Phosphoinositide localization tests
This protocol is extracted from research article:
Escherichia coli as a platform for the study of phosphoinositide biology
Sci Adv, Mar 27, 2019; DOI: 10.1126/sciadv.aat4872

To create the PLCδ-PH domain non–PIP2-binding mutant, we performed nonsynonymous mutations, substituting eight amino acids for alanine, such that the interaction would be impaired. Following the numbering of Ferguson and coworkers (39), whose paper we used to guide these mutations, the changes made were K30A, K32A, W36A, R40A, E54A, S55A, R56A, and K57A.

The cells for these experiments were visualized using a LEICA scanning confocal microscope with the pinhole aperture set at 1 Airy unit relative to the fluorescence of GFP at 507 nm (1 Airy = 128 μm). Excitation was done sequentially with a laser set at 488 and 587 nm for GFP and mCherry, respectively, and with acquisition on the range of 500 to 570 nm for GFP and 600 to 670 nm for mCherry. A 100× HC PL APO oil objective of numerical aperture 1.44 was used. To improve the speed of the measurement and minimize photobleaching, only a fraction of the image was scanned, a square of side 29 μm at a resolution of 512 × 512 pixels, which was scanned at 100 Hz.

To provide a quantitative measure of the change in localization of the reporter, we devised an index based on a line scan perpendicular to the major axis of the cell performed with the ImageJ software (60). This line is done on the center of the cell, to avoid the cell poles, where inclusion bodies can form, thus avoiding a confounding factor that could erroneously be interpreted as membrane localization if analyzing only the line. The expectation is then that, for the control channel (in red since it is given by mCherry), the line scan should show a unimodal distribution. For the reporter channel (in green), two different behaviors are expected: The line scan should also show a unimodal distribution overlapping with the red channel in the control experiments, but in the cells with the detected phosphoinositide, it should have a bimodal distribution with a dip in the middle. This behavior can be seen in Fig. 5A.

To minimize noise in the line scan analysis, we used a 10-pixel-thick line, such that random noise would be averaged. The lines were drawn manually on the red control channel such that there could be no experimental bias in the selection of the cells. The line scan was then standardized to its maximum so that variation is between 0 and 1, and differences in the levels of fluorescence between cells or channels did not affect the calculations. The first part of the index was then calculated for the green channel by splitting the line scan into three regions, the central 0.5 μm as the center region and the remainder of the line on each side as the left or right side accordingly. The scans were aligned manually. Once the line scan is performed, the maximum of the left side and that of the right side were averaged and divided by the minimum in the center region. For a curve that follows a normal or symmetric unimodal distribution, the maximum outside of the center area should be basically the same as the minimum inside, giving a value close to 1. On the other hand, if the curve is bimodal, then the value should be higher than 1 because the average of the maximums outside the center area should be higher than the minimum inside (where the dip occurs).

While the scan of the green channel would be enough to detect the change of the reporter localization, the alignment of the line scans when defining the center region could introduce noise into the calculations because errors in the alignment would cause the average of the outside regions to be higher than the minimum in the center region. While errors of alignment should be small, the index can be improved by also using the red control channel. The final index is the value for the green channel (first part of the index) divided by the red channel (second part of the index). In this way, any error caused by misalignment is corrected since the value for both green and red channels would show the same bias, thus taking the index back to 1 should both curves coincide. Figure S3 shows a schematic of the calculation of the index.

To obtain the distribution of the bimodality index for each population of cells, we scored 10 cells per field in four different fields for each experimental condition. As described earlier, the cells were chosen blindly in the red channel, and the only consideration for the selection was for the cell to be separate enough from other cells such that the line scan would reach the background level at its extremes, as opposed to rising again by going into a different cell. The values obtained in this way were then used to calculate a one-way ANOVA and a post hoc Tukey HSD test.

Note: The content above has been extracted from a research article, so it may not display correctly.

Please log in to submit your questions online.
Your question will be posted on the Bio-101 website. We will send your questions to the authors of this protocol and Bio-protocol community members who are experienced with this method. you will be informed using the email address associated with your Bio-protocol account.

We use cookies on this site to enhance your user experience. By using our website, you are agreeing to allow the storage of cookies on your computer.