Neurofilament Immunohistochemistry Followed by Luxol Fast Blue, for Staining Axons and Myelin in the Same Paraffin Section of Spinal Cord

Abstract

Many disorders of the central nervous system are characterized by both axonal pathology and demyelination. In assessing this concurrent pathology, techniques for staining axons or myelin are frequently used separately. Here we report the development of a combined immunohistochemical and tinctorial staining technique in which we have modified the Luxol fast blue myelin stain to be used in conjunction with a diaminobenzidine-based immunohistochemical stain for high molecular weight neurofilament (SMI-31). This modification of staining will have utility in experimental neuropathology laboratories investigating demyelination and axonal damage in human tissue and animal models.

Disorders of the nervous system including neurodegenerative disease, traumatic neuronal injury, and demyelinating disorders are characterized by different manifestations of axonal dysregulation and loss of myelin around affected axons. In the case of neurodegeneration, this axonal damage may precede or be directly related to progressive neuronal death in the disease process.^1,2 In amyotrophic lateral sclerosis (ALS), for example, a broad range of pathologies has been described that include neuronal cytoplasmic inclusions composed of neurofilaments, neuronal spheroids, and dystrophic neurites.^3–5 Axonal demyelination has also been observed in many neurodegenerative diseases including ALS in which pallor of the corticospinal tracts is a hallmark pathologic sign of the disease. Currently, stains used for the histopathologic assessment of ALS include hematoxylin and eosin and Luxol fast blue (LFB) (stain for myelin). Neither investigates axonal pathology itself or the relationship between axonal morphology and myelination. The ability to examine both of these processes (axonal loss and demyelination) to date has been largely undertaken using separate tissue sections. This requires use of twice as much tissue for experimental analysis, which in certain disease and experimental neuropathologic states can be scarce. For laboratories investigating both myelination and axonal health concurrently, immunofluorescence microscopy has typically been utilized.⁶ Although this approach is effective, it requires the use of 2 antibodies, a fluorescent microscope, long-term storage of slides at 4°C to preserve fluorescence, and is susceptible to photobleaching. More typically, a combination of silver-staining and a myelin stain are used to visualize axons and their myelin sheaths. Although these techniques remain in use with standardized protocols, they are perceived as being technically difficult and thus more difficult to attain in laboratories where they are not consistently used.⁷

Given this, we sought to develop an easily applied colorimetric approach to assess both myelination and axonal integrity concurrently. Given the primary focus of our research on understanding the pathogenesis of ALS, we have used ALS tissues as our prototypic neurodegenerative disease state. To do so we have modified the LFB stain to act as a counterstain for diaminobenzidine (DAB) stain for the axon-specific protein, neurofilament.

RESULTS

Procedures

For a detailed protocol, see Supplemental Protocol (Supplemental Digital Content 1, https://links.lww.com/AIMM/A245). To summarize, to costain for axon integrity and myelination, immunohistochemistry (IHC) was performed using SMI-31 (1:50,000 titer; BioLegend, San Diago, CA) to stain for phosphorylated neurofilament heavy polypeptide followed by a LFB counterstain. To perform counterstaining, slides were dehydrated into 95% ethanol following development with DAB using sequential ethanol baths (water, 50% ethanol, 75% ethanol, 95% ethanol). Following dehydration, slides were placed in 1% LFB (procured ready-made in solution, L0294; Sigma, St. Louis, MO) for 48 hours at 56°C. After differentiation with 0.05% lithium carbonate and 70% EtOH, slides were dehydrated fully and coverslipped.

Methodological Considerations

The combination of DAB-based IHC with LFB stain is rarely conducted. Performing LFB stain before IHC led to the loss of staining intensity of the LFB stain. Due to this issue, IHC was performed before LFB. The SMI-31 dilution required to reduce the background to levels allowing the LFB stain to remain visible was lower than previously utilized in our laboratory, and a 1:50,000 titer was needed to avoid saturation. SMI-31 was selected as an axonal marker because disruption of neurofilament protein has been widely described in ALS.^8–10 Furthermore, SMI-31 is a highly sensitive probe for both healthy and pathologic phosphorylated neurofilament which is an effective marker of axons in ALS.¹¹ Because of the deposition of the insoluble brown polymer formed by the peroxidase-catalyzed reaction with DAB, we observed that the propensity of the tissue to retain its blue hue from the LFB process was reduced. Because of this, LFB staining was performed for 48 hours to enhance the penetration of stain in the tissue. At low magnification, the blue coloration of the LFB is especially masked by the abundant brown deposition of the DAB reaction labeling axons. Despite this, the palor can still be observed at low magnification (Fig. 1B). Note that while LFB is an excellent marker of myelin, it also stains erythrocytes (Fig. 1C) and users should be aware of this additional reactivity when using to stain for myelin.

FIGURE 1: Representative photomicrographs of axon-myelin staining using combination Luxol fast blue and immunohistochemistry. A, Control spinal cord tissue demonstrating both axon and myelin staining. White matter clearly stains blue with a brown axonal stain. B, ALS case demonstrating palor of the lateral corticospinal tract (arrow) and anterior corticospinal tract (arrowhead). In this case, the right side exhibits degeneration while the left is shown substantially less pallor and axonal thinning. C, Control spinal cord tissue demonstrated healthy axons in both longitudinal (arrow) and cross-section (arrowhead). Note the additional stain of the erythrocytes (asterisks). D, High magnification (taken with ×40 objective) of lateral corticospinal tract in ALS demonstrating absence of myelin staining around axons and an apparent enlarged axonal diameter in cross-section. Note the tortuous axon (arrowhead) and dystrophic neurites can be observed. Scale bar=20 µm. E, High magnification image (taken with ×40 objective) of anterior corticospinal tract in ALS demonstrating absence of myelin and axons in this region. Scale bar=20 µm. F, Oil immersion image (taken with ×100 objective) of lateral corticospinal tract in ALS demonstrating neuritic pathology with enlarged axons (arrowhead). Scale bar=10 µm. G, Ventral horn and gray matter in rat spinal cord show intact myelin and axonal labeling. Motor neurons can be observed in the ventral horn (arrowheads). The image was taken with ×40 objective. H, Dorsal horn and gray matter in rat spinal cord showing intact myelin and axonal labeling. The image was taken with ×40 objective. I, Traditional Luxol fast blue methods counterstained with cresyl violet in rat spinal cord tissue demonstrating expected dark blue staining of white matter by Luxol fast blue along with purple staining of neuronal cell bodies by cresyl violet. The image was taken with ×20 objective. Scale bar=100 µm. J, Higher magnification of traditional Luxol fast blue staining counterstained with cresyl violet in rat tissue. Taken with ×40 objective. Scale bar=20 µm. ALS indicates amyotrophic lateral sclerosis; GM, gray matter; WM, white matter.

Cases

Archival paraffin-embedded spinal cord sections from 3 cases of ALS were used to investigate the utility of the axon-myelin stain on disease-related neuropathologic changes. One control case (clinically and neuropathologically normal) was used to verify that the observed neuropathologic changes were not due to inconsistency in the staining procedure itself (Fig. 1A). On low magnification examination, lateral corticospinal tract palor typical of ALS was evident. This was accompanied by extreme palor of the anterior corticospinal tract (Fig. 1B). Upon higher magnification examination, a qualitative reduction of the number and density of axons in the areas exhibiting demyelination was evident (Figs. 1D–F). Dystrophic neurites and tortuous (corkscrew) axons were readily visualized (Fig. 1D).

Healthy control spinal cord tissue staining revealed clearly myelinated axons along tracts in both longitudinal and cross-sections (Figs. 1A, C). No evidence of palor or demyelination was present on LFB stain, while SMI-31 stain revealed no evidence of beading, neuritic pathology, or inclusions of any kind.

Rats

All experimental protocols were approved by the University of Western Ontario Animal Care Committee (AUP #2017-108) in accordance with the policies established in the guide to Care and Use of Experimental Animals prepared by the Canadian Council on Animal Care. Parafin-embedded spinal cord sections from 3 normal adult wild-type Sprague-Dawley rats were stained using the same protocol as performed in human cases. As with the human tissue, robust labeling of axons and myelin was observed (Figs. 1G, H). A traditional LFB with cresyl violet counterstain was also conducted to compare myelin stain in the modified method to a common application of this stain (Figs. 1I, J).

DISCUSSION

The combination of stains utilized in this methodology allows for the analysis of both axonal integrity and myelination. We have used it for the assessment of descending corticospinal tracts in ALS to test its utility in human tissue and tested its efficacy in rat spinal cord tissue.

Beyond its utility in ALS, this method will have utility in other neurodegenerative diseases where axonal pathology is present, especially chronic traumatic encephalopathy, a disease hallmarked by axonal pathology including beading and spheroids.¹² Furthermore, we predict that this staining combination will have utility in multiple sclerosis in which myelination is progressively lost in neuronal tracts, preceding axonal damage and loss, and neurofilament pathology have been observed^13,14 as well as spinal cord injury models.¹⁵

The benefit of this method over fluorescence staining is the utility of only 1 antibody, half the tissue, and the stability over time at room temperature. This makes the method more easily accessible for facilities lacking access to fluorescent microscopy and improves the ability of the user to store slides indefinitely at room temperature rather than refrigerated. Although other staining techniques have previously been used to stain for axons and myelin, such as the Bodian silver stain counterstained with the Lissamine fast red solution,⁷ these techniques require specialized materials that are not commonly used in many laboratories. Use of such techniques therefore is made increasingly complex due to lack of expertise in the techniques, as well as challenges in ordering materials which are increasingly specialized, which may be associated with longer wait times and increased cost. The method we describe here is an effective technique relying on materials commonly housed in experimental laboratories making use of human and animal tissue. It may, therefore, be utilized for experimental purposes while requiring less financial and time investment.

CONCLUSIONS

This combination of stains will be useful for experimental neuropathology in human and animal models of diseases where axonal morphologic changes and myelination are affected.