Ex-vivo imaging of excised tissue using vital dyes and confocal microscopy

Abstract

Vital dyes routinely used for staining cultured cells can also be used to stain and image live tissue slices ex-vivo. Staining tissue with vital dyes allows researchers to collect structural and functional data simultaneously and can be used for qualitative or quantitative fluorescent image collection. The protocols presented here are useful for structural and functional analysis of viable properties of cells in intact tissue slices, allowing for the collection of data in a structurally relevant environment. With these protocols, vital dyes can be applied as a research tool to disease processes and properties of tissue not amenable to cell culture based studies.

INTRODUCTION

Staining and imaging cells with vital fluorescent dyes is a highly versatile approach widely used among cell biologists. The diversity of commercially available vital dyes allows for microscopic analysis of multiple cellular and subcellular parameters individually or combined. Parameters amenable to study using fluorescent probes include structure, organization, and localization of organelles and subcellular compartments, as well as functional measurements, such as reactive oxygen species (ROS) production, mitochondrial membrane potential, and pH. While vital dyes are popular tools among cell biologists, cell culture studies designed to model human disease are intrinsically limited; they lack the complex structure, cell type heterogeneity, and physical history of diseased tissues. Isolated reports of ex-vivo tissue staining appear in the literature, but these studies utilize few dyes and typically rely on two-photon microscopy for imaging.

In this unit methods are presented for staining and imaging live tissue slices ex-vivo for qualitative and quantitative data collection. The most basic procedure is suitable for collecting morphology and localization data from tissue using dyes that stain mitochondria and DNA simultaneously with detection of green fluorescent protein (GFP) tagged proteins (see Basic Protocol). Alternate dye sets may be used as a quality control measure for tissue viability and to examine cell death in tissue samples ex-vivo (see Alternate Protocols 1 and 2). Following these specific protocols is a more generalized method useful for a wide variety of dyes, which can be selected to fit the desired purpose (see Alternate Protocol 3). This protocol has been applied successfully to a range of tissue types, including human biopsy tissue. Two additional protocols present alternate methods for collecting quantitative data; one requires that images are collected at a constant depth from the surface of the tissue (see Alternate Protocol 4), while the other allows for depth independent acquisition, but requires empirical determination of the rates of signal decay as a function of depth within the tissue imaged (see Alternate Protocol 5). Data collected in this manner can be used to quantify changes in functional properties within three-dimensional structures within tissue. Physiological states such as cellular and mitochondrial ROS production can be examined using a slight variation on the basic technique (see Alternate Protocol 6). The staining steps in the presented protocols require little prior experience, while the image collection requires moderate experience with confocal microscopy, for which they are designed.

BASIC PROTOCOL 1 STAINING AND IMAGING MITOCHONDRIA, NUCLEI, AND GFP IN LIVE TISSUE SLICES EX-VIVO

A variety of transgenic mouse lines exist where protein targets of interest are tagged with fluorescent protein conjugates or fluorescent proteins are expressed under the control of regulatory units of interest. A range of spectral variants have arisen over the years since the original cloning of GFP to allow for greater versatility and the use of multiple transgenic markers at once. The use of transgenically expressed fluorescent proteins allows for temporal and spatial investigation of protein expression or promoter activation using standard techniques such as western blotting of homogenized tissue and immunostaining of sectioned fixed samples or direct imaging of aldehyde fixed samples. Samples from fluorescent protein expressing animals have also been subjected to imaging with two-photon microscopy for analysis of localization within intact tissue sections. The following protocol provides a method for staining live GFP expressing tissue slices with dyes specific for mitochondria and DNA. This method provides a simple and highly accessible method for analyzing the localization and level of GFP in living tissue in relation to nuclei and mitochondria in living tissue slices. This technique is particularly applicable to investigation of dynamic subcellular processes such as mitophagy, where autophagosomes, which are commonly labeled with transgenic GFP tagged LC3(1), are assembled around mitochondria to target them for destruction via the lysosome(2,3). In this method a slice of freshly excised tissue is viably stained and imaged by confocal microscopy.

Materials

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• 1mg/mL Hoechst 33342 in 1XPBS (see Reagents and Solutions)

• 200μM MitoTracker Deep Red in DMSO (see Reagents and Solutions)

• Ice for storing buffers, sectioning tissues, and incubating tissues in dyes on orbital shaker

• Weigh boat or small dish on ice

• 12 well tissue culture plates or equivalent size polystyrene plates for incubation on ice with shaking

• Orbital shaker capable of rotating at up to 100–150 rpm with ice bucket securely on top (or an orbital shaker in a walk-in cold room)

• Double edged razor blade (we prefer Fisher catalog number NC9873485)

• 2-well chambered coverglass slides (Thermo catalog number 155379)

• Aluminum foil folded to rigid square of ~2cm × 2cm × 8 folds thickness

• 1.8mL glass sample vial (Thermo catalog number 03-339-25A) or an object of similar size and a mass of 3–5g, suitable for gently compressing tissue.

• Confocal microscope with desired objectives and with lasers and filters for DAPI, FITC, and far red excitation and emission (see Table 1 for suggested laser and filter sets)

• Freshly excised tissue from an animal expressing GFP (see protocol)

• Recommended for mouse tissue: Mouse heart or brain slicing block (Zivic catalog numbers HSMS001-1, BSMS001.1, or BSMS001.2)

TABLE 1.

Properties of Selected Vital Dyes

Dye	Dye Excitation/Emission (nm)	Excitation Laser	Working Concentration	Suggested Emission Filter	Quality of Staining with Ex-Vivo Tissue Imaging*
10-n-Nonyl Acridine Orange (NAO)	500/525	488nm	400nM	505–530BP or 505–570BP	++
Calcein AM	490/525	488nm	5μM	505–530BP or 505–570BP	++
Cell Rox Deep Red	644/665	647nm	5μM	650LP	+++
Cell Tracker Blue	353/466	405nm	5μM	420–480BP	+++
Draq5	650/680	647nm	5μM	650LP	+++
Ethidium Homodimer-1 (EthD-1)	528/617	514nm	1μM	560–615BP	++
GFP (as LC3-GFP)	470/514	488nm	n/a	505–530BP or 505–570BP	+++
H2CMXRos	543/580	543nm	200nM	560–615BP	++
H2DCFDA	495/525	488nm	5μM	505–530BP or 505–570BP	+
Hoechst 33342	350/460	405nm	5μg/mL	420–480BP	+++
JC-1	Monomer - 485/530	488nm	10μg/mL	505–530BP	+
	Aggregate - 535/590	543nm		560–615BP
LysoSensor Blue DND-167	373/425	405nm	500nM	420–480BP	+
MCB	394/490	405nm	60μM	420–480BP	++
MitoPY1	543/560	543nm	5μM	560–615BP	+++
MitoSox	Non-Specific Product 510/590	514nm	5μM	560–615BP	++
	Superoxide Specific Product 396/590	405nm	5μM	560–615BP	++
MitoTracker Deep Red	640/662	647nm	200nM	650LP	+++
MitoTracker Green	490/516	488nm	200nM	Mitochondrial mass	+
Sytox Green	504/523	488nm	100nM	505–530BP	+++
TMRE	545/590	543nm	200nM	560–615BP	+++
Vybrant Ruby	638/686	647nm	5μM	650LP	+++

(B) *-Quality of staining refers to the general opinion of the authors of this paper of the utility of these dyes using this method. Specificity, brightness, ability to permeate tissue, and photostability are all included in this determination. The highest quality dyes are designated with +++, whereas dyes which were unfavorable are designated with +.

(C)

Protocol steps

1. Collect tissue from the source (animal expressing GFP) rapidly into ice-cold excision buffer (see Reagents and Solutions). The samples should be stored on ice and protected from light at all times.

   This protocol has been tested and works well with cardiac tissue, skeletal muscle, kidney, liver, brain, optic nerve, retina, and colon; its utility with other tissue types will have to be determined.

2. Prepare a tissue slice of approximately 2mm thickness. Direction of cutting should be based on imaging needs.

   Murine heart, for example, would typically be sliced perpendicular to the ventricular walls to form rings of tissue while liver can be cut into arbitrary slices given the liver’s uniform architecture. Small tissues such, as individual isolated nerves (i.e., optic nerve), can be stained without sectioning.

   Care in tissue handling is crucial to the success of this protocol. In addition to speed in collection and storage in cold buffer, sectioning the tissue must be performed with the sharpest instruments possible to allow for clean cutting and prevent damage associated with crushing or tearing of tissue. We recommend the use of cutting blocks, detailed in the materials section of this protocol, for high quality sectioning. Alternatively tissue can be embedded in 4–5 percent agarose to make a custom shaped blocks, an inexpensive alternative that allows for easy sectioning. Cutting blocks should be immersed in ice-cold buffer during sectioning.

3. Determine the volume of staining solution that will be needed. For a 12 well plate 2–3 tissue slices can be stained per well and each well will require 3 ml of solution.

4. Prepare the staining solution by diluting the stock dyes into the ice-cold HBSS mixture. Dilute 5μL of the 1mg/ml Hoechst stock solution and 1μL of the 200μM stock MitoTracker Deep Red into the staining buffer per milliliter of final volume. Aliquot accordingly into a 12 well plate on ice.

5. Use forceps to gently move the tissue slices into the staining solution, 1–3 slices per well.

6. Stain the tissues on an orbital shaker on ice or in a 4oC cold room, protected from light. Shake for 20 min at 150 rotations per

   minute, or at the lowest speed that keeps the tissue rotating around the sides of the well. Staining can be performed at ambient or physiological temperature (see Alternate Protocol 6), but the dyes listed here stain well on ice and maintaining cold temperatures results in greater tissue viability at the time of imaging.

7. While tissue is staining prepare for imaging. Place 2-well chambered coverglass slides filled roughly halfway with HBSS buffer on ice (~2 mL). Fold a piece of aluminum foil 8 times and cut to a size that will fit in one of the chambers, approximately 2cm × 2cm. Place this and a few glass sample vials on ice to cool.

8. Use forceps to gently move the tissue into fresh ice-cold buffer in a new 12 well plate. Remove excess dye by shaking as in step 5 for an additional 10 min. Tissue slices may be stored on in the 12 well plate in buffer on ice during imaging.

9. Image the tissues using standard confocal microscopy. To image, gently move a single slice of tissue into one side of a chambered coverglass slide, on ice. Select an appropriate microscope objective and apply immersion oil if needed. Place the foil over the tissue and the glass sample vial on top of the foil to gently press the tissue against the slide.

   40X oil objectives are typically recommended for this method. Higher power objectives allow for better subcellular imaging, but if this can be achieved using digital magnification it will allow for imaging of larger fields to facilitate histological comparisons. Non-immersion objectives can also be used for imaging larger structures, but condensation may need to be removed from the coverglass using a KimWipe.

10. Collect imaging using standard excitation wavelengths and emission filters for GFP/FITC, Hoechst/DAPI, and far red excitation and emission (see Table 1 for suggested laser and filter sets).

11. Images should be collected using a narrow optical slice thickness, such as 1–2μm, to allow for distinction of small subcellular objects.

12. Images must be collected deep to the cut surface, as damaged cells on the surface do not provide informative data. Example images are provided in Figure 1. The depth required is typically only ~2–5μm (see Figure 2). Penetration of excitation and collection of emission at depths up to 20–30μM is only slightly attenuated and does not require 2-photon microscopy (see Figure 2 and Alternate Protocol 4). Gain and compensation can be adjusted as needed to collect useful images.

    Note: This protocol describes the collection of data without quantification of intensity, and is therefore useful to discern spatial organization and relative fluorophore signals. Methods for collecting quantitative data are provided in Alternate Protocols 4 and 5.

13. Place tissue back on ice after every couple of images to prevent warming and loss of viability.

    Sections should be kept covered on ice in chamber slides while waiting for imaging. A cooled microscope sample chamber, if available, will allow for imaging without the need to monitor sample temperature.

Figure 1.

Sample Images Collected Using Vital Dyes and Ex-Vivo Imaging Sample images of tissues stained with vital dyes using the protocols in this unit. (A), (B), and (F) show, in order, mouse slow twitch and fast twitch skeletal muscle fibers and mouse optic nerve stained with TMRE and Hoechst 33342. Panel (C) is a sample image of human colon biopsy sample stained with TMRE and Cell Tracker Blue. (D) and (E) are mouse cerebellum and liver expressing LC3-GFP stained with TMRE and Hoechst 33343 (as in Basic Protocol 1). Tissue, cellular, and subcellular structures are easily distinguished using these methods.

Figure 2.

Tissue Viability Analysis with Depth Using Ethidium Homodimer-1 and Calcein AM or Sytox Green and Tetramethylrhodamine Ethyl Ester Tissue viability assessed using the staining methods in Alternate Protocols 1 and 2. (A) Liver tissue stained with Calcein AM and EthD-1. At the cut surface a small percentage of cells show lost membrane integrity by EthD-1 positivity (A1), but the frequency of EthD-1 positive cells quickly drops to nearly zero with depth. (B) Cardiac tissue stained with Calcein AM and EthD-1. Tissue cut in parallel with the cardiac fibers has few EthD-1 positive cells (B1-B2), demonstrating almost complete viability even just interior of the cut surface. Tissue cut transverse to the fibers has a slightly increased rate of membrane integrity loss as seen by a higher rate of EthD-1 positive staining (B1, B3). (C) Renal glomeruli stained with Sytox Green and TMRE. The rate of Sytox positive cells is highest near the surface and steadily drops with depth. By image analysis it is appears that the bulk of the Sytox positive cells do not contribute to the structures of interest, the glomeruli and renal tubules, which appear completely viable by Sytox staining. TMRE positive staining shows that the mitochondria in these cells are functional, further indicating the cells are viable.

ALTERNATE PROTOCOL 1 ASSESSING CELL VIABILITY IN LIVE TISSUE SLICES EX-VIVO USING ETHIDIUM HOMODIMER-1 AND CALCEIN AM

Damaged cells can be easily distinguished from intact cells with the use of cell-impermeable DNA binding dyes, such as ethidium homodimer-1 (EthD-1). Alternatively, dyes that are activated and retained only in intact cells provide a positive marker for cell viability. Calcein AM is a cell-permeable molecule that is converted by intracellular esterases into a calcium binding, highly fluorescent molecule. The use of these dyes together in staining of live tissue slices ex-vivo provides both positive and negative indicators of cell viability, allowing for the collection of data related to cell mortality in disease, as well as a powerful quality control method for tissue viability.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• 1mg/mL Hoechst 33342 in 1X PBS (see Reagents and Solutions)

• 2mM EthD-1 in DMSO (Sigma catalog number E1903)

• 4mM Calcein AM in DMSO (Sigma catalog number C1359)

Protocol steps

1. Collect tissue as in Basic Protocol 1.

   Note: This protocol can be used with GFP expressing tissue, as in Basic Protocol 1, but if the GFP is diffusely located within the expressing cells then Calcein AM may not provide useful data. If this is the case simply omit this dye and determine viability based on EthD-1 staining.

2. Calculate the staining solution volume needed as in step 3 of Basic Protocol 1.

3. Prepare the staining solution by diluting the stock dyes into the ice-cold HBSS mixture. Dilute 5μL of the 1mg/ml Hoechst stock solution, 0.5μL of the 2mM stock EthD-1, and 1.25μL of the 4mM stock Calcein AM into the staining buffer per milliliter of final volume. Aliquot accordingly into a 12 well plate on ice.

4. Stain and image as in steps 4–9 of Basic Protocol 1. See Table 1 for excitation and emission wavelengths.

5. To collect depth dependent cell viability data (for quality control of tissue handling and processing) collect images as z-section stacks with 1–5μm z-direction distance between optical slices.

6. Quantify dead cells as percent of Hoechst stained nuclei that are also positive for EthD-1. Cells can also be considered nonviable if they lack cytoplasmic Calcein AM staining. Nonviable cells are easily identified in a variety of tissues (see Figure 2).

ALTERNATE PROTOCOL 2 ASSESSING CELL VIABILITY IN LIVE TISSUE SLICES EX-VIVO USING SYTOX GREEN AND TETRAMETHYLRHODAMINE ETHYL ESTER

Alternate Protocol 1 provides a simple method for determining whether cell membranes have been compromised. This method is only one possible set of dyes that provides viability data, and may not be the best method for every scenario. Another useful dye set for examination of viability is a combination of the DNA dye Hoechst 33342, Sytox Green, and tetramethylrhodamine ester (TMRE). Hoechst 33342 is cell permeable and stains all nuclei regardless of membrane status. Sytox Green is similar to EthD-1 in that it cannot cross intact membranes and is a specific stain for the nuclei of cells with compromised membranes. TMRE is a membrane potential dependent mitochondrial dye. Mitochondrial membrane potential is often lost prior to cell death via apoptosis (REF), and when cells die via necrosis. Examining membrane potential provides an additional measure of viability not directly related to the cell membrane. Furthermore, staining with TMRE allows for examination of changes to mitochondrial function (as membrane potential) and mitochondrial structure in parallel with cellular viability. While the Sytox stain is extremely bright and highly specific, this method is not useful when tissue contains nuclear localized GFP.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• 1mg/mL Hoechst 33342 in 1XPBS (see Reagents and Solutions)

• 5mM Sytox Green in DMSO (Invitrogen catalog number S7020)

• 200μM TMRE in DMSO (see Reagents and Solutions)

Protocol steps

1. Collect tissue as in Basic Protocol 1.

   This protocol can be used with GFP expressing tissue, as in Basic Protocol 1, but if the GFP is nuclear localized than the Sytox stain will not provide informative data. In this case Alternate Protocol 1 is a better option for assessing cellular viability.

2. Calculate staining solution volume needed as in step 3 of Basic Protocol 1.

3. Prepare the staining solution by diluting the stock dyes into the ice-cold HBSS mixture. Dilute 5μL of the 1mg/ml Hoechst stock solution, 1μL of the 5mM stock Sytox Green, and 1μL of the 200μM stock TMRE into the staining buffer per milliliter of final volume. Aliquot accordingly into a 12 well plate on ice.

4. Stain and image as in steps 4–9 of Basic Protocol 1. See Table 1 for excitation and emission wavelengths.

5. To collect depth dependent cell viability data (for quality control of tissue handling and processing) collect images as z-section stacks with 1–5μm z-direction distance between the optical slices.

6. Quantify dead cells as percent of Hoechst stained nuclei that are also positive for Sytox Green. Loss of or decrease in TMRE indicates that the mitochondria within a cell have lost membrane potential, which can occur before or after cell membrane integrity loss, or can be indicative of a non-viability related change. Nonviable cells are easily identified in a variety of tissues (see Figure 2).

ALTERNATE PROTOCOL 3 GENERALIZED METHOD FOR TISSUE SLICE STAINING EX-VIVO

The method provided in Basic Protocol 1 is a specific staining set designed for imaging the subcellular localization of GFP and the general tissue structure and morphology of tissue expressing GFP. This approach can also be applied to non-fluorescent protein expressing tissues as well as tissues expressing any spectral variant of GFP. Using the more generalized approach described below, dye sets can be selected based on the parameters of interest, the spectral compatibility of the dyes, and the available laser and filter sets on the confocal microscope used.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• A mix of dyes chosen based on parameters of interest, spectral compatibility, available lasers and filter sets, etc. See Protocol Steps for details on selection of dye sets and Table 1 for the properties of dyes tested with this technique.

Protocol steps

1. Choose a dye set appropriate for collecting desired data. (See Critical Parameters/Troubleshooting)

   Note: Structurally similar or physically proximal structures should be stained with highly spectrally distinct dyes when possible for optimal differentiation (avoid staining ER red and mitochondria far-red, for example). This is unimportant if each of the dyes is bright and the available filters allow for short band-pass emission collection but becomes an issue when dyes are weak or when long pass filters are used.

2. Prepare stock solutions of all dyes as directed by the manufacturer of each dye. See Table 1 for suggested stock solution concentrations.

3. Collect tissue as in Basic Protocol 1.

4. Calculate staining solution volume needed as in step 3 of Basic Protocol 1. Plan for a volume sufficient to stain a set of slices with each of the dyes separately in addition to the desired combined staining set.

5. Prepare the staining solution by diluting the stock dyes into the ice-cold HBSS mixture. Prepare solutions with each of the dyes singly as well as the staining solution containing all dyes.

6. Stain and image as in steps 4–9 and 11–13 of Basic Protocol 1. See Table 1 for excitation and emission wavelengths.

   The dyes listed in Table 1 have been tested and work using this technique. Dyes not listed must be tested individually and their utility determined experimentally.

7. After setting all channel settings based on the sample stained with all dyes image the single dye stained slices. Collect data from all spectral channels. Signal crossover between channels is calculated from the apparent signal in empty channels in these images. Significant crossover would indicate that the overlapping dyes are not suitable for combined use. In this protocol “significant” would be any crossover sufficient to obstruct the ability to distinguish differentially stained structures. Crossover must be considered when collecting quantitative fluorescent intensity data, as discussed in Alternate Protocols 4 and 5, or if automated co-localization analysis is to be performed.

ALTERNATE PROTOCOL 4 DEPTH DEPENDENT QUANTITATIVE DATA COLLECTION

The methods in Basic Protocol 1 and Alternate Protocols 1–3 are all designed for the collection of data related to structure and localization but are not sufficient for collecting quantitative fluorescent intensity data. In order to collect quantitative data using confocal microscopy of a thick sample multiple variables must be accounted for. These include penetration of the dyes, signal attenuation of each dye as a property of depth, and signal crossover from one channel into the others. There are two approaches to collecting quantitative data: collecting data at constant depth in order to eliminate depth related variables, or collecting data at multiple, but defined, depths in order to empirically determine the relative rates of signal attenuation among the dyes and allow for normalized depth-independent collection of data. The first approach is utilized in this protocol.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• A mix of dyes chosen based on parameters of interest, spectral compatibility, available lasers and filter sets, etc. See Alternate Protocol 3 for details on selection of dye sets and Table 1 for the properties of dyes tested with this technique.

Protocol steps

1. Collect tissue as in Basic Protocol 1.

2. Choose dyes as Alternate Protocol 3. At least one dye should represent a repeated structure that can be used to normalize data between images. A common example would be a nuclear dye for DNA quantitation.

3. Calculate staining solution volume needed as in step 3 of Basic Protocol 1. Prepare a volume sufficient to stain a set of slices with each of the dyes separately in addition to the desired staining set.

   If a fluorescent protein is present in the tissue it should be considered a dye for the purposes of this protocol – an unstained sample slice should be taken through the staining procedure with the fluorescent protein as the only fluorophore present.

4. Prepare the staining solution by diluting the stock dyes into the ice-cold HBSS mixture. Prepare solutions with each of the dyes individually as well as the staining solution containing all dyes.

5. Stain and prepare for imaging as in steps 4–9 of Basic Protocol 1. See Table 1 for excitation and emission wavelengths, or refer to manufacture specifications for dyes not listed.

6. Select the slice stained with an individual dye that provides the best structural overview of the tissue and should not vary greatly between samples. DNA dyes usually provide a good reference. Use the z-scan function of the microscope to determine an optimal depth for image collection. This should be deep to cellular disruption due to processing but shallow enough to minimize signal attenuation; a typical depth might be between 5 and 20μm (see Figure 2). Set the coverslip/tissue interface to z = 0. This step allows image stacks to be collected at equivalent depths across the tissue.

   Steps 5–8 must be performed separately for each tissue type imaged, and must be repeated in each experiment.

7. After selecting an optimal depth for the imaging set for each singly stained tissue set the gain, offset, laser intensity, etc, using the same depth from surface for each channel.

   Note: In order to minimize the potential for signal crossover each fluorophore should be collected as a separate channel, unless the fluorophores can both be excited with a single dedicated laser but have unique emission properties. This reduces the amount of background fluorescence (signal present in a channel when the dye for that channel is not present) and spectral crossover (crossover caused by overlapping emission spectra when dyes are dually excited at separate wavelentghs, as compared to crossover caused by off target excitation due to overlapping excitation spectra).

8. Collect images from each singly stained tissue using the entire set of channels. The signal present in the channels that should be blank represents a combination of crossover, off target excitation, and autofluorescence.

   Note: Bright dyes and narrow band-pass emission filters typically limit background and crossover fluorescence to a tiny fraction of the real signal, but these data must be collected as quality control information and to provide an equation for quantifying changes between treatments.

9. Collect images from the tissues stained with the entire dye set. Each tissue type must be imaged using the parameters determined for that tissue. Images that are to be compared should be of similar regions of interest and tissue orientation.

10. Data collected this way is unaffected by depth-dependent factors (all images were collected at the same depth) and can be quantified without taking the tissue depth into account. Images should be analyzed using software such as NIH ImageJ (http://rsbweb.nih.gov/nih-image/).

11. If the crossover is demonstrated to be a tiny fraction of the total, and much less than observed changes between samples, the amount of crossover may simply be reported and this calculation avoided. This should be determined independently for each channel. If crossover occurs the measured signal of each affected channel must be adjusted to account for the crossover, as follows:

    Signal from a given channel a is quantified as follows:

    Actual signal of fluorophore a = signal collected – Σ(crossover from excitation of the other fluorophores)

    Where crossover from fluorophore x = signal in channel a resulting from excitation of fluorophore x (as a function of the signal in channel x) when fluorophore x is excited and no other dye is present. For the purpose of quantifying integrated signal intensity compensation need not be performed on a pixel by pixel basis. Pixel by pixel compensation is necessary for quantification between regions within an image if crossover is present. This type of calculation is out of the scope of this protocol and is thoroughly discussed in (4).

12. At this point the data collected is quantitative, but consideration must be given for proper normalization. Given that tissue slices can be highly heterogeneous, depending on the tissue, it is important that care is given to finding a normalization parameter that is not affected by any treatments being compared. Number of cells, number of nuclei, and nuclear fluorescence signal are examples of potential parameters to normalize the signal of interest against.

ALTERNATE PROTOCOL 5 DEPTH INDEPENDENT QUANTITATIVE DATA COLLECTION

The methods presented thus far are sufficient for collecting localization and structural information and for collecting quantitative data at a constant depth of imaging. In some circumstances it may be desirable to collect quantitative data across a set of z-plane images. In order to quantify data with depth of imaging as a variable the rate of signal attenuation with depth must be empirically determined for each of the dyes used. This information can be collected simultaneously with the experimental data and accounted for during image analysis.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• A mix of dyes chosen based on parameters of interest, spectral compatibility, available lasers and filter sets, etc. See Alternate Protocol 3 for details on selection of dye sets and Table 1 for the properties of dyes tested with this technique.

Protocol steps

1. Follow staining and imaging procedures as in Alternate Protocol 4 from step 1–8.

2. Collect images as z-stacks with the first image of each stack (z=0) set to the cut surface and each of the subsequent images a set distance inward (deep) in the z-direction. The first interior slice is slice 1. This is illustrated in Figure 3A.

3. Collect all images using the same z-stack settings.

4. In order to calculate the rates of signal decay with depth compile all of the images from the control treated sample. Calculate the signal of each fluorophore at each depth in each z-series and normalize the signal of each to slice 1. As stated in (2), do not including the cut surface in the analysis (the cut surface image should be kept as a marker for depth, but the data from the damaged surface of the tissue is not useful).

5. Determine, based on the plotted data, a function that describes the rate of signal decay with depth. This is approximately linear within the useful range of depth for most fluorophores tested (see Figure 3B1 and C).

   For each of the dyes listed in Table 1 the rate of signal decay in tissues tested is approximately linear within the useful imaging depth using confocal microscopy. If a two-photon microscope is used to image stained tissues at greater depths signal attenuation is likely to be non-linear.

6. Normalize the interior slices of each image stack to the first image using this calculated rate. After doing this normalization, images obtained at different depths may be quantitatively compared.

   Note: Some confocal microscope software programs allow users to set up collection parameters that include automatic signal attenuation correction. This method precludes the use of automated signal attenuation; settings such as these are useful for collecting nice images and comparing structures but should not be utilized when collecting data for quantification.

Figure 3.

Quantitative Data Collection and Analysis

(A) Schematic of the collection of a z-stack of images. A z-scan of the region of interest allows the first optical slice to be set at the cut surface (the tissue-coverglass interface) and the distance between optical slices to be specified. The first image in each stack, the cut surface image, is referred to as the z=0 slice. This optical slice is collected to allow for orientation of the other slices, but is not included in any analyses. (B1) Multiple z-stacks of images in a single sample are analyzed by depth. Average values for a given fluorophore across multiple depths are plotted and the approximate rate of signal loss is determined, as shown by (a) and (b). The calculated rate of signal loss is used to normalize the data by depth so that the data from all images of a given fluorophore in an optical stack, and across multiple stacks, can be directly averaged. The result of signal loss compensation is shown in (B2). For the sake of presentation each individual fluorophore is normalized to the relative z=1 signal. Each dye or fluorophore must be empirically examined. (B1) and (B3) show that fluorophore signal loss is dye dependent.

ALTERNATE PROTOCOL 6 STAINING OF TISSUES USING DYES DEPENDENT ON ACTIVE PHYSIOLOGICAL FUNCTIONS

The preceding protocols are designed for dyes that stain their respective targets without the need for active respiratory activity. While many stains that depend on functional cellular events may work while staining on ice it may be necessary for some targets to perform the staining at ambient or physiological temperature. Furthermore, the reactions necessary for the conversion of some probes into fluorescent molecules may not occur at a useful rate at 4°C. ROS sensitive dyes are one such set of probes that are poorly activated at low temperature and in the absence of active cellular metabolism. Staining at ambient or physiological temperatures requires that the staining buffer more closely match the natural environment of the tissue and provide appropriate nutrients and oxygen to support respiratory activity for the duration of the staining period. Raising the temperature also causes the tissue to become more fragile, especially weak and fatty tissues such as brain, requiring greater care during staining and handling.

Materials

• The equipment and labware listed in Basic Protocol 1

• Excision and incubation buffer (1X HBSS with additives, see Reagents and Solutions)

• A staining buffer of choice. The buffer should be selected based on the tissue type. For example: if brain slices or cranial nerves are stained, artificial cerebrospinal fluid would be the best choice. An easy, generic tissue staining buffer that has been tested and used with every tissue type listed in this paper is provided in Reagents and Solutions (Generic Staining Buffer).

• A mix of dyes chosen based on parameters of interest, spectral compatibility, available lasers and filter sets, etc. See Alternate Protocol 3 for details on selection of dye sets and Table 1 for the properties of dyes tested with this technique.

Protocol steps

1. Prepare a staining buffer, as noted above. Prepare a volume suitable for the desired staining and imaging.

2. Prior to excision warm Generic Staining Buffer (see Reagents and Solutions) or a suitable buffer of choice to desired temperature (ambient or physiological).

3. Excise and section tissue into excision and incubation buffer as in Basic Protocol 1.

4. Oxygenate the staining buffer by bubbling air (ambient) into the solution for 5–10 min prior to staining tissues.

5. Dilute dyes into staining buffer as in the preceding protocols or as directed by the manufacturer. Prepare them fresh just before moving the tissue slices for staining. Aliquot into the wells of a12 well plate pre-warmed to the desired temperature.

6. Gently move the tissue slices into their designated well.

7. Incubate for 20min with gentle orbital rotation.

8. Move to fresh incubation buffer on ice.

9. Incubate 10min on ice with gentle orbital rotation.

10. Store samples, image, and analyze data as outlined in the preceding protocols.

REAGENTS AND SOLUTIONS

Excision and incubation buffer

Requires 10X HBSS (purchased from Invitrogen, catalog number 14065056). Dilute to 1X with sterile water. Supplement with 8.9mM sodium bicarbonate (100X purchased from Sigma, catalog number S8761) and 1% (w/v) Bovine Serum Albumin (BSA – Sigma catalog number A2153) or 5% (v/v) Fetal Bovine Serum (FBS – Invitrogen catalog number 10082-147). pH to 7.4. If this is prepared in advance cell culture grade antiobiotics, such as a penicillin/streptomycin (Invitrogen catalog number 15070-063), can be added, as directed by the manufacturer, to prevent bacterial growth. Store in 4°C for up to 6 months.

1X Phosphate Buffered Saline (PBS)

Dilute the following into 500mL sterile deionized water: 8g of NaCl, 0.2g of KCl, 1.44g of Na₂HPO₄, and 0.24g of KH₂PO₄. pH to 7.4, add sterile deionized water to 1L. Can also be made and prepared as 10X by dissolving in a total of 100mL sterile water.

1mg/mL Hoechst 33342

Dilute Hoechst 33322 (Sigma catalog number B2261) to 1mg/mL in 1X PBS. Store protected from light at 4°C, up to 6 months, or long term in aliquots at −20°C.

200μM MitoTracker Deep Red

Dissolve the 50μg contents of one MitoTracker Deep Red FM Special Packaging vial (Invitrogen catalog number M22426) with 459.9μL of DMSO. Store at −20 or −80°C protected from light.

200μM TMRE

Dissolve the 25mg contents of one bottle of tetramethylrhodamine ethyl ester perchlorate (Sigma catalog number 87917) with 4.8mL of DMSO to make a 10mM (50,000X) master stock solution. Store this long term in aliquots at −20°C protected from light.

Further dilute this to 200μM (1,000X) by diluting 1:50 in DMSO. Also store at −20° protected from light.

Generic Staining Buffer

Requires 500mL low glucose DMEM with L-glutamine and pyruvate (Invitrogen catalog number 11885-084). Supplement 500mL bottle with 8.9mM sodium bicarbonate (100X purchased from Sigma, catalog number S8761) and 1% (w/v) Bovine Serum Albumin (BSA – Sigma catalog number A2153) or 5% (v/v) Fetal Bovine Serum (FBS – Invitrogen catalog number 10082-147).

Add cell culture grade antibiotics, such as a penicillin/streptomycin (Invitrogen catalog number 15070-063) to 1X to preserve unused solution for later use. Store at 4°C.

COMMENTARY

Background Information

A variety of techniques exist for the study of cell and molecular processes involved in the etiology or progression of disease. Standard approaches include observation of morphologic characteristics by light or electron microscopy and detection of markers of disease state using immunological or mass spectroscopy based methods. These approaches work well when suitable disease markers exist or when structural defects, such as physical aberrations or detectable mislocalization of proteins, are present. Conversely, cellular processes involving short lived or unstable mediators are often very difficult to study in the context of their role in disease. Stable markers may not exist for important dynamic properties of the process under examination, such as flux. Examples of such parameters involved in disease include the production and neutralization of ROS, the regulation or disruption of mitochondrial membrane potential, the localization and concentration of biologically active ions, and cellular events such as autophagy and lysosomal maturation. These processes have been implicated in a wide range of clinical pathologies, from neurodegenerative diseases such as Parkinson’s and Alzheimer’s(5–7), to metabolic diseases such as diabetic nephropathy and diabetic cardiac hypertrophy(8–12). Mitochondrial dysfunction, as measured by mitochondrial ROS production or decreased mitochondrial membrane potential, has also been strongly implicated in the processes of cellular and organism aging(13–16).

While ROS and mitochondrial dysfunction are not directly amenable to study using classic histological and immunological techniques, several approaches have been developed to study these dynamic cellular events. Vital dyes and fluorescent protein markers provide a robust method for studying a variety of parameters in live cells in culture. These dyes allow for quantitative and descriptive analyses, provided that a suitable cell culture system exists. Unfortunately, while cell culture models can provide powerful insights into kinetic or unstable biological processes they often lack important structural information and contextual relevance with regards to the tissue they model. Even the most robust models of specific cell types are often highly derived, and nearly all cell culture models necessarily ignore the heterogeneity of cell types that exist within tissue. Furthermore, cells in culture typically lack the structural properties of the corresponding cells in situ and nearly always lack the cell-to-cell interactions that occur in vivo.

Isolated reports of imaging live, excised tissues appear in the literature (17–21), but these examples are sporadic, have utilized relatively few dyes, and typically rely on the use of two-photon microscopy. The methods here provide an approach applicable to a range of tissues using a variety of vital dyes. These techniques provide a valuable way to investigate the role of complex biological properties such as mitochondrial function, ROS, and autophagy, in disease and aging and a versatile tool potentially applicable to a broad spectrum of parameters amenable to fluorescent microscopy. Furthermore, while two-photon microscopy would greatly enhance the depth of imaging, the methods described in this Unit do not require a two-photon microscope, thus making them more widely accessible.

Critical Parameters and Troubleshooting

Careful tissue handling is the most important parameter in each of these protocols. Quality of staining, imaging, and data all depend on the viability and structural integrity of the tissue slices; the ability to collect high-quality images of control tissue(s) should be demonstrated prior to attempting to collect experimental data. If viability is poor (as compared to those in Figure 2) a number of potential factors must be considered. These include the amount of time the tissue spent between separation from (or death of) the donor, time and temperature of incubation in buffer, the care with which the tissue was sectioned, and the total time before imaging. Potential solutions may include altering the collection or euthanasia protocol, utilizing tissue cutting blocks (see Basic Protocol 1 Materials) or a vibratome (non listed here), incubating in cold buffer, and limiting the size of the experiment in order to allow for quicker processing and imaging.

In order to select suitable dyes there are three major factors that must be considered: (1) the structures or parameters of interest, (2) the spectral compatibility of the dyes with the laser and filter sets available on the microscope, and (3) the spectral compatibility each dye with the other dyes in the set. A set of four dyes can generally be used in combination, filling the spectral space of blue, green, red, and far-red dyes.

Proper selection of excitation and emission settings is also crucial to success. A set of suggested excitation wavelengths and emission filter sets are provided in Table 1. If unlisted dyes are used, or if the microscope available is not equipped with the filters or lasers suggested, optical settings appropriate to the experiment should be carefully determined. Whenever possible, cultured cells should be relied upon to test dyes and determine optimal settings.

Autofluorescence is always a potential problem in fluorescent imaging experiments. Commercially available vital dyes are typically bright and live tissue tends to have much lower autofluorescence than fixed and processed tissue slices, but tissue autofluorescence can interfere with data collection and image analysis. Extracellular matrix(17,22), NADPH(23,24), lipofuscin(25,26), and advanced glycation end products (AGEs)(27,28) are a few examples of naturally occurring substances that can interfere with imaging. In order to prevent misinterpretation of data every experiment must include a negative control to demonstrate the lack of (or at least much smaller) signal in the absence of staining. Such a control should be included for every tissue type and condition (including age), as these confounding factors vary by tissue, treatment, and source (age, species, etc).

Anticipated Results

Staining and Imaging of Live Tissue Slices

The methods listed in basic protocol 1 and alternate protocols 1 and 3 provide high resolution,

Assessing tissue viability using EthD-1 and Calcein AM

The nuclei of non-viable cells stain brightly red with EthD-1. Non-viable cells do not stain with Calcein AM, whereas intact cells stain brightly green throughout their cytoplasm. Dead cells are positive for EthD-1 and negative for Calcein AM. All cell nuclei stain brightly blue-white after staining with Hoechst 33342. The dependence of viability on depth of optical sectioning is illustrated in Figure 2. As can be seen, high viability is typically seen at 5–10um depth, although results may depend on tissue type and even the direction of sectioning (see Figure 2B1).

Assessing tissue viability using Sytox Green and Tetramethylrhodamine

The nuclei of non-viable cells stain brightly with the green Sytox nuclear dye (Figure 2C1-2). Mitochondria of dead cells lose their red TMRE staining and become weakly fluorescent or undetectable. Cells may lose mitochondrial function without losing membrane integrity. In this case the cells would show weak or no mitochondrial TMRE signal but would not show Sytox Green staining. The reverse is also possible, allowing cells to be determined viable or nonviable by multiple criteria.

Time Considerations

Tissue Collection and Processing

Tissue collection and processing should take 30–60 min, but are dependent on the skills and tools available to the researcher and can vary greatly depending on circumstances and complexity of the experiment undertaken.

Tissue Staining

Staining and destaining tissues should take <45 min.

Imaging and data analysis

Image collection and data analysis are crucially dependent on the complexity of the experiment and the experience of the researcher. Imaging generally lasts 2–4 hours, but can last 6–8 hours for complicated imaging. Image analysis depends wholly on what is being analyzed and how the data is approached and can take as little as 30 min or if the analysis is difficult many hours.