Bacterial SurfaceTranslocation: Survey and a Classification · BACTERIAL SURFACETRANSLOCATION growth in capsulated aggregates (i.e., spread-ing byejection). The macroscopic appearance

BACTERIOLOGICAL REVIEWS, Dec. 1972, p. 478-503Copyright ( 1972 American Society for Microbiology

Vol. 36, No. 4Printed in U.S.A.

Bacterial Surface Translocation: a Survey and a

ClassificationJ0RGEN HENRICHSEN

Department of Diagnostic Bacteriology, Statens Seruminstitut, Copenhagen, Denmark

INTRODUCTION ............... .........

MATERIALS AND METHODS...Bacterial Strains ..................... . .. .. ....

Inspection of the Surface ColoniesAgar Plate Microscopy ..............

Wet Mounts ...................

Determination of Velocity Values .Staining of FlagellaMedia ... ........... ............

Incubation .. ... ... ... .. .....

Nomenclature ..... ...............OBSERVATIONS AND DISCUSSION.Swarming ...

Defimition of swarmingSwimming.

Definition of swimmingGliding...

Defimition of gliding .Twitching ..........................

Definition of twitchingSliding ....... .. ...... ... ... ..

Definition of slidingDarting ...... ....... .......... ..

Definition of dartingClassification of the Bacterial Spreading Phenomena

CONCLUDING COMMENTS ...........................

LITERATURE CITED ...... .................

478.................... 479..................... 479

479........ .......... 479

479479479

.................... 479.............. ..... 479

479..................... 479

479485485485485491

..................... 491494

.................... 494495

....495497497499

.. .....501

INTRODUCTIONThe swarming of strains of Proteus, Bacillus,

and Clostridium has been known to bacteriolo-gists for as long as it has been common practiceto culture bacteria on solid substrates. Like-wise, gliding motility of myxobacteria, blue-green algae, and other microorganisms is along-recognized phenomenon. In 1961, Lautrop(44) described yet another kind of bacterialmotility confined to substrate surfaces, and helater proposed to name it twitching (45). Bothswarming, gliding, and twitching lead to theproduction of what is usually referred to asswarming zones, although for reasons whichwill be obvious later on they will be calledspreading zones in this paper. A spreading zoneis a film, broad or narrow, of one or, at most, afew layers of cells extending from the edge of acolony or an area of confluent growth and,therefore, frequently so thin that it is barelyvisible to the naked eye.During a study of twitching it became evi-

dent that spreading zones may be produced on

surfaces of solid media in other ways than thethree above- mentioned and so, in order toestablish a sound basis for the further study ofdif'ferent kinds of' surface translocation and toutilize the different kinds of' surf'ace transloca-tion in the characterization of bacteria, it wasdeemed necessary to try, as f'ar as possible, todetect, describe, define, and classify all possi-ble kinds of surface spreading produced duringbacterial growth on solid media. This mightalso help to clarify what value such charactershave in taxonomy.

Six different types of surface translocationhave been recognized so far: (i) swarming,dependent on excessive development of flagellaand partly on cell to cell interaction; (ii)swimming, dependent on flagella and fluid;(iii) gliding, dependent on intrinsic motiveforces and partly on cell to cell interaction; (iv)twitching, dependent on intrinsic motive forces(and fimbriae?); (v) sliding, dependent ongrowth and reduced friction (i.e., spreading byexpansion; and (vi) darting, dependent on

478

on July 16, 2020 by guesthttp://m

mbr.asm

.org/D

ownloaded from

http://mmbr.asm.org/

BACTERIAL SURFACE TRANSLOCATION

growth in capsulated aggregates (i.e., spread-ing by ejection).The macroscopic appearance of a spreading

zone will only sometimes be so characteristicthat it directly reveals how the zone wasproduced. This may, however, in all cases bedecisively established by examining thespreading zone under the microscope.

In the following text, a survey is presented ofthe different kinds of bacterial surface translo-cation based on the micromorphological fea-tures of the spreading zone pattern and themanner of single-cell movement. The author'sobservations are reported in each section aftera review of the pertinent literature, and eachsection ends with a discussion and the author'sdefinition of the kind of surface spreading inquestion.

MATERIALS AND METHODSBacterial Strains

Table 1 lists the species and the number ofstrains of each species examined so far duringthe author's study. Strains specifically men-tioned or depicted, or both, in this paper arelisted in column 2 in Table 1.

Inspection of the Surface ColoniesBecause spreading zones may be inconspicu-

ous, the use of a hand lens is essential.Agar Plate Microscopy

A Carl Zeiss photomicroscope with phasecontrast equipment and a 6-mm auxiliarycondenser lens was used. Most of the observa-tions were made with low-power, dry lenses(x16, x25, or x40), but occasionally an oilimmersion objective was also used. In suchcases the observations were also checked with adry lens, because placing a cover slip on top ofthe agar plate culture inevitably alters theconditions by introducing another surface(glass) and extra fluid. Photomicrographs weretaken with a photomicroscope by using anelectronic flash (Ukatron UN 60) and KodakTri-X films.

Wet MountsWet mounts were used to reveal flagellar

motility in appropriate fluid media.

Determination of Velocity ValuesSingle cells, aggregates of cells, or the lead-

ing edge of the spreading zone were followeddirectly under the microscope. Distance cov-ered was measured by means of an eyepiecegraticule and time was recorded with a stopwatch.

Staining of FlagellaThe technique of Leifson (47) was used.

MediaMeat extract agar plates (MEA) were made

of 0.5% meat extract, 1.0% peptone (Orthana),0.3% sodium chloride, 0.2% primary sodium,and 1.8% agar.

Ascites agar plates (AA) were made of twovolumes of MEA (with an agar concentration of2.2%) and one volume of ascitic fluid.The 5% blood-agar plates were made of MEA

mixed with 5% (v/v) defibrinated horse blood.The 10% blood-agar plates were made of

MEA mixed with 10% (v/v) defibrinated horseblood.Cytophaga agar plates were used in the

following modifications. No. 62 was made of0.05% tryptone (Difco) 0.05% yeast extract(Difco), and 1.0% agar (Difco), with a final pHadjusted to 7.0. No. 70 was made as no. 62, butwith tryptone concentration of 0.5%. "Ordal"was made of 0.2% tryptone (Difco), 0.05% beefextract (Difco), 0.05% yeast extract (Difco),0.02% sodium acetate, and 1.0% agar (Difco),with a final pH adjusted to 7.2 (55).

IncubationClostridium tetani was incubated in an an-

aerobic jar (Baird and Tatlock, Ltd.). With theexception of Proteus mirabilis and Bacillusalvei the rest of the strains were incubatedaerobically in plastic bags in order to ensure ahumid atmosphere. The temperatures usedwere either 22, 30, 33, or 35 C, as indicated inthe text or the legends.

NomenclatureIn this paper the term "surface transloca-

tion" is synonymous with the terms "surfacespreading" and "motility on a solid surface."The term "swarming" as usually employed inthe literature covers any surface spreadingphenomenon but is here always qualified as"swarming in the broad sense" when used inthis way. "Swarming in the narrow sense" orjust "swarming" is here only used aboutswarming as specifically defined later in thispaper.

OBSERVATIONS AND DISCUSSION

SwarmingMany excellent and detailed observations of

swarming (i.e., swarming in the narrow sense)of bacteria can be found in the literature, and acomplete review of the many papers will not begiven; only certain facts pertinent to the prob-lem of flagella-dependent surface motility incontrast to other spreading phenomena will bediscussed. Classical swarming is shown by P.

VOL. 36, 1972 479


mbr.asm

.org/D

ownloaded from


480 HENRICHSEN BACTERIOL. REV.

TABLE 1. List of strains examined

No. of strains Designation of selected Origin of selectedSpecies studied studied strainsa strains'

Acinetobacter calcoaceticus .......... 265 BD-4 JuniAcinetobacter calcoaceticus 17905 ATCCAcinetobacter calcoaceticus AB 156 SSIAlcaligenes odorans ........... ....... 6 H 1079 SSIBacillus alvei .. ....................... 1 HV 56 SSIBacillus anthracis .................... 10 Ax 11 ThalBacillus cereus ....................... 3 HV 57 SSIBacillus circulans .................... 1 HV 61 SSIChondrococcus columnaris ........... 1 1-R 43 OrdalClostridium tetani ................... 2 An 551/71 SSIClostridium novyi .................... 1 An 73/70 SSICorynebacterium sp .................. 1 HV 59 SSICytophaga succinicans ......... ...... 1 RL-8 OrdalCytophaga sp. ....................... 5 U 67 SSIEscherichia coli ...................... 11 W 3703 Hfr 0rskovFlavobacterium meningosepticum .... 12 13253 ATCCFlavobacterium sp. ........... ....... 30 U 120 SSIKurthia zopfii ....................... 3 8603 NCIBMoraxella bovis ...................... 3Moraxella kingii ..................... 3Moraxella nonliquefaciens ....... ..... 7Moraxella osloensis ........... ....... 34 A 249 SSIMoraxella phenylpyruvica ....... ..... 1 23633/69 FrederiksenMoraxella urethralisc .......... ....... 1Proteus mirabilis .................... 10 H 1093 SSIProteus vulgaris ..................... 10Pseudomonas acidovorans ....... ..... 2Pseudomonas aeruginosa ........ ..... 13 AB 1421 SSIPseudomonas alcaligenes ........ ..... 10 PR 389 SSIPseudomonas mallei .......... ....... 15 23344 ATCCPseudomonas maltophilia ....... ..... 2Pseudomonas pseudoalcaligenes ...... 69Pseudomonas stutzeri ......... ....... 2Pseudomonas sp . .................... 1Salmonella newport ........... ....... 1Salmonella typhimurium ........ ..... 1Simonsiella crassa ................... 1 UWO 367 MurrayStaphylococcus albus ......... ....... 3 HV 54 SSIStreptococcus pneumoniae ........... 9Streptococcus sp. (nonhemolytic) ..... 15 4932/71 FrederiksenVitreoscillasp.2 UWO 390 Murray

a Indicates the designation of the strain that is specifically mentioned or depicted in this paper.5ATCC, American Type Culture Collection, Rockville, Maryland, U.S.A.; Frederiksen, W. Frederiksen,

Department of Medical Bacteriology, County Hospital of Aalborg, Aalborg, Denmark; Juni, E. Juni, Depart-ment of Microbiology, University of Michigan, Ann Arbor, Michigan, U.S.A.; Murray, R. G. E. Murray, De-partment of Bacteriology and Immunology, University of Western Ontario, London, Canada; NCIB, NationalCollection of Industrial Bacteria, Torry Research Station, Aberdeen, Scotland; Ordal, E. J. Ordal, Depart-ment of Microbiology, School of Medicine, University of Washington, Seattle, Wash., U.S.A.; SSI, Depart-ment of Diagnostic Bacteriology, Statens Seruminstitut, Copenhagen, Denmark; Thal, E. Thal, Statens Vet-erinairmedicinska Anstalt, Stockholm 50, Sweden; Orskov, Ida and F. 0rskov, International EscherichiaCentre (WHO), Statens Seruminstitut, Copenhagen, Denmark.

c An unofficial name, see reference 45a.

mirabilis and P. vulgaris (33, 38, 53), by some flagella and are able to swim in fluid media.species of Bacillus (see references 54 and 69 for Clark (13) states that nonflagellated variants ofreviews of the literature), and by some species B. alvei have lost the ability to swarm. Surface-of Clostridium, notably C. tetani and C. novyi active substances in certain concentrations(22, 82). inhibit swarming and swimming in liquid

All swarming organisms possess peritrichous media of Proteus by impairing the production


mbr.asm

.org/D

ownloaded from



of flagella (18, 42), and, furthermore, inhibitionof the swarming of P. mirabilis byethylenediaminetetraacetic acid (EDTA) hasrecently been reported to be associated withinhibition of the synthesis of flagella and pre-vention of the formation of the long-cell forms(87). The one serotype of C. tetani that does notswarm is nonflagellated (74). Orskov (54a)describes how the activity of the flagella of aproteus swarmer cell moving on an agar surfacemay be convincingly demonstrated by means ofhis "direct India ink-agar microscopy," be-cause the active flagella fling the ink particlesaway from the cell.Most flagellated bacteria, however, will not

swarm (in the narrow sense) and even motile,i.e., swimming, but nonswarming strains ofProteus have been described (15). These strainscould be made to swarm by treatment withphage lysates of other phenotypically similarstrains (14), indicating that something morethan the presence of normally functioningflagella, i.e., flagella capable of producingswimming motility, is needed for an organismto swarm. All available evidence demonstratesthat the special requirement is an excessivenumber of flagella. Thus, Roberts (64) de-scribes a swarming B. rotans as having numer-ous flagella; Hoeniger (39) shows photographsof long cells of P. mirabilis having more than1,000 flagella per cell; Turner and Eales (82)describe the long, filamentous cell forms of theswarming groups in the clostridia as being'richly endowed with luxuriant flagella"; andalso, Boltjes (9) provides evidence that swarm-ing bacteria have a large number of flagella.The curious developmental cycle of swarmingstrains of Proteus (38, 53) is a characteristicand unique feature of these organisms, but thelong swarmer cells have their counterpart inthe filamentous, heavily flagellated cells thatconstitute the advancing edge of swarmingstrains of Clostridium (82).

It is another characteristic of swarming thatthe cells move in the form of large rafts (53) or"bullet-shaped" colonies (54), whereas isolatedcells almost never move. The size of the rafts ormotile microcolonies is clearly dependent onthe amount of available moisture (73). Thedrier the agar, the bigger the rafts or themicrocolonies. On insufficiently dried agar,individual cells may be seen to move, whereasrotating and wandering colonies of Bacillus arebest demonstrated on either very dry plates oron plates with a high concentration of agar.Lowering the agar concentration results indiffuse spreading (37). Willis and Williams (89)have shown that inhibiting the swarming of C.tetani by adding antitoxin to the plates results

in the appearance of motile daughter coloniesat the periphery of the nonswarming colonies.Small, wandering colonies arising from themain colony were seen with Proteus grown onnutrient agar containing homologous 0 serum(67, 68).Murray and Elder (54) conclude that

"swarming is an expression of motility underspecial conditions-in the thin fluid layer over-lying the gel." "The swarming of Proteus is aconsequence of the movement of long cells andthis movement is only continuous when one cellis in contact with another," state Morrison andScott (53). And these authors advance thehypothesis "that the flagella of these very long,heavily flagellate, cells are able to obtain athrust on neighbouring cells which gives themsufficient impetus to propel themselves overthe surface of a solid medium."As examples of different kinds of swarming,

observations on P. mirabilis, C. tetani, and B.alvei are described.

Figure 1 shows the appearance of an agarplate 24 hr after streaking with a strain of P.mirabilis (to the far left in the picture) andincubating at 30 C. The typical, terracedspreading reflecting the cyclic growth andswarming of such organisms is clearly seen.Under the same conditions some strains willswarm extensively, covering the entire platewithin 24 hr, whereas other strains only pro-duce narrow spreading zones. The reason forthis variation is unknown. The spreading rateof a swarming strain depends upon the humid-ity of the plate, upon the growth rate of theorganism, and thus upon the incubation tem-perature.The outermost part of the spreading zone

seen in the microscope during active swarmingis depicted in Fig. 2. The cells, roughly measu-ring from 10 to 30 im, are mainly arranged inlarge bundles, or rafts, with areas of cell-freeagar surface in between. Under ordinary labo-ratory conditions (i.e., the agar surface notbeing too moist) motility is confined to therafts and isolated cells lie still. The pattern, asseen in the microscope, is constantly changing,the rafts moving quickly along in continuouscurves at a rate of approximately 10 to 15gm/sec. The isolated cells in a raft may move inopposite directions, and this enhances theimpression of high speed. For the sake ofconvenience, the velocity values determined byme are compiled in Table 2 where also all thevalues are converted into gm/min for com-parative purposes.

Flagella staining of the swarmer cells revealsan enormous number of flagella per cell (insetto Fig. 2). In broth, the long swarmer cells, as

481VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


482 HENRICHSEN

well as the shorter nonswarming cells, swim atthe same speed of 10 to 15 gm/sec. When theplates are insufficiently dried, isolatedswarmer cells will be seen to move, althoughnot as fast as rafts of cells, and the rate ofspreading may be so accelerated that the wholeplate is covered in one wave. On the otherhand, wandering colonies may be seen underconditions that tend to inhibit or slow downswarming as, for instance, on rather dry agarplates incubated at 22 C. Presumably, this isbecause the motile power of groups of cells isbigger than that of single cells and probably, toa certain extent, depends on the number ofcells in the group.The very delicate film of swarming produced

by a strain of C. tetani is seen in Fig. 3, which isa photograph of a 10% blood-agar plate takenafter 24 hr of incubation at 35 C in an anaerobicjar (Baird and Tatlock, Ltd.). Because the

BACTERIOL. REV.

spreading does not occur in cycles, it may bevery difficult to observe the film and determineits expanse. Using reflected light and scrapingoff the bacteria with a platinum loop will helpobservation. As with all other kinds of spread-ing, the rate of swarming of Clostridium de-pends on the dryness of the agar and the growthrate of the organisms. On ordinary laboratorymedia the swarming of C. tetani is so regularand characteristic that it can be used for theisolation of such strains from heavily con-

taminated material. This was realized as earlyas 1925 by Fildes (22).

Figure 4 shows the appearance of the spread-ing film as seen in the microscope. The very

long cells that measure up to 50 Am are

arranged in narrow bundles. Due to the ex-

treme sensitivity to oxygen of these organisms Ihave only been able to observe the last trace ofthe subsiding motility of the cells, which takes

TABLE 2. Compilation of approximate velocity values for the different spreading phenomenaa

Rate of movement Rate of progression

Type of spreading Name of organismb and designation of of cell rafts or of edge of spread-strains in parentheses aggregates in ing zone in

Am/min tsm/minc

Swarmingd P. mirabilis (H 1093) 700 75C. tetani (An 551/71) - 50B. alvei (HV 56) 60 20

Swimmingd B. cereus (HV 57) - 50Gliding Vitreoscilla sp. (UWO 390) 15 7

C. columnaris (1-R 43) 5 3.5Cytophaga sp. (U 67) 70 15

Twitching A. calcoaceticus (ATCC 17905) 5 5P. alcaligenes (PR 389) 4 4

Sliding A. odorans (H 1079) - 5Flavobacterium sp. (U 120) - 10A. calcoaceticus (BD-4) - 10Streptococcus sp. (4932/71) - 3B. anthracis (Ax 11) - 2

Darting Staphylococcus albus (HV 54) - 6

a The values vary with the strains and the conditions, which are those described in the text.' The nonabbreviated generic names may be found in Table 1.c These values should be compared with the value of 0.33 Am/min representing the rate of increase in colony

radius of Escherichia coli on a defined medium at 37 C (59).dThe swarming and swimming strains swim in wet mounts at velocities between 600 and 1,000 Mm/min.

FIG. 1. Proteus mirabilis strain H 1093 culture plate incubated on meat extract agar at 30 C for 18 hr; 1/4natural size.

FIG. 2. Swarming. Peripheral part of swarming zone in Fig. 1, ca. x300. Inset: swarm cell doubled up andwith numerous flagella. Flagella stain; x600.

FIG. 3. Clostridium tetani strain An 551/71 culture plate incubated on 10%o blood agar at 35 C for 18 hr; 3X4natural size.

FIG. 4. Swarming. Same strain as in Fig. 3 but incubated on ascites agar at 35 C for 18 hr. Shows peripheralpart of swarming zone, ca. x200. Inset: long swarm cell with a fur of flagella. Flagella stain; x600.

FIG. 5. Bacillus alvei strain HV 56 culture plate incubated on meet extract agar at 22 C for 48 hr; 3/4natural size.

FIG. 6. Swarming. Peripheral part of swarming zone in Fig. 5, ca. x300. Inset: flagella stain of heavilyflagellated cell from swarming zone; x600.


mbr.asm

.org/D

ownloaded from


FIGs. 1-6.483


mbr.asm

.org/D

ownloaded from


HENRICHSEN

place in the cell bundles. A flagella stained cellis shown in the inset to Fig. 4. Because of thelarge number of flagella the individual flagel-lum does not present itself clearly, but ratherthe cell looks as if being wrapped in a fur offlagella.The very fascinating swarming of certain

species of Bacillus that also exhibit wanderingand rotating colonies is examplified in Fig. 5.This figure shows the swarming of a B. alveistrain on MEA after 48 hr of incubation atroom temperature. Single colonies are easilyseen in the swarm, and in the microscope manysuch colonies are seen to rotate. The tracks ofthe curved paths taken by the, usually smaller,wandering colonies are rather inconspicuous ona photograph of this size. This strain swarms onthe ordinary laboratory plate media, but therate of spreading and the appearance of theswarm depend very much upon the humidity ofthe plate. The rotating and wandering coloniesare only seen clearly with the naked eye whenthe surface of the agar is relatively dry (or theagar content of the medium very high). Usingonly lightly dried agar plates and incubatingsuch plates in a humid atmosphere the swarm-ing will be faster and appear as a thin film.

Fig. 6 shows the microscopic picture of thespreading zone where "bullet-shaped" mi-crocolonies, rafts, and single cells are lying in aloose pattern, leaving large areas of agar sur-face uncovered. By direct observation, motilityis seen to be confined to cells in contact withother cells as, e.g., in the rafts which movesteadily over the agar surface at a constantspeed of about 1 gm/sec. The wandering mi-crocolonies may be much bigger than the onesshown here. The drier the agar, the bigger themicrocolonies. They travel in continuouscurves and nearly always counterclockwise(seen in the microscope). If a drop of water isput on top of such a colony, it disintegratesimmediately, showing that the cells are notbound very tightly together. Eventually a mi-grating colony might catch up with its own tailand thereby produce a rotating colony thatusually does not translocate but only rotates onthe spot. A rotating colony is not shown for, atthis enlargement, it would have filled the entirefield. It should be noted that the cells are shortin comparison with the long swarmer cells ofProteus and Clostridium but that, relatively,they are just as heavily flagellated as these(inset to Fig. 6).With a slight modification of the "direct

India ink-agar microscopy" of 0rskov (54a)further information on the swarming processcan be obtained. By placing a drop of diluted

India ink on the cell film and a cover glass ontop of the drop, so that only one corner coversthe drop, variations are often created in thethickness of the layer of fluid in which the inkparticles are uniformly distributed, and areas.nay usually be found where the layer of fluid issufficiently thick to allow the cells to swimfreely. They do so at a speed of about 10g,m/secand appear in the microscope surrounded by alight halo, produced on the dark backgroundwhere the particles are being whirled away bythe flagella. If, then, a cell swims into an areawhere the layer of fluid is thinner, the move-ment changes its character: the cell ceases torotate and moves steadily along on the agarsurface with a somewhat slower speed, and theIndia ink particles come closer to the body ofthe cell. The changes observed clearly indicatethat the flagella function differently in relationto the surroundings during swarming andswimming.

It should be stressed that the species de-scribed are only used as examples of types ofswarming which also occur in related speciesas, P. vulgaris, C. novyi, and B. circulans. Thefeatures that these types of swarming have incommon are essentially two, viz., the veryheavy flagellation of the cells and the fact thatthe translocation predominantly takes placewhere cells are in contact with other cells, i.e.,in cell bundles, rafts, microcolonies etc. Thissuggests that the motile power is bigger in cellaggregates than in single cells and this also fitsin with the fact that these aggregates becomebigger when conditions for swarming becomepoorer (drying of the agar, for example). Infact, migrating microcolonies can be foundunder conditions that are suboptimal forswarming in all three types of organisms,although to a varying extent, of course. Itseems most likely that there is a gradualtransition from the surface motility of a bundleof two to three cells to a macroscopically visiblemigrating colony.Smith (74), in writing that "the phenomenon

of motile colonies (of C. novyC) is not connectedwith swarming," probably misinterpretsTurner and Eales (82) when they say that "thephenomenon (of motile daughter colonies) mayand often does occur under conditions unsuita-ble for swarming." Turner and Eales' state-ment is, of course, perfectly true but it does notallow for the conclusion that Smith draws.The difference between the swarming of

Proteus, Clostridium, and B. alvei is primarilythat the former two have long s warmer cells,while the latter has swarmer cells of "normal"size. One could speculate on whether this has

484 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



any relation to the much more pronouncedtendency of the Bacillus strains to form migrat-ing colonies under conditions suboptimal forswarming because the motile power of the longswarmer cells of Proteus and Clostridium mustbe bigger than that of the much smaller Bacil-lus cells.The delimitation of swarming from swim-

ming is based on the observation that swarm-ing motility is qualitatively different fromswimming motility and should be looked uponas a special kind of bacterial locomotion de-pendent on the presence of a surface and withthe propulsive force generated by the manyflagella functioning in an as yet unknownmanner.The delimitation of swarming from spread-

ing phenomena other than swimming presentsno problems for these phenomena are all inde-pendent of flagella and, furthermore, have theirown characteristic micromorphological cellpattern and manner of movement.

Definition of swarming. Swarming is akind of surface translocation produced throughthe action of flagella but is different fromswimming. The micromorphological pattern ishighly organized in whirls and bands. Themovement is continuous and regularly followsthe long axis of the cells which are predomi-nantly aggregated in bundles during the move-ment.

SwimmingIn his studies on chemotaxis in bacteria Adler

(1), among other methods, uses semisolid agarplates to examine the influence of differentchemicals on the flagellar motility of Esche-richia coli. With a concentration of 0.2% agarthe bacteria swim on top of the agar but mayalso, if oxidative metabolic pathways arenot preferred, swim throughout the depth ofthe agar or at the bottom. The bacteria are saidto swarm, the meaning obviously being swarm-ing in the broad sense (see Nomenclature, p.479). Gard (24) uses the word "swarming" inthe same way when he states that all bacteriawhich swim in fluid medium will swarm undersuitable conditions, i.e., if the concentration ofagar is sufficiently low in the plate medium.Such a semisolid agar plate is often called"swarm agar"; a more appropriate designationwould be "motility agar." Because the term"swarming" has thus been used about thesurface swimming taking place in "swarmagar," it is understandable that confusion hassometimes arisen (see, e.g., references 5 and48).The spreading zone of a strain of B. cereus

produced by swimming motility on the surfaceof the solid cytophaga agar "Ordal" is shown inFig. 7. The most peripheral part of the zone isquite delicate and it may-macroscopically-be discerned that it consists of closely set fin-ger-like projections. This spreading is verysensitive to dryness of the medium and for thisreason it will only be seen on freshly poured,slightly moist plates or in semisolid motilityagar media. Most bacteriologists will knowthis kind of annoying phenomenon when insuf-ficiently dried media are used and B. cereusoccurs as a contaminant. Strains which spreadlike this, however, will also spread on cyto-phaga agar plates if incubated in plastic bags.

Fig. 8 shows the microscopic appearance ofthe peripheral part of the finger-like projec-tions. This picture clearly is different fromthose of P. mirabilis, C. tetani, and B. alveiand is, indeed, produced in an entirely differ-ent manner: the individual cells swim freely inthe surface film of liquid, little by little pres-sing projections further out over the agar sur-face. The inset shows that the cells haveperitrichous flagella but not nearly as many asseen on B. alvei, for example. Evidently thenthe only feature which the spreading of B.cereus has in common with the swarming of P.mirabilis, C. tetani, and B. alvei is the produc-tion of a spreading zone. The differences arethat B. cereus (i) actually does not move on theagar surface but swims in a preformed film ofsurface fluid of sufficient thickness, (ii) has arelatively small number of flagella, and (iii)does not spread on all media, and this lastpoint is related to the first.Here it might be appropriate to mention that

the spreading so often seen in strains ofPseudomonas aeruginosa is not usually, as hasbeen generally assumed, due to surface swim-ming but is due to twitching. This is apparentin all cases of spreading P. aeruginosa ex-amined in this study.

Definition of swimming. Swimming is akind of surface translocation produced throughthe action of flagella, but is different fromswarming and only takes place when the film ofsurface fluid is sufficiently thick. The mi-cromorphological pattern is unorganized. Thecells move individually and at random in thesame manner as flagellated bacteria in wetmounts.

GlidingThe kind of surface translocation designated

gliding has been observed in several groups ofmicroorganisms including some bacteria. It is,however, first of all a characteristic associated

485VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


HENRICHSEN BACTERIOL. REV.

;. '41i_t'S

IRY~~~~~~~~~~~~~~'. '

FIGS. 7-12.

486


mbr.asm

.org/D

ownloaded from



with filamentous blue-green algae, but unicel-lular blue-green algae also may have thisproperty (80).Among the bacteria, three main groups ex-

hibit gliding motility. According to Stanier etal. (79) they are: (i) filamentous, gliding bacte-ria consisting of the following principal genera,

Beggiatoa, Leucothrix, Simonsiella, Sapros-pira, Thiothrix, and Vitreoscilla; (ii) fruitingmyxobacteria comprising genera such as Ar-changium, Chondromyces, Chondrococcus,Myxococcus, Podangium, Polyangium, Soran-gium, and Stigmatella; and (iii) the cytophagagroup with the genera Cytophaga andSporocytophaga.

Besides in these groups of procaryotic cells,gliding has been demonstrated in Mycoplasmapneumoniae on glass surfaces in liquid media(4, 10).Members of the first group of gliding bacteria

are now generally accepted as apochloroticblue-green algae. The literature on the glidingphenomenon in this group is very extensive andwill not be surveyed here. Readers are referredto earlier reviews (40, 41, 60, 61, 62, 75).The descriptions of gliding motility in the

two other groups of bacteria are based mainlyon observations of suspensions of bacteria influid media (ordinary wet mounts or hangingdrop preparations) or agar plate cultures, al-though in individual cases it is not always quiteclear on what method the description is based.Most observers agree that gliding motility is

confined to solid surfaces, e.g., glass and agargel but, according to Bisset (8) and Drews andNultsch (19), the surface need not be solid butmight also be the surface film of a fluid (Bisset)or the air-water interface (Drews and Nultsch).In general, observations of wet mounts registerthe behavior of single cells as opposed to theobservation of aggregates of cells in agar platecolonies.The movements of single cells in wet mounts

are described as lashing around a fixed pole,flexing and gliding (49), and "a rolling anddarting jerky movement" (16) occurring only

when cells are in contact with slide or cover

slip. The flexibility of the myxobacterial cellwas stressed by Stanier (78) but questioned byAnderson and Ordal (3). Referring to the bend-ing of proteus swarmer cells, Dworkin (20)proposes that flexibility "merely is a conse-

quence of a high length-to-thickness ratio ofthe cell."The demonstration of gliding motility on

agar surfaces is very dependent on the humid-ity and much more so than that of swarming.Leaving the petri dish open, so that the film ofsurface liquid evaporates, leads to the cessationof motility (3). Therefore, the colony mor-

phology of gliding organisms is very muchinfluenced by the amount of moisture present(3, 62), and the concentration of nutrients alsoplays a decisive role, peptone concentrationshigher than 0.25 to 0.5% being inhibitory to thegliding of cytophagas (34, 78). Under condi-tions optimal for gliding, the colonies will beseen as "completely flat, rapidly spreading,almost invisible swarms" (78) or as a spread-ing, rhizoid growth with a honeycomb appear-

ance (2). Movement takes place mainly in"spearheads" (i.e., spearhead-shaped cell ag-gregates at the edge of the colony) (78), single,isolated cells very rarely being motile (76), andthe picture is one of a "changing dispersedborder" with "interlacing bands being continu-ously rearranged" (45). On nutrient agar, withthe usual concentration of nutrients, the colo-nies are convex, smooth, and glistening with an

entire edge and no sign of spreading (78). Manyof the gliding bacteria form rotating microcolo-nies (21, 29, 62) exhibiting a movement similarto that of rotating microcolonies of certainspecies of Bacillus, except that the direction ofrotation is a matter of chance (30) as opposedto the predominantly counterclockwise direc-tion observed in the rotating species of Bacillus(54).The rate of translocation in gliding locomo-

tion varies with the organisms and conditions(see reference 17 for a full review), but it has tobe pointed out that, when conditions are not

+-FIG. 7. Bacillus cereus strain HV57 culture plate incubated on cytophaga agar "Ordal" at 22 C for 18 hr; 3,4natural size.

FIG. 8. Swimming. Peripheral part of spreading zone in Fig. 7; ca. x200. Inset: flagella stain of cell fromspreading zone; x600.

FiG. 9. Vitreoscilla sp. strain UWO 390 culture plate incubated on meat extract agar at 30 C for 48 hr; 3,4natural size.

FIG. 10. Gliding. Peripheral part of spreading zone in Fig. 9. The colony to the left in the picture was slowlyrotating when the photograph was taken; ca. x200.

FIG. 11. Chondrococcus columnaris strain 1-R 43 culture plate incubated on cytophaga agar no. 62 at 22 Cfor 96 hr; 3/4 natural size.

FIG. 12. Gliding. Peripheral part of spreadihg zone in Fig. 11; ca. x200.

487VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


HENRICHSEN

specified, comparisons of velocity values do notmake sense. When Stanier (77) is cited forhaving observed a maximum speed of 150,um/min for soil cytophagas (17), it should bestressed that the movement in question is notthat of the swarming edge outwards, nor is itthat of a spearhead or a single cell on an agarsurface, but it is the movement seen in wetmounts, under uncontrolled temperature, ofsingle cells over an unspecified distance and foran unknown time. Jarosch (41) gives the valueof 6 to 8 ,um/sec as the gliding rate of Beggiatoasp. on glass slides at room temperature. Thegliding motility of a 10-cell trichome ofVitreoscilla on agar, at a rate of 23 ,um/min, wasfollowed for 15 min by Costerton et al. (16). Bymeasuring the distance covered by a Simon-siella crassa filament on agar at room tempera-ture (after 6 hr of incubation at 37 C) from thepictures of Steed (81), a rate of translocation of0.42 ,m/min can be calculated.Many theories about possible mechanisms of

movement have been put forward during theyears (see references 17, 40, 41 and 86 fordetailed reviews), but none of the theories haveas yet been proven. The most promising ap-proach at present seems to me to be electronmicroscopy. Some results have already beenobtained and because these have not beensurveyed before, these results are surveyedbelow.

Electron microscopy of thin sections ofVitreoscilla (16) and of Myxococcus xanthus(84) revealed no microscopical structures likelyto explain the gliding of these organisms. In1967 Pate and Ordal (56), however, reportedthe finding of fibrils intimately associated withthe inner layer of the outer-unit membrane ofChondrococcus columnaris. The fibrils wereseen to run parallel to each other in widebands, with a center-to-center distance of ap-proximately 16 nm, and they were found bothin preparations of lysates and in sections ofcells (56). The possible role of such fibrils ingliding motility cannot be assessed until moreorganisms have been studied by using the sametechnique of fixation. Schmidt-Lorenz andKiihlwein (72) in 1968 announced the findingof bundles of either filaments 4 to 5 nm thickor tubular structures with diameters of 10 to 16nm running parallel to the longitudinal axis ofthe cell. These filaments or microtubuli liefreely in the cytoplasm, i.e., under the plasmamembrane and, therefore, cannot be the sameas the peripheral fibrils found in C. columnaris.They were found in 12 different strains ofmyxobacteria representing seven differentnamed species and five unnamed species. They

had not been found in cells of Chondromycesand Sorangium. These authors categoricallystate that these elements are contractile andrepresent the long-sought motility organelles ofmyxobacteria. They give no arguments in favorof this hypothesis, however. In a later paperthese structures, with no further evidence infavor of their function as locomotor organelles,are called intracytoplasmic flagella (71). Onthe other hand, comparative fine-structurestudies of sections of a strain ofM. xanthus anda nongliding mutant of this strain showed nodifferences, suggesting a structural basis for amechanism of gliding (11).

Cytochalasin B (12) appears to inhibit rev-ersibly the function of contractile microfila-ments of many eucaryotic cell types (88), butthis drug had no inhibitory effect on glidingmotility of three strains of Cytophaga. Yet, thismight have been due to lack of penetration ofthe antibiotic into the cells (35).

Additionally, it should be mentioned that aparallel array of fibrils, 5 to 8 nm wide, hasbeen found wrapped around the trichome of theblue-green alga, Oscillatoria princeps, and it issuggested that the fibrils actually propel thetrichome (31).Jarosch (41) defines gliding as "the active

movement of an organism in contact with asolid substratum where there is neither a visi-ble organ responsible for the movement nor adistinct change in the shape of the organism."Lautrop (45) writes: "Gliding or creeping mo-tility in bacteria is defined as the ability of theorganism to perform active translocation whenit is in contact with a solid surface. It is impliedthat the cells have no flagella or any otherrecognizable organs of locomotion." Both ofthese definitions are unnecessarily broad, aswill be shown below.To represent the three major groups of glid-

ing bacteria mentioned previously, I havechosen for study: (i) a Vitreoscilla species, (ii)C. columnaris, and (iii) an unidentified speciesof Cytophaga.Only gliding motility on agar surfaces will be

commented on, for the interpretation of themovements seen in wet mounts is especiallycomplicated due to the influence of a numberof factors that are difficult to control, such asthe oxygen supply, the thickness of the fluidbetween the two glass surfaces, the cleanness ofthese surfaces and, thereby, their physico-chemical characteristics.

Figure 9 shows the spreading growth of astrain of a Vitreoscilla sp. on MEA after 48 hrof incubation at 30 C in a plastic bag. Thespreading zone looks rather massive compared

488 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



with that of the other gliding organisms, andthis is due to the rich growth on this medium,which would prevent spreading of the two othergliding strains examined here. A photomicro-graph taken from the spreading zone is shownin Fig. 10. Trichomes, many of which measure100 Aim or more, are seen to be arranged in raftsand bundles of varying size. Motility tracksappear light on the darker background of theagar. To the left in the Fig. 10 part of a mi-crocolony measuring 200 Am in diameter isseen. This colony rotates slowly. Gliding mo-tility otherwise is mainly seen in the rafts ofcells which move at a speed of 15 ,um/min onthis medium. This should be compared withthe rate of approximately 7 Am/min at whichthe edge of the spreading zone moves peripher-ally. Gliding continues until the whole plate iscovered, in 4 to 5 days' time, from the center ofthe dish (which measures 9 cm in diameter).The gliding of Vitreoscilla is less sensitive tothe composition of the medium and the humid-ity than that of the two other organisms used.On other media, such as the cytophaga agar"Ordal," the microscopy pattern is slightlydifferent, appearing somewhat more organized,and also single trichomes can be seen to move,often for minutes in the same direction, but thespeed of the gliding is essentially the same ason MEA. Cells of gliding Vitreoscilla strainsare nonmotile in wet mounts unless touchingthe glass surface.The one strain of S. crassa examined by me

showed gliding motility of single trichomes (onAA at room temperature after incubation at 35C for 18 hr) at a speed of only 0.6 Aim/minwhich, however, compares well with the valueof 0.42 Am/min calculated from the figure ofSteed (81).The beautifully structured, feather-like

spreading produced by a strain of C.columnaris on cytophaga agar no. 62 by meansof gliding motility is shown in Fig. 11. Thisphotomicrograph was taken after 96 hr ofincubation at 22 C. Under such conditions ittakes a week or more to cover the entire agarsurface. Spreading does not take place onmedia with a higher content of nutrients. Buton the different modifications of cytophagaagar plates, gliding is a constant trait, althoughthe microscopical appearance of the spreadingzones varies considerably. Slime production isnot evident on these media. In contrast to othergliding organisms, the growth of this strainadheres to the agar and is difficult to scrape off.The very orderly pattern of cells on medium no.62 is shown in Fig. 12. The formation of rafts isnot predominant, but clearly the cells are

arranged in strands and small bundles of five toten organisms, leaving parts of the agar surfaceuncovered. Movement is a rather slow butsteady gliding in the direction of the long axisof the cells at rates of approximately 5 ,um/minunder optimal conditions, including incubationin plastic bags to insure a high humidity.

Fig. 13 shows the appearance of a Cytophagasp. gliding on cytophaga agar no. 62. Thespreading zone is quite broad, and this strainwill, in fact, be able to cover the whole agarsurface of the dish in between 2 and 3 days.Figure 14 shows the outermost part of theadvancing edge. The cells are arranged in aloose pattern of interlacing bands and rafts ofcells, leaving areas of the agar surface uncov-ered but for single scattered cells. Groups ofcells resembling spearheads are seen projectingoutwards. The locomotion, which is principallyseen in the groups of cells, i.e., rafts andspearheads, and takes the direction of thelongitudinal axis of the bacteria, gives rise to aconstantly changing picture, steadily glidinggroups of cells uniting or dividing and othergroups slowly arising from individual cells.Single cells may also be seen to move, thedirection again being that of their long axis,but intermittently and more slowly thangroups of cells and often reversing or changingdirection. Single-cell gliding is seen only if thehumidity is kept high. Spearheads of cells fromthis strain will, under optimal conditions, movealong at speeds of 70 gm/min. In comparison,the edge of the spreading zone moves peripher-ally at a speed of 15 jm/min.On cytophaga agar no. 62 other Cytophaga

strains glide, at room temperature, with othervelocities and produce spreading zones whichmight vary slightly in micromorphology butwhich have in common all the principal fea-tures of gliding locomotion just described. Ve-locity values of 6 and 15 jim/min for thetranslocation of spearheads of another strain ofCytophaga sp. and a strain of Cytophagasuccinicans, respectively, compared with thatof 70 ,um/min for the depicted strain, U 67,could possibly be a consequence of a much lesspronounced tendency of forming rafts or spear-heads in the two slower strains.The general microscopical appearance of

gliding may, in some cases, be very similar tothat of swarming if the difference in speed isdisregarded.The primary and obvious difference between

the two processes, swarming and gliding, ofcourse, is that swarming is performed by bac-teria having flagella and gliding is performedby nonflagellated bacteria.

489VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


FIGS. 13-18.490


mbr.asm

.org/D

ownloaded from



But there are other differences. Gliding ismuch more dependent on the amount of mois-ture available on the surface of the medium.The trichome-forming bacteria like Vitreoscillawill glide on MEA, but the translocation ismore pronounced on media which are poorer innutrients and incubated in a humid atmos-phere. And the cytophagas will not glide at allon MEA or similar media in which the peptoneconcentration is too high, but instead they formdistinct colonies with a compact border. Withall three cytophaga agar media I found thatoptimal conditions for gliding also include theuse of freshly poured plates and incubation in a

humid atmosphere. Neither the concentrationof nutrients nor that of agar should be too high.An agar concentration just sufficiently high toprevent the cells from moving down into theagar is optimal.The delimitation of gliding from sliding and

darting is simple, as will be shown later.Gliding, in contrast to twitching motility,

usually results in a spreading growth thateventually covers the entire agar surface. Butthis is not always so. McMeekin et al. (50)mention at least one strain, allocated to thecytophagas, as exhibiting gliding motility (asjudged by direct microscopy of microcolonies)but not spreading growth. Under carefullycontrolled conditions similar strains, in myexperience, will produce at least perceptiblespreading zones. But, in any case, it seems

quite evident that at present the size of thespreading zone, or rather its continuous propa-gation, cannot be used to define gliding motili-ty. The differences between gliding and twitch-ing pertain mainly to the characteristic mannerof movement of the cells, necessitating a des-cription of this as part of the respective defini-tions. Returning now to the definitions ofgliding given by Jarosch and Lautrop, it can bestated immediately that they cover equallywell gliding and twitching. In both cases it is a

matter of translocation of nonflagellated bac-teria in contact-and only then-with a solidsurface. Both definitions use the term "active"in connection with the words "movement" and"translocation," respectively. This might or

might not be appropriate; nobody can knowwith certainty as long as explanations of the

two phenomena have not been found. There-fore, I shall omit this term. Also, I prefer toomit any statement about organs of locomotionother than flagella at this state of knowledge.Accordingly, the essential parts of these twodefinitions, with an additional description ofcell movement, will be adequate.

Definition of gliding. Gliding is a kind ofsurface translocation produced by an unknownmechanism occurring only in nonflagellatedbacteria. The micromorphological pattern ishighly organized in whirls and bands. Themovement is continuous and regularly followsthe long axis of the cells which are predomi-nantly aggregated in bundles during the move-ment.

Twitching

This kind of bacterial, surface-bound motil-ity, independent of flagella, was first demon-strated in strains of A. calcoaceticus in 1961by Lautrop (44) who also, in 1965, named it"twitching" (45). His original findings weresoon confirmed by others (26, 29, 32, 57, 70).The same kind of motility is reported to havebeen found in strains of M. lacunata and M.nonliquefaciens (57), in M. bovis, M. kingii,and M. nonliquefaciens (36), in nonflagellatedstrains of P. aeruginosa (45), in an unnamedMoraxella sp. (6), and in some gram-negativeyellow-pigmented rods (50).Most authors study twitching motility by

means of microscopy of cultures on thin agarplates poor in nutrients (i.e., some cytophagaagar modification or other) (29, 32, 44, 50, 70).Remarks about spreading zones are scarce (45,50), but Gianelli and Cabassi (26) present colorphotographs of strains of Acinetobacter andMoraxella with spreading zones varying inwidth, appearance, and distribution along thecentral streak of growth. They describe thezones as Proteus- and Bacillus-like or barelyperceptible and looking like acanthus leaves or

petals. Unfortunately, these authors do notseem to have examined their plate culturesunder the microscope and thus are not in a

position to know how the spreading actuallytook place. Others have used a specially de-signed "oil chamber" where the bacteria, in a

drop of broth, are placed between an oil and an

4-FIG. 13. Cytophaga sp. strain U 67 culture plate incubated on cytophaga agar no. 62 at 22 C for 48 hr; 3/4natural size.

FIG. 14. Gliding. Peripheral part of spreading zone in Fig. 13; ca. x300.FIG. 15. Acinetobacter calcoaceticus strain ATCC 17905 culture plate incubated on cytophaga agar no. 62

at 30 C for 24 hr; 3/4 natural size.FIG. 16. Twitching. Peripheral part of spreading zone in Fig. 15; ca. x300.FIG. 17. Pseudomonas alcaligenes strain PR 389 culture plate incubated on cytophaga agar no. 62 at 30 C

for 48 hr; 3/4 natural size.FIG. 18. Twitching. Peripheral part of spreading zone in Fig. 17; ca. x200.

VOL. 36, 1972 491


mbr.asm

.org/D

ownloaded from


HENRICHSEN

agar surface (6, 57). The descriptions of singlecell movements seen in an oil chamber or on anagar plate surface are consistent with eachother. The motility appears as small, intermit-tent jerks covering only short distances andoften changing the direction of movement,which is not regularly related to the long axis ofthe cell; the maximum speed of a group of cellsis about 1 to 2 ,um/min (45). The only othervelocity value found in the literature is one of 2to 5 ,m/min (29).Only very little information is available

about optimal experimental conditions. Pie-chaud (57) considers an agar concentration of0.5 to 0.7% superior to higher concentrations,and Gianelli and Cabassi (26) recommend thesame concentration with Difco agar. The latterfurthermore found that higher concentrationsof nutrients (1% peptone and 0.5% meat ex-tract) with some strains gave better resultsthan the medium used by Halvorsen (32),which they originally tried. Varying the pHfrom the original 7.2, altering the surfacetension by addition of Tween 80, and usingother energy sources did not lead to betterspreading of their strains (26).Some misunderstandings need to be cor-

rected. Samuels et al. (66) write: ". . we be-lieve that the alleged motility of Herellea andMima (now: Acinetobacter calcoaceticus) hasnot been proved. We believe that it is diagnos-tically correct to refer to Herellea and Mima asnonmotile organisms, as determined by theusual methods of motility examination in semi-solid agar or in hanging drop suspensions."Obviously the misunderstanding arises fromthe use of the word "motility." The authors donot realize that the motility they refer to in thefirst of the sentences quoted is twitching mo-tility which cannot be observed by the samemethods as swimming motility of flagellatedorganisms, to which they refer in the lastsentence quoted. Therefore the first sentence isfalse and the last sentence is true.The mechanism of twitching motility is un-

known. There has been a tendency to relate itto another, also unexplained, phenomenon,namely gliding motility (17, 32, 44, 70). Lau-trop (45), however, in 1965 presented argu-ments in favor of treating twitching and glidingas separate taxonomic characteristics. Ryterand Piechaud (65) concluded from an electronmicroscopy study, by using both negativestaining and thin sections, that the cell wall ofMoraxella has a structure like that of Vitreos-cilla, and they proposed that this cell wallstructure might be related to the characteristicmotility of these organisms. But Lautrop et al.

(46), on the other hand, found "a strikingoverall similarity" in the ultrastructure of 27species of gram-negative bacteria including,among-others, A. calcoaceticus, M. nonlique-faciens, P. aeruginosa, E. coli and P. mirabilis.Piechaud (58) in 1969 suggested that fila-

ments, called "proflagella," were the cause ofthe twitching motility of Moraxella. Thesefilaments apparently are the same as thosedescribed by Ryter and Piechaud earlier (65) as"filaments probablement d'origine cap-sulaire." No explanation for the mode of actionof these structures is proposed but for theallusion to flagellar activity (58). Henrichsen etal. (36) reported that twitching motility andfimbriation are correlated in colony variants ofM. nonliquefaciens, M. bovis, and M. kingii,and they conclude that fimbriation most likelyis a necessary but not a sufficient condition fortwitching motility in these organisms. Twitch-ing motility, like gliding, is not influenced bycytoclialasin B (35).The occurrence of twitching motility seems

to be limited to relatively few groups of gram-negative bacteria. Apart from A. calcoaceticus,P. aeruginosa, and the species of Moraxellaalready referred to above, I have at present onlyfound this kind of surface translocation instrains of M. osloensis and in different speciesof nonfluorescent Pseudomonas (J. Henrich-sen, unpublished observations).The spreading produced by a twitching

strain of A. calcoaceticus is shown on Fig. 15.Spreading may occur along the entire streak or,as is often the case, as a varying number offanshaped zones that may be unevenly dis-tributed. This is not a reflection of a directlyand genetically determined inhomogeneity inthe bacterial population, since a strain like theone shown here cannot be split up into aspreading and a nonspreading variant. Simi-larly, single colonies may be diffusely spreadingor produce fanshaped offshoots. The rate atwhich the leading edge moves peripherally isusually 2 to 5 gm/min, but it varies with thestrain and the conditions. In some isolates thespreading zone is rather inconspicuous.

Figure 16 shows the typical appearance of theoutermost part of the spreading zone as seen ina microscope. In contrast to what is seen withgliding and swarming organisms, the generalimpression is one of a much less organizeddistribution of the cells over the agar surface;interlacing strands of cells do not occur, andraftlike cell aggregates are not formed except atthe verge of the spreading zone. These cellaggregates, however, only move peripherallywith the speed of the advancing edge, and they

492 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



are formed at the verge where the movement ofthe individual cell is arrested, partly because ofthe collision among cells and partly becausethe surface film of fluid outside the spreadingzone presumably becomes so thin that condi-tions for twitching motility deteriorate. Withinthe area of randomly scattered cells, withuncovered agar surface in between, single cellsand occasionally small clusters of cells can beseen to move in jerks sideways or in thedirection of their longitudinal axis, covering inone jerk distances of up to 2 to 3 Am, butusually much shorter distances. The directionof movement of the individual cells, as well asthe number of jerks per time unit, is quitehaphazard. Sometimes a small number of cellslying together may be seen to make a jump, inany direction, in a body. With strains thatshow a high twitching activity it can be seenwith a magnification of x 1,000 or more that thecells, also in between the jerks, are continu-ously moving extremely slowly in shifting di-rections. From this it will be understood that itdoes not make sense to try to measure thespeed of single cell movements. Just as thespreading zones may be quite narrow, thedirectly observable twitching activity may beinconspicuous depending on the strain in ques-tion and the conditions used for studying thephenomenon.Some strains of A. calcoaceticus produce

relatively large amounts of extracellular slime,and the cells may then be seen to lie apparentlywithin a "common capsule" and exhibit whatmight be called abortive twitching motility notleading to any spreading. In some such casesquite active twitching may be seen if a coverglass is applied.At least one strain (AB 156) in my collection

of twitching A. calcoaceticus isolates producesmotile microcolonies when twitching motility ispartly inhibited by the addition of 10-2 M CaCl2to the medium. Such microcolonies move at aspeed of 2 Mm/min leaving tracks of scatteredsingle cells behind. The path is curved, but nodirection, clockwise or counterclockwise, seemsto predominate.

Spreading produced by a twitching strain ofP. alcaligenes is shown in Fig. 17, and Fig. 18gives the micromorphology of the spreadingzone (note that the enlargements of Fig. 16 and18 are not the same; they are x300 and 200,respectively). In comparison with Fig. 15 and16 the overall impression is one of great similar-ity. Likewise, the character of the cell move-ment is covered by the description alreadygiven for A. calcoaceticus, and the rate ofoutward locomotion of the verge of the spread-

ing zone is of the same order of magnitude,namely 4 ,um/min. In short, this strain alsoexhibits typical twitching motility. The cellshave, however, one polarly inserted flagellum,and normal flagellar activity, i.e., swimming,immediately becomes evident if a drop of fluidmedium is placed on the agar surface. Never-theless, it can be stated with a very high degreeof confidence that this spreading is the result oftwitching and not of swimming because: (i)nonflagellated variants of strains of differentspecies of Pseudomonas produce the samedegree of spreading and exhibit the samecharacteristic single cell movements as flagel-lated strains, if they spread at all; and (ii)occasionally swimming is seen at the same timeas twitching motility, but then it is alwayslocated to that part of the spreading zoneclosest to the densely packed single cell layer(bottom part of photograph in Fig. 18) wheresurface fluid presumably is most plentiful,whereas twitching alone can be seen peripher-ally to this. On direct observation the two kindsof cell movement are easy to distinguish.

P. mallei is the only species of Pseudomonasthat is constitutionally nonflagellated. None of15 examined strains of P. mallei showedtwitching motility (J. Henrichsen, unpublishedobservations).A systematic study of the composition of the

plate media showed that twitching, like glidingmotility, is very dependent on the availabilityof moisture, for the greatest activity was ob-served on media poor in nutrients, with arelatively low agar concentration, freshlypoured, dried for a very short period, andincubated in a humid atmosphere (J. Hen-richsen, unpublished observations). For thesame reason it is advantageous to use thick,rather than thin, agar plates, and this does notprevent plate microscopy if only an auxiliarycondenser lens is used. This means that formost purposes a modified cytophaga agar, likemedium no. 62, will be satisfactory. Moststrains of A. calcoaceticus exhibiting max-imum twitching activity on this medium do notspread at all on MEA.Some twitching strains, e.g., most species of

Moraxella, will not grow on medium no. 62,and other media, e.g., AA, must then be tried.

I have not succeeded in isolating spreadingand nonspreading variants of twitching A.calcoaceticus strains, but twitching and non-twitching wild-type strains of A. calcoaceticushave been compared in electron microscopystudies. The comparison shows that twitchingstrains have fimbriae and that nontwitchingstrains do not (J. Henrichsen and J. Blom,

VOL. 36, 1972 493


mbr.asm

.org/D

ownloaded from


HENRICHSEN

manuscript in preparation). It therefore seemsto be a necessary condition for twitching motil-ity that the bacteria have fimbriae and pre-sumably a special type of fimbriae. At least notwitching motility could be detected in 11strains of E. coli representing bacteria with atleast two different types of fimbriae (threestrains were Hfr, one strain was F+, two strainswere hemagglutinating, and the remaining fivewere nonmotile strains) (J. Henrichsen, unpub-lished observations).The similarities between twitching and glid-

ing lie primarily in the facts that both types ofsurface translocation are independent of fla-gella, are, by and large, dependent on the samecultural conditions, and also are similar in oneimportant feature of the micromorphology ofthe spreading zone: the cells are scattered overthe agar surface. But the patterns are different.That of gliding organisms is highly organized,whereas that of twitching organisms is onlyslightly so.

Other differences between twitching andgliding motility are found in the speed andtotal, or final, progression of the spreading andin the characteristic manner of single cellmovement. The spreading of a gliding strainusually continues until the entire agar surfaceis covered by bacteria; the spreading due totwitching motility stops before that. This dif-ference presumably has the following explana-tion: the speed of progression of the spreadingzone in gliding organisms primarily depends onthe speed of movement of the cell rafts,whereas in twitching organisms the speed ofprogression of the spreading zone, primarilydependent on growth activity, secondarilydetermines the speed of the movement of themarginal cell aggregates.The difference between spreading by twitch-

ing and by sliding is evident from the micro-scopic appearance of the spreading zones in-cluding the behavior of single cells, but therates of spreading and the resulting colony, orstreak, morphology are closely similar. And itis in fact possible that pronounced sliding mayin some cases prevent the demonstration of anexisting potential for twitching motility.

Definition of twitching. Twitching is akind of surface translocation produced by anunknown mechanism, occurring in both flagel-lated and nonflagellated bacteria, but neverdue to the action of flagella. The micromorpho-logical pattern is varying but not as organizedas in swarming and gliding. The cells movepredominantly singly, although smaller mov-ing aggregates occur. The movement appears

as intermittent and jerky and does not regu-larly follow the long axis of the cell.

SlidingBisset (7) in 1939 described strains of strep-

tococci giving rise to colonies almost indis-tinguishable microscopically from typicalmedusa-head colonies of B. anthracis. Dis-cussing his findings, Bisset states that "thequalities of a colony are merely the aggregate ofthe qualities of its constituent organisms," andlater on, "the growth and consequent extensionof the bacterial threads and chains across thesurface of the medium is impeded by fric-tion.. ." (7). In his textbook Bisset (8) states:"The long-chained type of streptococcus (also)produces a variety of medusa-head colony,"and he comments on the interaction between atendency to chain formation and the resistanceof the surface of the medium. He uses the term"frictionless surface" and describes what wouldhappen on such a surface, were it existing.The reason for quoting Bisset at length is

that his descriptions suggest a phenomenonthat comes close to the description I shall giveof sliding on the basis of my own observations.The definition of sliding will also cover

colony formation of B. anthracis as describedas early as in 1910 by Graham-Smith (28) andthe formation of colonies of Kurthia zopfii witha medusa-head appearance (25).Spreading zones produced by sliding by dif-

ferent species on different agar plate media(Fig. 19, 21, 23, 25) are macroscopically indis-tinguishable from spreading zones of twitchingstrains (Fig. 15 and 17). The correspondingphotomicrographs (Fig. 20, 22, 24, and 26 ascompared with Fig. 16 and 18) immediatelyshow one important difference: the spreadingzone of a sliding organism consists of a singlelayer of densely packed cells. In the microscopethis sheet of cells may be seen to slide peripher-ally without perceptible changes in the mutualdisposition of the cells. In other words, move-ment of single cells, or groups of cells, relativeto other cells cannot be observed, and this isthe other important difference. Velocity valuesfor the spreading of sliding strains vary from 2to 5 Am/min up to as much as 25 ,um/min withgiven the strains and the conditions. Thisshould be compared with the colony growthrate values of nonspreading bacteria, e.g., therate of increase in colony radius (Kr) is 0.33,m/min for E. coli on a defined medium at 37 C(59).

In watching sliding in the microscope, acertain amount of flickering is often noticed in

494 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



the field of vision. This effect is produced whenindividual cells are squeezed, so to say, out ofthe cell sheet on top of which they thereaftercome to rest. The process lasts a fraction of asecond and is therefore only perceived as aflickering. Such cells are seen as darker spots inthe photomicrographs. Their number evidentlydecreases peripherally where the generatingforces also must be considerably smaller.

Alcaligenes odorans is described as produc-ing spreading surface colonies (27, 51), but themechanism of this spreading has not beenstudied in detail before. One reason for thisclearly might be that flagella-dependent sur-face translocation, either swarming or swim-ming, probably would be an obvious and satis-factory guess to most bacteriologists for thisspecies is peritrichously flagellated. Figures 19and 20 show a strain of A. odorans which isrepresentative of the six strains I have studiedof this species. The observations and the pic-tures are consistent with the description ofsliding given above.

In search of gliding motility among strains ofFlavobacterium sp., spreading was found in 6out of 30 strains examined, but in all six casesactually was found to be due to sliding (Fig. 21and 22) (J. Henrichsen, unpublishedobservations). None of 12 strains of F.meningosepticum showed any kind of surfacetranslocation (J. Henrichsen, unpublishedobservations). In their study of gram-negative,yellow-pigmented rods, McMeekin et al. (50) ina group of 28 unidentified strains, fitting thegeneral description of Flavobacterium, foundfive strains showing spreading growth but nei-ther twitching nor gliding motility, and theysuggest that "this type of swarming could bepassive."Although A. calcoaceticus strains, like the

one shown in Fig. 23 and 24, have been followedfor days in the microscope right from the timeof plating, no signs of twitching motility what-soever were seen and the picture, indeed, wasclearly that of sliding.

Until now 12 strains of nonhemolytic strep-tococci that produce extensi-e spreading zoneshave been isolated. They cannot be classifiedserologically, but biochemically they are allvery much alike. One strain is seen in Fig. 25and 26. Although in appearance quite differentfrom Bisset's medusa-head colonies and inde-pendent of chain formation, the sliding of thesestreptococci may be looked upon as an extremedevelopment of the processes described byBisset (8). Strangely enough these strainsspread nicely on blood agar plates on primary

isolation, but in subculture spreading were onlyseen on AA and never on blood agar.

Extensive spreading due to sliding has alsobeen encountered in a strain of Moraxellaphenylpyruvica and in a Corynebacterium sp.

Fig. 27 shows the quite broad spreading zoneof a strain of B. anthracis on medium no. 62after 3 days of incubation at room temperature,and Fig. 28 gives the microscopical picture ofthe most peripheral part of growth. The pro-duction of the long, wavy strands of parallelchains of bacteria plays an important role inthe formation of the spreading zone, as was soaccurately described by Graham-Smith in 1910(28), but it fits nicely into the definition ofsliding to be given below. Spreading of B.mycoides and of K. zopfii is essentially similar.

Sliding is not a phenomenon restricted tocertain media. It has been met with on all theusual bacteriological media including bloodagar plates, but all sliding strains will not slideon any medium. Sliding, as the other kindsof surface translocation, is more or less in-fluenced by the available amount of moisture.

Sliding may be regarded as a special case ofthe normal process of colony formation of aparticular strain on the surface of a particularsolid medium. In fact one should expect allbacterial colonies to consist of a circular sheetof cells in one single layer were it not for theforces of friction between the. surface of the celland the surface of the substrate. From this itcan be concluded that sliding bacteria musthave other surface properties than bacteriathat do not slide.

Definition of sliding. Sliding is a kind ofsurface translocation produced by the expan-sive forces in a growing culture in combinationwith special surface properties of the cellsresulting in reduced friction between cell andsubstrate. The micromorphological pattern isthat of a uniform sheet of closely packed cellsin a single layer. The sheet moves slowly as aunit.

DartingThe isolation of spreading staphylococci was

reported in 1953 by a Danish veterinary sur-geon working on bacteriological control of dairyproducts and using solid media rather poor innutrients (43). The spreading colonies weredescribed as looking like a "bacillus colony"and designated "curly variants." On 1,976"ordinary bacterial count plates," 166 spread-ing colonies were observed, 109 of which wereStaphylococcus aureus and 57 Staphylococcus

495VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


ki.1

4 .'4 B _

FIGS. 19-24.

496


mbr.asm

.org/D

ownloaded from



albus. After a varying, although rather small,number of subcultures, spreading did not occur

any longer. An explanation is not given of thedescribed phenomenon, but it is thought to berelated to a particularly vigorous growth.A spreading S. albus, isolated as a contami-

nant on cytophaga agar no. 70, is shown in Fig.29. On MEA or blood agar plates spreading isinconspicuous, but the colonies are usuallyangular rather than circular. Conditions favor-able for this type of spreading also include a

high atmospheric humidity and, consequently,incubation in plastic bags was used.

Fig. 30 shows that the microscopical patternof the spreading zone is one of clusters of cocciscattered over the agar surface with cell-freeareas in between. Under the microscope motionof some kind is evident, but at first is difficultto analyze. It looks as if "a constant flickeringeffect (was produced) in the microscopic field,"as Puttlitz and Seeley (63) say in their de-scription of probably similar movements ob-served in strains of Lampropedia hyalina. Ob-serving the field carefully this "flickering ef-fect" is eventually seen to be due to pairs ofcocci being ejected from the clusters, whereas,in contrast, pairs of cocci lying separately on

the agar surface are never seen to move. Thiswould thus seem to indicate that these move-

ments do not depend on contact between cellsurface and agar surface and that the explana-tion should be sought in the conditions withinthe cell clusters. Hence, the same kind ofejection of cell pairs from clusters ought to befound in ordinary wet mounts. And in point offact they are.An explanation of the observed movements

of staphylococci could be the following: thecocci are cemented in clusters by a common

capsular material or by electrostatic interac-tion, or both. Counterforces tending to disrupta given cluster are created as a result of cellmultiplication and when these forces becomebigger than the cementing forces, a pair of coccimay be "shot" out of the cluster. As for the cellmotion of L. hyalina, this explanation seems to

be consistent with the description given (63).Definition of darting. Darting is a kind of

surface translocation produced by the expan-

sive forces developed in an aggregate of cellsinside a common capsule and resulting in theejection of cells from the aggregate. The mi-cromorphological pattern is that of cells andaggregates of cells distributed at random withempty areas of agar in between. Neither cellpairs nor aggregates move except during theejection which is observed as a flickering in themicroscope.

Classification of the Bacterial SpreadingPhenomena

The six different spreading phenomena are

easily distinguished by agar plate microscopy,and this direct observation is essential for theclassification described in this paper. Swarm-ing and gliding are the only two phenomenathat might be difficult to differentiate withoutfurther information, for their spreading zones

may look similar since motility mainly takesplace in rafts. But because swarming is flagelladependent and gliding organisms never are

flagellated, a wet mount, perhaps supple-mented by a flagella stain, is all that is neededin additional examinations. Twitching andsliding bacteria may or may not be flagellated,but the motive force is in either case inde-pendent of flagella. Twitching bacterial cellspredominantly move singly and swarming andgliding cells predominantly in aggregates. Slid-ing bacteria do not move actively, rather theyare moved, or pushed, whole sheets of cellstogether, by centrifugal forces arising from cellmultiplication. The motive force of darting islikewise generated in the cell community bygrowth. A key for the identification of thedifferent spreading phenomena is presented inTable 3. It is based on the following threefeatures that also constitute the main elementsof the definitions proposed above: (i) the mi-cromorphological pattern of the spreadingzone, including the degree of cell organization;

+-FIG. 19. Alcaligenes odorans strainH 1079 culture plate incubated on meat extract agar at 30 C for 24 hr; 3/4natural size.

FIG. 20. Sliding. Peripheral part of spreading zone in Fig. 19 consisting of a densely packed monolayer ofcells. The darker spots, increasing in number towards the bottom of the picture, represent cells lying on top ofthe monolayer; ca. x300.

FIG. 21. Flavobacterium sp. strain U 120 culture plate incubated on meat extract agar at 30 C for 24 hr; 3/4natural size.

FIG. 22. Sliding. Peripheral part of spreading zone in Fig. 21 consisting of a very broad monolayer of closelypacked cells; ca. x200.

FIG. 23. Acinetobacter calcoaceticus strain BD-4 culture plate incubated on cytophaga agar no. 62 at 30 Cfor 24 hr; '/4 natural size.

FIG. 24. Sliding. Peripheral part of spreading zone in Fig. 23; ca. x200.

497VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


I

FIGS. 25-30.

498

SIMM&I . -

f,0

444IVA

I


mbr.asm

.org/D

ownloaded from



(ii) the manner of cell movement; and (iii) themechanism of movement, whether known orunknown.

CONCLUDING COMMENTSThe value of bacterial spreading as a taxo-

nomic criterion at first sight appears to bedifficult to assess, and the existing literature onthis subject mirrors the present state of bewil-derment (see, e.g., 23, 34, 52, 85). One of themain obstacles obviously is that too little isknown about the underlying mechanisms.

TABLE 3. A key for the identification of kinds ofbacterial surface translocation manifested as

spreading zones

Kind of Motive forcesurfacetranslo- Flagella Not flagellacation

Swarming Spreading also pro-duced on relativelydry agar surfaces.The cells movetogether.

Swimming Spreading only pro-duced on very moistagar surfaces. Thecells move separate-ly.

Gliding Motive force generatedby the individualcell. High activity.The cells movetogether.

Twitching Motive force generatedby the individualcell. Low activity.The cells move sepa-rately.

Sliding Motive force generatedby the cell communi-ty. Centrifugal forcein expanding colony.

Darting Motive force generatedby the cell communi-ty. Tension force incell aggregates.

The results presented in this study clearlydemonstrate that spreading (swarming in thebroad sense) on the surface of solid substratesis not just one phenomenon but may actuallybe caused by six different kinds of bacterialsurface translocation that can be identifiedwith great certainty either as swarming, swim-ming, gliding and twitching motility, or assliding and darting motion. Therefore, thevalue of spreading phenomena in taxonomyshould now be reconsidered, especially becausegliding motility, often inferred simply from theoccurrence of spreading, always has been usedas a character of decisive importance in taxon-omy. Another obvious consequence of this newknowledge will be an attempt to straighten outthe existing inaccuracies in nomenclature.

If the different kinds of spreading phen-omena are to be used as taxonomic criteria, thefirst condition is that each of them be recog-nized for what it is. For instance, some of theconfusion found in the literature concerningthe relationship between cytophagas and flavo-bacteria and the use of gliding motility as acriterion in their distinction may stem frommisinterpretations of spreading as being alwaysindicative of gliding. Thus Hayes (34) onlyobserved distinct gliding motility microscopi-cally in 2 out of 20 spreading strains assumed tobe cytophagas because of the spreading, and itseems an obvious possibility, in the light of thepresent investigation and also of that ofMcMeekin et al. (50), that at least some ofthese strains actually do not glide but spreadby sliding.Keeping strictly to the definitions given, it

can be seen that it is, indeed, different organ-isms that perform different kinds of surfacetranslocation.The demonstration of the occurrence of a

specific kind of surface translocation in abacterial strain may then clearly represent apiece of information of some taxonomic valuewhile the opposite kind, i.e., no demonstrablesurface translocation, is of much less valuebecause too little is known about the factorscontrolling the different kinds of spreading.The taxonomic value of the demonstration of

4-FIG. 25. Streptococcus sp. strain 4932/71 culture plate incubated on ascites agar at 33 C for 72 hr; 3//4natural size.

FIG. 26. Sliding. Peripheral part of spreading zone in Fig. 25; ca. x30X.FIG. 27. Bacillus anthracis strain Ax 11 culture plate incubated on cytophaga agar no. 62 at 22 C for 72 hr;

3/4 natural size.FIG. 28. Sliding. Peripheral part of spreading zone in Fig. 27; ca. x200.FIG. 29. Staphylococcus albus strain HV54 culture plate incubated on cytophaga agar no. 70 at 30 C for 24

hr; 3/4 natural size.FIG. 30. Darting. Peripheral part of spreading zone in Fig. 29; ca. x300.

VOL. 36, 1972 499


mbr.asm

.org/D

ownloaded from


HENRICHSEN

swarming, gliding, or twitching depends on theknowledge about the occurrence of these phe-nomena in different bacterial species. Whatlittle is known at present has been given underthe appropriate sections above. Briefly summa-rized, swarming is found in spore-bearing gram-positive species and in two species of Proteus;gliding is found in two, also otherwise special-ized, groups of gram-negative rods and bydefinition in cytophagas; and twitching isfound among species of Acinetobacter, Morax-ella, and Pseudomonas.

Sliding, on the other hand, has been demon-strated in widely different bacterial species,gram-positive as well as gram-negative, andtherefore can be of no taxonomic value at thegeneric level. Whether it may be of somelimited value at species level is not known, butsome results (e.g., with the streptococci) mightsuggest so.

Darting most probably will play no role intaxonomy, but a possible value in differentia-tion at the subspecific level cannot be excludedon the basis of the presently available informa-tion."Swarming" is an old and accepted term

that is still used about any spreading phenome-non, so that a "swarming zone" is generallyused to indicate a "spreading zone," no matterhow the zone has arisen. It is obvious from theobservations of the present study that thisnomenclatorial want of precision is no longeracceptable. I therefore suggest that the use ofthe term "swarming" should be restricted tothe motility phenomenon covered by the defi-nition given above for swarming, and that"swarming zone" only should mean a spread-ing zone produced by swarming in the narrow

sense. Furthermore, "swimming" should beused for the flagella-dependent motility seen inwet mounts on insufficiently dried agar platesand in semisolid "motility agar," even if thishas been poured into a petri dish, and themedium should not be called "swarm agar."

"Gliding" is an old term, and "twitching" isa relatively recently proposed term. Confusionmay be avoided if they are used in accordancewith the definitions given above, i.e., twitchingshould not be called a special kind of glidingmotility (45).Furthermore, "sliding" and "darting" are

here suggested as new technical terms to desig-nate the two remaining processes describedand defined in the present paper. The termshave been chosen because they are consideredto reflect a characteristic trait in these twokinds of surface translocation phenomena, justas the terms "gliding" and "twitching" do.

As already mentioned, the use of the term"motility" may give rise to misunderstandingsunless the kind of motility in question isspecified. Since "motility" traditionally coversflagella-dependent swimming motility, theminimum requirement must be always tospecify when swarming, gliding, or twitchingmotility is at issue. Because the cell motion insliding and darting strains is basically differ-ent, the term "motion" is used here rather than"motility." For the same reason Puttlitz andSeeley (63) write about cell motion of L.hyalina.The expression of all six kinds of spreading

phenomena depends on the growth conditionsand therefore also on the composition of themedium used. Humidity enhances spreading inall cases. The available moisture is dependenton: (i) the concentration of agar and nutrients;(ii) drying of the plates before use; and (iii) thehumidity of the atmosphere in the incubationchamber. In this connection it is worth men-tioning that different agars vary greatly withregard to the strength of the gel. Therefore,information about the concentration of agar ina given medium is of no value unless furtherparticulars (e.g., trade name) are given.The curious phenomenon of motile mi-

crocolonies produced by swarming and glidingorganisms and by at least one twitching strainseems to be more pronounced when motility ispartly inhibited. The motile microcoloniesmove in curved paths running predominantlycounterclockwise in the flagellated organisms,whereas in the nonflagellated organisms thereis no predominant direction. An explanationfor this has never been found, but a seeminglyplausible one is that in the flagellated orga-nisms the direction is determined by the direc-tion of rotation of the flagella relative to thebacterial cell. This hypothesis assumes thatflagella rotate, as has been suggested by Vaitu-zis and Doetsch (83), and that they always, atleast within a given organism, rotate in thesame direction on all or most cells in a popula-tion. Reversal of direction of movement of a cellwould not alter the direction of rotation of theflagella relative to the direction of cell move-ment, for the flagella are assumed to reorien-tate so that their extremities are always point-ing away from the leading part of the cell.

ACKNOWILEDGMENTSThis work has been greatly influenced by the

inspiring and constructive criticism continuouslyoffered by Hans Lautrop. For this I owe him mysincere thanks. I also thank all those liste4 in Table1, who supplied me with strains. The photographical

500 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



assistance of A. G. Overgaard and F. Laursen isgratefully acknowledged.

LITERATURE CITED1. Adler, J. 1966. Chemotaxis in bacteria. Science

153:708-716.2. Anacker, R. L., and E. J. Ordal. 1959. Studies on

the myxobacterium Chondrococcus colmnaris.I. Serological typing. J. Bacteriol. 78:25-32.

3. Anderson, R. L., and E. J. Ordal. 1961. Cyto-phaga succinicans sp.n., a facultatively ana-erobic, aquatic myxobacterium. J. Bacteriol.81:130-138.

4. Andrewes, C. H., and F. V. Welch. 1946. A motileorganism of the pleuro-pneumonia group. J.Pathol. Bacteriol. 58:578-580.

5. Armstrong, J. B., J. Adler, and M. M. Dahl.1967. Nonchemotactic mutants of Escherichiacoli. J. Bacteriol. 93:390-398.

6. Bijsterveld, 0. P. van. 1970. New Moraxellastrain isolated from angular conjunctivitis.Appl. Microbiol. 20:405-408.

7. Bisset, K. A. 1939. The mode of growth ofbacterial colonies. J. Pathol. Bacteriol.48:427-435.

8. Bisset, K. A. 1970. The cytology and life historyof bacteria, p. 39-40 and p. 122-123, 2 ed.Livingstone, Edinburgh and London.

9. Boltjes, T. Y. K. 1948. Function and arrange-ment of bacterial flagella. J. Pathol. Bacteriol.60:275-287.

10. Bredt, W., K.-H. Hofling, H.-H. Heunert, and B.Milthaler. 1970. Messungen an beweglichenZellen von Mycoplasma pneumoniae. Z. Med.Mikrobiol. Immunol. 156:39-43.

11. Burchard, R. P. 1970. Gliding motility mutantsof Myxococcus xanthus. J. Bacteriol.104:940-947.

12. Carter, S. B. 1967. Effects of cytochalasins onmammalian cells. Nature (London) 213:261-264.

13. Clark, F. E. 1939. Non-motile variants of Bacil-lus alvei. J. Bacteriol. 38:491-497.

14. Coetzee, J. N. 1963. Transduction of swarming inProteus mirabilis. J. Gen. Microbiol. 33:1-7.

15. Coetzee, J. N., and T. G. Sacks. 1960. Morpho-logical variants of Proteus hauseri. J. Gen.Microbiol. 23:209-216.

16. Costerton, J. W. F., R. G. E. Murray, and C. F.Robinow. 1961. Observations on the motilityand the structure of Vitreoscilla. Can. J. Mi-crobiol. 7:329-339.

17. Doetsch, R. N., and G. J. Hageage. 1968. Mo-tility in procaryotic organisms: problems,points of view, and perspectives. Biol. Rev.43:317-362.

18. Doll, W. 1956. Hemmung des Schwarmens vonProteus-Bakterien durch oberflaichenaktiveSubstanzen (Pril und Rei). Zbl. Bakteriol. I.Abt. Orig. 166:43-47.

19. Drews, G., and W. Nultsch. 1962. SpezielleBewegungserscheinungen von Einzellern, p.883. In W. Ruhland (ed.), Handbuch der Pflan-zenphysiologie, vol. 12/2. Springer-Verlag,

Berlin.20. Dworkin, M. 1966. Biology of the myxobacteria.

Annu. Rev. Microbiol. 20:75-106.21. Faust, L., and R. S. Wolfe. 1961. Enrichment

and cultivation of Beggiatoa alba. J. Bacteriol.81:99-106.

22. Fildes, P. 1925. Tetanus. I. Isolation, morphologyand cultural reactions of B. tetani. Brit. J.Exp. Pathol. 6:62-70.

23. Floodgate, G. D., and P. R. Hayes. 1963. Theadansonian taxonomy of some yellow pig-mented marine bacteria. J. Gen. Microbiol.30:237-244.

24. Gard, S. 1938. Das Schwarmphanomen in derSalmonella-Gruppe und seine praktische Aus-nutzung. Z. Hyg. Infektionskr. 120:615-619.

25. Gardner, G. A. 1969. Physiological and morpho-logical characteristics of Kurthia zopfii iso-lated from meat products. J. Appl. Bacteriol.32:371-380.

26. Gianelli, F., and E. Cabassi. 1967. Sciamatura diMoraxella. Ig. Mod. 40:81-90.

27. Gilardi, G. L. 1967. Characteristics of Al-caligenes odorans var. viridans isolated fromhuman sources. Can. J. Microbiol. 13:895-900.

28. Graham-Smith, G. S. 1910. The division andpost-fission movements of bacilli when grownon solid media. Parasitology 3:17-53.

29. Hageage, G. J. 1963. A cytological study ofgliding microorganisms. Ph.D. Thesis, Uni-versity of Maryland, College Park, Maryland,U.S.A.

30. Hageage, G. J., and R. G. E. Murray. 1962.Random rotation among.Vitreoscilla species. J.Bacteriol. 84:1345-1346.

31. Halfen, L. N., and R. W. Castenholz. 1970.Gliding in a blue-green alga: a possible mecha-nism. Nature (London) 225:1163-1165.

32. Halvorsen, J. F. 1963. Gliding motility in theorganisms Bacterium anitratum (B5W), Mo-raxella Iwoffi and Alcaligenes haemolysans ascompared to Moraxella nonliquefaciens. ActaPathol. Microbiol. Scand. 59:200-204.

33. Hauser, G. 1885. (Jber Faulnissbacterien undderen Beziehungen zur Septicamie, p. 1-94. F.C. W. Vogel, Leipzig.

34. Hayes, P. R. 1963. Studies on marine flavobac-teria. J. Gen. Microbiol. 30:1-19.

35. Henrichsen, J. 1972. Gliding and twitching mo-tility of bacteria unaffected by cytochalasinB. Acta Pathol. Microbiol. Scand. 80:623-624.

36. Henrichsen, J., L. 0. Fr0holm, and K. Bovre.1972. Studies on bacterial surface transloca-tion. 2. Correlation of twitching motility andfimbriation in colony variants of Moraxellanonliquefaciens, M. bovis, and M. kingi. ActaPathol. Microbiol. Scand., section B. 80:445-452.

37. Henriksen, S. D., and M. Svendsen. 1946. Rotat-ing and wandering colonies in a strain of genusBacillus. Acta Pathol. Microbiol. Sand.23:387-394.

38. Hoeniger, J. F. M. 1964. Cellular changes accom-panying the swarming of Proteus mirabilis. I.Observations of living cultures. Can. J. Mi-

501VOL. 36, 1972


mbr.asm

.org/D

ownloaded from


HENRICHSEN

crobiol. 10:1-9.39. Hoeniger, J. F. M. 1965. Development of flagella

by Proteus mirabilis. J. Gen. Microbiol.40:29-42.

40. Jahn, T. L., and E. C. Bovee. 1965. Movementand locomotion of microorganisms. Annu. Rev.Microbiol. 19:21-58.

41. Jarosch, R. 1962. Gliding, p. 573-581. In R. A..Lewin (ed.), Physiology and biochemistry ofalgae. Academic Press Inc., New York.

42. Kopp, R., and J. Miller. 1965. Effects of relatedanionic detergents on flagellation, motility,swarming, and growth of Proteus. Appl. Mi-crobiol. 13:950-955.

43. Kruse, E. 1953. Varianter af Staphylococcusaureus og albus. Nord. Veterinaermed.5:982-994.

44. Lautrop, H. 1961. Bacterium anitratum trans-ferred to the genus Cytophaga. Int. Bull. Bac-teriol. Nomencl. 11:107-108.

45. Lautrop, H. 1965. Gliding motility in bacteria asa taxonomic criterion. Publ. Fac. Sci. Univ. J.E. Purkyne, ser. K, 35:322-327.

45a. Lautrop, H., K. B0vre, and W. Frederiksen.1970. A Moraxella-like microorganism isolatedfrom the genito-urinary tract of man. ActaPathol. Microbiol. Scand., section B,78:255-256.

46. Lautrop, H., A. Reyn, and A. Birch-Andersen.1964. A comparative electron microscope studyof the cell walls of gram-negative bacteria, p.157. Proc. Scand. Congr. Pathol. Microbiol.,14th, Oslo.

47. Leifson, E. 1960. Atlas of bacterial flagellation,p. 5. Academic Press Inc., New York.

48. Lev, M., R. H. Alexander, and H. J. Sobel. 1969.A study of Chromobacterium typhiflavumfrom the rat. J. Appl. Bacteriol. 32:429-433.

49. MacDonald, J. B., M. L. Knoll, and R. M.Sutton. 1953. Motility in a species of non-flagellated bacteria (Fusoscillus girans). Proc.Soc. Exp. Biol. 84:459-462.

50. McMeekin, T. A., J. T. Patterson, and J. G.Murray. 1971. An initial approach to the tax-onomy of some gram negative yellow pig-mented rods. J. Appl. Bacteriol. 34:699-716.

51. Mitchell, R. G., and S. K. R. Clarke. 1965. AnAlcaligenes species with distinctive propertiesisolated from human sources. J. Gen. Micro-biol. 40:343-348.

52. Mitchell, T. G., M. S. Hendrie, and J. M.Shewan. 1969. The taxonomy, differentiationand identification of Cytophaga species. J.Appl. Bacteriol. 32:40-50.

53. Morrison, R. B., and A. Scott. 1966. Swarming ofProteus-a solution to an old problem? Nature(London) 211:255-257.

54. Murray, R. G. E., and R. H. Elder. 1949. Thepredominance of counterclockwise rotationduring swarming of Bacillus species. J. Bacte-riol. 58:351-359.

54a. 0rskov, J. 1947. Method for the demonstrationof bacterial flagellum activity. Acta Pathol.Microbiol. Scand. 24:181-183.

55. Pate, J. L., and E. J. Ordal. 1967. The finestructure of Chrondrococcus columnaris. I.Structure and formation of mesosomes. J. CellBiol. 35:1-13.

56. Pate, J. L., and E. J. Ordal. 1967. The finestructure of Chondrococcus columnaris. III.The surface layers of Chondrococcuscolumnaris. J. Cell Biol. 35:37-51.

57. Piechaud, M. 1963. Mobilite chez les Moraxella.Ann. Inst. Pasteur 104:291-297.

58. Piechaud, M. 1969. Observations sur la mobilitedes Moraxella. Abstract, IX Conference on theTaxonomy of Bacteria, Brno.

59. Pirt, S. J. 1967. A kinetic study of the mode ofgrowth of surface colonies of bacteria andfungi. J. Gen. Microbiol. 47:181-197.

60. Pringsheim, E. G. 1949. The relationship be-tween bacteria and myxophyceae. Bacteriol.Rev. 13:47-98.

61. Pringsheim, E. G. 1963. Farblose Algen. EinBeitrag zur Evolutionsforschung, p. 132. Gus-tav Fischer Verlag, Stuttgart.

62. Pringsheim, E. G. 1964. Heterotrophism andspecies concepts in Beggiatoa. Amer. J. Bot.51:898-913.

63. Puttlitz, D. H., and H. W. Seeley, Jr. 1968.Physiology and nutrition of Lampropediahyalina. J. Bacteriol. 96:931-938.

64. Roberts, J. L. 1935. A new species of the genusBacillus exhibiting mobile colonies on the sur-face of nutrient agar. J. Bacteriol. 29:2,29-237.

65. Ryter, A., and M. Piechaud. 1963. Etude aumicroscope electronique de quelques souchesde Moraxella. Ann. Inst. Pasteur 105:1071-1079.

66. Samuels, S. B., B. Pittman, and W. B. Cherry.1969. Practical physiological schema for theidentification of Herellea vaginicola and itsdifferentiation from similar organisms. Appl.Microbiol. 18:1015-1024.

67. Sasarman, A. 1961. Contribution A l'etude de lamobilite des colonies bacteriennes. III. Le roledes cils. Arch. Roum. Pathol. Exp. Microbiol.20:215-218.

68. Sasarman, A., M. Dospina, and V. Ganea. 1959.Contribution A l'etude de la mobilite des colo-nies bacteriennes. Note II. Mecanisme. Arch.Roum. Pathol. Exp. Microbiol. 18:307-314.

69. Sasarman, A., V. Ganea, and M. Dospina. 1957.Contribution A l'etude de la mobilite des colo-nies bacteriennes. Note I. Arch. Roum. Pathol.Exp. Microbiol. 16:395-414.

70. Schlieber, E. 1964. Experimentelle Untersuchun-gen zur Systematik des sogenannten Bacter-ium anitratum. Inauguraldissertation, Bonn.

71. Schmidt-Lorenz, W. 1969. The fine structure ofthe swarm cells of myxobacteria. J. Appl.Bacteriol. 32:22-23.

72. Schmidt-Lorenz, W., and H. Kiihlwein. 1968.Intracellulare Bewegungsorganellen der Myx-obakterien. Arch. Mikrobiol. 60:95-98.

73. Shinn, L. E. 1940. The -'iotion of colonies ofBacillus alvei as shown by lapse-time cinema-tography. J. Bacteriol. 39:22.

502 BACTERIOL. REV.


mbr.asm

.org/D

ownloaded from



74. Smith, L. DS. 1955. Introduction to the patho-genic anaerobes, p. 93 and 101. University ofChicago Press, Chicago.

75. Soriano, S., and R. A. Lewin. 1965. Glidingmicrobes: some taxonomic reconsiderations.Antonie van Leeuwenhoek J. Microbiol. Serol.31:66-80.

76. Stanier, R. Y. 1940. Studies on the cytophagas. J.Bacteriol. 40:619-635.

77. Stanier, R. Y. 1942. The cytophaga group. Bac-teriol. Rev. 6:143-195.

78. Stanier, R. Y. 1947. Studies on nonfruiting myx-obacteria. I. Cytophaga Johnsonae, n.sp., achitin decomposing myxobacterium. J. Bacte-riol. 53:297-315.

79. Stanier, R. Y., M. Doudoroff, and E. A. Adel-berg. 1971. General microbiology, p. 545, 3 rded. Macmillan, London.

80. Stanier, R. Y., R. Kunisawa, M. Mandel, and G.Cohen-Bazire. 1971. Purification and proper-ties of unicellular blue-green algae (orderChroococcales). Bacteriol. Rev. 35:171-205.

81. Steed, P. M. 1962. Simonsiellaceae fam. nov.with characterization of Simonsiella crassa andAlysiella filiformis. J. Gen. Microbiol.29:615-624.

82. Turner, A. W., and C. E. Eales. 1941. Motiledaughter colonies in the Cl. oedematiens groupand some other clostridia (Cl. botulinum C, Cl.

tetani, Cl. septicum). Aust. J. Exp. Biol. Med.Sci. 19:167-174.

83. Vaituzis, Z., and R. N. Doetsch. 1969. Relation-ship between cell wall, cytoplasmic membrane,and bacterial motility. J. Bacteriol. 100:512-521.

84. Voelz, H., and M. Dworkin. 1962. Fine structureof Myxococcus xanthus during morphogenesis.J. Bacteriol. 84:943-952.

85. Weeks, 0. B. 1969. Problems concerning therelationships of cytophagas and flavobacteria.J. Appl. Bacteriol. 32:13-18.

86. Weibull, C. 1960. Movement, p. 153-205. In I. C.Gunsalus and R. Y. Stanier (ed.), The bac-teria, vol. 1. Academic Press Inc., New York.

87. Weiser, R., J. C. Wimpenny, A. W. Asscher, D.B. Moffat, and M. Sussman. 1971. Inhibitionby ethylenediamine tetra-acetic (EDTA) of theswarming of Proteus mirabilis. Zbl. Bakteriol.I. Abt. Orig. A 218:434-436.

88. Wessels, N. K., B. S. Spooner, J. F. Ash, M. 0.Bradley, M. A. Luduena, E. L. Taylor, J. T.Wrenn, and K. M. Yamada. 1971. Microfila-ments in cellular and developmental processes.Science 171:135-143.

89. Willis, A. T., and K. Williams. 1970. Somecultural reactions of Clostridium tetani. J.Med. Microbiol. 3:291-301.

VOL. 36, 1972 503


mbr.asm

.org/D

ownloaded from


Bacterial SurfaceTranslocation: Survey and a Classification · BACTERIAL SURFACETRANSLOCATION growth in capsulated aggregates (i.e., spread-ing byejection). The macroscopic appearance

Documents