|Home | About | Journals | Submit | Contact Us | Français|
Helicase are essential enzymes which are widespread in all life-forms. Due to their central role in nucleic acid metabolism, they are emerging as important targets for anti-viral, antibacterial and anti-cancer drugs. The development of easy, cheap, fast and robust biochemical assays to measure helicase activity, overcoming the limitations of the current methods, is a pre-requisite for the discovery of helicase inhibitors through high-throughput screenings. We have developed a method which exploits the optical properties of DNA-conjugated gold nanoparticles (AuNP) and meets the required criteria. The method was tested with the catalytic domain of the human RecQ4 helicase and compared with a conventional FRET-based assay. The AuNP-based assay produced similar results but is simpler, more robust and cheaper than FRET. Therefore, our nanotechnology-based platform shows the potential to provide a useful alternative to the existing conventional methods for following helicase activity and to screen small-molecule libraries as potential helicase inhibitors.
Helicases are ubiquitous enzymes, found in viruses, bacterial, archaeal and eukaryotic cells. They act as motor proteins to separate or remodel DNA or RNA substrates, using ATP as energy source; their activity is essential in nucleic acid metabolism, playing a key role in a variety of processes, such as DNA replication, repair, recombination, translation, RNA transport, etc. Not surprisingly, given their essential tasks in living organisms, they are emerging as an important class of targets for antiviral, antibiotic and anti-cancer drugs1,2. As an example of the latter, helicases (and in particular RecQ helicases) are essential to overcome the effect of chemotherapeutic drugs that damage DNA, making them attractive targets for inhibitors to make cancer cells more sensitive to chemotherapy3.
The search for specific inhibitors of helicases strongly relies on the development of easy, cheap, fast, reproducible biochemical assays, suitable for high-throughput (HT) screening. Currently used methods include both ATPase and helicase assays. ATP hydrolysis methods rely on the colorimetric detection of the phosphate4,5 or the detection of ADP through antibody-based ADP sensors or ADP-coupled reactions6. Nucleic acid unwinding assays are typically based on radioactive or fluorescent substrates7. Colorimetric ATPase assays are generally simpler, less expensive and more suited for HT screenings, but can miss inhibitors that abolish DNA/RNA unwinding without affecting ATP hydrolysis. On the other hand, strand separation assays are more complex and tend to require expensive and sophisticated reagents, such as labelled oligonucleotides8. Among them, the ones that can more easily be adapted for HT screens are those based on fluorescent resonance energy transfer (FRET), but are subjected to some drawbacks, including high costs, poor stability of the substrate and compound interference2,9,10,11. The latter is particularly serious, as compound libraries include many molecules that absorb or emit light at wavelengths that overlap with those of the fluorophores being monitored2. There is therefore the need to explore alternative methods to measure helicase activity.
Here we report a novel method for the measurement of helicase activity, based on the properties of functionalized gold nanoparticles (AuNPs). The ease of synthesis/functionalization and unique optical properties of AuNPs allow for their use in the development of newer diagnostic methods which are safer and easier than the conventional existing methods. A number of reports exist on the use of AuNP based structures to detect and measure the activities of many enzymes belonging to various classes such as hydrolases, transferases, oxidoreductases etc.12,13,14. Colorimetric techniques either follow the aggregation of AuNPs or the disintegration of AuNP aggregates, in response to the enzyme activity. In the present work, we have constructed nanoparticles conjugated to a specifically-designed DNA substrate, so as to monitor the DNA unwinding properties of a helicase. The current set-up is designed for a helicase with a 3′–5′ directionality, but it can easily be adapted for 5′–3′ helicases.
As a test case, we used the human RecQ4 helicase. RecQ helicases are ubiquitous nucleic acid unwinding enzymes, playing a vital role in maintaining genomic stability by acting at the interface of replication, repair and recombination. They are involved in DNA repair, homologous recombination, telomere maintenance, mitochondrial genome maintenance and DNA replication15,16,17. Three out of five human RecQ helicases are associated to genetic disorders, characterized by genomic instability, premature aging and predisposition to cancer18. Mutations in RecQ4 are associated to Rothmund-Thomson Syndrome (RTS), RAPADILINO and Baller-Gerold Syndrome19,20. Although RecQ4 deficient RTS patients have an elevated risk of developing osteosarcoma, overexpression of RecQ4 has been reported in human osteosarcoma, prostate and breast tumour samples21,22. The human RecQ4 helicase consists of 1208 amino-acids, including a conserved helicase core23,24. In most RecQ helicases the catalytic core is followed by a RecQ-C-terminal (RQC) domain, that has been proposed to have a crucial role in the helicase activity, by providing an aromatic residue acting as essential “pin” that physically disrupts the dsDNA base-pairing25,26; a bioinformatic analysis recently suggested the presence of a non-canonical RQC domain in RecQ427. Despite the role of RecQ4 in genetic disorders and carcinogenesis, not much information is available about its mechanism of action.
By using a novel AuNP helicase assay method, the catalytic core of human RecQ4 has been characterized and the results compare favourably with those obtained with standard FRET-based methods. A number of site-directed mutants have been analysed, confirming the importance of key residues in the putative RQC domain for the helicase activity.
To obtain the substrate for monitoring helicase activity, we prepared DNA-functionalized AuNPs which aggregate due to the annealing of the complementary DNA strands28. The assay relies on the disruption of the aggregates and the consequent solubilization of the particles upon DNA unwinding. The course of the reaction can be simply followed by monitoring the surface plasmon resonance (SPR) absorption spectra of AuNPs using a spectrophotometer. Surface plasmon is the collective oscillations of the free electrons on the surface of nanoparticles which is in resonance with the frequency of the visible light interacting with them. For 20nm sized monodispersed AuNPs, the SPR phenomenon causes the absorption of light at about 524nm29. The absorption peak shifts to longer wavelengths as the AuNPs aggregate. This shift in the absorbance can be used to monitor the change in the aggregation state of the AuNPs, due to the DNA unwinding catalysed by a helicase.
The protocol for preparing the substrate has been previously described30. Briefly, we used 20nm sized AuNPs functionalized with a mixed self-assembled monolayer (SAM) of DNA and (1-mercaptoundec-11-yl)hexa(ethylene glycol) referred to as TOEG6; (DNA/TOEG6@AuNP). The function of TOEG6 in the SAM is to provide salt-stability and specificity to the DNA conjugated AuNPs. The preparation of the substrate for the enzymatic assay was achieved by mixing two batches of DNA/TOEG6@AuNP (with DNA1 and DNA2 respectively, Table 1) in the presence of a duplex which has DNA strands partially complementary to the DNA strands on each batch of DNA/TOEG6@AuNP. The formation of DNA/TOEG6@AuNP based aggregates is schematically shown in Fig. 1a.
Melting experiments with the assembled substrates were carried out to confirm that the aggregate formation was based on DNA-DNA recognition and not due to non-specific interactions. For this, the aggregate was subjected to heating with a linear increase in temperature from 22 to 87°C while recording the spectrum of the solution, between 200nm and 800nm in a UV-vis spectrophotometer. The absorbance was measured at 524nm (corresponding to the DNA/TOEG6@AuNP) and at 260nm (corresponding to the DNA), and plotted against the temperature (Fig. 1b), showing a sharp transition in the absorbance at about 60°C. This is in agreement with previous studies on the melting patterns of DNA@AuNP based aggregates31. It was observed that the melting curves of DNA tagged with AuNPs are sharper than that of DNA alone. As expected, the absorbance variation at 260 and 524nm followed the same pattern, since heating leads to the disintegration of the aggregates, increasing the amount of DNA as well as separated nanoparticles in the solution. A more detailed analysis on the formation of the aggregate based on DNA recognition and annealing were reported in our previous work30. All studies supported the fact that the formation of the aggregate was guided by the DNA binding specificity and that it could serve as the substrate to follow DNA unwinding by a helicase.
As RecQ4 includes a long, partially unstructured N-terminal domain, which is apparently not involved in the helicase activity32, the enzyme used in the assay corresponds to a deletion mutant, which includes the helicase core and the putative RQC domain of the human RecQ4 helicase (Fig. 2). The protein was expressed and purified to homogeneity in bacterial cells33.
To monitor the RecQ4-catalysed enzymatic reaction, helicase assay buffer was added to the AuNP aggregate (hereafter referred to as substrate) that was then transferred to a Perkin Elmer UV-vis spectrophotometer cuvette, followed by the addition of the enzyme and ATP. The addition of the enzyme in the presence of ATP immediately causes a dramatic colour change: the solution turns into a purple suspension, followed by a slower colour variation towards red (Fig. 3a); this is caused by the disruption of the DNA-linked aggregates and the consequent solubilisation of the nanoparticles. To quantitatively monitor these events, the time dependent UV-vis spectra of the solution were recorded. The time dependent UV-vis graphs for the experiment, with an enzyme concentration of 60nM, are shown in Fig. 3. It can be observed that with increasing time, the absorbance spectra of the reaction showed an increase in the maximum absorbance at 524nm and a decrease at 650nm. This is due to RecQ4 unwinding the double-stranded DNA, breaking the substrate into smaller parts and consequently releasing the DNA/TOEG6@AuNPs in the solution, which absorbs in the visible region of the electromagnetic spectrum. The shoulder at 650nm corresponds to the assembled state of the nanoparticles and indeed decreases as the 524nm peak (corresponding to the monodispersed particles in solution) increases.
To derive quantitative data from these changes in the SPR spectra, we considered the ratio A@524nm:A@650nm at each time interval and converted it into the percentage of unwinding using the following relation:
In the above relation, the ratio for 100% unwound state corresponds to the value for the substrate DNA/TOEG6@AuNP boiled at 95°C for 10minutes (red curve in Fig. 3b).
The analyses were performed for five different concentrations of the enzyme, corresponding to 15nM, 30nM, 60nM, 100nM and 150nM respectively. The percentage of unwinding against the respective time interval for various concentrations of the enzyme is shown in Fig. 4a. As a control we used a protein mutant (WA) where the essential lysine 508 (the key residue within the Walker A motif, involved in ATP binding) has been mutated to an alanine.
As evident from the data, the extent of unwinding increases with time and is dependent on the concentration of the enzyme. The plots corresponding to higher enzyme concentrations (100nM and 150nM) initially show a relatively steep increase in unwinding with a tendency to reach a plateau (around 15–20mins) compared to the plots corresponding to the lower enzyme concentrations (15nM, 30nM and 60nM) where the value rises slowly and doesn’t reach any plateau after 20minutes of reaction time. As expected, the Walker A mutant (K508A) displays no detectable unwinding activity showing that the disaggregation of the substrate is specifically due to the RecQ4 unwinding activity and is not an artefact.
To explain the presence of the plateau, it is worth mentioning that beside the ATP-dependent DNA unwinding activity, RecQ4 can also catalyse DNA annealing in an ATP-independent manner33,34. The annealing reaction may thus compete with the DNA unwinding reaction, so as to reach equilibrium. In the present case the maximum % of unwinding observed for 100nM and 150nM enzyme is 25% to 30%.
In order to validate the flexibility of this method, we prepared two additional substrates: a longer blunt-ended duplex and a substrate with a 3′ tail (Supplementary Figure S1). The enzyme was able to unwind both substrates. These results confirm that the AuNPs-based assay is flexible enough to accommodate a variety of nucleic acid substrates.
As a control, DNA unwinding experiments in similar conditions were performed using a conventional FRET based assay. FRET data (Fig. 4b) show that the unwinding is slow and keeps increasing with time without reaching a plateau within the time range used for the AuNP-based assay. The comparative data of percentage unwinding at 4th, 10th, 14th and 20th minutes respectively (as obtained by the two methods) for various concentrations of the enzyme are shown in Fig. 5.
When the unwinding efficiency of the two methods is compared at 4, 10, 14 and 20minutes (Fig. 5), for various protein concentration, there is a good match at 10 and 14minutes, with some discrepancies at lower and higher time points. At the lower end (4min) a higher activity is measured in the AuNP-based assay, whereas at 20min the trend is reversed, with the FRET methods providing a higher estimate of the activity. A possible explanation of this behaviour is that the enzyme accumulates at/around the aggregated DNA-AuNPs, as described for protein-corona35,36, and once ATP is added, a rapid unwinding activity is observed while in the FRET-based assay the double-stranded DNA substrates are homogenously dispersed in the whole reaction volume, leading to a slow initial unwinding. Similar observations of increase in enzymatic activity in the presence of citrate stabilized AuNPs have been previously reported37. Within 10 to 14minutes the enzyme in both the systems might get access to equivalent amount of the available double-stranded DNA substrates and thus shows a similar unwinding activity. By 20min reaction time the enzyme in the FRET assay continues to unwind the soluble substrates, while in the AuNP based assay, the availability of multiple DNA strands per AuNP may favour re-aggregation due to strand annealing.
Despite these small differences in enzyme behaviour observed at the initiation of the reaction, this novel AuNP based colorimetric system provides a good alternative to other existing methods to detect the unwinding activity of a helicase and to quickly obtain comparative unwinding data. In contrast to FRET, our method does not require any lengthy and costly labelling step or specific storage conditions. Instead, our approach involves easy and fast steps for the preparation of the DNA-AuNP based substrate, which is stable in terms of time and buffer conditions (e.g. ionic strength, pH, etc.) and a very simple instrument like UV-vis spectrophotometer for the readout, making it competitive over the fluorescence based assay.
In order to further test the capabilities of the methods for biochemical characterization, we tested several site-directed mutants of RecQ4. The amino acid residues shown to play a role in enzyme function33, were mutated to alanine (shown in Fig. 2). The residues include the cysteines identified as putative Zn2+ ligands (C853, C855, C897, C925, C945 and C949) and the phenylalanine (F1077) which is predicted to act as a pin. Various mutant proteins (C853A/C855A, C897A, C925A, C945A, C949A and F1077A) were expressed and purified, and DNA unwinding was then tested at a protein concentration of 100nM. The time-dependent course of the unwinding reaction is shown for all the six mutants (Fig. 6a). All mutants lost the ability to unwind the substrate and thus disrupt the AuNP aggregates, as confirmed by the FRET assay (Fig. 6b). This indicates an essential role of all the cysteine residues in the putative Zn-binding domain and the phenylalanine at the tip of the predicted WH β-hairpin in DNA unwinding, hence validating the predictions made by the in-silico analysis27, providing important insights on the biochemistry and architecture of the protein.
We have developed a unique DNA-conjugated AuNP based method to detect and monitor the unwinding activity of DNA helicases, and we tested it by characterising the activity of the catalytic core of human RecQ4 helicase and site-directed mutants. The method is guided by the changes in the optical properties of the AuNP in the DNA–AuNP aggregated based substrate, caused by the DNA unwinding activity of the enzyme.
This is the first time that such a simple, easy, robust and cheap nanoparticle-based method for following the activity of a DNA helicase has been proposed. The method was validated by comparison with a conventional fluorescence-based assay and the results from the two methods were found to be in good agreement. In addition to the ease of preparation and stability of the substrate, our method provides a naked eye test for a comparative activity assay of the enzyme as well as its mutants, and can also be used to quantitatively determine enzymatic activity. The method was tested on a monomeric helicase belonging to the SF2 superfamily, with a 3′-5′ directionality. We expect the method to be less suitable for hexameric helicases (Superfamilies SF3-SF6) as their bulky structures may sterically interferes with the nanoparticles. However, we anticipate it to be easily adaptable to SF1/SF2 5′-3′ helicases, with the design and optimization of appropriate AuNP-DNA conjugates and DNA linkers. As superfamilies SF1/SF2 constitute by far the largest group of helicases, the methods is potentially adaptable to most enzymes of pharmacological interest. Further optimisation will make this method well-suited for high-throughput commercial screenings and robotic manipulations. Its simplicity, easy detection and robustness make it a competitive alternative to the conventional methods to assay helicase activity and to quickly and easily test the activity of small molecule inhibitors against helicases.
A detailed protocol for DNA-conjugated AuNPs preparation is reported in Deka et al.30. Briefly cit@AuNPs were prepared by a known method of citrate reduction of HAuCl4. A 1.1ml aliquote of a 1.73×10−2M HAuCl4 solution (Sigma-Aldrich Chemical Co.) was added to 43.6ml of MilliQ grade water in a volumetric flask and brought to boil on a Bunsen burner. Then 300μl of 72mg/ml trisodium citrate 2-hydrate (Merck) was added and the boiling was continued for another 30min to ensure complete reduction of HAuCl4. The solution turned from colorless to deep red, indicating the synthesis of AuNPs. The red solution was removed from the heat and allowed to cool down at room temperature. Freshly prepared AuNPs were used for every experiment. The diameter of the cit@AuNPs obtained was determined to be 19±2nm from Scanning Electron Microscopy (SEM) analysis.
All the oligonucleotides used for substrate preparation (Table 1) were purchased from biomers.net and are listed in Table 1. 5.22μl of 100μM ssDNA (DNA1 or DNA2) was added to 750.0μl of AuNPs mixed by vortexing and left for 10minutes. The pH of the mixture was subsequently adjusted to 4.3 by the addition of 1.5ml of 0.1M citrate-HCl buffer at pH 4.3. The mixture of ssDNA and AuNPs was incubated at room temperature for another 30minutes after which 15.0μL of 300.0μM (1-mercaptoundec-11-yl)hexa(ethylene glycol) (TOEG6) (Sigma-Aldrich Chemical Co.) was added. The mixture was then vortexed well and left to rest for another 10minutes. The solution was centrifuged in an eppendorf tube at a speed of 14,000rpm for 10minutes using a table top Beckman Coulter centrifuge and the supernatant was discarded to remove the unbound ssDNA and TOEG6 molecules from the functionalized AuNPs. The pellet was washed twice in 250μl deionised water using the same centrifugation process as described above. The final pellet was resuspended in 250μl of buffer A (10mM sodium phosphate buffer of pH 7.0 and 1M NaCl). Freshly prepared AuNPs were used for every experiment.
A duplex solution was prepared by mixing 5.22μl of 100μM each of dup-DNA1 and dup-DNA2 in 250.0μl of buffer A (mentioned above), the duplex DNA was annealed by incubating the reaction mixture at 50°C for 15minutes followed by slow cooling. The solution thus prepared was added to the mixture of DNA1/TOEG6@AuNP and DNA2/TOEG6@AuNP. The solution turned purple in few minutes and settled down in few hours. The mixture was left at room temperature overnight. The precipitate (aggregate) obtained was separated from the colourless supernatant by centrifuging at 5,000rpm for 10minutes.
500μl of Buffer A (mentioned above) was added to the AuNP-DNA aggregate in an eppendorf and kept in a water bath at 37°C for 5minutes. 50μl of the mixture were pipetted into a cuvette for UV-vis spectrophotometry. AThe measurement was repeated after every 5°C rising temperature till 87°C. The A@260nm (corresponding to DNA) and A@520nm (corresponding to AuNP-DNA) were recorded and plotted against the respective temperatures.
The helicase assay was performed in buffer (20mM Tris–HCl pH 7.5, 5mM MgCl2, 50mM KCl, 0.1mg/ml BSA and 5% Glycerol), added to the AuNP-DNA substrates, to a final volume of 400μl with 5mM ATP and protein concentration ranging from 0 to 150nM. Before starting the analyses, the instrument (Lambda 25 UV-Vis spectrometer, Perkin Elmer) was set to zero with the helicase assay buffer as the blank. The helicase activity of the enzyme was estimated by measuring the spectrum between 400 and 800nm of the reaction in the cuvette.
The nucleotide sequence encoding the helicase and RQC domain (residues 445 to 1112 of the human sequence) was PCR-amplified using the Platinum Pfx DNA polymerase (Invitrogen). The amplified fragment was subcloned into a pET-SUMO/CAT vector (Invitrogen) using the restriction free cloning method38. Site directed mutagenesis was carried out to mutate the lysine in the Walker A motif to alanine (WA: K508A) to be used as a control in the helicase assay. Several other mutations were produced (C853A/C855A, C897A, C925A, C945A, C949A and F1077A, Fig. 2). All the plasmids were sequence verified33.
The wild type and mutant proteins were expressed with a His-SUMO tag and were purified using Nickel affinity chromatography (HisTrap FF 5ml column) with 50mM Hepes pH 8.0, 0.5M NaCl, 1mM TCEP, 5% glycerol and 500mM Imidazole buffer using a linear gradient. The tag of the purified protein was later cleaved using SUMO protease and was further purified using Heparin affinity chromatography (HiTrap Heparin 5ml column) with 50mM Hepes pH 7.5 and 1M NaCl buffer using a linear gradient. Another step of purification was performed using a size exclusion chromatography (Superdex-200) with 20mM Tris pH 8.0, 250mM NaCl, 5% glycerol and 5mM β-mercaptoethanol. Protein concentration was determined by measuring the absorbance at 280nm, and protein purity was analyzed on SDS-PAGE. Purifications of all the mutant proteins were performed as described above for wild type protein33.
The double stranded DNA substrates were prepared by annealing the complementary fluorescent labelled oligonucleotides in annealing buffer (10mM Tris-HCl pH 7.5 and 50mM NaCl) by heating at 95°C for 8min and slowly cooling to room temperature. The fluorescent labelled DNA oligonucleotides (6FAM label at 5′ end and BHQ1 label at 3′ end) are shown in Table 2.
The helicase activity was measured by using a fluorescence resonance energy transfer (FRET) based assay7. The assay was performed in helicase assay buffer (20mM Tris–HCl pH 7.5, 5mM MgCl2, 50mM KCl, 0.1mg/ml BSA and 5% Glycerol) with 10nM Fork DNA substrate (F1:F2) and 5mM ATP with protein concentration ranging from 0 to 150nM. The fluorescence intensity was recorded using Infinite F200 PRO TECAN instrument. The reaction was incubated at 95°C in the absence of enzyme to provide a measurement of 100% unwinding. The assay was done in triplicate.
How to cite this article: Deka, J. et al. DNA-conjugated gold nanoparticles based colorimetric assay to assess helicase activity: a novel route to screen potential helicase inhibitors. Sci. Rep. 7, 44358; doi: 10.1038/srep44358 (2017).
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
The work was supported by FIRB 2011 grant “Nanotechnological approaches toward tumor theragnostic” (P.P. and L.C.). This research was also supported by the Associazione Italiana per la Ricerca sul Cancro [IG14718 to S.O.] and the Cross-Border Cooperation Programme Italy-Slovenia 2007–2013 by the European Regional Development Fund and national funds [PROTEO to S.O. and L.C.]. A.D. gratefully acknowledges the financial support of the ICTP TRIL fellowship programme.
The authors declare no competing financial interests.
Author Contributions J.D., A.M. and P.P. designed the experiments. J.D. and P.P. prepared DNA conjugated gold nanoparticles substrates. J.D. carried out the helicase assay using AuNP substrates. A.M. performed protein expression and purification and FRET helicase assay. J.D. and A.M. analysed the data with the help of P.P., S.O. and L.C.; J.D., A.M., S.O. and L.C. wrote the manuscript and P.P. contributed to the discussion of the results.