|Home | About | Journals | Submit | Contact Us | Français|
Integration of retroviral DNA into the host genome is an essential step in the viral replication cycle. The viral DNA, made by reverse transcription in the cytoplasm, forms part of a large nucleoprotein complex called the preintegration complex (PIC). The viral integrase protein is the enzyme within the PIC that is responsible for integrating the viral DNA into the host genome. Integrase is tightly associated with the viral DNA within the PIC as demonstrated by functional assays. Integrase protein catalyzes the key DNA cutting and joining steps of integration in vitro with DNA substrates that mimic the ends of the viral DNA. Under most in vitro assay conditions the stringency of the reaction is relaxed; most products result from “half-site” integration in which only one viral DNA end is integrated into one strand of target DNA rather than concerted integration of pairs of DNA as occurs with PICs and in vivo. Under these relaxed conditions catalysis appears to occur without formation of the highly stable nucleoprotein complexes that is characteristic of the association of integrase with viral DNA in the PIC. Here we describe methods for the assembly of nucleoprotein complex intermediates in HIV-1 DNA integration from purified HIV-1 integrase and substrates that mimic the viral DNA ends.
Integration of a DNA copy of the viral genome into chromosomal DNA of the host cell is an essential step in the replication cycle of HIV-1 and other retroviruses [1,2]. The chemical steps of DNA cutting and joining, which occur in two separate reactions, are catalyzed by the virally encoded integrase protein. In the first step of the integration process, two nucleotides are removed from each 3′-end of HIV-1 DNA in a reaction termed 3′-end processing. Cleavage occurs to the 3′-side of a CA dinucleotide that is conserved among many DNA transposons as well as retroviruses. This reaction exposes the terminal 3′-hydroxyl group that is to be joined to target DNA upon integration. In the second step, DNA strand transfer, the hydroxyl groups at the 3′-ends of the processed viral DNA attack a pair of phosphodiester bonds in the target DNA, to generate an integration intermediate in which the 3′-ends of the viral DNA are covalently joined to the 5′-ends of the target DNA at the site of integration. The integration intermediate then undergoes DNA repair to complete the integration process; the repair steps are likely to be carried out by cellular enzymes [3,4].
Both the 3′-end processing and DNA strand transfer activities of HIV-1 integrase can be recapitulated in vitro with short DNA substrates that mimic the viral DNA ends [5–7]. However, under those reaction conditions the strand transfer products mostly result from a “half-site” reaction in which only one viral DNA end is joined to one strand of target DNA, rather than concerted integration of a pair of viral DNA ends as occurs in vivo. Furthermore, highly stable complexes between integrase and viral DNA that resemble the association of integrase with the viral DNA in preintegration complexes (PICs) isolated from infected cells have not been observed with this reaction system.
Many enzymes that catalyze reactions at specific sites in DNA molecules have a high affinity for that DNA sequence. In some cases assembling nucleoprotein complexes of the enzyme and DNA is as simple as mixing the enzyme with DNA containing the appropriate recognition sequence. However, assembly of functional complexes of HIV-1 integrase with viral DNA substrate is a more complex process. The bacteriophage Mu transposase is an extensively studied model system that is useful in considering some of key features of complex assembly with HIV-1 integrase. Mu transposase binds specifically to three binding sites at each end of the Mu genome. Although specific, binding is quite weak and the transposase can be dissociated with 200 mM NaCl. However, in the presence of a divalent metal ion, and cofactors that we will not discuss here, a very stable complex called a transpososome (reviewed in ) is assembled. The heart of the transpososome is a tetramer of transposase that stably bridges a pair of Mu DNA ends. Once assembled, the transpososome is stable to challenge with high ionic strength, detergents, and heat. The assembly process is a downhill energetic pathway, but an energy barrier must first be climbed before the transpososome can be assembled. Assembly of fully functional complexes between HIV-1 integrase and viral DNA is a similar process. The primary binding of HIV-1 integrase to viral DNA ends is weak and exhibits little sequence specificity, but a pair viral DNA ends subsequently “lock” in with a tetramer of integrase to form a highly stable complex that is an intermediate on the integration pathway.
Improved in vitro reaction systems [9–12] have enabled concerted DNA integration to be studied in vitro. Under conditions that promote concerted DNA integration, highly stable complexes are formed between integrase and DNA substrate [13,14]. The first stable complex is the stable synaptic complex (SSC) in which integrase is stably associated with a pair of viral DNA ends. The SSC is a transient intermediate that integrates the pair of viral DNA ends into a target DNA. The integration product remains stably bound by integrase in a second stable complex, the stand transfer complex (STC) (Fig. 1). We describe protocols for generating these stable complexes between integrase and DNA substrate that appear to closely mimic the complexes formed in vivo as judged by functional assays.
Efficient assembly of stable complexes between HIV-1 integrase and viral DNA substrate occurs under reaction conditions that promote concerted DNA integration [11,13] and requires viral DNA substrate longer than several hundred base pairs. Sequence specificity does not extend beyond the terminal 20 bp and there are no sequence-specific requirements for the flanking DNA. We constructed a plasmid pSca355  (Fig. 2A) that when digested with ScaI and HincII generates a 1.5 kb linear fragment terminating with 32 bp of the blunt-ended U5 terminal DNA sequence (Fig. 2B). Additional restriction sites within this fragment allow substrates of different lengths to be made. Our typical reactions use a 1 kb fragment made by further cleavage with BanI (Fig. 2B). For some purposes it is necessary to manipulate the usually blunt terminal viral DNA sequence, for example to make a pre-processed viral DNA substrate with a 3′-dideoxyadenosine to trap the SSC. Such sequence modifications are conveniently made in oligonucleotides that are subsequently ligated to a longer linear DNA fragment to make the final DNA substrate (Fig. 2C). Although a U5 end normally pairs with a U3 end in vivo, a pair of U5 ends can support near wild type levels of HIV-1 infectivity  and are equally efficient at promoting concerted DNA integration and SSC and STC formation in vitro (see supplementary material in ). Note that viral DNA sequence is present at only one end of the substrate and pairing occurs between two separate DNA molecules.
Because the linear ~1 kb viral DNA substrate must be excised from plasmid pSca355 and purified by gel electrophoresis, it is advisable to start with a 500 μg or larger scale of plasmid preparation. Many standard protocols and commercial kits are available for this purpose. We use the QIAfilter Plasmid Maxi Kit (QIAGEN). The blunt-end viral DNA substrate is prepared from pSca355 as follows:
Ligation of oligonucleotides to non-specific flanking DNA is a convenient method to prepare viral DNA substrates with chemically modified ends because oligonucleotides are readily amenable to modification both during synthesis and post-synthesis.
The following purification method is essentially as previously described for purification of wild type His-tagged HIV-1 integrase . Briefly, integrase is expressed in Escherichia coli BL21 (DE3) and the cells are lysed in buffer containing 0.1 M NaCl. The lysate is centrifuged and integrase is extracted from the pellet in buffer containing 1 M NaCl. The protein is then purified by nickel-affinity chromatography. Following removal of the His-tag by thrombin digestion, the protein is further purified by gel filtration on a Superdex-75 column (GE Healthcare).
Efficient complex assembly occurs under reaction conditions that favor concerted integration . The SSC is a transient intermediate and is chased to the SSC unless the DNA strand transfer step is blocked (Fig. 3).
Reaction conditions and electrophoretic analysis products are essentially as described . Thaw frozen integrase and DNA stocks on ice and keep all freshly made buffer stocks, except dimethyl sulfoxide (DMSO), on ice.
The final concentrations of individual components are: 20 mM Hepes pH 7.5, 12% DMSO, 10 mM DTT, 10% PEG-6000, 10 mM MgCl2, 20μM ZnCl2, and 100 mM NaCl in a volume of 24.5 μl.
The SSC is a transient intermediate on the integration reaction pathway. The stable complexes formed in the above protocol are mainly STCs, which are product complexes containing a pair of viral DNA ends and target DNA stably associated with integrase (Figs. 1 and and3A).3A). Several strategies are available to trap and accumulate the intermediate SSC complex by preventing it from forming the STC and completing integration.
A major viral DNA substrate requirement for assembly of the SSC and STC, and for concerted integration of pairs of DNA ends, is that it be longer than about 300 bp; as the length is further reduced the efficiency of product formation drops off precipitously . The viral DNA sequence requirements do not extend beyond 10–20 bp and there appears to be no sequence specificity for the flanking DNA. The mechanistic basis of this requirement for “long” DNA substrate is unknown and is a question we are attempting to address. Recent results demonstrate that increasing the concentration of “short” viral substrates to that of integrase in the reaction significantly affects the efficiency of concerted foamy viral DNA integration in vitro .
SSCs can be efficiently assembled with pre-cleaved viral DNA ends and these complexes go on to form STCs. Therefore processing of the viral DNA ends by integrase does not appear to play a role in SSC assembly. However, in vitro, integrase can “by-pass” normal complex assembly and catalyze one-end integration without the formation of stable complexes. This more promiscuous pathway tends to be favored with pre-cleaved substrates . SSCs and STCs can be assembled with wild type, F185K, F185H, or W235F integrase. The efficiency differs by only a few-fold among these proteins but is generally higher with W235F and wild type integrase than with the F185K and F185H mutants.
Integrase stably pairs the viral DNA ends within the PIC and it is this complex that is the target of inhibitors such as Raltegravir, rather than free integrase protein. Ultimately high-resolution structural information on the SSC will be required to understand the detailed mechanism of action of this and other classes of integrase inhibitors and mechanisms of drug resistance. Unfortunately, the low abundance of PICs in infected cell extracts does not permit even low-resolution biophysical studies of complexes of integrase and viral DNA isolated from this source. In vitro assembly of nucleoprotein complexes of integrase and viral DNA that mimics the association of integrase and viral DNA in the PIC can be expected to facilitate future biophysical and structural studies of HIV-1 integrase and its interaction with inhibitors.
This work was supported by the Intramural Research Program of the National Institute of Diabetes and Digestive and Kidney Diseases and by the NIH AIDS Targeted Antiviral Program.