Introducing 12 new dyes for use with oligonucleotide functionalised silver nanoparticles for DNA detection with SERS

Abstract

Oligonucleotide functionalised metallic nanoparticles (MNPs) have been shown to be an effective tool in the detection of disease-specific DNA and have been employed in a number of diagnostic assays. The MNPs are also capable of facilitating surface enhanced Raman scattering (SERS) enabling detection to become highly sensitive. Herein we demonstrate the expansion of the range of specific SERS-active oligonucleotide MNPs through the use of 12 new Raman-active monomethine and trimethine chalcogenopyrylium and benzochalcogenopyrylium derivatives. This has resulted in an increased ability to carry out multiplexed analysis beyond the current small pool of resonant and non-resonant Raman active molecules, that have been used with oligonucleotide functionalised nanoparticles. Each dye examined here contains a variation of sulphur and selenium atoms in the heterocyclic core, together with phenyl, 2-thienyl, or 2-selenophenyl substituents on the 2,2',6, and 6' positions of the chalcogenopyrylium dyes and 2 and 2' positions of the benzochalcogenopyrylium dyes. The intensity of SERS obtained from each dye upon conjugate hybridisation with a complementary single stranded piece of DNA was explored. Differing concentrations of each dye (1000, 3000, 5000 and 7000 equivalents per NP-DNA conjugate) were used to understand the effects of Raman reporter coating on the overall Raman intensity. It was discovered that dye concentration did not affect the target/control ratio, which remained relatively constant throughout and that a lower concentration of Raman reporter was favourable in order to avoid NP instability. A relationship between the dye structure and SERS intensity was discovered, leaving scope for future development of specific dyes containing substituents favourable for discrimination in a multiplex by SERS. Methine dyes containing S and Se in the backbone and at least 2 phenyls as substituents give the highest SERS signal following DNA-induced aggregation. Principal component analysis (PCA) was performed on the data to show differentiation between the dye classes and highlight possible future multiplexing capabilities of the 12 investigated dyes.

Fig. 1. Schematic representation of each possible orientation of the oligonucleotide-NPs conjugates when hybridised to a complementary target sequence: (a) head to tail; (b) tail to tail; (c) head to head. A schematic representation of hybridisation of the conjugates in the presence of the target DNA is shown in (d).

Fig. 2. Dye structures.

Scheme 1. Synthesis of benzochalcogenopyrylium dyes 10 and 12: (a) synthesis of aldehydes 14-S and 14-Se; (b) synthesis of benzoselenopyrylium dye 10; (c) synthesis of benzothiopyrylium dye 12.

Fig. 3. Examples of the three behaviours previously described. SERS spectra were recorded using an excitation wavelength of 532 nm, 1 1 s exposure time, 1 accumulation. The spectra obtained adding target and control to the NPs-DNA conjugates coated with 5000 equivalents of dye are compared using dye 10, 1 and 9 respectively. Raman intensity vs. concentration of dye are described showing a linear trend for dye 10, a plateau at high concentrations of dye meaning the saturation of the NPs for dye 1 and a maximum followed by a decrease in intensity for dye 9. The error bars reported are the standard deviations calculated for the 4 replicated of the analyses.

Fig. 4. PCA plot demonstrating the discrimination of the 12 dyes into three classes. Monomethine dyes containing two phenyl groups as substituents (3–5 and 9) are shown in blue, methane methine dyes with four 2-thienyl or four 2-selenophenyl rings as substituents (1, 2, 8 and 11) are shown in green and trimethine dyes (6, 7, 10 and 12) are shown in red.