BIOLOGICAL OVERVIEW

The human tumor suppressor gene PTEN gets its name from its biochemical function, its domain structure and its chromosomal location: PTEN stands for the combination of phosphatase and tensin homolog on chromosome 10 (J. Li, 1997). The protein is also known as MMAC1 (Steck, 1997) or TEP1 (D. Li, 1997). The protein exhibits phosphatase activity against proteins and lipid phosphate residues. The PTEN protein contains a protein phosphatase domain with similarity to dual specificity phosphatases (D. Li, 1997; Li and Sun, 1997; Steck, 1997). Dual specificity phosphatases act both on serine/threonine and tyrosine phosphotyrosine residues. The phosphatase domain is embedded within a domain with homology to an actin-binding motif in the focal adhesion-associated protein tensin. Immunohistochemical evidence has suggested that like tensin, PTEN is associated with the cytoskeleton (D. Li, 1997).

Now that the domain structure of PTEN has been described, the known and potential functions of the PTEN homolog in Drosophila and mammals is presented. The lipid phosphatase function of PTEN places it in the middle of the insulin pathway, known to involve lipid signaling (Goberdhan, 1999; Huang, 1999). Drosophila Pten modulates cell size, and consequently tissue mass, by acting antagonistically to the lipid modifiying enzyme Phosphotidylinositol 3 kinase 92E, also known as Dp110, and its upstream activator Chico, an insulin receptor binding and signal transduction protein (Goberdhan, 1999). All signals from the Insulin-like receptor can be antagonized by Pten (Huang, 1999). In terms of its protein phosphatase function, mammalian PTEN targets focal adhesion kinase, a major effector of cytoskeletal function. Overexpression of wild-type mammalian PTEN and mutant PTEN that lacks lipid phosphatase activity can reduce levels of focal adhesion kinase (FAK: see Drosophila Focal adhesion kinase-like) phosphorylation and the formation of focal adhesions, thereby inhibiting cell migration and invasiveness (Tamura, 1998). In terms of its cytoskeletal connection, Drosophila Pten appears to regulate the subcellular organization of the actin cytoskeleton in multiple cell types. The bristle, hair, and rhabdomere phenotypes observed in Drosophila Pten mutant tissue have not been reported in flies defective in insulin or Dp110 signaling, indicating that unlike the Pten-linked growth defects, involving insulin signaling, these effects are probably not derived from alterations in lipid signaling but from direct influences on cytoskeletal function (Goberdhan, 1999).

The tumor suppressor gene PTEN is one of the most frequently mutated genes involved in the development of mammalian cancer. PTEN mutations are found in a wide variety of tumors such as glioblastomas, endometrial carcinomas, advanced prostate cancers and melanoma cells. Germ-line mutations in PTEN are linked to three rare autosomal dominant syndromes: Cowden Disease, Bannayan-Zonana syndrome and Lhermitte-Duclose disease. A common feature of these syndromes is a predisposition for the development of hamartomas: benign tumors that have differentiated but disorganized cells. Pten mutant mice exhibit early embryonic lethality (E7.5-E9.5) and the heterozygotes display a predisposition to tumor development. Consistent with the role for PTEN as a lipid phosphatase, Pten mutant cells exhibit increased lipid signaling. The C. elegans gene daf-18 encodes a distant PTEN homolog. In addition to a conserved phosphatase domain, daf-18 encodes a large non-homologous C-terminal region. Daf-18 is associated with the insulin signaling pathway. In C. elegans, daf-18 mutants can suppress the mutant phenotypes of daf-2, the C. elegans insulin-like receptor, and age-1, the C. elegans PI3K homolog (Huang, 1999 and references therein).

Drosophila Pten was identified in two laboratories using two different approaches, one based on homology to mammalian PTEN (Huang, 1999) and the second based on intensive analysis of a chromosomal region containing several other genes coding for signaling proteins (Goberdhan, 1999). This second approach will be examined in detail, since it foreshadows future approaches that will be taken to analyze Drosophila genes based on genomics. Previously, two Drosophila genes had been identified in the vicinity of the gene that was later to be identified as Drosophila Pten: Dror encodes a neural-specific receptor tyrosine kinase, and basket encodes the homolog of c-Jun amino-terminal kinase. These two genes map adjacent to each other at 31B/C on the second chromosome. Several deficiency chromosomes have been characterized that uncover one or both of these genes and also affect the adjacent gene chico , which codes for a homolog of mammalian insulin receptor substrates, IRS1-4 (Böhni, 1999). One of these deficiencies, Df(2L)170B, is of particular interest because it deletes sequences proximal to Dror and produces an overgrowth phenotype in homozygous clones. This effect was not observed with deficiencies [Df(2L)41C and Df(2L)147F] that only delete Dror, DJNK, and chico. Clones generated with these smaller deficiencies contain reduced numbers of small cells attributable to the loss of chico function (Böhni, 1999). During a chemical mutagenesis screen using the Df(2L)170B chromosome, a new lethal complementation group has been identified that maps proximal to Dror and affects tissue growth. Sequence analysis of genomic and cDNA clones reveals a novel gene at this locus encoding the Drosophila PTEN homolog, Pten (Goberdhan, 1999).

Drosophila Pten regulates cell number and size and affects assembly of specific cytoskeleton-dependent structures. Because animals transheterozygous for strong Pten alleles die with no obvious phenotypes, the functions of this gene have been elucidated further by generating homozygous mutant clones in heterozygous animals using the FLP/FRT system. Two Pten alleles, DPTEN¹ and DPTEN³, produce growth phenotypes slightly more severe than those generated by Df(2L)170B, the chromosomal deficiency deleting Pten, Dror, DJNK, and chico. DPTEN¹ and DPTEN³ also behave in a manner similar to the deficiency in combination with a weak, nonlethal Pten mutation; this suggests that they are strong or null Pten alleles (Goberdhan, 1999).

The adult compound eye in Drosophila is composed of ~700 regularly arranged unit eyes or ommatidia, each containing the same complement of photoreceptors and accessory cells. Mutant clones can be generated in the eye and other tissues by somatic recombination of homologous chromosomes in heterozygous animals. If only the nonmutant chromosome and not the mutant chromosome is marked with a copy of the white (w⁺) gene, mutant clones can be distinguished by the absence of pigment. The wild-type twin spot resulting from a recombination event with these chromosomes contains increased levels of pigment relative to the remainder of the eye, because the cells in this region carry two copies of the w⁺ gene. Ommatidia within clones homozygous for DPTEN¹, DPTEN³, or Df(2L)170B are enlarged and bulge from the surface of the eye. Ommatidial facets are occasionally fused, and there frequently are clusters of morphologically abnormal interommatidial bristles or clustered sockets in mutant regions, particularly (but not exclusively) in the dorsal half of the eye. To assess the effect of loss of Pten function on the regulation of cell number, the numbers of ommatidia in multiple mutant clones and their wild-type twin spots were compared. Mutant clones typically contain at least twice as many ommatidia as neighboring twin spots, suggesting that mutant cells overproliferate, a result supported by analysis of clones at larval stages. The largest Pten clones include more than half of the adult eye and form a hyperplastic tumor-like overgrowth. This hyperplastic phenotype, however, is mild compared to the effects of several other tumor suppressor mutations in Drosophila (Goberdhan, 1999).

Sections demonstrate that the number and identities of ommatidial cells are not obviously altered in Pten mutant clones, indicating that this gene does not have a major role in specifying cell fate in the eye. Analysis of nonpigmented mutant clones, however, reveals that the cell bodies of all mutant photoreceptors are greatly enlarged, but not those of their wild-type/heterozygous neighbors. By using a w⁺-marked mutant chromosome to visualize the pigment cell layer in clones, it is possible to show that the volume of this layer is also increased. These cellular growth defects account for the increased size of mutant ommatidia. Furthermore, in mutant photoreceptors, the light-sensing rhabdomeres are elliptical and not circular in cross-section. This phenotype is observed in several mutants that are believed to affect assembly of the actin cytoskeleton. Analysis of ommatidia that contain both wild-type and mutant photoreceptors reveals that both the cell size and rhabdomeric phenotypes are cell-autonomous (Goberdhan, 1999).

Cells in the wing, each of which normally possesses a single wing hair, are also significantly enlarged in mutant Pten clones. Wing hairs are occasionally duplicated, a defect observed under conditions where the cytoskeleton of wing epithelial cells is disrupted. In addition, crossvein formation in the wing, a process involving specialized contacts between wing epithelial cells and the extracellular matrix (ECM), is abnormal in mutant regions. In ~70% of cases, anterior cross-veins are absent or partially absent in mutant clones, whereas the posterior cross-vein is affected in ~35% of cases. Occasionally, extra wing vein material is observed. Longitudinal veins are affected only very rarely (Goberdhan, 1999).

Overexpression of Pten also produces enlargement of wing cells. Wild-type Pten cDNA is overexpressed in particular areas of the wing using the GAL4-UAS system. Initially flies carrying a dpp-GAL4 construct were used. This drives gene expression in cells that will normally populate the region between the third and fourth longitudinal wing veins (LIII and LIV). Overexpression of Pten reduces the size of these regions by nearly 25% compared with wild type. This is not a consequence of a general reduction in wing size in overexpressing flies, since an adjacent area of the wing between LIV and LV is essentially unaffected. The effect on wing area is similar to that produced by overexpression of Dp110^D954A, a dominant-negative, kinase-dead version of Pi3K92E. The reduction is caused by both a decrease in cell size and cell number and is opposite of the effect of overexpressing an activated, membrane-associated form of Pi3K92E, Dp110-CAAX, in the same region. To test whether Pten's growth regulatory functions are primarily mediated by its effects on the insulin receptor-Pi3K92E signaling pathway and not by an independent signaling cascade, the genetic effects of Pten alleles were sought using mutant phenotypes associated with chico and Pi3K92E (Goberdhan, 1999).

Interestingly, both overexpression of a dominant negative form of Phosphotidylinositol 3 kinase 92E (also known as Dp110 or Pi3K92E) and mutations either in the Drosophila insulin receptor or in chico/IRS1-4 also reduce cell size as well as proliferation. Furthermore, overexpression of wild-type and activated forms of Pi3K92E produces similar size and proliferation defects as those seen in Pten mutant cells. These observations are consistent with a model in which growth is regulated in Drosophila by specific phosphoinositides whose levels are controlled by the balance of Pten and Pi3K92E activities. Pten and Pi3K92E. At 25°C, flies do not survive to adult when wild-type Pi3K92E is ectopically expressed by means eyeless-GAL4. Interestingly, this lethality can be rescued by coexpression of Pten. Furthermore, the small eye phenotype of Pten overexpression is suppressed by overexpression of wild-type Pi3K92E and enhanced by overexpression of dominant negative Pi3K92E. These results clearly indicate that Pten and Pi3K92E function antagonistically in Drosophila. The recent characterization of chico, a Drosophila IRS1-4 homolog, has shown that chico, Pi3K92E and Insulin receptor(Inr) act as positive elements in a Drosophila insulin signaling pathway to regulate cell proliferation and cell size (Bohni, 1999). Consistent with the role of Pten as a negative regulator in this insulin pathway, removal of one copy of the chico gene genetically enhances the eye/Pten eye phenotype. Overexpression of Inr (eye/Inr) causes lethality at 25°C. At room temperature, few animals survive with overproliferated eyes. Strikingly, co-overexpression of Pten completely rescues lethality and the overproliferation phenotype. This suggests that all signals from the insulin receptor can be antagonized by Pten function. Together with the previous findings that mammalian and C. elegans PTEN molecules interact with components of the insulin pathway (Furnari, 1998; Maehama, 1998; Myers, 1998; Ogg , 1998; Stambolic, 1998; Sun, 1999), these genetic data argue that PTEN functions as a major conserved negative regulator in the insulin signaling pathway (Huang, 1999).

The importance of Drosophila Pten in negatively regulating the growth-promoting effects of insulin signaling in vivo, however, is best illustrated in homozygous clones mutant for both chico and Pten. In these clones, the reduced growth phenotype normally seen in chico mutant cells is masked completely by the overgrowth phenotype associated with loss of Pten function, suggesting that Pten normally has a critical role downstream of Chico in maintaining growth-promoting signals at nonhyperplastic levels (Goberdhan, 1999).

Pten also affects the organization of the actin cytoskeleton in differentiating adult cells. The bristle, hair, and rhabdomere phenotypes observed in Pten mutant tissue have not been reported in flies defective in insulin or Dp110 signaling, indicating that unlike the Pten-linked growth defects, these effects are probably not caused simply by an increase in PIP3 levels. Pten mutant clones have been analyzed in the larval imaginal discs to determine whether there is an underlying defect in microfilament organization during differentiation that might explain these phenotypes. Consistent with the adult eye phenotype, Pten mutant ommatidial preclusters in the eye imaginal disc were more widely spaced, and specific staining of photoreceptors indicates that these cells are enlarged in mutant clones. Phalloidin staining demonstrates that there are defects in the actin cytoskeletal network in mutant tissue. In particular, analysis of mutant photoreceptors reveals disorganization in the most basal part of the apical cytoskeleton, the scaffold on which apical photoreceptor projections are normally assembled in the disc. An altered distribution of neighboring actin microfilaments is also observed. Because rhabdomere assembly in photoreceptors requires the apical cytoskeleton, this defect may account for the rhabdomeric phenotype seen in adult Pten mutant ommatidia. Analysis of the wing imaginal disc also shows disturbed cytoskeletal organization in mutant tissue. Therefore, although Pten mutant cells can still assemble actin microfilaments, the subcellular regulation of this process appears to be abnormal, potentially accounting for the structural defects in adult bristles, rhabdomeres, and wing hairs. One interesting hypothesis that remains to be tested is that these effects on the cytoskeleton may have an important role in regulating cell growth by controlling spreading and shape change at the cell surface. It is intriguing to speculate that the loss of Pten function might alter communication between the peripheral actin cytoskeleton, the plasma membrane and the extracellular matrix, and therefore affect the invasiveness of cells (Goberdhan, 1999), as has been suggested from studies in mammalian cell culture (Tamura, 1998).

GENE STRUCTURE

There are three alternatively spliced PTEN mRNAs, producing proteins differing by a few amino acids at their carboxyl termini (Smith, 1999 and Goberdhan, 1999)

Bases in 5' UTR - 527

Exons - 12

Bases in 3' UTR - 921

PROTEIN STRUCTURE

Amino Acids - 509

Structural Domains

The predicted Pten protein was aligned and compared with other known PTEN homologs. The amino-terminal half of the Drosophila Pten protein has been particularly highly conserved over evolution, sharing ~65% identity with human PTEN in this region and ~40% and 30% identity, respectively, with the more divergent Caenorhabditis elegans (Ogg. 1998) and yeast (L. Li, 1997) PTEN homologs. Like its counterparts in other organisms, Drosophila Pten contains a putative protein phosphatase motif within a larger region related to an actin binding domain in the cytoskeletal protein tensin. A number of amino acids shared by all members of the PTEN family are not present in other protein phosphatases or tensin family members. These residues may have an important role in the specific functions of PTEN, such as its phosphoinositide 3-phosphatase activity. Sequence comparison of the carboxy-terminal half of Drosophila Pten with human PTEN also highlights a number of conserved regions, which may be involved in the regulation of this molecule (Goberdhan, 1999).

date revised: 30 January 2000

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.