Limits...
Rapid measurement of molecular transport and interaction inside living cells using single plane illumination.

Hedde PN, Stakic M, Gratton E - Sci Rep (2014)

Bottom Line: To demonstrate the advantages of our approach, we quantified the dynamics of several different proteins in the cyto- and nucleoplasm of living cells.For example, from a single measurement, we were able to determine the diffusion coefficient of free clathrin molecules as well as the transport velocity of clathrin-coated vesicles involved in endocytosis.Used in conjunction with dual view detection, we further show how protein-protein interactions can be quantified.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Fluorescence Dynamics, Department of Biomedical Engineering, University of California, Irvine, CA, USA.

ABSTRACT
The ability to measure biomolecular dynamics within cells and tissues is very important to understand fundamental physiological processes including cell adhesion, signalling, movement, division or metabolism. Usually, such information is obtained using particle tracking methods or single point fluctuation spectroscopy. We show that image mean square displacement analysis, applied to single plane illumination microscopy data, is a faster and more efficient way of unravelling rapid, three-dimensional molecular transport and interaction within living cells. From a stack of camera images recorded in seconds, the type of dynamics such as free diffusion, flow or binding can be identified and quantified without being limited by current camera frame rates. Also, light exposure levels are very low and the image mean square displacement method does not require calibration of the microscope point spread function. To demonstrate the advantages of our approach, we quantified the dynamics of several different proteins in the cyto- and nucleoplasm of living cells. For example, from a single measurement, we were able to determine the diffusion coefficient of free clathrin molecules as well as the transport velocity of clathrin-coated vesicles involved in endocytosis. Used in conjunction with dual view detection, we further show how protein-protein interactions can be quantified.

Show MeSH

Related in: MedlinePlus

Solution measurements.(A) From a series of fluorescence images (left) the spatiotemporal correlation is calculated (right). (B) The data (here, EGFP in solution) can be approximated with a Gaussian function. The width of the peak corresponds to the particle iMSD. Free diffusion is indicated by a linear increase of the iMSD with the lag time, τ. (C) Average iMSD resulting from solution measurements of Rhodamine110 (squares, n = 6), EGFP (dots, n = 10) and 20 nm beads (triangles, n = 6). A linear fit of the data results in average diffusion coefficients of 400 ± 20 μm2s−1 for Rhodamine110 (solid line), 87 ± 7.5 μm2s−1 for EGFP (dashed line) and 29 ± 2.7 μm2s−1 for the beads (dotted line). (D) Average amplitude measured in EGFP solutions with different concentrations in the nM range (n = 5). Linear fits of the data results in average relative particle numbers inside the observation volume of 0.27 ± 0.01 (solid line), 0.43 ± 0.02 (dashed line), 0.85 ± 0.05 (dotted line) and 2.6 ± 0.1 (dash-dotted line). All errors stated are standard deviations.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4231332&req=5

f1: Solution measurements.(A) From a series of fluorescence images (left) the spatiotemporal correlation is calculated (right). (B) The data (here, EGFP in solution) can be approximated with a Gaussian function. The width of the peak corresponds to the particle iMSD. Free diffusion is indicated by a linear increase of the iMSD with the lag time, τ. (C) Average iMSD resulting from solution measurements of Rhodamine110 (squares, n = 6), EGFP (dots, n = 10) and 20 nm beads (triangles, n = 6). A linear fit of the data results in average diffusion coefficients of 400 ± 20 μm2s−1 for Rhodamine110 (solid line), 87 ± 7.5 μm2s−1 for EGFP (dashed line) and 29 ± 2.7 μm2s−1 for the beads (dotted line). (D) Average amplitude measured in EGFP solutions with different concentrations in the nM range (n = 5). Linear fits of the data results in average relative particle numbers inside the observation volume of 0.27 ± 0.01 (solid line), 0.43 ± 0.02 (dashed line), 0.85 ± 0.05 (dotted line) and 2.6 ± 0.1 (dash-dotted line). All errors stated are standard deviations.

Mentions: The spatiotemporal correlations of an image series of immobile particles resembles the average shape of the particles convoluted with the microscope point spread function (PSF). For mobile molecules, the peak waist will broaden and the peak height will diminish with increasing lag time (Fig. 1A,B). In the case of free diffusion, the increase in peak width, i.e., the second order central moment, is proportional to the particle mean square displacement (MSD) (Eq. S2). By plotting the image MSD (iMSD) over the lag time, the average diffusion coefficient is determined by the slope of this linear relationship11. Notably, the slope is independent of the microscope PSF meaning that the calculation of the diffusion coefficient does not require any calibration of the instrument waist (Supplementary Fig. S5). Instead, the iMSD at zero lag time, i.e., the offset resulting from linear regression, points to the instrument PSF convoluted with the average particle size. To demonstrate that SPIM-iMSD is capable of precisely determining diffusion coefficients of fast moving molecules, we prepared nanomolar solutions of Rhodamine110, EGFP, and 20-nm red fluorescent beads. For each sample, SPIM image series were acquired and spatiotemporally correlated; the average iMSDs are plotted in Figure 1C. A comparison of the resulting diffusion coefficients with their literature values proves that SPIM-iMSD provides the correct numbers with high accuracy (Supplementary Table S1)121314. In addition to dynamics, particle concentrations can be measured with FFS methods, too. The amplitude of the STICS function is inversely proportional to the average particle number inside the observation volume and, hence, the concentration. For free, three-dimensional diffusion, the amplitude depends linearly on the inverse of the third power of the peak waist with the slope being proportional to the inverse particle number (Eq. S3). This relation was verified with EGFP solutions of different concentrations (Fig. 1D). Note however, that these values are relative particle numbers. With an EMCCD camera as detector, the measurement of absolute particle numbers requires calibration15.


Rapid measurement of molecular transport and interaction inside living cells using single plane illumination.

Hedde PN, Stakic M, Gratton E - Sci Rep (2014)

Solution measurements.(A) From a series of fluorescence images (left) the spatiotemporal correlation is calculated (right). (B) The data (here, EGFP in solution) can be approximated with a Gaussian function. The width of the peak corresponds to the particle iMSD. Free diffusion is indicated by a linear increase of the iMSD with the lag time, τ. (C) Average iMSD resulting from solution measurements of Rhodamine110 (squares, n = 6), EGFP (dots, n = 10) and 20 nm beads (triangles, n = 6). A linear fit of the data results in average diffusion coefficients of 400 ± 20 μm2s−1 for Rhodamine110 (solid line), 87 ± 7.5 μm2s−1 for EGFP (dashed line) and 29 ± 2.7 μm2s−1 for the beads (dotted line). (D) Average amplitude measured in EGFP solutions with different concentrations in the nM range (n = 5). Linear fits of the data results in average relative particle numbers inside the observation volume of 0.27 ± 0.01 (solid line), 0.43 ± 0.02 (dashed line), 0.85 ± 0.05 (dotted line) and 2.6 ± 0.1 (dash-dotted line). All errors stated are standard deviations.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4231332&req=5

f1: Solution measurements.(A) From a series of fluorescence images (left) the spatiotemporal correlation is calculated (right). (B) The data (here, EGFP in solution) can be approximated with a Gaussian function. The width of the peak corresponds to the particle iMSD. Free diffusion is indicated by a linear increase of the iMSD with the lag time, τ. (C) Average iMSD resulting from solution measurements of Rhodamine110 (squares, n = 6), EGFP (dots, n = 10) and 20 nm beads (triangles, n = 6). A linear fit of the data results in average diffusion coefficients of 400 ± 20 μm2s−1 for Rhodamine110 (solid line), 87 ± 7.5 μm2s−1 for EGFP (dashed line) and 29 ± 2.7 μm2s−1 for the beads (dotted line). (D) Average amplitude measured in EGFP solutions with different concentrations in the nM range (n = 5). Linear fits of the data results in average relative particle numbers inside the observation volume of 0.27 ± 0.01 (solid line), 0.43 ± 0.02 (dashed line), 0.85 ± 0.05 (dotted line) and 2.6 ± 0.1 (dash-dotted line). All errors stated are standard deviations.
Mentions: The spatiotemporal correlations of an image series of immobile particles resembles the average shape of the particles convoluted with the microscope point spread function (PSF). For mobile molecules, the peak waist will broaden and the peak height will diminish with increasing lag time (Fig. 1A,B). In the case of free diffusion, the increase in peak width, i.e., the second order central moment, is proportional to the particle mean square displacement (MSD) (Eq. S2). By plotting the image MSD (iMSD) over the lag time, the average diffusion coefficient is determined by the slope of this linear relationship11. Notably, the slope is independent of the microscope PSF meaning that the calculation of the diffusion coefficient does not require any calibration of the instrument waist (Supplementary Fig. S5). Instead, the iMSD at zero lag time, i.e., the offset resulting from linear regression, points to the instrument PSF convoluted with the average particle size. To demonstrate that SPIM-iMSD is capable of precisely determining diffusion coefficients of fast moving molecules, we prepared nanomolar solutions of Rhodamine110, EGFP, and 20-nm red fluorescent beads. For each sample, SPIM image series were acquired and spatiotemporally correlated; the average iMSDs are plotted in Figure 1C. A comparison of the resulting diffusion coefficients with their literature values proves that SPIM-iMSD provides the correct numbers with high accuracy (Supplementary Table S1)121314. In addition to dynamics, particle concentrations can be measured with FFS methods, too. The amplitude of the STICS function is inversely proportional to the average particle number inside the observation volume and, hence, the concentration. For free, three-dimensional diffusion, the amplitude depends linearly on the inverse of the third power of the peak waist with the slope being proportional to the inverse particle number (Eq. S3). This relation was verified with EGFP solutions of different concentrations (Fig. 1D). Note however, that these values are relative particle numbers. With an EMCCD camera as detector, the measurement of absolute particle numbers requires calibration15.

Bottom Line: To demonstrate the advantages of our approach, we quantified the dynamics of several different proteins in the cyto- and nucleoplasm of living cells.For example, from a single measurement, we were able to determine the diffusion coefficient of free clathrin molecules as well as the transport velocity of clathrin-coated vesicles involved in endocytosis.Used in conjunction with dual view detection, we further show how protein-protein interactions can be quantified.

View Article: PubMed Central - PubMed

Affiliation: Laboratory of Fluorescence Dynamics, Department of Biomedical Engineering, University of California, Irvine, CA, USA.

ABSTRACT
The ability to measure biomolecular dynamics within cells and tissues is very important to understand fundamental physiological processes including cell adhesion, signalling, movement, division or metabolism. Usually, such information is obtained using particle tracking methods or single point fluctuation spectroscopy. We show that image mean square displacement analysis, applied to single plane illumination microscopy data, is a faster and more efficient way of unravelling rapid, three-dimensional molecular transport and interaction within living cells. From a stack of camera images recorded in seconds, the type of dynamics such as free diffusion, flow or binding can be identified and quantified without being limited by current camera frame rates. Also, light exposure levels are very low and the image mean square displacement method does not require calibration of the microscope point spread function. To demonstrate the advantages of our approach, we quantified the dynamics of several different proteins in the cyto- and nucleoplasm of living cells. For example, from a single measurement, we were able to determine the diffusion coefficient of free clathrin molecules as well as the transport velocity of clathrin-coated vesicles involved in endocytosis. Used in conjunction with dual view detection, we further show how protein-protein interactions can be quantified.

Show MeSH
Related in: MedlinePlus