English

• Adenosine triphosphate (ATP), a multifunctional molecule in living entities, plays vital regulatory roles in various biological processes [1-3]. A disordered ATP level is closely related to many serious diseases including hypoglycemia, Parkinson's disease, hypoxia, ischemia, and angiocardiopathy [4-6]. To date, many classic ATP detection methods have been explored [7-10], such as electrophoresis, high performance liquid chromatography, and chemiluminescence. However, these approaches are usually time-consuming, labor-intensive, and not suitable for in situ imaging of ATP in living cells [11, 12]. Thus, engineering of a facile and reliable strategy for the real-time monitoring of intracellular ATP levels is of essential significance for clinical diagnosis and pathological analysis.

  In recent years, the aptamer-based ATP sensors have received extensive attention owing to the benefits of aptamer-recognizing components, including high affinity, long-term stability, and ease of synthesis and modification [13, 14]. Aptamers are single-stranded oligonucleotides capable of recognizing their specific targets [15-18]. A variety of aptamer sensors have been designed for ATP analysis by using different transduction models [19-23], such as fluorescence, colorimetry, and electrochemistry. Especially, the fluorescence-based detection has been widely applied in biological research due to its fast response and simple operation [19, 20]. Nevertheless, at least two issues remain to be resolved when using aptamer-based fluorescent probes for ATP imaging in living cells. The first problem is the effective delivery of probes into the cell, a critical step for intracellular applications. The other is the biostability and biocompatibility of nucleic acid probes. Generally, a single nucleic acid probe is susceptible to degradation by intracellular nucleases, leading to inevitable false-positive signals [24, 25]. To settle these problems, strategies incorporating exogenous nanomaterials (e.g., graphene and polymers) have been developed for intracellular ATP imaging [26, 27]. However, nanomaterials usually require complex preparation and functionalization steps, and even show some cytotoxicity to cells, which may interfere with the authentic expression level of ATP in living cells.

  DNA-based nanotechnology provides exciting opportunities to explore powerful biosensing strategies [28-32]. Especially, the DNA tetrahedron, an emerging nanophase biomaterial, exhibits several unique merits in the field of biology and medicine, such as excellent biocompatibility, nanoscale controllability, and editability [33-35]. Furthermore, it can be rapidly endocytosed into cells via a caveolin-dependent pathway, and maintain good stability within 48 h [36, 37]. These capabilities have significantly facilitated its applications in bioimaging, logic computing and drug delivery [38-40]. For example, Xing et al. have developed an accelerated DNA tetrahedron based molecular beacon for efficient detection and imaging of miRNA in living cells [41]. Su et al. have reported the first example of vertebral-shaped DNA tetrahedron nanostructures for accurate cancer identification and miRNA silencing induced therapy [42]. Despite these progresses, most of them are single-intensity sensing modes, which are more prone to false-positive signals from complex biological matrices. As such, there is still an urgent need to develop an ingenious DNA tetrahedron-based fluorescent sensor for accurate intracellular imaging.

  Inspired by the above challenges, we herein developed a DNA tetrahedron-based split aptamer probe (TD probe) for ratiometric fluorescence imaging of ATP in living cells, as illustrated in Scheme 1. The TD probe was composed of three modules: the DNA tetrahedron self-assembled by four DNA oligonucleotides (T1, T2, T3 and T4), the Cy3-labeled split ATP aptamer probe a (Apt-a) and the Cy5-labeled split ATP aptamer probe b (Apt-b). The DNA tetrahedron served as the backbone to immobilize two split aptamer probes via base complementary pairing. Meanwhile, the split aptamer probes as target recognition modules specifically bind to ATP molecules. In the absence of ATP, the fluorescent donor Cy3 and the fluorescent acceptor Cy5 were far apart due to the spatial separation of Apt-a and Apt-b, resulting in a low fluorescence resonance energy transfer (FRET) signal. In contrast, the presence of ATP will change the structure of the TD probe from the open to closed state, thus bringing the dual fluorophores into close proximity for high FRET signals. The proposed TD probe exhibited several unique properties, such as convenient preparation through a one-step procedure, excellent cell penetration and biological stability. Moreover, the FRET "off" to "on" signal output mode could effectively avoid false-positive signals from complex biological matrices, ensuring reliable ATP imaging in living cells.

  Scheme 1

  

  Scheme 1.  (A) Construction and (B) schematic illustration of the TD probe for ratiometric fluorescence imaging of ATP in living cells.

  DownLoad: Full-Size Img PowerPoint

  The synthesis of the TD probe was first investigated via 6% polyacrylamide gel electrophoresis (PAGE). As shown in Fig. 1A, with the addition of DNA strands (lanes 1-4), the electrophoretic mobility gradually decreases due to the increase in the molecular weight of the hybridization complex, suggesting the construction of the DNA tetrahedron. Then, after mixing with Apt-a and Apt-b, an extended band with lower mobility appeared in lane 5, indicating the successful formation of the TD probe. In addition, the TD probe was also characterized by dynamic light scattering (DLS), where the average hydrodynamic diameters of the DNA tetrahedron and TD probe were approximately 12.1 nm and 16.1 nm, respectively (Fig. 1B). Moreover, atomic force microscopy (AFM) image revealed that the TD probe exhibited a good monodispersion (Fig. S1 in Supporting information). All these results clearly verified the successful assembly of the TD probe.

  Figure 1

  

  Figure 1.  (A) PAGE characterization of the TD probe. Lane 1: S1; lane 2: S1+S2; lane 3: S1+S2+S3; lane 4: S1+S2+S3+S4; lane 4: S1+S2+S3+S4+Apt-a+Apt-b. (B) DLS analysis of the DNA tetrahedron and TD probe.

  DownLoad: Full-Size Img PowerPoint

  The feasibility of the TD probe for ATP detection was next investigated in a homogeneous solution. As shown in Fig. S2 (Supporting information), in the absence of ATP, a very weak FRET signal (F_A/F_D) of acceptor fluorophore (Cy5) to donor fluorophore (Cy3) was observed due to the spatial separation of Apt-a and Apt-b. Upon recognition of ATP, the TD probe showed a significant FRET signal, indicating that the designed TD probe was feasible for ATP sensing in vitro. Subsequently, real-time monitoring of the fluorescence emission intensity changes was performed to study the reaction kinetics. The excitation wavelength was fixed at 525 nm, and the emission wavelength was 667 nm (Fig. S3 in Supporting information). Without ATP, almost no change in fluorescence intensity was detected. Upon the addition of ATP, the fluorescence signal gradually increased and stabilized within 20 s, indicating that the TD probe could rapidly respond to the target. Then, the detection performance of the TD probe for ATP was investigated. As shown in Figs. 2A and B, the fluorescence spectra of the TD probe showed decreased Cy3 fluorescence at 560 nm and increased Cy5 fluorescence at 667 nm with the increasing ATP concentration from 0 to 3 nmol/L. Moreover, there was a good linear relationship between the FRET signal and ATP concentrations in the range of 0.02-0.2 mmol/L, and the limit of detection (LOD) was calculated to be 3.9 µmol/L (R² = 0.962) according to the blank signal plus 3σ (3 times the standard deviation). The performance of this strategy is better than some of the present nucleic acid-based fluorescence sensors [14, 43]. The selectivity of the TD probe was also evaluated using three analogue molecules of ATP, including cytidine triphosphate (CTP), guanosine triphosphate (GTP), and uridine triphosphate (UTP). The FRET signal in the presence of ATP was higher than that of these ATP analogues (Figs. 2C and D), demonstrating the excellent specificity of the TD probe for ATP detection.

  Figure 2

  

  Figure 2.  (A) Fluorescence spectra of the TD probe in response to different concentrations of ATP. (B) The relationship between the FRET signal (F_A/F_D) and ATP concentration. The inset shows the linear relationship from 0.02 to 0.2 mmol/L. Fluorescence spectra (C) and FRET signal (D) of the TD probe in response to 2 mmol/L ATP, CTP, GTP and UTP, respectively.

  DownLoad: Full-Size Img PowerPoint

  Good biological stability is quite important for DNA probes applied in complex physiological environment. To confirm the TD probe has good anti-degradation ability, the interaction of the TD probe with DNase I was first studied. As reported, DNase I is a powerful endonuclease that nonspecifically degrades single- and double-stranded DNA molecules by cleaving phosphodiester bonds [44, 45]. The TD probe was respectively treated with 0.25 U/mL and 2.5 U/mL of DNase I for different times at 37 ℃, then characterized by electrophoresis. As shown in Fig. 3A, almost no degradation was observed after incubation of the TD probe with 0.25 U/mL DNase I for 1 h, indicating that the TD probe had good biostability. However, after treatment with a higher concentration of DNase I (2.5 U/mL), the TD probe exhibited a time-dependent decrease in the electrophoretic band, and almost completely disappeared at about 50 min. Nevertheless, the FRET signal of the TD probe was not affected with the increase in the treatment time by 2.5 U/mL DNase I (Fig. S4 in Supporting information), which could avoid false-positive signals caused by nuclease degradation. The stability was further investigated by incubating TD probe with 10% (v/v) fetal bovine serum (FBS) (Fig. 3B). Compared with ssDNA probe, the TD probe exhibited longer-term stability, which was beneficial for intracellular bioimaging. Moreover, the biocompatibility of the TD probe was evaluated by the standard MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay. After incubating with different concentrations of TD probes (0, 50, 100, 200 and 400 nmol/L) for 12 h at 37 ℃, HeLa cells showed a good viability, with percentages over 88% (Fig. S5 in Supporting information). These results indicated that the TD probe was safe for cells and suitable for the following applications in living cells.

  Figure 3

  

  Figure 3.  (A) PAGE analysis of the low concentration of DNase I (0.25 U/mL) and high concentration of DNase I (2.5 U/mL) degradation assay products for the TD probe. (B) FBS (10%, v/v) degradation assay products for the TD probe and ssDNA at 37 ℃.

  DownLoad: Full-Size Img PowerPoint

  After demonstrating that the TD probe could achieve ATP sensing in vitro, we next explored its feasibility for ATP imaging in living cells. HeLa cells, which overexpressed ATP molecules, were selected as the research model. The cellular delivery ability of the TD probe was first investigated. As shown in Fig. 4, HeLa cells treated with free ssDNA probe showed very weak Cy3 and Cy5 fluorescence, suggesting that the free ssDNA probes were not suitable for ATP imaging in living cells due to their poor cell permeability and low biostability. In contrast, the cells incubated with the TD probe displayed obvious Cy3 and Cy5 fluorescence, indicating that the TD probe could be self-delivered into cells without any transfection agent, and used for ATP-imaging analysis in living cells. Then, the concentration and incubation time of the TD probe and HeLa cells were optimized for an optimal imaging performance. As shown in Fig. S6 (Supporting information), the FRET signal gradually increased during the incubation and reached saturated at about 4 h. Similarly, the FRET signal enhanced with increasing TD probe concentration and reached a maximal level at about 200 nmol/L (Fig. S7 in Supporting information). Therefore, 200 nmol/L TD probe and an incubation time of 4 h were chosen for subsequent experiments. Next, the intracellular distribution of the TD probe was examined. After incubating with the TD probe, HeLa cells were treated with the nuclear-specific dye Hoechst 33342 for 10 min at 37 ℃. The imaging date showed that the Cy3 and Cy5 fluorescence was mainly distributed in the cytoplasm, which did not overlap with the blue fluorescence from Hoechst 33342, suggesting that the TD probe could be used for cytoplasmic ATP detection (Fig. S8 in Supporting information).

  Figure 4

  

  Figure 4.  Confocal fluorescence imaging of HeLa cells incubated with 200 nmol/L TD probe (top) and ssDNA (bottom) for 4 h at 37 ℃.

  DownLoad: Full-Size Img PowerPoint

  To further confirm that the FRET signal was derived from endogenous ATP molecules in living cells, the TD probe was used to measure intracellular ATP changes upon different treatments. As reported, oligomycin can reduce intracellular ATP levels by inhibiting ATP synthase [46], and Ca²⁺ can increase intracellular ATP production via activating dehydrogenases [47, 48]. Before incubating with the TD probe, HeLa cells were treated with 10 µmol/L oligomycin or 5 mmol/L CaCl₂ for 30 min at 37 ℃. As shown in Fig. 5, compared with the control group (top), HeLa cells treated with oligomycin showed a decreased FRET signal (middle), while Ca²⁺-treated cells exhibited a decreased FRET signal (bottom), indicating that the observed FRET signal was indeed associated with the intracellular ATP concentration. These results illustrated that the TD probe could dynamically monitor the changes of ATP in living cells.

  Figure 5

  

  Figure 5.  Confocal fluorescence imaging of HeLa cells treated with medium (top), 10 µmol/L oligomycin (middle), and 5 mmol/L Ca²⁺ (bottom), followed by incubation with 200 nmol/L TD probe for 4 h at 37 ℃.

  DownLoad: Full-Size Img PowerPoint

  In summary, combining DNA tetrahedron with split aptamers, we have successfully developed a TD probe for ratiometric fluorescence imaging of ATP in living cells. The TD probe is easily prepared by one-step incubation, and shows excellent specificity and high sensitivity for in vitro detection of ATP with a detection limit of 3.9 µmol/L. Moreover, the TD probe presents improved cell internalization efficiency, prominent resistance to nuclease degradation and satisfactory biocompatibility. More importantly, the FRET "off" to "on" signal output mode effectively avoids false-positive signals from complex biological matrices, which is critical for intracellular applications, especially for accurate imaging over long periods of time. Furthermore, by replacing the split aptamers attached to DNA tetrahedron, the proposed strategy may be expanded to detect various intracellular targets. Thus, it provides a valuable sensing platform for accurate biomarkers analysis in living cells, which is of great significance for early clinical diagnosis and therapeutic evaluation.

  Declaration of competing interest

  The authors report no declarations of interest.

  Acknowledgments

  This work was supported by the Natural Science Foundation of China (Nos. 21877030, 21735002, 21778016 and 21521063).

  Supplementary materials

  Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2022.05.020.

  
  
• 1. [1]

     S. Biswas, K. Kinbara, T. Niwa, et al., Nat. Chem. 5 (2013) 613–620. doi: 10.1038/nchem.1681

  2. [2]

     L. Lu, B. Li, S. Ding, et al., Nat. Commun. 11 (2020) 4192–4203. doi: 10.1038/s41467-020-18051-1

  3. [3]

     Y. Zhou, L. Zou, G. Li, et al., Anal. Chem. 93 (2021) 13960–13966. doi: 10.1021/acs.analchem.1c03284

  4. [4]

     A.V. Gourine, E. Llaudet, N. Dale, et al., Nature 436 (2005) 108–111. doi: 10.1038/nature03690

  5. [5]

     B.S. Khakh, R.A. North, Nature 442 (2006) 527–532. doi: 10.1038/nature04886

  6. [6]

     Z. Duan, L. Tan, R. Duan, Anal. Chem. 93 (2021) 11547–11556. doi: 10.1021/acs.analchem.1c02049

  7. [7]

     M.J. Gordon, X. Huang, S.L. Pentoney, et al., Science 242 (1988) 224–228. doi: 10.1126/science.242.4876.224

  8. [8]

     C. Fu, L. Song, Y. Fang, Anal. Chim. Acta 399 (1999) 259–263. doi: 10.1016/S0003-2670(99)00466-3

  9. [9]

     F. Özogul, K.D.A. Taylor, P.C. Quantick, et al., Int. J. Food Sci. 35 (2000) 549–554. doi: 10.1111/j.1365-2621.2000.00405.x

  10. [10]

      H. Huang, Y. Tan, J. Shi, et al., Nanoscale 2 (2010) 606–612. doi: 10.1039/b9nr00393b

  11. [11]

      J. Dong, M. Zhao, Trends Anal. Chem. 80 (2016) 190–203. doi: 10.1016/j.trac.2016.03.020

  12. [12]

      Y. Zhou, L. Yang, J. Wei, et al., Anal. Chem. 91 (2019) 15229–15234. doi: 10.1021/acs.analchem.9b04126

  13. [13]

      S. Ng, H.S. Lim, Q. Ma, et al., Theranostics 6 (2016) 1683–1702. doi: 10.7150/thno.15850

  14. [14]

      Y.X. Wang, D.X. Wang, J.Y. Ma, et al., Analyst 146 (2021) 2600–2608. doi: 10.1039/D1AN00275A

  15. [15]

      L.C. Bock, L.C. Griffin, J.A. Latham, et al., Nature 355 (1992) 564–566. doi: 10.1038/355564a0

  16. [16]

      C.H.B. Chen, G.A. Chernis, V.Q. Hoang, et al., Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 9226–9231. doi: 10.1073/pnas.1332660100

  17. [17]

      N. Lin, L. Wu, X. Xu, et al., ACS Appl. Mater. Interfaces 13 (2021) 9306–9315. doi: 10.1021/acsami.0c11878

  18. [18]

      L. Li, J. Wan, X. Wen, et al., Anal. Chem. 93 (2021) 7369–7377. doi: 10.1021/acs.analchem.1c01445

  19. [19]

      X. Li, Y. Peng, Y. Chai, et al., Chem. Commun. 52 (2016) 3673–3676. doi: 10.1039/C6CC00110F

  20. [20]

      X. Chai, Z. Fan, M.M. Yu, et al., Nano Lett. 21 (2021) 10047–10053. doi: 10.1021/acs.nanolett.1c03732

  21. [21]

      C. Li, M. Numata, M. Takeuchi, et al., Angew. Chem. Int. Ed. 44 (2005) 6371–6374. doi: 10.1002/anie.200501823

  22. [22]

      X. Zhang, C. Song, K. Yang, et al., Anal. Chem. 90 (2018) 4968–4971. doi: 10.1021/acs.analchem.7b05442

  23. [23]

      X. Zuo, S. Song, J. Zhang, et al., J. Am. Chem. Soc. 129 (2007) 1042–1043. doi: 10.1021/ja067024b

  24. [24]

      A.K. Chen, M.A. Behlke, A. Tsourkas, Nucleic Acids Res. 35 (2007) e105. doi: 10.1093/nar/gkm593

  25. [25]

      L. He, D.Q. Lu, H. Liang, et al., ACS Nano 11 (2017) 4060–4066. doi: 10.1021/acsnano.7b00725

  26. [26]

      F. Gao, J. Wu, Y. Yao, et al., RSC Adv. 8 (2018) 28161–28171. doi: 10.1039/C8RA05193C

  27. [27]

      C.H. Zhang, H. Wang, J.W. Liu, et al., ACS Sens. 3 (2018) 2526–2531. doi: 10.1021/acssensors.8b00670

  28. [28]

      L.Y. Duan, J.W. Liu, R.Q. Yu, et al., Biosens. Bioelectron. 177 (2021) 112976. doi: 10.1016/j.bios.2021.112976

  29. [29]

      D. Zhu, Y. Wei, T. Sun, et al., Anal. Chem. 93 (2021) 2226–2234. doi: 10.1021/acs.analchem.0c04092

  30. [30]

      C. Zhu, J. Yang, J. Zheng, et al., Anal. Chem. 91 (2019) 15599–15607. doi: 10.1021/acs.analchem.9b03659

  31. [31]

      J. Deng, J. Xu, M. Ouyang, et al., Chin. Chem. Lett. 33 (2022) 773–777. doi: 10.1016/j.cclet.2021.08.046

  32. [32]

      Z. Qing, J. Xu, J. Hu, et al., Angew. Chem. Int. Ed. 58 (2019) 11574–11585. doi: 10.1002/anie.201812449

  33. [33]

      Z. Zhou, Y.S. Sohn, R. Nechushtai, et al., ACS Nano 14 (2020) 9021–9031. doi: 10.1021/acsnano.0c04031

  34. [34]

      Z. Qing, J. Hu, J. Xu, et al., Chem. Sci. 11 (2020) 1985–1990. doi: 10.1039/C9SC04916A

  35. [35]

      L. Meng, W. Ma, S. Lin, et al., ACS Appl. Mater. Interfaces 11 (2019) 6850–6857. doi: 10.1021/acsami.8b22444

  36. [36]

      A.S. Walsh, H. Yin, C.M. Erben, et al., ACS Nano 5 (2011) 5427–5432. doi: 10.1021/nn2005574

  37. [37]

      L. Liang, J. Li, Q. Li, et al., Angew. Chem. Int. Ed. 53 (2014) 7745–7750. doi: 10.1002/anie.201403236

  38. [38]

      J. Wang, D.X. Wang, J.Y. Ma, et al., Chem. Sci. 10 (2019) 9758–9767. doi: 10.1039/C9SC02281C

  39. [39]

      Q. Lin, A. Wang, S. Liu, et al., Chem. Commun. 56 (2020) 5303–5306. doi: 10.1039/D0CC01231A

  40. [40]

      X. Han, Y. Jiang, S. Li, et al., Nanoscale 11 (2019) 339–347. doi: 10.1039/C8NR05546G

  41. [41]

      C. Xing, Z. Chen, Y. Lin, et al., Chem. Commun. 57 (2021) 3251–3254. doi: 10.1039/D0CC08172H

  42. [42]

      J. Su, F. Wu, H. Xia, et al., Chem. Sci. 11 (2020) 80–86. doi: 10.1039/C9SC04823E

  43. [43]

      X. Zheng, R. Peng, X. Jiang, et al., Anal. Chem. 89 (2017) 10941–10947. doi: 10.1021/acs.analchem.7b02763

  44. [44]

      N. Li, M. Wang, X. Gao, et al., Anal. Chem. 89 (2017) 6670–6677. doi: 10.1021/acs.analchem.7b00889

  45. [45]

      J.W. Keum, H. Bermudez, Chem. Commun. 45 (2009) 7036–7038.

  46. [46]

      M.W. Bowler, M.G. Montgomery, A.G.W. Leslie, et al., Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 8646–8649. doi: 10.1073/pnas.0602915103

  47. [47]

      E.J. Griffiths, G.A. Rutter, Biochim. Biophys. Acta 1787 (2009) 1324–1333. doi: 10.1016/j.bbabio.2009.01.019

  48. [48]

      S. Hong, X. Zhang, R.J. Lake, et al., Chem. Sci. 11 (2020) 713–720. doi: 10.1039/C9SC04773E

• 

  Scheme 1  (A) Construction and (B) schematic illustration of the TD probe for ratiometric fluorescence imaging of ATP in living cells.

  下载: 全尺寸图片 幻灯片
  

  Figure 1  (A) PAGE characterization of the TD probe. Lane 1: S1; lane 2: S1+S2; lane 3: S1+S2+S3; lane 4: S1+S2+S3+S4; lane 4: S1+S2+S3+S4+Apt-a+Apt-b. (B) DLS analysis of the DNA tetrahedron and TD probe.

  下载: 全尺寸图片 幻灯片
  

  Figure 2  (A) Fluorescence spectra of the TD probe in response to different concentrations of ATP. (B) The relationship between the FRET signal (F_A/F_D) and ATP concentration. The inset shows the linear relationship from 0.02 to 0.2 mmol/L. Fluorescence spectra (C) and FRET signal (D) of the TD probe in response to 2 mmol/L ATP, CTP, GTP and UTP, respectively.

  下载: 全尺寸图片 幻灯片
  

  Figure 3  (A) PAGE analysis of the low concentration of DNase I (0.25 U/mL) and high concentration of DNase I (2.5 U/mL) degradation assay products for the TD probe. (B) FBS (10%, v/v) degradation assay products for the TD probe and ssDNA at 37 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 4  Confocal fluorescence imaging of HeLa cells incubated with 200 nmol/L TD probe (top) and ssDNA (bottom) for 4 h at 37 ℃.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Confocal fluorescence imaging of HeLa cells treated with medium (top), 10 µmol/L oligomycin (middle), and 5 mmol/L Ca²⁺ (bottom), followed by incubation with 200 nmol/L TD probe for 4 h at 37 ℃.

  下载: 全尺寸图片 幻灯片
•