Calmodulin

Recognition sites for calmodulin (CaM) are present in neuronal NOS (nNOS) (Zhang & Vogel, 1994), endothelial NOS (eNOS) (Lamas et al., 1992), and the inducible NOS in actived macrophages (macNOS) (Lyons et al., 1992; Xie et al., 1992). However, there are several differences in the consensus binding sites and proposed functions of CaM in these isoforms of NOS (Abu-Soud & Stuehr,

1993).

CaM is a single polypeptide with a molecular weight of 16,700 and is thought to be present in all eukaryotic cells so far studied. In mammalian tissues the concentration of CaM is thought to vary between 2 and 30pM (Carafoli, 1987). CaM has been found in a wide range of phylogenetically diverse species, yet the primary structure of CaM is so highly conserved that antibodies from bovine brain can cross-react with CaM in cotton seed (Wallace & Cheung, 1979). This suggests that CaM is one of the oldest members of the calcium-binding protein family. As a Ca2+ receptor, CaM is an integral part of many intra- and intercellular mechanisms that exist within any given tissue.

In NOS, CaM serves to regulate NO synthesis and does so through a variety of means (Schmidt et al., 1991). The eNOS isoforms in endothelial cells (Busse & Miilsch, 1990) and neuronal cells (Bredt & Snyder, 1990), have an absolute requirement for Ca2+ and CaM. In contrast, inducible macNOS is apparently not regulated by intracellular Ca2+ although it does contain CaM (Cho et al., 1992), and is regulated instead at the level of gene transcription (Xie et al., 1992). The activity of purified macNOS is therefore neither enhanced by exogenous Ca2+ and CaM, nor inhibited by chelators of divalent cations or drugs that block the binding of CaM to its targets (Stuehr et al., 1991; Yui etal., 1991a). Constitutive NOS is inhibited by such inhibitors (Bredt & Snyder, 1990; Busse & Miilsch, 1990). Not all inducible forms of NOS conform to this rule; Ca2+-dependent

(Palmer et al., 1992) and CaM-dependent (Iidas, 1992) isoforms of inducible NOS also exist.

Polyacrylamide gel electrophoresis has revealed that all cNOS peptides bind to CaM with 1:1 stoichiometry (Zhang & Vogel, 1994). On binding the CaM takes on an amphipathic helical conformation, similar to the CaM involved in binding to the domains of myosin light chain kinases (Olwin et al., 1984), ensuring an adequate interaction with its receptor proteins. The ability of CaM to assume a tertiary structure and display high flexibility may relate to the fact that NOS associated CaM does not contain cysteine or hydroxyproline (Cheung, 1980). Cysteine would restrict these conformational changes. Whether this binding pattern also exists for inducible NOS has not been addressed.

Urea gel electrophoresis and fluorescence spectroscopy have been used to identify the putative high affinity CaM binding sites associated with the NOS enzymes. The CaM binding domain is conserved in cNOS enzymes as a stretch of basic and hydrophobic amino acids and for nNOS this sequence exists between the 725-754 peptide residues (Vorherr et al., 1993). The eNOS consensus binding site is situated between the peptide residues 493-513 (Lamas et al., 1992). The eNOS consensus binding site is located at a similar amino acid sequence number to macNOS; the macNOS consensus binding site is at residues 503/4-532 (Lowenstein et al., 1992) however 16 out of the 29 residues in the eNOS CaM binding site are different to macNOS. This suggests that the CaM binding sequences are genetically distinct from one another and will most probably display different binding affinities for CaM. The cNOS binding site is thought to share the same or similar region on CaM as a CaM- dependent phosphodiesterase (Zhang & Vogel, 1994). This is evident by the competitive inhibition that exists between these two enzymes.

nNOS does not exist as an enzyme bound to CaM. It is therefore inactive, but can be constitutively activated following binding of CaM with transient

increases in intracellular Ca2+ levels. Spectral anaysis indicate that this binding is not a prerequisite for L-arginine binding to nNOS (Abu-Soud & Stuehr, 1993). In contrast CaM is very tightly bound to macNOS (Stuehr et al., 1991a). macNOS is one of the only four enzymes known to date to bind CaM in an apparently Ca2+-independent manner (Vorherr et al., 1993) and does not require the elevation of [Ca2+] above levels found in resting cells (400-1100 nM) (Olwin & Storm, 1985). Indeed, CaM is so tightly bound to macNOS that they copurify (Cho et al., 1992), which would explain why inhibitors of CaM are unable to detach or interrupt CaM’s interplay with macNOS. Three other enzymes which also display this characteristic are the K-subunit of phosphorylase kinase (Picton

et al., 1980), a cyclic nucleotide phosphodiesterase (Laemmli, 1970) and the adenylyl cyclase of Bordetella pertussis (Ladant, 1988).

Cho et al., (1992) were the first to suggest a role for CaM in regulating electron transfer and this was confirmed the following year by Abu-Soud & Stuehr, (1993). The idea is not altogether a novel one, nor is it unique to NOS. Evidence implies that the multimeric heme protein, sulfite reductase, may also use CaM to regulate electron transfer. As CaM is constitutively bound to macNOS it is continually primed to transfer NADPH-derived electrons onto its heme, requiring only the addition of NADPH to enable spontaneous reduction of the macNOS ferric-state heme (Abu-Soud & Stuehr, 1993). The addition of L- arginine to macNOS significantly increases the NADPH oxidation, however, in nNOS L-arginine binding slightly decreases NADPH oxidation. This suggests that electron transfer from NADPH in macNOS is dependent on substrate, in contrast to that of nNOS. This may enable nNOS to limit the amount of superoxide production in the absence of substrate. The substrate-dependent heme reduction in macNOS is also seen in cytochrome P450 enzymes (Gonzalez,

When considering the respective interactions of CaM with the different isoforms of NOS it is clear that the constitutive nature of CaM in macNOS is suited to release high levels of NO over prolonged periods of time, while the Ca2+/CaM regulated eNOS is better suited to release short bursts of NO in response to humoral signals mediated by Ca2+ transients.

CaM will undoubtedly influence pathways related to NO synthesis. For instance, it can stimulate phospholipase A2 activity in human platelets and may also regulate both the synthesis and degradation of 3’,5’-cyclic adenosine monophosphate (cAMP) in the brain (Cheung, 1980). Agonist-induced increases in cAMP can inhibit LPS-induced NO synthesis (Hon et al., 1995). CaM may also have an important role in cellular proliferation, as preventing it from binding and activating its target enzymes can attenuate the growth of malignant cells (Hait et al., 1995).