Pro. Natl. Aad. Si. USA Vol. 74, No. 5, pp.2045-2049, May 1977 Cell Biology Diretion of ative sliding of mirotubules in Tetrahymena ilia (dynein/ell motility/eletron mirosopy) WINFIELD S. SALE AND PETER SATIR Department of Physiology-Anatomy, University of California, Berkeley, Berkeley, California 94720 Communiated by Daniel Mazia, Marh 3, 1977 ABSTRACT Axonemes of protozoan (Tetrahymena thermophila BIII) ilia, isolated by the dibuaine method, were treated briefly with trypsin after removal of the iliary membranes by treatment with Triton X-100. After attahment to polylysine-oated surfaes, the partially digested axonemes remained mainly intat ylinders. Suh attahed axonemes an be treated with ATP, whih indues mirotubule sliding. ATPtreated preparations showed disrupted axonemes in whih doublets had telesoped out of the original ylinders. These ould be aptured in plae for eletron mirosopy after ritial point drying. Images of this type were used to determine relative movement between adjaent doublet mirotubules. Eah doublet atively slid relative to its neighbors in a single diretion, in whih the polarity of fore generation of the dynein arms was from base to tip. It is well established (1-4) that the soure of fore generation of beating eukaryoti ilia and flagella is an interation between the axonemal doublet mirotubules, this interation ausing some doublets to slide relative to their neighbors. The ative shear between doublet mirotubules is resisted in a ontrolled manner (5) to form the omplex, sometimes three-dimensional, bends that propagate along the organelle. When ATP is added appropriately to detergent-treated ilia or flagella, they will reativate with more or less normal movement depending on the system used (6). In suh iliary models, the membrane has been removed, leaving the 9 + 2 mirotubule array and intermirotubule onnetions intat. Of the latter, the three signifiant onnetions are: the dynein arms, the radially direted spokes, and the interdoublet or nexin links. In 1971, Summers and Gibbons (2) demonstrated that, in axonemes from sea urhin sperm, sliding of doublet mirotubules ould be unoupled from bending if the axonemes were treated briefly with trypsin. As revealed by dark-field mirosopy, the sliding doublets telesope outward, extending to five or more times the length of the original axonemal fragment. In these experiments, trypsin seletively digested the spoke heads and nexin links without appreiably digesting the dynein arms (7). These arms are the site of the major iliary ATPase, now termed dynein-1 (f. Gibbons et al. in ref. 4). It appears, therefore, that mirotubule sliding ours when the dynein arms of one doublet ylially interat with-i.e., "walk along"-sites of the mirotubule lattie of the adjaent B subfiber. The arms an be envisioned to at in a manner analogous to myosin ross bridges. The diretion in whih the arms generate fore, whih is related to the diretion of ative sliding in the iliary system, has not previously been identified. This is of onsiderable importane to our understanding of iliary motility and perhaps of mirotubule sliding in other situations (for general review, see ref. 4). The relative diretion of movement between mirotubules during ATP-indued sliding annot be resolved by dark-field mirosopy. We have therefore modified the original experiment so that the resulits ould be viewed by whole-mount eletron mirosopy. We reasoned that, with a omplete understanding of the fine struture, espeially those strutures that allow identifiation of the base and tip diretion along the axoneme, we ould determine the relative diretion of movement between doublets and thus the polarity of fore generation. MATERIAL AND METHODS Cilia from Tetrahymena thermophila (strain BIII) were isolated by the dibuaine proedure as desribed previously (8). Exept where stated otherwise, ilia and axonemes were kept at 0-4. Isolated ilia were sedimented at 10,000 rpm for 8 min in a Sorvall HB-4 rotor and suspended in a buffer (referred to as solution A hereafter) ontaining 30 mm N-hydroxyethylpiperazine-N'-2-ethanesulfoni aid (Hepes), 5 mm MgSO4, 20 mm KC1, 1 mm dithiothreitol, and 0.5 mm EDTA at ph 7.6. The ilia were sedimented again and suspended in solution A ontaining 0.5% (vol/vol) Triton X-100 at a iliary protein onentration of about 0.25 mg/ml. After 10 min, axonemes were sedimented by entrifugation at 10,000 rpm for 8 min, and the axonemal pellet was resuspended in solution A, without detergent, at a protein onentration of about 0.5 mg/ml. Trypsin (Sigma, bovine panrease, type III) digestion of the axonemal suspension was arried out in a 0.5-m uvette at room temperature and at a trypsin/axonemal protein ratio of approximately 1:500. The digestion proess was monitored turbidimetrially with a Zeiss spetrophotometer PMQII. Routinely, the digestion proess was halted with exess soybean trypsin inhibitor when the optial density at 350 nm dereased to 80% of the initial value. The digested axonemes were stored on ie for immediate use, but oasionally axonemes were pelleted and resuspended in fresh solution A. Protein determinations were by the method of Lowry et al. (9) with bovine serum albumin as a standard. For dark-field observations, the Zeiss ultraondensor with tungsten illumination was used, and photomirographs were made with 3- to 5-se exposures on Kodak Tri-X 35-mm film. For observation of ATP-indued disruption of the axonemes, a drop of the digestion suspension was added to a glass slide, whih sometimes was oated with polylysine (10), and a over slip was applied. A drop of 1 mm ATP (disodium salt from Sigma) in solution A was then added to the edge of the over slip, and the front of the diffusing ATP was followed. A reord of the effets indued by ATP was made by taking photomirographs of axonemes on the slide surfae before and after the diffusing ATP reahed them. For eletron mirosopy, digested axonemes were applied to a polylysine-oated, Formvar-arbon-stabilized, opper grid (10, 11). The grid was drained by touhing it to the edge of a piee of filter paper, and immediately a drop of solution A (ontrol) or a drop of the ATP solution was added to the grid and allowed to sit for 5-10 se at room temperature. (It is important that the grid surfae is never allowed to air dry.) At the end of this inubation, the grids were drained, quikly inverted, and floated on a drop of 3% (vol/vol) glutaraldehyde in 50 mm sodium aodylate buffer, ph 7.3, for fixation for 45 min. Axonemes and disrupted axonemes, whih adhered firmly to the polylysine surfae, were postfixed, stained and ritial point 2045
2046 Cell Biology: Sale and Satir Pro. Natl. Aad. Si. USA 74 (1977). -1 '5kt, IN 'L JL.- -* I--,. -.. 1-11 1% 1-1 I. 0 ki FIG. 1. (Inset) Cross setion of axoneme (for details of fixation and embedding, see ref. 11). (X55,000.) (a) Dark-field mirograph of digested axonemes prior to the addition of ATP. (X3300.) (b) Mirograph of the same axonemes after addition of ATP. Arrow depits an axoneme that disrupted by doublets sliding out of one end. (X3300.) dried, and viewed as desribed previously (11). Random fields in all quadrants of a grid were hosen at low power for ounting disrupted axonemes. Negative staining of axonemes or ATP-treated axonemes was aomplished by adding three suessive drops of 1% aqueous uranyl aetate to the grid. (Grids generally are not polylysineoated for negative staining.) The final drop was drained by touhing the grid edge to filter paper and the grid was plaed under over to dry. RESULTS AND DISCUSSION Axonemes before ATP addition Dibuaine-isolated ilia are very pure (8). Extration of these ilia with 0.5% Triton X-100 solubilizes the iliary membranes (Fig. 1 inset). Axonemes ould be reovered from the preparation and then treated with trypsin. The kinetis of the digestion proess followed approximately the same path as that desribed by Summers and Gibbons (7). We used a relatively high onentration of the enzyme so that the total digestion time was relatively short (<5 min). In dark field, the trypsin-treated axonemes were quiesent, luminous rods (Fig. la). We looked for effets of digestion on interdoublet links, the spokes (strutures digested in sea urhin flagella), and the dynein arms. In Tetrahymena ilia, morphologial disruption of the spokes is largely onfined to the spoke head. Triplet spoke groups were easily identified after trypsin treatment in ritial point dried (Figs. 2 and 3a) and negatively stained (Fig. 4) axonemes. Dynein arms were more learly visible after digestion (Fig. 3b) than before and probably were not seriously disrupted by treatment. Almost all polylysine whole-mounted axonemes prior to trypsin digestion appeared as entirely intat ylinders (Figs. 5a and 6). After treatment with trypsin, the number of axonemes that unroll (splay) (f. ref. 11) or disintegrate into doublets is inreased 10-fold (Fig. 6). When treated with trypsin, a few doublets (f. ref. 12) or even whole axonemes oil into a helix (Fig. 5b). Disintegration and oiling suggest that the interdoublet links and spokes are onsiderably weakened FIG. 2. Critial point dried axoneme in whih three doublets (X, Y, Z) have slid out of the axonemal bakbone toward the tip. The tip an be identified by several features, notably the triplet spoke arrangement shown in the Inset (see text for further desription). The arrows define the diretion of movement of the doublets. (X24,500; Insert, X41,000.) or digested by trypsin, in aord with the results for flagella. Axonemes after ATP addition Without trypsin treatment, Tetrahymena axonemes prepared by our methods were reativated by ATP to only a limited extent. Beating ilia were rarely seen, but many axonemes twithed briefly and stopped, or twithed and disintegrated by sliding. As might be expeted from the behavior in dark field, ATP addition, without prior trypsin treatment, resulted in an inrease in the perentage of splayed and disintegrated axonemes seen in whole-mount eletron mirographs, inluding some images that showed ontat and telesoping between doublets that we identified as the sliding onfiguration (see below). About one-quarter of the total axonemes were disrupted (Fig. 6). This dramati inrease in disruption without trypsin treatment is probably produed when ATP auses relaxation of the rigor state of the dynein arms (13). Spoke head attahments and interdoublet links in disintegrating axonemes may be disturbed by mehanial damage or proteolysis or plastiized by ATP. After trypsin treatment in dark field, as the ATP diffused slowly aross the slide, nearly every axoneme twithed briefly and then disintegrated (Fig. lb). When GTP was substituted for ATP, there was no notieable movement or disruption of the digested axonemes. During sliding disintegration, doublets or groups of doublets were extruded from the axonemal ylinders, whih elongated to several times their original length and dereased dramatially in diameter and luminosity. One disintegration ourred, the elongated, usually urved, thin rods remained in plae and no further ative movement- was seen. We have assumed that ATP added to polylysine-mounted axonemes on a grid will dupliate these light mirosope results, and that the polylysine will often bind the free produts of disintegration to the grid so that images of sliding mirotubules an be studied at the eletron mirosope level. Addition of ATP to digested axonemes attahed to the polylysine surfae resulted in disruption of 50% of the total axo-
Cell Biology: f Sale and Satir Pro. Natl. Aad. Si. USA 74 (1977) 2047..,1 t<j.~ r It 'V;' "V,. FIG. 4. (a) Negative stain of a portion of a splayed axoneme in whih the dynein arms tilt toward the base (markers along left doublet). The basal diretion an be determined by the spoke arrangement (brakets) of the right doublet. (X73,500.) (b) Uranyl aetate negative stain of a pair of doublets, showing tilt of the dynein arms. (X84,000.) a b6 - FIG. 3. (a) Pair of ritial point dried doublets in whih the left doublet (L) has moved toward the tip (arrows define relative movement) relative to the right doublet (R). The dynein arms of the right doublet (not seen) have provided the motive fore, and therefore the image fits ase I of the diagram in Fig. 7. The diretion of dynein fore generation is from base to tip. Spoke groups are braketed. (X52,500.) (b) Uranyl aetate negative stain of a pair of doublets that are displaed longitudinally. The A and B subfibers of eah doublet are distinguishable by the origin of the dynein arms. Using the information that the free, extended wing-shaped arms (e) of the right doublet point toward the base, we onlude that the image fits ase I. In the overlap region, the fore-generating arms appear more flattened (f) between doublets. (X80,500.) standard markers (5, 11, 14): the entral pair extension and ap; the spoke group spaing (inset, the spokes are arranged in triplet groups; 21 nm between the distal pair of spokes and 28 nm between the proximal baseward, pair); and the fragment of the A subfiber extension of the middle doublet (Y) of the three. The telesoping doublets were approximately the same length and about the same length as the other doublets of the axonemal fragment (entire length not shown in the figure). This is expeted if the image represents sliding; it is unexpeted if artifatual. It is diffiult to ount the total number of doublets ( nemes ounted (Figs. 5b and 6). The remaining intat axonemes probably were impeded from disintegrating by adherene to the surfae. There was an inrease in splayed axonemes and in isolated doublets ompared to the ontrol axonemes suh that the effets of ATP and trypsin alone were roughly additive. Furthermore, nearly one-fifth of the axonemes observed were in a speial onfiguration in whih the ends of some doublets of an axonemal fragment were displaed longitudinally while intimate lateral onnetion was maintained generally for 0.8-2.0,um (Fig. 5b, see S). This orresponded to the telesoping axonemes in dark field and the image that would be expeted as a result of doublet sliding, provided that the displaed doublets are aptured by the surfae. We onlude that this in fat represents the sliding onfiguration. In ritial point dried preparations, this image is never seen without the addition of ATP. Fig. 2 is an example of an axoneme showing the ATP-indued sliding onfiguration at higher magnifiation. The bulk of the axoneme, whih is firmly attahed to the polylysineoated surfae, ats as a trak on whih three doublets have moved toward the tip. The tip an be identified by numerous a - 'N N INsk IN 11I \~~~~0~- - - FIG. 5. (a) Low-power image of untreated ritial point dried axonemes attahed to a polylysine-oated grid. They are primarily intat. (X3,800.) (b) Axonemal fragments resulting from addition of ATP to the trypsin-treated axonemes prior to fixation. There is a great inrease in disrupted axonemes. Note image S in whih longitudinal telesoping of doublets has taken plae. Helially oiled fragment is seen at H. (X3,800.)
2048 Cell Biology: Sale and Satir Pro. Natl. Aad. Si. USA 74 (1977) AXONEMES AXONEMES D IGESTED DIGESTED + AXONEMES AXONEMES, ATP AT CASE I l 8 o Subfiber Subfiber A B r.~~~~~~~~~~r v BASE INCT SPIT SLID INCT SPLY SLID *CT SPIY DTS SLID I ltly VlID AXONEME CONFIGURATIONS FIG. 6. Axonemal onfigurations. Eah histogram represents more than 500 axonemes. Experiments ounted are untreated axonemes; axonemes plus ATP; trypsin-treated (digested) axonemes; and digested axonemes plus ATP. (Fig. 5a and b are examples of paired fields ounted.) Categories inlude intat axonemes (INCT), splayed axonemes (SPLY), individual doublets (DTS) divided by 10 (entral pairs look similar to doublets at low magnifiation), and the sliding onfiguration (SLID) that is found only after ATP addition. if >ig] BASE spokes dynein arms CASE _ w II C of the axoneme due to superimposition of strutures, but it does appear that six doublets and the entral pair remain in the intat portion of the axoneme. Presumably, the three "sliding" doublets have moved on the top half of the intat axoneme, fallen to the side, and beame attahed to the polylysine surfae, further ative sliding being impeded by the attahment. The residual overlap observed between sliding doublets may indiate that not enough fore for further sliding is generated when the number of dynein arm interations falls below a minimum. Polarity of sliding We are now in a position to examine the polarity of sliding. To do this, we restrit our attention to images where only two doublets are overlapping. In the simple ase of ative sliding between two doublets, one next to the other as in the diagram of Fig. 7, there are only two possible relative diretions of motion. As drawn, mehanohemial ativity in the form of yli dynein ross-bridging an only take plae between the dynein of the doublet on the right and attahment sites on the adjaent B subfiber of the doublet on the left. In the absene of attahment, the dynein arms on the left doublet would not ontribute to the sliding interation. In ase I, the diretion of ative fore generation of the dynein arms is from base to tip (small arrow) so that the left doublet will move toward the tip relative to the right doublet. In ase II, the diretion of fore generation is from tip to base and the left doublet will move toward the base relative to the right doublet. By examining images in whih there has been longitudinal displaement between any pair of doublets, and in whih the tip and base are learly identifiable, we were able to distinguish whether ase I or ase II or both exist after ATP-indued disruption. From Fig. 2, it is lear that the doublets have moved toward the tip relative to the position of the axoneme; in partiular, doublet Z has moved tipward relative to Y, and doublet Y has moved tipward relative to doublet X. To determine the diretion of fore generation between any two doublets by determining the diretion of relative movement between the doublets, we must also be able to distinguish the positions of the A and B subfibers of eah doublet. Due to superimposition and twisting, in Fig. 2 the positions of subfibers of the displaed pairs of doublets are not lear, and the diretion of fore prodution BASE FIG. 7. Diagram of the two possible diretions of relative movement between a pair of doublets, depending on whether the diretion of fore generation of dynein is toward the tip (ase I) or toward the base (ase II). The tip and base are defined by the spoke arrangement. an not be identified with ertainty. Also, Fig. 2 shows three interating doublets, whih ompliates our simple situation. However, both interations in this image are onsistent with ase I of Fig. 7 and also with a onstant polarity of sliding. In support of this interpretation, doublet Y -apparently hanges position in the two lateral interations with the two adjaent tubules (X and Z), and the spokes of doublet Y, and therefore its underlying dynein arms, are to the left. Other images an be used for unequivoal analysis. In Fig. ia, the A and B subfibers an be identified in eah doublet. The spokes mark the position of subfiber A and overlie the dynein arms. Subfiber A of the left doublet is to the outside and its arms are free, while the right doublet ontains the interating arms. The tip diretion an be identified by the A subfiber extension of the right doublet and the triplet spoke groups of both doublets (seleted spoke groups are braketed). This is supported by other markers from the same axoneme (not shown). By these riteria, the tip diretion is to the top of the image, and, therefore, the left doublet has moved toward the tip relative to the right doublet. This is an example of the polarity of fore generation desribed by ase I in Fig. 7, where the dynein arms of the doublet R generate fore from base to tip. Doublet L is then pushed toward the tip while R moves relatively baseward. In every pair of doublets we have examined in whih the longitudinal displaement and the fine struture markers are unequivoal, only ase I has been found,-this is true whether the ATP-indued sliding results are aptured by the fixation and ritial point proedure on a polylysine surfae or are ob- I
Cell Biology: MYOSIN THICK FILAMENT DYNEIN ON \\ \\ CILIARY DOUBLET Sale and Satir ADZ FIG. 8. Diagram of analogous polarity relationships of myosin and dynein. served in negative stain without fixation and without adherene to polylysine (Fig. 3b). Although axonemes are more disrupted in negative stain, and length measurements of individual doublets are not usually exat, the images onstrued as sliding onfigurations in negative stain agree ompletely with the ritial point results. Negative stain has the advantage that the dynein arms themselves are readily visible (Figs. 3b and 4). Strutural polarity of dynein Beause apparently there is a single diretion of dynein fore generation, the question arises: Is there a strutural polarity of the dynein arms to whih fore generation is related? Only reently have good images of dynein arms in negative stain beome available (15, 16). Notably in their study of the diretionality of mirotubule growth, Allen and Borisy (15) determined that the dynein arms of Chlamydomonas flagella have strutural polarity suh that, when the arms are viewed sitting along the edges of the A subfiber, they invariably point toward the distal end of the axoneme (plate 1 in ref. 15). We find that, in iliary axonemes of Tetrahymena, on the edges of the A subfiber, the dynein arms, viewed by negative stain, predominantly point basally. At this time we have no explanation for the differene in the strutural polarity between Chlamydomonas and Tetrahymena. Fig. 4a shows a portion of a splayed axoneme in whih the tip and base diretions an be defined by the spoke group organization of the right doublet. In Fig. 4, seleted dynein arms are marked with lines to show the baseward diretion of tilt. The idential tilt an be seen in intat axonemal ylinders both before and after trypsin digestion. Although there are two rows of dynein arms that may be superimposed in some images, wherever a single row is learly seen for either the inner or outer arm the tilt is the same. Therefore, in Tetrahymena, the tilt of the dynein arms provides an additional marker of base and tip. In Fig. 3b, for example, the tilt of the dynein arms of the right doublet shows that the base of the axoneme lies toward the bottom of the page. The subfibers in the overlap zone in Fig. 3b an be identified by the origin of the dynein arms. Suh images always fit ase I of Fig. 7, whih supports our finding that dynein fore generation is in a single diretion from base to tip. It is important to note that, in images suh as Fig. 4b, the tilted form of the unattahed dynein arms is hanged to a somewhat flattened form in the presumed attahment region. This information is preliminary and further work is needed to eliminate the question of artifat. Speulatively, these forms may represent the onformational hanges of dynein in the mehanohemial yle, but evidently suh images await further detailed analysis. In agreement with the funtional polarity of fore generation, the free dynein arms are, hypothetially, in a position prepared to move the adjaent doublet toward the tip when ativated (Fig. 8). This sheme is analogous to the events failitating the sliding of filaments in striated musle: the myosin head sits free, presumably pointing J Pro. Nati. Aad. Si. USA 74 (1977) 2049 away from the rigor position, prepared to advane the adjaent thin filament toward the enter, or M-line, of the saromere. Relation to iliary beat Cilia and eukaryoti flagella move by omplex bend formation and propagation, the bends produed by the restrition and ontrol of sliding and propagated probably by a mehanial feedbak system within the axoneme. If it is assumed that the fores involved in produing sliding during ATP-indued disruption of trypsin-treated axonemes are the same as those involved in normal beat, our results strongly suggest that eah doublet around the axoneme moves with a single polarity, its dynein arms produing fore from base to tip only. This would immediately imply that in produing bends all doublets do not generate the same amount of fore and slide atively at the same time. We must note ertain reservations in our data; in partiular, we have been unable to measure hains of telesoping doublets more than three or four doublets long. It ould be that we have inadvertently seleted a subset of dynein interations under our onditions of treatment. We think this unlikely. If the polarity of fore generation is onstant around the axoneme, then the dynein on doublets on one side of the axoneme must produe more fore in the effetive bend, while the doublets on the other side produe less or no fore and respond passively. In the reovery bend, the proess will be reversed. Surprisingly, sine doublets 5-6 lead during the effetive stroke, and hene move more tipward, our data are onsistent with the suggestion that it is doublets on the no. 1 side of the axoneme that are more ative during the effetive stroke. The unanswered question is: During normal iliary beat, what are the devies that ontrol ativation of ertain doublets and not others? This work was supported by U.S. Publi Health Servie Grants HL13849 and GM1021 and a Miller Professorship to P.S. The osts of publiation of this artile were defrayed in part by the payment of page harges from funds made available to support the researh whih is the subjet of the artile. This artile must therefore be hereby marked "advertisement" in aordane with 18 U. S. C. 1734 solely to indiate this fat. 1. Satir, P. (1968) J. Cell Biol. 39, 77-94. 2. Summers, K. E. & Gibbons, I. R. (1971) Pro. Natl. Aad. Si. USA 68, 3092-3096. 3. Satir, P. (1974) in Cilia and Flagella, ed. Sleigh, M. A. (Aademi Press, New York), pp. 131-142. 4. Goldman, R., Pollard, T. & Rosenbaum, J., eds. (1976) Cell Motility Book C. Mirotubules and Related Proteins (Cold Spring Harbor Laboratory, Cold Spring Harbor, NY), esp set. 8, pp. 841-932. 5. Warner, F. D. & Satir, P. (1974) J. Cell Biol. 63,35-63. 6. Gibbons, B. H. & Gibbons, I. R. (1972) J. Cell Biol. 54,75-97. 7. Summers, K. E. & Gibbons, I. R. (1973) J. Cell Biol. 58, 618-629. 8. Satir, B., Sale, W. S. & Satir, P. (1976) Exp. Cell Res. 97, 83-91. 9. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 10. Mazia, D., Sale, W. S. & Shatten, J. (1974) J. Cell Biol. 66, 198-200. 11. Sale, W. S. & Satir, P. (1976) J. Cell Biol. 71, 589-605. 12. Zobel, C. R. (1973) J. Cell Biol. 59, 573-594. 13. Gibbons, B. H. & Gibbons, I. R. (1974) J. Cell Biol. 63, 970-985. 14. Sale, W. S. & Satir, P. (1977) Cell Biology International Reports. 1, 45-49. 15. Allen, C. & Borisy, G. G. (1974) J. Mol. Biol. 90,381-402. 16. Warner, F. D. (1976)J. Cell Si. 20, 101-114.