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

Watch videos of the Lyme disease spirochete escaping from the bloodstream of live mice!

Most pathogenic microbes that cause systemic infections, regardless of their route of host entry, migrate to the circulatory system, which facilitates their spread throughout the body. These invasive microbes, which include the Lyme disease spirochete B. burgdorferi, eventually exit the bloodstream and penetrate into various organs of the host. Last June in the online journal PLoS Pathogens, a Canadian research group presented some fascinating microscopic video footage of Borrelia burgdorferi traveling within and escaping from the bloodstream of live mice. We may like to think that the unique shape of the spirochete allows it to simply drill through the vessel wall, but the videos suggest that escape from the bloodstream is a little more complex.

Because spirochetes are too thin to observe by light microscopy, Moriarty and colleagues made B. burgdorferi fluoresce by transforming the spirochete with a gfp (green fluorescent protein) plasmid. To prepare the animals, they lifted the skin of anesthetized mice for observation of the underlying dermal microvasculature by fluorescence intravital microscopy (IVM), which allows visualization of cellular events in a living animal. They next injected the fluorescent spirochetes into the bloodstream of the mice, and they examined dermal postcapillary venules under the microscope as the spirochetes traveled through the field of view within the vessels.

The black-and-white video reveals several types of interactions between the spirochetes and vessel wall. The bar graph displayed below the video indicates the proportion of each type of interaction observed. Almost 90% of the contacts are transient, lasting for less than a second. About 10% of the interactions involved crawling or dragging of the spirochete along the vessel wall for up to 20 seconds. As you can see from the bar graph, these short-term interactions, although common, rarely lead to escape of spirochetes from the bloodstream. Perhaps the spirochetes crawl along the wall probing for an escape route from the vessel. When their search fails, as it usually does, they detach and float (or swim) away and try again elsewhere along the vessel wall. Occasionally, a spirochete will remain stuck to the vessel wall for many minutes. One such spirochete can be seen in the video, near the center of the screen. More careful observation of stationary spirochetes in the bloodstream of several mice revealed at least one end deeply embedded with the vessel wall, usually between endothelial cells. It is unclear whether these stationary adhesions are a necessary prelude to exit of the spirochete of the vessel as consistent outward movement of embedded spirochetes was never observed during the observation period, which lasted up to 45 minutes.



The next two videos capture spirochetes in the process of escaping from the bloodstream. The endothelium was stained by injecting the bloodstream with red fluorescent antibody to PECAM-1, a protein found within endothelial junctions. The first video shows how difficult it is for B. burgdorferi to traverse the wall of the venule. The spirochete appears to be stuck as it moves back and forth (reciprocal translation) across the vessel wall for several minutes trying to free itself. The second video shows a spirochete successfully dislodging itself and fleeing from the venule. The average escape time was 10.8 minutes (N = 11 spirochetes). The authors could not clearly determine whether the spirochetes escaped between or through endothelial cells.





Here's the model illustrating the steps in the escape of B. burgdorferi from the bloodstream. The spirochete first contacts and crawls (drag) across the inside surface of the vessel wall. It then crosses the vessel wall end-first. After a long period of back-and-forth motion (reciprocal translation), the spirochete finally escapes into the tissue. It is unknown whether stationary adhesion is necessary for escape.


Moriarty et al. repeated the experiments with B. burgdorferi rendered noninfectious by long-term passage in culture. They found minimal interaction of noninfectious spirochetes with the vessel wall, and not a single spirochete could be found escaping from the bloodstream. This result indicates that specific Borrelia surface molecules that are missing on noninfectious B. burgdorferi mediate interaction with and escape from the bloodstream. What are these B. burgdorferi surface molecules, and which host molecules do they contact in the blood vessel? Past in vitro experiments with cultured mammalian cells by several research groups have revealed a few candidates for such bacterial and host factors. The authors described the roles of these candidates in transient, dragging, and stationary adhesions in live mice in a follow-up study, which I will write about in my next post.

Featured paper

Tara J. Moriarty, M. Ursula Norman, Pina Colarusso, Troy Bankhead, Paul Kubes, George Chaconas (2008). Real-Time High Resolution 3D Imaging of the Lyme Disease Spirochete Adhering to and Escaping from the Vasculature of a Living Host. PLoS Pathogens, 4 (6) DOI: 10.1371/journal.ppat.1000090

The origin of syphilis: a phylogenetic approach

K.N. Harper, P.S. Ocampo, B.M. Steiner, R.W. George, M.S. Silverman, S. Bolotin, A. Pillay, N.J. Saunders, and G.J. Armelagos. (2008). On the origin of the treponematoses: A phylogenetic approach. PLoS Neglected Tropical Diseases 2(1):e148.

The first recorded outbreak of syphilis occurred in Europe in 1495, a few years after Columbus sailed the ocean blue. Was syphilis a New World disease newly introduced into Europe by Columbus and his crew, or was it an Old World disease that simply was not noticed until 1495? A study from Harper and colleagues published last January described a molecular genetic analysis that may have yielded important clues hidden within the genetic material of Treponema pallidum.

The authors first examined the evolutionary relationships among Treponema pallidum strains from subspecies pertenue, endemicum, and pallidum, which are responsible for the diseases yaws, bejel, and syphilis, respectively. The small collection of strains or their DNA was obtained from different patients throughout the past century. The tree illustrated below (Figure 3 of Harper et al.) was constructed from the alignment of 70 SNPs (single nucleotide polymorphisms) and 12 indels (insertions/deletions). The branching pattern indicates that pertenue emerged the earliest. Subspecies endemicum later emerged from pertenue, and pallidum, the agent of syphilis, arose most recently.
The authors also obtained scrapings from yaws skin lesions on two aboriginal children living deep in the rainforests of Guyana. Since these children were members of a population that had been living for generations with minimal contact with the rest of the world, these pertenue strains may be closely related to those present in the Americas before the European explorers arrived. Unfortunately, the samples collected by the authors had degraded extensively by the time the DNA was extracted for analysis. Consequently, the two Guyanan strains could not be included in the phylogenetic analysis shown above; only regions encompassing 17 of the 70 SNPs could be sequenced from the degraded DNA. Nevertheless, they went ahead and aligned the 17 nucleotides with those from the strains used to construct the phylogenetic tree. The alignment revealed that among the nonveneral strains (pertenue and endemicum), only the Guyanan strains had as many as 4 nucleotides that were identical to those of the pallidum strains.

The world map illustrated below (Figure 4 of Harper et al.) depicts the path of sequence changes in the 4 SNPs among the Treponema strains. The dots mark the geographic source of each strain used in the analysis. The red and green colors demark areas of endemicity of the nonvenereal diseases yaws and bejel, respectively, around the year 1900. The map shows that T. pallidum first appeared as pertenue in the Old World and gave rise to the endemicum subspecies, which migrated with humans to the Middle East and Europe. The Old World pertenue or endemicum strain then eventually gave rise to the New World pertenue strain as humans crossed the Bering Land Bridge and spread throughout the Americas. The sequence identity of the Guyanan strains with the pallidum strains at all 4 positions is consistent with the New World strain being introduced back into the Old World as a progenitor of today's syphilis-causing pallidum strains, which are now found worldwide. Clinical evidence also supports the New World model: the nonvenereal skin lesions in the Guyanan yaws patients resembled syphilis chancres rather than the typical skin lesions found with yaws.I do not believe that the results presented in the paper support the New World origin of syphilis. As Harper et al. state in the Discussion of the paper, the close evolutionary relationship of the South American pertenue strains with the pallidum (syphilis) strains is based on a mere four nucleotides. Still, the authors concluded that pallidum arose from a descendant of the New World pertenue that was brought to Europe by Columbus. However, the results do not rule out the possibility that pallidum and New World pertenue strains evolved independently from a common ancestor, such as the "unknown" strain illustrated in the map. The New World strains would need to be included in the phylogenetic tree to distinguish the two possibilities. It was unfortunate that the entire set of 70 SNPs and 12 indels could not be examined in the Guyanan strains.

Chronic Lyme disease in mice?

E. Hodzic, S. Feng, K. Holden, K.J. Freet, and S.W. Barthold. (2008). Persistence of Borrelia burgdorferi following antibiotic treatment in mice. Infection and Immunity 52(5):1728-1736.

Controversy surrounds the management of those Lyme disease patients who continue to experience symptoms following treatment with the recommended course of antibiotics. These persisting symptoms, which include fatigue, sleep disturbances, and concentration difficulties, can be debilitating. The central question that underlies the controversy is whether the symptoms result from Borrelia spirochetes that survive treatment. If so, further treatment with oral or intravenous antibiotics, sometimes lasting for many months or even years, may be warranted, as Lyme disease advocates insist. However, the standard treatment guidelines do not recommend antibiotic treatment regimens lasting longer than one month. Supporters of the guidelines contend that post-treatment symptoms are not due to active Borrelia infection and that further treatment with antibiotics are not supported by clinical studies. Critics of the guidelines disagree with the interpretation of the clinical studies and cite cell culture and animal studies that suggest survival of Borrelia following antibiotic treatment. Insurance companies cite the guidelines in refusing to pay for costly long-term antibiotic treatment.

Early in 2008, Hodzic and colleagues at UC Davis published a study that examined the fate of Borrelia burgdorferi in infected mice treated for one month with ceftriaxone, an antibiotic commonly used to treat Lyme disease. The investigators succeeded in visualizing spirochetes in the tissues of a few mice. A typical example is presented in the image below (Figure 1 of the Hodzic et al. paper). Panel B shows a tissue section of a joint from a mouse that was treated with ceftriaxone initiated 4 months following infection with B. burgdorferi. The tissue was examined one month after antibiotic treatment was completed. Immunohistochemical staining clearly revealed a solitary spirochete (arrow). Panel A shows a joint from an infected mouse that was sham treated with saline. As expected, spirochetes were observed in tissues from most of the control mice, with up to four appearing in a section.

One could argue that the spirochetes observed in the ceftriaxone-treated mice were simply dead microbial carcasses awaiting removal. This interpretation appeared to be supported by the marked decrease in B. burgdorferi DNA copy number in mice tissues with time, as measured by quantitative PCR. However, ticks that fed on these mice were able to acquire and transmit the spirochetes to uninfected SCID mice, indicating that the spirochetes remained infectious despite ceftriaxone treatment. Yet all attempts to culture the bacteria from the treated mice, the ticks that fed on the mice, and even the SCID mice that the ticks fed on failed, even though B. burgdorferi was detected in the mouse tissues and ticks by PCR. In contrast, B. burgdorferi was successfully cultured from all control (saline treated) mice and the ticks that fed on the mice. These results suggest that B. burgdorferi that remained in treated mice were alive and infectious (and transmissible) but were impaired in their ability to replicate.

The most important question for those suffering from post-treatment symptoms is whether the spirochetes that survive antibiotic treatment cause clinical symptoms; the infectiousness of the spirochetes is less relevant. The authors microscopically examined the joints and hearts of the SCID mice for signs of inflammation. SCID mice are especially susceptible to developing severe inflammation when infected with B. burgdorferi. Nevertheless, inflammation was not detected in the SCID mice that acquired the disabled (yet infectious) spirochetes from ticks that previously fed on antibiotic-treated mice. This result does not rule out the possibility that disabled spirochetes contribute to post-treatment symptoms in humans by a microscopically undetectable mechanism. If residual spirochetes do indeed elicit clinical symptoms, then elimination of the spirochetes would be desired. Additional treatment with ceftriaxone may not be the best choice since it targets the cell walls of actively replicating bacteria. The mouse model developed by the UC Davis group will allow the investigators to test different treatment approaches for elimination of these persisting spirochetes.