Colorimetric and dynamic light scattering detection of DNA sequences by using positivelycharged gold nanospheres: a comparative study with gold nanorods

Abstract

We introduce a new genosensing approach employing CTAB (cetyltrimethylammoniumbromide)-coated positively charged colloidal gold nanoparticles (GNPs) to detect targetDNA sequences by using absorption spectroscopy and dynamic light scattering. Theapproach is compared with a previously reported method employing unmodifiedCTAB-coated gold nanorods (GNRs). Both approaches are based on the observation thatwhereas the addition of probe and target ssDNA to CTAB-coated particles results inparticle aggregation, no aggregation is observed after addition of probe and nontarget DNAsequences. Our goal was to compare the feasibility and sensitivity of both methods. A21-mer ssDNA from the human immunodeficiency virus type 1 HIV-1 U5 long terminalrepeat (LTR) sequence and a 23-mer ssDNA from the Bacillus anthracis cryptic proteinand protective antigen precursor (pagA) genes were used as ssDNA models. In the case ofGNRs, unexpectedly, the colorimetric test failed with perfect cigar-like particles butcould be performed with dumbbell and dog-bone rods. By contrast, our approachwith cationic CTAB-coated GNPs is easy to implement and possesses excellentfeasibility with retention of comparable sensitivity—a 0.1 nM concentration oftarget cDNA can be detected with the naked eye and 10 pM by dynamic lightscattering (DLS) measurements. The specificity of our method is illustrated bysuccessful DLS detection of one–three base mismatches in cDNA sequences for bothDNA models. These results suggest that the cationic GNPs and DLS can beused for genosensing under optimal DNA hybridization conditions without anychemical modifications of the particle surface with ssDNA molecules and signalamplification. Finally, we discuss a more than two–three-order difference in the reportedestimations of the detection sensitivity of colorimetric methods (0.1 to 10–100 pM) toshow that the existing aggregation models are inconsistent with the detectionlimits of about 0.1–1 pM DNA and that other explanations should be developed.