed in [3, 4]). Membrane-enveloped viruses can penetrate host cells by either viral glycoprotein-mediated fusion at the plasma membrane or inducing endocytic uptake (reviewed in [5–7]). Viral capsids released by fusion at the plasma membrane have to break through the actin matrix . Subsequent to this internalization process, free capsids are further transported towards the cell center along microtubules by hijacking cellular motor proteins like dynein or dynactin [3, 9, 10]. In contrast, endocytosed viruses are challenged to release their capsid into the cytosol before the endosomal content is delivered to lysosomes, where degradation occurs. To overcome this endosomal trap, some viruses take advantage of the pH conditions inside endosomes. The low pH of mature or late endosomes can trigger the fusion activity of viral glycoproteins, subsequently activating capsid release by merging the viral and the endosomal membranes .
Whereas the uptake pathways of some viruses are well-defined, we are only at the beginning of understanding how foamy viruses (FVs), a special type of complex retroviruses, infect host cells. Like all members of the Retroviridae, FVs integrate their genome into the host cell chromosomes. Besides this classical feature of retroviruses, other replication steps used by FVs are distinct from orthoretroviruses (e.g. HIV-1), but bear homology to hepadnaviruses. Therefore, FVs are classified into a separate subfamily, the Spumaretrovirinae. FVs are characterized by an extremely broad host-range. The nature of their ubiquitous receptor, which seems to be evolutionarily conserved, has not yet been conclusively determined . Recently, involvement of proteoglycans and heparan sulfate as attachment receptors for FVs was reported as they greatly enhance target cell susceptibility towards these viruses [13–15]. Previously, we had reported that all glycoproteins of a variety of FV species display a pH-dependent fusion activity peaking around pH 5.5, when examined using a cell-cell fusion assay . However, one species, the prototype FV (PFV), also displayed significant fusion activity at neutral pH. The pH-dependent fusion activity of all analyzed FV Env proteins and the sensitivity to lysosomotropic agents (e.g. Bafilomycin A1) suggest an endocytic entry mechanism of FVs . Additional support for this assumption comes from early reports showing endosomal SFVmac (macaque simian FV) uptake in infected cells by using electron microscopy .
Like other viruses, FVs can also hijack the cellular cytoskeleton for intracellular trafficking of incoming viral capsids. This piggyback transport along microtubules is thought to be achieved by direct interaction of the PFV Gag protein with the light chain 8 (LC8) of the dynein motor protein complex  after the fusion process. Furthermore, it was also reported that intact naked PFV capsids accumulate at the MTOC, which presumably disassemble later on at the centrosomes [19–21]. These observations led to the assumption that PFV capsid-envelope separation already occurs upon the route of viral particles to the centrosomes. Currently, it is not known whether the FV glycoprotein mediated fusion happens at the plasma membrane or after endocytosis.
To gain insights into the uptake processes leading to release of the capsid into the cytosol, we generated infectious double-tagged FV particles composed of eGFP-tagged PFV capsids and mCherry-labeled envelope proteins of PFV or SFVmac. The uptake and trafficking processes of these two types of fluorescent FV particles upon target cell entry were analyzed by biochemical methods in bulk populations as well as time-lapsed wide-field and confocal imaging analysis in individual living cells.
Development and characterization of single and double-labeled FV particles
The fusion of Ch to PFV or SFVmac Env only marginally influenced the relative infectivity of extracellular viruses (Figure 3B, bar 1, 3, 14, 16). In contrast, comparison of infectivities of particles harboring the authentic mCherry-tagged wild type PFV Env (Figure 3B, bar 1, 3) to those containing the respective fusion-defective PFV Env (iCS) variants (Figure 3B, bar 2, 4) revealed a 5,000-fold difference. Similarly a 1,000-fold difference was observed for the corresponding wild type (Figure 3B, bar 14, 16) and fusion-defective (Figure 3B, bar 15, 17) glycoproteins of SFVmac.
Double-tagged particles composed of Gag-eGFP and Ch-Env showed a 10 to 100-fold reduction of viral titers (Figure 3B, bar 1, 7 PFV; bar 14, 20 SFVmac). This is in accordance with previous reports that showed that different types of retroviral particles composed of only Gag-eGFP had a similarly decreased infectivity [13, 27]. We also could confirm our previous data , showing that cotransfection of untagged Gag:Gag-eGFP can rescue the infectivity defect up to almost wild-type levels in the case of PFV Env (Figure 3B, bar 9, 11) and up to 40% using SFVmac Env (Figure 3B, bar 22, 24). Thus, viral functions of double-tagged particles, and the infectivity in particular, are predominantly influenced by the modification of the Gag protein.
Next, we determined the fraction of double-labeled FV particles with respect to the total number of Gag-eGFP particles using fluorescence microscopy. Purified viral particles were allowed to settle on coverslips, and the fluorescent intensities of individual particles in the green channel (Gag-eGFP) and red channel (Ch-Env) were measured (Figure 3C). The percentage of Gag-eGFP signals colocalizing with Ch-Env was calculated for all double-labeled preparations used (Figure 3D). The fraction of double-tagged viruses (with respect to all particles having a Gag-eGFP signal) was found to be between 93 and 97% for all preparations. This is not surprising as the cognate Env protein is required for FV budding . In contrast, only about 50% of all red-labeled particles contained Gag-eGFP (Figure 3C). Attaching a fluorescent protein to the N-terminus of a FV glycoprotein strongly increased the release of capsidless subviral particles  that are characterized by Ch-Env only signals (data not shown).
Our approach relies on the fact that FV glycoproteins can be labeled with FP without affecting the functionality of the Env protein. The use of FP-labeled glycoproteins is rare (e.g. [2, 30, 31]) as Env-FPs are typically non-functional. Other approaches for the generation of double-labeled retroviruses using FP-tags have been reported in the past. For example, HIV-1 has been double-labeled by employing membrane-targeting signal-tagged FP proteins or FP-tagged Vpr co-packaged into HIV-1 particles [32, 33]. However, the FV Env-FP system developed in this study holds great promise for the further elucidation of the kinetics and dynamics of processes in the life cycle of retroviruses.
Differential cell-cell fusion characteristics of various FV Env containing virions
We used the double-labeled FVs to gain insights into viral entry and trafficking in infected cells. The entry pathway of foamy virus particles containing different Env proteins (PFV, SFVmac) in living cells was followed over 24 h using time-lapsed wide-field microscopy.
These observations point to a restriction of SFVmac fusion events to intracellular organelles. Fusion from mature and late endosomes would be consistent with the pH-dependent fusion properties of SFVmac Env described previously . In line with this, Bafilomycin A1 blocked SFVmac Env-mediated transduction of HeLa cells more efficiently than PFV Env-mediated transduction (see Additional file 2). As the PFV glycoprotein displays significantly higher fusion activity at neutral pH, we propose that PFV possesses a fusion-active state already at the plasma membrane where it then can induce syncytia formation [16, 34]. Since Chloroquine treatment, an inhibitor of lysosome acidification [35, 36], did not dramatically decrease FV infectivity, we assume that fusion from lysosomes does not play an important role in FV entry (see Additional file 2). The analysis of other drugs interfering with the function of cellular kinases (Genistein, ), dynamin (Dynasore, ) or drugs leading to cholesterol depletion (Nystatin, ) did not reveal strong infectivity-decreasing effects as was observed for Bafilomycin A1 (see Additional file 2). With these compounds, the infectivity was either slightly reduced (Chloroquine, Dynasore, Nystatin) or remained essentially unchanged (Genistein). For comparison, VSV-G enveloped HIV pseudoparticles were measured under identical conditions. The differential effects of all drugs on the VSV-G- and the FV Env-mediated infectivity suggest that FV entry triggers a different virus entry pathway than that of the VSV-G envelope.
Time-lapsed wide-field imaging of FV uptake in individual host cells
Next, we examined the potential uptake and fate of fusion-incompetent PFV Env (Figure 5C and Additional file 5) or SFVmac Env viral particles. The target cell attachment and initial uptake routes of both types of non-fusogenic viruses were quite similar to the corresponding wild-type viruses. This indicates that the Env fusion activity is not essential for FV attachment and early uptake into host cells and suggests that endocytic uptake pathways are exploited. Within the first hour after exposure of HeLa cells to double-labeled fusion-incompetent PFV particles (PE Ch iCS), we observed accumulation of viruses inside the cell in a similar manner to that of wild-type viruses (Figure 5C, 1 h). However, at 2–4 h pb, almost no separate Gag-eGFP signal was observable inside cells incubated with iCS virus and none of the cells (0%, n = 17) examined in detail showed aggregates of naked capsids at the MTOC, represented as punctate structures close to the nucleus (compare Figure 5C, 4 h to Figure 5A, 4 h). Furthermore, no tethering of Gag-eGFP to chromatin in mitotic cells was detectable (Figure 6B, see Additional file 4). This points to a rapid degradation of non-fusogenic viruses that are unable to release their capsid into the cytoplasm. Therefore, we assume that the productive fusion events occur predominantly within the first 4 h pb and particles that do not succeed in undergoing fusion within this time period are prone for rapid degradation.
In summary, the fusion-incompetent viruses were characterized by three major differences in comparison to the respective wild-type viruses. First, in the case of PFV Env, cells incubated with iCS virions never showed formation of cell-cell fusion (data not shown). Second, no capsid aggregates (Gag-eGFP only signals) were observed in cells exposed to both types of fusion incompetent viruses and the presence of individual capsids was rare (Figure 5C, 4 h, 12 h, 24 h). Third, Gag did not associate with the host cell chromosomes upon cell division (Figure 6B, see Additional file 4).
Incoming FV capsids accumulate at the MTOC
Degradation of FV
These results also indicate that the fusogenic entry pathway of FV occurs predominantly within the first 4–6 h pb. Viruses that were unable to undergo Env-mediated fusion with cellular membranes within this time period are prone to degradation.
Influence of Bafilomycin on the viral entry pathway
To further test whether SFVmac Env mediated entry, contrary to PFV Env, is restricted to endocytic uptake, we analyzed the affect of Bafilomycin A1 (BAF) on viral entry. BAF is an inhibitor of endosomal pH acidification  and would artificially retain SFVmac Env containing viruses in endosomes by preventing fusion on the one hand and viral degradation on the other hand. The fate of viral proteins of wild type or fusion-deficient SFVmac Env containing particles was examined biochemically upon infection of target cells treated with BAF. In these cell lysates, both Gag-eGFP and Ch-LP were detected by immunoblot until 24 h pb (Figure 8E, F). Furthermore, the occurrence of free mCherry proteins was neither observed in SE Ch nor SE Ch iCS viruses, indicating that the presence of free mCherry in BAF-untreated cells is observed concomitantly with degradation of the incoming Ch-LP of the virus.
Thus, inhibiting endosomal acidification and protein degradation through BAF incubation of cells results in retaining SE Ch particles in endosomal compartments, underlining the involvement of endocytosis in FV uptake.
If Env-mediate fusion is triggered by the lower pH value in endosomes, artificial elevation of the endosomal pH by BAF-treatment should also prevent escape of capsids into the cytoplasm and thus no aggregation of naked capsids at the MTOC in target cells should be observable. This was examined by time-lapsed wide-field imaging analysis of different FVs (PE Ch, SE Ch, SE Ch iCS) in BAF-treated target cells over 12 h (see Additional file 6). All three types of viruses were taken up into the cells, accumulated in the perinuclear region and dual-colored signals were observed until 12 h pb (compare Additional file 6 to Additional file 3, Additional file 5 and Figure 5). In BAF-treated cells incubated with SE Ch or SE Ch iCS particles, no accumulation of naked capsids at the MTOC was detectable. In contrast, cells incubated with PE Ch particles in the presence of BAF still showed the characteristic Gag-GFP punctate structures that represent centrosomal accumulations of naked capsids. Taken together, these results confirm the strictly pH-dependent fusion process of SFVmac Env containing particles after endocytic uptake. In addition, they demonstrate that PFV Env containing particles escape the inhibitory effect of BAF-treatment to a large extent, probably due to viral fusion at neutral pH.
Confocal time-lapsed analysis of colocalized capsid and Env signal in live cells
In order to get a general and more quantitative impression about the level of fusion activity of the different FV species and the time scale of viral entry and fusion, we evaluated the fraction of enveloped capsids in individual cells over time. Experiments were performed in live cells where the uptake of viruses could be followed in the same subset of cells, thereby decreasing one factor of variability within the measurement. We developed a software program to perform a global 3D colocalization analysis suitable for the data from the live-cell experiments. This program is based on the detection of signals in the Gag-eGFP channel and the presence of a corresponding signal in the opposite Ch-Env channel. In order to avoid artifacts arising from autofluorescence or channel crosstalk, 3D stacks of live-cell images were acquired with alternating excitation [43, 44] using a spinning-disk confocal microscope. Low particle numbers per cell were used to avoid random colocalization that would lead to false positive events (see Additional file 7). We also incorporated the full 3D image information obtained from the z-stack in the analysis. For more detailed information, see material and methods. From the analysis, the total number of particles in each channel was obtained as well as the number of colocalizing dual-color particles.
The percentage of colocalizing particles was always determined with respect to the number of detected Gag-eGFP particles. This approach minimizes artifacts arising from quenching of the eGFP signal at low pH. When the eGFP signal of the double-tagged virus particles is quenched, the particle will not be included in the analysis. In addition, quenching of the FP signal was not a significant issue for this sample. We measured the effect of low pH on our virus particles and only a slow decay over several minutes of the Gag-eGFP signal at pH 5.5 was observed (data not shown). No quenching of the Env-labeled mCherry was observed at low pH (data not shown).
A slight increase in the colocalization percentage of PE Ch particles was observable between 30 to 50 minutes pb. Although the increase is near the limit of statistical relevance, formation of large aggregates of both Gag and Env signal in the perinuclear region made it impossible to analyze individual particles in this region. Both the limited sensitivity in the perinuclear region as well as capsid disassembly would lead to a decrease in the detection of green only particles (Figure 5) and thus to an increase in the colocalization percentage. The second decay, starting after about 50 minutes, may result from a different entry pathway that becomes more prominent at this time point, e.g. fusion with (late) endosomes in contrast to fusion at the plasma membrane. However, this is currently speculation.
In summary, the observed colocalization percentage for PE Ch and SE Ch particles was always below the non-fusogenic control sample PE Ch iCS. For PE Ch particles, a significant drop in the colocalization percentage was observed within the first 10 minutes, whereas for SE Ch particles, the decay was slower and occurred throughout the length of the measurement. The slower kinetics observed for SE Ch particles can be, at least in part, attributed to the requirement of endosomal acidification to trigger the fusion process. In contrast, PFV particles were shown to already possess a significant fusion activity at neutral pH , which is consistent with our observation of early fusion events. Additionally, these findings are in agreement with the observation of syncytia formation, which was only observed for cells incubated with PE Ch particles (Figure 4A). Based on these findings, the timescale to expect most fusion events would be expected to occur during the first 30 min post attachment.
The generated FVs with FP-tagged capsids and glycoproteins provide an excellent tool for investigating the early steps of viral entry. We demonstrated that, in individual living cells, the attachment and uptake of viral like particles is independent of the fusion activity of the viral glycoprotein. The majority of fusion events appear to occur within the first two hours post entry. Virions that haven’t released their capsids into the cytosol within the first six hours are prone for degradation. In line with previous reports [18, 20], capsids released into the cytoplasm accumulate at the MTOC, and Gag proteins gain access to the cellular genome upon mitosis. Our results suggest that there are differences in the uptake pathways of various FV species determined predominantly by the type of FV glycoprotein utilized. PFV Env containing virions release their capsids, to a large extent, within the first few minutes after binding. This suggests that PFV can fuse with the plasma membrane, which is supported by the fusogenic activity of the Env protein at neutral pH and the high cell-cell fusion activity observed. In contrast, SFVmac Env containing viruses appear to require endocytosis and acidification for fusion to occur. Thus, the double-FP-tagged FVs introduced in this study provide a very powerful tool for detailed analyses of the early steps of FV entry and promise to be useful for visualizing the fusion process itself.
The human kidney cell line 293T , the human fibrosarcoma cell line HT1080  and the human cervical HeLa cell line  were cultivated in Dulbecco’s modified Eagle’s medium supplemented with 10% heat-inactivated fetal bovine serum (FBS) and antibiotics. During live-cell imaging assays, HeLa cells were kept in Leibovitz’s L15 medium, supplemented with 10% FBS.
The 4-component PFV vector system consisting of the PFV Gag expression vector pcoPG4, the PFV Pol expression vector pcoPP, the PFV Env expression construct pcoPE (PE), and the enhanced green fluorescent protein (eGFP)-expressing PFV transfer vector puc2MD9 or pMD11, encoding lacZ as a reporter gene, has been described previously [13, 48, 49]. In some experiments, the corresponding not codon-optimized expression constructs pcziGag4, pcziPol and pciSE (SE) [16, 48] were used. Expression vectors for FV mCherry-Env fusion proteins were cloned by fusing the FP tag sequence connected by a flexible glycine-serine (G/S) linker to the N-terminus of PFV Env in pcoPE (pcoPE Ch) or SFVmac Env in pciSE (pciSE Ch) (see Figure 2). Further modification of these Env expression constructs by insertion of the amino acid exchanges R571T in PFV Env  or RKRR570-573AAEA in SFVmac Env  resulted in the generation of the corresponding SU-TM cleavage site mutants (pcoPE Ch iCS, pciSE Ch iCS). All FP-tagged expression constructs were generated using standard PCR cloning techniques and mutagenesis primers and were verified by sequencing analysis. Details of the cloning procedure and primer sequences are available upon request.
The pgTubulin-DsRed expression plasmid encoding a gamma-tubulin-dsRed fusion protein was obtained from Euroscarf .
Recombinant PFV particles and HIV pseudoparticles were essentially produced and harvested from polyethylenimine (PEI) transfected cells as described previously [13, 52, 53]. Briefly, PFV containing supernatants were generated by cotransfection of 293T cells with transfer vector puc2MD9, Pol- (pcoPP), Env- (pcoPE, pcoPE Ch, pcoPE Ch iCS) and Gag packaging plasmid (pcoPG4, pcoPG4 CeGFP) at a ratio of 28:2:1:4. SFVmac containing supernatants were produced by cotransfection of puc2MD9, pcziPol, pcziGag4 (or pcziGag-CeGFP) and Env packaging plasmids (pciSE, pciSE Ch, pciSE Ch iCS) at a ratio of 1:1:1:1. For live-cell imaging experiments, the transfer vector pMD11 instead of puc2MD9 and the enzymatic inactive reverse transcriptase encoding Pol packaging plasmid pcoPP2  instead of pcoPP were used. At 24 h post-transfection, sodium butyrate (final concentration, 10 mM) was added to the growth medium. At 8 h post induction, the cell culture medium was replaced and, after an additional 16 h, viral supernatants were harvested.
Analysis of transduction efficiency
Transduction of host cells by HIV pseudoparticles, PFV or SFVmac Env containing viral supernatants was performed by infection of 2 x 104 HT1080 cells, plated 24 h in advance in 12-well plates. During the incubation period (4–6 h), target cells were covered with 1 ml of the viral supernatant or dilutions thereof prior to media replacement. The percentage of eGFP-positive cells was determined by flow cytometry analysis 72 h after infection. All transduction experiments were performed three times and, in each independent experiment, the titers obtained with the untagged wild-type viruses were arbitrarily set to 100% and those of the other samples expressed as values relative to the wt control, as described previously .
In some experiments, target cells (HeLa) were incubated at different concentrations with Bafilomycin A1 (Sigma-Aldrich), Chloroquin-Diphosphat (AppliChem), Genistein (Sigma-Aldrich), Nystatin (Merck), Dynasore (Sigma-Aldrich). After a 1 h preincubation period at 37°C, HeLa cells were exposed to PFV or SFVmac Env containing FVs or VSV-G enveloped HIV pseudoparticles for 4 h in the presence of drugs. The supernatant was substituted with fresh drug-containing cell culture medium for an additional hour before cultivation in medium without drug. Seventy-two hours later, the percentages of EGFP-expressing HeLa cells were determined by flow cytometry.
Viral particle purification for immunoblotting
Ten milliliters of cell-free viral supernatant were harvested by sterile filtration (pore size, 0.45 μm) and centrifuged at 4°C and 25,000 rpm for 3 h in an SW40 (Beckman) rotor through a 20% sucrose cushion. Subsequent to centrifugation, the supernatant was discarded and the viral pellet was resuspended in 100 μl 1xPPPC (sodium dodecyl sulfate (SDS) protein sample buffer).
Cell lysates, antisera, immunoblotting
A transfected 10 cm dish was prepared for cell lysis by incubation with 600 μl lysis buffer (10 mM Tris–HCl (pH 8), 140 mM NaCl, 0,025% NaN3, 1% TritonX-100) and subsequently centrifugation through QIAshredder (Qiagen) columns. Cell lysates and purified particles were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and analyzed by immunoblotting. The polyclonal antisera used were specific for PFV Gag , the LP of PFV Env (aa 1 to 86) , the LP of SFVmac Env (aa 2–69) or mCherry (399C). The monoclonal antisera used were raised against PFV SU (P3E10)  or eGFP (Roche). In some experiments, specific antibodies raised against the cellular housekeeping protein GAPDH (Glyceraldehyde 3-phosphate dehydrogenase) were used (Sigma). The chemiluminescence signal was digitally recorded using a LAS-3000 imager.
Production of antisera
SFVmac LP specific antiserum (SFVmac Env (aa 2–69)) was generated by insertion of a PCR fragment encoding SFVmac Env aa2-69 in-frame downstream of the maltose binding protein (MBP) ORF of the prokaryotic pMAL-C2 expression vector (New England Biolabs). The mCherry-specific antiserum (399C) was produced by insertion of a mCherry-decaHis fusion protein ORF (His, Histidine) into the prokaryotic pET11 expression vector (Novagen). The soluble fusion proteins were expressed in Escherichia coli TB1 or BL21(DE3) cultures after induction with 0.5 mM isopropylthiogalactopyranoside (IPTG) for 3 to 6 h respectively and affinity purified according to the manufacturer’s instructions.
Analysis of viral proteins in infected cells
HeLa cells were seeded at a density of 5 x 104 cells/well in 12 well plates one day prior to the experiments. If the assay was performed in the presence of Bafilomycin A1 (BAF), the target cells were preincubated with BAF (60 nM) for 1 h at 37°C. After pre-cooling to 12°C, the cells were incubated with 1 ml virus-containing cell culture media (+/− BAF (60 nM)), that had been harvested and concentrated (20x) by low speed centrifugation (14,000g, 1.5 h, 4°C). Following 30 minutes incubation at 12°C, the cells were either rinsed with PBS and prepared for cell lysis (0 h pb) or the media was replaced by fresh 10% DMEM (+/− BAF (60 nM)) prior to warming the cells to 37°C (1–24 h pb). At the given time points after the cell had reached 37°C, the cells were rinsed with PBS and incubated with 2xPPPC followed by centrifugation through QIAshredder (QIAgen) columns. Cell lysates were separated by SDS-polyacrylamide gel electrophoresis (PAGE) and analyzed by immunoblotting.
Purification and concentration of FV particles for imaging analysis
Fluorescent FV particles were produced as described above using the pMD11 transfer vector. Subsequent to ultracentrifugation (25,000 rpm) of 30 ml cell culture supernatant in a SW28 rotor (Beckman), the viral pellet was gently resuspended in 90 μl PBS supplemented with 10% FBS resulting in a 333x volume concentration. In some experiments, viral particles were harvested by using Pierce® concentrators (150K, Thermo Scientific). In that case, the cell culture media of transfected 293T cells was substituted with phenol red- and FBS-free DMEM after sodium-butyrate induction. With this method, an equal concentration factor was obtained. The viral particles were stored as aliquots at −80°C.
Wide-field live-cell microscopy
HeLa cells were seeded at a density of 0.5 x 104 cells/200 μl into one well of an eight-chamber slide (IBIDI, cat. No: 80826). The cell culture media was replaced 24 h later by cold L15 media supplemented with 10% FBS (+/− BAF (60 nM)) and the cells were cooled to 10°C (5 min). Afterwards, 3–5 μl of purified fluorescent FV particle preparations were added for 30 minutes at 10°C. Subsequently, cells were washed twice with cold L15 media (+/− BAF (60 nM)). Cells were then warmed to 37°C, and live-cell microscopy was started immediately. A z-stack was collected every 15 minutes using 600 nm spacing between consecutive planes and 10 to 12 planes total. Wide-field images were collected on a Nikon TE2000E using a Nikon Plan Fluor 40x (numerical aperture 1.3) oil immersion objective. The light of a mercury lamp was used to alternately excite eGFP (BP470/30) and mCherry (BP590/20). The emission signals were passed through a dichroic mirror and a 465/50 or 545/60 bandpass filter prior to detection with a CCD camera. The movies and images were evaluated with ImageJ (http://rsb.info.nih.gov/ij/).
Spinning-disk confocal microscopy
One day prior to the measurements, HeLa cells were seeded at a density of 2 x 104 cells/400 μl in one well of an eight-chamber slide (Lab-Tek). Prior to virus incubation, the cell-culture medium was replaced by L15 medium and the cells were cooled to ~10°C (10 min). 1–3 μl of purified virus were added in the vicinity of the HeLa cells to be measured and allowed to bind for an additional 10 minutes at ~10°C. Subsequently, the cells were rinsed with cold L15 medium and the imaging was started immediately after warming the cells to 37°C. The spinning-disk confocal microscope system (Revolution System; Andor Technology) utilized a Nikon microscope base (TE2000E) and the spinning-disk unit CSU10 from Yokogawa. Measurements were performed with an oil immersion total internal reflection fluorescence (TIRF) objective (60x, NA = 1.49, Nikon) in combination with a 1.5x tube lens. The detection path was equipped with an Optosplit II (Cairn Research Ltd.) for dual-color detection, a filter set for eGFP and mCherry (BS562, HC525/50 and ET605/70; AHF Analysentechnik AG) and a DU-897 Ixon EMCCD camera (Andor). In addition, a triple-band dichroic beam splitter was used to separate laser excitation from fluorescence emission (Di01-T405/488/568/647; Semrock). The excitation was controlled with an acousto-optic tunable filter (AOTF). The sample position was controlled with an xyz piezo stage (ProScan II, NanoScanZ; Prior Scientific). Multi-fluorescent beads (TetraSpeck microspheres, 0.1 μm, Invitrogen) immobilized on a coverslip were used to calibrate the overlap of the two detection channels. Multiple cells were measured sequentially during one experiment by recording the xy positions of several cells and automatically moving the xy stage to the appropriate positions during the experiment. The corresponding time interval between z-stacks for each cell was varied between 5 and 10 minutes.
2D colocalization analysis
An analysis program was developed in house to determine the amount of colocalization within a 2D image based on an intensity ratio of the particles. Particles in each channel were fitted to a 2D Gaussian and selected by the criteria of particle size, intensity threshold and minimal distance between two neighboring particles. Colocalization was determined based on an intensity ratio of the particles detected in the green channel with respect to the intensity in the red channel. Particles with a ratio around one were defined as colocalized.
3D colocalization analysis
For the 3D colocalization analysis, a software program was developed in house to determine the amount of colocalization from a z-stack of images based upon the minimum distance between particles detected in the green and red channels. Particles were initially detected using a spot-enhancing filter and an intensity threshold within the three-dimensional image volume. The position of each particle was estimated by calculating the center of mass of the fluorescence intensity for each spot. The image plane containing the highest intensity coming from the particle was taken and the lateral position of the particle determined by fitting the intensity to a 2D Gaussian function. The axial-position was taken as the position of the z-plane with the maximum intensity, which was accurate to within ±150 nm. The green and red channels were recorded alternatively. As autofluorescence has a very broad excitation spectrum, fluorescence structures that had a strong signal in the red channel after 488 nm excitation were assigned as cellular autofluorescence and excluded from the analysis to avoid false positive colocalizations. Particles were detected independently in the green and red channel for each z-plane and z-stack based on an intensity threshold, particle size criteria and a minimal distance between two neighboring particles. As the two channels were recorded alternatively with ~150 ms delay, the positions of the green and the red signals could differ slightly due to motion of the dual-color particle. This shift in position was taken into account by calculating the distance between the signals detected in the two channels and introducing a maximally allowed displacement of 2.2 μm in the xy plane and 600 nm between the z-planes. We chose the separation tolerance to be relatively high to ensure that colocalizing particles that are undergoing transport are identified as colocalizing particles and do not yield false positive fusion results. Thus, a detectible decrease in colocalization percentage would clearly imply that fusion is occurring. Colocalization of particles in different channels was based on the three-dimensional separation of the particles.
Confocal laser scanning microscopy
HeLa cells were seeded at a density of 2 x 104 cells/ml into 12-Well plates on glass cover slips. After 24 h, the cells were transfected with 0.5 μg of the pgTubulin-DsRed expression plasmid using FuGENE® HD transfection reagent according to the manufacturers instructions. Another 24 h later, the transfected cells were precooled and incubated on ice with Gag-eGFP fluorescent SFVmac particle preparations for 30 minutes. Subsequently, the cells were washed with cold PBS and either fixed with 3% PFA or incubated an additional 2 or 6 h at 37°C prior to fixation. Following DAPI staining, the samples were covered in Mowiol. Confocal laser scanning images were obtained on a Zeiss LSM 510 as described previously and evaluated by ImageJ .
Electron microscopy analysis
HeLa cells were seeded at a density of 1x106 cells/well in 6 well plates one day prior to measuring. After precooling, the cells were incubated with untagged wildtype PFV particles (MOI 10) produced as described above using the 4-component vector system. After 30 min incubation, the cells were either fixed or shifted to 37°C for an additional 10 or 30 min prior to fixation. The cells were harvested and processed for electron microscopy analysis as described previously .
We wish to thank Monika Franke for assistance with cell culture and Doreen Streichert for technical assistance for electron microscopy analysis. We gratefully acknowledge the financial support of the DFG through SPP1175 (D.L. LI 621/4-1, LI 621/4-2; D.C.L. LA1971/1-2) and Individual Grant (D.L. LI 621/3-3), the German Excellence Initiative via “Nanosystems Initiative Munich (NIM)” (D.C.L.) and the Ludwig-Maximilians-University Munich (LMUInnovativ BioImaging Network, D.C.L.).
- Smith AE, Helenius A: How viruses enter animal cells. Science. 2004, 304: 237-242. 10.1126/science.1094823.View ArticlePubMed
- Lehmann MJ, Sherer NM, Marks CB, Pypaert M, Mothes W: Actin- and myosin-driven movement of viruses along filopodia precedes their entry into cells. J Cell Biol. 2005, 170: 317-325. 10.1083/jcb.200503059.PubMed CentralView ArticlePubMed
- Brandenburg B, Zhuang X: Virus trafficking - learning from single-virus tracking. Nat Rev Microbiol. 2007, 5: 197-208. 10.1038/nrmicro1615.PubMed CentralView ArticlePubMed
- Thorley JA, McKeating JA, Rappoport JZ: Mechanisms of viral entry: sneaking in the front door. Protoplasma. 2010, 244: 15-24. 10.1007/s00709-010-0152-6.PubMed CentralView ArticlePubMed
- Kumari S, Mg S, Mayor S: Endocytosis unplugged: multiple ways to enter the cell. Cell Res. 2010, 20: 256-275. 10.1038/cr.2010.19.View ArticlePubMed
- Mercer J, Schelhaas M, Helenius A: Virus entry by endocytosis. Annu Rev Biochem. 2010, 79: 803-833. 10.1146/annurev-biochem-060208-104626.View ArticlePubMed
- Grove J, Marsh M: The cell biology of receptor-mediated virus entry. J Cell Biol. 2011, 195: 1071-1082. 10.1083/jcb.201108131.PubMed CentralView ArticlePubMed
- Taylor MP, Koyuncu OO, Enquist LW: Subversion of the actin cytoskeleton during viral infection. Nat Rev Microbiol. 2011, 9: 427-439. 10.1038/nrmicro2574.PubMed CentralView ArticlePubMed
- Dohner K, Nagel CH, Sodeik B: Viral stop-and-go along microtubules: taking a ride with dynein and kinesins. Trends Microbiol. 2005, 13: 320-327. 10.1016/j.tim.2005.05.010.View ArticlePubMed
- McDonald D, Vodicka MA, Lucero G, Svitkina TM, Borisy GG, Emerman M, Hope TJ: Visualization of the intracellular behavior of HIV in living cells. J Cell Biol. 2002, 159: 441-452. 10.1083/jcb.200203150.PubMed CentralView ArticlePubMed
- Linial ML: Foamy viruses are unconventional retroviruses. J Virol. 1999, 73: 1747-1755.PubMed CentralPubMed
- Lindemann D, Goepfert PA: The foamy virus envelope glycoproteins. Curr Top Microbiol Immunol. 2003, 277: 111-129. 10.1007/978-3-642-55701-9_5.PubMed
- Stirnnagel K, Luftenegger D, Stange A, Swiersy A, Mullers E, Reh J, Stanke N, Grosse A, Chiantia S, Keller H, et al: Analysis of prototype foamy virus particle-host cell interaction with autofluorescent retroviral particles. Retrovirology. 2010, 7: 45-10.1186/1742-4690-7-45.PubMed CentralView ArticlePubMed
- Nasimuzzaman M, Persons DA: Cell membrane-associated heparan sulfate is a receptor for prototype foamy virus in human, monkey, and rodent cells. Mol Ther. 2012, 20: 1158-1166. 10.1038/mt.2012.41.PubMed CentralView ArticlePubMed
- Plochmann K, Horn A, Gschmack E, Armbruster N, Krieg J, Wiktorowicz T, Weber C, Stirnnagel K, Lindemann D, Rethwilm A, Scheller C: Heparan sulfate is an attachment factor for foamy virus entry. J Virol. 2012, 86: 10028-10035. 10.1128/JVI.00051-12.PubMed CentralView ArticlePubMed
- Picard-Maureau M, Jarmy G, Berg A, Rethwilm A, Lindemann D: Foamy virus envelope glycoprotein-mediated entry involves a pH-dependent fusion process. J Virol. 2003, 77: 4722-4730. 10.1128/JVI.77.8.4722-4730.2003.PubMed CentralView ArticlePubMed
- Dermott E, Samuels J: Electron microscopic observations on the mechanisms of entry of simian foamy virus in HEp-2 cells. J Gen Virol. 1973, 19: 135-139. 10.1099/0022-1317-19-1-135.View ArticlePubMed
- Petit C, Giron ML, Tobaly-Tapiero J, Bittoun P, Real E, Jacob Y, Tordo N, De The H, Saib A: Targeting of incoming retroviral Gag to the centrosome involves a direct interaction with the dynein light chain 8. J Cell Sci. 2003, 116: 3433-3442.View ArticlePubMed
- Saib A, Puvion-Dutilleul F, Schmid M, Peries J, de The H: Nuclear targeting of incoming human foamy virus Gag proteins involves a centriolar step. J Virol. 1997, 71: 1155-1161.PubMed CentralPubMed
- Lehmann-Che J, Renault N, Giron ML, Roingeard P, Clave E, Tobaly-Tapiero J, Bittoun P, Toubert A, de The H, Saib A: Centrosomal latency of incoming foamy viruses in resting cells. PLoS Pathog. 2007, 3: e74-10.1371/journal.ppat.0030074.PubMed CentralView ArticlePubMed
- Lehmann-Che J, Giron ML, Delelis O, Lochelt M, Bittoun P, Tobaly-Tapiero J, de The H, Saib A: Protease-dependent uncoating of a complex retrovirus. J Virol. 2005, 79: 9244-9253. 10.1128/JVI.79.14.9244-9253.2005.PubMed CentralView ArticlePubMed
- Lindemann D, Pietschmann T, Picard-Maureau M, Berg A, Heinkelein M, Thurow J, Knaus P, Zentgraf H, Rethwilm A: A particle-associated glycoprotein signal peptide essential for virus maturation and infectivity. J Virol. 2001, 75: 5762-5771. 10.1128/JVI.75.13.5762-5771.2001.PubMed CentralView ArticlePubMed
- Campbell RE, Tour O, Palmer AE, Steinbach PA, Baird GS, Zacharias DA, Tsien RY: A monomeric red fluorescent protein. Proc Natl Acad Sci USA. 2002, 99: 7877-7882. 10.1073/pnas.082243699.PubMed CentralView ArticlePubMed
- Gross LA, Baird GS, Hoffman RC, Baldridge KK, Tsien RY: The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc Natl Acad Sci USA. 2000, 97: 11990-11995. 10.1073/pnas.97.22.11990.PubMed CentralView ArticlePubMed
- Verkhusha VV, Sorkin A: Conversion of the monomeric red fluorescent protein into a photoactivatable probe. Chem Biol. 2005, 12: 279-285. 10.1016/j.chembiol.2005.01.005.View ArticlePubMed
- Hendrix J, Flors C, Dedecker P, Hofkens J, Engelborghs Y: Dark states in monomeric red fluorescent proteins studied by fluorescence correlation and single molecule spectroscopy. Biophys J. 2008, 94: 4103-4113. 10.1529/biophysj.107.123596.PubMed CentralView ArticlePubMed
- Muller B, Daecke J, Fackler OT, Dittmar MT, Zentgraf H, Krausslich HG: Construction and characterization of a fluorescently labeled infectious human immunodeficiency virus type 1 derivative. J Virol. 2004, 78: 10803-10813. 10.1128/JVI.78.19.10803-10813.2004.PubMed CentralView ArticlePubMed
- Fischer N, Heinkelein M, Lindemann D, Enssle J, Baum C, Werder E, Zentgraf H, MÃÂ¼ller JG, Rethwilm A: Foamy virus particle formation. J Virol. 1998, 72: 1610-1615.PubMed CentralPubMed
- Stanke N, Stange A, LÃÂ¼ftenegger D, Zentgraf H, Lindemann D: Ubiquitination of the prototype foamy virus envelope glycoprotein leader peptide regulates subviral particle release. J Virol. 2005, 79: 15074-15083. 10.1128/JVI.79.24.15074-15083.2005.PubMed CentralView ArticlePubMed
- Koch P, Lampe M, Godinez WJ, Muller B, Rohr K, Krausslich HG, Lehmann MJ: Visualizing fusion of pseudotyped HIV-1 particles in real time by live cell microscopy. Retrovirology. 2009, 6: 84-10.1186/1742-4690-6-84.PubMed CentralView ArticlePubMed
- Sherer NM, Lehmann MJ, Jimenez-Soto LF, Ingmundson A, Horner SM, Cicchetti G, Allen PG, Pypaert M, Cunningham JM, Mothes W: Visualization of retroviral replication in living cells reveals budding into multivesicular bodies. Traffic. 2003, 4: 785-801. 10.1034/j.1600-0854.2003.00135.x.View ArticlePubMed
- Lampe M, Briggs JA, Endress T, Glass B, Riegelsberger S, Krausslich HG, Lamb DC, Brauchle C, Muller B: Double-labelled HIV-1 particles for study of virus-cell interaction. Virology. 2007, 360: 92-104. 10.1016/j.virol.2006.10.005.View ArticlePubMed
- Campbell EM, Perez O, Melar M, Hope TJ: Labeling HIV-1 virions with two fluorescent proteins allows identification of virions that have productively entered the target cell. Virology. 2007, 360: 286-293. 10.1016/j.virol.2006.10.025.PubMed CentralView ArticlePubMed
- Pietschmann T, Zentgraf H, Rethwilm A, Lindemann D: An evolutionarily conserved positively charged amino acid in the putative membrane-spanning domain of the foamy virus envelope protein controls fusion activity. J Virol. 2000, 74: 4474-4482. 10.1128/JVI.74.10.4474-4482.2000.PubMed CentralView ArticlePubMed
- Toole-Simms W, Sun IL, Morre DJ, Crane FL: Transplasma membrane electron and proton transport is inhibited by chloroquine. Biochem Int. 1990, 21: 761-769.PubMed
- Misinzo G, Delputte PL, Nauwynck HJ: Inhibition of endosome-lysosome system acidification enhances porcine circovirus 2 infection of porcine epithelial cells. J Virol. 2008, 82: 1128-1135. 10.1128/JVI.01229-07.PubMed CentralView ArticlePubMed
- Akiyama T, Ishida J, Nakagawa S, Ogawara H, Watanabe S, Itoh N, Shibuya M, Fukami Y: Genistein, a specific inhibitor of tyrosine-specific protein kinases. J Biol Chem. 1987, 262: 5592-5595.PubMed
- Macia E, Ehrlich M, Massol R, Boucrot E, Brunner C, Kirchhausen T: Dynasore, a cell-permeable inhibitor of dynamin. Dev Cell. 2006, 10: 839-850. 10.1016/j.devcel.2006.04.002.View ArticlePubMed
- Chen Y, Wang S, Lu X, Zhang H, Fu Y, Luo Y: Cholesterol sequestration by nystatin enhances the uptake and activity of endostatin in endothelium via regulating distinct endocytic pathways. Blood. 2011, 117: 6392-6403. 10.1182/blood-2010-12-322867.View ArticlePubMed
- Mullers E, Stirnnagel K, Kaulfuss S, Lindemann D: Prototype Foamy Virus (PFV) Gag nuclear localization - A novel pathway among retroviruses. J Virol. 2011, 85: 9276-9285. 10.1128/JVI.00663-11.PubMed CentralView ArticlePubMed
- Tobaly-Tapiero J, Bittoun P, Lehmann-Che J, Delelis O, Giron ML, de The H, Saib A: Chromatin tethering of incoming foamy virus by the structural Gag protein. Traffic. 2008, 9: 1717-1727. 10.1111/j.1600-0854.2008.00792.x.View ArticlePubMed
- Bayer N, Schober D, Prchla E, Murphy RF, Blaas D, Fuchs R: Effect of bafilomycin A1 and nocodazole on endocytic transport in HeLa cells: implications for viral uncoating and infection. J Virol. 1998, 72: 9645-9655.PubMed CentralPubMed
- Kapanidis AN, Laurence TA, Lee NK, Margeat E, Kong X, Weiss S: Alternating-laser excitation of single molecules. Acc Chem Res. 2005, 38: 523-533. 10.1021/ar0401348.View ArticlePubMed
- Ruthardt N, Lamb DC, Brauchle C: Single-particle tracking as a quantitative microscopy-based approach to unravel cell entry mechanisms of viruses and pharmaceutical nanoparticles. Mol Ther. 2011, 19: 1199-1211. 10.1038/mt.2011.102.PubMed CentralView ArticlePubMed
- DuBridge RB, Tang P, Hsia HC, Leong PM, Miller JH, Calos MP: Analysis of mutation in human cells by using an Epstein-Barr virus shuttle system. Mol Cell Biol. 1987, 7: 379-387.PubMed CentralView ArticlePubMed
- Rasheed S, Nelson-Rees WA, Toth EM, Arnstein P, Gardner MB: Characterization of a newly derived human sarcoma cell line (HT-1080). Cancer. 1974, 33: 1027-1033. 10.1002/1097-0142(197404)33:4<1027::AID-CNCR2820330419>3.0.CO;2-Z.View ArticlePubMed
- Gey G, Coffmann W, Kubicek M: Tissue culture studies of the proliferative capacity of cervical carcinoma and normal epithelium. Cancer Res. 1952, 12: 264-265.
- Heinkelein M, Dressler M, Jarmy G, Rammling M, Imrich H, Thurow J, Lindemann D, Rethwilm A: Improved primate foamy virus vectors and packaging constructs. J Virol. 2002, 76: 3774-3783. 10.1128/JVI.76.8.3774-3783.2002.PubMed CentralView ArticlePubMed
- Mullers E, Uhlig T, Stirnnagel K, Fiebig U, Zentgraf H, Lindemann D: Novel functions of prototype foamy virus gag glycine- arginine-rich boxes in reverse transcription and particle morphogenesis. J Virol. 2011, 85: 1452-1463. 10.1128/JVI.01731-10.PubMed CentralView ArticlePubMed
- Heinkelein M, Rammling M, Juretzek T, Lindemann D, Rethwilm A: Retrotransposition and cell-to-cell transfer of foamy viruses. J Virol. 2003, 77: 11855-11858. 10.1128/JVI.77.21.11855-11858.2003.PubMed CentralView ArticlePubMed
- Gerlich D, Beaudouin J, Gebhard M, Ellenberg J, Eils R: Four-dimensional imaging and quantitative reconstruction to analyse complex spatiotemporal processes in live cells. Nat Cell Biol. 2001, 3: 852-855. 10.1038/ncb0901-852.View ArticlePubMed
- Lindemann D, Bock M, Schweizer M, Rethwilm A: Efficient pseudotyping of murine leukemia virus particles with chimeric human foamy virus envelope proteins. J Virol. 1997, 71: 4815-4820.PubMed CentralPubMed
- Lindemann D, Rethwilm A: Characterization of a human foamy virus 170-kilodalton Env-Bet fusion protein generated by alternative splicing. J Virol. 1998, 72: 4088-4094.PubMed CentralPubMed
- Ho Y-P, Schnabel V, Swiersy A, Stirnnagel K, Lindemann D: A small-molecule-controlled system for efficient pseudotyping of Prototype Foamy Virus vectors. Molecular Therapy. 2012, In press
- Stange A, LÃÂ¼ftenegger D, Reh J, Weissenhorn W, Lindemann D: Subviral particle release determinants of prototype foamy virus. J Virol. 2008, 82: 9858-9869. 10.1128/JVI.00949-08.PubMed CentralView ArticlePubMed
- Duda A, Stange A, Luftenegger D, Stanke N, Westphal D, Pietschmann T, Eastman SW, Linial ML, Rethwilm A, Lindemann D: Prototype foamy virus envelope glycoprotein leader peptide processing is mediated by a furin-like cellular protease, but cleavage is not essential for viral infectivity. J Virol. 2004, 78: 13865-13870. 10.1128/JVI.78.24.13865-13870.2004.PubMed CentralView ArticlePubMed
- Krausslich HG, Facke M, Heuser AM, Konvalinka J, Zentgraf H: The spacer peptide between human immunodeficiency virus capsid and nucleocapsid proteins is essential for ordered assembly and viral infectivity. J Virol. 1995, 69: 3407-3419.PubMed CentralPubMed
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.