Crowding Alters the Dynamics and the Length of RecA Nucleoprotein Filaments in RecA-Mediated Strand Exchange
Article first published online: 26 NOV 2013
Copyright © 2014 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
Volume 15, Issue 1, pages 80–84, January 13, 2014
How to Cite
Wu, H.-Y. and Li, H.-W. (2014), Crowding Alters the Dynamics and the Length of RecA Nucleoprotein Filaments in RecA-Mediated Strand Exchange. ChemPhysChem, 15: 80–84. doi: 10.1002/cphc.201300835
- Issue published online: 8 JAN 2014
- Article first published online: 26 NOV 2013
- Manuscript Revised: 24 OCT 2013
- Manuscript Received: 10 SEP 2013
- National Science Council of Taiwan
- molecular crowding;
- strand exchange
The cellular environment contains high concentrations of crowding macromolecules which alter the biochemical parameters which have been determined under dilute conditions. We studied how the macromolecular crowding alters the dynamics and the length of RecA nucleoprotein filaments, and, in turn, affects the RecA-mediated DNA strand exchange, using a combination of a single-molecule outgoing strand experiment and magnetic pull-down and ATPase assays. Poly(ethylene glycol) (PEG) and bovine serum albumin (BSA) are used as the model macromolecular crowding molecules. We found that the RecA nucleoprotein filament is longer and more dynamic in the PEG-crowded environment owing to its dehydration character. In contrast, the lower ATPase rate and the shorter RecA nucleoprotein filaments under BSA crowding are correlated to the greater reduction in the observed strand exchange efficiency.
The cellular environment consists of many different macromolecules, including proteins, nucleic acids, and poly-saccharides within limited space, making it much more crowded than the typical in vitro condition.1 Therefore, enzymatic processes in the cellular matrix , that is, under crowded and less hydrated conditions, are thought to be carried out with slower diffusion. Within this context, the volume-occupying and the dehydration effect should be considered when adapting biochemical parameters to physiological activities. The volume-occupying effect comes from the exclusion volume of macromolecules, and energetically, it generally favors biochemical processes with smaller intermediates. The dehydration effect originates from the low water content in the crowded environment. Synthetic polymers such as PEG molecules have been shown to modify enzyme activities by altering molecular interactions among salt/water, protein molecules, and reaction substrates.1–3 It has been suggested that long-chain polymer crowding molecules alter biochemical processes through both the volume-occupying and the dehydration effect, whereas large macromolecules exhibit mostly a volume-occupying effect.4 We were interested in how the crowded environments alter the RecA-mediated strand exchange reaction, the central step in the homologous recombination repair pathway, aiming to correlate the kinetics and the efficiency of these biochemical steps in the cellular environment. To carry out the strand exchange reaction, RecA proteins first have to nucleate and extend on the single-stranded DNA (ssDNA) to form helical nucleoprotein filaments. This nucleoprotein filament searches and invades the homologous duplex DNA to carry out the strand exchange reaction coupled with ATP hydrolysis (Figure 1).5 To identify the major governing factors on how crowding alters the RecA-mediated strand exchange reaction, we studied two different crowding molecules: BSA, a large protein with defined conformation, and the long-chain PEG polymer.
We studied two model crowding molecules mimicking the potentially crowded conditions in the cellular environment. The polyether, PEG, is a commonly used crowding molecule to model hydrophobic interactions and flexible volume-occupying effects.2 BSA is a large water-soluble protein molecule used to mimic large proteins and organelles in cells with mainly the volume-occupying effect. To investigate how these crowding molecules alter the RecA-mediated biochemical process, we used single-molecule outgoing strand experiments to monitor the individual single-turnover RecA strand exchange reactions.6 In this type of outgoing strand experiment, the bead-labeled, double-stranded (ds) DNA was anchored on a glass slide through specific linkages (Figure 2 A). RecA nucleoprotein filaments with sequences identical to the bead-labeled strand were pre-incubated in the solution and were introduced to react with the surface-bound duplex DNA. The RecA nucleoprotein filament recognizes the homologous DNA and carries out the strand exchange reaction. When the strand exchange reaction is completed, the bead-labeled strand of the original dsDNA diffuses away. The disappearance of the bead signals the completion of the strand exchange reaction, and also eliminates the potential for a second round of reactions. The outgoing efficiency thus reflects the efficiency of the RecA-mediated strand exchange reaction.
There are several factors accounting for the bead disappearance, namely, DNA-surface linkage (digoxigenin/anti-digoxigenin), DNA-bead linkage (biotin-streptavidin), dsDNA instability, and the RecA-mediated strand exchange process. Previous work attributed the majority of the bead disappearance in this experiment to the RecA-mediated strand exchange under normal dilute conditions (conventional biochemical assay conditions contain 0.1 % BSA).7 In order to address how crowding alters the RecA reactions, the reported outgoing efficiency is the difference in the bead disappearance percentage between the experiments with and without the RecA nucleoprotein filaments under different crowded concentrations, and comparisons were made in reference to the conventional dilute condition (0.1 % BSA). Compared to the 31 % of outgoing efficiency observed under dilute conditions after 60 min of reaction time, the outgoing efficiency starts to decrease with 10 % BSA (w/v), and is reduced to further 10 % with 20 % BSA under crowded conditions (Figure 2 B). In the case of PEG-crowded conditions, the outgoing efficiency shows a reduction for the first time with 5 % PEG, with 10 % PEG it drops to 35 % efficiency, and finally, with 20 % PEG it is decreased to 13. All outgoing efficiencies measured reached their equilibrium values after 30 min of the addition of the RecA filaments both under dilute (0.1 % BSA) and crowded conditions (20 % PEG and BSA) (see the Supporting Information Figure S1). This suggests that the reduced outgoing efficiency cannot be attributed to the reduced diffusion in viscous crowded environments. Previous report showed that pure homologous ssDNA and dsDNA mixtures also are capable of displacing duplex DNA strands in a highly crowded environment (>40 %).8 Control experiments using bare ssDNA alone under different crowded concentrations (up to 20 %) showed limited outgoing efficiencies, confirming that RecA proteins are responsible for the observed strand exchange efficiency in our experiments.
As the outgoing efficiencies reported here (Figure 2 B) are the differences in bead disappearance between the conditions with and without RecA nucleoprotein filaments, the possibility of crowding molecules affecting the duplex DNA stability is not illustrated. Duplex DNA destabilization is strongly altered in crowded environments and depends on the DNA length, the crowding characteristics, and the concentrations.2, 9–11 To define how the duplex destabilization is affected by the crowded conditions, we monitored the bead disappearance percentage in the absence of RecA nucleoprotein filaments (no ssDNA and no RecA) for the normal duplex DNA and the psoralen cross-linked dsDNA substrates. The difference in the bead disappearance percentage between the normal and psoralen dsDNA directly reflects the crowding effect on the dsDNA stability. Using the dilute conditions (0.1 % BSA) as a reference, the dsDNA stability did not change significantly in PEG-crowded environments (up to 20 %). This result is consistent with the trend observed by Nakano et al., who showed that the longer the dsDNA, the less the change in melting temperature, Tm, under PEG crowding conditions.11 However, the dsDNA stability decreases as the BSA crowding concentration increases (Figure 2 C). In other words, BSA crowding destabilizes the duplex DNA. As a more readily unwound dsDNA should increase the observed outgoing efficiency, these results suggest that the reduced outgoing efficiency observed in BSA-crowded environments is dominated by RecA nucleoprotein filament dynamics, instead of the destabilization of the duplex DNA.
The observed RecA-mediated strand exchange efficiency includes contributions from the RecA nucleoprotein filament assembly step and the strand exchange step. We investigated how these crowding molecules alter the dynamics of RecA nucleoprotein filaments. The dynamic and the length of RecA nucleoprotein filaments dictate the functional component of the RecA-mediated strand exchange reaction and are key targets for regulating the RecA activity.12 X-ray crystal structures showed that ATP is located in the RecA monomer-monomer interface.13 Upon the ATP hydrolysis, RecA disassembles from the 5′ end of the filament, allowing the recycling of RecA proteins to the 3′ assembling end, so the RecA nucleoprotein filaments can dynamically progress with the direction preference during the strand exchange step.5 We specifically monitored the ATPase activity to see the dynamics of RecA recycling (Figure 3 B), and monitored the RecA-bound fraction to measure the RecA filament lengths (Figure 3 C) in different crowded environments.
The ATPase assay can be used as a measure of the filament dynamics (Figure 3 A). To make the comparison easier, we used dilute reaction conditions (0.1 % BSA) as a standard, and reported the changes in xfolds in different crowded environments. Unexpectedly, the BSA and PEG crowding molecules affect the ATPase activity of RecA differently. In BSA-crowded environments, the ATPase rate first decreased to one-half, and then gradually recovered to approximately 1.3 times the standard value with higher concentrations (20 %, Figure 3 B). However, as the concentration of PEG increases, the ATPase rate first increases to approximately 3–3.5 times the standard value, and then decreases again to approximately 1.5 times the standard value at high concentrations. Under crowded conditions with 30 % PEG, the ATPase rate drops to approximately 0.5 times the standard value. This PEG-stimulated ATPase increase has been observed with a RNA helicase, and was correlated to the structural compaction.14
Next we examined the RecA bound fraction by the magnetic pull-down assay to reflect the averaged RecA filament lengths in equilibrium (Figure 3 A). Short ssDNA-labeled magnetic beads were equilibrated with RecA in the presence of ATP and crowding molecules in solution. Using a magnet, the RecA-ssDNA-beads were isolated and bound RecA were eluted and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) quantification (Figure S2). In the case of BSA crowding molecules, we used an anti-histidine stain to visualize histidine-tag RecA. The RecA-bound fraction in BSA-crowded environments are about the same as under dilute conditions (Figure 3 C). With an increasing concentration of PEG molecules, the RecA-bound fraction increases up to six times that under typical dilute conditions.
The ATPase and bound-fraction experiments indicate that the dynamics of RecA and even more the bound fraction of RecA is enhanced in the nucleoprotein filaments in PEG-crowed environments. As a polyether and flexible long polymer, PEG induces hydrophobic interactions. Binding of RecA to DNA requires proton uptake,15 and the magnesium-ATP complex is likely to lose its hydration layer upon binding to the RecA/RecA interface.13 Both dehydration steps are favored in the presence of PEG crowding which induces additional hydrophobic interactions (Figure 4 A). This explains the sixfold increase in RecA filament length observed under PEG crowding (Figure 3 C). The enhanced ATPase rates observed under approximately 7–17 % PEG conditions (Figure 3 B) could be a result of the enthalpy–entropy compensation effect of water molecules participating in the ATP hydrolysis step, as observed in myosin motors.16
In contrast, the large-sized, negative surface-charged BSA proteins impose a bulky volume-occupying effect. Considering the rigidity of the RecA nucleoprotein filaments, bulky BSA are inhibited from accessing the RecA-coated ssDNA filament, as such configuration is entropically disfavored in volume-deficient environments.1 Knowing that magnesium ions could mediate the interactions between BSA and bare ssDNA through the “like-charge attraction”,17 this BSA-ssDNA structure is likely to reduce the RecA association–dissociation kinetics, and therefore leads to the reduced ATPase rate observed. Owing to the reduced RecA filament dynamics, the outgoing strand exchange efficiency is greatly reduced under high BSA crowding conditions (>10 %), as seen in Figure 2 B.
The molecular basis of the cellular crowded environment is definitely much more complex.18–20 Measurements to define the true cellular crowded environment will be useful in connecting the modeled crowding in vitro environment with the actual conditions within the cells. The experiments described here serve as an initial attempt to define how in vitro biochemical parameters could be altered in the presence of crowding.
In the eukaryotic recombinases Rad51 and Dmc1, there exist multiple pathways in regulating the dynamics and the length of recombinase nucleoprotein filaments,21 which, in turn, leads to the enhanced strand exchange efficiency. In this study, the lower ATPase rate and the shorter RecA filaments with BSA crowding is correlated to the greater reduction in the strand exchange efficiency. The volume-occupying effect of BSA and the dehydration effect of PEG are consistent with our experimental observations on how these crowding molecules alter the dynamics and the length in RecA nucleoprotein filaments. In addition to the dynamics and the length of RecA filaments, the pairing of the RecA filaments and homologous duplex DNA also contribute to the overall observed outgoing efficiency. How crowded environments alter these biochemical steps in detail remains to be explored.
DNA, Crowding Molecules, Enzymes, and Buffer
Biotin- and digoxigenin-labeled double-stranded DNA (dsDNA) was obtained by polymerase chain reaction (PCR) using biotin- and digoxigenin-labeled primers and pBR322 as the template. In the duplex stability assays, psoralen (Sigma–Aldrich) was used to crosslink the DNA. A saturated ethanol solution of psoralen was mixed with 500 nM dig-/bio-labeled dsDNA in equal volumes. The mixture was then irradiated by 312 nm UV for 30 min. The successfully cross-linked product was purified by alkaline denaturing agarose gel. BSA was purchased from Calbiochem. PEG (average molecular weight 200 g mol−1) was purchased from Sigma–Aldrich. Pyruvate kinase (PK), phospho(enol)pyruvate (PEP), lactate dehydrogenase (LDH), reduced nicotinamide adenosine dinucleotide (NADH), and adenosine triphosphate (ATP) in the ATPase activity assays were from Sigma–Aldrich. All RecA proteins used were from New England Biolabs (NEB) and contained a histidine-tag. The RecA reaction buffer contained 20 mM N-(2-acetamido)-2-aminoethanesulfonic acid (ACES), 3 mM potassium L-glutamate, and 10 mM magnesium acetate; pH 7.5.
Single-Molecule Outgoing Experiments
Polystyrene beads (220 nm in diameter, Bang′s Lab) with custom-made streptavidin labels, were tethered to a glass slide through a dsDNA and biotin/streptavidin, digoxigenin/anti-digoxigenin linkages, as previously described.6 ssDNA molecules were pre-incubated with RecA and ATP to form the active nucleoprotein filaments. This filament was then introduced into the microscope reaction chamber. The efficiency of reaction was defined as the percentage of bead disappearance. An Olympus IX71 microscope with a mercury lamp and a camera (Dage-MTI) was used in all single-molecule imaging experiments. Only DNA tethers with qualified Brownian motion and a symmetrical distribution were included for analysis. Subsequent analysis was done using custom-written LabVIEW and Matlab programs. Individual experiments were repeated for at least 2–3 times, with the error bars representing the standard errors of the mean (S.E.M.).
40-nucleotide random-sequence oligonucleotides without label were used for the ATPase activity assays. A Shimadzu UV-1800 UV-Vis Spectrophotometer was used during the PEG experiments. For the BSA experiments, a 96-well plate fluorescence reader (BioTek SynergyTM 4) was used. All components except for the RecA were pre-mixed at room temperature. Upon addition of RecA, the reaction mixture was pipetted into a cuvette or a 96-well plate and measured immediately. The error bars represent the standard error of the slope in linear regression of the spectroscopic data.
Magnetic Pull-Down Assay
Short oligonucleotides with biotin labels (20 nucleotides, Bio Basic Inc.) were anchored on the streptavidin-coated magnetic beads (1 μm, Roche Applied Science). This ssDNA-bead complex was then incubated with RecA, ATP, and the crowding molecules. After incubation, the beads were separated from the supernatant using a magnet. The supernatant was discarded, and fresh SDS buffer with dithiothreitol (DTT) was added to denature the ssDNA-bound RecA. Denaturation was done in 5 min at 100 °C. The remaining magnetic beads were removed using a magnet. The bound RecA was subject to the 10 % SDS-PAGE gel for quantification. For the PEG experiments, Coomassie Brilliant Blue G-250 was used for staining. For the BSA experiments, His-tag in-gel stain (Invitrogen, 302 nm excitation) was used to avoid BSA interference. Each experiment was subjected to at least two gels, and the error bars are S.E.M. of bands from different gels of the same experiment.
We acknowledge the support from the National Science Council of Taiwan, and from the National Taiwan University.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.