Dataset Information

Statistical deconvolution for superresolution fluorescence microscopy.

ABSTRACT: Superresolution microscopy techniques based on the sequential activation of fluorophores can achieve image resolution of ?10 nm but require a sparse distribution of simultaneously activated fluorophores in the field of view. Image analysis procedures for this approach typically discard data from crowded molecules with overlapping images, wasting valuable image information that is only partly degraded by overlap. A data analysis method that exploits all available fluorescence data, regardless of overlap, could increase the number of molecules processed per frame and thereby accelerate superresolution imaging speed, enabling the study of fast, dynamic biological processes. Here, we present a computational method, referred to as deconvolution-STORM (deconSTORM), which uses iterative image deconvolution in place of single- or multiemitter localization to estimate the sample. DeconSTORM approximates the maximum likelihood sample estimate under a realistic statistical model of fluorescence microscopy movies comprising numerous frames. The model incorporates Poisson-distributed photon-detection noise, the sparse spatial distribution of activated fluorophores, and temporal correlations between consecutive movie frames arising from intermittent fluorophore activation. We first quantitatively validated this approach with simulated fluorescence data and showed that deconSTORM accurately estimates superresolution images even at high densities of activated fluorophores where analysis by single- or multiemitter localization methods fails. We then applied the method to experimental data of cellular structures and demonstrated that deconSTORM enables an approximately fivefold or greater increase in imaging speed by allowing a higher density of activated fluorophores/frame.

SUBMITTER: Mukamel EA

PROVIDER: S-EPMC3353000 | biostudies-literature | 2012 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Statistical deconvolution for superresolution fluorescence microscopy.

Mukamel Eran A EA Babcock Hazen H Zhuang Xiaowei X

Biophysical journal 20120515 10

Superresolution microscopy techniques based on the sequential activation of fluorophores can achieve image resolution of ∼10 nm but require a sparse distribution of simultaneously activated fluorophores in the field of view. Image analysis procedures for this approach typically discard data from crowded molecules with overlapping images, wasting valuable image information that is only partly degraded by overlap. A data analysis method that exploits all available fluorescence data, regardless of ...[more]

PMID: 22677393

Similar Datasets

Project description:In nature, bacteria must survive long periods of nutrient deprivation while maintaining the ability to recover and grow when conditions improve. This quiescent state is called stationary phase. The biochemistry of Escherichia coli in stationary phase is reasonably well understood. Much less is known about the biophysical state of the cytoplasm. Earlier studies of harvested nucleoids concluded that the stationary-phase nucleoid is "compacted" or "supercompacted," and there are suggestions that the cytoplasm is "glass-like." Nevertheless, stationary-phase bacteria support active transcription and translation. Here, we present results of a quantitative superresolution fluorescence study comparing the spatial distributions and diffusive properties of key components of the transcription-translation machinery in intact E. coli cells that were either maintained in 2-day stationary phase or undergoing moderately fast exponential growth. Stationary-phase cells are shorter and exhibit strong heterogeneity in cell length, nucleoid volume, and biopolymer diffusive properties. As in exponential growth, the nucleoid and ribosomes are strongly segregated. The chromosomal DNA is locally more rigid in stationary phase. The population-weighted average of diffusion coefficients estimated from mean-square displacement plots is 2-fold higher in stationary phase for both RNA polymerase (RNAP) and ribosomal species. The average DNA density is roughly twice as high as that in cells undergoing slow exponential growth. The data indicate that the stationary-phase nucleoid is permeable to RNAP and suggest that it is permeable to ribosomal subunits. There appears to be no need to postulate migration of actively transcribed genes to the nucleoid periphery.IMPORTANCE Bacteria in nature usually lack sufficient nutrients to enable growth and replication. Such starved bacteria adapt into a quiescent state known as the stationary phase. The chromosomal DNA is protected against oxidative damage, and ribosomes are stored in a dimeric structure impervious to digestion. Stationary-phase bacteria can recover and grow quickly when better nutrient conditions arise. The biochemistry of stationary-phase E. coli is reasonably well understood. Here, we present results from a study of the biophysical state of starved E. coli Superresolution fluorescence microscopy enables high-resolution location and tracking of a DNA locus and of single copies of RNA polymerase (the transcription machine) and ribosomes (the translation machine) in intact E. coli cells maintained in stationary phase. Evidently, the chromosomal DNA remains sufficiently permeable to enable transcription and translation to occur. This description contrasts with the usual picture of a rigid stationary-phase cytoplasm with highly condensed DNA.

Dataset Information

Statistical deconvolution for superresolution fluorescence microscopy.

Publications

Statistical deconvolution for superresolution fluorescence microscopy.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure