Probes for Multiplexed Detection of GFP-Expressing Cells

The Green Fluorescent Protein (GFP) reporter has added a new dimension to the analysis of protein localization, allowing real-time examination in live cells of processes that have conventionally been observed through immunocytochemical "snapshots" in fixed specimens.ref Using spectrally distinct, organic fluorescent probes and markers (Table 1) adds extra data dimensions and reference points to these experiments (Figure 1).

Bacillus subtilis  Figure 1. The morphology of sporulating Bacillus subtilis in the early stages of forespore engulfment. The membranes and chromosomes of both the forespore and the larger mother cell are stained with FM 4-64 (red; T3166, T13320) and DAPI (blue; D1306, D3571, D21490), respectively. The small green-fluorescent patch indicates the localization of a GFP fusion to SPoIIIE, a protein essential for translocation of the forespore chromosome that may also regulate membrane fusion events (see Proc Natl Acad Sci U S A (1999) 96:14553). The background contains sporangia at various stages in the engulfment process stained with MitoTracker Green FM (green, M7514) and FM 4-64 (red).

The majority of the applications summarized in Table 1 involve live cells, tissues and organisms. There are many other instances where research objectives call for complementary use of immunochemical and GFP-based protein localization techniques. These experiments demand the combination of brightness, photostability and spectral separation provided by our Alexa Fluor dye–labeled secondary detection reagents. For two-color combinations with GFP, we recommend the Alexa Fluor 555, Alexa Fluor 568 or Alexa Fluor 594 dye–labeled secondary antibodies (Secondary Immunoreagents—Section 7.2, Summary of Molecular Probes secondary antibody conjugates—Table 7.1). The addition of Alexa Fluor 635 or Alexa Fluor 647 dye–labeled antibodies allows three-color detection. Some immunohistochemical procedures such as paraffin embedding of fixed tissue result in loss of the intrinsic fluorescence of GFP. In other cases, GFP expression levels may simply be too low for detection above background autofluorescence.ref Antibodies to GFP provide remedies for these problems (Figure 2). We offer unlabeled mouse and rabbit monoclonal GFP antibodies and unlabeled rabbit and chicken polyclonal GFP antibodies, as well as Alexa Fluor dye–labeled rabbit polyclonal GFP antibodies (Antibodies against Expression Tags—Section 7.5).

pShooter pCMV/myc/mito/GFP
Figure 2. HeLa cell transfected with pShooter pCMV/myc/mito/GFP, then fixed and permeabilized. Green Fluorescent Protein (GFP) localized in the mitochondria was labeled with mouse IgG2a anti-GFP antibody (A11120) and detected with orange-fluorescent Alexa Fluor 555 goat anti–mouse IgG antibody (A21422), which colocalized with the dim GFP fluorescence. F-actin was labeled with green-fluorescent Alexa Fluor 488 phalloidin (A12379), and the nucleus was stained with blue-fluorescent DAPI (D1306, D3571, D21490). The sample was mounted using ProLong Gold antifade reagent (P36930). Some GFP fluorescence is retained in the mitochondria after fixation (left), but immunolabeling and detection greatly improve visualization (right).

Alexa Fluor Dyes: Highly Fluorescent FRET Acceptors

Proximity-dependent fluorescence resonance energy transfer (FRET) allows detection of protein–protein interactions with much higher spatial resolution than conventional diffraction-limited microscopy.ref Alexa Fluor dyes with strong absorption in the 500–600 nm wavelength range are excellent FRET acceptors from GFP (Table 2). An assay to detect activation of GFP–GTPase fusions developed by researchers at Scripps Research Institute ref utilizes the GTPase-binding domain (PBD) of PAK1, a protein that binds to GTPases only in their activated GTP-bound form. GTPase activation is indicated by FRET from GFP to PDB labeled with Alexa Fluor 546 C5-maleimide at a single N-terminal cysteine residue. This assay has been used to determine the location and dynamics of rac and Cdc42 GTPase activation in live cells.ref

Normalizing Expression and Translation Signals

In 2002, researchers in Scott Fraser's laboratory at the California Institute of Technology reported a method of coinjecting Texas Red dye–labeled 10,000 MW dextran and GFP vectors into sea urchin embryos. This method overcomes a multitude of problems inherent in making intra- and inter-embryo comparisons of gene expression levels using confocal microscopy. In particular, laser excitation and fluorescence collection efficiencies vary with the depth of the fluorescent protein in the embryo, and the orientation of different embryos on the coverslip varies relative to the microscope objective. Measuring the ratio of GFP fluorescence to Texas Red dextran fluorescence corrects for these spatial factors, providing a gene expression readout that is 2–50 times more accurate than conventional confocal microscopy procedures depending on the localization of GFP within an embryo.ref A similar strategy was previously used to determine translation efficiencies of GFP-encoding mRNAs.ref

Table 1. Probes for multiplexed detection of GFP-expressing cells.*

TargetProbeCat. No.Ex/Em †GFP Fusion PartnerSpecimenReference
Physiological Indicators
Intracellular Ca2+Fura-2 AMF1201, F1221, F1225, F14185335/505 ‡Protein kinase C (PKC)BHK cellsBiochem J (1999) 337:211
Intracellular Ca2+X-Rhod-1 AMX14210580/602Trpm5 (melastatin-related cation channel)CHO cellsNat Neurosci (2002) 5:1169
Intracellular Ca2+Fura Red AMF3020, F3021488/650GFP expressed specifically in pancreatic β-cellsMouse pancreatic isletsAm J Physiol Endocrinol Metab (2003) 284:E177
Intracellular pH5-(and 6-)Carboxy SNARF-1 AM ester acetateC1271568/635Human growth hormone (hGH)RIN1046-38 insulinoma cellsAm J Physiol Cell Physiol (2002) 283:C429
Mitochondrial membrane potentialTMRMT668555/580Cytochrome cMCF-7 human breast carcinoma, HeLaJ Cell Sci (2003) 116:525
Superoxide (O2)DihydroethidiumD1168518/605Cytochrome cMCF-7 human breast carcinomaJ Biol Chem (2003) 278:12645
Synaptic activityFM 4-64T3166, T13320506/750 §VAMP (vesicle-associated membrane protein)Rat hippocampal neuronsNat Neurosci (2000) 3:445
Receptors and Endocytosis
Acetylcholine receptorTetramethylrhodamine α-bungarotoxinT1175553/577Rapsyn (receptor-aggregating protein)ZebrafishJ Neurosci (2001) 21:5439
Epidermal growth factor (EGF)Rhodamine EGFE3481555/581EGF receptorMTLn3 rat mammary adenocarcinomaMol Biol Cell (2000) 11:3873
EndosomesTransferrin from human serum, Alexa Fluor 546 conjugateT23364556/573β2-adrenergic receptor (β2AR)HEK 293, rat hippocampal neuronsBrain Res (2003) 984:21
EndosomesTransferrin from human serum, Alexa Fluor 568 conjugateT23365578/603PrPc (cellular prion protein)SN56 cellsJ Biol Chem (2002) 277:33311
EndosomesFM 4-64T3166, T13320506/750 §PrPc (cellular prion protein)SN56 cellsJ Biol Chem (2002) 277:33311
Endoplasmic reticulumER-Tracker Blue-White DPXE12353375/520 ‡HSD17B7 gene product (3-ketosteroid reductase)HeLa, NIH 3T3Mol Endocrinol (2003) 17:1715
Golgi complexBODIPY TR ceramideD7540589/617PrPc (cellular prion protein)SN56 cellsJ Biol Chem (2002) 277:33311
LysosomesLysoTracker RedL7528577/590HeparanasePrimary human fibroblasts, MDA-231 (human breast carcinoma)Exp Cell Res (2002) 281:50
MitochondriaMitoTracker RedM7512578/599Sam5p (mitochondrial carrier for S-adenosylmethionine)Yeast (Saccharomyces cerevisiae)EMBO J (2003) 22:5975
Nuclear DNADAPID1306, D3571, D21490358/461Histone H2BHeLaMethods (2003) 29:42
Nuclear DNAHoechst 33342H1399, H3570, H21492350/461Histone H1BALB/c 3T3 fibroblastsNature (2000) 408:877
Nuclear DNASYTO 17S7579621/634HIV-1 integraseHeLaJ Biol Chem (2003) 278:33528
Nuclear DNASYTO 59S11341622/645Microtubule plus-end binding proteinPorcine kidney epithelial cells (LLCPK)Mol Biol Cell (2003) 14:916
Nuclear DNATO-PRO-3T3605642/661Citron kinaseHeLaJ Cell Sci (2001) 114:3273
Plasma membraneDiID282, D3911, N22880549/565SynaptobrevinXenopus optic neuronsNat Neurosci (2001) 4:1093
Other Subcellular Structures
F-actinRhodamine phalloidinR415554/573ERM (ezrin-radixin-moesin) proteinsHuman peripheral blood T cells (PBT)Nat Immunol (2004) 5:272
F-actinAlexa Fluor 568 phalloidinA12380578/603CalponinNIH 3T3J Cell Sci (2000) 113:3725
Lipid raftsCholera toxin subunit B (recombinant), Alexa Fluor 594 conjugateC22842590/617Histocompatibility leukocyte antigen (HLA)-Cw4NK cell–B-cell immunological synapseProc Natl Acad Sci U S A (2001) 98:14547
Cell Classification Markers
Apoptotic cellsAnnexin V, Alexa Fluor 594 conjugateA13203590/617GRASP65 (Golgi stacking protein)HeLaJ Cell Biol (2002) 156:495
Transformed B lymphocytes (Raji cells)CellTracker Orange CMTMRC2927550/575ICAM-3 (intercellular adhesion molecule-3)T-lymphocytes and antigen-presenting B cellsNat Immunol (2002) 3:159
Cell-surface antigensR-Phycoerythrin (streptavidin conjugate)S866, S21388565/575GFP gene expressionNIH 3T3Cytometry (1996) 25:211
NeuronsNeuroTrace 530/615 red-fluorescent Nissl stainN21482530/620Tau microtubule-binding protein (Purkinje cell marker)Mouse brain sliceJ Neurosci (2003) 23:6392
NeuronsAlexa Fluor 594 hydrazideA10438, A10442588/613SynaptophysinAplysia californica sensory neuronsNeuron (2003) 40:151
* This list covers only Aequoria victoria GFP, optimized mutants (e.g., EGFP) and green-fluorescent proteins from other species (e.g., Renilla reniformis). Fluorescent proteins with distinctly different excitation and emission characteristics (CFP, YFP, dsRed, etc.) are not included. † Fluorescence excitation (Ex) and emission (Em) maxima, in nm. ‡ Simultaneous imaging of GFP with fura-2 or ER-Tracker Blue-White DPX requires excitation wavelength–switching capability, because the fluorescence emission spectra overlap extensively. Even under these conditions, signal bleedthrough from one detection channel to the other may still be problematic, depending on the expression level and localization of the GFP chimera. See Biochem J (2001) 356:345 for further discussion. § The fluorescence emission spectra of styryl dyes such as FM 1-43 and FM 4-64 are broad and extend into the green emission range of GFP. In some cases, FM dye emission can overspill into the GFP detection channel, causing degraded resolution of image features. The excitation and emission spectra of FM 1-43 overlap those of GFP more extensively than those of FM 4-64. Therefore, using FM 4-64 instead of FM 1-43 is recommended to minimize this problem.

Table 2. R0 values for FRET from EGFP to Alexa Fluor dyes.

Acceptor DyeR0 (Å)*
Alexa Fluor 546 dye57
Alexa Fluor 555 dye63
Alexa Fluor 568 dye54
Alexa Fluor 594 dye53
* R0 values in angstroms (Å) represent the distance at which fluorescence resonance energy transfer from the donor dye to the acceptor dye is 50% efficient. Values were calculated from spectroscopic data as outlined (Fluorescence Resonance Energy Transfer (FRET)—Note 1.2 ).