BIOLOGICAL OVERVIEW

Nicastrin is involved Notch signaling and along with Presenilin forms part of a large multiprotein complex. Drosophila nicastrin (nic) mutants display characteristic Notch-like phenotypes. Genetic and inhibitor studies indicate a function for Nicastrin in the gamma-secretase step of Notch processing, similar to Presenilin. Further, Nicastrin is genetically required for signaling from membrane-anchored activated Notch. In the absence of Nicastrin, Presenilin is destabilized and mature C-terminal subunits are absent. Partially processed Notch accumulates apically in nicastrin and presenilin mutant follicle cells. nicastrin and presenilin mutations disrupt the spectrin cytoskeleton, suggesting that the gamma-secretase complex has another function in Drosophila in addition to its role in processing Notch and the second target, ß amyloid protein precursor (APP). Nicastrin might recruit gamma-secretase substrates into the proteolytic complex as a prerequisite for Presenilin maturation and active complex assembly (Hu, 2002; López-Schier, 2002).

Recent studies have established that the crucial signal-generating cleavage of Notch depends upon a large protein complex containing Presenilin, with Presenilin itself likely performing the catalytic role in Notch proteolysis. The Presenilins are a family of polytopic membrane proteins having eight putative transmembrane domains and a large cytoplasmic loop, and they are localized predominantly in the ER/Golgi compartment. Mutations in the two known human Presenilin genes PS1 and PS2 account for most cases of early onset, autosomal dominant Alzheimer's disease. The PS mutations promote the generation of neurotoxic Aß peptide derivatives of amyloid precursor protein (APP) by altering the manner in which APP is cleaved at a site, termed the gamma-secretase site, within its transmembrane domain. These observations led to the hypothesis that Presenilin might itself be the long sought 'gamma-secretase' enzyme responsible for this cleavage. Several findings have now provided strong support for this idea. Elimination of PS1 activity in knockout mice leads to reduced gamma cleavage of APP, and mutagenesis of either of two conserved aspartate residues in PS1 interferes with APP cleavage. Moreover, PS1 and gamma-secretase activity can be biochemically copurified from solubilized HeLa cell membranes, and transition-state analog inhibitors of gamma-secretase label PS1 and PS2. Finally, database comparisons have led to the identification of a conserved motif in PS proteins exhibiting similarity to a short amino acid segment found in a family of bacterial aspartyl proteases (Hu, 2002 and references therein).

The assertion that Presenilin itself is a proteolytic enzyme has not been completely accepted, due to the difficulty in reconstituting in vitro the gamma-secretase proteolytic activity. Presenilin apparently functions as part of a large multiprotein complex (>2000 kD), which contains numerous other unidentified factors that might also be required for proteolytic activity of the intact complex. Another component of the complex, termed Nicastrin, has recently been identified, although the exact function of Nicastrin in the gamma-secretase cleavage of substrate proteins is unknown. Mammalian Nicastrin binds to C-terminal derivatives of APP that are substrates for gamma-secretase, and overexpression of engineered Nicastrin mutants influences cleavage of APP and subsequent Aß secretion (Yu, 2000). Nicastrin has also been implicated in the regulation of Notch proteolysis. Mutations in the C. elegans Nicastrin homolog aph-2 were isolated as mutants displaying Notch/Lin-12 pathway defects (Goutte, 2000), and overexpression of the engineered mammalian Nicastrin mutants has weak modulatory effects on Notch cleavage (Chen, 2001). Nicastrin also binds to membrane-tethered forms of Notch that are structurally analogous to the C-terminal derivatives of APP, suggesting that Nicastrin might generally bind to gamma-secretase proteolytic substrates (Chen, 2001). However, it has not been shown that Nicastrin is directly required for gamma-secretase-like proteolysis of the Notch receptor, or how the loss of Nicastrin might affect formation and/or function of the gamma-secretase complex (Hu, 2002).

Embryos that lack both maternal and zygotic nct display a strong neurogenic phenotype associated with a loss of Notch signaling. The cuticle fails to form except for a small patch on the dorsal side of the embryo, and there is a hyperplasia of the embryonic nervous system, as revealed by a large increase in the number of cells expressing the neuronal marker MAb22C10. nct^agro homozygous clones also produce the same phenotypes as Notch clones in other tissues. In the wing imaginal disc, Notch signaling establishes the dorsal-ventral compartment boundary, which is marked by the expression of Cut protein. In mosaic discs containing marked nct mutant clones, Cut is not expressed in mutant cells along the boundary. Mutant clones also produce other typical Notch phenotypes, such as large notches in the wing margin, thickening of the wing veins, and the loss or duplication of sensory bristles (López-Schier, 2002).

During stages 5-6 of oogenesis, Delta signals from the germline to activate Notch in the somatic follicle cells, inducing these cells to stop mitosis and differentiate, and this provides an excellent system for determining whether components of the Notch pathway are required in the signaling or responding cells. Whereas nct germline clones have no effect on follicle cell development, mutant follicle cell clones show an identical phenotype to clones of null mutations in Notch or psn. The cells continue dividing after stage 6 of oogenesis, as shown by the expression of the mitotic marker phospho-histone H3, and this causes an increase in cell number and decrease in cell size in mutant clones. Mutant cells also fail to differentiate, since they continue to express high levels of the immature follicle cell marker Fasiclin III (Fas III), and never express any markers of differentiated follicle cell types. For example, mutant clones at the posterior of the egg chamber cannot adopt a posterior fate in response to the Gurken signal from the oocyte, and these cells therefore fail to signal back to the germline to polarize the anterior-posterior axis of the oocyte. As a consequence, the oocyte does not repolarize at stage 7, and the germinal vesicle fails to migrate from the posterior to the anterior pole. This requirement for Nct is strictly cell autonomous, since mutant follicle cells always express Fas III, even in very small clones. Thus, nct is a novel neurogenic gene that is required cell autonomously for the transduction of the Notch signal in responding cells (López-Schier, 2002).

By isolating and characterizing nicastrin loss-of-function mutants in Drosophila, it has been demonstrated that Nicastrin is required both genetically and biochemically for proteolysis of membrane-tethered forms of Notch and associated Notch signaling activity. The Drosophila nicastrin mutants exhibit defects in Notch signaling and intramembranous cleavage that are indistinguishable from those seen in presenilin mutants. Notch-like phenotypes are observed in various tissues at all stages of development, and may be attributed to a specific failure in the gamma-secretase-like cleavage of Notch that normally generates an ~120 kDa intracellular signaling fragment from the activated receptor. Identical biochemical effects are observed for Drosophila Notch with the gamma-secretase inhibitor DFK-167, supporting the role of Nicastrin in this proteolytic event and suggesting that Drosophila Notch processing might be a useful model for investigating pharmacological inhibitors of gamma-secretase. Analysis of Notch immunoreactivity in genetic mosaics reveals that homozygous loss of nicastrin activity results in a subtle overaccumulation of Notch proteins at the apical intracellular surface. Similar observations have been reported for Drosophila Psn mutants, and are consistent with the idea that nic and Psn are necessary for cleavage and release of membrane-anchored Notch C-terminal fragments from the cell surface (Hu, 2002).

In mammalian cells, expression of engineered mutants of Nicastrin with amino acid substitutions or internal deletions in a conserved domain have been reported to have dramatic modulatory effects on APP processing and modest effects on Notch proteolysis. In an RNAi-based Drosophila S2 cell assay, expression of analogous engineered forms of Drosophila Nicastrin does not lead to any observable proteolytic activity toward Notch or any dominant-negative effects. These results indicate that some aspects of Nicastrin function might not be conserved among different species, or that differences in the assay systems might underlie the contradictory results (Hu, 2002).

The function of Nicastrin in the large gamma-secretase proteolytic complex is not well understood. Presenilins from different species show high levels of sequence conservation in their transmembrane domains, including conserved aspartyl residues thought to be important for catalysis, consistent with the proposed catalytic role in the intramembranous cleavage of membrane-bound substrates. In contrast, Nicastrins from various organisms display little conservation in their single transmembrane segment and short intracellular domain, suggesting that they are unlikely to participate directly in the proteolytic reaction within the lipid bilayer. The most highly conserved segments of Nicastrin are located in the large extracellular domain, while only a very small portion of Presenilin consisting of the short loops connecting the transmembrane domains is topologically exposed at the extracellular surface. Since Nicastrin binds to C-terminal membrane-anchored substrates of gamma-secretase (Yu, 2000; Chen, 2001), one speculative idea is that the extracellular domain of Nicastrin might act as a sensor that measures the size of the extracellular domain of bound substrates, a critical determinant of cleavage efficiency. Further functional characterization of the gamma-secretase complex will be necessary to evaluate this possibility (Hu, 2002).

Data showing that mature Presenilin protein is unstable in Drosophila cells lacking Nicastrin might simply be due to a general degradation of Presenilin in the absence of an essential interacting protein. However, other studies have suggested that steady-state levels of mature Presenilin are regulated by a specific 'stabilization/cleavage' mechanism in which most nascent full-length Presenilin is degraded before incorporation into the complex, while a small pool becomes stabilized by unknown factors and undergoes subsequent endoproteolytic maturation. Nicastrin might be a limiting cellular factor which influences the amount of Presenilin that is either stabilized or degraded. Some data indicate that binding of Nicastrin to Presenilin occurs prior to assembly of the large multiprotein gamma-secretase complex. For example, the aspartyl mutant PS1-D385A is unable to be incorporated into the complex, but is still able to bind human Nicastrin (Yu, 2000). Since Nicastrin is known to bind to Notch and APP C-terminal derivatives, one possibility is that the inactive PS holoprotein might only be stabilized and converted into an active protease following association with Nicastrin bound to an appropriate substrate (Hu, 2002).

GENE STRUCTURE

cDNA clone length - 2903

Bases in 5' UTR - 49

Exons - 7

Bases in 3' UTR - 772

PROTEIN STRUCTURE

Amino Acids - 695

Structural Domains

The nicastrin locus is composed of seven exons, spanning 3.4 kb of genomic DNA. The nct^agro allele contains a two-base pair insertion in the fourth exon of the gene, creating a frameshift that truncates the protein at amino acid 373 before the putative transmembrane domain. The Nicastrin protein has a conserved signal peptide and a C-terminal transmembrane domain. It shows a low overall sequence similarity to its mammalian and nematode counterparts; however, some small regions of the protein show higher sequence conservation, including the DYIGS domain, which is deleted in the dominant-negative constructs that inhibit APP processing in human cell lines (López-Schier, 2002).

date revised: 15 February 2002

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.