Method of identifying hairpin DNA probes by partial fold analysis

#The present invention generally relates to the use of DNA hairpins as molecular beacon probes. More specifically, the present invention is directed to methods of generating highly specific and highly selective molecular beacon probes by using naturally occurring DNA hairpins present in organisms of interest.

#Methods for the rapid detection and serotyping of pathogens are of high interest, due in part to the dramatic improvement in treatment efficacy for a bacterial or viral infection diagnosed early relative to one diagnosed at a later stage (Inglesby et al., “Anthrax as a Biological Weapon: Medical and Public Health Management,” J. Am. Med. Assoc. 281:1735-1745 (1999)). Unfortunately, most current methods of pathogen identification rely on some level of sample manipulation, (enrichment, fluorescent tagging, etc.) which can be costly in terms of both time and money. Thus, eliminating sample labeling will result in a significant savings and has the potential to speed diagnosis. The use of DNA hairpins as “molecular beacons” (Broude, “Stem-loop Oligonucleotides: a Robust Tool for Molecular Biology and Biotechnology,” Trends Biotechnol. 20:249-256 (2002)), either in solution (Tyagi et al., “Molecular Beacons: Probes that Fluoresce upon Hybridization,” Nature Biotech. 19:365-370 (2001); Dubertret et al., “Single-mismatch Detection Using Gold-quenched Fluorescent Oligonucleotides,” Nature Biotech. 19:365-370 (2001)) or immobilized on a solid surface (Fang et al., “Designing a Novel Molecular Bacon for Surface-Immobilized DNA Hybridization Studies,” J. Am. Chem. Soc. 121:2921-2922 (1999); Wang et al., “Label Free Hybridization Detection of Single Nucleotide Mismatch by Immobilization of Molecular Beacons on Agorose Film,” Nucl. Acids. Res. 30:61 (2002); Du et al., “Hybridization-based Unquenching of DNA Hairpins on Au Surfaces: Prototypical “Molecular Beacon” Biosensors,” J. Am. Chem. Soc. 125:4012:4013 (2003); Fan et al., “Electrochemical Interrogation of Conformational Changes as a Reagentless Method for the Sequence-specific Detection of DNA,” Proc. Natl. Acad. Sci. USA 100:9134-9137 (2003)), has proven to be a useful method for “label-free” detection of oligonucleotides. Molecular beacons consist of DNA hairpins functionalized at one terminus with a fluorophore and at the other terminus with a quencher. In the absence of their complement, they exist in a closed, “dark” conformation. Hybridization occurs on introduction of complementary oligonucleotides, which concomitantly forces open the hairpin and allows for a fluorescent, “bright” state.

#Traditionally, as illustrated in FIG. 1, molecular beacons have been designed by supplementing the targeted DNA sequence at both termini with additional self-complementary nucleotides to force the formation of a hairpin (Monre et al., “Molecular Beacon Sequence Design Algorithm,” Biotechniques 34:68-73 (2003)). While generally successful, the addition of non target-derived nucleotides increases the potential for non-specific binding, thus potentially reducing both the sensitivity and selectivity of the probe beacon. Modifications of this discovery protocol, such as the “shared stem” methodology of Bao and coworkers (Tsourkas et al., “Structure-function Relationships of Shared-Stem and Conventional Molecular Beacons,” Nucl. Acids Res. 30:4208-4215 (2002)), still incorporate several bases unrelated to the target sequence. Thus, the latter approach potentially suffers from the same deficiencies. It would be desirable to identify a reliable approach for identifying DNA hairpins that overcomes the above-noted deficiencies.

#FIG. 1 shows the prior art method of DNA hairpin probe design demonstrating the section of non-pairing sequences present in the final complex.

#FIG. 2 shows the final probe/target complex of the present invention when the probe is selected based on total sequence complementarity.

#FIG. 3 shows the predicted secondary structure and hairpin regions selected from Bacillus anthracis. A partial gene sequence of the Bacillus anthracis pag gene (isolate IT-Carb3-6254) (Adone et al., J. Appl. Microbio. 92:1-5 (2002), which is hereby incorporated by reference in its entirety) was obtained from GenBank Accession AJ413936, which is hereby incorporated by reference in its entirety. The secondary structure of ˜1000 nucleotide fragment (SEQ ID NO: 1) of the aforementioned sequence was then computationally predicted using RNAstructure v. 3.7 (Mathews et al., J. Mol. Biol. 288:911-940 (1999), which is hereby incorporated by reference in its entirety).

#The method of the invention involves obtaining or providing a probe nucleotide sequence from a molecular target. The target nucleotide sequence can be sequenced from an isolated cDNA or obtained from an online database such as GenBank. Regardless of the source of the target nucleotide sequence, a partial fold analysis is performed on the nucleotide sequence using any of a variety of suitable folding software such as, e.g., RNAStructure program (available from D. Turner at the University of Rochester, Rochester, N.Y.), Mfold software package (available from M. Zucker at the Rensselear Polytechnic Institute, Rensselear, N.Y.), and Vienna RNA software package, including RNAfold, RNAeval, and RNAsubopt (available from I. Hofacker at the Institute for Theoretical Chemistry, Vienna, Austria). With respect to the RNAStructure program, applicants have discovered that segments larger than approximately 1000 bases would crash the program RNAstructure v. 3.7. Thus, it may or may not be possible to predict the secondary structure of an entire nucleic acid molecule depending on the length thereof. Ideally, the secondary structure of the entire sequence would be predicted, but as demonstrated in the examples that is not necessary.

#The resulting folded structure may or may not be the true active conformation of the RNA molecule in a cellular environment; however, it represents the lowest free energy state as predicted using such software. It is believed that more often than not, the predicted lowest free energy state of the nucleic acid molecule sufficiently resembles the true active conformation. Nonetheless, the resulting folded structure is analyzed to identify hairpin regions thereof.

#Having identified hairpin structures within the folded structure of the prospective target nucleic acid molecule, the hairpin sequences are isolated from the larger sequence (i.e., that was used as input to the folding software). The isolation can be performed in silico. Once isolated, the hairpin sequence is subjected to a second structural prediction as was performed on the prospective target nucleic acid molecule.