Imaging of living mammalian retina ex vivo by confocal laser scanning microscopy

Abstract

The present paper describes a new procedure for labelling living retinas, ex vivo, with fluorescent dyes, in order to image them by microscopy techniques by a procedure that maintains the physiological features of the tissue and its association with the retinal pigmented epithelium (RPE).

---

Introduction

The technique presented here is a new ex vivo protocol for the imaging of retinas by Confocal Laser Scanning Microscopy (CLSM), after exposure to fluorescent dyes. The incubation of retinas is performed in the living eye-semicup without fixation, to maintain the physiological and morphological features of the tissue. No other protocols are currently available to label living retinas still attached to the RPE. In fact, imaging of the photoreceptors and of the whole retina has been previously reported, however, exclusively after embedding and fixation. The immunohistochemical analysis of retina has basically been performed by inclusion or on a previously fixed tissue,^1,2 therefore with the exclusion of vital dyes. The structure of isolated mammalian fixed retinas was analysed by immunohistochemistry and CLSM.^2,3 By this protocol we have shown that in a whole bovine retina, only retinal mitochondria and rod outer segments (OSs), that are devoid of mitochondria, were stained with lipophilic fluorescent vital mitochondrial dyes MitoTracker (MT) Deep Red 633 and tetrachloro- 1,2,3,30-tetraethylbenzimidazolylcarbocyanine iodide (JC-1).⁴ These probes, essentially carbocyanine and rhodamines, interact with actively respiring membranes and show a potential-dependent partitioning across membranes.⁵ The selective association of dyes with both retinal rod OSs and mitochondria was strongly inhibited by treatment with inhibitors of electron transport or protonophores.⁴ Data suggested that a proton potential (Δψ) of the same order of magnitude of that existing in the mitochondrial inner membranes exists across the disk membranes. This is consistent with our new hypothesis on the mechanism of ATP supply for phototransduction.⁶

In this communication we explore another classical lipophilic fluorescent vital mitochondrial dye, i.e. Rhodamine-123 (RH-123). CLSM imaging experiments were performed after the exposure of retinas ex vivo to classical mitochondrial vital dyes (RH-123, MT Deep Red 633 and JC-1).

Experimental section

Solutions

Mammalian Ringer (MR), 0.157 M NaCl, 5 mM KCl, 7 mM Na₂HPO₄, 8 mM NaH₂PO₄, 0.5 mM MgCl₂, and 2 mM CaCl₂, pH 6.9, plus protease inhibitor cocktail (Sigma-Aldrich, St Louis, MO, USA) and ampicillin (50 µg ml⁻¹), and mitochondrial dyes: RH-123, MitoTracker Deep Red 633 and JC-1 were dissolved in DMSO to make 200 mM stock solutions and kept at −20 °C in dark bottles.

Procedure

It is advisable to work in sterile conditions and to add ampicillin (100 µg ml⁻¹) to the solutions. All solutions are prepared using Milli-Q® Biocel System water (Millipore, http://www.millipore.com/). Freshly enucleated bovine eyes must be used within 1.5 h of animal death. All operations are carried out at room temperature, in dim red light. Eyes are cut into half and the eyeballs divided into two eyecups. The cornea, lens and vitreous are carefully removed and discarded. The vitreous must be removed with great care, by gently squeezing the eye-semicup, in order not to detach retina while pulling it away.

The semi-cup containing choroid and retina is filled with 6–9 ml of MR (depending on the dimensions of the eye) containing 2 mM glucose, protease inhibitor cocktail and ampicillin (100 mg ml⁻¹) and incubated for 5 min. Then the desired fluorescent dye is added to the MR and incubated in the dark for 5–10 more min. Rhodamine-123, MitoTracker Deep Red 633 and JC-1 from a stock solution are added to the MR in the eye-semicup in order to reach the desired final concentration (20–50 nM for RH-123, 50–500 nM for MT and 3–5 µM for JC-1). Then samples are mounted for imaging. Eye-semicups must not be washed. Depending on whether it is preferred to image isolated retinas, incubation can be prolonged to 20 min or more so that rods spontaneously detach from the RPE and the rim of the retina floats. Retinas can be removed by rapidly turning the eye-semicup inside out, letting the MR to drip out and cutting the retina free from the optic nerve. Retinas are then put on the coverslips. If the incubation volume is too large, the retina detaches too early and floats in the medium. In some cases, for example, if the metabolism of retina is to be studied, attention should be paid to avoid an early detachment of the photoreceptors from the RPE. Otherwise, if the rods must be left embedded into the RPE, the overall incubation time must be set to no more that 10 min. In this case, the eye-semicup can be cut at the rim three times and mounted on the coverslip chamber with the face containing the retina downwards. Alternatively, the intact whole semi-cup can be immersed in a small pool arranged all over the immersion objective of the microscope. A sealed chamber filled with the MR in which to immerse the eye upside-down was created around the immersion objective by a made-on-purpose plastic partition. By this set-up it was possible to hold the eye-semicup, meanwhile the objective is free to move up and down in order to reach the proper focal planes, while no liquid drips out. In all cases retinas remain alive. No differences were noticed in either procedure. Indeed, a limiting passage of the procedure consists in the removal of the retina from the RPE without ROS loss. In fact, if the retina is pulled away too early or by a too harsh procedure ROSs are torn apart from their ellipsoid and remain immersed into the RPE. In this case rods can no more be imaged in the isolated retinas.

Imaging of the labelled retinas

CLSM imaging was performed on the above samples, at 23 °C. The measurements were acquired by means of a Leica TCS SP5-AOBS (Leica Microsystems, Mannheim, Germany) inverted confocal laser scanning microscope equipped with a 457–476–488–514–543–633 nm laser lines. Specimens are normally examined with an HCX APO L U-V-I 63×/0.9 NA (Leica Microsystems, Mannheim, Germany) water immersion objective with a working distance of 2.2 mm. This objective guarantees a good balance between the confocal sectioning and the penetration depth.⁷ Retinas can be either left in the eye-semicup or detached from the choroid. This procedure was adopted when incubation was longer than 10 min, as in this case retina detaches spontaneously from the pigmented epithelium (RPE).

Retinas stained with RH-123 were excited at 488 nm, the emission was collected in the channel covering the spectral range from 500 to 560 nm, while retinas stained with MitoTracker Deep Red 633 were excited at 633 nm, and the emission was collected in the spectral range from 650 to 700 nm. The laser power and the detection settings are kept equal in all the experiments in order to avoid artefacts. Retinas stained with JC-1, as already described,⁴ were analysed exciting at 488 nm. The emission was collected by a sequential scan in two channels, using the same detector at the same gain power and moving sequentially the spectral range, the green one 520 ± 20 nm and the red one 580 ± 20 nm (representing monomer and J-aggregate fluorescence, respectively). The derivation of the ratio of the monomer to aggregate fluorescence provides a measure directly related to the membrane potential. Images were acquired, stored and visualized with the Leica Confocal Software (LCS, Leica Microsystems, Mannheim, Germany). Images elaboration, analysis and tridimensional rendering were realized by ImageJ software (US National Institutes of Health, Bethesda, Maryland, USA).

Results and discussion

Fig. 1 is the CLSM images of the fluorescence of a portion of an isolated whole bovine retina after incubation with RH-123 within the eye-semicup. Fluorescent rod outer segments (ROSs), still attached, are imaged in perspective from their apical zone.⁸ No other neuron of the retina was visible in any other picture acquired. In both cases ROSs appear fluorescent in their distal tract (about 30 µm), i.e. a half of their length, which is 60 µm. Similar results were obtained using MT (Fig. 2). In this case, the whole eye was imaged while immersed in a small pool arranged all over the immersion objective of the CLSM. JC-1, a membrane potential carbocyanine probe, alters reversibly its colour from green to red with increasing membrane potential.⁹Fig. 3 shows the merge of the green and red JC-1 fluorescence in an isolated ROS, that has spontaneously detached from the retina. The red colour of the tip shows that J-aggregates are prevalent there. Some mitochondria, less fluorescent belonging to other retinal cells, are visible on a deeper focus plane.

Fig. 1 Confocal Fluorescence (CLSM) image of Rhodamine-123 fluorescence in a living bovine retina. CLSM image of Rhodamine-123 (RH-123) fluorescence, excited at 488 nm, in a living bovine retina. The image is a maximum projection (z = 84 µm) of serial confocal sections on a part of a whole bovine retina exposed to RH-123 ex vivo. The rod OSs are clearly visible.

Fig. 2 Confocal Fluorescence (CLSM) image of MitoTracker Deep Red 633 fluorescence in a living bovine retina. CLSM image of MitoTracker Deep Red 633 fluorescence, excited at 633 nm, and collected in the spectral range from 650 to 700 nm, in a living bovine retina. The image is a maximum projection (z = 40 µm) of serial confocal sections on an undulate part of a whole bovine retina exposed to MT ex vivo.

Fig. 3 Confocal Fluorescence (CLSM) image of JC-1 fluorescence in rods detached from a living bovine retina. Maximum projection (z = 54 µm) of serial confocal sections of some isolated rods detached from an undulate part of the retina. Green (520 ± 20 nm) and red (580 ± 20 nm) channels were acquired sequentially and merged. The image is a representative of at least 30 from at least three different experiments for each probe.

This protocol provides us with a new way to visualize rod OSs in a mammalian retina. In fact, considering the similarities among mammalian retinas,¹⁰ the protocol may be also applied to other mammalian, and even human retinas or part of them. The procedure may also be used in a wider range of applications than presented, in fact other vital dyes, fluorescent probes, or metabolic probes may be utilised, to investigate the aspects of retinal metabolism. The procedure may also be useful to study the function and morphology of retinal mitochondria. In fact, at concentrations which do not inhibit mitochondrial function, for example, RH-123 is a sensitive and specific probe of Δψ in isolated mitochondria. The protocol can also be used if fixation of the retina is necessary before imaging. A fixation step can be conducted inside the eye-semicup by substituting the incubation medium with a proper fixative, after the probe incubation.

Acknowledgements

We thank Pietro Calissano (CNR and University of Roma, Italy) and I. Mario Pepe (University of Genova, Italy) for their invaluable contribution. The work was supported by grants from Italian Ministry for research and Technology (MIUR). For A.D. and P.B. the work was granted by IFOM-2007-2011 and IIT.

References

1. L. Vitanova, S. Haverkamp and H. Wässle, Cell Tissue Res., 2004, 137, 227–235.
2. S. Haverkamp and H. Wässle, J. Comp. Neurol., 2000, 42, 1–23.
3. N. Cueca, I. Pinilla, Y. Sauvé and R. Lund, Eur. J. Neurosci., 2005, 22, 1057–1072.
4. P. Bianchini, D. Calzia, S. Ravera, G. Candiano, A. Bachi, A. Morelli, M. Bruschi, I. M. Pepe, A. Diaspro and I. Panfoli, J. Biomed. Opt., 2008, 13, 054017.
5. O. Uckermann, I. Iandiev, M. Francke, K. Franze, J. Grosche, S. Wolf, L. Kohen, P. Wiedemann, A. Reichenbach and A. Bringmann, Glia, 2004, 45, 59–66.
6. I. Panfoli, D. Calzia, P. Bianchini, S. Ravera, A. Diaspro, G. Candiano, A. Bachi, M. Monticone, M. G. Aluigi, S. Barabino, G. Calabria, M. Rolando, C. Tacchetti, A. Morelli and I. M. Pepe, Int. J. Biochem. Cell Biol., 2009, 41, 2555–2565.
7. A. Diaspro, Confocal and Two-Photon Microscopy, Wiley-Liss, Inc., New York, 2002.
8. S. Nickell, P. S. Park, W. Baumeister and K. Palczewski, J. Cell Biol., 2007, 177, 917–925.
9. M. Reers, T. W. Smith and L. B. Chen, Biochemistry, 1991, 30, 4480–4486.
10. H. Issle and B. B. Boycott, Physiol. Rev., 1991, 71, 447–480.

---