Recombinant retroviral integrase and DNA oligomers mimicking viral DNA ends can form an enzymatically active complex known as an intasome. Intasomes may be used for biochemical, structural, and kinetic studies. This protocol details how to assemble and purify prototype foamy virus intasomes.
A defining feature and necessary step of the retrovirus life cycle is the integration of the viral genome into the host genome. All retroviruses encode an integrase (IN) enzyme that catalyzes the covalent joining of viral to host DNA, which is known as strand transfer. Integration may be modeled in vitro with recombinant retroviral IN and DNA oligomers mimicking the ends of the viral genome. In order to more closely recapitulate the integration reaction that occurs in vivo, integration complexes are assembled from recombinant IN and synthetic oligomers by dialysis in a reduced salt concentration buffer. The integration complex, called an intasome, may be purified by size exclusion chromatography. In the case of prototype foamy virus (PFV), the intasome is a tetramer of IN and two DNA oligomers and is readily separated from monomeric IN and free oligomer DNA. The integration efficiency of PFV intasomes may be assayed under a variety of experimental conditions to better understand the dynamics and mechanics of retroviral integration.
Integration of the viral genome into the host genome is a mandatory step in the life cycle of all retroviruses1. The viral enzyme integrase (IN) catalyzes the covalent joining of each end of the viral DNA genome to the host DNA. During a cellular infection, IN is part of a pre-integration complex that mediates integration. Recombinant IN complexed with double stranded DNA oligomers mimicking the viral DNA ends can also perform integration into a target DNA in vitro2. A common integration assay in vitro utilizes a supercoiled plasmid as the target DNA. Integration of both viral DNA oligomers (vDNA) to the plasmid results in a linear product and is termed concerted integration (Figure 2A). The integration assay in vitro may also yield products with only one vDNA covalently joined to the target plasmid resulting in a relaxed circle. This half-site integration product appears to be an artefact of the assay in vitro.
Recombinant IN and vDNA may perform integration in vitro, but they are not ideal reagents for the study of the dynamics or structure of integration complexes when monomeric IN would obscure relevant visualization. Purified integration complexes, or "intasomes," are required for dynamic single molecule analysis or structural studies. PFV IN and vDNAs may be assembled by dialysis from a relatively high salt concentration buffer to a lower salt concentration3,4. During dialysis, a precipitate forms. This precipitate is removed from dialysis and the salt concentration is increased. The higher salt concentration solubilizes the precipitate containing PFV intasomes. The intasomes are then purified by size exclusion chromatography (SEC). Recombinant prototype foamy virus (PFV) IN has been shown to exist as a monomer in solution at concentrations up to 225 µM5. SEC fractionation effectively separates the PFV intasomes (225.5 kDa), which includes a tetramer of PFV IN and two vDNAs, from monomeric PFV IN (44.4 kDa) and free vDNA (24.0 kDa). The PFV intasomes may be frozen and retain integration activity for at least six months of storage at -80 ˚C.
Recombinant PFV intasomes may also be modified to include IN amino acid substitutions or truncation mutations or vDNAs labeled with fluorophores or biotin4,6. The purified PFV intasomes readily perform integration into a supercoiled plasmid target DNA in vitro. Bulk biochemical integration assays with intasomes may test the effects of IN mutations, IN inhibitors, or other chemical additives. Biotinylated intasomes can be used to probe affinity with nucleic acids or proteins. PFV intasomes are functional at ambient temperature allowing for single molecule microscopy analysis by magnetic tweezers to measure the time between joining of the two vDNA ends or total internal reflection fluorescence to visualize the intasome search on target DNA6. In addition, PFV intasomes were the first to be structurally characterized significantly impacting the field of retroviral integration3.
1. Annealing of vDNA
2. Intasome Assembly
3. Intasome Solubilization
4. Intasome Purification
5. Integration Strand Transfer Assay, Fraction Selection, and Storage
PFV intasomes are assembled from recombinant IN and vDNA. After assembly, intasomes are purified by SEC (Figure 1). Integration activity of each fraction is assayed with a supercoiled DNA target and agarose gel electrophoresis (Figure 2). This gel is imaged with a fluorescent scanner set to detect EtBr (and fluorophore, if fluorophore-labeled vDNA is used). Using image analysis software, band pixel volumes can be used to calculate the integration efficiency (Figure 3). Fractions with the highest integration efficiency are snap frozen for later use. Frozen fractions retain integration activity (Figure 4), allowing for long term storage of assembled intasomes. Any label molecule and its position within the vDNA may affect intasome assembly efficiency (Figure 5).
Figure 1: Size exclusion chromatogram. A PFV intasome assembly was separated with an agarose SEC column. This column allows for the separation of the aggregate (1), PFV intasome consisting of a tetramer of PFV IN and two vDNA (2), and monomeric PFV IN and vDNA (3). This example included vDNA with an internal Cy5 fluorophore label detected by 650 nm excitation. The y-axis denotes optical density (OD) in milli-absorbance units (mAU). Please click here to view a larger version of this figure.
Figure 2: Integration assay of SEC fractions. (A) Schematic of integration strand transfer reaction. Left, cartoon of a PFV intasome including a tetramer of IN (blue circles) and two vDNA (thick black lines), and a target plasmid (thin black lines). The supercoils (SC) of the target plasmid are not drawn. Center, cartoon of a half-site integration (HSI) product when one vDNA has been covalently joined to the target plasmid. The joining reaction introduces a nick and relaxes the supercoils. Right, cartoon of a concerted integration (CI) product where two vDNAs have been joined to the target DNA. This integration product results in a linear DNA with vDNAs at the ends. (B) Supercoiled plasmid DNA was added to each SEC fraction #49-55. Following incubation, the integration reactions were deproteinated and separated by 1% agarose gel electrophoresis containing EtBr. Negative control was plasmid DNA with no protein (T). DNA size markers are shown in kb. Supercoiled plasmid (SC), concerted integration products (CI), half-site integration products (HSI), and vDNA are indicated. (C) The agarose gel was scanned for the presence of Cy5. Only the vDNA, CI, and HSI products are visible. Please click here to view a larger version of this figure.
Figure 3: Quantitation of integration assay. The agarose gel from Figure 2 was scanned and quantified for the presence of both (A) EtBr and (B) Cy5. (A) For the EtBr calculation, HSI, CI, and SC band pixel volumes were obtained using gel analysis software. Integration efficiency was calculated by dividing the pixel volume of concerted integration products (CI) by the total pixel volume of all three bands (HSI+CI+SC). For example, the pixel values in the EtBr channel for the bands of fraction #49 are: 110565.2 SC, 25152.56 CI, and 6313.04 HSI. The total DNA pixel value for fraction #49 is the sum of these values, 142030.8. The integration efficiency is the CI pixel volume of 25152.56 divided by the total DNA value 142030.8 equaling 0.18. (B) Cy5 CI band pixel volumes are graphed as arbitrary units. Fractions with peak integration activity are individually aliquoted and frozen. In this example, fractions #50-53 were selected. Aliquots are snap frozen with liquid nitrogen and stored at -80 ˚C for future use.
Figure 4: Effect of freezing on intasome activity. Half of one SEC fraction was flash frozen with liquid nitrogen, stored at -80 ˚C for 1 h, and then slowly thawed on ice. The remaining half of the SEC fraction was kept on ice in a cold room while the first half was frozen and thawed. Intasomes were tested for activity without (-) and with (+) freeze/thawing. Integration efficiency was measured as described above. (A) EtBr stained agarose gel of integration reaction products. Supercoiled plasmid (SC), concerted integration products (CI), half-site integration products (HSI), and unreacted vDNA are indicated. DNA size markers are shown in kb. (B) Quantitation of the integration efficiency from the EtBr image. Calculations are described in Figure 3 legend. There is no loss of integration activity following freezing and thawing. The average integration efficiency is shown for 5 independent experiments with 2 intasome preparations. Error bars indicate the standard deviation. Paired t-test analyses yielded a two-tailed P = 0.011, suggesting that the freeze/thaw may have slightly enhanced integration activity. Please click here to view a larger version of this figure.
Figure 5: Label molecule and position impacts intasome assembly. PFV intasome assemblies included vDNAs that were unlabeled (red), end labeled on Oligo 2 with Cy5 (blue), internally labeled on Oligo 1 with Cy5 (black), or end labeled on Oligo 2 with biotin (orange). All assemblies were separated with an agarose SEC column. The y-axis denotes OD in mAU. Chromatograms are representative of at least 2 independent assemblies of each intasome type. End labeling reduces the yield of PFV intasomes with Cy5 10-fold and biotin 1.8-fold. End Cy5 and biotin assemblies have noticeably more precipitate remaining after high salt resolubilization (step 3.5), resulting in apparent loss of material on the size exclusion column. Internal labeling of the vDNA with the Cy5 fluorophore leads to an equal yield of intasomes as unlabeled vDNA. Please click here to view a larger version of this figure.
Retroviral INs form a multimeric complex with viral genomic DNA to perform integration and continue the viral life cycle. The number of IN monomers per intasome may be tetramers, octamers, or possibly higher order multimers11,12,13,14. PFV intasomes are a tetramer of recombinant IN with double-stranded DNA oligomers that mimic viral genomic DNA ends3. These intasomes have been used in structural and dynamic studies3,5,13,14.
The assembly of PFV intasomes occurs during dialysis from a relatively high salt concentration buffer to a lower salt concentration. This results in the formation of a precipitate which includes the PFV intasomes. The intasomes are solubilized by increasing the salt concentration. PFV IN has been shown to be monomeric at concentrations up to 225 µM5. The complexes are then isolated from monomeric IN and free vDNA by SEC. We have modified the purification of PFV intasomes to include glycerol in the SEC buffer6. The addition of glycerol allows the PFV intasomes to be frozen for future use without loss of activity (Figure 4).
There are several key features of this protocol. Although protein purification is often performed at low temperatures, we have found empirically that PFV intasome assembly is inefficient at 4 ˚C. Instead, the assembly dialysis should occur in the range of 18-22 ˚C. It is also important to use recombinant PFV IN that is free of contaminating bacterial nuclease7,8. Finally, the ratio of IN to vDNA is a key factor of the assembly. If the starting IN protein concentration is lower than recommended here, the vDNA concentration must be proportionally decreased. While it is possible to assemble PFV intasomes at lower concentrations, the complexes must be at a high enough concentration for detection during SEC separation.
This method may be used for PFV intasomes with unlabeled or labeled vDNA. We have found that the vDNA may be efficiently labeled with a fluorophore or biotin6. Other small molecule labels may also be possible. In the case of Cy5 fluorophore, we find that end labeled vDNA reduces the yield of assembled intasomes with the peak concentration reduced 10-fold. However, internally Cy5 labeled vDNA has an equivalent yield to unlabeled vDNA. This assembly method may also be used with point mutations or truncation mutations of PFV IN4. Since PFV IN is inhibited by clinically relevant IN strand transfer inhibitor (INSTI) drugs targeting HIV-1 IN, it is possible to use these PFV intasomes to study the mechanism of INSTIs3. In addition, some INSTI resistant mutations of HIV-1 IN occur at conserved residues in PFV IN, allowing studies of drug resistance with engineered PFV intasomes.
The authors have nothing to disclose.
This work was supported by NIH AI099854 and AI126742 to KEY.
DNA Oligomers | IDT | N/A | Custom DNA Oligos |
Tris Ultra Pure | Gojira Fine Chemicals | UTS1003 | |
NaCl | P212121 | RP-S23020 | |
UltraPure EDTA | Invitrogen/Gibco | 15575 | |
Amicon Ultra 0.5 mL centrifugal filters | Sigma-Aldrich | Z677094-24EA | 3 kDa MWCO |
DTT | P212121 | SV-DTT | |
BIS-TRIS propane,>=99.0% (titration) | Sigma-Aldrich | B6755-500G | |
ZnCl2 | Sigma-Aldrich | 208086 | |
MgSO4 | Amresco | 0662 | |
Glycerol | Thermo Fisher Scientific | G37-20 | |
Gel-loading tips, 1 – 200 μL | Corning | CLS4853-400EA | |
Razor blade; Single-edged; 100/Pk.; Pack of 100 | Fisher Scientific | 12-640 | |
Sterile Disposable Filter Units with PES Membrane > 250mL | Thermo Fisher Scientific | 568-0020 | |
Dialysis tubing clips | Spectrum Labs | 132734 | |
6-8 kDa 10 mm Dialysis Tubing | Spectrum Medical | 132645 | |
Superose 6 10/300 GL | GE Healthcare Life Sciences | 17517201 | |
Hi-Res Standard Agarose | AGTC Bioproducts | AG500D1 | |
Ethidium bromide | Thermo Fisher Scientific | BP1302 | |
Orange G | Fisher Scientific | 0-267 | |
Hyladder 10kb, 500 lanes | Denville Scientific | CB4225-4 |