Prototype Foamy Virus Integrase Displays Unique Biochemical Activities among Retroviral Integrases

Integrases of different retroviruses assemble as functional complexes with varying multimers of the protein. Retroviral integrases require a divalent metal cation to perform one-step transesterification catalysis. Tetrameric prototype foamy virus (PFV) intasomes assembled from purified integrase and viral DNA oligonucleotides were characterized for their activity in the presence of different cations. While most retroviral integrases are inactive in calcium, PFV intasomes appear to be uniquely capable of catalysis in calcium. The PFV intasomes also contrast with other retroviral integrases by displaying an inverse correlation of activity with increasing manganese beginning at relatively low concentrations. The intasomes were found to be significantly more active in the presence of chloride co-ions compared to acetate. While HIV-1 integrase appears to commit to a target DNA within 20 s, PFV intasomes do not commit to target DNA during their reaction lifetime. Together, these data highlight the unique biochemical activities of PFV integrase compared to other retroviral integrases.


Introduction
Retroviruses require reverse transcription and integration to complete the viral life cycle [1]. Reverse transcriptase copies the viral genomic RNA to a linear double-stranded DNA (cDNA). Integrase (IN) acts on the viral cDNA with two necessary catalytic activities: terminal cleavage and strand transfer. Terminal cleavage, also called 3 end processing, is the removal of two nucleotides from the 3 ends of nascent reverse transcripts yielding recessed 3 hydroxyls. Strand transfer is the covalent joining of these 3 hydroxyls to the host DNA. The 3 ends are joined across one major groove of host DNA with 4-6 base pairs of intervening host sequence. The product of IN activities in vivo is the integrated provirus flanked by 4-6 base gaps of the host sequence and 5 dinucleotide flaps of the viral sequence. The integration intermediate is likely repaired by host DNA repair enzymes [2]. Retroviral IN may also catalyze intramolecular strand transfer, termed autointegration. This phenomenon effectively ends a cellular infection and may also appear during biochemical assays. Half site integration (HSI) products where a single viral DNA is joined to a target DNA have also been observed in vitro. It is unclear if these aberrant integration products occur under normal conditions in vivo, although observed altered integration sites suggest they may occur with mutant IN or in the presence of suboptimal concentrations of clinically relevant IN strand transfer inhibitors [3,4]. Retroviral INs catalyze 3 processing and strand transfer by single step transesterification chemistry [5][6][7]. Each step may be independently assayed with DNA oligomers mimicking blunt viral cDNA ends (vDNA) to test 3 processing or preprocessed vDNA to test strand transfer. These enzymes do not require an energetic co-factor such as adenosine In spite of structural similarities among intasomes of different retroviruses, these enzymatic properties distinguish PFV IN from other retroviral INs.

Expression and Purification of PFV IN
PFV IN was purified as previously described [27,28]. Briefly, hexahistidine tagged PFV IN was induced in E. coli strain BL21(DE3) pLysS with 250 µM IPTG at 25 • C for 4 h. Soluble cellular lysate was fractionated by nickel affinity chromatography. Fractions with PFV IN were treated with HRV 3C protease to remove the hexahistidine tag. PFV IN was further purified by heparin affinity chromatography. Fractions containing highly concentrated PFV IN were combined, dialyzed against 50 mM Tris HCl, pH 7.5, 500 mM NaCl, 5 mM DTT, and 10% glycerol, aliquoted, snap frozen with liquid nitrogen, and stored at −80 • C.

Assembly and Purification of PFV Intasomes
Intasomes were assembled as previously described [23]. Briefly, 50 mM Bis-tris propane, pH 7.5, 500 mM NaCl, 120 µM PFV IN, and 50 µM vDNA were combined in a total volume of 150 µL and dialyzed overnight at 18 • C against 20 mM Bis-tris propane, pH 7.5, 200 mM NaCl, 2 mM DTT, and 25 µM ZnCl 2 . The intasome aggregates were solubilized by increasing the concentration of NaCl from 200 to 320 mM and incubating on ice. The intasomes were purified by size exclusion chromatography using a Superose 12 10/300 (GE Healthcare, Chicago, IL, USA) equilibrated with 20 mM Bis-tris propane, pH 7.5, 320 mM NaCl, and 10% glycerol. Fractions containing intasomes were aliquoted, snap frozen with liquid nitrogen, and stored at −80 • C. PFV intasomes appear to retain activity for one year at −80 • C.

Integration Assays
Standard integration reactions were performed in 30 mM Bis-tris propane, pH 7.5, 110 mM NaCl, 5 mM MgSO 4 , 4 µM ZnCl 2 , 10 mM DTT, indicated concentration of PFV intasomes, and 1.8 nM supercoiled plasmid target DNA in a final volume of 15 µL. Reactions were incubated for 5 min at 37 • C, stopped with the addition of 0.5% SDS, 0.5 mg/mL proteinase K, 25 mM EDTA (pH 8.0), and incubated for 1 h at 37 • C. Integration products were resolved with 1.25% agarose gel electrophoresis. Gels were stained with ethidium bromide and imaged for ethidium bromide and Cy5 fluorescence (Sapphire Biomolecular Imager, Azure Biosystems, Dublin, CA, USA). Cy5 fluorescence was quantified using AzureSpot gel analysis software (Azure Biosystems, Dublin, CA, USA). Where indicated, 5 mM PCA was included in reactions. Acetate buffer included 25 mM Tris-HCl, pH 7.4, 125 mM NaOAc, 5 mM MgOAc, 10 µM ZnCl 2 , 1 mM DTT. In assays comparing buffers, 20 nM PFV intasome was added to reaction buffer and incubated on ice for indicated times before addition of 1.8 nM 3 kb supercoiled pGEMT plasmid. At the addition of target DNA, reactions were incubated at 37 • C for 5 min and analyzed as standard integration reactions. p values were determined by two tail paired t test. Error bars indicate standard deviation between at least three independent experiments performed with at least two independent intasome preparations.

PFV Intasome Requirements for Divalent Cations
Retroviral intasomes may be assembled with a preprocessed vDNA and purified recombinant PFV IN [22]. The vDNA may be labeled with a fluorophore, such as Cy5, to identify integration products by fluorescence imaging [23]. The assembled complexes are purified by size exclusion chromatography. To assay integration activity, the PFV intasomes are diluted in a buffer containing a divalent cation and target DNA. Addition of supercoiled plasmid as the target DNA allows facile visualization and quantitation of the integration products. Concerted integration (CI 1 ) of both vDNAs to a circular plasmid yields a linear product with vDNA at the ends ( Figure 1). Additional CI (CI 2 ) to the linear product results in fragments shorter than the linear product. Half site integration (HSI) is the joining of a single vDNA to the plasmid and results in a fluorescently tagged plasmid with the mobility of a relaxed circle. Autointegration (AI) products result from integration of one vDNA to another vDNA and have slightly slower mobility compared to unreacted vDNA. These reaction products are resolved by agarose gel electrophoresis. The agarose gel is imaged for ethidium bromide and Cy5 fluorescence and quantified. Integration products are distinguished by their mobilities and presence of the fluorophore. es 2022, 12, x FOR PEER REVIEW 4 of 14 integration products. Concerted integration (CI1) of both vDNAs to a circular plasmid yields a linear product with vDNA at the ends ( Figure 1). Additional CI (CI2) to the linear product results in fragments shorter than the linear product. Half site integration (HSI) is the joining of a single vDNA to the plasmid and results in a fluorescently tagged plasmid with the mobility of a relaxed circle. Autointegration (AI) products result from integration of one vDNA to another vDNA and have slightly slower mobility compared to unreacted vDNA. These reaction products are resolved by agarose gel electrophoresis. The agarose gel is imaged for ethidium bromide and Cy5 fluorescence and quantified. Integration products are distinguished by their mobilities and presence of the fluorophore. PFV intasomes with preprocessed vDNA were assembled without a divalent metal ion. The intasomes were purified by size exclusion chromatography in the continued absence of a divalent metal ion, indicating that the complexes are stable without cations. The use of a preprocessed vDNA allows analysis of strand transfer activity. PFV integration assays are commonly performed in the presence of magnesium sulfate [25,29]. PFV intasome integration to a supercoiled plasmid was performed with a titration of magnesium sulfate (Figure 2A). In the absence of cation, PFV intasomes are unable to perform strand transfer. Maximal concerted integration activity was observed in the presence of 5 mM magnesium sulfate. Similar results were seen with a titration of magnesium chloride (Figure 2B). There was no statistically significant difference between PFV integration activity in the presence of magnesium sulfate or magnesium chloride (p > 0.05 at all equivalent concentrations). PFV intasomes with preprocessed vDNA were assembled without a divalent metal ion. The intasomes were purified by size exclusion chromatography in the continued absence of a divalent metal ion, indicating that the complexes are stable without cations. The use of a preprocessed vDNA allows analysis of strand transfer activity. PFV integration assays are commonly performed in the presence of magnesium sulfate [25,29]. PFV intasome integration to a supercoiled plasmid was performed with a titration of magnesium sulfate ( Figure 2A). In the absence of cation, PFV intasomes are unable to perform strand transfer. Maximal concerted integration activity was observed in the presence of 5 mM magnesium sulfate. Similar results were seen with a titration of magnesium chloride ( Figure 2B). There was no statistically significant difference between PFV integration activity in the presence of magnesium sulfate or magnesium chloride (p > 0.05 at all equivalent concentrations).  PFV intasomes were also assayed for integration activity in manganese chloride an calcium chloride. Intasomes were most active in the presence of 1 mM manganese chlori ( Figure 3). In contrast to magnesium, increasing the concentration of manganese chlori led to decreased integration activity. Previous studies of HIV-1 IN also revealed an inh bition of strand transfer, but only at 64 mM manganese chloride [30]. PFV intasomes we also assayed with a titration of calcium chloride ( Figure 4). Many retroviral integrases a not active in the presence of calcium. However, PFV intasomes were active in the presen of calcium chloride and activity increased with increasing concentrations of calcium chl ride. In the presence of this divalent metal cation, integration products were mostly H Employing a higher concentration of PFV intasomes revealed the formation of CI produc in the presence of Ca. These data highlight the unique characteristics of PFV IN divale metal requirements compared to other retroviral INs. PFV intasomes were also assayed for integration activity in manganese chloride and calcium chloride. Intasomes were most active in the presence of 1 mM manganese chloride ( Figure 3). In contrast to magnesium, increasing the concentration of manganese chloride led to decreased integration activity. Previous studies of HIV-1 IN also revealed an inhibition of strand transfer, but only at 64 mM manganese chloride [30]. PFV intasomes were also assayed with a titration of calcium chloride ( Figure 4). Many retroviral integrases are not active in the presence of calcium. However, PFV intasomes were active in the presence of calcium chloride and activity increased with increasing concentrations of calcium chloride. In the presence of this divalent metal cation, integration products were mostly HSI. Employing a higher concentration of PFV intasomes revealed the formation of CI products in the presence of Ca. These data highlight the unique characteristics of PFV IN divalent metal requirements compared to other retroviral INs.

PFV Intasome Mediated Integration Is Quick
PFV intasomes were previously shown to complete integration within 5 min at 37 • C with either a supercoiled plasmid target or a nucleosome target [31,32]. However, there was only one earlier time point assayed in those studies. For a more accurate determination of intasome kinetics, we extended the time course with the same buffer conditions and found that integration is complete within 2 min at 37 • C ( Figure 4).

PFV Intasome Mediated Integration Is Quick
PFV intasomes were previously shown to complete integration within 5 min at 37 °C with either a supercoiled plasmid target or a nucleosome target [31,32]. However, there was only one earlier time point assayed in those studies. For a more accurate determination of intasome kinetics, we extended the time course with the same buffer conditions and found that integration is complete within 2 min at 37 °C ( Figure 4).
The PFV intasome footprint appears to be ~10 bp of target DNA [33]. A 3000 bp plasmid target offers 300 possible integration sites. Integration assays were performed with a molar excess of target sites to PFV intasomes ( Figure 4) [31,34]. Reactions with 25 nM PFV intasomes are at a 20-fold molar excess of target sites to intasomes. The intasome concentration was reduced to 2.5 nM in order to evaluate a 200-fold molar excess of target sites. The 200-fold molar excess reactions yielded less CI products, particularly CI2 products, compared to 20-fold molar excess of target sites. All reactions were complete by 2 min. The amount of target plasmid included in the reactions was increased to obtain a 1000fold molar excess of target sites to 2.5 nM intasomes. The data with altered molar ratios of target substrate to intasomes consistently show the reaction to be complete at 2 min. Neither the substrates nor the intasomes were completely consumed during these reactions.
The short reaction time of PFV intasome-mediated integration makes classic Michaelis-Menten analysis of enzyme kinetics difficult. Previous studies showed that PFV intasomes incubated at 37 °C for 5 min before the addition of target lost integration activity [31]. However, addition of the small molecule PCA significantly rescued the activity of the preincubated PFV intasomes. This was attributed, at least in part, to the ability of PCA to prevent aggregation of PFV intasomes at physiologically relevant ionic strength buffer conditions [31]. PCA was added to PFV intasomes to test the ability of this small molecule to lengthen the time that intasomes are active beyond 2 min ( Figure 5). Two PFV intasome concentrations were assayed in the presence or absence of PCA. The presence of PCA had The PFV intasome footprint appears to be~10 bp of target DNA [33]. A 3000 bp plasmid target offers 300 possible integration sites. Integration assays were performed with a molar excess of target sites to PFV intasomes ( Figure 4) [31,34]. Reactions with 25 nM PFV intasomes are at a 20-fold molar excess of target sites to intasomes. The intasome concentration was reduced to 2.5 nM in order to evaluate a 200-fold molar excess of target sites. The 200-fold molar excess reactions yielded less CI products, particularly CI 2 products, compared to 20-fold molar excess of target sites. All reactions were complete by 2 min. The amount of target plasmid included in the reactions was increased to obtain a 1000-fold molar excess of target sites to 2.5 nM intasomes. The data with altered molar ratios of target substrate to intasomes consistently show the reaction to be complete at 2 min. Neither the substrates nor the intasomes were completely consumed during these reactions.
The short reaction time of PFV intasome-mediated integration makes classic Michaelis-Menten analysis of enzyme kinetics difficult. Previous studies showed that PFV intasomes incubated at 37 • C for 5 min before the addition of target lost integration activity [31]. However, addition of the small molecule PCA significantly rescued the activity of the preincubated PFV intasomes. This was attributed, at least in part, to the ability of PCA to prevent aggregation of PFV intasomes at physiologically relevant ionic strength buffer conditions [31]. PCA was added to PFV intasomes to test the ability of this small molecule to lengthen the time that intasomes are active beyond 2 min ( Figure 5). Two PFV intasome concentrations were assayed in the presence or absence of PCA. The presence of PCA had no effect on the accumulation of CI products over time and reaction products saturated by 2 min. These data suggest that while PCA prevents the aggregation of PFV intasomes, the completion of integration at 2 min was not due to aggregation of complexes.
Biomolecules 2022, 12, x FOR PEER REVIEW 8 of 14 no effect on the accumulation of CI products over time and reaction products saturated by 2 min. These data suggest that while PCA prevents the aggregation of PFV intasomes, the completion of integration at 2 min was not due to aggregation of complexes.

PFV Intasomes Are Less Active in the Presence of Acetate Buffer
Studies of PFV intasome activity are routinely performed in the presence of NaCl and MgCl2 or MgSO4 [25,29,31,[34][35][36][37][38][39][40][41]. PFV IN strand transfer showed no difference in the presence of MgSO4 or MgCl2 indicating no significant effects of sulfite or chloride co-ions ( Figure 2) [25]. A recent study employed sodium acetate (NaOAc) and magnesium acetate (MgOAc) during assays of PFV intasome activities [42]. This led to results contradicting previously published results obtained with buffer containing NaCl and MgSO4 [34]. While Jones et al. were able to measure the time between PFV concerted strand transfer events by magnetic tweezers in the presence of NaCl and MgSO4, Vanderlinden et al. were only able to observe HSI in the presence of NaOAc and MgOAc. To address the difference in experimental outcomes, PFV intasome activity was directly compared in the presence of a buffer with NaOAc and MgOAc to a buffer containing NaCl and MgSO4.
PFV intasome integration activity was measured over time in two different buffer conditions ( Figure 6A). PFV intasomes were active in both buffers generating readily visualized CI products. However, the intasomes displayed more CI in buffer with NaCl and MgSO4. The relative increase in CI product accumulation in the presence of NaCl and MgSO4 is apparent at 0.5 min. In addition, accumulation of CI products plateaus in the presence of NaCl and MgSO4 at 2 min incubation. However, CI is complete at 1 min in the acetate buffer. These data indicate that PFV intasomes are less active in acetate buffers.

PFV Intasomes Are Less Active in the Presence of Acetate Buffer
Studies of PFV intasome activity are routinely performed in the presence of NaCl and MgCl 2 or MgSO 4 [25,29,31,[34][35][36][37][38][39][40][41]. PFV IN strand transfer showed no difference in the presence of MgSO 4 or MgCl 2 indicating no significant effects of sulfite or chloride co-ions ( Figure 2) [25]. A recent study employed sodium acetate (NaOAc) and magnesium acetate (MgOAc) during assays of PFV intasome activities [42]. This led to results contradicting previously published results obtained with buffer containing NaCl and MgSO 4 [34]. While Jones et al. were able to measure the time between PFV concerted strand transfer events by magnetic tweezers in the presence of NaCl and MgSO 4 , Vanderlinden et al. were only able to observe HSI in the presence of NaOAc and MgOAc. To address the difference in experimental outcomes, PFV intasome activity was directly compared in the presence of a buffer with NaOAc and MgOAc to a buffer containing NaCl and MgSO 4 .
PFV intasome integration activity was measured over time in two different buffer conditions ( Figure 6A). PFV intasomes were active in both buffers generating readily visualized CI products. However, the intasomes displayed more CI in buffer with NaCl and MgSO 4 . The relative increase in CI product accumulation in the presence of NaCl and MgSO 4 is apparent at 0.5 min. In addition, accumulation of CI products plateaus in the presence of NaCl and MgSO 4 at 2 min incubation. However, CI is complete at 1 min in the acetate buffer. These data indicate that PFV intasomes are less active in acetate buffers.
The quick reaction kinetics of PFV intasomes suggested that practical considerations may also play a role in integration assays. In the case of single-molecule magnetic tweezer assays, PFV intasomes are typically diluted in reaction buffer before loading to a flow cell. In some cases, the diluted intasomes may remain on ice before being exposed to target DNA within the flow cell. The two different reaction buffers were tested for their effects on intasomes incubated on ice ( Figure 6B). PFV intasomes were diluted to a working concentration in reaction buffer and incubated on ice for a variable time. Following variable incubation time on ice, target DNA was added and the reactions were immediately transferred to 37 • C for 5 min. PFV intasome incubation on ice led to reduced activity over time in both buffers. However, the intasomes in acetate buffer lost 50% of their integration activity after 30 min and 55% after 60 min on ice. Intasomes in buffer with NaCl and MgSO 4 lost 37% of their activity after 30 min and 42% after 60 min of incubation on ice. These data indicate that PFV intasomes are more prone to loss of activity in acetate buffer compared to buffer with NaCl and MgSO 4 .

PFV Intasomes Do Not Commit to Target DNA
HIV-1 IN was previously shown to quickly commit to a target DNA [26]. In these assays, HIV-1 IN and vDNA were added to a plasmid and incubated at 37 • C. At variable times a second plasmid of differing size was added to the reaction. HIV-1 IN performed equivalent integration to both plasmids when they were added simultaneously. However, HIV-1 IN appeared to fully commit to the first plasmid within 20 s with no integration to a second plasmid. In contrast, similar experiments showed that PFV IN integrated to a second plasmid up to 60 min after the addition of the first plasmid [25]. Both of these experimental approaches employed free IN and vDNA rather than assembled intasomes.
We tested the commitment of PFV intasomes to two plasmid DNA targets, 3 kb and 6 kb (Figure 7). Reactions were performed with 2.5 nM PFV intasomes to reduce the amount of CI 2 products which would confound quantitation. When the plasmids were added to the reaction simultaneously, integration to the plasmids was equivalent. As seen with a single plasmid target, integration was complete by 2 min. At times shorter than 2 min, PFV intasomes integrated to either plasmid. Over time, the fraction of integration to the first plasmid increased and integration to the second plasmid decreased. For the entire time that PFV intasomes are active, they are able to integrate to either plasmid. The integration dynamics were unaffected by whether the smaller or larger plasmid was added first. These results suggest that PFV intasomes do not fully commit to a target DNA early, as seen with free HIV-1 IN.
We tested the commitment of PFV intasomes to two plasmid DNA targets, 3 k 6 kb (Figure 7). Reactions were performed with 2.5 nM PFV intasomes to reduc amount of CI2 products which would confound quantitation. When the plasmids added to the reaction simultaneously, integration to the plasmids was equivalent. As with a single plasmid target, integration was complete by 2 min. At times shorter t min, PFV intasomes integrated to either plasmid. Over time, the fraction of integrat the first plasmid increased and integration to the second plasmid decreased. For the time that PFV intasomes are active, they are able to integrate to either plasmid. The gration dynamics were unaffected by whether the smaller or larger plasmid was a first. These results suggest that PFV intasomes do not fully commit to a target DNA as seen with free HIV-1 IN.

Discussion
Many enzymes that utilize two divalent metal cations for catalysis display activ Mg or Mn, but not Ca. It has been noted that obvious factors such as the different si the ions or coordination number do not correlate well with the observed differen enzymatic activity [14]. The mechanism of cation preference is not clear, but sever potheses have been proposed. The reaction energy barrier in the presence of Ca ap to be raised for both RNaseH and BamHI [9,14]. Another factor may be the larger a radius of Ca compared to Mg [10]. Observations of RNaseH suggested that the arc ture of the active site was the same in the presence of Ca or Mg, but changes in ion dination geometry were observed [9]. This may be a key factor considering the stru

Discussion
Many enzymes that utilize two divalent metal cations for catalysis display activity in Mg or Mn, but not Ca. It has been noted that obvious factors such as the different sizes of the ions or coordination number do not correlate well with the observed differences of enzymatic activity [14]. The mechanism of cation preference is not clear, but several hypotheses have been proposed. The reaction energy barrier in the presence of Ca appears to be raised for both RNaseH and BamHI [9,14]. Another factor may be the larger atomic radius of Ca compared to Mg [10]. Observations of RNaseH suggested that the architecture of the active site was the same in the presence of Ca or Mg, but changes in ion coordination geometry were observed [9]. This may be a key factor considering the structural similarities of retroviral intasomes at the active site in what has been termed a conserved intasome core (CIC) [43]. Structural studies of PFV intasomes in the presence of Mg or Mn described the octahedral coordination for both binding sites, but Ca was not included in the analysis [44]. Instead of a single difference between cations defining catalytic function, a combination of features may account for the ability-or not-of an enzyme to function in the presence of Ca [15].
It is unclear why PFV IN, of all retroviral INs, appears uniquely able to utilize Ca in strand transfer catalysis. HIV-1 IN has been the most widely studied for its activity in the presence of various divalent cations [45]. Divalent metals were shown to be required for assembly of the integration complex, 3 end processing, and strand transfer. HIV-1 IN was shown to efficiently use Mg or Mn for all three activities [19,30]. However, HIV-1 IN could only use Ca for assembly, not 3 processing or strand transfer [19]. It could use Co for strand transfer, but not assembly or 3 processing [19]. Similarly, ASV IN performs strand transfer in the presence of Mg or Mn, with better efficiency in the latter [24]. HIV-1 preintegration complexes (PICs) derived from cells and recombinant MMTV intasomes are inactive in the presence of Ca [46][47][48]. MMTV intasomes display greater strand transfer activity in Mn compared to Mg at concentrations ≤10 mM, but also perform more AI in Mn at these concentrations, suggesting a lack of specificity [48]. PFV intasomes do not require any divalent cation for assembly [22,33]. By comparing integration with blunt or preprocessed vDNA, PFV IN was shown to use Mg or Mn for 3 processing, but not Ca [25]. These experiments also revealed that PFV IN can use Ca for strand transfer ( Figure 5) [25]. In addition to the difference in ability to perform strand transfer in the presence of Ca, PFV intasomes also differed from other retroviral INs and intasomes by their activity in Mn. Retroviral INs have been reported to have equal activity in Mg and Mn or better activity in Mn [19,24,30]. HIV-1 IN was shown to be inhibited at >32 mM Mn [30]. MMTV intasome strand transfer activity correlated with increased Mn concentrations up to 15 mM [48]. In contrast, PFV intasome strand transfer activity displays an inverse correlation with Mn concentration ( Figure 5).
Multiple transposases have also been characterized for their ability to utilize different divalent cations. Similar to INs, transposases assemble complexes and perform the same single-step transesterification reactions. Phage Mu transposase MuA forms a tetramer and can also perform strand transfer in the presence of Ca [11]. MuA must cleave the Mu DNA ends to form a cleaved donor complex, but is not able to utilize Ca for this activity [11]. This is not true for all transposases, as Tn10 transposase has no activity in the presence of Ca [12]. Thus, PFV IN activities in the presence of Ca are more similar to MuA transposase than to other retroviral Ins or transposases.
In addition to divalent metal cations, retroviral INs require sodium (Na) or potassium (K) cations. PFV intasomes were previously characterized for their strand transfer activity in a titration of NaCl [32]. Analysis of integration to alternative target DNAs revealed maximal strand transfer activity at variable NaCl concentrations. Integration to a supercoiled plasmid was maximal at 200 mM NaCl and to a mononucleosome at 150 mM NaCl [32]. The kinetics of PFV or MMTV intasome integration to plasmid or nucleosome DNA were shown to be the same [32,48]. These data suggest that the observations of intasome integration to plasmid DNA will be indicative of integration to more physiologically relevant nucleosome targets.
The use of co-ion by retroviral INs has not been thoroughly studied. The divalent metal cation is necessary for IN catalysis, but the role of the co-ion during a reaction is unclear. Interestingly, the presence of OAc co-ions results in less strand transfer activity of PFV intasomes compared to Cl co-ions. The OAc co-ion also led to earlier saturation of reaction products. These results suggest that PFV intasomes are more active in the presence of Cl co-ions. Future studies of retroviral intasomes should consider co-ion effects on enzymatic activity.
PFV IN also appears to be distinguishable from HIV-1 IN by a lack of commitment to a target DNA. HIV-1 IN and gel-filtered HIV-1 PICs derived from infected cells were previously shown to commit to a target DNA within 20 s [26]. Interestingly, addition of a cellular extract to gel-filtered HIV-1 PICs allowed integration to a second target [26]. It is unknown what cellular factor might allow HIV-1 PICs to display reduced target commitment. It is possible that the known host co-factor for HIV-1 integration lens epithelium-derived growth factor 75 kD splice variant (LEDGF)/p75 was removed from PICs during gel filtration. In stark contrast to the early target commitment of HIV-1 IN, PFV IN was previously shown to efficiently integrate to a second target DNA during a 60 min reaction [25]. Here we observed that PFV intasomes are also able to integrate to a second target DNA at any time while they are active (Figure 7). It should be noted that PFV intasomes do not have a known host co-factor. The lack of PFV IN or PFV intasome early target commitment could be due to 3D searching of target DNAs or relatively slow binding followed by fast reaction kinetics. These experiments did not distinguish these two models. A combination of 1D and 3D searching of DNA can enhance the ability of an enzyme to identify a target site [49]. Regardless of the mechanism, the PFV IN promiscuous interaction with target DNA is demonstrably different from the observed HIV-1 IN or gel-filtered HIV-1 PIC fast commitment to target DNA. HIV-1 PICs may require a host co-factor to allow 3D searching and a more efficient search for a preferred integration site.
These results highlight the unique biochemical characteristics of PFV IN compared to other retroviral INs. All retroviruses display unique patterns of integration to the genome during infection [50]. PFV displays minimal preferences for genomic elements, such as transcription start sites and cytosine guanine (CpG) islands [51]. Most retroviral integration is directed by a host integration co-factor, such as LEDGF/p75 for HIV-1 integration [52,53]. However, PFV integration appears to be targeted by the viral Gag protein [51]. Thus, PFV integration in cells differs from other retroviruses by the use of a viral protein to tether the integration complex to chromatin, rather than a host protein. Indeed, the foamy viruses are non-pathogenic and exhibit a life cycle unique among retroviruses [54]. These results extend those observations by showing that PFV intasomes also display unique biochemical characteristics. Our data suggests that retroviral intasomes display variable activity depending on the buffer conditions, which may not be predictable despite the structural similarities at the IN active site observed in multiple retroviral intasome structures [20][21][22]33,[55][56][57]. Further biochemical studies of additional retroviral intasomes will reveal their unique properties.