TALKING POINT TIBS 24 – MARCH 1999

to a mutually beneficial association), and the increasingly evident mosaic features of eukaryotic genomes8,9, nobody had Metabolic at the origin proposed other symbiosis hypotheses until recently. In 1998, the 30th anniver- of sary of the endosymbiosis proposal, we, and Martin and Müller, independently, published two novel symbiosis hypoth- eses: the hypothesis10 and the Purificación López-García and David Moreira syntrophy hypothesis11. The hypotheses are different but share striking similarities. Thirty years after Margulis revived the endosymbiosis theory for the origin of mitochondria and chloroplasts, two novel symbiosis hypotheses for the Syntrophy and interspecies hydrogen transfer origin of eukaryotes have been put forward. Both propose that eukaryotes The hydrogen hypothesis arose through metabolic symbiosis (syntrophy) between eubacteria and The hydrogen hypothesis proposed methanogenic . They also propose that this was mediated by inter- by Martin and Müller10 states that eu- hydrogen transfer and that, initially, mitochondria were anaerobic. karyotes arose through a symbiotic These hypotheses explain the mosaic character of eukaryotes (i.e. an metabolic association (or syntrophy) in archaeal-like genetic machinery and a eubacterial-like ), as well anaerobic environments between a fer- ␣ as distinct eukaryotic characteristics (which are proposed to be products mentative -proteobacterium that gen- of symbiosis). Combined data from comparative genomics, microbial ecol- erated hydrogen and carbon dioxide as ogy and the fossil record should help to test their validity. waste products, and a strict anaerobic autotrophic archaeon that depended on hydrogen and might have been a OUR ORIGINS HAVE always concerned in the eubacterial branch4. This ex- (Fig. 2). The authors follow us – from the speciation of Homo sapiens plained the archaea– similari- a metabolic top-down approach from the to the origins of life itself. Between these ties but also refined our view of the ori- observation that amitochondriate eukary- events, a crucial midpoint is the origin of gin of eukaryotes: these would share a otes possess eubacterial-like metabolic the eukaryotic cell, the nature of which common (prokaryotic) ancestor with enzymes (in addition to other known eu- is still controversial and elusive. Until the archaea. This idea permeated the scien- bacterial-like genes) and that hydrogeno- mid-1970s, two possibilities were conceiv- tific community rapidly and is now somes (hydrogen-producing organelles able: either eukaryotes were ancestral, widely accepted. present in some anaerobic eukaryotes) and thus their origin was intimately Many protein trees, however, contain share a common ancestry with mitochon- linked to the origin of life itself, or they discrepancies that consistently relate dria. They propose that both symbionts derived from prokaryotes (organisms either archaea (mostly on the basis of the first met in anaerobic environments rich lacking a true nucleus that encloses the genetic machinery) or Gram-negative bac- in hydrogen and carbon dioxide, but genetic material). The evolution of eu- teria (mostly on the basis of metabolism) that soon the host changed its depend- karyotes from simple prokaryotes (then to eukaryotes. Such discrepancies have ence on an exogenous source of these called Monera) is implicit in the phy- led several investigators to propose dif- products, becoming dependent on the logeny proposed by Haeckel in the 19th ferent chimera hypotheses (Fig. 1). Of eubacterium supplying them. The ar- century. However, after the demise of these, fusion and engulfment models are chaeon increased the cell-contact sur- the eukaryote/prokaryote dogma (the mechanistically problematic, although face with the symbiont and ended up commonly held belief that all that is not they can explain the mosaic distribution importing membrane transport systems eukaryote is similar) – which followed of many genes. By contrast, symbiosis and carbohydrate metabolism. Finally, to the recognition (based on rRNA-sequence models rely on intimate relationships over avoid futile cycling of metabolites in its comparison) that there are two distinct extended periods of time that allowed cytoplasm, the host lost its autotrophic phylogenetic lineages within the prokary- symbionts to co-evolve and become de- pathway. The final organism in this evo- otes, eubacteria () and archaea1,2 pendent on each other. Indeed, the first lutionary process is an irreversible – the situation became more complex. detailed symbiosis proposal for the ori- that contains ancestral mito- These lineages appeared to be as differ- gin of eukaryotes, the endosymbiosis chondria and has lost its dependence ent from each other as they were from hypothesis for the origin of plastids and on hydrogen – hence, the need for eukaryotes; yet, later, the use of several mitochondria proposed by Margulis5, al- anaerobiosis. The more efficient oxi- protein gene markers established that though harshly criticized initially, is genic respiration was then adopted by archaea are, strikingly, more similar to supported by extensive evidence6 and is many organisms: aerobic mitochondria eukaryotes3. Furthermore, studies of para- now accepted as mainstream science. evolved. Secondary reduction or loss of logous duplications allowed us to place Later, Margulis7 further proposed that organelles would explain present-day the root of the tree of life tentatively eukaryotes originated through symbio- amitochondriate protists. sis between spirochetes and wall-less The idea that the origin of mitochon- P. López-García is at the Institut de Génétique archaea (Fig. 1), but compelling evidence dria is the key to the origin of the eu- et Microbiologie, Université Paris-Sud, to support this hypothesis is lacking. karyotic cell is not new in scientific bâtiment 409, 91405 Orsay Cedex, France; and D. Moreira is at the Institut de Biologie Surprisingly, despite the potential of thought. Another hypothesis published Cellulaire BC4, Université Paris-Sud, bâtiment symbiosis to account for mixed charac- in 1998, proposes that an aerobic 444, 91405 Orsay Cedex, France. ters (which would be a consequence of ␣-proteobacterium was engulfed by an Email: [email protected] the contribution of at least two partners archaeon prior to the establishment of 88 0968 – 0004/99/$ – See front matter © 1999, Elsevier Science. All rights reserved. PII: S0968-0004(98)01342-5 TIBS 24 – MARCH 1999 TALKING POINT classical endosymbiosis12. The brilliant Common ancestor novelty is that Martin and Müller offer a (a) plausible explanation of the process in terms of metabolism, and conclude that the origins of mitochondria and eukary- otes are identical but that anaerobic mitochondria came first.

The syntrophy hypothesis Our syntrophy hypothesis11 is based on similar metabolic considerations (i.e. we propose that symbiosis was medi- ated by interspecies hydrogen transfer), but we speculate that the organisms involved were ␦-proteobacteria (ances- Eubacteria Eukaryotes Archaea tral sulphate-reducing myxobacteria) and methanogenic archaea (Fig. 2). The (b) (c) (d) hydrogen and syntrophy hypotheses share several common features (Fig. 3), Fusion Engulfment Symbiosis despite our use of a different approach. Wall-less archaeon From , we know that Spirochetes the most widespread symbiotic associ- ation between archaea and eubacteria is 37 3 Lake and Rivera syntrophy between sulphate-reducing Zillig A bacterium engulfs an An archaeon and a eocyte (kingdom Crenarchaeota) bacteria and methanogenic archaea. bacterium amalgamate to Furthermore, the biotopes where these generate the eukaryotic cell organisms exist are ubiquitous. This gave us a likely initial driving force for sym- biosis: syntrophy. The methanogen con- 38 sumed the hydrogen and carbon dioxide Gupta and Golding ␣-Proteobacteria A wall-less Gram-negative Cyanobacteria liberated from the sulphate reducer by bacterium engulfs an eocyte . The sulphate reducer also benefited: it could speed up its meta- bolic rate because it now had a ready hydrogen sink. We based our arguments on a variety of molecular features that, in addition to the classical characteris- Sogin38 tics that link Gram-negative bacteria or An RNA-based proto-eukaryote Margulis7 archaea to eukaryotes, connect myxo- phagocytoses an archaeon Serial endosymbiosis theory bacteria and certain with eukaryotes. Myxobacteria display very Figure 1 complex social behaviour and develop- (a) A schematic view of the evolution of Eubacteria, eukaryotes and Archaea. (b–d) Previous mental cycles, and many of the genes chimeric models for the origin of eukaryotes [colour coded as in (a)]3,7,37–39. involved have specific homologues in eukaryotic signalling pathways13. Meth- The latter are endowed instead with coded proteins involved in metabolism). anogen candidates (Fig. 2) share some small DNA-binding proteins analogous Finally, would have been homologous or -synthesis to those of eubacteria14. lost in favour of a versatile heterotrophy. pathways with eukaryotes, and their con- We envisage an evolutionary pathway tent with respect to many enzymes that in which close cell–cell contacts and ex- Common features interact with DNA (e.g. topoisomerases) tensive membrane development in well- Although their starting points differ, the is similar to that of eukaryotes. How- established symbiotic consortia led to two hypotheses agree in several respects. ever, their most remarkable feature is more-highly evolved structures that had Their arguments are therefore comple- the presence of true histones and nucleo- primitive eukaryotic features, such as a mentary. Central to both is the meta- somes. Archaeal nucleosomes not only protonuclear region (old archaeal cyto- bolic nature of the original symbiotic are homologous in sequence and three- plasm) defined by membranous struc- event; in this sense, both hypotheses in- dimensional structure to eukaryotic tures (Fig. 3). Eubacterial genome extinc- volve hydrogen and syntrophy. Martin

(H3–H4)2 tetramers, but also experience tion could have occurred by progressive and Müller have provided a detailed similar dynamics (D. Musgrave et al., transfer to the archaeal genome, where explanation of the metabolic context pers. commun.). This strongly suggests genes adapted to a new genetic environ- during the process and give a good rea- that these organisms have eukaryotic- ment. Many redundant eubacterial genes son for the loss of the host autotrophic like chromatin (both at structural and (mostly those that encoded the genetic pathway, whereas we have tried to regulatory levels), which is found nei- machinery) would have been lost, construct a more global picture that ther in the kingdom Crenarchaeota nor whereas others would have replaced also covers the formation of the eukary- in the halophilic methanogens (Fig. 2). archaeal genes (mainly those that en- otic genome and membrane systems. 89 TALKING POINT TIBS 24 – MARCH 1999

Eukaryotes

␣ Syntrophy hypothesis Hydrogen hypothesis ␤ ␥ ␦ ⑀ Euryarchaeota Methanosarcina Eubacteria Proteobacteria Methanothrix Halobacterium Spirochetes Chlamydiae Bacteroides Haloferax Green sulphur bacteria Methanococcus Archaeoglobus Cyanobacteria Archaea Thermoplasma Methanopyrus Low G+C ThermococcusPyrobaculum Gram-positive bacteria High Thermoproteus G+C Thermophilum Thermus Deinococcus Sulpholobus Crenarchaeota Aquifex Thermotoga Desulphurococcus bacteria Pyrodictium

Planctomyces Green non-sulphur

Figure 2 An rRNA-based phylogenetic tree showing the locations of the prokaryotic partners that are at the origin of eukaryotes in the two novel symbiosis hypotheses. Myxobacteria and most sulphate reducers belong to the ␦-Proteobacteria.

Another common feature is the sugges- eubacterial types were involved. First, anaerobic. This contradicts the classical tion that a methanogen was the archaeal sulphate-reducing ␦-proteobacteria, which endosymbiosis theory, which assumes partner. Martin and Müller believe that also produce hydrogen from fermentation that the predecessors of mitochondria any anaerobic hydrogen-dependent auto- and form syntrophic consortia with were efficient aerobes. As Smith and trophic archaeon (e.g. a sulphur-depen- methanogens. Second, either at the same Szathmary15 first pointed out, and dent archaeon) could have started the time or shortly after, ␣-proteobacterial Martin and Müller10 emphasize, in the process. Methanogens, however, appeal (the progenitors of mito- endosymbiosis theory the initial benefit to them because they appear early in chondria) took part in the symbiotic for the host is not clear. No bacterium the archaeal tree, are autotrophic, community. These methanotrophs fed on gives free ATP to the medium. anaerobic, and widespread, and can use the produced by the methano- Nevertheless, whereas in the hydrogen hydrogen, carbon dioxide and gen, producing carbon dioxide and hypothesis the presumptive ␣-proteo- (waste products of the presumptive thereby permitting an increase in the rate bacterial ancestor of mitochondria is a symbiont)10. We have several additional of methanogenesis (Fig. 3). Everybody fermentative anaerobe, we suggest that reasons for suggesting that the archaeal was happy. it is an anaerobic . Again, symbiont must have been a methan- It might be difficult to find out which choosing between the two possibilities ogen11. Finally, both proposals give a hypothesis is correct, given that all eu- might be difficult, but some indicative plausible explanation for the mosaic bacterial candidates belong to Proteo- evidence should be investigated further. nature of eukaryotic genomes without bacteria and that phylogenetic signals that For instance, it is widely assumed that proposing any dramatic event but simply could allow us to differentiate between methanotrophs are strict aerobes, be- by postulating gene transfer and re- the two hypotheses might have been ef- cause the enzyme that converts methane placement over a long symbiotic life. faced with time. However, some pieces to methanol, methane monooxygenase, of evidence, which link ␦-proteobacteria requires . However, anaerobic One or two eubacterial symbionts at the (especially myxobacteria) to eukaryotes, methanotrophs that might use sulphate origin? are worth exploring. or nitrate instead exist. Interestingly, The critical difference between the two these are linked to methanogen sulphate- hypotheses is the nature of the eubac- An anaerobic origin for mitochondria reducer consortia16. Also intriguing is the ␣ terial partners (Fig. 2). According to the Regardless of whether the -proteo- recent discovery of several C1-transfer hydrogen hypothesis, ␣-Proteobacteria bacterium was the primary symbiont enzymes and coenzymes (which are re- established the symbiosis and, on the (hydrogen hypothesis) or a secondary quired for the interconversion of one- way to becoming mitochondria, pro- symbiont (syntrophy hypothesis), we carbon compounds) that link methylo- duced eukaryotes. In our proposal, two agree that ancestral mitochondria were trophic bacteria (which feed on C1 90 TIBS 24 – MARCH 1999 TALKING POINT compounds in general) and methano- Anaerobic community genic archaea17. These enzymes and co- factors previously were thought to be ␦-Proteobacteria unique to methanogens and the sul- ␣-Proteobacteria phate reducer Archaeoglobus (a derived Methanogens methanogen). Interdomain horizontal Others transfer of these genes between such in- timately associated groups of organisms could be a good explanation. Mitochondria are believed to be derivatives of Rickettsia-like ancestors18. The Rickettsia are intracellular parasites and seem to be phylogenetically related to mitochondria18. However, because of their adaptation to the intracellular environ- ment, endosymbiotic organelles and cyto- plasmic parasites have accelerated their evolutionary rates and, consequently, display long branches in phylogenetic trees. The fact that fast-evolving lin- eages tend to cluster together artefactu- ally because of the long-branch attraction phenomenon is well known19. Therefore, the phylogenetic positions of such or- ganisms should be regarded with caution. Syntrophy Methanotrophs are interesting alterna- Hydrogen hypothesis hypothesis tives to Rickettsia as mitochondrial pro- genitors. Methanotrophy is widely dis- ␣ 16 tributed among the -Proteobacteria Anaerobic and probably represents an ancestral phenotype in this group. Hence, even if proto-eukaryote mitochondria have a rickettsial origin, Organics their ancestor might have been en- CH4 dowed with this metabolic ability. CO2 Furthermore, note that methanotrophs are commonly ecto- or endo-symbionts. Acetate They are found in the cytoplasm of a CO2 wide variety of eukaryotes (e.g. they are H2 H2 Acetate abundant in the tissues of deep-vent- associated invertebrates) and eubacteria, CO2 such as Beggiatoa (a ␥-proteobacterium that usually forms mats around deep hydrothermal vents)16,20. CH4 Organics Insights from comparative genomics The impressive developments in gen- Figure 3 ome sequencing over the past few years Evolutionary pathways proposed by the hydrogen and syntrophy hypotheses for the origin of have already produced enough data to eukaryotes. The exchange of products between the different prokaryotic partners is shown support a mixed heritage for the eukary- at an advanced state of the metabolic symbiosis. otic genome, which contains archaeal-like DNA-processing (informational) genes and Of course, the fact that these symbi- are being sequenced). Comparative analy- Gram-negative-bacterial-like metabolic otic models explain the mosaic character sis would show whether methanogens (operational) genes8,9. This can only be of eukaryotic genomes does not mean are more closely related to eukaryotes explained either by a massive horizontal that they are correct. Extensive compari- than are other archaea. A myxobacterial gene transfer from Gram-negative bacteria son of the increasingly available genome genome sequence (the Myxococcus xan- to eukaryotic ancestors9 or by a chimeric sequences might, however, help to test thus genome project is also under way) origin8,9. The two symbiosis hypotheses the hypotheses’ robustness. Two genome could support one symbiosis hypoth- marry both possibilities: the chimerism types would be particularly interest- esis rather than the other, given that the they propose is directional. Either a selec- ing: (1) a non-methanogenic archaeal syntrophy hypothesis predicts that tive transfer of metabolic genes towards genome, preferably belonging to the eukaryotes contain a mixture of ␣- and an archaeal host occurred (the hydrogen Crenarchaeota (Pyrobaculum aerophilum ␦-proteobacterial-like genes. hypothesis), or a progressive transfer and should shortly be released, and others Increased evolutionary rates and the replacement of non-informational genes will follow); (2) an extreme halophile (the generation of innovative properties are occurred (the syntrophy hypothesis). genomes of two Halobacterium species associated with symbiosis7. Indeed, 91 TALKING POINT TIBS 24 – MARCH 1999 symbiosis can be regarded as an eco- myxobacteria exist24; an anaerobic temperatures of ~50–70°C) are aerobic at logical association that provides most sulphate-reducing myxobacterium has the surface and possess a cyanobacte- mechanisms (genomic and spatial com- even been identified25. rial layer (the first cyanobacterial mats partmentation, genomic and physio- Communities that might be especially would have given eukaryotes the poten- logical redundancy and specialization, interesting to study are microbial mats, tial to acquire chloroplasts). However, the and evolutionary flexibility) for cir- where sulphate reducers and methano- earliest probably were anaerobic and cumventing or reducing selective con- gens frequently dominate in the anoxic possessed anaerobic photosynthetic eu- straints21. It allows an increase in com- layers, and strong metabolic interac- bacteria instead. In the deeper, anaerobic plexity, the hallmark of eukaryotes, that tions exist. Interestingly, myxobacteria layers, methanogens and sulphate otherwise would be selected against21. are present in microbial mats at active reducers dominate23,34. At 22–55°C, The detection, through comparative gen- hydrothermal vents26. sulphate reduction and methanogenesis omics, of eukaryotic molecules that also predominate in sea-floor commu- have adapted to perform functions dif- Insights from the fossil record nities associated with geothermal regions, ferent from those of their prokaryotic Not only are the biotopes where environments thought to be among the predecessors (and have therefore in- methanogens, sulphate reducers and oldest on the planet35. creased their evolutionary rates) is methanotrophs coexist ubiquitous on particularly interesting. One example the planet today, but some might be as Conclusions might be the evolution of cytoskeletal ancient as the first living organisms. The two symbiosis hypotheses for proteins. Tubulin shares some sequence Around 3500 million years ago, the the origin of eukaryotes try to explain similarity with, and is structurally Earth supported complex prokaryotic as much as possible with the minimum homologous to, prokaryotic FtsZ, which communities that have left us fossil number of assumptions. Thus, although is involved in cell division. Interestingly, stromatolites and microfossils27, whereas they differ in the nature and number of eubacteria have a single ftsZ copy, but the first eukaryotic fossils date from eubacterial original symbionts that are Euryarchaeota at least (and thus meth- 1800–2100 million years ago28. If the proposed, both hypotheses convincingly anogens) have two different copies22. analysis of microfossils alone does not account for the mosaic character of eu- Duplicated genes, which are released reveal decisive information about the karyotic genomes and are based on from functional constraints, can evolve origins of eukaryotes, the combination metabolic interactions that are wide- faster and adapt to new needs, and of this approach and physico-chemical spread in nature. Remarkably, both pro- eukaryotes obviously needed a well- measurements of biogenic markers might. pose that a methanogen was the archaeal developed cytoskeleton. For example, Kral and co-workers29 have partner and that mitochondria have an produced evidence for hydrogen con- anaerobic origin. Insights from microbial ecology sumption by methanogens on the early To prove the Martin and Müller hy- New life is unlikely to be originating Earth. Chappe and co-workers30 have re- pothesis directly could be difficult. We nowadays, because proto-organisms ported the presence of glycerol tetra- already know that the ancestors of mito- would be outcompeted by efficient life ethers, which are characteristic of chondria transferred many genes to their forms. By analogy, eukaryotic life is methanogens, at that time. Ohmoto and host, but how can we identify whether unlikely to be forming anew: proto- Felder31 report that eubacterial sulphate the latter was an archaeon or a member eukaryotes would be outcompeted by reduction occurred in Archaean oceans, of a third lineage (the classical model2)? modern well-adapted eukaryotes. None- which were rich in sulphate, at tempera- To test the syntrophy hypothesis could theless, the study of present-day anaero- tures up to 50°C. On the basis of the iso- be easier, although a recent adaptation bic communities might provide interesting topic composition of the organic carbon of the classical view would explain the clues to eukaryote evolution. contained in sediments, Hayes32 has presence in eukaryotes of eubacterial Molecular ecology might be of further pointed out the existence of a historical genes from different taxonomic groups help. We have identified an enormous peak of methanotrophic activity, which (the ‘you are what you eat’ version)36. diversity of uncultured would be linked to methanogenesis in However, in our model, we expect traits by this means and have confirmed the global carbon cycling, at the Archaean– common to only a restricted range idea that sulphate reducers and meth- Proterozoic transition. Interestingly, the of organisms (␦-proteobacterial, ␣-pro- anogens predominate in many anaero- isotopic signal of methanotrophy ap- teobacterial methanotrophs and some bic meso- and thermophilic environ- pears first and most strongly in stromato- methanogenic archaea) to be found to- ments23. In fact, many organisms that litic units32. By contrast, the extremely gether. We must learn to look at the are candidates for the eubacteria and ar- low abundance of steranes (biomarkers molecular level but, at the same time, at chaea in the syntrophy hypothesis had for eukaryotes), compared with that of the ecological context and at the fossil not been identified until recently, be- hopanes (steroid surrogates that are record. Only if data from different ap- cause they could not be cultivated considered biomarkers for eubacteria), proaches converge, will we be able to (many would be difficult to grow in pure in mid-Proterozoic sediments supports construct a plausible answer to the cultures) or simply because they had a later rise of eukaryotes33. question of the origin of eukaryotes. not been looked for. This was the case Finally, because the early Earth was for methanotrophic anaerobes, which probably warmer and had much less free Acknowledgements are now known to form symbiotic con- atmospheric oxygen, it is interesting to We thank Miklós Müller and William sortia with methanogens and sulphate study living fossil ecosystems, such as Martin for critical reading of the reducers in anoxic environments16. thermophilic microbial mats (laminar manuscript and helpful comments, and Similarly, although myxobacteria gener- stromatolites) or geothermally heated the European Community and the ally are thought to be strict aerobes, we biotopes in the oceanic crust. Most pres- Spanish Ministerio de Educación y have known for a long time that anaerobic ent-day thermophilic mats (which are at Cultura for financial support. 92 TIBS 24 – MARCH 1999 TALKING POINT

References 15 Smith, J. M. and Szathmary, E. (1995) The 28 Han, T. M. and Runnegar, B. (1992) Science 1 Woese, C. R. and Fox, G. E. (1977) Proc. Natl. Major Transitions in Evolution, Freeman, Oxford 257, 232–235 Acad. Sci. U. S. A. 74, 5088–5090 16 Hanson, R. S. and Hanson, T. E. (1996) 29 Kral, T. A., Brink, K. M., Miller, S. L. and 2 Woese, C. R., Kandler, O. and Wheelis, M. L. Microbiol. Rev. 60, 439–471 McKay, C. P. (1998) Orig. Life Evol. Biosph. 28, (1990) Proc. Natl. Acad. Sci. U. S. A. 87, 17 Chistoserdova, L., Vorholt, J. A., Thauer, R. K. and 311–319 4576–4579 Lidstrom, M. E. (1998) Science 281, 99–102 30 Chappe, B., Albrecht, P. and Michaelis, W. 3 Zillig, W. (1991) Curr. Opin. Genet. Dev. 1, 18 Andersson, S. G. E. et al. (1998) Nature 396, (1982) Science 217, 65–66 544–551 133–143 31 Ohmoto, H. and Felder, R. P. (1987) Nature 4 Doolittle, W. F. and Brown, J. R. (1994) Proc. 19 Felsenstein, J. (1978) Syst. Zool. 27, 401–410 328, 244–246 Natl. Acad. Sci. U. S. A. 91, 6721–6728 20 Larkin, J. M. and Henk, M. C. (1996) Microsc. 32 Hayes, J. M. (1994) in Early Life on Earth 5 Margulis, L. (1970) Origin of Eukaryotic Cells, Res. Tech. 33, 23–31 (Bengtson, S., ed.), pp. 220–236, Columbia Yale University Press 21 Kirschner, M. and Gerhart, J. (1998) Proc. Natl. University Press 6 Gray, M. W. (1989) Trends Genet. 5, 294–299 Acad. Sci. U. S. A. 95, 8420–8427 33 Jackson, M. J., Powell, T. G., Summons, R. E. 7 Margulis, L. (1993) Symbiosis in Cell Evolution, 22 Faguy, D. M. and Doolittle, W. F. (1998) Curr. and Sweet, I. P. (1986) Nature 322, 727–729 W. H. Freeman Biol. 8, R338–R341 34 Ward, D. M., Tayne, T. A., Anderson, K. L. and 8 Rivera, M. C., Jain, R., Moore, J. E. and 23 Fenchel, T. and Finlay, B. J. (1995) in Ecology Bateson, M. M. (1987) in Ecology of Microbial Lake, J. A. (1998) Proc. Natl. Acad. Sci. U. S. A. and Evolution in Anoxic Worlds (May, R. M. and Communities (Fletcher, M., Gray, T. R. G. and 95, 6239–6244 Harvey, P. H., eds), pp. 85–98, Oxford University Jones, J. G., eds), pp. 179–210, Cambridge 9 Ribeiro, S. and Golding, G. B. (1998) Mol. Biol. Press University Press Evol. 15, 779–788 24 Sanzhieva, E. U. (1970) Mikrobiologiia 39, 35 Delaney, J. R. et al. (1998) Science 281, 10 Martin, W. and Müller, M. (1998) Nature 392, 817–820 222–230 37–41 25 Cole, J. R., Cascarelli, A. L., Mohn, W. W. and 36 Doolittle, W. F. (1998) Trends Genet. 14, 11 Moreira, D. and López-García, P. (1998) J. Mol. Tiedje, J. M. (1994) Appl. Environ. Microbiol. 60, 307–311 Evol. 47, 517–530 3536–3542 37 Lake, J. A. and Rivera, M. C. (1994) Proc. Natl. 12 Vellai, T., Takacs, K. and Vida, G. (1998) J. Mol. 26 Moyer, C. L., Dobbs, F. C. and Karl, D. M. Acad. Sci. U. S. A. 91, 2880–2881 Evol. 46, 499–507 (1995) Appl. Environ. Microbiol. 61, 1555–1562 38 Gupta, R. and Golding, G. B. (1996) Trends 13 Dworkin, M. (1996) Microbiol. Rev. 60, 70–102 27 Knoll, A. H. (1990) in Paleobiology: A Synthesis Biochem. Sci. 21, 166–171 14 Grayling, R. A., Sandman, K. and Reeve, J. N. (Briggs, D. E. G. and Crowther, P. R., eds), 39 Sogin, M. L. (1991) Curr. Opin. Genet. Dev. 1, (1994) Syst. Appl. Microbiol. 16, 582–590 pp. 9–16, Blackwell Scientific Publications 457–463

http://tto.trends.com Editor: Adrian Bird Institute for Cell and Molecular at the University of Edinburgh

Protocols are now to be featured in Technical Tips Online, in addition to publishing peer-reviewed Technical Tips articles (novel applications or significant improvements on existing methods). Protocols incorporate all the features that are currently available in Technical Tips articles: comment facility; links to Medline abstracts; product information and so on.

New Technical Tips articles published recently in Technical Tips Online include:

■ Riva, P. et al. (1999) A rapid and simple method for the generation of locus-specific probes for fish analysis Technical Tips Online (http://tto.trends.com) T01618 ■ Sebastian Möller and Lars Edvinsson (1999) A strategy for the generation of RNA competitors in competitive RT-PCR Technical Tips Online (http://tto.trends.com) T01604

New products featured in Technical Tips Online include:

Technical Tips Online also features press releases on new products. Click on the ‘product news’ button and a simple reader- response facility allows you to email the relevant company for more information. Recently featured new products include: Efficient sequencing-reaction clean-up with new 96-well format kit The new Quantum Prep® SEQueaky Kleen terminator removal kit from Bio-Rad enables the effective removal of unincorporated fluorescent dye-terminators from up to 96 sequencing reactions simultaneously in under four minutes. A new confocal technique for in vivo double labelling of microvessels without fixation Researchers at the University of Davis in California have used confocal fluorescence and reflection methods simultaneously to produce crisp three-dimensional images of endothelial cells, with no bleed-through problems. This ability to visualize intact vascular architecture and endothelial cell morphology in three-dimensions provides a new way of investigating the response to inflammatory mediators in microvessels.

93