Nature Cell Biology 5, S1–S7 (2003)

Review: Lighting up cells: labelling proteins with fluorophores

During the past decade, rapid improvements have been made in the tools available for labelling proteins within cells, which has increased our ability to unravel the finer details of cellular events. One significant reason for these advances has been the development of fluorescent proteins that can be incorporated into proteins by genetic fusion to produce a fluorescent label. In addition, new techniques have made it possible to label proteins with small organic fluorophores and semiconductor nanocrystals.

Atsushi Miyawaki, Asako Sawano and Takako Kogure
Laboratory for Cell Function Dynamics, Advanced Technology Development Group, Brain Science Institute, RIKEN, 2-1 Hirosawa, Wako-city, Saitama, 351-0198, Japan.
correspondence to: matsushi@brain.riken.go.jp

Published online: 1 September 2003
doi:10.1038/ncb1031

The conventional techniques that are used to fluorescently label proteins and image them in their native environment can be laborious. For example, reliable protein labelling requires expertise in protein chemistry, and the successful microinjection of labelled products into cells with minimal damage requires much technical experience. Moreover, it is difficult to target fluorescently labelled proteins directly to specific sites within a cell, because the distribution and targeting of most proteins is regulated by their in vivo translation and post-translational modifications.

By contrast, de novo synthesis is much more likely to result in native patterns of protein localization. Gene transfer techniques, including liposome-mediated transfection, which use various viral vectors, electroporation and the GENE GUN, make this possible and have shown significant progress in recent years. As a result, proteins can now be expressed within cells as fusions to fluorescent proteins or to small tags that can react with specialized fluorophores. Although more traditional methods such as protein microinjection are not without their advantages, these new methods for fluorescently labelling proteins by genetic fusion are opening new windows for our understanding of cellular function.

Labelling with fluorescent proteins

Expanding the colour repertoire. Green fluorescent protein (GFP) was originally isolated from the light-emitting organ of the jellyfish Aequorea victoria¹ by Shimomura et al. in 1962, although more than 30 years passed before the complementary DNA encoding the protein was subsequently characterized^2,3. As Aequorea GFP is spontaneously fluorescent, chimeric GFP fusions offer the great advantage that they can be expressed in situ by gene transfer into cells, thereby circumventing the need for high-level heterologous production, purification, in vitro labelling and microinjection of recombinant proteins. In addition, these GFP fusions can be localized to particular sites within the cell by appropriate targeting signals.

Although spectral variants with blue, cyan and yellowish-green emissions have been successfully generated from the Aequorea GFP⁴, none exhibit emission maxima longer than 529 nm¹. Fortunately, the discovery of novel 'GFP-like proteins' from Anthozoa (coral animals) have significantly expanded the range of colours available for cell biological applications. As a result, the family of 'GFP-like proteins' deposited in sequence databases now includes approximately 30 significantly different members^5,6. Despite only a modest degree of sequence similarity, these GFP-like proteins probably share a -can fold structure that is central to the fluorescence of GFP (Fig. 1). In this review, the term 'fluorescent proteins' is used to describe those proteins that can become spontaneously fluorescent through the autocatalytic synthesis of a CHROMOPHORE. Although most GFP-like proteins fall into this category, a subset display only intense absorption without fluorescence emission, and are dubbed chromoproteins⁶. Most of the fluorescent proteins discussed in this review are summarized in Fig. 2 (also see Ref. 7, which covers fluorescent proteins more broadly).

Fig. 1
		Figure 1 | Schematic of protein labelling with a variety of fluorophores.

Fig. 2
		Figure 2 | Wild-type fluorescent proteins and the spectral variants discussed in this review.

Proteins that fluoresce at red or far-red wavelengths (red fluorescent proteins or RFPs) are of specific interest, as eukaryotic cells and tissues display reduced AUTOFLUORESCENCE at these longer wavelengths. Also, RFPs can be used in combination with other fluorescent proteins that fluoresce at shorter wavelengths for both multicolour labelling and fluorescence resonance energy transfer (FRET) experiments. At present, the commercially available RFPs are derived from two wild-type GFP-like proteins. The first, DsRed (drFP583), has excitation and emission maxima at 558 nm and 583 nm, respectively⁵. It retains an impressive brightness and remains stable despite pH changes, the presence of denaturants and photobleaching⁸. The second, a far-red fluorescent protein, was generated by mutagenesis of a chromoprotein that absorbs at 571 nm⁹. The resulting protein, HcRed1 (Clontech), has excitation and emission maxima at 588 nm and 618 nm, respectively. HcRed1, however, exhibits a low MOLAR EXTINCTION COEFFICIENT () at 588 nm (₅₈₈ = 20,000 M^-1 cm^-1) and a low FLUORESCENCE quantum yield ( = 0.015). As the brightness of a fluorophore depends on both the at a certain wavelength and , which define the absorption and the ratio of photons emitted to photons absorbed, respectively, HcRed1 fluoresces weakly. So far, the fluorescent protein which emits fluorescence at the longest wavelength (without any mutations being introduced) is eqFP611, cloned from the sea anemone Entacmaea quadricolor¹⁰. This protein absorbs at 559 nm and emits at 611 nm. As many spectral variants have emerged, more investigators are becoming interested in the simultaneous imaging of multiple fluorophores and/or FRET signals.

Making a successful fusion protein. There are three important points to consider when creating a functional fluorescent protein (Fig. 1a,b): the fluorescent protein must fold correctly to fluoresce, the host protein also needs to fold correctly to be functional, and the integrity of the chimeric protein must be maintained.

The length and sequence of the linker between the fluorescent protein and host protein should be optimized for each specific application. In many cases, steric hindrance or folding interference can occur between the fluorescent protein and host protein if the linker is not sufficiently long and flexible. The amino acid that confers the most flexibility to a peptide chain is glycine (Gly), which has the smallest side chain of all amino acids. A small number of glycine residues is therefore the most common method used to link two protein domains. The most widely used linker designs have sequences that primarily consist of Gly and serine (Ser) stretches, Ser residues being interspersed to improve the solubility of a poly-Gly stretch.

The decision of whether to fuse a fluorescent protein to the amino or carboxyl terminus of a protein depends on the properties of the protein. For example, a particular terminus might need to be preserved to retain proper protein function or to ensure correct localization. This decision might also be made on the basis of structural aspects of the particular fluorescent protein. For example, Aequorea GFP has a floppy carboxyl terminal tail of approximately ten amino acids¹¹, which makes its fusion to the amino terminus of other proteins possible without the addition of a linker (Fig. 1a). By contrast, DsRed is more successfully fused to the carboxyl terminus of proteins of interest (Fig. 1b), because the amino termini project fully from a tetrameric complex of DsRed (see below).

In rare cases in which neither end of a host protein can be modified, it is possible to insert the fluorescent protein into the middle of the protein. A highly flexible portion such as a -turn should, theoretically, be tolerant to such an insertion. For example, enhanced GFP (EGFP; Clontech) has been inserted into the cytoplasmic domain of a non-conducting mutant of the Shaker K⁺ channel¹² and fluorescence was visible in cells. Moreover, the intensity of fluorescence was altered only slightly by the voltage-dependent change in the domain structure.

In general, to preserve the original structure of a host protein, the resulting amino and carboxyl termini of the inserted fluorescent protein should be in close proximity; this also allows efficient folding of the fluorescent protein. A circularly permuted GFP (cpGFP), the amino and carboxyl portions of which have been interchanged and reconnected by a short spacer between the original termini^13,14, could be used for such a purpose. As the two resulting termini are close to each other, it might even be possible to insert this molecule into the middle of secondary structures (Fig. 1c). However, this remains to be tested.

Maturation of fluorescent proteins. Poor folding of a fluorescent protein variant results in a non-fluorescent chimaera. Accumulation of a large amount of such a protein inside cells will decrease the fluorescent signal, and potentially perturb cellular homeostasis if the labelled host protein retains its original function. It is therefore imperative to use fluorescent protein variants that mature efficiently. In this respect, it is important to note that the folding efficiencies of both the fluorescent protein and the host protein are likely to be interdependent. Fluorescent proteins tend to fold less efficiently when fused to other proteins, although fusion to a well-folded host protein can facilitate proper fluorescent protein folding¹⁵.

After protein synthesis, many GFP variants mature quite slowly, involving a multi-step folding process that consists of cyclization, dehydration and oxidation. However, citrine¹⁶ and Venus¹⁷, two bright versions of a yellow-emitting mutant of GFP (YFP) that mature efficiently, have recently been developed. These variants can also facilitate host protein folding¹⁷. Moreover, their rapid maturation allows the immediate detection of fluorescent signals after the introduction of genes to freshly prepared biological samples, such as brain slices¹⁷. Similarly to GFP variants, the red chromophore of DsRed also undergoes these maturation steps, but requires an additional autocatalytic modification of its GFP-like chromophore¹⁸; incomplete maturation gives rise to residual green fluorescence, which might be a disadvantage for separation from green signals. Two recently developed varieties of DsRed, known as T1 (Ref. 19) and E57 (Ref. 20), display improved maturation, making them preferable for use in dual-colour experiments.

Conversely, a long-lived green state can be advantageous if the intention is to analyse the history of the synthesis of a protein in a cell. A new mutant of DsRed, E5, is particularly useful for this because it changes its colour from green to red over a predictable time course²¹. This feature makes it possible to use the ratio of green-to-red emission as a measure of the time that has elapsed since the initiation of protein synthesis. Therefore, E5 functions as a fluorescent timer that yields both temporal and spatial information about target protein age. For example, this property of E5 maturation has been applied in experiments with adrenal chromaffin cells expressing a peptide hormone fused to E5, and has been used to demonstrate that secretory vesicles are segregated functionally and spatially according to age²².

It is emerging that the maturation and subsequent fluorescence of some GFP variants can be 'photoactivated' by specific illumination, which provides the advantage that fluorescence can be turned on at a chosen time point. During the past year, three new fluorescent proteins that undergo photochemical modification in or near the chromophore have been developed: PA-GFP²³, Kaede²⁴ and KFP1²⁵ (see also the review on page S7 of this supplement). They enable selective activation of fluorescence signals after specific illumination, and can be used to fluorescently mark individual cells, organelles or proteins⁷.

The advantages and disadvantages of oligomerization. The propensity for a fluorescent protein to form oligomers is an important consideration, as such interactions can interfere with the function of the host protein to which it is fused. Unfortunately, all of the Anthozoan GFP-like proteins characterized so far form obligate oligomers¹⁴. Although oligomerization does not prevent their use for reporting gene expression or marking cells, it does preclude their use in fusion protein applications. Similarly, fusion of DsRed, which normally forms a tetramer, to host proteins often disrupts their normal behaviour, although there are some exceptions (Fig. 3A). Campbell et al. recently reported the successful engineering of monomeric RFP (mRFP1)²⁶ from DsRed. As mRFP1 matures ten times faster than its parental protein, it exhibits similar brightness to DsRed in living cells despite its lower molar extinction coefficient, fluorescence quantum yield and photostability. Because it is monomeric, mRFP1 has enabled red-fluorescence labellings that were not possible before with DsRed (Fig. 3B). Also, the excitation and emission maxima of mRFP1 are 584 nm and 607 nm, respectively (Fig. 2b,c), which gives good spectral separation from other fluorescent protein signals. This work provides hope that other oligomeric fluorescent proteins might also be converted into monomers. Indeed, the far-red variant HcRed1, which is made from a parent chromoprotein that seems to form obligate tetramers, has also been engineered to form dimers⁹. Similarly, the anemone fluorescent protein epFP611 can function as a monomer, but only at low concentrations and in the presence of detergent¹⁰.

Fig. 3
		Figure 3 | Fluorescence images of cells containing proteins labelled with fluorophores by various techniques.

A further problem is the potential aggregation of fluorescent proteins, which impedes any cellular application and leads to cellular toxicity. Although the molecular mechanisms of fluorescent protein aggregation remain unclear, there are two possible explanations. First, aggregation might be due to electrostatic or hydrophobic interactions between fluorescent proteins. The possible contribution made by electrostatic interactions has been supported by recent work in which non-aggregating mutants were successfully generated by removing basic residues located near the amino termini of several fluorescent proteins²⁷, including DsRed. So, it might also be possible to make non-aggregating mutants by removing hydrophobic side chains on the surface of oligomeric complexes. It should be noted that Renilla GFP becomes soluble as a result of its dimerization; a hydrophobic patch becomes hidden at the dimerization interface and allows the surface of the dimer to become hydrophilic.

The second possibility is that aggregation might follow fluorescent protein oligomerization. So, the problem might be made worse still if host proteins are also oligomeric, as fusion to fluorescent proteins might result in crosslinking into massive aggregates. Indeed, DsRed tends to produce more serious aggregation when fused to a host protein, although, in an exception to this trend, fusion of DsRed to protein kinase C- ( PKC-) retains the dynamic redistribution of the enzyme after stimulation (Fig. 3a). Overall, this aggregation problem would most easily be solved by using monomeric fluorescent proteins (Fig. 3B)²⁶.

In vitro labelling with organic dyes

Despite the numerous advantages afforded by in vivo imaging with fluorescent proteins, these approaches, as discussed above, have limitations. Another disadvantage is that the known fluorescent proteins are relatively large (27 kDa in monomeric form) tags for protein labelling. Although there is keen interest in 'downsizing' fluorescent proteins, mutagenesis studies have not yet been successful. So far, the most promising results for smaller protein labels have come from the use of small organic fluorophores such as fluorescein and rhodamine (<1 kDa), which can be placed at specific sites in proteins using elaborate protein chemistry labelling techniques (Fig. 1d). An important benefit of using small organic fluorophores is that it minimizes possible steric hindrance problems that can interfere with protein function. In addition, site-specific attachment of the fluorophores to proteins can permit changes in the local environment, or the distances between labelled sites, to be assessed when these molecules are used for FRET^28,29. For example, FRET was used as a spectroscopic ruler to determine the distances between protein regions labelled with small fluorophores in the Shaker K⁺ channel, and demonstrated that the voltage-sensing segment of the channel twists during activation^30,31.

If membrane permeabilization or microinjection is possible, then this approach also allows fine control of the quantity of introduced fluorescently labelled proteins. Recent FRET studies have used small fluorophores as acceptors in combination with a GFP donor. For example, the combination of Cy3.5- and Cy3-labelled phospho-specific antibodies with GFP-fusion proteins has allowed the visualization of PKC-³² autophosphorylation and epidermal growth factor receptor ( EGFR)³³ activation, respectively, in fixed and permeabilized cell samples. The EGFR studies provided the first evidence for ligand-independent lateral propagation of receptor activation in the plasma membrane, a surprising result to many who had believed that ligand binding was essential for receptor activation. In another approach used to visualize activation of the RhoGTPase Rac, cultured cells expressing a GFP�Rac fusion were injected with a fragment of p21-activated kinase labelled with an Alexa-546 dye, which selectively binds GTP-bound GFP�Rac³⁴. This study revealed the spatial control of growth factor-induced Rac activation in membrane ruffles, which forms an activation gradient at the leading edge of motile cells.

In vivo labelling with organic dyes

Recently, two innovative techniques have been developed for labelling specific recombinant proteins with small organic fluorophores within live cells^35,36: the bi-arsenic fluorophore labelling of proteins that have been genetically altered to contain tetra-cysteine motifs, and the labelling of proteins fused to O⁶-alkylguanine-DNA alkyltransferase with enzymatic substrate derivatives.

In the first technique, Tsien and colleagues made use of the well-known affinity of arsenoxides for closely spaced cysteine pairs^35,37. Two arsenoxide groups were introduced into fluorescein to form FlAsH, which binds with high affinity to tetra-cysteines containing the rare sequence CCXXCC (Fig. 1e). Therefore, a host protein of interest can be genetically fused to a short peptide of 6�20 amino acids containing the CCXXCC motif, and this construct can then be produced inside cells. The FlAsH label is membrane-permeant and non-fluorescent, acquiring fluorescence only on binding to the CCXXCC motif. Importantly, this property of the compound significantly decreases the background signal generated by unbound fluorophores.

Various derivatives of FlAsH can be designed by chemical modification of the original compound¹⁴. For example, a red analogue of FlAsH has been synthesized using the red fluorophore resorufin, and is termed ReAsH. Using the combination of FlAsH and ReAsH, Gaietta et al. determined the mechanism by which connexin 43 (Cx43), a subunit of gap junction channels, is added to and removed from gap junction plaques (Fig. 3C)^38,39. By engineering the FlAsH/ReAsH-binding motif into the Cx43 protein and then alternately labelling cells with FlAsH or ReAsH, different pools of the protein could be followed over time. This partitioning between red and green fluorescence revealed the novel characteristics of Cx43 transport, assembly into channels and turnover. This study demonstrates the benefits of the FlAsH/ReAsH technique for studying protein ageing over any time frame, which is highly versatile compared with the green-to-red shifting E5 protein that matures in a fixed time frame.

Among the other potential applications of these fluorophores, new derivatives can be synthesized to incorporate other functionalities, such as photosensitizing groups, into recombinant proteins. In addition to its role in fluorescence labelling, a recent study of the synaptic vesicle protein synaptotagmin in Drosophila melanogaster neuronal synapses demonstrated that FlAsH itself is useful for the fluorophore-assisted light inactivation (FALI) of recombinant proteins⁴⁰.

More recently, there has been a report of a second technique that uses the enzymatic activity of human O⁶-alkylguanine-DNA alkyltransferase ( hAGT). hAGT irreversibly transfers the substrate alkyl group (an O⁶-benzylguanine (BG) derivative) to one of its cysteine residues³⁶. The mutant ^W160hAGT demonstrates increased activity against BG derivatives. Following the expression of a chimeric fusion of ^W160hAGT and a protein of interest, a membrane-permeable derivative of BG containing fluorescein, BGFL (O⁶-benzylguanine fluorescein), is added. Once inside the cells, BGFL is acted on by the ^W160hAGT-containing protein, which leads to specific substrate labelling with fluorescein (Figs 1f and 3D). Although this method seems to produce reliable labelling, there are two drawbacks. First, hAGT, at 207 amino acids in length, might be too large a fusion tag for many applications. Second, experiments on mammalian cells would need to be performed using AGT-deficient cell lines to avoid labelling of the endogenous AGT.

Other approaches have used the selective binding of a chemical ligand to its receptor protein to study pH regulation in different compartments along the secretory pathway^41,42. For example, synthesized membrane-permeable conjugates of a hapten and fluorescent pH probes were trapped by a single-chain antibody that had been expressed in the lumen of the organelles⁴¹. In another approach, biotin conjugates of fluorescent pH probes were targeted to the secretory compartments by the localized expression of chicken avidin⁴², which binds biotin tightly. However, further development of these chemical probes that are genetically targetable will need the exchange of more information and ideas between chemists and biologists.

Labelling with quantum dots

In addition to small organic fluorophores, semiconductor nanocrystals (quantum dots) represent a promising new fluorescent label, owing to their photostability and wide range of excitation and emission wavelengths⁴³. But despite their advantages over organic fluorophores and fluorescent proteins, the use of quantum dots has so far been limited by their lack of biocompatibility. New advances in surface coating chemistry, however, have helped to overcome these problems to allow long-term, multi-colour imaging of live cells^44,45,46.

Quantum dots are semiconductor nanocrystalline particles, typically measuring 2�10 nm in size (roughly the size of typical proteins). They provide several important advantages over organic fluorophores and fluorescent proteins, including narrow, symmetrical and tuneable emission spectra that can be varied according to the size and material composition of the particles. This property allows flexible and close spacing of different quantum dots without substantial spectral overlap. In addition, their absorption spectra are broad, which makes it possible to excite all quantum dot colour variants simultaneously using a single excitation wavelength, thereby minimizing sample autofluorescence. Last, they have exceptional photostability.

Quantum dots are initially synthesized with hydrophobic organic ligands at their surface. For use in aqueous biological conditions, however, these organophilic species must be exchanged for ones that are more polar to prevent their aggregation and nonspecific adsorption in biological samples. Recent advances in nanomaterials have allowed quantum dots to be conjugated to biorecognition molecules^44,45, such as streptavidin (Figs 1g and 3E) and antibodies; these conjugates have been used on both fixed cells and tissue sections. In addition, cell-surface proteins and the endocytic compartments of live cells have been labelled with quantum dot bioconjugates. More recently, quantum dots encapsulated in phospholipid micelles were injected into Xenopus laevis embryos, and their fluorescence was followed until the tadpole stage in a cell-autonomous manner, which illustrates that quantum dots are stable and non-toxic inside cytosolic compartments⁴⁶.

Labelling for cellular imaging

Live cell imaging has become more accessible to researchers, largely as a result of recent advances in the techniques for fluorescence labelling of proteins by gene transfer. However, general properties must be considered when developing new methods for labelling proteins. Quantitative and physiological imaging requires that cells and subcellular structures are loaded with amounts of fluorescence labelled proteins that elicit only a minimal perturbation of normal cell processes, while maintaining a favourable signal-to-noise ratio. In this context, the integrity of the labelled protein is crucial. It is also essential that the fluorophores fluoresce at a high efficiency and that the act of labelling does not disrupt the biochemical function or cellular localization of the host protein.

Each of the techniques that we have reviewed has advantages and disadvantages, which must be considered and weighed for each application. For example, the inability to control the amount of protein produced by exogenous means can be a disadvantage of gene-transfer techniques. In this regard, the more traditional introduction of fluorescence-labelled soluble proteins through a permeabilized plasma membrane or a glass pipette is superior, particularly for FRET analyses, for which a 1:1 stoichiometry between two distinct fluorescently labelled proteins is desirable. Clearly, it is imperative to evaluate the potential and limitations of each fluorescence-labelling technique so as to use them to their fullest capacity and to derive the greatest possible benefit from them. With these conditions met, the acquisition of images of cells labelled with two or more fluorophores might be a key tool to decipher the complexity of the spatial and temporal behaviour of cellular events.

Acknowledgements

We thank Dr Ulrich Nienhaus for providing the spectral data of eqFP611. This work was partly supported by grants from CREST of JST (Japan Science and Technology), and the Japanese Ministry of Education, Science and Technology.

References

1.	Shimomura, O., Johnson, F. H. & Saiga, Y. Extraction, purification and properties of aequorin, a bioluminescent protein from the luminous hydromedusan, Aequorea. J. Cell Comp. Physiol. 59, 223�240 (1962).
2.	Prasher, D. C., Eckenrode, V. K., Ward, W. W., Prendergast, F. G. & Cormier, M. J. Primary structure of the Aequorea victoria green fluorescent protein. Gene 111, 229�233 (1992). | Article | | PubMed |
3.	Chalfie, M., Tu, Y., Euskirchen, G., Ward, W. W. & Prasher, D. C. Green fluorescent protein as a marker for gene expression. Science 263, 802�805 (1994). | PubMed |
4.	Tsien, R. Y. The green fluorescent protein. Annu. Rev. Biochem. 67, 509�544 (1998). | Article | | PubMed |
5.	Matz, M. V. et al. Fluorescent proteins from nonbioluminescent Anthozoa species. Nature Biotechnol. 17, 969�973 (1999). | Article | | PubMed |
6.	Labas, Y. A. et al. Diversity and evolution of the green fluorescent protein family. Proc. Natl Acad. Sci. USA 99, 4256�4261 (2002). | Article | | PubMed |
7.	Lippincott-Schwartz, J. & Patterson, G. H. Development and use of fluorescent protein markers in living cells. Science 300, 87�91 (2003). | Article | | PubMed |
8.	Baird, G. S., Zacharias, D. A. & Tsien, R. Y. Biochemistry, mutagenesis, and oligomerization of DsRed, a red fluorescent protein from coral. Proc. Natl Acad. Sci. USA 99, 11984�11989 (2000). | Article |
9.	Gurskaya, N. G. et al. GFP-like chromoproteins as a source of far-red fluorescent proteins. FEBS Lett. 507, 16�20 (2001). | Article | | PubMed |
10.	Wiedenmann, J. S. et al. A far-red fluorescent protein with fast maturation and reduced oligomerization tendency from Entacmaea quadricolor (Anthozoa, Actinaria). Proc. Natl Acad. Sci. USA 99, 11646�11651 (2002). | Article | | PubMed |
11.	Orm�, M. et al. Crystal structure of the Aequorea victoria green fluorescent protein. Science 273, 1392�1395 (1996). | PubMed |
12.	Siegel, M. S. & Isacoff, E. Y. A genetically encoded optical probe of membrane voltage. Neuron 19, 735�741 (1997). | PubMed |
13.	Baird, G. S., Zacharias, D. A. & Tsien, R. Y. Circular permutation and receptor insertion within green fluorescent proteins. Proc. Natl Acad. Sci. USA 96, 11241�11246 (1999). | Article | | PubMed |
14.	Zhang, J., Campbell, R. E., Ting, A. Y. & Tsien, R. Y. Creating new fluorescent probes for cell biology. Nature Rev. Mol. Cell Biol. 3, 906�918 (2002). | Article | | PubMed |
15.	Sacchetti, A. & Alberti, S. Protein tags enhance GFP folding in eukaryotic cells. Nature Biotechnol. 17, 1046 (1999). | Article | | PubMed |
16.	Griesbeck, O., Baird, G. S., Campbell, R. E., Zacharias, D. A. & Tsien, R. Y. Reducing the environmental sensitivity of yellow fluorescent protein. J. Biol. Chem. 276, 29188�29194 (2001). | Article | | PubMed |
17.	Nagai, T. et al. A variant of yellow fluorescent protein with fast and efficient maturation for cell-biological applications. Nature Biotechnol. 20, 87�90 (2002). | Article | | PubMed |
18.	Gross, L. A., Baird, G. S., Hoffman, R. C., Baldridge, K. K. & Tsien, R. Y. The structure of the chromophore within DsRed, a red fluorescent protein from coral. Proc. Natl Acad. Sci. USA 97, 11990�11995 (2000). | Article | | PubMed |
19.	Bevis, B. J. & Glick, B. S. Rapidly maturing variants of the Discosoma red fluorescent protein (DsRed). Nature Biotechnol. 20, 83�87 (2002). | Article | | PubMed |
20.	Terskikh, A. V., Fradkov, A. F., Zaraisky, A. G., Kajava, A. V. & Angres, B. Analysis of DsRed mutants. Space around the fluorophore accelerates fluorescence development. J. Biol. Chem. 277, 7633�7636 (2002). | Article | | PubMed |
21.	Terskikh, A. V. et al. 'Fluorescent timer': protein that changes color with time. Science 290, 1585�1588 (2000). | Article | | PubMed |
22.	Duncan, R. R. et al. Functional and spatial segregation of secretory vesicle pools according to vesicle age. Nature 422, 176�180 (2003). | Article |
23.	Patterson, G. H. & Lippincott-Schwartz, J. A photoactivatable GFP for selective photolabeling of proteins and cells. Science 297, 1873�1877 (2002). | Article | | PubMed |
24.	Ando, R., Hama, H., Yamamoto-Hino, M., Mizuno, H. & Miyawaki, A. An optical marker based on the UV-induced green-to-red photoconversion of a fluorescent protein. Proc. Natl Acad. Sci. USA 99, 12651�12656 (2002). | Article | | PubMed |
25.	Chudakov, D. M. et al. Kindling fluorescent proteins for precise in vivo photolabeling. Nature Biotechnol. 21, 191�194 (2003). | Article |
26.	Campbell, R. E. et al. A monomeric red fluorescent protein. Proc. Natl Acad. Sci. USA 99, 7877�7882 (2002). | Article | | PubMed |
27.	Yanushevich, Y. G. et al. A strategy for the generation of non-aggregating mutants of Anthozoa fluorescent proteins. FEBS Lett. 511, 11�14 (2002). | Article | | PubMed |
28.	Stryer, L. Fluorescence energy transfer as a spectroscopic ruler. Annu. Rev. Biochem. 47, 819�846 (1978). | Article | | PubMed |
29.	Lakowicz, J. R. in Principles of Fluorescence Spectroscopy (ed. Lakowicz, J. R.) 368�391 (Kluwer Academic/Plenum, New York, 1999).
30.	Cha, A., Snyder, G. E., Selvin, P. R. & Bezanilla, F. Atomic scale movement of the voltage-sensing region in a potassium channel measured via spectroscopy. Nature 402, 809�813 (1999). | Article | | PubMed |
31.	Glauner, K. S., Mannuzzu, L. M., Gandhi, C. S. & Isacoff, E. Y. Spectroscopic mapping of voltage sensor movement in the Shaker potassium channel. Nature 402, 813�817 (1999). | Article | | PubMed |
32.	Ng, T. et al. Imaging protein kinase C- activation in cells. Science 183, 2085�2089 (1999). | Article |
33.	Verveer, P. J., Wouters, F. S., Reynolds, A. R. & Bastiaens, P. I. H. Quantitative imaging of lateral ErbB1 receptor signal propagation in the plasma membrane. Science 290, 1567�1570 (2000). | Article | | PubMed |
34.	Kraynov, V. S. et al. Localized Rac activation dynamics visualized in living cells. Science 290, 333�337 (2000). | Article | | PubMed |
35.	Griffin, B. A., Adams, S. R. & Tsien, R. Y. Specific covalent labeling of recombinant protein molecules inside live cells. Science 281, 269�272 (1998). | Article | | PubMed |
36.	Keppler, A. et al. A general method for the covalent labeling of fusion proteins with small molecules in vivo. Nature Biotechnol. 21, 86�89 (2003). | Article |
37.	Griffin, B. A., Adams, S. R. & Tsien, R. Y. Fluorescent labeling of recombinant proteins in living cells with FlAsH. Methods Enzymol. 327, 565�578 (2000).
38.	Gaietta, G. et al. Multicolor and electron microscopic imaging of connexin trafficking. Science 296, 503�507 (2002). | Article | | PubMed |
39.	Falk, M. M. Genetic tags for labeling live cells: gap junctions and beyond. Trends Cell Biol. 12, 399�404 (2002). | Article |
40.	Marek, K. W. & Davis, G. W. Transgenically encoded protein photoinactivation (FlAsH-FALI): acute inactivation of synaptotagmin I. Neuron 36, 805�813 (2002).
41.	Farinas, J. & Verkman, A. S. Receptor-mediated targeting of fluorescent probes in living cells. J. Biol. Chem. 274, 7603�7606 (1999). | Article | | PubMed |
42.	Wu, M. M. et al. Mechanisms of pH regulation in the regulated secretory pathway. J. Biol. Chem. 276, 33027�33035 (2001). | Article | | PubMed |
43.	Chan, W. C. et al. Luminescent quantum dots for multiplexed biological detection and imaging. Curr. Opin. Biotechnol. 13, 40�46 (2002). | Article | | PubMed |
44.	Wu, X. et al. Immunofluorescent labeling of cancer marker Her2 and other cellular targets with semiconductor quantum dots. Nature Biotechnol. 21, 41�46 (2003). | Article | | PubMed |
45.	Jaiswal, J. K., Mattoussi, H., Mauro, J. M. & Simon, S. M. Long-term multiple color imaging of live cells using quantum dot bioconjugates. Nature Biotechnol. 21, 47�51 (2003). | Article | | PubMed |
46.	Dubertret, B. et al. In vivo imaging of quantum dots encapsulated in phospholipid micelles. Science 298, 1759�1762 (2002). | Article | | PubMed |