Download Swimming Behavior of Borrelia burgdorferi: Eukaryotic Flagella-like Waveform and more Study notes Microbiology in PDF only on Docsity!
Proc. Natl. Acad. Sci. USA Vol. 91, pp. 3433-3437, April 1994 Microbiology
Borrelia burgdorferi swims with a planar waveform similar to that of
eukaryotic flagella
(motilibty/perplasmic flagelia/spirochetes)
STUART F. GOLDSTEINt, NYLES W. CHARON, AND JILL A. KREILINGt
*Department of Genetics and Cell Biology, University of Minnesota, St. Paul, MN 55108; and tDepartment of Microbiology and Immunology, P.O. Box 9177, Health Sciences Center, West Virginia University, Morgantown, WV 26506-
Communicated by Allan M. Campbell, November 1, 1993 (received for review August 10, 1993)
ABSTRACT Borrelia burgdorferi is a motile spirochete
with multiple internal periplasmic flagella (PFs) attached near each end of the cell cylinder; these PFs overlap in the cell center. We (^) analyzed the (^) shape and motion of wild (^) type and PF-deficient mutants using both photomicrography and video microscopy. We found that swimming cells resembled the dynamic movements of eukaryotic flagella. In contrast to helically shaped spirochetes, which propagate spiral waves,
translating B. burgdorferi swam with a planar waveform with
occasional axial twists; waves had a peak-to-peak amplitude of 0.85 am and a wavelength of 3.19 (^) jAm. Planar waves began full-sized at the anterior end and propagated toward the back end of the cell. Concomitantly, these waves gyrated counter- clockwise as viewed from the posterior end along the cell axis. In nontranslating cells, wave propagation ceased. Either the waveform of nontranslating cells resembled the translating form, or the cells became markedly contorted. Cells of the PF-deficient mutant isolated by Sadziene et al. [Sadziene, A., Thomas, D. D., Bundoc, V. G., Holt, S. C. & Barbour, A. G. (1991) J. Clin. Invest. 88, 82-92] were found to be relatively straight. The results suggest that the shape of B. burwdorferi is dictated by interactions between the cell body and the PFs. In addition, the PFs from opposite ends of the cell are believed to interact with one another so that during the markedly distorted nontranslational form, the PFs from opposite ends rotate in opposing directions around one another, causing the cell to bend.
Spirochetes have several attributes that probably contribute to their ability to swim in gel-like media (1-6). Most-but not all (see below)-species are helical or have helical portions (4, 7); this morphology allows them to bore their way through these media in a corkscrew-like manner (1, 4). In addition, these bacteria have periplasmic flagella (PFs) between the outer membrane sheath and the^ cell cylinder (4, 7). Several lines of evidence indicate that the PFs are (^) directly involved in (^) motility (4, 8), and recent results with (^) protruding PFs indicate that the PFs rotate similarly to the external flagella of (^) rod-shaped bacteria (^) (9-11). During translation, the PFs and cell body interact with one another and probably form a more rigid propeller than external flagella alone (1, 11, 12). We report here that Borrelia burgdorferi, the causative agent of Lyme disease (13), swims by using helical PFs to produce a nonhelical cell shape. The PFs of B. burgdorferi are left-handed helices of defined helix pitch (1.48 ,um) and diameter (0.28 ,um) (10). Other spirochetes have been shown to have left-handed PFs but ofdifferent helix dimensions (10). B. burgdorferi has multiple PFs attached at each end that overlap in the cell center (13, 14). However, whereas most other spirochetes have cell bodies that are clearly helical, we report here that swimming cells of B. burgdorferi are planar.
Moreover, cell translation is accomplished by producing posteriorly propagating planar waves resembling those found in eukaryotic flagella.
MATERIALS AND METHODS
Organisms and Culture Conditions. Strains HB19, B31, and both avirulent and fresh isolates of strain 297 (fewer than three in vitro passages from hamsters) ofB. burgdorferi were provided by R. C. Johnson (University of Minnesota, Min- neapolis). The spontaneously occurring motility mutant of strain HB19 lacking PFs was provided by Alan Barbour (University of Texas, San Antonio) (8). Cells were grown in BSK medium at 350C (15).
Light Microscopy. Approximately 5 jA of a cell suspension
was placed on a slide and covered with a cover glass (22 X 22 mm) supported by a mixture of paraffin and Vaseline. For observations of cells swimming in methylcellulose, a drop of cells in culture medium was mixed on the slide with an approximately equal volume of 1% or 2% methylcellulose
(2% =^4 N-s m-2 =^4000 cP; Matheson) in salts buffer (SB;
14.1 mM NaCl/12.6 mM NaHCO3/5.4 mM KCl/1.5 mM
MgCl2/0.1 mM CaCl2/46 mM Na2HPO4/3.5 mM NaH2PO4,
pH 7.7). For measurements of swim speeds in a pure liquid, cells were centrifuged for 10 min at -745 x g at 40C, rinsed twice in cold SB, and resuspended in the original volume of cold SB. Because cells in SB tended to adhere to glass, for measurements of swim speed, cell suspensions in SB were mixed on the slide with an approximately equal volume of a 1% solution of Ficoll (400 kDa; Sigma) in SB. This concen- tration of Ficoll slightly increases viscosity without produc- ing a^ gel-like structure (3). More importantly, Ficoll inhibited the attachment of cells to the glass. To avoid surface effects related to cell motion, swim speeds were measured on cells translating at^ least^ a^ few^ micrometers from glass surfaces. Tethered cells were obtained either by using latex beads coated with (^) antibody H6831 (16, 17) or (^) by screening for cells spontaneously attached to the glass surface. Photomicrography. Cell^ motions^ were^ recorded^ at^ room temperature, with either dark-field multiple-exposure photo- graphs or video sequences and using either Zeiss or Leitz optics (12, 18). Multiple-exposure photographs displaying nonoverlapping images were taken with the film moving through the camera (18). Video sequences were taken in field
mode with a model 72X CCD camera (DAGE-MTI, Michigan
City, IN) and recorded in Super VHS with a model AG-6720A recorder (Panasonic). Either stroboscopic (18) or mercury arc lamp (12) illumination was used. As determined with a stage micrometer, the images were not reversed. Prints of video fields were made with a model UP-910 printer (Sony). Video sequences were taken with phase, dark-field, or Nomarski optics. Computer-assisted measurements ofwave-
Abbreviations: (^) PF, periplasmic flagellum; CCW, counterclockwise. tTo whom reprint requests should be addressed.
3433
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
3434 Microbiology: Goldstein et al.
I/it
fv (^) '~- _
FIG. 1. Strain B31 in BSK medium. (^) (a) Dark-field photograph of cell without protrusion. (^) (b) Cell with (^) protrusion. Reverse-contrast dark-field video fields showing progression of a wave (^) away from the cell body. Line indicates same winding in each field. Interval between images, 1/60 s. (Bar = (^2) pm.)
forms were made (^) from video and multiple-exposure prints (12). Data^ points were spaced 0.15 pm apart. Measurements of swim (^) speeds were made on the monitor screen from phase-contrast (^) sequences. Results are expressed as means + SD.
RESULTS
Swim Pattern. B. burgdorferi cells of all strains (^) including fresh isolates often swam for a few seconds (exhibiting a "translational" form), stopped briefly (exhibiting a "non- translational" form), and resumed swimming in either the same or the reverse direction. Tethered cells exhibited (^) pat- terns of stopping and starting similar to those of freely swimming cells and had similar waveforms. As with other bacteria, these tethered cells were more convenient to record and analyze for extended periods than cells that were swim- ming freely. Motion of Protruding PFs. We have previously observed occasional protrusions on motile cells; these protrusions were composed of PFs surrounded by an outer membrane sheath; in other species these protrusions were found to rotate (10). We have now found that protrusions on B. burgdorferi also propagated waves, indicating rotation of the internal PFs (Fig. 1). Protrusions were rarely seen on strains 297 and HB19 but were frequently seen on strain B31. Form of Translating Cells. The (^) morphology of free- swimming translating cells and tethered cells was (^) analyzed in detail. Cells looked the same in both methylcellulose and a
).-
It& IK
4
l:
.i
Ib F
I
FIG. 2. Cell of (^) strain 297 in 0.5% methylcellulose, tethered to a cover (^) glass at end near (^) bottom offigure. The cell is gyrating as waves propagate to bottom (^) of figure. Cell is seen flat-on in a and edge-on in b. Axial length of cell is equal in all images. (Bar = 1 pum.)
pure liquid. Although we concentrated on nonvirulent strain 297, similar results were observed for strains (^) B31, HB19, and fresh isolates of strain 297. Free-swimming translating cells and tethered cells exhibited a characteristic form. (^) Several lines of evidence suggest that the cells had a flattened waveform rather than being a circular helix. First, the wave- form could be oriented parallel (Fig. 2a), perpendicular (Fig. 2b), or oblique to the focal plane of the microscope. When a cell was oriented (^) parallel to the focal plane, the entire waveform usually went in and out of focus as a unit as the plane of the (^) microscope's focus was varied. Second, when the waveform was oriented perpendicular to the focal plane, the cell (^) appeared as a line. (^) Such cells had a beaded or a dashed appearance due to a series of alternating lighter and darker (^) regions (Fig. 2b) rather than being a line of uniform density; a circular helix would appear as a series of slanted lines (19). Similar observations have been (^) made on planar eukaryotic flagella (20). Finally, the length of the cell, as measured along the axis of the (^) waveform, was the same whether the waveform was parallel to or perpendicular to the focal plane (Figs. 2 and (^) 3). These results indicate that (^) the observed variations of the waveform were not due to a periodic straightening and bending of the cell. However, occasionally (<5% ofcells) a slight helical form was detected, and both right- and left-handed forms were observed. Even
(-44 (- (
(II
FIG. 3. Same cell as in Fig. 2. Waves are (^) propagating toward bottom of (^) figure. Bend indicated (^) by arrowhead is (^) moving from right to left. Focus is slightly above the cell, so that the bend is (^) moving upward in the first three (^) images and downward in the last three (^) images. It (^) is therefore rotating CCW as viewed from the bottom of the figure. (Bar =^1 pm.)
Proc. Natl. Acad. Sci. USA (^91) (1994)
-f.
a
.,
3436 Microbiology: Goldstein et al.
around one another and consistently did so in a^ right-handed sense (n >^10 pairs of cells; Fig. 8).
f /
t f:
FIG. 6. Cell of strain 297 in BSK medium. Nontranslating cell beginning to^ resume^ translation^ toward top of^ figure.^ Propagation^ of waves is^ beginning at^ the anterior end^ (arrows),^ with^ no^ noticeable change at the^ other^ end.^ Interval^ between^ images, 1/60^ s.^ (Bar^ =^1
Am.)
rapidly moved closer together (Fig. 5). The^ deformation
could become^ extreme^ enough to^ produce a^ markedly^ dis-
torted shape, causing the cell to bend in the midregion
resembling a U shape (Fig. 5). These^ distorted^ cells^ appeared
to retain some of the planar bends. Video sequences viewed
in slow motion suggest that the changes in^ cell^ shape involved
a relative gyration of the cell ends. Also, an apparent un-
winding could be seen in pairs of intertwined cells, which, as
noted above, assumed^ a^ helical form^ as^ they^ wound around
one another. When translation stopped, these intertwined
cells separated from^ one^ another^ as^ their helix^ diameters
increased (Fig. 4b). As soon as waves began to be propagated
along the cells, they resumed the^ translational^ mode. The end
that was to become the anterior one could often be seen to
initiate wave propagation (Fig. 6); propagating waves^ were
apparent at^ the other end^ within^ 0.1-0.2^ s.
PF-Deficient Nonmotile Mutants. Sadziene et^ al.^ (8) re-
cently reported the^ isolation^ and^ characterization of^ a^ PF-
deficient nonmotile mutant ofB. burgdorferi strain^ HB19 (8).
This mutant was reported to be straighter than wild-type
HB19 yet appeared helical with lower amplitude and greater
pitch. To^ characterize^ the cell^ bodies of^ the^ mutant^ in^ detail, we examined cells by high-magnification dark-field^ and
Nomarski microscopy. Cells of wild-type HB19 resembled
those of strain 297 (Fig. 7a). Individual mutant cells were straight or slightly curved and often had a slightly hooked form at^ one end^ (Fig. 7b). These results^ suggest that the^ PFs are involved in determining the shape of the entire cell. In contrast to the observations of Sadziene et al. (8), no helical component was evident in the individual mutant cells. How-
ever, the PF-deficient mutants tended to form chains of
unseparated cells. The cells within a chain often wrapped
DISCUSSION
How is the shape of B. burgdorferi determined? Spirochetes are generally characterized as being helically shaped^ orga- nisms (7, 24). However, there have been conflicting reports, especially in the older literature, about the presence of flat waveforms in spirochetes (25-32). We propose that the intact B. burgdorferi cell conforms to its varied waveforms as a consequence of two different mechanical interactions. First, we suggest that the PFs interact with the cell cylinder. The motility mutant of Sadziene et al. (8) that lacks PFs had cell bodies that were relatively straight. Although it is conceiv- able that this mutant suffered a pleiotropic mutation, which altered both cell^ cylinder shape and PF^ synthesis,^ the results suggest that the PFs influence the^ shape of the entire^ cell.^ In addition, this hypothesis is consistent with models explaining the hook- and spiral-shaped ends of Leptospira (4, 5) and the bent-end morphology of Treponema phagedenis (11, 12). The PFs in these organisms are relatively short, do not overlap, and influence the shape of the cell in the specific domain where the PFs reside. We propose that the second major type of interaction occurs between the two PF bundles. This interaction causes the shape of the cells to be altered and is most obvious during the (^) flexing, which leads to the U-shaped form. Central to this hypothesis is the^ evidence^ that the PFs^ overlap^ in the midregion of the cell (14). Electron microscopy reveals that
these overlapping PFs form a bundle and are not dispersed
around the cell in the periplasmic space (14). Because B.
burgdorferi PFs have a defined helix shape (10) and protru- sions derived from the PFs propagate helical waves as do
those of other spirochetes (10), the PFs apparently rotate.
During flexing, we propose that the PFs from opposite ends wind around one another and rotate in opposing directions. Consequently, the^ cell^ is^ distorted, sometimes^ bending^ in^ the midregion and^ forming the^ U^ form. We have^ seen^ similar
distortions in^ nontranslating cells of^ Treponema pallidum
(N.W.C., S.F.G., and S.^ Norris, unpublished data), and similar forms have been reported in Spirochaeta aurantia (1, 33). All three species have^ overlapping PF^ bundles (7). In contrast, Leptospira and T. phagedenis fail to form markedly distorted shapes during flexing; their PFs are relatively short and do not overlap. Thus, we suggest that for the ability to form contortions, the PFs must overlap and interact with those of the opposite end. The waves on individual cells were observed to gyrate
CCW. This direction of gyration is consistent with the
observed wrapping ofpairs ofmotile cells around one another in the right-handed sense (a wave gyrating CCW as it prop- agates posteriorly would wrap around another cell in the right-handed sense). The torque generated by these gyrating waves must be balanced (4, 34). We propose that as the propagating planar waves gyrate CCW, the cell cylinder rolls clockwise (CW) about the cell axis. This proposal is analo- gous to the swimming motion of Leptospira: the CCW gyrating left-handed spiral wave is balanced by the CW roll of the right-handed cell cylinder (4, 5).
The observed swim speed of 4.25 jm/s of B. burgdorferi
in pure liquids can be compared to expected values. For long,
FIG. 7. Strain HB19 in growth medium. (a) Wild-type cell. (b) Mutant cell lacking PFs. (Bar =^1 pm.)
Proc. Natl. Acad Sci. (^) USA (^91) (1994)
Proc. Natl. Acad. Sci. USA 91 (1994) 3437
FIG. 8. Intertwined mutant cells of strain HB19 lacking PFs, wound around one another in the right-handed sense in growth medium. Focus is slightly above the cells. (Bar =^1 gm.)
thin cells, the ratio of swim speed (relative to the medium) to speed of wave propagation (relative to the cell) can be predicted for planar sine waves (35), planar meander-like waves consisting of circular arcs and straight segments (36), and helical waves (37). For an infinitely thin cell, the pre- dicted values of swim speed/wave speed for the five cells whose speeds were measured are 0.21 ±^ 0.02 for a planar sine wave (35), 0.24 ±^ 0.02 for a planar meander-like wave (36), and 0.23 ±^ 0.02 for a circular helix (37). Assuming a diameter of 0.2 ,um and a wavelength of 3.35 ,um and using the correction of Cox (38) for cells of finite thickness as adapted
by Chwang et al. (37) gives predicted values of swim speed/
wave speed =^ 0.13 + 0.01 for a planar meander-like wave (36) and 0.14 ±^ 0.01 for a circular helix (37). These latter estimates are in good agreement with the measured values of 0. 0.02. The similarity in the values predicted for planar and helical waveforms suggests that swim speed is not a major selective factor in the determination of cell shape. However, these calculations and measurements are for swimming in pure liquids, which may be different than the in vivo condi- tion. Kimsey and Spielman (6) have shown that the swim speed ofB. burgdorferi increases markedly as the medium becomes more gel-like, as it does for other spirochetes (1). However, these authors report a swim speed of only 1.7 ,um/s in^ growth medium. Their cells swam at 34.9 pm/s in^ 1%^ methylcellu- lose, indicating a^ wave speed of -34.9 (^) Am/s. Assuming^ a similar wave speed and cell shape in growth medium, the expected swim^ speed would^ be^ 0.12^ 34.9^ um/s =^ 4. ,um/s, which is similar to the results^ reported here.^ The reason for their low measured value^ in^ growth medium^ is^ not clear. (^) However, B. (^) burgdorferi can make (^) frequent, brief stops without^ a^ change in^ cell^ shape, so^ that its^ average^ swim speed can^ be less than its^ swim^ speed during^ wave^ propaga- tion. Because B. burgdorferi stops and reverses, care must be exercised in making speed measurements. The waveform of B. burgdorferi resembles the waveforms of many eukaryotic flagella. Invertebrate sperm flagella typ- ically have a flat meander-like waveform (21), as do many algal flagella (39) and some mammalian sperm tails (40, 41). B. burgdorferi and eukaryotic flagella both produce planar propagating waves but use very different mechanisms. B. burgdorferi generates traveling waves with rotary motors at the base of the PFs. Eukaryotic flagella propagate waves via microtubular sliding actively generated all along their length (e.g., see ref. 42). Indeed, planar propagating undulations of the sort seen in eukaryotic flagella require energy input all along the flagellar length (43). B. burgdorferi thus^ achieves a feat of which eukaryotic flagella are not capable: propagation of planar waves of constant amplitude along its entire length with energy input only in its terminal regions. Much remains to be learned about how B. burgdorferi is able to achieve this remarkable feat.
We thank R. C. Johnson for (^) strains, encouragement, and help with cultures, and A. Barbour for strains and monoclonal antibodies. We appreciate the advice of S. Block concerning video equipment. This
research was supported by Public Health Service Grant AI29743, and by a Grant-in-Aid from the University of Minnesota Graduate School.
- Canale-Parola, E. (1978) Annu. Rev. Microbiol. 32, 69-99.
- Greenberg, E. P. & Canale-Parola, E. (1977) J. Bacteriol. 131, 960-969.
- Berg, H. C. & Turner, L. (1979) Nature (London) 278,349-351.
- Goldstein, S. F. & Charon, N. W. (1988) Cell Motil. Cytoskel- eton 9, 101-110.
- Berg, H. C., Bromley, D. B. & Charon, N. W. (1978) Symp. Soc. Gen. Microbiol. 28, 285-294.
- Kimsey, R. B. & Spielman, A. (1990) J. Infect.^ Dis. 162, 1205-1208.
- Canale-Parola, E. (1984) in Bergey's Manual of Systematic Bacteriology, eds. Krieg, N. R. &^ Holt, J. G. (Williams & Wilkins, Baltimore),^ pp. 38-70.
- Sadziene, A., Thomas, D. D., Bundoc, V. G., Holt, S. C. & Barbour, A. G. (1991) J. Clin. Invest. 88, 82-92.
- Berg, H. C. & Anderson, R. A.^ (1973) Nature (London) 245, 380-382.
- Charon, N. W., Goldstein, S. F., Block, S. M., Curci, K., Ruby, J. D., Kreiling, J. A. & Limberger, R. J. (1992) J. Bacteriol. 174, 832-840.
- Charon, N. W., Greenberg, E. P., Koopman, M. B. H. & Limberger, R. J. (1992) Res. Microbiol. 143, 597-603.
- Charon, N. W., Goldstein, S. F., Curci, K. & Limberger, R. J. (1991) J.^ Bacteriol.^ 173,^ 4820-4826.
- Johnson, R. C., Schmid, F. W., Hyde, F. W., Steigerwalt, A. (^) G. & Brenner, D. J. (1984) (^) Int. J. Syst. Bacteriol. 34, 596-597.
- Hovind Hougen, K. (1984) Yale J. Biol. Med. 57, 543-548.
- Barbour, A. G. (1984) Yale J. Biol. Med. 32, 818-824.
- Barbour, A. G., Tessier, S. L. & Hayes, S. F. (1984) Infect. Immun. 45, 94-100.
- Charon, N. W., Daughtry, G. R., McCluskey, R. S. & Franz, G. N. (1984) J. Bacteriol. 160, 1067-1073.
- Goldstein, S. F. & Charon, N. W. (1990) Proc. Natl. Acad. Sci. USA 87, 4895-4899.
- Shimada, K., Kamiya, R. & Asakura, S. (1975) Nature (Lon- don) 254, 332-334.
- Gray, J. (1955) J. Exp. Biol. 32, 775-801.
- Brokaw, C. J. & Wright, L. (1963) Science 142, 1169-1170.
- Gibbons, B. H. & Gibbons, I. R. (1972) J. Cell Biol. 54, 75-97.
- Goldstein, S. F. (1977) J. Exp. Biol. 71, 157-170.
- Stephan, D. E. & Johnson, R. C. (1981) Infect. Immun. 32, 937-940.
- Schellack, C. (1908) Arb. Kaiserl. Gesundh. 27, 364-387.
- Neumann, F. (1929) Klin. Wochenschr. 8, 2081-2085.
- Pijper, A. (1949) Symp. Soc. Gen. Microbiol. 1, 144-179.
- Cox, C. D. (1972) J. Bacteriol. 109, 943-944.
- Jeantet, P. & Kermorgant, Y. (1925) C. R. Sceances Soc. Biol. (Paris) 92, 1036-1038.
- DeLamater, E. D., Wiggall, R. H.^ & Haanes, M.^ (1950) J. Exp. Med. 92, 239-246.
- (^) Sequeira, P. J. L. (1956) Lancet 271, 749 (and figure).
- (^) Jarosch, R. (1967) Oesterr. Bot. Z. 114, 255-306.
- (^) Fosnaugh, K. & Greenberg, E. P. (1988) J. Bacteriol. 170, 1678-1774.
- Taylor, G. (1952) Proc. R. Soc. London A 211, 225-239.
- Gray, J. & Hancock, G. J. (1955) J. Exp. Biol. 32, 802-814.
- Brokaw, C. J. (1965) J. Exp. Biol. 43, 155-169.
- Chwang, A. T., Winet, H. & Wu, T. Y. (1974) J. Mechano- chem. Cell Motil. (^) 3, 69-76.
- (^) Cox, R. G. (^) (1970) J. Fluid Mech. 44, 791-810.
- Goldstein, S. F. (1992) in Algal Cell Motility, ed. Melkonian, M. (Chapman & Hall, New York), pp. 99-153.
- (^) Ishijima, S. & (^) Mohri, H. (^) (1985) J. (^) Exp. Biol. (^) 114, 463-475.
- (^) Ishijima, S. & (^) Witman, G. B. (^) (1987) Cell Motil. (^) Cytoskeleton 8, 375-391.
- Summers, K. E. & Gibbons, I. R. (1971) Proc. Natl. Acad. Sci. USA 68, 3092-3096.
- Machin, K. E. (1958) J. Exp. Biol. 35, 796-806.
Microbiology: Goldstein^ et^ al.
