Subcellular Euclidean distance measurements with multicolor fluorescence localization imaging in cultured cells

Abstract

This protocol measures the 3D Euclidean distance (Δ_3D) between two/three fluorescently labeled kinetochore components in fixed samples using Kinetochore Delta software (KiDv1.0.1, MATLAB based). Overestimation of mean Δ_3D is corrected through a Bayesian algorithm, with Δ_EC distances reflecting the ensemble average positions of fluorophores within a kinetochore population. This package also enables kinetochore categorization, which can be used to sub-sample kinetochores and measure Δ_EC. Together, this allows the dynamic architecture of human kinetochores to be investigated (tested in hTERT-RPE1 cells). For complete details on the use and execution of this protocol, please refer to Roscioli et al. (2020).

Keywords: Cell Biology; Cell culture; Microscopy.

Graphical abstract

Figure 1

3D Delta Euclidian distance measurements of CenpC to CenpC (0 nm true distance) Comparison between the measurement before and after BEDCA inflation correction. (A) Example images of CenpC labeled with anti-CenpC primary antibody and mix of A488, A568 and A647 secondary antibodies. Scale bar 500 nm. (B) Delta 3D distance measurements of the CenpC-A488-to-CenpCA568, CenpC-A488-to-CenpC-A647, and CenpC-A568-to-CenpC-A647 distances. The measured 3D Euclidian distance distribution (Δ3D) prior to BEDCA distance inflation correction is shown in black. The 3D Euclidian distance distribution after BEDCA distance inflation correction (ΔEC) is shown in red. Mean, standard deviation and sample size for each distribution is indicated on the right. (Adapted from Roscioli et al., 2020 with permission)

Figure 2

Workflows for Delta Euclidian distance inflation corrected measurements (ΔEC) between kinetochore components and kinetochore intensity measurements The protocol contains the steps for four workflows depending on the desired output: (1) For ΔEC measurement between two fluorescently labeled kinetochore proteins, follow steps 1–6, 9, 10, 12–14. (2) For ΔEC measurements between three fluorescently labeled kinetochore proteins (in pairs), follow steps 1–7, 9, 10, 12–14. (3) For ΔEC measurements between two fluorescently labeled kinetochore proteins at kinetochore marker positive and negative kinetochores, follow steps 1–7, 9, 10, 12–14. (4) For kinetochore intensity measurements, perform step 11 after any of the previous workflows.

Figure 3

Kinetochore Delta (KiDv1.0.1) software graphic user interphase (GUI) for data analysis set up The parameters of data anlysis are set up and saved in a jobset as described in steps 3 and 4. The figure displays the jobset analysis set up of the provided Nnf1-CenpC-Ndc80N sample dataset as described in step 4. (See also Figures S1 and S2)

Figure 4

Kinetochore sister-sister pairing and spot quality control in the Coordinate system channel (CSC) (A) Kinetochore sister-sister pairing. Left: zoom-in at the kinetochore spot under consideration (white cross) and spots within two microns that are suggested for pairing (green cross), already paired (red cross) or previously ignored for pairing (yellow cross). The image is projection of 5 z-slices around the spot under consideration. Right: image region of interest as set by the user in step 4, projection of 5 z-slices around the spot under consideration. (B) Examples of kinetochores that should be ignored in the analysis due to spot distortion in the kinetochore under consideration or its kinetochore sister pair (2 and 4), poor detection of the kinetochore spot center (1), and no sister kinetochore spot detection (3, 5 and 6).

Figure 5

Kinetochore spot selection for spot quality control in ΔEC measurements (A and B) show examples of spot selection GUI using kitSelectData function in two analyzed cells from the exemplar dataset after KiDv1.0.1 spot detection and manual sister-sister pairing. The user is required to evaluate the spot shape and center detection, and to discard non-Gaussian shaped asymmetric spots and spots with poor center detection. A. Spots 1, 31, 40 and 47 are to be discarded. Spot 1 is distorted. Spots 31, 40 and 47 are too close to a neighboring spot and the center may not be detected appropriately. B. Spots 4, 18, 28, 33 and 49 are to be discarded. These spots are distorted.

Figure 6

Quality control of BEDCA run Posterior distributions of the ΔEC mean, taux, and tauz (blue) and the corresponding prior distributions (red). (A) Example of figure saved as “Posterior_mu” in “MCMC_EuclDistMargFigs” folder using Nnf1-to-Ndc80N distance correction from provided dataset. Histogram of corrected mean (mu) distance (blue). Red line is the corresponding prior. Prior mean is indicated in the legend (60 nm). The mean and sd. of the posterior mu distribution are stated above (mean 57.4496 nm, sd. 1.214 nm). (B) Example of figure saved as "Posterior_taux” in “MCMC_EuclDistMargFigs” folder, dataset as in A. Histogram of posterior precision in x(y) (taux) (blue). Red line is the corresponding taux prior. Prior mean is stated in the legend (0.0025 nm^-2). The mean and sd. of the posterior taux distribution are stated above (mean 0.002594 nm^-2, sd. 0.00021498 nm^-2). (C) Example of figure saved as "Posterior_tauz” in “MCMC_EuclDistMargFigs” folder, dataset as in A. Histogram of precision in z (tauz) (blue). Red line is the corresponding tauz prior. Prior mean is stated in the legend (0.000625 nm^-2). The mean and sd. of the posterior tauz distribution are stated above (mean 0.00036346 nm^-2, sd. 0.000023406 nm^-2).

Figure 7

Quality control of BEDCA run Δ3D and ΔEC distributions, and convergence. (A) 3D Delta Euclidian distance distribution of the Nnf1-to-Ndc80N distance from the provided dataset. The figure is automatically saved by the algorithm under the name “DataWithPostEsts”. The figure shows histogram of the 3D distance measurements before (nm, blue) and after (nm, red) BEDCA correction.The uncorrected 3D Delta distribution (blue) lies to the left of the 200 nm (default) cutoff imposed in BEDCA. (B) Convergence of the mean, taux (precision in x), and tauz (precision in z) of the Nnf1-to-Ndc80N distance measurements from the provided dataset (see A.). The figure is automatically saved as “ConvPosteriorSummary”. BEDCA is ran with default parameters. GRc numbers below 1.01 indicate successful convergence. Taux has order of magnitude 10^-3 nm^-2. Tauz has order of magnitude 10^-4 nm^-2.

Figure 8

BEDCA output table illustrated with analysis of the Nnf1-to-Ndc80N distance from the exemplar dataset The table shown here illustrates the obtained parameters after analysis of the Nnf1-to-Ndc80N distance from the exemplar dataset, and the format of the output .csv file that the user can expect. In the .csv file, BEDCA automatically saves the mean, median and standard deviation of the following inferred parameters of the measured distance: the distribution mean (mu, nm), taux and tauz (nm^-2, see step 14 b i.), sdx and sdz (nm). The .csv file has a general name of “SummaryStats_chain1” in MATLAB\MCMCruns\AnalysisName, where AnalysisName refers to the name under which the analysis is saved (see step 14 b. v.).