Use of Fluorescent Probes: Their Effect on Cell Biology and Limitations

Fluorescent molecules, known as fluorophores, respond distinctly to light. Each fluorophore has distinct characteristics, which can be used to determine which fluorophore to use for a given application or experimental system. Some proteins or small molecules in cells are naturally fluorescent; this is called intrinsic fluorescence or autofluorescence [e.g., green fluorescent protein (GFP)]. Proteins, nucleic acids, lipids, or small molecules can be labeled with an extrinsic fluorophore—a fluorescent dye—which can be a small molecule, a protein or a quantum dot (Fig. 1). This article discusses pitfalls and problems with a variety of fluorescent compounds that are currently used. Confocal image of a fixed, embedded temporal lobe from a preterm lamb that was supported by nasal high-frequency ventilation for 3 weeks. Immature oligodendrocytes (green; labeled with “O1” antibody) are shown. Proliferating cells are red (proliferating cell nuclear antigen). Many astrocytes have yellow nuclear regions, indicating colocalization of the cell-specific marker (green) with the proliferating cell marker (red). Blue shows the nuclear stain, DAPI. Laser scan time of 10 μs, using an Olympus BX41 confocal microscope. Scale bar is 20 μm. The image was provided by Jeremy Alvord (third-year medical student) and Kurt H. Albertine, Ph.D., University of Utah, School of Medicine, Salt Lake City, UT. There are many types of fluorescent proteins of different colors, which are currently used as fluorescent tags for bioimaging (Stepanenko et al., 2011). These fluorescent proteins have been successful in biological imaging because of their unique chemical structure. For example, a key feature of all fluorescent proteins is their ability to self-generate the intrinsic chromophore from three amino acids at positions 65–67 (numbering is given according to Aequorea victoria GFP), and this does not involve cofactors or enzymatic components. Fluorescent proteins and their engineered variants can be used for routine monitoring of gene activation, as well as the selective labeling and analysis of single proteins, cellular organelles, and even whole cells. Creation of a genetic in-frame fusion of a fluorescent protein to a protein of interest enables localization of that protein to specific tissues, cells or subcellular compartments. Tracking cellular proteins in vivo using fluorescent tags began with the development of GFP and its family members (Shimomura et al., 1962; Prasher et al., 1985; Chaflie et al., 1994; Matz et al., 1999; Shu et al., 2009). Other fluorescent proteins with emissions across the spectrum are still being discovered (Wallace and Muirhead, 2007). For choosing a fluorescent protein to use for imaging, a researcher must decide on which color variant to use. The most common fluorescent protein used in imaging of cells is GFP (Fig. 2). The GFP family proteins are relatively small, compact, and chemically inert. GFP emits bright green light with stimulation with ultraviolet or blue light. GFP is used as an indicator for gene activity, labeling proteins and subcellular compartments in live cells, and GFP-labeled cells can be tracked in tissues. Red fluorescent GFP-like proteins are naturally found in sea anemones. The first genes of GFP-like proteins, such as the red emitter dsRed, were isolated from anthozoa species (Wiedenmann et al., 2009). The color palette of these GFP-like proteins has been considerably extended, especially toward the far-red end of the spectrum, by protein engineering. Imaging of biological material with red fluorescent proteins allows better penetration of cells and tissues by long wavelength light and decreased cellular autofluorescence in the red emission range. Therefore, red fluorescent proteins are useful for whole body imaging; for example, for monitoring tumor progression in mouse models. Other commonly used fluorescent proteins are mCitrine/mVenus (green–yellow), TagBFP (blue), tdTomato (orange), mCerulean3 (cyan), mCherry (Fig. 2) and mApple (red), and mKate2 and mNeptune (far-red). Wound healing in live zebrafish. The images show zebrafish expressing GFP in neutrophils and mCherry in the entire blood cell lineage, which were wounded and imaged using a cooled couple device (CCD) camera every 10 min over a period of 12 hr. Macrophages (red) and neutrophils invade the wound area in the first few hours, and eventually clear the area of bacteria and debris to allow wound healing. The left panel is at time zero, and subsequent panels are 4 hr apart. Images were acquired using Metamorph (Universal Imaging), an Olympus IX81 microscope and a Hamamatsu Orca ER CCD camera. Images are courtesy of Mike Redd (Huntsman Cancer Institute) and Christopher Rodesch (University of Utah Fluorescence Microscopy Core Facility). The imaging of fluorescent proteins discussed in the previous paragraph relies on the autocatalytic formation of chromophores with oxygen present. However, the fluorescence properties of some fluorescent proteins are able to be modified by irradiation using light of specific wavelengths. These fluorescent proteins with light-modulated spectral properties are collectively termed photoactivatable fluorescent proteins (Patterson and Lippincott-Schwartz, 2002). Photoactivatable fluorescent proteins, for example, the PAmCherries, are nonfluorescent, but they show bright red fluorescence after ultraviolet–violet light illumination. One of the problems with GFP is that it may lose its fluorescence during tissue fixation or subsequent processing (Swenson et al., 2007). Therefore, immunostaining with commercial antibodies is often used for detecting GFP. GFP expression can vary considerably, even among similar cell types in a single animal. For example, GFP expression levels can vary among mice, and these levels can be increased or decreased with selective breeding (Brazelton and Blau, 2005). Before conducting transplantation or lineage-marking experiments, a thorough analysis of the expression pattern of GFP in appropriate control animals should be performed (Swenson et al., 2007). Different fluorescent proteins have different effects depending on the cell type. For example, enhanced GFP is useful for an ideal long-term expression tracer for hematopoietic cells, but DsRed-express fluorescent protein impairs the viability or growth of hematopoietic stem and progenitor cells (Tao et al., 2007). There are conflicting results on whether GFP is toxic to cells. First, aggregation of fluorescent proteins can lead to cellular toxicity. Second, exciting GFP for an extended time may generate free radicals that are toxic to cells. It has been claimed that most GFP applications have not resulted in overt toxicity (Brazelton and Blau, 2005). However, it has also been shown that GFP can induce apoptosis, which indicates a possible reason for the difficulty in establishing stable cell lines expressing GFP (Liu et al., 1999). The expression temperature can affect the cellular brightness of a fluorescent protein (Wiedenmann et al., 2009). For example, mEosFP can be employed successfully as a bright cellular marker in a range of organisms, including plants, drosophila or zebrafish, but no fluorescence is observed in mammalian cells cultured at 37.8°C. The experimental setup used might affect fluorescent proteins. A requirement for detection of fluorescent proteins in living cells and tissues is excitation with a light or laser source. Light, especially that of short wavelengths, can induce phototoxic effects in cells. Additionally, phototoxicity also depends on the total dose of incident light. Some of these negative effects associated with the exposure of cells to light can be minimized by using microscopy methods, such as two-photon microscopy, which uses infrared light for imaging and manipulation of fluorescent proteins, or microlens-enhanced dual spinning disk microscopy. The attachment of a fluorescent protein to a protein of interest often has no identifiable effects on the function, structure, and localization of a protein (Crivat and Taraska, 2012). However, in some cases, it can impair protein function and expression of this construct can adversely affect cellular function. Therefore, the nature of the expression system should be considered. A potentially large increase in protein expression level can have a large effect on the subcellular localization of a fusion protein and its cellular function. It is preferable for cells to express the smallest amount of fusion protein that can still be clearly imaged. Expression of a plain fluorescent protein in a cell may induce cytotoxic effects depending on what type of fluorescent protein is used. In conclusion, a large amount of fluorescent proteins are available, but none of them are well suited for all imaging purposes. Comparative tests should be performed to determine the ideal marker for an experiment. As well as using intrinsically fluorescent proteins related in structure or sequence to GFP, there are three other methods of covalently labeling a protein (Crivat and Taraska, 2012). First, self-labeling enzymes (Halo tags and SNAP/CLIP tags) can be used. These are used to overcome the limitations of fluorescent proteins, such as the fact that the photophysical properties of fluorescent proteins are not as good as organic dyes. For example, (1) fluorescent proteins can “blink” or emit light intermittently, (2) fluorescent proteins are not very bright, (3) some are not photostable, and (4) fluorescent proteins have limited colors and chemistry. This class of protein-based fusion tags can catalyze the covalent autoattachment of an organic fluorophore inside living cells. The enzymes are fused to a protein, and the pair is labeled by introduction of a fluorescent ligand, which covalently reacts with the fusion tag. These tags are similar in size to a fluorescent protein. The proteins are not innately fluorescent when expressed but become fluorescent when the cells are exposed to a fluorescent ligand. Second, an enzyme can be used to covalently attach a fluorophore ligand to another protein or peptide. An example of this is the small 9-kDa acyl carrier protein (ACP) tag, which can be covalently labeled with CoA derivatives by the enzyme phosphopantetheine transferase (PPTase) (George et al., 2004). The ACP fusion protein is incubated with the CoA derivatives, which are labeled with fluorescence. However, since substrates for this system are not membrane-permeable, use of ACP is restricted to extracellular proteins. Third, small cell-permeable biarsenical dyes have been developed. The smallest and most successful genetically encoded tag for covalent small-fluorophore labeling is the tetracysteine-biarsenical dyes, FLAsH and ReAsH. A short peptide sequence is genetically introduced into the sequence of a target protein. This sequence can specifically react with a membrane-permeable biarsenical dye. There are two advantages of biarsenical dyes. First, the targeting motifs are small. Therefore, there is less opportunity for introduced sequences to disrupt the overall fold and function of the labeled protein. The second advantage is that pulse-labeling procedures are easy. There are three problems when using biarsenical dyes: the level of background fluorescence, cellular toxicity of the ligands, and the effect of the tag on protein function or localization. Covalent labeling of cellular proteins can be achieved with aminoreactive molecules, such as the succinimide ester of carboxyfluorescein (CFSE) (Wallace and Muirhead, 2007). One of the most common uses of CFSE labeling has been in combination with antibodies against phenotypic and functional markers (e.g., lymphocyte subsets, intracellular cytokines, and hematopoietic lineage) to allow monitoring of proliferative responses in complex populations in vitro. CFSE has several advantages for in vitro or ex vivo tagging of immune cells. This method can be applied to most any cell type as long as toxicity due to overlabeling is avoided. Very high-intensity labeling can be achieved because a large fraction of cellular proteins becomes labeled. A limitation is that the specificity, affinity, and/or function of labeled proteins may be compromised if critical residues involved in receptor/coreceptor binding or antigen recognition sites are modified. This is because covalent protein labeling is random (any succinimide reactive residue that is accessible has the potential to become labeled). Another limitation is that, unlike stable genetic tags that have the same intensity in all daughter cells, CFSE intensity is halved with each round of cell division, which limits the length of time that daughter cells can be distinguished from unlabeled cells. Additionally, fixation, or the presence of necrotic or apoptotic cells, can lose labeled protein as fragmentation or shedding of apoptotic vesicles occurs. This can considerably distort CFSE fluorescence intensity distributions and complicate proliferation analysis. Reducing the concentration of CFSE used for labeling can limit some of the above problems, but as a result, the number of cell divisions that can be distinguished is reduced. Quantum dots are brighter and more photostable than organic dyes for labeling macromolecules and can therefore be used to complement the fluorophore and fluorescent protein-fusion techniques (Bruchez et al., 1998; Alivasatos, 2004; Ballou et al., 2004). Additionally, imaging by using quantum dots overcomes limitations of fluorescent proteins, such as possession of a narrow excitation spectrum and a broad emission spectrum. Quantum dots are nanometer-sized (1–10 nm) crystals composed of semiconductors. Fluorescent quantum dots are made of cadmium, an inorganic semiconductor material. Quantum dots are powerful probes used for fluorescence imaging including detection of disease, fluorescent assays for drug discovery, single protein tracking, and intracellular reporting. These dots possess several unique optical properties that are useful for in vivo imaging as follows. Quantum dots are very photostable, possess size-controlled fluorescence [i.e., the larger size dots (4 nm) emit red fluorescence, while the smaller size dots (2 nm) emit green fluorescence], have a long fluorescence lifetime, and are available in many colors, including in the near infrared region (Li et al., 2011). Another important property of quantum dots is their higher fluorescence intensity compared with organic dyes, which enables single molecular imaging, and they are more suitable for imaging in vitro and in vivo. They were one of the first nanotechnologies to be integrated with biology (Rosenthal et al., 2010). Quantum dots are considerably better than other methods for delivering a gene-silencing tool, called siRNA, into cells. Quantum dots must first be synthesized and their surface modified so that they are biologically compatible. Generally, a quantum dot has a three-layer structure composed of a core, shell and polymer coatings. The most common quantum dots used have a CdSe core that is usually coated with a zinc sulfide shell to improve photoluminescence of the quantum dots. Polymer coatings that surround the shell are used to make the materials soluble to water. The surface of quantum dots should be further modified so that the dots can be directed to a target. There are three main methods to target a biocompatible quantum dot: (1) using antibodies; (2) with peptides; and (3) using small molecules. Each of these approaches has advantages and disadvantages, and none of the approaches is universal for all applications. Quantum dots can be used for specific labeling of individual cell surface biomolecules, as well as for targeting intracellular single biomolecules, such as mRNA and molecular motors. The application of quantum dots is similar to that of organic fluorophores. After a ligand is attached to the label, this conjugate binds to its target receptor, which is then viewed by the fluorescence of the label using an optical microscope. For labeling only a few cells, fluorescence-labeled ligands can be injected into live cells using micropipettes. Electroporation can also be used (Parak et al., 2005). It is important to minimize nonspecific binding with quantum dots because this can lead to the misinterpretation of experimental results. Rigorous controls must also be performed in quantum dot experiments (Rosenthal et al., 2010). The largest problem associated with quantum dots is toxicity (Li et al., 2011). Because the nanomaterials of quantum dots are composed of heavy metals, they are potentially toxic during in vitro and in vivo imaging. Quantum dot toxicity depends on many parameters, such as size, shape, composition, charge, redox activity, concentration, surface coating, photostability, solubility, and exposure time. Toxicity is generally caused by two sources: (1) the composition of the semiconductor materials, especially the core composed of heavy metals; and (2) the generation of free radicals during excitation. The toxicity of quantum dots to cells and animals is time and dose dependent. Therefore, when imaging with a low concentration of quantum dots within a certain time, they can be ideal probes. Another issue with quantum dots is that the of coatings that the is larger than GFP or and larger than small organic dyes such as CFSE and (Wallace and Muirhead, 2007). Therefore, quantum dots are to cells by naturally (e.g., or to targeting molecules that in cellular Fluorescent dyes are usually to as but labeling of can also be dyes become fluorescent in a labeling with fluorescent dyes has the advantages of being and for any type of cell et al., Fluorescent probes can important on the structure and of structure, or et al., 2010). Fluorescent probes can be introduced in the It is possible to (e.g., and or to properties (e.g., and molecular within the Many different fluorescent dyes are available depending specific properties Some of these dyes are related to molecules, while not molecules. The most commonly used and are with long by into the of cell which the of function. The dyes are in the because of the of the dyes and of for example, has and and is a common for in vivo of ex vivo labeled stem cells and or to and fluorescence are stable for in cell However, in fluorescence are halved with each which allows proliferation monitoring in populations on A potential problem with fluorescent dyes is related to the fact that, because dyes exposure of labeled cells to that (e.g., and may also these dyes (Wallace and Muirhead, 2007). However, to many of these dyes has should be when using tissue fixation and processing Because labeling by concentration and cell concentration during the affect the each should determine the results of and to the than when with antibodies or dyes have some limitations that are similar to For example, long-term use is limited because intensity in to the number of cell intensity as several of than can complicate and color when other probes are used. Additionally, monitoring of proliferation by is by from the presence of apoptotic and necrotic cells. It has also been shown that the dyes and which are used to cells, blood cells, and different populations of hematopoietic stem and/or progenitor cells, et al., 2010). et the of the for each cell type and whether the fluorescent was still associated with the cells of interest in vivo and in vitro. They found that the of imaging is a key in In vitro that this can be large during short indicating the for of controls to the of A variety of should be when using dyes, such as the concentration its and of the used during It is also important to the viability of the cells, fluorescence and of the over time. In the of it is important to whether cell function is by the et the effect of fluorescent dyes on by on different fluorescent and any in the of the main by the of an such as a fluorescent The found that the effect of these probes on properties is However, the main the different effects was that some of them a of due to This issue can be limited by a in fluorescent of using and confocal microscopy is which can the of fluorescent images is called an because it can in the or when into the et the structure of and at the and using They found that the total area and of the and their were by the which that should be when the structure and properties of using extrinsic probes. can be used for These have many including into the which enables of genetic and the ability to and cells. A can be which becomes when a stem cell into other cells, and then imaging of a fluorescent protein, and is Fluorescent of live cells using has been in such as cells et al., and et al., 2009). In cells, the fluorescent genes enhanced green fluorescent protein and tdTomato are cytotoxic in some cells with are but the level of fluorescence is not the et al., 2010). It can be that certain gene are more appropriate for some cell lines than In has been observed in several cell that the gene combination must be for the cell type. might because of toxicity of the fluorescent protein. Other limitations of this system low and low The Christopher Mike Jeremy and Kurt H. for the