Mycoplasmas are exceptional subjects for imaging. The specialized manner in which Mycoplasma pneumoniae, as an example, interacts with host cells requires the attachment (or terminal) organelle, a polar extension of the cell. This important virulence-associated feature couldn’t have been predicted de novo from any technique; microscopy was essential. Knowing about that structure and similar ones in other mycoplasma species has led to advances in understanding interactions with the host, cell movement, and cell division, and has even contributed to nanotechnology.
SEM of Mycoplasma genitalium G37. We can see the outside, but not the inside! Courtesy of J.M. Hatchel and the Miami University Center for Advanced Microscopy and Imaging.
What’s the best way to approach imaging of mycoplasmas? The answer depends heavily upon what the question being asked is. Continuing with the example of the attachment organelle, Biberfeld and Biberfeld (Journal of Bacteriology, 1970, 102:855-861) established its presence using both scanning electron microscopy (SEM) and transmission electron microscopy (TEM), and both those techniques provided different kinds of information. TEM images of thin sections of M. pneumoniae cells clearly showed the highly differentiated, electron-dense, cytoskeletal core of the attachment organelle, and although we now use techniques for processing of cells for SEM that result in greater reproducibility than back in 1970, their SEM images of whole cells strongly indicated the differentiated nature of this organelle.
Compared to SEM, TEM is much better suited to the task of identifying internal structures. However, when these structures occupy very small volumes, they may be missed unless a lot of thin sections are examined. Moreover, one can’t reliably extrapolate the shape of a cell from thin sections. An early description of Mycoplasma iowae (Jordan et al., 1982, International Journal of Systematic Bacteriology, 32:108-115) showed TEM images that suggested a coccobacillary shape, which has been refuted numerous times by SEM (e.g., Gallagher and Rhoades, 1983, Avian Diseases, 27:211-217). And SEM images of attachment organelle cores prepared by detergent extraction (Hatchel and Balish, 2008, Microbiology, 154:286-295), though subject to less consistent extraction of peripheral components, lead to greater reproducibility in measuring dimensions of internal structures than TEM, since sectioning is unlikely to be strictly parallel to the object being measured.
The bottom line: use the technique that answers your question with the greatest reproducibility, but don’t reinvent the wheel; excellent protocols for imaging of mycoplasmas by TEM and SEM, as well as newer and fancier techniques, have been published. And don’t rely on the wrong kind of data to interpret your results.