Predation and eukaryote cell origins: A coevolutionary perspective

Predation and eukaryote cell origins: A coevolutionary perspective

The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322 Contents lists available at ScienceDirect The International Journal of Bi...

1MB Sizes 25 Downloads 305 Views

The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

Contents lists available at ScienceDirect

The International Journal of Biochemistry & Cell Biology journal homepage: www.elsevier.com/locate/biocel

Review

Predation and eukaryote cell origins: A coevolutionary perspective T. Cavalier-Smith Department of Zoology, University of Oxford, South Parks Road, Oxford, OX1 3PS, UK

a r t i c l e

i n f o

Article history: Available online 18 October 2008 Keywords: Retrotranslocation Endomembrane system Cytoskeleton Nucleus Quantum evolution Phagocytosis

a b s t r a c t Cells are of only two kinds: bacteria, with DNA segregated by surface membrane motors, dating back ∼3.5 Gy; and eukaryotes, which evolved from bacteria, possibly as recently as 800–850 My ago. The last common ancestor of eukaryotes was a sexual phagotrophic protozoan with mitochondria, one or two centrioles and cilia. Conversion of bacteria ( = prokaryotes) into a eukaryote involved ∼60 major innovations. Numerous contradictory ideas about eukaryogenesis fail to explain fundamental features of eukaryotic cell biology or conflict with phylogeny. Data are best explained by the intracellular coevolutionary theory, with three basic tenets: (1) the eukaryotic cytoskeleton and endomembrane system originated through cooperatively enabling the evolution of phagotrophy; (2) phagocytosis internalised DNA-membrane attachments, unavoidably disrupting bacterial division; recovery entailed the evolution of the nucleus and mitotic cycle; (3) the symbiogenetic origin of mitochondria immediately followed the perfection of phagotrophy and intracellular digestion, contributing greater energy efficiency and group II introns as precursors of spliceosomal introns. Eukaryotes plus their archaebacterial sisters form the clade neomura, which evolved from a radically modified derivative of an actinobacterial posibacterium that had replaced the ancestral eubacterial murein peptidoglycan by N-linked glycoproteins, radically modified its DNA-handling enzymes, and evolved cotranslational protein secretion, but not the isoprenoid-ether lipids of archaebacteria. I focus on this phylogenetic background and on explaining how in response to novel phagotrophic selective pressures and ensuing genome internalisation this prekaryote evolved efficient digestion of prey proteins by retrotranslocation and 26S proteasomes, then internal digestion by phagocytosis, lysosomes, and peroxisomes, and eukaryotic vesicle trafficking and intracellular compartmentation. © 2008 Elsevier Ltd. All rights reserved.

Contents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Dating eukaryote origins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 Eukaryote phylogeny and the properties of the earliest eukaryotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308 Neomuran and ␣-proteobacterial precursors of eukaryotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 The neomuran revolution and immediate archaebacteria/eukaryote divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311 Transition analysis resolves key difficult problems of bacterial phylogeny . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 The neomuran ancestor of eukaryotes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 312 Prerequisites for evolving phagotrophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313 Internal digestion first by retrotranslocation and 26S proteasomes? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 Pseudopodia, prey uptake, and membrane recycling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 315 DNA internalisation, copII and the origin of permanent endomembranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 316 Peroxisomes as endomembrane digestive differentiations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317 Phagotrophy and novel protein targeting enabled symbiogenetic organelle additions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318 Cilia, a novel compartment not delimited by membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 The botched, piecemeal nature and rapidity of megaevolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320

Abbreviations: LGT, lateral gene transfer; ERAD, ER-associated degradation of proteins; SR, signal-recognition-particle receptor; NE, nuclear envelope. E-mail address: [email protected]. 1357-2725/$ – see front matter © 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.biocel.2008.10.002

308

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

The origin of eukaryotes (eukaryogenesis) was the largest reorganization of cell structure ever. To explain it we must answer six questions: (1) When did they evolve? (2) What was the nature of the last common ancestor of all eukaryotes (the cenancestral eukaryote)? (3) What were its ancestors? (4) What were the physical mechanisms of the changes? (5) What were the major steps involved? (6) What triggered such exceptionally disruptive but seminal changes? This cannot be done in detail here, as ∼5000 novel genes originated during the origin of eukaryotes, the most dramatic genetic explosion in history (Makarova et al., 2005; Yang et al., 2005; Cavalier-Smith, 2007a), and I have only ∼9000 words. Fortunately, not all genes are equally fundamental, and understanding eukaryogenesis is not nearly as difficult as often made out. Elements of a sound explanation already exist through advances in ultrastructure, molecular and cell biology, genetics, phylogeny, and theoretical analyses over 40 years. Persisting problems are that bacteria are so small that their cell biology lags greatly behind that of eukaryotes and we do not know the functions or 3D structure of many important structural proteins; lipid membrane dynamics that endow cells with form and integrity are also insufficiently understood. Eukaryogenesis poses the problem how and why, just once in the history of life, cells radically spatially reorganized their membrane, skeleton, and chromosomal relationships. The origin of an endoskeleton, membrane budding and fusion, and a novel mode of feeding were fundamental. Table 1 summarises 60 innovations that eukaryogenesis theories must explain. Rather than trying to explain each in detail I focus on five things: (1) setting the phylogenetic scene; (2) showing how the different changes were logically interconnected; (3) emphasizing that twin themes of disruption and continuity underlie a coherent explanation; (4) explaining in more detail than hitherto the earliest steps in endomembrane origins associated with the evolution of phagotrophy; (5) arguing that these were probably preceded by a simpler predatory stage with internal digestion mediated by retrotranslocation and improved proteasomes. I use the classical term bacteria as a simpler synonym for prokaryote (Cavalier-Smith, 2007b), i.e. embracing both classical bacteria and cyanobacteria, which prior to invention of that name (Stanier, 1974) were called blue-green algae or Cyanophyta, and also archaebacteria, renamed archae (Woese et al., 1990). See also discussion in Cavalier-Smith, 1991a,b.

mals date from ∼550 My, but most phyla only appeared after 530 My during the Cambrian explosion (most animals, some protozoa, e.g. Foraminifera, Radiozoa, green algae) or substantially later (land plants). A few poorly dated Melanocyrillium-like fossils date from ∼800 My, but relatively numerous deposits dated 850 My are devoid of them or anything definitely eukaryotic. Thus the most conservative estimate of the age of eukaryotes is 850–800 My ago (Cavalier-Smith, 2002a,c). That they are as old as bacteria (Kurland et al., 2006) is disproved by the fossil evidence. In marked contrast there is unequivocal evidence for oxygenic photosynthetic prokaryotes as early as 2.45 Gy ago; most palaeontologists think cyanobacteria arose earlier, 2.9–2.7 Gy ago. No convincing evidence shows life before 3.5 Gy, currently the best estimate of when life began. Thus bacteria are probably four times as old as eukaryotes, making it certain that eukaryotes evolved from bacteria, not the reverse (Cavalier-Smith, 2006a). Even were optimistic identifications of a few meagre fossils ∼1.5 Gy ago as eukaryotic (Javaux et al., 2001) justified (I think not), bacteria would be 2.3× as old as eukaryotes. No molecular biological ‘clock’ ticks constantly throughout geological time. Proteins evolve at rates differing over many orders of magnitude. As new proteins all evolve from old ones by gene duplication, rates must change dramatically over time. They change systematically among different branches of the tree and also episodically. When new paralogues arise, evolution is initially very fast as novel functions are acquired, e.g. during eukaryogenesis the ancestral RNA polymerase evolved into RNA polymerases I, II and III divergently adapted for transcribing rRNA, mRNA and tRNA; as distinct functions became perfected initial fast evolution gave way to much slower more trivial divergence. The bigger the functional shift the more dramatic the transient initial acceleration, a >10,000 fold increase in rate being likely for molecules like tubulins and actins that arose from bacterial FtsZ and MreB by multiple duplications (Amos et al., 2004; Erickson, 2007). Similar transient increases occur in rRNA. Many evolutionary misinterpretations stem from treating sequence divergence as clock-like (Cavalier-Smith, 2002c). Averaging rates of change in local parts of the tree allows useful interpolation between known fossil dates, but extrapolating backwards beyond fossil calibration points is extremely unreliable, providing no useful information beyond what fossils directly say, yet giving false confidence in inferences (Graur and Martin, 2004; Roger and Hug, 2006). It is scientifically unsound to use a ‘clock’ of tick-rate unknown by a factor of 10,000.

1. Dating eukaryote origins The oldest indubitably eukaryotic fossils are vase-shaped, e.g. Melanocyrillium; their oldest secure date for numerous wellpreserved specimens is 760 My ago (Porter and Knoll, 2000). They are almost certainly shells of testate amoebae constructed by pseudopodial activity that never occurs in bacteria. Claims that they are arcellinid amoebae (phylum Amoebozoa) and euglyphid amoebae (phylum Cercozoa) (Porter et al., 2003) are highly questionable. None are confidently morphologically euglyphids (those suggested to be could be another group with agglutinated shells); no marine arcellinids are known, yet these fossils are all marine. Moreover, they apparently became extinct before the Phanerozoic. Most likely they were an extinct group of testate amoebae that flourished before Foraminifera evolved (Cavalier-Smith, in press). Possibly, Cryogenian glaciations that largely or entirely covered the globe in several kilometres of ice periodically from ∼710 to 635 My ago extinguished them (snowball earth). The only fossils confidently assignable to a modern eukaryotic phylum all postdate the melting of snowball earth (Cavalier-Smith, 2006a). The first are red algae (Rhodophyta) about 600 My old. Earliest ani-

2. Eukaryote phylogeny and the properties of the earliest eukaryotes To infer the nature of the first eukaryote rigorously we must locate the root of the eukaryotic tree confidently. This has been difficult, with many false trails. For a century, the first eukaryotes were variously postulated to be algae, fungi or protozoa; if protozoa, anaerobic or aerobic amoebae, flagellates or amoeboflagellates have each been considered primitive. It is well established that mitochondria evolved by symbiogenetic cell enslavement (Cavalier-Smith, 2006b, 2007a) from ␣-proteobacteria, which have the most mitochondrion-like respiratory chain (John and Whatley, 1975) and include purple non-sulphur photosynthetic bacteria. Sequence phylogeny (Keeling et al., 2005; Rodríguez-Ezpeleta et al., 2007) has revealed major clades of the eukaryote tree that are congruent with much ultrastructural data and helps position its root (Fig. 1). All known groups of anaerobic eukaryotes had ancestors that were at least facultatively aerobic with oxidative phosphorylation in mitochondria; in various protozoa and fungi mitochondria were subsequently polyphyletically modified as hydrogenosomes

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

309

Table 1 Sixty groups of key innovations in the origin of the eukaryotic cell and cell cycle. 1. Phagocytosis (Cavalier-Smith, 1987c, 2002e) 2. Actin and actin-related proteins (Arps 1,2, 3) (Cavalier-Smith, 2006a) 3. Pinocytosis (Field et al., 2007) 4. SNAREs (Koumandou et al., 2007; Yoshizawa et al., 2006) 5. 3+ myosins (Odronitz and Kollmar, 2007; Richards and Cavalier-Smith, 2005) 6. Tubulin evolution from FtsZ and subsequent triplication: ␥ for centrosome and ␣ and ␤ for microtubules fixing it to cell surface (Cavalier-Smith, 1987b, 2002e, 2006a) 7. 13 kinesins (Cavalier-Smith, 2006a; Wickstead and Gull, 2006)) 8. Dynein for sliding surface-attached astral microtubules and related midasin for ribosome export (Cavalier-Smith, 2006a) 9. Exocytosis and exocysts (Koumandou et al., 2007) 10. Arf1 and Sar1 GTPases (Dacks and Field, 2004) 11. Rab GTPases (Jékely, 2004) 12. Rho GTPases (Jékely, 2004) 13. Ras GTPases (Jékely, 2004) 14. Centrioles and ␦, ␧, and ␩ tubulins (Beisson and Wright, 2003; Feldman et al., 2007; Keller et al., 2005; Marshall, 2007; Zamora and Marshall, 2005) 15. Cilia (nine doublets, dynein arms and centre pair spokes, ciliary transport) (Jékely and Arendt, 2006; Mitchell, 2007) 16. Clathrin coats and adaptins 17. COPI vesicle coats (Stagg et al., 2008) 18. COPII coats (Gürkan et al., 2006; Stagg et al., 2008) 19. Peroxisomes (Tabak et al., 2006) 20. Endosomes (early, late and multivesicular bodies) (Leung et al., 2008) 21. Lysosomes 22. Golgi complex 23. Sphingolipid synthesis 24. Trans-Golgi network 25. Centrin (Ca++ contractility) (Salisbury, 2007) 26. Phosphatidylinositol/kinase signaling 27. Calmodulin, Ca++ and inositol triphosphate second messenger systems. 28. Delrin protein extrusion channel for ER-associated degradation (ERAD) 29. Ubiquitin and polyubiquitin labelling system 30. 26S proteasomes with 19S regulatory subunit (Cavalier-Smith, 2006c) 31. SR␤ of signal-recognition-particle receptor (Schwartz and Blobel, 2003; Schwartz et al., 2006) 32. Plasma membrane phosphatidylinositol anchor proteins (Oriol et al., 2002) 33. Cell cycle resetting by anaphase proteolysis (Cavalier-Smith, 2005, 2006a,c; de Lichtenberg et al., 2007) and mitosis 34. Massive expansion of serine/threonine kinase controls (Shiu and Bleecker, 2001) 35. Formins for positioning actomyosin (Grunt et al., 2008; Higgs and Peterson, 2005) 36. AAA lysine biosynthetic pathway (Cavalier-Smith, 1987b; Sumathi et al., 2006) 37. Mcm replication licensing system controlled by cyclins (de Lichtenberg et al., 2007; Krylov et al., 2003; Nasmyth, 1995) 38. Internalisation of DNA attachment sites as protoNE/roughER (Cavalier-Smith, 1975, 1987b, 1991a, 2002e) 39. Cell division by actomyosin not FtsZ (Cavalier-Smith, 1975, 1981, 1987b, 1992, 2002e, 2006a) 40. 4-histone nucleosomes 41. Separate RNA polymerases I, II and III 42. Four-module 30-subunit mediator complex regulating polII transcription (Bourbon, 2008) 43. Chromatin condensation cycle: histone phosphorylation, methylation, acetylation; heterochromatin (Cavalier-Smith, 2005) 44. Centromeres/kinetochores (CenpA from core histone) for attaching DNA to microtubules (Cavalier-Smith, 1987b, 2002e, 2006a) 45. Meiosis and synaptonemal complex (Cavalier-Smith, 1981, 1987a, 1995a, 2002b,d,e) 46. Telomerases and telomeres (Cavalier-Smith, 1981, 1987b, 1988, 2002e) 47. Post-transcriptional gene silencing, dicer and argonaut nucleases (Molnar et al., 2007) 48. Proteinaceous interphase nuclear matrix with bound DNA-topoisomerase II and its ability to reorganize as mitotic chromosome cores (Cavalier-Smith, 1982a) 49. Nuclear lamina (Cavalier-Smith, 1982a, 2005) 50. Nuclear pore complexes (NPCs) (Bapteste et al., 2005; Cavalier-Smith, 2004b, submitted for publication; Devos et al., 2004, 2006; Mans et al., 2004) 51. Nucleolus and more complex rRNA processing (e.g. 5.8S rRNA) (Cavalier-Smith, 2002e) 52. Ran GTP/GDP cycle for directionality of NE export/import (Bapteste et al., 2005; Jékely, 2004; Mans et al., 2004) 53. Karyopherins (Bapteste et al., 2005; Mans et al., 2004) 54. Ribosome subunit nuclear export machinery (Bapteste et al., 2005; Mans et al., 2004) 55. mRNA capping and export machinery (Bapteste et al., 2005; Cavalier-Smith, 1981; Mans et al., 2004) 56. polyA transcription termination system (Cavalier-Smith, 1981) 57. Nuclear envelope fusion and syngamy (Cavalier-Smith, 1995a, submitted for publication) 58. Mitochondria (Cavalier-Smith, 2006b, 2007a) 59. Spliceosomes and spliceosomal introns (Cavalier-Smith, 1981, 1985, 1987b, 1991c, 1993) 60. Nonsense-mediated mRNA decay (Maquat, 2004; Cavalier-Smith, submitted for publication) The references discuss their evolution or molecular basis in more detail. This list includes only characters that evolved before the cenancestor and excludes later additions like chloroplasts or intermediate filaments. It is probably not exhaustive. Another possible one is contractile vacuoles, if, contrary to the deduction that they were originally marine (Cavalier-Smith, in press), eukaryotes originated in soil or freshwater. Note that only the last three innovations depended on symbiogenesis; by focusing on them to the exclusion of the 57 primary, purely autogenous, and often far more complex, innovations, too many speculators have put the cart before the horse and evaded the key issues. These are explaining coordinated radical structural change in almost every non-metabolic aspect of the cell. Molecular cell biology, not metabolic biochemistry or population genetics or sequence phylogeny, is required to understand the transition.

or mitosomes, double membrane organelles typically lacking DNA and ribosomes but retaining the Tom/Tim protein-import machinery that evolved in the last common ancestor of all eukaryotes (Embley, 2006; Gill et al., 2007; Stechmann et al., 2006). The idea of a photosynthetic ancestor of all eukaryotes was abandoned when the symbiogenetic origin of chloroplasts by permanent

intracellular enslavement of a cyanobacterium (Mereschkovsky, 1905) was proved in the late 1970s by discovering that chloroplasts of the glaucophyte alga Cyanophora retain a typically eubacterial peptidoglycan wall. One primary cyanobacterial symbiogenesis in a biciliate host created the plant kingdom, whereas other eukaryotic algae arose by secondary symbiogenesis: enslavement of a pre-

310

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

Fig. 1. The tree of life, emphasizing the fundamental difference between the ancestral negibacteria, with two bounding membranes, and the advanced unibacteria (Posibacteria plus Archaebacteria), which evolved from them by losing the outer membrane before giving rise to eukaryotes. Marked changes in membrane and wall chemistry were of key importance in bacterial evolution and in preparation for eukaryogenesis. The neomuran revolution, when core histones evolved and N-linked glycoproteins that are glycosylated during translation replaced the peptidoglycan murein, created a fundamental division between the ancestral eubacteria (Negibacteria plus Posibacteria), which are a grade of organization not a taxon, and the derived clade neomura (Archaebacteria plus Eukaryota), which is also not a taxon. Symbiogenetic enslavements of two negibacteria (␣-proteobacteria, cyanobacteria) to form mitochondria and chloroplasts, are shown by thin grey arrows; their organellar envelope outer membranes evolved from the negibacterial outer membranes, not from the food vacuoles membrane which was lost, liberating them into the cytosol (Cavalier-Smith, 1982c, 1983, 2000, 2006a,b,c). A secondary symbiogenesis generated all chromophyte chloroplasts from an enslaved red alga (thick grey arrow), adding two extra membranes across which proteins have to be targeted; the food vacuole membrane was retained and signal peptides added upstream of transit peptides for crossing it; additionally the ERAD extrusion channel duplicated and was recruited for protein import across the extra periplastid membrane (the former red algal plasma membrane) (Maier, in press). Two secondary symbioses not shown by arrows implanted green algal chloroplasts into chlorarachnean Cercozoa and euglenoid excavate flagellates (asterisks). For more on the bacterial tree and its rooting see Cavalier-Smith (2006a, 2006c); for more on the eukaryote tree see Cavalier-Smith (2003b, submitted for publication, in press).

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

existing eukaryotic plant cell by a phylogenetically distinct host, which adds extra membranes (requiring additional translocation machinery) and sometimes even a foreign nucleus (Cavalier-Smith, 1995b, 1999, 2000, 2002a, 2003a). Demonstration of shared Toc/Tic protein-targeting machinery in plastids throughout kingdom Plantae (Glaucophyta, Rhodophyta (red algae) and Viridiplantae, i.e. green plants) proved a common origin for Plantae by a unique cyanobacterial enslavement. The most important of four known secondary symbiogeneses was the enslavement of a red alga by a biciliate host to form chromalveolates, a major clade comprising all chlorophyll-c containing algae (chromophytes) and numerous heterotrophic descendants whose ancestors secondarily lost photosynthesis, e.g. ciliate protozoa, malaria parasites, oomycete pseudofungi, (Cavalier-Smith, 1999, 2004a). Common proteintargeting mechanisms in chromalveolates (Cavalier-Smith, 2003a) plus compelling evidence of shared plastid protein replacements by duplicate host genes (Keeling, in press) firmly establishes the single secondary symbiogenetic acquisition of chromalveolate chloroplasts. Three independent secondary symbiogeneses implanted green algal plastids into chlorarachnean Cercozoa, euglenoids, and one small dinoflagellate clade. Another small dinoflagellate clade acquired haptophyte plastids by tertiary symbiogenesis. The root of the eukaryote tree cannot be within Plantae because eukaryote cells must have existed before one enslaved a cyanobacterium to make the first plant. Likewise, the root cannot be within chromalveolates, as the enslaved red alga evolved after Plantae. Just as unique plastid enslavement excludes the root from these clades, so can a unique lateral gene transfer (LGT); several shared by the anaerobic diplomonads and parabasalids show that the root cannot lie between them (Andersson et al., 2005), where early rRNA trees subject to gross long-branch artefacts falsely put them. The root is thought to be between unikonts and bikonts (Fig. 1); two alternative positions, within unikonts (between opisthokonts and Amoebozoa) or within bikonts (between excavates and others), are less likely but not completely unreasonable. Whichever is correct, we can conclude that the eukaryote cenancestor had one or two cilia and aerobic mitochondria, and was unquestionably a phagotroph that ate other cells. This necessarily follows from the absence of any branch on the eukaryotic tree below the last common ancestor of animals and plants and the presence of these homologous characters on both sides of the deepest bifurcation. As secondary loss of photosynthesis does occur and is not firmly excluded even for the ancestral unikont, it is formally possible that the eukaryotic cenancestor was photosynthetic; it might have already enslaved a plastid, in which case coadaptation and syntrophy among plastids, mitochondria and peroxisomes now seen in plants dates from it (Cavalier-Smith, 1987c). However, later origin of plastids in an early bikont is more likely; arguably eukaryotes were initially benthic heterotrophs, plastids only evolving with the first plankton (Cavalier-Smith, in press).

3. Neomuran and ␣-proteobacterial precursors of eukaryotes Eukaryote cells are evolutionary chimaeras of an ancestrally phagotrophic cell with nucleus, endomembranes, and endoskeleton (Cavalier-Smith, 2002e) and an enslaved ␣-proteobacterium converted into a mitochondrion (Fig. 1) (Cavalier-Smith, 2006b). Archaebacteria are closely related to the eukaryote host (together forming a clade called neomura: Cavalier-Smith, 1987b, 2002c). Archaebacteria are not direct ancestors of eukaryotes. Instead they are their sisters (Cavalier-Smith, 2002c, 2006a,c). Besides those earlier arguments, comprehensive 136-gene analysis convincingly places archaebacteria as a holophyletic clade, sister to eukaryotes

311

not ancestral to them (Yutin et al., 2008). Thus >20 features shared by both groups but absent from eubacteria (e.g. N-linked glycoproteins, more complex RNA polymerases, core histones) are not specifically archaebacterial, but neomuran characters evolved by their common ancestor, which itself arose during the ‘neomuran revolution’ caused by loss of the ancestral eubacterial murein wall, arguably from an actinobacterium (Cavalier-Smith, 1987b, 2002c, 2006a,c). Calling the host for mitochondrial enslavement ‘archaebacterial’ (Searcy, 1992; Yutin et al., 2008) is therefore wrong; such uncritical terminology has extremely distorted much thinking about eukaryogenesis. Archaebacterial holophyly disproves the numerous archaebacteria-as-host hypotheses of the origin of mitochondria and eukaryotes, e.g. López-García and Moreira (2006); Martin and Müller (1998), which are also all explanatorily empty for nearly all eukaryotic characters (Table 1). The ancestor was not a eubacterium either, but an extinct missing link – an early neomuran with all characters shared by both archaebacteria and eukaryotes, but no uniquely archaebacterial properties. Purely archaebacterial characters (isoprenoid-ether lipids, archaeosine modified rRNAs, unique flagella, duplicate versions of DNA polymerase B (Rogozin et al., 2008)) are most parsimoniously interpreted as having evolved in the ancestral archaebacterium after it diverged from the prekaryote lineage (Cavalier-Smith, 2002c, 2006c). Moreover, genes shared by eukaryotes and eubacteria, but not archaebacteria (e.g. MreB that became actin (Cavalier-Smith, 2002e, 2006a), eubacterial surface molecules that became nuclear envelope lamin B receptors (Bapteste et al., 2005), cytochrome P450 ancestors of ER respiration, and enzymes making acyl ester phospholipids and sterols), were probably lost by the ancestral archaebacterium, which apparently lost ∼1000 genes when adapting to hyperthermophily (CavalierSmith, 2002c, 2006c, 2007b).

4. The neomuran revolution and immediate archaebacteria/eukaryote divergence To reconstruct the nature of the transient early neomuran vertical ancestor of eukaryotes (i.e. excluding what came from the ␣-proteobacterium), we must consider not only what genes were present in the ancestral archaebacterium, but also those of the eubacterial ancestors of neomura. When arguing that neomura are a clade derived from eubacteria by radical changes in cell surface and mode of DNA supercoiling, Cavalier-Smith (1987b) deduced that their closest eubacterial relatives were posibacteria. Posibacteria (Actinobacteria plus Endobacteria: Cavalier-Smith, 1987b, 2002c) are the only eubacteria with one bounding membrane like neomura and thus the potential to evolve into neomura merely by replacing the murein wall by N-linked glycoproteins. Other eubacteria (collectively Negibacteria) all have an extra outer membrane that would have to be lost, which is extremely difficult (Cavalier-Smith, 1987b). I proposed that actinobacteria, often morphologically complex aerobes, with cell differentiation, aerial spores and very diverse lipids, including invariably phosphatidylinositol that was crucial for eukaryote origins but absent from all other bacteria, were most likely ancestral to neomura. Endobacteria (Firmicutes in its confused modern sense minus Eurybacteria: see Cavalier-Smith, 2006c) have specialised endospores that could not have evolved into protozoan cysts and fungal and plant spores (unlike actinobacterial exospores) and lack phosphatidylinositol and sterols. For deducing that actinobacteria are the closest relatives of neomura, a key consideration was the evolution of proteasomes, cylindrical proteolytic chambers found in neomura and actinobacteria, with simpler precursors in other bacteria (Gille et al., 2003). Actinomycete proteasomes were considered ancestral to those of

312

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

neomura, and derived from HslV of other eubacteria (CavalierSmith, 2006c). Two recent findings complicate that inference. First, a second eubacterial proteasome precursor was discovered: Anbu (Valas and Bourne, 2008), three-dimensionally somewhat more similar to both 20S proteasomal subunits than is HslV, earlier considered the ancestor of actinobacterial/archaebacterial 20S proteasomes (Cavalier-Smith, 2006c). Anbu, not HslV, was probably ancestral to 20S proteasomes. Second, actinobacterialtype proteasome genes occur in environmental metagenomes and metaproteomes dominated by Leptospirillum negibacteria (De Mot, 2007); unless they come from contaminating actinobacteria, proteasome genes probably underwent lateral gene transfer from actinobacteria to Leptospirillum, a lineage of uncertain phylogenetic position, likely a very deep-branching proteobacterium (not a separate phylum, as many assume because its relatives are uncertain) of no significance for eukaryote origins. Possibility of LGT does not invalidate the conclusion that actinobacterial proteasomes are probably ancestral to those of neomura, as many other characters finger actinobacteria as ancestral to neomura; concordance of disparate characters is convincing. Among these are phosphatidylinositol, with pervasive roles in the evolution of phagocytosis and endomembrane trafficking but absent from all other bacteria, and sterols (never in archaebacteria). The uniqueness of archaebacterial lipids was a major (not sole) reason for rejecting the idea of an archaebacterial ancestry of eukaryotes (Van Valen and Maiorana, 1980) in favour of the actinobacterial/neomuran interpretation where archaebacteria are simply the sisters of eukaryotes (Cavalier-Smith, 1987b). Replacement of archaebacterial lipids by acyl ester lipids from the enslaved ␣-proteobacterial ancestor of mitochondria is a formal possibility (Martin and Müller, 1998), but evolutionarily extremely onerous and unlikely; phylogeny gives no reason to assume it in the first place. Far more likely, archaebacteria and eukaryotes evolved from a common prokaryotic ancestor with acyl ester lipids. Shared neomuran characters are best interpreted as derived characters stemming from two key changes to eubacteria after murein was lost: origin of cotranslational synthesis of N-linked glycoproteins and evolution of passive negative DNA supercoiling by core histones (thermal ratchets possibly derived from MreB or FtsZ: Gardiner et al., 2008) instead of active negative supercoiling by eubacterial DNA gyrase (Cavalier-Smith, 2002c). Both changes were arguably logical novelties adapting cells to hotter environments, ones in the reverse direction being incomprehensible. Likewise all novel features of archaebacteria, especially lipids and flagella, are explicable as adaptations to even higher temperatures (hyperthermophily). Archaebacterial extremophily is a derived character compared with eubacterial mesophily; a reverse transition makes no sense (Cavalier-Smith, 2002c).

5. Transition analysis resolves key difficult problems of bacterial phylogeny Critical analysis of major transitions involving complex characters can polarise them and distinguish ancestral from derived states. Such ‘transition analysis’ is important for understanding major evolutionary transitions, because sequence analysis generally cannot distinguish ancestral and derived states (CavalierSmith, 1991a, 2006c). Applying transition analysis to major prokaryotic transitions roots the tree of life more reliably than paralogue rooting, which gave two contradictory results; a root in eubacteria (most metabolic enzymes) or between eubacteria and neomura (DNA/ribosome-related enzymes) (Cavalier-Smith, 2006c). It strongly indicates that eubacteria are not a clade but are ancestors of archaebacteria, confirming suspicions that

DNA/ribosome-related paralogue trees (Gogarten et al., 1989; Iwabe et al., 1989) misrooted the tree because of long-branch artefacts. It supports actinobacteria as the closest relatives of neomura, and strongly indicates that posibacteria are derived from, not ancestral to negibacteria (Cavalier-Smith, 2006c); contrary arguments (Lake et al., 2007; Skophammer et al., 2007) are extremely weak, the indel data being explicable by convergent deletions or unrelated insertions. I concluded that the root of all life is within negibacterial Eobacteria (Hadobacteria plus Chlorobacteria), specifically between Chlorobacteria (e.g. the filamentous non-sulphur green bacterium Chloroflexus) and all other organisms or possibly within Chlorobacteria (Cavalier-Smith, 2006c). Critical evaluation of fossil evidence also fits this conclusion, showing negibacteria as over four times older than eukaryotes; archaebacteria are probably no older than eukaryotes, thus the youngest bacterial phylum (CavalierSmith, 2006a), not the oldest as often assumed. It is necessary to use transition analysis to root the tree and, together with discrete character cladistic analysis, to deduce the bacterial tree topology, including which eubacteria are closest to neomura, because sequence trees are indecisive (Cavalier-Smith, 2006a). No single gene retains enough deep phylogenetic information for reliably reconstructing relationships among bacterial phyla. This is true even for eukaryotes where deepest divergences were only about 600–500 My ago. Unsurprisingly for bacteria the task is even harder, as their major divergences were probably 2.5–3.5 Gy ago. Of the few thousand genes in most free-living bacteria, only ∼136 are conserved enough for multiphylum multigene trees (Yutin et al., 2008); even these have less resolution than the greater number used for deep eukaryotic phylogeny. Biased evolutionary rates are another problem. Because of marked changes to DNA-handling enzymes and ribosome-related enzymes of neomura compared with metabolic enzymes (Cavalier-Smith, 2002c), these two molecular classes give discordant results, e.g. rRNA greatly exaggerates the evolutionary distance between neomura and eubacteria, introducing misleading long-branches into trees. Although posibacteria were probably ancestors to neomura, and actinobacteria plus neomura together are probably a clade: the proteates (Cavalier-Smith, 2006c), it is less clear whether actinobacteria are direct ancestors of neomura as I argued is most likely (Cavalier-Smith, 1987b, 2002c,e, 2006c) or just their sisters. A gyrase insertion shows that Actinomycetes are a clade, so neomura do not branch within them, but could be their sisters; contrary to Servin et al. (2008) that insertion is not in all Actinobacteria, being absent in Rubrobacter.

6. The neomuran ancestor of eukaryotes Thus the immediate ancestor of eukaryotes was neither an archaebacterium, nor a eubacterium with murein wall and DNA gyrase, but a transitional prokaryotic intermediate with a unique character combination: a surface coat or wall of cotranslationallymade N-linked glycoproteins, acyl ester phospholipids, including phosphatidylinositol, sterols, H3/H4 core histones and neomurantype DNA-handling enzymes. It was probably a large facultatively aerobic heterotrophic cell with huge secretome of hundreds of proteins (actinobacteria have 800) including numerous digestive enzymes. As no bacteria have this combination of characters, it probably persisted but briefly historically, surviving only through splitting into two further-altered daughter lineages: the radically divergent eukaryotes and archaebacteria. Thus the first neomuran had an unstable phenotype unable to compete effectively with standard mesophilic bacteria. The special strength of its surviving daughter lineages was arguably that each entered a novel adaptive zone totally free of eubacterial competitors: phagotrophy for

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

eukaryotes; hyperthermophily for archaebacteria. This ancestral neomuran itself probably arose in a thermophilic (not hyperthermophilic) environment for which histones and cotranslational rather than post-translational secretion were specific adaptations (Cavalier-Smith, 2002c). Many actinomycetes are thermophilic, a habitat in which their proteasomal destruction of denatured proteins is useful. Whether the habitat was terrestrial, freshwater or marine is unclear, but from the distribution across the eukaryote tree of habitat preferences, a marine ancestry is somewhat more parsimonious; a benthic ancestry is much more likely than a planktonic one (Cavalier-Smith, in press). The most likely habitat for the origin of neomura was a benthic thermal gradient near a submarine hydrothermal vent, where extremes of temperature and redox potential closely juxtaposed and the transitional population was small. There, one daughter lineage could have rapidly become the first (hyperthermophilic) archaebacterium, the other the first phagotroph. I suggest that this occurred within a dense microbial mat in shallows of the photic zone, where cyanobacteria, actinomycetes, and proteobacteria abounded, making competition for resources intense with marked chemical warfare by antibiotics. In a dense marine mat high salt and organic solutes exuded by phototrophs would have been osmoprotectants, enabling the ancestral neomuran to survive loss of the murein wall, possibly favoured by consequent resistance to peptidoglycan-targeted antibiotics like penicillin. Moreover, in dense mats, prey would be immediately available all around, so phagocytosis could evolve without efficient eukaryotic motility; they just had to bathe in their food not chase or fish it, as necessary for plankton. At the opposite extreme, freshwater plankton would be the most difficult environment for losing a cell wall, with higher likelihood of osmotic rupture, or acquiring prey or evolving hyperthermophily. Soil would be osmotically dangerous, requiring surviving rain and/or extreme drying. Prekaryotes happened to undergo gene duplications in the surface skeletal protein Mreb, making actin and actin-related proteins (Arps) able to cross-link actin in a 3D meshwork and osmotically stabilise the cell. This preadapted them for amoeboid locomotion by pseudopodia, ingestion of prey by phagocytosis and cytokinesis by actomyosin, freeing bacterial FtsZ protein (Osawa et al., 2008) to become tubulin and then duplicate to six paralogues (Beisson and Wright, 2003; McKean et al., 2001; Tuszynski et al., 2006) enabling mitosis and cilia to evolve. An intriguing idea that possibly facilitated eukaryogenesis is that tubulins evolved not from bacterial cell division FtsZ, but from related proteins that segregate posibacterial plasmids, with properties more like tubulin (Chen and Erickson, 2008); this could have enabled cell division by FtsZ to persist temporarily whilst microtubules evolved for chromosome segregation. By contrast, prearchaebacteria became miniaturized and increased their surface stability by making a rigid wall from their glycoproteins and evolving a membrane with a covalently bonded unilayer of more rigid C40 isoprenoid-ether lipids (Cavalier-Smith, 2002c). Arguably, sterols in their membranes helped rigidify them during the transitional stage when ancestral neomuran sterols and acyl ester phospholipids were replaced by isoprenyl ethers. Archaebacteria ancestrally retained FtsZ for division, but lost MreB. A handful of archaebacteria have MreB-related proteins, probably all acquired by LGT from eubacteria, possibly via plasmids, which have a related protein ParM for their DNA segregation, as proposed for the ‘actin-like’ protein of Thermoplasma and Ferroplasma (Hara et al., 2007). These protein sequences are closer to those of thermophilic endobacteria than to actin or actinobacterial MreBs, so probably came by LGT from an endobacterium long after archaebacteria and eukaryotes evolved. Contrary to Searcy (1987), seized on by Margulis et al. (2005), Thermoplasma actin-like proteins

313

are therefore irrelevant to eukaryogenesis. Ancestral hyperthermophilic Archaebacteria also lost the related Hsp70; secondarily mesophilic lineages reacquired it from eubacteria by LGT.

7. Prerequisites for evolving phagotrophy Phagotrophy requires binding prey to the predator’s surface glycoproteins, pseudopodial engulfment by actomyosin and lipid modification, successive processing of phagosome contents by vesicle fusion, e.g. of acidosomes, fusion with lysosomes, digestion by lytic enzymes, active import of products to the cytosol, recycling many membranes to other compartments by vesicle budding and fusion, and of the residual digestive vacuoles to the cell surface by exocytosis. Evolution of efficient phagocytosis depended on thousands of new genes making the endomembrane system and cytoskeleton, many of unknown functions. The complexity is such that we will never reconstruct in detail every pathway by which it evolved, because so many things had to happen in parallel. The best we can do is to distil its central logic and suggest plausible key intermediates (Fig. 2). Much of its present complexity concerns efficiency that would not have been important or even possible initially, when core functions began. Central is well-controlled vesicle budding and fusion and actomyosin motility; neither ever occurs in bacteria. Both so intertwine that each must have started simply and become more complex together. Both are controlled by small GTPases, whose multiplication into numerous paralogues was a key concomitant and partial cause of endomembrane differentiation. Contrary to past ideas, small GTPases are not restricted to eukaryotes, but occur throughout bacteria as four different paralogues (Dong et al., 2007b). As eukaryotes arose from an early neomuran descendant of actinobacteria, all most likely evolved from small RarD GTPases (present only in actinobacteria and archaebacteria) and thus were vertically inherited by the eukaryote host. The idea that they came from the less similar BglA GTPase family found widely in negibacteria via cell fusion with a ␦-proteobacterium (López-García and Moreira, 1999, 2006; Moreira and López-García, 1998) is unparsimonious, superfluous and entirely implausible. The fact that eukaryotic paralogues form two branches on a GTPase tree (Dong et al., 2007b) is probably a phylogenetic artefact caused by their markedly contrasting divergence from a common ancestor, not reliable evidence that only Sar/Arf came from actinobacterial RarD and Ras/Rab/Rho from ␣-proteobacterial BglA (less like Rabs), or for the cytologically untenable fusion theory or even LGT. No evidence contradicts the idea that phagocytosis and endomembranes evolved purely autogenously in an early neomuran prior to mitochondria, without input from LGT or any more complex history. There is a marked chicken-and-egg problem: specific membrane budding would lack an obvious advantage without specific fusion to target the vesicle, and vice versa. Since each involves several coadapted proteins, how did either start? The explanation previously adopted was that actomyosin accidentally internalised surface membrane, and crude membrane re-fusion with the surface membrane arose prior to the origin of endomembrane vesicle budding (Cavalier-Smith, 1987b, 2002e). Because phagocytosis internalises membrane it cannot have evolved without means for recycling it to the plasma membrane, i.e. exocytosis. The whole process was so complex that the neomuran that started evolving phagotrophy for predation must have been preadapted as a non-phagotrophic predator that digested its prey externally. Several such bacterial predators exist today, e.g. myxobacteria, Vampirococcus, Daptobacter (Guerrerro et al., 1987), but all are negibacteria, bounded by a murein wall and a second outer membrane, whose rigidity and complexity prevented

314

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

Fig. 2. Logic and phagotrophic origin of the eukaryotic endomembrane system. (a) Modern cells. Membrane lipids and proteins are made in the nuclear envelope/ER, which is divided into rough regions (RER) bearing signal-recognition-particle receptors and therefore ribosomes, where alone cotranslational protein synthesis/glycosylation occurs, and smooth regions whence Golgi-destined COPII vesicles bud. From the trans-Golgi network (TGN) bud clathrin-coated vesicles for supplying digestive enzymes for lysosomes (L) and endosomes (E); phagosomes (P) enclosing prey phagocytosed by actomyosin pseudopodia (A) are acidified by acidosomes (a) and fuse with lysosomes; endosome-destined clathrin-coated vesicles bud from plasma membrane coated pits; TGN non-clathrin vesicles bud for exocytosis (secretion and plasma membrane growth). Peroxisomes (Per) divide and/or form from vesicles budded from ER; COPI vesicles budding from Golgi cisternae mediate retrograde transport towards the ER. (b–c) Three hypothetical stages in evolution of prekaryote endomembranes. (b) Predation probably began by prey adhesion to the surface and secretion of externally bound digestive enzymes; prey lysis liberated proteins (Pr). The first simplest form of intracellular digestion perhaps involved just surface membrane protein-import channels (Derlin; grey ovals) and associated 26S proteasome digestion chambers (top right). Soon thereafter, actin-driven pseudopodia (A) sometimes accidentally fused, enclosing prey within a phagosome (P). After digestion, the phagosomal membrane could be re-fused with the plasma membrane by ancestral V-SNAREs (black) and t-SNAREs (grey). (c) The second phase of endomembrane stabilisation involved evolution of the first undifferentiated coated vesicle (cv) budding by a novel protocoat (ancestral to COPI and adaptins) and pre-existing SNAREs (not shown). By chance some protoendomembranes carried DNA, some TAT protein translocation machinery (black rectangle; protoperoxisomes: pPer) and some Sec61 channels, SR and Derlin (protoER/Golgi: pER/Golgi). As coated vesicle driven growth of the plasma membrane increasingly predominated over direct return of phagosomal membrane, the plasma membrane became depleted of ribosome receptors (SR) and DNA attachment proteins; when totally devoid of them peroxisomes, endomembranes, peroxisomes and DNA were permanently internal and absent from the surface membrane. Proteasomes associated with the ER and ERAD/ubiquitin controls evolved, quality and cell cycle control superseding their original digestive function. (d) In the third phase of endomembrane differentiation, divergence of vesicle coats into COPII and adaptin/clathrin (and of cognate SNAREs) made separate ER and protoGolgi. Evolution of retrograde transport by COPI sharpened that differentiation. Finally homotypic COPII fusion (H) generated Golgi cisternae (specialising in complex saccharide and lipid synthesis) distinct from the TGN; COPI duplications and divergences to COPIa,b differentiated proximal and distal cisternae (Donohoe et al., 2007), and new structural proteins stacked them. Separate homotypic copII fusion made the nuclear envelope and pore complexes (Fig. 3), after which many proteasomes functioned within the nucleus also. For additional discussion see Dacks and Field (2007), Gürkan et al. (2007).

any of them evolving phagocytosis and internal digestion. If however a neomuran prekaryote evolved a similar capacity to bind and digest bacteria externally, it would have been preadapted for the difficult transition to phagocytosis. Initially its glycoprotein surface coat could be modified to bind prey, enabling secreted digestive enzymes and membrane destabilisers to kill and digest the prey and surface membrane transporters to import the products. At that stage digestive enzymes would have been secreted cotranslationally across its single surface membrane, and cotranslationally glycosylated by oligosaccharyl transferase associated with the ribosome, transferring oligosaccharides from their isoprenoid carriers to N-asparagines on the nascent protein, this cooperation having evolved in the ancestral neomuran before prekaryotes and

prearchaebacteria diverged. Digestive enzymes probably attached to the cell surface by membrane anchors to prevent waste through diffusing away uselessly. A central inefficiency of such external digestion is the loss of much digested material to the environment and competitors before absorption. This is reduced if the predator is smaller than its host, bores into its cytoplasm, and bathes in a soup of digestion products still enclosed by prey membranes, as does Daptobacter (Guerrerro et al., 1987). Though not a generalised answer to the problem as it limits predator size, constraining it to extra large prey, when most potential prey is too small, it could have enabled efficient evolution of prey binding and external digestion before phagocytosis arose, facilitating internal digestion and size increase later.

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

One can envisage parallel ways for predation to improve gradually by evolving several characteristic eukaryotic properties. One is to control cotranslational secretion, secreting enzymes only in response to prey adhesion to the membrane carrying ribosomes making them, not constitutively. Conformational changes in integral membrane glycoproteins binding prey externally could achieve this by directly affecting the signal-recognition-particle receptor (SR) on the cytosolic face of the membrane. Possibly this was why eukaryotes, but not archaebacteria, evolved SR␤, a small GTPase which can switch between binding to SR␣, allowing it to transfer a secretory protein into the trimeric Sec61 ER membrane channel, or alternatively forming homodimers that prevent translocation because they cannot bind SR␣ (Schwartz et al., 2006). SR␤ has an N-terminal hydrophobic helix attaching it to the membrane. SR␤ could have been ancestral to related GTPases, Sar1 and ARF, which initiate coat assembly on membrane transport vesicles (Kahn et al., 2006; Pasqualato et al., 2002), but evolved only after phagocytosis later internalised parts of the cell surface as protoendomembranes; both modified the membrane-embedded helix, Sar1 only attaching to membranes in its GTP, not GDP state, and Arfs adding an N-terminal myristoyl fatty acid also as membrane anchor. Arfs positively curve membranes by GTP-driven insertion of their Nterminal amphipathic helix into the lipid bilayer (Lundmark et al., 2008). I suggest that prior to phagocytosis four other eukaryotic novelties evolved to improve digestion and protein absorption from surface-bound prey: actomyosin-based pseudopodia, membrane tubulation, ER-associated degradation of proteins (ERAD), and eukaryotic 26S proteasomes. Any of these would benefit predation via many fewer innovations than the endomembrane system and phagocytosis, and be easier to get started.

8. Internal digestion first by retrotranslocation and 26S proteasomes? The ERAD system and proteasomes are coadapted for extruding unfolded proteins from ER lumen into the cytosol, concomitantly polyubiquitinating them, then passing them to 26S proteasomes for degradation within (Brodsky, 2007; Li et al., 2008; Lipson et al., 2008; Raasi and Wolf, 2007; Tamura et al., 2008). If this coadaptation evolved whilst ribosomes were still attached to the cell surface membrane, it would take partly digested prey proteins from outside the cell and digest them inside with 100% retention, necessarily far more efficient than external digestion with some inevitable loss. This idea provides the first obviously extremely strong selective force for evolving ERAD, 26S proteasomes, and polyubiquitination, much stronger and more basic than quality control over accidentally misfolded proteins or temporal controls over the cell cycle that they now mediate. Moreover, as the function was just converting food protein into amino acids, digestion alone sufficed. There would be no need also to evolve all or any other processes necessary for use of this machinery in cellular controls, previously postulated as their original function (Cavalier-Smith, 2002e). The basic proteasomal degradative machinery probably had to evolve its present complexity before its co-option for multifarious controls. What is more basic than simple digestion of prey proteins? ERAD depends on a small integral membrane protein Derlin1, possibly constituting the ER membrane channel for extrusion; the 19S regulatory subunit of the 26S proteasome is the only cytosolic factor essential for translocating non-ubiquitinated substrates (Wahlman et al., 2007). Any soluble external protein could be imported and digested just by Derlin and cytosolic proteasomes. Much simpler than phagocytosis, this would efficiently digest extracellular proteins internally. Modern cells use numerous different motifs to select different substrates, mostly polyubiq-

315

uitinated before entry into the proteasome. But such complex polyubiquitin selection is necessary only because ERAD now just exerts quality control over the cell’s own newly made proteins and must not digest the wrong thing. It would have been easier to evolve the basic machinery when any external protein should be digested. Adding polyubiquitin tags by polyubiquitin kinases helped by ATPase Cdc48 was a refinement only necessary after the endomembrane system formed, when ER became distinct from endosomes and lysosomes and phagocytosis replaced retrotranslocation/proteasomes for internal prey digestion. Relieved of that responsibility, ubiquitin-controlled proteasomes were recruited piecemeal for disparate cellular controls by proteolysis. An origin of ubiquitin-controlled degradation simply to make predation initially more efficient, explains why archaebacteria, which never became predators, did not evolve it, why the 19S proteasomal regulatory subunit became complex (Chen et al., 2008; da Fonseca and Morris, 2008; Lipson et al., 2008; Nickell et al., 2007), and why protease subunits underwent gene duplication and divergence in eukaryotes alone: increased protease diversity quickly increased digestive capabilities for utilising more prey proteins. Later, after phagocytosis and lysosomes took over prey digestion, different subunits could digest different cellular proteins, enabling full modern control complexity to evolve easily. Derlin probably evolved from distantly related bacterial membrane proteins (e.g. ABJ85348) of mostly unknown function.

9. Pseudopodia, prey uptake, and membrane recycling By evolving actin and Arps to branch it, the prekaryote could extend pseudopods partially around the prey cell, much increasing contact area and thus absorption efficiency of digestion products. Today Rho small GTPases (e.g. Rac) control actin remodelling, both for extending phagocytic pseudopods and prey internalisation. Rho can weakly group on sequence trees with related Rab GTPases involved in transport vesicle target specificity (Jékely, 2004), which may also have evolved from SR␤ by losing its membrane-attaching peptide, getting lipid tails instead for membrane attachment. Rabs all are prenylated and associate with diverse endomembranes; Rac is attached to the plasma membrane by phosphatidic acid. Outer leaflet phosphatidylinositol was a key preadaptation of the prekaryote, derived from actinobacteria (the only bacteria preadapted for phagocytosis by having phosphatidylinositol) as diverse phosphoinositides are vital for membrane traffic. Phosphatidylinositol 3-phosphate is essential for phagocytosis and endosomal trafficking, phosphatidylinositol 4-phosphate regulates trans-Golgi secretion, phosphatidylinositol 4,5-bisphosphate exocytosis, and phosphatidylinositol 3,5-bisphosphate late endosome/multivesicular body trafficking (Deli et al., 2008). Surface membrane ARF and actomyosin could probably together tubulate the membrane even before microtubules evolved (ARF, microtubules and kinesin alone are sufficient for tubulation: Jékely, 2004). Invaginating surface membrane as narrow tubules would usefully magnify surface area for protein uptake by Derlin and amino acid active import. It is implausible that surface invaginations evolved instead to increase efficiency of secreting digestion enzymes to the prey (Jékely, 2004); direct secretion at the surface would be more efficient. It seems odd to accept the prekaryote as predatory, but deny that internal digestion (entirely novel for eukaryotes) had the key creative role in eukaryogenesis, and to argue instead that enzyme secretion, which prokaryotes have done perfectly well for 3.5 Gy, was the primary force in eukaryogenesis (Jékely, 2004). Jékely (2004) did that because he interpreted his GTPase sequence tree as implying that secretion evolved before phagocytosis; that conclu-

316

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

sion was unjustified (see Cavalier-Smith, submitted for publication for more details). Actin evolved from MreB, present in the neomuran ancestor, but not in most archaebacteria, whose ancestor lost it as well as phosphatidylinositol, sterols, and Hsp70 chaperones (related Hsc70 is essential for clathrin uncoating), all prerequisites for evolving endomembranes, making archaebacteria implausible ancestors for eukaryotes. Phagocytosis also needs myosin. Together with its distant relative kinesin, myosin probably arose by domain fusion between ATPases derived from a bacterial GTPase (Leipe et al., 2002), forming the head, and a coiled-coil protein forming the tail. Plausible coiled-coil tail ancestors are Smc proteins universally important for chromosome segregation, from which eukaryotic cohesins (holding sister chromatids together for proper disjunction) and condensins (compacting chromatin) evolved (Cortes-Ledesma et al., 2007; Hirano, 2005, 2006; Nasmyth and Haering, 2005). Myosin tails diversified into at least three kinds before the eukaryotic cenancestor (Odronitz and Kollmar, 2007; Richards and Cavalier-Smith, 2005); myosin II mediates phagocytosis in unikonts, but myosin I does in bikonts; probably the first phagocytic myosin was less differentiated. Coiled-coil motifs are rare in bacteria but important in many eukaryotic structural proteins, e.g. tethers anchoring vesicles to target membranes before fusion, SNAREs that mediate specific fusion (Antonin et al., 2002; Bassham and Blatt, 2008; Cai et al., 2007; Fasshauer et al., 1998; Liu et al., 2008; Mima et al., 2008), formins for positioning actin, and intermediate filaments. I suggest that SNAREs first evolved in the prekaryote to help phagosomes re-fuse more efficiently with the plasma membrane (Fig. 2b). Accidental fusion of pseudopodia would inevitably often internalise prey. After digestion by enzymes made by ribosomes attached to its bounding membranes, the spent phagosome could accidentally re-fuse with the plasma membrane, but this would be inefficient without fusogenic proteins (SNAREs), probably a key initial innovation to make internal digestion more efficient. Thus, primitive exocytosis to recycle phagosome membrane to cell surface probably evolved before internal membrane budding. It was easier for SNAREs and Rab-based GTPase controls to evolve then with only one possible target membrane, the plasma membrane. Rabs themselves are controlled by activators, with three classes present in the cenancestor; the activator of Rab4 that controls the last step in exocytosis (Novick et al., 2006), possibly first to evolve, was a catalytic coiled-coil, Sec2p (Dong et al., 2007a), reemphasizing the importance of novel coiled-coil proteins in eukaryogenesis. As soon as phagosome formation by actomyosin and recycling by SNAREs evolved, coated vesicles could evolve for pinocytosis of proteins liberated by external lysis of prey by enzymes made by ribosomes still present on the inner face of the surface membrane (Fig. 2c). Thus, early predators had three ways of internalising food. The first endocytic vesicle coats were probably the common ancestor of COPI coats (now used for retrograde transport from Golgi (Donohoe et al., 2007), which did not then exist), and clathrin coats (now used for plasma membrane, endosome and trans-Golgi budding); COPI coats are homologous to adaptins that bind clathrin (Stagg et al., 2007). The target membrane for vesicles generated by these protocoats was the primitive phagosome. Even random fusion of endocytosed vesicles promiscuously with phagosomes and plasma membrane, despite waste if the latter, would have been beneficial, the more so the higher the proportion of internal membrane. For explaining evolutionary transitions it is important that simple uncontrolled versions of processes have strong advantages; efficiency is initially irrelevant, but could immediately improve if duplicated SNAREs that preferentially targeted phagosome rather than surface membrane associated specifically with the coated

vesicles. If different membranes to be targeted increased one at a time and selectivity is advantageous, as on my scenario, its evolution is comprehensible. This stage would have worked well, with temporary phagosomes ancestral ultimately to endosomes, lysosomes (which now digest proteins), peroxisomes (where D-amino acids and fatty acids are digested) and ER (where proteins and phospholipids and sterols are now made, and to which DNA is attached) and Golgi (where complex carbohydrates and sphingolipids are made). Final constriction of clathrin-coated vesicles requires dynamin GTPase, already present in bacteria (Low and Löwe, 2006), including posibacteria. Spiral arrays of dynamins divide membranes (Mears et al., 2007); duplications generated many paralogues, eventually recruited for dividing endosomes, vacuoles, peroxisomes, and finally mitochondria. Prey was probably not only bacteria; viruses are more abundant (Danovaro et al., 2008) and protozoa eat them (González and Suttle, 1993; Hennemuth et al., 2008); probably some manipulated early endocytosis for entering the host, which perhaps even got useful genes from them. 10. DNA internalisation, copII and the origin of permanent endomembranes Recycling phagosomes to the surface was originally inevitably inefficient. Some membranes failed to re-fuse, becoming potential endomembranes, but without division would not be perpetuated genetically; dynamin provided the division mechanism – the more they fragmented into numerous small pieces the more random segregation could maintain them. Some of these membranes would bear chromosomes; until mitosis evolved, random segregation would make some DNA-less cells that would die and many multigenomic cells (ineffective for making numerous paralogues in the absence of efficient multichromosomal segregation, so individual gene duplication, not whole genome duplication, probably mainly supplied the burst of new paralogues). Primitive dynamins dividing membranes and coats budding them, neither as specific as now, would produce a plethora of internal membranes of disparate protein composition (and lipid composition unless lipid synthesising enzymes were distributed uniformly). Poorly specific coat proteins and SNAREs would equilibrate lipids and proteins among these endomembranes. As specificity developed through initial random preferences among paralogues of the dividing, budding, and fusion machineries, different endomembrane types would be stabilised. Such differentiation also required physical links among the evolving membranous organelles, because such attachments, and between them and the cytoskeleton, are central to eukaryotic cell heredity. Attachments linking all organelles are striking exemplified in the small flagellate Sainouron (Cavalier-Smith et al., 2008a), and by specific binding of centromeres by nuclear envelope protein to microtubules in yeast (King et al., 2008). Not only clathrin vesicle adaptins and COPI coats, but also nuclear pore complexes (NPCs) and COPII coats that bud from ER have a common origin, as they all belong to a major superfamily of ␣-solenoid/␤-propeller proteins (Devos et al., 2004, 2006), never found in bacteria, whose origin as novel membrane-curving machines arguably began in early phagotrophs as outlined above. What made endomembranes permanent was the origin of secretory vesicle budding from endomembranes such that the vesicles excluded both ribosome- and DNA-attachment sites (CavalierSmith, 1987b, 2002e). If vesicles carried SNAREs for plasma membrane fusion with them, they would be an alternative means of recycling phagosome membrane to the cell surface. The combination of recycling biased against ribosome- and DNA-attachment sites and continued phagocytosis would steadily deplete the sur-

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

face of ribosome- and DNA-binding sites until all were lost from it, remaining only on endomembranes. New phagosomes would thereafter lack such binding sites; therefore, whether their membrane was recycled to the surface by direct re-fusion or vesicle budding, the surface membrane could never regain its prokaryotic state bearing ribosomes and DNA. This differentiation between plasma membrane and endomembrane was not directly coded by genes, but was a membrane mutation caused by permanent receptor loss from the cell surface that no DNA mutation could reverse (Cavalier-Smith, 2004b). A new era of cell evolution had begun, but the cell was still not eukaryotic, as a nuclear envelope did not yet exist. One may not be able to equate the original vesicle coat with any modern component because the system must then have been much less differentiated. There are two obvious possibilities. Ribosome depletion might have been achieved by the very first coat to evolve, as suggested above for pinocytotic feeding. Even if initially inefficient and more indiscriminate for both budding and fusion, with vesicles fusing with and budding from both surface membrane and phagosome, this would provide a quantitative balance between both membranes; just by mass action an excess of internal membrane caused by phagocytosis would bias transfer, ensuring net movement from endomembrane to cell surface. The second possibility is that duplication of coats and SNAREs and divergence in binding properties separated more polarised secretory and endocytic paths. If such duplication of vesicle targeting did not itself make a permanent endomembrane, it probably almost immediately followed it to make digestion more efficient and ensure unimpeded plasma membrane growth after ribosome internalisation prevented them inserting membrane proteins directly into the plasma membrane. The logic of evolutionary endomembrane differentiation requires mechanistic linkage of coat-biogenesis of compartmentspecific trafficking vesicles and of their targeting machinery (SNAREs); otherwise their evolution would be impossible. The simplest explanation is that SNAREs bind directly and specifically to cognate coats, ensuring their automatic targeting to membranes already having the same SNAREs; this perpetuated target identity – a form of membrane heredity depending on protein complementarity (Cavalier-Smith, 2004b). T-SNAREs would remain in the target membrane, but V-snares must be recycled to the donor. Heinrich and Rapoport (2005) showed that such complementarity can automatically generate distinct membrane compartments if binding strength of cognate SNAREs and coats is substantially greater than for non-cognate ones. Probably >23 different coat/SNARE partners evolved before the cenancestor (Kloepper et al., 2007). I suggest that primary endomembrane differentiation was into internal membranes bearing ribosomes and DNA (protoER/NE/Golgi making proteins and lipids) and some not (protoendosomes), both initially generated automatically by phagocytic internalisation. Once surface ribosome depletion made endomembranes permanent, phagosomes would have to fuse with protoendosomes bearing digestive enzymes. To make this efficient two different kinds of membrane budding arose from the protoER/Golgi: clathrin-coated vesicles targeted to endosomes and a largely uncharacterised system for exocytotic secretory vesicles. One widespread exocytosis method involves Avl9, a universal eukaryotic protein with distant relatives involved in wall attachment of mainly posibacteria (Harsay and Schekman, 2007). Unfortunately it is unknown if Avl9 acts in trans-Golgi network vesicle budding, as does Chs5/6 coat complex for budding secretory chitosomes in higher fungi (Sanchatjate and Schekman, 2006). This supposes a relatively later origin of COPII budding (Fig. 2d) than once suggested (Cavalier-Smith, 2002e, 2004b) because relatively efficient SNAREs for exocytosis must exist before COPII could evolve (otherwise it would lack selective advantage) and the

317

simplest way of evolving them is for recycling phagosomal membranes, outlined above. This exemplifies for membrane trafficking the established principle that complex metabolic pathways evolve more easily backwards (Horowitz, 1945), because this requires no evolutionary foresight. Differentiated endomembrane compartments become permanent only if they divide, e.g. by dynamin, or can be made de novo by homotypic membrane fusion. They must also segregate to both daughters. Protoendosomes perhaps initially randomly segregated with sufficient accuracy given high copy numbers. The low copy genetic/biosynthetic compartment evolved specific binding to microtubules or the ␥-tubulin-containing centrosomes that nucleate them: most important for the chromosome, but rough ER could bind to chromatin surfaces by specific proteins for segregation. ER segregation was probably an early reason for thus originating a nuclear envelope, which does not merely surround but binds to chromatin. Another was probably protection of DNA from shearing by the new molecular motors, initially myosin that first helped phagocytosis, then vesicle transport and cytokinesis. Origins of the nucleus and mitosis are discussed in detail elsewhere (CavalierSmith, submitted for publication); I lack space to repeat arguments, but Fig. 3 summarises key points. Homotypic fusion of COPII vesicles not only generated the nuclear envelope and NPCs, but also the Golgi (Fig. 2d–a); as can still happen in modern cells. Homotypic COPII fusion can generate the compartment intermediate between ER and Golgi in many opisthokonts and plants (Bentley et al., 2006). However, this intermediate compartment possibly did not exist in the cenancestor. Many small protozoa of diverse phyla lack it; their Golgi attaches directly to a smooth region of the NE, acting as transitional ER (Cavalier-Smith et al., 2008a,b), possibly the ancestral state for eukaryotes rather than multiple secondary simplifications. Retrograde COPI transport probably evolved later to increase retention efficiency for ER functions, as did numerous other coated-vesicle budders and cognate SNAREs.

11. Peroxisomes as endomembrane digestive differentiations Peroxisomes form both by dynamin-powered division of preexisting peroxisomes (Hoepfner et al., 2001) and by ER budding of precursors containing the membrane protein PEX3 (Hoepfner et al., 2005), which imports other peroxisomal proteins, endowing peroxisome proteins with their organelle identity. PEX3 is also the basis of peroxisomal membrane heredity when they form by division (Cavalier-Smith, 2004b). Therefore, peroxisomes probably originated from the protoendomembrane by budding and differentiation as originally proposed (Cavalier-Smith, 1975) and discussed in detail (Tabak et al., 2006). An origin by endosymbiosis (Cavalier-Smith, 1990) is wrong; they are probably not irreplaceable genetic membranes (Cavalier-Smith, 2004b; De Duve, 2007), unlike mitochondria, plastids and ER. Peroxisomes made prey digestion more efficient by segregating fatty acid ␤-oxidation enzymes and D-amino acid oxidases (for digesting hydrolytic products of eubacterial murein walls) plus catalase for destroying their harmful byproduct, hydrogen peroxide. The PEX system for importing such soluble proteins probably evolved from the eubacterial TAT system (Cavalier-Smith, 2006a), accidentally removed from the cell surface and segregated into ancestral peroxisomes (not ER) by phagocytosis, as TAT and PEX both import proteins post-translationally fully folded after recognising related C-terminal targeting signals. Thus, the two contrasting ancestral neomuran protein-targeting systems Sec and TAT segregated respectively into ER and peroxisomes, two different autogenously evolved respiratory organelles; the ER retained the actinobacterial surface membrane cytochrome

318

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

Fig. 3. Origin of the cytoskeleton, mitosis (left) and nucleus (right). (a) By the time peroxisomes (P) and permanent endomembranes (EM) evolved and DNA was thereby permanently internalised (Fig. 2c), early mitosis must also have evolved to segregate the chromosome/EM/P complex, generating a prekaryote with separate kinesin and myosin motors. (b) Dynein evolution improved segregation and preadapted the cell for ciliary origins; EM attachments around DNA made a primitive NE. (c) Adding nuclear pore complexes (NPCs), mitochondria and cilium made the cenancestral eukaryote. (d–f) Two-phase origin of the nuclear envelope from protoER and of trans-envelope transport. (d) Phase I: nucleoporins (Nups) forming the octagonal cylindrical scaffold evolved by duplications of coat proteins of COPII secretory vesicles with ␣-solenoid and/or ␤-propeller domains, being attached by integral membrane Nups descended from actinobacterial membrane proteins; the nucleoporins separated the outer membrane domain bearing ribosome receptors (SR) and the inner membrane domain bearing integral membrane chromatin binding proteins (grey rectangles). (e) Phase II: NPC lumens were narrowed by FG-repeat-rich Nups, preventing passive diffusion of macromolecular complexes and mediating active specifically targeted nucleocytoplasmic exchange by three carrier complexes (karyopherins; ribosome subunit and mRNA transporters). (f) Phase I in surface view, showing complete Ran GTPase-mediated fusion of RER cisternae prevented by COPII coat proteins (black blobs) remaining in place to become octagonal NPC scaffolds. Modified from Cavalier-Smith (submitted for publication), which gives the detailed rationale.

P450 system for oxidative sterol synthesis. A novel coat type might soon be discovered for budding PEX3 precursors. Peroxisomes also evolved a novel multispanning integral membrane ADP/exchange protein to use fatty acid oxidation energy better. 12. Phagotrophy and novel protein targeting enabled symbiogenetic organelle additions Arguably the peroxisomal ADP/ATP exchanger was the ancestor of related mitochondrial, and thereby secondarily of plastid, inner membrane carriers, which had the primary role in later enslaving ␣-proteobacteria and cyanobacteria. Central to mitochondriogenesis was the origin of novel protein-import machinery, partly from ␣-proteobacterial outer membrane proteins and partly from host proteins; see Cavalier-Smith (2006b, 2007a). Initially the host probably had oxidative phosphorylation in its surface membrane, which it could lose after enslaving the ␣-proteobacterium. Acquisition of mitochondria permanently added no novel metabolism to the host, but made its phagotrophy and respiration more efficient through compartmentation (Cavalier-Smith, 2002e, 2006b, 2007a).

Incidentally it added harmful genetic parasites: group II introns that after moving to the nucleus became spliceosomal introns that rampantly invaded genes (Cavalier-Smith, 1991c). Though increasing transcriptional costs, introns did not increase replication costs because nuclear genome sizes are not minimized by selection (unlike bacterial genomes) but coadapted to cell volume by positive skeletal functions of nuclear non-coding DNA caused by nuclear envelope origins (Cavalier-Smith, 2005). Ubiquitin is essential for spliceosomal assembly (Bellare et al., 2008), showing that spliceosomes evolved after endomembranes, so mitochondria also evolved later. Only prior evolution of phagotrophy, cytoskeleton and endomembranes, as Stanier (1970) cogently argued, made endosymbiosis and mitochondrial enslavement possible. Later phagocytosis allowed a bikont eukaryote to enslave a cyanobacterium to make chloroplasts and the plant kingdom, in which as for mitochondria the negibacterial outer membrane was retained for ever, its envelope protein-export apparatus being modified for protein import with transit sequences mediating targeting via Toc/Tic machinery. In contrast to mitochondria, plastids did add

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

novel metabolism (oxygenic photosynthesis) but also increased compartmentation efficiency, e.g. by replacing cytosolic fatty acid synthetases by plastid/cyanobacterial ones. Soon after plants diversified into three lineages, enslavement of a red alga by another bikont host generated chromalveolates, whose algal members (e.g. diatoms, haptophytes, brown algae) dominate the oceans and have the most complex cell structures known (Cavalier-Smith, 2003a, 2004a, 2007c). Their chloroplasts are not in the cytosol but inside the former phagosomal membrane (alveolates) or RER (chromists). The host signal and transit machineries were co-opted to target nuclear-coded proteins to plastids by bipartite targeting signals (Cavalier-Smith, 1999, 2003a). Intriguingly the ancestral chromalveolate duplicated the ancestral ERAD machinery, which I argued above first made prekaryotes efficient predators, and then apparently co-opted it for exporting nuclear-coded proteins from the ER lumen across the periplastid membrane (former red algal plasma membrane) into the periplastid space (its former cytosol) (Maier, in press). Thus throughout eukaryote megaevolution basic principles of membrane heredity and protein targeting into and across membranes apply. Innovations in them were keys for making more complexly compartmented cells, irrespective of whether innovations were purely autogenous (e.g. origin of endomembranes, peroxisomes and nucleus: the primary eukaryotic features) or symbiogenetic, as in the later refinements of mitochondria, chloroplasts, and chromalveolate periplastid membrane.

13. Cilia, a novel compartment not delimited by membranes Cilia, centriole-nucleated organelles requiring 1000 genes for biogenesis, also evolved in the ancestral eukaryote together with the nucleus (Cavalier-Smith, 1987b, 2002e) autogenously – without any contribution from symbiogenesis, contrary to the hypothesis of Sagan (1967). Her spirochaete theory of the origins of cilia and mitosis is incompatible with all we know of their cell biology; earlier predictions are disproved and nothing explanatory or testable remains in the latest version (Margulis et al., 2005). A recent absurdly complicated symbiotic ‘explanation’ (Li and Wu, 2005) is phylogenetically unmerited and cell biologically implausible. Mechanistically how a spirochaete could lose its membranes is unexplained as are any advantages of transitional changes. Planctobacterial tubulins are irrelevant to eukaryogenesis, being LGTs from eukaryotes (their membrane invaginations are also irrelevant: Cavalier-Smith, submitted for publication). Ciliary origin by gene duplication of pre-existing tubulin to ␦, ␧, ␩ tubulins for centrioles and of kinesins, dyneins and numerous other cytoskeletal proteins and recruitment of centrosomal microtubules that evolved autogenously for mitosis in the earliest eukaryote is phylogenetically, mechanistically and selectively almost infinitely simpler. Very likely kinesin-based ciliary gliding preceded swimming and feeding by cellwards movement by dynein of prey trapped on the cilium preceded motility of the cell (Cavalier-Smith, in press). Dynein evolved from a bacterial ATPase (Iyer et al., 2004); duplication made 20 dyneins in the eukaryote cenancestor largely for cilia (Wickstead and Gull, 2007; Wilkes et al., 2008); higher plants lost cilia, then all dyneins. Nine transition fibres attach centrioles orthogonally to the plasma membrane, ensuring that outer doublet microtubule growth makes a projecting cilium. They also make the ciliary lumen a separate compartment, gate the entry of ciliary proteins into it, and exert quality control over them (Stephan et al., 2007). Ciliary transport particles that move precursors up the cilium are kinesin-driven and consist of a complex of ␣-solenoid/␤-propeller proteins related to and doubt-

319

less evolutionarily derived from vesicle COP/adaptins (Jékely and Arendt, 2006). Thus prior endomembrane evolution was prerequisite for the origin of a separate ciliary compartment, despite it not being delimited by membrane, invalidating the hypothesis that symbiotic origin of cilia initiated eukaryogenesis (Margulis et al., 2005). For the autogenous origin of cilia see: Cavalier-Smith (1978, 1982b, 1987b, 1992), Jékely and Arendt (2006), Mitchell (2007). 14. The botched, piecemeal nature and rapidity of megaevolution ‘When this adaptation [to ‘a new and peculiar line of life’] had once been effected, and a few species had thus acquired a great advantage over other organisms, a comparatively short time would be needed to produce many divergent forms, which would spread rapidly and widely throughout the world’ Darwin (1872). Darwin would have been fascinated how disparate bacterial molecular machines were modified so dramatically during eukaryogenesis to create the first predators on earth to digest food internally. He rightly believed that all evolution was by piecemeal modification of inherited structures, sometimes generating bizarre awkwardness as no intelligent designer would. The four membranes around chromalveolate chloroplasts are relics of successive historical accidents and makeshift jobs, not elegant design. So is the oddly complicated endomembrane system, with dozens of different vesicle-targeting machineries that evolved to make the best of a bad job when the cell surface budded off as protoendomembranes and phagocytosis-driven ribosome depletion risked preventing surface-membrane growth. These and other cack-handed properties, e.g. the conserved sorting pathway in mitochondria, where periplasmic proteins cross two membranes, then recross the second one instead of being imported directly across one as any rational designer would effect, evidence the botched way that descent with modification generates complexity. Darwin’s quotation shows he did not believe evolution to be steadily paced, but thought it must be especially fast during perfection and early radiation of a new body plan. Simpson (1944, 1953) found strong evidence in the fossil record for just such extremely rapid quantum evolution, as he called it, and geologically sudden radiations for virtually every new animal group. Quantum evolution markedly applies to the origin of major body plans distinguishing taxa ranked as order of higher: what Simpson (1944) named megaevolution to distinguish it from qualitatively less dramatic normal macroevolutionary divergence. 15. Conclusions Darwin’s and Simpson’s ideas on quantum and megaevolution apply exceptionally forcefully to eukaryogenesis (Cavalier-Smith, 2006a). I have tried to explain here how eukaryotic intracellular digestion, though now very complicated, could have evolved in simple stages following a few initial key mutations that allowed proteins from externally lysed prey to be individually internalised for proteasomal digestion prior to the origin of the endomembrane system as the indirect consequence of the accidental ingestion of whole prey cells following the origin of the actomyosin motility system. Once started, the postulated cascade of events could have gone to completion and generated a fully functional eukaryotic cell having all the properties of Table 1 with remarkable speed (likely well under a million years: over 1010 generations), with all the less efficient intermediates rapidly becoming extinct (Cavalier-Smith, 1987b). In my view, the key to understanding this dramatic innovation was the specific nature of the precursor cells and the initiating

320

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

mutations and the logic of cytoskeleton/membrane interactions and of stepwise increases in their specificity, all under the overarching selective forces of improving the efficiency of phagotrophy and of the consequentially changed cell cycle. Eating, not sex, generated the complexity of the eukaryotic cell; sex was probably an indirect consequence of the novel cell cycle and of nakedness allowing cell fusion (Cavalier-Smith, 2002b,d,e). No specific environmental stimuli or external interventions are needed to explain that uniquely revolutionary internal cellular upheaval. References Amos LA, van den Ent F, Löwe J. Structural/functional homology between the bacterial and eukaryotic cytoskeletons. Curr Opin Cell Biol 2004;16:24–31. Andersson JO, Sarchfield SW, Roger AJ. Gene transfers from nanoarchaeota to an ancestor of diplomonads and parabasalids. Mol Biol Evol 2005;22:85– 90. Antonin W, Fasshauer D, Becker S, Jahn R, Schneider TR. Crystal structure of the endosomal SNARE complex reveals common structural principles of all SNAREs. Nat Struct Biol 2002;9:107–11. Bapteste E, Charlebois RL, MacLeod D, Brochier C. The two tempos of nuclear pore complex evolution: highly adapting proteins in an ancient frozen structure. Genome Biol 2005;6:R85. Bassham DC, Blatt MR. SNAREs: cogs and coordinators in signaling and development. Plant Physiol 2008;147:1504–15. Beisson J, Wright M. Basal body/centriole assembly and continuity. Curr Opin Cell Biol 2003;15:96–104. Bellare P, Small EC, Huang X, Wohlschlegel JA, Staley JP, Sontheimer EJ. A role for ubiquitin in the spliceosome assembly pathway. Nat Struct Mol Biol 2008;15:444–51. Bentley M, Liang Y, Mullen K, Xu D, Sztul E, Hay JC. SNARE status regulates tether recruitment and function in homotypic COPII vesicle fusion. J Biol Chem 2006;281:38825–33. Bourbon HM. Comparative genomics supports a deep evolutionary origin for the large, four-module transcriptional mediator complex. Nucleic Acids Res 2008;36:3993–4008. Brodsky JL. The protective and destructive roles played by molecular chaperones during ERAD (endoplasmic-reticulum-associated degradation). Biochem J 2007;404:353–63. Cai H, Reinisch K, Ferro-Novick S. Coats, tethers, Rabs, and SNAREs work together to mediate the intracellular destination of a transport vesicle. Dev Cell 2007;12:671–82. Cavalier-Smith T. The origin of nuclei and of eukaryotic cells. Nature 1975;256:463–8. Cavalier-Smith T. The evolutionary origin and phylogeny of microtubules, mitotic spindles and eukaryote flagella. BioSystems 1978;10:93–114. Cavalier-Smith T. The origin and early evolution of the eukaryotic cell. In: Carlile MJ, Collins JF, Moseley BEB, editors. Molecular and cellular aspects of microbial evolution. Cambridge: Cambridge University Press; 1981. p. 33–84. Cavalier-Smith T. The evolution of the nuclear matrix and envelope. In: Maul GG, editor. The nuclear envelope and matrix. New York: Alan R. Liss; 1982a. p. 307–18. Cavalier-Smith T. The evolutionary origin and phylogeny of eukaryote flagella. In: Amos WB, Duckett JG, editors. Prokaryotic and eukaryotic flagella. 35th symposium of the society of experimental biology. Cambridge University Press; 1982b. p. 465–93. Cavalier-Smith T. The origins of plastids. Biol J Linn Soc 1982c;17:289–306. Cavalier-Smith T. Endosymbiotic origin of the mitochondrial envelope. In: Schwemmler W, Schenk HEA, editors. Endocytobiology II. Berlin: de Gruyter; 1983. p. 265–79. Cavalier-Smith T. Selfish DNA and the origin of introns. Nature 1985;315:283–4. Cavalier-Smith T. The origin of cells: a symbiosis between genes, catalysts, and membranes. Cold Spring Harb Symp Quant Biol 1987a;52:805–24. Cavalier-Smith T. The origin of eukaryotic and archaebacterial cells. Ann N Y Acad Sci 1987b;503:17–54. Cavalier-Smith T. The simultaneous symbiotic origin of mitochondria, chloroplasts, and microbodies. Ann N Y Acad Sci 1987c;503:55–71. Cavalier-Smith T. Origin of the cell nucleus. BioEssays 1988;9:72–8. Cavalier-Smith T. Symbiotic origin of peroxisomes. In: Nardon P, Gianinazzi-Pearson V, Grenier AM, Margulis L, Smith DC, editors. Endocytobiology IV. Paris: Institut National de la Recherche Agronomique; 1990. p. 515–21. Cavalier-Smith T. The evolution of cells. In: Osawa S, Honjo T, editors. Evolution of life. Tokyo: Springer-Verlag; 1991a. p. 271–304. Cavalier-Smith T. The evolution of prokaryotic and eukaryotic cells. In: Bittar GE, editor. Fundamentals of medical cell biology. Greenwich, Connecticut: J.A.I. Press; 1991b. p. 217–72. Cavalier-Smith T. Intron phylogeny: a new hypothesis. Trends Genet 1991c;7:145–8. Cavalier-Smith T. Origin of the cytoskeleton. In: Hartman H, Matsuno K, editors. The origin and evolution of the cell. Singapore: World Scientific Publishers; 1992. p. 79–106. Cavalier-Smith T. Evolution of the eukaryotic genome. In: Broda P, Oliver SG, Sims P, editors. The eukaryotic genome. Cambridge University Press; 1993. p. 333–85.

Cavalier-Smith T. Cell cycles, diplokaryosis, and the archezoan origin of sex. Archiv Protistenk 1995a;145:189–207. Cavalier-Smith T. Membrane heredity, symbiogenesis, and the multiple origins of algae. In: Arai R, Kato M, Doi Y, editors. Biodiversity and evolution. Tokyo: The National Science Museum Foundation; 1995b. p. 75–114. Cavalier-Smith T. Principles of protein and lipid targeting in secondary symbiogenesis: euglenoid, dinoflagellate, and sporozoan plastid origins and the eukaryotic family tree. J Euk Microbiol 1999;46:347–66. Cavalier-Smith T. Membrane heredity and early chloroplast evolution. Trends Plant Sci 2000;5:174–82. Cavalier-Smith T. Chloroplast evolution: secondary symbiogenesis and multiple losses. Curr Biol 2002a;12:R62–4. Cavalier-Smith T. Meiosis. In: The Oxford encyclopaedia of evolution. New York: Oxford University Press; 2002b. p. 700–8. Cavalier-Smith T. The neomuran origin of archaebacteria, the negibacterial root of the universal tree and bacterial megaclassification. Int J Syst Evol Microbiol 2002c;52:7–76. Cavalier-Smith T. Origins of the machinery of recombination and sex. Heredity 2002d;88:125–41. Cavalier-Smith T. The phagotrophic origin of eukaryotes and phylogenetic classification of Protozoa. Int J Syst Evol Microbiol 2002e;52:297–354. Cavalier-Smith T. Genomic reduction and evolution of novel genetic membranes and protein-targeting machinery in eukaryote–eukaryote chimaeras (meta-algae). Phil Trans Roy Soc Lond B 2003a;358:109–34. Cavalier-Smith T. Protist phylogeny and the high-level classification of Protozoa. Eur J Protistol 2003b;39:338–48. Cavalier-Smith T. Chromalveolate diversity and cell megaevolution: interplay of membranes, genomes and cytoskeleton. In: Hirt RP, Horner DS, editors. Organelles, genomes and eukaryote phylogeny. London: CRC Press; 2004a. p. 75–108. Cavalier-Smith T. The membranome and membrane heredity in development and evolution. In: Hirt RP, Horner DS, editors. Organelles, genomes and eukaryote phylogeny. London: Taylor & Francis; 2004b. p. 335–51. Cavalier-Smith T. Economy, speed and size matter: evolutionary forces driving nuclear genome miniaturization and expansion. Ann Bot 2005;95:147–75. Cavalier-Smith T. Cell evolution and earth history: stasis and revolution. Phil Trans Roy Soc B 2006a;361:969–1006. Cavalier-Smith T. Origin of mitochondria by intracellular enslavement of a photosynthetic purple bacterium. Proc Roy Soc B 2006b;273:1943–52. Cavalier-Smith T. Rooting the tree of life by transition analysis. Biol Direct 2006c;1:19. Cavalier-Smith T. The chimaeric origin of mitochondria: photosynthetic cell enslavement, gene-transfer pressure, and compartmentation efficiency. In: Martin WF, Müller M, editors. Origin of mitochondria and hydrogenosomes. Berlin: Springer-Verlag; 2007a. p. 161–99. Cavalier-Smith T. Concept of a bacterium still valid in prokaryote debate. Nature 2007b;446:257. Cavalier-Smith T. Evolution and relationships of algae: major branches of the tree of life. In: Brodie J, Lewis J, editors. Unravelling the algae. Boca Raton: CRC Press; 2007c. p. 21–55. Cavalier-Smith T. Intracellular coevolution and the origin the cell nucleus and sex, submitted for publication. Cavalier-Smith T. Megaphylogeny, cell body plans, adaptive zones: causes and timing of eukaryote basal radiations. J Euk Microbiol, in press. Cavalier-Smith T, Lewis R, Chao EE, Oates B, Bass D. Morphology and phylogeny of Sainouron acronematica sp. n. and the ultrastructural unity of Cercozoa. Protist 2008a;159:591–620. Cavalier-Smith T, Chao EE, Stechmann A, Oates B, Nikolaev S. Planomonadida ord. nov. (Apusozoa): ultrastructural affinity with Micronuclearia podoventralis and deep divergences within Planomonas gen. nov. Protist 2008b;159:535–62. Chen C, Huang C, Chen S, Liang J, Lin W, Ke G, et al. Subunit-subunit interactions in the human 26S proteasome. Proteomics 2008;8:508–20. Chen Y, Erickson HP. In vitro assembly studies of FtsZ/tubulin-like proteins (TubZ) from Bacillus plasmids: evidence for a capping mechanism. J Biol Chem 2008;283:8102–9. Cortes-Ledesma F, de Piccoli G, Haber JE, Aragon L, Aguilera A. SMC proteins, new players in the maintenance of genomic stability. Cell Cycle 2007;6:914–8. da Fonseca PC, Morris EP. Structure of the human 26S proteasome: subunit radial displacements open the gate into the proteolytic core. J Biol Chem 2008;283:23305–14. Dacks J, Field MC. Eukaryotic cell evolution from a comparative genomic perspective: the endomembrane system. In: Hirt RP, Horner DS, editors. Organelles, genomes and eukaryote phylogeny. Systematics association special vol. 68. CRC Press; 2004. p. 309–34. Dacks JB, Field MC. Evolution of the eukaryotic membrane-trafficking system: origin, tempo and mode. J Cell Sci 2007;120:2977–85. Danovaro R, Dell’Anno A, Corinaldesi C, Magagnini M, Noble R, Tamburini C, et al. Major viral impact on the functioning of benthic deep-sea ecosystems. Nature 2008;454:1084–7. De Duve C. The origin of eukaryotes: a reappraisal. Nat Rev Genet 2007;8:395–403. de Lichtenberg U, Jensen TS, Brunak S, Bork P, Jensen LJ. Evolution of cell cycle control: same molecular machines, different regulation. Cell Cycle 2007;6: 1819–25. De Mot R. Actinomycete-like proteasomes in a Gram-negative bacterium. Trends Microbiol 2007;15:335–8.

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322 Deli D, Leondaritis G, Tiedtke A, Galanopoulou D. Deficiency in lysosomal enzyme secretion is associated with upregulation of phosphatidylinositol 4-phosphate in Tetrahymena. J Eukaryot Microbiol 2008;55:343–50. Devos D, Dokudovskaya S, Alber F, Williams R, Chait BT, Sali A, et al. Components of coated vesicles and nuclear pore complexes share a common molecular architecture. PLoS Biol 2004;2:e380. Devos D, Dokudovskaya S, Williams R, Alber F, Eswar N, Chait BT, et al. Simple fold composition and modular architecture of the nuclear pore complex. Proc Natl Acad Sci U S A 2006;103:2172–7. Dong G, Medkova M, Novick P, Reinisch KM. A catalytic coiled coil: structural insights into the activation of the Rab GTPase Sec4p by Sec2p. Mol Cell 2007a;25:455–62. Dong JH, Wen JF, Tian HF. Homologs of eukaryotic Ras superfamily proteins in prokaryotes and their novel phylogenetic correlation with their eukaryotic analogs. Gene 2007b;396:116–24. Donohoe BS, Kang BH, Staehelin LA. Identification and characterization of COPIaand COPIb-type vesicle classes associated with plant and algal Golgi. Proc Natl Acad Sci U S A 2007;104:163–8. Embley TM. The multiple secondary origins of anaerobic lifestyles in eukaryotes. Phil Trans Roy Soc B 2006;361:1055–67. Erickson HP. Evolution of the cytoskeleton. BioEssays 2007;29:668–77. Fasshauer D, Sutton RB, Brunger AT, Jahn R. Conserved structural features of the synaptic fusion complex: SNARE proteins reclassified as Q- and R-SNAREs. Proc Natl Acad Sci U S A 1998;95:15781–6. Feldman JL, Geimer S, Marshall WF. The mother centriole plays an instructive role in defining cell geometry. PLoS Biol 2007;5:e149. Field MC, Gabernet-Castello C, Dacks JB. Reconstructing the evolution of the endocytic system: insights from genomics and molecular cell biology. Adv Exp Med Biol 2007;607:84–96. Gardiner J, McGee P, Overall R, Marc J. Are histones, tubulin, and actin derived from a common ancestral protein? Protoplasma 2008;233:1–5. Gill EE, Diaz-Trivino S, Barbera MJ, Silberman JD, Stechmann A, Gaston D, et al. Novel mitochondrion-related organelles in the anaerobic amoeba Mastigamoeba balamuthi. Mol Microbiol 2007;66:1306–20. Gille C, Goede A, Schloetelburg C, Presissner R, Kloetzel PM, Gobel UB, et al. A comprehensive view on proteasomal sequences: implications for the evolution of the proteasome. J Mol Biol 2003;326:1437–48. Gogarten JP, Kibak H, Dittrich P, Taiz L, Bowman EJ, Bowman BJ, et al. Evolution of the vacuolar ATPase: implications for the origin of eukaryotes. Proc Natl Acad Sci U S A 1989;86:6661–5. González JM, Suttle CA. Grazing by marine nanoflagellates on viruses and virus sized particles: ingestion and digestion. Mar Ecol Prog Ser 1993;94:1–10. Graur D, Martin W. Reading the entrails of chickens: molecular timescales of evolution and the illusion of precision. Trends Genet 2004;20:80–6. Grunt M, Zársky V, Cvrcková F. Roots of angiosperm formins: the evolutionary history of plant FH2 domain-containing proteins. BMC Evol Biol 2008;8:115. Guerrerro R, Esteve I, Pedros-Alio C, Gaju N. Predatory bacteria in prokaryotic communities. Ann N Y Acad Sci 1987;503:238–50. Gürkan C, Koulov AV, Balch WE. An evolutionary perspective on eukaryotic membrane trafficking. Adv Exp Med Biol 2007;607:73–83. Gürkan C, Stagg SM, Lapointe P, Balch WE. The COPII cage: unifying principles of vesicle coat assembly. Nat Rev Mol Cell Biol 2006;7:727–38. Hara F, Yamashiro K, Nemoto N, Ohta Y, Yokobori S, Yasunaga T, et al. An actin homolog of the archaeon Thermoplasma acidophilum that retains the ancient characteristics of eukaryotic actin. J Bacteriol 2007;189:2039–45. Harsay E, Schekman R. Avl9p, a member of a novel protein superfamily, functions in the late secretory pathway. Mol Biol Cell 2007;18:1203–19. Heinrich R, Rapoport TA. Generation of nonidentical compartments in vesicular transport systems. J Cell Biol 2005;168:271–80. Hennemuth W, Rhoads LS, Eichelberger H, Watanabe M, Van Bell KM, Ke L, et al. Ingestion and inactivation of bacteriophages by Tetrahymena. J Eukaryot Microbiol 2008;55:44–50. Higgs HN, Peterson KJ. Phylogenetic analysis of the formin homology 2 domain. Mol Biol Cell 2005;16:1–13. Hirano T. SMC proteins and chromosome mechanics: from bacteria to humans. Philos Trans R Soc Lond B Biol Sci 2005;360:507–14. Hirano T. At the heart of the chromosome: SMC proteins in action. Nat Rev Mol Cell Biol 2006;7:311–22. Hoepfner D, Schildknegt D, Braakman I, Philippsen P, Tabak HF. Contribution of the endoplasmic reticulum to peroxisome formation. Cell 2005;122:85–95. Hoepfner D, van den Berg M, Philippsen P, Tabak HF, Hettema EH. A role for Vps1p, actin, and the Myo2p motor in peroxisome abundance and inheritance in Saccharomyces cerevisiae. J Cell Biol 2001;155:979–90. Horowitz NH. On the evolution of biochemical syntheses. Proc Natl Acad Sci U S A 1945;31:153–7. Iwabe N, Kuma K, Hasegawa M, Osawa S, Miyata T. Evolutionary relationship of archaebacteria, eubacteria, and eukaryotes inferred from phylogenetic trees of duplicated genes. Proc Natl Acad Sci U S A 1989;86:9355–9. Iyer LM, Leipe D, Koonin EV, Aravind L. Evolutionary history and higher order classification of AAA+ ATPases. J Struct Biol 2004;146:11–31. Javaux EJ, Knoll AH, Walter MR. Morphological and ecological complexity in early eukaryotic ecosystems. Nature 2001;412:66–9. Jékely G. Small GTPases and the evolution of the eukaryotic cell. BioEssays 2004;25:1129–38. Jékely G, Arendt D. Evolution of intraflagellar transport from coated vesicles and autogenous origin of the eukaryotic cilium. BioEssays 2006;28:191–8.

321

John P, Whatley FR. Paracoccus denitrificans and the evolutionary origin of mitochondria. Nature 1975;254:495–8. Kahn RA, Cherfils J, Elias M, Lovering RC, Munro S, Schurmann A. Nomenclature for the human Arf family of GTP-binding proteins: ARF, ARL, and SAR proteins. J Cell Biol 2006;172:645–50. Keeling P. Chromalveolates and the evolution of plastids by secondary endosymbiosis. J Euk Microbiol, in press. Keeling PJ, Burger G, Durnford DG, Lang BF, Lee RW, Pearlman RE, et al. The tree of eukaryotes. Trends Ecol Evol 2005;20:670–6. Keller LC, Romijn EP, Zamora I, Yates 3rd JR, Marshall WF. Proteomic analysis of isolated Chlamydomonas centrioles reveals orthologs of ciliary-disease genes. Curr Biol 2005;15:1090–8. King MC, Drivas TG, Blobel G. A network of nuclear envelope membrane proteins linking centromeres to microtubules. Cell 2008;134:427–38. Kloepper TH, Kienle CN, Fasshauer D. An elaborate classification of SNARE proteins sheds light on the conservation of the eukaryotic endomembrane system. Mol Biol Cell 2007;18:3463–71. Koumandou VL, Dacks JB, Coulson RM, Field MC. Control systems for membrane fusion in the ancestral eukaryote; evolution of tethering complexes and SM proteins. BMC Evol Biol 2007;7:29. Krylov DM, Nasmyth K, Koonin EV. Evolution of eukaryotic cell cycle regulation: stepwise addition of regulatory kinases and late advent of the CDKs. Curr Biol 2003;13:173–7. Kurland CG, Collins LJ, Penny D. Genomics and the irreducible nature of eukaryote cells. Science 2006;312:1011–4. Lake JA, Herbold CW, Rivera MC, Servin JA, Skophammer RG. Rooting the tree of life using nonubiquitous genes. Mol Biol Evol 2007;24:130–6. Leipe DD, Wolf M, Koonin EV, Aravind L. Classification and evolution of P-loop GTPases and related ATPases. J Mol Biol 2002;317:41–72. Leung KF, Dacks JB, Field MC. Evolution of the multivesicular body ESCRT machinery; retention across the eukaryotic lineage. Traffic 2008;9:1698–716. Li G, Zhao G, Schindelin H, Lennarz WJ. Tyrosine phosphorylation of ATPase p97 regulates its activity during ERAD. Biochem Biophys Res Commun 2008;375:247–51. Li JY, Wu CF. New symbiotic hypothesis on the origin of eukaryotic flagella. Naturwissenschaften 2005;92:305–9. Lipson C, Alalouf G, Bajorek M, Rabinovich E, Atir-Lande A, Glickman M, et al. A proteasomal ATPase contributes to dislocation of endoplasmic reticulumassociated degradation (ERAD) substrates. J Biol Chem 2008;283:7166–75. Liu T, Wang T, Chapman ER, Weisshaar JC. Productive hemifusion intermediates in fast vesicle fusion driven by neuronal SNAREs. Biophys J 2008;94:1303– 14. López-García P, Moreira D. Metabolic symbiosis at the origin of eukaryotes. Trends Biochem Sci 1999;24:88–93. López-García P, Moreira D. Selective forces for the origin of the eukaryotic nucleus. BioEssays 2006;28:525–33. Low HH, Löwe J. A bacterial dynamin-like protein. Nature 2006;444:766–9. Lundmark R, Doherty GJ, Vallis Y, Peter BJ, McMahon HT. Arf family GTP loading is activated by, and generates, positive membrane curvature. Biochem J 2008;414:189–94. Maier UG. Nucleomorphs and protein trafficking. J Euk Microbiol, in press. Makarova KS, Wolf YI, Mekhedov SL, Mirkin BG, Koonin EV. Ancestral paralogs and pseudoparalogs and their role in the emergence of the eukaryotic cell. Nucleic Acids Res 2005;33:4626–38. Mans BJ, Anantharaman V, Aravind L, Koonin EV. Comparative genomics, evolution and origins of the nuclear envelope and nuclear pore complex. Cell Cycle 2004;3:1612–37. Maquat LE. Nonsense-mediated mRNA decay: splicing, translation and mRNP dynamics. Nature Rev Mol Cell Biol 2004;5:89–99. Margulis L, Dolan M, Whiteside JH. “Imperfections and oddities” in the origin of the nucleus. Paleobiology 2005;31:175–91. Marshall WF. Centriole assembly: the origin of nine-ness. Curr Biol 2007;17:R1057–9. Martin W, Müller M. The hydrogen hypothesis for the first eukaryote. Nature 1998;392:37–44. McKean PG, Vaughan S, Gull K. The extended tubulin superfamily. J Cell Sci 2001;114:2723–33. Mears JA, Ray P, Hinshaw JE. A corkscrew model for dynamin constriction. Structure 2007;15:1190–202. Mereschkovsky C. Über Natur und Ursprung der Chromatophoren im Pflanzenreiche. Biol Zentralbl 1905;25:593–604. Mima J, Hickey CM, Xu H, Jun Y, Wickner W. Reconstituted membrane fusion requires regulatory lipids, SNAREs and synergistic SNARE chaperones. EMBO J 2008;27:2031–42. Mitchell DR. The evolution of eukaryotic cilia and flagella as motile and sensory organelles. Adv Exp Med Biol 2007;607:130–40. Molnar A, Schwach F, Studholme DJ, Thuenemann EC, Baulcombe DC. miRNAs control gene expression in the single-cell alga Chlamydomonas reinhardtii. Nature 2007;447:1126–9. Moreira D, López-García P. Symbiosis between methanogenic archaea and deltaproteobacteria as the origin of eukaryotes: the syntrophic hypothesis. J Mol Evol 1998;47:517–30. Nasmyth K. Evolution of the cell cycle. Philos Trans R Soc Lond B Biol Sci 1995;349:271–81. Nasmyth K, Haering CH. The structure and function of SMC and kleisin complexes. Annu Rev Biochem 2005;74:595–648.

322

T. Cavalier-Smith / The International Journal of Biochemistry & Cell Biology 41 (2009) 307–322

Nickell S, Mihalache O, Beck F, Hegerl R, Korinek A, Baumeister W. Structural analysis of the 26S proteasome by cryoelectron tomography. Biochem Biophys Res Commun 2007;353:115–20. Novick P, Medkova M, Dong G, Hutagalung A, Reinisch K, Grosshans B. Interactions between Rabs, tethers, SNAREs and their regulators in exocytosis. Biochem Soc Trans 2006;34:683–6. Odronitz F, Kollmar M. Drawing the tree of eukaryotic life based on the analysis of 2,269 manually annotated myosins from 328 species. Genome Biol 2007;8:R196. Oriol R, Martinez-Duncker I, Chantret I, Mollicone R, Codogno P. Common origin and evolution of glycosyltransferases using Dol-P-monosaccharides as donor substrate. Mol Biol Evol 2002;19:1451–63. Osawa M, Anderson DE, Erickson HP. Reconstitution of contractile FtsZ rings in liposomes. Science 2008;320:792–4. Pasqualato S, Renault L, Cherfils J. Arf, Arl, Arp and Sar proteins: a family of GTPbinding proteins with a structural device for ‘front-back’ communication. EMBO Rep 2002;3:1035–41. Porter SM, Knoll AH. Testate amoebae of the Chuar Group, Grand Canyon. Paleobiology 2000;27:345–70. Porter SM, Meisterfeld R, Knoll AH. Vase-shaped microfossils from the Neoproterozoic Chuar Group, Grand Canyon: a classification guided by modern testate amoebae. J Paleont 2003;77:409–29. Raasi S, Wolf DH. Ubiquitin receptors and ERAD: a network of pathways to the proteasome. Semin Cell Dev Biol 2007;18:780–91. Richards TA, Cavalier-Smith T. Myosin domain evolution and the primary divergence of eukaryotes. Nature 2005;436:1113–8. Rodríguez-Ezpeleta N, Brinkmann H, Burger G, Roger AJ, Gray MW, Philippe H, et al. Toward resolving the eukaryotic tree: the phylogenetic positions of jakobids and cercozoans. Curr Biol 2007;17:1420–5. Roger AJ, Hug LA. The origin and diversification of eukaryotes: problems with molecular phylogenetics and molecular clock estimation. Phil Trans Roy Soc B 2006;361:1039–54. Rogozin IB, Makarova KS, Pavlov YI, Koonin EV. A highly conserved family of inactivated archaeal B family DNA polymerases. Biol Direct 2008;3:32. Sagan L. On the origin of mitosing cells. J Theoret Biol 1967;14:225–74. Salisbury JL. A mechanistic view on the evolutionary origin for centrin-based control of centriole duplication. J Cell Physiol 2007;213:420–8. Sanchatjate S, Schekman R. Chs5/6 complex: a multiprotein complex that interacts with and conveys chitin synthase III from the trans-Golgi network to the cell surface. Mol Biol Cell 2006;17:4157–66. Schwartz T, Blobel G. Structural basis for the function of the beta subunit of the eukaryotic signal recognition particle receptor. Cell 2003;112:793–803. Schwartz TU, Schmidt D, Brohawn SG, Blobel G. Homodimerization of the G protein SR␤ in the nucleotide-free state involves proline cis/trans isomerization in the switch II region. Proc Natl Acad Sci U S A 2006;103:6823– 8. Searcy D. Origins of mitochondria and chloroplasts from sulfur based symbioses. In: Hartman H, Matsuno K, editors. The origin and evolution of the cell. Singapore: World Scientific; 1992. p. 47–78. Searcy DG. Phylogenetic and phenotypic relationships between the eukaryotic nucleocytoplasm and thermophilic archaebacteria. Ann N Y Acad Sci 1987;503:168–79. Servin JA, Herbold CW, Skophammer RG, Lake JA. Evidence excluding the root of the tree of life from the actinobacteria. Mol Biol Evol 2008;25:1–4. Shiu SH, Bleecker AB. Receptor-like kinases from Arabidopsis form a monophyletic gene family related to animal receptor kinases. Proc Natl Acad Sci U S A 2001;98:10763–8. Simpson GC. Tempo and mode in evolution. New York: Columbia University Press; 1944.

Simpson GC. The major features of evolution. New York: Columbia University Press; 1953. Skophammer RG, Servin JA, Herbold CW, Lake JA. Evidence for a gram-positive, eubacterial root of the tree of life. Mol Biol Evol 2007;24:1761–8. Stagg SM, LaPointe P, Balch WE. Structural design of cage and coat scaffolds that direct membrane traffic. Curr Opin Struct Biol 2007;17:221–8. Stagg SM, LaPointe P, Razvi A, Gürkan C, Potter CS, Carragher B, et al. Structural basis for cargo regulation of COPII coat assembly. Cell 2008;134:474–84. Stanier RY. Some aspects of the biology of cells and their possible evolutionary significance. In: Charles HP, Knight JG, editors. Organization and control in prokaryotic and eukaryotic cells (society for general microbiology symposium no. 20). Cambridge: Cambridge University Press; 1970. p. 1–38. Stanier RY. Division I. The Cyanobacteria. In: Buchanan RE, Gibbons NE, editors. Bergey’s manual of determinative bacteriology. 8th ed. Baltimore: Williams and Wilkins; 1974. p. 22. Stechmann A, Baumgartner M, Silberman JD, Roger AJ. The glycolytic pathway of Trimastix pyriformis is an evolutionary mosaic. BMC Evol Biol 2006;6:101. Stephan A, Vaughan S, Shaw MK, Gull K, McKean PG. An essential quality control mechanism at the eukaryotic basal body prior to intraflagellar transport. Traffic 2007;8:1323–30. Sumathi JC, Raghukumar S, Kasbekar DP, Raghukumar C. Molecular evidence of fungal signatures in the marine protist Corallochytrium limacisporum and its implications in the evolution of animals and fungi. Protist 2006;157:363–676. Tabak HF, Hoepfner D, Zand A, Geuze HJ, Braakman I, Huynen MA. Formation of peroxisomes: present and past. Biochim Biophys Acta 2006;1763:1647–54. Tamura T, Cormier JH, Hebert DN. Sweet bays of ERAD. Trends Biochem Sci 2008;33:298–300. Tuszynski JA, Carpenter EJ, Huzil JT, Malinski W, Luchko T, Luduena RF. The evolution of the structure of tubulin and its potential consequences for the role and function of microtubules in cells and embryos. Int J Dev Biol 2006;50:341–458. Valas RE, Bourne PE. Rethinking proteasome evolution: two novel bacterial proteasomes. J Mol Evol 2008;66:494–504. Van Valen LM, Maiorana VC. The archaebacteria and eukaryotic origins. Nature 1980;287:248–50. Wahlman J, DeMartino GN, Skach WR, Bulleid NJ, Brodsky JL, Johnson AE. Real-time fluorescence detection of ERAD substrate retrotranslocation in a mammalian in vitro system. Cell 2007;129:943–55. Wickstead B, Gull K. A “holistic” kinesin phylogeny reveals new kinesin families and predicts protein functions. Mol Biol Cell 2006;17:1734–43. Wickstead B, Gull K. Dyneins across eukaryotes: a comparative genomic analysis. Traffic 2007;8:1708–21. Wilkes DE, Watson HE, Mitchell DR, Asai DJ. Twenty-five dyneins in Tetrahymena: a re-examination of the multidynein hypothesis. Cell Motil Cytoskeleton 2008;65:342–51. Woese CR, Kandler O, Wheelis ML. Towards a natural system of organisms: proposal for the domains Archaea, Bacteria, and Eucarya. Proc Natl Acad Sci U S A 1990;87:4576–9. Yang S, Doolittle RF, Bourne PE. Phylogeny determined by protein domain content. Proc Natl Acad Sci U S A 2005;102:373–8. Yoshizawa AC, Kawashima S, Okuda S, Fujita M, Itoh M, Moriya Y, et al. Extracting sequence motifs and the phylogenetic features of SNARE-dependent membrane traffic. Traffic 2006;7:1104–18. Yutin N, Makarova KS, Mekhedov SL, Wolf YI, Koonin EV. The deep archaeal roots of eukaryotes. Mol Biol Evol 2008;25:1619–30. Zamora I, Marshall WF. A mutation in the centriole-associated protein centrin causes genomic instability via increased chromosome loss in Chlamydomonas reinhardtii. BMC Biol 2005;3:15.