TY - JOUR
AU1 - Liu,, Yan
AU2 - Peters, Winfried, S
AU3 - Froelich, Daniel, R
AU4 - Howell, Alexander, H
AU5 - Mooney,, Sutton
AU6 - Evans, James, E
AU7 - Hellmann, Hanjo, A
AU8 - Knoblauch,, Michael
AB - Abstract Forisomes are protein bodies known exclusively from sieve elements of legumes. Forisomes contribute to the regulation of phloem transport due to their unique Ca2+-controlled, reversible swelling. The assembly of forisomes from sieve element occlusion (SEO) protein monomers in developing sieve elements and the mechanism(s) of Ca2+-dependent forisome contractility are poorly understood because the amino acid sequences of SEO proteins lack conventional protein–protein interaction and Ca2+-binding motifs. We selected amino acids potentially responsible for forisome-specific functions by analyzing SEO protein sequences in comparison to those of the widely distributed SEO-related (SEOR), or SEOR proteins. SEOR proteins resemble SEO proteins closely but lack any Ca2+ responsiveness. We exchanged identified candidate residues by directed mutagenesis of the Medicago truncatula SEO1 gene, expressed the mutated genes in yeast (Saccharomyces cerevisiae) and studied the structural and functional phenotypes of the forisome-like bodies that formed in the transgenic cells. We identified three aspartate residues critical for Ca2+ responsiveness and two more that were required for forisome-like bodies to assemble. The phenotypes observed further suggested that Ca2+-controlled and pH-inducible swelling effects in forisome-like bodies proceeded by different yet interacting mechanisms. Finally, we observed a previously unknown surface striation in native forisomes and in recombinant forisome-like bodies that could serve as an indicator of successful forisome assembly. To conclude, this study defines a promising path to the elucidation of the so-far elusive molecular mechanisms of forisome assembly and Ca2+-dependent contractility. Introduction Vascular plants translocate the products of photosynthesis from organs of net production, the sources, to organs of net consumption or storage, the sinks (Knoblauch and Peters 2013, de Schepper et al. 2013, Savage et al. 2016). Transport occurs as cytoplasmic bulk flow in the sieve tubes of the phloem (Knoblauch et al. 2016), which consist of anucleate cells, the sieve elements (Behnke and Sjolund 1990). As suggested by Ernst Münch (1930), the mechanism of sieve tube transport depends on photosynthate accumulation driving osmotic water influx into the transport system in the sources, and photosynthate unloading leading to osmotic water efflux in the sinks (Knoblauch and Peters 2017). The mechanisms of flow regulation, however, and the function of several structural components of mature sieve elements still are largely obscure (Knoblauch and Peters 2010). For example, most sieve elements contain phloem-specific, or P-proteins (Cronshaw and Sabnis 1990, Sabnis and Sabnis 1995). A large family of P-proteins are the sieve element occlusion-related (SEOR) proteins, which form either irregular slime masses or so-called nondispersive protein bodies. P-proteins including SEORs are widely assumed to block sieve tubes in response to wounding (e.g. Ernst et al. 2012) and in the defense against phloem-feeding insects (e.g. Kunierczyk et al. 2008). Unfortunately, the evidence supporting these ideas is inconclusive (Knoblauch et al. 2014). Forisomes, contractile P-protein bodies that swell and shrink in dependence on the concentration of Ca2+ ions (Knoblauch and Peters 2004), may be an exception. After initial doubts about the capacity of swollen forisomes for blocking sieve tubes (Peters et al. 2006), forisomes were demonstrated to plug artificial microtubes of the size of natural sieve tubes when triggered by Ca2+ (Knoblauch et al. 2012). Interestingly, forisomes in Vicia faba responded to attacks by two generalist aphid species (Medina-Ortega and Walker 2015) but not to those by a host-specific specialist species (Walker and Medina-Ortega 2012). Moreover, forisomes in live plants transiently swell in response to turgor changes induced by extracellular osmotica (Knoblauch et al. 2001), suggesting that forisomes may function in the regulation of sieve tube flow independently of their probable role in the defense against phloem-feeding pests. Forisomes consist of sieve element occlusion (SEO) proteins, which closely resemble SEOR proteins but still are unambiguously distinguishable based on their amino acid sequences (Pélissier et al. 2008). Forisomes are restricted to the Fabaceae, and SEO proteins seem to have evolved from SEOR proteins in the last common ancestor of the family (Peters et al. 2010). Members of the Fabaceae examined so far possess several SEO genes each, and the proteins translated from these genes assemble into heteromeric forisomes (Pélissier et al. 2008, Zielonka et al. 2014). Forisome assembly appears driven by similar mechanisms in different species, as indicated by the ready integration of three different SEO proteins from Medicago trunculata into forisomes of V. faba and Glycine max (Pélissier et al. 2008). Forisome-like protein bodies may form in non-phloem plant cells (Zielonka et al. 2014) and in yeast (Groscurth et al. 2012) when SEO genes controlled by appropriate promoters are introduced into these cells. In this way, several SEO proteins could be prompted to form homomeric forisome-like bodies (Müller et al. 2010), suggesting that the heteromeric character of native forisomes is a nonessential feature. Homomeric forisome-like protein bodies swell in response to Ca2+ and alkaline pH (Müller et al. 2010, Zielonka et al. 2014). While the reported Ca2+ responses are much weaker than those observed in fully functional native forisomes under physiologically relevant conditions (Knoblauch et al. 2012), homomeric forisome-like bodies appear a useful experimental system for studying the molecular basis of SEO protein function. Intracellular movements and contractility generally are driven by ATP-dependent processes, such as the polymerization of actin filaments and microtubules and the motion of motor proteins along these fibrillar structures (Bray 2001). In contrast, Ca2+-driven mechanic activity is rare. It occurs especially in protozoa, most impressively in the flight responses mediated by the contractile spasmoneme of sessile ciliates (Mahadevan and Matsudaira 2000, Misra et al. 2010, Chung and Ryu 2017). Ca2+-dependent cellular actuation relies on centrins and spasmins, proteins characterized by their Ca2+-binding motifs of the EF-hand family (Gogendeau et al. 2007, Zhang and He 2012, Aubusson-Fleury et al. 2017, Ryu et al. 2017). In remarkable contrast, characteristic motifs of conventional divalent cation-binding sites such as EF hands or zinc fingers (Arias-Moreno et al. 2011, Yáñez et al. 2012) are absent from SEO proteins (Pélissier et al. 2008, Rüping et al. 2010). As an interesting possibility, Ca2+-binding sites in forisomes could be formed by amino acids of more than one peptide chain, which would make them undetectable by conventional sequence analysis. Due to these problems, our understanding of the mechanism(s) of Ca2+-dependent forisome swelling and contraction has remained on a theoretical level for over a decade (Pickard et al. 2006). We tackled the problem of Ca2+ binding to SEO proteins in forisomes by an alternative approach. Rather than searching for sequence similarities with known domains, we compared amino acid sequences of SEO and SEOR proteins that had been characterized in living plants, aiming at the identification of residues that were differentially conserved in one protein group but not in the other. We reasoned that differentially conserved amino acids had an increased probability to be involved in specific protein functions, for instance the Ca2+ responsiveness of SEO proteins that is absent from SEOR proteins. Following selective mutation of some of these residues, we observed characteristic phenotypes in homomeric forisome-like bodies assembled from mutated Medicago truncatula SEO1 proteins in transformed yeast. As a result, we identified several amino acids critical for the assembly of the protein bodies and their responsiveness to Ca2+ and alkaline pH. Results Comparative analysis of SEO and SEOR amino acid sequences To identify promising candidate amino acids of functional significance, we compared SEO and SEOR proteins that had been characterized structurally and functionally in live plants. The selection comprised eight forisome-forming SEO proteins (MtSEO1–3; GmSEO1–4; VfSEO1) and six SEOR proteins, of which five were slime-forming SEORs (AtSEOR1 and -2; NtSEOR1 and -2; CmSEOR1), while one was a SEOR involved in the formation of nondispersive protein bodies (PtSEOR1). The amino acid sequences of these 14 proteins are available in Supplementary Data S1. Comparison with average amino acid frequencies in plant proteins (extracted from the UniProtKB/Swiss-Prot database) revealed characteristic patterns of deviation from the average in both SEO and SEOR proteins (Fig. 1; see also Supplementary Tables S2, S3). Most importantly, SEO and SEOR proteins differed significantly (P < 0.01) from each other in the proportions of six amino acids, with aspartate, lysine, and asparagine enriched in SEO as compared to SEOR proteins (Fig. 1). This hinted at possible roles of these residues in SEO-specific functions, e.g. involvement of the anionic aspartate in Ca2+ binding. Fig. 1 Open in new tabDownload slide Comparison of the amino acid compositions of cytologically characterized SEOR proteins (orange bars; from left to right AtSEOR1, AtSEOR2, NtSEOR1, NtSEOR2, CmSEOR1, PtSEOR1) and SEO proteins (blue bars; from left to right MtSEO1, MtSEO2, MtSEO3, VfSEO1, GmSEO1, GmSEO2, GmSEO3, GmSEO4) with the average composition of all reviewed plant proteins in the UniProtKB/Swiss-Prot database. Bar height indicates the factor by which the proportion of an amino acid in a given protein deviates from the average proportion of that amino acid (value 1, highlighted by gray line, indicates no deviation). Significant enrichment and depletion (P < 0.01) of an amino acid in the SEOR or SEO groups relative to the average composition are marked by letters e and d, respectively, at the bottom of the bars. Significant differences (P < 0.01) between the proportions of an amino acid in the SEOR and SEO protein groups are highlighted by asterisks above the bars. Fig. 1 Open in new tabDownload slide Comparison of the amino acid compositions of cytologically characterized SEOR proteins (orange bars; from left to right AtSEOR1, AtSEOR2, NtSEOR1, NtSEOR2, CmSEOR1, PtSEOR1) and SEO proteins (blue bars; from left to right MtSEO1, MtSEO2, MtSEO3, VfSEO1, GmSEO1, GmSEO2, GmSEO3, GmSEO4) with the average composition of all reviewed plant proteins in the UniProtKB/Swiss-Prot database. Bar height indicates the factor by which the proportion of an amino acid in a given protein deviates from the average proportion of that amino acid (value 1, highlighted by gray line, indicates no deviation). Significant enrichment and depletion (P < 0.01) of an amino acid in the SEOR or SEO groups relative to the average composition are marked by letters e and d, respectively, at the bottom of the bars. Significant differences (P < 0.01) between the proportions of an amino acid in the SEOR and SEO protein groups are highlighted by asterisks above the bars. Next, we aligned the sequences of our sample of 14 previously characterized proteins and found 67 residues highly conserved in the entire set (i.e. present at the same position in at least 13 of the sequences; Fig. 2). More interesting for our purposes, 77 residues were differentially conserved in SEO proteins (defined as being present in at least seven of the eight SEO proteins but absent from all SEOR proteins) while 59 residues were differentially conserved in SEOR proteins (present in at least five of the six SEOR proteins but absent from all SEO proteins). In both SEO and SEOR proteins, cysteine and particularly tryptophan were much more frequently found among the conserved residues than would have been expected based on their proportion among the amino acids in the respective group of proteins (Supplementary Fig. S4). Fig. 2 Open in new tabDownload slide Amino acid sequence alignment of cytologically characterized SEOR (six cases: Nt, Nicotiana tabaccum; At, Arabidopsis thaliana; Cm, Cucurbita maxima; Pt, Populus trichocarpa) and SEO proteins (eight cases: Mt, Medicago truncatula; Vf, Vicia faba; Gm, Glycine max). Positions 1–130 and 907–926 of the alignment were dominated by gaps and are omitted. Residues conserved in more than 10 proteins appear in the increasingly dark print, with residues conserved in all proteins shown in black. Residues differentially conserved in SEO proteins (present in at least seven SEOs but not in SEORs) appear in blue, while residues differentially conserved in SEOR proteins (present in at least five SEORs but not in SEOs) are marked orange. The two negatively charged residues conserved in proteins that form part of protein bodies (SEO proteins and PtSEOR1) appear in negative print. Fig. 2 Open in new tabDownload slide Amino acid sequence alignment of cytologically characterized SEOR (six cases: Nt, Nicotiana tabaccum; At, Arabidopsis thaliana; Cm, Cucurbita maxima; Pt, Populus trichocarpa) and SEO proteins (eight cases: Mt, Medicago truncatula; Vf, Vicia faba; Gm, Glycine max). Positions 1–130 and 907–926 of the alignment were dominated by gaps and are omitted. Residues conserved in more than 10 proteins appear in the increasingly dark print, with residues conserved in all proteins shown in black. Residues differentially conserved in SEO proteins (present in at least seven SEOs but not in SEORs) appear in blue, while residues differentially conserved in SEOR proteins (present in at least five SEORs but not in SEOs) are marked orange. The two negatively charged residues conserved in proteins that form part of protein bodies (SEO proteins and PtSEOR1) appear in negative print. The density of conserved amino acids was higher in the C-terminal than in the N-terminal half of the sequences. This was true for SEO and SEOR proteins evaluated separately and for the combined sample (Fig. 2). Since we were hoping to find hints at the Ca2+-binding mechanism in forisomes, we decided to target anionic residues, particularly aspartic acid (D), that were differentially conserved in the C-terminal half of SEO proteins, and other conserved residues in their immediate environment. We selected aspartate at alignment positions 641 and 708 and within three conserved motifs for site-directed mutation experiments (Fig. 3; to see how alignment positions correspond to positions in the MtSEO1 sequence, consult Supplementary Table S5). Fig. 3 Open in new tabDownload slide Parts of the sequence alignment of SEOR and SEO proteins, showing amino acids targeted by site-directed mutagenesis of the MtSEO1 gene on a red background. Color coding as in Fig. 2; residues conserved in both SEOR and SEO proteins appear dark, while residues differentially conserved in SEO and SEOR proteins are shown in blue and orange, respectively. Two acidic residues strictly conserved in SEO proteins and the protein body-forming PtSEOR1 appear in negative print. Fig. 3 Open in new tabDownload slide Parts of the sequence alignment of SEOR and SEO proteins, showing amino acids targeted by site-directed mutagenesis of the MtSEO1 gene on a red background. Color coding as in Fig. 2; residues conserved in both SEOR and SEO proteins appear dark, while residues differentially conserved in SEO and SEOR proteins are shown in blue and orange, respectively. Two acidic residues strictly conserved in SEO proteins and the protein body-forming PtSEOR1 appear in negative print. Structure of MtSEO1-based, recombinant forisome-like bodies Following Groscurth et al. (2012), we set up a yeast expression system for recombinant forisome-like bodies by transforming Saccharomyces cerevisiae with vectors containing MtSEO1 and MtSEO1 N-terminally fused to the eYFP gene under the control of the constitutive glyceraldehyde-3-phosphate dehydrogenase (GPD) promoter. The transformed yeast cells produced elongate forisome-like bodies that exhibited eYFP fluorescence (Fig. 4a). Length (L) and width (W) of most protein bodies were 5–9 and 0.7–1.6 µm, respectively, resulting in aspect ratios (W L−1) typically between 0.13 and 0.23 (for the original measurements, see Supplementary Table S6). The cross-sectional shape of the bodies resembled rounded squares (Fig. 4b), and the estimated volumes (L W2) lay between 3 and 20 µm3 in most cases. Fig. 4 Open in new tabDownload slide Structure of recombinant forisome-like bodies assembled from wild-type MtSEO1 combined with eYFP-tagged MtSEO1 in the yeast, S. cerevisiae. (a) Confocal micrograph of a yeast cell containing a fluorescent forisome-like body. (b) Isolated forisome-like body attached to the tip of a glass micropipette, seen in bright-field (top) and fluorescence (bottom) micrographs. In the fluorescence micrographs, the direction of view is perpendicular to the body’s long axis (left) and along the axis (right). (c) SEM micrograph of an isolated forisome-like body exhibiting transverse surface striations of 45–50 nm periodicity. The scale bar covers 40 ridges. Fig. 4 Open in new tabDownload slide Structure of recombinant forisome-like bodies assembled from wild-type MtSEO1 combined with eYFP-tagged MtSEO1 in the yeast, S. cerevisiae. (a) Confocal micrograph of a yeast cell containing a fluorescent forisome-like body. (b) Isolated forisome-like body attached to the tip of a glass micropipette, seen in bright-field (top) and fluorescence (bottom) micrographs. In the fluorescence micrographs, the direction of view is perpendicular to the body’s long axis (left) and along the axis (right). (c) SEM micrograph of an isolated forisome-like body exhibiting transverse surface striations of 45–50 nm periodicity. The scale bar covers 40 ridges. In the SEM, yeast-derived forisome-like bodies exhibited a regular, transverse surface striation of 45–50 nm periodicity (Fig. 4c). Because this feature had not been reported before, we isolated forisomes from our established experimental species, the sword bean (Canavalia gladiata), and prepared them for SEM observation in the same way (Fig. 5a). These native forisomes also showed a faint but unmistakable transverse striation with a periodicity of about 45 nm (Fig. 5b). Fig. 5 Open in new tabDownload slide Surface structure of native forisomes. (a) Tailed forisome isolated from Canavalia gladiata with the main body of 25 µm length. (b) Surface of the forisome’s main body at higher magnification, showing a transverse striation of about 45 nm periodicity underlying the more pronounced longitudinal groves. Fig. 5 Open in new tabDownload slide Surface structure of native forisomes. (a) Tailed forisome isolated from Canavalia gladiata with the main body of 25 µm length. (b) Surface of the forisome’s main body at higher magnification, showing a transverse striation of about 45 nm periodicity underlying the more pronounced longitudinal groves. Two aspartate residues are critical for forisome assembly Replacing the aspartate residues at alignment position 708 and at position 542 in motif I by alanine (D708A and D542A, respectively) abolished the ability of MtSEO1 produced in yeast to assemble into forisome-like bodies. The proteins formed more or less spherical masses instead (Table 1, Fig. 6b, c). When motif I was more severely modified by the exchange of five conserved residues (Q527A–K539A–K540A–D542A–K544A), the protein became soluble in the cytosol (Fig. 6d). Evidently, D708, D542 and probably other residues in motif I play important roles in the spontaneous formation of forisome-like bodies. These residues might still function in Ca2+ binding, but the possibility obviously could not be evaluated by studying Ca2+-dependent swelling responses in mutant protein bodies. Fig. 6 Open in new tabDownload slide Fluorescence micrographs of yeast cells producing mutated MtSEO1 proteins. (a) D641A is an example of a mutation that allows the protein to assemble into normal, elongate forisome-like bodies. In contrast, the mutation D708A causes the protein to form amorphous masses in the cells (b), and a similar phenotype is produced by D542A in motif I (c). The quintuple mutation in motif I, Q527A–K539A–K540A–D542A–K544A, renders the protein soluble in the cytosol (d). The scale bar in (b) applies to all micrographs. Fig. 6 Open in new tabDownload slide Fluorescence micrographs of yeast cells producing mutated MtSEO1 proteins. (a) D641A is an example of a mutation that allows the protein to assemble into normal, elongate forisome-like bodies. In contrast, the mutation D708A causes the protein to form amorphous masses in the cells (b), and a similar phenotype is produced by D542A in motif I (c). The quintuple mutation in motif I, Q527A–K539A–K540A–D542A–K544A, renders the protein soluble in the cytosol (d). The scale bar in (b) applies to all micrographs. Table 1 Overview of the phenotypes of forisome-like bodies consisting of MtSEO1 wild-type or single and double mutant proteins Motif (compare Fig. 3) . Mutations . Assembly . Striation . Ca2+ response . pH response . Alignment ID . MtSEO1 ID . 1 mM . 5 mM . 10 mM . Wild-type MtSEO1 Yes Yes + + – D641A D420A Yes Yes + + – D708A D481A No Motif I D542A D334A No Motif II D783A, C784A D517A, C518A Yes Yes (+) + Q787A, D799A Q521A, D533A Yes Yes + (+) D783A, D799A D517A, D533A Yes No − − E785A E519A Yes Yes + + Motif III E840A E573A Yes Yes + + D843A D576A Yes Yes − (+) + E848A E581A Yes Yes + + Motif (compare Fig. 3) . Mutations . Assembly . Striation . Ca2+ response . pH response . Alignment ID . MtSEO1 ID . 1 mM . 5 mM . 10 mM . Wild-type MtSEO1 Yes Yes + + – D641A D420A Yes Yes + + – D708A D481A No Motif I D542A D334A No Motif II D783A, C784A D517A, C518A Yes Yes (+) + Q787A, D799A Q521A, D533A Yes Yes + (+) D783A, D799A D517A, D533A Yes No − − E785A E519A Yes Yes + + Motif III E840A E573A Yes Yes + + D843A D576A Yes Yes − (+) + E848A E581A Yes Yes + + Mutated residues are identified by their numbers in the sequence alignment (Fig. 2) as well as in the MtSEO1 sequence (Supplementary Table S5). Structural phenotypes listed are the assembly of forisome-like bodies from the monomeric proteins in transgenic yeast, and the occurrence of transverse surface striations on these bodies. Functional phenotypes are the ability of the bodies to respond to Ca2+ (1, 5 or 10 mM) and to pH 10.3 by changes in their aspect ratios and estimated volumes. Responses that were significantly weaker than those in the wild type appear in parentheses (for quantitative assessments, see Fig. 7). pH 10.3 caused irreversible disintegration of Q787A–D799A protein bodies, which showed a much-reduced response to pH 9.8. Open in new tab Table 1 Overview of the phenotypes of forisome-like bodies consisting of MtSEO1 wild-type or single and double mutant proteins Motif (compare Fig. 3) . Mutations . Assembly . Striation . Ca2+ response . pH response . Alignment ID . MtSEO1 ID . 1 mM . 5 mM . 10 mM . Wild-type MtSEO1 Yes Yes + + – D641A D420A Yes Yes + + – D708A D481A No Motif I D542A D334A No Motif II D783A, C784A D517A, C518A Yes Yes (+) + Q787A, D799A Q521A, D533A Yes Yes + (+) D783A, D799A D517A, D533A Yes No − − E785A E519A Yes Yes + + Motif III E840A E573A Yes Yes + + D843A D576A Yes Yes − (+) + E848A E581A Yes Yes + + Motif (compare Fig. 3) . Mutations . Assembly . Striation . Ca2+ response . pH response . Alignment ID . MtSEO1 ID . 1 mM . 5 mM . 10 mM . Wild-type MtSEO1 Yes Yes + + – D641A D420A Yes Yes + + – D708A D481A No Motif I D542A D334A No Motif II D783A, C784A D517A, C518A Yes Yes (+) + Q787A, D799A Q521A, D533A Yes Yes + (+) D783A, D799A D517A, D533A Yes No − − E785A E519A Yes Yes + + Motif III E840A E573A Yes Yes + + D843A D576A Yes Yes − (+) + E848A E581A Yes Yes + + Mutated residues are identified by their numbers in the sequence alignment (Fig. 2) as well as in the MtSEO1 sequence (Supplementary Table S5). Structural phenotypes listed are the assembly of forisome-like bodies from the monomeric proteins in transgenic yeast, and the occurrence of transverse surface striations on these bodies. Functional phenotypes are the ability of the bodies to respond to Ca2+ (1, 5 or 10 mM) and to pH 10.3 by changes in their aspect ratios and estimated volumes. Responses that were significantly weaker than those in the wild type appear in parentheses (for quantitative assessments, see Fig. 7). pH 10.3 caused irreversible disintegration of Q787A–D799A protein bodies, which showed a much-reduced response to pH 9.8. Open in new tab Aspartate residues in motifs II and III are essential for pH and Ca2+ responsiveness Isolated forisome-like bodies consisting of the wild-type MtSEO1 responded to Ca2+ and alkaline pH by contracting longitudinally and expanding transversely, resulting in increased aspect ratios and estimated volumes. Deformations induced by Ca2+, which developed to the full extent when the protein bodies were exposed to 1 mM Ca2+, were less pronounced than those caused by alkaline pH (Fig. 7a, d). The pH-induced swelling developed gradually over a range of pH 9.3–10.6, and we chose pH 10.3 as a standard for subsequent tests in protein bodies assembled from mutated forms of MtSEO1. Fig. 7 Open in new tabDownload slide Ca2+- and pH-induced deformations in isolated forisome-like bodies. (a) Fluorescence micrographs of a wild-type MtSEO1 protein body in Ca2+-free medium (top) and after transfer to 1 mM Ca2+ at constant pH 7.2 (bottom) show the typical Ca2+-induced longitudinal shortening and radial swelling (the scale bar applies to all micrographs). This response was quantified as the change in aspect ratio (AR) (b) and estimated volume (c) of protein bodies assembled from wild-type and various mutated MtSEO1 proteins, as indicated on top of (b); ≥20 forisome-like bodies of each protein type were tested. Changes were normalized by setting initial values to 1 (horizontal gray lines), and the distribution of the factors by which AR and volume changed are visualized as box-plots (boxes represent the 2nd and 3rd quartiles, and whiskers the 5th and 95th percentiles of the distributions). Datasets in which the scatter of final values around the initial value 1 did not indicate a consistent increase or decrease in aspect ratio or volume (P > 0.001 in two-tailed Wilcoxon signed-rank tests) are shown in red. The Ca2+ effect on volume in the D783A–D799A mutant (blue) was consistent, but it was a decrease rather than an increase as in the wild type. Asterisks mark distributions that appeared significantly different from the wild-type response (P ≤ 0.001, two-tailed Mann–Whitney test). (d) Wild-type forisome-like body in Ca2+-free medium at pH 7.2 (top) and after transfer to pH 10.3 (bottom). pH effects were analyzed as described for Ca2+ responses above; relative changes in AR and estimated volume are displayed in (e) and (f), respectively. Original data are available in Supplementary Table S6. Fig. 7 Open in new tabDownload slide Ca2+- and pH-induced deformations in isolated forisome-like bodies. (a) Fluorescence micrographs of a wild-type MtSEO1 protein body in Ca2+-free medium (top) and after transfer to 1 mM Ca2+ at constant pH 7.2 (bottom) show the typical Ca2+-induced longitudinal shortening and radial swelling (the scale bar applies to all micrographs). This response was quantified as the change in aspect ratio (AR) (b) and estimated volume (c) of protein bodies assembled from wild-type and various mutated MtSEO1 proteins, as indicated on top of (b); ≥20 forisome-like bodies of each protein type were tested. Changes were normalized by setting initial values to 1 (horizontal gray lines), and the distribution of the factors by which AR and volume changed are visualized as box-plots (boxes represent the 2nd and 3rd quartiles, and whiskers the 5th and 95th percentiles of the distributions). Datasets in which the scatter of final values around the initial value 1 did not indicate a consistent increase or decrease in aspect ratio or volume (P > 0.001 in two-tailed Wilcoxon signed-rank tests) are shown in red. The Ca2+ effect on volume in the D783A–D799A mutant (blue) was consistent, but it was a decrease rather than an increase as in the wild type. Asterisks mark distributions that appeared significantly different from the wild-type response (P ≤ 0.001, two-tailed Mann–Whitney test). (d) Wild-type forisome-like body in Ca2+-free medium at pH 7.2 (top) and after transfer to pH 10.3 (bottom). pH effects were analyzed as described for Ca2+ responses above; relative changes in AR and estimated volume are displayed in (e) and (f), respectively. Original data are available in Supplementary Table S6. To determine the effects of the mutation of selected amino acids on the Ca2+ and pH responses of recombinant protein bodies, we measured the length and width before and after any response in at least 20 isolated bodies for each protein type (original data are available in Supplemental Table S6). The results were evaluated in two steps. First, for every protein type, the consistency of the shifts in aspect ratio and estimated volume was evaluated using the Wilcoxon signed-ranks test (two-tailed). Second, the magnitude of the observed changes in mutant protein bodies was compared to that exhibited by wild-type protein bodies by the Mann–Whitney test (two-tailed). Before this test, data were normalized by setting aspect ratios and estimated volumes before the response to 1, to allow for quantitative comparisons between the responses of protein bodies that differed in size. We adopted a relatively low P-value, 0.001, as our criterion for the potential significance of any observed differences. Among the analyzed single-mutant proteins that assembled into forisome-like bodies, D641A, E785A in motif II and E840A and E848A in motif III exhibited wild-type-like responses to 1 mM Ca2+ and alkaline pH (Table 1, Fig. 7b, c, e, f). In contrast, D843A in motif III caused a significant decrease in the sensitivity to Ca2+. The aspect ratios and estimated volumes of D843A protein bodies did not respond consistently to 5 mM Ca2+, whereas the aspect ratio but not the estimated volume did under exposure to 10 mM Ca2+, 10 times the concentration that evoked the full effect in wild-type forisome-like bodies (Fig. 7b, c). The pH response appeared unimpaired in this mutant (Fig. 7e, f). The situation was more complex in motif II. When the aspartate at alignment position 783 was replaced together with the neighboring cysteine (D783A–C784A), wild-type-like pH responses (Fig. 7e, f) and somewhat reduced Ca2+ responses (Fig. 7b, c) were observed. In contrast, when D799 was replaced in addition to D783 (D783A–D799A), both the Ca2+ and pH responses were lost (Fig. 7b, c, e, f). There was no consistent response of the aspect ratio even to 10 mM Ca2+ (Fig. 7b), while the estimated volume appeared to decrease rather than increase (Fig. 7c). Intriguingly, the responsiveness to 1 mM Ca2+ remained unaffected by D799A in combination with Q787A (D799A–Q787A; Fig. 7b, c). However, these protein bodies disintegrated when exposed to pH values above 10, while very slight increases in aspect ratio and estimated volume were evoked by pH 9.8 (Fig. 7e, f). Surface striations in forisome-like bodies Fine surface striations as in wild-type forisome-like bodies (Fig. 4c) were also found in all protein bodies that assembled from mutated MtSEO1 proteins, with one exception. The Ca2+- and pH-insensitive protein bodies carrying the D783A–D799A double mutation lacked striations (Table 1, Fig. 8). Fig. 8 Open in new tabDownload slide Scanning electron micrographs highlighting surface striations or the lack thereof in forisome-like bodies produced in yeast. (a) Surface striations on a protein body assembled from MtSEO1 with the Q787A–D799A double mutation. (b) Forisome-like bodies assembled from the protein with the D783A–D799A double mutation lack surface striations. (c) Striations occur on protein bodies formed from MtSEO1 with the D843A mutation. The scale bar in (a) applies to all subfigures. Fig. 8 Open in new tabDownload slide Scanning electron micrographs highlighting surface striations or the lack thereof in forisome-like bodies produced in yeast. (a) Surface striations on a protein body assembled from MtSEO1 with the Q787A–D799A double mutation. (b) Forisome-like bodies assembled from the protein with the D783A–D799A double mutation lack surface striations. (c) Striations occur on protein bodies formed from MtSEO1 with the D843A mutation. The scale bar in (a) applies to all subfigures. Discussion Attempting to identify amino acids involved in the Ca2+ responsiveness of forisomes, we specifically targeted residues that were differentially conserved in SEO as compared to SEOR proteins. The validity of this approach is demonstrated by our finding of the critical function of several amino acids not only in the Ca2+ response but also in the reaction to alkaline pH and in the assembly of forisome-like bodies. Ca2+ responses in MtSEO1-based forisome-like bodies The conservation in close vicinity of aspartate and cysteine residues, known for their potential to participate in the coordination of cations (Arias-Moreno et al. 2011), had suggested motif II as an interesting target (Fig. 3). D799 in motif II was separated from a strictly conserved cysteine–leucine–lysine group by a single amino acid, either glutamine or arginine. Another aspartate residue, D783, occurred as part of a conserved DCEIQ sequence (positions 783–787) in all SEO proteins except MtSEO3. When the two aspartate residues were mutated together (D783A–D799A), responsiveness even to 10 mM Ca2+ was lost (Fig. 7b, c). Replacement of any of the two as such did not have this effect, as the double mutants D783A–C784A and Q787A–D799A consistently responded to 1 mM Ca2+ (Fig. 7b, c). Evidently, residues of motif II functioned cooperatively in mediating Ca2+ responses. Replacing the aspartate residue with alignment number D843 in motif III by alanine abolished the ability of the protein bodies to swell in response to Ca2+ concentrations below 10 mM (Table 1, Fig. 7b, c). We had chosen motif III because it included the anionic D843 and E848, the only two residues conserved in all SEO proteins and also in PtSEOR1, but not in other SEORs (Fig. 3). This was interesting because PtSEOR1 was the only SEOR protein in our sample known to participate in the formation of nondispersive protein bodies. Importantly, native protein bodies consisting of PtSEOR1 do not respond to Ca2+ (Mullendore et al. 2018). These facts indicated that D843 cooperates with other residues conserved in SEO proteins but absent from PtSEOR1 in mediating the Ca2+ response. It seems of interest in this context that several additional, not strictly conserved charged residues were present in the vicinity of D843 in all SEO proteins (e.g. K839, E840, K847 and D851 in MtSEO1). Moreover, positions 842, 849, and 850 in all SEO proteins were occupied by aromatic residues, which may connect to positively charged partners through cation–π interactions (Gallivan and Dougherty 1999). A previous hypothesis had suggested a ‘potential thioredoxin fold’ as a possible Ca2+-binding structure in forisome proteins. In a broad in silico analysis, this structure had been detected in all SEO and SEOR protein sequences examined, most of which represented putative proteins derived from genome data (Rüping et al. 2010; somewhat confusingly, these authors called SEO proteins ‘SEO-F’ and SEOR proteins ‘SEO’). Related domains exist in calsequestrins (Wang et al. 1998), proteins that bind large amounts of Ca2+ and buffer its concentration in the sarcoplasmic reticulum of muscle (Royer and Ríos 2009). Rüping et al. (2010) therefore hypothesized that the ‘potential thioredoxin folds’ in SEO and SEOR proteins might be involved in Ca2+ binding, similarly as their apparent counterparts are in calsequestrin. The idea has various problems. First, in contrast to forisome-forming SEO proteins, SEOR proteins are unresponsive to Ca2+ (Knoblauch et al. 2014, Mullendore et al. 2018). Therefore, the conservation of motifs or domains in SEOR as well as SEO proteins does not speak, by itself, for the involvement of these structures in Ca2+ binding. Second, Ca2+ binding in calsequestrins depends on acidic amino acids, which contribute over 27% of all residues in these proteins (Yano and Zarain-Herzberg 1994). While the proportion of acidic residues is increased in SEO proteins compared to plant proteins in general (Pélissier et al. 2008), it never exceeds 15% in SEOs actually studied in live cells (Supplementary Table S2). Third, acidic residues tend to form clusters in calsequestrins, which may act as Ca2+ sensors and in Ca2+ binding (Kumar et al. 2013). SEO and SEOR proteins do not possess similar structural features. Fourth, among the numerous cations that efficiently compete with Ca2+ for binding on calsequestrin, Mg2+ and Sr2+ show similar affinities (Beard et al. 2004). On the contrary, forisomes respond to Ca2+ and Sr2+, but not to Mg2+ (Knoblauch et al. 2001, Knoblauch et al. 2003). Fifth, calsequestrin polymerizes in the presence of Ca2+ while it depolymerizes when Ca2+ is depleted in situ (Perni et al. 2013, Manno et al. 2017), and it assembles into highly ordered crystals when Ca2+ is present in vitro (Maurer et al. 1985). In forisomes, Ca2+ binding has the opposite effects. Only when Ca2+ is absent do SEO proteins arrange into dense arrays of crystalline character, as indicated by their birefringence (Peters et al. 2007b). Ca2+ binding to SEO proteins induces their dispersion into irregular, non-birefringent masses of low molecular order (Knoblauch et al. 2001, Peters et al. 2007b). In conclusion, the interactions of calsequestrin with Ca2+ appear unlikely to provide a valid model for SEO action in forisomes. Neither can our present results serve as supporting evidence for the hypothesis because the aspartate residues on alignment positions 783, 799 and 843, which we have identified as essential for full Ca2+ responsiveness (Fig. 7), are located outside of the postulated ‘potential thioredoxin folds’, which correspond roughly to our alignment positions 535–640 (Rose et al. 2020). pH responses and their relation to Ca2+ responsiveness Ca2+ effects on forisome geometry can, to some degree, be mimicked by pH values above and below the physiological range (Knoblauch et al. 2003). Forisome-like bodies from transformed yeast showed changes in their aspect ratios in response to pH 10.3 that were much stronger than those evocable by Ca2+ (Fig. 7). Intriguingly, the loss of Ca2+ responsiveness was accompanied by a loss of pH responsiveness in the double mutant D783A–D799A, whereas the mutant D843A, which had exhibited reduced sensitivity to Ca2+, showed a wild-type-like response to pH 10.3. This supported the idea that Ca2+ and high pH affect forisome conformation through different mechanisms, which may partially depend on the same amino acids. The interaction between the two stimuli can be complex, as suggested by the phenotypes resulting from the double mutations we generated in motif II (Fig. 7). Despite the unambiguous effects of D783A–D799A on the responsiveness to Ca2+, the mutation of any of the two aspartates by itself did not seem to interfere with the protein bodies’ ability to undergo Ca2+-induced deformation, as shown by the D783A–C784A and Q787A–D799A mutants (Fig. 7b, c). Therefore, it seems that part(s) of the mechanism(s) underlying the Ca2+ response is redundant, allowing the response to occur without D783 or D799, but not without both. Similarly, mutating D783 as such did not impair the responsiveness to pH 10.3 (see the D783A–C784A mutant in Fig. 7e, f), suggesting that the abolishment of the pH responsiveness in the D783A–D799A mutant might be due to the mutation of D799. On the other hand, the Q787A–D799A double mutation affected the stability of the protein bodies, which disintegrated at pH values above 10. Apparently, either Q787 is essential for protein body stability or D799 is. The latter seems less likely because no pH-dependent stability effect occurred in D783A–D799A mutants. In any case, motif II clearly is a sequence element in which various functions—Ca2+ responsiveness, pH responsiveness, stability of the protein body—overlap mechanistically, and which therefore calls for in-depth characterization in the future. Assembly of homomeric MtSEO1-based forisome-like protein bodies We identified two aspartate residues that were critical for the assembly of the protein bodies. For one of them, D708 (Fig. 6b), we could hardly have predicted this finding, while for the other, D542 (Fig. 6c), the result was not completely unexpected. This was because D542 in motif I (replaced by glutamate, E542, in GmSEO3) was neighbored by another anionic amino acid (E541) on its N-terminal side in six of the eight SEO proteins (Fig. 3). Its C-terminal neighbor was a strictly conserved phenylalanine (F543), which may form cation–π linkages with cationic groups. These three amino acids, (E)DF, were sandwiched between three strictly conserved cationic lysine moieties, K539 (replaced by arginine, R539, in MtSEO3), K540 and K544, with another lysine, K535, a few positions away. Thus, motif I was characterized by a high density of charged residues, and the resulting hydrophilicity and electrostatic interactions could be expected to be important for the structure of this part of the protein. The relatively small modification brought about by the D542A mutation abolished the protein’s ability to assemble normally (Fig. 6c). Intriguingly, the more radical quintuple mutation Q527A–K539A–K540A–D542A–K544A rendered the protein soluble (Fig. 6d) despite the decreased proportion of hydrophilic residues in the mutant protein. These findings are noteworthy as no conventional protein–protein interaction motifs have been identified in SEOs so far. Rose et al. (2020) recently analyzed protein–protein interactions involving randomly mutagenized segments and full-length MtSEO1. They detected 43 ‘conserved sites’ that were apparently necessary for MtSEO1 dimerization among the 267 N-terminal amino acids, and 287 ‘conserved amino acids’ that the authors deemed critical for forisome assembly in the full-length sequence of 647 residues. Unexpectedly, the smaller of the two sets was only partially included in the larger one, so that the total count of ‘conserved’ amino acids required for the formation of homomeric forisome-like bodies from monomeric MtSEO1 amounted to 299, or 46% of the complete sequence (Supplementary Table S5). Such a large proportion seems difficult to interpret in functional terms. Moreover, of the residues that we have identified as differentially conserved in SEO proteins (Fig. 2; Supplementary Table S5), only 51% are among the ‘conserved amino acids’ of Rose et al. (2020). This proportion differs insignificantly from that in the complete sequence (P = 0.62, Pearson’s χ2 test). In other words, there is no correlation between the amino acids identified as ‘conserved’ based on the random-mutagenesis approach applied to MtSEO1 by Rose et al. (2020) on the one hand and those identified as conserved by conventional sequence alignment (Fig. 2) on the other hand. In this context, the fact that D542 and D708, critical amino acids for protein body assembly in our tests, were among the ‘conserved’ residues of Rose et al. (2020; compare our Supplementary Table S5) has limited significance. With 46% of all amino acids ‘conserved’, the probability that both members of any randomly chosen pair of residues possess this attribute is P = 0.21. Ultrastructure of forisome-like bodies Surface striations of 45–50 nm periodicity were an unexpected feature of wild-type forisome-like bodies produced in yeast (Fig. 4c). With the exception of the D783A–D799A double mutant, we consistently found these striations in all mutants (Table 1, Fig. 8) and in native forisomes from C. gladiata (Fig. 5). To produce high-resolution SEM images of forisome-like bodies, we had applied unusually thin coats of platinum–palladium and later did the same when we searched for surface striations in native forisomes. We assume that this is the reason why similar striations have not been detected in previous SEM studies of forisome ultrastructure. Since the D783A–D799A double mutation did not only abolish surface striations (Fig. 8b) but also caused a loss of pH and Ca2+ responsiveness (Table 1, Fig. 7), one might conclude that internal features manifesting themselves as surface striations represent structural requirements for the protein bodies’ ability to respond to alkaline pH and Ca2+. The structural feature as such is certainly not sufficient, though, as neither the decreased Ca2+ sensitivity in the D843A mutant (Fig. 7b, c) nor the strong reduction in the pH response in the Q787A–D799A double mutant (Fig. 7e, f) was linked to a loss of surface striations (Table 1, Fig. 8a, c). As has long been known, SEO proteins form fibrils that in condensed forisomes are oriented parallel with the forisome’s long axis. These fibrils are tightly co-aligned as indicated by a prominent, regular cross-striation of 12–15 nm periodicity (Laflèche 1966, Palevitz and Newcomb 1971, Lawton 1978, Knoblauch et al. 2001). Whether there is a structural connection linking this internal molecular order to the regular surface pattern on forisomes remains to be established. Materials and Methods Plant material Canavalia gladiata plants were grown in a greenhouse in 20-l pots at 23°C, 60–70% relative humidity, and a 14/10-h light/dark schedule (daylight plus lamp light; model PL 90; P.L. Light Systems, Beamsville ON, Canada) with a minimum irradiance of 150 mE m−2 s−1. Sequence data analysis Amino acid sequences of SEO and SEOR proteins that previously had been characterized in live plants were obtained from the NCBI database (https://www.ncbi.nlm.nih.gov/). These proteins were AtSEOR1 and -2 from Arabidopsis thaliana (Froelich et al. 2011, Anstead et al. 2012), NtSEOR1 and -2 from Nicotiana tabaccum and CmSEOR1 from Cucurbita maxima (Ernst et al. 2012), PtSEOR1 from Populus trichocarpa (Mullendore et al. 2018), GmSEO1–4 from G. max (Zielonka et al. 2014), MtSEO1–3 from M. truncatula (Pélissier et al. 2008) and VfSEO1 from V. faba (Noll et al. 2007). Amino acid sequences of these proteins and the proportions of each amino acid are available in Supplementary Data S1 and Supplementary Table S2, respectively. Amino acid composition profiles of the reviewed plant (Viridiplantae) proteins in the UniProtKB/Swiss-Prot database (release 2019_01; https://www.uniprot.org/statistics/Swiss-Prot) and those of SEOR and SEO proteins are provided in Supplementary Table S3. Sequence alignments were generated using CLC Sequence Viewer v.7.8.1 (Qiagen, Aarhus, Denmark) with default settings. Amino acids differentially conserved between SEO and SEOR proteins were identified manually. Site-directed mutagenesis of MtSEO1 and yeast expression vectors Site-directed mutagenesis in MtSEO1 (Genbank accession number: EU016204) was performed with the QuikChange Multi Site-Directed Mutagenesis Kit (Agilent Technologies, Santa Clara, CA, USA). The vector pDONR-Zeocin with MtSEO1 was used as the original template. Amino acids targeted are shown in Fig. 3. MtSEO1 and mutated MtSEO1 were introduced into pAG425GPD-ccdB and pAG424GPD-eYFP-ccdB yeast expression vectors (Addgene, Watertown, MA, USA) by the Gateway LR reaction. pAG424GPD-eYFP-ccdB (Addgene plasmid # 14344; http://n2t.net/addgene: 14344; RRID: Addgene_14344) and pAG425GPD-ccdB (Addgene plasmid # 14154; http://n2t.net/addgene: 14154; RRID: Addgene_14154) were generated originally by Susan Lindquist. All constructs were confirmed by sequencing (Eurofins Genomics, Louisville, KY, USA). The recombinant proteins that were expected to be produced by the transformed yeast were wild-type or mutated MtSEO1 linked to eYFP, together with the same protein without eYFP. Saccharomyces cerevisiae strain EGY48 was transformed using the LiAc/SS carrier DNA/PEG method (Gietz and Schiestl 2007). Correct auxotrophy (–leucine for pAG425GPD and –tryptophan for pAG424GPD-eYFP) on minimal synthetic defined medium (SD medium) was used to select yeast transformants. Yeast colonies appeared on the selection plates after incubation at 28°C for 2–3 d. Isolation of forisome-like protein bodies from yeast To isolate recombinant forisome-like bodies from yeast (S. cerevisiae), a minimum of 100 ml SD medium with selections was incubated at 28°C with shaking at 160 rpm for 2 d. Yeast cells were harvested by centrifugation and washed twice with Ca2+-free medium (100 mM KCl, 10 mM EDTA, 10 mM HEPES, pH 7.3) before cell wall digestion by 600 U Arthrobacter luteus Lyticase (L2524; Sigma-Aldrich). Spheroplasts were broken with a grinding glass pestle. The resulting mixture was loaded onto a Nycodenz density gradient (20–80% Nycodenz) produced with a custom-built gradient mixer (see Knoblauch et al. 2003, for details). The gradient was centrifuged at 2060 × g for 3 h (Centrifuge 5810 R; Eppendorf, Hauppage, NY, USA). Forisome-like bodies accumulated in a defined band. Light microscopy Yeast cells expressing MtSEO1-eYFP fusion proteins and eYFP-tagged forisome-like bodies isolated from these cells were examined under a Leica TCS SP8 confocal laser-scanning microscope (Leica Microsystems, Wetzlar, Germany) at the 514-nm excitation wavelength. To photograph protein bodies from their pointed ends, glass micropipettes produced for microinjection experiments (Howell et al. 2020) were used. When protein bodies were touched with a pipette mounted on a micromanipulator, they became attached to the pipette and could be rotated into the orientation desired for imaging. Electron microscopy For scanning electron microscopy (SEM) of native forisomes, forisomes were isolated from C. gladiata as described by Peters et al. (2007a). In short, phloem tissue was scratched from peeled stems, ground in liquid nitrogen and passed through 60-µm mesh after thawing in Ca2+-free medium (100 mM KCl, 10 mM EDTA, 10 mM HEPES, pH 7.3). About 0.5 ml of the resulting crude forisome preparation was dropped onto SEM specimen mounts covered with carbon tabs. Solid particles were allowed to settle for 5 min before the solution was washed three times with Ca2+-free medium. While most particles such as cell wall fragments wash off easily, forisomes stick fairly well to the carbon tab; usually, 10–20 forisomes were present on a sample holder after washing. The samples then were shock-frozen in liquid nitrogen, freeze-dried and sputter-coated with 2 nm platinum–palladium. The samples were imaged using an FEI (Hillsboro, OR, USA) Quanta 200 field emission gun SEM. A similar procedure was used for recombinant forisome-like bodies from yeast. Following isolation as described above, the protein bodies were washed three times with double-distilled water and about 10 µl of the resulting mixture was applied to a carbon tab on a specimen mount. The samples were flash-frozen in liquid nitrogen, freeze-dried and then coated with 3 nm platinum–palladium. Functional analyses of recombinant forisomes Functional analyses of recombinant forisomes were performed immediately after isolation. Isolated forisomes extracted from the Nycodenz gradient were washed three times in Ca2+-free medium, and 3 µl medium containing isolated forisomes was placed on a microscope slide with a rectangular glue wall to hold the liquid in place. Various test media were applied using a custom-made flow-through system. We routinely tested 1, 5 and 10 mM CaCl2 in 100 mM KCl and 10 mM HEPES, pH 7.2, and calcium-free media of various pH values (10 mM EDTA, 100 mM KCl, with either 10 mM HEPES for pH 7.2 or 10 mM CAPSO for pH 9.6, 9.8, 10.0 or 10.3). Length (L) and width (W) of forisome-like bodies were measured on confocal micrographs using the ImageJ software (https://imagej.nih.gov/ij/), and aspect ratios (W L−1) and estimated volumes (L W2) were calculated from these data. Statistical analyses Statistical evaluations of the differences in the amino acid compositions of SEO and SEOR proteins were conducted with Composition Profiler (http://www.cprofiler.org/; Vacic et al. 2007), using the total set of 39,787 reviewed plant (Viridiplantae) proteins in the UniProtKB/Swiss-Prot database (release 2019_01; https://www.uniprot.org/statistics/Swiss-Prot) as a background (compare Fig. 1). Aspect ratios and estimated volumes of forisome-like bodies often showed skewed, i.e. non-normal distributions; their pH and Ca2+ responses, therefore, were evaluated with nonparametric tests. The Wilcoxon signed-ranks test (two-tailed) was used to determine whether the forisome-like bodies tested in one particular treatment responded consistently, while differences in the responses of wild-type and mutant protein bodies were compared with the Mann–Whitney test (two-tailed). Both tests and other exploratory tests mentioned in the text were executed online at vassarstats.net, a recommendable tool. Supplementary Data Supplementary data are available at PCP online. Funding This work was supported by the US National Science Foundation (NSF grant number 0818182) and by the Army Research Office (ARO grant 64108-LS-II). Acknowledgments We thank Chuck Cody for maintaining the experimental plants and the Franceschi Microscopy and Imaging Center (WSU) for technical support. Disclosures The authors have no conflicts of interest to declare. References Anstead J.A. , Froelich D.R. , Knoblauch M. , Thompson G.A. ( 2012 ) Arabidopsis P-protein filament formation requires both AtSEOR1 and AtSEOR2 . Plant Cell Physiol . 53 : 1033 – 1042 . Google Scholar Crossref Search ADS PubMed WorldCat Arias-Moreno X. , Abian O. , Vega S. , Sancho J. , Velazquez-Campoy A. ( 2011 ) Protein-cation interaction: structural and thermodynamic aspects . Curr. Protein Peptide Sci. 12 : 325 – 338 . Google Scholar Crossref Search ADS WorldCat Aubusson-Fleury A. , Balavoine G. , Lemullois M. , Bouhouche K. , Beisson J. , Koll F. ( 2017 ) Centrin diversity and basal body patterning across evolution: new insights from Paramecium . Biol. Open 6 : 765 – 776 . Google Scholar Crossref Search ADS PubMed WorldCat Beard N.A. , Laver D.R. , Dulhunty A.F. ( 2004 ) Calsequestrin and the calcium release channel of skeletal and cardiac muscle . Prog. Biophys. Mol. Biol . 85 : 33 – 69 . Google Scholar Crossref Search ADS PubMed WorldCat Behnke H.-D. , Sjolund R.D. , eds. ( 1990 ) Sieve Elements . Springer , Berlin . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC Bray D. ( 2001 ) Cell Movements (2nd edn.). Garland Publishing , New York . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Chung E.G. , Ryu S. ( 2017 ) Stalk-length-dependence of the contractility of Vorticella convallaria . Phys. Biol . 14 : 066002 . Google Scholar Crossref Search ADS PubMed WorldCat Cronshaw J. , Sabnis D.D. ( 1990 ) Phloem proteins. In Sieve Elements . Edited by Behnke H.-D. , Sjolund R.D. pp. 257 – 283 . Springer , Berlin . Google Scholar Crossref Search ADS Google Preview WorldCat COPAC de Schepper V. , de Swaef T. , Bauweraerts I. , Steppe K. ( 2013 ) Phloem transport: A review of mechanics and control . J. Exp. Bot . 64 : 4839 – 4850 . Google Scholar Crossref Search ADS PubMed WorldCat Ernst A.M. , Jekat S.B. , Zielonka S. , Müller B. , Neumann U. , Rüping B. , et al. ( 2012 ) Sieve Element Occlusion (SEO) genes encode structural phloem proteins involved in wound sealing of the phloem . Proc. Natl. Acad. Sci. USA 109 : E1980 – E1989 . Google Scholar Crossref Search ADS WorldCat Froelich D.R. , Mullendore D.L. , Jensen K.H. , Ross-Elliott T.J. , Anstead J.A. , Thompson G.A. , et al. ( 2011 ) Phloem ultrastructure and pressure flow: sieve-element-occlusion-related agglomerations do not affect translocation . Plant Cell 23 : 4428 – 4445 . Google Scholar Crossref Search ADS PubMed WorldCat Gallivan J.P. , Dougherty D.A. ( 1999 ) Cation-π interactions in structural biology . Proc. Natl. Acad. Sci. USA 96 : 9459 – 9464 . Google Scholar Crossref Search ADS WorldCat Gietz R.D. , Schiestl R.H. ( 2007 ) Quick and easy yeast transformation using the LiAc/SS carrier DNA/PEG method . Nat. Protoc. 2 : 35 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat Gogendeau D. , Beisson J. , de Loubresse N.G. , Le Caer J.-P. , Ruiz F. , Cohen J. , et al. ( 2007 ) An Sfi1p-like centrin-binding protein mediates centrin-based Ca2+-dependent contractility in Paramecium tetraurelia . Eukar. Cell 6 : 1992 – 2000 . Google Scholar Crossref Search ADS WorldCat Groscurth S. , Müller B. , Schwan S. , Menzel M. , Diekstall F. , Senft M. , et al. ( 2012 ) Artificial forisomes are ideal models of forisome assembly and activity that allow the development of technical devices . Biomacromolecules 13 : 3076 – 3086 . Google Scholar Crossref Search ADS PubMed WorldCat Howell A.H. , Peters W.S. , Knoblauch M. ( 2020 ) The diffusive injection micropipette (DIMP) . J. Plant Physiol . 244 : 153060 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Froelich D.R. , Pickard W.F. , Peters W.S. ( 2014 ) SEORious business: structural proteins in sieve tubes and their involvement in sieve tube occlusion . J. Exp. Bot . 65 : 1879 – 1893 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Knoblauch J. , Mullendore D.L. , Savage J.A. , Babst B.A. , Beecher S.D. , et al. ( 2016 ) Testing the Münch hypothesis of long distance phloem transport in plants . eLife 5 : 315341 . Google Scholar Crossref Search ADS WorldCat Knoblauch M. , Noll G.A. , Müller T. , Prüfer D. , Schneider-Hüther I. , Scharner D. , et al. ( 2003 ) ATP-independent contractile proteins from plants . Nat. Mater. 2 : 600 – 603 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Peters W.S. ( 2004 ) Forisomes, a novel type of Ca2+-dependent protein motor . Cell Motil. Cytoskel. 58 : 137 – 142 . Google Scholar Crossref Search ADS WorldCat Knoblauch M. , Peters W.S. ( 2010 ) Münch, morphology, microfluidics – our structural problem with the phloem . Plant Cell Environ . 33 : 1439 – 1452 . Google Scholar PubMed OpenURL Placeholder Text WorldCat Knoblauch M. , Peters W.S. ( 2013 ) Long-distance translocation of photosynthates: a primer . Photosynth. Res. 117 : 189 – 196 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Peters W.S. ( 2017 ) What actually is the Münch hypothesis? A short history of assimilate transport by mass flow . J. Integr. Plant Biol. 59 : 292 – 310 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Peters W.S. , Ehlers K. , van Bel A.J.E. ( 2001 ) Reversible calcium-regulated stopcocks in legume sieve tubes . Plant Cell 13 : 1221 – 1230 . Google Scholar Crossref Search ADS PubMed WorldCat Knoblauch M. , Stubenrauch M. , van Bel A.J.E. , Peters W.S. ( 2012 ) Forisome performance in artificial sieve tubes . Plant Cell Environ . 35 : 1419 – 1427 . Google Scholar Crossref Search ADS PubMed WorldCat Kumar A. , Chakravarty H. , Bal N.C. , Balaraju T. , Jena N. , Misra G. , et al. ( 2013 ) Identification of calcium binding sites on calsequestrin 1 and their implications for polymerization . Mol. Biosyst. 9 : 1949 – 1957 . Google Scholar Crossref Search ADS PubMed WorldCat Kunierczyk A. , Winge P.E.R. , Jrstad T.S. , Troczyska J. , Rossiter J.T. , Bones A.M. ( 2008 ) Towards global understanding of plant defence against aphids – timing and dynamics of early Arabidopsis defence responses to cabbage aphid (Brevicoryne brassicae) attack . Plant Cell Environ . 31 : 1097 – 1115 . Google Scholar Crossref Search ADS PubMed WorldCat Laflèche D. ( 1966 ) Ultrastructure et cytochimie des inclusions flagellées des cellules criblées de Phaseolus vulgaris . J. Microscopie 5 : 493 – 510 . Google Scholar OpenURL Placeholder Text WorldCat Lawton D.M. ( 1978 ) Ultrastructural comparison of the tailed and tailless P-protein crystals respectively of runner bean (Phaseolus multiflorus) and garden pea (Pisum sativum) with tilting stage electron microscopy . Protoplasma 97 : 1 – 11 . Google Scholar Crossref Search ADS WorldCat Mahadevan L. , Matsudaira P. ( 2000 ) Motility powered by supramolecular springs and ratchets . Science 288 : 95 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat Manno C. , Figueroa L.C. , Gillespie D. , Fitts R. , Kang C. , Franzini-Armstrong C. , et al. ( 2017 ) Calsequestrin depolymerizes when calcium is depleted in the sarcoplasmic reticulum of working muscles . Proc. Natl. Acad. Sci. USA 114 : E638 – E647 . Google Scholar Crossref Search ADS WorldCat Maurer A. , Tanaka M. , Ozawa T. , Fleischer S. ( 1985 ) Purification and crystallization of the calcium binding protein of sarcoplasmic reticulum from skeletal muscle . Proc. Natl. Acad. Sci. USA 82 : 4036 – 4040 . Google Scholar Crossref Search ADS WorldCat Medina-Ortega K.J. , Walker G.P. ( 2015 ) Faba bean forisomes can function in defence against generalist aphids . Plant Cell Environ. 38 : 1167 – 1177 . Google Scholar Crossref Search ADS PubMed WorldCat Misra G. , Dickinson R.B. , Ladd A.J.C. ( 2010 ) Mechanics of Vorticella contraction . Biophys. J . 98 : 2923 – 2932 . Google Scholar Crossref Search ADS PubMed WorldCat Mullendore D.L. , Ross-Elliott T. , Liu Y. , Hellmann H.H. , Roalson E.H. , Peters W.S. , et al. ( 2018 ) Non-dispersive phloem-protein bodies (NPBs) of Populus trichocarpa consist of a SEOR protein and do not respond to cell wounding and Ca2+. PeerJ 6 : e4665 . Google Scholar Crossref Search ADS PubMed WorldCat Müller B. , Noll G.A. , Ernst A.M. , Rüping B. , Groscurth S. , Twyman R.M. , et al. ( 2010 ) Recombinant artificial forisomes provide ample quantities of smart biomaterials for use in technical devices . Appl. Microbiol. Biotechnol. 88 : 689 – 698 . Google Scholar Crossref Search ADS PubMed WorldCat Münch E. ( 1930 ) Die Stoffbewegungen in der Pflanze . Gustav Fischer , Jena . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Noll G.A. , Fontanellaz M.E. , Rüping B. , Ashoub A. , van Bel A.J.E. , Fischer R. , et al. ( 2007 ) Spatial and temporal regulation of the forisome gene for1 in the phloem during plant development . Plant Mol. Biol. 65 : 285 – 294 . Google Scholar Crossref Search ADS PubMed WorldCat Palevitz B.A. , Newcomb E.H. ( 1971 ) The ultrastructure and development of tubular and crystalline P-protein in the sieve elements of certain papilionaceous legumes . Protoplasma 72 : 399 – 425 . Google Scholar Crossref Search ADS WorldCat Pélissier H.C. , Peters W.S. , Collier R. , van Bel A.J.E. , Knoblauch M. ( 2008 ) GFP tagging of sieve element occlusion (SEO) proteins results in green fluorescent forisomes . Plant Cell Physiol . 49 : 1699 – 1710 . Google Scholar Crossref Search ADS PubMed WorldCat Perni S. , Close M. , FranziniArmstrong C. ( 2013 ) Novel details of calsequestrin gel conformation in situ . J. Biol. Chem. 288 : 31358 – 31362 . Google Scholar Crossref Search ADS PubMed WorldCat Peters W.S. , Haffer D. , Hanakam C.B. , van Bel A.J.E. , Knoblauch M. ( 2010 ) Legume phylogeny and the evolution of a unique contractile apparatus that regulates phloem transport . Amer. J. Bot . 97 : 797 – 808 . Google Scholar Crossref Search ADS WorldCat Peters W.S. , Knoblauch M. , Warmann S.A. , Schnetter R. , Shen A.Q. , Pickard W.F. ( 2007 a) Tailed forisomes of Canavalia gladiata: a novel model to study Ca2+-driven protein contractility . Ann. Bot . 100 : 101 – 109 . Google Scholar Crossref Search ADS PubMed WorldCat Peters W.S. , Schnetter R. , Knoblauch M. ( 2007 b) Reversible birefringence suggests a role for molecular self-assembly in forisome contractility . Funct. Plant Biol. 34 : 302 – 306 . Google Scholar Crossref Search ADS PubMed WorldCat Peters W.S. , van Bel A.J.E. , Knoblauch M. ( 2006 ) The geometry of the forisome-sieve element-sieve plate complex in the phloem of Vicia faba L. leaflets . J. Exp. Bot . 57 : 3091 – 3098 . Google Scholar Crossref Search ADS PubMed WorldCat Pickard W.F. , Knoblauch M. , Peters W.S. , Shen A.Q. ( 2006 ) Prospective energy densities in the forisome, a new smart material . Mater. Sci. Eng . 26 : 104 – 112 . Google Scholar Crossref Search ADS WorldCat Rose J. , Visser F. , Müller B. , Senft M. , Groscurth S. , Sicking K.F. , et al. ( 2020 ) Identification and molecular analysis of interaction sites in the MtSEO-F1 protein involved in forisome assembly . Int. J. Biol. Macromol . 144 : 603 – 614 . Google Scholar Crossref Search ADS PubMed WorldCat Royer L. , Ríos E. ( 2009 ) Deconstructing calsequestrin. Complex buffering in the calcium store of skeletal muscle . J. Physiol . 587 : 3101 – 3111 . Google Scholar Crossref Search ADS PubMed WorldCat Rüping B. , Ernst A.M. , Jekat S.B. , Nordzieke S. , Reineke A.R. , Müller B. , et al. ( 2010 ) Molecular and phylogenetic characterization of the sieve element occlusion gene family in Fabaceae and non-Fabaceae plants . BMC Plant Biol. 10 : 219 . Google Scholar Crossref Search ADS PubMed WorldCat Ryu S. , Pepper R.E. , Nagai M. , France D.C. ( 2017 ) Vorticella: a protozoan for bio-inspired engineering . Micromachines 8 : 4 . Google Scholar Crossref Search ADS WorldCat Sabnis D.D. , Sabnis H.M. ( 1995 ) Phloem proteins: structure, biochemistry and function. In The Cambial Derivatives (Encyclopedia of Plant Anatomy, Vol. 9 Part 4) . Edited by Iqbal M. pp. 271 – 292 . Gebrüder Borntraeger , Berlin . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Savage J.A. , Clearwater M.J. , Haines D.F. , Klein T. , Mencuccini M. , Sevanto S. , et al. ( 2016 ) Allocation, stress tolerance and carbon transport in plants: how does phloem physiology affect plant ecology? Plant Cell Environ. 39 : 709 – 725 . Google Scholar Crossref Search ADS PubMed WorldCat Vacic V. , Uversky V.N. , Dunker A.K. , Lonardi S. ( 2007 ) Composition Profiler: a tool for discovery and visualization of amino acid composition differences . BMC Bioinformatics 8 : 211 . Google Scholar Crossref Search ADS PubMed WorldCat Walker G.P. , Medina-Ortega K.J. ( 2012 ) Penetration of faba bean sieve elements by pea aphid does not trigger forisome dispersal . Entomol. Exp. Appl. 144 : 326 – 335 . Google Scholar Crossref Search ADS WorldCat Wang S. , Trumble W.R. , Liao H. , Wesson C.R. , Dunker A.K. , Kang C. ( 1998 ) Crystal structure of calsequestrin from rabbit skeletal muscle sarcoplasmic reticulum . Nat. Struct. Mol. Biol. 5 : 476 – 483 . Google Scholar Crossref Search ADS WorldCat Yáñez M. , Gil-Longo J. , Campos-Toimil M. ( 2012 ) Calcium binding proteins . Adv. Exp. Med. Biol. 740 : 461 – 482 . Google Scholar Crossref Search ADS PubMed WorldCat Yano K. , Zarain-Herzberg A. ( 1994 ) Sarcoplasmic reticulum calsequestrins: structural and functional properties . Mol. Cell. Biochem. 135 : 61 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat Zhang Y. , He C.Y. ( 2012 ) Centrins in unicellular organisms: functional diversity and specialization . Protoplasma 249 : 459 – 467 . Google Scholar Crossref Search ADS PubMed WorldCat Zielonka S. , Ernst A.M. , Hawat S. , Twyman R.M. , Prüfer D. , Noll G.A. ( 2014 ) Characterization of five subgroups of the sieve element occlusion gene family in Glycine max reveals genes encoding non-forisome P-proteins, forisomes and forisome tails . Plant Mol. Biol. 86 : 51 – 67 . Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2020. Published by Oxford University Press on behalf of Japanese Society of Plant Physiologists. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Aspartate Residues in a Forisome-Forming SEO Protein Are Critical for Protein Body Assembly and Ca2+ Responsiveness
JF - Plant and Cell Physiology
DO - 10.1093/pcp/pcaa093
DA - 2020-10-01
UR - https://www.deepdyve.com/lp/oxford-university-press/aspartate-residues-in-a-forisome-forming-seo-protein-are-critical-for-B2z3JjRTxr
SP - 1699
EP - 1710
VL - 61
IS - 10
DP - DeepDyve
ER -