Cell Motility and the 14:251-262 (1989)

Characterization of Renatu red ProfiI in Purified by Urea Elution From Poly-L-Proline Agarose Columns

Donald A. Kaiser, Pascal J. Goldschrnidt-Clerrnont, Barry A. Levine, and Thomas D. Pollard

Department of Cell Biology and Anatomy, The Johns Hopkins Medical School, Baltimore (D.A.K., P.J. G. -C., T.D. P.); Inorganic Chemistry Laboratory, Oxford University, Oxford, England (B.A.L.)

We present evidence that native can be purified from cellular extracts of Acanrhamoeba, Dictyostelium. and human platelets by affinity chromatography on poly-L-proline agarose. After applying cell extracts and washing the column with 3 M urea, homogeneous profilin is eluted by increasing the urea concentra- tion to 6-8 M. Acanrhamoebu profilin-I and profilin-I1 can subsequently be separated by cation exchange Chromatography. The yield of Acunthamoebu pro- filin is twice that obtained by conventional methods. Several lines of evidence show that the fully renature after removal of the urea by dialysis: 1) dialyzed Acanthamoeba and human profilins rebind quantitatively to poly- L-proline and bind to in the same way as native, conventionally purified profilin without urea treatment; 2) dialyzed profilins form 3-D crystals under the same conditions as native profilins; 3) dialyzed Acanthamoeba profilin-I has an NMR spectrum identical with that of native profilin-I; and 4) dialyzed human and Acanthamoeba profilins inhibit actin polymerization. We report the discovery of profilin in Dictyxtelium cell extracts using the same method. Based on these observations we conclude that urea elution from poly-L-proline agarose followed by renaturation will be generally useful for preparing profilins from a wide variety of cells. Perhaps also of general use is the finding that either -11 or alpha- in crude cell extracts can be bound selectively to the poly-L-proline aga- rose column depending on the ionic conditions used to equilibrate the column. We have purified myosin-I1 from both Acanrhamoebu and Dicfyosrelium cell extracts and alpha-actinin from Acanrhnmoebu cell extracts in the appropriate buffers. These are retained as complexes with actin by the agarose and not by a specific interaction with poly-L-proline. They can be eluted by dissociating the complexes with ATP and separated from actin by gel filtration if necessary.

Key words: Acanthamoeba, affinity chromatography, Dictyostelium, NMR spectroscopy, platelets, myosin

INTRODUCTION al., 19881 has added further to their importance. At the present time a variety of detailed biochemical [Pollard Profilins are small actin monomer binding proteins originally discovered in vertebrate tissues [Carlsson et al., 19761 but now known to exist in many, if not all, cells [reviewed by Pollard and Cooper, 19861. The ob- Received March 7, 1989: accepted May 25, 1989. servation that profilins are inhibited by acidic phospho- Address reprint requests to Donald A. Kaiser. Department of Cell lipids [Lassing and Lindberg, 198.51 and that profilin is Biology and Anatomy. The Johns Hopkins Medical School. 725 N. probably necessary for cellular viability [ Magdolen et Wolfe Street. Baltimore. MD 21205. 0 1989 Alan R. Liss. Inc. 252 Kaiser et al. and Cooper, 1984; Kaiser et al., 1986; Lindberg et al., Platelets were washed by three successive centrifuga- 1988; Vandekerckhove et al., 19891, molecular biologi- tions (I ,000~for 30 min at 22°C) in 135 mM NaCI, 8.6 cal [Kwiatkowski and Bruns, 1988; Magdolen et al., mM trisodium citrate, 5.3 mM citric acid, and 11.1 mM 19881, physiological [Lind et al., 19871, and biophysical glucose, 10 mM Tris, pH 6.5. After the last wash the [Magnus et al., 19861 studies are being done with pro- pellet was resuspended in 5 volumes of ice-cold lysing filins in a number of laboratories. buffer containing 2 mM Tris (pH 7.2), 0.1 mM ATP, 0.5 Tanaka and Shibata [ 19851 discovered that profilin mM dithiothreitol, 1% Triton X-100, 1% dimethyl sul- binds to poly-L-proline immobilized on agarose beads. foxide, and 50 Fgiml each of chymostatin, leupeptin, The association is strong since at least 6 M urea is re- antipain, and pepstatin. After sonicating for 60 s, the quired to elute the bound profilin from the column. This extract was clarified by centrifugation at 12,000~for 15 property of profilin has already been used as an assay for min at 4°C. profilin in crude extracts [Lind et al., 19871. These re- sults suggested affinity Chromatography on immobilized Monoclonal Antibodies poly-L-proline would be a valuable method for purifying Mouse monoclonal antibodies to myosin-I, myo- profilin directly from crude extracts providing that the sin-11, alpha-actinin, profilin-I, and profilin-I1 from profilin eluted from poly-L-proline with urea could be Acanthamoeba were prepared as previously described renatured. [Kiehart et al., 19841. A mouse monoclonal anti-actin Here we show that Acanthamoeba profilins-I and antibody, c4D6, was kindly provided by Dr. James -11, Dictyostefium profilin, and human platelet profilin Lessard [Lessard, 19881. can be purified by affinity chromatography on poly-L- proline agarose. The profilins are eluted from the colzmn Reagents with 6-8 M urea, renatured by dialysis against a low I-Ethyl-3-(3-dimethylaminopropyl)carbodiimide ionic strength buffer without urea and, in the case of (EDC) and N-hydroxysulfosuccinimide (NHS) were pur- Acanthamoeba, separated into the isoforms by cation ex- chased from Pierce Chemicals (Rockford, IL). Both change chromatography. The yield is higher than con- EDC and NHS were dissolved at a concentration of 50 ventional methods by a factor of about 2, and the dia- mM in 2 mM potassium phosphate, pH 7.5, immediately lyzed profilins have native structure by several criteria. prior to use. The pH of the NHS was adjusted with Independently, Lindberg et al. [ 19881 showed that affin- NaOH. Cyanogen bromide and dimethylsulfoxide were ity chromatography on poly-L-proline-Sepharosewith from Fisher Chemicals, (Pittsburgh, PA). Sepharose-4B dimethylsulphoxide elution can be used to purify a mix- was from Pharmacia (Piscataway , NJ). Poly-L-proline ture of spleen profilin and profilactin complex. We have (MW 14,200), glycine , MES (2-[ N-morpholino]ethano- found that in addition to profilin, the poly-L-proline aga- sulfonic acid), tris base, ATP, EDTA, dithiothreitol, rose column also binds myosin-I1 and alpha-actinin from sodium dodecylsulphate, beta-mercaptoethanol, phe- crude extracts of Acanthamoeba and myosin-I1 from Dic- nylenediamine, Triton X- 100, chymostatin, leupeptin, tyostefium cell extracts. These proteins are retained as antipain, and pepstatin were from Sigma Chemical Co. complexes with actin, and it is the agarose beads rather (St. Louis, MO). Sucrose was obtained from J.T. Baker than the poly-L-proline that traps them. They can be Co. (Phillipsburg, NJ). Urea and guanidine-HCI were eluted by dissociating the complexes with ATP. from Schwarz-Mann Biotech (Cleveland, OH). Car- boxymethylcellulose ion exchange resin (CM-CM) was purchased from Amicon Corp. (Danvers, MA). MATERIALS AND METHODS Cell Extracts Poly-L-Proline Agarose Columns Acanthamoeha cell extracts were prepared by cen- One gram of poly-L-proline was dissolved in 100 trifuging homogenates of whole cells in 0.34 M sucrose, ml of distilled water by stirring at 4°C for 2 days and then 1 mM EDTA, 1 mM dithiothreitol, and 10 mM imida- coupled to 50 ml of Sepharose-4B activated with cyano- zole-HCI, pH 7.0, with or without 1 mM ATP at gen bromide by stirring for 20 h at 4°C [Tuderman et al., 100,000g for 90 min at 4°C [Tseng et al., 19841. Dicfy- 19751. After each use the beads were washed with 6 or 8 ostelium Ax-3 (wild-type) cells in 0.20 M sucrose, 0.2 M urea or 5 M guanidine for 1-2 h to remove all bound mM EDTA, and 10 mM tris-HCI, pH 8.0, were lysed by material. The columns have been used more than 20 forcing through a 5 pm filter and centrifuged at times without loss of activity or specificity. 100,000g for 60 min at 4°C [Devreotes et al., 19871. Cells were developed 4-5 h prior to lysis. Outdated hu- Biochemical Methods man platelets (7-9 days post-phlebotomy) were provided Proteins were separated by polyacrylamide gel by the American Red Cross in Baltimore, Maryland. electrophoresis in sodium dodecyl sulfate [Laenimli, Purification and Renaturation of Profilin 253 19701 and either stained with Coomassie blue or trans- tract binds in a high salt buffer (Fig. IA) but little or none ferred to nitrocellulose paper as previously described binds in high salt buffer with ATP or in a low salt buffer [Kiehart et al., 1984; Towbin et al., 19791. con- (Fig. IB). In low salt, a substantial amount of alpha- centrations were measured by Coomassie blue dye bind- actinin is retained and little or no myosin binds to the ing [Bradford, 19761 using ovalbumin as a standard. column (Fig. IA,B). Profilin binds quantitatively in both buffers. Only a small fraction of the total actin binds in lmmunochemical Methods either buffer (Fig. IA). Control experiments showed that Proteins were detected on nitrocellulose paper with myosin-I1 (Fig. 2) is retained by the agarose beads alone, monoclonal antibodies followed by reaction with goat but that the profilin binds to the poly-L-proline. Purified anti-mouse second antibodies which were labeled with myosin-I1 is not retained by the agarose unless actin is either horseradish peroxidase (Hy-clone Laboratories present under conditions which favor the formation of [Logan, UT]) or iodine- 125 (Cappel Laboratories [Bur- actomyosin (not illustrated). lingame, CAI) [Kiehart et al., 19841. For semi-quanti- The myosin-I1 and alpha-actinin retained on tative analysis, fractions were diluted in 10 mM imida- columns of agarose or poly-L-proline-agarose are eluted zole-HCI, pH 7.0, and dried onto polyvinyl-chloride 96- quantitatively by 5 mM MgATP (Figs. 1-4). Depletion well plates for ELISA with mouse monoclonal antibodies of the endogenous ATP in the extract with hexokinase and goat anti-mouse second antibodies labeled with and glucose did not improve the yield of either horseradish peroxidase (Tago Inc., Burlingame, CA). myosin-I1 or alpha-actinin significantly. However, in The substrate O-phenylenediamine was used to detect extracts where 5 mM ATP or 30 mM pyrophosphate antibody binding by measuring absorbance at 490 nm in were added prior to chromatography, neither myosin-I1 a Biotech EL-310 plate reader. nor alpha-actinin bound (data not shown). The bound actin is eluted partially by MgATP and completely by 3 'H-NMR Spectral Characterization M urea which denatures the actin irreversibly. Profilin is For NMR studies Acuntharnoeba profilin-I purified not necessary for retention of actin and myosin on either by conventional methods [Kaiser et al., 19861 or by elu- column, since actin and myosin-I1 can be purified from tion from poly-L-proline with 8 M urea were concen- extracts depleted of profilin (not illustrated). The trated by precipitation with 0.4 giml ammonium sulphate profilin is quantitatively retained on poly-L-proline and dialyzed against 10 mM NaHC03, 20 mM NaCl in agarose in 3 M urea, but is all eluted by 6 M urea (Figs. 'H20, pH 7.4. The proteins were further concentrated 1, 3) or 5 M guanidine (not illustrated). The profilin and exchanged into 5 mM NaHC03 in 'H20 by ultrafil- elutes in a sharper peak with 8 M urea, presumably due tration on an Amicon 8MC membrane to a final concen- to some weak interaction with the column even in 6 M tration of approximately 10 mg/ml. 'H-NMR spectra urea. were obtained at 500 and 600 MHz at an ambient probe The myosin-I1 is the major protein eluted from the temperature of 303°K. Two-dimensional correlated high salt column with MgATP (Figs. lA, 2, 3). The (COSY) spectra were recorded using the TPPI method contaminating actin is easily removed from the myosin-I1 [Bodenhausen et al., 19801 to produce phase-sensitive by gel filtration using standard methods [Pollard et al., spectra with irradiation of the residual OH resonance 19781. This highly purified myosin-I1 has normal Ca" during the relaxation time. Spectra were recorded with a ATPase activity. The yield of myosin-I1 from 1 g of cells sweep width of 6,024 Hz in both dimensions with 192 is about 0.25 mg, about 30% of the total of 2 pmol/kg scans per TI increment. The resulting data matrix (512 [Pollard et al., 19781. This yield is about the same as that x 4K) was filtered by shifted sine bell multiplication obtained by conventional methods. and zero filled to 2K in T, before complex Fourier trans- The alpha-actinin bound to the column in low salt formation. One dimensional nuclear Overhauser en- buffer elutes with 5 mM ATP (Figs. lB, 4). It has the hancement (NOE) measurements were carried out using same complex electrophoretic pattern of multiple bands an 0.8 s irradiation pulse with the decoupler in the HG as Acantharnoeba alpha-actinin purified in other ways mode. [Pollard et al., 19861. It is positively identified as alpha-actinin by binding to anti-alpha-actinin mono- clonal antibodies (Fig. I, 4). The yield is about 0.38 mg RESULTS per g of cells, about 4 times higher than conventional Affinity Chromatography of Crude Extracts of methods [Pollard et al., 19861. We presume that like Acanthamoeba on Poly-L-Proline myosin-11, alpha-actinin is retained as a complex with Profilin, myosin-11, alpha-actinin, and actin bind to actin. We do not yet understand why this association of columns of poly-L-proline immobilized on agarose alpha-actinin with actin occurs more favorably in the beads. A substantial amount of the myosin-I1 in the cx- low salt buffer. Column A Column B Acanthamoeba extract in high salt Acantharnoeba extract in low salt

- 2.0 6.0 t t a I- 1.6 -

3 1.2 - 'I s ng 2 v)E 0.8 - 88

2.0t/\1.o 1 1 1 1 00 0 10 20 30 40 50 60 70

0.3 -

0.2 -

0.1 -

0.0 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70

0.8 r

0 10 20 30 40 50 60

0.5 I 1 0.5 s

04

0.3

02

0.1

00 0 10 20 30 40 50 60 Fraction NLnrber Fraction Nm&r Purification and Renaturation of Profilin 255 3 to 4 M urea, while the elution of the profilin requires ut- a 6 M urea (Fig. 5). Although actin was irreversibly de- [r t- natured by this treatment, purified profilin was renatured X W 5mM MgATP by dialysis against 0.2 mM ATP, 0.1 mM CaCI2, 0.5 12 34 56 mM dithiothreitol, 2 mM Tris pH 7.2 for 48 h. The yield

200- is about 0.5 mg profilin per gram of packed platelets. Renatured profilin has normal functional properties in- cluding binding to actin, inhibition of actin polymeriza- tion, and binding to poly-L-proline-agarose. 68 - 60- 55 - Affinity Chromatography of Crude Extracts of 43 - Dicfyosfelium discoideum on PoI y- L-Pro1 i n e Chromatography of Dictyostelium extracts on poly- L-proline equilibrated with high salt gives the same re- sults as the Acanthamoeba extracts (Fig. 6);both myo- 29 - sin-I1 and a polypeptide the size of profilin bind and are eluted with MgATP and 6 M urea, respectively. The myosin-I1 was positively identified by its cross-reactivity with 2 monoclonal antibodies to Acanthamoeba myosin- 14- I1 (Fig. 6). The polypeptide eluting in a highly purified form with 6 M urea coelectrophoreses with Acan- thamoeba profilin and is positively identified as Dictyo- stelium profilin by cross-reactivity in ELISA with 2 of our 14 monoclonal antibodies to Acanthamoeba profi- Fig. 2. Retention of Acunthumoebu myosin-I1 and actin on a control lins. None of these antibodies bound to the Dictyostelium column of agarose without poly-L-proline. Twenty milliliters of ex- protein on immunoblots after gel electrophoresis in SDS tract containing 57 1 mg protein was diluted into 20 ml high salt buffer (data not shown). and applied to a 2.0 X 1.5 cm column of Sepharose 4B. The column was rinsed with 1 L of high salt buffer. Samples were analyzed by Large-Scale Purification Acanthamoeba SDS-polyacrylamide gel electrophoresis. Shown are the extract sam- of ple applied to the column (lane l), the unbound fraction (lane 2), and Profilins by Affinity Chromatography fractions eluted with 5 mM MgATP in high salt buffer (lanes 3-6). Substitution of poly-L-proline affinity chromatog- Approximately 6.5 mg of actomyosin was recovered in this experi- raphy for hydroxylapatite and gel filtration chromatog- ment. raphy [Reichstein and Korn, 19791 results in improved yield and purity of Acanthamoeba profilins. Crude ex- Affinity Chromatography of Extracts of Human tracts can be applied directly to the affinity column (Figs. Platelets on Poly-L-Proline I, 3), but we usually start with a large DEAE-cellulose column, since it is the initial step in the purification of Chromatography of extracts of human platelets on several other proteins including actin and myosin-I1 poly-L-proline-agarose retains virtually all detectable [Tseng et al., 19841. The flow-through from the DEAE- profilin present in the extract and a small fraction of the column containing the profilins is run on the affinity actin (corresponding to approximately one third of the column. The capacity of the poly-L-proline agarose col- profilin concentration, Fig. 5). The actin was eluted with umn is 5-10 mg of profilin per ml, so that the 1.5 x 15 cm column is large enough for 1.3 kg of cells. After

~~ washing with buffer and 3 M urea, highly purified pro- Fig. I. Column A: A 3 ml extract of A~~~rr,~tlztrtnorbtrcontaining 20 filins are eluted with 6 or 8 M urea (Fig. 7). The profilins mg of protein was applied to a 5 X 0.75 cm column of poly-L-proline are renatured by dialyzing out the urea and the two iso- agarose equilibrated in high salt (100 mM NaCl, 100 mM glycine, 1 forms separated by cation exchange chromatography mM dithiothreitol. and 10 mM Tris-HCI. pH 8.0).Column B: 4 ml of Acuiithumoohli extract containing 35 mg of protein was applied to a (Fig. 7). The yield of total profilin is about 0.1 mg per 15 x I .5 cm column ofpoly-L-proline agarose equilibrated in low salt gram of cells compared with 0.05 mg by the Kaiser et al. (0.34 M sucrose. I .O mM EDTA, I .O mM dithiothreitol. and 10 mM [ 19861 modification of the method of Reichstein and imidarole, pH 7.0). Unbound protein was eluted with column buffer Korn [ 19791. The ratio of purified profilin-I to profilin-I1 and bound material was eluted successively with 5 mM MgATP. 3 M is about 2 to 1 compared with 4 to 1 by the method of urea, and 6 M urea all in column buffer. The top panels show the protein concentration in the fractions. The lower panels show ELISA Kaiser et al. [ 19861. In the conventional method, part of assays for profilin. actin. myosin-Il. and alpha-actinin the profilin-I1 is lost during hydroxylapatite chromatog- 256 Kaiser et al.

COOMASSIE BLUE ANTI-PROFILIN ANTI-ACTIN ANTI-MYOSINII ANTI-ALPHA-ACTININ ANTI-MYOSIN I

200 - 116 - 95 - 68 - 60 - 55 - 43 - 40 -

29 -

18 -

11 -

Fig. 3. Affinity chromatography of Acanrhamoeba extract on poly- mouse second antibodies labeled with '"1. The lanes in each panel are L-proline agarose in high salt buffer. Samples were analyzed by SDS- extract sample applied to column, unbound fraction, fraction eluting polyacrylamide gel electrophoresis and immunoblotting. Left: A gel with 5 mM MgATP, fraction eluting with 3 M urea, and fraction stained with Coomassie blue. Right: Autoradiograms of five identical eluting with 6 M urea. (The immunoreactive bands migrating ahead of gels transferred to nitrocellulose and reacted with monoclonal anti- the myosin-I1 heavy chain in the anti-myosin-11 panel are breakdown bodies specific for profilin, actin, myoain-If, alpha-actinin and myo- products of the myosin-11 heavy chain. sin-I. Monoclonal antibody binding was visualized with goat anti-

raphy where none of the profilin-I, but part of the pro- Comparative 'H-NMR Spectral Study of Profilin-l filin-11 binds to the column (D.A. Kaiser and T.D. Pol- Prepared Using Poly-L-Proline Purification and lard, unpublished observations). Conventional Procedures The two Acanrhamoeba profilin isoforms purified by urea elution from poly-L-proline and renatured by The spectra of the two profilin-I preparations (Fig. dialysis are native by several criteria. In a chemical 9) are virtually indistinguishable. The identity in chem- cross-linking assay (Fig. S), both bind to actin monomers ical shift position and resonance lineshape for signals as well as profilins purified without urea [Vandekerck- that provide a fingerprint for the conformation of the hove et al., 19891. They also inhibit actin polymerization protein proves that the renatured profilin-I adopts the as described previously [Pollard and Cooper, 19841 and native structure. Some 30 amide hydrogens slow to ex- rebind to poly-L-proline. Both renatured profilin iso- change with solvent (t1,2 > 20 h) are readily resolved forms also bind to anti-profilin monoclonal antibodies (7.5 < 6 < 9.7 ppm, Fig. 9) while downfield-shifted equally well [Kaiser and Pollard, 19881. The affinity backbone-CalphaHsignals, typical of beta-sheet segments purified preparations also form crystals under the same [Dalgarno et al., 19831, also show spectral homology. conditions as conventional preparations [Magnus et al., Correlation of these indicators of beta-structure and sev- 19861 (L. Machesky and T.D. Pollard, unpublished ob- eral of the slowly exchanging -NH resonances is shown servations). by the cross peaks in the identical COSY spectra ob- Purification and Renaturation of Profilin 257

COOMASSIE ALPHA-ACTININ ACTIN MYOSIN II 12 34

200 - 116 - 95 - 68 - 60 - 55 - 200 - 116 - 43 - 95 - 68 - 60 = 55 29 - 43 - 40 -

18 - 29 - 11 -

Fig. 5. Affinity chromatography of a human platelet extract on poly- L-proline-agarose. Samples were analyzed by SDS-polyacrylamide gel electrophoresis and staining with Coomassie blue. Lane 1: Triton soluble extract (10 pg protein). Lane 2: Material that did not bind to 18- the column (10 pg protein). Lane 3: Pooled fraction eluted from the column with 4 M urea (2 pg protein). Lane 4: Pooled fractions eluted 11- from the column with 8 M urea (3 pg protein). This fraction contained a single protein of about 15 kD.

Fig. 4. Affinity chromatography of Acanrharnoeha extract on poly- L-proline agarose in low salt buffer. Samples were analyzed by SDS- polyacrylamide gel electrophoresis and immunoblotting. Left: A gel stained with Coomassie blue. Right: Autoradiograms of 3 identical gels transferred to nitrocellulose and reacted with antibodies to alpha- lished observations); 4) structure from NMR spectra actinin, actin, and myosin-11. They were prepared as in Figure 2, (Figs. 9, 10); and 5) binding of anti-profilin monoclonal except that the gel sample of the fraction eluting with 5 mM ATP was not boiled. This accounts for slower niigration of the upper band of antibodies (Figs. I, 3) [Kaiser and Pollard, 19881. alpha-actinin [Pollard et al., 19861. This work establishes the utility of poly-L-proline affinity chromatography for the simple and rapid purifi- cation of profilin in high yield from a variety of cells. Contaminating actin can be removed by a wash with 3 M urea during the affinity chromatography or by prior ad- sorption to DEAE (Fig. 7). In either case the procedures tained for both profilin-I preparations (Fig. 10). Reten- are somewhat simpler than the methods recently de- tion of the native structure by the renatured profilin-I is scribed by Lindberg et al. [I9881 since an additional also indicated by diagnostics of more localized folding hydroxylapatite column was required to obtain pure pro- characteristics seen from the ring-shifted methyl group filin. Our yields of profilin from both Acnnthamoeha signals (0.7 < 6 < I ppm) and the homologous aromatic (see Fig. 3) and human platelets are nearly quantitative. sidechain spectral region (6.3 < 6 < 7.5 ppm). The Our recovery of 0.5 mg of profilin from 1 g of platelets conserved proximity of the corresponding groups was is in the same range (0.4-0.5 mg per gram of spleen) verified using nuclear Overhauser enhancement effects. reported by Lindberg et al. [ 19881, although most of their profilin was bound to actin. Their use of DMSO to elute profilin from the poly-L-proline column has the DISCUSSION proven advantage of producing native profilactin. Profilin eluted from poly-L-proline agarose in 6-8 We have already used the Acunthumoeha and hu- M urea and dialyzed extensively into low salt buffers man profilins for a variety of structural and functional without urea appears to be native by the following cri- studies. Based on our discovery of profilin in Dictyoste- teria: 1) rebinding to poly-L- proline; 2) binding to actin liutn, we expect that the method can be used effectively (Fig. 8): 3) formation of 3D crystals [Magnus et al., for many other cell types. It is still possible that more 19861 (L. Machesky and T.D. Pollard, 1988, unpub- detailed studies of the mechanism of action or structure 258 Kaiser et al.

COOMASSIE BLUE IMMUNOBLOTS

200 -

116 - 95 -

68 - 60 - 55 -

43 - 40 -

29 -

18 -

11 -

Fig. 6. Affinity chromatography of DicQosreliurn extract on poly- with monoclonal antibodies to Acunthuinorba myosin-I1 and actin. L-proline agarose in high salt buffer. Left panel: An SDS-polyacry- Monoclonal antibody binding was visualized with horseradish perox- lamide gel stained with Coomassie blue showing the composition of idase-labeled goat anti-mouse second antibodies. Only M2.17 and the extract applied to the column. the unbound fraction. the fractions M2.42 cross-react appreciably with myosin-I1 from Dicryostoliurn. eluting with 5 mM MgATP. 3 M and 6 M urea. Last panel: A mixture None of 14 different monoclonal antibodies to Acanthurnorhu profilin of Acanthurnoubu profilin-1 and profilin-11. The irnmunoblots shown bound to the polypeptide with the same electrophoretic mobility as to the right show the reaction of samples eluted with 5 mM MgATP Actrnthuinoebu profilin that eluted in 6 M urea (data not shown).

of the profilins will reveal some difference between the Insight into the mechanism of binding of the vari- profilins recovered from the poly-L-proline columns by ous profilins to the poly-L-proline will have to await our denaturation with high concentrations of urea, but our current efforts to determine the structure of Acan- efforts to date have not uncovered any such differences. rhamorha profilin [ Magnus et al., 19861 and the complex Consequently, we feel it safe to conclude that both hu- of Actitithumorha profilin with poly-L-proline by X-ray man and amoeba profilins are capable of refolding into a crystallography. Since all of the known profilins have native conformation after denaturation with urea or similar amino acid sequences [Ampe et al., 1985; Ampe guanidine. and Vandekerckhove, 1987; Ampe et al., 1988; Kwiat- Purification and Renaturation of Profilin 259

12345 12345

200- 200 - 116 116- 95 - 95 - 68 - 60 - 68- 55 - 60- 55- 43 - 40 - 43- 40- 29 -

18 -

11 -

18-

11-

Fig. 8. Comparison by chemical cross-linking of the actin monomer binding activities of profilins purified by a conventional method with- out urea [Kaiser et al.. 19861 and by urea elution from poly-L-proline. Mixtures of 20 pM actin and 20 pM profilin were dialyzed vs. 2 mm Fig. 7. Large-scale purification of profilins from Acantharnorba by KP, buffer, pH 7.5, and cross-linked with I mM EDC and 1 mM NHS affinity chromatography on poly-L-proline agarose. The samples were for 30 min at 22°C [Vandekerckhove et al., 19893. Samples were analyzed by SDS-polyacrylamide gel electrophoresis and stained with analyxd by SDS-polyacrylamide gel electrophoresis and stained with Coomassie blue. Lane 1: The flow-through fractions from DEAE- Coomassie blue. Lane 1: Rabbit skeletal muscle actin. Lane 2: Con- cellulose chromatography of a low-salt extract of Accor//~amoebathat ventional profilin-1 plus actin. Lane 3: Poly-L-proline purified profi- lin-1 plus actin. 4: Conventional profilin-I1 plus actin. 5: was derived from 970 g of cells. This material (3 L) was applied to a Lane Lane Poly-L-proline purified profilin-11 plus actin. The 55 kD covalently 15 X 1.5 cm column of poly-L-proline agarose. The column was of washed with 10 volumes of column buffer (100 mM NaCI, 100 mM cross-linked I: I complex profilin and actin is indicated. glycine, I mM dithiothreitol, 10 mM Tris-HCI, pH 8.0)and 1 volume of 3 M urea in column buffer. Lanes 2 and 3: Profilin eluted with 6 M urea. This mixture of 2 profilin isofomis was then fractionated by cation-exchange chromatography on Amicon CM-CM in 10 mM MES, pH 6.0. Lane 4: Purified profilin-1. Lane 5: Purified profilin- 11. A total of 103 mg profilins was obtained with approxirnately twice to actin. We suspect, but cannot yet document, that there as much profilin-l as profilin-11. are proteins in cells with segments of poly-L-proline that bind to profilin. Such interactions could modulate the activity of one of the members of the complex or possi- bly serve as an anchor for localizing prof'ilin in the cell. We have not carefully investigated the mechanism kowski et al., 19881, it is not possible to identify unique of the binding of alpha-actinin and myosin-I1 to the aga- regions of homology that might endow all of the proteins rose columns. but we point out the phenomenon here, in this family with the ability to bind to poly-L-proline. because the method appears to be very useful for the Probably the most interesting question raised by rapid. high-yield purification of these proteins from our work and the related studies of Tanaka and Shibata crude extracts. We hope that it will be useful to others, 119851, Lind et al. 119871 and Lindberg et al. I19881 is, particularly where a simple method is required to sepa- Why do profilins bind to poly-L-proline with such high rate these proteins from the bulk of cytoplasmic compo- affinity'? It is obvious from the conditions required to nents for analytical purposes. Since actin appears to be elute actin and profilin from the affinity column that required for the retention of both proteins by agarose, we profilin binds to poly-L-proline much more strongly than suspect that aggregates of the proteins are being trapped 260 Kaiser et al.

X

B

A

1's'' -"'I"" "'*I'.'' '7rT1171z z"'l"" '.''I"'' '."1''" "'lIt'*x ~"'~~"~~"~'~~"~x1'117s1 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1 .o 0.0 PPn

Fig. 9. Proton magnetic resonance spectra of A) profilin-I prepared by agarose. C: A fourfold expansion of the low-field region of spectrum conventional procedures and B) profilin-I renatured from 6 M urea B. Conditions: 303°K. pH 7.4. 10 mgiml protein. The signal marked following purification by affinity chromatography on poly-L-proline by an x derives from residual Trk.

rather nonspecifically. Such a mechanism is also consis- ACKNOWLEDGEMENTS tent with the observed elution of myosin-I1 by ATP, since dissociation of the actinmyosin complex could This work was supported by NIH research grant well release the individual proteins trapped in the column GM-26338. The authors are very grateful to Dr. Masa- as a large complex. hiko Sato and Sachiko Karaki for their help in preparing Purification and Renaturation of Profilin 261

a i. h.9 f:

11,r 1 I I,I I I I I I I I I lppM 9.5 9.0 8.5 PPM

B

A TII ' " ' " ' ' a I ''1' I' '-1 " ''I ' " I " ~'"~""1'"'~""I""~""I''"r'"1"''j''''I''' 9.0 8.0 7.0 6.0 5.0 4.0 3.0 2.0 1.0 0.0 PPM

Fig. 10. A: Filter 'H-NMR spectrum of native profilin-1 showing the slowly exchanging amide proton resonances. B: The aromatic sidechain and downfield C alpha H signals together with C a COSY section depicting NH-C,,,,,H cr-ma-peaks. Conditions as in Figure 9. monoclonal antibodies to profilins and alpha-actinin and Devreotes, P., Fontana, D., Klein, P., Sherring, J., and Theibert, A. to John Sinard and Dr. Richard Adams for many useful (1987): In Spudich, J.A. (Ed.): "Methods in Cell Biology." suggestions. We also thank Pamela Lilly and Dr. Peter Vol. 28. New York, NY: Academic Press, pp. 299-331. Kaiser, D.A., Sato, M., Ebert, R.F., and Pollard, T.D. ( 1986): Pu- Devreotes for providing Dicfyosteliurn cell extracts. rification and characterization of two isoforms of Acun- thumoebu profilin. J. Cell Bid. 102:221-226. Kaiser, D.A., and Pollard, T.P. (1988): Characterization of mono- REFERENCES clonal antibodies to profilin isoforms of Accinthamoebcr. J. Cell Bid. 107:30a. Ampe,C., Vandekerckhove. J., Brenner, L., Tobacrnan. L., and Kiehart, D.P., Kaiser, D.A.. and Pollard, T.D. (1984): Monoclonal Korn, E.D. (1985): The amino acid sequence of Acurzthrrmoeha antibodies demonstrate limited structural homology between profilin. J. Biol. Chem. 260:834-840. myosin isozymes from Acunrhamorbu. J. Cell Biol. 99: 1002- Ampe. C., and Vandekerckhove, J. (1987): The F-actin capping pro- 1014. teins of Physcrrum polwcyJherlum: cap42(a) is very similar, it (1988): not identical, to fragmin and is structurally and functionally Kwiatkowski, D.. and Bruns, G. Human profilin: Molecular very homologous to ; cap42(b) is Physcirutn actin. cloning, sequence comparison and chromosomal analysis. J. EMBO J. 6:4149-4157. Biol. Chem. 263:5910-5915. Anipe, C., Sato, M.. Pollard. T.D.. and Vandekerckhove. J. (1988): Lassing, I.. and Lindberg, U. ( 1985): Specific interaction between The prirnary structure of the basic isoform of Accrritherrnoehti phosphatidyl-inositol 4.5-bisphosphate and profilactin. Nature profilin. Eur. J. Biochem. 170:597-601. 3 14:472-474. Bodenhausen, G.. Vold. R.L.. and Vold.. R.R. (1980): Multiple Laernmli. U.K. ( 1970): Cleavage of structural proteins during the quantum spin-echo spectroscopy. J. Magn. Reson. 3793-106. assembly of the head of bacteriophage T4. Nature 227:680- Bradford. M. (1976): A rapid and sensitive method for the quantitation 685. of microgram quantities of protein utilizing the principle o! Lessard. J. (1988): Two monoclonal antibodies to actin: One muscle protein-dye binding. Anal. Biochem. 72:248-254. selective and one generally reactive. Cell Motil. Cytoskeleton Carlsson. L.. Nystrom, L.E.. Lindberg. U., Kannan. K. K.. Cid-Dresd- I0:349-362. ner. H., Lovgren. S.. and Jornvall, H. (1976): Crystallization Lind. S.E., Janmey , P.A.. Chaponnier. C.. Horbort. T.-J.. and of a non-muscle actin. J. Mol. Biol. 105:353-366. Stosscl, T.P. ( 1987): Reversible binding of gelsolin and profilin Ihlgarno. D.C.. Levine. B.A., and Williams, R.J.P. 11983): Struc- in human platelet extracts. J. Cell Biol. 105:833-842. tural information from MNR secondary chemical shifts of pep- Lindberg, U., Schutt, C.E.. Hellston. E.. Tjader, A.C., and Hult, T. tide alpha C-H protons in proteins. Biosci. Rep. 3:433. 1 19XX): The use of poly(L-proline)-sepharose in the purification 262 Kaiser et al.

of profilin and profiliactin. Biochim. Biophys. Acta 967:391- protein of low molecular weight from Acunrhumoeha c.ustdk- 400. nii that inhibits actin nucleation. J. Biol. Chem. 254:6174- Magdolen, V., Oechsner, U., Muller G., and Bandlow. W. (1988): 6179. The intron-containing for yeast profilin (PFY) encodes a Tanaka, M., and Shibata. H. (1985): Poly (L-proline)-binding proteins vital function. Mol. Cell. Biol. 8:5108-51 15. from chick embryos are a profilin and a profilactin. Eur. J. Magnus, K.A.. Lattman. E.E., Sato, M.. and Pollard, T.D. ( 1986): Biochem. IS I :291-297. Crystallization of Acunrhomoeho profilin-I. J. Biol. Cheni. Towbin. H.. Staehelin, T., and Gordon. J. (1979): Electrophoretic 261 : 13360-1 3361. transfer of proteins from polyacrylamide gels to nitrocellulose: Pollard, T.D., Stafford, W.F., and Porter, M.E. (1978): Chardcter- Procedure and some applications. Proc. Natl. Acad. Sci. USA ization of a second myosin from Acuiirhumoehu cusrelluiiii. J . 76:4350-4354. Biol. Chem. 253:4798-4808. Tseng, P.C.-H., Runge, M.S., Cooper, J.A., Williams. R.C., and Pollard, T.D., and Cooper, J.A. (1984): Quantitative analysis of the Pollard. T.D. (1984): Physical, immunochemical and func- effect of Acurirhamodm profilin on actin filament nucleation tional properties of Acunthamoeha profilin. J. Cell Biol. 98: and elongation. Biochemistry 23:663 1-6641. 2 14-22 I. Pollard, T.P., and Cooper, J.A. (1986): Actin and actin-binding pro- Tuderman, L.. Kuutti. E.. and Kivirikko, K. (1975): An affinity- teins. A critical evaluation of mechanisms and functions. column procedure using poly (L-proline) for the purification of Annu. Rev. Biochem. SS:987-1035. prolylhydroxylase. Eur. J. Biochem. S2:9-16. Pollard. T.D.,Tseng. P.C.-H., Rimm, D.L., Bichell, D.P., Williams, Vandekerckhove, J.. Kaiser, D.A., and Pollard, T.D. (1989): Acun- R.C., and Sinard, J. (1986): Characterization of alpha-actinin rhamoebu actin and profilin can be crosslinked between glu- from Acuizthumoeba. Cell Motil. 6649-661. tamic acid 364 of actin and lysine 115 of profilin. J. Cell Biol. Reichstein, E., and Korn, E.D. (1979): Acunrhomoehu profilin: A (in press).