Cholesterol regulates membrane binding and aggregation by annexin 2 at submicromolar Ca(2+) concentration.

Annexin 2 is a member of the annexin family which has been implicated in calcium-regulated exocytosis. This contention is largely based on Ca(2+)-dependent binding of the protein to anionic phospholipids. However, annexin 2 was shown to be associated with chromaffin granules in the presence of EGTA. A fraction of this bound annexin 2 was released by methyl-beta-cyclodextrin, a reagent which depletes cholesterol from membranes. Restoration of the cholesterol content of chromaffin granule membranes with cholesterol/methyl-beta-cyclodextrin complexes restored the Ca(2+)-independent binding of annexin 2. The binding of both, monomeric and tetrameric forms of annexin 2 was also tested on liposomes of different composition. In the absence of Ca(2+), annexin 2, especially in its tetrameric form, bound to liposomes containing phosphatidylserine, and the addition of cholesterol to these liposomes increased the binding. Consistent with this observation, liposomes containing phosphatidylserine and cholesterol were aggregated by the tetrameric form of annexin 2 at submicromolar Ca(2+) concentrations. These results indicate that the lipid composition of membranes, and especially their cholesterol content, is important in the control of the subcellular localization of annexin 2 in resting cells, at low Ca(2+) concentration. Annexin 2 might be associated with membrane domains enriched in phosphatidylserine and cholesterol.

Cholesterol enhances phospholipid binding and aggregation of annexins by their core domain.

Annexins are Ca(2+)-dependent phospholipid-binding proteins composed of two domains: A conserved core that is responsible for Ca(2+)- and phospholipid-binding, and a variable N-terminal tail. A Ca(2+)-independent annexin 2-membrane association has been shown to be modulated by the presence of cholesterol in the membranes. Herein, the roles of the core and the N-terminal tail on the cholesterol-enhancement of annexin 2 membrane binding and aggregation were studied. The results show that (i) the cholesterol-mediated increase in membrane binding and in the Ca(2+) sensitivity for membrane aggregation were not modified by a N-terminal peptide (residues 15-26), and were conserved in mutants of the N-terminal end (S11 and S25 substitutions); (ii) cholesterol induced an increase in the Ca(2+)-dependent membrane binding and aggregation of the N-terminally truncated protein (Delta 1-29); and (iii) annexins 5 and 6, two proteins with unrelated N-terminal tails and homologous core domains showed a cholesterol-mediated enhancement of the Ca(2+)-dependent binding to membranes. These data indicate that the core domain is responsible for the cholesterol-mediated effects. A model for the cholesterol effect in membrane organisation, annexin binding and aggregation is discussed.

Modulation by Ca(2+) and by membrane binding of the dynamics of domain III of annexin 2 (p36) and the annexin 2-p11 complex (p90): implications for their biochemical properties.

The modulation of the local structure and dynamics of domain III of annexin 2 (Anx2), in both the monomeric (p36) and heterotetrameric forms (p90), by calcium and by membrane binding was studied by time-resolved fluorescence intensity and anisotropy measurements of the single tryptophan residue (W212). The results yield the same dominant excited-state lifetime (1.4 ns) in both p36 and p90, suggesting that the conformation and environment of W212 are very similar. The fluorescence anisotropy decay data were analyzed by associative (two-dimensional) as well as nonassociative (one-dimensional) models. Although no statistical criterion is decisive for one model versus the other, only the associative model allows recovery of a physically relevant value of the Brownian rotational correlation of the protein. Using the associative model, a nanosecond flexibility is detectable in p90 but not in p36. When Ca(2+) binds in the millimolar concentration range to both forms of Anx2, a conformational change takes place leading to an increase of the major excited-state lifetime (2.6 ns) and to a suppression of the W212 local flexibility of p90. Binding to membranes of either p36 or p90 in the presence of Ca(2+) does not induce any conformational change other than that provoked by Ca(2+) binding alone. The W212 local flexibility in both proteins increases significantly, however, in their membrane-bound forms. In the presence of membranes, the conformation change of domain III in p90 displays a sensitivity to Ca(2+) 2 orders of magnitude higher than that of p36, reaching intracellular sub-micromolar concentration ranges. This higher Ca(2+) sensitivity correlates with the Ca(2+)-dependent membrane aggregation but not with their Ca(2+)-dependent binding to membranes. The significance of these structural and dynamical changes for the function of the protein is discussed.

N-Terminal domain of annexin 2 regulates Ca(2+)-dependent membrane aggregation by the core domain: a site directed mutagenesis study.

Annexin 2 binds and aggregates biological membranes in a Ca(2+)-dependent manner. This protein exists as a monomer (p36) or as a heterotetramer (p90) in which two p36 chains are associated with a dimer of p11, a member of the S100 protein family. Protein kinase C phosphorylates the protein at the level of the N-terminal tail on serines 11 and 25, thereby modifying its oligomeric structure and its properties of membrane aggregation. To analyze these effects, the properties of a series of mutants in which serines 11 and 25 were replaced by alanine and/or glutamic acid were investigated. The affinity for p11 light chain was decreased in the S11E mutants. Glutamic acid residues in positions 11 or 25 did not change membrane binding, either in the tetrameric or in the monomeric form. On the other hand, these mutations affected the aggregation properties of the two forms. For the tetramer, the aggregation efficiency was decreased but not the Ca(2+) sensitivity, whereas the latter was affected in the case of the monomer. The effects were stronger in the S11E mutants, and they were cumulative in the double mutant. They suggest a different conformation of the N-terminal domain in the mutants (and in the phosphorylated protein), a hypothesis which is supported by proteolysis experiments. This conformational change would affect aggregation by the monomer through a dimerization step.

Stability of ColE1-like and pBR322-like plasmids in Escherichia coli.

The average copy number, the level of ampicillin resistance conferred by one plasmid, and the degree of plasmid multimerization were determined for several ColE1-like and pBR322-like plasmids. From the results obtained, the variance of the units of partition corresponding to each plasmid studied was calculated. Experimentally determined plasmid stability was compared with that calculated using the variance of the units of partition and the ratio between the generation times of plasmid-free and of plasmid-carrying cells, assuming that the units of partition are distributed randomly between daughter cells. Stability of the pBR322-like plasmids present mainly as monomers in the bacterial host was consistent with random partitioning, whereas pBR322-like plasmids, present mainly as dimers, and the ColE1-like plasmid showed greater stability than that predicted with random partitioning at cell division.