APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Jan. 1989. p. 261-2630099-2240/89/010261-03$0'2.00/0Copyright © 1989, American Society for Microbiology

Chemokinetic Motility Responses of the CyanobacteriumOscillatoria terebriformis

LAURIE L. RICHARDSONT* AND RICHARD W. CASTENHOLZ

Departmenit *JfBiology, Univ'ersitv of Or egonI, Euigenie, Or egoni 97403

Received 1 August 1988/Accepted 11 October 1988

Oscillatoria terebriformis, a gliding, filamentous, thermophilic cyanobacterium, exhibited an inhibition ofgliding motility upon exposure to fructose. The observed response was transient, and the duration ofnonmotility was directly proportional to the concentration of fructose. Upon resumption of motility, the rateof motility was also inversely proportional to the concentration of fructose. Sulfide caused a similar response.

The effect of sulfide was specific and not due to either anoxia or negative redox potential. Exposure to glucose,acetate, lactate, or mat interstitial water did not elicit any motility response.

Cyanobacteria exhibit complicated motility patterns inmicrobial mats (14), mudflats (13, 16), and other benthicenvironments. Daytime motility responses of filamentouscyanobacteria have been explained by photokinesis, photo-taxis, and step-up and step-down photophobic reactions (6,9). The light responses exhibited by cyanobacteria appar-

ently are adaptations aimed at maintaining optimal lightregimens for support of photosynthesis, as well as avoidanceof burial by sedimentation (13). In addition to light re-

sponses, there have been three reports of cyanobacterialchemotaxis, or chemotaxis-related, behavior. Fechner (7)demonstrated a negative tactic response to acids by some

members of the family Oscillatoriaceae. Malin and Walsby(12) reported a positive light-dependent chemotaxis to CO2,HCO3-, and 0, again in an Oscillatoria sp. We (14) haverecently reported a negative chemokinetic response to sul-fide by Osc illatoria terebriforiniiis.Our results on 0. terebriforinis were part of an in-depth

study (14, 15) concerning motility of this cyanobacterium inthe natural environment during darkness, when chemotaxiswould be expected to be important. Populations of 0.

terebriformis migrate during darkness to a position below theO2/sulfide interface within a microbial mat to a microenvi-ronment which is both anaerobic and highly reducing. In thelaboratory, this species was shown to be capable of main-

taining viability for periods up to 1 month by fermentingexogenous fructose to lactic acid, but only at a negativeenvironmental redox potential. When respiring aerobicallyin the dark, 0. terebrifjrinis rapidly depleted intracellularstored glycogen and often lysed in 24 to 48 h. Therefore, thedownward migration appears to be adaptive.

In this report, we present the results of laboratory exper-iments on chemokinetic responses of 0. terebrifoirnis toseveral organic compounds and interstitial mat water andadditional results on the specificity of sulfide in eliciting thenegative chemokinetic response.An axenic clone of 0. terebriiformiis (OH-80-Ot-D) was

isolated in 1980 from Hunter's Hot Springs, Lakeview,Oreg. (15). This culture was maintained at 45°C on Dmedium, a mineral medium with 0.52 mM nitrilotriaceticacid (a chelator) as its only organic constituent (5).

Instead of monitoring the behavior of individual tri-

Corresponding author.t Present address: NASA Ames Research Center. MS 242-4.

Moffett Field, CA 94035.

chomes, laboratory studies involved the behavior of popu-

lations of 0. terebriforinis. A dispersed suspension of 0.

terebriforinis was quickly mixed into a test solution. The rateof gliding motility was then measured by the relative rate ofmass clumping. Trichomes, when dispersed, will begin toform clumps within a few seconds. If the suspension is denseto the point at which all trichomes are interconnected, onlyone clump will form (Fig. 1). Castenholz (3, 4) has demon-strated that the clumping rate is directly related to the rate ofgliding motility and that gliding continues in the dark,although at a somewhat lower rate. The maximum rate ofgliding motility by 0. terebriforinis (strain OH-51-Ot) is 8.6p.m/s (10).To quantify the rate of clumping, we drew targets on the

bottoms of small glass petri dishes. A point of inert siliconecement was placed in the center of the target which allowedthe clump to retract evenly to the center of the dish. The rateof clumping was measured as the rate of edge withdrawal, or

distance traveled by the edge of the contracting aggregationover an arbitrarily prescribed distance.

Test solutions were made up in D medium buffered withEPPS (N-[2-hydroxyethyl]-piperazine-N'-3-propanesulfonicacid; 1.2 g/liter at pH 8.0). To investigate motility responses,we made up a range of concentrations of test substances inthe targeted dishes, and an inoculum of 0. terebrifoirmis wasquickly dispersed into each solution by pumping with a

syringe. Equal volumes of 0. terebrif)rtnis were used in theinocula. Figure 2 illustrates the effect of fructose concentra-tion on clumping. In this experiment, done at 45°C in thelight, concentrations of 0 to 7 mM fructose were used. Asfructose concentration increased, the time before the onsetof clumping increased. In additional experiments, fructoseconcentrations of up to 30 mM were used, which preventedthe onset of clumping for 25 min.The experiment of Fig. 2 is depicted graphically in Fig. 3,

which shows the rate of motility (measured as centimeterstraveled by the retracting edge of the clump) as a function offructose concentration. When plotted in this manner. twoeffects are evident, response A and response B. Response Ais an initial inhibition of clumping (visually observed) whichis proportional to the concentration of fructose. Withoutfructose, there was no delay in clumping. With 7 mMfructose. trichomes were nonmotile for approximately 9 min.Response B was a measurable change in the rate of clumpingwith varying fructose concentration. The rate of edge with-drawal was inversely related to the concentration of fruc-

261

Vol. 55, No. 1

on July 10, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

APPL. ENVIRON. MICROBIOL.

A

C

E FFIG. 1. Clumping of 0. terebriformis at 47°C in a glass petri dish.

Time after dispersal: A, t = 0; B, 20 s; C, 40 s; D, 60 s; E, 90 s; F,4 min. (Reprinted by permission of the author [3] and the Journal ofPhycology.)

tose. While the experiment shown was done in the light, asimilar response occurred in darkness, although rates ofwithdrawal were slower.When trichomes were redispersed in a fresh fructose

solution (7 mM) after clumping in the presence of fructose,the duration of the lag (response A) decreased. Multipleexposures (up to four, including initial exposure) resulted ina final clumping rate comparable to an initial exposure of 1mM (Fig. 3). When trichomes which had been exposed to 7mM fructose were redispersed in basal D medium, the onsetand rate of clumping were as rapid as if no exposure tofructose had occurred. No further experiments were done inthis line of investigation.Clumping experiments were also done with glucose, ace-

tate, and lactate (all at concentrations of 10, 15, and 30 mM),as well as interstitial water from the natural microbial matfrom which 0. terebriformis was isolated. None of thesecompounds had any effect on either the onset of clumping orthe rate of clump edge retraction. The carbohydrates wereselected because both glucose and fructose can serve as anextracellular carbon source to support fermentation in 0.terebrifor-mis (15) and because acetate and lactate are com-mon organic compounds produced in hot spring mats (1).

In a previous report (14), we demonstrated a negativemotility response by 0. terebriformis to sulfide. Artificial

FIG. 2. Effect of fructose concentration on clumping by 0.terebriformis. Bottom dishes in each panel (left to right): 0, 1, and 3mM fructose. Top dishes: 5 (left) and 7 (right) mM fructose. (A) t =1.5 min. (B) t = 6 min. (C) t = 16 min.

gradients of sulfide were set up in vials. An overlay ofdispersed 0. terebrifo)rmis in soft (0.7%) agar was inoculatedon top of a 0.7 mM sulfide plug (also soft agar). A discreteband of trichomes accumulated just below the boundary ofthe sulfide plug.An additional series of experiments was done to determine

whether the negative effect of sulfide on motility was, in fact,specific. It is possible that the observed accumulation oftrichomes in the sulfide zone was due to the negative redoxpotential or to anoxia per se, both caused by sulfide. To testsulfide specificity, we compared clumping response in thepresence of sulfide with clumping responses under condi-tions of anoxia (generated by a means other than the additionof sulfide) and the presence of an alternate strong reducingagent (sodium thioglycolate).

262 NOTES

on July 10, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from

NOTES 263

4

NO FRUCTOSE ~

D D o< <~

11!<Q 3mM M

0 4 8 12 16 20 24minutes

FIG. 3 Motility rate versus fructose concentration (O cm = edgeof test chamber).

Test solutions were prepared in Hungate tubes in Dmedium buffered with EPPS (pH 8.0). Tubes for aerobicexperiments were gassed (20 min) with sterile air, whileanaerobic tubes were prepared by gassing for 20 min with N,passed through copper filings heated to 300°C. After gassing,tubes were capped with serum caps and test substances were

added by syringe through the serum cap into prepared testsolutions. The tubes were then shaken and incubated in thelight or dark, and trichomes were observed over time formotility. Motility was assessed as the time taken for clumpsto form. Clumping rates were not quantified (the first visibleresponse in the tubes was the formation of numerous smallclumps).

Sulfide alone (0.7 mM) inhibited the formation of clumps,but only for 5 min. Sulfide inhibition occurred in the dark inthe presence or absence of DCMU [3-(3,4-dichlorophenyl)-1,1-dimethyl urea, an inhibitor of photosystem II; used at 5,uM]. Under illumination, however, sulfide elicited the neg-ative motility response only when DCMU was present; inthe absence of DCMU, there was no inhibitory effect.Possibly the lack of inhibition by sulfide in the light withoutDCMU was due to partial oxidation of the sulfide byphotosynthetically produced O.

Neither anoxia alone (flushed with N,) nor lowering theenvironmental redox potential with sodium thioglycolate(4.4 mM) delayed the onset of clumping. These resultsindicate that the sulfide effect is specific.

Similar experiments were also done with the addition of 25mM fructose. In all cases in which fructose was present, no

clumps formed for at least 15 min.Fructose and sulfide, when present together. acted syner-

gistically (data not shown). With 0.7 mM sulfide and 7 mMfructose, clumping was inhibited for 1 h both in the light withDCMU and in the dark. This effect was much greater thanthe sum of the lag-inducing effects of each compound.Fructose (25 mM) stopped motility for 20 min. With theaddition of 0.3 and 0.7 mM sulfide (no fructose), the effectwas comparable to fructose at 3 and 5 mM, i.e., the onset ofclumping was inhibited for 4 and 6 min, respectively.

Despite overall differences, there are two similaritiesbetween the negative chemokinesis of 0. terebriforin is andthe common type of bacterial chemotaxis such as that seen

in Escherichia( coli. First, in both motility response types,the duration of the response is directly proportional to theconcentration of substance which elicits the response (11;this study). Second, the ability to metabolize a substance is

unrelated to the success of that substance in eliciting amotility response. Thus, in bacterial chemotaxis, both at-tractants and repellents can be metabolizable or nonmetabo-lizable substances (9). 0. terebriformis responds to fructosebut not glucose, each of which supports fermentation.

It has been suggested by Glagolev (8) that a simple,efficient, and probably primordial form of chemotaxis wouldbe the halting of movement when favorable conditions areencountered. A substance which elicited stopping would beconsidered an attractant. For 0. terebrifjrinis, both fructoseand sulfide, each of which is beneficial to the physiology ofthe organism in darkness, elicit such a response. Thus,negative chemokinesis is more similar to the primordial formof chemotaxis postulated by Glagolev than is the response ofmodern chemotactic bacteria, such as E. coli, to an attract-ant, in which individual cells continue to move forward (2,11).

We are grateful to William R. Sistrom and Bo Barker J0rgensenfor helpful discussions. We thank two anonymous reviewers fortheir comments on the manuscript.

LITERATURE CITED

1. Anderson, K. L., T. A. Tayne, and D. M. Ward. 1987. Forma-tion and fate of fermentation products in hot spring cyanobac-terial mats. Appl. Environ. Microbiol. 53:2343-2352.

2. Boyd, A., and M. Simon. 1982. Bacterial chemotaxis. Annu.Rev. Physiol. 44:501-517.

3. Castenholz, R. W. 1968. The behavior of Oscillatoria terebri-forinis in hot springs. J. Phycol. 4:132-139.

4. Castenholz, R. W. 1973. Movements, p. 320-339. In N. G. Carrand B. A. Whitton (ed.), The biology of blue-green algae.University of California Press, Berkeley.

5. Castenholz, R. W. 1981. Isolation and cultivation of thermo-philic cyanobacteria, p. 236-246. In M. P. Starr et al. (ed.), Theprokaryotes. Springer-Verlag KG, B3erlin.

6. Castenholz, R. W. 1982. Motility and taxes, p. 413-440. In N. G.Carr and B. A. Whitton (ed.), The biology of cyanobacteria.University of California Press, Berkeley.

7. Fechner, R. 1915. Die chemotaxis der Oscillatorien und ihreBewegungserscheinungen uberhaupt. Z. Bot. 7:289-362.

8. Glagolev, A. N. 1984. Motility and taxis in prokaryotes. InSoviet Scientific Reviews Supplement, vol. 3. Harwood Acad-emy Publishers, Chur, Switzerland.

9. Hader, D. P. 1987. Photosensory behavior in procaryotes.Microbiol. Rev. 51:1-21.

10. Halfen, L. N., and R. W. Castenholz. 1971. Gliding motility inthe blue-green alga Oscillatoria princeps. J. Phycol. 7:133-145.

11. Macnab, R. M. 1979. Chemotaxis in bacteria, p. 310-334. In W.Haupt and M. E. Feinleib (ed.), Encyclopedia of plant physiol-ogy, vol. 7. Springer-Verlag, New York.

12. Malin, G., and A. E. Walsby. 1985. Chemotaxis of a cyanobac-terium on concentration gradients of carbon dioxide, bicarbon-ate and oxygen. J. Gen. Microbiol. 131:2643-2652.

13. Pentecost, A. 1984. Effects of sedimentation and light intensityon mat-forming Oscillatoriaceae with particular reference toMicrocoleius lyngbyacetts Gomont. J. Gen. Microbiol. 130:983-990.

14. Richardson, L. L., and R. W. Castenholz. 1987. Diel verticalmovements of the cyanobacterium Oscillatoria terebriformis ina sulfide-rich hot spring microbial mat. Appl. Environ. Micro-biol. 53:2142-2150.

15. Richardson, L. L., and R. W. Castenholz. 1987. Enhancedsurvival of the cyanobacterium Oscillattoria terebriforimis indarkness under anaerobic conditions. Appl. Environ. Micro-biol. 53:2151-2158.

16. Whale, G. F., and A. E. Walsby. 1984. Motility of the cyano-bacterium Microcoleuis chthonoplastes in mud. Br. Phycol. J.19:117-123.

VOL. 55. 1989

on July 10, 2018 by guesthttp://aem

.asm.org/

Dow

nloaded from