This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 146/2/772    most recent pp.107.109645v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Wagner, V. Articles by Kreimer, G. Search for Related Content PubMed PubMed Citation Articles by Wagner, V. Articles by Kreimer, G. Agricola Articles by Wagner, V. Articles by Kreimer, G.

SYSTEMS BIOLOGY, MOLECULAR BIOLOGY, AND GENE REGULATION

The Phosphoproteome of a Chlamydomonas reinhardtii Eyespot Fraction Includes Key Proteins of the Light Signaling Pathway^1,[W]

Institut für Allgemeine Botanik und Pflanzenphysiologie, Friedrich-Schiller-Universität Jena, 07743 Jena, Germany (V.W., M.K., M.M.); and Department Biologie, Friedrich-Alexander-Universität Erlangen, 91058 Erlangen, Germany (K.U., A.M., G.K.)

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
Flagellate green algae have developed a visual system, the eyespot apparatus, which allows the cell to phototax. In a recent proteomic approach, we identified 202 proteins from a fraction enriched in eyespot apparatuses of Chlamydomonas reinhardtii. Among these proteins, five protein kinases and two protein phosphatases were present, indicating that reversible protein phosphorylation occurs in the eyespot. About 20 major phosphoprotein bands were detected in immunoblots of eyespot proteins with an anti-phosphothreonine antibody. Toward the profiling of the targets of protein kinases in the eyespot fraction, we analyzed its phosphoproteome. The solubilized proteins of the eyespot fraction were treated with the endopeptidases LysC and trypsin prior to enrichment of phosphopeptides with immobilized metal-ion affinity chromatography. Phosphopeptides were analyzed by nano-liquid chromatography-electrospray ionization-mass spectrometry (MS) with MS/MS as well as neutral-loss-triggered MS/MS/MS spectra. We were able to identify 68 different phosphopeptides along with 52 precise in vivo phosphorylation sites corresponding to 32 known proteins of the eyespot fraction. Among the identified phosphoproteins are enzymes of carotenoid and fatty acid metabolism, putative signaling components, such as a SOUL heme-binding protein, a Ca²⁺-binding protein, and an unusual protein kinase, but also several proteins with unknown function. Notably, two unique photoreceptors, channelrhodopsin-1 and channelrhodopsin-2, contain three and one phosphorylation sites, respectively. Phosphorylation of both photoreceptors occurs in the cytoplasmatic loop next to their seven transmembrane regions in a similar distance to that observed in vertebrate rhodopsins, implying functional importance for regulation of these directly light-gated ion channels relevant for the photoresponses of C. reinhardtii.

Due to the elaborate structures of algal eyespot apparatus and the known presence of rhodopsins in some lineages, algae are thought to play an important role in the evolution of photoperception and eyes (Gehring, 2004). Therefore, the structural components forming this early visual system and its associated signaling cascades are not only of special interest to plant biologists. Until 2005, only six components of the eyespot of C. reinhardtii were known at the molecular level. These included EYE2 and MIN1, two proteins important for eyespot assembly (Roberts et al., 2001; Dieckmann, 2003), two splicing variants of the abundant retinal binding protein COP (Chlamydomonas opsin), and two unique seven-transmembrane domain (TMD) photoreceptors, COP3 and COP4 (Deininger et al., 1995; Fuhrmann et al., 2001; Nagel et al., 2002; Sineshchekov et al., 2002; Suzuki et al., 2003). COP3 and COP4, widely known as channelrhodopsin-1 and channelrhodopsin-2 (ChR-1 and ChR-2), are directly light-gated cation channels that contain a planar all-trans, 6-S-trans retinal chromophore, which undergoes 13-trans to cis isomerization upon illumination. They are involved in the phototactic and photophobic behavior of C. reinhardtii. However, details of the signaling cascades initiated upon their excitation are still obscure (Sineshchekov et al., 2002; Nagel et al., 2002, 2005b; Govorunova et al., 2004; Kateriya et al., 2004). Notably, ChR-2 has been successfully expressed for light stimulation of different systems, including Caenorhabditis elegans and mammalian neurons (Boyden et al., 2005; Nagel et al., 2005a). ChR-2 was even delivered to retinal ganglion cells in a rodent model of inherited blindness (Bi et al., 2006). In this way, genetically engineered surviving retinal neurons were generated to take on the lost photoreceptive function.

A recent conducted proteomic analysis of a fraction enriched in eyespot apparatuses of C. reinhardtii, including the lipid globuli and the associated parts of chloroplast and plasma membranes, resulted in the identification of 202 proteins that were covered by at least two peptides in the mass spectrometry (MS) analysis (Schmidt et al., 2006). Besides the already above-mentioned proteins, this analysis revealed the presence of proteins from diverse functional groups in the eyespot. These include, for example, calcium-sensing and binding proteins, channels, membrane-associated/structural proteins such as proteins with PAP-fibrillin domains, proteins involved in retinal, carotenoid, and chlorophyll biosynthesis as well as in lipid metabolism, and thylakoid membrane-associated proteins. Care must be taken to conclude that thylakoid membrane-associated proteins of the chloroplast, which are also present in the eyespot, have the same localization and/or function there. The -, β-, and -subunits of the soluble CF1 complex of the ATP synthase, for example, probably have specialized localization within the eyespot and are thermolysin resistant (Schmidt et al., 2007). Of special interest is also the presence of a SOUL heme-binding protein (SOUL/HBP) in the eyespot. The first member of this group of HBPs was found to be specifically located in the retina and pineal gland in chicken (Zylka and Reppert, 1999). Because Descartes considered the pineal gland as the soul, Zylka and Reppert named the protein accordingly. Further on, kinases and phosphatases were found in the eyespot proteome, indicating that light signaling cascades(s) may involve regulation by reversible protein phosphorylation. The two detected protein phosphatases (PPs) belong to the PP2C family of Ser/Thr PPs. In higher plants, it was shown that PP2Cs often regulate signaling negatively (Schweighofer et al., 2004). The five identified kinases in the eyespot proteome include a cyclic nucleotide-dependent kinase II, two unusual protein kinases with AarF domains, the blue light photoreceptor phototropin with its Ser/Thr kinase domain, and casein kinase1 (CK1). CK1 was functionally analyzed by an RNAi approach (Schmidt et al., 2006). Its silencing has multiple effects and results in severe disturbances in hatching, flagellum formation, and circadian control of phototaxis. These first data already point out the significance of reversible phosphorylation events in the signaling pathways within the eyespot and toward the flagella. Therefore, we decided to find the targets of these kinases and phosphatases in the eyespot by a phosphoproteome approach.

In C. reinhardtii, several cellular subfractions have been unraveled in the past years by large-scale proteome approaches, such as the flagella (Pazour et al., 2005), centrioles (Keller et al., 2005), or the eyespot (Schmidt et al., 2006). First phosphoproteome approaches have also been carried out, for example, with crude extracts (Wagner et al., 2006) or thylakoids isolated from C. reinhardtii cells grown under different environmental conditions (Turkina et al., 2006a, 2006b). However, phosphoproteome analysis has been and still is a challenging task (for review, see Mann et al., 2002; Reinders and Sickmann, 2005). This is due to a few facts. (1) Phosphoproteins can have more than one phosphorylation site and the phosphorylation status of these sites can fluctuate, depending on the physiological conditions of the cells. (2) Only a small portion of a given protein in the cell can be phosphorylated. (3) Furthermore, phosphoproteins, especially those of signaling pathways, are often low-abundance proteins anyway. Therefore, enrichment of phosphoproteins/peptides from the cell or a subcellular compartment is a prerequisite for efficient phosphoproteome analysis. Different methods can be used for this purpose (for review, see Reinders and Sickmann, 2005). One of them, immobilized metal-ion affinity chromatography (IMAC), is based on the presence of the negatively charged phosphate groups and enriches for phosphorylated Ser, Thr, and Tyr. This method has already been applied for phosphoproteome analyses in different systems like lymphoma cells (Shu et al., 2004), higher plants (Nühse et al., 2004), and C. reinhardtii (Turkina et al., 2006a, 2006b; Wagner et al., 2006). It relies on the direct identification of phosphopeptides in MS in contrast to other methods that chemically substitute phosphate residues. However, in tandem MS (MS/MS; hereafter MS²), phosphopeptide precursor ions can exhibit neutral loss of phosphoric acid (–98 D). The reason for this loss is that phosphopeptides (phospho-Ser and phospho-Thr) can undergo gas-phase β-elimination when subjected to collision-induced fragmentation (Mann et al., 2002). Because mass/charge (m/z) values and not absolute masses are measured in a mass spectrometer, doubly and triply charged peptide ions show an apparent loss of 49 and 32.66, respectively. In the MS² spectrum, the presence of neutral loss therefore indicates phosphorylation and can be used as a selection parameter for phosphopeptides. There are some types of electrospray ionization (ESI)-MS (e.g. linear ion trap LTQ [Thermo Electron]) that allow the acquisition of data-dependent neutral loss. Thus, neutral-loss analysis can be executed during MS measurements (MS² scan). If a pair of peaks for one of the most prominent ions of the MS² spectrum versus the full-scan MS spectrum is found with a mass difference of 98, 49, or 32.66 (depending on the charge of the peptide ion), a phosphorylation-specific neutral loss is indicated. This prominent ion of the MS² spectrum (representing the dephosphorylated peptide) will then be selected for an automatically triggered MS/MS/MS (MS³) scan so that sufficient fragment ion information can be obtained if not provided by the MS² spectrum. Thus, data from MS² spectra can be additionally confirmed by MS³ spectra. This method was used, for example, for phosphoproteomics for the yeast (Saccharomyces cerevisiae) pheromone signaling pathway (Gruhler et al., 2005), in the recent whole-cell phosphoproteome approach in C. reinhardtii (Wagner et al., 2006), but also for a selected protein (Ouelhadj et al., 2007).

In this work, we analyzed the phosphoproteins of eyespot preparations to get information about its in vivo kinase targets and thereby insights into its signaling network. Due to the elaborate structure of the eyespot and the rather hydrophobic character of many of its proteins, we had to apply a special protocol to bring the proteins in proper solution for efficient proteolytic digest followed by IMAC. Multiple liquid chromatography (LC)-ESI-MS analyses from independent eyespot preparations were then conducted via MS² and neutral-loss-triggered MS³ spectra. Thus, 68 phosphopeptides, belonging to 32 proteins that were already identified in former eyespot proteome analyses, as well as 15 phosphopeptides that do not correspond to yet-known proteins from this fraction, were identified. Analysis of phosphorylation sites revealed a bias toward certain amino acids in their surroundings and a tendency to occur outside of known functional domains. The eyespot phosphoproteome includes proteins of (potential) light signaling pathways, chloroplast and thylakoid components, carotenoid and fatty acid metabolism, but also several proteins with unknown function. Notably, two photoreceptors, ChR-1 and ChR-2, were also found with three and one phosphorylation sites, respectively. Localization of these sites in a cytoplasmatic loop with close proximity to the channel-forming region implies functional relevance for the regulation of these unique directly light-gated ion channels.

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

The Eyespot Fraction Contains a Significant Number of Thr-Phosphorylated, But Only a Few Tyr-Phosphorylated Proteins

Detection of five kinases and two Ser/Thr PPs of the PP2C family in the eyespot proteome of C. reinhardtii (Schmidt et al., 2006) underlined the potential importance of reversible protein phosphorylation for signaling pathways in this complex cell organelle. For phosphoproteome analyses, preparation of the eyespot fraction was basically done according to Schmidt et al. (2006). Additionally, a set of seven phosphatase inhibitors (microcystin LR, cantharidin, (–)-p-bromotetramisole, vanadate, molybdate, tartrate, imidazole) was used during cell rupture to reduce potential dephosphorylation of the proteins. For analysis of phosphoproteins, we first prepared immunoblots with two different commercially available phospho-amino acid-specific antibodies. Western analysis of the proteins from the purified eyespot fraction with a polyclonal anti-phospho-Thr antibody revealed about 20 major and some minor phosphorylated protein bands in the apparent molecular mass range 25 kD (Fig. 1A ). The labeling pattern was reproducibly observed in several independent eyespot isolations. Coomassie staining of protein bands transferred to the polyvinylidene difluoride (PVDF) membrane showed that several phosphoproteins (e.g. >116 kD) belong to low-abundance proteins that are slightly, if at all, detectable by Coomassie staining (Fig. 1A). Specificity of the labeling was confirmed by alkali treatment of the blotted proteins prior to immunoanalysis, which caused massive reduction in both labeling intensity and amount of detected phosphoproteins (Fig. 1B). This treatment is known to significantly reduce the amount of phosphorylation at Ser and Thr residues of blotted proteins without significant protein loss from the PVDF membrane (Kamps and Sefton, 1989; Rudrabhatla et al., 2006). To analyze whether Tyr-phosphorylated proteins were also present in the final eyespot fraction, a monoclonal anti-phospho-Tyr antibody was used. In contrast to the results obtained with the anti-phospho-Thr antibody, in this case only two major protein bands at apparent molecular masses of 37 and 40 kD were consistently labeled (Fig. 1C). These results clearly demonstrate that (1) considerable amounts of phosphoproteins are still present in the eyespot fraction after the isolation procedure; and (2) phosphorylation apparently occurs mainly at Thr and probably also at Ser residues, whereas phosphorylation at Tyr is minor.

View larger version (34K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Western-blot analysis of proteins from the eyespot fraction of C. reinhardtii with phospho-amino acid-specific antibodies. A, Proteins (4 µg) were separated by 11% SDS-PAGE, transferred to a PVDF membrane, and either analyzed with a polyclonal anti-phospho-Thr antibody (Anti-pThr; 1:1,000) or stained with Coomassie Brilliant Blue R250 (Coomassie). B, The membrane, to which proteins (4 µg) had been blotted, was treated with 1 M KOH according to Kamps and Sefton (1989) prior to incubation with the antiserum to reduce the amount of phosphorylated Ser and Thr residues. The developing time was identical to that in A (3 min). C, Proteins (4 µg) from two independent eyespot isolations were separated by SDS-PAGE (11%), blotted, and probed with a monoclonal anti-phospho-Tyr antibody (Anti-pTyr; 1:1,000, developing time 15 min). Arrowheads indicate the two major labeled protein bands. Positions of molecular mass markers are indicated on the left (kD).

Because only a small fraction of the proteins prone to phosphorylation exists in the phosphorylated state at any given time and proteins regulated by phosphorylation can be very low abundant, enrichment of the phosphopeptides is an essential initial step prior to their subsequent analysis. Removal of the nonphosphorylated peptides increases selectivity and detection sensitivity of phosphopeptides in the MS analysis. Different approaches are currently used for phosphopeptide enrichment from complex mixtures of peptides (for review, see Reinders and Sickmann, 2005). We applied IMAC based on Ga³⁺ as a metal ion. Application of the membrane-shaving method with trypsin, which is often used to release surface-exposed phosphopeptides of membrane proteins prior to IMAC (e.g. Vener et al., 2001; Turkina et al., 2006a, 2006b), was not chosen here for the following reason. We could not exclude that the eyespot globules form a steric barrier that prevents complete hydrolysis of the proteins. Because the eyespot fraction has a low protein and high carotenoid/lipid content, samples were extracted by chloroform:methanol:water (4:8:3 [v/v/v]) to concentrate the proteins and to extract the lipids/carotenoids in parallel. Precipitated proteins were solubilized with 6 M guanidine hydrochloride. We then applied the protocol of Wagner et al. (2006), which has successfully been used for phosphoproteins of crude extracts from C. reinhardtii. In this protocol, proteins are digested with trypsin and afterward phosphopeptides are enriched with Ga³⁺-based IMAC. For efficient digestion with trypsin, it is necessary to dilute the guanidine hydrochloride solution containing the solubilized proteins down to a concentration of 0.6 M. However, in the eyespot sample, massive protein precipitation occurred upon dilution of guanidine hydrochloride prior to tryptic digestion. Therefore, a new protocol was developed for getting a representative phosphopeptide yield from eyespot fractions. For this purpose, the precipitated proteins were solubilized in 4 M urea instead of 6 M guanidinium hydrochloride and then treated with the endoproteinase LysC, which specifically cleaves peptide bonds C-terminally at Lys and still has an enzyme activity of 86% in 4 M urea (www.roche-applied-science.com, datasheet LysC). Because the proteolytic digest with LysC results in peptides with relatively high molecular masses, which are not suited for routine LC-ESI-MS analysis, the reduced and desalted peptides were subjected to a second digest with trypsin prior to enrichment of the phosphopeptides with IMAC. By using this modified procedure (summarized in Fig. 2 ), aggregation and precipitation of the proteins of the eyespot fraction, which is enriched in proteins with a moderate hydrophobic and amphipathic character (Schmidt et al., 2006), were avoided to a large extent.

View larger version (14K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Schematic overview of phosphopeptide enrichment from the eyespot fraction of C. reinhardtii for LC-ESI-MS analysis. Details are described in "Materials and Methods." Nano-LC-ESI-MS (MS2 and MS3) analysis was carried out in a mass spectrometer with a linear ion trap that permits the acquisition of data-dependent neutral loss (Finnigan LTQ; Thermo Electron).

Phosphopeptides from the eyespot fraction were enriched and analyzed in three independent experiments according to the protocol presented in Figure 2. To increase the confidence of the final phosphoprotein-peptide assignment, the following four criteria had to be fulfilled. (1) Phosphopeptides could be identified within the protein sequence of a predicted gene model of the analyzed genome. We used version 3 (Vs3) and the older version 2 (Vs2) of the C. reinhardtii genome in our analysis because the eyespot proteome was based on Vs2 (Schmidt et al., 2006). (2) Phosphopeptides belong to proteins that have already been identified in former eyespot analyses. An exception was only made in one case, which is explained later. (3) They showed a significant Xcorr factor and probability score (see "Materials and Methods" for details). (4) Phosphoproteins were identified in at least two of the three independent experiments. Thus, one phosphopeptide purification experiment always covers the proteins from two independent eyespot isolations. In total, we identified 68 phosphopeptides fulfilling these criteria. Among them, 19 peptides (indicated by an arrow in Table I ) had overlapping sequences or the same, but existed with a different number of phosphorylation sites. The latter has most likely functional implications for the respective protein. Table I summarizes the phosphopeptides along with the protein ID numbers and biological function (if known) of the corresponding protein models. One further phosphopeptide belonging to a protein of yet-unknown function (protein ID 187713; Vs3) was also included in the list despite the fact that it was not found in our former eyespot analyses (Schmidt et al., 2006; M. Mittag and G. Kreimer, unpublished data). This protein has a distribution of trypsin sites that results in peptides being either too small or too large for routine LC-ESI-MS analysis and thus could have easily been missed in those analyses. It is specially marked in Table I. In two cases, the exact protein model could not be identified with certainty because the same peptide was present in different protein models. This concerns, for example, a peptide found in different members of the light-harvesting complex II (LHCII) family, which are highly conserved. In these cases, all possible protein IDs and National Center for Biotechnology Information (NCBI) BLASTp results are listed. In statistical analyses, these cases were taken as one protein. Further details (charges, Xcorr values, and probability scores) of the identified 69 phosphopeptides are listed in Supplemental Table S1. Moreover, the MS² and MS³ spectra of these phosphopeptides are publicly available under the following Web site (http://www.uni-jena.de/Protein_network_of__C__reinhardtii-lang-en.html), where spectra from each peptide resulting in the highest Xcorr are shown.

The 69 phosphopeptides presented in Table I belong in total to 33 different proteins. They correspond to 27 gene models of Vs2 and 29 models of Vs3 of the C. reinhardtii genome. For three models present in Vs2, but missing or reduced in Vs3, ESTs as well as proteomic support for the Vs2 model exist (Table I; Schmidt et al., 2006; Wagner et al., 2006; this article). Seventeen of the phosphoproteins were identified by a single phosphopeptide and 16 were covered by more than one, with up to nine phosphopeptides (Table I), where slightly different peptides with overlapping sequences or identical peptides with different phosphorylation sites are also counted. From the identified 33 phosphoproteins in this study, only six were previously described in phosphoproteomic approaches to whole-cell extracts and thylakoids of C. reinhardtii (Table II ; Turkina et al., 2006b; Wagner et al., 2006) and an additional three were found when comparing the phosphoproteins from Supplemental Table S2. Thus, most phosphorylation sites described here are novel. Additionally, none of the proteins known or putatively involved in light signaling is among the previously identified phosphoproteins, whereas at least two key elements of the phototactic signaling cascade have been identified in this study (see below). Thus, subcellular fractionation prior to specific phosphoprotein enrichment increases the probability of identifying signaling-related and low-abundance phosphoproteins.

View larger version (24K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. Identification of a phosphopeptide from ChR-1 (ID 166415; Vs2) by nano-LC-ESI-MS2 and by neutral-loss-triggered MS3. A, Identification of the phosphopeptide ASpLDGDPNGDAEANAAAGGK by the MS2 fragmentation pattern of peptide ion 942.02; "p" indicates the phosphorylation site on Ser. B, Identification of the peptide ADhaLDGDPNGDAEANAAAGGK by the MS3 fragmentation pattern of the detected neutral-loss fragment 892.6; "Dha" indicates the site of the neutral loss of phosphoric acid from phospho-Ser. Only prominent y- and b-fragment ions have been labeled.

Characterization and Functional Analysis of the Phosphoproteins from the Eyespot

To get functional information about the identified phosphoproteins, we did NCBI protein BLAST homology searches and analyzed for putative TMDs (Table I). Predictions using three different programs (see "Materials and Methods") revealed that the majority of the phosphoproteins in the eyespot fraction have at least a partial hydrophobic character. For 30% of the phosphoproteins, all three programs predicted at least one TMD. For an additional 54%, two of the used programs predicted TMDs.

The results of the NCBI searches were complemented by additional functional information from the Joint Genome Institute (JGI) Web site and annotation notes (Supplemental Table S1). As already detailed above, with one exception, all proteins in Table I were identified in large-scale proteome analyses of eyespot fractions carried out previously by our laboratories. A significant group of the proteins on the list (10) represents novel and conserved proteins of as-yet unknown function. The highest phosphorylated protein identified in our analyses falls in this group (ID 167609; Vs2). It was identified with six different phosphopeptides along with nine phosphorylation sites. Among the proteins with functional clues (23), those (potentially) involved in light signaling show significant representation (five) as will be discussed later. The most interesting candidates among these are the two photoreceptors ChR-1 and ChR-2, which are involved in phototactic and photophobic responses of C. reinhardtii. ChR-1 was identified by three different and ChR-2 by a single phosphopeptide. A detailed analysis will be presented in the next paragraph. A small number of phosphoproteins represent most likely contaminants, such as a putative RNA helicase and a putative mitochondrial external rotenone-insensitive NADPH dehydrogenase.

Analysis of Phosphorylation Sites Reveals a Bias for Surrounding Amino Acids and a Tendency for Clustering Outside Known Functional Domains

Specificity of the active sites of kinases toward their substrates is primarily based on the linear sequence surrounding the phosphorylation site and several typical motifs are known for plant and animal kinases (e.g. Blom et al., 1999; Nühse et al., 2004). To detect the general frequency of selected single amino acid residues in the surrounding phosphorylation sites and to find a potential bias, we analyzed the –6 to +6 positions of all unambiguously assigned phospho-Ser and phospho-Thr sites from phosphopeptides in Table I, including those of Supplemental Table S2. Our analysis was restricted to the group of acidic, basic, and aromatic amino acids. Additionally, Gly was included in this analysis. In total, 135 acidic, 79 basic, 69 aromatic amino acids, and 71 Gly residues were found to surround the phosphorylation sites. The frequency of these amino acids in positions –6 to +6 relative to their total frequency in these 12 positions is shown in Figure 4 . Clear indications of a bias were found for acid and basic residues as well as for Gly. Whereas acidic residues cluster in the +1 to +6 positions with a peak at the +2 position, basic residues are overrepresented in positions –2, –3, and –6. A disproportionate frequency was also found for Gly. Overall, 35.2% of the total Gly residues are either present at the –1 or the +1 position, but were virtually excluded from the –2 and +5 positions. Except for clear exclusion from position –4, clear bias for the aromatic residues is less evident. Indication of a bias toward acidic and basic residues is in agreement with the observation that acidic- or basic-directed site specificity is common among characterized kinases. Examples are the basic motif [R,K]-X_1-2-Sp or the acidic motifs Sp-[D,E] and [D,E]-Sp-[D,E] (e.g. Blom et al., 1999; Nühse et al., 2004, and refs. cited therein).

View larger version (26K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. Frequency of selected amino acids in single positions relative to their total frequency in positions –6 to +6 surrounding the Ser/Thr phosphorylation sites. Only those phosphorylation sites, which were unambiguously assigned, were included in the analysis. For the analyses, sequences of Vs2 were used when models were present in both genome versions.

Functional importance of the location of the phosphorylation sites can be deduced for the two ChRs. In both vertebrates and invertebrates, rapid phosphorylation and dephosphorylation of rhodopsin critically controls the visual transduction cascade (Ohguro et al., 1998; Shukolyukov, 1999; Kennedy et al., 2001; Lee et al., 2002). Thus, the identification of phosphopeptides from both ChRs of C. reinhardtii in this study is of special interest. Phosphorylation of the vertebrate rhodopsins occurs at multiple sites at the C terminus. Besides two Tyr phosphorylation sites directly at the end of the last TMD, five additional sites are clustered in a distance of 28 to 37 amino acids away from it (Ser-334 to Ser-343; PhosphoSite; http://www.phosphosite.org/Login.jsp). We therefore analyzed the localization of the phosphorylation sites in the rhodopsins of C. reinhardtii in detail. Three sites were identified in ChR-1 (Ser-358, Thr-373, Ser-376; Fig. 5A ). Two of these sites are also predicted by NetPhos2.0 (http://www.cbs.dtu.dk/services/NetPhos) with high scores. For ChR-2, we identified a single site at Ser-321. In both photoreceptors, phosphorylation occurred in the first cytoplasmic loop after the seven TMDs (Fig. 5, A and B), which form the light-gated channel (Nagel et al., 2002, 2003; Sineshchekov et al., 2002). Additionally, the distance from the last TMD to the first phosphorylation site is almost identical in both ChRs (54 amino acids in ChR-1 and 55 amino acids in ChR-2). Both ChRs have unusually long C-terminal extensions. Thus, the relative position of the phosphorylation sites with respect to the seven TMD regions is remarkably similar to those of the vertebrate rhodopsins, underlining the potential functional significance of phosphorylation at these sites also in the algal ChRs. The amino acid sequences surrounding the first phosphorylated Ser residue in ChR-1, ChR-2, and the closely related ChR-1 of Volvox are highly conserved (Fig. 5C). No additional putative ortholog sequences of this motif were found by NCBI Protein BLAST searches against the nonredundant protein sequence database and the Chlamydomonas genome using the 60-amino acid sequence motif surrounding the phosphorylated Ser-358 of ChR-1. Because specificity of the active sites of kinases toward their substrates is primarily based on the linear sequence surrounding the phosphorylation site (e.g. Nühse et al., 2004), this might indicate that green algal ChRs are targets of a specialized kinase.

View larger version (49K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. Identified phosphorylation sites in the two photoreceptors ChR-1 and ChR-2. A, Partial sequences of ChR-1 and ChR-2 are depicted. In both cases, parts of the C-terminal sequences are missing. For ChR-1, only the first 420 and for ChR-2 only the first 328 amino acids are shown, respectively. Peptides identified by MS analyses have been labeled in bold and phosphorylation sites are marked by a "p" with gray background. All TMDs are underlined. B, Localization of the phosphorylation sites in relation to the TMD arrangement of ChR-1 and ChR-2 (modified from Kateriya et al., 2004). Phosphorylation sites are marked by an asterisk. C, ClustalW alignment of the experimentally determined first phosphorylation site of the two ChRs from C. reinhardtii (CrChr-1 and CrChR-2) and the orthologous sequence of the ChR of Volvox carterii (VcChR-1). *, Perfect matches; :, high similarity; ·, low similarity. Phosphorylated Ser is highlighted by a gray background and charged amino acid residues (+/–) at identical positions are indicated above the sequence.

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In large-scale proteomics using peptide-enrichment strategies, such as IMAC, protein identification often has to rely on identification by a single peptide. Thus, it is very important that the single peptide is identified with high stringency. We therefore did three independent isolations with two pooled eyespot fractions for each (Fig. 2) and chose only peptides that were present in at least two of the three isolations fulfilling the Xcorr and P values (see "Materials and Methods"). In the central list (Table I), we included only proteins (with one exception, discussed below) that had also been depicted in former proteome analyses of eyespot fractions (Schmidt et al., 2006; M. Mittag and G. Kreimer, unpublished data). From the 33 identified phosphoproteins that we found in this way, 16 were covered with more than one phosphopeptide. In the case of 13 phosphoproteins, there is only information available from one phosphopeptide based either on MS² (eight) or on neutral-loss-triggered MS³ (five) spectra. Four phosphoproteins were depicted by one phosphopeptide based on information from both MS² and MS³ spectra. Combining the information obtained in MS² and MS³ spectra further increases confidence in individual peptide identification (Nielsen et al., 2005). In total, 22 phosphopeptides are based on information from both MS² and MS³ spectra in our approach.

By this procedure, we were able to obtain insight into the phosphoproteome of the eyespot of C. reinhardtii, a subcellular compartment that is complex structured, as mentioned before, and involves thylakoid membranes subtending the two lipid globule layers. Comparisons to the currently known phosphoproteins from C. reinhardtii obtained by phosphoproteomic approaches with whole-cell extracts (Wagner et al., 2006) or thylakoid membranes (Turkina et al., 2006b) revealed that only a few of these phosphoproteins (six) are overlapping with the eyespot phosphoproteome (Table II). Thus, a significant number of phosphoproteins (27) could only be identified when the subcellular eyespot fraction was enriched prior to enrichment of its phosphopeptides by IMAC. These data further underline the importance of purifying subcellular fractions, such as the eyespot, as an efficient and necessary approach to get a rather complete set of phosphoproteins from any subcellular fraction. However, one should not neglect the possibility that the significant increase in phosphoproteins from the eyespot might in parallel be due to application of the modified phosphopeptide enrichment procedure involving LysC treatment. This is especially valid for the rather hydrophobic proteins that may simply not have been brought in solution in the former whole-cell phosphoproteomic approach (Wagner et al., 2006). Still, we cannot be sure whether several phosphoproteins from the eyespot may have been missed. Immunoblots with anti-Thr and anti-Tyr antibodies, along with proteins from purified eyespot apparatuses, revealed about 20 and two protein bands, respectively, that were significantly and reproducibly labeled. These analyses indicate that Tyr phosphorylation is rare in the eyespot fraction and Ser/Thr phosphorylation dominates. This agrees notably well with the outcome that all precisely determined phosphorylation sites (55) in the central list (Table I) are found on Ser and Thr. Only three Tyr phosphorylation sites were found with proteins, which were not depicted in former eyespot proteome analyses (listed in Supplemental Table S2). This is additionally in agreement with the currently identified kinases of the eyespot fraction, which are all predicted Ser/Thr kinases. Only some of them (e.g. cyclic nucleotide-dependent protein kinase II) have, in addition, predicted Tyr-kinase activity. Further, dominating PPs identified in the large-scale proteomic approach to the eyespot fraction are type PP2C Ser/Thr PPs (Schmidt et al., 2006).

Immunoblots also revealed the presence of very low-abundance phosphoproteins hardly detectable by this technique. Such phosphoproteins are of special interest because signaling components often occur in a very low amount in the cell. Among the identified 33 phosphoproteins, at least some of them should have been detected. The functional classification of the identified phosphoproteins indeed shows several (putative) components of light signaling pathways. Among them is a Ca²⁺-binding protein with an EF hand motif (ID 162748; Vs2) with a C-terminally located single phosphorylation site. In the behavioral responses of C. reinhardtii to light, extracellular Ca²⁺ and Ca²⁺ fluxes are intricately involved (Schmidt and Eckert, 1976; Morel-Laurens 1987; Pazour et al., 1995; for review, see Witman, 1993). Both photoreceptors ChR-1 and ChR-2 are directly light-gated ion channels, which are able to conduct Ca²⁺ under physiological conditions. Their excitation initiates fast inward-directed complex currents in the region of the eyespot, which finally produce a Ca²⁺ influx into the flagella (Harz and Hegemann, 1991; Holland et al., 1996; Nagel et al., 2002, 2003; Sineshchekov et al., 2002; for review, see Sineshchekov and Govorunova, 2001; P. Hegemann, personal communication). Notably, changes in the free concentration of Ca²⁺ have been shown to strongly affect rapid protein phosphorylation and dephosphorylation in isolated eyespot apparatuses of the green alga Spermatozopsis similis (Linden and Kreimer, 1995). No Ca²⁺-dependent protein kinases have been detected in the core eyespot proteome of C. reinhardtii (Schmidt et al., 2006). However, in an extended eyespot proteome analysis, two kinases of this type have been detected recently (M. Mittag and G. Kreimer, unpublished data). Their calculated molecular masses (67 and 54 kD) are similar to proteins exhibiting Ca²⁺-dependent autophosphorylation in eyespot preparations of S. similis (77, 48, and 47 kD; Linden and Kreimer, 1995; Schlicher et al., 1995). Due to the outlined central role of Ca²⁺ in eyespot-related signaling, it seems likely that the identified Ca²⁺-binding phosphoprotein might be one member of this light-driven signaling cascade.

Another protein that might be involved in this cascade represents a SOUL/HBP (ID 185703; Vs3) annotated as SOUL3. A SOUL/HBP was found in a screen for chicken mRNAs specifically expressed in the retina and pineal gland, where it might play a role in light signaling or in protecting the retina from damage by reactive oxygen species (Zylka and Reppert, 1999; Sato et al., 2004). In the meantime, several HBPs have been found in bacterial, animal, and plant systems and several proteins of this family have been annotated in Vs3 of the Chlamydomonas genome (Merchant et al., 2007). Functional information on plant HBPs is still lacking, whereas the former ones are involved in different processes. For example, mammalian NPAS2 has two heme-binding sites and is functionally similar to CLOCK. Both are transcription factors involved in the oscillatory loop of the circadian clock that is entrained by light-dark cycles (DeBruyne et al., 2007). In Escherichia coli, a heme-regulated phosphodiesterase was characterized (Sasakura et al., 2005). In cultured rat hepatocytes, HBP23 is an antioxidant protein, which is induced by various stress stimuli (Immenschuh et al., 2002). Phosphorylation of the C. reinhardtii SOUL/HBP at two sites of its N terminus (Thr-42 and Ser-44) might play a role for its heme-binding capacity, since His and Cys possibly involved in the necessary complexation of Fe²⁺ in the heme are also situated at the N terminus (T. Schulze and M. Mittag, unpublished data). A second Chlamydomonas SOUL/HBP (ID 154433; Vs2), which is not present in the eyespot proteome, is also phosphorylated at the N terminus (Ser-2 and Ser-3; Wagner et al., 2006).

Also a predicted unusual protein kinase (ID 153985) could be involved in signaling events. It is well known that protein kinase activity can be regulated by phosphorylation or dephosphorylation (Krupa et al., 2004). The relatively strong bias of amino acids surrounding the phosphorylation sites that has been found in eyespot phosphoproteins (Fig. 4) might indicate that a limited set of kinases was active under the chosen physiological conditions. This would be in concert with the identification of only a few kinases in the eyespot proteome (Schmidt et al., 2006). The second kinase identified here (stt7) is clearly not involved in the photoresponses. It regulates state transitions in C. reinhardtii (Fleischmann et al., 1999).

The most striking candidates that are clearly involved in light signaling are the directly light-gated ChR-1 and ChR-2. In both cases, the phosphorylation sites are located after the seven TMDs in the cytoplasmatic loop region at the C termini. Whereas ChR-2 has only one phosphorylation site, ChR-1 has three sites that are situated close together. In vertebrates, the C terminus of rhodopsin has a cluster of Ser and Thr residues that are considered hallmarks for multiple phosphorylations. Thus, phosphate groups are mostly found at Ser-334, Ser-338, and Ser-343 (Lee et al., 2002; Maeda et al., 2003). In vertebrates, phototransduction starts with the absorption of light, which causes photoisomerization of 11-cis-retinal to all-trans-retinal and initiates a phototransduction signaling cascade. Signal transduction in rods is terminated by a series of molecular events. These are initiated by phosphorylation of rhodopsin by a G-protein-coupled receptor kinase followed by binding of a regulatory protein, arrestin (Maeda et al., 2003). As detailed before in "Results," the relative position of the phosphorylation sites of the C. reinhardtii ChRs with respect to the seven TMD regions is remarkably similar to those of vertebrate rhodopsins. This might point to some evolutionarily conserved signaling mechanism with regard to the phosphorylation sites, albeit the involved kinases and components of the pathways are probably quite different. Because the amino acid sequences surrounding the first phosphorylated Ser residue in ChR-1 and ChR-2 are highly conserved and cannot be found in other proteins from C. reinhardtii, the green algal ChRs might well be targets of a specialized kinase. From the five kinases identified so far in the eyespot fraction, the two predicted unusual kinases with AarF domains could be involved. The other identified kinases (CK1, the cyclic nucleotide-dependent kinase II, and phototropin) are not likely candidates for such a specialized function because they are not restricted to the eyespot. All three can also be found, for example, in the flagellum (Pazour et al., 2005) and may be involved in eyespot flagellum-directed signaling processes. In the case of CK1, there are already data available supporting this hypothesis (Schmidt et al., 2006). However, we cannot also rule out that the responsible kinase has been lost during the eyespot isolation procedure and is not among the currently known kinases of this fraction.

Among the identified phosphoproteins, there were also some enzymes that are involved in carotenoid and lipid metabolism, components of thylakoid membranes, or chloroplast outer membranes and other membrane-related components. 1-Deoxy-D-xylulose-5-P synthase (DXS) catalyzes the first step in the methylerythritol phosphate pathway leading finally to isopentenyldiphosphate, a precursor of carotenoids (Lohr et al., 2005). Several phosphorylation sites were found in this protein, indicating that phosphorylation might play a role (e.g. for the activity or targeting of the enzyme). At least one of these sites (MSYT) is conserved in several other DXSs from different organisms. Furthermore, two proteins (Vs2; IDs 158846 and 161417) grouped in the eyespot proteome by Schmidt et al. (2006) under membrane-associated/structural proteins were identified in this study as phosphoproteins. One (Vs2; ID 161417) contains a motif with weak homology to the PAP/fibrillin domain and the other has homologies to MORN repeat proteins. These proteins carry multiple membrane occupation and recognition nexus motifs, which function in attaching proteins to membranes and forming junctional complexes between membranes (Takeshima et al., 2000; Shimada et al., 2004). In the MORN repeat protein 1 found in the eyespot preparation, two phosphorylation sites were detected. One of them was already found in the cellular phosphoproteome and could be confirmed here. The other one is novel. Interestingly, Drosophila retinophilin, which contains MORN repeats, is conserved in humans (Mecklenburg, 2007). Drosophila retinophilin is expressed in fly photoreceptor cells and the human homolog was demonstrated to be expressed in the retina. Despite the fact that significant homology of the C. reinhardtii protein to these proteins is restricted to the MORN motif region, the presence of proteins with such motifs in visual systems from algae to humans is notable.

Several of the phosphoproteins found in this study show no significant hit in NCBI BLASTp so their potential function is unknown. Some of them have multiple phosphorylation sites. One example is a protein with nine phosphorylation sites (ID 167609). Such candidates are also relevant for future functional analysis because these proteins could be specific to green algal eyespots and, thus, no homology to other organisms may yet be available in the databases.

Our phosphoproteome analysis has also identified proteins that have not been found in previous eyespot proteome analyses (one protein in Table I and 13 in Supplemental Table S2). In these analyses, the proteins from the eyespot were separated on one-dimensional SDS-PAGE and slices from the gels were used for in-gel tryptic digestion. Thus, there was no specific enrichment for phosphopeptides. Within these 14 proteins, 13 were identified by only one phosphopeptide. Three of them were also found in former phosphoproteome analysis (Turkina et al., 2006b; Wagner et al., 2006). Seven of them possess more than one phosphorylation site. As reported previously, multiple phosphorylated peptides are preferentially enriched by IMAC (Mann et al., 2002). Therefore, it seems possible that very low-abundance phosphoproteins/peptides of the eyespot, which have not been detected in recent proteome analyses, were significantly enriched by the IMAC procedure, especially when they have multiple phosphorylation sites. Another possibility would be an unfavorable distribution of tryptic sites in these proteins, resulting in peptides too long or short for routine LC-ESI-MS analysis. We checked the tryptic sites in silico for all 14 proteins, but only in one case (ID 170226; Vs2) were the tryptic sites found to be very unfortunate. Therefore, this one protein was included with a special mark in Table I. Notably, for most of the proteins in Supplemental Table S2, the function is fully unknown.

The phosphoproteins detected in our analyses represent only one analyzed physiological condition. As pointed out earlier, the phosphorylation state of a given protein can strongly depend on physiological conditions. For example, it was shown that state 1 to 2 transitions induced phosphorylation of the PSII core components D2 and PsbR and quadruple phosphorylation of a minor LHCII antennae subunit, CP29, as well as phosphorylation of constituents of a major LHCII complex, Lhcbm1 and Lhcbm10 in C. reinhardtii (Turkina et al., 2006b). Examinations of various physiological conditions, especially with respect to various light conditions (e.g. intensity and quality), will be interesting to study in the future. However, this will be time consuming and work intensive. For this identification of the eyespot phosphoproteome, 40 L of cells were grown for one enrichment of phosphopeptides (two pooled samples). To get reliable results, we repeated this procedure two times and used only candidates that came up in at least two of the three experiments. Thus, 120-L cell cultures and six eyespot isolations were needed in total for the final experiments for only one physiological condition. Furthermore, only analyses of steady-state conditions will be amenable by the chosen experimental approach. Expected transient and/or rapid changes in the in vivo phosphorylation status of eyespot proteins (e.g. in response to sudden changes in the light conditions) cannot be analyzed in this way due to the time-consuming concentration of large volumes of cells and the subsequent isolation of eyespots. Nevertheless, these data underline that phosphorylation plays an important role in eyespot signaling. Having a significant number of targets of the eyespot kinases in hand creates an efficient basis for studying the molecular components of the eyespot signaling pathways in detail. For example, silencing of the eyespot kinases, as was already done with CK1 (Schmidt et al., 2006), can show which targets and phosphorylation sites in the eyespot are specific for a given kinase. Of special interest will be the identification of the most likely specialized kinases involved in phosphorylation of the two ChRs because it might have a critical role in guiding phototransduction in the eyespot. It is striking that the relative position of the functional sites of phosphorylation within the algal and vertebrate rhodopsins are so much conserved. The identification of these sites paves the way for future studies aimed to dissect the exact role of phosphorylation of these unique directly light-gated ion channels in the phototactic signaling cascade.

	MATERIALS AND METHODS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Isolation of Eyespot Apparatuses

Growth of Chlamydomonas reinhardtii strain cw15 and isolation of fraction A enriched in eyespot apparatuses was done as previously described (Schmidt et al., 2006), except that the homogenization buffer was supplemented with phosphatase inhibitor cocktails 1 and 2 (Sigma; dilution 1:50). The concentrated eyespot fraction was extracted overnight with chloroform:methanol:water (4:8:3 [v/v]) at –20°C. Precipitated proteins were washed several times with methanol:chloroform (2:1 [v/v]) and finally dried in a speed-vac (1 min at room temperature) and stored at –80°C.

Protein Digestion and Reverse-Phase Chromatography

The precipitated proteins of two independent eyespot isolations (0.5–0.7 mg) were digested with endoproteinase LysC (Roche Diagnostics) and the resulting peptides were reduced and carboxyamidomethylated according to Nielsen et al. (2005) prior to a tryptic digest of the purified peptides. Briefly, proteins were resuspended in 250 µL of LysC buffer (4 M urea, 100 mM Tris-HCl, pH 8.0). Five micrograms of endoproteinase LysC (in 50 µL of LysC buffer) were added and the mixture was incubated overnight at 25°C. After addition of further 200 µL of LysC buffer, 10 µL of 0.5 M dithiothreitol were added. The solution was incubated for 30 min on ice. Subsequently, 9 mg of iodoaceteamide were added to the solution to reach a final concentration of 100 mM. After 2 h of incubation at 25°C, peptides were separated from the insoluble material by centrifugation (25,000g; 15 min; 4°C). Peptides were fractionated on a reverse-phase column (SOURCE 15 RPC; GE Healthcare) on a FPLC system. After adding acetonitrile (2% final concentration) and formic acid (0.1% final concentration) to the peptide mix, peptides were loaded onto the column and eluted with 0.5 mL 80% acetonitrile and 0.1% formic acid. The eluent was dried in a vacuum concentrator and the resulting pellet was dissolved in 100 µL of 100 mM NH₄HCO₃ followed by overnight incubation at 37°C with 10 µg of trypsin (Promega). Tryptic peptides were again desalted with a SOURCE RPC 1-mL column (see above) and applied to the IMAC procedure.

Enrichment of Proteins by IMAC

IMAC was done according to Wagner et al. (2006), using self-made micro-columns prepared according to Erdjument-Bromage et al. (1998). Briefly, 50 µL of POROS 20 MC (Applied Biosystems) metal-chelating resin (66% [w/w] slurry) was transferred into Eppendorf gel-loading tips. For sufficient column packing, 2 x 50 µL of 0.1% acetic acid were passed through the column. Charging of the micro-tip column was carried out by applying 150 µL of 100 mM GaCl₃, followed by a washing step (50 µL of 0.1% acetic acid). The following procedure was performed according to Shu et al. (2004) with slight modifications. After loading the peptides onto the activated IMAC column, the column was subsequently washed with 50 µL of 0.1% acetic acid, 50 µL of 50% acetonitrile/0.1% acetic acid, 50 µL of 50% acetonitrile/0.1% acetic acid/100 mM sodium chloride, and, finally, with 50 µL of 0.1% acetic acid. The phosphopeptides were eluted with 3 x 20 µL of 200 mM Na₂HPO₄ and desalted using a ZipTip micro tip (Perseptive Biosystems) according to Stauber et al. (2003). The ZipTip contained C18 reverse-phase chromatographic medium and was additionally loaded with 50 µL of a POROS R2 (Applied Biosystems) mixture (10 µL of POROS R2 + 50 µL of methanol). To improve binding of the phosphopeptides to the reverse-phase column, formic acid (final concentration 5%) was added to the sample. After two washing steps with 5% methanol/5% formic acid, the acidified sample was applied to the ZipTip, followed by two washes with 50 µL of 5% methanol/5% formic acid. Peptides were eluted with 2 x 50 µL of 60% methanol/5% formic acid. The eluate was dried in a speed-vac for 2 to 3 h and stored at –80°C.

Peptide Identification by LC-ESI-MS

After dissolving the dried pellet in 5 µL of 5% acetonitrile/0.1% formic acid, phosphopeptides were subjected to nano-LC-ESI-MS using an UltiMate 3000 nano-HPLC (Dionex) with a flow rate of 300 nL/min coupled online with a linear ion trap ESI mass spectrometer (Finnigan LTQ; Thermo Electron). Peptides were concentrated on a µ-precolumn cartridge (LC Packings) using a flow rate of 2 µL/min for 2 min. After applying the peptides to the analytical reverse-phase C18 column (LC Packings), the successive steps of the gradient for the elution of the peptides were as follows: 5 min 96% A/4% B (v/v); within 30 min gradually to 50% A/50% B (v/v); within 1 min gradually to 10% A/90% B (v/v); 10 min 10% A/90% B (v/v); within 1 min gradually to 96% A/4% B (v/v); and 15 min 96% A/4% B (v/v), whereby A consists of 5% acetonitrile/0.1% (v/v) formic acid and B consists of 80% acetonitrile/0.1% (v/v) formic acid. The instrument was run by the data-dependent neutral-loss method, cycling between one full MS and MS² scans of the four most abundant ions. After each cycle, these peptide masses were excluded from analysis for 1 min. Detection of a neutral-loss fragment (98, 49, or 32.66 D) in the MS² scans immediately triggered a MS³ scan of the precursor ion representing the dephosphorylated peptide. Peptide mass tolerance was set to 1.5 D in MS mode. In MS² and MS³ mode, fragment ion tolerance was set up to 1 D.

Data Analysis

Data analysis was done according to Schmidt et al. (2006) with some modifications using Bioworks software (version 3.2) from Thermo Electron including the SEQUEST algorithm (Link et al., 1999). Searches were done for tryptic peptides allowing two missed cleavages. Software parameters were set to detect a modification of 79.96 D in Ser, Thr, or Tyr in MS² and MS³ spectra. When phospho-Ser and phospho-Thr undergo gas-phase β-elimination, Dha and 2-aminodehydrobuturic acid (methyldehydroalanine), respectively, are produced. Thus, modifications of –18.00 D on Ser and Thr residues were additionally used for the database searches with MS³ data. Further, detection of a modification of 16 D on Met representing its oxidized form was enabled. Scores for the cross-correlation factor Xcorr (Eng et al., 1994) were set to the following limits: Xcorr > 1.5 if the charge of the peptide is 1; Xcorr > 2 if the charge of the peptide is 2; and Xcorr > 2.5 if the charge of the peptide is 3. Only peptides that fulfilled the Xcorr limits and had, in addition, a probability score P < 0.05 were included into the result tables. P was recently introduced with the Bioworks (version 3.2) software from Thermo Electron and represents the statistical likelihood of finding an equally good peptide match by chance. By default, peptide probabilities are reported as probabilities normalized to 1 and a lower probability value represents a better match (Bioworks, version 3.2). Application of a P-value screening, in addition to the Xcorr settings, strengthens the quality of positive hits. Furthermore, three independent phosphoprotein purifications (one purification combines the proteins from two independent eyespot isolations) were analyzed and for all databases a significant hit was only considered when the identified phosphoprotein was present in at least two of these samples.

Data were searched against the following C. reinhardtii databases: final model database from the JGI (versions 2 and 3 [genome.jgi-psf.org/chlre2/chlre2.home.html and genome.jgi-psf.org/chlre3/chlre3.home.html]), mitochondrial database available from the NCBI (NC001638 [gi:11467088]), and the chloroplast database (www.chlamy.org/chloro/default.html). Data from all runs were combined and further evaluated using a program developed in-house (Schmidt et al., 2006). Additionally, EST information from the Kazusa DNA Research Institute (Asamizu et al., 1999, 2000, 2004) and the JGI were used for data evaluation as well as previous C. reinhardtii proteome data from the eyespot and the cellular phosphoproteome deposited on a protein network site of C. reinhardtii (available under http://www2.uni-jena.de/biologie/chlamy/index.php?page=search&cmd=search). Protein sequences of the gene models were compared to the NCBI protein database using BLAST (Altschul et al., 1997). For positive identification of both protein and functional domain prediction, an internal cutoff E value of 1e-05 was used. TMD information was based on predictions by the programs TMHMM (Krogh et al., 2001), TMpred (Hofmann and Stoffel, 1993), and TopPred (von Heijne, 1992). Theoretical predictions of phosphorylation sites were done with NetPhos 2.0 software (Blom et al., 1999).

Electrophoretic Methods

For SDS-PAGE analyses, a modified high-Tris system was used. Lipid removal, protein precipitation, and SDS-PAGE were conducted as described (Kreimer et al., 1991; Calenberg et al., 1998). Western-blot analyses followed basically the protocol of Schlicher et al. (1995) with the following modifications: Skim milk powder was replaced by bovine serum albumin, the first blocking step was reduced to 2 h, and incubations with the polyclonal anti-phospho-Thr antibody and the monoclonal anti-phosphor-Tyr antibody (Cell Signaling Technologies; both used at a dilution of 1:1,000) were done overnight at 4°C. Images of blots were taken with a Coolpix 990 (Nikon) and processed with Photoshop (Adobe Systems).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure S1. Full MS scan in the m/z range 300 to 2,000 with the prominent peptide ion 942.02

Supplemental Table S1. Detailed information about the identified phosphopeptides from the Chlamydomonas eyespot fraction.

Supplemental Table S2. Identified phosphopeptides from the Chlamydomonas eyespot fraction that belong to proteins that have not been identified in former eyespot proteome analyses.

	ACKNOWLEDGMENTS

 
We thank Wolfram Weisheit and Sascha Schäuble for their help with the Web site containing the spectra. We appreciate the free delivery of EST and genome sequences from the genome projects of C. reinhardtii in the U.S. Department of Energy (JGI) and Japan (Kazusa DNA Research Institute).

Received September 24, 2007; accepted November 28, 2007; published December 7, 2007.

	FOOTNOTES

 
¹ This work was supported by the Deutsche Forschungsgemeinschaft (grant no. Kr 1307/7–1 to G.K. and grant nos. Mi 373/7–3 and Mi 373/8–3 to M.M.).

² These authors contributed equally to the article.

The authors responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantphysiol.org) are: Georg Kreimer (gkreimer{at}biologie.uni-erlangen.de) and Maria Mittag (m.mittag{at}uni-jena.de).

^[W] The online version of this article contains Web-only data.

www.plantphysiol.org/cgi/doi/10.1104/pp.107.109645

^* Corresponding author; e-mail gkreimer{at}biologie.uni-erlangen.de.

	LITERATURE CITED

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
Altschul SF, Madden TL, Schäffer AA, Zhang J, Zhang Z, Miller W, Lipman D (1997) Gapped BLAST and PSI-BLAST: a new generation of protein database search programs. Nucleic Acids Res 25: 3389–3402[Abstract/Free Full Text]

Asamizu E, Miura K, Kucho K, Inoue Y, Fukuzawa H, Ohyama K, Nakamura Y, Tabata S (2000) Generation of expressed sequence tags from low-CO₂ and high-CO₂ adapted cells of Chlamydomonas reinhardtii. DNA Res 7: 305–307[Abstract]

Asamizu E, Nakamura Y, Sato S, Fukuzawa H, Tabata S (1999) A large scale structural analysis of cDNAs in a unicellular green alga, Chlamydomonas reinhardtii. I. Generation of 3433 non-redundant expressed sequence tags. DNA Res 6: 369–373[Abstract]

Asamizu E, Nakamura Y, Miura K, Fukuzawa H, Fujiwara S, Hirono M, Iwamoto K, Matsuda Y, Minagawa J, Shimogawara K, et al (2004) Establishment of publicly available cDNA material and information resource of Chlamydomonas reinhardtii (Chlorophyta) to facilitate gene function analysis. Phycologia 43: 722–726[Web of Science]

Bi A, Cui J, Ma YP, Olshevskaya E, Pu M, Dizhoor AM, Pan ZH (2006) Ectopic expression of a microbial-type rhodopsin restores visual responses in mice with photoreceptor degeneration. Neuron 50: 23–33[CrossRef][Web of Science][Medline]

Blom N, Gammeltoft S, Brunak S (1999) Sequence- and structure-based prediction of eukaryotic protein phosphorylation sites. J Mol Biol 294: 1351–1362[CrossRef][Web of Science][Medline]

Boyden ES, Zhang F, Bamberg E, Nagel G, Deisseroth K (2005) Millisecond-timescale, genetically targeted optical control of neural activity. Nat Neurosci 8: 1263–1268[CrossRef][Web of Science][Medline]

Bruce VG (1970) The biological clock in Chlamydomonas reinhardtii. J Protozool 17: 328–334

Calenberg M, Brohsonn U, Zedlacher M, Kreimer G (1998) Light- and Ca²⁺-modulated heterotrimeric GTPases in the eyespot apparatus of a flagellate green alga. Plant Cell 10: 91–103[Abstract/Free Full Text]

DeBruyne JP, Weaver DR, Reppert SM (2007) CLOCK and NPAS2 have overlapping roles in the suprachiasmatic circadian clock. Nat Neurosci 10: 543–545[CrossRef][Web of Science][Medline]

Deininger W, Kröger P, Hegemann U, Lottspeich F, Hegemann P (1995) Chlamyrhodopsin represents a new type of sensory photoreceptors. EMBO J 14: 5849–5858[Web of Science][Medline]

Dieckmann CL (2003) Eyespot placement and assembly in the green alga Chlamydomonas. Bioessays 25: 410–416[CrossRef][Web of Science][Medline]

Eng J, McCormack AL, Yates JR (1994) An approach to correlate tandem mass spectral data of peptides with amino acid sequences in a protein database. J Am Soc Mass Spectrom 5: 976–989[CrossRef][Web of Science]

Erdjument-Bromage H, Lui M, Lacomis L, Grewal A, Annan RS, McNulty DE, Carr SA, Tempst P (1998) Examination of micro-tip reversed-phase liquid chromatographic extraction of peptide pools for mass spectrometric analysis. J Chromatogr A 826: 167–181[CrossRef][Web of Science][Medline]

Fleischmann MM, Ravanel S, Delosme R, Olive J, Zito F, Wollman FA, Rochaix JD (1999) Isolation and characterization of photoautotrophic mutants of Chlamydomonas reinhardtii deficient in state transition. J Biol Chem 274: 30987–30994[Abstract/Free Full Text]

Fuhrmann M, Stahlberg A, Govorunova E, Rank S, Hegemann P (2001) The abundant retinal protein of the Chlamydomonas eye is not the photoreceptor for phototaxis and photophobic responses. J Cell Sci 114: 3857–3863[Abstract/Free Full Text]

Gehring W (2004) Historical perspective on the development and evolution of eyes and photoreceptors. Int J Dev Biol 48: 707–717[CrossRef][Web of Science][Medline]

Govorunova EG, Jung K-H, Sineshchekov OA, Spudich JL (2004) Chlamydomonas sensory rhodopsins A and B: cellular content and role in photophobic responses. Biophys J 86: 2342–2349[Web of Science][Medline]

Gruhler A, Olsen JV, Mohammed S, Mortensen P, Gaergeman NJ, Mann M, Jensen ON (2005) Quantitative phosphoproteomics applied to the yeast pheromone signaling pathway. Mol Cell Proteomics 4: 310–327[Abstract/Free Full Text]

Harz H, Hegemann P (1991) Rhodopsin-regulated calcium currents in Chlamydomonas. Nature 351: 489–491[CrossRef]

Hofmann K, Stoffel W (1993) TMbase—a database of membrane spanning proteins segments. Biol Chem Hoppe Seyler 374: 166

Holland EM, Braun FJ, Nonnengäßer C, Harz H, Hegemann P (1996) The nature of rhodopsin triggered photocurrents in Chlamydomonas. I. Kinetics and influence of divalent cations. Biophys J 70: 924–931[Web of Science][Medline]

Hubbard MJ, Cohen P (1993) On target with a new mechanism fort he regulation of protein phosphorylation. Trends Biochem Sci 18: 172–177[CrossRef][Web of Science][Medline]

Immenschuh S, Iwahara S, Schwennen B (2002) Induction of heme-binding protein23/peroxiredoxin I gene expression by okadaic acid in cultured rat hepatocytes. DNA Cell Biol 21: 347–354[CrossRef][Web of Science][Medline]

Kamps MP, Sefton BM (1989) Acid and base hydrolysis of phosphoproteins bound to immobilon facilitates analysis of phosphoamino acids in gel-fractionated proteins. Anal Biochem 176: 22–27[CrossRef][Web of Science][Medline]

Kateriya S, Nagel G, Bamberg E, Hegemann P (2004) "Vision" in single-celled algae. News Physiol Sci 19: 133–137[Abstract/Free Full Text]

Keller LC, Romijn EP, Zamora I, Yates JR III, Marshall WF (2005) Proteomic analysis of isolated Chlamydomonas centrioles reveals orthologs of ciliary-disease genes. Curr Biol 15: 1090–1098[CrossRef][Web of Science][Medline]

Kennedy MJ, Lee KA, Niemi GA, Craven KB, Garwin GG, Saari JC, Hurley JB (2001) Multiple phosphorylation of rhodopsin and the in vivo chemistry underlying rod photoreceptor dark adaptation. Neuron 31: 87–101[CrossRef][Web of Science][Medline]

Kreimer G (2001) Light reception and signal modulation during photoorientation of flagellate algae. In M Lebert, D-P Häder, eds, Comprehensive Series in Photosciences. Elsevier/North Holland, Amsterdam, pp 193–227

Kreimer G, Brohsonn U, Melkonian M (1991) Isolation and partial characterization of the photoreceptive organelle for phototaxis of a flagellate green alga. Eur J Cell Biol 55: 318–327[Web of Science][Medline]

Krogh A, Larsson B, von Heijne G, Sonnhammer ELL (2001) Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol 305: 567–580[CrossRef][Web of Science][Medline]

Krupa A, Preethi G, Srinivasan N (2004) Structural modes of stabilization of permissive phosphorylation sites in protein kinases: distinct strategies in Ser/Thr and Tyr kinases. J Mol Biol 339: 1025–1039[CrossRef][Web of Science][Medline]

Lee KA, Craven KB, Niemi GA, Hurley JB (2002) Mass spectrometric analysis of the kinetics of in vivo rhodopsin phosphorylation. Protein Sci 11: 862–874[CrossRef][Web of Science][Medline]

Linden L, Kreimer G (1995) Calcium modulates rapid protein phosphorylation/dephosphorylation in isolated eyespot apparatuses of the green alga Spermatozopsis similis. Planta 197: 343–351[Web of Science]

Link AJ, Eng J, Schieltz DM, Carmack E, Mize GJ, Morris DR, Garvik BM, Yates JR III (1999) Direct analysis of protein complexes using mass spectrometry. Nat Biotechnol 17: 676–682[CrossRef][Web of Science][Medline]

Lohr M, Im C-S, Grossman AR (2005) Genome-based examination of chlorophyll and carotenoid biosynthesis in Chlamydomonas reinhardtii. Plant Physiol 138: 490–515[Abstract/Free Full Text]

Maeda T, Imanishi Y, Palczewski K (2003) Rhodopsin phosphorylation: 30 years later. Prog Retin Eye Res 22: 417–434[CrossRef][Web of Science][Medline]

Mann M, Ong SE, Grønborg M, Steen H, Jensen ON, Pandey A (2002) Analysis of protein phosphorylation using mass spectrometry: deciphering the phosphoproteome. Trends Biotechnol 20: 261–268[CrossRef][Web of Science][Medline]

Marchler-Bauer A, Bryant SH (2004) CD-Search: protein domain annotations on the fly. Nucleic Acids Res 32: 327–331[CrossRef]

Mecklenburg KL (2007) Drosophila retinophilin contains MORN repeats and is conserved in humans. Mol Genet Genomics 277: 481–489[CrossRef][Web of Science][Medline]

Merchant SS, Prochnik SE, Vallon O, Harris EH, Karpowicz SJ, Witman GB, Terry A, Salamov A, Fritz-Laylin LK, Maréchal-Drouard L, et al (2007) The evolution of key animal and plant functions is revealed by analysis of the Chlamydomonas genome. Science 318: 245–251[Abstract/Free Full Text]

Mittag M, Kiaulehn S, Johnson CH (2005) The circadian clock in Chlamydomonas reinhardtii. What is it for? What is it similar to? Plant Physiol 137: 399–409[Free Full Text]

Morel-Laurens N (1987) Calcium control of phototactic orientation in Chlamydomonas reinhardtii: sign and strength of response. Photochem Photobiol 45: 119–128[Web of Science][Medline]

Nagel G, Brauner M, Liewald JF, Adeishvili N, Bamberg E, Gottschalk A (2005a) Light activation of channelrhodopsin-2 in excitable cells of Caenorhabditis elegans triggers rapid behavioral responses. Curr Biol 15: 2279–2284[CrossRef][Web of Science][Medline]

Nagel G, Ollig D, Fuhrmann M, Kateriya S, Musti AM, Bamberg E, Hegemann P (2002) Channelrhodopsin-1: a light-gated proton channel in green algae. Science 296: 2395–2398[Abstract/Free Full Text]

Nagel G, Szellas T, Huhn W, Kateriya S, Adeishvili N, Berthold P, Ollig D, Hegemann P, Bamberg E (2003) Channelrhodopsin-2, a directly light-gated cation-selective membrane channel. Proc Natl Acad Sci USA 100: 13940–13945[Abstract/Free Full Text]

Nagel G, Szellas T, Kateriya S, Adeishvili N, Hegemann P, Bamberg E (2005b) Channelrhodopsins: directly light-gated cation channels. Biochem Soc Trans 33: 863–866[CrossRef][Web of Science][Medline]

Nielsen PA, Olsen JV, Podtelejnikov AV, Andersen JR, Mann M, Wisniewski JR (2005) Proteomic mapping of brain plasma membrane proteins. Mol Cell Proteomics 4: 402–408[Abstract/Free Full Text]

Nühse TS, Stensballe A, Jensen ON, Peck SC (2004) Phosphoproteomics of the Arabidopsis plasma membrane and a new phosphorylation site database. Plant Cell 16: 2394–2405[Abstract/Free Full Text]

Ohguro H, Yoshida N, Shindou H, Crabb JW, Palczewski K, Tsuda M (1998) Identification of a single phosphorylation site within octopus rhodopsin. Photochem Photobiol 68: 824–828[Web of Science][Medline]

Ouelhadj A, Kaminski M, Mittag M, Humbeck K (2007) Receptor-like protein kinase HvLysMR1 of barley (Hordeum vulgare L.) is induced during leaf senescence and heavy metal stress. J Exp Bot 58: 1381–1396[Abstract/Free Full Text]

Pazour G, Sineshchekov OA, Witman GB (1995) Mutational analysis of the phototransduction pathway of Chlamydomonas reinhardtii. J Cell Biol 131: 427–440[Abstract/Free Full Text]

Pazour GJ, Agrin N, Leszyk J, Witman GB (2005) Proteomic analysis of a eukaryotic cilium. J Cell Biol 170: 103–113[Abstract/Free Full Text]

Reinders J, Sickmann A (2005) State-of-the-art in phosphoproteomics. Proteomics 5: 4052–4061[CrossRef][Web of Science][Medline]

Roberts DGW, Lamb MR, Dieckmann CL (2001) Characterization of the eye2 gene required for eyespot assembly in Chlamydomonas reinhardtii. Genetics 158: 1037–1049[Abstract/Free Full Text]

Rudrabhatla P, Reddy MM, Rajasekharan R (2006) Genome-wide analysis and experimentation of plant serine/threonine/tyrosine-specific protein kinases. Plant Mol Biol 60: 293–319[CrossRef][Web of Science][Medline]

Sasakura Y, Kanda K, Yoshimura-Suzuki T, Matsui T, Eukuzono S, Shimizu T (2005) Investigation of the relationship between protein-protein interaction and catalytic activity of a heme-regulated phosphodiesterase from Escherichia coli (Ec DOS) by protein microarray. Biochemistry 44: 9598–9605[CrossRef][Web of Science][Medline]

Sato E, Sagami I, Uchida T, Sato A, Kitagawa T, Igarashi J, Shimizu T (2004) SOUL in mouse eyes is a new hexameric heme-binding protein with characteristic optical absorption, resonance raman spectral, and heme-binding properties. Biochemistry 43: 14189–14198[CrossRef][Web of Science][Medline]

Schlicher U, Linden L, Calenberg M, Kreimer G (1995) G proteins and Ca²⁺-modulated protein kinases of a plasma membrane-enriched fraction and isolated eyespot apparatuses of Spermatozopsis similis (Chlorophyceae). Eur J Phycol 30: 319–330[CrossRef]

Schmidt JA, Eckert R (1976) Calcium couples flagellar reversal to photostimulation in Chlamydomonas reinhardtii. Nature 262: 713–715[CrossRef][Medline]

Schmidt M, Geßner G, Luff M, Heiland I, Wagner V, Kaminski M, Geimer S, Eitzinger N, Reißenweber T, Voytsekh O, et al (2006) Proteomic analysis of the eyespot of Chlamydomonas reinhardtii provides novel insights into its components and tactic movements. Plant Cell 18: 1908–1930[Abstract/Free Full Text]

Schmidt M, Luff M, Mollwo A, Kaminski M, Mittag M, Kreimer G (2007) Evidence for a specialized localization of the chloroplast ATP-synthase subunits alpha, beta, and gamma in the eyespot apparatus of Chlamydomonas reinhardtii (Chlorophyceae). J Phycol 43: 284–294[CrossRef][Web of Science]

Schweighofer A, Hirt H, Meskiene I (2004) Plant PP2C phosphatases: emerging functions in stress signaling. Trends Plant Sci 9: 236–243[CrossRef][Web of Science][Medline]

Shimada H, Koizumi M, Kuroki K, Mochizuki M, Fujimoto H, Ohta H, Masuda T, Takamiya K-I (2004) ARC3, a chloroplast division factor, is a chimera of prokaryotic FtsZ and part of a eukaryotic phosphatidylinositol-4-phosphate 5-kinase. Plant Cell Physiol 45: 960–967[Abstract/Free Full Text]

Shu H, Chen S, Bi Q, Mumby M, Brekken DL (2004) Identification of phosphoproteins and their phosphorylation sites in the WEHI-231 B lymphoma cell line. Mol Cell Proteomics 3: 279–286[Abstract/Free Full Text]

Shukolyukov SA (1999) Key reactions of visual cycle in vertebrates and invertebrates: activation and regeneration of rhodopsin. J Evol Biochem Physiol 35: 579–595[Web of Science]

Sineshchekov OA, Govorunova EA (2001) Electrical events in photomovement of green flagellated algae. In M Lebert, D-P Häder, eds, Comprehensive Series in Photosciences. Elsevier/North Holland, Amsterdam, pp 245–280

Sineshchekov OA, Jung K-H, Spudich JL (2002) Two rhodopsins mediate phototaxis to low- and high-intensity light in Chlamydomonas reinhardtii. Proc Natl Acad Sci USA 99: 8689–8694[Abstract/Free Full Text]

Stauber EJ, Fink A, Markert C, Kruse O, Johanningmeier U, Hippler M (2003) Proteomics of Chlamydomonas reinhardtii light-harvesting proteins. Eukaryot Cell 2: 978–994[Abstract/Free Full Text]

Suzuki T, Yamasaki K, Fujita S, Oda K, Iseki M, Yoshida K, Watanabe M, Daiyasu H, Toh H, Asamizu E, et al (2003) Archael-type rhodopsins in Chlamydomonas: model structure and intracellular localization. Biochem Biophys Res Commun 301: 711–717[CrossRef][Web of Science][Medline]

Takeshima H, Komazaki S, Nishi M, Iino M, Kangawa K (2000) Junctophilins: a novel family of junctional membrane complex proteins. Mol Cell 6: 11–22[CrossRef][Web of Science][Medline]

Turkina MV, Blanco-Rivero A, Vainonen JP, Vener AV, Villarejo A (2006a) CO₂ limitation induces specific redox-dependent protein phosphorylation in Chlamydomonas reinhardtii. Proteomics 6: 2693–2704[CrossRef][Web of Science][Medline]

Turkina MV, Kargul J, Blanco-Rivero A, Villarejo A, Barber J, Vener AV (2006b) Environmentally modulated phosphoproteome of photosynthetic membranes in the green alga Chlamydomonas reinhardtii. Mol Cell Proteomics 5: 1412–1425[Abstract/Free Full Text]

Vener AV, Harms A, Sussman MR, Vierstra RD (2001) Mass spectrometric resolution of reversible protein phosphorylation in photosynthetic membranes of Arabidopsis thaliana. J Biol Chem 276: 6959–6966[Abstract/Free Full Text]

von Heijne G (1992) Membrane protein structure prediction: hydrophobicity analysis and the ‘positive inside’ rule. J Mol Biol 225: 487–494[CrossRef][Web of Science][Medline]

Wagner V, Geßner G, Heiland I, Kaminski M, Hawat S, Scheffler K, Mittag M (2006) Analysis of the phosphoproteome of Chlamydomonas reinhardtii provides new insights into various cellular pathways. Eukaryot Cell 5: 457–468[Abstract/Free Full Text]

Witman GB (1993) Chlamydomonas phototaxis. Trends Cell Biol 3: 403–408[CrossRef][Medline]

Zylka MJ, Reppert SM (1999) Discovery of a putative heme binding protein family (SOUL/HBP) by two-tissue suppression subtractive hybridization and database searches. Brain Res Mol Brain Res 74: 175–181[Medline]

This article has been cited by other articles:

				C. Desplats, F. Mus, S. Cuine, E. Billon, L. Cournac, and G. Peltier Characterization of Nda2, a Plastoquinone-reducing Type II NAD(P)H Dehydrogenase in Chlamydomonas Chloroplasts J. Biol. Chem., February 13, 2009; 284(7): 4148 - 4157. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 146/2/772    most recent pp.107.109645v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (5) Citing Articles via Google Scholar Google Scholar Articles by Wagner, V. Articles by Kreimer, G. Search for Related Content PubMed PubMed Citation Articles by Wagner, V. Articles by Kreimer, G. Agricola Articles by Wagner, V. Articles by Kreimer, G.