Purification and Sequencing of Radish Seed Calmodulin ... · Wheat embryo CDPK phosphorylates a...

Plant Physiol. (1993) 101: 545-551

Purification and Sequencing of Radish Seed Calmodulin Antagonists Phosphorylated by Calcium-Dependent

Protein Kinase'

Gideon M. Polya*, Shubhra Chandra, and Rosemary Condron

Department of Biochemistry, La Trobe University, Bundoora, Victoria, 3083, Australia

A family of radish (Raphanus sativus) calmodulin antagonists (RCAs) was purified from seeds by extraction, centrifugation, batch-wise elution from carboxymethyl-cellulose, and high performance liquid chromatography (HPLC) on an SP5PW cation- exchange column. l h i s RCA fraction was further resolved into three calmodulin antagonist polypeptides (RCA1, RCAZ, and RCA3) by denaturation i n the presence of guanidinium HCI and mercaptoethanol and subsequent reverse-phase HPLC on a Ca column eluted with an acetonitrile gradient in the presence of 0.1% trif lu- oroacetic acid. The RCA preparation, RCA1, RCA2, RCA3, and other radish seed proteins are phosphorylated by wheat embryo Caz+-dependent protein kinase (CDPK). The RCA preparation contains other CDPK substrates in addition to RCAl, RCAZ, and RCA3. The RCA preparation, RCAl, RCA2, and RCA3 inhibit chicken gizzard calmodulin-dependent myosin light chain kinase assayed with a myosin-light chain-based synthetic peptide substrate (fifty percent inhibitory concentrations of RCA2 and RCA3 are about 7 and 2 ~LM, respectively). N-terminal sequencing by sequential Ed- man degradation of RCAl, RCA2, and RCA3 revealed sequences having a high homology with the small subunit of the storage protein napin from Brassica napus and with related proteins. l h e deduced amino acid sequences of RCA1, RCA2, RCA3, and RCA3' (a subform of RCA3) have agreement with average molecular masses from electrospray mass spectrometry of 4537, 4543, 4532, and 4560 kD, respectively. The only sites for serine phosphorylation are near or at the C termini and hence adjacent to the sites of proteolytic precursor cleavage.

Ca2+ acts as a "second messenger" in plants as well as in animal cells. Externa1 signals transiently elevate cytosolic free Ca2+ concentration, allowing the activation of enzymes by Ca2+-calmodulin or by Ca2+ (Hepler and Wayne, 1985). CDPKs that are activated at micromolar free Ca2+ concentrations are present in plant cells as well as in other eukaryote cells and are involved in Ca2+-mediated signal transduction through the (reversible) phosphorylation of proteins (Polya and Chandra, 1990; Roberts and Harmon, 1992).

Plant CDPKs phosphorylate a number of exogenous, animal-derived proteins, including casein, phosvitin, reduced BSA, and histones (e.g. histone 111-S, a histone H1-rich com- mercial calf thymus histone preparation) (Polya et al., 1989). Wheat CDPK catalyzes the phosphorylation of a number of synthetic peptides having a Basic-X-X-Ser(Thr) amino acid

sequence but not synthetic peptides with the Basic-Basic-X- Ser(Thr) sequence recognized by cyclic AMP-dependent protein kinase (Polya et al., 1989).

Wheat embryo CDPK phosphorylates a number of endogenous protein substrates including histone H1, histone H2B (Polya et al., 1989), an a-amylase inhibitor (Polya and Chan- dra, 1990), and a wheat putative phospholipid transfer protein (Polya et al., 1992). Wheat CDPK phosphorylates a phospholipid transfer protein from barley (Polya and Chan- dra, 1990) as well as highly homologous proteins from wheat and pine (Polya et al., 1992). Some plant enzymes are plant CDPK substrates, namely NAD+-quinate oxidoreductase (Ranjeva and Boudet, 1987), H+-ATPase (Schaller and Suss- man, 1988), and phosphoenolpyruvate carboxylase (Echev- arria et al., 1988). The definition of CDPK-mediated signaling pathways in plants requires the resolution of the protein substrates for CDPKs. The present paper describes the purification and amino acid sequencing of three radish (Raphanus sativus) seed calmodulin antagonist proteins that are substrates for plant CDPKs.

MATERIALS AND METHODS

Materials

Radish (Raphanus sativus L.) seeds were obtained from Selected Foods Company, Waterloo, New South Wales, Aus- tralia. [T-~~PIATP (4 Ci/mmol) was obtained from Bresatec, Adelaide, Australia. A synthetic peptide substrate of smooth muscle MLCK (KKRAARATSNVFA-NH2) was obtained from Auspep, Melbourne, Australia. Calf thymus histones (type 111-S) and ATP were obtained from the Sigma Chemical Co. Affi-Gel blue (100-200 mesh) was obtained from Bio-Rad, DEAE-Sephacel from Pharmacia, and carboxymethyl-cellulose (CM-52) from Whatman. A Protein Pak SP5PW column (75 mm X 7.5 cm) was obtained from Waters, and an Aqua- pore RP300 (C,) reverse-phase HPLC guard column (4.6 mm X 30 mm; 7 pm particle size) was obtained from Brownlee Laboratories.

Protein Kinase and Calmodulin lsolation

Wheat embryo CDPK was purified as described previously (Lucantoni and Polya, 1987). MLCK was extensively purified

This work was supported by a grant from the Australian Research Council to G.M.P.

* Corresponding author; fax 61-3-479-2467.

545

Abbreviations: CDPK, Ca*+-dependent protein kinase; MLCK, myosin light chain kinase; PTH, phenylthiohydantoin; RCA, radish calmodulin antagonist.

www.plantphysiol.orgon July 18, 2020 - Published by Downloaded from Copyright © 1993 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

546 Polya et al. Plant Physiol. Vol. 1 01, 1993

from chicken gizzard smooth muscle (excised from freshly killed chickens) by a procedure involving muscle extraction and chromatography on Affi-Gel blue and DEAE-Sephacel (Walsh et al., 1983). The active MLCK was dialyzed against 10 mM K phosphate (pH 6) containing 250 mM NaCl and 1 mM DTT and then stored at -7OOC. Calmodulin was purified to homogeneity from wheat embryo as previously described (Jinsart et al., 1991).

Protein Phosphorylation

Wheat germ CDPK activity was determined radio- chemically by precipitation on phosphocellulose (P-81) disks employing radish proteins or 1.0 mg/mL of calf thymus histones (Sigma type 111-S) as substrate as previously described (Lucantoni and Polya, 1987). Phosphoamino acid analysis of phosphorylated polypeptides by high voltage electrophoresis, SDS-PAGE of phosphorylated proteins, and autoradiography were conducted as described previously (Polya et al., 1983). MLCK was routinely assayed in a reaction mixture (final volume 120 pL) containing 7 mM Hepes (Na+, pH 7), 1 mM Mg acetate, 0.1 mM CaCI2, 0.2 mg mL-' BSA, 0.2% (w/v) Tween 80, 17 p~ ATP ([y-32P]ATP, specific activity about 30 Ci mol-'), smooth muscle MLCK peptide substrate (21 KM), 0.2 PM wheat germ calmodulin, and MLCK. Reactions were terminated by spotting onto phosphocellulose (P-81) paper disks, which were then washed successively three times in 500 mL of 75 mM H3P04, dried, and Cerenkov counted. Protein was determined by the Coomassie blue method (Sedmak and Grossberg, 1977) using crystalline BSA as a standard.

Purification of Radish Seed Proteins

Unless specified otherwise, a11 operations were conducted at O to 4OC. One hundred grams of radish seeds were ho- mogenized (Omnimixer blender, full power for 1.5 min) in 500 mL of an extraction medium containing 10 mM sodium phosphate (pH 6.0), 0.5 mM PMSF, and 0.25% (v/v) ethanol. The homogenate was filtered through muslin and then cen- trifuged at 16,OOOg for 25 min. The supernatant was added to 25 g of carboxymethyl-cellulose (CM-52). The CM-52 was washed batchwise with 1 L of 10 mM sodium phosphate (pH 6.0) and then basic proteins were eluted in 200 mL of 1 M NaCl in 10 mM sodium phosphate (pH 6.0). This eluate was concentrated to 1 mL by pressure filtration (Amicon YM-10 membrane), diluted to 10 mL with 10 mM Na phosphate (pH 6.0), and reconcentrated to 5 mL. This solution was filtered (Millipore, 0.45 pm HA filter), concentrated to 1 mL by pressure filtration, and applied to a SP5PW cation-exchange column coupled to a Varian HPLC system. The SP5PW column (7.5 mm x 7.5 cm; 10-pm particle size; O.l-pm pore size) was eluted over 100 min (flow rate 1 mL min-') at 25OC with a linear gradient of O to 1 M NaCl in 10 mM Na phosphate (pH 6.0) (Fig. 1).

RCA preparation (0.3 mg) was denatured in 3 M guanidinium HCl, 100 mM 2-mercaptoethanol at 100°C for 10 min. Denatured RCA was then subjected to reverse-phase HPLC on a Cs column (Aquapore RP-300 guard column; 4.6 mm X 30 mm; 7-pm particle size) using an acetonitrile gradient in

the presence of aqueous 0.1% (v/v) TFA (a linear gradient increasing from O-25% acetonitrile in 25 min and then to 50% in a further 100 minutes; flow rate 1 mL min-').

Amino Acid Sequencing and Electrospray lonization MS

Amino acid sequencing by sequential Edman degradation was conducted with an Applied Biosystems 470A gas-phase peptide sequenator. PTH amino acid analysis employed a Waters HPLC fitted with a Zorbax Cs column (DuPont, Wilmington, DE), which was eluted with a discontinuous Na acetate buffer (pH 5.0)/acetonitrile gradient (Zimmerman et al., 1977). The RCA polypeptides (50-350 pmol) in 10 pL of methanol:H20:acetic acid (49.5:49.5:1) was introduced at a flow rate of 2 pL min-' into a VG BIO-Q electrospray spec- trometer (VG Biotech, Cheshire, UK). The quadrupole spec- trometer was scanned from m/z 600 to m/z 1400 (10 s per scan) over 5 min to obtain the final spectra, and the quadrupole was set for unit mass resolution. The mass scale was calibrated by using the multiple charged ions from the sepa- rate introduction of myoglobin.

RESULTS A N D DISCUSSION

Purification of RCA Polypeptides

Radish seeds contain a number of basic proteins that can be resolved by HPLC on a SP5PW cation-exchange column and that are phosphorylated by wheat embryo CDPK (Fig. 1). Basic radish proteins resolved in this fashion include proteins eluted at 0.17, 0.25, 0.35, and 0.82 M NaCl, respectively, from the SP5PW column (Fig. 1).

Because a calmodulin antagonist protein preparation had been previously resolved from radish seeds (Cocucci and Negrini, 1988), we examined the fractions eluting from the SP5PW column for calmodulin antagonist activity. Calmod- ulin antagonist activity was measured by determining inhibition of chicken gizzard Ca2+-ca1modulin-dependent MLCK activity, employing as a protein substrate a synthetic peptide substrate (KKRAARATSNVFA-NH2) of this enzyme (the sequence of this peptide relating to the phosphorylateable region of myosin light chains). A major zone of inhibitory activity corresponds to a protein peak eluting at 0.8 M NaCl concentration (Fig. 1). The yield of this protein from radish seeds by the protocol described in "Materials and Methods" was 26 mg kg-' fresh weight.

SDS-PAGE of the peak fractions of this RCA fraction revealed two major polypeptide bands migrating between the 14.3- and 3.5-kD standards (estimated molecular mass values of about 7 and 6 kD, respectively) (Fig. 2). This polypeptide pattern is similar to that found for a calmodulin antagonist preparation from radish (Cocucci and Negrini, 1988). In view of this heterogeneity, the RCA preparation was subjected to denaturation by heating in the presence of 3 M guanidinium HCl and 0.1 M 2-mercaptoethanol and subsequent reverse-phase HPLC on a Cs matrix, as described in "Materials and Methods." This procedure revealed consid- erable heterogeneity in the RCA preparation, some nine peaks of material absorbing at 230 nm being resolved (Fig. 3A). Calmodulin antagonist activity was associated with three major, closely migrating peaks (RCA1, RCA2, and RCAB)



Calmodulin Antagonist Phosphorylation 547

eluting at about 26% acetonitrile in 0.1% TFA. No other zoneof calmodulin antagonist activity was found in the profile.SDS-PAGE of RCA1, RCA2, and RCA3 revealed single bandswith mobilities corresponding to molecular mass values ofabout 6 kD (data not shown). The yields of RCA1, RCA2,and RCA3 from the overall isolation procedure were 0.4, 1.2,and 0.9 mg per 100 g fresh weight, respectively.

RCA1, RCA2, and RCA3 all inhibit Ca2+-calmodulin-de-pendent MLCK-catalyzed phosphorylation of a myosin lightchain-based synthetic peptide substrate (Fig. 3B). Thirty mi-crograms per milliliter of RCA2 and 11 jug/mL of RCA3inhibit MLCK by 55 and 56%, respectively. Thus, with respectto MLCK the concentrations for 50% of inhibition for RCA2and RCA3 are about 7 and 2 ^M, respectively, assuming molwts for both RCA2 and RCA3 of about 4540 (see below).

Phosphorylation of RCAs

RCA is phosphorylated by wheat embryo CDPK (Fig. 1),as are the RCA polypeptides RCA1, RCA2, and RCAS (Fig.3A). Experiments involving SDS-PAGE and autoradiographyconfirmed that phosphorylated proteins in the RCA prepa-ration comigrate with the RCA polypeptide bands (Fig. 2).Similar experiments confirmed the phosphorylation of RCA1,RCA2, and RCA3 (data not shown). The CDPK rate with thenear-saturating concentration of 0.4 mg mL"1 of RCA prep-aration (Fig. 4) is 8% of that with 1 mg mL"1 of histoneIII-S. The RCA preparation Km is 0.23 ± 0.12 mg mL"1 (asdetermined from fitting kinetic data to the Michaelis-Mentenequation by a least squares curve-fitting program), corre-sponding to 51 HM if one assumes a mol wt of these proteinsof about 4540 (see below). By way of comparison, Km valuesfor histone HI (Polya et al., 1989) and wheat basic protein(Polya et al., 1992) (two of the better endogenous wheatCDPK substrates) are 6 and 2 /UM, respectively.

Wheat embryo CDPK phosphorylates a variety of proteinand oligopeptide substrates on either Ser or Thr or on bothSer and Thr residues (Polya et al., 1989). The heterogeneousRCA preparation contains polypeptides phosphorylated bywheat embryo CDPK on both Ser (35%) and Thr (65%)residues as determined by the phosphoamino analysis de-scribed in "Materials and Methods." However, sequencing ofthe individual polypeptides RCA1, RCA2, and RCA3 indi-

kD

66.0 —45.0 —

29.0 —

20.1 —14.3 —3.4 —

1 2 3Figure 2. SDS-PACE and autoradiography of a 32P-phosphorylatedRCA preparation purified by SP5PW cation-exchange HPLC. Lane1, Position of molecular mass standards; lane 2, Coomassie blue-stained gel; lane 3, autoradiograph of lane 2 electrophoretogram.

cates that these are likely to be phosphorylated on Ser resi-dues near the C terminus, as discussed below.

Amino Acid Sequences of RCA1, RCA2, and RCA3

Amino acid sequences of RCA1, RCA2, and RCA3 weredetermined by sequential Edman degradation (Fig. 5). Thethree polypeptides have very similar N-terminal sequencesthat differ in positions 7 (Y, I), 12 (K, R), 29 (K, R), and 32(R, M). These proteins are highly homologous to the Brassicanapin small subunit variants and related proteins (Fig. 5). Aminor sequence detected in the sequencing of RCA1 anddenoted RCA1' appears to derive from deletion of the firstsix residues of RCA3 by an Arg-recognizing protease. Nonew residues (except for a PTH-Pro signal) were observed incycles 39, 40, 41 (RCA1); 39, 40, and 41 (RCA2); and 39 and40 (RCAS), consistent with termination of the polypeptidechains at residues 39. Positions 26, 34, 36, and 39 are uncer-tain in RCA1 and RCA2, and positions 26, 36, and 39 areuncertain in RCA3 (Fig. 5). Residue 26 is likely to be W for

25 30 35 40 45 50 55 60 65 70 75 80 85

Time (min)

Figure 1. Resolution of radish basic proteinsubstrates for wheat CDPK by cation-exchangeHPLC on an SP5PW column. The RCA fractionelutes at between 75 and 80 min. CDPK activityin the standard assay conditions with aliquotsof the indicated fractions as substrate (25 pi inthe range 32-51 min; 10 ̂ L elsewhere) is indi-cated by the hatched areas. The wavelengthwas changed at about 74 min to keep absorb-ance on scale. CDPK-catalyzed ["PJphosphorylincorporation of 10,000 cpm in an assay periodof 120 min corresponds to a CDPK activity of0.2 nmol min~' mL"'.




, I 1 , , I I I I I I 1.6 I 2

65 70 75 80 85 30 35 40.' 55 60

Time (min)

o

O 2 0.8- a

0.4 -

Y o -I

20

O O 30 32 34 36 38 40

Time (min)

Figure 3. Purification of RCA1, RCA2, and RCA3 by reverse-phase HPLC of the RCA preparation on a C8 column. HPLC was conducted as described in "Materials and Methods." A, Resolution of RCAs 1 , 2, and 3 (at 31-35 min) from other proteins. The CDPK activity obtained in the standard assay with 25-fiL aliquots of the indicated fraction as protein substrate is indicated by the hatched area. CDPK-catalyzed [3ZP]phosphoryl incorporation of 1000 cpm in the assay period of 80 min corresponds to a CDPK activity of 1.4 pmol min-' mL-'. 6, lnhibition of MLCK by RCA fractions is indicated by the hatched area.

reasons of homology (Fig. 5), noting that a PTH-W peak is masked by another signal in the analytical system used. Residue 34 is probably S in RCA2 (based on the trace occurrence of a PTH-S signal and of a PTH-dehydroalanine signal deriving from S in the Edman degradation). The remaining assignments as S residues (Fig. 5) are consistent with poor S detectability after 39 Edman cycles, location of S in homologous proteins (Fig. 5), the absence of CDPK-phosphorylateable Ser/Thr residues elsewhere, RCA termination after po-

I I I I5

6 7 b 10-

E below. X sition 39, and precise molecular mass data from MS, described

9 The average molecular masses of polypeptides in the RCA1, RCA2, and RCA3 preparations were determined by electrospray ionization MS as follows ( ~ s D ; average molecular masses of minor components in parentheses): RCA1, 4537.2 f 0.8 (4613.4 f 0.8); RCA2, 4542.8 f 0.4 (4619.2 f 0.6); RCA3, 4532.1 f 0.7 (4560.3 +. 0.2). The deduced RCAI, RCAZ, and RCA3 amino acid sequences (indicated in square brackets, Fig. 5) give average mo1 wts (i.e. taking C, O, N, H, and S isotope natural abundances into account) of 4537.1, 4543.1, and 4531.2, respectively, in very close agreement with the 'pectrometric data. The RCA3 preparation contains a minor component (mo1 wt 4560.3) exactly consistent with the Edman sequencing evidence of K or R at position 12. The average mo1 wt of RCA3' (the form with R12) is

o s

O 400 Kx) 200 300

P c ~ (pg.tn13

Figure 4. CDPK-catalyzed phosphorylation of the RCA preparation purif,ed by sp5pw cation-exchange HPLC, CDPK-catalyzed [32p].

phosphoryl incorporation of 10,000 cpm in an assay period of 80 min corresponds to a CDPK activity of 10 pmol min-' mL-'.



Calmodulin Antagonist Phosphorylation 549

1 5 10 15 20 25 30 35 39

(H) (RI 1 . RCAl - P A G P F R Y P(R)(C)R K E F Q Q A Q H L R A(C)Q Q(Q)L(L) K Q A(A)Q X G X G(P)(P)

P A G P F R Y P R [RCAl, MW 4537.11 C R K E F Q Q A Q H L R A C Q Q W L H K Q A R Q S G S G P

(R) (H) (P) 2. RCAl‘ - I X(T) X R K E F Q Q A Q(P)L X A X X X X X X X(H)X M X X

(H) (A) 3 . RCAZ - P A G P F R I P R (C)R R E F Q Q A Q H L R A(C)Q Q X L(L) R Q A(R)Q(S)G X G(P)(P)

P A G P F R I P R 5 [RCA 2, MW 4543.11 C R R E F Q Q A Q H L R A C Q Q W L H R Q A R Q S G S G P

(K) (H) (M) P A G P F R I P R (C)R(R)E F Q Q A Q H L R A(C)Q Q X L(L) R Q A(A)Q Q x G(P)(p) - 4. RCA3

[RCA 3, MW 4531.21 P A G P F R I P R C R K E F Q Q A Q H L R A C Q Q W L H €7 Q A Q Q G s G p 2 5. [RCA3‘, MW4559.21 P A G P F R I P R C R R E F Q Q A Q H L R A C Q Q W L H R Q A H Q Q G S G P 5

P A G P F R I P K C R K E F Q Q A Q H L R A C Q Q W L H K Q A M Q S G S G P 5 G - 6. S.alba

7. R.communis T R T N P S Q Q G C R G Q I Q E Q Q N L R Q C Q E V I K Q Q V s G Q G P R R

8 . d.thaliana 2S1 P I G P - K M R K C R K E F Q K E Q H L R A C Q Q L M L Q Q A R Q G R S D E F

9 . A.thaliana 252 P M G P - R - Q K C Q K E F Q Q S Q H L R A C Q K L M R M Q M R Q G R G G G p

10. d.thaliana 2S3 P

11. d.thaliana254 P I G P - - 1 Q K C Q K E F Q Q D Q H L R A C Q R W M R K Q M W Q G R G G G p

12. &?.napus pN1 S A G P F R L P K C R K E F Q Q A Q H L R A C Q Q W L H K Q A M Q S E G G P s 13. B.napus pNZ P A G P F R I P K C R K E F Q Q A Q H L K A C Q Q W L H K Q A M Q S G S G P s 14. &?.napus gNa P A G P F R I P K C R K E F Q Q A Q H L R A C Q Q W L H K Q A M Q P G G G S G

15. B.napuS nIa, nIb - Q P Q K C Q R E F Q Q E Q H L R A C Q Q W I R Q Q L A G s P F

16. &?.excelsa Bn-S

- P V G P - R - Q R C Q K E F Q Q S Q H L R A C Q R W M S K Q M R Q G R G G G

K

Q Q Q C R E Q M Q R Q Q M L S H C R M Y M R Q Q M E E Z - 17. E. excelsa Bn-2SS M Q K Q Q M L S H C R M Y M R Q M M K E L P Y Q T M

Figure 5. N-terminal sequences of RCA1, RCAl’, RCA2, RCA3, RCA3’, and homologous sequences of related napin small subunits from Sinapsis aba, Ricinus communis, Arabidopsis thaliana, Brassica napus, and Bertholletia excelsa. C, S, and T residues are in bold, as are residues where the assignment from Edman sequencing alone is uncertain. N-terminal and C-terminal residues of processed proteins are underlined. X, Nonallocated residues in the Edman sequencing; -, gaps introduced for alignment. RCA residues in parentheses indicate the presence of the PTH-amino acid signal but uncertain assignment. The C’’ and Cz3 residues of the RCA polypeptides were assigned from the absence of a PTH- amino acid derivative and the presence of a PTH-dehydroalanine signal (Polya et al., 1992). The RCA1’ sequence is a minor sequence detected in the sequencing of RCAl (i.e. I’ of RCA1’ was present in cycle 1 of RCAl sequencing). The RCA3‘ sequence differs from RCA3 in having an R’’. Deduced sequences of the RCAs consistent with molecular masses from ms are also given, the additional assigned amino acids corresponding to residues marked X or in parentheses in the Edman sequencing results; the average mo1 wts of the indicated deduced sequences of RCA1, RCAZ, RCA3, and RCA3’ are given in square brackets. The references to the numbered napin-related sequences are as follows: 6, Menéndez- Arias et al., 1988; 7, Sharief and Li, 1982; lrwin et al., 1990; lrwin and Lord, 1990; 8-1 1, Krebbers et al., 1988; 12, Crouch et al., 1983; Ericson et al., 1986; Josefsson et al., 1987; 13, Baszczynski and Fallis, 1990; 14, Schofield and Crouch, 1987; 15, Monsalve et al., 1991; 16, Ampe et a!., 1986; 17, De Castro et al., 1987.

4559.2. The equivalent of one Q deamidation to E per chain would yield average mo1 wts for RCA3 (Fig. 5, protein 4) and RCA3‘ (Fig. 5, protein 5) of 4532.2 and 4560.2, in exact agreement with the mass spectrometric data. The occurrence of minor species in the RCAl and RCA2 preparations that differ in mass from the major entity by 76.2 and 76.4, respectively, is consistent with the existence of two variant forms of RCAl and RCA2, both involving substitution of F for an A (mass increase 76.1). Although a minor signal of elevated PTH-Phe was not observed in cycles 2, 17, 22, and 31 from sequencing of RCAl and RCA2, the substantial but declining PTH-Phe signal due to F14 would have masked such a minor signal at cycle 17 (Fig. 5).

Because wheat CDPK is Ser/Thr-specific (Polya et al., 1989) and there are no other Ser/Thr sites, we infer that

CDPK-catalyzed phosphorylation occurs on Ser residues deduced to be present in the C-terminal region as discussed above (Fig. 5). Ser residues variously corresponding to 534, 536, or S39 are present in a number of RCA homologs, notably in protein 6 (Fig. 5), which has a very high homology to the RCA proteins (Fig. 5).

A phosphorylation site amino acid sequence motif recognized by wheat germ CDPK is Basic-X-X-Ser(Thr) as determined using synthetic peptide substrates (Polya et al., 1989). However the sequence around the site (Serlo3) on histone H1 that is phosphorylated by wheat CDPK, although not con- forming to this motif, does have functional similarities to the inferred C-terminal phosphorylation region on the RCAs and other napin-related proteins in that it is A/G- and S/T-rich (Polya et al., 1989):




97 103 Histone H1: G T G A S G S F K Deduced RCAs 1 and 2: Q A M Q S G Deduced RCAs 2 and 3': Q A M Q Q G

S S

G P S (Fig. 5) G P S (Fig. 5)

Most of the RCA-related polypeptides have G-, P-, and S- rich C-terminal domains (Fig. 5). The P residues in particular could well determine a non-a-helical, open conformation of this segment, making it more accessible to the CDPK in either the precursor protein or as the C-terminal segment of the proteolytically processed proteins studied here. The location of the CDPK phosphorylation site close to the point of precursor protein processing suggests a possible function of signal-regulated phosphorylation in this process.

The RCAs are very basic proteins that inhibit calmodulin function and are similar in these properties and in molecular size to calmodulin antagonist proteins previously partially purified from radish seeds (Cocucci and Negrini, 1988). These calmodulin-inhibiting proteins disappear during germination, and ABA inhibits both germination and the decrease in the levels of the inhibitor proteins (Cocucci and Negrini, 1988). Germination is associated with an increase in calmodulin and a decrease in calmodulin inhibitory activity (Cocucci and Negrini, 1988). The RCA polypeptides resolved and charac- terized in the present study would have contributed to the set of low mo1 wt, basic calmodulin inhibitors previously described (Cocucci and Negrini, 1988). The functional con- sequences of phosphorylation of the RCA inhibitor proteins by signal-regulated CDPK are not known. However, such phosphorylation of these napin homologs could conceivably affect small subunit precursor processing, polypeptide chain folding, interaction with other proteins such as calmodulin, and proteolysis in early germination.

The mechanism of inhibition of Ca2+-calmodulin-dependent MLCK is likely to be through RCA binding to Ca2+- calmodulin and prevention of MLCK activation. Thus, a preparation of radish calmodulin antagonist also inhibits Ca2+-calmodulin-dependent cyclic nucleotide phosphodies- terase (Cocucci and Negrini, 1988). Further, the sequences of the purified RCA proteins (Fig. 5 ) have the potential to form a-helical structures, as do other calmodulin-binding polypeptides such as melittin (Terwilliger and Eisenberg, 1982; An- derson and Malencik, 1986; Vogel and Jahnig, 1986) and polypeptides corresponding to the calmodulin-binding domains of a number of other calmodulin-interacting proteins including MLCK (O'Neil and De Grado, 1990). The RCA sequences 1-14 listed in Figure 5 have homologous regions rich in helix-forming amino acids (corresponding to C'O to Q33 of RCAI) bounded by C-terminal and N-terminal sequences rich in the helix-breaking amino acids Gly and Pro, respectively. Analysis of these potential a-helical domains by helical wheel projections (see Lukas et al., 1986; O'Neil and De Grado, 1990) reveals a surprising confinement of a11 basic (H, K, and R) residues to one "side" of the a-helices. Likewise, of the potential 24 amino acid a-helical domains of proteins 1-6 and 8-14 (Fig. 5) six to seven basic amino acids in each domain are confined to one sector constituting 200 to 240' of the projection circle and are completely absent from the remaining sector. In the putative helix of protein 7 (Fig. 5 ) three basic residues are confined to a 120' sector of the

projection circle. This arrangement is very similar to the disposition of basic residues in calmodulin-binding polypeptides deriving from the calmodulin-binding domains of smooth muscle MLCK (Lukas et al., 1986; Pearson et al., 1988) and skeletal muscle MLCK (Blumenthal et al., 1985, 1988). Thus, a helical wheel projection applied to the 20- amino acid RS20 calmodulin-binding peptide based on the calmodulin-binding domain of avian MLCK (Lukas et al., 1986) reveals seven basic amino acids confined to a 220° sector of the projection circle. The same analysis applied to the first 17 residues of M13 peptide, a calmodulin-binding peptide corresponding to the calmodulin-binding domain of skeletal muscle MLCK (Blumenthal et al., 1985, 1988), simi- larly reveals seven basic amino acids confined to a 220' sector of the projection circle.

In contrast with the results obtained with the RCA and MLCK sequences, this analysis applied to the first 21 amino acids of melittin (which have been determined to form an amphiphilic a-helix in membranes [Vogel and Jahnig, 19861) reveals 14 nonpolar residues confined to a 240' sector containing no polar residues and 5 polar residues (including 2 K residues) confined to the remaining 120° sector. In comparison, the potential a-helical domains of the RCAs and related plant proteins (Fig. 5 ) are much more similar to the amphiphilic a-helices of the RS20 and M13 MLCK-derived calmodulin-binding peptides, in which a11 of the basic residues (and not simply a11 of the polar residues as in melittin) are confined to one sector of the helical projection wheel.

The retention of RCA calmodulin antagonist activity after the heat treatment and chromatography in CHKN involved in isolation (Fig. 3B) could suggest nonspecific interaction of denatured basic RCA polypeptides with acidic calmodulin. However, in this connection, the MLCK-derived RS20 peptide that exhibits similarities to the RCA proteins in terms of amphiphilic helix determinants exists as a random coil in aqueous solution but has about 60% a-helix secondary structure in the presence of Ca2+-calmodulin or in 70% trifluoro- ethanol (Lukas et al., 1986). The low RCA concentrations required for MLCK inhibition and the conservation in the RCAs and related proteins (Fig. 5) of amphiphilic a-helix- related structural features found in other calmodulin-binding proteins (O'Neil and De Grado, 1990) suggest physiological calmodulin-antagonist functions for these proteins. Such functions could include control of seed germination, as pro- posed by Cocucci and Negrini (1988).

ACKNOWLEDCMENTS

We are grateful to Ian Thomas and Dr. P.B. H6j for electrospray MS and interpretative advice.

Received September 8, 1992; accepted October 21, 1992. Copyright Clearance Center: 0032-0889/93/lOl/0545/07

LITERATURE ClTED

Ampe C, Van Damme J, De Castro LAB, Sampaio MJAM, Van Montagu M, Vanderckhove J (1986) The amino-acid sequence of the 25 sulphur-rich proteins from seeds of Brazil nut (Bertholletia excelsa H.B.K.). Eur J Biochem 159 597-604

Anderson SR, Malencik DA (1986) Peptides recognizing calmodu-



Calmodulin Antagonist Phosphorylation 55 1

lin. 1n Cheung WY, ed, Calcium and Cell Function, Vol VI. Academic Press, New York, pp 1-42

Baszczynski CL, Fallis L (1990) Isolation and nucleotide sequence of a genomic clone encoding a new Brassica napus napin gene. Plant Mo1 Biol 14: 633-635

Blumenthal DK, Charbonneau H, Edelman AM, Hinds TR, Ro- senberg GB, Storm DR, Vincenzi FF, Beavo JA, Krebs EG (1988) Synthetic peptides based on the calmodulin-binding domain of myelin light chain kinase inhibit activation of other calmodulin- dependent enzymes. Biochem Biophys Res Commun 1 5 6 860-865

Blumenthal DK, Takio K, Edelman AM, Charbonneau H, Titani K, Walsh KA, Krebs EG (1985) ldentification of the calmodulin- binding domain of skeletal muscle myosin light chain kinase. Proc Natl Acad Sci USA 82: 3187-3191

Cocucci M, Negrini N (1988) Changes in the levels of calmodulin and of a calmodulin inhibitor in the early phases of radish (Ra- phanus sativus L.) seed germination. Plant Physiol88 910-914

Crouch ML, Tenbarge KM, Simon AE, Ferl R (1983) cDNA clones for Brassica riapus seed storage proteins: evidence from nucleotide sequence analysis that both subunits of napin are cleaved from a precursor polypeptide. J Mo1 Appl Genet 2: 273-283

De Castro LAB, Lacerado L, Aramayo RA, Sampaio MJAM, Gan- der ES (1987) Evidence for a precursor molecule of Brazil nut 2 s seed proteins from biosynthesis and cDNA analysis. Mo1 Gen Genet 206: 338-343

Echevarria C, Vida1 J, Le Marechal P, Brulfert J, Ranjeva R, Gadal P (1988) The phosphorylation of sorghum leaf phosphoenolpyruvate carboxylase is a Ca’+-calmodulin dependent process. Biochem Biophys Res Commun 1 5 5 835-840

Ericson ML, Rodin J, Lenman M, Glimelius K, Josefsson L-G, Rask L (1986) Structure of the rapeseed 1.7s storage protein napin, and its precursor. J Biol Chem 261: 14576-14581

Hepler PK, Wayne RO (1985) Calcium and plant development. Annu Rev Plant Physiol36 397-439

Irwin SD, Keen JN, Findlay JBC, Lord JM (1990) The Ricinus communis 2 s albumin precursor: a single preprotein may be processed into two different heterodimeric storage proteins. Mo1 Gen Genet 222: 400-408

Irwin SD, Lord JM (1990) Nucleotide sequence of a Ricinus communis 25 albumin precursor gene. Nucleic Acids Res 1 8 5890

Jinsart W, Ternai B, Polya GM (1991) Inhibition and activation of wheat embryo calcium-dependent protein kinase and inhibition of avian myosin light chain kinase by long chain aliphatic amphi- philes. Plant Sci 78: 165-175

Josefsson L-G, Lenman M, Ericson ML, Rask L (1987) Structure of a gene encoding the 1.7s storage protein, napin, from Brassica napus. J Biol Chem 262: 12196-12201

Krebbers E, Herdies L, De Clerco A, Seurinck J, Leemans J, Van Damme J, Segura M, Gheysen G, Van Montagu M, Vandek- erckhove J (1988) Determination of the processing sites of an Arabidopsis 2s albumin and characterization of the complete gene family. Plant Physiol87: 859-866

Lucantoni A, Polya GM (1987) Activation of wheat embryo Ca2+- regulated protein kinase by unsaturated fatty acids in the presence and absence of calcium. FEBS Lett 221: 33-36

Lukas TJ, Burgess WH, Prendergast FG, Lau W, Watterson DM

(1986) Calmodulin binding domains: characterization of a phosphorylation and calmodulin binding site from myosin light chain kinase. Biochemistry 2 5 1458-1464

Menendez-Arias L, Moneo I, Dominguez J, Rodriguez R (1988) Primary structure of the major allergen of yellow mustard (Sinapsis alba L.) seed, SinaI. Eur J Biochem 177: 159-166

Monsalve RI, Lopez-0th C, Villalba M, Rodriguez R (1991) A new distinct group of 2 s albumins from rapeseed. Amino acid sequence of two low molecular weight napins. FEBS Lett 295:

ONei l KT, De Grado WF (1990) How calmodulin binds its targets: sequence independent recognition of amphiphilic a-helices. Trends Biochem Sci 15: 59-64

Pearson RB, Wettenhall REH, Means AR, Hartshorne DJ, Kemp BE (1988) Autoregulation of enzymes by pseudosubstrate proto- topes: myosin light chain kinase. Science 241: 970-973

Polya GM, Chandra S (1990) Ca’+-dependent protein phosphorylation in plants: regulation, protein substrate specificity and product dephosphorylation. Curr Top Plant Biochem Physiol9 164-180

Polya GM, Chandra S, Chung R, Neumann GM, Hbj PB (1992) Purification and characterization of wheat and pine small basic protein substrates for plant calcium-dependent protein kinase. Biochim Biophys Acta 1120: 273-280

Polya GM, Davies JR, Micucci V (1983) Properties of a calmodulin- activated Ca’+-dependent protein kinase from wheat germ. Biochim Biophys Acta 761: 1-12

Polya GM, Morrice NA, Wettenhall REH (1989) Substrate specificity of wheat embryo calcium-dependent protein kinase. FEBS Lett

Ranjeva R, Boudet AM (1987) Phosphorylation of proteins in plants: regulating effects and potential involvement in stimulus/response coupling. Annu Rev Plant Physiol38 73-93

Roberts DM, Harmon AC (1992) Calcium-modulated proteins: targets of intracellular calcium signals in higher plants. Annu Rev Plant Physiol Mo1 Biol 43: 375-414

Schaller GE, Sussman MR (1988) Phosphorylation of the plasma membrane H+-ATPase of oat roots by a calcium-stimulated protein kinase. Planta 173: 509-518

Schofield SR, Crouch ML (1987) Nucleotide sequence of a member of the napin storage protein family from Brassica napus. J Biol Chem 262: 12202-12208

Sedmak JJ, Grossberg SE (1977) A rapid, sensitive and versatile assay for protein using Coomassie brilliant blue G250. Anal Biochem 791: 544-552

Sharief FS, Li SS-L (1982) Amino acid sequence of small and large subunits of seed storage protein from Ricinus communis. J Biol Chem 257: 14753-14759

Terwilliger TC, Eisenberg D (1982) The structure of melittin 11. Interpretation of the structure. J Biol Chem 257: 6016-6022

Vogel H, Jahnig F (1986) The structure of melittin in membranes. Biophys J 5 0 573-582

Walsh MP, Hinkins S, Dabroniska R, Hartshorne DJ (1983) Smooth muscle myosin light chain kinase. Methods Enzymol 9 9

Zimmerman CL, Appella E, Pisano JJ (1977) Rapid analysis of amino acid phenylthiohydantoins by high performance liquid chromatography. Anal Biochem 77: 569-573

207-210

253 137-140

279-288



Purification and Sequencing of Radish Seed Calmodulin ... · Wheat embryo CDPK phosphorylates a...

Documents

Transcript of Purification and Sequencing of Radish Seed Calmodulin ... · Wheat embryo CDPK phosphorylates a...