Published: February 28, 2011 r2011 American Chemical Society 2476 dx.doi.org/10.1021/ac1026034 | Anal. Chem. 2011, 83, 24762484 ARTICLE pubs.acs.org/ac Temperature Modulation and Quadrature Detection for Selective Titration of Two-State Exchanging Reactants K. Zrelli, T. Barilero, E. Cavatore, H. Berthoumieux, T. Le Saux,* , V. Croquette, A. Lemarchand,* ,§ C. Gosse,* ,|| and L. Jullien* , Departement de Chimie, Ecole Normale Superieure, UMR CNRS-ENS-UPMC Paris 06 8640 Pasteur, 24 rue Lhomond, 75231 Paris Cedex 05, France Laboratoire de Physique Statistique, Departement de Physique, Ecole Normale Superieure, UMR CNRS 8550, 24 rue Lhomond, 75231 Paris Cedex 05, France § Laboratoire de Physique Theorique de la Mati ere Condensee, Universit e Pierre et Marie Curie Paris06, UMR CNRS-UPMC 7600, 4 place Jussieu, 75252 Paris Cedex 05, France ) Laboratoire de Photonique et de Nanostructures, LPN-CNRS, route de Nozay, 91460 Marcoussis, France b S Supporting Information ABSTRACT: Biological samples exhibit huge molecular diversity over large concentration ranges. Titrating a given compound in such mixtures is often dicult, and innovative strategies emphasizing selectivity are thus demanded. To overcome limitations inherent to thermodynamics, we here present a generic technique where discrimination relies on the dynamics of interaction between the target of interest and a probe introduced in excess. Considering an ensemble of two-state exchanging reactants submitted to temperature modulation, we rst demonstrate that the amplitude of the out-of-phase concentration oscillations is maximum for every compound involved in a reaction whose equilibrium constant is equal to unity and whose relaxation time is equal to the inverse of the excitation angular frequency. Taking advantage of this feature, we next devise a highly specic detection protocol and validate it using a microfabricated resistive heater and an epiuorescence microscope, as well as labeled oligonucleotides to model species displaying various dynamic properties. As expected, quantication of a sought for strand is obtained even if interfering reagents are present in similar amounts. Moreover, our approach does not require any separation and is compatible with imaging. It could then benet some of the numerous binding assays performed every day in life sciences. INTRODUCTION Quantifying a given species in a complex medium of living origin is currently encountered in many medical and pharma- ceutical activities such as molecular diagnosis 1,2 and drug screening. 3 Moreover, extensive characterization of the cellular content is the workhorse of scientic elds such as genomics, proteomics, and metabolomics. Yet these analytical tasks represent a tremendous challenge, as crude biological samples typically contain 10 4 -10 6 components of the same kind (e.g., RNAs, proteins) at concentrations covering up to 10 orders of magnitude. 1,4,5 Improvement in assay selectivity is therefore required to avoid as much as possible the cumbersome pre- treatments today necessary prior to any specic detection. Directly countingintrinsic labels such as functional groups is not appropriate to investigate the considered mixtures because they are mostly made of biopolymers that share the same building blocks in the same proportions. In fact, this type of approach results in spectral crowding, and unambiguous species identication can be achieved only through coupling with methods emphasizing physicochemical criteria (e.g., separation for mass spectrometry 6 or correlation analysis for IR absorption 7,8 ). Besides such complicated multidimensional analyses, titration is then a valuable strategy because one just has to observe a single reagent selectively probing the compound of interest. Furthermore, in the present context there it benets from the high specicity that characterizes molecular recognition in biology. 9 Binding assays, which rely on thermodynamics for discrimina- tion, are actually widespread. Whereas mix-and-readtests are homogeneous and consequently convenient, the advent of micro- arrays has yielded unprecedented multiplexing capabilities, with thousands of titrations being performed in parallel. 10-14 How- ever, species concentrations are often overestimated because label- ing reagents are insuciently rinsed away or because nonspecic adsorption onto solid supports is poorly controlled. 9,12,14,15 Addi- tionally, in a mixture of similar components, a single probe is anticipated to display close anities for several targets. 11 Thus, Received: October 2, 2010 Accepted: January 16, 2011