Systematic Subcellular Localization of Novel Proteins
The completion of several genome sequencing projects now reveals many thousand open reading frames (ORFs) encoding novel proteins of unknown function. One of the major challenges in the next years will be to allocate functional data to each of these new proteins and to determine how they interact with each other to form the complex regulatory networks underlying fundamental processes of life and disease. Determining the subcellular localisation of these novel ORFs is one important step to be taken in order to bridge the gap between known sequence and unknown function. One way to achieve this may be to systematically raise protein specific antibodies and use them subsequently to determine localisation by immunofluorescence microscopy. However, raising antibodies on such a scale is laborious and expensive. A good alternative to subcellular localisation by immunofluorescence using antibodies is the tagging of the novel ORFs with the green fluorescent protein (GFP) or its spectral variants (Tsien, 1998, 1998, 2002; Zhang et al., 2002) followed by subcellular localisation of the GFP-tagged fusion proteins in living cells or multicellular organisms (Ding et al., 2000; Simpson et al., 2000). This approach has now become much simpler and faster by advances in restriction enzyme-free cloning methods, such as the Gateway system from Invitrogen (Walhout et al., 2000), which enable hundreds of defined ORFs to be transferred into GFP vectors in a matter of days. GFP tagging is not only less expensive and perhaps faster, but has the critical advantage compared to immunofluorescence that the expressed GFP fusion proteins can be localised in living samples, which reduces the risk of artefacts caused by fixation and subsequent permeabilisation as is necessary for immunofluorescence. The GFP tag further enables determination of the dynamics of the fluorescent protein, e.g., by time-lapse fluorescence microscopy or fluorescence recovery after photobleaching (FRAP; Bastiaens and Peppperkok, 2000), and thus permits a further level of functional characterisation. This article describes the basic methodology used to systematically determine the subcellular localisation of novel human proteins as they have been derived by past and current cDNA sequencing projects worldwide.
II. EXPERIMENTAL STRATEGY
Our experimental strategy to systematically localise novel proteins is summarised schematically in Fig. 1. It is based on tagging of the respective cDNAs with the GFP and subsequent expression and localisation of the GFP fusion proteins in living and fixed cells. We start the subcellular localisation procedure with bioinformatic analyses of the sequences under study in order to identify organelle-specific targeting sequences or related proteins of which the localisation has already been determined. These data are always considered alongside the final experimental results obtained.
Because localisation is wholly dependent upon targeting sequences within the protein of interest, tagging of a protein with GFP always carries the risk that these targeting sequences become masked, which will finally lead to a mislocalisation of the tagged protein. We address this problem by tagging the proteins separately at their N and C termini and determining the localisation of both fusion proteins. If N- and C-terminal fusions show identical localisation patterns, one can be confident that the subcellular localisation determined is correct. If the N- and C-terminal fusions give different results, data are considered with respect to the bioinformatic predictions. The localisation that best matches the bioinformatic data is then considered as the correct one. Finally, data are verified by colocalisation of the GFP fusion proteins with established endogenous organelle-specific markers.
Localisation studies can be performed in a variety of cultured cell lines, the choice of which should preferably match the source of the ORFs. However, we prefer to use the monkey kidney fibroblast cell line, Vero (ATCC CCL-81), as these cells have the advantages that they are large in diameter (about 60µm), display a very clear subcellular morphology (see examples shown in Fig. 2), and are particularly flat, which makes them ideal for imaging using wide-field fluorescence microscopy. Furthermore, we have so far observed no discrepancies of protein localisations in these cells compared to HeLa (ATCC CCL-2) cells, which are of human origin. In cases where a clear localisation of the GFP-tagged fusion protein to a cellular compartment or structure is difficult to achieve in Vero or HeLa cells, we use more specialised cell types for the localisation experiments, such as rat primary hippocampal neurons or SH-SY5Y human neuroblastoma cells (ATCC CRL-2266), when the protein under investigation is, for example, derived from a brain-specific cDNA library.
III. MATERIALS AND INSTRUMENTATION
Vero cells (ATCC CCL-81) are grown in minimal essential medium (MEM) containing Earle's salts (Cat. No. 21090-022) with the addition of 2mM L-glutamine (Cat. No. 25030-024) and 100U/ml penicillin/100µg/ml streptomycin (Cat. No. 15140-122) all from Invitrogen and 10% foetal calf serum (FCS) (Cat. No. A15-043) from PAA Laboratories. Trypsin-EDTA (Cat. No. 25300-054) is from Invitrogen. Live cell imaging is performed in "Imaging Medium," consisting of MEM containing Earle's salts but lacking phenol red, FCS, and antibiotics (Cat. No. M3024) from Sigma. For transfections, OptiMEM with Glutamax (Cat. No. 51985-026) is from Invitrogen and the FuGENE6 transfection reagent (Cat. No. 1814443) is from Roche. Highpurity cycloheximide powder (Cat. No. 239764) is from Calbiochem. Methanol (Cat. No. 106009) and glycine (Cat. No. 104201) are from Merck. Paraformaldehyde (PFA) (Cat. No. P6148) and Triton X-100 (Cat. No. T9284) are from Sigma. Cy5-conjugated secondary antibodies (antimouse, Cat. No. PA45002, and antirabbit, Cat. No. PA45004) are from Amersham Biosciences. Alexa Fluor647-conjugated secondary antibodies (antisheep, Cat. No. A-21448, and antigoat, Cat. No. A-21447) are from Molecular Probes. Standard cell culture plasticware is from Falcon/BD Biosciences.
Glass-bottomed live cell imaging dishes (35 mm, with 10mm number 1.5 coverglass) (Cat. No. P35G-1.5-10- C) are available from MatTek, and Lab-Tek 8-well chambered cover glass dishes (Cat. No. 155411) are from Nalge Nunc. Cells are imaged on a Zeiss Cell Observer System consisting of a Zeiss Axiovert 200 inverted microscope equipped with a Zeiss Planapochromat 63x/1.4NA objective and standard CFP (Cat. No. 1196-682), GFP (Cat. No. 1114-459), YFP (Cat. No. 1196-681), and Cy5/Alexa647 (Cat. No. 488026) filter sets. Images are captured with a CCD camera (Zeiss Axiocam) using Zeiss Axiovision 3.1 software. Images are contrast adjusted and merged using Photoshop 6.0 from Adobe.
A. Localisation of GFP-Fusion Proteins in Living and Fixed Cells
1. Plating Cells on Live Cell Imaging Dishes
2. Transfection of Cells
3. Imaging of Living Cells
Localisations of GFP-tagged proteins may change due to the expression levels and therefore we use cells expressing low and moderate levels of the protein under investigation for our localisation experiments.
Therefore, cells are imaged at various times, typically 14, 20, and 40h after transfection. This results in cells expressing low, moderate, and high amounts of the GFP-tagged proteins and gives further information how the expression level might influence localisation.
4. Imaging of Fixed Cells
In parallel cultures or following image acquisition from live cells it is important to fix the cells and remove any soluble GFP signal that may be obscuring more subtle localisation patterns. The choice of fixation reagent is largely determined by the localisation pattern observed in the live cells. Paraformaldehyde, for example, allows for better fixation of small membrane structures such as endosomes. However, when appropriate, we prefer methanol as the fixative, as this is rapid and effectively removes soluble cytoplasmic GFP-tagged proteins, but leaves structures largely intact.
5. Classification of Localisations
Images obtained from living and fixed cells are then inspected manually and compared to images obtained in living and fixed cells with already established GFPtagged organelle-specific markers (see examples in Fig. 2; more examples for organelle-specific localisations can be seen at http://gfp-cdna.embl.de.
B. Integration of Localisations with Bioinformatic Predictions
Having classified localisation of the GFP-tagged protein in live and fixed cells (Section IV, A), and if the N- and C-terminal fusions give the same localisation pattern, one can be relatively confident that this represents the localisation of this protein and therefore one can proceed with confirmation of the results by colocalisation of the GFP-tagged proteins with endogenous organelle-specific markers (Section IV, C). When the N- and C-terminal localisation patterns differ, bioinformatic data about the protein of interest should be consulted. In our experience, for over two-thirds of the proteins we have screened, combination of bioinformatic predictions with experimental cell localisations allows a final localisation to be concluded.
C. Verification of Results by Colocalisation with Endogenous Organelle-Specific Markers
1. Immunostaining of Cells
Before staining, cells are fixed with methanol or paraformaldehyde as described in Section IV, A. For verification of the localisations, we use commercially available primary antibodies recognising organellespecific marker proteins. The suppliers of these antibodies, the host animals in which they have been raised, the preferred cell fixation method giving best results, and the required antibody dilutions are summarised in Table I. The secondary antibodies we use are conjugated with Cy5 or Alexa647, which can be separated easily from YFP or CFP fluorescence using standard filter sets.
2. Analysing Colocalisation
For colocalisation of the double-labelled samples we use a Zeiss Cell Observer System for image acquisition. It is equipped with filter sets for CFP, GFP, YFP, and Cy5/Alexa647. Imaging of the two colour channels (GFP-tagged protein and Cy5/Alexa647-stained organelle marker) is performed sequentially, which has the advantage of minimising bleed through of the channels. We analyse colocalisation by merging the images acquired for the GFP-tagged protein (green) and the organelle marker (red). This is usually sufficient to accurately determine whether the suspected localisation of the GFP-tagged protein matches the one of the reference marker. However, for reasons of reliability, it is important that during image acquisition the exposure time is set such that the camera is not saturated and that the range of grey levels of the captured images covers the entire dynamic range of the imaging system (e.g., 256 on an 8-bit camera). Some image acquisition software (e.g., Axiovision) contains an autoexposure feature to ensure that this occurs.
Bastiaens, P. I., and Pepperkok, R. (2000). Observing proteins in their natural habitat: The living cell. Trends Biochem. Sci. 25, 631-637.
Ding, D. Q., Tomita, Y., Yamamoto, A., Chikashige, Y., Haraguchi, T., and Hiraoka, Y. (2000). Large-scale screening of intracellular protein localization in living fission yeast cells by the use of a GFP-fusion genomic DNA library. Genes Cells 5, 169-190.
Simpson, J. C., Wellenreuther, R., Poustka, A., Pepperkok, R., and Wiemann, S. (2000). Systematic subcellular localisation of novel proteins identified by large-scale cDNA sequencing. EMBO Rep. 1, 287-292.
Tsien, R. Y. (1998). The green fluorescent protein. Annu. Rev. Biochem. 67, 509-544.
Walhout, A. J. M., Sordella, R., Lu, X., Hartley, J. L., Temple, G. F., Brasch, M. A., Thierry-Mieg, N., and Vidal, M. (2000). Protein interaction mapping in C. elegans using proteins involved in vulval development. Science 287, 116-122.
Zhang J., Campbell, R. E., Ting, A. Y., and Tsien, R. Y. (2002). Creating new fluorescent probes for cell biology. Nature Rev. Mol. Cell Biol. 3, 906-918.
© 2018 Biocyclopedia | All rights reserved.