Ca2+-dependent Actin-binding Phosphoprotein in Ph ysarum polycephalum SUBUNIT b IS A DNase I-BINDING AND F-ACTIN CAPPING PROTEIN*

Physarum contains at least two distinct DNase I-binding proteins, i.e. actin and Cap 42 (a + b). The latter, a tight (1:1) complex of Cap 42 (a) and Cap 42 (b) (Maruta, H., Isenberg. G., Schreckenbach, T., Hallmann, R., Risse, G., Schibayama, T., and Hesse, J. (1983) J. Biol. Chem. 258, 10144-10150), is a Ca2+-dependent F-actin capping protein. DNase I binds to Cap 42 (b) but not to Cap 42 (a). Consequently, DNase I-agarose was used for an affinity-purification of Cap 42 (a + b), after its separation from actin by DEAE-cellulose chromatography. Cap 42 (a + b) was dissociated into its subunits when released from DNase I-agarose by 8.8 M formamide. The two subunits were subsequently separated from each other on hydroxylapatite. Both Cap 42 (a) and Cap 42 (b) were Ca2+-dependent F-actin capping proteins that cap the fast growing end of actin filaments and block actin polymerization at this end. Like Cap 42 (a + b), Cap 42 (b) required Ca2+ for its capping activity only when phosphorylated. The phosphorylation of Cap 42 (b) was completely blocked by DNase I or a tertiary complex of Cap 42 (a), actin, and Ca2+. Cap 42 (b) is not identical with native (= polymerizable) actin because (i) Cap 42 (b) was unable to form filaments, (ii) the Cap 42 (b) kinase did not phosphorylate native actin, and (iii) fragmin formed a tight (1:1) complex with native actin but not with Cap 42 (b). Although it is unlikely that Cap 42 (b) is simply a denatured form of actin that has lost its polymerizability during the preparation, it still remains to be clarified whether Cap 42 (b) is a nonpolmerizable actin variant derived from a distinct actin gene or a post-translationally modified form of polymerizable actin.

cellulose chromatography. Cap 42 (a + b) was dissociated into its subunits when released from DNase Iagarose by 8.8 M formamide. The two subunits were subsequently separated from each other on hydroxylapatite. Both Cap 42 (a) and Cap 42 (b) were Ca2+dependent F-actin capping proteins that cap the fast growing end of actin filaments and block actin polymerization at this end. Like Cap 42 (a + b), Cap 42 (b) required Ca2+ for its capping activity only when phosphorylated. The phosphorylation of Cap 42 (b) was completely blocked by DNase I or a tertiary complex of Cap 42 (a), actin, and Ca2+. Cap 42 (b) is not identical with native (=polymerizable) actin because (i) Cap 42 (b) was unable to form filaments, (ii) the Cap 42 (b) kinase did not phosphorylate native actin, and (iii) fragmin formed a tight (1:l) complex with native actin but not with Cap 42 (b). Although it is unlikely that Cap 42 (b) is simply a denatured form of actin that has lost its polymerizability during the preparation, it still remains to be clarified whether Cap 42 (b) is a nonpolymerizable actin variant derived from a distinct actin gene or a post-translationally modified form of polymerizable actin.
Physarum contains at least four proteins of 42,000 daltons which are functionally distinguishable from each other (1-5). The major protein is actin which forms filaments by self assembly. The other three proteins, Le. fragmin, Cap 42 (a), and Cap 42 (b), do not form filaments by themselves but regulate actin polymerization in the following manner. Fragmin is a Ca2+-dependent F-actin severing protein which forms a 1:1 complex with actin monomer and causes a rapid fragmentation of actin filaments only in the presence of Ca2+ (1-3). Cap 42 (a) and Cap 42 (b) are two distinct subunits of a Ca2+-dependent F-actin capping protein called Cap 42 (a + b) which caps or binds to the fast growing end of actin filaments, * The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
blocks actin polymerization at this end, and induces rapid depolymerization of the filaments at the opposite end (4, 5 ) .
Cap 42 (a + b) requires Caz+ for its capping activity only when Cap 42 (b) is phosphorylated (5). When Cap 42 (a) and Cap 42 (b) are separated from each other in the presence of 7 M urea, the F-actin capping activity resides in Cap 42 (a) but not in Cap 42 (b) ( 5 ) . Cap 42 (a) alone requires Ca2+ for its capping activity ( 5 ) . However, unlike fragmin, Cap 42 (a) has no F-actin severing activity ( 5 ) . Cap 42 (b) is a phosphoprotein whose phosphorylation is completely inhibited by an equimolar complex of Cap 42 (a) and actin only in the presence of Ca2+ (4). Neither Cap 42 (a), actin, or fragmin are phosphorylated by Cap 42 (h) kinase (4). When Cap 42 (b) is dephosphorylated, the capping activity of Cap 42 (a + b) becomes Ca2+-independent, indicating that Cap 42 (a) requires two alternative activators, i.e. Ca2+ or the dephosphorylated Cap 42 (b), for its capping activity (5).
Since actin was identified as a DNase I inhibitor and forms a tight 1:l complex with DNase I (6, 7), DNase I-agarose has been widely used for the rapid purification of actin from a variety of organisms (8-10). Here we provide evidence for the occurrence of a second DNase I-binding protein in Physarum, i.e. Cap 42 (b), which is distinct from actin. DNase I completely inhibited the phosphorylation of Cap 42 (b) but in a Caz+-independent manner. Cap 42 (b) itself bound to DNase I-agarose even in the absence of Cap 42 (a) and actin. In addition, we show that Cap 42 (b) alone has an F-actin capping activity which becomes Ca2+-dependent when Cap 42 (b) is phosphorylated.

MATERIALS AND METHODS
Cell Culture-Microplasmodia of Physarum polycephalum Colonia CL were used as the source for Cap 42 (a + b), Cap 42 (b) kinase, and actin preparations as previously described (4).
Preparation of Hemoglobin-Agarose-200 mg of hemoglobin (Sigma, H-2500) was covalently coupled with 20 ml of Affi-Gel IO (Bio-Rad) in HC 100 buffer according to the manufacturer's manual.
sequently through a DEAE-cellulose column (10 ml) equilibrated with HC 10 buffer in order to be separated from the proteases which were bound to either hemoglobin-agarose or DEAE-cellulose.
Preparation of DNase I-Agarose-DNase I recovered in the flow through fractions from DEAE-cellulose was practically free of the protease activity, and was covalently coupled with 25 ml of Affi-Gel 10 (Bio-Rad) in HC 100 buffer according to essentially the same procedure as described for the preparation of hemoglobin-agarose.
Separation of Cap 42 (a + b) and Cap 42 (b) Kinase from Actin-400 g of microplasmodia were extracted by 800 ml of sucrose buffer as previously described (4). The sucrose extract was then fractionated on 300 ml (bed volume) of DEAE-cellulose equilibrated with TEDA buffer by a stepwise elution with 75 mM, 150 mM, and 300 mM NaCl in TEDA buffer as described by Maruta et al. (4). All the Cap 42 (a + b) and the Cap 42 (b) kinase activity were recovered in the 150 mM NaCl eluate, whereas all the actin was recovered in the 300 mM NaCl eluate.
DNase I-Agarose Affinity Chromatography of Cap 42 (a + b) and Purification of Physarum Actin-Actin in the 300 mM NaCl eluate from DEAE-cellulose was precipitated with 3.0 M ammonium sulfate in the presence of 2 mM M F -A T P and 0.2 mM CaC1,. After two cycles of polymerization-depolymerization, the actin was further purified by gel filtration on Sephadex G-150 (4). The final actin preparation was a t least 99% pure as judged by SDS-polyacrylamide gel electrophoresis and did not contain any detectable F-actin capping activity.
Assay for Cap 42 (3) Kinase Activity-10 r g of purified Cap 42 (b) were incubated with various kinase fractions a t 35°C in AM buffer, and the incorporation of 32P into a trichloroacetic acid-insoluble fraction was measured by the filter paper assay as previously described (4).
Assay for F-actin Capping Actiuity-The falling ball viscometer was used to monitor the F-actin capping activity of Cap 42 (a) and/ or Cap 42 (b) which reduces the low shear viscosity of rabbit skeletal muscle actin (0.5 mg/ml) solutions as previously described (4). Short fragments of S-1 decorated and fixed actin filaments were used as nuclei to test for actin filament growth in order to determine which end of the acin filaments is capped by Cap 42 (b) according to the procedure described previously (5).

Further Evidence That Cap 42 63) Is Not Identical with
Actin-After prolonged storage, G-actin from both Physarum and rabbit skeletal muscle usually loses the ability to form filaments. Even in the presence of Ca2+, an aged and unpolymerizable preparation of Physarum actin showed no significant F-actin capping activity, when compared with Cap 42 (b) under the same conditions (data not shown). This result supported the functional differences between Cap 42 (b) and a denatured form of Physarum actin.
Actin from vertebrate smooth muscle and a few other sources was shown to be phosphorylated by the catalytic subunit of CAMP-dependent protein kinase (13). However, the catalytic subunit of CAMP-dependent protein kinase phosphorylated only the denatured (aged or digested), unpolymerizable actin preparations from Physarum or rabbit skeletal muscle and not the native, polymerizable actin (Fig. 10). Interestingly, the same kinase was unable to phosphorylate Theoretically actin could be denatured and become unpolymerizable in a number of different ways during its storage, preparation, and even within cells. Hence, the question has been raised if Cap 42 (b) is one of these denatured forms of actin. However, there is no way of knowing precisely which type of denaturation of actin, if any, might create Cap 42 (b) until a conversion of actin to Cap 42 (b) could be demonstrated in uitro under defined conditions. Since the preparation of Cap 42 (b) was carried out mostly a t 0 "C, we simply asked whether actin could be converted to Cap 42 (b) under identical conditions, e.g. by prolonged storage of actin a t 0 "C. This treatment always creates a denatured, nonpolymerizable, form of actin which becomes phosphorylatable by CAMP-dependent protein kinase. However, unlike Cap 42 (b), the aged (or denatured), nonpolymerizable, actin is unable to cap the fast growing end of actin filaments and is not phosphorylated by Cap 42 (b) kinase. Furthermore, Cap 42 (b) is never phosphorylated by CAMP-dependent protein kinase, nor loses its capping activity even after prolonged storage. Therefore, we can exclude the possibility that Cap 42 (b) is identical with this aged (or denatured), nonpolymerizable form of actin. Although also highly unlikely, however, we can not absolutely exclude the possibility that actin has been converted into Cap 42 (b) in vivo by a post-translational modification(s). This modification, however, can not be simply an in uiuo phosphorylation, for example, because the treatment of Cap 42 (b) by a variety of protein phosphatases does not restore any POlymerizability.
It was previously shown that F-actin capping activity resides in Cap 42 (a) (1-3).
We have found that Cap 42 (b) required Ca2+ for its F-actin capping activity when phosphorylated, but that the F-actin capping activity becomes Ca2+-independent when it is dephosphorylated. This phosphorylation-dependent change in the Ca2+-requirement of Cap 42 (b) for its capping activity appears to be the same whether Cap 42 (b) is complexed with Cap 42 (a) or not. Since Cap 42 (a), Cap 42 (b), and the equimolar complex show almost the same specific F-actin capping activities in the presence of Ca", it is conceivable that both Cap 42 (a) and Cap 42 (b) are equally active in the complex. However, if we assume that in the dephosphorylated state and in the absence of Ca", the F-actin capping activities of both Cap 42 (a) and Cap 42 (b) are independently expressed in the complex, we would expect that Cap 42 (a) is inactive and only Cap 42 (b) is active, resulting in half of the specific F-actin capping activity of the complex, in comparison with the activity expressed in the presence of Ca2+. In fact, however, the F-actin capping activity of the dephosphorylated complex is absolutely Ca2+ independent. This suggests the activation of the F-actin capping activity of one subunit by the other in the following two alternative ways: (i) dephosphorylated Cap 42 (b) fully activates the F-actin capping activity of Cap 42 (a) even in the absence of Ca2+ as we previously suggested (5) or (ii) Cap 42 (a) stimulates 2-fold the F-actin capping activity of dephosphorylated Cap 42 (b) at least in the absence of Ca2+, probably due to an increase in the affinity of the latter to actin.
It is also worth-while to note that Cap 42 (a) forms a very tight (1:l) complex with Cap 42 (b) (4, 5 ) as fragmin does with actin (1-3). However, Cap 42 (a) never formed such a tight complex with actin nor did fragmin with Cap 42 (b). This clearly indicates that Cap 42 (a) is not identical with fragmin nor Cap 42 (b) with actin, respectively. On the other hand, however, our preliminary data obtained from immunological analysis, tryptic peptide maps, and direct photoaffinity labeling by ATP of these four proteins of 42,000 daltons strongly suggest that Cap 42 (a) is, structurally and functionally, closely related to fragmin, while Cap 42 (b) is closely related to actin (14).* Nevertheless, in order to establish the H. Maruta