A Pollen-specific Novel Calmodulin-binding Protein with Tetratricopeptide Repeats*

Calcium is essential for pollen germination and pollen tube growth. A large body of information has established a link between elevation of cytosolic Ca2+ at the pollen tube tip and its growth. Since the action of Ca2+ is primarily mediated by Ca2+-binding proteins such as calmodulin (CaM), identification of CaM-binding proteins in pollen should provide insights into the mechanisms by which Ca2+regulates pollen germination and tube growth. In this study, a CaM-binding protein from maize pollen (maizepollen calmodulin-bindingprotein, MPCBP) was isolated in a protein-protein interaction-based screening using 35S-labeled CaM as a probe. MPCBP has a molecular mass of about 72 kDa and contains three tetratricopeptide repeats (TPR) suggesting that it is a member of the TPR family of proteins. MPCBP protein shares a high sequence identity with two hypothetical TPR-containing proteins fromArabidopsis. Using gel overlay assays and CaM-Sepharose binding, we show that the bacterially expressed MPCBP binds to bovine CaM and three CaM isoforms from Arabidopsis in a Ca2+-dependent manner. To map the CaM-binding domain several truncated versions of the MPCBP were expressed in bacteria and tested for their ability to bind CaM. Based on these studies, the CaM-binding domain was mapped to an 18-amino acid stretch between the first and second TPR regions. Gel and fluorescence shift assays performed with CaM and a CaM-binding synthetic peptide further confirmed MPCBP binding to CaM. Western, Northern, and reverse transcriptase-polymerase chain reaction analysis have shown that MPCBP expression is specific to pollen. MPCBP was detected in both soluble and microsomal proteins. Immunoblots showed the presence of MPCBP in mature and germinating pollen. Pollen-specific expression of MPCBP, its CaM-binding properties, and the presence of TPR motifs suggest a role for this protein in Ca2+-regulated events during pollen germination and growth.

germination apertures of hydrated pollen (7). Studies with 45 Ca 2ϩ (3,8) and later direct measurement of cytosolic Ca 2ϩ concentration ([Ca 2ϩ ] cyt ) 1 using fluorescence ratio imaging of Ca 2ϩ indicator dyes established that, in pollen tubes, [Ca 2ϩ ] cyt accumulates at the growing tip and forms a steep-tip focused gradient with about 3.0 M at the tip to about 0.2 M within 20 or 65 m from the tip (6, 9 -16). Pollen tube elongation is inhibited by chemicals (e.g. ion channel blockers or ionophores) that interfere with Ca 2ϩ homeostasis in the growing pollen tubes (3,11,12,14,(17)(18)(19). Disrupting the Ca 2ϩ gradient or blocking Ca 2ϩ influx inhibited pollen tube growth, dissipated the cytoplasmic streaming associated with the pollen tube elongation and eliminated the clear zone found at the tip of actively elongating pollen (7,12,20). Extracellular Ca 2ϩ influx at the pollen tube tip was reported and this influx was correlated with the intracellular Ca 2ϩ gradient and polarized growth of the tip (12,(21)(22)(23). Ca 2ϩ influx at the apex is speculated to be due to stretch-activated Ca 2ϩ channels at the tip (12,22), while Ca 2ϩ -ATPase pumps are implicated in maintaining the Ca 2ϩ gradients in the cytosol (7,17). Recently, the relationship between [Ca 2ϩ ] cyt and pollen tube growth was further strengthened by the discovery that the tip-focused [Ca 2ϩ ] cyt gradients oscillate with tip-high [Ca 2ϩ ] cyt corresponding to peaks in pollen tube growth (24 -27). Furthermore, Ca 2ϩ has been shown to influence the direction of pollen tube growth. Induced Ca 2ϩ fluctuation in the cytosol or modifying extracellular Ca 2ϩ influx redirected pollen tube growth toward the high Ca 2ϩ concentration (13,27,28), suggesting a role for pistil-derived Ca 2ϩ in the directed growth of pollen tubes toward the ovary. These studies showed that the establishment and maintenance of a precise tip-focused intracellular Ca 2ϩ gradient, possibly through regulating Ca 2ϩ influxes at the tip, is essential for pollen tube elongation and directional growth. How the tip-focused Ca 2ϩ gradient at the tip regulates pollen tube growth and direction is poorly understood. Some studies suggest that Ca 2ϩ interacts directly or indirectly, through Ca 2ϩ -binding proteins, with the cytoskeleton to regulate cytoplasmic streaming, vesicle fusion, and the function of cytoskeletal elements required for tube emergence and growth (6, 12, 18, 19, 28 -35).
Calcium has been implicated in regulating diverse physiological processes in plants (36 -40). Calcium-regulated physiological responses are often mediated directly or indirectly by Ca 2ϩ -modulated proteins of which CaM is ubiquitous in all eukaryotes. Calmodulin, a Ca 2ϩ -modulated protein with four Ca 2ϩ -binding EF-hands, is considered to be the primary intra-cellular Ca 2ϩ receptor in all eukaryotes. In plants, over the last 10 years CaM and CaM isoforms have been identified, and their involvement in transducing Ca 2ϩ signals into a variety of cellular responses has been reported (41)(42)(43). Since CaM acts by modulating the activity of a variety of other proteins directly by interacting with them, research in recent years has focused on identifying the CaM-target proteins and analyzing the function of these proteins in cellular processes. Whereas a wide variety of CaM-activated proteins were described and characterized in animals only a few CaM-binding proteins have been identified in plants and their function in regulating plant growth and development in response to elevated Ca 2ϩ signals is still in its infancy (42)(43)(44). Furthermore, very little is known about CaM-target proteins in pollen (37).
Considering the widespread association between CaM and Ca 2ϩ -sensitive cellular processes, it is reasonable to expect CaM to mediate Ca 2ϩ action in pollen tubes. Some reports suggest the involvement of CaM in mediating Ca 2ϩ effect on pollen tube growth (37). Exogenous CaM enhances pollen germination and pollen tube growth (30,45), whereas CaM antagonists and anti-CaM serum inhibit pollen germination and tube growth (18,45,46) and stop cytoplasmic streaming (17) in a concentration-dependent manner. In addition, upon CaM antagonist treatment of pollen tubes, Ca 2ϩ remains in the tip membranes (45) or its level increases behind the tip (17), suggesting that CaM is involved in maintaining Ca 2ϩ gradients in the pollen tubes through Ca 2ϩ influx from the plasma membrane channels and/or sequestration of Ca 2ϩ into the internal organelles by CaM-regulated Ca 2ϩ -ATPases (34,47). Ca 2ϩ -ATPase activity has been detected in the plasma membrane of the pollen tube tip as well as in the endoplasmic reticulum and mitochondria behind the tip and these ATPases are stimulated by CaM (7,43,48). Calmodulin localization in the pollen during various stages of pollen growth has produced contradicting results. Calmodulin is localized to the region of germinal apertures of the hydrated pollen, the plasma membrane, and the cytoplasm in the vicinity of the germination bubble and in the plasma membrane and the cytosol of the growing pollen tube where it forms an apically focused gradient similar to the tip-focused Ca 2ϩ gradient (7,49). A similar localization of CaM is observed in the Fucus rhizoid tip pre-emergence location (37), suggesting a role for CaM in polar growth and tip extension. In other localization studies, however, CaM was found to be diffuse and uniformly distributed in the pollen tube (50,51) and no tip high gradient of the protein was observed (51). Recent studies on effects of exogenous CaM on pollen tube germination and growth concluded that CaM acts extracellularly exerting its effect on pollen possibly through a signal transduction pathway involving a receptor-mediated stimulation of a G protein (30,46). Although these studies implicate the involvement of CaM in pollen germination and tube growth, little is known about proteins that bind to CaM in pollen. Hence studies on pollen-specific CaM-binding proteins should help us understand the role of CaM in Ca 2ϩ -mediated signal transduction pathways in pollen. Here we report the isolation of a gene encoding a maize pollen calmodulin-binding protein (MPCBP) using a protein-protein interaction-based screening of an expression library with radiolabeled CaM. The MPCBP is expressed specifically in mature and germinating pollen and binds CaM in a Ca 2ϩ -dependent manner. The MPCBP contains three tetratricopeptide repeats (TPRs) that are known to function in protein-protein interaction. The region of the protein that binds to CaM is mapped to an 18-amino acid stretch between the TPR1 and TPR2. Pollen-specific expression of MPCBP and its CaM-binding ability suggest a role for this protein in pollen germination and tube growth.

EXPERIMENTAL PROCEDURES
Materials-Maize (Zea mays L.) inbred lines KYS and A632 seeds were germinated on moist filter paper, and the tissues (roots, hypocotyls, and leaves) were collected after 10 days of germination. Mature pollen was collected from tassels of field or greenhouse-grown maize. Freshly collected maize pollen was germinated on a medium containing 12% sucrose, 300 mg/liter CaCl 2 , 100 mg/liter boric acid, and 0.7% agarose. Triton X-100-free nitrocellulose filter discs were obtained from Millipore. Easy tag 35 S-isotope labeling mixture was obtained from PerkinElmer Life Sciences. Exassist helper phage and Escherichia coli SOLR cells were obtained from Stratagene. Nitro blue tetrazolium, 5-bromo-4-chloro-3-indolyl phosphate, IPTG, and Trizol were obtained from Life Technologies, Inc. pET vectors and E. coli strain BL21(DE3) were purchased from Novagen. Gelatin and diaminobenzidine were obtained from Sigma. Biotinylated CaM and bovine CaM were from Calbiochem. Complete protease inhibitor mixture was from Roche Molecular Biochemicals. Phenyl-Sepharose CL-4B, bovine CaM Sepharose-4B, and CNBr-activated Sepharose-4B were obtained from Amersham Pharmacia Biotech. Vanadyl ribonucleoside RNase inhibitor was from BioLabs. Expression and purification of the carboxyl-terminal region of KCBP (1.4C) was described in Reddy et al. (52). All other chemicals were of reagent grade.
Expression and Purification of Recombinant Arabidopsis thaliana CaM Isoforms-pET expression vectors containing CaM-2, -4, or -6 isoforms were kindly provided by Dr. Raymond Zielinski (53). The expected molecular weights for AtCaM2, -4, and -6 are 16,808, 16,824, and 16,822, respectively. The AtCaM isoforms were induced and purified as described earlier with some modifications (53). The E. coli BL21(DE3) cells containing the recombinant pET CaM expression clone were grown to A 600 of 0.6 and induced by 1 mM IPTG for 3 h at 37°C in 1 liter of NZY medium containing 50 g/ml ampicillin as described by Fromm and Chua (54). All the following steps were performed at 4°C. The cells were harvested, washed in buffer A (50 mM Tris-HCl, pH 7.5), and resuspended in extraction buffer (Buffer A with 2 mM EDTA, 1 mM DTT, 200 g/ml lysozyme, and complete protease inhibitor mixture). After treatment with DNase to remove DNA, the cell extract was clarified by centrifugation and the supernatant fraction was precipitated with 55% ammonium sulfate. The proteins in the supernatant were precipitated with 50% H 2 SO 4 (pH 4) for 30 min with stirring. After centrifugation, the pellet was resuspended in buffer A containing 1 mM DTT, dialyzed first in distilled water, and then in buffer A containing 100 mM NaCl, 0.5 mM EGTA, and 1 mM DTT. After adjusting CaCl 2 concentration to 5 mM, the protein was loaded onto a phenyl-Sepharose column CL-4B (10 ml bed volume) pre-equilibrated with buffer B (buffer A containing 0.1 mM CaCl 2 and 0.5 mM DTT). The column was washed with buffer B containing 5 mM NaCl and the AtCaM protein was eluted with elution buffer containing 50 mM Tris-HCl, pH 7.5, 1 mM EGTA, and 0.5 mM DTT. The eluates containing the proteins were dialyzed in water.
Production and Purification of 35 S-labeled AtCaM Isoforms-The 35 S-AtCaM isoforms were prepared and purified as described (54) with slight modifications. Initially, the cells were grown in M9 medium with 10 g/liter tryptone and 50 g/ml ampicillin overnight and then the cells were concentrated in 10 ml of M9 medium with ampicillin. One mM IPTG was added to the cells (0.6 A 600 ), 2 mCi of Easy tag 35 S-labeling mixture was added after 15 min, and the cultures were grown for 3 h at 37°C. The cells were pelleted, resuspended in 1.5 ml of buffer A (see above), lysed with lysozyme (0.2 mg/ml), and DNase-treated (50 units) in the presence of 3 mM MgCl 2 . After centrifuging the lysate at 30,000 rpm for 30 min, the supernatant was heated for 3 min at 90°C, centrifuged, and the resulting supernatant was used to purify the radiolabeled CaM on a 1-ml phenyl-Sepharose CL-4B column as described above.
Screening of Maize Pollen Expression Library with 35 S-CaM-A cDNA library from maize pollen constructed in the EcoRI site of Zap II vector was used for screening. About 120,000 recombinants were screened with a mixture of 35 S-labeled AtCaM isoforms 4 and 6. Approximately 9000 pfu per 15-cm plate were plated on NZCYM plates using E. coli XL1-blue MRA (Stratagene) as the host strain. The plates were incubated at 42°C until the plaques appeared, at which time the plates were overlaid with nitrocellulose filters that were previously soaked in 10 mM IPTG. Plaques were allowed to resume growth overnight at 37°C. The plates were then placed at 4°C for 1 h. The nitrocellulose filters were removed and washed briefly in TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM CaCl 2 ). The filters were blocked in TBS containing 1% nonfat milk for 15 min with gentle shaking at room temperature and then incubated in a mixture of 35 S-labeled At-CaM isoforms at 5 g/ml for 12 h. The membranes were rinsed in TBS/Ca 2ϩ (TBS containing 5 mM CaCl 2 ) three times for 5 min each and dried between 2 sheets of 3MM paper for 24 h before exposing to a x-ray film. The putative positive plaques were plaque-purified by two additional rounds of screening. The cDNA inserts from Zap II were excised in vivo in a plasmid form using the Exassist helper phage and by infecting E. coli SOLR cells with phage recombinant. The insert was excised by digesting the plasmid DNA with EcoRI. The positive clones were confirmed for Ca 2ϩ dependence in binding to CaM by expressing the purified clones as above, incubating the membrane in 35 S-CaM probe, and washing the membrane in TBS buffer containing either 2 mM Ca 2ϩ or 5 mM EGTA, a Ca 2ϩ chelator.
Isolation of the Genomic Clone-Maize genomic library in EMBL3 (CLONTECH) was screened with a 32 P-labeled cDNA (1.2 kb) according to the standard (55). About 650,000 recombinants were screened and about 20 positives were isolated after two rounds of screening. Restriction mapping and Southern analysis of the positive clones indicated that all of the isolated clones were derived from the same gene. The clone with the largest hybridizing fragment was sequenced by primer walking.
DNA Sequencing and Analysis-Both strands of cDNAs were sequenced by dideoxynucleotide chain termination using double stranded DNA. Sequence analysis was performed using MacVector and Sequencher programs. BLAST searches were performed at the National Center for Biotechnology Information web site.
Construction and Expression of Truncated cDNAs in E. coli-The 1.2-kb cDNA insert (named P3) containing the coding region for 242 amino acids in the carboxyl-terminal region was cloned in-frame into pET 28b expression vector. For mapping the CaM-binding domain, several truncated constructs were prepared in pET vector. The P3 cDNA in pET 28 vector was digested with either NcoI or BamHI to release the 168-or 717-bp fragment corresponding to the 5Ј end of the cDNA. The NcoI and BamHI fragments were cloned into pET 28b vector digested with respective enzymes to generate P3/NcoI and P3/BamHI clones. The P3 clones lacking either NcoI or BamHI fragments were religated to generate P⌬NcoI and P⌬BamHI expression clones. A 531-bp fragment representing the BamHI fragment without the NcoI fragment was also cloned in pET 28b vector. All the clones were introduced into E. coli BL21(DE3) and expressed. Expression of the protein was achieved by growing the bacterial cells containing the subclones at 37°C to an A 600 of 0.6, IPTG was then added to the cultures to a final concentration of 1 mM and the cells were allowed to grow for 3 more hours at 30°C. The cells from induced and uninduced cultures were collected, resuspended in lysis buffer (50 mM Tris-HCl, pH 8.0, 2 mM EDTA, 100 mM lysozyme, 0.1% Triton X-100), vortexed, and incubated on ice for 1 h. The mixture was then sonicated 3 times, 10 s each, using a Virsonic digital 475 ultrasonic cell disruptor (Virtis, NY). The lysate was then centifuged at 12,000 ϫ g for 20 min to separate the soluble (supernatant) from the insoluble (pellet) fractions. The pellet was dissolved in either 6 M urea containing buffer or sample buffer, and the supernatant and the pellet were electrophoresed on 12% SDS-polyacrylamide gels. The gels were blotted onto a nitrocellulose membrane using a Bio-Rad transfer cell. Expressed proteins were detected by T7 tag monoclonal antibody conjugated to alkaline phosphatase (T7-AP) (Novagen). Briefly, the blots were blocked for 2 h in TBST buffer (10 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% Tween-20) containing 3% gelatin. The blots were then incubated for 30 min in T7-AP (1:10,000 dilution) in TBST containing 1% gelatin, washed three times with TBST, rinsed briefly in AP buffer (50 mM Tris-HCl, pH 9.5, 100 mM NaCl, 1 mM MgCl 2 ), and then developed in AP buffer containing 0.3 mg/ml nitro blue tetrazolium and 0.15 mg/ml 5-bromo-4-chloro-3-indolyl phosphate.
Calmodulin Binding to Fusion Proteins-Soluble and insoluble proteins from the induced and uninduced cultures were separated on 12% SDS-polyacrylamide gels and blotted as described above. To detect the binding of the expressed proteins to 35 S-labeled AtCaM, duplicate blots were blocked for 15 min in TBS (50 mM Tris-HCl, pH 7.5, 150 mM NaCl) containing 5 mM of either CaCl 2 or EGTA and 1% nonfat dry milk. The blots were then incubated for 12 h in a mixture of 35 S-labeled AtCaM isoforms at 5 g/ml in the same buffers as above. The filters were then washed with the corresponding buffer and dried before exposing to x-ray film. Binding to biotinylated CaM was detected by blocking the blots in 3% gelatin in TBS/Ca/Mg or TBS/EGTA/Mg (50 mM Tris-HCl, pH 7.5, 50 mM MgCl 2 , 150 mM NaCl, and 5 mM of either CaCl 2 or EGTA) at 30°C. The blots were then incubated in 5 g/ml biotinylated CaM for 2 h in TBS/Ca/Mg or TBS/EGTA/Mg containing 0.1% Tween 20 and 1% gelatin followed by three washes of 10 min each in the corresponding buffers above. The blots were then incubated in Vectastain ABC.HRP in TBS/Ca/Mg or TBS/EGTA/Mg for 30 min, washed 3 times, 10 min each, with the above buffer and the CaM-binding proteins were detected colorimetrically in a substrate solution (0.8 mg/ml diaminobenzidine, 0.4 mg/ml NiCl 2 , and 0.009% H 2 O 2 in 100 mM Tris-HCl, pH 7.5).

Preparation of Horseradish Peroxidase (HRP)-labeled CaM and Blot
Overlay Assay-HRP was conjugated to CaM and used in blot overlay assays according to the procedure described by Lee et al. (56). Briefly, the Arabidopsis CaM2 was incubated in 25 mM Tris-HCl, pH 7, 2 mM EGTA, and 0.1 M DTT at 55°C for 1.5 h and the reduced CaM was extensively dialyzed against phosphate-buffered saline at 4°C. Maleimide-activated HRP (Pierce) and reduced CaM were incubated at a molar ratio of 1:1 in phosphate-buffered saline at room temperature for 1.5 h. The resulting HRP⅐CaM complex was verified on a denaturing gel and used in blot-overlay assays. Protein blots were prepared as described above, rinsed in TBST (TBS containing 0.1% (v/v) Tween 20), and blocked by incubating in TBST plus 7% (w/v) non-fat dry milk overnight. The blocked membranes were washed three times, 5 min each, with TBST. After equilibration in the overlay buffer (50 mM imidazole-HCl, pH 7.5, and 150 mM NaCl) for 1 h, the membranes were incubated for 1 h in overlay buffer containing 0.1% gelatin and 1 g/ml HRP-CaM. The HRP-CaM overlay blots were washed sequentially in three buffers. Wash with each buffer was performed five times, 5 min each. The composition of the buffers was: 1) TBST, 50 mM imidazole-HCl, pH 7.5, 1 mM CaCl 2 ; 2) 20 mM Tris-HCl, pH 7.5, 0.5% Tween 20, 50 mM imidazole-HCl, pH 7.5, 0.5 M KCl, 1 mM CaCl 2 ; 3) 20 mM Tris-HCl, pH 7.5, 0.1% Tween 20, 0.5 M KCl, 1 mM MgCl 2 . The proteins that bound to HRP-CaM were detected by immersing the blots in a substrate solution as above.
Slot Blot Analysis-The interaction between synthetic peptides and AtCaM isoforms was analyzed by applying of synthetic peptides to a nitrocellulose membrane in a slot blot apparatus, and incubating the membrane with 35 S-labeled AtCaM isoforms individually using the method described above (see "Calmodulin Binding to Fusion Proteins").
Calmodulin Shift Assays-The interaction of CaM with the synthetic peptide was also analyzed using electrophoretic mobility shift of CaM in the presence of a synthetic peptide (57). Each of the AtCaM isoforms and the bovine CaM (166 pmol) was incubated with increasing concentrations of the MPCBP-synthetic peptide (166, 332, and 664 pmol) in the presence of 4 M urea, in a buffer containing 100 mM Tris-HCl, pH 7.5, and 1 mM CaCl 2 or 5 mM EGTA at room temperature for 1 h in a total volume of 20 l. Then, 10 l of sample buffer (0.375 M Tris-HCl, pH 6.8, 30% glycerol, and 0.023% bromphenol blue) was added to the samples and the mixture was electrophoresed in 12% polyacrylamide gels containing 4 M urea or 7.5% glycerol, 0.375 M Tris-HCl, pH 8.8, and either 1 mM CaCl 2 or 5 mM EGTA. The gels were run at a constant voltage of 25 V per gel in an electrode buffer (25 mM Tris-HCl, pH 8.3, 192 mM glycine, and either 1 mM CaCl 2 or 5 mM EGTA). The gels were stained with 0.25% Coomassie Blue R-250 in 7.5% acetic acid and 50% methanol for 1 h and then destained with 30% methanol and 7% acetic acid.
Fluorescence Spectroscopy Assay-The tryptophan fluorescence spectra of free and CaM-bound synthetic peptide were recorded with a Hitachi-F-3010/4010 spectrofluorimeter as described in Reddy et al. (52). MPCBP peptide and AtCaMs (166 pmol each) were mixed in a 600-l reaction volume. The Trp (W) residue in free and CaM-bound synthetic MPCBP peptide was excited at 290 nm and the emission wavelength values were recorded from 300 to 400 nm with a bandwidth of 5 nm in a 5-mm quartz cell at 25°C. Samples were incubated for 1 h at 25°C prior to spectroscopic measurements. Corrections were made for the protein and solvent blanks.
Binding of Fusion Proteins to CaM-Sepharose Columns-Bovine CaM Sepharose-4B was obtained from Amersham Pharmacia Biotech. AtCaM isoforms were conjugated to Sepharose-4B. Briefly 1 g of CNBractivated Sepharose-4B was rehydrated in 1 mM HCl and then washed with 1 mM HCl on sintered glass for 15 min to remove additives. Ten mg of each of the CaM isoforms 2, 4, and 6 were dialyzed in the coupling buffer (0.1 M NaHCO 3 , pH 8.3, 0.5 M NaCl) and incubated with the washed Sepharose in 5 ml of coupling buffer. The mixture was rotated at room temperature for 2 h in a 15-ml tube after which it was centrifuged at low speed (1000 ϫ g) and washed 3 times with 5 gel volumes of coupling buffer to remove unconjugated ligand. After centrifugation, the washing solution was replaced with blocking solution (0.1 M Tris-HCl, pH 8.0) and incubated for 2 h at room temperature to block remaining active groups. The beads were then washed three times, each time with 0.1 M acetate buffer containing 0.5 M NaCl followed by 0.1 M Tris-HCl, pH 8.0, containing 0.5 M NaCl. The solution was then replaced by CaM-Sepharose binding buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 5 mM CaCl 2 ) and the slurry was degassed and packed into a column. Insoluble fusion protein from P3 clone was dissolved in binding buffer containing 6 M urea. To purify these proteins on CaM-Sepharose column, the column was equilibrated in the binding buffer with 6 M urea and the dissolved proteins were loaded. The unbound protein was washed with the same buffer, and the bound protein was eluted with binding buffer containing 6 M urea except that CaCl 2 was replaced with 7 mM EGTA. The fractions were analyzed on 12% SDSpolyacrylamide gels and detected with T7-AP, biotinylated CaM, or 35 S-labeled AtCaM in the presence or absence of CaCl 2 . Northern Blot Analysis-Total RNA was extracted from maize pollen and kernels as described earlier (58). Pollen was ground in liquid nitrogen and sand, then homogenized in 3 volumes of lysis buffer (50 mM EGTA, 100 mM NaCl, 1% SDS, 100 mM Tris-HCl, pH 7.6, 50 mM ␤-mercaptoethanol containing 10 mM vanadyl ribonucleoside. The homogenate was extracted several times with an equal volume of phenol/ chloroform and centrifuged for 10 min at 10,000 rpm until there was no detectable interphase. RNA was precipitated with sodium acetate and ethanol, dissolved in diethyl pyrocarbonate-treated water, and reprecipitated twice with equal volume of 5 M LiCl on ice for 1 h. RNA from other maize tissues was extracted using the Trizol method according to manufacturer's instructions. RNA was loaded on denaturing 1% agarose containing formaldehyde and blotted and probed according to standard procedures (55).
Southern Blot Analysis-Maize genomic DNA was digested with different restriction enzymes, electrophoresed in 0.8% agarose gel, and transferred onto a Hybond nylon membrane. The DNA was fixed to the membrane by UV cross-linking. The blot was hybridized to the radiolabeled 1.2-kb cDNA at 65°C and washed under high stringency conditions (55).
Antibody Production-About 150 g of fusion protein induced from the 1.2-kb cDNA (P3) clone and purified on CaM-Sepharose column was electrophoresed on a preparative SDS-polyacrylamide mini-gel. The gel was washed 3 times with distilled water, stained in water-based 0.05% Coomassie R-250 for 30 min, and rinsed in several changes of water for 1 h. The protein band (32 kDa) was cut, rinsed briefly with 50 ml of water, and then overnight with 5 ml of buffer (50 mM Tris-HCl, pH 7.5, at 4°C). The gel pieces were solubilized, mixed with Freud's incomplete adjuvent (1:1 ratio), and injected intradermally at multiple spots into New Zealand White rabbits. Booster injections were performed on 14, 28, 42, and 56 days after initial injection. Bleeds were performed before the initial injection and on the day of each booster injection. The terminal bleed was collected on day 140. Serum from day 126-bleed was used for antibody purification and immunodetection studies.
Anti-MPCBP antibodies were affinity purified using a modified method of the Millipore technical protocol TP015 (Millipore Corp., Bedford, MA). About 1 mg of the purified MPCBP protein was electrophoresed on an SDS-polyacrylamide gel and transblotted onto a polyvinylidene difluoride membrane. The membrane was stained briefly with Ponceau stain (0.1% (w/v) in 1% (v/v) acetic acid), destained in distilled water, and a strip was cut with the band corresponding to MPCBP protein. The strip was cut into 1 ϫ 2-cm pieces, equilibrated in 0.5 M potassium phosphate buffer, pH 7.4, rinsed in phosphate-buffered saline containing 0.1% Tween 20, and then incubated in 10% monoethanolamine in 1 M NaHCO 3 for 2 h. Following two 30-min rinses in phosphate-buffered saline containing 0.1% Tween 20, the strips were incubated with 2 ml of serum for 3.5 h. The strips were rinsed three times and incubated in 0.9 ml of 100 mM glycine, pH 2.5, for 10 min. The solution was removed and neutralized with 0.1 ml of 1 M Tris, pH 0.8 (59).
Immunodetection of MPCBP-Proteins from pollen grains, germinating pollen grains, and other maize tissues were extracted by grinding tissues in liquid nitrogen and homogenizing in extraction buffer (50 mM Tris-HCl, pH 7.4, 5 mM EDTA, 250 mM sucrose, 5 mM DTT, and complete protease inhibitor mixture). After centrifuging at 14,000 ϫ g for 20 min, the supernatant was collected and used for electrophoresis. Microsomal and soluble fractions of the pollen extract were separated by centrifuging at 100,000 ϫ g, and washing the pellet twice with extraction buffer. The microsomal fraction was dissolved in sample buffer. Proteins were separated on SDS-polyacrylamide gel and transblotted onto nitrocellulose membrane using a Bio-Rad transfer cell. After blocking with 3% gelatin in antibody buffer (20 mM Tris-HCl, pH 7.5, 0.5 M NaCl) for 2 h, the membranes were incubated with the affinity-purified MPCBP antibody (1:5,000) in antibody buffer containing 1% gelatin for 2 h at 30°C. The membrane was then washed with antibody buffer containing 0.05% Tween 20 followed by incubation for 1 h at 30°C in 1:4000 dilution of goat anti-rabbit IgG conjugated to alkaline phosphatase (Stratagene). Immunoreactive bands were detected colorimetrically by immersing the filter in substrate solution (0.3 mg/ml nitro blue tetrazolium and 0.15 mg/ml 5-bromo-4-chloro-3-indolyl phosphate in AP buffer).
Purification of Native MPCBP from Pollen Extract Using CaM-Sepharose Column-Maize pollen proteins were extracted in a buffer containing 50 mM Tris-HCl, pH 7.4, 250 mM sucrose, 5 mM DTT, and complete protease inhibitor mixture. The extract was first centrifuged at 14,000 ϫ g for 20 min and then at 100,000 ϫ g for 15 min at 4°C to obtain the supernatant containing soluble proteins. Prior to loading the supernatant onto the CaM-Sepharose column (Amersham Pharmacia Biotech) the concentration of Ca 2ϩ was adjusted to 1.25 mM. The flowthrough was collected and saved. The column was washed thoroughly with binding buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1.25 mM CaCl 2 ) to remove unbound proteins. Washes equivalent to 1-bed volume of the column were collected and saved. The CaM-binding proteins were eluted with elution buffer (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 5 mM EGTA). The absorbance of the eluted fractions was recorded at 235 and 280 nm to compare the elution profile of protein with EGTA. Initial soluble extract, flow-through, wash fraction, and eluted protein were separated on three denaturing gels. One gel was stained with Coomassie Blue and the other two were blotted onto nitrocellulose membrane and probed with MPCBP-specific antibody or HRP-CaM.
Pull-down Assay with CaM-Sepharose Beads-Five-hundred l of pollen soluble proteins prepared as above were mixed and incubated with 0.2 ml of CaM-Sepharose beads at room temperature for 30 min. After a brief centrifugation, the supernatant (designated as flowthrough) was collected and saved. The beads were washed three times with the binding buffer. The first 0.5 ml of wash was saved. The bound proteins were eluted with the elution buffer. The flow-through, wash, and eluted proteins were separated on three SDS gels and processed as above.

Isolation of a Maize Pollen
CaM-binding Protein-We used a protein-protein interaction-based screening to isolate cDNAs encoding CaM-binding proteins (54). A cDNA library from maize pollen prepared in Zap II expression vector was screened with plant CaM isoforms (53). Screening of about 120,000 recombinants resulted in the isolation of 6 clones coding for putative maize pollen MPCBP. To confirm the Ca 2ϩ -dependent binding of CaM, the putative clones were probed with 35 S-CaM in the presence of Ca 2ϩ or EGTA, a Ca 2ϩ chelator. All clones showed binding to 35 S-CaM in the presence of Ca 2ϩ but not in the presence of EGTA (Fig. 1). Restriction enzyme analysis and the length of cDNAs revealed that they were all derived from the same clone. Sequencing of the cDNAs indicated that all six cDNAs are identical. The protein encoded by this gene is termed MPCBP.
Isolation of the Full-length Genomic Clone and Its Structural Analysis-Since the isolated cDNA (1.2 kb) is much smaller than the transcript size (see section on expression analysis), we screened a genomic library to isolate the full-length gene. Screening of maize genomic library with the partial cDNA clone (1.2 kb) yielded several positives. All positives were characterized by restriction mapping and Southern analysis. The clone (number 2) containing the largest hybridizing band was sequenced by primer walking. Introns in MPCBP were predicted by comparing the gene sequence with the cDNA and by using the NetPlantGene and NetGene2 programs. The nucleotide sequence of the MPCBP gene and its predicted amino acid sequence is shown in Fig. 2A. The MPCBP gene has 5 exons and 4 introns (Fig. 2, A and B). The open reading frame of MPCBP starts with a translation initiation codon in exon 1 (nucleotide position 1087 in Fig. 2A) and ends with a stop codon (TGA) in exon 5 (nucleotide position 4572 in Fig. 2A). The size of the added exons or cDNA (1931 bp) is in agreement with the estimated size of the transcript on Northern blot analysis (see below). The predicted protein has 659 amino acid residues with an estimated molecular mass of about 72 kDa. A search of sequence data bases with the predicted amino acid sequence using BLAST searches revealed that the protein is highly similar to two hypothetical proteins found in the Arabidopsis genome data base. A protein with accession number AC006224 showed 54% identity and 68% similarity, and another protein (accession number AC00457) showed 39% identity and 56% similarity. Both of these proteins have similar gene structure with five exons and the CaM-binding domain in MPCBP is conserved in both of these proteins (Fig. 2C). Based on this high degree of sequence identity we named these APCBP1 (accession number AC006224) and APCBP2 (accession number AC00457). APCBP1 and APCBP2 are located on chromosome 2 and chromosome 1, respectively. In addition, the MPCBP showed limited sequence similarity to SPINDLY, a protein involved in GA signal transduction from Arabidopsis (U62135; 34% identity and 57% similarity) (60), and barley (AF035820; 19% identity and 35% similarity) (61), and to an O-linked GlcNAc transferase from Methanobacterium thermoautotrophicum (62) with 19% identity and 37% similarity. Fig. 2D shows the alignment of the deduced amino acid sequence of MPCBP with the amino acid sequence of APCBP1 and APCBP2, two hypothetical proteins identified in the Arabidopsis genome sequencing project. SPINDLY was not included in the alignment as it is a much larger protein (914 amino acids) and showed limited homology. Analysis of the predicted amino acid sequence of MPCBP using SMART (Simple Modular Architecture Research Tool), a program that predicts functional domains (63) revealed the presence of three TPRs, one in each of exons 3, 4, and 5 (Fig. 2, C and D). TPR domains consist of degenerate consensus sequences and are implicated in protein-protein interaction (64). The proteins that are similar to MPCBP contain one or more TPRs. The alignment of TPR1, -2, and -3 of MPCBP with TPRs from Arabidopsis proteins is shown in Fig. 2D. TPR1 of MPCBP showed 63% identity and 69% similarity with TPR1 of APCBP1. TPR2 of MPCBP showed 43% identity and 59% similarity with TPR2 of APCBP2. TPR3 of MPCBP has 79% identity and 85% similarity with TPR2 of APCBP1. A Blast search also revealed sequence similarities between MPCBP TPRs and TPRs from a variety of other proteins.
Mapping of the CaM-binding Domain-The protein encoded by the partial cDNA (1.2 kb) showed CaM-binding, suggesting that the CaM-binding domain is present in the region between 418 and 659 amino acids. To map the location of the CaMbinding domain, four truncated versions of the cDNA were made (Fig. 3A) and expressed in E. coli as His-Tag fusions using pET28 expression vector. Fig. 3B shows a Coomassiestained gel of the fusion proteins and the corresponding blots detected with various probes. The presence of the fusion protein was detected with T7 tag antibody whereas the binding of the fusion protein to CaM was tested with either radiolabeled or biotinylated CaM (Fig. 3B). The expression of the expected size fusion proteins was verified by probing the blot with T7 tag antibody (Fig. 3B). Probing a duplicate blot with 35 S-CaM showed that the protein (16 kDa) expressed from the N-terminal 168-bp region of the cDNA contained the putative CaMbinding domain (Fig. 3B, 35 S-CaM/CaCl 2 , lane 6I). Truncated clones P3/BamHI and P3/NcoI which include the 168-bp region of the cDNA also produced CaM-binding peptide (see Fig. 3B, 35 S-CaM/CaCl 2 , lanes 1I and 4I). Therefore, the protein encoded by the 168-bp truncated cDNA contains the CaM-binding domain. The protein was mainly detected in the insoluble inclusion bodies although under some conditions a small amount was expressed in the soluble fractions (data not shown). No binding to CaM was observed in the presence of EGTA confirming the Ca 2ϩ dependence of the binding of MPCBP to CaM (Fig.  3B, 35 S-CaM/EGTA). Interestingly, the protein did not bind to biotinylated CaM when a duplicate blot was probed with biotinylated CaM in the presence of Ca 2ϩ . This is probably related to the biotin moiety on CaM that may interfere with its binding to MPCBP. The protein expressed from clone 1 (P3) was solubilized in 6 M urea and successfully purified on a bovine CaM-Sepharose affinity column (Fig. 3C), further confirming the binding of the MPCBP to CaM.
A Synthetic Peptide Corresponding to Amino Acids 421 to 438 Binds to CaM-In many CaM target proteins in animals, the CaM-binding domain has been shown to reside in a stretch of 18 -20 amino acid residues (42)(43)(44). Although the amino acid sequence in the CaM-binding domain of different CaM target proteins is not conserved, the binding region is predicted to form a basic, amphiphilic ␣-helix in which hydrophobic residues are segregated from hydrophilic residues along the helix (65). In addition, studies using synthetic peptides confirmed the speculation that CaM recognizes amphiphilic peptides (65). CaM binding studies with truncated proteins of MPCBP have shown that the CaM-binding domain is located in a 56-aa stretch (418 -474). Analysis of this stretch of amino acids using a helical wheel program has revealed a region from 421 to 438 forms basic amphiphilic ␣Ϫhelical structure. To test if this 18-aa stretch binds CaM, a synthetic peptide containing these amino acids was synthesized and used for binding studies. As shown in Fig. 4A, the synthetic peptide (MP-1, VSKGWRLLA-LILSAQQRF) bound to the bovine CaM and CaM isoforms 2, 4, and 6 from Arabidopsis at concentrations as low as 0.5 g indicating that this region is indeed involved in CaM binding (Fig. 4A). Another synthetic peptide (MP2, AKLDQGSLL-RVKAKLKVAQSSPM) corresponding to a different region of MPCBP but lacking typical features of CaM-binding domains was used as a negative control in the binding studies. Neither MP2 nor BSA showed any binding to the labeled CaM isoforms (Fig. 4A, MP-2, and B). These results indicate the specificity of the synthetic peptide in binding to CaM. A previously characterized CaM-binding peptide of a microtubule motor protein (KCBP) bound to all three CaM isoforms (52) (Fig. 4A, K). To determine the stoichiometry of the CaM-peptide complex, we performed binding studies using the synthetic peptide (MP-1) and CaM in the presence of Ca 2ϩ or EGTA. The binding of the synthetic peptide to CaM was detected by a gel mobility shift assay in polyacrylamide gels containing 4 M urea (Fig. 4B). At 4 M urea, low affinity and nonspecific complexes dissociate while high affinity complexes remain intact. In the presence of  (Fig. 4B). No change in CaM mobility was observed in the presence of EGTA (data not shown) suggesting that Ca 2ϩ is required for the formation of CaM-peptide complex. At a molar ratio of 1:1 (peptide:CaM) about 50% of CaM showed a shift (Fig. 4B, lane 2). At a molar ratio of 2:1 and 4:1 (peptide:CaM) the entire CaM migrated as a complex, and the band corresponding to the free CaM disappeared (Fig. 4B, lanes  3 and 4). The synthetic peptide retarded mobility of CaM similarly with bovine CaM and the three AtCaM isoforms, indicating that the peptide binds to these CaMs in the same stoichiometry (Fig. 4B). Peptide-CaM complexes that do not dissociate in 4 M urea have dissociation constants of less than 100 nM (57). These mobility assays suggest that the binding between the peptide and CaM is strong and does not dissociate in the presence of 4 M urea.
The binding of a peptide to CaM can also be tested by fluorescence spectroscopy since the peptide contains a tryptophan residue which is absent in CaM. Tryptophan-containing peptides have been shown to, upon binding to CaM, shift their fluorescence spectrum and change the intensity of fluorescence (52, 57, 66 -68). In our study, The MP1 peptide contains a tryptophan and therefore was tested for fluorescence shift at equimolar ratios of peptide and CaM. As shown in Fig. 4C cation of the MPCBP fusion protein by affinity chromatography columns of Sepharose-4B conjugated to each of the AtCaM 2, 4, and 6 isoforms. The affinity column purified MPCBP fusion protein to a high degree of purity as indicated by the single band on Coomassie-stained gels as well as on blots probed with T7 tag antibody (see Fig. 5).
Expression of MPCBP in Maize-To determine the expression of MPCBP, total RNA from maize roots, shoots, kernels, and pollen was isolated and RNA gel blot analysis was performed. Single transcript of about 2 kb was detected only in the pollen and the germinated pollen (Fig. 6). The transcript is absent in all other tissues. These results were further confirmed by reverse transcription-polymerase chain reaction where the MPCBP transcript was not detected in maize roots, shoots, and kernels (data not shown). To demonstrate the presence of first strand cDNA in reverse transcriptase-polymerase chain reaction, another maize gene (CBP-1) that is expressed in root, shoot, and kernel was amplified (69). The CBP-1 transcript was detected in all of the above tissues (data not shown). These Northern and reverse transcriptase-polymerase chain reaction results indicate that MPCBP is expressed only in pollen. Southern analysis of maize DNA digested with different restriction enzymes, revealed a single hybridizing band indicating that MPCBP in maize is encoded by a single gene (Fig. 7).
Immunodetection of MPCBP in Different Tissues-To determine the expression of MPCBP and the native size of the protein, total protein was extracted from maize roots, leaves, shoot tips, and pollen tissues, separated on a SDS-polyacrylamide gel electrophoresis, transblotted onto polyvinylidene difluoride membranes, and MPCBP was detected using affinitypurified MPCBP antibody. As shown in Fig. 8A, a band of about 72 kDa is detected only in pollen and was absent in all other maize tissues indicating pollen-specific expression of MPCBP. The size of the immunoreactive protein is the same as the predicted size for the MPCBP gene. Two faint smaller size bands on immunoblots are most likely the degradation product of MPCBP as the intensity of these bands varied depending on the presence or absence of protease inhibitors in the extraction buffer. The protein was found in both the soluble and the microsomal fractions of the pollen (Fig. 8A). To examine the presence of the protein during germination and growth of the pollen tubes, maize pollen was germinated on media for 30, 60, 120, and 240 min. Protein extracts from germinated pollen were separated and detected as above. The level of MPCBP protein was fairly constant throughout germination. There was some decline, however, in the amount of MPCBP after 30 min (Fig. 8B). This may indicate that the protein is stored in the mature pollen and may be depleted in the first half hour of germination and is resynthesized during germination and pollen tube growth. These results may indicate that this protein is expressed in mature pollen and during germination, suggesting a role for MPCBP in pollen germination and tube growth.
Native MPCBP Binds CaM in a Ca 2ϩ -dependent Manner-Our studies with bacterially expressed fusion proteins indicate that the truncated C-terminal region of MPCBP binds to CaM in the presence of Ca 2ϩ (Figs. 3 and 5). However, these results do not show that the full-length MPCBP binds CaM. To demonstrate that the native MPCBP interacts with CaM in a Ca 2ϩdependent manner, we isolated MPCBP from maize pollen extract either by using a pull-down assay with CaM-Sepharose beads or by passing protein through a CaM-Sepharose column as described under "Experimental Procedures." The proteins that were isolated with both of these methods were blotted and probed with either affinity purified MPCBP-specific antibody or HRP-CaM (Fig. 9). As shown in Fig. 9A, pollen CaM-binding proteins bound to CaM-Sepharose column in the presence of Ca 2ϩ and eluted in a buffer containing EGTA, a Ca 2ϩ chelator. The spectral curves at 235 and 280 nm clearly show the elution of CaM-binding proteins with EGTA (Fig. 9A). We then analyzed the initial soluble protein extract, the flow-through, the wash, and the eluted proteins from the column as well as the pull-down assay with CaM-Sepharose beads. We separated these proteins along with the bacterially expressed truncated MPCBP (P3) and Arabidopsis KCBP 1.5C (70) on three gels. One gel was stained with Coomassie Blue. The other two were blotted onto nitrocellulose membranes. One blot was probed with MPCBP-specific antibody to detect MPCBP and the second one was subjected to HRP-CaM overlay assay to detect CaM-binding proteins. The majority of the pollen proteins did not bind CaM-Sepharose column (Fig. 9B, compare lane 1 to lane 4). MPCBP-specific antibody detected a single band (72 kDa) protein in crude extract and EGTA eluted fraction (Fig.  9B). Detection of MPCBP only in EGTA eluted fractions (Fig.  9B, lanes 4 and 9) and not in either flow-through or wash fractions (Fig. 9B, lanes 2, 3, 7, and 8) by MPCBP-specific antibody and CaM overlay assay clearly indicates that native MPCBP binds CaM in the presence of Ca 2ϩ . The positive control, P3 (Fig. 9B, lane 5), was also immunodetected by MPCBP antibody. However, HRP-CaM detected several other unknown CaM-binding proteins in both crude extract (Fig. 9B, lane 1) and elution fractions from CaM-Sepharose (Fig. 9B, lanes 4 and  9). No CaM-binding proteins were detected in flow-through and wash fractions (Fig. 9B, lanes 2, 3, 7, and 8), suggesting that all CaM-binding proteins in the pollen extract were bound to CaM-Sepharose and eluted with EGTA. We used KCBP, a well characterized CaM-binding protein from Arabidopsis (lane 6 in Fig. 9B) as another positive control. KCBP was detected by HRP-CaM but not by MPCBP antibody. Since several proteins were bound to CaM-Sepharose, it is possible that MPCBP binds CaM-Sepharose indirectly through other proteins bound to the column. However, two lines of evidence eliminate this possibility. First, the protein detected by MPCBP antibody is also detected by HRP-CaM in a blot overlay assay (Fig. 9B). Second, bacterially expressed truncated MPCBP bound to CaM-Sepharose in a Ca 2ϩ -dependent manner (Figs. 3, 5, and 9B). These results clearly show that the native MPCBP binds CaM in the presence of Ca 2ϩ . DISCUSSION To understand calmodulin action in pollen germination and tube growth, we screened an expression library from maize with calmodulin, one of the primary calcium sensors in eukaryotes. This screening has resulted in isolation of a cDNA clone encoding a CaM-binding protein. Using several different approaches, we have demonstrated that the partial cDNA-encoded peptide binds to CaM with high affinity in a Ca 2ϩ -dependent manner. First, the cDNA-encoded protein bound CaM only in the presence of CaCl 2 but not in the presence of EGTA (Fig. 1). Second, in experiments using the truncated proteins of MPCBP in CaM binding studies, MPCBP bound to 35 S-CaM in a Ca 2ϩ -dependent manner in an SDS gel blot overlay assay (Fig. 3, A and B). The Ca 2ϩ -dependent CaM binding of this protein was further confirmed by purifying the bacterially expressed protein on a CaM affinity chromatography column (Fig. 3C). In addition, by purifying MPCBP from maize pollen using CaM-Sepharose, we have shown that native MPCBP binds CaM in a calcium-dependent manner (Fig. 9). Finally, the CaM-binding domain was narrowed to a stretch of 18 aa ( Figs. 2A and 4). The synthetic peptide corresponding to the CaM-binding domain bound CaM in gel and fluorescence shift assays only in the presence of Ca 2ϩ (Fig. 4). This property of CaM-peptide complexes in gel shift assays and fluorescence spectrometry has been described for many CaM-binding proteins and has been used to determine the binding stoichiometry between CaM-binding proteins and CaM (57). Binding of the protein and the peptide to CaM at a concentration as low as 0.18 M in mobility shifts indicates the high affinity between the MPCBP and CaM. The concentration of CaM used in screening and binding studies is within the physiological levels in the cell indicating that this protein plays a physiological role in pollen (71).
Slot blots, gel mobility, and fluorescence shift assays using the synthetic peptide corresponding to the putative CaM-binding domain from MPCBP and the three CaM isoforms CaM2, CaM4, and CaM6, indicated that MPCBP binds to the three AtCaM isoforms in a Ca 2ϩ -dependent manner. The truncated protein from the P3 clone also complexed with bovine CaM and AtCaM isoforms and shifted their mobility. Purification of the truncated protein was also achieved on AtCaM isoform affinity columns (Figs. 3C and 5). In addition to these studies with truncated proteins, we have shown that the native protein binds calmodulin in a calcium-dependent manner (Fig. 9). Gel shift assays showing the formation of a complex between MPCBP CaM-binding peptide and CaM isoforms in the presence of 4 M urea indicate that MPCBP binds to CaMs with a dissociation constant less than 100 nM (57). Fluorescence shift assay showed some differences in CaM isoform shifts where CaM2 exhibits the highest shift followed by CaM6, bovine CaM, and CaM4. CaM2 shifted to a higher fluorescence and its maximum emission shifted to a lower wavelength than that of CaM6, -4, or bovine CaM. These results agree with our previous work on the binding affinity of CaM isoforms to another CaMbinding peptide, where CaM2 had the highest affinity (71). However, further studies are needed to estimate the affinity of MPCBP to the various CaM isoforms and to determine the dissociation constants.
Data base searches revealed that a homologue of MPCBP is not present in animals. However, the MPCBP amino acid sequence has a high degree of sequence similarity with two hypothetical proteins from Arabidopsis (APCBP1 and APCBP2) indicating the conservation of the gene across monocotyledonous and dicotyledonous plants. The number and position of exons and introns of the Arabidopsis hypothetical proteins (AC006224, APCBP1 and AC00457, APCBP2) are also closely related to MPCBP indicating that they are evolutionarily conserved. The amino acid sequence of MPCBP and ACBP1 also shows a high degree of identity suggesting that APCBP1 is a homologue of MPCBP. This is further supported by a high degree of conservation of the CaM-binding domain between MPCBP and APCBP1. The sequence similarities between these proteins is not confined to the TPR repeats but extends to non-TPR regions. However, SPINDLY (60) another protein with limited sequence similarity to MPCBP, shows protein sequence homology to MPCBP only in the TPR motifs while sequences outside these motifs are not related.
The presence of three TPR domains in MPCBP suggests that it is a member of the TPR family of proteins (64). The TPR motif is a degenerate 34-amino acid region that is repeated up to 16 times and has been identified across the biological kingdoms in a large number of proteins of diverse functions (64). TPR-containing proteins are believed to function through protein-protein interaction and to modulate diverse cellular processes including cell cycle (64,72), transcription (64,(73)(74)(75), protein transport across mitochondria and peroxisomes (76,77), dephosphorylation of proteins (78), heat shock protein (79), and muscle development by interacting with myosins (80). TPRs from different proteins have loosely conserved consensus residues that are conserved in terms of their size, hydrophobicity, and spacing in a way to allow the motif to form a pair of antiparallel amphipathic ␣ helices (64,75,78). These helices are predicted to form the so-called "knob and hole" model which is responsible for protein-protein interaction either intramolecularly among the multiple repeats or intermolecularly by interacting with TPR or non-TPR target proteins (64,75,78). One of the interesting features of TPR sequences is that even though they are loosely conserved with a consensus sequence, there is variation among different TPRs, implying that TPR motifs interact with diverse proteins. In support of this, TPR proteins have been shown to interact with different target proteins (78 -80). Furthermore, mutation or deletion of the TPRs has, in all cases, resulted in loss of function (64,80).
MPCBP showed significant similarities to proteins from Arabidopsis (APCBP1 and APCBP2). Although the function of the hypothetical proteins is not known, the close similarity between MPCBP and Arabidopsis hypothetical protein APCBP1 in regions other than TPR regions indicate that APCBP1 may be an orthologue of MPCBP. The sequence similarity between MPCBP and APCBP2 suggests that APCBP2 is also related to MPCBP. SPINDLY from plants and O-linked glucose N-acetyltransferase from bacteria have limited sequence similarities in certain TPRs and are diverse in functions indicating that they are not related to MPCBP. However, it is interesting to note that each of the 3 TPR domains of MPCBP aligned with specific TPRs from the four proteins suggesting functional specialization for each of the TPR motifs. This specificity of TPR function may be expressed by TPR binding to specific target proteins, or distinct combination of different TPRs may provide a specific binding site for a target protein. For example, only two out of 10 TPRs in SPINDLY share sequence similarities with MPCBP TPR3. This may indicate some specificity of substrate binding of an individual TPR.
The observation that TPRs are present in proteins with a wide array of functions and in various combinations and numbers has led researchers to assume (and prove in some cases) that TPR functions as a scaffold in binding to specific substrates depending upon the secondary structure assumed by the individual or combination of TPRs (64,78,80). The functions reported in the literature all require that the TPR-containing proteins form a complex with another component to regulate their function (64, 75, 76, 79 -82). Genetic analysis of SPINDLY in Arabidopsis and barley predicted the protein to be a negative regulator of the GA signal transduction possibly by interacting with other proteins in the pathway or a protein factor that binds to GA-regulated promoter through its TPR motifs. Mutations in two of the TPR motifs activated the GA transduction pathway indicating the importance of these domains for the function of the protein (60,61). MPCBP as well as SPINDLY showed sequence homology with N-acetylglucoseamine transferases (O-linked GLcNAc transferase) from bacteria and animals. However, various O-linked GLcNAc transferases have a range of 9 -13 tandemly repeated TPR domains clustered in their NH 2 -terminal region while the catalytic domain of transferase activity is in the COOH-terminal (83,84). MPCBP TPR motifs that are homologous to O-linked GLcNAc transferase are COOH-terminal and no sequence similarity in the catalytic domain of the transferase with MPCBP is observed. It is therefore unlikely that MPCBP has a similar enzyme activity.
Northern and Western analyses indicated accumulation of the transcript and the protein in the mature pollen grain of maize. Immunodetection with proteins from germinated pollen indicated that the protein may be utilized in the first half-hour of germination and then is resynthesized. It is known that as the pollen develops it accumulates mRNA required for its rapid germination and tube growth upon landing on the stigma (85). MPCBP may be one of the proteins that is required immediately after hydration especially since it is related to a Ca 2ϩmediated pathway that is implicated early in the germination process and tube growth (27).
Since the MPCBP protein homologues from Arabidopsis are newly identified in the Arabidopsis genome sequencing project, functional analysis of these proteins has not been studied. The presence of TPR domains implies a role of these proteins in regulating cellular processes by interacting with themselves or other proteins. The presence of a CaM-binding domain in MPCBP between TPR-1 and TPR-2 makes this protein unique and suggests that the function of MPCBP is regulated by Ca 2ϩ / CaM. It is known that intracellular tip-focused Ca 2ϩ gradients together with the tip-directed inward currents of extracellular Ca 2ϩ are major events in the process of pollen tube growth and directionality of tip growth (47). It is speculated that Ca 2ϩsensing proteins couple changes in Ca 2ϩ levels to growth response. CaM, a ubiquitous receptor of Ca 2ϩ that is involved in mediating Ca 2ϩ action, is a strong candidate for this role. CaM has been localized in the tip region of growing pollen (49) and in the extracellular vicinity of the pollen tip (30). Tip extension relies upon vesicle migration to the pollen tube tip that is driven by cytoplasmic streaming. Cytoplasmic streaming and tip growth are inhibited by reagents that interfere with Ca 2ϩ homeostasis and antagonize CaM (17,18,45,46). Hence, part of Ca 2ϩ action in controlling tip growth is likely to be mediated by CaM and its target proteins. Cytoplasmic streaming is known to be accomplished by the actomyosin cytoskeletal system and the maintenance of this directional streaming has been reported to be controlled by microtubules (86). Several reports indicated that these two cytoskeletal elements and the associated proteins are regulated by Ca 2ϩ and CaM (87). Calmodulin is known to stabilize the microtubules (43) and to bind to some of the microtubule-associated proteins. It is possible, therefore, that MPCBP after being modulated by CaM may be involved in binding to some component of the cytoskeletal system through its TPR proteins. TPR proteins were reported to bind to all four heavy chain myosin isoforms in a manner necessary for their proper assembly into thick filaments (80). Similarly, Ca 2ϩ /CaM-modulated MPCBP can be associated with the microtubules or microtubule-associated proteins to effect integrity of the microtubule network necessary for the polar organization of pollen tube growth. On the other hand, the TPR domain may be involved in transporting protein complexes across mitochondrial and peroxisomal membranes. The association of MPCBP with microsomes suggests that it may interact with other membrane proteins and this association could be modulated by Ca 2ϩ /CaM. The identification and isolation of a pollen-specific protein with a CaM-binding domain and TPR motifs, both of which are involved in signal transduction, implicate a role for this protein in Ca 2ϩ signaling in pollen. Localization of MPCBP and generation of knock-out mutants should help elucidate the function of MPCBP in pollen.