Research

Epithelia, an Evolutionary Novelty of Metazoans

Description
Epithelia, an Evolutionary Novelty of Metazoans
Categories
Published
of 10
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Share
Transcript
  Epithelia, an Evolutionary Noveltyof Metazoans SALLY P. LEYS 1   AND  ANA RIESGO 2 1 Department of Biological Sciences, University of Alberta, Edmonton, Alberta, Canada 2 Department of Organismic and Evolutionary Biology, Harvard University, Cambridge, Massa-chusetts At the point in animal evolution when cells began to adhere to each other they presumably initiallyfunctioned as colonies. The formation of an epithelium that enclosed and controlled an internalmilieu would have been the first event to distinguish an individual animal from a colony. To betterunderstand when the first epithelium arose and what its characteristics were, we evaluate themorphological, functional, and molecular characters of epithelia in sponges, considered here theextant representatives of the first metazoans. In particular, we show new claudin-like sequencesfrom sponges align most closely with sequences from  Drosophila   that have a barrier function inseptate junctions. We also show that type IV collagen, the main component of the basementmembrane (BM), is present in calcareous sponges, and we confirm the presence of type IV-likecollagen (spongin short chain collagen) in other sponges. Though in sponges as in other metazoansthe epithelium has grades of specialization with varying complexity of junctions and the BM, themain character of a functional epithelium, the ability to seal and control the ionic composition of the internal milieu, is a property of even the simplest sponge epithelium, and therefore the firstmetazoans likely also had epithelia with these characteristics, which we consider a ‘‘true’’epithelium.  J. Exp. Zool. (Mol. Dev. Evol.) 314B, 2011 .  &  2011 Wiley Periodicals, Inc. How to cite this article: Leys SP, Riesgo A. 2011. Epithelia, an evolutionary novelty of metazoans. J. Exp. Zool. (Mol. Dev. Evol.) 314B:[page range]. BIOLOGY OF THE INTEGUMENT  The first animal epithelium was a novel structure that had thedistinct (and also novel) function of compartmentalizing regionsof the body and controlling the passage of molecules betweentwo environments. The epithelium is, therefore, a good exampleof an evolutionary novelty, a feature that is ‘‘not homologous toa feature in an ancestral taxon’’ (Hall, 2005). Oddly, it is still notaccepted that the first multicellular animals, Porifera, have a trueepithelium, and this perspective seems to depend on the level of organization at which homology is required.Tyler (2003) summarizes a widely held view that epithelia arethe ‘‘default state of cells in all eumetazoans,’’ which arose in ‘‘thestem to the Cnidaria.’’ He says Porifera are not considered to havetrue epithelia by ‘‘established criteria,’’ but admits that Porifera‘‘likely have developed many, if not all, of the mechanismsdeemed specific to true epithelium.’’ This statement highlightsa long-standing equivocation on the part of comparativemorphologists on the question of whether Porifera can beconsidered epithelial, and therefore whether epithelia are anovelty of Metazoa or only Eumetazoa. Even in the introductionto  The Biology of the Integument  , K. Sylvia-Richards (’84) writesthat ‘‘Porifera and Cnidaria are both essentially epithelial animalscomposed of two layers of cells surrounding a central,environment-filled cavity’’; yet, Porifera were omitted from that volume primarily based on work that had suggested that spongeepithelia are of ‘‘predominantly one sort and are apparently very loosely associated’’ (Mackie, ’84). Cnidarian epithelia in contrastare ‘‘histologically more advanced’’ with both striated andsmooth muscle, nerves, and chromatophores, but with the absenceof a middle layer (mesoderm) which prevents specialization(Mackie, ’84). By this definition, to count as epithelia the tissueought to contain a diversity of cell types which adhere tightly to Published online in Wiley Online Library (wileyonlinelibrary.com).DOI: 10.1002/jez.b.21442Received 22 March 2011; Revised 6 August 2011; Accepted 24 August2011Grant Sponsor: NSERC Canada; Grant number: 222863; Grant Sponsor:SSHRC Canada; Grant number: 410-2008-0140; Grant Sponsor: NSF; Grantnumber: DEB  ] 0844881.Additional supporting information may be found in the online version of thisarticle.  Correspondence to: Sally Leys, CW 405 Biological Sciences Building,University of Alberta, Edmonton, AB, T6G 2E9. E-mail: sleys@ualberta.ca ABSTRACT  J. Exp. Zool.(Mol. Dev. Evol.) 314B, 2011 &  2011 WILEY PERIODICALS, INC. RESEARCH ARTICLE  one another, and therefore presumably also seal and control theionic constitution of the internal compartment.Tyler’s ‘‘established criteria’’ of epithelia are morphologicalcharacteristics (adherens and sealing junctions and a basementmembrane (BM)) and the presence of molecular componentsknown from studies in model systems to be involved in theformation of the above junctions and in their physiologicaladhering, sealing, or signaling functions. Although theultrastructure of all animal tissues has been known for sometime, the accounting of the molecules involved is very much stillin progress. What Tyler does not take into account, and aboutwhich there is less information, is the physiology of epithelialfunction, because it is only known for a few model systems(mammalian cell lines, fish, amphibians,  Drosophila melanogaster  and other insects,  Caenorhabditis elegans , and recently a sponge, Spongilla lacustris  [see references in Adams et al., 2010]).Sponges have traditionally been united in onephylum—Porifera—that is characterized by having a porous body containing canals and chambers and a specific cell type—choanocyte—that generates the feeding and respiratory current(Grant, ’36). Monophyly of Porifera is supported by one set of molecular analyses (Philippe et al., 2009), but another set findsthat sponges are paraphyletic (Sperling et al., 2009). In bothinstances, the four classes of sponges fall into a clade of siliceoussponges (Silicea) containing Demospongiae and Hexactinellida, anda grouping of calcareous and homoscleromorph sponges. Recentwork has shown that the latter group, Homoscleromorpha, has a BMstructure containing type IV collagen (Boute et al., ’96), which somesayisevidenceofa‘‘trueepithelium’’andwhichhasprompteduseof the term ‘‘Epitheliozoa’’ (Boury-Esnault et al., 2003; Sperling et al.,2009). But what exactly is a ‘‘true’’ epithelium?The term ‘‘true,’’ in this context, seems to come from ‘‘good’’(Greek  5 eu) as in ‘‘eumetazoa,’’ a term used by Hyman to refer to‘‘animals of the tissue or organ system grade of construction withmouth and digestive tract,’’ including Cnidaria, Ctenophora, andbilaterians (Hyman, ’40). Hyman followed Sollas’ (1884) who feltsponges were more similar to  infusiora  (protists) than to other metazoans, based on the idea that the sponge ciliated larvalepithelial cells were protist characters (‘‘choanoflagellate-cells,’’p 612) that arose earlier during development than the metazoancharacters (gastrulation by in-folding, cellular migration), whichhappened at settlement. Sollas proposed the term Parazoa for sponges, and Hyman (’40) therefore used the term Eumetazoa for the rest of multicellular animals. [Sollas arrived at this conclusionby studying development in a homoscleromorph sponge ( Hali-sarca lobularis , the species now referred to as  Oscarella lobularis ),and ironically it is now homoscleromorphs that are proposed to becloser to other metazoans than other sponges.] Therefore, a ‘‘trueepithelium’’ is one that has the characteristics of epithelia inEumetazoa: adherens and occluding junctions and a BM.In this article we ask whether it is important to consider thehierarchy of traits in determining ‘‘novelty’’ (that is, whether there are ‘‘grades’’ of epithelia, but ultimately all characteristicsare necessary for a ‘‘true epithelium’’) or whether functionality rather than just type of morphological trait is importantfor ‘‘novelty.’’ The challenge is that although formerly anevolutionary novelty was identified only by its morphology,today molecular data is used to account for the ability to producea structure with a known function and the two do not alwaysagree. For example, although all the genes for muscle and,therefore, mesoderm are in the  Hydra  genome (Chapman et al.,2010), cnidarians are not understood to have a mesoderm, in thesense of a structure that arises as a third layer via embryogenesis(see Siepel and Schmid, 2006, for a discussion of cnidarianmesoderm). Conversely, ctenophores seem to have proper mesoderm that arises as a third layer during embryogenesis,but possess none of the genes for mesoderm (M Q Martindale,Kewalo Marine Laboratory, personal communication). Nevertheless,more often than not gene presence and absence is now used topredict the presence or absence of a structure. In relation to thefocus of this article, however, the fact that we lack completemolecular data for more than one species of Porifera remains amajor obstacle.Here we use the definition of novelty as ‘‘a feature nothomologous to a feature in an ancestral taxon’’ (Hall, 2005), toassess whether the epithelium was a novel character for allmetazoans or only eumetazoans or whether it is a novel structurethat evolved within the Porifera. In contrast to the more conceptualarticles in this issue (e.g., Hall and Kerney, this issue), we feel thatin order to assess whether epithelia are a novelty, importantquestions need to be settled regarding molecular, functional, andmorphological changes. It is these aspects we address. CELL–CELL JUNCTIONS: MORPHOLOGY AND PHYSIOLOGY  The ultrastructure of adherens junctions, of tight and septate junctions (SJs) (both occluding junctions), and of a BM, is wellcharacterized from both vertebrates and invertebrates (Fig. 1), butnot all epithelia show equally good morphology when preserved.Junctions in Porifera are at times clear and at others very indistinct, so much so that, although there are many reports of adherens and SJs in Porifera, some authors still find theultrastructure ‘‘less than convincing’’ (Fahey and Degnan,2010). In sponges, adherens junctions seem less electron opaquethan in Cnidaria, Platyhelminthes, and in textbook views of  vertebrate junctions, although not all images of vertebrate junctions show equally perfect preservation (e.g., Farquhar andPalade, ’63). Both zonula and macula adherens junctions, beltand spot adhesion plaques, respectively, contain cadherin-based junctional proteins. Although numerous reports on invertebratetissues describe the presence of desmosomes (macula adherentes),strictly speaking desmosomes are defined by the presence of desmosomal cadherins and plakoglobins, which are only in vertebrates. Hence, ‘‘desmosomes’’ (properly spot adherens junctions) have been reported in all groups of sponges (Lethias LEYS AND RIESGO 2 J. Exp. Zool. (Mol. Dev. Evol.)  et al., ’83; reviewed in Leys et al., 2009), but in none of the groups(hexactinellids, demosponges, calcarea, homoscleromorphs) doesthe ultrastructure appear as it does in bilaterians. In fact, equally ‘‘convincing’’ adherens junctions can be seen by TEM in the slimemold  Dictyostelium  (Grimson et al., 2000; Dickinson et al., 2011),where they link cells in the inner region of the fruiting body.Those junctions have even been shown to label with actin and the Dictyostelium  protein homolog of   b -catenin, Aardvark. No dataare yet available on the proteins used in sponge adherens junctions, but thanks largely to the presence in the spongegenome of the catalog of molecular components known to berequired for making adherens junctions (including severalcadherins), the presence of adherens junctions in sponges is notquestioned (Fahey and Degnan, 2010). As to occluding junctions, the ladder-like structure of SJsis particularly faint in demosponge and homoscleromorphtissues and yet can be very distinct in calcareous sponges. Thisis attributed to the presumed need to have a sealed compartmentto secrete calcium carbonate in calcareous sponges, but not insiliceous sponges (including homoscleromorphs). Nevertheless,based on these ultrastructural data, all sponges have beenconsidered to have SJs (see Leys et al., 2009, and referencestherein), and new physiological data now show that demospongeepithelia do seal and control the ionic composition of their internal milieu (Adams et al., 2010).The physiological ability of an epithelium to seal isdetermined either by measuring transepithelial resistance acrossthe cell sheet (Shaw, ’58) or block of tracers, e.g., 10kd dextrans(Asano et al., 2003; Nelson et al., 2010) or 0.85kd ruthenium red(Hori, ’87). The extent of sealing depends on the specialization of the tissue, but in all instances there is at least a resistance of  4 10 O cm 2 or block of 10kd molecules (see Supplementary Table 1 in Adams et al., 2010). Sponge tissues tested with thesame techniques had a resistance of 500–1500 O cm 2 and blockedentry of ruthenium red, just as it is blocked by SJs in theepithelium of the planarian  Dugesia  (Fig. 2A–C). Yet, it is still notunderstood how the block occurs in sponges not only because theseptae appear so ‘‘faint’’ in the transmission electron microscope(TEM) images (for siliceous sponges and homoscleromorphs), butalso because the molecular components for SJs known in other invertebrates are not so readily found in the sponge genome. Anintriguing possibility is that sponge cell junctions may combinemolecules used in septate and tight junctions (TJs) to generate theseal. TJs are the sealing junctions of chordates, but cell biologistshave also considered sponges to have TJs based on TEM images(Revel, ’66), and similar ‘‘kissing points’’ (Tsukita and Furuse, ’99)can be found in sponge and mammalian epithelia (compare Figs.14, 15, and 23 in Farquhar and Palade, ’63; with Figs. 3 and 4C, Din Adams, 2010). If sponge epithelia seal and the SJs have anunusual ultrastructure, perhaps the junctions used in sealing in Figure 1.  Cartoons depicting cell junction and basement membrane structures in sponges, other invertebrates, and vertebrates. A selectionof the genes known to be present in each is shown in parentheses. Sequence data from  1 Fahey and Degnan, 2010;  2 Nichols et al. 2006;  y thisarticle. EPITHELIA: A NOVELTY OF METAZOANS  3 J. Exp. Zool. (Mol. Dev. Evol.)  Figure 2.  Sealing junctions. Block of ruthenium red by septate junctions in  A , the planarian  Dugesia   and  in B  and  C ,  Spongilla lacustris  . RC,ruthenium coat; L, lumen; V, vesicle; Ex, external medium; Me, mesohyl. Black arrows mark the point at which ruthenium is blocked frompenetrating the junction (A, from Hori,’87, with permission; and B and C from Adams, 2010, with permission). C, Representative multiplealignment of claudin-like sequences from sponges (AquClaudinSF,  A. queenslandica  : Fahey and Degnan, 2010; Pfi  Petrosia ficiformis  ;Cca  Corticium candelabrum ), anemone (Nve  N. vectensis  ), and human claudin 3. Alignments were generated with MAAFT version 6(http://mafft.cbrc.jp/alignment/server/) and viewed in BioEdit 7.0.5.3. Dashes indicate gaps. The W-GLW and two conserved cysteines aremarked with asterisks. The PDZ domain is marked with a box. Black shading is used to show  4 75% similarity in the residues.  D , Phylogeneticanalysis by maximum likelihood of epithelial membrane proteins (EMP) and claudin protein sequences. The phylogenetic tree was generatedusing an amino acid alignment of EMP and claudin proteins from 22 metazoans with PhyML (Guindon and Gascuel, 2003; http://atgc.lirmm.fr/phyml/). The JTT substitution model was optimized with ProtTest (Abascal et al., 2005; http://darwin.uvigo.es/software/prottest_server.html)with a  g -shaped parameter of 5.186. Statistics for the ML tree: log-likelihood 5  8447.09, unconstrained likelihood 5  1347.81. Thecladogram was generated with FigTree 1.3.1 and rooted using EMPs. Bootstrap values are shown only when they are greater than 50%. LEYS AND RIESGO 4 J. Exp. Zool. (Mol. Dev. Evol.)  the first epithelia used a mix of proteins to seal, and these later diverged to take on the specific functions of those forming tightand SJs in bilaterians.Genomic accounting of the epithelium has been carried out ingeneral assessments of the genomes of   Hydra, Nematostella,Trichoplax, Amphimedon , and  Monosiga  (Putnam et al., 2007;King et al., 2008; Srivastava et al., 2008, 2010; Chapman et al.,2010), but more specifically in comparative analyses of thedomain structure of genes known to be involved in forming theepithelia and junctions (Sakaraya et al., 2007; Alie´ and Manuel,2010; de Mendoza et al., 2010; Fahey and Degnan, 2010); someof these are shown in Figure 3. Among the animals at the base of the metazoan tree, however, drift in sequence is often so greatthat, as noted by Aouacheria et al. (2006), domains, such ascollagen IV, are not recovered by automatic domain detectionsoftware (e.g., Pro-Dom, PFAM, SCOP, PROSITE, and Interpro).Sequence similarity may also be so low between species of distantphyla that phylogenetic analysis does not offer much insight, butthe proteins of these phyla are still estimated to form nearly identical 3D structures. Until the ligands are also known, theactual interactions of these proteins can only be speculated upon. With more data from additional poriferan species, however, itbecomes apparent that even though the sequences are divergent,they do seem to be present in a wider selection of Porifera. Wefocus here on specific examples of genes involved in formingsealing junctions (septate and tight) and the basal lamina. SEALING JUNCTIONS: GENE ACCOUNTING Because sponge epithelia do seal, either SJs or TJs are expected tobe present (it is less likely that sponges have a unique way of forming seals). Given the ultrastructure of faint septae in junctions seen in demosponges and homoscleromorphs and of strong septae in Calcarea (Fig. 2 in Leys et al., 2009), we wouldexpect some of the proteins and genes involved in SJs to bepresent in sponges. SJs are found in the apical side of junctioncomplexes in epithelia of invertebrates, whereas in vertebratesthey are only found at the paranodal junctions of axons and glia(Hortsch and Margolis, 2003).Proteins involved in forming SJs are best known from Drosophila , but exactly which proteins are involved in each junction and what their functions are is still being determined. In Drosophila , there are two types of SJ: smooth SJ (sSJ) whichoccur in the midgut and malpighian tubules, and pleated sheet SJ(pSJ) which occur in the epidermis, salivary gland, andphotoreceptors (so-called primary epithelia). The molecular components of pSJ are best characterized and consist of the adhesion proteins neuroglian, contactin, and neurexin IV (a contactin associated protein or Caspr). All three have beenshown to be necessary for barrier function at pSJ in the fly, but of these only neurexin IV is needed to retain the ultrastructure of septae that characterize this junction (Baumgartner et al., ’96;Genova and Fehon, 2003; Faivre-Sarrailh et al., 2004). NeurexinIV is not found at sSJ, which also lack strong ladder-like septae inelectron micrographs (Baumgartner et al., ’96; Baumann, 2001).sSJ are characterized by fasciclin III, ankyrin, and  a , b -spectrin.Neuroglian is not localized to sSJ, but rather is found aroundmuscle and regenerative cells in the midgut (Baumann, 2001). So,the proteins that one might be looking for in order to assesswhether SJ are in sponges or arose as a novelty of cnidarians,depend on what kind of SJ we expect sponges to have. A neurexin IV homolog, found in the homoscleromorphsponge  Oscarella carmela  (Nichols et al., 2006; Nichols, UCBerkeley, personal communication), aligns well with twofragments found in the genome of the demosponge  Amphimedonqueenslandica  (Fahey and Degnan, 2010). The sponge sequencesare short, so it is not known if they are able to function in thesame way as other neurexin IVs. Neither neuroglian nor contactinwere found in  Amphimedon , but other genes associated with pSJin  Drosophila  do have homologs in this sponge, such as thoseencoding for the linker proteins Scribble and Discs-large,although these may also be associated with other aspects of cellpolarity (Fahey and Degnan, 2010). Genes encoding for proteinsof the 4.1m family, which has been shown to be essential for barrier function in  Drosophila  (Lamb et al., ’98), were not assessedin the study by Fahey and Degnan (2010), nor was fasciclin III,but blast searches identify potential fragments in the  A. queenslandica  sponge genome. The ultrastructure of smoothSJ in  Drosophila  is very similar to that of SJ seen in sponges, andgiven that the sponge epithelium functions more like the tubulesof a gut (for absorption and secretion and slow peristalticcontractions, rather than for maintaining strength or robustness),it would not be at all surprising if a similar combination of proteins to those in sSJ make up the sponge junctions. On theother hand, we might expect a neurexin-based complex incalcareous sponges where the ultrastructure shows firm septaebetween choanocytes and between spicule secreting sclerocytes(Ledger, ’75; Green and Bergquist, ’79). In sum, although there aremany proteins involved in SJ for which homologs of genes havenot yet been sought in sponges, with those already identifiedthe sponge might operate a SJ and use this to seal. However,immunocytochemistry is needed to determine the location of theproteins and RNAi or equivalent tests are needed to confirmfunction. It would not be surprising to find some of these proteinslocalized to SJs, but sponges may also use other novel proteins inbarrier function, and one possibility is claudins.It is a fairly recent discovery that claudins are involved in SJbarrier function in both  Drosophila  and  C. elegans  (Asano et al.,2003; Nelson et al., 2010). Claudins are members of the PMP-22/EMP/MP-20/Claudin superfamily, which together with occludinsand junctional adhesion molecules are involved in integrity andcontrol of paracellular transport in vertebrate TJs (Heiskala et al.,2001). This complex interacts with MAGUK homologs (ZO 1–3)and other proteins to bind to the actin cytoskeleton of the cell,and variation in TJ permeability occurs by variability in claudin EPITHELIA: A NOVELTY OF METAZOANS  5 J. Exp. Zool. (Mol. Dev. Evol.)
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks
SAVE OUR EARTH

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!

x