Love:Protocols:Counting

Cell Counting (in the metamorph era) [Cut]

Staining

Scan slide

1.place slide clean a glass slide thoroughly with ethanol; let dry and then carefully place the grid on the slide (open wells facing down). leave some PBS after the last wash on top of the grid, so that the seal is uniform and air is not trapped place glass slide on microscope use fast transmission settings (TRANS fast - transmission light; 2ms exp time; live) to find a rough focus make sure grid is slightly angled, so that the top-left corner of any given block is higher in the image than the top-right corner -to move the grid around, take advantage of the thin liquid layer that should be left and tap lightly the grid to move it.

launch 'scan slide' application. set binning to 1 and wavelengths to 1+#of wavelengths to scan for the setup process. Transmission is used as the additional wavelength for calibration and alignment.

2.adjust alignment move to right end top position and adjust focus use dapi for the actual focus as fluorescence and transmission plane may not be confocal - use live DAPI mode to adjust focus: set z= 0 in the software; adjust z height on microscope, then if necessary fine tune it in the software; set the z position in order to obtain the highest values in the image histogram keep z position on microscope and software constant from now on. move to right end bottom position and adjust the focus (always in DAPI live mode), this time using the corresponding alignment screw on the slide holder. check alignment of the right end by scrolling up and down move to left end, center of the slide, and adjust focus using the corresponding alignment screw

3.check illumination settings move to a random position in the center of the grid adjust illumination settings for all channels; take snaps and try minimizing the number of repetitive exposures to reduce scan time: keep the exposure time to the minimum necessary in every channel to improve image recognition: use exposure time that gives the best spread in the image histogram usually there is a threshold exposure, above which the improvement in image quality is not worth the increase in overall time

4.Calibration snap a transmission picture (to make sure calibration process will be performed using this wavelength) start step by step calibration procedure following software instruction. pick a corner of one of the blocks as feature, in order to easily individuate it. Don't use wells alone, as it may be confusing later on.

5. Set scan area using transmission live mode, go to outermost corner of PDMS (not only the corner of the features, but further on) make sure that you are really on outermost corner by scrolling a little around the edges both in the horiz and vert directions set first corner position to current repeat for the bottom right corner, paying attention to exagerate(!) the extent of your area. metamorph tends to scan a smaller area than the one inputted.

6.start scanning set binning to 2, reduce wavelengths to #of wavelengths to scan set description and directory for file save. every single scan attempt should be in a different directory. cycle once more through all illumination settings to check focus and exp times. press 'scan' button

once done (usually ~3hrs), backup all files before proceeding

Counting algorithm - step 1 - parse into blocks

in MM offline, open scan to be parsed. then open journal for step 1, making sure all references to files are respected (sometimes copies on the local and network drives may conflict) run journal save temp file to directory where all parsed blocks should be saved align grid to DAPI Fast Scan preview image (keeping the magenta feature on the bottom right corner) click on 'run journal for every grid posn' nb: metamorph usually crashes on the last journal loop; make sure all 52 blocks have been saved in the directory, then end program and restart.

Counting algorithm - step 2 - align regions open directory where parsed blocks are saved open journal for step 2 run journal (make sure all journals are on a local drive, for best performance) adjusting angle and size, align 40x40 grid with the features in the DAPI image press 'run journal for each grid posn' repeat for every block backup region files before proceeding

Counting algorithm - step 3 - counting and logging cells

using randomly selected files for all channels, calibrate settings for 'multiwavelength cell scoring' application after repeating for several blocks, save settings open journal for step 3 go to line of code where multiwavelength cell scoring is called; edit settings for command: load settings you just calibrated save journal run journal

Check

by checking the log numbers, the original tiffs for the blocks and comparing with the segmentation images in the directory control if results make sense and seem reliable use journal for step 4 and input number of blocks for which count has to be repeated recall to save in separate logs

import data into MATLAB

open journal for step 3