Fig 1.
Parameters necessary for calculation of ratio factor A for Cerulean and Venus fluorescent proteins.
(A) Spectra of the optical components in the excitation path of the imaging system. Ssrc(λ), Sex(λ), Sdichr(λ), Sobj(λ) are spectra for the light source (Polychrome V monochromator, TILL Photonics, yellow), excitation filter (Chroma 69008x, blue), dichroic mirror (Chroma 69008bs, green) and objective (Olympus UAPO 40XOI3/340, black), respectively, mounted on Olympus IX71 microscope. Point-by-point multiplication of spectra for each optical element yields an optical function of microscope excitation light path shown by a violet trace. This excitation optical function of the particular imaging system can be used for calculation of a ratio factor A for a wide range of different fluorescent labels. (B) Further point-by-point multiplication of the optical function (violet), normalized Cerulean absorption spectrum (blue), and a spectrum of monochromator slit chosen for a given experiment (orange) gives a function of excitation path for Cerulean (black). Integration of this function and multiplication by the extinction coefficient for Cerulean absorbance results in
, necessary for estimation of the ratio factor A (see Eq 1). (C) Spectra of the optical components in the emission path of the imaging system. Sobj(λ),Sdichr(λ),Sem(λ),Sdet(λ) are objective (Olympus UAPO 40XOI3/340, black), dichroic mirror (Chroma 69008bs, green), and emission filter (Chroma 69008m, dotted red) transmittance and detector (QImaging ExiBlue, brown) sensitivity, respectively. Point-by-point multiplication of spectra for each optical element gives an optical function of microscope emission path shown by a red bold trace. This emission optical function of the particular imaging system can be used for calculation of a ratio factor A for a wide range of different fluorescent labels. (D) Further point-by-point multiplication of the optical function (red dashed trace; left Y axes) and Cerulean emission spectrum, integral of which is normalized to 1,
(blue dashed trace; right Y axes) gives the function of emission path for Cerulean (black trace, right Y axes). Integration of this function and multiplication by Cerulean quantum yield results in
.
Table 1.
Reference data on optical elements of the imaging system.
Fig 2.
Estimation of approach accuracy using a tandem of Cerulean and Venus.
(A) Images of representative PC12 cell expressing the tandem: Cerulean fluorescence (a), Venus fluorescence (b), and a ratio image of Venus to Cerulean fluorescence (c). A color and intensity of each pixel in (c) represent the ratio of Venus/Cerulean fluorescence and averaged intensity of respective pixels in the images (a) and (b) (Favr = (FC + FV)/2), respectively. A scale bar in (a) is 5μm. (B) A linear regression of correlation plot between Cerulean and Venus fluorescence intensities for each pixel within the PC12 cell image shown in (A). A strong linear correlation between the intensities (slope = 3.922±0.004, intercept = 11.1±0.01, R2 = 0.99; the slope is significantly different from zero at the 0.05 level) demonstrates co-localization of fluorescent protein labels. (C) Linear regressions of correlation plots similar to one shown in (B) for five PC12 cells. Cells having different levels of tandem expression were chosen for this plot in order to demonstrate that the ratio of fluorescence intensities remains unchanged in the wide range of tandem expression levels. (D) Expected (Expec.) and apparent (Appar.) ratios of Venus to Cerulean concentrations in the tandem ([LV]/[LC]). The histogram demonstrates that the apparent ratio of Venus to Cerulean concentrations estimated based on the ratio factor (1.33±0.06, mean±S.E.M., n = 5) is close to the expected ratio, which is equal to 1. It indicates that an error associated with inaccurate determination of spectral properties of labels and equipment is about 30%.
Fig 3.
Estimation of ECFP concentration in PC12 cells.
ECFP expressing PC12 cells were filled with Alexa Fluor 594 via patch pipette. (A) Images of two PC12 cells obtained in ECFP (a, b) and Alexa Fluor 594 (c,d) channels right before the cell membrane rupture (a, c) and in 10 min after the rupture (b, d). Images were taken at the moments indicated by black dashed lines in (C). A position of patch pipette are depicted in (A) by white dotted triangles. Scale bar is 20 μm. (B) A correlation plot of ECFP and Alexa Fluor 594 fluorescence intensities for a cytoplasmic part of cell. Each dot in the correlation plot represents a pixel located in a red oval depicted within the cytoplasmic part of cell, of which an image is shown in the insert (ECFP channel). ECFP and Alexa Fluor 594 images used for building the correlation plot were obtained by averaging all frames in ECFP channel taken before the cell rupture and all frames in Alexa Fluor 594 channel between 10 and 15 min. A linear regression without offsets for this plot indicates to reasonable correlation between the fluorescent values (R2 = 0.67). (C) Time courses of cell loading with Alexa Fluor 594 (red traces) and washing out of ECFP (blue traces). Fluorescence intensities for both labels were calculated in ROIs shown in (A) as an average intensity of all pixels within each ROI. Solid and dashed lines in ROIs (A) and time courses (C) correspond to the patched and intact cell, correspondingly. Cell rupture is indicated by black arrowheads in (C) and (D). Alexa Fluor 594 fluorescence (FAlx) reached a plateau in about 5 min after the membrane rupture indicating an equilibrium state at which the concentration of Alexa Fluor 594 inside the cell and in the patch pipette are expected to be equal. Residual fluorescence in ECFP channel indicates that some ECFP molecules (about 30%) are retained in intracellular compartments. Images were taken at the rate of 1 min per frame for each channel in order to minimize photo-bleaching. ECFP photobleaching in a control cell at this frame rate was negligible (dashed blue trace in Fig 3B); thus, it could be also neglected in the patched cells. (D) Time courses from five cells were normalized and averaged, representing a perfusion dynamics of Alexa Fluor 594 (red traces) and washing out of ECFP (blue traces), analogous to those in (C).
Fig 4.
Estimation of HPCA-EYFP concentration in a dendritic tree of cultured hippocampal neurons.
Hippocampal neurons (DIV 10–20) expressing HPCA-EYFP were patched with pipettes filled with an intracellular solution containing Alexa Fluor 594 dye (20–50 μM). (Aa) HPCA-YFP image of dendrite taken at the beginning of experiment, before the membrane rupture. (Ab). Alexa Fluor 594 image of the same area taken in 15 min after cell rupture. Analogous images were taken in other tested neurons and used for obtaining graphs shown in (D) and (F). Arrows denote regions of interest, ROIs, where the fluorescence intensity was measured; the corresponding time courses shown in (B). (Ac). A color-coded ratio of Alexa Fluor 594 to HPCA-YFP fluorescence intensities. Scale bar is 10μm. (B) Intensity of Alexa Fluor 594 fluorescence was measured over time at ROIs indicated by the arrows in (Ab). Intensity profiles are normalized by their maximal values. Colors of the traces correspond to the colors of the arrows in (Ab). (C) A ratio profile of Alexa Fluor 594 to HPCA-YFP fluorescence taken along the dendrite. The plot demonstrates a plateau region up to 100μm from soma where Alexa Fluor 594 wash-in was competed in 15 min after the membrane rupture thus far indicating a range of distances from soma where HPCA-EYFP concentration can be accurately estimated. (D) Ratios of Alexa Fluor 594-to-HPCA-EYFP fluorescence in the proximal and distal parts of the apical dendrites in 15 min after the membrane rupture for 8 tested neurons. The graph demonstrates substantial wash-in of Alexa Fluor 594 in the distal parts of dendrites. A ratio of Alexa/EYFP fluorescence in a distal part normalized to the same ratio in a proximal part was 0.75±0.04. (E) A correlation plot of HPCA-EYFP and Alexa Fluor 594 fluorescence intensities generated for all pixels in the dendrite corresponding to the plateau region in (C) (green region in (Ac)). A linear regression of this plot indicates to a significant linear correlation between HPCA-YFP to Alexa Fluor 594 fluorescence (R2 = 0.87). (F) Estimated dendritic concentrations of HPCA-EYFP in the cytosol pulled from eight tested neurons.