Significance for Endosymbiosis Theory

7reb. Soc. C'a'. Biol., Vol. 41 (1990) 5-16

KINETOSOME-CENTRIOLAR DNA: SIGNIFICANCE FOR ENDOSYMBIOSIS THEORY

LYNN MARGULIS' and MARK MCMENAMIN2

'Department of Botany, University of Massachussetts. Amherst, USA. 'Department of Geology and Geography, Mt. Holyoke College. South Hadley, USA.

SUMMARY

The "extreme version" of the serial endosymbiosis theory (SET) postulates three types of integration of bacterial symbionts into what became the nucleocytoplasm of nucleated cells. The first, a symbiosis between motile bacteria and less motile archaeobacterial hosts is thought to have resulted in the origin of nucleocytoplasm capable of both internal movement and motility by undulipodia. That is, microtubule-based motility systems of eukaryotic cells (kinetosome- centrioles, undulipodia including cilia, mitotic spindles, etc.) are hypothesized to have evolved from symbiotic associations of spirochetes with Therinoplasma-like archaeobacteria (protonu- cleocytoplasm). The second, a symbiosis between aerobically respiring bacteria and the nucleocytoplasm, resulted in mitochondria of aerobic eukaryotes. The third association led to photosynthetic plastids from undigested cyanobacteria. Whereas the symbiotic origins of plastids and mitochondria are firmly established, the recent discovery by David LUCK and his colleagues of centriole-kinetosome DNA greatly enhances the likelihood that the last postulate concerning motility is valid. We can thus anticipate a rapid, definitive test of the "extreme version" of the SET. Unknown to most Western scientists, the hypothesis of the origin of eukaryotic cell motility from symbiotic bacteria has a Russian antecedent: the concept of symbiogenesis or origin of evolutionary novelty via hereditary symbiosis was well developed by Russian biologists late in the last century and early in this one. Indeed, Boris M. Kozo-POLYANSKI, who wrote only in Russian, suggested that cilia derived from 'flagellated cytodes", by which he meant motile bacteria. New results from molecular biology lend strong support to the SET, a theory which forms the foundation of the concept of symbiogenesis. Both the SET (which explains the genesis of the first eukaryotic cells) and symbiogenesis (a general evolutionary process involving heritable transmis- sion of symbiotically acquired characteristics) require that all eukaryotic organisms be viewed, at the cellular level, as complex microbial communities. 6 L)':NN.t1.-IRG(LI.S.LVI) .t1.IRA t1ct1L;AAMIA

Personal recollections ses was named Serial Endosvrrtbiosis Theory (SET) by TAYLOR (1974). F. J. R. Impressed with the frequency of cases of (Max) TAYLOR, Canadian marine biologist nonmendelian heredity in a world of nu- and expert on dinomastigotes, attempted clear inheritance (e. g., photosynthesis mu- -with limited success- to develop the de- tants of plants and algae, "petites" in tails of the contrasting nonsymbiotic ori- yeast, cortical inheritance in Paramecium, gin of eukaryotes. As intellectual exercise etc.), as a genetics student at the Universi- he described the origin of organelles by ties of Wisconsin (1957-60) and Califor- "direct filiation" or autogenesis (TAYLOR, nia, Berkeley (1960-65), I (L.M.), explored 1976). He also distinguished different ver- the early literature for clues for the expla- sions of the views presented here: the xe- nation of cytoplasmic heredity. That there nogenous view of organelle origin, i.e., the were no "naked genes" in eukaryotic cells serial endosymbiosis idea. He collected became clear, and evidence for the presen- data concerning the possible origins of mi- ce of bacterial genetic systems inside them tochondria (from respiring bacteria) and became equally obvious. On the basis of plastids (from cyanobacteria). The concept literature reviews of cytoplasmic heredity of the origin of plastids (but not mitochon- and my consequent predictions of organe- dria) or the origin of both plastids and llar DNA, I wrote the statement of the mitochondria by symbiosis was labeled the origin of nucleated ("mitosing") cells from "mild version" of the SET. TAYLOR called bacterial symbiotic associations in 1965. It my theory the "extreme version" of the was first published by James DANIELL], SET for I insisted that the kinetosomes, co-originator of the lipoprotein bilayer cilia and related microtubule organelles membrane theory, in the Journal oJ'Theo- (Fig. 1) were also of symbiotic origin from retical Biology after approximately 15 re- motile bacteria. jections of the manuscript (MARGULIS, un- As far as I knew at the time, mine was an der the previous name of SAGAN, 1967). entirely original contribution. In the 60's The expanded version of the theory of the cell symbiosis ideas were to some ex- origin of eukaryotic cells organelles (mito- tent still disreputable. Some scientists, like chondria, plastids, centrioles in nucleocy- Hans (who was my professor at the toplasm) was developed into a book origin- University of Wisconsin, Madison), knew ally under contract, and then rejected, well E.B. WILSON'S (1928) masterpiece by Academic Press. Eventually published by The Cell in Development and heredity, and Yale University Press (MARGULIS, 1970), saw to it that articles written describing the theory itself was supported during the the discovery of DNA outside the cell nu- decades of the 70's and 80's by many new cleus in the early 1970's mentioned ideas results derived from molecular biological, of organellar origin by symbiosis. Al- genetic and ultrastructural studies. The re- though some authors did cite historical vised version of the work became the mo- views of cell symbiosis, and noted ideas of nograph Symbiosis in Cell Evolution the origin of mitochondria and plastids as (MARGULIS, 1981). Because of the impor- presented by American (e. g.: WALLIN, tance of the discovery of kinetosome/ 1927) and French authors (e. g.: PORTER, centriole DNA described here, a second 1918), these scientists writing in the 1960's edition of the 1981 book is scheduled for and 1970's were extremely skeptical of he- publication in 1992. reditary endosymbiosis (Chapter 2 in The theory of the derivation of nuclea- MARGULIS, 1981). Indeed TAYLOR and I ted cells from a series of bacterial symbio- were conscious of both prejudice against KINETOSOME-CENTRIOLAR DNA: SIGNIEIC.ANCE FOR ENDOSYMBIOSIS THEORY 7

KINETOSOME CENTRIOLE Microtubule organelles Figure I. Microtubule-based organelles of eukaryotes from heterokont cells of algae and other protoctists, ciliated epithelium of animals, centrioles of animal and protoctist cells and mitotic spindles. If a shaft (axoneme) is present the basal structure is called a kinetosome, if absent it is a centriole. and ignorance of cell symbiosis theories by sequence data (Fox, 1985), it has become "mainstream" experimental scientists. clear that the acquisition of centriole/ Yet, neither of us had knowledge of our kinetosome/undulipodia preceded that of Eastern European predecessors: the mitochondria and that mitochondria, deri- "symbiogeneticists" such as MERESCH- ved from several types of respiring bacte- KOVSKII and Kozo-POLYANSKI in Russia ria, are most likely polyphyletic within the who, from the late 19th century until the protoctista (Fig. 2). 1950's (Kozo-POLYANSKI died in 1957) The recent spectacular discovery of ki- developed detailed proposals for the origin netosomal-centriolar DNA, reported re- of evolutionary novelty, including cell or- cently from the laboratory of David LUCK ganelles, by establishment of hereditary at Rockefeller University (HALL et symbiosis. al., 1989) clarifies and refines the experimental evidence favoring the "extreme SET", and presents a severe challenge for Undulipodia before mitochondria competing hypotheses. Our purpose here is The original statement of the SET, on to place this work, achieved by genetic, the basis of a preconception of monophyly cytological and molecular biological stu- of mitochondria, hypothesized the acquisi- dies in the chlorophyte protist Chlamydo- tion of these organelles prior to that of monas, in its appropriate historical con- centriole/kinetosome/undulipodia. Becau- text. The genetics of the motility system in se of the explosion of information on the Chlamydomonas needs to be better eukaryotic microorganisms (protists and known; we hope to integrate these arcane other protoctists, including so many lack- genetic discoveries with the history of cell ing mitochondria altogether, MARGULIS evolution concepts both in the USSR and cl al., 1990) and from ribosomal RNA the West. S L1"h'N .11.4KG ('l./.5 .4^'/) ,M1f.^IRl^^ A1r .11F,^'.4AfLti' S.E.T. Serial Endosymbiosis Theory

PROTOCTISTS ANIMALS FUNGI

o""1""/ ,,^,.,",^...n`

Figure 2. Serial endosymbiosis theory: diagram of the extreme version of eukaryotic cells origins with the order of organellar acquisition superimposed.

Kinetosomal DNA Chlaml^domonas smithii wildtypes. In shit hybridization using cloned segments from Recently, the presence of a large quan- two regions of the uni DNA as probes loca- tity of DNA, 6-8 megabases, in the centrio- lized, with fluorescence microscopy, the le/kinetosomes of Chlamvdomonas rein- uni DNA to the two kinetosomes. In hardtii has been reported (HALL et Chlair^t^dornonas undulipodial motility is a/., 1989). Taking advantage of the fact lost during mitosis; these kinetosomes al- that in Chlarnvdartonas rc^inhar^dtii mu- ter their morphology during the process of tants affecting kinetosomal assembly and cell division and become the centrioles at undulipodia formation are clustered on a the mitotic poles. By using known DNA single linkage group, LUCK and his collea- standard chromosomes from yeast and gues first showed that the position of these Neurospora, HALL et a/. (I 989) estimate motility markers follows a circular pattern. the size of the kinetosome/centriolar DNA Using pulsefield gel electrophoresis to re- to be very large, about 6 megabases. solve chromosome-sized DNA, they were Although the importance of this molecu- able to identify the DNA corresponding lar analysis is evident, it is also worthy to the uni (also called "chromosone XIX") note that geneticists and cytologists had linkage group. Restriction fragment length anticipated this discovery. For some time, polymorphisms detected with subclones of it has been observed that centrioles repli- this DNA showed the expected 2:2 segre- cate in the cells (Fig. 3). They underlie and gation patterns in crosses of C i^ernhardtii are required for the development of all bearing linked mutant markers with undulipodia (i.e., cilia and eukaryotic "fla- h7NETOSO.'L1L'-CENTRlOL.4R DNA: S/GNIFIC'ANC'6 FOR ENDOS}'A/B/OS/S THEOR }" 9

.^. ^^^t• ^^^+.'^° d^^^ ^^^^ ^ ^ .^^a m'I^. . <:nv^/^k h^ a .y®^ "^`^

KINETOSOME REPRODUCTION Figure 3. Kinetosome reproduction diagram based on electron micrographs of Paramecium. The [9(3)+0] transverse section is shown above and the longitudinal section below. gella"). In many organisms , such as the protist Lophomonas ( Fig. 4), "centrioles" and centromeres (= kinetosomes) can be seen to grow directly from undulipodial bases. When a shaft or axoneme is present they are called kinetosomes ( or inadvi- sably basal bodies ; MARGULIS and SAGAN, 1985), in the absence of an axoneme they arc centrioles . centrioles , as [9(3 )+ O] mi- crotubular structures , often ( i.e., in most animals ), but not necessarily ( i.e., never in plants ), lie at each pole of the mitotic spindle . centrioles have been called the "central enigma of cell biology" (WHe- ATLEti , 1985). Although often called flagella, all axoneme - bearing eukaryotic structures are entirely different from the flagella of bacteria . The term " flagellum" should therefore be avoided and undulipodium used . Figure 5 shows a comparative diagram of both the eukaryotic undulipodium and bacterial flagellum. Especially in ciliate protists, kinetosomes and the pattern they make to form the cortex ( outer 1µm or so of the surface) are implicated in a weird inheritance system. [n paramecia certain cortical mutants ("siamese-twin" or double-bodied cells) Figure 4. The connection between the kinetosomes were induced using antisera and surgical and the mitotic spindle is clear in this drawing of the of termites, Lophomonas. Ba- needles by extremely skillful investigators parabasalid symbiont sed on electron micrographs of Andre Hollande, one (BF1SSt ) N and SONNEBORN , 1965). DOU- can see the nuclear membrane-embedded kinetocho- bled-bodiness was inherited through hun- res and extranudear spindle attached to kinetosomes dreds of generations . Genetic studies com- by striated fibers. lU L1'.A'.1'.t/.^IR (iCLLS_ i^'D SI.-iRh^.tlrSlti .1'A.tf/.^"

UNDULIPODIUM FLAGELLUM

f {

9(2)^2^ 4;'. J ^ ^ m,crotobok I B-tubule _ai rGi ^^ .,.. _ radial apakt profeinr CuDUIin Q,Q ^ 1 `^ cenfra/ ahealh yaltina axonemai d, flagellin proteins 1 ^ :^, 250 .^ .i, ^. '^"..^-

Pia:ma brnne ^^^^,^^ 1 ^^^ II \ 350 30

5 flagellar KINETOSOME basal structure

Figure 5. Comparison between the intracellular undulipodium and its kinetosomal base, that contain well over 500 proteins, with the much smaller and less complex extracellular bacterial flagellum, that contains fewer than 40 proteins. pletely ruled out the possibility of nuclear the perfect haploid organism, very familiar gene control. Furthermore, because mito- to geneticists, in which to have sought chondria) inheritance patterns of parame- kDNA. During each mitotic division kine- cia also can be manipulated by resear- tosomes resorb their axonemes and each chers, mitochondria) DNA control was naked kinetosome becomes a mitotic cen- excluded as the explanation for cytoplas- triole in Chlarnrdornonas, therefore unlike mic inheritance of the doublet form. Other in many other protoctists and anitnals, morphological mutants of paramecia each centriole/kinetosome is capable of ("swimmer", "snaky", etc.) also proved genetic continuity. We do not expect all to be cytoplasmically -but not mitochon- mature kinetosomes to contain DNA; in- drially- inherited. Cortical inheritance is deed those not capable of reproduction not under nuclear gene control. In spite of probably lack DNA at all times (Ycnlrv^etz the fabulous quality of ciliate cortical ge- et al., 1972). The use of Luck's DNA pro- netics these remarkable cases of the inheri- bes, highlighted with a fluorescent label, tance of acquired characteristics are unap- allowed the visualization of the two tiny preciated in vade mecum-type, enormous kinetosomes (each approximately 0.2^ µm genetics textbooks. in diameter) per Chlamt^domonas cell. As discussed in Microcosmos (MARGU- They were highlighted on the cover of the us and SAGAN, 1986b), the discovery of journal Cell (HALL et al., 1989). Such pro- kinetosomal DNA (kDNA) -or if not bes must now be used in other centriole- DNA, at least some remnant nucleic acid kinetosome containing organisms. system- was anticipated. We are delighted If this centriolar-kinetosomal DNA that Luca and his colleagues performed turns out to be the "spirochetal secret such elegant experiments. Chlamydorrro- agents" predicted by M.xkc;tl^ls and S.^- nas, the green protist with two kinetoso- c^^tv (1986a), then all microtubule-making mes per cell, well-marked nucleus, single eukaryotes must have some DNA homolo- chloroplast and controllable life cycle is gous to Chlamvdomonas kDNA within KIA'EI'OSO,i1E-CENiRIOLAR I)A I .SYGAHAAACE FOR LADOS}'tIBIOSIS THEORY 1 I their cells even if, at certain times, the DNA takes up nuclear residence. If the intracellular motility system of all animal and plant cells evolved from symbiotic bacteria, such as members of the genu,, Spirochaeta (BERMUDES et al., 1990), it iti logical that a remnant bacterial DNA i,, detectable in all descendants of this event. even if it is 1500 million years after the symbioses were established (MARGULIS. 1981; BERMUDES et al., 1987a). A strong case can be made that the evolution of the nucleated cell was a consequence of integration of motile spirochete symbionts which contained microtubules like that in Plate I a. Such symbiotic spirochetes with lithe fast-swimming habits conferred rapid motility upon their sluggish partners. Spi- rochete-host genetic integration including membrane proliferation-initiated eukaryo- of mi- sis: evolutionary steps in the origin Plate Ia. Cytoplasmic tubules, approximately 24 nm tosis and eventually meiotic sex (M.ARGU- in diameter, found by A. Hollande and I. Gharazo- LIS and SAGAN, 1986b). Epibiotic attached glou in the protoplasmic cylinder of Piltotina spiro- spirochetes and other bacteria are known chetes (mu: microtubules; in. cytoplasmic membrane; ri. ribosomes; n: nuclear region). today; in some cases these bacterial symbionts are difficult to distinguish from undulipodia (Plate lb). centers were likely to be products of intracellular symbiosis (Kozo-POLYANSKI, 1924). They were intracellular structures, Time of origin of centriole - kinetosomes "organoids " [ organelles ], derived from "flagellated cytodes", by which he meant Frond-like organic films occur on bed- motile bacteria. ding planes from the Belt Supergroup, Montana. These compressions are remains of algal thalli, possibly chlorophytes or Symbiogenesis phaeophytes (STROTHER, 1989). Thus 1300-1500 million years is a minimum age Symbiogenesis, the evolution of novelty estimate for the origin of eukaryotes at the by integration of partners belonging to dif- cent riolar-k1netosomal grade of evolution, ferent taxa which remain in protracted i.e., protoctista (M.ARGULIS et al., 1990). physical association, had been a principle The Russian biologist Boris Michailo- of evolution -at least in Russia- since the vich KOZO-POLYANSKI (1890-1957) did late 19th century. Symbiogenesis, a term not consider centriolar origin an enigma. coined by Konstantin Sergeivich ME- To him, the centrioles -or their manifesta- RESCHKOVSKII (1865-1921), explained the tions- could actually be visualized in liv- presence of greenish photosynthetic units ing cells or iron-hematoxylin stained pre- in heterotrophs as diverse as hydra, dia- parations: blepharoplasts and other cell toms and plant cells. MERESCHKOVSKII 1 2 L KYN AtI IRGt 7.1.1' .-1,17) MARK AhMENAMIN

Plate lb. The tendency to establish microbial symbionts is apparent in this photograph of staurojoenina, a parabasalid protist containing both undulipodia (u) and unidentified surface rod-shaped symbiotic bacteria (h).

(University of Kazan) and his theory of only are they in animals, but, apparently, two plasms, together with his senior collea- also in plants. gue from St. Petersburg, Andrei Sergeivich Many consider it demonstrated that ble- FAMINTZYN (1835-1916), was the most pharoplasts [kinetosomes] manifest them- successful articulator of the theory of chlo- selves as a variety of centrosomes (or cen- roplast origins as specific example of his trioles): the first may he transformed into general principle of symbiogenesis. These the second and vice versa as seen microsco- keen Russian biologists suggested -in con- pically in live cells. From the views stated temporary terms- that plastids originated here, the divisions of cells are synchroni- as captive cyanobacterial symbionts in het- zed... with the divisions of blepharoplasts erotrophic cells. Although there was dialo- (in the role of centrosomes or centrioles) gue, mutual criticism and disagreement -i.e., the motile flagella (orflagella-produ- between these professors of biology, both cing, or flagellated organelles or partners of down-played natural selection and both the cell)... At least the suspicion of the bac- thought symbiosis to be crucial as an evo- terial nature of these kinetoplasmatic fmo- lutionary mechanism (KHAKHINA, 1979). tilitvJ organelles without question appears But it is KOZO-POLYANSKI who, in entirety legitimate. DYSON's terms, is our most "illustrious (Kozo-POLYANSKI, 1924; pp. 56-57). predecessor" in claiming the symbiotic origin of centrioles (DYSON, 1987). The Furthermore, in his book New Concepts position of Kozo-POLYANSKI on eukaryo- of Biology, unlike his own illustrious pre- tic cell motility was clear: decessors MERESCHKOVSKII and F.A- MINTZYN, Kozo-POLY.ANSKI did not reject Bodies known as blepharoplasts, immersed Charles Darwin. According to Liya Nico- in the plasma of the cell and bearing a laevna KHAKHINA, Soviet historian of ,flagellum fundulipodiumJ or several fagel- science in Leningrad who researches his- la, come out from the cell interior. Blepha- tory of symbiosis theories in evolution, roplasts occur in mastigotes, flagellated KOZO-POLYANSKI recognized that natural cells of sponges and also spermatozoa; not selection plays a crucial part in symbiont K1NETOSOME-CENTRIOI IR DNA: SIGNIFICANCE. FOR L'NDOSYMBIOSIS THEORY 13

integration : if symbiosis is "author", natu- ce of interaction with the Russian symbio- ral selection is still " editor". Admitting geneticists. WALLIN was heartily disdain- natural selection , Kozo-POLYANSKi deed; his last article in 1965 was rejected by scribed the power of symbiogenesis using the journal Science (MEHOS, 1983). Al- the term with the same original meaning of though Paul PORTIER (1918, in France), its inventor. In her book (1979) KHAKHI- Umberto PIERANTONI (1948, in Italy), and NA Cites MERESCHKOVSKII: Paul BUCHNER (1965, in Germany) were all sympathetic to various degrees of "I named this process symbiogenesis, which symbiosis as a mechanism for the genera- means: the origin of organisms by means of tion of evolutionary innovation, these au- combination or conjunction of two or more thors were not fully "mainstream" scien- beings, joined in symbiosis " (Meresehkovs- tists. The Russians, however, held impor- kii, 1920). Si'mbiogenesis was considered tant positions: MERESCHKOVSKII was pro- by him as an evolutionary principle, which fessor at Kazan University, to Moscow the permitted a new approach toward resolu- second most important university in Rus- tion of the question of the origin of orga- sia, and FAMINTZYN was director of the St. nisms. Working from his foundation, Afe- Petersburg (Leningrad) Plant Physiology reschkovskii formulated evolutionary con- Institute. Nevertheless, symbiogenesis was cepts which were called the "theory of virtually unknown outside the Russian- symhioi^enesis" (Alereschkovskii, 1909). speaking world where the concept, if not "These factual materials are based only on ignored, was labeled "controversial" or ri- the accumulation of new facts from cyto- diculed (e.g.: LIJMIERE, 1919). logy, from biochemistry , and from physio- For the serial endosymbiosis theory logy, principally from lower organisms, the (SET) of cell origins to be definitively pro- aim of these attempts being to raise anew the ven, Luck's centriolar DNA must be found veil obscuring secrets of the origin of orga- generally in eukaryotes. Because it follows nisms. I have resolved to make such at- the distribution of microtubule organizing tempts, and my work is the original... juxta- centers, this motility-associated DNA may position of previous statements of new theo- be universal in meiotic organisms (e.g.: ries of the origin of organisms, in which the fungi, animals and plants), and nearly uni- phenomenon of symbiosis apparently plays a versal in protoctista, excepting those which prominent role; I propose to call the theory lack microtubules, centrioles and kinetoso- symbiogenesis " (91creschkovskii, 1909). mes (MARGULIS and SAGAN, 1986a, MAR- (KHAKHINA, 1979; p. 53) (UUS et al., 1990). If DNA homologies are established between the appropriate spiro- But the language barrier was definitive: chetal ancestors and these eukaryotes, then although works of both MERESC:HKOVSKII symbiogenesis must be considered anew. and F.AMINTZYN were cited by German The inescapable inference is that animals, and English speakers (WILSON, 1928), Ko- fungi and most protoctists had at least zO-POLv .ANSKI's contribution is virtually three microbial ancestors and all plants unknown outside the USSR. Fascinated, and algae had at least four. Evolutionary we note that the American anatomist innovation for all eukaryotes involved far (University of Colorado, Denver Medical more than accumulation of mutations: it School) Ivan E. WALLIN (1883-1969) deve- required integration of heterologous geno- loped svmbionticism, an extremely similar mes. If all animal cells have at least three theory of the role of symbiosis in specia- ancestors and all plant cells have at least tion and cell organelle origin in the absen- four, how many heterologous ancestors has 14 L }"NN ,M11. JRGL'LIS .^iND 1/,1 HA

Leognathid fishes. But Luck's revelation of our spirochetal secret agents (if indeed ki- ^> netosomal-centriolar DNA's are present and homologous) uncovers the "anima" in all of us (MARGULIS and SAGAN, 1986a). No longer will the type of statement: "al/ ,:^m^^¢. lichens are symbioses between algae and «^^,^^^,^o^= czm¢oc rrngi" be limited to lichens. Future cell T¢i05i5 4 biology texts must begin with a description plastids of how all eukaryotic cells are derived from co-evolved communities of symbio- PROTOCTISTA, kg- tic bacteria. Future general biology text- ^^-^ books, in addition to explaining why "protozoa" (as "sing/c^-ce//ed anifnals') is no ^^ 3 longer a valid taxon, must laud the enor- mitoehondria mous success of the phototrophic, oxygen- undulipodia ^ is bacterial endos ymbionts that color our nucleocytoplasm planet green. 1 In the seminal discovery P^^^o=r„^^^=^s of kinetosomal/ centriolar DNA by Hnt.t, et al., we see the potential of the powerful techniques of r^ molecular biology to ferret out evolution- c- ary relationships between our bacterial an- C, MONERA cestors. The strength of their contribution is the brilliant Figure b. Serial endosymbiosis theory showing the application of these techni- origin of nucleated cells by bacterial symbioses: dia- ques to fundamental biological problems gram emphasizing the polyphyly of the microbial an- extending back more than a century and cestors. across several language barriers. Note added in proof. The term undulipodia a man, a cow or a weeping willow tree'? has a long history of usage in the Russian Not only will the concept of "individual" literature, but the origins of the term are be replaced with that of "symbiont" for all obscure and the scientist who originated it animals, but since all eukaryotes harbor is unknown to us. The following passage, heterologous DNA's from various sources excerpted with condensation from L.N. both the sciences of eukaryotic evolution SERAVIN (pp. 10-11, 1967, Advanced and of developmental biology transform: svsterns ^^/^ Protozoa. Structure, mechano- they become very special cases of applied chemistrv and physiology. USSR Academy microbial community ecology (Fig. 6). of Sciences, Scientific Council on Prob- Svmbiogenesis as an evolutionary prin- lems of Cytology. "Science" Publishing ciple is under reconsideration (MAYNARD House, Leningrad Division, Leningrad; SMITH, 1989; LAW, 1989; MARGULIS and terms in square brackets are ours) demons- FesTeK, 1991). Of course, that new orga- trates how a Russian biologist in the nisms evolve by symbiont integration is 1960's understood the term undulipodia to not entirely new to the English speaking refer both to cilia and to eukaryotic flagel- world: it is the explanation of choice for la: the origin of lichen structure, termite wood Cilia and (eukarvoticJ /lagella are thread- digestion, and the luminous organs in or filament-/i^r attachments to the surfaces A7,%'EIOSO.iIE-(Yi.V7RIOLAR DA'A: S/GN'IEICA.tiY'E I"OR F„VDOSY.NBIOS/S TI/FORY 15 of protozoan [protoctist] cells. These orga- chusetts and Mt. Holyoke College. L.M. is gra- nelles are capable of rhythmic motion, use- teful to Jennifer Margulis for expert Russian exchange ful either for propelling the cell or for stiring translation aid and to the scientist Office of International Affairs surrounding fluid. program of the the of the U.S. National Academy of Sciences, to elongated objects are numerous on if these Soviet colleagues and to the Soviet Academy of short, they are called the cell surface and Sciences for the opportunity to learn about the cilia; if they are long and few in number history of symbiogenesis in a visit to Leningrad they are called flagella. The activity of cilia and Moscow in April, 1989. M.M. acknowled- of a single cell is coordinated, whereas ges support from the National Science Founda- ,flagella of a single cell function relatively tion (EAR 8857995). independently. The distinctions, however, between these two types are not clear cut and there are a series ofgradations between the two types. Electron microscope studies BIBLIOGRAPHY demonstrate that both organelles have a shared structure, and a single concept can BEISSON, J. and SONNEBORN, T.M. (1965) Cytoplas- inheritance of the organization of the cell cor- denote both types. A.P. Shma- mic be utilized to tex in Paramecium aurelia. Proceedings ofNational gina (1948) calls them undulipodia (unda- Academy ofSciences 53:275-282. wave, podos-foot), the organelles of motion BERMUDES, D., MARGULIS, L. and TZERTZINIS, G. of the protozoans (protoctistsJ. (1987a) Prokaryotic origins of undulipodia: Appli- Undulipodia consist of two parts -the part cation of the panda principle to the centriole enig- York Academy Sciences as "proper" ma. Annals of the New of outside of the cell (referred to 503:187-197. undulipodia) and the basal body (the part of BERMUDES, D., FRACEK, S.P. Jr., LAURSEN, R.A., the undulipodium within the cell). In some MARGULIS, L., OBAR R. and TZERTZINIS, G. organisms fibrous structures attach to the (1987b) Tubulinlike protein from Spirochaeta baja- Annals the New York Academy of basal body. californiensis. of Sciences 503:515-527. Undulipodia length varies greatly in differ- BERMUDES, D.. TZERTZINIS, G.. OBAR, R. and Bosco, ent groups (ranging from 3 microns to 150 G. (1990) Localization and immunological charac- microns), hut undulipodia length has an teristics of tubulin-like protein in Spirochaeta. Ar- approximately constant value for all chives of-tlicrobiology (submitted). P. (1965) Endosvmhiosis of.lniinals with undulipodia BuCHNER, lengths. The distal section of Plant Microorganisms. Interscience Publishers called achronemes is often observed to nar- (John Wiley Sons, Inc.), New York. row distally, however. DYSON, F. (1985) Origins ofLife. Cambridge Univer- (SMAGINA, A.P. 1948. Ciliarv Movement. sity Press, Cambridge England. Fox, G. (1985) Insights into the phylogenetic posi- State Publishing House for Medical Litera- tions of photosynthetic bacteria obtained from 5S ture "MEDGIZ", Moscow.) rRNA and 16S rRNA sequence data. In: D. Sagan (ed). The Global Sulfur C'yclc. NASA Technical Memorandum 87570, Washington, DC. Acknowledgements HALL, J.L., RAMANIS, Z. and Lu

Lt INHERE, A. (1919) Le Myth des Svinhiotes. Masson Typo-lithography of the Imperial University, Pro- et Cie., Paris. ceedings of Studies of the Imperial Kazan Univer- MARGuus, L. (1970) Origin ofEukarvotic Cells. Yale sity, Kazan (In Russian). University Press, New Haven, CT. MERE SCHKOVSKIL K.S. (= MEREJKOVSKY , C.) (1920) MARGULis, L. (1981) Symbiosis in Cell Elvolution. La plante consideree comme un complexe symbio- W.H. Freeman Publishing Co., San Francisco. tique..Societe des Sciences A'aturelles de l'Ouest de MARGULis, L. and SAGAN, D. (1985) Order amidst la France, Nantes. Bulletin 6:17-98. animalcules: the protoctista kingdom and its undu- PIERANTONI, U. (1948 ) l'rattato di Biologia e Zoolo- lipodiated cells. BioSystems 18:141-147. gia Generale Casa edit . Humus, Naples. MARGULis, L., and SAGAN, D. (1986a) Origins ofSex. PORTIER, P . (1918) Les Symbiotes. Masson et Cie, Pa- Yale University Press, New Haven, CT. (Paper- ris. back, second printing, January 1990). SAGAN, L. (= MARGULis, L.) (1967) On the origin of MARGUUS, L., and SAGAN, D. (1986b) ificrocostnos: mitosing cells. J. Theoretical Biology 14:225-274. Four Billion Years of Evolution From Our Bacterial STROTHER , P.K. (1989 ) Pre-metazoan life. In K. C. ,Ancestors. Summit Books, New York. Allen and D.E.G. Briggs ( eds.). Evolution and the MARGULIS, L., CORLiSS, J.0., MELKONIAN, M. and Fossil Record. Belhaven Press, London . p. 51-72. CHAPMAN, D. (eds.). (1990) Handbook of Protoctis- TAYLOR, F.J.R. (1974) Implications and extensions of ta: The Structure, Cultivation, Habitats, and Life the serial endosymbiosis theory of the origin of Histories of the Eukarvotic Microorganisms and eukaryotes . la-von 23:229-258. Their Descendants Exclusive ofAnimals, Plants and TAYLOR, F.J.R. (1976) Autogenous theories for the Fungi. Jones and Bartlett Publishers, Inc., Bos- origin of eukaryotes . Taxon 25:377-390. ton. WALLIN, I.E. (1927) Symhionncimi and the Origin of' MARGULis, L. and FESTER, R. (eds) (1991). Evolution Species. Williams & Wilkins, Baltimore. and Speciation: Symbiosis as a Source of Evolutio- WHEATLEY, D.N. (1982) The Centriole: A Central nary Innovation. MIT Press, Cambridge. (In Enigma ofCell Biology. Elsevier Biomedical Press, press.). New York. MAYNARD SMITH, J. (1989) Generating novelty by WILSON, E.B. (1928 ) The Cell in Development and symbiosis. Nature 341:284-285. Herediti. 3rd edition. Macmillan , New York. MEHOS, D. (1983) The symbionticism principle of YOUNGER, K.B., BANERJEE , S., KELLEHER , J.K., WINS- Ivan E. Wallin. Master's Thesis, Boston University TON, M. and MARGUUS, L. (1972) Evidence that Graduate School. the synchronized production of new basal bodies is MERES(HKOVSKH, K.S. (1909) The theory of two not associated with DNA synthesis in Stentor coer- plasms as the foundation of symhiogenesis, new uleus. Journal ofCell Science 11:621-637. knowledge concerning the origins of organisms.