Modulation of calmodulin function by ubiquitin-calmodulin ligase and identification of the responsible ubiquitylation site in vertebrate calmodulin.

Calmodulin is the universal calcium modulator in eukaryotic cells. Its biological activity is closely regulated by the second messenger Ca2+. Previous studies in cell-free extracts [Laub, M. & Jennissen, H. P. (1997) Biochim. Biophys. Acta 1357, 173-191] have shown that calmodulin is reversibly ubiquitylated by ubiquityl-calmodulin synthetase (ubiquitin-calmodulin ligase, EC 6.3.2.21) in the presence of Ca2+ without being channeled to degradation by the 26S proteasome. As shown here monoubiquitylation strongly decreases the biological activity of calmodulin towards phosphorylase kinase by reducing its affinity approximately threefold and the maximal degree of activation approximately twofold. Thus, a structural clarification of the ubiquitylation site on calmodulin has become crucial for advancing our knowledge in this field on a molecular level. As demonstrated by sequence analysis and mass spectrometry of conjugates, the ubiquitylation site is located in the first Ca2+-binding loop of calmodulin and has the octapeptide structure -L-F-D-K21-D-G-D-G- with Lys21 being the ubiquitylated residue in vertebrate and other calmodulins. This catalytic recognition sequence is, however, not the only structural requirement for calmodulin ubiquitylation by ubiquityl-calmodulin synthetase. Removal of the 41 C-terminal amino acids (fourth Ca2+-binding loop) separated by several nanometers from Lys21 drastically decreases the affinity and reactivity of the synthetase for calmodulin, indicating a more extensive structural requirement for the substrate binding site i.e. binding recognition. This allows the enzyme to discriminate in a site-specific manner between two nearly identical catalytic recognition sites in vertebrate calmodulin of which the second site -V-F-D-K94-D-G-N-G- in the third Ca2+-binding loop is apparently not ubiquitylated by the synthetase.

The ubiquityl-calmodulin synthetase system from rabbit reticulocytes: isolation of the calmodulin-binding second component and enzymatic properties.

Ubiquitin-calmodulin ligase (uCaM synthetase: EC 6.3.2.21), which has been detected in all tissues so far examined, catalyzes the Ca2+-dependent reversible synthesis of ubiquityl-calmodulin which is not directed to degradation by the ATP-dependent 26-S protease [Laub, M. & Jennissen, H. P. (1997) Biochim. Biophys. Acta 1357, 173-191]. As has been shown in the preceding paper in this journal, the uCaM synthetase holosystem can be separated into two essential protein components: uCaM Syn-F1, a ubiquitin-binding protein belonging to the ubiquitin-activating enzyme family (E1) and uCaM Syn-F2 which bestows the reaction specificity leading to the covalent modification of calmodulin with ubiquitin. UCaM Syn-F2, which binds to calmodulin-Sepharose in a Ca2+-dependent manner, has been purified over 3500-fold in seven steps from rabbit reticulocytes and has a native molecular mass of approximately 620 kDa. It binds calmodulin with a Km of 5 microM and to uCaM Syn-F1, i.e. ubiquitin-activating enzyme (E1), with a Km of 3 nM. The maximal specific activity obtained in enriched uCaM Syn-F2 is 6-8 pkat/mg. The pH optimum of uCaM synthetase lies at pH 8.5. In kinetic experiments the Km values for 125I-ubiquitin and ATP/Mg2+ were determined to be 8 microM and 16 nM, respectively, for the uCaM synthetase holosystem. The existence of a third separable protein component of uCaM synthetase, as is the case in E1, E2, E3 systems, is very unlikely since affinity chromatography on calmodulin-Sepharose, two ion-exchange chromatography steps and finally a gel-filtration step failed to indicate any additional protein component essential for synthetase activity. We therefore propose a two-component model for uCaM synthetase. This model is also supported by simple hyperbolic velocity curves in kinetic experiments based on the variation of these two components. The data suggests that uCaM Syn-F2 is neither an E2 nor an E3 but evidently combines the properties of both, making the Ca2+-dependent uCaM synthetase the member of a group of two-component ubiquitin ligase systems.

Control of root growth and development by cyclin expression.

Root development is plastic, with post-embryonic organogenesis being mediated by meristems. Although cell division is intrinsic to meristem initiation, maintenance and proliferative growth, the role of the cell cycle in regulating growth and development is unclear. To address this question, we examined the expression of cdc2 and cye genes, which encode the catalytic and regulatory subunits, respectively, of cyclin-dependent protein kinases that control progression through the cell cycle. Unlike cdc2, which is expressed not only in apical meristems but also before lateral root initiation in quiescent, pericycle cells arrested in the G2 phase of the cell cycle, cyc1At transcripts accumulate specifically in dividing cells immediately before cytokinesis. Ectopic expression of cyc1At under the control of the cdc2aAt promoter in Arabidopsis plants markedly accelerates growth without altering the pattern of lateral root development or inducing neoplasia. Thus cyclin expression is a limiting factor for growth, which in turn drives indeterminate development of the root system.

Structure of a functional geranylgeranyl pyrophosphate synthase gene from Capsicum annuum.

A geranylgeranyl pyrophosphate synthase (GGPPS) gene from Capsicum annuum (bell pepper) was cloned. The nucleotide sequence shows that this gene, like the capsanthin/capsorubin gene but unlike the phytoene synthase gene from C. annuum, is not interrupted by an intron. Southern blot analysis of C. annuum genomic DNA suggests the presence of a single gene highly similar to the cDNA and also of additional related sequences. The present data suggest that this cloned gene is functional.

Structure and expression of two plant genes encoding chromoplast-specific proteins: occurrence of partially spliced transcripts.

The genes encoding fibrillin and capsanthin-capsorubin synthase are specifically expressed during fruit ripening in Capsicum annuum, leading to the accumulation of these two proteins in chromoplasts. Here, we report for the first time the cloning of genomic DNA fragments encoding these two enzymes, as well as DNA fragments containing upstream regions which are potentially involved in the regulation of the expression of these genes. While the capsanthin-capsorubin synthase gene is uninterrupted, the fibrillin gene is interrupted by two introns, the first one being inefficiently spliced. Occurrence of unspliced transcripts is apparently not related to a post-transcriptional mechanism controlling the synthesis of fibrillin or an alternative polypeptide. This work provides tools for studies on gene activation and intron splicing in plants.

The nuclear-encoded polypeptide Cfo-II from spinach is a real, ninth subunit of chloroplast ATP synthase.

Proton-translocating F-ATP synthases from chloroplasts contain a nuclear-coded subunit, CFo-II, that lacks an equivalent in the corresponding E. coli complex. Three recombinant phages that code for the entire precursor of this subunit have been isolated from lambda gt11 cDNA expression libraries made from polyadenylated spinach RNA using a two-step strategy. The reading frame of 222 amino acid residues includes 147 residues for the mature protein (M(r) 16.5 kDa) and a transit sequence of 75 residues (M(r) 8.0 kDa). Secondary structure predictions indicate a bitopic protein, anchored by a single N-terminal transmembrane segment and a C-terminal hydrophilic region that probably reaches into CF1. CFo-II precursor made in vitro can be imported into isolated, intact chloroplasts and assembled into ATP synthase. This protein is a real subunit of the plastid enzyme and a distinctive characteristic of ATP synthases involved in photosynthetic processes. Unique features are (i) that the gene for CFo-II (atpG) appears to be a duplication of atpF encoding CFo-I, the homologues of the genes for subunits b' and b in photosynthetic bacteria, (ii) that it represents the first instance that one copy of the various duplicated loci found in plastid chromosomes has been phylogenetically translocated to the nucleus, and (iii) that it operates with a bipartite (import/thylakoid-targeting) transit peptide but without an intermediate cleavage site for the stroma protease, suggestive of a way of membrane integration different from that of its plastome-encoded counterpart CFo-I. With these data, the first complete sequence for a chloroplast ATP synthase of a higher plant (spinach) is available.

Insertion and assembly of the precursor of subunit II into the photosystem I complex may precede its processing.

The biogenesis and assembly of subunit II of photosystem I (PSI) (psaD gene product) were studied and characterized. The precursor and the mature form were produced in vitro and incubated with intact plastids or isolated thylakoids. Following import of the precursor into isolated plastids, mostly the mature form of subunit II was found in the thylakoids. However, when the processing activity was inhibited only the precursor form was present in the membranes. The precursor was processed by a stromal peptidase and processing could occur before or after insertion of the precursor into the thylakoids. Following insertion into isolated thylakoids, both the precursor and the mature form of subunit II were confined to the PSI complex. Insertion of the mature form of subunit II was much less efficient than that of the precursor. Kinetic studies showed that the precursor was inserted into the membrane. Only at a later stage, the mature form began to accumulate. These results suggest that in vivo the precursor of subunit II is inserted and embedded in the thylakoids, as part of the PSI complex. Only later, it is processed to the mature form through the action of a stromal peptidase.

The 'positive-inside rule' applies to thylakoid membrane proteins.

In integral membrane proteins, regions that span the lipid bilayer alternate with regions that are exposed on either side of the membrane. For proteins from the plasma membrane of both prokaryotic and eukaryotic cells it has been shown that the exposed parts follow a 'positive-inside rule': on average, segments that are translocated across the membrane have a 2-4-fold lower frequency of positively charged residues than non-translocated segments. We now present an analysis of proteins from the thylakoid membrane of chloroplasts. It is shown that these proteins have the same charge asymmetry as has been reported for proteins from other membrane systems, with their more highly charged regions facing the stromal compartment.

Nucleotide sequences of cDNA clones encoding the entire precursor polypeptide for subunit VI and of the plastome-encoded gene for subunit VII of the photosystem I reaction center from spinach.

Recombinant phage which encode the entire precursor polypeptide for subunit VI of the photosystem I reaction center have been selected from a lambda gt11 cDNA expression library made from polyadenylated RNA of spinach seedlings. The sequence predicts a precursor polypeptide of 144 amino acids (Mr = 15.3 kDa), a mature protein of 95 residues (Mr = 10.4 kDa) that lacks methionine, histidine and cysteine, and a transit peptide of 49 residues (Mr = 4.9 kDa). The corresponding gene(s) is (are) designated psaH. The gene for subunit VII, psaC, has been located in the small single-copy region of the spinach plastid chromosome using a synthetic oligonucleotide and a heterologous hybridization probe. It is part of a polycistronic transcription unit that is constitutively expressed and processed. Putative processing products include a monocistronic RNA for psaC. The polypeptide chain of 18 (deduced) amino acids is highly conserved and strikingly resembles bacterial-type ferredoxins. It harbours cysteine residues that appear to be involved in the ligation of the two 4Fe4S centres A and B in photosystem I. None of the two subunits appears to be membrane-spanning, and subunit VI, as subunit VII, is located at the reducing (stromal) side of the reaction center. All available information on the major subunits of photosystem I from spinach has been combined into a (revised) topographic model. Evidence that the innermost - plastome-encoded - core of photosystem I represents an old bacterial heritage in present day chloroplasts is discussed.

Domain structure of mitochondrial and chloroplast targeting peptides.

Representative samples of mitochondrial and chloroplast targeting peptides have been analyzed in terms of amino acid composition, positional amino acid preferences and amphiphilic character. No highly conserved 'homology blocks' are found in either class of topogenic sequence. Mitochondrial-matrix-targeting peptides are composed of two domains with different amphiphilic properties. Arginine is frequently found either at position -10 or -2 relative to the cleavage site, suggesting that some targeting peptides may be cleaved twice in succession by two different matrix proteases. In stroma-targeting chloroplast transit peptides three distinct regions are evident: an uncharged amino-terminal domain, a central domain lacking acidic residues and a carboxy-terminal domain with the potential to form an amphiphilic beta-strand. Targeting peptides that route proteins to the mitochondrial intermembrane space or the lumen of chloroplast thylakoids have a mosaic design with an amino-terminal matrix- or stroma-targeting part attached to a carboxy-terminal extension that shares many characteristics with secretory signal peptides.

Nucleotide sequences of cDNAs encoding the entire precursor polypeptides for subunits II and III of the photosystem I reaction center from spinach.

Several cDNA clones encoding the complete subunit II and III precursor polypeptides of the photosystem I reaction center were isolated from two spinach lambda gt1 1 expression libraries by immunoscreening and homologous hybridization. The identity of the recombinant cDNAs was verified by an N-terminal amino acid sequence of 14 and 20 residues for the respective mature spinach proteins. The ca. 880 nucleotide long sequence and derived amino acid sequence for subunit II predict a precursor of 23.2 kDa (212 residues) and a positively charged, mature protein of 17.9 kDa (162 residues). The corresponding data for subunit III are ca. 710 nucleotides (cDNA), 13.4 kDa (125 residues, precursor polypeptide) and, again, a positively charged, mature protein of 9.7 kDa (91 residues). Secondary structure predictions indicate that both subunits are extramembraneous components of photosystem I. Subunit II is probably located on the matrix-side, subunit III in the lumen of stroma lamellae which is consistent both with biochemical findings and the proposed roles of these proteins in the electron transition from and to photosystem I, respectively. Major transcripts of 1.1 kb (subunit II) and 0.8 kb (subunit III) have been observed by RNA-DNA hybridization.

Nucleotide sequence of cDNA clones encoding the entire precursor polypeptides for subunits IV and V of the photosystem I reaction center from spinach.

Using lambda gt11 expression cloning and immunoscreening, cDNA-containing recombinant phages for subunits IV and V of the photosystem I reaction center were isolated, sequenced and used to probe Northern blots of polyadenylated RNA prepared from spinach seedlings. The mRNA sizes for both components are approximately 1000 and 850 nucleotides, respectively. The 968 nucleotide cDNA sequence and derived amino acid sequence for subunit IV predict a single open reading frame of 231 amino acid residues (25.4 kDa). Comparison with a 13-residue N-terminal amino acid sequence determined for subunit IV suggests a mature protein of 17.3 kDa (154 residues) and a transit sequence of 77 amino acids (8.1 kDa). The corresponding data for subunit V are 677 bp (cDNA), 167 residues for the precursor protein (18.2 kDa), 98 residues for the mature polypeptide (10.8 kDa) and 69 residues for the transit peptide (7.4 kDa). Secondary structure predictions indicate that both proteins possess greatly different transit sequences and that none is membrane-spanning.

Analysis of cDNA clones encoding the entire precursor-polypeptide for ferredoxin:NADP+ oxidoreductase from spinach.

In this paper, we report the structural characterization of several spinach ferredoxin-NADP+ oxidoreductase (FNR) cDNAs ranging in size from 0.9 to 1.5 kilobases. A comparison of the deduced amino acid sequence with the known amino acid sequence determined for the spinach protein establishes that 1.4-1.5 kpb inserts span the full length of the mature protein (314 amino acid residues; Mr = 35,382). These also include an N-terminal 55 amino acid transit peptide as well as maximally 171 and 214 nucleotide 5' and 3' untranslated sequences, respectively. Evidence has been obtained that various forms of FNR arise from at least two similar genes. The FNR precursor (369 amino acid residues) has a calculated molecular mass of 41.2 kDa. Comparison of the transit peptide with transit peptides from two other stromal proteins shows little similarity at the level of primary sequence but some common features in secondary structure predictions.

Nucleotide sequence of cDNA clones encoding the complete precursor for subunit delta of thylakoid-located ATP synthase from spinach.

The nucleotide sequence of the entire nuclear-encoded precursor for subunit delta of the ATP synthase from spinach thylakoid membranes was determined by cDNA sequencing. Appropriate recombinant DNAs were selected from pBR322 and lambda gt11 libraries made from polyadenylated RNA of greening spinach seedlings. The mature protein consists of 187 amino acid residues corresponding to a molecular weight of 20468. The precursor protein (257 amino acid residues; M r=27676) is probably processed between a Met-Val bond. The predicted secondary structure of the transit sequence (70 residues; 7.2 kDa) resembles that of the Rieske Fe/S polypeptide, but shows little similarity with those of stromal or luminal proteins. The comparison of the chloroplast delta amino acid sequence with the published delta sequences from respiratory ATP synthases of bacterial and mitochondrial sources and from the thylakoid ATP synthase of the cyanobacterium Synechococcus suggests substantial divergence at the genic level although structural elements appear to be remarkably conserved.