Item

Abstract

Background: Frequency-domain fluorescence lifetime imaging microscopy (FLIM) is finding increasing use in the analysis of biological systems. However, the calibration, determination of resolvable lifetime differences, and evaluation of artifacts have not been extensively treated. We describe a multi-point method for calibrating a frequency-domain FLIM system, characterize the minimum detectable heterogeneity and intra- and inter-image lifetime differences, discuss the statistical treatment of FLIM data, and suggest methods for minimizing artifacts. Methods: A set of solutions exhibiting single-component lifetimes suffice for accurately calibrating a reference material with a single-component lifetime, even in the absence of accurate data on the lifetimes of the individual solutions or the reference material. We used a set of rhodamine 6G solutions quenched with varying concentrations of iodide, leading to lifetimes of 0.5-4.0 ns, to calibrate a 1 M reference solution of rhodamine 6G in water. Results: We measured a value of 4.11 ns with an estimated absolute error of ±0.05 ns for the rhodamine 6G reference solution. With 57.7 MHz modulation, the minimum detectable inter-image lifetime difference was 0.1-0.15 ns and the minimum detectable intra-image lifetime difference was 4-5 ps, allowing solutions differing in lifetime by 40 and 70 ps to be easily distinguished. The minimum detectable lifetime heterogeneity was 50-80 ps. Evaluation of replicate measurements of single solutions demonstrated that inter-image instrument errors exceeded those predicted from intra-image statistics by more than an order of magnitude. We also measured lifetimes and heterogeneity in 4 GFP variants (WTGFP, EGFP, S65T, and EYFP) with the technique. Conclusion: The multi-point calibration method is applicable to any system consisting of single-component lifetimes. Applying the method in our FLIM microscope allowed us to demonstrate a previously unreported degree of lifetime resolution in a FLIM microscope.