Calcium storage and release properties of F-actin: evidence for the involvement of F-actin in cellular calcium signaling.

Preceding studies have shown that the bulk of the ATP-dependent, inositol 1,4,5-trisphosphate (IP3)-sensitive Ca2+ store of hamster insulinoma (HIT) cells is located in microvilli on the cell surface. Similar results were obtained with isolated rat hepatocytes. Moreover, in vesicles of microvillar origin, passive fluxes of Ca2+, ATP, and IP3 occur through cation and anion channels, respectively, suggesting that Ca2+ storage is due to ATP-dependent Ca2+ binding to an intravesicular component. Here we demonstrate that F-actin may be a possible candidate for this function. ATP-actin monomers bind Ca2+ with high affinity (Kd = 2-8 nM) to their divalent cation binding sites. Polymerization of actin monomers decreases the rate constant for divalent cation exchange at this binding site by more than 3 orders of magnitude rendering bound cations nearly unavailable. F-actin-bound Ca2+ can be released by depolymerization and dissociation from Ca(2+)-ADP-actin monomers (Kd = 375 nM). We now provide additional evidence for the possible involvement of actin in Ca2+ storage. (1) Preincubation of surface-derived Ca(2+)-storing vesicles from HIT cells with the F-actin stabilizer, phalloidin, strongly inhibited ATP-dependent Ca2+ uptake, reducing the IP3-sensitive Ca2+ pool by 70%. Phalloidin, when added after the loading process, affected neither the amount of stored Ca2+ nor IP3 action on the store. (2) F-actin polymerized in the presence of Mg2+ in nominally Ca(2+)-free buffer still contained about half of the high affinity sites occupied with Ca2+ (Mg/Ca-F-actin). (3) Using the fura-2 technique, we found that in the presence of ATP, Mg/Ca-F-actin incorporated free Ca2+ at a relatively low rate. Short pulses of ultrasound (3-10 s) strongly accelerated Ca2+ uptake, decreasing free Ca2+ from 500 nM to below 100 nM. (4) In the presence of physiological levels of Mg2+ (0.5 mM), sonication liberated large amounts of Ca2+ from Mg/Ca-F-actin. (5) Ca-F-actin released bound Ca2+ at a very slow rate. Short ultrasonic pulses rapidly elevated free Ca2+ from about 50 nM up to 500 nM. (6) Small amounts of profilin, an actin-binding protein, released Ca2+ both from Ca- and Mg/Ca-F-actin and also inhibited uptake of Ca2+ into Mg/Ca-F-actin. (7) Phalloidin completely inhibited Ca-uptake into Mg/Ca-F-actin even during ultrasonic treatment. These findings suggest that Ca2+ storage may occur by addition of Ca-ATP-actin monomers to reactive ends of the polymer and emptying of this store by profilin-stimulated release of Ca-ADP-actin. Thus, receptor-operated Ca2+ signaling, initiated by phospholipase C activation, may proceed via the well-known phosphatidylinositol phosphate-regulated profilin/gelsolin pathway of actin reorganization/depolymerization. The importance of the proposed microvillar Ca2+ signaling system for living cells remains to be established.