BIOLOGICAL OVERVIEW

Amalgam (Ama), a secreted member of the immunoglobulin superfamily, has been identified as a ligand for the Neurotactin (Nrt) receptor. Nrt is a member of the serine esterase-like family of membrane proteins. Ama is necessary for Nrt-expressing cells both to aggregate with one another and to associate with embryonic primary culture cells. Aggregation assays performed with truncated Nrt molecules reveal that the integrity of the cholinesterase-like extracellular domain is not required either for Ama binding or for adhesion, with only amino acids 347-482 of the extracellular domain being necessary for both activities. Moreover, the Nrt cytoplasmic domain is required for Nrt-mediated adhesion, although not for Ama binding. Using an ama-deficient stock, it has been found that ama function is not essential for viability. Pupae deficient for ama do exhibit defasciculation defects of the ocellar nerves similar to those found in nrt mutants. Although the specific roles of Nrt and Ama during Drosophila development remain quite obscure, different aspects of the puzzle are beginning to fall into place. A link has been established between these two enigmatic proteins. One of the critical missing pieces is what lies downstream of the Nrt receptor and how Ama binding to Nrt affects these downstream components. Addressing this issue and identifying orthologs in other organisms are probable key steps in a further understanding of this intriguing receptor-ligand interaction (Fremion, 2000).

Nrt functions as a heterophilic cell adhesion protein in various cell culture assays (Barthalay, 1990), suggesting that its cholinesterase domain functions as a part of a recognition/signaling transmembrane receptor-like protein. Further studies have delineated a specific region of the cholinesterase-like domain located near the membrane that is involved in this recognition process (Darboux, 1996). Ama is able to promote the aggregation of Nrt-expressing S2 cells; Ama interacts specifically with Nrt, and Ama is necessary for the interaction of Nrt-expressing S2 cells with dissociated embryonic primary culture cells, the original assay that established the heterophilic adhesion properties of Nrt. The nearly identical patterns of Ama and Nrt protein accumulation throughout embryogenesis are consistent with a receptor-ligand relationship in vivo. Furthermore, the finding that Ama protein fails to exhibit its normal pattern of accumulation in an nrt mutant is compatible with Nrt acting as the key receptor for Ama during embryogenesis. These are intriguing observations: that during embryogenesis, cells that accumulate Ama on their surface are sometimes different from the cells that are expressing ama transcripts. Certainly these differences can be explained easily for a secreted protein such as Ama. However, the functional significance, if any, of these disparities between spatial patterns of protein expression versus protein accumulation is unknown (Fremion, 2000).

How does Ama facilitate Nrt-mediated adhesion? Observations that an artificially generated membrane-anchored form of Ama can mediate homophilic aggregation in the absence of Nrt indicate that Ama can interact with itself. This may suggest one mechanism whereby Ama could function as a linker between Nrt molecules on opposing cell surfaces. Nrt would bind Ama, which would bind to a second Ama molecule that could interact with Nrt on the surface of a second cell. Thus, Ama would serve as a linker protein between Nrt-expressing cells. Alternatively, Ama could alter the Nrt protein upon binding, such that it could now recognize in a homophilic manner other Nrt molecules (Fremion, 2000).

The binding of Ama to Nrt alone is not sufficient for cell aggregation. The finding that the Nrt cytoplasmic domain is essential for adhesion indicates that important events are taking place in the cytoplasm that are critical for this adhesion process. The amino acid sequence of the Nrt cytoplasmic domain does not reveal any clear motifs such as kinase or phosphatase domains. However, it does encode several putative phosphorylation sites, and Nrt is found to be phosphorylated on serine and threonine residues. Interestingly, Nrt is enriched at sites of cellular contacts in neuronal and epithelial tissues and within Nrt transfectant aggregates (Barthalay, 1990). This recruitment to specific regions of the membrane suggests that Nrt may bind to cytoskeleton components. Nrt-mediated cell adhesion might be considered as a two-step process: (1) binding of the ligand (Ama) to the extracellular domain, a step that can occur without the Nrt cytoplasmic domain; and (2) stabilization and strengthening of the interaction through clustering of Nrt to sites of cell contact and interactions dependent upon the Nrt cytoplasmic domain. This latter step was not achieved with transfectants lacking the Nrt cytoplasmic domain, demonstrating the critical role of the Nrt cytoplasmic domain in this process (Fremion, 2000).

The isolation and characterization of nrt mutants have demonstrated that Nrt is required for outgrowth, fasciculation and guidance of axons (Speicher, 1998). Mutations in nrt produce quite subtle and variable defects in axon guidance within the embryonic CNS. These phenotypes are greatly enhanced when nrt mutations are combined with mutations in other cell adhesion or signaling molecules such as nrg. These results have led to the suggestion that nrt plays a cooperative or partially redundant role in axon guidance. One of the most robust phenotypes of nrt mutants alone is disruption of ocellar nerve axon pathfinding in pupae. In nrt mutants, misrouting, stalling and defasciculation of the ocellar pioneer nerves are observed at high frequencies (Fremion, 2000).

One might expect that if Ama and Nrt indeed interact in vivo as indicated by the co-immunoprecipitation experiment, mutations in these genes would lead to similar phenotypes. Since the ocellar pioneer nerve disruptions are the most consistent phenotype for nrt null mutants, a determination was made whether a loss-of-function ama mutant displays comparable phenotypes. Similar to nrt mutations, a deficiency for ama results in highly penetrant defasciculation phenotypes of the ocellar pioneer nerves in developing pupae. However, there are significant differences in the expressivity of ocellar nerve defects. Stalling or misrouting of these axons as is found in nrt mutants is not observed in ama mutant pupae. This result suggests that rather than having an instructive role in pathfinding, Ama maintains proper cohesion between these axons and contributes to the organization of these fascicles (Fremion, 2000).

The role of Ama during embryonic axon guidance cannot be addressed due to the lack of specific ama mutations. The deficiency that was used in these studies [Df(3R)ama] is also deleted for the embryonic dorsal-ventral patterning gene zerknullt. The lethality associated with zerknullt mutations could not be rescued with a wild-type zerknullt transgene in order to generate an adult viable strain that is deficient for Ama. The difficulty is that partial rescue of zerknullt embryonic functions leads to variable defects in germ band extension and retraction. These disruptions in morphogenetic movements make it impossible to assay reliably for subtle defects in formation of axon pathways in the embryonic CNS. Therefore, it is possible that loss-of-function mutations in ama could result in embryonic axon pathfinding defects that are similar to nrt loss-of-function mutations (Fremion, 2000).

GENE STRUCTURE

amalgam lies between bicoid and Deformed in the Antennapedia complex (Seeger, 1988).

cDNA clone length - 1.5 kb

Bases in 5' UTR - 324

Exons - 2

Bases in 3' UTR - 281

PROTEIN STRUCTURE

Amino Acids - 333

Structural Domains

The N-terminal 23 amino acids constitute a hydrophobic region characteristic of signal sequences for secreted or membrane-bound proteins. There are 3 N-linked glycosylation sites. The Ig domains are approximately 100 amino acids in length and contain two distantly spaced cysteines per domain (Seeger, 1988).

date revised: 21 October 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.