Home Browse SearchHelp DownloadGroup

Endosite: Help page
Endosite Help Index:

Endocytic assay:

The endocytic assay was adopted from standardised protocol to suit the requirements of the screen. dsRNA was plated on slides and dried down. Cells were then spotted on the slides, at a concentration of approximately 1000 cells per well, and allowed to grow and take up dsRNA for 4 days. Following this cells were subjected to the endocytic assay. Cells were pulsed with FITC-conjugated dextran and Alexa568-conjugated transferrin for 20 minutes. Following this any remaining Tfr on the surface was stripped using an acid-wash on ice, and cells were then labelled on ice with Okt9, an antibody to the transferrin receptor, to differentially label steady-state levels on the cell surface. Following this cells were labelled with Hoechst to label nuclei and fixed in 4% paraformaldehyde. A coverslip was sealed onto the glass slide. The slide was imaged through the coverslip.

Parameters assayed:
  • FDex and Tfr intensity measurements

    Four direct intensity measurements were made in both the FDex and Tfr channels. (Fint1-4 for FDex and Tint1-4 for Tfr) Three of these related to per-cell intensity following different degrees of local background subtraction. Local background subtraction was carried out using the ‘tophat’ method, available in the MATLAB image processing toolbox as an inbuilt function. Tophat subtraction was carried out using a disk as the structuring element. The size of the disk determines the final output; the intensity of objects smaller than the disk will not be significantly affected. We used three different disk sizes to emphasize different aspects of the intensity distribution within a single cell. The large disk(Fint1 and Tint1) is larger than the biggest cell in the field, resulting in preservation of all cellular fluorescence, punctate or diffuse. The medium disk(Fint2 and Tint2) preserves bright punctae and collections of punctae. The small disk(Fint3 and Tint3), roughly the same size as one puncta, eliminates all but individual bright punctae. For the medium and small disks, it was observed that after subtraction, much of the image consisted of residual background (intensities common to all images). To remove this, only intensities above a certain percentile were considered. All measurements were normalised to cell size. Fint4 and Tint4 represent the ratio of the non-zero fluorescence after small disk background subtraction to the cell area, for FDex and Tfr respectively.

  • Okt9 and Tfr/Okt9 intensity measurements

    Okt9 intensity per cell was measured using a large disk (Okt). Since Tfr endocytosis is receptor-mediated, it follows that internalised levels will depend on steady state surface levels. Therefore, an additional variable of interest for the transferrin channel is the ratio of internalised Tfr to Okt9. This was calculated for large-disk (Rto1), medium-disk (Rto2) and small-disk Tfr (Rto3).

  • Tfr and FDex geometry measurements

    For each cell, 4 endosome geometry measurements were made in each of the two endocytic channels (Fmph1-4 for FDex and Tmph1-4 for Tfr). Putative endosomes were identified through very tight local background subtraction (smaller disk than the small disk mentioned above) and thresholding to produce a binary mask. It should be noted that at 20X-0.75NA, it is impossible to differentiate between single endosomes and clumps of endosomes, due to insufficient resolution. However, for the purposes of this study, punctae identified by this method are referred to as endosomes. It should also be noted that the emprically selected threshold was intentionally high, to prevent merging of nearby punctae. As an unfortunate consequence, however, some out-of-focus or dim punctae were sometimes ommitted from the final binary mask. Fmph1 and Tmph1 measure number of endosomes. Fmph2 and Tmph2 measure the average size of endosome per cell. Fmph3 and Tmph3 refer to the ratio of total endosomal area to cell area (similar to Fint4 and Tint4). Fmph4 and Tmph4 measure circularity (perimeter^2/area). Fclc and Tclc measure the number of colocalised pixels between the two channels normalised to total numbers of FDex endosome pixels and Tfr pixels respectively.

  • Cell and nuclear geometry measurements

    Although not directly linked to endocytosis, these parameters were calculated because it made sense to extract as much information from the images as possible. For the cell, the only parameter extracted was cell size. For the nucleus, four parameters were extracted; size(NucSize), circularity(NucCirc), border fluctuation(NuFluct) (standard deviation of the radius calculated for every point on the nuclear perimeter) and the distance of the centroid of the nucleus from the centroid of the cell(NucCellD).


Experimental parameters:
Fluid Intensity
  • Fint1
  • Fint2
  • Fint3
  • Fint4
Fluid Morphology
  • Fnum
  • Fmph1
  • Fmph2
  • Fmph3
  • Fclc
Transferrin Intensity
  • Tint1
  • Tint2
  • Tint3
  • Tint4
Ratio Intensity
  • Rto1
  • Rto2
  • Rto3
  • Okt
Transferrin Morphology
  • Tnum
  • Tmph1 Tmph2
  • Tmph3
  • Tclc
Cell Morphology
  • CellSize

Search Endosite

Users can search EndoSite using the following options:

  • Simple text search (Gene names, symbols, CG identifiers, Fbgn identifiers, protein domains and GO terms)
    Users can search and retrieve genes using simple keywords. Simple text search enables search using offciial gene names, symbols, CG identifiers, Fbgn identifiers, protein domains and GO terms.

Browse Endosite

Users can browse EndoSite using the following options:

  • Browse genes assayed in "Primary Screen"
  • Browse genes assayed in "Classification Assay"

Endosite Team :
Prof. Satyajit Mayor (Contact : mayor@ancbs.res.in)
Prof. R. Sowdhamini (Contact : mini@ncbs.res.in)