Viewing the arrangement of Borrelia burgdorferi flagella by electron cryotomography

The most peculiar feature of spirochetes may be the location of their flagella, the thin motility structures that propel bacteria through liquids. Flagella typically extend out from the surface of bacteria into the surroundings. Spirochetes, being not so typical, keep their flagella hidden in the periplasm between the cytoplasmic and outer membranes (see figure). For example, the Lyme disease spirochete Borrelia burgdorferi has 7-11 flagella attached near each end of the "protoplasmic" or cell cylinder, with each flagellum extending through the periplasm towards the center of the spirochete. The flagella impose a flat-wave shape (not a spiral shape!) on B. burgdorferi by wrapping around its protoplasmic cylinder.

from Rosa et al., 2005

How do flagella that are located in the periplasm drive the spirochete through the medium? B. burgdorferi motility is thought to require the rotation of its flagella against the cell cylinder, causing the cell body to gyrate.

B. burgdorferi flagella often appear as a bundle when observed by standard transmission electron microscopy. Here is one such image from a 2000 study revealing at least 10 flagella in a cross section of B. burgdorferi. With the flagella arranged in this manner, it is difficult to imagine how the flagella that are not in direct contact with the cell cylinder could contribute to its gyration.

A study by Charon and colleagues in the January 2009 issue of Journal of Bacteriology suggests that the flagellar bundle is an artifact of the standard techniques used to prepare the samples for electron microscopy. They employed the emerging technique of electron cryotomography to avoid the fixation and staining procedures that often introduce artifacts into samples. Electron cryotomography consists of the following steps:

1. To preserve structure, the specimen is plunge frozen at -165°C or less. Fixing or staining is not necessary.
2. While maintaining the sample at the ultralow temperature, 2D projections of the sample are obtained at different angles by transmission electron microscopy.
3. Computer software assembles the 3D structure of the specimen from the 2D projections.

The software also permits slices of the specimen to be observed without having to actually perform thin sectioning.

Here's a cross-section of B. burgdorferi as viewed by electron cryotomography. Note that the flagella are arranged in a single layer within the periplasm, not in a bundle.

Figure 1 of Charon et al. Bar, 50 nm.
PFs, periplasmic flagella; PS, periplasmic space; PM, plasma (or cytoplasmic) membrane; OM, outer membrane.

A longitudinal slice through the periplasm of B. burgdorferi reveals nine flagella neatly arranged in a parallel fashion along the surface of the protoplasmic cylinder. The authors refer to this array as a "flat ribbon." Each flagellum in the ribbon is separated by ~3 nm, allowing each to rotate in the same direction without interference from neighboring flagella.

Figure 5 of Charon et al. Bar, 200 nm.

3D reconstruction of a section of the spirochete illustrates the flat ribbon of flagella (in red) wrapping around the cell cylinder (in blue). Only a section of the cell cylinder is shown, and the outer membrane has been removed from the image.



These new images support a model for for B. burgdorferi motility that was first described back in the 1990s. In this model, the rotation of the flagella against the cell cylinder generates gyrating waves that progress backwards along the cell body. As explained in the discussion of the Charon et al. paper, it is conceivable that all 7-11 flagella must lie against the cell cylinder as a flat ribbon to exert the force necessary to generate the waves; a flagella bundle may not exert enough force. The torque generated by the rotating flagella causes a counter rotation of the cell cylinder (panel a below). The backward-propagating, gyrating waves push the spirochete through the medium. Flagella arranged in a bundle would not generate enough torque because of potential interference between rotating flagella (panel b).

Figure 8 of Charon et al. a. Flagella arranged in a flat ribbon. b. Flagella arranged in a bundle.

This model also explains why B. burgdorferi moves so well through viscous gel-like material such as the extracellular matrix; the gel provides traction for the backward-progressing waves to drive the spirochete through the medium.

Here's a movie animating B. burgdorferi motility, first presented at a meeting in 2001 .


SOURCE

You can also see real B. burgdorferi gyrating and generating backward-moving waves in a movie embedded in Dr. Nyles Charon's website.

Reference

N. W. Charon, S. F. Goldstein, M. Marko, C. Hsieh, L. L. Gebhardt, M. A. Motaleb, C. W. Wolgemuth, R. J. Limberger, N. Rowe (2009). The Flat-Ribbon Configuration of the Periplasmic Flagella of Borrelia burgdorferi and Its Relationship to Motility and Morphology Journal of Bacteriology, 191 (2), 600-607 DOI: 10.1128/JB.01288-08

The Lyme disease spirochete hijacks fibronectin to escape from the bloodstream

The glycoprotein fibronectin is a component of the molecular mesh known as the extracellular matrix, which not only provides physical support for our cells but also directs cellular activities during embryonic development, tissue repair, and other processes. High levels of soluble fibronectin (300 μg/ml) are also found in our bloodstream, where it quietly circulates until it is recruited to stabilize clots and promote wound repair.

Fibronectin has a modular organization consisting of binding sites for various matrix and cell surface molecules. Examples include attachment sites for integrins (labeled "Cell" in the figure below) and glycosaminoglycans (labeled "Heparin"), which are exposed on the surface of the endothelial cells that line our blood vessels.

from Wierzbicka-Patynowski, 2003

B. burgdorferi is injected into the skin by an infected tick and spreads outward within the dermis, causing the familiar "bulls-eye" rash in some Lyme disease patients. The spirochete may eventually enter the bloodstream so that it can spread to other tissues. While in the bloodstream, the spirochete is surrounded by fibronectin. In fact, B. burgdorferi attaches to fibronectin in vitro, suggesting a role for fibronectin in Lyme disease.

It turns out that B. burgdorferi exploits the adhesive properties of plasma fibronectin to bind to the vessel wall before escaping into the surrounding tissue. Like their earlier work, which I described in my last post, this follow-up study by Norman and colleagues was conducted with fluorescent B. burgdorferi injected into the veins of live mice. The interactions of the spirochete with the capillary wall were observed by fluorescent intravital microscopy. The earlier study revealed that most of the interactions were transient, lasting for less than a second. B. burgdorferi was also seen crawling (dragging) along the vessel wall, which was followed by escape into the tissue or by stationary adhesion, a more intimate association with the vessel wall. Stationary adhesion could also be followed by extravasation and escape of the spirochete into the tissue. Both stationary adhesion and escape usually occurred between the endothelial cells lining the vessel wall. Each type of interaction (transient, dragging, and stationary adhesion) was quantitated by counting.

In their follow-up study, the authors demonstrated that coinjection of anti-fibronectin antibody and B. burgdorferi into the bloodstream of the mice diminished all catagories of interactions (transient, dragging, and stationary adhesion) by at least 90%. Control antibody (goat IgG) had no effect. These results indicate that fibronectin has a key role in mediating the attachment of B. burgdorferi to the microvasculature. Since fibronectin could potentially bind to GAGs and integrins on endothelial cells, the investigators also coinjected B. burgdorferi with peptides or antibodies known to block attachment of fibronectin to these targets. They found that the GAG-specific peptide reduced the interaction of B. burgdorferi with the vessel wall, whereas the integrin-specific peptide and antibodies had little effect. Thus, fibronectin may serve as a molecular bridge linking B. burgdorferi to GAGs displayed on the endothelial cells lining the blood vessel.

Which B. burgdorferi factor is involved in adherence to the vessel wall in vivo? Past studies had shown that the borrelial protein BBK32, a known fibronectin binding protein, mediated attachment of B. burgdorferi to fibronectin in vitro. Therefore, BBK32 was a logical candidate. To determine whether BBK32 was involved in vascular interactions in the mouse model, the research team employed a noninfectious B. burgdorferi strain that had lost bbk32 and other genes during long-term culture. The noninfectious strain failed to interact with the vasculature in the mouse. However, expression of BBK32 restored the ability of the noninfectious strain to transiently interact and drag along the vessel wall but only partly restored stationary adhesion. This results suggest that although BBK32 plays a role in vascular adherence, other bacterial factors are also involved.

Many pathogens have been shown to interact with fibronectin and GAGs in vitro. However, this study is highly significant as it is the first to demonstrate a role for these host molecules in bacterial adherence to the microvasculature in a living animal. Other spirochetes such as Treponema pallidum and Leptospira also express fibronectin binding proteins. Hence, the mechanism employed by B. burgdorferi to escape from the bloodstream may be similar for all disease-causing spirochetes. Moreover, other microbial pathogens have been shown to stick to fibronectin and GAGs in vitro. Thus, a large number of invasive pathogens may employ similar mechanisms to spread to different tissues via the bloodstream.

M. Ursula Norman, Tara J. Moriarty, Ashley R. Dresser, Brandie Millen, Paul Kubes, George Chaconas (2008). Molecular Mechanisms Involved in Vascular Interactions of the Lyme Disease Pathogen in a Living Host PLoS Pathogens, 4 (10) DOI: 10.1371/journal.ppat.1000169