Long genomic DNA amplicons adsorption onto unmodified gold nanoparticles for colorimetric detection of Bacillus anthracis

Abstract

Unmodified gold nanoparticles (GNPs) can be wrapped with long genomic single- and double-stranded DNA (ssDNA and dsDNA) molecules produced by asymmetric polymerase chain reaction (As-PCR). More importantly, the DNA–Au interaction can be utilized for colorimetric detection of a specific nucleic acid sequence in clinical samples.

---

Gold nanoparticles (GNPs) as biosensors for molecular recognition and nucleic acid detection have attracted considerable attention in recent decades because of their unique optical and physicochemical properties. Typically, the colloidal gold (15 nm) appears ruby-red and exhibits a maximum absorbance wavelength around 520 nm in the UV-visible spectrum. Whereas, the color turns to blue-purple in the solution containing aggregated nanoparticles corresponding to the absorbance peak of GNPs shifting to longer wavelengths (red-shift). The dramatic change in color of gold colloid can be easily monitored with the naked eye which provides a very simple and cost-effective methodology in biodiagnostics. However, most of the previous assays require covalent modification of GNPs in advance or additional instruments for implementation with practical limitations.^1–4

Recently, Li and Rothberg presented an interesting phenomenon that single- and double-stranded DNA (ssDNA and dsDNA) have different affinities to negatively charged GNPs. More importantly, ssDNA electrostatically adsorbed onto naked GNPs stabilizes them against salt-induced aggregation, whereas dsDNA does not.⁵ In addition, the rate of adsorption onto the GNPs surface was related to sequence length of ssDNA and reaction temperature. Shorter oligonucleotides (oligo) bind to GNPs faster and higher temperature speeds up this process.⁶ The essential difference in adsorption behavior between ssDNA and dsDNA can be ascribed to their different electrostatic properties. Since ssDNA is sufficiently flexible and uncoils in solution, exposed bases in ssDNA stick to nanoparticles via van der Waals attraction. In contrast, dsDNA with double-helix structure is more stable and rigid, thereby failing to expose bases. The repulsion between the negatively charged phosphate backbone of dsDNA and the negative ions of the GNPs surface prevents the interaction. More recently, however, an alternative explanation was proposed that a strong hydrophobic effect may be involved in the ssDNA–Au interaction.⁷ More studies are necessary that may help us to have a deeper understanding of this complex process. Other studies have investigated the binding affinity of various DNA bases or nucleotides (nt) to GNPs.^8–11 Nevertheless, all of the employed ssDNA in previous studies were synthetic short fragments (5–50 nt), or even single DNA bases such as A/T/G/C. What about genomic ssDNA (>100 nt) with completely random sequences of bases interaction with GNPs? There are no investigations to address this issue yet.

It was reported that long genomic ssDNA can wrap around carbon nanotubes and improve their solubility.^12,13 On the basis of these studies, we speculate that the large ssDNA fragment may also adsorb onto GNPs. Since larger amplicons with sizes of several hundred base pairs (bp) are routinely used in many PCR-based diagnostic protocols, we hypothesize that it might be possible to use such genomic ssDNA–Au interaction for identification of targeted DNA sequences. In our study, a novel sensing strategy was proposed for colorimetric detection of a specific nucleic acid sequence. Asymmetric polymerase chain reaction (As-PCR) generates genomic ssDNA with biologically genuine sequences and expected sizes from clinical samples. In the presence of a targeted template, successful As-PCR produces a large amount of amplified ssDNA. These ssDNA amplicons wrap around GNPs and stabilize nanoparticles against salt-induced aggregation, the red color of gold colloid remains unchanged. Conversely, in the negative control (NC), GNPs rapidly aggregate in salt solution in the absence of ssDNA products. The aggregation of GNPs corresponding to a red-to-blue color shift serves as a simple indicator for specific DNA sequence detection (Scheme 1).

Scheme 1 Schematic illustration of the colorimetric detection of the DNA sequence with naked GNPs. The genomic DNA serves as a template for As-PCR generating ssDNA–dsDNA mixtures. Dispersed nanoparticles remain red in color in salt solution after amplicons adsorption. Non-target control fails to produce amplified products, aggregated nanoparticles turn to blue-gray upon adding NaCl.

In our preliminary assays, a synthetic oligo (71 nt) was found to protect GNPs against aggregation when NaCl was introduced, which inspired us to try longer fragments (>100 nt) for further experiments. To get longer ssDNA fragments, we applied a simple but effective method, As-PCR, in this study. As a proof-of-concept, this system was used to identify B. anthracis in clinical samples. According to previous studies on B. anthracis genome,^14,15 we designed primer sets to amplify specific sequences with the size of 116 nt, 242 nt, 345 nt and 508 nt (Fig. 1A and D). The primers for B. anthracis fragments amplification are listed in Table 1.

Fig. 1 (A) Agarose gel electrophoresis of the As-PCR product. A series of primer ratios were optimized to produce more ssDNA amplicons. (B) Colorimetric detection of B. anthracis. After simple and quick column purification to remove primers and salt, the As-PCR product (lane 5, 100 : 1) was incubated with gold colloid (15 nm spherical GNPs, BBInternational) at RT for 10 min followed by NaCl addition (50 mM final). (C) UV-visible spectral analysis of the reaction mixture. (D) Electrophoresis analysis of other amplified fragments by As-PCR with various sizes (242 nt/bp, 345 nt/bp, 508 nt/bp).

Table 1 Primer sets used in this study

	Sequence (5′–3′)	Length (nt)	Product size (nt)
116F	aga taa atg cgt aag gac aa	20	116
116R	aca tag aag gac gat aca gac	21	116
242F	taa atg tct gta tcg tcc ttc	21	242
242R	ctg tgg gtg tac ctt tgg	18	242
345F	aag cta gat aaa tgc gta agg	21	345
345R	ttt agc cgc tgt ggg tgt	18	345
508F	cgc aag ttg aat agc aag c	19	508
508R	gat acc agg atg ggt ctc g	19	508

---

As-PCR is an economical and simply executable route to produce ssDNA targets, and it often requires unbalanced primer pairs in the mixture.^16–18Fig. 1A shows the electrophoresis of the As-PCR product of expected size (116 nt). The amplified product after brief purification (10 μl) was mixed with GNPs (10 μl, 20 nM) at room temperature (RT) for 10 min followed by the addition of NaCl. A marked color change from ruby-red to blue-purple immediately occurred in NC, whereas the original red color was retained in experiments with the 116 nt product from the B. anthracis sample (Fig. 1B). The 116 nt ssDNA amplicons with random coil structure could effectively bind to GNPs via coordination between the nitrogen atoms of exposed bases and GNPs, thus rendering more negative charges to GNPs. This process enhances the electrostatic repulsion of GNPs and stabilizes them in salt solution. To confirm this optical change of GNPs, a characteristic broadening and red-shifting of the surface plasmon resonance (SPR) adsorption band only occurring in the NC was recorded on a UV-vis spectrophotometer (Fig. 1C).

As shown in Fig. 1A, the As-PCR product (lane 5) actually contained major ssDNA and minor dsDNA amplicons. To examine which fraction contributed to nanogold stability, a 116 nt fragment was synthesized to get pure ssDNA, conventional PCR followed by simple purification to get pure dsDNA amplicons and PCR without a DNA template denoted NC was performed, respectively. As little as 0.1 picomole (pmol) ssDNA (116 nt) can prevent GNPs from aggregation. A slight color change consistent with a small spectral shift (average 3.7 nm) is shown in Fig. 2 when using the same molar amount of dsDNA (116 bp). More quantity (∼0.3 pmol) of 116 bp dsDNA has a similar effect to that of 116 nt ssDNA. We also validated our observation with other dsDNA fragments, 345 bp and 508 bp dsDNA amplicons stabilized unmodified GNPs against aggregation. These results support that both ssDNA and dsDNA can stabilize nanogold in high salt solution (50 mM NaCl), and the ssDNA works better than dsDNA. Xia and Plaxco et al. reported their finding that short dsDNA (<40 bp) adsorbed onto GNPs and prevented aggregation of GNPs at low ionic strength (<12 mM NaCl).¹⁹ To our surprise, the long dsDNA (>100 bp) functions as a protectant against NaCl-induced aggregation of GNPs. Ion-induced dipole dispersive interactions might be a possible explanation for dsDNA–Au binding.²⁰ As previously described, the coordination interaction between DNA and GNPs is quite complicated.²¹ Actually, this process is still poorly understood. Since linear amplification of As-PCR can minimize the potential carry-over of PCR products. In addition, ssDNA plays a more effective role in protecting GNPs against aggregation. Taken together, we applied As-PCR to produce ssDNA–dsDNA mixtures in this study.

Fig. 2 (A) Photograph of gold colloid after incubation with 116 nt ssDNA, 116 bp dsDNA and NC followed by addition of salt. (B) UV-visible spectral analysis of the reaction mixture.

Furthermore, sensitivity and specificity of this approach were also examined. The starting genomic DNA template (1 ng, 0.1 ng, 10 pg and 1 pg) of B. anthracis was used to produce amplicons with a size of 508 nt/bp by using As-PCR. Obviously, as little as 10 pg of initial genomic DNA is identifiable without visual aids suggesting the picogram detection level of this method (Fig. 3A). Reaction mixtures containing the B. anthracis sample with specific amplicons retained red color. However, the solution color of three closely-related species of B. anthracis remained purple-gray even after loading 5 or 10 times (50–100 ng) more genomic DNA templates (Fig. 3B). Herein, the amplified ssDNA–dsDNA with various fragments (116 nt/bp, 242 nt/bp, 345 nt/bp, 508 nt/bp) worked quite well, which cover the commonly used sizes of PCR products. These results clearly support our hypothesis and indicate the practical applicability of this approach.

Fig. 3 (A) Initial genomic DNA (1 ng–1 pg) of B. anthracis was used to run As-PCR for colorimetric detection as mentioned before. Correlation of starting template DNA quantities and maximum absorbance shift with triple replicates. (B) B. anthracis (B. a), B. cereus (B. c), B. subtilis (B. s) and Escherichia coli (E. coli) samples were employed in this selectivity test by visual inspection, UV-vis spectral analysis and gel electrophoresis.

In summary, we demonstrated that the long genomic DNA amplicons with biologically genuine sequences bind to unmodified GNPs enabling colorimetric detection of B. anthracis. Naked GNPs were used here to obviate the need for complicated, time-consuming modification and labelling of GNPs. In practice, targeted ssDNA–dsDNA can be simply prepared by taking on the selectivity and sensitivity of As-PCR. To the best of our knowledge, this is the first report of long genomic ssDNA–dsDNA adsorption onto GNPs with potential clinical application. Here, we focused on detecting B. anthracis as an example and this novel approach would be universal for identification of any DNA sequence.

This work was supported by grants from the Natural Science Foundation of China (No. 30970784 and 81171455), National Key Basic Research Program of China (2009CB930200), Chinese Academy of Sciences (CAS) “Hundred Talents Program” (07165111ZX) and the CAS Knowledge Innovation Program.

Notes and references

1. R. K. Jhansi, A. Tahir, A. K. Sadia, N. Adria, C. Perry, V. Birsen, W. Marla, M. Shardae, R. Britinia, B. Santanu, K. S. Anant, S. Dulal and C. R. Paresh, Angew. Chem., Int. Ed., 2009, 121, 9848–9851.
2. P. V. Baptista, M. K. Montewka, J. P. Oles, G. Doria and R. Franco, Clin. Chem., 2006, 52, 1433–1434.
3. J. Aveyard, M. Mehrabi, A. Cossins, H. Braven and R. Wilson, Chem. Commun., 2007, 4251–4253.
4. T. Liu, J. Zhao, D. M. Zhang and G. X. Li, Anal. Chem., 2010, 82, 229–233.
5. H. X. Li and L. Rothberg, Proc. Natl. Acad. Sci. U. S. A., 2004, 101, 14036–14039.
6. H. X. Li and L. Rothberg, J. Am. Chem. Soc., 2004, 126, 10958–10961.
7. E. M. Nelson and L. J. Rothberg, Langmuir, 2011, 27, 1770–1777.
8. J. J. Storhoff, R. Elghanian, C. A. Mirkin and R. L. Letsinger, Langmuir, 2002, 18, 6666–6670.
9. A. Gourishankar, S. Shukla, K. N. Ganesh and M. Sastry, J. Am. Chem. Soc., 2004, 126, 13186–13187.
10. L. M. Demers, M. Ostblom, H. Zhang, N. H. Jang, B. Liedberg and C. A. Mirkin, J. Am. Chem. Soc., 2002, 124, 11248–11249.
11. H. K. Suda, D. Y. Petrovykh, M. J. Tarlov and L. J. Whitman, J. Am. Chem. Soc., 2003, 125, 9014–9015.
12. W. A. Zhao, Y. Gao, M. A. Brook and Y. F. Li, Chem. Commun., 2006, 3582–3584.
13. B. Gigliotti, B. Sakizzie, D. S. Bethune, R. M. Shelby and J. N. Cha, Nano Lett., 2006, 6, 159–164.
14. L. Radnedge, P. G. Agron, K. K. Hill, P. J. Jackson, L. O. Ticknor, P. Keim and G. L. Andersen, Appl. Environ. Microbiol., 2003, 69, 2755–2764.
15. M. H. Antwerpen, P. Zimmermann, K. Bewley, D. Frangoulidis and H. Meyer, Mol. Cell. Probes, 2008, 22, 313–315.
16. H. Mao and K. S. Rosenthal, BioTechniques, 1999, 27, 674–678.
17. S. K. Poddar, Mol. Cell. Probes, 2000, 14, 25–32.
18. H. Deng, Y. Xu, Y. H. Liu, Z. J. Che, H. L. Guo, S. X. Shan, Y. Sun, X. F. Liu, K. Y. Huang, X. W. Ma, Y. Wu and X. J. Liang, Anal. Chem., 2012, 84, 1253–1258.
19. F. Xia, X. Zuo, R. Yang, Y. Xiao, D. Kang, A. Vallée-Bélisle, X. Gong, J. D. Yuen, B. B. Y. Hsu, A. J. Heeger and K. W. Plaxco, Proc. Natl. Acad. Sci. U. S. A., 2010, 107, 10837–10841.
20. P. Sandstrom, M. Boncheva and B. Akerman, Langmuir, 2003, 19, 7537–7543.
21. H. Li, J. Huang, J. An, X. Zhang, Z. Zhang, C. Fan and J. Hu, Angew. Chem., Int. Ed., 2005, 44, 5100–5103.

---