Sensitive detection of methylated DNA and methyltransferase activity based on the lighting up of FAM-labeled DNA quenched fluorescence by gold nanoparticles

DNA methylation of cytosine bases, which is catalyzed by methyltransferase enzymes, involve biochemical processes that contribute to gene expression and gene regulation in cells. Detection of abnormal patterns of both methylated DNA and methyltransferase enzyme activity at early stages could be considered as promising targets for early cancer diagnosis. In the present study, a novel and facile method is introduced for the sensitive detection of the M.SssI methyltransferase (M.SssI MTase) enzyme and methylated DNA based on the fluorescence recovery of FAM-labeled DNA coupled with gold nanoparticles (AuNPs). Thiol-modified probes were functionalized with AuNPs, which brought the FAM fluorophore into the close proximity of the AuNPs. This led to the overlap between the FAM fluorescence emission and AuNPs absorption spectra, introducing a FRET occurrence and causing fluorescence quenching. The hybridization of the probe and its complementary target provided specific CpG sites for M.SssI MTase enzyme activity. The methylation process gradually converted the quenched FAM fluorophore into an emissive fluorophore upon the addition of the MTase enzyme, and the observed fluorescence recovery proved the efficiency of the assay for the detection of MTase enzyme. The fluorescence intensity showed an increasing trend with M.SssI MTase enzyme activity in the range of 1–8 U mL−1 with a detection limit of 0.14 U mL−1. The addition of methylated ssDNA targets to a ssDNA FAM-labeled probe resulted in a DNA duplex formation, leading to a strong fluorescence signal emission due to the recovery of the fluorophore signal. Conversely, the unmethylated ssDNA target caused no changes in the fluorescence signal. In the presence of methylated DNA targets, the biosensor could specifically recognize it and accordingly trigger the methylated targets through a fluorescence enhancement in the range of 5–100 pM by monitoring the increase in the fluorescence intensity with a detection limit of 2.2 pM. The obtained results showed that the assay could realize the detection of M.SssI MTase and methylated DNA effectively in diluted human serum samples. Human serum conditions showed no significant interference with the assay performance, indicating that the present method has great potential for further application in real samples.


Introduction
The detection of methylated DNA is the most widely studied epigenetic modication. This process is known as hypermethylation, which consists of the covalent addition of a methyl group from the methyl donor S-adenosylmethionine (SAM) to the cytosine within the CpG dinucleotide. 1 Hypermethylation in the promoter region of tumor suppressor genes have been implicated in downregulated or silenced genes and leads to some cancers. [2][3][4] Regulation of DNA methylation is catalyzed by the M.SssI MTase enzyme in the presence of SAM. The abnormal expression and activity of this enzyme is associated with hypermethylation of the promoter region of the tumor suppressor genes. Therefore, the detection of MTase enzyme activity and methylated DNA would be a crucial and efficient predictive biomarker for the early detection of cancers. 5,6 Recent attempts in biosensor technology could be a determining factor for the development of medical detection systems to reach the maximum assay limit of biomolecules, such as protein and nucleic acid molecules. Previous studies showed that exploiting the advantages of nanomaterials for biosensor fabrication while enhancing their performance improved their detection limit as demonstrated for DNA detection. [7][8][9][10] The conventional methods for the detection of MTase enzyme activity and methylated DNA include bisulte treatment, 11 high performance liquid chromatography (HPLC), 12 electrochemistry, [13][14][15][16][17] enzyme-linked immunosorbent assay (ELISA), 18 surface plasmon resonance (SPR) spectroscopy, 19,20 surface-enhanced Raman scattering (SERS), 21,22 and use of microuidic based biosensors 23,24 and uorescence based biosensors. [25][26][27][28] Although several attempts have been made for the detection of methylated DNA in recent years, there are still some drawbacks that limit their sensitivity and efficacy in clinical applications.
Among the above-mentioned approaches, the uorescence based assays (i.e., utilization of uorophore molecules in a detection system) present remarkable characteristics, including high and specic sensitivity and selectivity with relatively rapid and simple operations, which introduce them as efficient and reliable detection markers. The presence of donor uorophore and acceptors in specic distances leads to uorescence resonance energy transfer (FRET) between the uorescent donors and acceptors. Nowadays, FRET has been considered for the design and fabrication of novel biosensors and detection systems. 29,30 FAM is a well-known and widespread uorophore used for the labeling of DNA as a uorescent marker and can be attached to either the 5 0 or 3 0 end of oligonucleotides. It can be protonated and has decreased uorescence below pH 7.
Au nanoparticles (AuNPs) are regarded as one of the most applicable nanoparticles for the detection of biological analytes. Their unique properties include surface attachment to biological elements, such as DNA, proteins, and enzymes. Their absorption properties and high electron conductivity makes them one of the most frequently used biological probes. The absorption and electron transfer capacity properties of AuNPs are dependent on their diameter size, which can be controlled through the parameters of their synthesis. These tunable characteristics introduce them as a very efficient candidate for being an acceptor element in a FRET based detection system for biological studies. [31][32][33] In the present study, a rapid and sensitive FRET based DNA methylation detection method was developed. AuNPs (serving as uorescence acceptors) were conjugated with a 6-carboxy-uorescein (FAM)-labeled probe employed as the uorescence donor. Through this process, the proximity of the uorophore and AuNPs resulted in a FRET occurrence, leading to the uorescence quenching of the FAM uorophores. The DNA hybridization of the probe with a complementary target formed a double stranded DNA with a specic inside recognition site for M.SssI MTase enzyme activity. The presence of a 5 0 -CCGG-3 palindromic site provided the reaction site for initiating the M.SssI MTase activity and DNA methylation process. The activity of the MTase enzyme caused a linear uorescence recovery of the FAM uorophore, which showed a relationship with the enzyme concentration and contributed to the enzyme assay (Scheme 1). Meanwhile, for the detection of methylated DNA, the unmethylated and methylated ssDNA targets were designed to hybridize with the functionalized probe during the complementary pairing. In the presence of ssDNA targets, the DNA probes hybridized with the targets and achieved uorescence recovery for the methylated targets, whereas no changes were observed in the presence of unmethylated targets (Scheme 2). Therefore, it was concluded that the uorescence recovery of FAM-labeled DNA was closely related to the M.SssI MTase enzyme activity and methylated DNA level. The present novel detection strategy could determine the DNA methylation levels of a specic site of DNA through the sensitive and rapid assay of MTase enzyme and methylated ssDNA targets.
The DNA probe and target strands were synthesized according to the CpG sequence sites of the P53 tumor suppressor gene promoter by Shanghai Generay Biotech Co. All oligonucleotide stock solutions were puried by PAGE, prepared with TE buffer, and kept frozen until used. For the TE buffer preparation, 1 mL of 1 M Tris-HCl (pH 7.5) and 0.2 mL EDTA (0.5 M) were added to deionized water and diluted to a total volume of 100 mL. Human serum samples were obtained from Tehran University of Medical Science, Tehran, Iran and used for the determination of method efficiency in real sample conditions.

Instrumentation
All uorescence measurements were performed on a Perkin-Elmer LS-55 uorescence spectrometer with the excitation and emission slits set at 10 nm bandpass. UV-vis spectroscopy was performed by a Specord 250 spectrophotometer (Analytik Jena, Germany). The morphology of the AuNPs was characterized by a transmission electron microscope (TEM) (Zeiss, EM10C, 100 kV, Germany). Deionized water for the preparation of the oligonucleotide stock solution was acquired from a Milli-Q ultrapure water system (Millipore, Z18 MU cm).

Synthesis of gold nanoparticles
AuNPs with a mean diameter in the range of 10 AE 3 nm were synthesized by the bottom up method as previously described. 34 Briey, 50 mL of an aqueous solution of tetrachloroauric acid (1 mM) was heated to reach the boiling point while being stirred with the magnetic stir bar in a round-bottom ask with reux modes. Then, 10 mL of trisodium citrate (38.8 mM) was poured into the solution and le to boil for another 10 min. The solution color changed from yellow to purple, and then nally turned wine-red, which was stored in the refrigerator at 4 C.

Construction of the FAM-oligo-Au probe
Au nanoparticles were functionalized by a uorophore (FAM) modied 5 0 -ssDNA probe, which also had a thiol group at the 3 0 end. Prior to use, a solution of the reducing agent DTT was added to the lyophilized thiolated DNA probe to reduce the disulde bonds and prevent the formation of probe dimers. It was then incubated at room temperature for 1 hour (0.1 M DTT, 0.18 M phosphate buffer (PB), pH 8.0). Freshly prepared AuNPs and oligonucleotides probes were added to the solution with phosphate buffer (PB) and sodium dodecyl sulfate (SDS) at 0.01 M and 0.01% concentration, respectively. The ssDNA/AuNP solution was allowed to incubate overnight at room temperature. In order to remove the unbound probes through a washing process, the AuNPs were centrifuged at 14 000 rpm and the supernatant was removed, leaving a red pellet of AuNPs at the bottom of the ask. The precipitated nanoparticles were resuspended in a solution containing 0.01% SDS. This process was repeated 3 times to obtain the FAM-oligo-Au probe.

Detection of M.SssI enzyme activity
In order to assay the M.SssI MTase, the probe strand (FAMlabeled ssDNA) and complementary strand (target 1) were rst hybridized at the same concentration in 50 mM Tris-HCl buffer (pH 7), which was incubated at 37 C for 1 h. Following the incubation, the SAM substrate and buffer solution (10 mM potassium phosphate pH 7.0, 400 mM KCl, 1 mM DTT, 1 mM EDTA, 0.2 mg mL À1 BSA and 50% v/v glycerol) with various concentrations of M.SssI MTase (from 1 to 8 U mL À1 ) were added to the solution. The mixture was then incubated at 37 C for 30 min, and nally the reaction was terminated by the inactivation of the enzymes by further incubating for 20 min at 80 C. Subsequently, the uorescence spectra of the newly formed methylated DNA assembly were recorded.

Detection of methylated DNA
To determine the effect of methylated ssDNA target concentrations on the uorescence recovery efficiency, a xed amount (1 mg mL À1 ) of the FAM-oligo-Au capture probe was incubated with both unmethylated and methylated ssDNA in the range of 5 pM to 100 pM oligonucleotide concentrations for 2 h to allow hybridization at room temperature. The above mixture would allow the formation of a hybridized sandwich complex structure. Aer incubation, the uorescence intensity was measured using a spectrouorometer. The obtained uorescence intensities were recorded under the same experimental conditions.

Real sample assay
In order to determine the efficiency of the experiment in a real sample condition, the assay was performed in three different media, namely, (1) buffer (optimum condition) as a control, (2) human serum sample and (3) in the presence of human serum plus interfering DNA. 50 pM of P probes, 50 pM of hybridized P/ S3 and 50 pM of hybridized P/S2 were added to each sample, respectively. DNA was then extracted from the above mixtures by a QIAamp DNA Blood Mini Kit (Qiagen) and diluted 2 fold in Tris-HCl buffer. The extracted P/S2 hybrids were analyzed by a spectrouorometer, while the puried P/S1 hybrids were treated with the M.SssI MTase enzyme for the enzyme assay process.

Assay strategy based on FRET between FAM and AuNPs
It was assumed that FAM and AuNPs acting as the donor and acceptor, respectively, would be a favorable pair for FRET. As illustrated in Scheme 1, the FAM-labeled probe that emitted a strong uorescence was conjugated with AuNPs through the 3 0 thiolated end of the probe. The obtained results showed that the uorescence signal of FAM was quenched efficiently in the presence of certain amounts of AuNPs. This result was due to a signicant overlap of the uorescence emission of FAM with the absorption spectrum of the AuNPs (Fig. 1a). The spectrum of AuNPs showed the absorption peak at 520 nm and its overlap with the FAM functionalized DNA at 540 nm, suggesting that FRET could be occurring between the FAM-labeled DNA (donors) and the AuNPs (acceptors) (Fig. 1b).
The FAM uorescence quenching resulted from a resonance energy transfer from the FAM to the AuNPs. The uorescence quenching efficiency reached a uorescence intensity plateau of 82% when a concentration higher than 1.9 mM of AuNPs was dispersed into the solution. Increasing the AuNPs concentration to more than 1.9 mM did not lead to a decrease in the uorescence intensity and showed that this concentration is capable of quenching the uorescence of the FAM-labeled probes. The estimated FRET efficiency or uorescence quenching efficiency at the same excitation wavelength is dened as: where IDA and ID are the uorescence intensities of an excited donor uorophore D in the presence and absence of an acceptor A, respectively. So, the nal concentration of AuNPs was determined as 1.9 mM.

Characterization of FAM and AuNPs
The TEM image of the synthesized AuNPs showed that the nanoparticles were well dispersed and exhibited uniform sizes with a diameter ranging from 8 to 12 nm (Fig. 1c), and showed a visible red color when dispersed in ultrapure water (inset of Fig. 1c). The UV-vis spectra of bare AuNPs and FAM-labeled DNA-AuNPs are compared in Fig. 1d. The absorption peak related to the AuNPs-DNA sample at approximately 260 nm was observed, conrming Au functionalization with ssDNA on their surfaces. 35 Also, as shown in Fig. 1d, the absorption peak of the DNA/AuNPs sample shied slightly to the right by about 7 nm (from 522 to 529 nm). The observed shi for the DNA/AuNPs sample is attributed to its functionalization with the surface attached DNA probe. 35,36

Detection of DNA methylation induced by M.SssI MTase
In order to determine the effect of methylation on the FAM-ssDNA-AuNP probes, an unmethylated target was hybridized with the probe to provide -CCGGsites as a substrate for DNA methylation by the M.SssI MTase enzyme in the presence of SAM. The methylation induction showed the conversion of the quenched FAM uorophore to a green emitting uorophore upon DNA methylation. Fig. 2 shows that the uorescence emission of the treated dsDNA increases with the increase in the M.SssI MTase enzyme concentration from 1 to 8 U mL À1 . The inset of Fig. 2 clearly shows that the DF was proportional to  the M.SssI MTase concentration from 1 to 8 U mL À1 with a correlation coefficient (R 2 ) of 0.9334. The regression equation was y ¼ 52.119x + 247.46 and the detection limit of this assay was determined to be 0.14 U mL À1 . Previously, it was reported that the activities of enzymes such as proteases and nucleases would be detectable using "switchon" uorescent nanoprobes in order to determine their enzyme kinetics or biological activity. [37][38][39][40] Most of the mentioned methods are based on the cleavage function of enzymes on their substrate, which separate the uorescence donor and acceptor, leading to uorescence emission from the quenched uorophore. Also, many reports used oligonucleotide cleavage based methods for MTase enzyme detection. 41,42 It is worth noting that we reported the cleavage free MTase enzyme detection for the rst time. The induction of a methyl group on the DNA structure may contribute to the recovery of the quenched FAM uorophore through its electron donating activity. Recently, Pongor et al. reported that the electron-donor properties of the methyl group possibly enhances the basestacking interactions that stabilize strand pairing in the stretched state. 43 They found that in the overstretched state, methylated DNA is longer than the nonmethylated form, suggesting a novel extended S-form. In our experience, the abovementioned changes in the DNA structure and length would have crucial roles in the observed uorescence recovery of the FAM uorophore.
Meanwhile, it was illustrated that DNA methylation reduced the FRET efficiency while it was in the nucleosome structure, and indicated that the extent of DNA wrapping or the conformation of DNA was altered upon methylation. 44 Prior ndings indicate that upon the excitation of the sample with polarized light, a favorable dipole orientation (i.e., parallel) of the uorophore and acceptors and the relative angle between them are determining factors in the FRET efficiency. 45,46 The FRET orientation factor is called k 2 and different donor/acceptor conformations can lead to k 2 values in the 0 # k 2 # 4 range. 47 A value of zero for k 2 means that the process of FRET is forbidden. We assumed that the covalent addition of a methyl group to the DNA could also signicantly change the DNA exibility and conformation. This could lead to different FAM uorophore and AuNP dipole orientations, resulting in a k 2 value of zero and subsequently interfering in the FRET occurrence and uorescence recovery.

Detection of methylated DNA based on FAM/Au NPs FRET system
In order to determine the effect of hybridization on FRET in the present assay, two complementary target DNAs (methylated and unmethylated) were designed. Interestingly, the uorescence was recovered when the probe was hybridized with the perfectly matched methylated DNA. This effect was not observed for the unmethylated target DNA.
As discussed previously, the uorescence of the FAM labeled oligonucleotide was recovered aer the addition of the methylated DNA. The uorescence intensity pattern of the FAM was monitored relative to the amount of methylated ss-DNA target introduced into the FAM-oligo-AuNPs system. It seems that the presence of a methyl group on the complementary target strand inhibited the energy transfer from FAM to AuNPs and resulted in uorescence recovery. As depicted in Fig. 3, the uorescence emission gradually enhanced as the methylated DNA target concentration increased from 5 pM to 100 pM. For the determination of the detection limit of the present FRET system for ssDNA targets, the calibration curve of the detection procedure was plotted in the inset of Fig. 3. The observed results showed that the increase in the methylated ssDNA targets is proportional to the increase in the uorescence intensity in the linear regression equation y ¼ 8.56x + 82.45, with a correlation coef-cient (R 2 ) of 0.9642. Finally, the detection limit was determined to be 2.2 pM. Thus, it was concluded that the methylated DNA target can be detected based on the recovered uorescence intensity and it would be possible to discriminate the methylated DNA from the unmethylated DNA by comparing their uorescence recovery results.

Application of the assay in a human serum sample
Determining the applicability of the assay in a real sample is a challenging factor to evaluate the performance of the method. Hence, the human serum as a complex and real substrate was applied in the experiment in the presence of hybridized P/S2 for the detection of methylated DNA and P/S1 strands, as well as for the detection of the MTase enzyme. Meanwhile, noncomplementary strands of DNA were used as another interfering material present in human serum as an additional sample. As shown in Fig. 4, the obtained results from the puried human serum DNA (series 2) showed about 84% and 90% uorescence recovery for the methylated DNA detection method (red columns) and enzyme assays (black columns), respectively. Upon the addition of non-specic DNA (series 3) to Fig. 3 The fluorescence spectra of (1 mg mL À1 ) of FAM-oligo-Au ssDNA probe upon incubation with a series of concentrations of methylated ssDNA targets (a-g: 0, 5 pM, 10 pM, 20 pM, 40 pM, 70 pM, 100 pM). The inset shows a standard curve of the relative fluorescence intensity versus the methylated ssDNA target concentration.
the above samples, the uorescence recovery of the methylated DNA and M.SssI enzyme assay reached 81% and 63%, respectively, which proved the considerable efficiency of the present methods in real samples. Although the presence of non-specic DNA decreased the uorescence recovery, it did not interfere signicantly with the experiment.
As we anticipated, our assay system was able to detect DNA methylation and enzyme activity in a real sample condition. However, further studies are needed in order to fully evaluate and validate the clinical utility of this method.

Sensitivity of experiment
To explore the sensitivity for the detection of methylated DNA, the effects of a complementary methylated sequence (P/S2), a single base mismatched sequence (P/S3), a single base mismatched methylated sequence (P/S4) and a non-complementary sequence (P/S5) were examined at the same concentration for comparison. As depicted in Fig. 5, a comparison of the aboveformed oligonucleotides was performed and the obtained results showed that the uorescence intensity of FAM just recovered with a mismatched target at a very low efficiency of around 29%. The uorescence recovery was not observed with the addition of non-complementary and single base unmethylated targets. It is worth noting that the presently described FRET based assay has an efficient capacity for the determination of a methylated sequence. The obtained results also suggested that there was no obvious inuence for the detection of a methylated DNA in the presence of the potential interfering sequences at the same concentrations.

Conclusion
Here, we exploited a new strategy, in which the FAM uorescence donor could de-excite through a methylation process for the detection of DNA methylation. To construct the biosensor, FAM uorophore labeled probes containing the thiol modied tails were designed. Aer conjugation with AuNPs, the complementary unmethylated targets were used as targets to form a dsDNA duplex and provide a substrate for the M.SssI MTase enzyme activity assay. The MTase enzyme activity resulted in the uorescence recovery upon an increase in the enzyme concentration. A similar behavior was also observed upon the addition of methylated ssDNA targets. In our assumption, the DNA orientation aer DNA methylation may cause the k 2 value to approach zero, forbidding the FRET occurrence and resulting in the uorescence recovery. This FRET based system provided an efficient and simple method rather than other conventional multistep methods for the detection of DNA methylation. The overall results clearly illustrated that this introduced novel FRET-based strategy was sensitive enough with great potential to detect MTase activity and determine the presence of the methylated DNA. The performance of this experiment in human serum had no signicant effect on the assay performance, holding great potential for further clinical application.

Conflicts of interest
There are no conicts to declare.