Article pubs.acs.org/Biomac
Measuring in Situ Primary and Competitive DNA Hybridization Activity on Microspheres James O. Hardin† and Valeria T. Milam*,†,‡,§ †
School of Materials Science and Engineering, ‡Wallace C. Coulter Department of Biomedical Engineering, and §Petit Institute for Bioengineering and Bioscience, Georgia Institute of Technology, 771 Ferst Drive NW, Atlanta, Georgia 30332-0245, United States S Supporting Information *
ABSTRACT: Microspheres serve as convenient substrates for studying DNA activity on surfaces. Here, in addition to employing conventional sample preparation involving multiple wash and resuspension steps prior to flow cytometry measurements, we also directly sampled the reaction volume to acquire in situ measurements of primary and competitive hybridization events. Even in the absence of post hybridization wash steps, nonspecific binding events were negligible and thus allowed for direct, quantitative assessment of hybridization events as they occurred on colloidal surfaces. The in situ results indicate that primary duplex formation between immobilized probes and soluble targets on microsphere surfaces is less favorable than predicted by solution models. The kinetics of competitive displacement of primary hybridization partners by secondary targets measured in situ or post washing also deviate from expectations based on theoretical solution thermodynamics, but are consistent with predicted kinetic trends stemming from differences in either the toehold base length or branch migration.
■
INTRODUCTION DNA has emerged as a key player in numerous schemes ranging from planar detection platforms1−3 to colloid-based therapeutic4,5 and assembly6−13 schemes. For any of these scenarios, optimizing a design strategy that employs DNA as a macromolecular linker or as a displacement agent requires an understanding of its hybridization activity. Although oligonucleotide solutions are well studied yielding a functional thermodynamic model,14 hybridization kinetics at surfaces has been examined with limited DNA sequences on select surfaces such as planar gold.15−19 The association rate constants for primary hybridization at flat surfaces are reportedly ∼104 M−1 s−1 or less, depending on base length and composition20,21 with comparable rate constants reported on microsphere22 surfaces. In oligonucleotide solutions, on the other hand, primary duplex formation is reportedly faster23−25 (∼106 M−1 s−1) and largely independent of composition and length26,27 of the hybridization partners. Separate studies indicate that hybridization activity between DNA-functionalized gold nanoparticles can be less 28 or more 29 thermodynamically favorable than in oligonucleotide solutions. Notably, while solution kinetic models typically use fits identical in form to a first-order Langmuir equation,23,30 analytical approaches for surfaceimmobilized oligonucleotides range from employing a second-order Langmuir equation31 inferring significant interaction between binding sites to biexponential fits22 involving two types of binding sites. Considering the diverse reports for surface studies, solution studies involving unperturbed duplex formation between soluble strands help elucidate surface-induced deviations in © 2013 American Chemical Society
hybridization activity. In solution, primary hybridization or duplex formation between single-stranded DNA begins with the formation of a stable nucleus of ∼3 base pairs.25 Hybridization then propagates at a rate of ∼106−107 base pairs per second as the duplex “zips up”.32 This fast propagation rate implies that the lifetime of most intermediates or partially zipped duplexes is too short to measure with most experimental methods, but hybridization activity itself can be approximated with a twostate model33 in which strands are either completely hybridized or dissociated. Compared to primary hybridization, competitive DNA hybridization involving the replacement of one partner stand in a duplex by another partner strand has received less attention, but two experimental scenarios have been examined for immobilized duplexes. In the first scenario, multiple different DNA targets are simultaneously introduced to probe strands. This scenario is relevant to oligonucleotide detection scenarios involving various target candidates.34,35 The second scenario involves the two distinct hybridization stages depicted in Figure 1 in which single-stranded probes are first exposed to multiple copies of a DNA target to allow primary hybridization events to occur (i.e., target strands hybridize to unoccupied probes with unhybridized target typically removed through washing steps). In the second stage, secondary targets are introduced and allowed to compete with the original binding partner or hybridized primary target. Received: November 9, 2012 Revised: February 7, 2013 Published: February 12, 2013 986
dx.doi.org/10.1021/bm3017466 | Biomacromolecules 2013, 14, 986−992
Biomacromolecules
Article
toeholds can impact competitive displacement rates by increasing the stability of the secondary duplex nucleus. Competitive hybridization involving immobilized sequences has recently served as a route to achieve DNA-mediated colloidal assembly and disassembly under isothermal conditions.45 In our prior work, the extent and rate of competitiondriven particle release from oligonucleotide-linked colloidal assemblies was tuned through choice in sequence composition, length, and concentration.44,46−49 In this work and previous work, washing steps prior to flow cytometry analysis were performed to remove unassociated, weakly adsorbed, or displaced target strands. While primary duplexes appeared stable after multiple washes in our prior studies,44 this assumption is not necessarily valid for all sequences, especially for weak binding partners.43 To evaluate this assumption and provide greater temporal resolution in the current work, the primary and competitive hybridization reaction volumes were also directly sampled as hybridization progressed.
■
Figure 1. Schematic diagram of hybridization pathways and associated rate constants examined in this work. Only one immobilized probe and one relevant target species are depicted here for clarity. The rate constant for primary hybridization between initially unoccupied probes and soluble primary targets is k1. Following the addition of secondary targets, two pathways for secondary duplex formation are indicated. In the displacement pathway, k2 is the rate constant for formation of intermediate complexes between probes and both primary and secondary targets. Following the formation of complexes, each base in the primary duplex is sequentially replaced by a base in the secondary target until the primary target is fully displaced at a rate constant of k3. In the dissociative pathway, k4 is the rate constant for complete dissociation of primary duplexes, and k5 is the rate constant for duplex formation between the now unoccupied probes and secondary targets. Fluorescent labels on the primary target are indicated by a green star. Unpaired bases in the primary duplex are shown in gray. This single stranded or toehold region of the immobilized probe is complementary to the green segment of the secondary target. Due to experimental conditions involving weaker primary targets and an excess concentration of stronger secondary targets [T2], all reactions are assumed to be biased in the forward direction as indicated by arrows. Similar to prior approaches,39,43,44 the next step in each pathway following primary duplex formation is assumed to be the rate-limiting step (i.e., k2 [T2] ≫ k3; k4 ≫ k5 [T2]).
EXPERIMENTAL SECTION
DNA Sequences. Sequences (Integrated DNA Technologies, Coralville, IA) listed in Table 1 were based on previous work.44,48
Table 1. List of the Function and Nomenclature for DNA Sequences Studieda function probe NC target perfectly matched targets
mismatched targets
sequence
ΔGhyb kcal/mol
A20= 5′-TTT TTT GGA TTG CGG CTG AT-3′ NC = 3′-GGA TTG CGG CTG AT-5′ P11= 3′-AAC GCC GAC TA-5′
NA −4.4 −15.3
P13= 3′-CT AAC GCC GAC TA-5′ P15=3′-A CCT AAC GCC GAC TA-5′ M11=3′-AAC GCG GAC TA-5′ M13=3′-CT AAC GGC GAC TA-5′ M15=3′-A CCT AAC CCC GAC TA-5′
−17.9 −20.9 −13.3 −13.9 −14.2
The theoretical change in Gibb’s free energy of hybridization, ΔGhyb, between single-stranded A20 probes and various target strands in 150 mM NaCl at 25 °C was determined using Zuker’s Mfold server for oligonucleotide solutions.50−53 Mismatches are indicated in bold, underlined text. Noncomplementary targets are indicated with “NC” abbreviation. a
Among relevant oligonucleotide solution studies,36−38 Reynaldo and co-workers39 developed a kinetics model for competitive hybridization in oligonucleotide solutions based on the dissociative and displacement pathways depicted on particles in Figure 1. The dissociative pathway involves complete dissociation of a primary duplex followed by hybridization between the original probe and a new target strand. The displacement pathway involves conversion of a primary duplex to an intermediate three-stranded complex prior to secondary duplex formation. Considering the fast displacement activity (
