A simple method for discriminating between cell membrane and cytosolic proteins. New Phytologist (2005) 165:

Transcription

1 Research Methods Blackwell Publishing, Ltd. A simple method for discriminating between cell membrane and cytosolic proteins Laura Serna Facultad de Ciencias del Medio Ambiente, Universidad de Castilla-La Mancha, E Toledo, Spain Summary Author for correspondence: Laura Serna Tel: ext 5467 Fax: laura.serna@uclm.es Received: 5 July 2004 Accepted: 27 September 2004 Transgenic plants expressing either green fluorescent protein (GFP)-genomic DNA or GFP-cDNA fusions have been used as powerful tools to define the subcellular localization of many proteins. Because most plant cells are highly vacuolated, the cytosol is confined to a thin layer at the periphery of the cells, making it very difficult to distinguish among cell wall, cell membrane and cytosolic GFP-fusion proteins. Plasmolysis tests inform about cell-wall localization of GFP-tagged proteins, but they do not discriminate between its cell membrane and/or cytoplasmic localization. By observing the GFP signal in transgenic protoplasts placed at a hypotonic solution, it was possible to distinguish between cell membrane and cytosolic GFP-tagged proteins. The osmotic disruption of the protoplast vacuole in the hypotonic solution allows the diffusion of the GFP signal from the cell periphery to the central part of the cell volume when the GFP is fused to a soluble protein. By contrast, such diffusion does not occur when the protein under study is attached to the cell membrane. The present method is easier, faster and cheaper than subcellular fractionating studies and/or immunoelectron microscopy, which have been traditionally used to discern between cell membrane and cytosolic proteins. Key words: Arabidopsis, bioinformatic analysis, cell membrane, green fluorescent protein (GFP), osmotic rupture, peripheral cytoplasm, plasmolysis. New Phytologist (2005) 165: New Phytologist (2004) doi: /j x Introduction The understanding of the function of a given protein requires its assignment to a specific cell compartment. The green fluorescent protein (GFP) initiated a revolution in molecular cell biology by allowing the visualization of any protein of interest at a subcellular level. Plant biology researchers have made an important use of the GFP by studying, under conventional fluorescence or confocal microscopy, its subcellular patterns in transgenic plants expressing either GFP genomic DNA or GFP cdna fusions. This effort has allowed defining the subcellular localization of many proteins, which is rendering us a better and deeper understanding of the plant cell biology. The subcellular and even suborganelar localization of some GFP-tagged proteins can be precisely defined because they locate to cellular or organellar compartments that are easily identified by using fluorescence or confocal microscopy. For example, because of the red autofluorescence of the chloroplast membranes, covisualization experiments with red and green fluorescence revealed that chlorophyllide an oxygenase locates to the periphery of mature chloroplast and on the thylakoid membranes when it is transiently expressed in pea leaves (Eggink et al., 2004). This enzyme catalyses the synthesis of chlorophyll b in Arabidopsis (Espineda et al., 1999). Similarly, transient expression and covisualization experiments with red fluorescence also showed that the phosphate transporter of Medicago truncaluta, MtPHT2;1, locates to the chloroplast envelope (Zhao et al., 2003). The network pattern of the endoplasmic reticulum membranes has also allowed its identification with relative ease. The cellulose synthase-like 947

2 948 Research Methods protein encoded by the Arabidopsis KOJAK gene, which is required for root hair morphogenesis, draws very well such a reticulate pattern of membranes when it is transiently expressed in the leaf cells of Nicotiana benthamiana (Favery et al., 2001). The nucleus has been also easily visualized because its spherical shape, and many transcriptional factors have been located there. For example, ZIM, an Arabidopsis thaliana putative transcription factor with a motif for a zinc finger involved in inflorescence and flower development, is targeted to the nucleus (Nishii et al., 2000). Because some subcellular compartments are visually indistinguishable at the level of fluorescence or confocal microscopy, visualizing the GFP signal in plants expressing GFP-fusion proteins has been combined with additional techniques. For example, the presence of a large vacuole in many plant cells confines the cytosol to a thin layer at the cell periphery, making very difficult to distinguish among cell wall, cell membrane and cytosolic proteins by visualizing the GFP signal. Plasmolysis experiments have allowed confirming/ discarding the cell wall localization of GFP-tagged proteins. The ELONGATION DEFECTIVE1 gene of Arabidopsis encodes a serine-rich protein that localizes to the cell wall in plasmolysed cells (Lertpiriyapong & Sung, 2003). However, in many highly vacuolated cells the plasmolysis tests have been insufficient for discerning between cell membrane and soluble proteins, and additional techniques have been usually used to distinguish between these two possibilities. For example, immunoblot analysis of soluble and microsomal membrane fractions and GFP imaging have solved the membrane association of several isoforms of calcium-dependent protein kinases (Dammann et al., 2003), and immunoelectron microscopy has shown that an associated acyl-coa binding protein of Arabidopsis named ACBP1 locates to the cell wall, to the plasma membrane and to the cell vesicles (Chye et al., 1999). To date, cell fractionation and/or immunoelectron microscopy have been indispensable tools to precisely locate cell surface proteins. However, I describe a new technique to distinguish cell membrane and cytosolic GFP-tagged proteins. This method is based on observing GFP signal in protoplasts suspended in a hypotonic medium. The osmotic disruption of the vacuole in such a medium allows the diffusion of the GFP fluorescence when the GFP protein is fused to a soluble protein. This diffusion does not occur in protoplasts bearing a GFP-tagged protein attached to the cell membrane. Although I use confocal microscopy to distinguish between cell membrane and cytosolic proteins, it is likely that this goal can be also achieved by using conventional fluorescence microscopy. Materials and Methods Plant material and growth conditions Arabidopis plants expressing cell surface GFP-fusion proteins were used. Plants bearing a CPC promoter CPC-GFP construct have been previously described (Wada et al., 2002). The line defining the RCI2A protein localization was identified in a random GFP cdna fusion screening (Cutler et al., 2000). Seeds were vernalized at 4 C for several days, after which they were sterilized in 5% sodium hypochlorite, and plated on Petri dishes containing Murashige and Skoog salts (Sigma, Aldrich Química SA, Madrid, Spain) supplemented with 1% sucrose and solidified with 1% agar. Seedlings were grown for 6 d on horizontally oriented dishes at 22 C in light (16 h light : 8 h dark cycle). Preparation of plasmolysed cells Transgenic plants were incubated in 1 M KNO 3 for 5 min, leading to plasmolysis. Mechanical isolation of protoplasts and its osmotic rupture After plasmolysis, plants were placed in slides and their hypocotyls were cross-sectioned by hand using a surgical blade (BB 536). A single section was performed per hypocotyl, and a total of 15 organs of each transgenic plant (CPC promoter-cpc-gfp plant and line) were sectioned. After sectioning, a drop of distilled water was added to the sample. This induced deplasmolysis, resulting in the expansion and release of the protoplasts from the cut ends of the cells. Fifteen protoplasts of each transgenic line (CPC promoter CPC-GFP plant and line) were monitored from its release to its osmotic disruption, studying in each the dynamic of GFP localization. Visualization of the GFP subcellular localization The fluorescence of the GFP was observed with a DMIRB inverted Leica TCS SP2 confocal microscope (Leica Microsystems GmbH, Mannheim, Germany) using a 488 nm laser line of an argon (Ar) laser, with the emission window set at nm. The plant autofluorescence was also visualized by setting the emission window at nm. The GFP fluorescence and autofluorescence acquisition were performed simultaneously using either a 10/NA 0.30 PL fluotar objective or a 40/NA PL APO oil immersion objective. Either optical sectioning along all cells to capture the entire fluorescent signals or single sections were performed depending of the samples. When the fluorescent signals were captured along all cells, the sections were overlaid in a single image. Images were recorded with picture size of pixels, arranged and labelled using Adobe Photoshop 6.0. Computational prediction of transmembrane helices Both CPC protein and RCI2A protein were analysed for the presence of plasma membrane segments using HMMTOP2.0 ( PSORT ( New Phytologist (2005) 165: New Phytologist (2005)

3 Methods Research 949 Fig. 1 Schematic representation of the method that distinguishes between cell membrane and cytosolic proteins. Plasmolysed tissues are handcut and placed in a hypotonic medium, which facilitates the releasing of a few protoplasts and which induces the disruption of their vacuoles. The latter allows the diffusion of the green fluorescent protein (GFP) signal along all cell volume when the GFP is fused to cytosolic proteins. In proteins targeted to the cell membrane, GFP fluorescence after vacuole disruption continues but is restricted to the cell periphery. The osmotic rupture of the plasma membrane creates green fluorescent curves in protoplasts bearing plasma membrane GFP-tagged proteins. and TMHMM2.0 ( servers. Results and Discussion In highly vacuolated cells where the cytosol is confined to a thin layer at the extreme cell periphery, the present method distinguishes between cell membrane and cytosolic GFP-tagged proteins by following two easy steps (Fig. 1): (1) the mechanical isolation of the protoplasts, and (2) observation of the GFP signal during its osmotic rupture. The protoplasts isolation starts with the plasmolysis of the plant (or organ) under study, resulting in movement of plasma membranes away from the cell walls. This allows division of the plant into two sections by disrupting only some protoplasts. The deplasmolysis step induces the movement of the living protoplasts, which facilitates its release to the external medium. This method of isolating protoplasts is a better alternative to those based on enzymatic reactions, because (1) the small number of protoplasts released (from one to two in every hypocotyl cross-section; n = 15), which facilitates its monitoring, and (2) absence of enzymatic buffers that may induce some effects on the stability/localization of the GFP. It is important to note that observation of the GFP signal must be combined with the simultaneous visualization of the red fluorescence, which informs about the position of the chloroplasts in the cell volume. This allows easy inference of the moment in which the vacuole disruption takes place; such a moment coincides with the beginning of the free movement of the chloroplast from the cell periphery to the central volume of the cell. Two GFP-tagged cell surface proteins (Fig. 2) were used to test the efficacy of the present method: CAPRICE (CPC) and RARE COLD INDUCIBLE 1A (RCI1A). CPC belongs to the MYB family of transcriptional factors and it positively regulates root hair formation (Wada et al., 1997). Previous studies have shown that CPC locates to the nucleus of root epidermal cells (Wada et al., 2002). RCI2A is a hydrophobic protein, whose gene expression is induced by cold (Capel Fig. 2 Confocal images of the hypocotyl epidermis. (a) In CPC promoter CPC-GFP plants, the green fluorescent protein (GFP) signal locates to both the cell surface and the nucleus (arrows). (b) In the line, the GFP fluorescence is restricted to the cell surface. Images are single optical sections. Bar, 60 µm. et al., 1997). Plants bearing either the CPC promoter CPC- GFP construct or a random GFP cdna fusion that solves the RCI2A protein localization (37-26 line), showed GFP fluorescence at the cell surface of hypocotyl epidermal cells (Fig. 2). The nucleus of CPC promoter-cpc-gfp plants, as it might be expected for a putative transcriptional factor, also exhibited fluorescence (Fig. 2a). Plasmolysis experiments showed that these proteins, CPC and RCI2A, do not localize to the cell wall (Fig. 3). Plasmolysed CPC promoter CPC-GFP protoplasts showed the GFP signal restricted to the nucleus and to the cell surface (Fig. 4a). From 5 to 35 min, after placing the protoplasts in a hypotonic solution, the chloroplasts started to move freely New Phytologist (2005) New Phytologist (2005) 165:

4 950 Research Methods Fig. 3 Confocal images of plasmolysed hypocotyl epidermal cells. (a) CPC promoter CPC-GFP plants; (b) line. The green fluorescent protein (GFP) signal does not localize to the cell wall in no line. Images are single optical sections. Bar, 40 µm. along the cytoplasm because of osmotic disruption of the tonoplast. In parallel, the GFP signal of the cell surface diminished because of diffusion of the GFP-tagged CPC in the absence of the vacuole; such a GFP signal was absent some seconds/minutes after vacuole disruption. A total of 15 protoplasts were monitored, and in all of them the GFP dynamic was identical. These results indicate that CPC is not targeted to the cell membrane, but that it is a soluble protein. It is likely that the localization of GFP signal in the cytosol is caused by an excess of CPC protein, perhaps saturating its movement to the nucleus. The dynamic of GFP movement in protoplasts of the line (n = 15; Fig. 4b) differed from that of the CPC promoter CPC-GFP protoplasts. A free movement of chloroplasts, from 2 to 30 min after placing the protoplast in a hypotonic medium, was also observed in this line. By contrast, the GFP signal at the cell surface did not decrease, indicating that RCI2A protein is targeted to the plasma membrane. The osmotic disruption of the plasma membrane provided fluorescent curves confirming such subcellular localization. Proteins interact with the cell membranes through specific sequences. Both CPC and RCI2A sequences were analysed for the presence of plasma membrane segments using three programs (THMM2.0, PSORT and HMMTOP). The three programs showed that the CPC sequence has no predicted membranespanning domain (Fig. 5a). Although some transcriptional factors are inserted into the membrane by interaction with other proteins, thereby preventing or reducing their access to the nucleus and thus the transcription regulation of their Fig. 4 Dynamics of the green fluorescent protein (GFP) fluorescence in protoplasts suspended in a hypotonic medium. (a) In the CPC promoter CPC-GFP line, 5 min after placing the protoplast under a hypotonic solution, the chloroplasts move freely from the cell periphery to the central part of the cell volume, which reveals the osmotic disruption of the vacuole. Simultaneously, the GFP signal diffuses from the cell periphery, discarding the CPC attachment to the cell membrane and indicating that it locates to the cytoplasm. Arrows show the nuclei. (b) The osmotic disruption of the plasma membrane of a protoplast of the line, 3 min after placing it under a hypotonic medium, reveals that RCI2A is a plasma membrane protein. Images are single optical sections. Bars, (a) 25 µm, (b) 35 µm. New Phytologist (2005) 165: New Phytologist (2005)

5 Methods Research 951 Fig. 5 Posterior probabilities for CPC and RCI2A proteins. (a) The whole sequence of CPC is labelled as inside (or outside), predicting that it contains no membrane helices. (b) For RCI2A, the TMHMM2.0 server predicts two transmembrane segments between amino acids 4 and 26, and between 31 and 53. Between such transmembrane segments the server predicts the presence of four cytosolic residues (from 27 to 30 amino acids), and it also predicts a few extracellular amino acids in the C- and N-termini of the protein. target genes (Hoppe et al., 2001), observation of the CPC promoter CPC-GFP protoplasts under a hypotonic medium indicated that CPC is not sequestrated in the plasma membrane. The nuclear localization of the protein was identified using PSORT. The program THMM2.0, which is considered the best performing transmembrane prediction program (Möller et al., 2001), predicted two transmembrane helices between amino acids 4 and 26, and between amino acids 31 and 53, for RCI2A protein (Fig. 5b). A previous method for prediction of membrane-spanning regions (Klein et al., 1985), also suggested that RCI2A contains two potential transmembrane domains (Capel et al., 1997). The THMM2.0 program also predicted that the region between the amino acid 27 and 30 locates to the cytoplasm, and that the N (amino acid between 1 and 3) and C (amino acid 54) terminus of the protein locates to the extracellular medium (Fig. 5b). Both the HMMTOP and the PSORT programs also predicted two transmembrane helices for RCI2A protein. In summary, the predictions through bioinformatic analysis are consistent with the presence of a fraction of soluble CPC, and with the attachment of RCI2A to the plasma membrane. The present method provides an easy and inexpensive way to solve the subcellular localization of GFP-tagged proteins by using confocal/fluorescence microscopy. It then offers an interesting alternative to immunoelectron microscopy and subcellular fractionating studies, which are therefore greatly restricted by the available financial and technological resources of the laboratories. The analysis of the genomic sequence of Arabidopsis revealed that 8613 genes contain at least a transmembrane domain (The Arabidopsis Genome Initiative, 2000) and the localization of most of them is unknown. However, this might be experimentally addressed easily by using this method. This will require the availability of transgenic plants with fusions of such genes to the GFP. Cutler et al. (2000), in a screen for random GFP cdna fusions, isolated five Arabidopsis lines showing the GFP fluorescence at the extreme periphery of the cell. One of them revealed the cell surface localization of RCI2A, which has been studied here, showing that it is attached to the plasma membrane. The precise localization of the GFP in the remaining four lines remains unknown, although the gene identity responsible for such localization suggests that the proteins encoded by them are targeted to the plasma membrane (Cutler et al., 2000). The precise localization of such proteins can therefore be investigated using this method. Because subcellular localization will become an increasingly important tool in determining the function of the many unknown proteins predicted by plant genome projects, and GFP fusions are already indispensable, the present method might also extend not only to Arabidopsis thaliana but also to any plant species. Acknowledgements I am grateful to Takuji Wada for generously providing CPC promoter CPC-GFP transgenic plants. I thank David Ehrhardt for kindly providing the line, and for sharing unpublished results. Thanks also to Carl Ng for the critical reading of the manuscript. This work was supported by grants from the Universidad de Castilla-La Mancha (UCLM), the Ministerio de Ciencia y Tecnología (MCyT) (BMC C) and the Junta de Comunidades de Castilla-La Mancha (JCCM) (GC, ). References Capel J, Jarillo JA, Salinas J, Martínez-Zapater JM Two homologous low-temperature-inducible genes from Arabidopsis encode highly hydrophobic proteins. Plant Physiology 115: Chye M-L, Huang B-Q, Zee SY Isolation of a gene encoding Arabidopsis membrane-associated acyl-coa binding protein and immunolocalization of its gene product. Plant Journal 18: New Phytologist (2005) New Phytologist (2005) 165:

6 952 Research Methods Cutler SR, Ehrhardt DW, Griffitts JS, Somerville CR Random GFP::cDNA fusions enable visualization of subcellular structures in cell of Arabidopsis at a high frequency. Proceedings of the National Academy of Sciences, USA 97: Dammann C, Ichida A, Hong B, Romanowsky SM, Hrabak EM, Harmon AC, Pickard BG, Harper JF Subcellular targeting of nine calcium-dependent protein kinase isoforms from Arabidopsis. Plant Cell 132: Eggink LL, LoBrutto R, Brune DC, Brusslan J, Yamasato A, Tanaka A, Hoober JK Synthesis of chlorophyll b: localization of chlorophyllide a oxygenase and discovery of a stable radical in the catalytic subunit. Biomed Central Plant Biology 4:5 doi: / Espineda CE, Linford AS, Devine D, Brusslan JA The AtCAO gene, encoding chlorophyll a oxygenase, is required for chlorophyll b synthesis in Arabidopsis thaliana. Proceedings of the National Academy of Sciences, USA 96: Favery B, Ryan E, Foreman J, Linstead P, Boudonck K, Steer M, Shaw P, Dolan L KOJAK encodes a cellulose synthase-like protein required for root hair cell morphogenesis in Arabidopsis. Genes and Development 15: Hoppe T, Rape M, Jentsch S Membrane-bound transcription factors: regulated release by RIP or RUP. Current Opinion in Cell Biology 14: Klein P, Kanehisa M, DeLisi C The detection and classification of membrane-spanning proteins. Biochimica et Biophysica Acta 815: Lertpiriyapong K, Sung ZR The elongation defective1 mutant of Arabidopsis is impaired in the gene encoding a serine-rich secreted protein. Plant Molecular Biology 53: Möller S, Croning MDR, Apweiler R Evaluation of methods for the prediction of membrane spanning regions. Bioinformatics 17: Nishii A, Takemura M, Fujita H, Shikata M, Yokota A, Kohchi T Characterization of a novel gene encoding a putative single zinc-finger protein, ZIM, expressed during the reproductive phase in Arabidopsis thaliana. Bioscience, Biotechnology, and Biochemistry 64: The Arabidopsis Genome Initiative Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: Wada T, Kurata T, Tominaga R, Koshino-Kimura Y, Tachibana T, Goto K, Marks MD, Shimura Y, Okada K Role of a positive regulator of root hair development, CAPRICE. Arabidopsis root epidermal cell differentiation. Development 129: Wada T, Tachibana T, Shimura Y, Okada K Epidermal cell differentiation in Arabidopsis determined by a Myb homolog, CPC. Science 277: Zhao L, Versaw WK, Liu J, Harrison MJ A phosphate transporter from Medicago truncatula is expressed in the photosynthetic tissues of the plant and located in the chloroplast envelope. New Phytologist 157: About New Phytologist New Phytologist is owned by a non-profit-making charitable trust dedicated to the promotion of plant science, facilitating projects from symposia to open access for our Tansley reviews. Complete information is available at Regular papers, Letters, Research reviews, Rapid reports and Methods papers are encouraged. We are committed to rapid processing, from online submission through to publication as-ready via OnlineEarly the 2003 average submission to decision time was just 35 days. Online-only colour is free, and essential print colour costs will be met if necessary. We also provide 25 offprints as well as a PDF for each article. For online summaries and ToC alerts, go to the website and click on Journal online. You can take out a personal subscription to the journal for a fraction of the institutional price. Rates start at 109 in Europe/$202 in the USA & Canada for the online edition (click on Subscribe at the website). If you have any questions, do get in touch with Central Office (newphytol@lancaster.ac.uk; tel ) or, for a local contact in North America, the USA Office (newphytol@ornl.gov; tel ). New Phytologist (2005) 165: New Phytologist (2005)