BIOLOGICAL OVERVIEW

The Drosophila Ten-m (also called Tenascin-major, or odd Oz (odz)) gene has been associated with a pair-rule phenotype. This study identified and characterized new alleles of Drosophila Ten-m to establish that this gene is not responsible for segmentation defects but rather causes defects in motor neuron axon routing. In Ten-m mutants the inter-segmental nerve (ISN) often crosses segment boundaries and fasciculates with the ISN in the adjacent segment. Ten-m is expressed in the central nervous system and epidermal stripes during the stages when the growth cones of the neurons that form the ISN navigate to their targets. Over-expression of Ten-m in epidermal cells also leads to ISN misrouting. It was also found that Filamin, an actin binding protein, physically interacts with the Ten-m protein. Mutations in cheerio, which encodes Filamin, cause defects in motor neuron axon routing like those of Ten-m. During embryonic development, the expression of Filamin and Ten-m partially overlap in ectodermal cells. These results suggest that Ten-m and Filamin in epidermal cells might together influence growth cone progression (Zheng, 2011).

Ten-m was thought to be the first non-transcription factor pair-rule gene. Three P-element insertion alleles were reported to lie in the same complementation group and be embryonic lethal. One allele, 5309, showed a severe pair-rule phenotype, one allele showed a moderate version of the same phenotype, whereas other lethal alleles did not have a significant cuticle phenotype. This study showed that the cuticle phenotype associated with the original 5309 stock segregates with the balancer chromosome. Indeed, the Df(3L)Ten-m^AL1 and Df(3L)Ten-m^AL29 deletion alleles exhibit fusions similar to the 5309 allele when balanced with the same balancer chromosome as the one used in 5309. The data show that the 'Ten-m' cuticle phenotype analyzed previously is an odd-paired mutation on the balancer chromosome and is not due to the Ten-m mutation itself. It should be noted that the embryonic CNS longitudinal connective discontinuity phenotype documented for Ten-m⁵³⁰⁹ is also due to a mutation on that original balancer (Zheng, 2011).

When Ten-m is not expressed, or is ectopically expressed in the ectoderm, aberrant motor axon growth cone guidance occurs. Growth cones use various kinds of substrates and guidance cues to navigate through a specific path to find their targets. In Drosophila and C. elegans, genetic screens have identified many secreted or transmembrane guidance cues including Netrins, Semaphorins, Slits, Nephrins, and classic morphogens that also act as guidance molecules. Most of these cues are expressed either by axonal tracts themselves or along the axonal trajectory by peripheral tissues, such as muscles. A transmembrane protein affecting migration, Ten-m, is expressed in epidermal stripes and in neurons. In both vertebrate and invertebrate embryos, axons must first exit the CNS, and then migrate along stereotyped pathways to reach their specific targets in the periphery. Since motor axons in both Ten-m loss of function mutants, and Ten-m ectopic expression embryos, exit the CNS normally but do not migrate along their specific paths, Ten-m appears to be dispensable for axonal extension but necessary for pathfinding decisions in the periphery (Zheng, 2011).

Ten-m is often required for correct choice point determination. During embryonic development, axons preferably extend along the surface of other axons to form axon bundles or fascicles (selective fasciculation), and exit those fascicles to navigate into their targets (selective defasciculation). These processes are regulated by both attractive and repulsive cues. These attractive and repulsive molecules can originate from the axonal tracts themselves or from the surrounding peripheral tissues. A guidance cue disruption of a repulsive molecule should cause abnormal defasciculation. In the case of Ten-m loss of function mutations, or gain of function, the ISN branches fail to maintain their segmental boundaries and invade the adjoining segments, occasionally to fuse with the adjacent segmental nerve. In mutants, loss of Ten-m activity in the motor axons could lead to failing to respond properly to cues, among other possible failings of the neurons. However, epidermal ectopic, or Paired-driven Ten-m expression leading to the same phenotypes suggest that Ten-m impacts peripheral cues, assuming that the overexpression effects were specific (Zheng, 2011).

Axon guidance disruptions observed when Ten-m is ectopically expressed suggests that Ten-m either: maintains peripheral cells to allow them to reach a stage to express a cue for motor axons; or more directly regulates the expression of cues to which motor axons respond. It is speculated that Ten-m might itself act as a peripheral cue for migration. Epidermal Ten-m, and perhaps specifically its expression spatially restricted to stripes, could help position a repulsive guidance cue for the ISN axons and prevent them from crossing into the adjacent segments. Given the collection of defects observed in motor neuron axons, Ten-m might induce a gene product that is, or might itself be, both a repulsive and an attractive guidance molecule, a situation that is not uncommon. For example, Netrins were first found as a chemoattractant for vertebrate commissural axons and circumferential axons in C. elegans. However, Netrins can also repel some axons, as demonstrated in unc-6/netrin mutant C. elegans and Drosophila. DCC/frazzled (deleted in colorectal carcinoma), a netrin receptor, mediates both attraction and repulsion while UNC-5, another netrin receptor, functions exclusively in repulsion (Zheng, 2011).

Filamins are very large proteins with an actin binding domain and more than 20 Ig-like repeats, that self associate as dimers. They act as Actin crosslinking proteins that are also scaffolds for a very large number and variety of binding partners. As such, they are involved in many functions, but especially relevant are cell adhesion and migration. This includes interactions with different cytoskeletal complexes. In flies, Filamin affects peripheral motor axon navigation in a manner similar to that of Ten-m. This function echoes vertebrate filamin activities. In contrast to Drosophila, mammals have three filamin proteins, A, B and C. Loss of function mutations in Filamin A are found in the human disease periventricular heterotopia, which is a defect in axonal migration that has been associated with the dynamic regulation of actin (Eksioglu, 1996; Kakita, 2002; Fox, 1998; Moro, 2002). However, detailed studies of patients carrying mutations shows evidence for a more complex regulation of axonal navigation than can be explained by simply an effect on growth cone motility. The current studies in flies suggest that in addition to growth cone motility, the context and restricted expression pattern of Filamin might influence axon guidance. In Drosophila, Filamin associates with the seg1 domain of Ten-m, a highly conserved domain within the Ten-m/Odz family that this study has named the FID. These two proteins are expressed in the epidermis, including co-expression in a series of 'belts' of epidermal cells, strongest laterally (Zheng, 2011).

How might Ten-m, together with Filamin, regulate axonal guidance? It is hypothesized that anchored epidermal Ten-m and Filamin might influence lateral motor axon navigation. These two proteins might set the stage for proper development leading to the expression of spatially restricted epidermal axon guidance cues, or directly impact the regulation of such cues, as ISN motor axon projections start to migrate out of the CNS and begin to reach their lateral and dorsal muscle targets. One speculation is that Ten-m linked to Filamin could itself be a candidate cue for motor axons (Zheng, 2011).

Ten-m, a Drosophila gene related to tenascin, is a new pair-rule gene

Ten-m is the first example of a known pair-rule gene product acting from outside the cell. Every odd-numbered body segment is deleted in Ten-m mutants. The protein is also a putative transmembrane tyrosine kinase substrate. The protein is under control of fushi tarazu and even-skipped (Baumgartner, 1994 and Levine, 1994). Ten-m alters wingless and engrailed expression through odd-paired's effects on paired and sloppy paired (Baumgartner, 1994).

Splice variants of the Drosophila PS2 integrins differentially interact with RGD-containing fragments of the extracellular proteins tiggrin, ten-m, and D-laminin 2

Two new potential ligands of the Drosophila PS2 integrins have been characterized by functional interaction in cell culture. These potential ligands are a new Drosophila laminin alpha2 chain encoded by the wing blister locus and Ten-m, an extracellular protein known to be involved in embryonic pattern formation. As with previously identified PS2 ligands, both contain RGD sequences, and RGD-containing fragments of these two proteins (DLAM-RGD and TENM-RGD) can support PS2 integrin-mediated cell spreading. In all cases, this spreading is inhibited specifically by short RGD-containing peptides. As previously found for the PS2 ligand Tiggrin (and the Tiggrin fragment TIG-RGD), TENM-RGD induces maximal spreading of cells expressing integrin containing the alphaPS2C splice variant. This is in contrast to DLAM-RGD, which is the first Drosophila polypeptide shown to interact preferentially with cells expressing the alphaPS2 m8 splice variant. The betaPS integrin subunit also varies in the presumed ligand binding region as a result of alternative splicing. For TIG-RGD and TENM-RGD, the beta splice variant has little effect, but for DLAM-RGD, maximal cell spreading is supported only by the betaPS4A form of the protein. Thus, the diversity in PS2 integrins due to splicing variations, in combination with diversity of matrix ligands, can greatly enhance the functional complexity of PS2-ligand interactions in the developing animal. The data also suggest that the splice variants may alter regions of the subunits that are directly involved in ligand interactions, and this is discussed with respect to models of integrin structure (Graner, 1998).

Curiously, the ten-m gene is expressed in an embryonic pair-rule pattern, and ten-m mutants display pair-rule patterning defects. Since the protein influences expression of downstream genes, it must communicate its presence to the cell nucleus. However, it does not appear that integrin signal transduction is important in early embryonic segmentation. PS integrins are not strongly expressed at this time, and, more importantly, mutations in integrin subunit genes do not cause segmentation phenotypes (Graner, 1998 and references).

Ten-m is later localized (among other places) at muscle attachment sites, where integrins are known to accumulate. This localization of Ten-m in vivo, as well as the demonstration of TENM-RGD interactions with PS2 integrins in vitro, suggests that Ten-m may function with PS2 integrins in muscle attachment. Interestingly, the heparan sulfate-containing protein D-syndecan also localizes to muscle attachments, and Ten-m contains a consensus heparin-binding sequence near the RGD, suggesting the potential of a Ten-m-syndecan-integrin complex. Syndecan proteoglycans recently have been shown to be important for signal transduction in vertebrate cell focal adhesions (Graner, 1998 and references).

The available data, although very suggestive, do not demonstrate unequivocally that Ten-m serves as an integrin ligand at muscle attachment sites. However, other potential PS2 ligands, such as Tiggrin, also accumulate at muscle attachment sites, and genetic studies of tiggrin suggest considerable functional redundancy among the extracellular matrix components there. Because of this redundancy, a direct genetic demonstration of a role for Ten-m in muscle attachment may require simultaneous disruption of multiple genes encoding matrix proteins, and the early embryonic phenotype of ten-m mutants will further complicate such an analysis. One potential approach might be to demonstrate a dominant genetic effect of ten-m mutations in a background that has been sensitized for loss of function phenotypes by viable mutations in other genes that encode proteins important for muscle attachment or other integrin-dependent processes (Graner, 1998).

Drosophila Ten-a is a maternal pair-rule and patterning gene

The Ten-a gene of Drosophila encodes several alternative variants of a full length member of the Odz/Tenm protein family. A number of Ten-a mutants created by inexact excisions of a resident P-element insertion are embryonic lethal, but show no pair-rule phenotype. In contrast, these mutants, and deficiencies removing Ten-a, do enhance the segmentation phenotype of a weak allele of the paralog gene odz (or Ten-m) to the odz amorphic phenotype. Germ line clone derived Ten-a⁻ embryos display a pair-rule phenotype which phenocopies that of odz. Post segmentation eye patterning phenotypes of Ten-a mutants establish it as a pleiotropic patterning co-partner of odz (Rakovitsky, 2007).

A recently produced deletion mutation of the Drosophila melanogaster gene Ten-a removes the entire gene, yet does not lead to lethality (Zheng, 2011). Thus the lethality and the pair-rule gene roles reported for Ten-a in Mech. Dev. 124, 911-­924 (2007), cannot be attributed to this gene, but to second site loci in mutants studied. Given the centrality of the pair-rule claim for this paper, it has been retracted. The other findings of the paper remain as reported, particularly molecular and biochemical characterizations of the Ten-a gene, its alleles and its transcripts Rakovitsky, 2011.

Unlike other pair-rule (P-R) genes, the Drosophila zygotic P-R gene odd Oz (odz or Ten-m) encodes a membrane anchored cell surface protein, and not an obvious transcription factor, with many indications that it is involved in patterning. These appear to be receptor like proteins, with discordant yet convergent evidence that discrete domains of the proteins might be processed into elements involved in transcription. The first phenotypes for non-Drosophilan Odz/Tenm mutations were recently described, further supporting this family's requirement for proper patterning and development. Phenotypes of mouse Odz4 mutations include failure in gastrulation and somite formation, and a small deletion of the C. elegans ortholog, Ten1, leads to early embryo arrest and gonadal defects (Lossie, 2005; Drabikowski, 2005; Rakovitsky, 2007 and references therein).

Ten-a was initially identified as a gene encoding a relatively short protein with EGF-like repeats at its C-terminus. Genomic sequence made it clear that the D. melanogaster genome project Ten-a gene model, once 'merged' with gene model annotations lying adjacent and 'downstream' to Ten-a on the X chromosome (gene annotations CG2590 and CG2578), constitutes a full length gene paralog of Ten-m/odz. An examination of cDNAs verified that a Ten-a transcript exists that encodes a full length 300 kDa protein of the Odz/Tenm family, with 48% amino acid similarity to Drosophila odz. A number of the expression sites of Ten-a in embryos suggest extensive overlap with odz expression (Rakovitsky, 2007).

This study has described lethal and semi-lethal mutations caused by small deletions that remove a region containing an exon of an alternative splice form of Ten-a transcripts. It is likely that it will be necessary to examine mutations that disrupt exons shared between all of the splice forms in order to see truly amorphic phenotypic consequences. Nonetheless, the levels of all tested Ten-a transcripts are reduced significantly in the small lethal deletions. This is documented by RT-PCRs performed on RNA from post-gastrulation male embryos, a stage when any maternal stores should be exhausted. A lack of transcripts was shown using a primer pair bridging the fourth to fifth exons, and a significant reduction was shown using a primer pair bridging the eighth to ninth exons, in coding regions shared by essentially all protein forms of Ten-a. The loss of transcripts, and the phenotypes affecting viability, are not likely due solely to the consequences of the loss of exon 1d. Rather, it is assumed that some crucial regulatory region or elements must be impacted by these small deletions (Rakovitsky, 2007).

There are predicted full length Ten-a proteins that span the original CG15733, CG12720, and Ten-a annotations, as well as full length Ten-a proteins that are derived from only the original Ten-a annotation's exons. Yet other transcripts encode both a full length Ten-a protein as well as exons in the transcript's 5' UTR with capacity to direct attenuated translation of smaller upstream gene products. This implies the possible alternative translation of cryptic regulatory short polypeptides from the transcript versus the full Ten-a protein reading frame, through regulation of the utilization of IRES sequences. Perhaps most intriguing is the existence of protein domains (domains of the original CG12720 annotation's exons) that appear alternatively in the attenuated polypeptide products from some transcripts, and at other times as the N-terminus of full length Ten-a protein from other transcripts. It should be noted that these N-terminal regions are among the most poorly conserved among proteins in the family, so that inferences for other homologs can not immediately be drawn. What is shared among all homologs are proportionately very large 5' introns among all sequenced species (except for C. elegans, which employs trans-splicing). This implies further extensive regulation in the 5' region of the genes, as do IRES sequences resident in the most upstream exons (Rakovitsky, 2007).

Among the alternatively spliced messages uncovered, there are encoded proteins lacking the hydrophobic domain close to the N-terminus present in all family homologs, and others with only a subset of the EGF-like repeats. These many variations are reminiscent of reports of Odz/Tenm protein variants in other metazoans. The existence of varied validated Mouse Odz4 proteins, from different regions of the gene and of different lengths, emphasizes the complexity of the gene products deriving from Odz/Tenm family loci (Lossie, 2005). Ultimately, the many optional messages support new predictions for membrane-bound and other deployments for different variants of these proteins. Perhaps these variants and isoforms can explain the differing and sometimes mutually exclusive results that have been reported for proteins of this family in the past. The multiple splice variants and resulting protein forms can explain the different biochemical, immunocytochemical, and membrane deployment observations that have been made for this family's homolog gene products (Rakovitsky, 2007).

Ten-a impacts proper segmentation through both zygotic and maternal contributions. It zygotically enhances a weak odz cuticle phenotype with high penetrance. In germ line clones it leads to a very strong, if not canonical, pair-rule phenotype likely caused by maternal and zygotic contributions. This occurs in a minority of the embryos, in a population that includes those with weak and no segmentation phenotypes. Both maternal and zygotic activities appear necessary, but a clearcut 'ratio' of weights of these two contributions will likely only be clear when null Ten-a mutations will be examined (Rakovitsky, 2007).

The germ line mosaic derived embryos deprived of Ten-a contribution provide evidence that activity from the Ten-a gene locus is necessary for segmentation at the pair-rule stage of segmentation. These embryos display a full-blown pair-rule phenotype in the same register as odz mutants. This case of a gene displaying a maternally dependent P-R phenotype is rare, but has been seen previously in kismet. It is likely that Ten-a acts in a direct manner, and not through downstream events, such as is the case for kismet. This is most evident from the disruption of slp expression in the odd parasegments of embryos mutant for odz and Ten-a. Striped slp RNA expression is not initiated, most likely due to the involvement of these paralogous gene products in transcription (Rakovitsky, 2007).

Ten-a⁻ mutants and deficiency chromosomes zygotically enhance weak odz segmentation phenotypes in a highly specific manner. There are indications that Odz and Ten-a proteins form heterodimers, as has been well established for the four vertebrate homologs (Feng, 2002). The segmentation phenotypes of each gene's mutations are in the same segmental register, and essentially are identical. Therefore, whether or not they prove to act as dimer proteins in this context, they are at least involved in the same process at the same time. It is thought that both of the two gene products are necessary for a concerted, coordinated activity that contributes to proper segmentation at the P-R stage. While each alone can cause the 'odd-fused' phenotype in the proper zygotic or maternal mutant context, it is assumed that mutants of each deplete the same active complex. Given the Ten-a maternal contribution, it is possible that the Ten-a protein is ubiquitous throughout the early embryo. It is therefore envisioned that the region of activity of both family-paralogs as spatially delimited to the striped domain of deployment of the Odz protein in the context of segmentation (Rakovitsky, 2007).

slp expression in odd parasegments requires Runt and Odd paired (Opa) activity, as opposed to slp expression's dependence on Hairy and Ftz in even parasegments. This study showns that the initiation of the odd parasegment expression is also dependent on Odz and Ten-a. The Odz striped domain covers odd parasegments, whereas Runt is expressed in the posterior half of the odd parasegment and anterior half of the even parasegment, and whereas Opa and Ten-a are ubiquitous. The domain of Odz and Runt expression overlap is therefore in the posterior half of the odd parasegment. This overlap corresponds to the unique domain of slp odd parasegment transcriptional activation. A complex, or confluence of activities, of: the two Odz/Tenm family homologs; Runt; and Opa can be envisioned turning on slp expression and provide an excellent context to be followed in order to establish the nature of downstream outcomes of odz homologs' activities. Whether the two Tenm/Odz paralogs' proteins interact directly with Runt and Opa in transcription, or whether they initiate a chain of interactions that results in these transcription factors' alterations, must still be clarified. It is also an alternative possibility that Odz and Ten-a influence the levels of odd parasegment slp RNA post-transcriptionally, for instance through an effect on slp transcript stability (Rakovitsky, 2007).

The atypical nature of odz and Ten-a as P-R genes extends to more than their proteins' structures and subcellular deployment. These are extremely large genes, transcribing larger than 120 kb nascent messages. This is essentially unheard of for early segmentation genes in Drosophila, where their nascent transcript sizes cluster around 2 kb. At a transcription rate of 1.4 kb per minute, interphase lengths as short as 3.5 min during syncitial blastoderm, and evident transcription abortion at each mitosis, transcription units are thought to be limited to 5 kb in order to function at early embryonic stages. This size restriction was shown to be the causal difference between knirps acting as a segmentation gap-gene, versus knirps-related only functioning at later stages, due to its larger transcript size (Rakovitsky, 2007).

This renders Ten-a and odz zygotic gene function complex to invoke, given the 1.5 h needed to transcribe them. They are proposed to act first at the cellular blastoderm, given the first appearance of Odz protein, given the timing of appearance of genes proven downstream to them, and given their cell-transmembrane deployment. Yet even the cycle 14 interphase is not clearly adequate to allow for synthesis of their full transcription unit. Instead it is most likely that the key regulatory step to produce these proteins, when needed just post-cellularization, is translation of partial messages or pre-existing messages. For odz, RNA has never been observed in a stripewise manner in the cellular blastoderm, unlike the seven observed stripes of protein. This implies possible regulated translation, which can be consistent with the presence of IRES sequences in Odz transcripts. This study showns that Ten-a has a maternally provided component, but have not examined whether or not odz does. In both cases, the zygotic contribution is unquestionably critical, raising the possibility that some length transcript form is adequately synthesized at cellular blastoderm. How this compliments gene activity afforded by the maternal contribution will need to be clarified in future work (Rakovitsky, 2007).

Phenotypes of Ten-a mutant eyes, and of Ten-a plus odz transhetereozygous combinations, attest to the importance of Ten-a in this second tissue system. Given the nature of the phenotypes, and within the context of what is known about Odz function, the importance of Ten-a in these cases is likely to center on its patterning roles. The strong phenotypes of transheterozygous combinations of odz and Ten-a alleles, that each alone display no eye phenotypes, suggest that they cooperate closely in eye patterning. Thus Ten-a is a maternally required pair-rule gene with likely far reaching patterning dimensions in many contexts. Understanding of Odz/Tenm family contributions to metazoan patterning can now be furthered in a system in which all (both) members can now be coordinately manipulated and studied (Rakovitsky, 2007).

GENE STRUCTURE

Exons - six

PROTEIN STRUCTURE

Amino Acids - 2515

Structural Domains

The extracellular domain contains eight tenascin like EGF repeats. These are followed by a putative transmembrane region (Levine, 1994). The putative intracellular domain contains putative fibronectin domains (Baumgartner, 1994).

Ten-m possesses some, but not all, of the features common to most vertebrate tenascins. For example, Ten-m is a secreted glycoprotein with eight tenascin-type EGF-like repeats and putative fibronectin-type III repeats (Baumgartner, 1994). Ten-m lacks a tenascin C-terminal fibrinogen-like domain, and the Ten-m RGD sequence is found 72 amino acids from the C terminus. Recombinant protein fragments containing this RGD sequence promote RGD-dependent, PS2 integrin-mediated cell spreading better with cells expressing the PS2C splice variant than with cells expressing the PS2m8 variant (Graner, 1998).

Levine (1994) reported a partial cDNA sequence from the ten-m gene; this partial sequence stops short of the final 325 amino acids and thus does not include the RGD tripeptide near the C terminus; it also includes 216 N-terminal residues not reported by Baumgartner. Levine ascribed properties to the presumed polypeptide that are significantly different from those deduced by Baumgarter; for example, Levine suggests that Odd Oz is a transmembrane phosphoprotein with tenascin homology in its putative extracellular domain, and the earlier study also proposes that the polypeptide is cleaved into smaller mature proteins. These apparent discrepancies have yet to be resolved, and it is possible that the protein functions in different forms. In any case, Baumgartner found that a Ten-m polypeptide could be found in conditioned media from Drosophila cells, and so a secreted form is present in at least some instances (Graner, 1998).

date revised:  13 September 2011

Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.

The Interactive Fly resides on the
Society for Developmental Biology's Web server.