An excitation-scanning hyperspectral microscope for biomedical imaging of GFP in highly autofluorescent lung tissue.
Fluorescence microscopy has traditionally been performed using bandpass filters to isolate peak fluorescence excitation and emission wavelengths for fluorescent labels. However, the emission spectra of many labels share similar peak emission wavelengths, resulting in poor spectral separation. Additionally, highly autofluorescent signals commonly possess broad emission spectra that can obfuscate dimly fluorescent labels .
Hyperspectral fluorescence microscopy combines spectroscopy and imaging techniques to quantify and separate multiple fluorescent probes. Hyperspectral imaging samples many wavelength bands to collect a contiguous emission spectrum. Consequently, hyperspectral imaging is capable of separating fluorophores based on a unique spectrum, notably in the separation of fluorescent probes from autofluorescence . The traditional method of hyperspectral imaging filters the fluorescence emission in set increments across a broad wavelength range (emission scanning). Filtering can be accomplished using prisms , diffraction gratings , and tunable filters . We have previously shown the efficacy of thin-film tunable filters for filtering fluorescence emission of GFP from lung autofluorescence . However, emission-scanning hyperspectral imaging provides inherently low signal because the emission is sampled using narrow-bandwidth tunable filters. Consequently, emission scanning requires high acquisition times, making traditional hyperspectral approaches prohibitive for high-speed applications such as live- cell imaging or cell signaling.
We have developed a novel approach to hyperspectral imaging that overcomes the limitations of emission scanning by filtering the fluorescence excitation (excitation scanning), rather than fluorescence emission. Excitation scanning increases the available signal because the fluorescence emission is not filtered by a narrow-bandwidth filter. The increased signal results in shorter acquisition times, permitting high-speed hyperspectral imaging. Additionally, excitation scanning provides complimentary spectral information to emission scanning that can be used to better discriminate among multiple labels. In this work, we demonstrate the implementation of a novel excitation-scanning hyperspectral imaging microscope and compare the new system to an emission-scanning hyperspectral microscope using detection of GFP in highly autofluorescent lung tissue.
Cell, animal, and sample preparation was performed as described previously , . Briefly, cell samples were acquired from pulmonary microvascular endothelial cells (PMVECs) and transfected with a lentivirus encoding GFP. CD adult rats were infected intratracheally with P. aeruginosa one week prior to injection of GFP-positive PMVECs in the jugular vein. Rats were euthanized after one week and samples removed from the most injured portions of the lung. Lungs were paraffin-embedded, sectioned, and placed on microscope slides. Rats not injected with GFP were injected with saline and euthanized for autofluorescence controls. Tissue samples were stained with Hoechst-33342 (Life Technologies, Carlsbad, California). Confluent monolayers of GFP-positive PMVECs were prepared on coverslips to obtain a pure GFP spectrum. Samples of wild-type PMVECs were stained with Hoechst to collect a pure Hoecsht spectrum. A pure autofluorescence spectrum was obtained from unstained tissue slides.
An inverted fluorescent microscope (TE2000-U, Nikon Instruments) with a 40X oil-immersion objective (S Fluor, 40X/1.30 Oil DIC H/N2, Nikon Instruments) was used for imaging. Excitation scanning was performed using an array of thin-film tunable filters (VersaChrome, Semrock, Inc.) placed after a 300 watt Xe arc lamp (Titan 300, Sunoptic Technologies) in the optical lightpath (Fig. 1). A separate tunable filter array was positioned for emission scanning. A long-pass dichroic beamsplitter (BLP01-495R, Semrock, Inc.) and a long-pass emission filter (FF-495Di03, Semrock, Inc.) separated excitation from emission light. Excitation scanning image acquisition was performed from 360-480 nm, in 5 nm increments. At each excitation wavelength, a 495 nm long-pass emission filter separated excitation from emission light. Emission scanning image acquisition was performed from 470 to 700 nm, in 5 nm increments. Excitation light was filtered with a 420/17 nm bandpass filter. An acquisition time of 300 ms and an EMCCD gain of 3800 were used for both systems, for all samples, except for the GFP control. An EMCCD gain of 3600 was used due to the high signal intensity of the confluent monolayer of GFP. Background subtraction and wavelength-dependent attenuation were characterized for all images.
Excitation scanning provided higher sensitivity for detection of GFP than emission scanning (Fig. 2a-d). Additionally, excitation scanning resulted in increased delineation of nuclear regions and lung structure. Excitation scanning also provided complimentary spectral information to emission scanning spectra.
In this work, we have demonstrated a novel hyperspectral imaging system that filters the fluorescence excitation instead of the emission (excitation scanning), and compared this technique to an emission-scanning hyperspectral microscope. Excitation scanning provided higher sensitivity for detection of GFP, and increased delineation of nuclear regions and lung structure. Time-dependent spectral measurements are possible with excitation scanning due to higher signal detection. Additionally, excitation scanning provided complimentary spectral information to emission scanning. Combining both spectral data sets would increase available spectral information unique to different labels, and consequently increase spectral discrimination of several fluorophores present in a single study.
Current fluorescence microscopy techniques utilizing several bandpass filters result in poor discrimination of fluorophores when multiple labels are present. Hyperspectral imaging is a recent technique that permits quantitative delineation of multiple labels by filtering and measuring the fluorescence emission spectrum of each fluorophore. However, filtering the emission results in reduced signal detection and long acquisition times, prohibiting time-lapse hyperspectral imaging studies. In this work, we have developed a novel excitation scanning hyperspectral microscope. We have compared excitation scanning to emission scanning by determining the efficacy for detecting GFP in highly autofluorescent lung tissue. Excitation scanning provided higher sensitivity for GFP than emission scanning, and complimentary spectral information to emission spectra. Our future work will involve time-lapse fluorescence imaging studies of multiple labels using excitation scanning. We anticipate increased temporal resolution for live-cell imaging.
The authors would like to acknowledge support from NIH grant P01 HL066299, the Abraham Mitchell Cancer Research Fund, the Alabama Space Grant Consortium, and the University Committee on Undergraduate Research (UCUR). VersaChrome filters and tuning hardware for this study were provided by Semrock, Inc., a Unit of IDEX.
 J. R. Mansfield, K. W. Gossage, C. C. Hoyt, and R. M. Levenson, "Autofluorescence removal, multiplexing, and automated analysis methods for in-vivo fluorescence imaging," J. Biomed. Opt., vol. 10, no. 4, p. 041207, 2005.
 S. J. Leavesley, N. Annamdevula, J. Boni, S. Stocker, K. Grant, B. Troyanovsky, T. C. Rich, and D. F. Alvarez, "Hyperspectral imaging microscopy for identification and quantitative analysis of fluorescently-labeled cells in highly autofluorescent tissue," J. Biophotonics, vol. 5, no. 1, pp. 67-84, Jan. 2012.
 D. T. Dicker, J. M. Lerner, and W. S. El-Deiry, "Hyperspectral Image Analysis of Live Cells in Various Cell Cycle Stages," Cell Cycle, vol. 6, no. 20, pp. 2563-2570, Oct. 2007.
 J. N. Meyer, C. A. Lord, X. Y. Yang, E. A. Turner, A. R. Badireddy, S. M. Marinakos, A. Chilkoti, M. R. Wiesner, and M. Auffan, "Intracellular uptake and associated toxicity of silver nanoparticles in Caenorhabditis elegans," Aquat. Toxicol., vol. 100, no. 2, pp. 140-150, Oct. 2010.
 N. Gupta, "Acousto-optic-tunable-filter-based spectropolarimetric imagers for medical diagnostic applications--instrument design point of view," J. Biomed. Opt., vol. 10, no. 5, pp. 051802-051802, Sep. 2005.
 P. Favreau, T. Rich, A. Lindsey, D. Alvarez, P. Prashant, and S. J. Leavesley, "Thin-Film Tunable Filters for Hyperspectral Imaging of Lung Tissue," J. Biomed. Opt., Feb. 2013.
Peter Favreau (1,2), Thomas Rich (2,3), Ashley Stringfellow (2), Diego Alvarez (2,4), Prashant Prabhat (5), Silas Leavesley (1,3)
(1) Chemical and Biomolecular Engineering, University of South Alabama, (2) Center for Lung Biology, University of South Alabama, (3) Pharmacology, University of South Alabama, (4) Internal Medicine, University of South Alabama, (5) Semrock, Inc., A Unit of IDEX Corp.
|Printer friendly Cite/link Email Feedback|
|Author:||Favreau, Peter; Rich, Thomas; Stringfellow, Ashley; Alvarez, Diego; Prabhat, Prashant; Leavesley, Si|
|Publication:||Journal of the Mississippi Academy of Sciences|
|Date:||Apr 1, 2014|
|Previous Article:||Assessing the morphology of human gingival fibroblasts upon exposure to dental adhesives in the presence of porphyromonas gingivalis...|
|Next Article:||Hyperspectral FRET imaging and analysis approaches to determine cAMP compartmentalization in PMVECs.|