Báo cáo khoa học: Kinetic study of the HIV)1 DNA 3¢-end processing Single-turnover property of integrase docx

Kinetic study of the HIV)1 DNA 3¢-end processing Single-turnover property of integrase Maksim Smolov1, Marina Gottikh1, Vadim Tashlitskii1, Sergei Korolev1, Ilya Demidyuk2, Jean-Claude Brochon3, Jean-Francois Mouscadet3 and Eric Deprez3 ¸ Belozersky Institute of Physico-Chemical Biology, Moscow State University, Russia Institute of Molecular Genetics, Russian Academy of Science, Moscow, Russia ´ LBPA, UMR 8113 CNRS, IFR121, Ecole Normale Superieure de Cachan, France Keywords 3¢-processing; fluorescence anisotropy; integrase; protein–DNA interactions; singleturnover kinetics Correspondence E Deprez, LBPA, UMR 8113 CNRS, ´ IFR121, Ecole Normale Superieure de ´ Cachan, 61 avenue du President Wilson, 94235 Cachan cedex, France Fax: +33 47 40 76 84 Tel: +33 47 40 23 94 E-mail: deprez@lbpa.ens-cachan.fr (Received 17 October 2005, revised 20 December 2005, accepted 16 January 2006) doi:10.1111/j.1742-4658.2006.05139.x The 3¢-processing of viral DNA extremities is the first step in the integration process catalysed by human immunodeficiency virus (HIV)-1 integrase (IN) This reaction is relatively inefficient and processed DNAs are usually detected in vitro under conditions of excess enzyme Despite such experimental conditions, steady-state Michaelis–Menten formalism is often applied to calculate characteristic equilibrium ⁄ kinetic constants of IN We found that the amount of processed product was not significantly affected under conditions of excess DNA substrate, indicating that IN has a limited turnover for DNA cleavage Therefore, IN works principally in a singleturnover mode and is intrinsically very slow (single-turnover rate constant ¼ 0.004 min)1), suggesting that IN activity is mainly limited at the chemistry step or at a stage that precedes chemistry Moreover, fluorescence experiments showed that IN–DNA product complexes were very stable over the time-course of the reaction Binding isotherms of IN to DNA substrate and product also indicate tight binding of IN to the reaction product Therefore, the slow cleavage rate and limited product release prevent or greatly reduce subsequent turnover Nevertheless, the time-course of product formation approximates to a straight line for 90 (apparent initial velocity), but we show that this linear phase is due to the slow single-turnover rate constant and does not indicate steady-state multiple turnover Finally, our data ruled out the possibility that there were large amounts of inactive proteins or dead-end complexes in the assay Most of complexes initially formed were active although dramatically slow Integration of a DNA copy of the human immunodeficiency virus (HIV)-1 RNA genome into the human genome is an essential step in the viral replication cycle This process is catalysed by a viral protein, integrase (IN) The first key reaction in the overall integration process is the cleavage of a dinucleotide from each 3¢-end of the viral DNA (substrate DNA), a reaction termed 3¢-end processing In the second step, DNA strand transfer, a pair of processed DNA ends of the same viral DNA is inserted into the host cellular DNA (target DNA) The 3¢-end processing reaction requires a conserved nucleotide sequence at the viral DNA ends, but the second reaction does not absolutely require specific sequences within the host DNA For integration, IN has to bind simultaneously to the substrate and target DNA, and although the organization of the functional ternary complex IN–viral DNAtarget DNA is not yet known, the concerted integration mechanism very likely involves a multimeric active IN [1,2] Abbreviations HIV, human immunodeficiency virus; IN, integrase; LTR, long terminal repeat; PIC, preintegration complex; r, anisotropy; RSV, Rous sarcoma virus FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1137 Single-turnover kinetics of HIV-1 integrase M Smolov et al The 3¢-processing and strand transfer reactions can be carried out in vitro using purified recombinant IN, a divalent metal cation such as Mg2+ or Mn2+ and an oligonucleotide duplex that mimics one of the viral DNA ends The strand transfer reaction can also be studied by adding heterologous target DNA These simplified in vitro systems have been widely used to study the biochemical mechanism of DNA integration, and many IN inhibitors were initially characterized in such in vitro systems They include the so-called diketo acids, which preferentially inhibit strand transfer, and styrylquinolines, which inhibit 3¢-processing [3,4] Recently, it was shown that recombinant IN alone is able to perform the concerted joining reaction (i.e complete integration process involving two viral DNA ends) [5–7] Nevertheless, in all the reactions mentioned above, recombinant IN displays low catalytic activities and the reasons for such low activities remain unknown There have been significant advances in recent years regarding the ability of recombinant IN to use Mg2+ as a cofactor This divalent cation, which is believed to be the IN cofactor in vivo, is not equivalent to Mn2+ in vitro The specificity of catalysis is greater with Mg2+ and the choice by the enzyme of the nucleophile for the 3¢-processing reaction strongly depends on the nature of the cationic cofactor [8–11] Moreover, several drugs have different activities in Mn2+- and Mg2+-based activity assays, and Mg2+-dependent activities are usually more predictive of physiological behaviour [12–14] In addition, some mutations that confer resistance to inhibitor in vivo may have parallel effects in vitro in the presence of Mg2+, but not Mn2+ [15] Recently, we developed a new protocol for IN preparation, without using detergent during the purification, leading to substantial beneficial effects on many of the properties of IN, including its multimeric state and the ability to use Mg2+ as a cofactor [16] Although Mg2+-competent IN seems more specific and relevant to physiological activity, the Mg2+dependent activity remains low and does not significantly exceed the Mn2+-dependent activity With either cation, a high enzyme-to-DNA ratio (typically > 30 : 1) is required in vitro for efficient catalysis, although the reasons for such a high ratio are unclear Despite the abundant literature on IN, very little quantitative data on the kinetic properties of this protein are available Moreover, in some studies, Michaelis– Menten equations are applied, despite assays usually containing a large excess of IN over DNA substrate, an experimental condition that normally precludes such analytical treatment Quantitative evaluation of IN performances in vitro and the characterization of 1138 Mg2+-competent IN at the catalytic level, as well as the identification of rate-limiting steps in the overall catalytic process, are thus important, especially for pharmacological purposes Here, we describe a detailed kinetic analysis of HIV-1 IN under specific Mg2+ conditions using singleturnover formalism We found that IN is intrinsically very slow (single-turnover rate of DNA cleavage of 0.004 min)1) and works in a single-turnover mode even in the presence of an excess of DNA substrate Steadystate multiple turnover cannot be achieved for several reasons, including low cleavage rate and tight binding of IN to the processed DNA product The stability of IN during the time-course of the reaction and the influence of protein aggregation are discussed Results General features of the 3¢-end processing kinetics The kinetics of 3¢-end processing was studied using the U5-duplex that mimics the U5 long terminal repeat (LTR) sequence First, we used experimental conditions typical of those described in the literature, i.e an excess of enzyme over DNA substrate (3 nm DNA, 100 nm IN) in the presence of the physiologically relevant Mg2+ cofactor The time-course for the 3¢-processing reaction displayed three distinct phases (Fig 1B, upper) Phase I was a lag phase lasting $ 15– 20 (see also Fig 4B) and was followed by an apparent linear phase in which product formation versus time approximated to a straight line (phase II), although the experimental conditions, involving a high E : S ratio, were obviously non-Michaelis–Menten During phase III, as the substrate became depleted, the product concentration reached a plateau This plateau did not correspond to the complete conversion of DNA substrate to the cleaved product ($ 80% of the substrate was cleaved); this point is discussed further below The kinetic characteristics of the two first phases were then addressed Interpretation of the lag phase (phase I) Phase I may be due to slow binding of IN to DNA, as suggested by earlier studies [17,18] Thus, we studied the DNA-binding step using steady-state fluorescence anisotropy [17] with Fl–U5B ⁄ U5A duplex under conditions similar to those used in the activity assay As shown in Fig 1C, IN-bound DNA gave a higher anisotropy value than free DNA and equilibrium was reached after 20 of IN incubation with DNA substrate (the first-order kinetic constant k¢on was FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS M Smolov et al 2.5 [Product] (nM ) T S I II III 2.0 1.5 1.0 0.5 0.0 P 50 100 150 200 250 300 Time (min) N S P 15 30 60 90 120 150 180 240 300 10 [Product] (nM) 2.5 2.0 1.5 1.0 0.5 0.0 50 100 150 200 250 300 Time (min) C 0.25 0.2 r 0.23 min)1), suggesting that the lag phase corresponds to the DNA-binding step Indeed, phase I was not observed when IN was preincubated with DNA and Mg2+ for 30 at 20 °C (permissive temperature for DNA-binding but nonpermissive for activity) prior to the incubation at 37 °C (data not shown) or when the reaction was allowed to proceed by addition of Mg2+ after preincubation of IN with DNA at 37 °C (Fig 1B, lower) Under these two conditions, product formation was approximately two- to threefold higher during the first 20 compared with the experiment without preincubation (Fig 1B, upper) However, except for the absence of phase I, preincubation did not significantly influence the overall time-course of product formation (phases II + III) Interestingly, the DNA-binding step as measured by steady-state anisotropy was not strongly influenced by the DNA sequence or by the absence of a metal ion cofactor (data not shown): HIV-specific or random sequences in the presence or absence of Mg2+ gave similar DNAbinding kinetics However, in the absence of Mg2+, a higher limit anisotropy value (10% higher) was systematically obtained, suggesting the presence of high-order multimeric forms of IN, possibly aggregates, bound to DNA Furthermore, under conditions compatible with the 3¢-processing activity of IN, the steady-state anisotropy value which is related to the fractional saturation function remained stable throughout the activity experiment, i.e 300 (Fig 1D), suggesting that the processed DNA product has a strong affinity for IN (see also Fig 5) B A 0.15 0.1 0.05 0 10 20 30 Time (min) D 0.25 0.2 r Fig Kinetic study of the 3¢-processing reaction (A) Analysis of reaction products showing the weak nonspecific endonucleolytic activity of IN under the Mg2+ condition S, substrate (21-mer); P, 3¢-processing product (19-mer); N, nonspecific products (£ 18mer) Strand transfer products (T) are estimated to be £ 5% of total products Incubation time was 160 (lane 1) or 190 (lane 2) (B) Time-course of cleaved product formation (Upper) DNA substrate was mixed at t ¼ with IN in the presence of MgCl2 Substrate (21-mer) and product (19-mer) were separated by gel electrophoresis (below the curve) and quantified as indicated in Experimental procedures (Lower) DNA substrate was first preincubated with IN for 30 at 37 °C in the absence of Mg2+ MgCl2 was then added to the mixture to start the reaction (t ¼ 0) (C) Binding of IN to DNA at 37 °C as monitored by steady-state fluorescence anisotropy (r) IN was added to fluorescein-labelled U5-duplex in the reaction buffer and r-values were recorded at time intervals of 50 s for 30 Initial r-value (0.055) corresponds to free DNA The DNA binding at 20 °C (not shown) was only slightly slower than that at 37 °C (k’on, 20 °C ¼ 0.18 min)1; k’on, 37 °C ¼ 0.23 min)1) (D) IN–DNA complexes are stable throughout the time-course of 3¢-processing Binding of IN to fluorescein-labelled U5-duplex was monitored by fluorescence anisotropy for h at 37 °C In all experiments, DNA and IN concentration were and 100 nM, respectively Single-turnover kinetics of HIV-1 integrase 0.15 0.1 0.05 0 50 100 150 200 250 300 Time (min) Determination of equilibrium and catalytic constants (phase II) During phase II ($ 90 min), the product concentration increased linearly with time suggesting that this FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1139 Single-turnover kinetics of HIV-1 integrase M Smolov et al 1140 A 0.005 2000 0.003 1500 1/kobs (min) kobs (min-1) 0.004 0.002 0.001 1000 500 0.00 0.000 50 0.08 0.16 -1 1/[IN] (nM ) 100 150 0.24 200 [IN] (nM) B 0.8 [ P r o d u c t ] ( n M) phase is comparable with the initial velocity (vi) in a Michaelis–Menten enzyme reaction although E ) S (single-turnover conditions) As standard Michaelis– Menten formalism is not applicable in the case of IN for quantitative analysis of the 3¢-processing kinetics, we used a modified formalism corresponding to concentration conditions in which E ) S, according to the enzymatic model presented in the Experimental Procedures As the equilibrium and kinetic parameters were not steady-state parameters, we called them Km¢ and kcat¢, respectively Under conditions of catalyst excess, kcat¢ is a single-turnover rate corresponding to the actual cleavage reaction which is not affected by subsequent steps such as, for instance, limited release of product (the cleavage reaction accounts for all events that precede and include the chemistry step) Despite the apparent linearity of phase II, the time-course of product formation (phases II + III) actually corresponds to an exponential law (see Eqn 11; the reason why phase II displays apparent linearity is explained below) (Equations to 13 are given in the Experimental Procedures below.) The first-order kinetic constant (called kobs in Eqn 9) is obtained by directly fitting data to a single exponential equivalent to Eqn (11) [19–21] The dependence of kobs values on [IN] allows parameters Km¢ and kcat¢ to be determined according to Eqns (9) and (10) Directly fitting the hyperbolic curve (Fig 2A, kobs as the function of [IN]0) according to Eqn (9) gave Km¢ and kcat¢ values of 26 nm and 0.004 min)1, respectively The plot of ⁄ kobs as a function of ⁄ [IN]0 gave a straight line (Fig 2A, inset) in agreement with Eqn (10) and similar Km¢ and kcat¢ values were derived from this plot (30 nm and 0.0045 min)1, respectively) Comparison between the first-order kinetic constants for DNA binding and catalytic steps (0.23 min)1, Fig 1C and 0.004 min)1, Fig 1B, respectively) indicates that the catalytic reaction was very slow compared with the DNA-binding step Moreover, because the kcat¢ value is very low, Km¢ is a good estimation of the Kd value (quasiequilibrium assumption) Indeed, 26– 30 nm is close to values obtained previously in DNA-binding assays [17] The IN concentration was varied between and 200 nm in Fig 2A because, in this concentration range, the 3¢-processing activity increased as the IN concentration increased In fact, we found that the activity was maximal at 250 nm and then decreased dramatically as concentration increased (Fig 2B) The low 3¢-processing activities of IN at protein concentrations > 250 nm can be ascribed to IN aggregation, as suggested previously [16,22] 0.6 0.4 0.2 0.0 1000 2000 3000 4000 5000 [IN] (nM) Fig Influence of the IN concentration on the 3¢-end processing efficiency (A) Determination of Km¢ and kcat¢ parameters by plotting kobs ¼ f([IN]) IN concentration was varied between and 200 nM The 3¢-processing reaction was allowed to proceed for either 36 (filled circles) or 60 (unfilled circles) DNA product concentration was measured as described in Experimental Procedures Km¢ (26 nM) and kcat¢ (0.004 min)1) were estimated using Eqn (9) The inset shows ⁄ kobs ¼ f(1 ⁄ [IN]) This plot was also fitted according to Eqn (10) to estimate Km¢ (30 nM) and kcat¢ (0.0045 min)1) (B) Influence of the IN concentration on the processing activity IN (5 nM to lM) was incubated with nM DNA substrate for h at 37 °C The linear regression shown in (A) was obtained from the bell-shaped dose–response, by selecting data from the increasing phase (corresponding to IN concentrations between and 200 nM inclusive) Insights into the linear phase (phase II) Because the curve in Fig 1B approximates to a straight line in phase II, we were interested in understanding the apparent Michaelis–Menten behaviour of IN, despite the enzyme concentration being so much higher than the DNA substrate concentration This is mathematically possible if kcat¢ is sufficiently low given d[product] ⁄ dt ¼ constant even in the absence of a FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS M Smolov et al Single-turnover kinetics of HIV-1 integrase steady-state (Eqn 13) Therefore, we investigated the influence of the kcat¢ value on the linear phase (phase II) using simulation analysis to determine whether the experimental kcat¢ was compatible with this hypothesis The dependences of ES (IN*DNAS) and product (DNAP) concentrations on time obey Eqns (7) and (11), respectively, and were simulated using two different kcat¢ values (Fig 3) The kcat¢ value used in the first simulation (light grey) was the value found experimentally (0.004 min)1), whereas an arbitrary higher kcat¢ value of 0.02 min)1 was used in the second simulation (dark grey) All other parameters were identical in both simulations (see legend to Fig 3) The rst simulation using kcat ẳ 0.004 min)1 shows that the timecourse of product formation approximates to a straight line during approximately the first hour (Fig 3B) This approximation is not valid beyond 80 (Fig 3A) Taking into account that the lag phase was not simulated, this result is consistent with the experimental duration of phase II (Fig 1B) The expected concentration of IN*DNAS complex as a function of time, is shown in Fig 3C,D These plots show that the product concentration increases linearly with time as long as the ES concentration does not decrease below 20% of the initial value For larger decreases (when ES cannot be considered as constant with time), the time-course of product formation becomes strongly nonlinear We verified that the time range for which the approximation ES ¼ constant is valid depends directly on the value of kcat¢ In the second simulation with kcat ẳ 0.02 min)1 the resulting product formation over time was clearly nonlinear (Fig 3A,B, dark grey), corresponding to more standard first-order behaviour as expected under single-turn- over conditions This nonlinear behaviour is related to a rapid decrease in the ES concentration which exceeds 20% at t ¼ 10 (Fig 3C,D) In conclusion, the apparent linear phase, as found experimentally under conditions of enzyme excess, originates in the low single-turnover rate constant (kcat¢) because ES can be considered constant in this phase This is compatible with the rapid formation of ES (0.23 min)1) compared with the product formation (0.004 min)1) The exponential term of Eqn (12) can be neglected when kcat¢ is low and d[DNAP] ⁄ dt can be considered constant according to Eqn (13) The simulations indicate that simplification of Eqn (12) to Eqn (13) and the phase II duration are compatible with the experimental kcat¢ value In conclusion, the linear phase does not necessarily indicate a steady-state multiple-turnover mechanism In the case of IN, it corresponds to a singleturnover reaction with a slow rate for the chemical step This ‘linear’ phase is apparent and actually corresponds to an exponential phase We verified that under our experimental conditions, inactive enzymes were not in excess over active enzymes In such a case, the situation could become similar to a standard Michaelis–Menten condition in which S ) E This possibility has been carefully addressed in Mn2+-dependent reactions previously [23,24] Here, the presence or otherwise of large amounts of inactive IN was assessed using the general Cornish–Bowden relation [25]: kcat ẵIN0 ẵDNA0 ẵDNAP ị dẵDNAP ẳ dt Km ỵ ẵIN0 ỵ ẵDNA0 ẵDNAP ị Eqn (14) is valid in two cases: either [IN]0 ) [DNA]0 (single turnover) or [DNA]0 ) [IN]0 (multiple turn- A Fig Simulation analysis of the processed DNA product formation and of the change in the IN–DNA substrate complex concentration during the reaction Product formation (A, B) and IN–DNA substrate complex concentration (C, D) were simulated using Eqns (11) and (7), respectively, covering the time range 0–400 (A, C) or 0–60 (B, D) Kinetic simulations were performed using either kcat¢ as found experimentally (0.004 min)1; light grey) or an arbitrary higher kcat¢ value (0.02 min)1; dark grey) All the other parameters were identical in both simulations: IN and DNA substrate concentrations were 150 and nM, respectively The Km¢ value was 30 nM (estimated from Fig 2A) ð14Þ B C D FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1141 Single-turnover kinetics of HIV-1 integrase M Smolov et al over) Under single-turnover conditions, this expression leads to the first-order kinetic constant kobs as defined in Eqn (9) In that case, kobs depends hyperbolically on enzyme concentration Nevertheless, one possibility is that the concentration of active IN (INa) in reaction mixtures could be much lower than the con< centration of total IN Assuming [INa]0 < [DNA]0, the observed kinetics of IN could be due to steadystate multiple turnovers of a small number of active < : is required for optimal activity (Fig 2B) The increasing phase for ratio up to 70 : is simply due to the fractional saturation function that increases as the IN concentration increases Nevertheless, for very high ratio (> 70), the activity is lower, most probably due to aggregation [22] This is also consistent with previous data showing that aggregation occurs mainly above an IN concentration of 200–300 nm [16] FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1145 Single-turnover kinetics of HIV-1 integrase M Smolov et al Multiple turnover of Rous sarcoma virus (RSV) IN using Mn2+ as the cofactor has been described previously for 3¢-processing [32] In contrast, in the presence of Mg2+, another study with RSV IN showed that only 5% of DNA substrates are processed in h using an excess of enzyme over substrate [10], which is more consistent with our data Using our HIV-1 IN preparation under either Mg2+ or Mn2+ conditions, we did not find any significant quantitative difference in the time-course of the 3¢-processing product formation, suggesting similar single-turnover rate constants (data not shown), although we confirmed that the Mn2+dependent activity is less specific than the Mg2+dependent activity Moreover, the anisotropy study suggesting the tight binding of IN to DNA product (as shown in Fig 1D) displays no difference between Mg2+ and Mn2+ experiments (data not shown) Therefore, it appears that IN exhibits single-turnover properties irrespective of the metallic cofactor However, catalytic turnover of IN has been described in several studies of the disintegration reaction using Y or dumbbell substrates and Mn2+ [33–35] The apparent discrepancy may be a consequence of the difference in the nature of the DNA substrate Disintegration is known to be less specific than 3¢-processing because: (a) single mutants (for instance K156E or K159E) [36] and truncated proteins, inactive for 3¢-processing or the joining reaction, remain competent for disintegration; and (b) a strict requirement of Mn2+ or a large preference for Mn2+ over Mg2+ is generally observed in disintegration tests with truncated or full-length IN, respectively In addition, Gerton & Brown [37] have shown that the core domain of IN can turnover faster than full-length IN in disintegration assays Thus, it appears that multiple turnover of IN is dependent on conditions that disfavour stringency or reaction specificity We now consider the molecular basis for the absence of steady-state turnover by IN for 3¢-processing The steady-state rate constant kcat (not measurable in the case of IN) is a composite constant including binding and docking of the ES complex, chemistry, product dissociation and subsequent recycling steps, whereas the single-turnover rate constant kcat¢ monitors only events that precede and include chemistry The kcat¢ value is therefore independent of the product release, which is not monitored under single-turnover conditions The low value of kcat¢ (0.004 min)1) is compatible with most studies of 3¢-processing activity In some cases, faster reaction rates have been reported (up to 10· faster) and this is generally related to the presence of cosolvent or detergent in the reaction buffer There are several possible explanations for such a 1146 slow catalysis rate, which represents, together with the limited product release, the main limiting step for multiple turnover First, it is unlikely that DNA binding is limiting even if it is slow This step is responsible for the sigmoidal shape of the product formation versus time plot (phase I) but is over after 15–20 Therefore, all the IN–DNA substrate complexes are already formed at the beginning of the linear phase (phase II) Indeed, preformation of complexes at a temperature nonpermissive for activity or at 37 °C in the absence of Mg2+ abolished the lag phase without changing significantly the kcat¢ value calculated from phase II Moreover, the anisotropy approach reveals the nonspecific nature of IN–DNA interactions because DNA-binding kinetics are similar, irrespective of the sequence and, consequently, cannot discriminate between specific (or catalytically active) and nonspecific (or catalytically inactive) complexes when IN binds to the HIV DNA substrate Taken together, these results suggest that the DNA-binding step corresponding to phase I accounts for the formation of nonspecific complexes and that, at the beginning of phase II, the majority of complexes are nonspecific The limiting step occurs after the nonspecific DNA binding and may correspond to a step before the chemistry or the chemistry itself It is important to underline that the single-turnover behaviour of IN (as shown in Fig 1B) highlights a distribution of reaction velocities in which kcat¢ represents the average singleturnover rate constant: Fig 1B shows that 50% of the DNA substrate is converted into product before t ¼ 100 and 50% is converted after this time It is unlikely that the chemistry step itself (i.e the nucleophilic attack of the phosphodiester bond by a water molecule) would be responsible for such a distribution Most likely, this distribution of velocities corresponds to an equilibrium displacement ([inactive complex] « [ ] « [active complex]) with the presence of a minority of active complexes at the beginning of the reaction (multiple turnovers of the most efficient complexes are then limited by the high stability of the IN– DNA product complexes) Thus, it is hypothesized that a step following DNA binding of IN but prior to cleavage is rate limiting This step corresponds to a relaxation step that leads to a specific and catalytically competent conformation of the IN–DNA complex and the low single-turnover rate constant (kcat¢) may then originate from the slow conversion from the inactive state to the active state We propose three different models to explain slow kinetics in one round of catalysis (Fig 7) Models and are based on the intrinsic ability of IN to bind DNA in two binding modes, specific and nonspecific Model does not address the FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS M Smolov et al problem of nonspecific ⁄ specific partition but, rather, a slow conformational change of IN or ⁄ and DNA within the complex that may be rate limiting It is well known that there is no sequence specificity for the DNA-binding step of IN although the 3¢-processing reaction requires specific sequences within the U3 or U5 viral DNA ends [9] Sequence specificity is thus strictly required at the catalytic level but not for the DNA-binding step Consequently, the nonspecific binding mode of IN which is essential for the integration reaction may be detrimental to the 3¢-processing reaction, at least in vitro The nonlinear dependence of kobs on IN concentration (Fig 2A) precludes the possibility that the assay contains a large proportion of denaturated forms of IN that originate from the purification procedure, although this approach cannot rule out the presence of a binding but catalytically inactive form of IN [21] However, our data show that IN bound in a nonspecific manner is not definitively trapped as most complexes initially formed, even if nonproductive on a short timescale, are potentially active The search for the cleavage site could be either slow linear diffusion along the DNA substrate (model 1, Fig 7A), in a similar manner to that described for restriction endonucleases and methyltransferases [38,39] or via a slow relaxation process between two states (model 2, Fig 7B), one corresponding to viral DNA bound at the ‘nonviral’ site (normally occupied by the target DNA), the other corresponding to viral DNA bound at the ‘viral’ site (specific site) Several studies on mutants and chimeric IN strongly support the existence of these two sites in the catalytic core domain [40,41] Using a bifunctional diketo acid derivative efficiently inhibiting both 3¢-processing and strand transfer reactions, Pommier and co-workers have shown that the two sites probably overlap [42] in agreement with molecular docking studies suggesting that these sites are close to each other within the catalytic core [43,44] Hence, they are close enough inside the active site to allow viral DNA displacement to the viral site when initially bound to the nonviral site Models and are both compatible with the ability of IN to bind nonspecific sequences of DNA tightly and are consistent with a previous study suggesting a rapid and nonspecific DNA binding of IN followed by a slow and specific catalysis step [45] Moreover, faster single-turnover kinetic rates were found for the disintegration reaction than for 3¢-processing (2–3 orders of magnitude higher) [46], reinforcing the idea that the limiting step is related to specificity requirements However, it seems unlikely that either of these mechanisms occurs in vivo because IN is probably already positioned at the ends of the viral genome in the Single-turnover kinetics of HIV-1 integrase A IN CAGT inactive B Catalytic core of IN viral nonviral active viral nonviral C IN IN inactive active Fig Schematic diagram of three models compatible with the formation of the catalytically competent viral IN–DNA complex (A) DNA scanning by linear diffusion (model 1) Owing to the ability of IN to bind nonspecifically but tightly to DNA, the majority of complexes initially formed are nonspecific and nonproductive; linear diffusion allows the appropriate positioning of IN onto DNA extremity for catalysis (B) Equilibrium between two DNA-bound forms of IN (model 2) The viral DNA extremity is bound either to the nonspecific site (nonviral) (upper, inactive complex) or the specific site (viral) (lower, active complex) (C) Induced fit of the IN onto the DNA substrate that leads to the active complex conformation (model 3) preintegration complex (PIC) context Model (Fig 7C) involves a rate-limiting conformational change within the IN–DNA complex Such an induced fit may be strictly required before catalytic cleavage can proceed Cross-linking data suggest that complexes obtained with DNA substrate and those with product have different conformations [30] This suggests that a conformation change may occur within the IN–DNA complex, although there is no clear evidence that such FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1147 Single-turnover kinetics of HIV-1 integrase M Smolov et al a conformational change is a prerequisite or a consequence of the cleavage activity and then actually occurs when the complex is initially formed with the DNA substrate Aggregation properties or the multimeric status of IN may be also critical factors that explain the low kcat¢ value found in vitro In this case, the rate-limiting step in model may correspond to any modification in the quaternary structure of IN prior to catalysis process such as the organization of the competent dimeric form for the 3¢-processing reaction Our analysis of binding isotherms indicates that IN binds the viral DNA substrate and the cleaved DNA product equally well The strong affinity of IN for its reaction product also limits enzyme turnover in vitro although it may be a functional advantage in vivo Numerous enzymes are characterized by rate-limiting product dissociation Usually, this causes a burst of product formation (presteady-state phase which accounts for the first turnover) under multiple-turnover conditions Consistently, the discrepancy between kcat (multiple turnover) and kchemistry (single turnover, also called kcat¢ here) values, where kchemistry exceeds kcat, indicates that a step that follows chemistry such as product release is limiting For instance, the DNA repair enzymes Salmonella endonuclease V and Escherichia coli mismatch uracil glycosylase are able to support multiple-turnover only to a limited extent [47,48] These enzymes possess high affinity for their reaction products, which strongly restricts any cycling mode of catalysis In most cases, the functional reason for the tight binding of the reaction product to the catalyst is not clear In the case of IN, tight binding of the protein to recessed viral DNA ends is obligatory in the context of the HIV replication cycle: the 3¢-processing of viral DNA ends is only the first reaction of the overall integration process In the cell, this reaction can occur in the cytoplasm, whereas integration is obviously nuclear Consequently, for concerted integration, there must be a ternary complex with IN simultaneously bound to both the processed viral DNA (donor) and the target DNA (acceptor) Thus, the complex involving IN and viral DNA extremities must be stable enough for the PIC to enter the nucleus and subsequently integrate, although the 3¢-processing reaction has already been completed The fact that the 3¢-processing reaction product is trapped inside the active site is functionally beneficial, as it optimizes integration yield Moreover, in addition to catalysis and functional considerations, the stable nucleoprotein complex might protect the DNA from degradation by nucleases and also minimize the number of free DNA ends, thereby limiting apoptotic responses 1148 With its inefficient cleavage and tight binding to DNA product, IN is similar to other members of the polynucleotidyl transferase family, including, for example, transposases These enzymes share a common catalytic property, they have evolved to catalyse multiple-sequential steps (two in the case of IN and four in the case of Tn5 transposase) in a single active site A multiple-step process implicitly requires tight binding of reaction product after each chemical step to optimize the overall process but dramatically limits turnover The poor efficiency or velocity of these enzymes is, however, not detrimental to their function because a single transposition or integration event is biologically sufficient Experimental procedures Oligodeoxyribonucleotides and integrase Complementary oligonucleotides U5B, 5¢-GTGTGGAAAA TCTCTAGCAGT-3¢ and U5A, 5¢-ACTGCTAGAGATTT TCCACAC-3¢, were synthesized using a 380B Applied Biosystems synthetizer by the standard cyanoethyl phosphoramidite procedure Oligonucleotides Fl–U5B, Fl)5¢-GTGTGG AAAATCTCTAGCAGT-3¢ and Fl–U5B-2 Fl)5¢-GTGTGG AAAATCTCTAGCA-3¢ (where Fl designates fluorescein) were purchased from Eurogentec (Liege, Belgium) Terminal nucleotides removed by IN during the 3¢-processing reaction are underlined All oligonucleotides were further purified on an 18% denaturing acrylamide ⁄ urea gel The detergent-free recombinant IN protein was produced and purified as previously described [16] 32 P-labelling of U5-DNA Ten picomoles of U5B oligonucleotide was 5¢-end labelled with 32P using 25 activity units of T4 polynucleotide kinase and 50 lCi of [32P]ATP[cP] (3000 CiỈmmol)1) T4 polynucleotide kinase was inactivated by EDTA and heating at 65 °C for followed by enzyme extraction with phenol ⁄ chlorophorm ⁄ isoamyl alcohol (25 : 24 : v ⁄ v ⁄ v) An equimolar quantity of complementary U5A oligonucleotide was then added The mixture was heated to 90 °C for and the U5-duplexes were annealed by slow cooling to room temperature The U5-duplexes were purified on Micro Bio-Spin columns P-6 (Bio-Rad, Munich, Germany) 3¢-End processing activity IN activity was studied by mixing IN and 32P-labelled U5-duplex in 20 lL of a buffer containing 20 mm Hepes (pH 7.5), 10 mm dithiothreitol, 7.5 mm MgCl2 at 37 °C Various concentrations of both DNA and IN, and incubation times were used (see indications in figure legends) The reaction was stopped with 80 lL of a buffer containing FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS M Smolov et al Single-turnover kinetics of HIV-1 integrase mm Tris ⁄ HCl (pH 7.5), mm EDTA, 0.125 mgỈmL)1 glycogene, 400 mm NaOAc DNA fragments were precipitated with ethanol, then suspended in loading dye (80% formamide, 0.05% bromophenol blue, 0.05% xylene cyanol) and separated on a 20% polyacrylamide denaturing gel Gels were analysed on a STORM 840 PhosphorImager (Molecular Dynamics, Sunnyvale, CA, USA) and quantified using image quant 4.1 software Eqn (7) predicts the IN–DNA substrate complex concentration during phase II From Eqns (3) and (7), ln ẵDNA0 ẳ kobs t ẵDNA0 ẵDNAP 8ị with kobs ẳ kcat0 Km ẵIN0 ỵ 9ị Eqn (9) is equivalent to Analysis of kinetic data and determination of apparent Km and kcat values (Km and kcatÂ) 1 K ẳ ỵ 0m kobs kcat kcat ẵIN0 For 3Â-processing kinetic study, the amount of 19-mer reaction products was measured and quantified by gel electrophoresis and PhosphoImager scanning as described above Only a weak nonspecific endonucleotytic activity of IN was observed under Mg2+ conditions and strand transfer products were estimated to be ¼ 5% of total products (Fig 1A) Single turnover of the 3¢-end processing reaction with excess of enzyme over DNA substrate was analysed according to the following model: kcat k1 ) IN ỵ DNAS ( IN DNAS ! DNAP kÀ1 ð1Þ According to Eqn (8), the function that predicts product formation over time is given by: ð2Þ where [DNA]0 represents the total concentration of DNA Moreover, vi0 ẳ dẵDNAP is constant in phase II dt (pseudo-initial velocity phase) as observed experimentally (see Fig and text) The observation of a linear phase means that the ES complex can be considered to be constant during this period Thus, vi ẳ k0cat ẵIN DNAS t cat Km0 ỵ1 ẵIN0 k ẵDNAP ẳ ẵDNA0 e ị 11ị Thus, k t cat Km0 dẵDNAP ẵDNA0 kcat0 ỵ1 ẳ e ẵIN0 vi ẳ Km0 dt ỵ1 12ị ẵIN0 when where S and P designate substrate and product, respectively The conservation relationship for DNA substrate can be written as: ẵDNA0 ẳ ẵDNAS ỵ ẵIN DNAS ỵ ẵDNAP 10ị kcat0 ! e; dẵDNAP ẵDNA0 kcat0 ẳ Km0 dt ỵ1 13ị ẵIN0 Eqns (9) and (10) were used to determine Km¢ and kcat¢ experimentally The kobs values were obtained by fitting kinetics according to Eqn (11) (with [DNA]0 ẳ [DNAp]+Ơ when the reaction is not total) Note that kobs approximates to kcat¢ when the IN concentration is high compared with Km¢ The 3¢-processing reaction was performed with various IN concentrations (5–200 nm) and nm 32P-labelled DNA substrate (U5-duplex) for 36 or 60 3ị Steady-state uorescence anisotropy and dẵIN DNAS %0 dt 4ị Thus, 4ị k1 ẵINẵDNAS k1 ỵ kcat ị ẵIN DNAS ẳ and Km ẳ k1 ỵ k0cat ẵINẵDNAS ẳ ẵIN DNAS k1 5ị Because [IN]0 ẳ [IN] + [IN*DNA] % [IN] (when [IN]0 ) [DNA]0), Eqn (5) can be rearranged as: Km0 ẳ k1 ỵ kcat ẵIN0 ẵDNAS ẳ ẵIN DNAS k1 6ị From Eqns (2) and (6) ẵIN DNAS ẳ ẵDNA0 ẵDNAP Km0 ẵIN0 ỵ1 7ị Steady-state anisotropy was used to estimate the fractional saturation function ([IN*DNA] ⁄ [DNA]0) The measurements were performed as previously described [17] using U5-duplexes, Fl–U5B ⁄ U5A or Fl–U5B-2 ⁄ U5A, composed of U5A oligonucleotide and a 5¢-end fluorescein-labelled complementary strand, either Fl–U5B or Fl–U5B-2 Fl– U5B ⁄ U5A and Fl–U5B-2 ⁄ U5A mimic blunt and processed U5 viral DNA ends, respectively Pseudo first-order kinetic constant (k¢on) for DNA-binding was determined using an exponential fit Acknowledgements This work was supported by the TRIoH European project (FP6 grant 503480), the Russian Foundation for Basic Research (grants 04-04-22000 and 05-0448743), the French National Agency for Research FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1149 Single-turnover kinetics of HIV-1 integrase M Smolov et al against AIDS (ANRS) and the PICS program (n°271) ´ We thank Dr Gerald Peyroche for critically reading the manuscript and Francoise Simon for technical ¸ assistance 14 References Craigie R (2001) HIV integrase, a brief overview from chemistry to therapeutics J Biol Chem 276, 23213–23216 Chiu TK & Davies DR (2004) Structure and function of HIV-1 integrase Curr Top Med Chem 4, 965–977 Hazuda DJ, Felock P, Witmer M, Wolfe A, Stillmock K, Grobler JA, Espeseth A, Gabryelski L, Schleif W, Blau C et al (2000) Inhibitors of strand transfer that prevent integration and inhibit HIV-1 replication in cells Science 287, 646–650 Mekouar K, Mouscadet JF, Desmaele D, Subra F, Leh H, Savoure D, Auclair C & D’Angelo J (1998) Styrylquinoline derivatives: a new class of potent HIV-1 integrase inhibitors that block HIV-1 replication in CEM cells J Med Chem 41, 2846–2857 Sinha S, Pursley MH & Grandgenett DP (2002) Efficient concerted integration by recombinant human immunodeficiency virus type integrase without cellular or viral cofactors J Virol 76, 3105–3113 Sinha S & Grandgenett DP (2005) Recombinant human immunodeficiency virus type integrase exhibits a capacity for full-site integration in vitro that is comparable to that of purified preintegration complexes from virusinfected cells J Virol 79, 8208–8216 Li M & Craigie R (2005) Processing of viral DNA ends channels the HIV-1 integration reaction to concerted integration J Biol Chem 280, 29334–29339 Engelman A & Craigie R (1995) Efficient magnesiumdependent human immunodeficiency virus type integrase activity J Virol 69, 5908–5911 Esposito D & Craigie R (1998) Sequence specificity of viral end DNA binding by HIV-1 integrase reveals critical regions for protein–DNA interaction EMBO J 17, 5832–5843 10 Bushman FD & Wang B (1994) Rous sarcoma virus integrase protein: mapping functions for catalysis and substrate binding J Virol 68, 2215–2223 11 Skinner LM, Sudol M, Harper AL & Katzman M (2001) Nucleophile selection for the endonuclease activities of human, ovine, and avian retroviral integrases J Biol Chem 276, 114–124 12 Hazuda D, Felock P, Hastings J, Pramanik B, Wolfe A, Goodarzi G, Vora A, Brackmann K & Grandgenett D (1997) Equivalent inhibition of half-site and full-site retroviral strand transfer reactions by structurally diverse compounds J Virol 71, 807–811 13 Neamati N, Lin Z, Karki RG, Orr A, Cowansage K, Strumberg D, Pais GC, Voigt JH, Nicklaus MC, 1150 15 16 17 18 19 20 21 22 23 24 Winslow HE et al (2002) Metal-dependent inhibition of HIV-1 integrase J Med Chem 45, 5661–5670 Marchand C, Johnson AA, Karki RG, Pais GC, Zhang X, Cowansage K, Patel TA, Nicklaus MC, Burke TR Jr & Pommier Y (2003) Metal-dependent inhibition of HIV-1 integrase by beta-diketo acids and resistance of the soluble double-mutant (F185K ⁄ C280S) Mol Pharmacol 64, 600–609 Grobler JA, Stillmock K, Hu B, Witmer M, Felock P, Espeseth AS, Wolfe A, Egbertson M, Bourgeois M, Melamed J et al (2002) Diketo acid inhibitor mechanism and HIV-1 integrase: implications for metal binding in the active site of phosphotransferase enzymes Proc Natl Acad Sci USA 99, 6661–6666 Leh H, Brodin P, Bischerour J, Deprez E, Tauc P, Brochon JC, LeCam E, Coulaud D, Auclair C & Mouscadet JF (2000) Determinants of Mg2+-dependent activities of recombinant human immunodeficiency virus type integrase Biochemistry 39, 9285–9294 Deprez E, Barbe S, Kolaski M, Leh H, Zouhiri F, Auclair C, Brochon JC, Le Bret M & Mouscadet JF (2004) Mechanism of HIV-1 integrase inhibition by styrylquinoline derivatives in vitro Mol Pharmacol 65, 85–98 Pinskaya M, Romanova E, Volkov E, Deprez E, Leh H, Brochon JC, Mouscadet JF & Gottikh M (2004) HIV-1 integrase complexes with DNA dissociate in the presence of short oligonucleotides conjugated to acridine Biochemistry 43, 8735–8743 Uter NT & Perona JJ (2004) Long-range intramolecular signaling in a tRNA synthetase complex revealed by pre-steady-state kinetics Proc Natl Acad Sci USA 101, 14396–14401 Etzkorn C & Horton NC (2004) Ca2+ binding in the active site of HincII: implications for the catalytic mechanism Biochemistry 43, 13256–13270 Brocklehurst K, Resmini M & Topham CM (2001) Kinetic and titration methods for determination of active site contents of enzyme and catalytic antibody preparations Methods 24, 153–167 Hawkins ME, Pfleiderer W, Mazumder A, Pommier YG & Balis FM (1995) Incorporation of a fluorescent guanosine analog into oligonucleotides and its application to a real time assay for the HIV-1 integrase 3¢-processing reaction Nucleic Acids Res 23, 2872–2880 Tramontano E, Colla PL & Cheng YC (1998) Biochemical characterization of the HIV-1 integrase 3¢-processing activity and its inhibition by phosphorothioate oligonucleotides Biochemistry 37, 7237–7243 Bugreev DV, Baranova S, Zakharova OD, Parissi V, Desjobert C, Sottofattori E, Balbi A, Litvak S, Tarrago-Litvak L & Nevinsky GA (2003) Dynamic, FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS M Smolov et al 25 26 27 28 29 30 31 32 33 34 35 thermodynamic, and kinetic basis for recognition and transformation of DNA by human immunodeficiency virus type integrase Biochemistry 42, 9235–9247 Cornish-Bowden A (1976) Principles of Enzyme Kinetics Butterworth, London Engelman A, Hickman AB & Craigie R (1994) The core and carboxyl-terminal domains of the integrase protein of human immunodeficiency virus type each contribute to nonspecific DNA binding J Virol 68, 5911– 5917 Mazumder A, Gupta M, Perrin DM, Sigman DS, Rabinovitz M & Pommier Y (1995) Inhibition of human immunodeficiency virus type integrase by a hydrophobic cation: the phenanthroline–cuprous complex AIDS Res Hum Retroviruses 11, 115–125 Deprez E, Tauc P, Leh H, Mouscadet JF, Auclair C, Hawkins ME & Brochon JC (2001) DNA binding induces dissociation of the multimeric form of HIV-1 integrase: a time-resolved fluorescence anisotropy study Proc Natl Acad Sci USA 98, 10090–10095 Deprez E, Tauc P, Leh H, Mouscadet JF, Auclair C & Brochon JC (2000) Oligomeric states of the HIV-1 integrase as measured by time-resolved fluorescence anisotropy Biochemistry 39, 9275–9284 Gao K, Butler SL & Bushman F (2001) Human immunodeficiency virus type integrase: arrangement of protein domains in active cDNA complexes EMBO J 20, 3565–3576 Faure A, Calmels C, Desjobert C, Castroviejo M, Caumont-Sarcos A, Tarrago-Litvak L, Litvak S & Parissi V (2005) HIV-1 integrase crosslinked oligomers are active in vitro Nucleic Acids Res 33, 977–986 Jones KS, Coleman J, Merkel GW, Laue TM & Skalka AM (1992) Retroviral integrase functions as a multimer and can turn over catalytically J Biol Chem 267, 16037–16040 Reinke RA, Lee DJ, McDougall BR, King PJ, Victoria J, Mao Y, Lei X, Reinecke MG & Robinson WE Jr (2004) 1-Chicoric acid inhibits human immunodeficiency virus type integration in vivo and is a noncompetitive but reversible inhibitor of HIV-1 integrase in vitro Virology 326, 203–219 King PJ, Lee DJ, Reinke RA, Victoria JG, Beale K & Robinson WE Jr (2003) Human immunodeficiency virus type-1 integrase containing a glycine to serine mutation at position 140 is attenuated for catalysis and resistant to integrase inhibitors Virology 306, 147–161 Gerton JL, Ohgi S, Olsen M, DeRisi J & Brown PO (1998) Effects of mutations in residues near the active site of human immunodeficiency virus type integrase on specific enzyme–substrate interactions J Virol 72, 5046–5055 Single-turnover kinetics of HIV-1 integrase 36 Jenkins TM, Esposito D, Engelman A & Craigie R (1997) Critical contacts between HIV-1 integrase and viral DNA identified by structure-based analysis and photo-crosslinking EMBO J 16, 6849–6859 37 Gerton JL & Brown PO (1997) The core domain of HIV-1 integrase recognizes key features of its DNA substrates J Biol Chem 272, 25809–25815 38 Surby MA & Reich NO (1996) Contribution of facilitated diffusion and processive catalysis to enzyme efficiency: implications for the EcoRI restrictionmodification system Biochemistry 35, 2201–2208 39 Jeltsch A, Wenz C, Stahl F & Pingoud A (1996) Linear diffusion of the restriction endonuclease EcoRV on DNA is essential for the in vivo function of the enzyme EMBO J 15, 5104–5111 40 Harper AL, Skinner LM, Sudol M & Katzman M (2001) Use of patient-derived human immunodeficiency virus type integrases to identify a protein residue that affects target site selection J Virol 75, 7756–7762 41 Appa RS, Shin CG, Lee P & Chow SA (2001) Role of the nonspecific DNA-binding region and alpha helices within the core domain of retroviral integrase in selecting target DNA sites for integration J Biol Chem 276, 45848–45855 42 Marchand C, Zhang X, Pais GC, Cowansage K, Neamati N, Burke TR Jr & Pommier Y (2002) Structural determinants for HIV-1 integrase inhibition by betadiketo acids J Biol Chem 277, 12596–12603 43 Perryman AL & McCammon JA (2002) AutoDocking dinucleotides to the HIV-1 integrase core domain: exploring possible binding sites for viral and genomic DNA J Med Chem 45, 5624–5627 44 Adesokan AA, Roberts VA, Lee KW, Lins RD & Briggs JM (2004) Prediction of HIV-1 integrase ⁄ viral DNA interactions in the catalytic domain by fast molecular docking J Med Chem 47, 821–828 45 van Gent DC, Elgersma Y, Bolk MW, Vink C & Plasterk RH (1991) DNA binding properties of the integrase proteins of human immunodeficiency viruses types and Nucleic Acids Res 19, 3821–3827 46 Bao KK, Wang H, Miller JK, Erie DA, Skalka AM & Wong I (2003) Functional oligomeric state of avian sarcoma virus integrase J Biol Chem 278, 1323–1327 47 Feng H, Klutz AM & Cao W (2005) Active site plasticity of endonuclease V from Salmonella typhimurium Biochemistry 44, 675–683 48 O’Neill RJ, Vorob’eva OV, Shahbakhti H, Zmuda E, Bhagwat AS & Baldwin GS (2003) Mismatch uracil glycosylase from Escherichia coli: a general mismatch or a specific DNA glycosylase? J Biol Chem 278, 20526– 20532 FEBS Journal 273 (2006) 1137–1151 ª 2006 The Authors Journal compilation ª 2006 FEBS 1151 ... to the processed DNA product The stability of IN during the time-course of the reaction and the influence of protein aggregation are discussed Results General features of the 3¢-end processing kinetics... clear In the case of IN, tight binding of the protein to recessed viral DNA ends is obligatory in the context of the HIV replication cycle: the 3¢ -processing of viral DNA ends is only the first... Taking into account the standard duration of 3¢ -processing assays, the absence of turnover in the case of IN originates primarily in the low kcat¢ value The amount of IN? ?DNA complex, which did