Insight into the Integrase-DNA Recognition Mechanism
2008; Elsevier BV; Volume: 283; Issue: 41 Linguagem: Inglês
10.1074/jbc.m803257200
ISSN1083-351X
AutoresOlivier Delelis, Kévin Carayon, Elvire Guiot, Hervé Leh, Patrick Tauc, Jean‐Claude Brochon, Jean‐François Mouscadet, Eric Deprez,
Tópico(s)Glycosylation and Glycoproteins Research
ResumoIntegration catalyzed by integrase (IN) is a key process in the retrovirus life cycle. Many biochemical or structural human immunodeficiency virus, type 1 (HIV-1) IN studies have been severely impeded by its propensity to aggregate. We characterized a retroviral IN (primate foamy virus (PFV-1)) that displays a solubility profile different from that of HIV-1 IN. Using various techniques, including fluorescence correlation spectroscopy, time-resolved fluorescence anisotropy, and size exclusion chromatography, we identified a monomer-dimer equilibrium for the protein alone, with a half-transition concentration of 20-30 μm. We performed specific enzymatic labeling of PFV-1 IN and measured the fluorescence resonance energy transfer between carboxytetramethylrhodamine-labeled IN and fluorescein-labeled DNA substrates. FRET and fluorescence anisotropy highlight the preferential binding of PFV-1 IN to the 3′-end processing site. Sequence-specific DNA binding was not observed with HIV-1 IN, suggesting that the intrinsic ability of retroviral INs to bind preferentially to the processing site is highly underestimated in the presence of aggregates. IN is in a dimeric state for 3′-processing on short DNA substrates, whereas IN polymerization, mediated by nonspecific contacts at internal DNA positions, occurs on longer DNAs. Additionally, aggregation, mediated by nonspecific IN-IN interactions, occurs preferentially with short DNAs at high IN/DNA ratios. The presence of either higher order complex is detrimental for specific activity. Ionic strength favors catalytically competent over higher order complexes by selectively disrupting nonspecific IN-IN interactions. This counteracting effect was not observed with polymerization. The synergic effect on the selection of specific/competent complexes, obtained by using short DNA substrates under high salt conditions, may have important implications for further structural studies in IN·DNA complexes. Integration catalyzed by integrase (IN) is a key process in the retrovirus life cycle. Many biochemical or structural human immunodeficiency virus, type 1 (HIV-1) IN studies have been severely impeded by its propensity to aggregate. We characterized a retroviral IN (primate foamy virus (PFV-1)) that displays a solubility profile different from that of HIV-1 IN. Using various techniques, including fluorescence correlation spectroscopy, time-resolved fluorescence anisotropy, and size exclusion chromatography, we identified a monomer-dimer equilibrium for the protein alone, with a half-transition concentration of 20-30 μm. We performed specific enzymatic labeling of PFV-1 IN and measured the fluorescence resonance energy transfer between carboxytetramethylrhodamine-labeled IN and fluorescein-labeled DNA substrates. FRET and fluorescence anisotropy highlight the preferential binding of PFV-1 IN to the 3′-end processing site. Sequence-specific DNA binding was not observed with HIV-1 IN, suggesting that the intrinsic ability of retroviral INs to bind preferentially to the processing site is highly underestimated in the presence of aggregates. IN is in a dimeric state for 3′-processing on short DNA substrates, whereas IN polymerization, mediated by nonspecific contacts at internal DNA positions, occurs on longer DNAs. Additionally, aggregation, mediated by nonspecific IN-IN interactions, occurs preferentially with short DNAs at high IN/DNA ratios. The presence of either higher order complex is detrimental for specific activity. Ionic strength favors catalytically competent over higher order complexes by selectively disrupting nonspecific IN-IN interactions. This counteracting effect was not observed with polymerization. The synergic effect on the selection of specific/competent complexes, obtained by using short DNA substrates under high salt conditions, may have important implications for further structural studies in IN·DNA complexes. Retroviral integrases (INs) 3The abbreviations used are: IN, integrase; INT; TAMRA-labeled integrase; FCS, fluorescence correlation spectroscopy; Fl, fluorescein; FRET, fluorescence resonance energy transfer; HIV-1, human immunodeficiency virus, type 1; LTR, long terminal repeat; ODN, oligodeoxynucleotide; PFV-1, primate foamy virus type-1; TFA, time-resolved fluorescence anisotropy; TGase, transglutaminase; 3′-P, 3′-processing; Mo, monomeric; Di, dimeric; Te, tetrameric; TAMRA, carboxytetramethylrhodamine. 3The abbreviations used are: IN, integrase; INT; TAMRA-labeled integrase; FCS, fluorescence correlation spectroscopy; Fl, fluorescein; FRET, fluorescence resonance energy transfer; HIV-1, human immunodeficiency virus, type 1; LTR, long terminal repeat; ODN, oligodeoxynucleotide; PFV-1, primate foamy virus type-1; TFA, time-resolved fluorescence anisotropy; TGase, transglutaminase; 3′-P, 3′-processing; Mo, monomeric; Di, dimeric; Te, tetrameric; TAMRA, carboxytetramethylrhodamine. catalyze the integration of viral DNA into the host genome ensuring its perpetuation in the host cell. The integration process requires two catalytic steps. During the first step, titled 3′-processing (3′-P), IN specifically removes two nucleotides from each viral DNA end. IN then transfers both ends into target DNA by a one-step transesterification reaction, resulting in full-site integration. IN alone is competent for the insertion process (1Sinha S. Grandgenett D.P. J. Virol. 2005; 79: 8208-8216Crossref PubMed Scopus (71) Google Scholar). Retroviral INs consist of three functional domains. The core domain contains the DDE catalytic triad and is flanked by the N-terminal domain (involved in multimerization) and the C-terminal DNA-binding domain. The integrity of the DDE triad and a metallic cofactor are strictly required for enzymatic activity. Biochemical and structural studies of HIV-1 IN have been severely impeded because of solubility. No structural data are available for the full-length protein (free or bound to DNA) to date, although x-ray structures of DNA-free single or double domains have been successfully solved (2Chen J.C. Krucinski J. Miercke L.J. Finer-Moore J.S. Tang A.H. Leavitt A.D. Stroud R.M. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 8233-8238Crossref PubMed Scopus (379) Google Scholar, 3Wang J.Y. Ling H. Yang W. Craigie R. EMBO J. 2001; 20: 7333-7343Crossref PubMed Scopus (311) Google Scholar, 4Goldgur Y. Dyda F. Hickman A.B. Jenkins T.M. Craigie R. Davies D.R. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 9150-9154Crossref PubMed Scopus (364) Google Scholar, 5Cherepanov P. Ambrosio A.L. Rahman S. Ellenberger T. Engelman A. Proc. Natl. Acad. Sci. U. S. A. 2005; 102: 17308-17313Crossref PubMed Scopus (345) Google Scholar). Hence, the sequence specificity of the 3′-P reaction remains poorly understood and little is known about the mechanism of IN/DNA substrate recognition. Point mutations or the use of detergent may improve the solubility but only to a limited extent (6Jenkins T.M. Engelman A. Ghirlando R. Craigie R. J. Biol. Chem. 1996; 271: 7712-7718Abstract Full Text Full Text PDF PubMed Scopus (265) Google Scholar, 7Deprez E. Tauc P. Leh H. Mouscadet J.F. Auclair C. Brochon J.C. Biochemistry. 2000; 39: 9275-9284Crossref PubMed Scopus (128) Google Scholar) and may also cause changes to IN properties. For instance, detergent is detrimental to 3′-P activity in the presence of the physiological cofactor Mg2+ (8Leh H. Brodin P. Bischerour J. Deprez E. Tauc P. Brochon J.C. LeCam E. Coulaud D. Auclair C. Mouscadet J.F. Biochemistry. 2000; 39: 9285-9294Crossref PubMed Scopus (120) Google Scholar), and one soluble mutant was found to be resistant to diketo-acid compounds (9Marchand C. Johnson A.A. Karki R.G. Pais G.C. Zhang X. Cowansage K. Patel T.A. Nicklaus M.C. Burke Jr., T.R. Pommier Y. Mol. Pharmacol. 2003; 64: 600-609Crossref PubMed Scopus (133) Google Scholar). Endonucleolytic cleavage depends primarily on the presence of the canonical CA sequence preceding the removed dinucleotide. Other positions are crucial for 3′-P activity under Mg2+ conditions, despite the absence, in vitro, of IN preference/specificity for the cognate U5 or U3 terminal sequence at the DNA binding level (10Agapkina J. Smolov M. Barbe S. Zubin E. Zatsepin T. Deprez E. Le Bret M. Mouscadet J.F. Gottikh M. J. Biol. Chem. 2006; 281: 11530-11540Abstract Full Text Full Text PDF PubMed Scopus (37) Google Scholar, 12Engelman A. Hickman A.B. Craigie R. J. Virol. 1994; 68: 5911-5917Crossref PubMed Google Scholar). Thus, it is generally hypothesized that IN specificity is fully explained at the catalytic level. However, due to poor solubility, it is a difficult task to determine which enzymatic properties of HIV-1 IN actually correspond to intrinsic properties of retroviral INs and which properties are related to aggregation. During the course of our study on HIV-1 IN, we have observed significant differences relating to solubility between HIV-1 IN and another retroviral IN, primate foamy virus-1 (PFV-1) IN. PFV-1 is the prototype of foamy viruses belonging to the retrovirus family and differ from lentiviruses in some aspects. For instance, PFV-1 reverse transcription occurs at late stages of the replication cycle (13Moebes A. Enssle J. Bieniasz P.D. Heinkelein M. Lindemann D. Bock M. McClure M.O. Rethwilm A. J. Virol. 1997; 71: 7305-7311Crossref PubMed Google Scholar, 14Yu S.F. Sullivan M.D. Linial M.L. J. Virol. 1999; 73: 1565-1572Crossref PubMed Google Scholar). Interaction with host cellular cofactors is also distinct, since LEDGF/p75 only interacts with lentiviral INs (15Cherepanov P. Nucleic Acids Res. 2007; 35: 113-124Crossref PubMed Scopus (142) Google Scholar). Furthermore, only the U5 end is processed due to the asymmetric nature of the integration process in the case of PFV-1 (16Enssle J. Moebes A. Heinkelein M. Panhuysen M. Mauer B. Schweizer M. Neumann-Haefelin D. Rethwilm A. J. Gen. Virol. 1999; 80: 1445-1452Crossref PubMed Scopus (40) Google Scholar). However, integration is an obligatory step for replicating PFV-1, a hallmark of retrovirus life cycle (16Enssle J. Moebes A. Heinkelein M. Panhuysen M. Mauer B. Schweizer M. Neumann-Haefelin D. Rethwilm A. J. Gen. Virol. 1999; 80: 1445-1452Crossref PubMed Scopus (40) Google Scholar). PFV-1 IN shares common structural features with other retroviral INs, such as three-domain organization as well as the three major catalytic activities, 3′-P, strand transfer, and disintegration (17Pahl A. Flugel R.M. J. Biol. Chem. 1995; 270: 2957-2966Abstract Full Text Full Text PDF PubMed Scopus (40) Google Scholar, 19Oh Y.T. Shin C.G. Biochem. Mol. Biol. Int. 1999; 47: 621-629PubMed Google Scholar). To get deeper insight into the IN/DNA recognition mechanism, we studied and compared solubility, oligomeric status and catalytic efficiency of PFV-1 and HIV-1 IN. We used various methods, such as time-resolved fluorescence anisotropy (TFA), fluorescence correlation spectroscopy (FCS), and size exclusion chromatography, and found that free PFV-1 IN was significantly more soluble than HIV-1 IN. This enabled us to successfully apply transglutaminase (TGase)-mediated TAMRA labeling to a soluble IN at a single predefined position. Fluorescence resonance energy transfer (FRET) experiments and a DNA-binding anisotropy-based assay were conducted to assess the specificity of IN-DNA interactions. A specific mode of interaction was clearly shown for PFV-1 IN but not for HIV-1 IN. DNA promotes the catalytically competent dimeric state. The dimerization process is in competition with other processes (which were found to be detrimental to activity), such as IN polymerization and aggregation, mediated by nonspecific IN-DNA and IN-IN interactions, respectively. A selection of specific/competent complexes may be achieved using short DNA substrates under conditions of high ionic strength. Purification and IN Labeling by Guinea Pig TGase—For DNA-binding and 3′-P assays, unlabeled wild-type HIV-1 and PFV-1 IN were purified as previously described (8Leh H. Brodin P. Bischerour J. Deprez E. Tauc P. Brochon J.C. LeCam E. Coulaud D. Auclair C. Mouscadet J.F. Biochemistry. 2000; 39: 9285-9294Crossref PubMed Scopus (120) Google Scholar). The plasmid encoding PFV-1 IN with the specific C-terminal PKPQQFM tag for TGase-mediated labeling was constructed by PCR amplification of pET15b-PFV-IN (20Delelis O. Petit C. Leh H. Mbemba G. Mouscadet J.F. Sonigo P. Retrovirology. 2005; 2: 31-49Crossref PubMed Scopus (21) Google Scholar) with P1 and P2 primers (P1, 5′-ACA TAT GTG TAA TAC CAA AAA ACC AAA CCT GG-3′; P2, 5′-AGG ATC CTA CAT AAA CTG CTG AGG TTT TGG CTC GAG TTC ATT TTT TTC-3′). The corresponding plasmid for tagged HIV-1 IN was constructed by amplification of pET15b-HIV-IN (8Leh H. Brodin P. Bischerour J. Deprez E. Tauc P. Brochon J.C. LeCam E. Coulaud D. Auclair C. Mouscadet J.F. Biochemistry. 2000; 39: 9285-9294Crossref PubMed Scopus (120) Google Scholar) with P3 and P4 primers (P3, 5′-CCA TAT GTT TTT AGA TGG AAT AGA TAA-3′; P4, 5′-CGG ATC CTA CAT AAA CTG CTG AGG TTT TGG GTC CTC ATC CTG TCT ACT-3′). The underlined nucleotides correspond to the TGase tag (PKPQQFM) located at the C terminus of IN. The resulting DNAs encoding IN-PKPQQFM were inserted into pET15b. For labeling, IN attached to Ni2+-beads was incubated overnight at 4 °C with 0.5 mm TAMRA cadaverine (Molecular Probes) and 1 unit of guinea pig TGase (Sigma) in the labeling buffer (20 mm Hepes, pH 7.5, 10 mm CaCl2). Beads were extensively washed with Tris buffer (50 mm, pH 8, 1 m NaCl) to remove free fluorophores. Elution was done with 1 m imidazole. TAMRA-labeled or unlabeled PFV-1 IN was further purified by cation exchange chromatography on a Mono S column (Bio-Rad) previously equilibrated with 20 mm Hepes, pH 7.5, 0.1 m NaCl. IN was diluted before loading (1 ml/min) and a linear salt gradient (0.1-1 m NaCl) was applied at 1 ml/min. IN was eluted at 450-550 mm NaCl (cation exchange chromatography was unsuccessful with HIV-1 IN for solubility reasons). It was possible to further concentrate PFV-1 IN using Amicon Ultra centrifugal filters (Millipore). Concentrations up to 250 μm in 20 mm Hepes, pH 7.5, 75 mm NaCl were obtained. Ionic strength was adjusted by buffer exchange during the concentration step. Dialysis was not used, since protein aggregation was reproducibly observed during the dialysis step of concentrated samples (significant precipitation occurred at PFV-1 IN concentrations of 12 mg/ml, whichever protocol was used). Size Exclusion Chromatography—100 μl of varying concentrations of purified PFV-1 IN (from 2 to 60 μm) were applied at room temperature onto a Superdex 200 preparation grade column (Hiload 16/60; Amersham Biosciences), using an AKTA FPLC system (Amersham Biosciences) previously equilibrated with the protein buffer (20 mm Tris-HCl, pH 8, 0.05 or 0.5 m NaCl, 10% glycerol, 50 μm ZnSO4, and 4 mm β-mercaptoethanol). The column was calibrated with the following globular proteins: sweet potato β-amylase (200 kDa), yeast alcohol dehydrogenase (150 kDa), bovine serum albumin (66 kDa), ovalbumin (45 kDa), carbonic anhydrase (29 kDa), and horse heart cytochrome c (12.5 kDa) (MW-GF-200 kit; Sigma). Fractions were collected at a flow rate of 0.3 ml/min, under a pressure of 0.5 megapascals. Absorbance was measured at 280 nm. The exclusion volume of the column V0 was measured by calibration with blue dextran (2000 kDa). Characterization of TAMRA-labeled IN (INT)—We evaluated the number of TAMRAs per labeled protein by mass spectrometry analysis; INT and the unlabeled tagged IN were subjected to SDS-PAGE. After gel staining, the gel slides containing the corresponding proteins were destained in 25 mm NH4CO3, 50% acetonitrile. A proteolysis using sequence grade trypsin (Promega) was then performed overnight at 37 °C. The resulting peptides were recovered by the addition of 50% acetonitrile, 5% trifluoroacetic acid and dried in a SpeedVac. They were reconstituted in 5% acetonitrile, 0.1% trifluoroacetic acid and applied on a SEND-ID protein array according to Ciphergen's protocol. The mass spectra were obtained by reading the protein chip using a SELDI mass spectrometer. The resulting spectra were compared with the in silico profile obtained with GPMAW software (Lighthouse Data). The labeling yield of purified INT was estimated by both absorbance spectroscopy and FCS. Given the molar extinction coefficient of TAMRA-cadaverine at 556 nm (62,000 m-1·cm-1) and the results from the Bradford protein assay, the labeling yield was estimated to be 45% by absorbance spectroscopy. The FCS autocorrelation curve (Fig. 1E) was also used to estimate the labeling yield, since it is possible to deduce from the inverse autocorrelation amplitude the mean number of labeled molecules in the detection volume. Here, g(0)-1 was equal to 0.025 and the detection volume was 1.27 μm3 (the lateral ω0 and axial z0 dimensions are equal to 0.45 and 1.5 μm, respectively) giving an estimated concentration of INT (52 nm). Considering the concentration of total IN (100 nm), the deduced labeling yield (∼50%) is in good agreement with that determined by absorbance spectroscopy. This calculation is evidently correct only if a single fluorophore is attached to the protein (SELDI analysis of INT confirmed that a single TAMRA moiety was covalently and specifically attached to the protein C-terminal end). INT, as obtained after the labeling procedure, was also tested for 3′-P at the U5 end (21-mer oligonucleotide) using a standard assay (8Leh H. Brodin P. Bischerour J. Deprez E. Tauc P. Brochon J.C. LeCam E. Coulaud D. Auclair C. Mouscadet J.F. Biochemistry. 2000; 39: 9285-9294Crossref PubMed Scopus (120) Google Scholar) and was compared with both the untagged and the unlabeled tagged proteins. 3′-P and half-transfer products were about 23-27% and 3.5-4.5%, respectively, for a 3-h incubation (data not shown), and no significant difference was observed between the three samples, showing that neither the PKPQQFM tag nor the attached fluorophore influences the catalytic activity. PFV-1 ODNs for DNA Binding/3′-P Anisotropy-based Assays and FRET Experiments—All ODNs (≤65-mer) were purchased from Eurogentec (Liège, Belgium) and further purified by gel electrophoresis. ODNs of various lengths (15-65-mer) were tested for 3′-P activity, all mimicking the specific (cognate) terminal U5 sequence: PFV-CAAT (sequence of the 65-mer, 5′-GAA CTA CAC TTA TCT TAA ATG ATG TAA CTC CTT AGG ATA ATC AAT ATA CAA AAT TCC ATG ACA AT-3′. The 3′-dinucleotide cleaved by IN is underlined). The processed strand and its complementary nonprocessed strand were denoted A and B, respectively. Fluorescein (Fl) was attached at the 5′- or 3′-end of strand A for DNA-binding experiments. Fl was attached at the 3′-end of strand A for the activity assay. The PFV-GTAT sequence was identical, except that the four 3′-term bases of strand A were GTAT instead of the canonical sequence CAAT. The specific HIV-1 ODN (U5 sequence) is provided in Ref. 21Guiot E. Carayon K. Delelis O. Simon F. Tauc P. Zubin E. Gottikh M. Mouscadet J.F. Brochon J.C. Deprez E. J. Biol. Chem. 2006; 281: 22707-22719Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar. The three nonspecific (random) sequences used for competition experiments (Fig. 4) were as follows: 5′-ACA TAA TCT AAA ATA ATT GCC-3′ (21mer), 5′-ACC TAT GCG CCG CTA GAT TCC-3′ (21-mer), and 5′-TCA AGC TAG AAG ATT ATC TCA AGT ACA TAA TCT AAA ATA ATT GCC-3′ (45-mer). Annealing of oligonucleotides was obtained by mixing equimolar amounts of complementary strands in 20 mm Hepes, pH 7.5, 100 mm NaCl, heating to 85 °C for 5 min, and slow cooling to 25 °C. For FRET experiments, ODNs of various lengths (21-300-mer) were studied, all mimicking the PFV-1 U5 end. The double-stranded ODN was Fl-labeled at the 5′-end (A or B strand) (Fig. 3A). Fl-labeled 100- and 300-mer double-stranded ODNs were obtained by PCR. All PCRs were performed using the pHSRV13 plasmid encoding the proviral PFV-1 genome. The 100-mer BFl5 and 300-mer BFl5 were obtained using, as primers, the single-stranded 21-mer BFl5 ODN and a second ODN hybridizing to an internal sequence located, respectively, 100 or 300 bp from the U5 extremity. The 100-mer AFl5 and 300-mer AFl5 were obtained using, as primers, the nonfluorescent single-stranded 21-mer B ODN and a 5′-Fl-labeled ODN hybridizing to an internal sequence located, respectively, 100 or 300 bp from the U5 extremity. After amplification, PCR products were purified on agarose gel using the Qiagen gel extraction kit. Steady-state Fluorescence Anisotropy-based Assay—Anisotropy-based assay is a quantitative method to determine both DNA binding and catalytic parameters of IN using the same sample (21Guiot E. Carayon K. Delelis O. Simon F. Tauc P. Zubin E. Gottikh M. Mouscadet J.F. Brochon J.C. Deprez E. J. Biol. Chem. 2006; 281: 22707-22719Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar). Briefly, IN binding to Fl-labeled DNA increases the steady-state anisotropy value (r) (measured on a Beacon instrument (Panvera, Madison, WI)), allowing the calculation of the fraction of DNA sites bound to IN. The DNA-binding step can be recorded at 25 °C using Fl-labeled ODNs at the 5′- or 3′-end of either the A or B strand. By shifting the temperature to 37 °C, the activity-dependent decrease in the r value allows quantification of the 3′-P reaction. This occurs only if Fl is initially linked to the AT dinucleotide (the released PFV-1 dinucleotide). The fraction of released dinucleotides (Fdinu) can be calculated using the relative decrease of r compared with the initial value obtained at the end of the DNA-binding step (rt = 0) (real time assay) (21Guiot E. Carayon K. Delelis O. Simon F. Tauc P. Zubin E. Gottikh M. Mouscadet J.F. Brochon J.C. Deprez E. J. Biol. Chem. 2006; 281: 22707-22719Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar). No decrease in the r value was observed with the PFV-GTAT control sequence, in contrast to the PFV-CAAT canonical sequence (data not shown). The activity can be also calculated in fixed time experiments (3′-P is stopped by SDS), Fdinu = (rNP - r)/(rNP - rdinu), where rNP and rdinu are the anisotropies for pure solutions of nonprocessed double-stranded ODN and dinucleotide, respectively (rdinu was measured using the 5′-AT-3′-Fl dinucleotide). The formation of IN·DNA complexes and the subsequent 3′-P catalytic reaction were performed by incubating Fl-labeled ODNs with unlabeled IN in buffer A (20 mm Hepes, pH 7.5, 1 mm dithiothreitol, 10 mm MgCl2). IN, DNA, and NaCl concentrations are explicitly mentioned in the figure legends. FRET Experiments—FRET between Fl-labeled DNA (donor) and TAMRA-labeled IN (acceptor) was monitored with an Eclipse (Varian) spectrofluorimeter. DNA concentration was constant (20 nm) in buffer A + 50 mm NaCl. INT was progressively added, and the decrease in the steady-state emission intensity of the donor was recorded at 25 °C. The excitation wavelength was 500 nm. The donor quenching (qD) was measured at 520 nm. qD was estimated using the equation, qD = 1 - (ID,[INT]/ID,0), where ID,0 and ID,[INT] represent the donor emission intensity at 520 nm in the absence (DNA alone) and in presence of acceptor (along the titration), respectively. Excitation and emission slit widths were 10 and 5 nm, respectively. For each DNA/INT mixture, a corresponding control experiment was done using unlabeled IN to check the signal stability. Fluorescence Correlation Spectroscopy—FCS measurements were performed under two-photon excitation (840 nm) on a home-built system using an 80-MHz mode-locked Tsunami Ti:Sapphire laser (pulse 100 fs) pumped by a Millenia solid state laser (Spectra Physics) and a Nikon TE2000 inverted microscope. Before entering through the epifluorescence port of the microscope, the laser beam was expanded with a two-lens afocal system to fill the back aperture of the objective (Nikon, Plan Apo, ×100, numerical aperture 1.4, oil immersion). The setup was optimized to obtain a diffraction-limited focal spot. Measurements were typically carried out in a 50-μl solution dropped on the coverslip (buffer A + 50 mm NaCl). Fluorescence was collected by the same objective, separated from the excitation by a dichroic mirror (Chroma 700DCSPXR) and focused onto an avalanche photodiode (SPCM-AQR-14; PerkinElmer Life Sciences). An additional filter was used to reject the residual excitation light (Chroma 2P-Emitter E700SP). The detector was connected to a digital correlator (ALV 6000) that calculates the normalized correlation function g(τ) of fluorescence fluctuations according to Equation 1, where I(t) is the number of detected fluorescence photons per time unit. Assuming a three-dimensional Gaussian distribution of the excitation intensity, the FCS function for a free Brownian diffusion process is given by Equation 2, where N is the mean number of fluorophores in the excitation volume, and τD is the translational diffusion time. ω0 and z0 are the lateral and axial dimensions of the excitation volume, respectively. The mean translational diffusion time was determined by fitting autocorrelation curves using a Levenberg-Marquardt nonlinear least-squares fitting algorithm according to the analytical model (Equation 2). For two-photon excitation, the diffusion coefficient D is then calculated according to the equation, D = ω2/8τD. The calibration of the excitation volume was performed using a 10 nm water solution of TAMRA (D = 2.8 × 10-10 m2 s-1). The lateral ω0 and axial z0 dimensions of the excitation volume were 0.45 and 1.5 μm, respectively. The excitation power was adjusted using a neutral density filter. For our setup, we determined that an excitation power of 10 milliwatts was suitable for two-photon excitation of both TAMRA alone and INT (fluorescence intensity exhibits a quadratic dependence, and τD value is constant (absence of photobleaching) as a function of the incident power below 16 milliwatts). Recording times were typically 5 min (average of 10 cycles of 30 s). Time-resolved Fluorescence Anisotropy—Fluorescence of Trp or Fl was used for hydrodynamic studies of free or DNA-bound PFV-1 IN, respectively (in buffer A + 50 mm NaCl). IN·DNA complexes were obtained using unlabeled PFV-1 IN and Fl-labeled 15- and 21-mer U5 PFV-1 ODN (Fl at the 5′-end of strand A). Correlation times (θ) were calculated from the two polarized fluorescence decays I⊥(t) and I∥(t), using time-correlated single photon counting (21Guiot E. Carayon K. Delelis O. Simon F. Tauc P. Zubin E. Gottikh M. Mouscadet J.F. Brochon J.C. Deprez E. J. Biol. Chem. 2006; 281: 22707-22719Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar, 22Laboulais C. Deprez E. Leh H. Mouscadet J.F. Brochon J.C. Le Bret M. Biophys. J. 2001; 81: 473-489Abstract Full Text Full Text PDF PubMed Scopus (26) Google Scholar). Briefly, the time scaling was 19.5 ps/channel, and 4096 channels were used. The excitation light pulse source was a Ti:Sapphire laser (Millennia-pumped Tsunami femtosecond laser; Spectra Physics) (repetition rate, 8 MHz) associated with a second or third harmonic generator tuned to 490 or 296 nm for Fl or Trp, respectively. The emission monochromator (ARC SpectraPro-150) was set to 530 or 340 nm (Δλ = 15 nm) for Fl or Trp, respectively. The two polarized components were collected alternately over a period of 30 s (total count of I∥(t) was 15 millions, which is a condition compatible with the recovery of correlation times up to ∼80 ns using Trp fluorescence (7Deprez E. Tauc P. Leh H. Mouscadet J.F. Auclair C. Brochon J.C. Biochemistry. 2000; 39: 9275-9284Crossref PubMed Scopus (128) Google Scholar)). I⊥(t) and I∥(t) were analyzed by the maximum entropy method (23Brochon J.C. Methods Enzymol. 1994; 240: 262-311Crossref PubMed Scopus (251) Google Scholar, 24Livesey A.K. Brochon J.C. Biophys. J. 1987; 52: 693-706Abstract Full Text PDF PubMed Scopus (425) Google Scholar) to determine the distributions of both lifetimes (τ) and rotational correlation times (θ), according to Equations 3 and 4. where αi represents the relative population characterized by the lifetime τi, and ρj represents the initial anisotropy related to a motion characterized by the rotational correlation time θj. r0′ is the apparent fundamental anisotropy value (typically 0.24-0.26 in the present study, which is below the fundamental anisotropy value, 0.3, for Trp at λex = 296 nm, suggesting that a fast component in the PFV-1 IN anisotropy decay cannot be resolved by the instrument as previously found for many proteins (25Lakowics J. Principles of Fluorescence Spectroscopy. 3rd Ed. Springer-Verlag New York Inc., New York2006Crossref Scopus (17137) Google Scholar)). The decay of the total fluorescence intensity, IT(t), and the fluorescence anisotropy decay, r(t), are then calculated from both polarized components according to Equations 7 and 8, respectively, where G represents the correction factor for the difference in the monochromator transmission between parallel and perpendicular polarized components. TFA and FCS Reveal Distinct Solubility Properties between HIV-1 and PFV-1 INs—TFA and FCS are complementary methods used to study the hydrodynamic properties of proteins, measuring the rotational and translational diffusion of molecules, respectively (25Lakowics J. Principles of Fluorescence Spectroscopy. 3rd Ed. Springer-Verlag New York Inc., New York2006Crossref Scopus (17137) Google Scholar). Overall rotational diffusion or flexibility are major causes of light depolarization, and anisotropy decay analysis allows the determination of rotational correlation time (θ) distribution. Long correlation time (θlong) represents the overall tumbling motion related to the protein hydrodynamic volume. In FCS, the temporal behavior of fluorescence fluctuations within a small excitation volume, described by the autocorrelation function, allow
Referência(s)