Gelatin Embedding and LED Autofluorescence Reduction for Rodent Spinal Cord Histology

Abstract

We present two innovations in histological technique for rodent spinal cord: gelatin embedding and LED photobleaching. Gelatin embedding uses liquid gelatin solution to permeate delicate biological structures then solidify to provide mechanical support throughout dissection, vibratome sectioning, and staining. LED photobleaching uses high-intensity visible light during blocking and primary incubations to reduce autofluorescence in tissue sections before fluorescent secondaries are added. We found gelatin embedding improved mechanical stability without interfering with immunohistochemical staining. Gelatin embedding also preserved some spinal roots and provided an opportunity for dye-less and cut-less tracking of left/right orientation during free-floating staining, which is valuable for tissue samples that have no spare areas that can be marked. LED photobleaching greatly reduced autofluorescence and added essentially no extra time or labor to the process. Descriptions of the techniques and characterization data are provided.

1. Introduction

Histology is a collection of techniques that allow for microscopic examination of cells and tissues such as rodent spinal cords. One popular technique is immunofluorescence, which uses fluorescently-tagged antibodies to visually label proteins of interest. Successful histology requires that the tissue remain physically intact throughout the sectioning and staining process, but this can be difficult for certain tissues. Furthermore, the fluorescent signal must be brighter than background autofluorescence. In this paper, we present two techniques to help researchers keep fragile tissues intact and reduce autofluorescence.

The first technique is a gelatin embedding process that supports the long tubular structure of spinal cord tissue. Rat spinal cords are delicate, especially if injured [1] or toward the cauda equina. Gelatin embedding was pursued because of gelatin’s ability to diffuse into tissues while molten, to form an aqueous gel when cooled, and to remain strongly adhered to tissue once treated with formaldehyde [1-4]. The second technique is a light-emitting diode (LED) photobleaching technique that reduces autofluorescence by exposing tissue sections to high-intensity visible light before fluorescent secondaries are added. Autofluorescence comes from a variety of molecules including collagen and elastin, lipofuscin, ferritin, hemosiderin, vitamins, metabolites, and aldehyde fixation products [5]. Even if overall background levels are low, there may be localized concentrations of autofluorescence that can be misinterpreted as signals [6]. Photobleaching reduces autofluorescence [7-9] without the risks of degrading, discoloring, or quenching subsequent antibody fluorescence that can occur for chemical techniques [10, 11]. Since photobleaching can be performed concurrently with blocking and primary incubation, it adds essentially no additional time or labor to the process. The custom-built photobleaching machine in this study improved upon a previous design [9] by incorporating a more powerful LED array and operating in a cold room without the need for a dedicated cooling unit.

2. Materials and Methods

2.1. Overview

The combined procedure was as follows:

1. Perfuse animal

2. Remove vertebral column from animal

3. Postfix

4. Wash out fixative

5. Embed in gelatin

6. Remove excess gelatin

7. Extract spinal cord

8. Re-embed in gelatin (optional)

9. Cut sections

10. Block and incubate in primary antibodies while performing LED photobleaching

11. Incubate in secondary antibodies

12. Mount and image

2.2. Surgery, Perfusion, and Necropsy

All procedures involving animals were approved by the University of Connecticut Health Campus Institutional Animal Care and Use Committee and the United States Army Medical Research and Materiel Command Animal Care and Use Review Office. Male Sprague Dawley rats under isoflurane anesthesia had approximately 15 mm of lumbar spinal cord exposed by laminectomy and were used for experiments involving epidural ultrasound stimulation and electrophysiology. After surgery, rats were deeply anesthetized with isoflurane then transcardially perfused with 200 ml PBS followed by 200 ml 4% w/v paraformaldehyde depolymerized in PBS (4% PFA). The entire vertebral column was removed and post-fixed in 4% PFA at 5 °C overnight. Next, a piece of the vertebral column extending approximately 20 mm caudal and rostral from the surgery site was cut and most of the paraspinal muscles and connective tissues removed.

2.3. Gelatin Embedding

Prior to gelatin embedding, samples were stored in 5° C PBS for several days to allow unreacted formaldehyde to diffuse out, then warmed to room temperature before handling. Type B gelatin (SKU G9391, Millipore Sigma, St. Louis, MO, USA) was dissolved 10% w/v into PBS containing 0.1% sodium azide while stirring on a 140 °C hotplate until clear (except for air bubbles). This solution could be kept liquid in a 37 °C incubator for multiple days but not beyond a week or gelling performance was affected. Pieces of the vertebral column were submersed in liquid gelatin at 37 °C for 24 h. Then, samples were moved to 5 °C and allowed to completely solidify for several hours. Excess gelatin was trimmed away and the gelatin-infiltrated vertebral column was fixed in 4% PFA at 5 °C overnight (Fig. 1A). Finally, the processes and laminae were removed using rongeurs until the gelatin-infiltrated spinal cord came free (Fig. 1B-C).

Fig. 1.

Gelatin embedding process. (A) Dorsal surface of the vertebral column after embedding and removal of excess gelatin. Dotted rectangle indicates surgery site. (B) Spinal cord removed and cut to shorter size. Field of view matches dotted rectangle in A. (C) Ventral surface. (D) Compresstome^® sectioning. The spinal cord has been re-embedded in gelatin, loaded into the sample tube with agarose, and is actively being cut by the vibratome blade. (E) Partially-sectioned sample removed from tube to provide clear view of gelatin/agarose arrangement.

2.4. Gelatin Re-Embedding for Orientation Tracking

To track left/right orientation, gelatin-infiltrated spinal cords were placed again into 37 °C gelatin solution, moved to 5 °C, and positioned in the middle of the gelatin solution in an upright position. This was accomplished by using forceps to periodically reposition the cord while the gelatin set, most importantly when the gelatin was highly viscous but not quite solid. The gelatin was then allowed to set, trimmed into an asymmetrical shape (e.g. an off-center rectangle), and fixed overnight in 4% PFA at 5 °C.

2.5. Sectioning

For standard vibratome sectioning, a re-embedded spinal cord sample can be secured to the stage using cyanoacrylate glue and sectioned normally. The block of gelatin must be wide enough to provide adequate support during sectioning. For this reason, for specimens longer than 1 cm, we preferred to use a VF-130-0Z Compresstome^® vibratome (Precisionary Instruments, LLC, Greenville, NC, USA). The device is similar to a vibratome, but instead of holding the sample on a flat stage, the sample is held in an agarose-filled tube and extruded out one end (Fig. 1D-E). To set up the sample, the gelatin block was attached to the tissue advancement plunger using cyanoacrylate glue, the plunger was loaded into the specimen tube, and the tube was filled with molten 4% w/v agarose (SKU BP160-100, Thermo Fisher Scientific, Hampton, NH, USA) in PBS and rapidly cooled using the chilling block. Agarose was used because it physically separates from the gelatin after cutting, allowing sections to have the desired size and asymmetrical shape. 40 μm sections were obtained at a variety of settings, with best results at speed setting 1.5 and oscillation setting 3.

2.6. LED Photobleaching Device

To reduce autofluorescence, an LED photobleaching device was created (Fig. 2A). Our device used an 864 W LED full-spectrum grow light (HTG Supply, Orange, CT, USA) containing six different types of LEDs with peak emissions at 390, 430, 460, 630, 660, and 850 nm plus a seventh phosphor LED emitting broad-spectrum blue-white light with color temperature 10,000 K. The light was unfiltered except by the protective clear glass. A metal frame built from slotted angles and screws held the light above the samples and allowed unobstructed airflow to the built-in cooling fans. A 3D-printed sample holder held three 24-well plates and could be positioned any distance from the LEDs using a pair of standard laboratory stands/clamps. A shield made from metal sheets and hinges prevented bright light from bothering users (Fig. 2B). The system was operated in a large 5 °C cold room.

Fig. 2.

Photobleaching device and characterization. (A) Interior of device. (B) Device surrounded by metal shield. (C) Green-channel images of autofluorescent tissue sections treated with different durations of photobleaching and imaged using the same settings. (D) Quantification of average pixel intensities in C. (E-H) Immunofluorescent staining results using gelatin embedding and LED photobleaching. Cell nuclei (DAPI, blue) were present throughout the spinal cord. Neuronal cell bodies (NeuN, yellow) were found in the gray matter. Macrophages and microglia (IBA-1, cyan) were present throughout the spinal cord, especially in the gray matter and meninges. Spinal roots are indicated by white arrowheads.

2.7. Immunohistochemistry and Microscopy

Free-floating sections were stained similarly to [9] except with PBS/azide instead of tris-buffered saline, no antigen retrieval, 4% goat serum, 0.5% Triton X-100, all blocks and incubations overnight, and post-antibody rinse steps overnight. Additionally, biotin blocking was performed beforehand using 1% w/v avidin with 0.5% Triton X-100 then 5% w/v biotin. Primary antibodies included biotinylated mouse anti neuronal nuclei (NeuN) (Millipore Sigma, St. Louis, MO, USA) and rabbit anti ionizing calcium-binding adaptor molecule 1 (IBA-1) (Midland Scientific, La Vista, NE, USA) diluted 1:1000. Secondary antibodies included goat anti- rabbit Alexa Fluor 568 and streptavidin Alexa Fluor 647 diluted 1:1000, with 4’,6-diamidino-2-phenylindole (DAPI) added as a nuclear stain (Thermo Fisher Scientific, Gaithersburg, MD, USA). All overnight block and primary incubations were performed in the photobleacher 18 cm from the LED panel.

Images were taken with a 10× objective on an Olympus BX43 microscope (Olympus Life Science, Waltham, MA, USA) with fluorescent filters (SKU 49000, 49006, 49008, and 49011, Chroma Technology Corp., Bellows Falls, VT, USA) and an Infinity2-1RM monochrome camera (Teledyne Lumenera, Ottawa, Canada) or with a Nikon A1R confocal microscope (Nikon Instruments Inc., Melville, NY, USA). Basic brightness and contrast adjustment, cropping, and pseudocoloring was performed after imaging.

2.8. Photobleacher Characterization

To quantify the effectiveness of the photobleacher, mouse spines were obtained from the Neuroengineering and Pain Research Lab at the University of Connecticut (Storrs, CT, USA). The spinal cord was extracted by hydraulic extrusion, post-fixed overnight in 4% PFA, embedded in agarose, and sectioned 40 μm thick using a VT1200 vibratome (Leica Biosystems, Buffalo Grove, IL, USA). Six free-floating sections were exposed to 0, 1, 6, 12, 24, or 48 h of LED irradiation in PBS with 0.1% w/v sodium azide before mounting on glass slides and imaging. The sections were protected from light when not in the photobleacher to avoid confounding effects of bleaching by room light. Images were taken in gray matter on the Olympus microscope using the same exposure and imaging settings, with no digital brightness/contrast adjustments made. The average pixel intensity for each image was calculated using Fiji [12].

3. Results

3.1. Gelatin Embedding

The gelatin embedding technique was developed in order to improve the structural integrity of rat spinal cords during tissue processing and allow non-destructive tracking of right/left orientation of sections. Fig. 1A shows the dorsal surface of a piece of lumbar rat vertebral column after gelatin embedding, and removal of excess gelatin. The surgical site is within the dotted rectangle, and has gelatin intentionally left in place over the surgery site. Fig. 1B shows a closer view of the gelatin-embedded spinal cord after removal and trimming. The dotted rectangle in Fig. 1A corresponds to the entire field of view in Fig. 1B. Adherent blood products, gelatin, and spinal roots are visible. Fig. 1C shows the ventral surface with spinal roots. The gelatin stabilized the roots, allowing many of them to stay intact if handled gently. After dissection, the spinal cord was cut into shorter pieces and re-embedded in gelatin. Fig. 1D shows the Compresstome^® sectioning process in which the sample is incrementally pushed out of a tube, with agarose filling the space between sample and tube. Fig. 1E shows what remained of the sample after sectioning all but the last millimeter of spinal cord, illustrating the arrangement of agarose, gelatin, and spinal cord during sectioning. It was important to match the mechanical properties of the gelatin and agarose to avoid skipped or inconsistent-thickness vibratome sections (fixed 10% gelatin and unfixed 4% agarose in this case). The gelatin remained intact throughout the remainder of the staining process, and its intentionally asymmetric shape allowed for determination of left/right orientation during slide mounting. We recommend marking the underside of the glass slide or otherwise documenting the sections’ orientation when the section is drying on the slides just before coverslipping (when the gelatin is most visible), and later digitally flipping the images as necessary.

3.2. LED Photobleaching

A photo of our custom-built LED photobleacher is shown in Fig. 2A. Fig. 2B shows the device in typical operation, with metal shield blocking the bright light from view. To quantitatively assess the device’s effectiveness, six identically-treated, unstained sections were exposed to different durations of photobleaching. Fig. 2C shows the results imaged using the 488 nm filter set while maintaining the same camera exposure and illumination settings for each section. Quantification of the average pixel intensity in these images is shown in Fig. 2D. Autofluorescence decreased in a roughly exponential fashion.

We also tested the photobleacher’s effect on temperature in the well plate. Since we normally perform incubations at room temperature, we did not want temperatures to exceed 25 °C, nor did we want them to approach cold room temperature, which would slow down antibody diffusion. We found that a distance of 18 cm achieved the ideal steady-state temperature of 25 °C. The most extreme values tested were 2.5 cm (33 °C) and 22 cm (19 °C). It was noted that the overall temperature in the cold room was unaffected, even when operating the photobleacher continuously. Another version of the photobleacher incorporated two 864 W LED panels, one above and one below, but this resulted in too high of temperatures.

Finally, we tested whether gelatin embedding and LED photobleaching could be combined to produce high quality images. Fig. 2E-H show good signal-to-noise ratio labeling of DAPI, NeuN, and IBA-1. Preservation of several spinal roots was noted. We did not observe any UV damage to the tissue. In fact, the LED panel was found to emit very little UV. Using a 280-400 nm UVA/B meter, we measured 1-3 μW/cm² at 18 cm distance, more than an order of magnitude below the irradiance coming in through a closed glass window in our lab during the day.

4. Discussion

Gelatin has been used to provide mechanical support for delicate tissues such as spinal cord [1], brain [3], cochlea [4], and others [2]. The novelty of our technique is in performing the first embedding as early as possible (before even finishing dissection) and performing a second embedding afterward to give the sections an asymmetric shape for left/right orientation tracking. We found early embedment to be important to reduce the risk of damaging samples during dissection. Preservation of some spinal roots was an added benefit that could likely be improved with more careful dissection if desired. It might be possible to adapt our method for thick sections, cryostat sections, or paraffin sections, but we did not attempt to do so in this study. Differences in techniques in the literature include differences in fixatives, gelatin concentration, sectioning method, decalcification [4], and enzymatic digestion [1]. Common alternatives to gelatin are paraffin wax or epoxy embedding, but these require different cutting equipment and the processing can sometimes interfere with antibody binding. Aqueous hydrogels besides gelatin could be used, for example those used in tissue clearing protocols [13], but since intracellular infiltration is not needed, gelatin maintains advantages over these gels in terms of cost, ease of use, and non-toxicity. Overall, our gelatin embedding technique was very useful in our studies of rat spinal cord.

Autofluorescence has long been a problem for immunofluorescence histology, and many solutions have been developed over the years, but most investigators ignore the problem and focus instead on stronger signals and using context to identify any localized concentrations of autofluorescence. This is probably because the most well-known techniques to reduce autofluorescence are chemical treatments, which are labor-intensive and can in some cases degrade, discolor, or quench antibody fluorescence in the sections [6, 10, 11]. Tissue clearing protocols can partially reduce autofluorescence by removing lipids, but other sources may remain [13]. Light-based photobleaching offers an easier, safer, more comprehensive option. The previous iteration of our design demonstrated successful photobleaching of rabbit and cat brain sections in 24 h while preserving high-quality multi-label staining [9]. Sun and Chakrabartty created a similar system to ours using a cheaper and lower-powered 6 W white phosphor LED desk lamp while also achieving satisfactory reduction of autofluorescence within 48 h [7]. Their use of thinner (10 μm) sections and positioning the lamp closer to the sections may explain their comparable performance. Neumann and Gable reported successful photobleaching of 5 μm sections within 24 h using a combination of neon, ultraviolet, and fluorescent lamps, however their system required ice packs for cooling and could not bleach all types of autofluorescence in human brain tumor sections (though perhaps this was a particularly challenging tissue type) [8]. Lin et al. used white phosphor LEDs and dilute hydrogen peroxide for the different, yet similar, purpose of quenching secondary antibodies so that additional rounds of staining could be performed [14]. Photobleaching their Alexa-Fluor-stained cell monolayers only took 1 h and did not interfere with subsequent staining, though it is unclear whether this technique would be as effective on autofluorescent tissue sections. Ultraviolet light exposure has also been explored, but in general has been unsuccessful at suppressing autofluorescence across the full spectrum (e.g. [15]). In summary, photobleaching using broad-spectrum visible light to reduce autofluorescence is a promising alternative to chemical treatments and we widely recommend its use.

Highlights.

• Rat spinal cord embedded in gelatin before dissection to support delicate tissue.

• Re-embedding after dissection non-destructively marks left/right orientation.

• LED photobleaching reduces autofluorescence with almost no added time/labor.

• High-quality immunofluorescent staining obtained