Quantitative Approaches for Studying Cellular Structures and Organelle Morphology in Caenorhabditis elegans

The ability to quantify changes in the morphology of tissues and organelles is important for understanding gene function as well as the impact of genetic mutations. Our protocols explain how freely available software can be used for non-subjectively assessing the morphology of synapses, muscles, and mitochondria in a nematode, C.elegans. Many previous studies have relied upon qualitative methods for comparing morphological changes.

However, these can be problematic as they may not capture subtle phenotypic differences, might over and underestimate changes, and are assessed subjectively. Our quantitative methods provide more robust and less biased means for assessing morphological changes. These methods could easily be used for studying morphological changes in other cellular structures or model organisms in order to define the gene function and the consequences of disease-associated mutations.

Start by preparing slides for imaging. Place a drop a of melted agarose onto a microscope slide and immediately press down the droplet with another microscope slide, gently flattening the agarose. Leave the agarose to dry for one minute then carefully separate the two slides and leave the solidified pad on one of the slides.

Place the slide on the microscope stage and add a five to 10-milliliter drop of anesthetic to the center of the agarose pad. Use a heat-sterilized pick to quickly transfer 10 to 15 animals from the stock plate to the anesthetic droplet before the solution dries out. Then apply a coverslip by positioning it just above the agarose pad and gently dropping it down.

Image the synaptic regions with a line-scanning confocal microscope coupled to a 488 and 552 nanometer optically-pumped semi-conductor diode laser and equipped with image capture software. After locating the worms, switch to a 40X magnification and apply immersion medium. Visual the synaptic region by exciting the fluorophores with a 552 nanometer laser, 1%power for the tagRFP fluorophore, and 488 nanometers, 2%power for the GFP fluorophore.

Capture images using a hybrid photodetector and set the gains to prevent overexposure of the fluorophores. Collect Z-stack images to cover the entire synapse using the optimal Z-step size depending on the objective. To analyze the synaptic region begin by loading the images into the CellProfiler 3.1.5 software.

Set up the NameAndType module and determine the synaptic region by hand definition using the diffuse tagRFP expressed from the UIS-115 transgene. Measure the size and fluorescence of the hand-defined synapse by adding the MeasureObjectSizeShape and MeasureObjectIntensity modules to the pipeline. Then calculate the relative integrated fluorescence intensity by adding the CalculateMath module and dividing the integrated intensity units obtained from JSIS37 by those obtained from UIS-115.

When finished export all measurements and calculations. To measure muscle cell area open the image in Fiji software and use polygon selection to carefully trace around a single oblique muscle cell. Adjust the line of the polygon at the end by dragging the anchor dot to improve tracing.

Navigate to the Analyze tab at the top of the software and click Measure to calculate the selected area. For muscle cells with a degenerated or missing region trace the missing area with the polygon selection tool and click Measure again. If there are multiple gaps, trace each one separately.

Calculate the ratio of the gap area to the area of the entire cell. A high ratio indicates a higher extent of muscle degeneration. And if there are no missing regions, the ratio is calculated as zero.

Open Elastic and choose Pixel Classification under Create New Project. Then rename and save the project. Navigate to the input data tab and click Add New, and then Add separate images.

Direct the pop-up window to the folder with the TIF files and select the desired image. In the Feature Selection tab click on Select Features to select all pixel features indicated by the green ticked boxes. The pixel selection in this step is used to distinguish the different classes of pixels in the next step.

Use a Training panel to distinguish between the individual GFP-expressing myosin filaments and unwanted background. Click on Add Label and scrawl unwanted backgrounds and spaces between the filaments to begin classifier training. Add a second label, rename it to Filament, and scrawl over a number of filaments.

Click on Live Update to make sure that the classification has been done correctly. If necessary, add more scrawls to fine tune the training. Once training classifier has been performed, go to the Prediction Export panel and click on Choose Export Image Settings with probabilities as the source.

Make sure that the cutout subregions x and y are ticked, and c is unticked. Select zero and one as start and stop values for c, and Convert Data Type to unsigned 8-bit. Tick Renormalize and then change the output file video to png, renaming the file and directory as desired.

Close the Export settings window and export the image that has just been segmented. Then open the png file in Fiji software. Adjust the threshold of the image using Default settings and apply the skeletonization plug-in, which will create a separate table of results.

This protocol has been used to explore the function of alpha-tubulin acetyltransferase MEC-17 in synapse development. Quantitative analysis of images of the posterior/lateral microtubule or PLM synapses indicates that overexpression of MEC-17 disrupts normal neuron function. Compared to the wild type control animals overexpressing MEC-17 have significantly reduced presynaptic regions and synaptic integrity.

These techniques have also been used to analyze C.elegans models of Charcot-Marie-Tooth type 2 or CMT2 carrying mutations in CMT2-associated genes. Visual assessment of body wall muscle morphology revealed a 2.5 to 3.5-fold increase in defects in the test animals compared to the wild type control. Animals with mutations in fzo-1 and unc-116 experienced loss of muscle striations, accumulation of cellular debris, and muscle fiber degeneration.

While animals with mutations in dyn-1 experienced significantly less defects in muscle morphology. The fzo-1 and unc-116 mutants displayed a five to six-fold increase in the ratio of gap to total single cell area and a dramatically shorter average fiber length compared to wild type animals. For comparisons to be made between samples it is important to determine the most optimal conditions for image acquisition, and to ensure that these conditions are used consistently.

These techniques will allow researchers to compare morphological changes across different genetic backgrounds with quantitative approaches. This reduces the subjectivity and therefore help to enhance the representability of the data generated.