Quantitative linear unmixing of CFP and YFP from spectral images acquired with two-photon excitation.

BACKGROUND: Spectrally distinct fluorescent proteins (FPs) have been developed permitting the visualization of several proteins simultaneously in living cells. The emission spectra of FPs, in most cases, overlap, making signal separation based on filter technology inefficient and in cases of bleed-through, inaccurate. Spectral imaging can overcome these obstacles through a process called linear unmixing. Given a complex spectra composed of multiple fluorophores, linear unmixing can reduce the complex signal to its individual, weighted, component spectra. Spectral imaging with two-photon excitation allows the collection of nontruncated emission spectra. The accuracy of linear unmixing under these conditions needs to be evaluated. METHODS: Capillaries containing defined mixtures of CFP and YFP were used to test the accuracy of linear unmixing using spectral images obtained with two-photon excitation. RESULTS: Linear unmixing can be accurate when wavelength and power-matched reference spectra are provided to the algorithm. Linear unmixing errors can occur due to (1) excitation laser contamination of emission signals, (2) the presence of FRET, (3) poor selection of excitation wavelength, and (4) failure to background subtract reference spectra. CONCLUSIONS: Linear unmixing, when judiciously performed, can accurately measure the abundance of CFP and YFP in mixed samples, even when their relative intensities range from 90:1.     

Simultaneous Quantitative Live Cell Imaging of Multiple FRET-Based Biosensors

We have developed a novel method for multi-color spectral FRET analysis which is used to study a system of three independent FRET-based molecular sensors composed of the combinations of only three fluorescent proteins. This method is made possible by a novel routine for computing the 3-D excitation/emission spectral fingerprint of FRET from reference measurements of the donor and acceptor alone. By unmixing the 3D spectrum of the FRET sample, the total relative concentrations of the fluorophores and their scaled FRET efficiencies are directly measured, from which apparent FRET efficiencies can be computed. If the FRET sample is composed of intramolecular FRET sensors it is possible to determine the total relative concentration of the sensors and then estimate absolute FRET efficiency of each sensor. Using multiple tandem constructs with fixed FRET efficiency as well as FRET-based calcium sensors with novel fluorescent protein combinations we demonstrate that the computed FRET efficiencies are accurate and changes in these quantities occur without crosstalk. We provide an example of this method’s potential by demonstrating simultaneous imaging of spatially colocalized changes in [Ca²⁺], [cAMP], and PKA activity.     

Back Cover: J. Biophotonics 1/2012

Hyperspectral imaging microscopy of rat lung cryoslices can be used to identify individual pulmonary microvascular endothelial cells (PMVECs) in the presence of a high lung autofluorescence of the same peak fluorescence emission wavelength. PMVECs expressing green fluorescent protein are shown in green, lung autofluorescence is shown in red, and nuclei are shown in blue. A plot of the spectral library used for linear unmixing is also shown. (Picture: S. Leavesley et al., pp. 67–84 in this issue)     

Efficient Blind Spectral Unmixing of Fluorescently Labeled Samples Using Multi-Layer Non-Negative Matrix Factorization

The ample variety of labeling dyes and staining methods available in fluorescence microscopy has enabled biologists to advance in the understanding of living organisms at cellular and molecular level. When two or more fluorescent dyes are used in the same preparation, or one dye is used in the presence of autofluorescence, the separation of the fluorescent emissions can become problematic. Various approaches have been recently proposed to solve this problem. Among them, blind non-negative matrix factorization is gaining interest since it requires little assumptions about the spectra and concentration of the fluorochromes. In this paper, we propose a novel algorithm for blind spectral separation that addresses some of the shortcomings of existing solutions: namely, their dependency on the initialization and their slow convergence. We apply this new algorithm to two relevant problems in fluorescence microscopy: autofluorescence elimination and spectral unmixing of multi-labeled samples. Our results show that our new algorithm performs well when compared with the state-of-the-art approaches for a much faster implementation.     

Spectral imaging and its applications in live cell microscopy

In biological microscopy, the ever expanding range of applications requires quantitative approaches that analyze several distinct fluorescent molecules at the same time in the same sample. However, the spectral properties of the fluorescent proteins and dyes presently available set an upper limit to the number of molecules that can be detected simultaneously with common microscopy methods. Spectral imaging and linear unmixing extends the possibilities to discriminate distinct fluorophores with highly overlapping emission spectra and thus the possibilities of multicolor imaging. This method also offers advantages for fast multicolor time-lapse microscopy and fluorescence resonance energy transfer measurements in living samples. Here we discuss recent progress on the technical implementation of the method, its limitations and applications to the imaging of biological samples.     

Evaluation and optimization of multiple fluorophore analysis of a Pseudomonas aeruginosa biofilm

Conventional laser scanning microscopy for multiple fluorescent stains can be a useful tool if the problems of autofluorescence and cross-talk are eliminated. The technique of spectral imaging was employed to unmix five different fluorophores - ranging in emission from 435 to 665 nm - applied to a Pseudomonas aeruginosa biofilm with overlapping spectra and which was not possible using traditional channel mode operation. Using lambda scanning and linear unmixing, the five fluorophores could be distinguished with regions of differentiation apparent.     

Spectral Unmixing: Analysis of Performance in the Olfactory Bulb In Vivo

Background

The generation of transgenic mice expressing combinations of fluorescent proteins has greatly aided the reporting of activity and identification of specific neuronal populations. Methods capable of separating multiple overlapping fluorescence emission spectra, deep in the living brain, with high sensitivity and temporal resolution are therefore required. Here, we investigate to what extent spectral unmixing addresses these issues.

Methodology/Principal Findings

Using fluorescence resonance energy transfer (FRET)-based reporters, and two-photon laser scanning microscopy with synchronous multichannel detection, we report that spectral unmixing consistently improved FRET signal amplitude, both in vitro and in vivo. Our approach allows us to detect odor-evoked FRET transients 180–250 µm deep in the brain, the first demonstration of in vivo spectral imaging and unmixing of FRET signals at depths greater than a few tens of micrometer. Furthermore, we determine the reporter efficiency threshold for which FRET detection is improved by spectral unmixing.

Conclusions/Significance

Our method allows the detection of small spectral variations in depth in the living brain, which is essential for imaging efficiently transgenic animals expressing combination of multiple fluorescent proteins.     

A new configuration of the Zeiss LSM 510 for simultaneous optical separation of green and red fluorescent protein pairs.

The power and simplicity of genetically encoded fluorophores (fluorescent proteins, FPs) have drawn many molecular biologists to light microscopy. First generation FPs suffered from overlapping excitation and emission spectra, which limited their use together in pairs (Patterson et al., J Cell Sci 2001;114 (Part 5):837-838). Image acquisition and processing techniques, collectively known as linear unmixing, have been developed to separate overlapping fluorescence signals encountered in the imaging of FP pairs and also in FRET. These specialized techniques are not without their potential drawbacks, including limitations on sensitivity and time-resolution for live cell imaging, and the risk of artifact in the hands of nonspecialists. With the advent of a new generation of red-shifted FPs (Shaner et al., Nat Biotechnol 2004;22:1567-1572; Verkhusha and Lukyanov, Nat Biotechnol 2004;22:289-296) careful selection of excitation sources and emission filters obviate the need for linear unmixing when simple two channel imaging of FPs is required. Here we introduce a new configuration of the Zeiss LSM 510 laser scanning confocal microscope, optimized for live cell imaging of green fluorescent protein (GFP) together with spectral variants such as mRFP1 and mCherry using standard photo-multipliers. A 2 mW, 594 nm HeNe laser was chosen as the excitation source for the red FP. This wavelength efficiently excites the aforementioned red variants without limiting the detection range of GFP emission during simultaneous two-channel imaging. Compared to excitation of GFP and mCherry at 488 and 543 nm, excitation at 488 and 594 nm approximately doubles the sensitivity of GFP detection and eliminates bleed-through of GFP into the mCherry channel. However, sensitivity of mCherry detection is decreased by 30%, suggesting the need for red FPs having longer emission peaks. Practical advantages to the simultaneous optical separation of FPs with nonoverlapping emission spectra include simplicity, robustness, reduced risk of artifact, and increased sensitivity during live cell imaging.     

Simultaneous detection of multiple green fluorescent proteins in live cells by fluorescence lifetime imaging microscopy

The green fluorescent protein (GFP) has proven to be an excellent fluorescent marker for protein expression and localisation in living cells [1] [2] [3] [4] [5]. Several mutant GFPs with distinct fluorescence excitation and emission spectra have been engineered for intended use in multi-labelling experiments [6] [7] [8] [9]. Discrimination of these co-expressed GFP variants by wavelength is hampered, however, by a high degree of spectral overlap, low quantum efficiencies and extinction coefficients [10], or rapid photobleaching [6]. Using fluorescence lifetime imaging microscopy (FLIM) [11] [12] [13] [14] [15] [16], four GFP variants were shown to have distinguishable fluorescence lifetimes. Among these was a new variant (YFP5) with spectral characteristics reminiscent of yellow fluorescent protein [8] and a comparatively long fluorescence lifetime. The fluorescence intensities of co-expressed spectrally similar GFP variants (either alone or as fusion proteins) were separated using lifetime images obtained with FLIM at a single excitation wavelength and using a single broad band emission filter. Fluorescence lifetime imaging opens up an additional spectroscopic dimension to wavelength through which novel GFP variants can be selected to extend the number of protein processes that can be imaged simultaneously in cells.     

A fluorescent imaging method for analyzing the biodistribution of therapeutic monoclonal antibodies that can distinguish intact antibodies from their breakdown products

Many monoclonal antibodies have been developed for therapy over the last 2 decades. In the development of therapeutic antibodies, the preclinical assessment of an antibody's biodistribution is important for the prediction of the antibody's efficacy and safety. For imaging analyses of such biodistributions, radioisotope (RI) labeling and fluorescence labeling methods are typically used, but the resulting data are limited because these methods cannot distinguish breakdown products from intact antibodies. To resolve this problem, we developed a novel method using fluorescent resonance energy transfer (FRET)-type labeling and a spectral unmixing tool. With FRET-type labeling (labeling with 2 species of fluorophore), different fluorescence properties of labeled intact antibodies and their breakdown products (the hydrolyzed/digested type of breakdown products) are made visible. With the spectral unmixing tool, the fluorescence of a solution containing the intact antibody and its breakdown products could be unmixed in proportion to their contents. Moreover, when labeled antibodies that targeted either human epidermal growth factor receptor-2 or epidermal growth factor receptor were injected into nude mice implanted subcutaneously with tumor cells, the accumulation of the injected labeled antibodies and their breakdown products in the tumor could be separately analyzed by both whole-mouse imaging and a tumor homogenate analysis. These results suggest that our method using FRET-type labeling and a spectral unmixing tool could be useful in distinguishing breakdown products from intact antibodies.     

Seven-color fluorescence imaging of tissue samples based on Fourier spectroscopy and singular value decomposition.

Seven-color analyses of immunofluorescence-stained tissue samples were accomplished using Fourier spectroscopy-based hyperspectral imaging and singular value decomposition. This system consists of a combination of seven fluorescent dyes, three filtersets, an epifluorescence microscope, a spectral imaging system, a computer for data acquisition, and data analysis software. The spectra of all pixels in a multicolor image were taken simultaneously using a Sagnac type interferometer. The spectra were deconvolved to estimate the contribution of each component dye, and individual dye images were constructed based on the intensities of assigned signals. To obtain mixed spectra, three filter sets, i.e., Bl, Gr, and Rd for Alexa488 and Alexa532, for Alexa546, Alexa568, and Alexa594, and for Cy5 and Cy5.5, respectively, were used for simultaneous excitation of two or three dyes. These fluorophores have considerable spectral overlap which precludes their separation by conventional analysis. We resolved their relative contributions to the fluorescent signal by a method involving linear unmixing based on singular value decomposition of the matrices consisting of dye spectra. Analyses of mouse thymic tissues stained with seven different fluorescent dyes provided clear independent images, and any combination of two or three individual dye images could be used for constructing multicolor images.     

Reliability of confocal microscopy spectral imaging systems: use of multispectral beads.

BACKGROUND: There is a need for a standardized, impartial calibration, and validation protocol on confocal spectral imaging (CSI) microscope systems. To achieve this goal, it is necessary to have testing tools to provide a reproducible way to evaluate instrument performance. METHODS: We evaluated fluorescent spectral beads (FocalCheck) from Molecular Probes/Invitrogen that consist of four pairs with emissions between 500 and 725 nm and a europium macrocycle quantum dye bead. These bead tools compliment our previously published protocol for testing spectral imaging systems that used an inexpensive multi-ion discharge lamp (MIDL) that contains Hg(+), Ar(+), and inorganic fluorophores that emits distinct, stable spectral features. RESULTS: We acquired the spectra of the FocalCheck beads on a Zeiss 510 Meta, a Leica TCS-SP1, a Leica SP2 AOBS, an Olympus FV 1000, and a Nikon C1Si confocal systems and a PARISS microscopic spectral system and of the europium beads on the Leica TCS-SP1 and PARISS spectral imaging systems. A lack of performance with some equipment between 650 and 750 nm was identified using the far red pair of the FocalCheck beads. The position of the slider in front of PMT 2 that reflects light into PMT 1 and PMT 3 affected the measurement of all bead intensities. Unmixing algorithms were used to separate beads with different fluorochromes and separate two fluorochromes on the same bead. CONCLUSIONS: The FocalCheck multi-spectral beads yielded similar profiles on four CSI systems and a PARISS spectral system. The utilization of the spectral FocalCheck beads is helpful to evaluate proper spectral performance, especially in the far red region. Europium beads provide a very narrow spectrum that can help to identify machines that have spectral problems. The dyes located on individual beads or mixed together in ring-core configuration can be used as test particles to demonstrate spectral unmixing with various algorithms.     

Improved in vivo whole-animal detection limits of green fluorescent protein-expressing tumor lines by spectral fluorescence imaging

Green fluorescent protein (GFP) has been used for cell tracking and imaging gene expression in superficial or surgically exposed structures. However, in vivo murine imaging is often limited by several factors, including scatter and attenuation with depth and overlapping autofluorescence. The autofluorescence signals have spectral profiles that are markedly different from the GFP emission spectral profile. The use of spectral imaging allows separation and quantitation of these contributions to the total fluorescence signal seen in vivo by weighting known pure component profiles. Separation of relative GFP and autofluorescence signals is not readily possible using epifluorescent continuous-wave single excitation and emission bandpass imaging (EFI). To evaluate detection thresholds using these two methods, nude mice were subcutaneously injected with a series of GFP-expressing cells. For EFI, optimized excitation and emission bandpass filters were used. Owing to the ability to separate autofluorescence contributions from the emission signal using spectral imaging compared with the mixed contributions of GFP and autofluorescence in the emission signal recorded by the EFI system, we achieved a 300-fold improvement in the cellular detection limit. The detection limit was 3 x 10(3) cells for spectral imaging versus 1 x 10(6) cells for EFI. Despite contributions to image stacks from autofluorescence, a 100-fold dynamic range of cell number in the same image was readily visualized. Finally, spectral imaging was able to separate signal interference of red fluorescent protein from GFP images and vice versa. These findings demonstrate the utility of the approach in detecting low levels of multiple fluorescent markers for whole-animal in vivo applications.     

Cerulean, Venus, and VenusY67C FRET reference standards

F?rster's resonance energy transfer (FRET) can be used to study protein-protein interactions in living cells. Numerous methods to measure FRET have been devised and implemented; however, the accuracy of these methods is unknown, which makes interpretation of FRET efficiency values difficult if not impossible. This problem exists due to the lack of standards with known FRET efficiencies that can be used to validate FRET measurements. The advent of spectral variants of green fluorescent protein and easy access to cell transfection technology suggests a simple solution to this problem: the development of genetic constructs with known FRET efficiencies that can be replicated with high fidelity and freely distributed. In this study, fluorescent protein constructs with progressively larger separation distances between donors and acceptors were generated and FRET efficiencies were measured using fluorescence lifetime spectroscopy, sensitized acceptor emission, and spectral imaging. Since the results from each method were in good agreement, the FRET efficiency value of each construct could be determined with high accuracy and precision, thereby justifying their use as standards.     

Microspectroscopic imaging of nodulation factor-binding sites on living Vicia sativa roots using a novel bioactive fluorescent nodulation factor.

A novel bioactive fluorescent nodulation (Nod) factor, NodRlv-IV(BODIPY FL-C16), has been synthesized by attaching a BODIPY FL-C16 acyl chain to the primary amino group of chitotetraose deacetylated at the nonreducing terminus by recombinant NodB. The binding of the fluorescent Nod factor to root systems of Vicia sativa was investigated with fluorescence spectral imaging microscopy (FSPIM) and fluorescence ratio imaging microscopy (FRIM). Spatially resolved fluorescence spectra of living and labeled Vicia sativa root systems were measured by FSPIM. Strong autofluorescence, inherent to many plant systems when excited at 488 nm, was corrected for by utilizing the difference in fluorescence emission spectra of the autofluorescence and NodRlv-IV(BODIPY FL-C16). A methodology is presented to break down the in situ fluorescence emission spectra into spatially resolved autofluorescence and BODIPY FL fluorescence spectra. Furthermore, an FRIM method was developed for correcting autofluorescence in fluorescence micrographs for this system. After autofluorescence correction it was shown that NodRlv-IV(BODIPY FL-C16) was concentrated in the root hairs, but was also bound to other parts of the root surface.     

Surveillance of siRNA integrity by FRET imaging

Techniques for investigation of exogenous small interfering RNA (siRNA) after penetration of the cell are of substantial interest to the development of efficient transfection methods as well as to potential medical formulations of siRNA. A FRET-based visualization method including the commonplace dye labels fluorescein and tetramethylrhodamin (TMR) on opposing strands of siRNA was found compatible with RNA interference (RNAi). Investigation of spectral properties of three labelled siRNAs with differential FRET efficiencies in the cuvette, including pH dependence and FRET efficiency in lipophilic environments, identified the ratio of red and green fluorescence (R/G-ratio) as a sensitive parameter, which reliably identifies samples containing >90% un-degraded siRNA. Spectral imaging of siRNAs microinjected into cells showed emission spectra indistinguishable from those measured in the cuvette. These were used to establish a calibration curve for assessing the degradation state of siRNA in volume elements inside cells. An algorithm, applied to fluorescence images recorded in standard green and red fluorescence channels, produces R/G-ratio images of high spatial resolution, identifying volume elements in the cell with high populations of intact siRNA with high fidelity. To demonstrate the usefulness of this technique, the movement of intact siRNA molecules are observed after introduction into the cytosol by microinjection, standard transfection and lipofection with liposomes.     

Multicolor Whole-Cell Bacterial Sensing Using a Synchronous Fluorescence Spectroscopy-Based Approach

The wide collection of currently available fluorescent proteins (FPs) offers new possibilities for multicolor reporter gene-based studies of bacterial functions. However, the simultaneous use of multiple FPs is often limited by the bleed-through of their emission spectra. Here we introduce an original approach for detection and separation of multiple overlapping fluorescent signals from mixtures of bioreporters strains. The proposed method relies on the coupling of synchronous fluorescent spectroscopy (SFS) with blind spectral decomposition achieved by the Canonical Polyadic (CP) decomposition (also known as Candecomp/Parafac) of three-dimensional data arrays. Due to the substantial narrowing of FP emission spectra and sensitive detection of multiple FPs in a one-step scan, SFS reduced spectral overlap and improved the selectivity of the CP unmixing procedure. When tested on mixtures of labeled E. coli strains, the SFS/CP approach could easily extract the contribution of at least four overlapping FPs. Furthermore, it allowed to simultaneously monitor the expression of three iron responsive genes and pyoverdine production in P. aeruginosa. Implemented in a convenient microplate format, this multiplex fluorescent reporter method provides a useful tool to study complex processes with different variables in bacterial systems.     

Quantitative lifetime unmixing of multiexponentially decaying fluorophores using single-frequency fluorescence lifetime imaging microscopy

Fluorescence lifetime imaging microscopy (FLIM) is a quantitative microscopy technique for imaging nanosecond decay times of fluorophores. In the case of frequency-domain FLIM, several methods have been described to resolve the relative abundance of two fluorescent species with different fluorescence decay times. Thus far, single-frequency FLIM methods generally have been limited to quantifying two species with monoexponential decay. However, multiexponential decays are the norm rather than the exception, especially for fluorescent proteins and biological samples. Here, we describe a novel method for determining the fractional contribution in each pixel of an image of a sample containing two (multiexponentially) decaying species using single-frequency FLIM. We demonstrate that this technique allows the unmixing of binary mixtures of two spectrally identical cyan or green fluorescent proteins, each with multiexponential decay. Furthermore, because of their spectral identity, quantitative images of the relative molecular abundance of these fluorescent proteins can be generated that are independent of the microscope light path. The method is rigorously tested using samples of known composition and applied to live cell microscopy using cells expressing multiple (multiexponentially decaying) fluorescent proteins.     

Four-color, 4-D time-lapse confocal imaging of chick embryos

Our understanding of the molecular mechanisms that direct cell motility, cell division, and cell shaping has benefited from innovations in cell labeling and the ability to resolve intracellular dynamics with multispectral, high-resolution imaging. However, due to difficulties with in vivo cell marking and monitoring, most studies have been restricted to fixed tissue or cells in culture. Here, we report the delivery of multiple (up to four), multicolor fluorescent protein (FP) constructs and four-dimensional (4-D), multispectral time-lapse confocal imaging of cell movements in living chick embryos. Cell cytoskeletal components are fluorescently tagged after microinjection and electroporation of a cocktail of FP constructs into specific regions of chick embryos. We tested 11 different FP constructs in various two-, three-, and four-color combinations using multispectral imaging and linear unmixing to limit the crosstalk between different emission spectra. We monitored intracellular dynamics in individual multicolored migrating cells in vivo and developed a set of advantageous imaging parameters for 4-D time-lapse confocal microscopy. We find that the number of four-color labeled cells in a typical embryo is approximately 10% of the total number of fluorescently labeled cells; this value consistently increases showing that approximately 50% of the total labeled cells have only one-color. We find that multicolored cells are photostable for time-lapses of approximately 2-3 h. Thus, cell labeling with up to four FP color schemes combined with multispectral, 4-D confocal time-lapse imaging offers a powerful tool to simultaneously analyze cellular and molecular dynamics during chick embryogenesis.     

Spectral imaging microscopy web sites and data.

The Internet is enabling greater access to spectral imaging publications, spectral graphs, and data than that was available a generation ago. The spectral imaging systems discussed in this issue of Cytometry work because reagent and hardware spectra are reproducible, reusable, and provide input to spectral unmixing and spectral components recognition algorithms. These spectra need to be readily available in order to determine what to purchase, how to use it, and what the output means. We refer to several commercially sponsored and academic spectral web sites and discuss our spectral graphing and data sites. Sites include fluorescent dye graph servers from Invitrogen/Molecular Probes, BD Biosciences, Zeiss/Bio-Rad Cell Sciences, and filter set servers from Chroma Technology and Omega Optical. Several of these sites include data download capabilities. Recently, two microscope manufacturers have published on their web sites transmission curves for select objective lenses-crucial data for anyone doing multiphoton excitation microscopy. Notable among the academic sites, PhotoChemCAD 2.0 has over 200 dyes and a downloadable database/graphing program, and the USC-A Chemistry UV-vis Database displays absorption spectra of many dyes and indicators used in clinical histology and pathology. Our Fluorescent Spectra graphing/calculator site presents dyes, filters, and illumination data from many of these and additional sources. PubSpectra is our free download site which uses Microsoft Excel files as standardized human/machine readable format with over 2,000 biomedical spectra. The principle that data is not subject to copyright provides a framework in which all scientific data should be made freely accessible.