Dispatch R655 cellular or symbiotic functions, cannot genome reduction in the symbiont, 10. Nakabachi, A., Ishida, K., Hongoh, Y., Ohkuma, M., and Miyagishima, S. (2014). be lost without harmful or fatal results. HGT from various sources to the host Aphid gene of bacterial origin encodes a But these genes can be replaced, genome to maintain symbiont function, protein transported to an obligate and perhaps ever more easily as and now the targeting of protein endosymbiont. Curr. Biol. 24, R640–R641. 11. Moran, N.A., and Jarvik, T. (2010). Lateral the interaction networks of the products from host to symbiont has transfer of genes from fungi underlies endosymbiont reduce in complexity, even been found [4,10]. These make carotenoid production in aphids. Science 328, 624–627. thus reducing pressures on proteins clean separation of endosymbiont from 12. Jorgenson, M.A., Chen, Y., Yahashiri, A., to co-evolve. In a sense, when genes organelle more difficult to see, Popham, D.L., and Weiss, D.S. (2014). The are transferred from symbiont or prompting us not to look for the point bacterial septal ring protein RlpA is a lytic transglycosylase that contributes to rod organelle to the host nuclear genome when a symbiont ‘becomes’ an shape and daughter cell separation in and their proteins targeted back, this is organelle, but rather to ask, ‘Is there Pseudomonas aeruginosa. Mol. Microbiol. 93, 113–128. a compensatory change [4,9,14,17]. really anything so special about 13. Acuna, R., Padilla, B.E., Florez-Ramos, C.P., But other compensatory transfers that organelles?’ Rubio, J.D., Herrera, J.C., Benavides, P., affect a symbiont or organelle can also Lee, S.J., Yeats, T.H., Egan, A.N., Doyle, J.J., et al. (2012). Adaptive horizontal transfer of a involve genes transferred from References 1. Mu¨ ller, M., Mentel, M., van Hellemond, J.J., bacterial gene to an invasive insect pest of unrelated donors [8,9,18]. In the case of Henze, K., Woehle, C., Gould, S.B., Yu, R.-Y., coffee. Proc. Natl. Acad. Sci. USA 109, organelles the identity of potential van der Giezen, M., Tielens, A.G.M., and 4197–4202. Martin, W.F. (2012). Biochemistry and evolution 14. Timmis, J.N., Ayliffe, M.A., Huang, C.Y., and donors of transferred genes is not of anaerobic energy metabolism in eukaryotes. Martin, W. (2004). Endosymbiotic gene transfer: always clear, but in insects gene Microbiol. Mol. Biol. Rev. 76, 444–495. organelle genomes forge eukaryotic chromosomes. Nat. Rev. Genet. 5, 123–135. donors often appear to be pathogens, 2. Gray, M.W. (2014). The pre-endosymbiont hypothesis: a new perspective on the origin 15. Keeling, P.J. (2013). The number, speed, and specifically reproductive manipulators and evolution of mitochondria. Cold Spring impact of plastid endosymbioses in eukaryotic [7–9]. These bacteria skew the number Harb. Perspect. Biol. 6, a016097–a016097. evolution. Annu. Rev. Plant Biol. 64, 583–607. 3. Keeling, P.J., and Archibald, J.M. (2008). 16. Larkum, A.W.D., Lockhart, P.J., and Howe, C.J. and sex ratio of offspring in infected Organelle evolution: what’s in a name? Curr. (2007). Shopping for plastids. Trends Plant Sci. populations in different ways, and often Biol. 18, R345–R347. 12, 189–195. 17. Keeling, P.J., and Palmer, J.D. (2008). reside in insect germ line cells. The 4. Nowack, E.C.M., and Grossman, A.R. (2012). Trafficking of protein into the recently Horizontal gene transfer in eukaryotic evolution. frequency of transfer from these established photosynthetic organelles of Nat. Rev. Genet. 9, 605–618. groups is therefore probably due to Paulinella chromatophora. Proc. Natl. Acad. 18. Stairs, C.W., Eme, L., Brown, M.W., Mutsaers, C., Sci. USA 109, 5340–5345. Susko, E., Dellaire, G., Soanes, D.M., van der simple chance: their cell biology 5. Cavalier-Smith, T., and Lee, J.J. (1985). Giezen, M., and Roger, A.J. (2014). A SUF includes infection of the germ line, Protozoa as hosts for endosymbioses and Fe-S cluster biogenesis system in the mitochondrion-related organelles of the which provides ample opportunity for the conversion of symbionts into organelles. J. Protozool. 32, 376–379. anaerobic protist Pygsuia. Curr. Biol. 24, gene transfer that can be passed to 6. Martin, W. (2010). Evolutionary origins of 1176–1186. future generations. The same is true of metabolic compartmentalization in eukaryotes. 19. Theissen, U., and Martin, W. (2006). The Philos. Trans. R. Soc. Lond. B Biol. Sci. 365, difference between organelles and single-celled eukaryotes (protists), 847–855. endosymbionts. Curr. Biol. 16, R1016–R1017. where all newly acquired genes are taken 7. Nikoh, N., McCutcheon, J.P., Kudo, T., 20. Bhattacharya, D., and Archibald, J.M. (2006). The difference between organelles and into the germline automatically [17]. Miyagishima, S.Y., Moran, N.A., and Nakabachi, A. (2010). Bacterial genes in the endosymbionts - response to Theissen and As the prevalence of intimate and aphid genome: absence of functional gene Martin. Curr. Biol. 16, R1017–R1018. stable endosymbiotic associations transfer from Buchnera to its host. PloS Genet. 6, e1000827. 1Division of Biological Sciences, University of has become more clear, the degree 8. Husnik, F., Nikoh, N., Koga, R., Ross, L., to which host and endosymbiont Duncan, R.P., Fujie, M., Tanaka, M., Satoh, N., Montana, Missoula, MT 59801, USA. 2 are integrated has been revealed to Bachtrog, D., Wilson, A.C.C., et al. (2013). Canadian Institute for Advanced Research, Horizontal gene transfer from diverse bacteria Botany Department, University of British be far less discontinuous than to an insect genome enables a tripartite nested Columbia, 3529-6270 University Boulevard, previously believed. Accordingly, the mealybug symbiosis. Cell 153, 1567–1578. Vancouver, BC, V6T 1Z4, Canada. 9. Sloan, D.B., Nakabachi, A., Richards, S., Qu, J., characteristics separating ‘symbiont’ Murali, S.C., Gibbs, R.A., and Moran, N.A. E-mail: [email protected], from ‘organelle’ have become less (2014). Parallel histories of horizontal gene [email protected] clear [3,4,19,20]. There is an transfer facilitated extreme reduction of endosymbiont genomes in sap-feeding insects. understandable desire to draw a Mol. Biol. Evol. 31, 857–871. http://dx.doi.org/10.1016/j.cub.2014.05.073 distinct line between the two for simplicity, but first we must ask, does this line exist? If so, it is best drawn by evolutionary and mechanistic distinctions, not by perceived Animal Evolution: Looking for the differences born of tradition, definitions, or historical contingency. First Nervous System Organelles were discovered first, have been studied for decades, and their bacterial origins dominated the The human brain is easily the most baffling bit of biology on the planet. How did discussion about endosymbiosis and the nervous system evolve? What came first: neurons or synaptic proteins? A new Trichoplax evolution for many years. They enjoy paper studying the pancake-shaped suggests it was not the neurons. a status apart from other biological entities: derived from bacteria, but Erik M. Jorgensen rise of complex, unfamiliar looking so different as to be given their own multicellular marine life forms, ended name. But the list of their ‘unique’ Something bad must have happened with an extinction that wiped out most characteristics is shrinking: stable around 542 million years ago: the of these creatures. Unfortunately, this endosymbioses promote extensive Ediacaran period, which had seen the event obscured our view on the Current Biology Vol 24 No 14 R656 evolution of the first nervous systems. speck of an animal. It was originally has a protein closely related to the We know that some Ediacaran discovered in 1883 crawling on the synaptic synaptotagmin and thus animals must have had nervous glass of a fish tank containing perhaps only Trichoplax will exhibit systems, as they left tracks and seawater. Its simple circular body plan, true synaptic behavior, i.e. tunnels, and thus could move in an only 1 mm in diameter, was described calcium-activated vesicle fusion. active, organized fashion. But at the time but the organism was largely Surprisingly, only ctenophores because they left no directly forgotten thereafter until its possess panexin gap-junction recognizable successors, we can’t rediscovery in the 1970s. proteins, which allow rapid infer how those first nervous systems Can this dog do tricks? Sure, the communication between cells by direct arose. The Cambrian explosion that essential ones: the pancake that is electrical conductance. All three phyla, followed gave rise to the familiar Trichoplax can glide along a surface, placozoa, porifera and ctenophora, animal phyla. Examination of their which implies coordination of beating lack pentameric ligand-gated ion genomes suggests that the nervous cilia. Moreover, when it comes across a channels altogether, and therefore system arose once, in a perfect state, patch of algae, it pauses over the food show no evidence of fast as all of the key molecular components source, presumably digesting them by neurotransmission by acetylcholine, of the nervous system are conserved secreting enzymes onto its ventral serotonin, dopamine, GABA or glycine. in the genomes of animals: the ion surface, for absorption via microvilli. It is possible that variants will channels used by excitable cells, Trichoplax also changes shape from a eventually be found when these draft an extensive repertoire of circular to an elongated disk genome sequences are fully neurotransmitters and synaptic sometimes. So is there evidence of a assembled. But the lack of most machinery. In fact at the level of nervous system underlying these classical neurotransmitter receptors neurotransmitter systems, the human behaviors? Now, through the work of is further supported by a lack of nervous system looks like a simplified Smith et al. [1], a full understanding of the vesicular transporters for version of the primordial nervous the anatomy of the w50,000 cells in acetylcholine, monoamines, GABA, system. How then did the basic unit Trichoplax has become available. glycine, or nucleotides. These are the of the nervous system — the Trichoplax resembles a sandwich proteins that make a sophisticated neuron — arise? What came first? composed of two epithelial layers and nervous system, that provide Synaptic proteins, or the specialized some loose cells in between; there is no excitatory or inhibitory inputs, and anatomy of neurons? What was the gut, no muscles or neurons—instead diversity of neurotransmitter responses first neurotransmitter? What kind of there are only six specialized cell types. in complicated networks — these behavior did it mediate? In the last It moves by means of cilia covering the primitive animals are deprived of few years, evidence has arisen that ventral surface. Combined with access to this toolkit. some of the more primitive soft-bodied previous genome sequence data, this What was the first fast creatures around today, including the study suggests that the genes for a neurotransmitter? Probably glutamate, odd little Trichoplax, might offer a nervous system are present, but based on the absence of any other glimpse of Pre-Cambrian life. A new neurons are not. candidates. The genomes of all three study by Smith and colleagues [1] in phyla encode subunits of the glutamate this issue of Current Biology describes Pre-Cambrian Proteins receptor family. In fact, transmission the anatomy of Trichoplax in What are the important genes for using these receptors may be quite unprecedented detail and understanding nervous system elaborate in these organisms, as the demonstrates a complicated repertoire evolution? Recently the genomes of ctenophore Pleurobrachia bachei of cells, but they are unfamiliar and representative species from the phyla encodes 14 different ionotropic bizarre. of porifera [2,3], ctenophora [4,5], and glutamate receptors [5]. However, it placozoa (Trichoplax) [6] have been remains to be seen whether these Pre-Cambrian Animals sequenced. All three phyla encode receptors form glutamate-gated ion In the last few years the genomes of the components for fast synaptic channels, and there is reason to doubt. some of the more insignificant transmission: voltage-gated ion Only Trichoplax has a protein weakly creatures on the planet have been channels and specialized SNARE similar to the vesicular glutamate completed — porifera, ctenophores, proteins that mediate synaptic vesicle transporter; sponges and ctenophores and placozoans. These animals are all fusion. Importantly, these genomes only have homologs of the related sialic soft-bodied and presumably sit at the also encode complexin, a acid transporter sialin. Thus, the base of the animal tree of life, though SNARE-binding protein, which may molecule activating these receptors their exact relationship with the rest of interrupt constitutive fusion of vesicles may not be glutamate at all, and given animals is still contentious. The [7], and introduce a regulatory step. the absence of a synaptotagmin-1 creatures in these phyla are quite But the key protein for synaptic ortholog it is not even clear that simple, they have soft bodies, they lack function is synaptotagmin, the calcium transmission in sponges and three germ layers, and lack a true gut. binding protein that drives regulated ctenophores will be fast. But they can coordinate cells to move vesicle fusion. The homolog of or contract. These phyla may be synaptotagmin in ctenophores and Pre-Cambrian Cells representatives of Precambrian animal sponges is not closely related to the Based on the genomic data, it appears life and could provide insight into the synaptic version. Thus, it is not clear that the most primitive organism with ancient origins of the nervous system. if calcium-dependent vesicle fusion all of the true components for synaptic The sole representative of the can take place in sponges and transmission is the pancake-shaped Placozoans, Trichoplax, is a mere ctenophores. By contrast, Trichoplax Trichoplax. So does Trichoplax Dispatch R657 possess the cellular components of a genome. Moreover, these cells do not recognizable nervous system? In the express the specialized SNARE light of the new work by Smith et al. [1], proteins used at synapses. the answer is a resounding ‘No’. Antibody staining for synaptic Grossly, Trichoplax is composed of SNAREs instead highlighted the gland two epithelial layers, it has no cells. The gland cells, like the ventral endoderm, glands, musculature or epithelial cells, are ciliated and recognizable nervous system. Smith therefore contribute to locomotion. et al. used cutting-edge methods to Feeding animals pause over patches of preserve cells and searched for algae, making it possible that these evidence of neurons and synapses. cilia are not just used for locomotion Specifically, they used high-pressure but perhaps also sensing food. Gland freezing to preserve tissues, thus cells also contain potential secretory avoiding the substantial artifacts vesicles, some vesicles even appear to caused by slow aldehyde fixation and be docked to the plasma membrane. dehydration, and the view is stunning: But the cells are oriented toward the Figure 1. Trichoplax adhaerens. there is nothing resembling a typical ventral surface, making it possible that Micrograph of Trichoplax in partially polar- nervous system; there are no axons, these cells secrete material onto the ized light. Crystal cells appear as white dots around edges. The dark granules in the substrate to aid in either locomotion or there are no synapses. Nevertheless, center are inclusions in fiber cells. (Photo: the authors find evidence for digestion rather than mediating Carolyn Smith.) specialized cells in these simple cell–cell communication. creatures: six different cell types can Alternatively, of course, the gland be identified. On the ventral surface cells may be bona fide neurosecretory There is one other somewhat are three cell types: the ventral cells releasing fast-acting disturbing possibility: glutamate is not epithelial, lipophil and gland cells. The neurotransmitters onto other cells. a neurotransmitter — and Trichoplax cells on the lower surface are ciliated There are three potential does not possess a primitive nervous and provide for gliding motility. These neurotransmitters encoded in system that uses any classical cells are also covered with a shag the genome of Trichoplax: the neurotransmitters. The sole evidence carpet of microvilli, which serves as an neuropeptide FMRFamide, the that glutamate is acting as a transmitter external stomach to digest algae purinergic transmitter ATP, and is the presence of ionotropic glutamate growing on the substrate. The dorsal the classical neurotransmitter receptors. But these receptors may surface is covered by a thin epithelium. glutamate. Smith and colleagues simply be relaying the presence of a Between the ventral and dorsal demonstrated that the gland cells nutritive signal from the algae — amino surfaces are the rare crystal cells and a express the neuropeptide FMRFamide. acids. Glutamate activation of the loose collection of fiber cells. In most organisms, neuropeptides gland cells could direct movement The most bizarre cells are these activate metabotropic receptors and toward a food source. In this crystal cells. They are found at the not fast-acting ion channels. But in model, amino acids evolved to margin of the animal (Figure 1) invertebrates FMRFamide can also become the predominant classical between the epithelial cell layers. Each activate ion channels of the ASIC neurotransmitters because they were contains a specialized organelle family. Trichoplax has a particularly rich originally extrinsic molecules signaling containing a crystal that can polarize collection of genes encoding ASIC the most important thing to the animal: light. Are they eyes? There are reports channel family members, introducing the presence of something to eat. Only that Trichoplax responds to light, and the possibility that fast later did they become adapted for opsins are encoded in the genome [8]. neurotransmission in Trichoplax acts internal communication. But the crystal cells do not look like via peptide-gated ion channels rather It is all a big surprise. Its genome neurons by ultrastructure (or anything than classical neurotransmitter-gated says Trichoplax is an animal with a seen before, for that matter) and do ion channels. full-blown nervous system, including not stain for antibodies for SNARE The gland cells could also use synaptic proteins, neurotransmitter proteins — which would be required glutamate or ATP as a systems, and muscle proteins; it even for secretory communication with neurotransmitter, as the genome engages in behaviors we expect would other cells. encodes subunits of ionotropic require a neuromuscular system. But The fiber cells are also intriguing. glutamate and P2X purinergic instead we find a foreign and Since they have long processes receptors. If glutamate or ATP is unrecognizable cellular anatomy, no reaching among the other cells, one released from vesicles, then there must axons, no synapses, no muscles. There could imagine them to be the be a vesicular transporter. The are three possibilities: Trichoplax had a predecessors of neurons or muscle glutamate and nucleotide transporters nervous system but lost its anatomical cells. Indeed, they express P2X are encoded by the SLC17 family of specializations during the process of purine-activated ion channels and anion transporters. The Trichoplax acquiring a simpler lifestyle. might therefore receive excitatory genome does not encode orthologs of Alternatively, our expectations were input. However, the contacts with other the conserved glutamate or ATP wrong and a nervous system can look cells lack evidence of synaptic transporters, but does encode several much different than we expected. specializations, and the processes lack other SLC17 anion transporters; one of Finally, it may have never had one, and a contractile apparatus, despite the these others may load these molecules we really are looking at what animals presence of muscle myosins in the into secretory vesicles. did with all the pieces before they were Current Biology Vol 24 No 14 R658 assembled into a fabulous biological transcriptomes from all poriferan classes reveals the nature of placozoans. Nature 454, surprising genetic complexity in sponges. Mol. 955–960. contraption. Biol. Evol. 31, 1102–1120. 7. Hobson, R.J., Liu, Q., Watanabe, S., and 4. Ryan, J.F., Pang, K., Schnitzler, C.E., Jorgensen, E.M. (2011). Complexin maintains References Nguyen, A.-D., Moreland, R.T., Simmons, D.K., vesicles in the primed state in C. elegans. Curr. 1. Smith, C.L., Varoqueaux, F., Kittelmann, M., Koch, B.J., Francis, W.R., Havlak, P., et al.; Biol. 21, 106–113. Azzam, R.N., Cooper, B., Winters, C.A., Eitel, M., NISC Comparative Sequencing Program (2013). 8. Feuda, R., Hamilton, S.C., McInerney, J.O., and Fasshauer, D., and Reese, T.S. (2014). Novel cell The genome of the ctenophore Mnemiopsis Pisani, D. (2012). Metazoan opsin evolution types, neurosecretory cells, and body plan of the leidyi and its implications for cell type evolution. reveals a simple route to animal vision. Proc. early-diverging metazoan Trichoplax adhaerens. Science 342, 1242592. Natl. Acad. Sci. USA 109, 18868–18872. Curr. Biol. 24, 1565–1572. 5. Moroz, L.L., Kocot, K.M., Citarella, M.R., 2. Srivastava, M., Simakov, O., Chapman, J., Dosung, S., Norekian, T.P., Povolotskaya, I.S., Fahey, B., Gauthier, M.E.A., Mitros, T., Grigorenko, A.P., Dailey, C., Berezikov, E., Dept. of Biology, University of Utah, Richards, G.S., Conaco, C., Dacre, M., Buckley, K.M., et al. (2014). The ctenophore Howard Hughes Medical Institute, Hellsten, U., et al. (2010). The Amphimedon genome and the evolutionary origins of neural Salt Lake City, UT, USA. systems. Nature 510, 109–114. queenslandica genome and the evolution E-mail: [email protected] of animal complexity. Nature 466, 6. Srivastava, M., Begovic, E., Chapman, J., 720–726. Putnam, N.H., Hellsten, U., Kawashima, T., 3. Riesgo, A., Farrar, N., Windsor, P.J., Giribet, G., Kuo, A., Mitros, T., Salamov, A., Carpenter, M.L., and Leys, S.P. (2014). The analysis of eight et al. (2008). The Trichoplax genome and http://dx.doi.org/10.1016/j.cub.2014.06.036

Tumor Models: Tumor–Stroma nuclear protein [8], cooperates with EGF receptor to elicit malignant Interactions Drive Neoplastic neoplastic growth of imaginal Drosophila primordia. The multilayered tumor Transformation in induces the expression of MMP1 and the consequent degradation of the basement membrane and becomes Stromal cells play a supportive role in the initiation and progression of highly metastatic, as transformed cells carcinomas. A new study in Drosophila implicates mesenchymal cells in are found in distant internal organs supporting EGF receptor-driven tumor growth and cellular transformation of such as the gut and malphigian tubules. epithelial tissues. Remarkably, Pipsqueak behaves as a tumor-promoting gene in a Drosophila Marco Mila´ n additional oncogenic mutations are Notch-driven epithelial tumor model required for disease progression. In [9]. This observation thus reinforces the Carcinomas, malignant neoplasms of this regard, the imaginal primordia of context-dependent tumor suppressor epithelial origin, are the most common Drosophila — monolayered epithelia or tumor-promoting roles of many form of human cancer. Mesenchymal within the feeding larvae that grow cancer genes. cells in the stroma regulate the one-thousand fold in cell number and As is often the case, the initial expression and remodeling of the tissue size — have been used to identify observation made by Herranz et al. extracellular matrix (ECM) and produce new molecular elements that cooperate [1] that allowed the identification growth factors that support the with these two oncogenes in driving of a mesenchymal cell population survival and proliferation of epithelial tumor growth, epithelial transformation, supporting EGF-receptor-driven tumor transformed cells. As they reported basement membrane degradation, growth was unexpected, but a key recently in Current Biology, Herranz and invasive behavior [2–5]. Mutations finding. During the characterization of et al. [1] have used the genetic model that affect the Scribbled–Disc the oncogenic cooperation between species Drosophila to dissect the Large–Lgl cell polarity complex or EGF receptor and Pipsqueak, underlying molecular and cellular those causing mitochondrial GFP-positive EGF-receptor-expressing mechanisms driving tumor–stroma dysfunction, or overexpression of cells were found to intermingle interactions. This work underscores certain miRNAs, cooperate with with ‘‘groups of cells not expressing the contribution of resident EGF receptor/Ras dysregulation in GFP’’ (Figure 1). Further functional mesenchymal cells in promoting promoting tumorigenesis. In all these characterization of this population the neoplastic transformation of cases, tumorigenesis relies on a indicated that the GFP-negative cells EGF-receptor-expressing epithelial JNK-dependent transcriptional are resident myoblasts that proliferate cells and identifies Dpp and Wingless program that regulates the invasion of in response to Dpp — a member of the as the signaling molecules driving transformed cells, drives the expression TGF-b superfamily — produced by the growth of these two cell populations. of the mitogenic molecules responsible tumor. Thus, those myoblasts abutting EGF-receptor gene amplification for tumor growth, and induces the the transformed cell population showed has been reported in a wide range of expression of matrix metalloproteins strong activation of the Dpp pathway carcinomas, and mutations that activate (MMPs) involved in basement and elevated mitotic activity. the small G protein Ras are found membrane degradation, a prerequisite Genetically elegant experiments in 20–25% of all human tumors. for tissue invasiveness [6,7]. performed by Herranz et al. [1] However, neither EGF-receptor The work of Herranz et al. [1] stems demonstrated that the proliferative overexpression nor the presence of from the observation that depletion myoblast population plays a major role activated Ras is sufficient to drive of the Polycomb group epigenetic in driving neoplastic tumor growth. malignant transformation, and silencer Pipsqueak, a BTB-containing Thus, selective ablationof the myoblasts