Cell–cell interactions via cell adhesion are the basis for the evolution of all multicellular organisms. The first experiments to understand cell adhesion were performed at the beginning of the last century by Wilson (1907), who dissociated sponges and allowed them to fuse and to reconstitute. Fifty years later, Townes and Holtfreter (1955) demonstrated that dissociated cells from amphibian embryos adhere to form random aggregates of their origin, with ectoderm forming an outer surface layer, endoderm forming a compact central ball, and mesoderm producing a loose array of cells. In 1963, Roger Sperry (1963) proposed that different cells bear distinct cell surface proteins that serve as markers or tags. One of these markers, the Neural Cell Adhesion Molecule (NCAM) was first described in 1974 by Elizabeth Bock (Jacque et al. 1974) and designated as D2 antigen. Three years later, the group of Edelman (Thiery et al. 1977) approved that NCAM mediates cell–cell adhesion and established it as one of the first cell adhesion molecules. When NCAM was sequenced and cloned during the mid-1980s, it became obvious that NCAM belongs to the immunoglobulin superfamily (Barthels et al. 1987) and that NCAM is not only responsible for static adhesion, but also transmits signals across the cell membrane. With the help of transgenic knockout techniques, Cremer and colleagues could demonstrate 1994 (Cremer et al. 1994) that NCAM is involved in learning and memory.
Structure and Function
NCAM mediates preferentially homophilic NCAM–NCAM interactions, which are Ca2+-independent. The homophilic NCAM–NCAM interactions mediate adhesion between neurons and neurons, neurons and glia cells, and between glia cells and glia cells. NCAM140 is known to be responsible for the fasciculation of axons or neurites and for promotion of axonal regeneration after injury. NCAM-deficient mice have defects in learning and memory and have a smaller brain compared to wild-type animals. Furthermore, the nervous system–specific NCAM180 is thought to stabilize cell–cell contacts at synapses. A recent model of NCAM–NCAM interaction suggests that two NCAMs form dimers on the cell curface (cis-interaction) and that these dimers bind to existing dimers on opposing cells (trans-interaction) via their Ig-domains 2 and 3 in antiparallel orientation (“zipper”-like) (Soroka et al. 2003).
Please note that there are several and sometimes confusing abbreviations for polysialic acid or polysialylated NCAM used in the literature (PSA = polysialic acid; E-NCAM = embryonic NCAM; polySia = polysialic acid).
The last two decades of research on NCAM-mediated signal transduction focuses mainly on NCAM-mediated promotion of neurite outgrowth, which will be discussed in the following. One central pathway of the NCAM-mediated signal transduction is the mitogen-activated protein kinase pathway (MAPK pathway). The MAPK pathway can be activated by two mechanisms: First, via the Fibroblast Growth Factor Receptor (FGF-receptor) (Doherty and Walsh 1996) and second via non-receptor kinases such as FAK and Fyn (Beggs et al. 1997).
The Mitogen-Activated Protein Kinase Pathway
Phospholipase C and Calcium
NCAM-mediated signaling has been demonstrated to involve an increase in intracellular Ca2+-concentration (Kolkova et al. 2000). Several studies suggested that phospholipase Cγ (PLCγ) is responsible for the NCAM-mediated increase in Ca2+-concentration. Upon NCAM–NCAM interaction, PLCγ cleaves PIP2 into inositol trisphosphate (IP3) and diacylglycerol (DAG). IP3 Receptors binds to its intracellular receptor at the ER-membrane and releases Ca2+ from the ER store. Furthermore arachidonic acid (AA) might be released from DAG by the DAG lipase, which further activates Ca2+-channels at the plasma membrane. All this leads to an increase of intracellular Ca2+. The increased Ca2+ binds to calmodulin and this leads to the activation of the Ca2+-calmodulin-dependent protein kinase II. Ca2+-calmodulin-dependent protein kinase II phosphorylates several target proteins, which are involved in neurite outgrowth (see Fig. 3).
Cyclic Adenosine Monophosphate and Cyclic Guanosine Monophosphate
The involvement of the two second messengers cyclic adenosine monophosphate (cAMP) and cyclic guanosine monophosphate (cGMP) in NCAM signaling has been demonstrated in several studies by the use of specific inhibitors (Shimomura et al. 1998). However, the exact role of the respective heterotrimeric G-proteins (in the case of cAMP) or nitric oxid synthases (in the case of cGMP) is not clear yet.
NCAM may transmit signals into the cell by several mechanisms. This is, like in many other cases, a complex network of signal transduction pathways. The MAPK pathway seems to be very crucial for NCAM-mediated signal transduction. However, little is known about relative roles of the individual players and further research is necessary to elucidate the total network of NCAM signaling.