The goal of this protocol is to specifically tag and selectively isolate viral DNA from infected cells for the characterization of viral genome associated proteins.
The goal of this protocol is to isolate herpes simplex virus type 1 (HSV-1) DNA from infected cells for the identification of associated viral and cellular proteins by mass spectrometry. Although proteins that interact with viral genomes play major roles in determining the outcome of infection, a comprehensive analysis of viral genome associated proteins was not previously feasible. Here we demonstrate a method that enables the direct purification of HSV-1 genomes from infected cells. Replicating viral DNA is selectively labeled with modified nucleotides that contain an alkyne functional group. Labeled DNA is then specifically and irreversibly tagged via the covalent attachment of biotin azide via a copper(I)-catalyzed azide-alkyne cycloaddition or click reaction. Biotin-tagged DNA is purified on streptavidin-coated beads and associated proteins are eluted and identified by mass spectrometry. This method enables the selective targeting and isolation of HSV-1 replication forks or whole genomes from complex biological environments. Furthermore, adaptation of this approach will allow for the investigation of various aspects of herpesviral infection, as well as the examination of the genomes of other DNA viruses.
Viruses have a limited capacity to carry out essential functions and therefore depend on host factors to facilitate critical aspects of infection including viral gene expression, replication, repair, recombination, and transport. The activities of these host factors are often augmented by virally encoded proteins. In addition, viruses must avoid detection and interference by cellular responses to viral infection. Therefore, virus host interactions dictate the outcome of infection. Of paramount importance is understanding how viruses alter the cellular environment to adapt the cellular machinery to facilitate viral processes. Of particular interest is identifying which factors and processes act on viral genomes throughout the infectious cycle.
Herpes simplex virus type 1 (HSV-1) is a double stranded DNA virus that infects a substantial proportion of the human population. Within the first hour of infection, the viral genome enters the nucleus, where an ordered cascade of viral gene expression ensues in coordination with viral DNA (vDNA) replication1. In the nucleus, genomes are subject to epigenetic regulation, undergo repair and recombination, and are packaged into capsids, such that the first progeny virions are produced within less than six hours. The comprehensive evaluation of viral genome associated proteins throughout the course of infection will lay the groundwork to investigate the molecular details of processes that act on viral genomes, and will provide insight into which viral and cellular factors are involved in various stages of infection.
Previous methods for the investigation of host factors involved in viral infection include affinity purification of viral proteins for the analysis of associated cellular proteins2,3,4,5,6,7,8,9. These assays have been instrumental for the identification of cellular factors involved in host antiviral responses, as well as viral chromatin modification, gene expression, and DNA repair. However, it is difficult to ascertain whether interactions depend on the association with vDNA, and proteomics only provide insight into interactions that occur as a function of a specific viral factor. Chromatin immunoprecipitation (ChIP) has been used to identify where specific viral and cellular proteins bind to viral genomes10,11,12,13,14,15 and fluorescent in situ hybridization (FISH) combined with immunocytochemistry has enabled the visualization of cellular factors that colocalize with vDNA16,17,18,19,20. These assays allow for spatial and temporal analysis. However, limitations include the need for highly specific antibodies, limited sensitivity, and the need for previous insight into virus host interactions. We therefore developed a method based on iPOND (isolation of proteins on nascent DNA)21 and aniPOND (accelerated native iPOND)22 to selectively label and purify vDNA from infected cells for the unbiased identification of viral genome associated proteins by mass spectrometry. iPOND has been instrumental for the investigation of cellular replication fork dynamics.
For the selective purification of viral genomes from infected cells, replicating vDNA is labeled with ethynyl modified nucleosides, 5-ethynyl-2´-deoxyuridine (EdU) or 5-ethynyl-2´-deoxycytidine (EdC) (Figure 1), followed by covalent conjugation to biotin azide via click chemistry to facilitate single step purification of viral genomes and associated proteins on streptavidin-coated beads (Figure 2B). Importantly, infections are carried out in stationary cells, which are not engaged in cellular DNA replication to enable specific labeling of vDNA. Furthermore, HSV-1 infection causes cell cycle arrest and inhibits cellular DNA replication23,24. Virus can be prelabeled before infection for the analysis of proteins associated with incoming viral genomes (Figure 1A) or labeled during DNA replication for the analysis of proteins associated with newly synthesized vDNA (Figure 1B)25. Furthermore, pulse chase analysis can be used to investigate the nature of proteins associated with viral replication forks (Figure 1C)26. In addition, ethynyl-modified vDNA can be covalently conjugated to a fluorophore for spatial investigation of protein dynamics (Figure 2A and Figure 3). Imaging allows for the direct visualization of vDNA and is a complimentary approach for the validation of vDNA-protein interactions, and can be adapted to track viral genomes throughout infection. We anticipate that these approaches could be further modified to study any aspect of herpesviral infection, including latency and reactivation, and to study other DNA viruses. Moreover, labeling with 5-ethynyl uridine (EU) could allow for the analysis of RNA viral genomes.
1. Cell Culture, Viral Infection, and EdC Labeling (Figure 1)
The following protocol involves working with viruses. Please refer to your institution's bio-safety protocols regarding safe handling of viruses and other biological agents. This protocol was approved by the Institutional Review Board of the University of Pittsburgh.
2. Imaging of EdC labeled DNA (Figure 2A)
NOTE: It is useful to carry out imaging in tandem with the viral genome purification method to verify that cellular DNA is not labeled in experiments. Imaging can also be used to visualize the nature of labeled vDNA and to validate proteomics data. Examples of cellular and viral DNA staining patterns are shown in Figure 3.
3. Purification of vDNA and Associated Proteins
NOTE: Several aspects of this protocol have been adapted from Leung et al.22
NOTE: All buffers and reagents should be chilled on ice before use and all steps should be carried out on ice unless otherwise indicated.
The use of click chemistry for the purification of DNA from cells was first accomplished by the iPOND method21. The purpose of iPOND is to purify cellular replication forks for the identification of associated proteins. We have adapted this technique to specifically study vDNA protein interactions during infection. Manipulation of the approach to label viral genomes with EdC (Figure 1), combined with synchronized infections, has allowed for the selective isolation and investigation of separate populations of vDNA. HSV-1 DNA is labeled with EdC (Figure 1) to facilitate covalent attachment to a fluorophore for imaging or biotin for purification (Figure 2). Virus can be prelabeled before infection for the analysis of proteins associated with incoming genomes (Figure 1A) or labeled during DNA replication for the analysis of proteins associated with newly synthesized vDNA (Figure 1B)25. Furthermore, pulse chase analysis can be used to investigate the nature of proteins associated with viral replication forks (Figure 1C)26. DNA can be visualized within cells to provide spatial information about the nature of the labeled DNA and support for identified vDNA-protein interactions (Figure 3). Taken together, the protocols described here allow for the proteomic investigation of multiple aspects of productive infection to provide insight into dynamic changes that occur on HSV-1 genomes. These approaches could be further adapted to investigate latency and reactivation, to examine phenotypes associated with viral mutants, and to study other DNA and RNA viruses.
Aspects of this protein purification method were adapted from Leung et al.22. Here a major improvement for the purification of EdC labeled DNA is the use of Streptavidin T1 beads instead of streptavidin-coated agarose beads. To optimize DNA purification, several types of streptavidin-coated beads were tested for nonspecific binding when infection was carried out in the absence of EdC and for maximum protein recovery in the presence of EdC (Figure 4A). For these experiments, western blotting was carried out for the vDNA binding protein ICP4. Purifications carried out on Streptavidin T1 beads resulted in the least amount of background binding in the absence of EdC and the maximum protein recovery in the presence of EdC.
Proteins associated with vDNA can be identified by proteomic methods including western blotting and mass spectrometry. Western blotting can be used to investigate the dynamics of known interactions and mass spectrometry can be used as an unbiased approach to identify novel vDNA associated proteins. An example of protein yield as a function of increasing EdC labeling is shown in Figure 4B. A smear, rather than a clear banding pattern, is usually observed by Coomassie Blue staining. This is likely because there is DNA present in the protein sample or because there is an abundance of protein. It is also important to note that the copper catalyst used in the click reaction can cause partial degradation of protein and nucleic acids28,29. By western blotting, similar levels of ICP4 are typically detected in eluates (Step 3.5.5.) after purification of vDNA that was labeled for 60 min with EdC as that present in the same amount of input sample prepared from nuclear lystates (Step 3.3.6.). This information can be used as a guide to determine whether recovery is sufficient to send the remaining sample for mass spectrometric analysis.
Nano liquid chromatography with tandem mass spectrometry (LC-MS/MS) can be carried out as described previously25 to identify proteins associated with purified vDNA. LC-MS/MS data are analyzed by comparing spectral count (SpC) values between the EdC-labeled experimental sample and the corresponding unlabeled negative control. Proteins are considered to be enriched in the experimental sample based on the following criteria: 1) protein has at least 5 spectral counts (SpC) in the experimental sample, 2) protein is not detected in the control or is enriched over the control by at least four-fold based on dividing SpC values, and 3) the protein is enriched over the control in two or more biological replicates. The normalized spectral abundance factor (NSAF: ) can be used to account for differences in molecular weight (MW) and total protein yield. This equation determines the relative abundance of individual proteins within a sample and allows for the direct comparison of two different experimental conditions in which different amounts of vDNA are purified (for example, comparing input viral genomes (Figure 1A) to replicated genomes (Figure 1B)).
There are multiple ways to present protein enrichment data. Some examples include pie charts, Venn diagrams, and protein interaction networks. Pie charts can be used to illustrate the proportion of proteins identified that are known to function in a specific biological process (Figure 5A). However, problems can arise when one protein has more than one biological function, which is often the case. It is also difficult to compare data across different pie charts. Venn diagrams are useful to demonstrate relationships between proteins identified under different experimental conditions, but do not provide any functional information about the proteins identified (Figure 5B). Protein interaction networks portray a visual illustration of potential interactions between identified proteins and provide information about the types of biological processes that are involved (Figure 5C). Several online resources are available for mapping predicted protein-protein interactions including STRING (Search Tool for the Retrieval of Interacting Genes/proteins)30. Online tools are generally equipped to handle proteomics data from a variety of species and provide valuable information about host proteins involved in viral genome mechanics. However, viral proteins are generally not included in databases, which can complicate analysis. Therefore, it is important to consider the major conclusions one would like to convey when deciding how to present proteomics data.
Figure 1: Approaches to label HSV-1 DNA. (A) To assay the state of incoming virus, resting MRC-5 cells in G0 are infected with prelabeled HSV-1 and genomes are assayed at less than 4 hpi to study unreplicated viral DNA. (B) To assay the state of replicated virus, MRC-5 cells are infected with unlabeled HSV-1 and EdC (orange stars) is added to the growth medium of infected cells and incubated for 2 – 4 h after the onset of viral DNA replication (≥4 hpi). (C) To assay viral replication forks, infected cells are pulse labeled with EdC for 5-20 min. Replication forks can then be chased with deoxyC. EdC labeled DNA is orange. This figure has been modified from Dembowski and DeLuca25. Please click here to view a larger version of this figure.
Figure 2: Procedures described in this paper for analysis of EdC labeled DNA. (A) An outline of the method for imaging EdC labeled DNA. Tagged DNA is represented in green. (B) An outline of the method for the purification and downstream analysis of EdC labeled vDNA. This figure has been modified from Dembowski and DeLuca25. Please click here to view a larger version of this figure.
Figure 3: Visualization of EdC labeled DNA. Cells were infected, labeled, and imaged as described in Figure 1 and Figure 2. Representative images of labeled cellular and viral DNA are shown. Cellular DNA colocalizes with Hoechst stain and vDNA with ICP4. Scale bar: 5 µm.This figure was modified from Dembowski and DeLuca25 and Dembowski et al.26. Please click here to view a larger version of this figure.
Figure 4: Representative protein purification results. (A) Comparison of protein yield when purification steps were carried out using several different types of streptavidin coated beads. Infection was carried out in the presence of EdC (+) to compare protein yield or in the absence of EdC (-) to compare background binding. Infections and EdC labeling were carried out as described in Figure 1B and DNA-protein complexes were purified according to the steps outlined in Figure 2B using comparable amounts of different types of streptavidin coated beads. Top panel: Comparison of protein yield after purification on streptavidin coated agarose beads or Streptavidin M-280. Bottom panel: Comparison of protein yield after purification on Streptavidin M-270, M-280, Streptavidin C1, or Streptavidin T1. Western blotting was carried out with an antibody against the viral protein ICP4. Purified ICP4 is shown in the first lane for comparison. Longer exposure of the bottom panel was required to observe a similar intensity of ICP4 in lanes "280" as in the top panel. (B) Representative protein purification results as a function of time of EdC labeling are shown. Infections and EdC labeling were carried out as described in Figure 1B and DNA-protein complexes were purified according to the steps outlined in Figure 2B using Streptavidin T1 beads. EdC labeling was carried out for 0, 20, 40, or 60 min and Coomassie Blue staining (top) and western blotting (bottom) results are shown. Western blotting was carried out with antibodies specific for ICP4, UL42, or GAPDH. The eluate sample was taken from Step 3.5.5. and lysate from Step 3.3.6. The arrow indicates streptavidin, while L indicates protein ladder. Please click here to view a larger version of this figure.
Figure 5: Examples of different ways to present proteomics data. (A) Pie charts summarize proteins that were identified by mass spectrometry of protein eluates associated with viral genomes purified at 6 hpi. Values indicate the number of proteins identified for each functional category. T indicates the total number of proteins identified. Colors indicate the following categories: purple – RNA processing, red – transcription, green – chromatin remodeling, orange – DNA replication, yellow – nuclear transport, teal – cytoskeleton, dark blue – HSV structural proteins, and gray – other/unknown. (B) Venn diagrams depict the overlap of proteins identified to be associated with viral genomes at 6, 8, or 12 hpi. (C) A STRING map depicts proteins enriched on viral replication forks after a 5 min EdC pulse. Human proteins enriched by 5-fold compared to the unlabeled negative control are shown in the functional interaction map, which was generated using STRING30 with data settings to display only high confidence interactions. Gene names were used to map interactions. Circles indicate proteins that function in the same biological process. This figure was modified from Dembowski and DeLuca25 and Dembowski et al.26. Please click here to view a larger version of this figure.
This protocol includes multiple steps which, if not followed carefully, can result in significantly reduced protein yield or contamination with cellular DNA. It is critical that stationary cells are used for all experiments to ensure that cellular DNA is not labeled and purified. This can be confirmed by the absence of cellular DNA polymerases in the protein sample because HSV-1 does not utilize cellular DNA polymerase for genome synthesis. During EdC labeling and nuclei harvesting steps, samples should be staggered to ensure that each is processed equally. During purification steps, care should be taken not to manipulate nuclei by pipetting too much. This can lead to premature lysis of nuclei, as well as sample loss due to sticking to the sides of pipet tips. Furthermore, the sonication step should be optimized for individual sonicators and the amount of streptavidin-coated beads used for biotin binding should be titrated for optimal protein yield.
Several modifications can be made to this protocol and include the use of different viruses or cell types or the addition of a step to crosslink protein to DNA. When choosing a cell type to use for proteomic investigation, it is important to check for the availability of a protein sequence database for that species. A lack of proteomics resources will significantly limit downstream analysis of protein data. Another limitation is the need for a large number of cells, therefore it may be difficult to prepare enough primary cells, such as neurons, for this experiment. This method should be compatible with other DNA or RNA viruses as long as the viral infection does not depend on cellular DNA replication. Furthermore, it would be interesting to explore changes in DNA-protein interactions that occur in HSV-1 mutants. Another modification to this protocol could be the addition of a formaldehyde crosslinking step before isolation of nuclei, which would allow for more stringent wash conditions during purification21. The addition of formaldehyde will not affect the ability to harvest nuclei but may result in decreased protein yield because of difficultly reversing crosslinks during protein elution.
Results from this protocol represent the average state of viral DNA being examined. Therefore, it is difficult to distinguish if complexes are together on the same or separate DNA molecules. vDNA labeling methods described here in combination with high resolution imaging can be used to verify proteomics data and may help to distinguish between different populations of labeled DNA within a cell. Furthermore, ChIP-seq could be used as a follow up method to identify the specific locations of proteins on the viral genome.
The authors have nothing to disclose.
We acknowledge Hannah Fox for help in the preparation of this manuscript. This work was supported by the NIH grant R01AI030612.
MRC-5 cells | ATCC | CCL-171 | |
Fetal Bovine Serum (FBS) | Gibco | 26140-179 | |
Dulbecco's Modified Eagle Medium (DMEM) | Gibco | 12800-082 | substituted with 10% FBS, 2 mM L-glutamine, 12 mM (for growth in flasks) or 30 mM (for growth in dishes) sodium-bicarbinate |
600 cm2 tissue culture dish | Thermo Fisher Scientific | 166508 | |
Tris Buffered Saline (TBS), pH 7.4 | 137 mM NaCl, 5 mM KCl, 491 mM MgCl, 680 mM CaCl, 25.1 mM Tricine | ||
HSV-1 stock | stocks with titers greater than 1×109 PFU/mL work best | ||
Sephadex G-25 column (PD-10 Desalting Column) | GE Healthcare | 17085101 | |
Dimethyl sulfoxide (DMSO) | Fisher Scientific | D128-1 | |
5´-Ethynyl-2´-deoxycytidine (EdC) | Sigma-Aldrich | T511307 | Dissolve in DMSO to prepare 40 mM stock, aliquot, and store at -20 °C |
2´-deoxycytidine (deoxyC) | Sigma-Aldrich | D3897 | Dissolve in water to prepare 40 mM stock, aliquot, and store at -20 °C |
Nuclear Extraction Buffer (NEB) | prepare fresh (20 mM Hepes pH 7.2, 50 mM NaCl, 3 mM MgCl2, 300 mM Sucrose, 0.5% Igepal) | ||
Cell scraper | Bellco glass | 7731-22000 | Autoclave before use |
Trypan blue solution | Sigma-Aldrich | T8154 | |
PBS, pH 7.2 (10x) | 1.37 M NaCl, 27 mM KCl, 100 mM Na2HPO4, 18 mM KH2PO4 (dilute to 1x in sterile water before use) | ||
Copper (II) sulfate pentahydrate (CuSO4-5H2O) | Fisher Scientific | C489 | Prepare 100 mM stock and store at 4 °C for up to 1 month |
(+) Sodium L-ascorbate | Sigma-Aldrich | A4034 | Freshly prepare 100 mM stock and store on ice until use |
Biotin azide | Invitrogen | B10184 | Prepare 10 mM stock in DMSO, aliquot, and store at -20 °C for up to 1 year |
Click Reaction Mix | prepare immediately before use by adding reagents in the indicated order (10 mL: 8.8 mL 1x PBS, 200 mL 100 mM CuSO4, 25mL 10 mM Biotin Azide, 1 mL 100 mM sodium ascorbate) | ||
cOmplete Protease Inhibitor Cocktail | Roche | 11697498001 | Dissolve in 1 mL water to prepare 50x stock, can store at 4 °C for up to 1 week, or directly add 1 pill to 50 mL buffer |
freezing buffer | prepare fresh (7 mL 100% glycerol, 3 mL NEB, 200 μL 50x protease inhibitor) | ||
Buffer B1 | prepare fresh (25 mM NaCl, 2 mM EDTA, 50 mM Tris-HCl pH 8, 1% Igepal, 1x protease inhibitor) | ||
Buffer B2 | prepare fresh (150 mM NaCl, 2 mM EDTA, 50 mM Tris-HCl pH 8, 0.5% Igepal, 1x protease inhibitor) | ||
Buffer B3 | prepare fresh (150 mM NaCl, 2 mM EDTA, 50 mM Tris-HCl pH 8, 1x protease inhibitor) | ||
Vibra Cell Ultra Sonic Processer equipped with a 3 mm microtip probe | Sonics | VCX 130 | |
Cell strainer | Falcon | 352360 | |
Dynabeads MyOne Streptavidin T1 | Life Technologies | 65601 | |
DynaMag-2 Magnet | Life Technologies | 12321D | |
Mini-Tube Rotator | Fisher Scientific | 260750F | |
2X Laemmli sample buffer | Mix 400 mg of SDS, 2 mL 100% glycerol, 1.25 mL of 1 M Tris (pH 6.8), and 10 mg of bromophenol blue in 8 mL water. Store at 4 °C for up to 6 months. Before use add 1 M DTT to a final concentration of 200 mM. | ||
cap locks for 1.5 mL tube | Fisher Scientific | NC9679153 | |
Standard western blotting reagents | |||
NOVEX Colloidal Blue Staining Kit | Invitrogen | LC6025 | |
12-well tissue culture dish | Corning | 3513 | |
Coverslips | Fisher Scientific | 12-545-100 | Autoclave before use |
Microscope slides | Fisher Scientific | 12-552-5 | |
16% paraformaldehyde | Electron Microscopy Sciences | 15710 | dilute to 3.7% in 1x PBS before use |
Bovine serum albumin (BSA) | Fisher Scientific | BP1605 | prepare 3% in 1x PBS |
Permeabilization buffer (0.5% Triton-X 100) | Sigma-Aldrich | T8787 | prepare 0.5% Triton-X 100 in 1x PBS |
Click reaction cocktail – Click-iT EdU Alexa Fluor 488 Imaging Kit | Molecular Probes | C10337 | prepare according to manufactorer's protocol |
Hoechst 33342 | Life Technologies | H1399 | prepare 10 mg/mL in water, can store at 4 °C for up to 1 year |
mouse anti-ICP8 primary antibody | Abcam | ab20194 | Use a 1:200 dilution in 1x PBS |
mouse anti-UL42 primary antibody (2H4) | Abcam | ab19311 | Use a 1:200 dilution in 1x PBS |
Goat anti-mouse 594-conjugated secondary antibody | Life Technologies | a11005 | Use a 1:500 dilution in 1x PBS |
Immu-mount | Thermo Fisher Scientific | 9990402 | |
2x SDS-bicarb solution | 2% SDS, 200 mM NaHCO3 | ||
Phenol:chloroform:isoamyl alcohol | mix 25:24:1 at least 1 day before use, store at 4 °C in the dark | ||
chloroform:isoamyl alcohol | mix 24:1 at least 1 day before use, store at room temperature in the dark | ||
10x Tris EDTA (TE), pH 8.0 | 100 mM Tris, 10 mM EDTA (dilute to 1x before use) | ||
Qiaquick PCR purification kit | Qiagen | 28104 | |
3 M sodium acetate pH 5.2 | |||
Qubit 2.0 Fluorometer | Invitrogen | Q32866 | |
Qubit dsDNA HS Assay Kit | Invitrogen | Q32851 | |
Qubit assay tubes | Invitrogen | Q32856 | |
Fluorescence microscope equipped with imaging software | |||
Microcentrifuge for 1.5 mL tubes | |||
Tabletop centrifuge for 15 and 50 mL tubes | |||
Cell culture incubator | |||
Biosafety cabinet | |||
Heat blocks | 65 °C and 95 °C |