Purification and Characterization of a Novel Ca2+-binding Protein (CBP- 18) from Bovine Brain*

A novel Ca2+-binding protein (CBP-18) has been identified and purified from bovine brain. On sodium dodecyl sulfate-polyacrylamide gel electrophoresis, the purified protein consists of a single band of apparent M, 18,000 in the presence of Ca2+ or 20,000 in the presence of EGTA. CBP-18 contains one high affinity Ca2+-binding site, measured at M Ca2’ in the pres- ence of 1 mM M e + and 0.1 M K+. The amino acid composition and UV absorption spectrum distinguish CBP-18 from other Ca2+-binding proteins identified in brain. The protein has an extinction coefficient c!i%.e., = 4.9 and contains 1 tryptophan/mol, 5 tyrosines/mol, and no trimethyllysine. CBP-18 does not interact with or activate calmodulin-stimulated phosphodiesterase. However, available evidence suggests that CBP- 18 binds to other component(s) present in the brain extract in a Ca2+-dependent manner. calcineurin sequence homology, in the hand”

They also undergo conformational changes on binding Ca2+ (1-3) and, even under denaturing conditions, exhibit alterations in electrophoretic mobility in the presence or absence of Ca2+ (6). The characteristic Ca2+-dependent electrophoretic shift can be used to detect these Ca2+-binding proteins in crude protein mixtures. The present report describes the purification and characterization of a novel Ca2+-binding protein detected in bovine brain extracts using this method.
UV absorption spectra were measured with a Cary Model 118 C spectrophotometer. Protein concentrations were determined by absorbance assuming an &&, , , , , of 10 for the crude fractions or 6;znrn = 4.9 for the purified protein. Amino acid analyses were performed by a modification of the method of Spackman et al. (13), using a Waters High Performance Liquid Chromatography Amino Acid Analysis System, equipped with a CAT EX Resin column (0.4 X 25 cm). The ionic strength gradient system was as described in the manufacturer's manual. At 58 "C, trimethyllysine elutes at 59 min between phenylalanine (retention time 51 min) and lysine (retention time 63 min).
Ca" binding was measured by the method of Hummel and Dreyer as described in Ref. 14. The specific activity of '%az+ in buffers and eluting fractions was based on determinations of Ca2+ concentration using a Perkin-Elmer Model 5000 atomic absorption spectrometer. All buffers were prepared using Analar reagents (BDH Chemicals). A solution of purified CBP-18 (1.1 mg/ml) made 1 mM ["CJEDTA (2000 cpm/nmol) was dialyzed overnight against 1000 volumes of 0.05 M Tris-HC1 buffer, pH 7.5, containing 0.1 M KCl, 1 mM MgClz, and 0.1 mM dithiothreitol, with two changes of dialysis fluid. After dialysis, [14C]EDTA was less than 1 p~ in the protein solution. A column of Sephadex G-25 (1 X 17.5 cm) was equilibrated at 0-4 "C with the above buffer containing 13 p~ '6CaC12. A 0.2-ml sample of the CBP-18 solution, adjusted to the same total 45CaC12 concentration as that of the column buffer, was applied and eluted with identical buffer at a flow rate of 3 ml/h, collecting 0.4-ml fractions.

RESULTS
A new Ca2+-binding protein (CBP-18) was initially identified in fractions obtained by DE23 chromatography of a bovine brain extract (Fig. 1). When column fractions were subjected to SDS-polyacrylamide gel electrophoresis, three low molecular weight proteins were found to exhibit Ca2+induced alterations in mobility. Two of these proteins were calcineurin B and calmodulin. The third protein, which exhibited apparent M , 18,000 in the presence of Ca2+ or 20,000 in the presence of EGTA, had not yet been described. This protein (CBP-18) had a larger M, and eluted earlier from DE23 than calcineurin B or calmodulin. When a calmodulin-Sepharose column was deliberately overloaded with pooled material from the DE23 column, CBP-18 was recovered in flow-through fractions eluting in the presence of Ca2+. This flow-through material (fraction I), depleted of calcineurin and calmodulin, was used as a source of CBP-18 for subsequent purification and characterization.
Effect of Cu2+ on Ammonium Sulfate Precipitation of CBP-I8"When aliquots of fraction I were subjected to ammonium sulfate fractionation in the presence of Ca", approximately two-thirds of CBP-18 was precipitated at 35% saturation ( Table I)  but was quantitatively precipitated in the presence of Ca2+ ( Table I). The differential effect of Ca2+ on the ammonium sulfate fractionation was exploited in the purification of CBP-18. Fraction I (10.6 liters derived from 9.5 kg of bovine brain) in 0.04 M Tris-HC1 buffer, pH 7.5, containing 0.2 mM CaC12, 3 mM MgC12, 0.05 M NaCl, 0.1 mM dithiothreitol, 1 pg/ml of leupeptin, and 10 pg/ml of soybean trypsin inhibitor, was brought to 45% saturation with ammonium sulfate. The precipitate, which contained CBP-18, was isolated by centrifugation, resuspended in the original volume of buffer containing 2 mM EGTA instead of Ca", and again brought to 45% saturation with ammonium sulfate. CBP-18 was recovered in the supernatant fraction. This material was dialyzed overnight against 10 volumes of 0.05 M (NH), HCOs with one change of dialysis fluid. The protein was lyophilized, resuspended in a minimal volume of 0.04 M Tris-HC1 buffer, pH 8.0, containing 0.1 M NaCl and 0.5 mM dithiothreitol, and centrifuged at 8000 x g for 30 min to remove insoluble material.
The partially purified CBP-18 was then applied to a column (2.6 X 79 cm) of Sephadex G-200 (superfine) which had been equilibrated with 0.04 M Tris-HC1 buffer, pH 7.5, containing 0.1 M NaCl, 1 mM MgClz, 0.1 mM EGTA, and 0.5 mM dithiothreitol. The column was eluted with the same buffer at a flow rate of 4 ml/h, collecting 4.8-ml fractions.
Interaction of CBP-18 with CAPP Affi-Gel-Fractions from the Sephadex G-200 column which contained CBP-18 were pooled, made 2 mM CaC12, and applied to a column (1.5 x 3.5cm) of CAPP Affi-Gel equilibrated with 0.04 M Tris-HC1 buffer, pH 7.5, containing 0.05 M NaCl, 1 mM MgC12, 2 mM CaClz, and 0.05 mM dithiothreitol (buffer 1). The column was washed with 70 ml of buffer 1 followed by 70 ml of buffer 1 made 0.5 M NaCl. A small amount of calmodulin was removed from the column on substitution of 2 mM EGTA for CaC12 in the eluting buffer (Fig. 1). CBP-18 was also selectively bound to the column but, unlike calmodulin, CBP-18 was retained on the column in the absence of Ca2+ and was recovered only when 6 M urea was added to the eluting buffer (Fig. 1). 7 430 a CBP-18 was quantified by densitometric analysis of Coomassie blue stained gels, as described under "Experimental Procedures." Fraction I, which contained 0.2 mM CaC12, was brought to 45% saturation with ammonium sulfate. Precipitated material was resuspended in buffer containing 2 mM EGTA instead of Ca2+, again brought to 45% saturation with ammonium sulfate, and the supernatant used for the next step.   HPLC Purification of CBP-18"Reverse phase HPLC on an alkylphenyl column was used as the final step in the purification of CBP-18. Aliquots (3.5 ml) of dialyzed eluate from the CAPP Affi-Gel column were applied to an alkylphenyl column (3.9 mm X 30 cm, 10-pm beads, Waters Associates) which had been equilibrated with 0.05 M potassium phosphate buffer, pH 6.5, containing 1 mM EGTA (buffer A). Elution was performed at room temperature with a 20-min linear CH&N gradient from 100% buffer A to 30% buffer A and 70% CH3CN. The flow rate was maintained at 1.5 ml/ min. CBP-18 eluted as a single peak with a retention time of approximately 18 min. CBP-18 was incompletely recovered under these conditions, and additional CBP-18 could be eluted from the same column by repeated washes with CH,CN. Fractions containing CBP-18 from serial injections were pooled, evaporated to remove the solvent, dissolved in 0.05 M NH4HC03, and desalted by chromatography on a 9-ml column of Sephadex G-25 (PD-10 column, Pharmacia) equilibrated and eluted with 0.05 M NH4HC0,. Fractions containing CBP-18 were pooled and stored at -70 "C. On SDS-gel electrophoresis, the purified CBP-18 consisted of a single band of apparent M, 18,000 in the presence of Ca2+ and 20,000 in the presence of EGTA (Fig. 1). The purification procedure is summarized in Table 11.
Characterization of the Purified Ca2+-binding Protein-The UV absorption spectrum of CBP-18 is shown in Fig. 2. Using a protein concentration determined by amino acid analysis, an extinction coefficient = 4.9 was calculated. The UV absorption spectrum exhibits a shoulder at 290 nm, suggesting the presence of tryptophan. The amino acid composition (Table 111) indicates that CBP-18 is rich in acidic residues and contains 12 lysines, 2 histidines, and 5 tyrosines/mol. No trimethyllysine was detected. A molar extinction coefficient = 8900), based on a M, of 18,000, is consistent with the presence of one tryptophan and 4 to 5 tyrosines/mol. Direct evidence of Ca2+-binding was obtained by gel filtration of the purified protein in the presence of Ca2+ by the method of Hummel and Dreyer (14), as shown in Fig. 3. The protein eluted in the void volume associated with a peak of bound 45Ca2+. Binding of ''CaZ+ to the protein resulted in depletion of Ca2+ from the eluting buffer, manifested bp a Ca2+ trough in the included volume. Based on radioactivity and absorbance at 279 nm, CBP-18 bound 0.9 f 0.3 mol of Ca2+/mol at a Caz+ concentration of 13 p~.

DISCUSSION
A new Ca2+-binding protein (CBP-18) has been identified in bovine brain. Comparison of the amounts of CBP-18, calcineurin B, and calmodulin identified in DE23 fractions suggests that CBP-18 is present at a substantial level in brain. Uncertainty in recovery at early steps of extraction and ion-exchange chromatography precludes a reliable estimate of the total amount of the protein present.
CBP-18 exhibits properties typical of other intracellular Ca'+-binding proteins, including Ca'+-induced changes in electrophoretic mobility (6), and binding to CAPP Affi-Gel (15). However, based on present results, CBP-18 has several characteristics which distinguish it from the other Ca'+-binding proteins identified in brain (4)(5)(6)(7)(8)(9)(16)(17)(18)(19)(20). In addition to its unique electrophoretic mobility and amino acid composition, the protein fails to elute from CAPP Affi-Gel in the presence of EGTA. CBP-18 contains one high affinity Ca2+-binding site, measured at M Ca'+ in the presence of physiological concentrations of Mg2+ and K+. In crude extracts, CBP-18 is quantitatively precipitated on heating to 100 "C in the presence of EGTA under conditions in which added calmodulin remains soluble (data not shown). CBP-18 does not activate calmodulin-regulated cyclic nucleotide phosphodiesterase, nor does it inhibit activation of the enzyme by calmodulin (data not shown). Thus, CBP-18 does not interact with the calmodulin-binding site of this enzyme. In crude extacts, the Ca'+-dependent precipitation of CBP-18 with low concentrations of ammonium sulfate suggests a Ca'+-dependent interaction with another constituent in the brain extract. Chromatographic behavior of CBP-18 provides further evidence of such an interaction. First, on gel filtration of a brain extract, CBP-18 elutes in the void volume in the presence of Ca'+, and in the included volume in the presence of EGTA. In contrast, free CBP-18 elutes in the included volume even in the presence of Ca'+, thus excluding Ca'+dependent self-association as an explanation for this behavior (data not shown). Second, purified CBP-18 does not interact with calmodulin-Sepharose. However, if a crude extract containing CBP-18 is applied to a calmodulin-Sepharose column under conditions in which immobilized calmodulin is in excess, Ca'+-dependent retention of CBP-18 can be detected (data not shown). These observations suggest that CBP-18 interacts with a low affinity calmodulin-binding component. The possibility that CBP-18 is a Ca2+-binding myosin light chain cannot be ruled out. Myosin is known to be present in mammalian brain (21), and interaction of myosin heavy chains with calmodulin has been observed by 1251-calmodulin gel overlay.' Alternatively, CBP-18 may interact with a yet undiscovered calmodulin-binding protein. Such an interaction is reminiscent of phosphorylase kinase (22) and calcineurin (6), which are capable of simultaneous interaction with two different Ca'+-binding proteins. Efforts are in progress to identify constituents which bind to and are potentially regulated by this novel Ca'+-binding protein. Note Added in Proof-After this paper was accepted, the purificaby Waisman et al. (FEBS Lett. (1983) 164, 80-84). The protein tion of a similar Ca2+-binding protein from bovine brain was reported described by Waisman et al. exhibits characteristics similar to those of CBP-18, including UV absorption spectrum and early elution from DEAE-cellulose but the M, determined by SDS-gel electrophoresis and the amino acid compositions of the two proteins appear to be different. Waisman et al. have suggested that their Ca2+-binding protein may be present only in brain tissue. Recent experiments in our laboratory indicate that CBP-18 is brain-specific. Using a polyclonal antibody to CBP-18, the protein was detected in a crude homogenate of bovine brain by Western blot analysis. It was not detected in homogenates of bovine heart, kidney, skeletal muscle, spleen, liver, stomach, lung, or thymus, even with a 10-fold increase in the amount of homogenate applied.