Abstract 3982: Super-resolution imaging of drug delivery to live cells using real-time structured illumination microscopy

Abstract

Abstract Imaging the effects of drug delivery in live cells is most often performed by confocal microscopy, which has a resolution limited by the diffraction of light. Here, we report the development of a novel method for imaging drug delivery to live cells using real-time super-resolution microscopy in wide-field mode. Structured illumination microscopy (SIM) is one of several super-resolving microscopy techniques and is capable of imaging beyond the resolution of conventional microscopy to less than 100 nm. In SIM, each super-resolution (SR) image is computationally constructed from a dataset of numerous raw images. Due to the computational time required for SR image construction, datasets are usually post-processed to avoid slowdowns during the acquisition process. Although post-processing is convenient, this approach has limited use since SR images are not available during image acquisition and thus cannot be utilized for real-time evaluation of the sample. Utilizing SR images in real-time is particularly important when objects of interest are smaller than the resolution limit of conventional microscopy. Applications such as determining optimal focus and finding desirable regions of interest are depending on real-time feedback to user. When performing real-time SIM imaging live cells, image acquisition, processing, and the display of SR images must occur in rapid succession thereby reducing the effects of molecular movement. To achieve the data processing speed necessary for real-time imaging of live cells, we utilized a graphics processing unit (GPU) to perform high-speed SR image construction. Each function for processing SIM data was optimized for parallel operation using GPU kernels. Excluding data transfer overhead, the actual GPU processing time for synthesizing the SR image was 45 milliseconds. Compared with parallel processing on a multi-core computer (CPU), using the GPU achieved a 45-fold increase in processing speed. Overall, data throughput using the GPU attained an 8-fold speed increase over CPU processing, thus allowing real-time high resolution imaging. We applied this SIM platform to record the action of nocodazole on microtubule depolymerization in HeLa cells. We are currently developing this real-time imaging method to investigate the delivery and release mechanism of therapeutic cargoes contained within nanoparticles. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 3982.