Description

How large are C. elegans’ tissues? To our surprise, we could not find any quantification of C. elegans’ tissue volumes. By reaching out to the community, we received the following suggestions on how to quantify tissue volumes: First, using tissue-specific fluorescent reporter transgenes, confocal microscopy, and automated segmentation. Second and third, using electron microscopy serial sections (Altun et al. 2020) or methylene blue/pararosaniline-stained cross-sections (McGee et al. 2011) and manual or semi-automated segmentation. Here, we decided to use an alternative approach by directly measuring cell and tissue volumes from the Virtual Worm model (Grove and Sternberg 2011, http://caltech.wormbase.org/virtualworm/). The Virtual Worm is a three-dimensional model of a young adult C. elegans containing cell shape and position based on various sources, including the WormAtlas (Altun et al. 2020), the “Mind of the Worm” (White et al. 1986), and the “C. elegans Atlas” book (Hall and Altun 2007). We determined tissue volumes in the free software Blender by measuring individual cell volumes using the plugin “3D print toolbox” and directly with the plugin “Mesh volume calculator” (see Methods) (Figure 1a). Our quantifications revealed that the intestine was the largest tissue, followed by the hypodermis, gonads (including the volume of the rachis containing the non-somatic germ cells), body wall muscles, cuticle, pharynx, and neurons (Figure 1b, Extended Data). The majority of the total model volume was contributed by the four tissues intestine, hypodermis, body wall muscle, and gonads (including the rachis and germ cells), and each of these roughly contributed by a quarter (Figure 1c, bottom). However, several non-cellular structures were not included in these measurements, for example, the pericellular pseudocoelom, additional spaces in the uterus, and especially the large extracellular cuticle that together could take up 1/6^th or more of a total animal. The original Virtual Worm was based on resources that mostly relied on electron microscopy and dehydrating fixation protocols that reduce volume. This likely leads to a general underestimation of volumes. This is also indicated by a comparison between the total model volume of 1.4 million μm³ and the 2.2 million μm³ estimated for a live adult hermaphrodite (Grove 2012). Taken together, our calculations are a zeroth-order approximation of tissue size, however, we can infer relative tissue sizes. Thus, we conclude that intestine > hypodermis > gonads > body wall muscles > pharynx > neurons.

C. elegans contain several syncytial cells, for example, the pharyngeal pm1 cell (6 nuclei) or the hypodermal hyp7 cell (139 nuclei). Additionally, intestinal cells are polyploid and contain 16 diploid genomes (“32C”) at the young adult stage. We collected this information from the literature for all cells (Figure 1d, Extended Data). This allowed us to make several observations. For example, roughly one-third of somatic nuclei were contributed by neurons (Figure 1d, top, dark blue), but neurons minimally contributed to the final volume, with the caution that their volume might be underestimated due to the difficulty of modeling fine axonal structures (Figure 1c, top, dark blue). Germ cell nuclei are tightly packed, and although they accounted for more than half of all nuclei (Figure 1d, bottom, light blue), they took up only 1/8^th of the animal’s volume (Figure 1c, bottom, light blue).

We then analyzed individual cell volumes. The largest cell was the syncytial hypodermal hyp7, more than 10x larger than the second-largest cell, the polyploid intestinal int5L/R (Figure 1e). Intestinal cells were the largest cells on average (Figure 1e). To account for multiple genomic copies per cell (in syncytial- and polyploid cells), we normalized the cell volumes by the number of diploid genomes. This revealed that, on average, body wall muscle cells have a more than 5x larger volume per genomic copy than other cells (Figure 1f). mRNA and protein synthesis or their turnover might differ in these cells to support a larger cell volume with the same genomic copies. Additional observations could be made from the data. For instance, tissues that contributed most to the final animal volume also contained larger cells on average (see body wall muscle, intestine, hypodermis, and gonads in Figure 1f). The cell volume per diploid genome for 86% of cells was below 1,000 μm³, and for 70% of cells even below 200 μm³ (Figure 1h). Absolute estimates for single cells might be underestimated, as already noted above. Their comparison also has to be taken with caution because their modeling does not contain fine morphological structures.

In conclusion, we provide an approximation of tissue volumes of young adult C. elegans based on the Virtual Worm project. These values are rough estimates, and future experiments could determine precise volumes and variability between animals. Our results might be useful for interpreting relative tissue contributions of whole C. elegans for transcriptomics, proteomics, epigenomics, or when performing gene-specific detection by qRT-PCR and Western blot. The approximations of cell volumes and their relationship to genomic copies might be useful for the planning and interpretation of single-cell/single-nuclei methods in C. elegans and related nematodes.