IEEE JOURNAL OF SELECTED TOPICS IN QUANTUM ELECTRONICS, VOL. 11, NO. 4, JULY/AUGUST 2005 873 Distribution Analysis of the Photon Correlation Spectroscopy of Discrete Numbers of Dye Molecules Conjugated to DNA Samantha Fore, Ted A. Laurence, Yin Yeh, Rod Balhorn, Christopher W. Hollars, Monique Cosman, and Thomas Huser Abstract—The formation of protein complexes with other pro- teins and nucleic acids is critical to biological function. Although it is relatively easy to identify the components present in these com- plexes, it is often difficult to determine their exact stoichiometry and obtain information about the homogeneity of the sample from bulk measurements. We demonstrate the use of single molecule photon-pair correlation spectroscopy to distinguish between dis- crete numbers of molecules in biological complexes. Fluorescence photon antibunching is observed from a single molecule by employ- ing time-correlated single photon counting in combination with a Hanbury-Brown and Twiss coincidence setup. In addition, pulsed laser excitation and time-tagged time-resolved data collection al- low for the measurement of photon arrival times with nanosecond time resolution. The interphoton time distribution between consec- utively arriving photons can be calculated and provides a measure of the second-order temporal correlation function. Analysis of this function yields an absolute measure of the number of molecules, N , present in a given complex. It is this ability to measure N that renders this technique powerful for determining stoichiometries in complex biological systems at the single molecule level. We investi- gate the counting efficiency and statistics of photon antibunching of specifically designed biological samples labeled with multiple copies of the same fluorescent dye and derive conclusions about its use in the analytical evaluation of complex biological samples. Index Terms—DNA-protein complexes, Hanbury-Brown and Twiss, photon antibunching, protein complexes, single molecule fluorescence, time-correlated single photon counting. I. INTRODUCTION D NA-PROTEIN interactions play a critical role in DNA damage repair. The nucleotide excision repair (NER) path- way is the major pathway for the repair of bulky DNA lesions and involves at least two dozen proteins that have been identi- fied. [1], [2] However, the role these proteins play and the order Manuscript received January 7, 2005; revised August 3, 2005. This work was supported by funding from the National Science Foundation (NSF). The work of S. Fore was supported by the student employee graduate fellowship program at Lawrence Livermore National Laboratory (LLNL). The Center for Biopho- tonics, an NSF Science and Technology Center, is managed by the University of California, Davis, under Cooperative Agreement PHY 0120999. Work at LLNL was performed under the auspices of the U.S. Department of Energy by the University of California, Lawrence Livermore National Laboratory, under Contract W-7405-Eng-48. S. Fore, T. A. Laurence, R. Balhorn, C. W. Hollars, M. Cosman, and T. Huser are with Lawrence Livermore National Laboratory, Livermore, CA 94550 USA, and also with the NSF Center for Biophotonics Science and Technol- ogy, University of California-Davis, Sacramento, CA 95817 USA (e-mail: huser1@llnl.gov). Y. Yeh is with the Department of Applied Science, University of California, Davis, and also with the NSF Center for Biophotonics Science and Technology, University of California-Davis, Sacramento, CA 95817 USA. Digital Object Identifier 10.1109/JSTQE.2005.857738 in which they interact is poorly understood and highly debated. In fact, even the identity of some of the proteins, particularly the DNA damage recognition proteins, has become controver- sial [3]–[6]. Several of the proteins are known to form com- plexes, but their individual stoichiometries in these complexes are not well characterized [7]. The challenge here would be the ability to measure these stoichiometries at all times during the repair process with the possibility to distinguish the numbers of proteins in a given complex bound to DNA. Single molecule techniques have revolutionized the ability to probe these ques- tions and to unmask the heterogeneity that is not accessible in bulk measurements. Since the mid-1990s, advances in single molecule mi- croscopy and spectroscopy techniques have made it possible to study the structure and interactions of complex biological systems one molecule or complex at a time [8]. Several single molecule methods have been adapted to studying biological systems, and indeed many of them have inherent properties that lend themselves ideal for studying biomolecular mechanisms. The more traditional atomic force microscopy (AFM) and transmission electron microscopy (TEM) have been skillfully applied to imaging single molecules of DNA, proteins, and DNA-protein complexes. However, AFM is typically slow and chemically insensitive, and it is difficult to conduct experiments in situ/in vitro, requiring specially prepared probe tips (i.e., carbon nanotube tips) to resolve these complexes [9]. Basically, it is an inherently static technique at these resolutions. TEM is also inherently static and is ultimately destructive to the sample. Optical microscopies, although limited in their spatial resolution compared with the other techniques, offer a solution to their static nature. In particular, innovations in single molecule fluorescence microscopy and spectroscopy have enabled measurements on the dynamics of interactions [e.g., by employing fluorescence correlation spectroscopy (FCS)], and have provided structural information despite the diffraction limit of optical light [e.g., through single molecule fluorescence resonance energy transfer (FRET)] [8], [10]. Through the introduction of novel, more photostable fluorescent dyes, interactions of single, fluorescently tagged biological molecules can now be monitored using what have become standard fluorescence microscopy and spectroscopy techniques. FCS, for example, yields information on the mobility, concentrations, and reaction kinetics of biological molecules in systems. However, it typically accesses rather long time scales (∼μs to ms) and requires averaging more than ∼10 6 1077-260X/$20.00 © 2005 IEEE