We demonstrate three different tissue preparation techniques for immunohistochemical visualization of rat retinal microvascular pericytes, i.e., cryo-sections, whole-mounts, and hypotonic isolation of the vascular network.
Retinal pericytes play an important role in many diseases of the eye. Immunohistochemical staining techniques of retinal vessels and microvascular pericytes are central to ophthalmological research. It is vital to choose an appropriate method of visualizing the microvascular pericytes. We describe retinal microvascular pericyte immunohistochemical staining in cryo-sections, whole-mounts, and hypotonic isolated vasculature using antibodies for platelet-derived growth factor receptor β (PDGFRβ) and nerve/glial antigen 2 (NG2). This allows us to highlight advantages and shortcomings of each of the three tissue preparations for the visualization of the retinal microvascular pericytes. Cryo-sections provide transsectional visualization of all retinal layers but contain only a few occasional transverse cuts of the microvasculature. Whole-mount provides an overview of the entire retinal vasculature, but visualization of the microvasculature can be troublesome. Hypotonic isolation provides a method to visualize the entire retinal vasculature by the removal of neuronal cells, but this makes the tissue very fragile.
Retinal pericytes are the focus of many research laboratories as these cells play a major role in the integrity of the vasculature. Pathological conditions such as diabetic retinopathy1, ischemia2, and glaucoma3 have vascular characteristics that involve the function of pericytes. Pericytes are found in the inner retinal capillary plexuses. The central retinal artery that supplies the inner retina branches into two layers of capillary plexuses. The inner vascular bed is situated between the ganglion cell and inner nuclear layers. The deeper layer is more dense and complex and is localized between the inner and outer nuclear layers4,5. In addition, some parts of the retina also contain a third network termed the radial parapapillary capillaries. These are long, straight capillaries that lie among the nerve fibers and rarely anastomose with one another or the other two plexuses6. Within the capillary wall, pericytes are embedded in the basement membrane and line the abluminal side of vascular endothelial cells.
To this date, there is no unique biological marker of these pericytes that can differentiate them from other vascular cells. Platelet-derived growth factor receptor β (PDGFRβ) and nerve/glial antigen 2 (NG2) are commonly used markers which both present on pericytes but also other vascular cells. Identification of pericytes is further complicated by the existence of pericyte subsets that vary in morphology and protein expression7. Currently, the best identification relies on a combination of protein markers and the characteristic positioning of the pericyte in the vascular wall. We demonstrate here three different tissue preparation techniques for immunohistochemical PDGFRβ/NG2 staining of rat retinal microvascular pericytes, i.e., cryo-sections, whole-mounts, and hypotonic isolation of the vascular network.
With cryo-sections, the retina and sclera are cut through the optic nerve. This allows for the visualization of all layered structures of neurons. The distinct ten layers of the retina are apparent as interchanging nuclear and axonal/dendritic structures that can be visualized with stains such as hematoxylin/eosin or fluorescent nuclear 4′,6-diamidino-2-phenylindole (DAPI)8. The metabolic requirements differ between the layers9 and it provides a method to determine the thickness or total absence of a specific layer (e.g., the loss of retinal ganglion cells is one of the hallmarks of retinal ischemia10,11). The vasculature is evident as transverse cuts through the retina, making it possible to separately study the capillary plexuses within the respective retinal layers12,13.
More traditionally, the investigations of the retinal vasculature network are performed in retinal whole-mounts. With this tissue preparation, the retina is cut and flattened as a flower-shaped structure. The method is a relatively fast tissue preparation technique that can highlight the overall architecture retinal vasculature and it is therefore often applied in the investigation of neovascularization in the murine retina. Successful visualization of the microvasculature in whole-mounted retinas is also reported in the developing neonatal mouse and rat retina14,15,16,17,18,19. These studies reveal a more defined pericytic activity with larger capillary-free areas in the adult compared to the neonatal retina14.
Another way of visualizing is the retinal microvasculature after hypotonic isolation. This tissue preparation technique results in retinal blood vessels and capillaries being freed of the neuronal cells. This type of two-dimensional imaging of the isolated retinal vascular network is usually performed after retinal trypsin digestion20 and used to assess the vascular abnormalities of diabetic retinopathy including pericyte loss and capillary degeneration20,21,22. The hypotonic isolation method offers the investigations of retinal vascular gene and protein regulatory responses as they has been done with RT-PCR and western blotting23,24,25. We provide here a protocol for the free-float immunohistochemical staining of the hypotonic isolated retinal vasculature as an alternative to trypsin digestion to examine the microvascular pericytes.
The protocol was optimized and demonstrated on adult male albino rats. In all experimental procedures, animals were treated according to the regulations in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Animals were euthanized by carbon dioxide and subsequent cervical dislocation.
1. Rat Retinal Tissue Preparations
2. Immunohistochemistry
The successful protocols provide three different retinal preparations for visualizing microvascular pericytes. Each of these methods uses the PDGFRβ and NG2 immunoreactivity co-localization and the unique position of the pericytes that wrap around the capillary endothelium foridentification.
With cryo-sections, the neuronal layers can be identified by the fluorescent density of DAPI-labelled nuclei and the inner and deep capillary plexuses contain pericytes that display PDGFRβ and NG2 immunoreactivity (Figure 1). The unique position of pericytes in the vascular wall can be seen as circular or horseshoe-shaped immunoreactivity in the periphery of the vessels (Figure 1, arrows) or occasional longitudinally cut vessels (Figure 1, arrowhead).
Successful visualization of the microvascular pericytes in whole-mount preparation of adult rat can be challenging14,20. PDGFRβ immunoreactivity in the whole-mounted adult retina resulted only in a very weak signal (not shown). NG2 staining outlined the inner vascular network with intense staining in abluminal pericytes (Figure 2).
To obtain an overview of the vascular network of the adult rat retina, immunohistochemistry of hypotonic isolated rat retinal vascular network provides an alternative method to whole-mount. When successfully performed, this tissue preparation technique provides an overview of the entire retinal microvasculature including inner and deep capillary plexuses. Immunohistochemical staining with pericytic markers results in the entire microvascular network showing NG2 immunoreactivity with an intense response in specific vascular cells (Figure 3). Some of these cells also display PDGFRβ immunoreactivity (Figure 3).
A careful unfolding of the tissue is important to obtain a good overview of the retinal vasculature. Insufficient unfolding can result in the vasculature being arranged in multiple focus layers which makes fluorescent imaging difficult (Figure 4). Unfolding can be facilitated by gentle shaking, tilting the chamber slide back and forth or aspirating drops of liquid onto the retinal vasculature. It should be noted that the anatomy of the retinal capillary plexuses will result in some overlay of the vascular beds. Hence, it is problematic to distinguish the inner from the deeper capillary plexuses when flattened onto the glass slide.
Figure 1: Double immunohistochemical pericyte staining of cryo-section of the inner rat retina. (A) NG2 (green) immunohistochemistry reveals positive vessels within the inner part of the retina. Arrows indicate circular and horseshoe-shaped immunoreactivity in the vessels. The arrowhead points out a longitudinally cut vessel. (B) PDGFRβ (red) immunohistochemistry show reactivity similar to NG2 immunoreactivity. Arrows and arrowhead point at the same structures as for NG2 staining. (C) The image shows a merge of NG2, PDGFRβ and DAPI (blue) by superimposing the three different filters. It is revealed that the two antibodies are co-localized (green + red = yellow). NF: nerve fiber layer, GCL: ganglion cell layer, IPL: inner plexiform layer, INL: inner nuclear layer, OPL : outer plexiform layer, ONL: outer nuclear layer. Please click here to view a larger version of this figure.
Figure 2: Immunohistochemical pericyte staining of whole-mounted rat retina. NG2 (red) immunoreactivity is visible along the superficial vascular networkwith intense staining in abluminal pericytes. The images show merged NG2 and DAPI (blue). The neuronal cells are visible as DAPI-blue nuclei between the NG2-stained microvasculature. PDGFRβ immunoreactivity in the whole-mounted adult retina resulted only in a very weak signal and is therefore not included in this figure. Please click here to view a larger version of this figure.
Figure 3: Double immunohistochemical pericyte staining of hypotonic isolated rat retinal vasculature. (A). The image shows hypotonic isolated rat retinal vascular network immunohistochemical stained with NG2 (green) and PDGFRβ (red). The complete network showed NG2 immunoreactivity. However, PDGFRβ immunoreactivity was found in cell somas. The insert of A is an overview of retinal vasculature network stained with DAPI (blue). (B) The image shows the same specimen as (A) at a different location and higher magnification. It is clear that PDGFRβ staining is only displayed in cell somas in the vascular network indicating pericyte immunoreactivity. The insert of B is the image that shows high magnification of vasculature stained with NG2 (green), PDGFRβ (red) and merged with DAPI. Including DAPI staining together with NG2 and PDGFRβ reveals three pericytes (PDGFRβ-positive) and two unidentified vessel wall cells. Please click here to view a larger version of this figure.
Figure 4: Example of an inadequate unfolded hypotonic isolated retinal vasculature. The careful unfolding of a hypotonic isolated vasculature is important as insufficient unfolding can result in the vasculature being stacked. The pictures show DAPI-stained retinal vasculature at the same site but taken in different focus layers. (A) The DAPI-stained cells in one focus layer show an approximately 25 µm thick blood vessel in the retinal vasculature. (B) The DAPI-stained cells in a second focus layer elucidate thin branching capillaries at the same site that has been stacked on the larger blood vessel displayed in (A). Please click here to view a larger version of this figure.
We present three retinal preparation techniques that can be applied in the study of microvascular retinal pericytes. Below, we provide a comparison between each of the methods and highlight critical steps in the protocols.
With cryo-sectioning, the retina is cut in sagittal sections and hence, it is possible to obtain numerous specimens from the same retina. The numeral sections resulting from this method make it an ideal choice for antibody specificity and titration testing as it prevents unnecessary animal sacrifice. The preparatory work is decisive for good results. It is crucial while dissecting the eye not to pull the eye but to cut all the muscles around the eye. There is a risk of retinal detachment if force is used. Because the eye cup is thin, one should not fix too long. Immersion with 4% paraformaldehyde in PBS for 2–4 h will fix the tissue. If fixed longer, there is a risk of destroying the antigen of interest. In order to generate the best possible immunostainings, it is critical for each section to be as flat as possible against the glass. Because the sections contain only a few occasional transverse cuts of the microvasculature, this tissue preparations technique offers only intermittent visualization of the microvasculature. Therefore, this method is not suitable for the studies of the overall vascular architecture and quantifiable measures of pericytes.
The overall vasculature can be stained after whole-mounting to provide an overview of the vasculature. When mastering this technique, it can be a fast tissue preparation procedure that requires relatively little workload. However, there are a few points to be particularly aware of. The retina is fragile and it is paramount to handle them with care and avoid rips and cuts. Furthermore, while carefully handling the retina, it is of key importance to avoid folds and other crevasses during mounting that may affect the immunohistochemical imaging later on. The mounted retina has a thickness of 250-300 µm and antibody penetration of the neuronal layers might be problematic. In regard to pericyte visualization, the NG2 staining in the whole-mount protocol could only be achieved with a Cy3-linked secondary antibody, and PDGFRβ immunoreactivity in the whole-mounted adult retina resulted in such a very weak signal that it was not shown in the figure. Hence, the usefulness of the whole-mount staining technique is strongly hampered by the successful staining depending mainly on the specific primary and secondary antibodies used. Further suggestions to optimize the protocol for visualized microvascular plexuses of an adult rat retina is by the incubation in primary antibody for a longer period of time and/or changing the incubation temperature26.
Due to the thickness of the whole-mounted, it is important to be aware of the direction of the retina after making the four incisions from the retinal periphery towards the optic nerve head to avoid accidental mounting with the inner retina containing the capillaries facing down. Still, detailed visualization of the deeper microvascular plexuses and the pericytes can be troublesome with this method14,20. As whole-mount staining of the adult retina is difficult in the deeper capillary plexus, this might result in the retina appearing avascular in some parts. The immunohistochemical staining of whole-mounted adult retina provides only a fragmented overview of the vasculature with pericytes lining the vessels in the superficial capillary plexus and therefore, this method can lead to false negative results in the visualization of the rat retinal vasculature.
Trypsin digestion has long been considered the gold standard technique for isolation and visualization of the retinal vasculature20. Hypotonic isolation provides an alternative method for the visualization of the entire complex three-dimensional vascular network and faces many of the same challenges working with very fragile tissue which is difficult to handle. Due to the resemblance in the retinal vascular product between the two methods, the hypotonic isolation faces many of the same challenges as trypsin digestion20 including careful dissection to prevent major tears and cautious handling by avoiding pipetting and touching with forceps. An important difference between the methods is the fixation. Trypsin digestion is typically done after fixation and can be applied successfully on preserved retinas that have been in fixative for several years20. Hypotonic isolation is performed before fixation and thus allows for the implementation in various other assays as previously described in literature23,24,25. The trypsin digestion is performed over two days20, whereas the hypotonic isolation can be completed within hours. Furthermore, the hypotonic isolation protocol does not include the risk of over-digestion and enzymatic cleavage of membrane markers required for pericyte recognition, e.g. PDGFRβ and NG2, making this method preferred for subsequent immunohistochemical microvascular staining.
Finally, although out of the immediate scope of this manuscript, there is a less widely known but very useful technique for isolating large, intact microvascular plexi from rodent out of the retina27. The retina is positioned in a special glass chamber and coverslip. Upon the removal of the coverslip, the microvasculature remains attached to the coverslip to create a fully alive (≈ 98% of cells) tissue print. The biggest advantage of this method is that it allows for the physiological study of the vessels and subsequent fixation for immunostaining. An obvious limitation of the method is that it does not provide the visualization/isolation of the entire microvascular network as retina flat mount and hypotonic preparations do. Still, pericytes can be visualized by application of this method27.
The three tissue preparation techniques described here are complementary in the sense that each method comprises advantages and shortcomings for visualization of the microvascular pericytes. The assessment of each methods potentials and weaknesses is important to select the optimal method for investigating the microvascular pericytes under specific pathological conditions. The demonstrated methods can be expanded to staining not only for pericyte markers but also for the visualization of microvascular structures in general. Eventually, it is for the individual researcher to select the appropriate visualization method based on the research hypothesis in question.
The authors have nothing to disclose.
The research was funded by The Lundbeck Foundation, Denmark.
Geletin from porcine skin | Sigma-Aldrich | G2625-500G | |
Albumin from chicken egg white | Sigma-Aldrich | A5253-500G | |
Deoxyribonuclease (DNAse) I from bovine pancreas | Sigma-Aldrich | D5025-15KU | Dissolved in 0.15 M NaCl |
Bovine serum albumin (BSA) | VWR | 0332-100G | |
Normal donkey serum | Jackson ImmunoResearch | 017-000-121, lot 129348 | |
Rabbit anti-PDGFRβ | Santa Cruz | sc-432 | 1:100 |
Mouse anti-NG2 | Abcam | ab50009 | 1:500 |
Alexa Fluor 594 AffiniPure Donkey Anti-Rabbit IgG | Jackson ImmunoResearch | 711-585-152 | 1:100 |
Fluorescein (FITC) AffiniPure Donkey Anti-Mouse IgG (H+L) | Jackson ImmunoResearch | 715-095-151 | 1:100 |
Cy2 AffiniPure Donkey Anti-Rabbit IgG (H+L) | Jackson ImmunoResearch | 711-225-152 | 1:100 |
Cy3 AffiniPure Donkey Anti-Mouse IgG (H+L) | Jackson ImmunoResearch | 715-165-150 | 1:100 |
4',6-diamidino-2-phenylindole (DAPI) | Sigma-Aldrich | D9542-1MG | Dissolved in DMSO |
Anti-fading mounting medium | Vector Laboratories | H-1000 | |
Anti-fading mounting medium with DAPI | Vector Laboratories | H-1200 | |
Nunc Lab-Tek II 4-well chamber slide | Thermo Fisher Scientific | 154526 |