Helical transformations of Salmonella flagella in vitro

J. Mol. Biol. (1976) 106, 167-186

Helical Transformations RITSU KAMIYA

of SaZmonella Flagella in vitro AND SHO ASAKURA

Institute of Molecular Biology, Faculty of Science Nagoya University, Nagoya, Japan (Received 9 February 1976) Helical transformations of reconstituted Salmonella flagella were visualized by dark-field light microscopy. Flagella from 55670 strain were lefthanded helices with a pitch of 2.3 pm at neutral pH. When, however, the pH of the solution wss lowered to 4.7, they were discontinouously transformed into close-coils with a pitch of 0.5 pm and a diameter of 1.2 pm, and a further lowering of the pH converted these coiled flagella into so-called curly ones, righthanded helices with a pitch of 1.1 pm. The transformation was rapid and reversible. Two other kinds of flagella (5525 and 5530) also underwent such polymorphic conversions. Thus pH is an important factor in the control of flagellar transformation. As a result of the transformation, the degree of flow birefringence of a flagellar solution depends strongly on pH. Measurements of this parameter were useful in the study of the effects on the transformation of salt concentration and temperature.

1. Introduction Evidence has been presented that in bacteria such as Escherichia and Salmonellu, each organism swims by rotation of the helical flagella (Berg & Anderson, 1973; Silverman & Simon, 1974; Larsen et al., 1974; Berg, 1974). The flagellum, however, must not be considered as simply a static, rigid screw, since it displays remarkable polymorphism, i.e. it can assumea variety of distinct helices. The transformation of flagella was first definitely reported by Pijper & Abraham (1954). They observed by dark-field light microscopy that flagellar bundles of swimming bacteria were suddenly transformed from a normal helix to a curly one, with a pitch of half the normal value, and called this phenomenon “biplicity”. On the other hand, Leifson et al. (1955), examining stained preparations of bacteria with a light microscope, found that when some cultures of Proteus specieswere exposed to an acid environment, their flagella were rapidly and reversibly transformed into several forms, called by them curly, wiled and semi-coiledetc. Important as their findings were, it was not clear whether the environmental factors acted directly on the flagella, or through some cellular functions. Abram & Koffler (1964) found that reconstituted flagella themselves can assumetwo configurations : when Bacillus flagella were reconstituted at low pH values and observed with an electron microscope, they took on coiled appearances,as well as normal ones. In addition, Asakura et al. (1966) observed that Salmonellaflagella took on a curly configuration after a long period of dialysis against cold water, and that this transformation was reversible: the transformed flagella again becamenormal when 1 mM-ATP was added to the specimens.Another flagellin 167

IW

It.

li;\.lIIY.\

.\SI)

S. .\S;\lil’K;A

copolymerization experiment (Asakura & lim+ 1972) a,lso strongly suggest,ed that. flagella could assume srvera.1 distinct hrGal configurations. thcl condit,ions under which the flagellar transIn spite of these early findings: formation occurs have been unclear. This results from the fact) that until recentI? we have had no means of observing individual flagella in solutions. In 1974. however, Macnab & Koshland demon&rated that, dark-field light microscopy could he used for this purpose. With this method we examined the flagellar transformation it{, vitro, and found that flagella from certain stra.ins of Salmonelln rapidly changed t,heir shapes when the pH was changed. Also, \\‘e mea,sured the degree of flow hirefringenco of flagellar solutions, since this parameter sensitively reflects the over-all shape of individual flagella. This paper describes the ir/. r!itro transformations of flagella as revealed b&y light microscope observations and flo\r. hirefringence measurements. A part of this work has heen quoted by Oosa\va & ,&akura (1975).

2. Materials (a)

Preparation.

and Methods

and

reconstitution~

qf

jagella

Table 1 shows the 3 strains of bacteria used in this study. The strains were supplied by Dr T. Iino (Laboratory of Genetics, E’aculty of Science, University of Tokyo). In this paper, flagella derived from strains SJ670, SJ25 and SJ30 will be referred to as i, n and c, respectively, as in previous papers (Asakura & Iino, 1972; Kamiya & Asakura, 1974). Cultivattion of the bacteria of each strain and the isolation and purification of flagella were carried out by the method of Asakura et al. ( lQ64) and Asakura & Iino (1972). As longer flagella were advantageous both for obesrvation by microscopy and for measurement of flow birefringence, reconstitution was performed in the following way. Purified flagella were suspended in 0.15 &%-NRC1 to give a concentrat,ion of about 10 mg protein/ml, and then they were broken into short fragments ill a sonic vibrat,or (KSMlOO, Kubota,

TABLE Salmonella Abbreviation

straitl.8 used it? this study Antigen type

Strain

i

55670

,t

(’

I

Flagellar shape under physiological conditions

i

,,OWM/l

SJ25

1,Z

?lOTmd

SJ30

?,,I ,c

curly

Tokyo). 95% of the product was heated at 65°C for 3 min, followed by centrifugation at 105,000 g for 1 h. The clear supernatant fraction, which contained monomeric flagellin, was mixed with the unheated fraction (5% of the total prot,ein) and left at 26°C for one day, when polymerization was almost complete. The product was composed of long flagellar filaments, the average length being about 10 pm (Hotani &. Asakura, 1974). It was stored for subsequent use. (b)

Specimens

were prepared

to give a concentration used aa a negative 4.0 and 7.0. A JEM

Electron

by diluting

microscopy

t,he stock solution

of about 0.1 mg protein/ml. stain, with the pH adjusted 100 C microscope was used.

of flagella with

0.5% phosphotungstate with NaOH to various

distilled

water

solution was values between

HELT(‘AL

TH.~SSFORhlATIOSS

OF

FLAGELI,.\

IFY

(c) Dark-field light microscopy Usually, 0.1 ml of the flagellar stock solution was mixed with 100 ml of a solution COIItaining an appropriate concentration of NaCl and 1 mM-acetate buffer or 1 rnx-MacIlvaine’s buffer (1 mM-citric acid and 2 mM-Na,HPO* ; pH 4.9). After the pH of the mixture was adjusted to a given value by tire addition of a few drops of 1 N-NaOH or 1 N-HCl, a fraction of it was sampled for observation. The final caoncrntration of flagella wax about 0.01 mg/ml. The optical system used collsistetl of an Ustlio 100 \V mr~rc~lry arc lamp, a Nikon SKK microscope equipped with all Olympus DC condenser (numerical aperture = 1.2 to 1.33), rlrunerical aperture = l.O), a Nikon an Olympus Apo 40 x objecti1.r (oil immersion. 1 o x rayopiece, and a Nikon EFM camera. Thr slide wax 1.0 mm thick and was washed in !)00,, ethyl alcohol before use. T11r rise in temperatllre of tile specimen due to illumina~ tion was not measured. bllt Hotatti (1976) has suggest,rti that. in his optical system wllich was very similar to ours, t,lle temperature of specimerls did not rise above 35°C. As flagella Ila\~ a weak tetldency t,o stick to tilt> glass aurfacr at Iliph salt concentration (above 20 mM). we could focus and photograph them with an (exposure of about 1 s (Hotani, 1976). Ho\levPr, due t,o Rrownian motion, frtae-floating flagella in solnt,ioll could not be photographed under tile microscope. Pllotographs lverc usl~nll\taken at a total magnificatioli of 200 (calibratctl with a grating replica \vith 600 litlcs,lmm) on Kodak Tri-X film. tile effccti\.e speed of whicll was incrrasrd to ASA 3200 t)y tlc~r~lopmr~lt I\-itll Pandol drvelopr~ (Fuji plloto film (‘0.. Tokyo). (d) Jleasctremed of helical pura,raefers an<1 tleterTrrinatio)l of handedness Tllr pitch P and diameter U of flapellar helices wer(’ measured on a number of enlarged plmtopraphs. As a rule, flagella with more than 5 Ilrlical turns were chosen. Ca.re was taken to exchldr flagella whicll Ilad apparently snfiered deformation through interaction with the glass surface. In tllis paper we define pitch angle, 0, a,s the angle between a tangent to the filament alrd tile axis of the hrlix, i.e. tan 8 = nl)/Z’. Tllis parameter was calculated. The method ilsed for the detrrlnination of lLelica1 handedness was similar t,o tile opticalsectioning method of Weibull (1949) and our own (Shimada, et al., 1975). The difference is that, in tllis stlldy we used specimens containing nrithrr ammonium sulphate nor methylcellulose. Several photographs were taken of a flagellum while successively raising the spc~cimrrl stage, and the liaritletliicw was directly tlctc~rmincd from the pictures. ((1)

Flow

birefringence

measwemevts

An apparattls of concetltric cylinder type was used which had a rapacit,y of 3.0 ml and aI1 optical path-lcnpth of 20 min. All measrmments WPW performed at a velocity gradient of 160/s. The temperature of tlte specimen was controllc~d within _+ 0.5 deg C by circulating water around the outer c\vlinder. \vit,h H telnperaturr-controlled circulator. In many experiments, samples were prepared by mixing t,hr fla,gellar stock solution and a buffer solution with the desired pH and containing an appropriate concentration of Na(‘l to give a concentration of about 3.0 mg protein/ml. When very low concentrations of salt were required for the sample, the flagellar stock solution was dialyzed against distilled water. The pH of t,he sample was measured before and after the measurement of flow birefringellce. at the same temperature. In a few experiments. samples were prepared in another way as follows. A large volume of the stock wollltion was mixed with a solution containing an appropriate concrrrt,ration of MacIlvainr’s buffer and NaCl to give a final concentration of about 3.0 mg prot)ein/ml and I xn>l-bllKer. While adjusting the pH with the addition of a few drops of 1 N-NaOH or 1 N-H(‘1. WY’ took samples at different pH values for the birefringence measurement. Under tlltl c~xprrimental conditions employed, the degree of flow birefringenre depended on t,llr flapellar mean length. Since we could not st,rictly control t,he latter I-nine, the data oht,ainmi wets semi-ri~lantitative. (f) ,Vomenclature of the puolymorph.s The flagellar config~~rations described in tllra text w
170

R.

KAMIYA

AND

S. AHAKURA

TABLE 2 Terminology

Leifson (1960)

Shape aft)er drying

of the JEagelEar polymorphs

Present

Normal

Normal

Circle

Coiled

Coiled

Short wavelength large amplitude

Semi-coiled

Semi-coiled

Curly

Curly curly

Long large

Short small

wavelength amplitude

wavelength amplitude

Asakura (1970); Asakure & Iino (1972)

study

or curly II

I

Calladine (1975)

TYPO 1

f

Circular

e

Type Type

II or II’ III or III’

d c

3. Results (a) Electron

microscope

observations

The possibility that pH may have some effect on the flagellar shape was first suggestedby the electron microscope study in which the pH of the negative stain (0.5% phosphotungstate) was adjusted to various values. When i-flagella from SJ670 strain were negatively stained at neutral pH they always appeared normal, while they became coiled when stained with phosphotungstate at lower pH values (Fig. l(a) and (b)). Also, n-flagella from SJ25 strain took on a curly appearance when stained with an acidic phosphotungstate solution (Fig. I(c) and (d)). Since the unusual shapeswere brought about solely by the lowered pH value of the negative stain, we supposedthat a conversion of flagellar configuration was induced by H+ ion on the electron microscope grids, and that the conversion was so rapid that it was completed in the courseof drying the specimen,which took a few minutes at the most. (b) Flagellar

transformations

observed by light microscopy

The use of light microscopy enabled us to observe remarkable transformations of flagella when the environmental pH was lowered from 7 to 4. At pH values above 6.0, all the i-flagella in a solution assumedthe normal configuration, a lefthanded helix with a pitch of about 2.3 pm (Shimada et al., 1975) (Fig. 2(a)). If, however, the pH of the solution was gradually lowered by the addition of a small volume of HCl, some flagella were converted into a coiled form with a diameter of 1.2 pm and a pitch of 0.5 pm (Fig. 2(b)). Th e ratio of the coiled flagella to the total increased at lower pH, reaching a maximum at pH 4.7, when most flagella assumedthe wiled configuration unless the salt concentration exceeded 0.2 M. A further lowering of pH brought another conversion; from the wiled to the curly configuration (Fig. 2(c)). The latter configuration was a righthanded helix with a pitch of 1.1 pm (Shimada et al., 1975). The conversion into the curly form was favoured by a salt concentration higher than about 0.05 M: most flagella assumedthis configuration in the presence of 0.1 M-NaCl and 2 mM-MacIlvaine’s buffer of pH 4.2. All these conversions were reversible. Indeed, curly i-flagella at pH 4.2 were again rapidly transformed into wiled or normal flagella, when the pH was raised to 4.7

HELICAL

Pm. 1. Electron stained with 0.6% 8000 >

FL.AGELLA

171

(a) and (b), and n-flagella (c) and (d). Negatil 7.0 (a) and (c). and pH 4.5 (b) and (d). Magnifical

rely ;ion

TRANSFORMATIONS

micrographs of i-flagella phosphotongstat,e at pH

OL:

172

light micrographs of ,i-flagella. (a) Sormnl form found at pH 7.0, (b) c&oil, PI< 4. 2. Dark-field were suspended in a solvent, containil form at pH 4.7, and (c) curEy form at pH 4.1. Flagella acid and 1 rn>I-sodium acetate. and the pH of the suspension was change 0.1 M -NaCl, 1 mM-acetic of 1 N-HCI or 1 X-NaOH. Ilept~ifir~~tion Z540 by th le addition

HELICAL

TR~4SSPORMATTONS

OF

I 73

FLAGELL.4

or 7.0. It should be noted that under these experimental conditions, no appreciable amount of flagellar depolymerization or aggregation took place. Through observations of i-flagella conducted while both pH and ionic strength were varied systematically, we obtained a “phase diagram” of the flagellar state (Fig. 3(a)). As the temperature of the specimen (presumably about 30°C) could not be controlled under the microscope, this diagram must be considered as rather a rough one. Moreover, the boundaries were not distinctly drawn: two, or sometimes three types of configurations appeared simultaneously in rather broad regions near the boundaries. In such a region, flagella with two blocks of different forms were occasionally found (Fig. 4). However, it can be said that a lower concentration of salt was favourable to the occurrence of the coiled form, and the pH range where this form emerged became narrower when the ionic strength was increased. Though effects of salts other than NaCl have not been examined in detail, this tendency remained unchanged when NaCl was replaced by KC1 or MgCl,.

05,

(b)

(0) 0.4

-

s d z”

Cc)

I: II ; : -\I

c;

:

03-G!

i :

: ,

% =02-1 z 0

Normal I

curty 1

Normot t

\ \ ‘, I

: I OI-’ /

0-

4

Coiled

\

\

\

\ ‘t

5

\ 6

7

4

5

6 PH

FIG. 3. Diagrams showing the range over (a) i-flagrlla, (h) u-flagella, and (c) r-flagella.

which

particular

rwlymorphs

were

predominant.

Xext we describe transformations of n-flagella. When pH was lowered at low ionic strength. most of the n-flagella were converted into the so-called semi-coiled form (Fig. 5). The pitch of this form is as small as 1.2 pm, but it is distinguishable from the curly one by its large amplitude. Apart from the semi-coiled form, a small fraction of n-flagella took the curly and coiled forms, which were very similar to those of i-flagella. A curly n-flagellum is shown in Figure 5(c). Though the appearance of curly or coiled n-flagella was rare and could not, be controlled completely, they emerged most frequently under the conditions where most flagella had the semicoiled form. As Figure 3(b) shows, n-flagella had shapes other than the normal one only when the ionic strength was very low. We have menbioned that most ,n-flagella appeared curly in the electron microscope when stained with an acidic phosphotungstate solution (Fig. l(d)). The curly appearance possibly emerged from the transformation of the semi-coiled flagella on the grid (see Discussion). In a preliminary experiment we found that i-flagella also assumed t)he semi-coiled configuration at low pH in the presence of a, high concentration of citrate ion.

FI a. 4. Occurrence of different, forms iu a single i-flagellum. (a) ~Vornurl-rvilec( at pH 4.3, and (c) rrorrrmZ-curly at pH 4.7. Experimental conditions (b) clSiltYi- curly a8 in Fig. 2 except that the concn of Na(‘I was increaartl to 0.5 I in (c). ~lagnificatim~

at p1 WHI’B the 4000 8

HELIC’AL

TRANSFORMATIONS

Ol!

FLAGELLA

l! ‘IG. 6. I)ark-field light micrographs of n-flagella. (a) Nomaul form occurred at pH 7.0, (b) s( ?rnisd form at pH 4.1, and (c) co-occurrence of semi-coiled and curly forms at pH 4.1. Flag :elle e suspended in a solvent containing 0.01 M-Nacl, 1 mar-acetic acid and 1 m&r-sodium acet ,ate, WW by the addition of 1 w-HCl or 1 N-NaOH. Magnification: 2640x and the pH was adjusted (a) and (b); 4000 x (c). COill

K

liti

I\: A hl I \1’ .A .A S I ) S

.\ S A li I7 H ;I

wf

1s reported by Asakura & tine (1972). c-flagella assumed two configurations. lile these flagella formed helices wit)11 a pitch of 0.93 pm at a pH range higher than 6.0 and an ionic strength lower than 042. they were converted into the other type if the pH was lowered or the of 1lelix with a pitch of 0.90 pm and a smaller amplitude in pitch. the two ion ic strength was raised (Figs 6 and 3(c)). In spite of the similarity under the microscope: the formel tY1 )es of helices could be clearly distinguished

PH the

F‘IC. 6. Dark-field light micrographs of(~-flagella. (a) (‘trrly I form at pH X,tI, (b) m a flagellum at pH 6.0. Expwirncmtal 4.1, and (c) occurrence of the 2 forms same as in Fig. 5. Magnification: 2540 J: (a) and (b) : 4000 A (c).

rurly II conditions

form

ate wmx

HELICAL

TRANSFORMATIONS

OP

FLAGELLA

177

type with a large amplitude was called curly I and the latter curly II. They may correspond, respectively, to the configurat,ions called type II’ and type III’ by Asakura & Iino. (c) Flow

birefringence

of figella

The degree of flow birefringence of a flagellar solution strongly depends on the pH. For example, Figure 7(a) shows the dependence on pH of this parameter for i-flagella at different ionic strengths. When the pH value was decreased, the degree of birefringence diminished in the pH range between 7.0 and 4.7, and again increased at lower values of pH. Since depolymerization or aggregation was scarcely detected during the experiment,, the change in degree of hirefringence was interpreted as (b)

(0)

,

120

loo

01

4

’

I

I

5

6

7

1

4

5

6 PH

5

7

6

I

PH

FIN. 7. pH-dependence of the flow birefringence of flagellar solutions at various ionic strengths. (a) i-flagella in the presence of 3 concns of NaCl, i.e. 0.5 M ( l ), 0.2 I+X (a>) and 0.04 M (0). Each sample was prepared by mixing a stock solution of flagella and an acetate buffer solution with the desired pH and containing an appropriate concn of NaCI. Final concns: 2.7 mg flagella/ml and 2 mM-fWetate buffer. Temperature 22”C, and velocity gradient 160/s. (b) n-flagella. Final concn of flagella was 3.7 mg/ml, and the concns of NaCl were 0.1 M (a), 0.01 M (a) and 0 M (0). Other conditions were the same as in (a). (c) c-flagella. Final concn of flagella was 3.3 mg/ml, and the concns of NaCl were 0.15 M (0) and 0 M (0). Other conditions were the same as in (a).

178

K.

KAMIY.1

<\Sl)

S.

,%Rc\KUHA

representing the change in the over-all shape of individual flagella : it would be most reasonable to attribute these changes seen in Figure 7(a) to t,he transformations from the normal to the coiled form at, pH values between 7.0 and 4-T. and from t:hcl coiled to the curly form at pH values below 4.7. The increase in the minimum hirefringence value at around pH 4.7 when t,he ionic strength was raised indicates that, in agreement with the result of microscope observations, the formation of the coiled configuration was favoured by low ionic strength and a smaller fraction of flagella was converted into this form in the presence of relat’ively high concentrat,ions of salt. Figure 7(b) shows the data obtained for n-flagella. In this case, the low degree of birefringence at low pH values must be due to the occurrence of the semi-coiled form, as microscopy revealed. We should mention at this point that since semicoiled flagella have a large pitch angle (0 = 543, they must show a considerably low degree of birefringence even if they were well oriented in the flow. In the presence of 0.1 M-Nacl, the decline in birefringence at low pH was slight, indicating that only a small proportion of flagella was converted into the semi-coiled form. This again agrees with the results of the microscope observations. In the case of c-flagella, the degree of flow birefringence decreased when the pH was raised above 6-O at very low ionic strength (Fig. 7(c)). This keefold decrease must represent the transformation from curly II to curly I, for this transformation accompanies a considerable increase in pitch angle of the flagellar helix. Figure 8 shows that the transformations were dependent on temperature. When the temperature was raised, the pH value at which the flow birefringence began to decrease became lower and the minimum degree of birefringence was increased. These results suggest that the formation of the coiled and semi-coiled helices is favoured by lower temperature, (a) 120 -

100

80 : D

(b)

100 -

60

4 4c 402c

20-

C PH FIG. 8. pH dependence of the flow birefringence of flagellar solutions measured at various and 12.6% (0). A stock solution of flagella temperatures. (a) i-flagella at 33.O”C ( l ), 20.6’C (a) was mixed with an acetate buffer solution containing NaCl and the mixture was divided into parts for measurements at the 3 temperatures. Final concns: 3.1 mg flagella/ml, 0.2 M-NaCl and 2 maa-acetate buffer. Velocity gradient 160/s. (b) n-flagella at 34,O”C (a), 20.5”C ((I)) and ll.O”C (0). Experimental conditions were the same as in (a) except that final concn of flagella was 3.7 mg/ml and NaCl was absent. In each Figure, the disagreement of the value near neutral pH among the 3 curves may be due to the temperature-dependence of the orientation factor of flagella, and not to the difference in flagellar shape.

HELICAL

TRANSFORMATIONS

OF

FLAGELLA

179

In the measurements of flow birefringence so far mentioned, the specimens were prepared by lowering the pH of the stock solution, the pH of which was near 7.0. This means that the degree of flow birefringence of i and n-flagella obtained in these experiments was that of flagella which had been converted from the normal form. Another experiment was then performed to examine the reverse process, the conversion of coiled i-flagella into the normal form. In this experiment, specimenswere prepared from two stock solutions, each having been left at room temperature. at pH 4.4 and pH 6.7, respectively, for several hours, and the measurement was done within one minute after the pH was adjusted to the final value. As Figure 9 shows, the degree of flow birefringence of the flagellar solutions was independent of their previous pH values, at least one minute after the pH was changed. This implies that t’he transformation between the normal and coiled configurations is reversible, and proceeds without hysteresis. Transformations between the other kinds of polymorphs seem t’o have similar features, though detailed experiments have not heen done

45

50

55

60

65

PH FIG. 9. Flow birefringence data showing reversibility of the transformation of flagella. The open circles were obtained for i-flagella which had been left at pH 6.7, and the closed circles for those which had been left at pH 4.4. Each value of the degree of flow birefringence was measured within 1 min after the pH was re-adjusted with the addition of l/6 vol. of 70 miw-MacIlvaine’s buffer solution with an appropriate pH value. Final concns: 2.1 mg flagella/ml, 86 mM-N&l, 10 miw-citric acid and 20 mrvr-NazHPO,. Temperat’ure 22’C. and velocity gradient, 160/s.

(d) Helical

parameters

of the polymorphs

Recently we reported that, whereas normal flagella were lefthanded helices, curly flagella were righthanded, whether the flagella were from a mutant (SJ30) or a wildtype (55670) strain (Shimada et al., 1975). We present here evidence that the coiled configuration of i-flagella and the semi-coiledone of ,n-flagella were left- and righthanded, respectively. Figure 10(a) showsthree pictures of the same coiled i-flagellum taken with varied focussing, i.e. on the upper, middle and lower surfaces in (l), (2) and (3), respectively. It’ can be seenthat the image in (1) is made up of stripes which

FIG. 10. Determination of‘ helical handetlncss. (a) A set of’ micrographs of a co&l i-flagellum. This flagellum suffered deformation through int~craction with t,hr glass surface, and was markedly elongated. Focus was on the upper tiurface of the flagellum in (I), the middle portion in (2) and of a Cwmi-c*oiZed flagellum the lower surface in (3). Magnification 4700 x (b) A scxt of micrographs taken and arranged in the smn~ way as in (a). Magnification 4700 i(

tilt from the upper left to the lower right direction, while the stripes in (3) tilt in the . . . opposite direction: this is the case in a lefthanded helix. Another observation confirmed that the coiled form of n-flagella is also a lefthanded helix. A similar set of pictures was taken of a semi-coiled n-flagellum (Fig. 10(b)) showing that this helix is right-handed. In addition to handedness, other helical parameters of the polymorphs are listed in Table 3. In this Table, P and D represent, respectively, the pitch and the diameter of a helix measured on micrographs, and L denotes the contour length of the filament contained in one pitch of the helix, which was calculated from the formula,

HELICAL

TRANSFORMATIONS

OF

TABLE

Helical Strain

parameters

Handedness

Type

Left

SJ25

Nt

polymorphs

P(P)

D(P)

LbJm)

e(deg.1

5.32 7.10

20 20

2.27(0.03) 2.30(0.03)

0.42(0.03)

2.65

30

0.53(0.03)

1.16(0,07)

3.68

82

0.69

1.53

4.9

82

Coiled

Left

4.50

14

Coiledf

Left

5.90

5

Cztdy

Right

4.15 4.46

20 6

1~15(0~01) 1~14(0~01)

0.31(0.06)

1.51

40

Normcd

Left

4.60 6.98

16 10

2.28(0.02) 2.28(0.02)

O.SS(O.04)

2.57

28

Coiled$

Left

4.10

4

0.9

1.4

4.5

78

x 6

1.16(0.02) 1.14(0.02)

0.51(0.04)

1.98

64

1.1

&r&coiled

Curly

Numbers t M the 3 These

3

of flagellur

pH

1x1

FLAGELLA

II

Right

4.20 4.60

Right

4.10

4

0.3

1.5

40

Right

7.00 8.14

15 15

0.32(0.01)

1.37

47

Right

4.90 7.00 8.14

15 15 15

O.lR(0.02)

1.06

32

in parentheses are standard number of flagella used. polymorphs occurred rawly

deviations. and

data

show

largcb

scat,tw

L2 = (TD)~ + P2; and 0 is the pitch angle given by tan 6’ = TDIP. The value of P could be fairly precise, because it was obtained by dividing the length of a flagellum by the number of helical turns it contained, both of which could be measured easily and precisely. However, this was not the case in the measurement of the diameter: the broad image of a flagellum in comparison with its helical diameter might cause considerable errors. Thus the values of D, L and @listed must be considered approximate. In general, the helical parameters of each polymorph depended little on the environment,al conditions. For example, Table 3 shows that the pitch of a given polymorph at different pH values was constant within experimental error. This means that the flagellar transformation proceeds not continuously but in a discrete manner. However, a question remains in the case of coiled flagella. At pH values between 5.3 and 5.9, a small fraction of i-flagella assumed a coiled form. the pitch and diameter of which were considerably larger than those of the ordinary form observed at pH values between 4.3 and 5.0. As the coiled form occurs in a very small proportion at pH values above 5.3, and flagella with this shape often tend to be deformed when attached to the glass surface, it is difficult to be certain whether this larger coiled form is a disstinct, polymorph or simply an elongated state of the ordinary one. (e) Dynamic

aspects of the transformatr:on

The flagellar transformation could be followed under the microscope. As a typical case, we describe here the transformation of i-flagella from the normal configuration to the curly one. When pH was lowered rapidly, normal ,&flagella were found to be

K.

182

KAMIYA

:4X1)

s.

.\SBI
converted into the curly form wit,hout passing through the coiled state. IL drop of’ a!~ i-flagella solution containing WI M-,l’aCl and 1 InNI-phosphate In&r (pH 6.0) was placed on a microscope slide. A cover-slip was pub on it, two opposite sides of which were sealed with nail varnish, and the specimen was observed under the microscopcb. Then, all flagella had the ~zorr~znl form and many of thrm were sticking by one end to the glass surface. Hotani (personal communication) has shown that this end corresponds to the proximal end of flagella in intact, cells. When a drop of 10 ~RIMacllvaine’s buffer solution (pH 4.0) and a filter paper disc, respe&ively, were placed beside the two opposite open sides of the cover-slip, flow occurred in the specimen, orienting the flagella in the direction of the flow. Though many flagella were swept away in this process, some remained sticking to the glass. Continuous monitoring of these flagella occasionally revealed their dynamic behaviour during the transformation, a model of which is seen in Figure 11. In this model, the two types of helices have opposite handedness. As the pH of the specimen n-as lowered sufficiently, a small portion of the flagellum near the free end (the dista.1 ond) was suddenly wound up into a curly helix. The axis of the curly portion then appeared to move rapidly around the axis of the wrmal portion, while their junction travelled to the proximal end at a rate of a few pm/s. This kind of observation showed that the transformation proceeded unidirectionally, but the directionalit,y shown when both ends of a flagellum are free in solution and no shear forces are present is to be studied further. Also, more details of the dynamics of this and other kinds of tIransformations await further investigation with microcinematography.

4. Discussion Asakura & Iino (1972) have shown that copolymers of i or n-flagellin and flagellin from a straight-flagellar mutant (SJ814) assume six stable configurations including the circular (c&d) form, and have concluded t’hat homogeneous polymers themselves must be polymorphic. Although some of the polymorphs they found have not been observed in the present study, their general conclusion is confirmed in this study. Table 4 shows the pitch P and the contour length L of four of the polymorphs

Proximal

FIG. formation.

11.

A

schematic

end

illustration

of a flagellum

in

t,hr

procms

of the

normnl

to

curl?~

trans-

HELICAL

TRANSFORMATIONs

UP

183

FLAGELLA

TABLET Comparison

between the helical parameters measured by light and electron microscopy Flagellin

Type

*vormnz (type I) Coiled (circular) (‘urly (type II) Curly curzj/

i

denotes

2.49

2.70 4.3

1.36 1.17 0.99

1.54 1.48 1.12

i

i i 1 st ( c

z (typo 11’) 1z (type III’)

t s-tlagnllin

Asakura & Iino (1972) PW) L(w)

the

flagellin

from

a straight-flagellar

mutant,

Present PW)

study LW)

2.30 0.53 1.15

2.65 3.68 1.51

0.93 0.89

1.37 1.06

55814.

found by Asakura & Iino (1972), along with those of the polymorphs observed in this study. The disagreement between the values for P may be due to the fact that Asakura & Iino measured the wave-length of flattened flagella as an approximate value of a helical pitch: flagella seem to elongate with a constant value of L when they are dried on a surface (Hot,ani, 1976). The type II form of the copolymer and the curly form of i or n-flagella have quite similar L values and have the same handedness (Shimada et al., 1975). So we assume the two forms to be homologous. On the other hand. Asakura & Iino have shown, using copolymers of i and c-flagellins, that the type II form of copolymer and type II’ (curly I) form of c-flagella are homologous. Therefore, we may regard the curly I form of c-flagella as homologous to the curly form of i or n-flagella, though their helical parameters differ to some extent. When Asakura & Iino compared two kinds of specimens of i-flagella prepared by shade\\--casting and negative staining, a circular (coiled) form was predominant in the shadow-cast’ed specimen, whereas it was rarely found in the negatively stained specimen. This can be understood if we assume that the pH of the specimen in the formrr situation was decreased since the specimen was prepared by diluting a flagellar solution with a large volume of distilled water, while in the latter case the negative st,ain (pH 7) acted as a buffer. Another misleading observation by Asakura et al. (1966) on the reversible transformation (see Introducbion) must have been caused also bv changes in pH of the electron microscope specimens. Hot,ani (1976) has shown the occurrence of the semi-coiledform in the product of flagellin copolymerization, while this form was missed by Asakura & Iino. Xemicoiled flagella seem to have a tendency to transform as they are dried on a surface. We have shown that the curly shape observed by rlectron microscopy when n-flagella were negatively stained wibh phosphotungstate at acid pH have originated from the bransformatJion of the semi-coiledflagella. Resides the curly form, an aberrant shape with an alternating regularity was occasionally found in such a specimen sta,ined witJh acidic phosphotungstate (Fig. 12). The same aberrant shape has been found by Leifson et al. (1955), who described it as one of the flagellar polymorphs. Howcbver. we have confirmed that no flagella had such alternating regularity even in the presence of 1 o/,, phosphotungstate. The aberrant, shape, which might be an alternating combination of curly and semi-coiled forms. is probably an artifact brought about by drying.

R.

184

KAMIYA

rlND

8.

ARAKUKA

The bacterial flagellum may be represented by the surface of a cylinder whose axial line is deformed into a large-scale helix (Lowy & Spencer, 1968; Asakura. 1970). The surface comprises t,wo particular helical lines. the outermost line and t,he innermost line. Since the two helical lines have the same pitch but, different diameters, the length per pitch of the outermost line L, is larger than t’hat of the innermost line L,. Denoting the mean value of L, and Li as L, cx = (L, -- L,)/L corresponds to the curvature of the cylinder. It is given by CI = ZT~~DIL~, where d is the diameter of the cylinder and n is that of the large-scale helix. Also, the inclination Af? of the outermost line or the innermost line to the axial line of the cylinder is given by A0 =: 180 dP/L2 (deg.), where P is the pitch of the helix. When the helix is righthanded (or lefthanded), the outermost and innermost lines are twisted about the axial line of the cylinder in righthanded (or lefthanded) sense. So AtJ has a plus or minus sign depending on the handedness of the helix (Calladine, 1975), though Asakura (1970) and Asakura & Iino (1972) paid no attention to this point. Table 5 shows values of a and A6 calculated assuming d = 0.02 pm (O’Brien & Bennett, 1972). Tt is of interest’ to note that in each species of flagella. a form which is twisted more in t,he righthanded direction occurs as the pH becomes lower. Calladine (1975) has predicted that in the relations between the flagellar polymorphs, the co&d configuration should be between the normal one and the curly one in order. However, he did not take into account the presence of the semi-coiled form, which is between t,hc coiled and the curly forms. The simultaneous occurrence of two or three configurations has been described. TABLE

Morphological

twisting

Flagella

5

(AO) and bending TYP

(a)

in jlagellar

AtJ (deg.)

E ( x 100)

i

Normal Coiled Curly

- 1.3 ~ 0.2 1.8

2.4 3.3 5.4

?I,

Normal Semi-coiled

- 1.3 1.0

2.3 5 I

c

Curly Curly

1.8 2.9

6.8 6.4

1 II

helices

At pH 5.5 in the presence of 0.1 M-N&~, i-flagella assumed the normal form and the coiled form: the numbers of the two types of flagella were approximately equal, and a small number of flagella were of the mixed type like those shown in Figure 4(a). In this solution, for example, we observed no flagella fluctuating between the two forms. Even in the mixed-type flagella, rapid fluctuation was absent. On the other hand, we have obtained evidence that the normal-coiled transformation is a rapid process that involves no hysteresis (Fig. 9). This means that the ratio of the numbers of the two types of flagella rapidly reaches an equilibrium value. In order to satisfy this condition, individual flagella must continue rapid fluctuations, which, however, have not been observed in this study. To solve this inconsistency, more detailed analyses of fluctuation and hysteresge are now in proress. As illustrated in Figure 11, the flagellar transformation takes place in a transitiwtial

186

K.

KAMIY.1

.\SI)

s.

,\s.\I
manner. In this respect the transformation is Gnilar to t’hth (~ont~ra<~t~i(~n of thcb IJ~c~t,eriophage sheath (Moody, 1973): 1lo\Yrvcr, this pJw!WS is kno\vn to 1~. irrcvtkrsit)l(b. Among large ordered biological struct~urcs that arci known to fw polymorphic*. onl! bacterial flagella have been shown to undergo reversible transition between the polymorphs. The mechanism involved in this transition is st,ill to bc determined. but it must be correlat,ed with the problem of packing identical suimnits in a supercoiled structure (Klug, 1967; Asakura. 1970: Harris, 1974; Callsdine, 1975). Flagellar transformation actually occurs in living bacteria (Pijper & Abraham, 1954), but its physiological meaning is uncertain at present,. If bacteria swim by rotation of the normal flagella in an appropriat*e sense. then the bacteria must be “tumbled” by t’ransformation of the flagella into another form (for example CUTZ~), as well as by reversal of the sense of flagellar rotation (Silverman & Simon, 1974; Larsen et aZ., 1974; Berg, 1974). In relation to these circumstancc,s, Macnab’s report (First Intersectional Congress of the Itafernatiwzl ilsrociation of ,Mdcrof&ogical Societies, Tokyo, September, 1974) is noteworthy. Hc observed that, during bacterial tumbling, the flagella were converted from t,he normal configuratiou to th(b curl!/. This finding suggestis that, the flagellar transformation pIa?. 7~ a role in t.hc: mechanism of tumbling and therefore tactic responses. We are indebted to Dr T. Iino for strailts ofSalmonella. Dr H. Hotani (Kyoto Urliversity) first suggested to us the use of light microscopy. In the early course of this work, Dr K. Maruyama (Kyoto University) kindly offered facilities for the measurement, of flov birefringence. We also thank Dr F. Oosawa for stimulating discussions, and Dr It. M. Macnab for critically reading and correct,ing our manuscript. REFERENCES Abram, D. & Koffler, H. ( 1964). J. Mol. Biol. 9, 168 -185. Asakura, S. (1970). A&an. Biophys. (Japan), 1, !19-155. Asakura, S. & Iino, T. (1972). J. Mol. Biol. 64, 251 268. Asakura, S. Eguchi, G. & Iino, T. (1964). ,/. Mol. Biol. 10, 42-56. Asakura, S., Eguchi, G. R: Iino, T. (1966). ,I. iVo2. Riol. 16, 30% 316. Berg, H. C. ( 1974). Nature (London), 249, 77 79. Berg, H. C. & Anderson, R. A. (1973). A’ature (London,), 245, 38G382. Calladine, C. R. (1975). Nature (London), 255, 121 -124. Harris, W. F. (1974). J. Theoret. BioZ. 47, 295--308. Hotani, H. (1976). J. Mol. Biol. 106, 149-164. Hotani, H. Asakura, S. (1974). J. Mol. liliol. 86, 285-302. Kamiya, R. & Asakura, S. (1974). J. ;Vol. Biol. 87, 55-62. Klug, A. (1967). Symp. Int. Sot. Cell Riol. 6, 1 18. Larsen, S. H., Reader, R,. W., Kort, E. N., Tso, W.-TV. & Adler, ,J. (1974). ;Vature (Lon,don) 248, 74-77. Leifson, E. (1960). Atlas of Bacterial E’lagellation, Academic Press, New I-nrk Ji London. Leifson, E., Carhart, S. R. & Fulton, M. (1955), J. &cbevioZ. 69, 73 82. Lowy, J. & Spencer, M. (1968). Symp. Sot. Exp. Biob. 22, 215-236. Macnab, R. M. & Koshland, D. E., Jr (1974). .I. i$fol. Biol. 84, 399-406. Moody, M. F. (1973). J. Mo2. BioZ. 80, 613-635. O’Brien, E. J. & Bennett, P. M. (1972). J. ,VoL. Biol. 70, 133-152. Oosawa, F. & Asakura, S. (1975). Thermodynamics of the Polymerization of Prvteiq Academic Press, London, New York and San Francisco. Pijper, A. & Abraham, G. (1954). J. Gen. ,Vicrobiol. 10, 452-456. Shimada, K., Kamiya, R. & Asakura, S. (1975). Xature (London), 254, 332-334. Silverman, M. & Simon, M. (1974). Future (I,ondonf. 249, 73-74. Weibull, C. (1949). Arkiw Remi, 1, 573.-575.