Quantitative Analysis of Membrane Phosphoinositides

All intensity-based image processing and analysis were performed using ImageJ/(Fiji) software (Schindelin et al., 2012 (link)). The Imaris software (Bitplane AG, Switzerland) was specifically used to generate 3D projections and volume renderings. For analysis of the microtubule patterning, z-stacks were deconvolved with Huygens Essential version 22.04 (Scientific Volume Imaging, The Netherlands), using the CMLE algorithm, with Acuity: –60 and 20 iterations.
FLIM-based images were analyzed using LAS X SP8 Control Software (Leica Microsystems GmbH). Global fitting of the intensity decay profile using n-exponential reconvolution was performed to separate major fluorescent components within each channel and calculate their lifetime. The number of components (n) used for curve fitting was determined according to the evaluation of the chi-squared (χ2) value (Lakowicz, 2006 (link)) and a threshold of 30 photons was applied to generate the final images, unless otherwise stated. Components with lifetime value <1 ns representing autofluorescent species accumulating in plant cell walls (Donaldson, 2020 (link); Heskes et al., 2012 (link)) were subtracted from each channel.
All data on reconvolution of decay profiles of images presented in the article are provided as source files.
Since global fitting does not always allow to resolve and characterize the decay profile of pixels constituting minor populations within an image (Ranjit et al., 2018 (link)), the decay profile of fluorescent emission of regions of interest (ROIs) selected on membrane domains was additionally analyzed for the analysis and quantification of PI(4,5)P₂ membrane enrichment. n-exponential fitting of the decay curve was carried out to resolve lifetime values associated with each ROI (Figure 5—figure supplement 1). Reconvolution data from all images and selected ROIs analyzed are provided as source files. Further, the raw decay profile of each ROI was analyzed using the Phasor approach to obtain a Phasor fingerprint whose position describes the fluorescent emission at that region and reflects its composition with respect to the relative abundance of fluorescent species (Figure 5—figure supplement 1) (see Ranjit et al., 2018 (link) and Malacrida et al., 2021 (link) for a detailed description of the Phasor approach). Phasor plots were generated using a second harmonic, a threshold of 22 photons, and a median of 15. To visualize and compare the Phasor fingerprint of ROIs selected in multiple images from different genotypes, the center of mass of Phasor images depicting the pixel populations originating from each ROI was calculated images using ImageJ/(Fiji) (Schindelin et al., 2012 (link)) and the obtained XM and YM coordinates were plotted on a graph (Figure 5—figure supplement 1). A Mann-Whitney non-parametric test was performed to detect statistically significant differences between XM and YM values.
For the quantification of PI(4,5)P₂ and PI4P membrane enrichment, the mean fluorescence intensity of freehand ROIs drawn on the IT tip and on the plasma membrane of infected root hairs were measured on images of single fluorescent components from each channel, obtained from global fitting of the intensity decay profile (threshold of 20 photons) and subtraction of components with lifetime <1 ns (both channels) and >1.7 ns (channel 2). The mean intensity of a 2.5 × 2.5 μm square ROI outside of the infected root hair cell was measured as a background value, subtracted from each ROI and the mean gray value on the IT tip was divided by the mean gray value on the plasma membrane [(Int_{IT tip} - Int_bkgd) / (Int_PM- Int_bkgd)] for each component to obtain an index of the enrichment of each phosphoinositide at the IT tip relative to the plasma membrane. Statistically significant differences were detected using a Mann-Whitney non-parametric test.