In 1676 Antoni van Leeuwenhoek discovered microorganisms using his own-made single-lens microscope. His self-fabricated biconvex or planoconvex lenses could resolve structures of 700 nm; this resolution has been proven by taking pictures through an original Leeuwenhoek microscope (Ford, 2011). Over the centuries light microscopy has been optimized, e.g., by use of two-lens microscopes, introduction of phase-contrast microscopy and developed into new and very special fields (for reviews see e.g., Keller et al., 2009; Kasper and Huang, 2011; Coltharp and Xiao, 2012). In addition, a combination of various microscopic techniques can be used to obtain high specificity and detailed structural information (Smith, 2012). Special techniques today overcome the resolution limit (~200 nm) of light microscopy (Keller et al., 2009; Kasper and Huang, 2011). This is especially true for electron microscopy (EM) with its various disciplines like transmission EM (TEM), scanning EM (SEM), cryo EM, focused ion beam EM, etc. Modern techniques allow e.g., high-resolution EM of cellular components, initially with a resolution of ~7 Å (Yu et al., 2012), and today even better, down to near atomic resolution of ~4 Å (e.g., Bernecky et al., 2016; Braun et al., 2016; Taylor et al., 2016; Zoued et al., 2016). Even whole small bacterial cells meanwhile have been analyzed with a resolution of ~4 Å (Oikonomou and Jensen, 2016).

Spirochetes are—due to their very special cell body plan—a fascinating group of bacteria; most interestingly they can swim by use of so-called endoflagella. These endoflagella are located between the cell body and the outer membrane, i.e., in the periplasmic space of the Gram-negative bacteria. Flagella are inserted into the cell body at both ends of a cell, they can overlap in the middle region—if only one flagellum is inserted per cell pole one speaks of a 1:2:1(end:middle:end) flagellar arrangement (Leschine et al., 2006). In some cases, like Leptonema illini, the endoflagella are too short to overlap; those cells have, depending on their motility status, hook-shaped ends, (for further details see e.g., Figures 2, 3 in Wolgemuth, 2015). Spirochetes come in two different shapes: either a helix, like e.g., L. illini (Wolgemuth, 2015) or a flat wave, like e.g., Borellia burgdorferi (Charon et al., 2012). A scheme on the cell body plan of B. burgdorferi, L. illini, and Spirochaeta thermophila—the subject of this study—is given in Figure 1.

S. thermophila is a member of the phylum Spirochaetes which contains important disease causing species like Treponema pallidum (syphilis), B. burgdorferi (lyme disease), and Leptospira interrogans (leptospirosis, resulting in renal failure). Recently a phylogenomic analysis identified a 3 amino acid conserved signature indel in the FlgC protein as a unique molecular marker for this phylum, resulting in a proposal for a taxonomic revision (Gupta et al., 2013). The genus Spirochaeta is comprised of anaerobic and facultative aerobic spirochetes that are indigenous to aquatic environments. All cultivated species of the genus Spirochaeta are mesophiles with the exception of Spirochaeta caldaria and Spirochaeta thermophila having optimal growth temperatures of 48–52°C and 66–68°C, respectively. For mesophilic spirochetes, especially pathogenic ones like L. illini (Goldstein et al., 1996) and B. burgdorferi many data are available, including e.g., the structure and assembly of the endoflagella (Zhao et al., 2013, and references therein), or their mode of motility (Goldstein et al., 1994; Motaleb et al., 2000; Charon et al., 2012). To the best of our knowledge, however, no data as to the motility of any thermophilic spirochete is available.

In many cases members of the phylum Spirochaetes are found in habitats characterized by an increased viscosity like mucous membranes. It has been reported that swimming speed increases if such spirochetes are transferred from aqueous media to more viscous ones (see Canale-Parola, 1978 for details). This was taken as explanation for the unique cell body plan of spirochetes with their flagella contained within the outer sheath of the cell body. The argument is that a screw-like movement of the spiral shaped bacteria through highly viscous media like slime overlying mucous membranes should be superior to the attempt to move bacteria in viscous environments by rotation of extracellular flagella.

The very unusual cell body plan of spirochetes has attracted much attention and various hypotheses upon their motility mechanisms have been formulated, see e.g., (Canale-Parola, 1978). The most popular one, namely Berg's model for spirochetal motility states that the endoflagella inserted at the cell poles do rotate and cause counter-rotation of the helical protoplasmic cylinder within the outer sheath. Since flagella can rotate clockwise or counterclockwise or not at all, one can state that a total of 9 different rotation states should be possible. Of those only one—namely no rotation at both cell poles—should result in immotile cells, whilst the 8 other potential rotation states should result in at least four, but very probably five different functional rotation states. Figure 2 characterizes these functional states, which are based on the assumption that the left and right ends of the spiral shaped cells are functionally equivalent.

We analyzed here the swimming behavior of S. thermophila, using a newly developed device (Mora et al., 2014) which can be added onto any upright light microscope to allow analyses at high temperatures under anaerobic conditions. Our data prove the extraordinary flexibility of the cells during swimming, with rapid changes between linear cells, cells possessing at least three different spiral forms and flexing cells, bending by 180°. Our combination of TEM, SEM, and light microscopy studies gives a hint how the various cell morphologies can be correlated with the various flagellar rotation states possible for this spirochete.

The data obtained here and earlier (Herzog and Wirth, 2012; Mora et al., 2014) clearly show that statements on the motility of anaerobic (hyper) thermophilic microorganisms cannot be made if cells are analyzed under non-physiological conditions (=aerobic light microscopy at room-temperature). In the case of S. thermophila light microscopy under such conditions results in the existence of immotile cells of primarily one morphotype—this very clearly is an artifact.

Our swimming movies indicated the existence of at least three different cell morphologies; in combination with SEM analyses even five different cell morphologies could be identified: cells being approximately linear over their total length; cells exhibiting a half-linear half-spiral form; and cells possessing one of three different spiral forms. Very interestingly the cells can rapidly switch between these different morphotypes during swimming. At the moment it would be pure speculation which of the five different cell morphologies corresponds with the five different endoflagella rotation states outlined in Figure 2. As mentioned above, the 1.25 × 0.25 μm (wave length × amplitude) cells very probably represent the state in which both flagella are not rotating, i.e., the immotile state of Figure 2.

Our data show that S. thermophila exhibits a spiral form; its cell body plan therefore does not resemble B. burgdorferi which has flat wave morphology. On the other hand S. thermophila also does not resemble L. illini in its cell body plan, because the endoflagella of S. thermophila exhibit a 1:2:1 arrangement, i.e., they overlap in the middle of the cell. In addition we never observed the hook-like ends, being characteristic of L. illini or L. interrogans (Wolgemuth, 2015). It was brought to our attention that in theory one might be able to estimate the rotational rate of the cell using the cell's shape and their swimming speed. One even might be able to predict the cell stiffness; for such calculations certain assumptions (like: swimming speed is determined mainly by rotation of the helical cell body) have to be made about which we feel not sure. Therefore, these theoretical calculations are not included here. We want to stress, however, the fact that the morphotype of spirochetes well can change. In the case of B. burgdorferi the flat wave form of the cells alters to a rod form, if the endoflagella are not expressed (Motaleb et al., 2000), or the motor does not rotate (Sultan et al., 2015). In the case of Treponema denticola with a 2:4:2 endoflagella arrangement an irregular twisted morphology bas been observed. Both planar and helical regions of the cells were observed with endoflagella wrapped around planar regions or lying along the cell axis in helical regions. In addition, however, also some T. denticola cells with completely right-handed morphology have been observed (Ruby et al., 1997). All these data point to the fact that only a combination of light microscopy (done under physiological conditions to observe motility) with electron microscopy—as we have done here—will allow one to define the cell body plan of spirochetes.

The addition of PVP or MC as gelling agent in our experiments led to a decrease in the swimming speed of S. thermophila; these agents, however not only increase the viscosity, but also elasticity. Interestingly, an increase in viscosity alone—by addition of Ficoll—slowed down, both B. burgdorferi and T. pallidum (Harman et al., 2013). On the other hand the swimming speed of B. burgdorferi did increase by an increase of viscoelasticity via the addition of 1% MC from <1 μm/s to ~10 μm/s (Li et al., 2010). Similar viscoelasticity effects were reported for T. denticola (Ruby and Charon, 1998). We therefore conclude that S. thermophila does not react to a gel-like habitat like the pathogenic spirochetes do by altering their swimming speed. The normal habitat of S. thermophila indeed, is not characterized by marked viscosity or viscoelasticity: the type strain was isolated from a marine hot spring in Kamchatka (Aksenova et al., 1992).

RW planned the study, took some of the movies and wrote the manuscript. MU took most of the movies. GW made all electron microscopic pictures.