Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2010 Feb 17;98(4):715-23.
doi: 10.1016/j.bpj.2009.10.035.

Expanding Two-Photon Intravital Microscopy to the Infrared by Means of Optical Parametric Oscillator

Affiliations
Free PMC article

Expanding Two-Photon Intravital Microscopy to the Infrared by Means of Optical Parametric Oscillator

Josephine Herz et al. Biophys J. .
Free PMC article

Abstract

Chronic inflammation in various organs, such as the brain, implies that different subpopulations of immune cells interact with the cells of the target organ. To monitor this cellular communication both morphologically and functionally, the ability to visualize more than two colors in deep tissue is indispensable. Here, we demonstrate the pronounced power of optical parametric oscillator (OPO)-based two-photon laser scanning microscopy for dynamic intravital imaging in hardly accessible organs of the central nervous and of the immune system, with particular relevance for long-term investigations of pathological mechanisms (e.g., chronic neuroinflammation) necessitating the use of fluorescent proteins. Expanding the wavelength excitation farther to the infrared overcomes the current limitations of standard Titanium:Sapphire laser excitation, leading to 1), simultaneous imaging of fluorophores with largely different excitation and emission spectra (e.g., GFP-derivatives and RFP-derivatives); and 2), higher penetration depths in tissue (up to 80%) at higher resolution and with reduced photobleaching and phototoxicity. This tool opens up new opportunities for deep-tissue imaging and will have a tremendous impact on the choice of protein fluorophores for intravital applications in bioscience and biomedicine, as we demonstrate in this work.

Figures

Figure 1
Figure 1
Intravital three-color 4D imaging of compartmentalized fluorescent proteins in the brain stem of anesthetized mice. (A) Dual NIR/IR excitation TPLSM allows the visualization of cellular dynamics and interactions in the brain stem of an EAE-affected CerTN L15 mouse with a tdRFP-expressing immune compartment, i.e., by bone-marrow chimera. The neurons of CerTN L15 mice express a Ca2+ biosensor based on a Cerulean (ECFP-derivative)/Citrine (YFP-derivative) FRET pair. Thus, the simultaneous Ti:Sa (850 nm) and OPO (1110 nm) excitation allows monitoring of Ca2+ responses of neurons (blue and yellow) during the interaction with immune cells (red). (B) A representative Ca2+ image of a neuron; blue and yellow represent low and high Ca2+ levels , respectively, on the false color scale.
Figure 2
Figure 2
Two-photon excitation spectra of EGFP, tdRFP, and mCherry. (A) Excitation spectrum of EGFP. (B) Excitation spectra of tdRFP (solid line) and mCherry (dashed line). The spectra were measured in T cells expressing EGFP, tdRFP or mCherry. Data were corrected for the peak photon flux in photons per excitation volume and time at the sample and for the endogenous cellular fluorescence at each wavelength. Thus, the spectra represent relative wavelength-dependent, two-photon cross sections of the investigated fluorescent proteins. The absolute two-photon cross sections, measured relative to Rhodamin B in methanol (30) amounted to δEGFP = 23 ± 1 GM at λexc = 850 nm, and δtdRFP = 20.2 ± 0.8 GM at λexc = 1110 nm in 20 mM Tris-HCl/150 mM NaCl (for further details see the Supporting Material and Results); ϕEGFP = 0.6 and ϕtdRFP = 0.68 (18,31).
Figure 3
Figure 3
DdSR in agarose films and in acute brain slices. The lateral and axial resolution of the setup was measured on 200 nm red fluorescent (λexc = 1110 nm, λem = 605 nm) and yellow-green fluorescent (λexc = 850 nm or λexc = 920 nm, λem = 515 nm) polystyrene beads. (A) Typical axial and lateral profiles of the experimentally determined ePSF of a representative yellow-green (solid line) and red (dotted line) fluorescent bead at 70 μm depth in agarose (top) and brain tissue (bottom) are shown. The lateral and axial resolution is given by the full width at half-maximum of the Gaussian fit. Representative xz-projections of a red (bottom) and yellow-green (top) microsphere, respectively, at the surface and at 70 μm depth illustrate the spatial resolution for 850 nm excitation by Ti:Sa as compared to 1110 nm excitation by OPO. (B) Mean axial and lateral resolution (n = 5–10 beads) for 850 nm (solid line), 920 nm (dashed line), and 1110 nm (dotted line) excitation is plotted at various imaging depths. Error bars are mean ± SE.
Figure 4
Figure 4
DdSNR and maximal penetration depth in cocultures of acute brain slices with T cells and explanted lymph nodes. (A) Representative ddSNR plots (right) measured at the same region of a brain slice invaded by EGFP- and tdRFP-expressing T cells and corresponding perspective 3D fluorescence images (left) recorded in 2 μm z-steps. (B) Typical 3D fluorescence images of explanted lymph nodes derived from Rag 1−/− mice reconstituted with EGFP- and tdRFP-expressing T cells. Square size: (A) 30 μm × 30 μm; (B) 20 μm × 20 μm. EGFP was excited with either 3.13 × 105 mW peak power of Ti:Sa at 850 nm (solid line, green cells in 3D images) or 3.13 × 105 mW peak power of Ti:Sa at 920 nm (dashed line, blue cells in 3D images). tdRFP was excited with 3.03 × 105 mW peak power at 1110 nm of OPO (dotted line, red cells in the 3D images). For spectral separation, a 560 nm dichroic mirror and 525/50 (EGFP) as well as 593/40 (tdRFP) interference filters were used. In all graphs, error bars are mean ± SE.
Figure 5
Figure 5
DdSNR and maximal penetration depth in the brain stem and cortex of anesthetized mice (intravital imaging). ddSNR plots (A and B, right), corresponding fluorescence images of yz-projections (A, left) and perspective 3D fluorescence images (B, left) recorded under intravital conditions in the brain stem (A) and cortex (B) of Thy1-21 EGFP or Nex-cre tdRFP mice, i.e., expressing EGFP or tdRFP in neurons, respectively, are shown. Square size in B is 30 μm × 30 μm. EGFP was excited with either 3.13 × 105 mW peak power of Ti:Sa at 850 nm (green graph, green cells in yz-projection (A) and in 3D image (B)) or 3.13 × 105 mW peak power of Ti:Sa at 920 nm (blue graph, blue cells in 3D image (B)). tdRFP was excited with 3.03 × 105 mW peak power of OPO at 1110 nm (red graph, red cells in yz-projection (A) and in 3D image (B)). For spectral separation, a 560 nm dichroic mirror and 525/50 (EGFP) as well as 593/40 (tdRFP) interference filters were used. In all graphs, error bars are mean ± SE.
Figure 6
Figure 6
Photobleaching behavior in hippocampal brain slices and brain stem of living anesthetized mice. (A) Photobleaching of EGFP- and tdRFP-expressing T cells in brain slice cocultures. Representative fluorescence images summed up over 100 μm (z-step: 2 μm) are shown for simultaneous excitation by Ti:Sa and OPO at 5.21 × 105 mW and 4.96 × 105 mW peak laser power, respectively. The corresponding relative fluorescence (F(t = n min)/F(t = 0 min)) decay curves caused by photobleaching of EGFP at λexc = 850 nm (solid line) and tdRFP at λexc = 1110 nm (dotted line) are shown in B. Fluorescence images were acquired each minute for >30 min. (C) Representative images summed up over 100 μm 3D stacks display photobleaching during intravital imaging on the brain stem of anesthetized Thy1-21 EGFP and CNP tdRFP mice at 850 nm (top), 920 nm (middle), and 1110 nm (bottom) excitation. The peak laser power was 5.21 × 105 mW at 850 nm and 920 nm, and 4.96 × 105 mW at 1110 nm. The acquisition time of fluorescence images (z-step: 2 μm) was >1 h, with one 3D stack per minute. The corresponding quantification of fluorescence signal loss over time is shown in D for 850 nm (solid line), 920 nm (dashed line), and 1110 nm (dotted line). In all graphs, error bars are mean ± SE.
Figure 7
Figure 7
Spectral separation of mCherry- and tdRFP-expressing T cells. mCherry-expressing T cells, generated by lentiviral gene transfer of in vitro activated and differentiated 2d2 T cells, were mixed with tdRFP-expressing T cells. Simultaneous excitation of both T cell types at 1170 nm by OPO and spectral analysis using a 605 nm dichroic mirror and 593/40 (tdRFP) and 624/30 (mCherry) interference filters allows spectral separation of the RFPs with yellow-appearing cells as tdRFP (arrow) and red-appearing cells as mCherry-expressing T cells (arrowhead).

Similar articles

See all similar articles

Cited by 27 articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback