Experimental autoimmune encephalomyelitis (EAE) serves as an animal model of multiple sclerosis. This article describes an approach for scoring spinal cord inflammation, demyelination, and axonal injury in EAE. Additionally, a method to quantify soluble neurofilament light levels in the mice serum is presented, facilitating the assessment of axonal injury in live mice.
Experimental autoimmune encephalomyelitis (EAE) is a common immune-based model of multiple sclerosis (MS). This disease can be induced in rodents by active immunization with protein components of the myelin sheath and Complete Freund's adjuvant (CFA) or by the transfer of myelin-specific T effector cells from rodents primed with myelin protein/CFA into naïve rodents. The severity of EAE is typically scored on a 5-point clinical scale that measures the degree of ascending paralysis, but this scale is not optimal for assessing the extent of recovery from EAE. For example, clinical scores remain high in some EAE models (e.g., myelin oligodendrocyte glycoprotein [MOG] peptide-induced model of EAE) despite the resolution of inflammation. Thus, it is important to complement clinical scoring with histological scoring of EAE, which also provides a means to study the underlying mechanisms of cellular injury in the central nervous system (CNS).
Here, a simple protocol is presented to prepare and stain spinal cord and brain sections from mice and to score inflammation, demyelination, and axonal injury in the spinal cord. The method for scoring leukocyte infiltration in the spinal cord can also be applied to score brain inflammation in EAE. A protocol for measuring soluble neurofilament light (sNF-L) in the serum of mice using a Small Molecule Assay (SIMOA) assay is also described, which provides feedback on the extent of overall CNS injury in live mice.
Experimental autoimmune encephalomyelitis (EAE) is the most common murine model for the human demyelinating disease, Multiple Sclerosis (MS)1. Classic MS inflammatory pathology, including the infiltration of IFN-γ (gamma) and IL-17-producing T helper cells2, the infiltration of inflammatory monocytes3, the formation of perivascular and sub-meningeal inflammatory demyelinating lesions4, and the occurrence of axon injury4 in the central nervous system (CNS), is also observed in EAE5,6,7,8,9. The similarity in immune mechanisms between EAE and MS has made EAE a suitable pre-clinical model for testing the efficacy and mechanisms of action of a number of approved immune-based therapies for MS, including natalizumab, fingolimod, dimethyl fumarate, and glatiramer acetate (reviewed in1,5). Certain EAE regimens model other aspects of progressive MS pathology beyond axonal injury, including the development of sub-meningeal inflammation in the brain, chronic demyelination, spinal cord atrophy, synapse, and neuron loss6,10,11,12. Thus, EAE has utility for screening the efficacy of neuroprotective therapies for MS.
EAE is induced in rodents in a number of ways. Active immunization is the most common induction method and involves immunization of rodents with myelin antigens (either whole proteins or peptides) emulsified in CFA supplemented with heat-killed Mycobacterium tuberculosis13. Depending on the strain of mouse, pertussis toxin (PTX) is also administered on day 0 and day 2 of immunization to increase the penetrance of disease13. EAE can also be induced by adoptively transferring myelin-specific T cells obtained from myelin/CFA-primed mice into healthy mice14 or can develop spontaneously in mice that overexpress T cell receptors specific for the major myelin antigens5.
EAE disease severity and progression are commonly scored using a discrete 5-point clinical scale: 1 – tail limpness, 2 – hindlimb or foot weakness, 3 – complete paralysis in one or both hindlimbs, 4 – forelimb weakness, 5 – moribund or dead13. This clinical scoring system is sound in documenting the progression of ascending paralysis that occurs at disease onset but is less sensitive at capturing the extent of recovery from CNS inflammatory attacks. For example, both mice that ambulate with difficulty and mice that ambulate easily but exhibit foot-grasping weakness are assigned a score of 2 on the EAE scale. Scores can remain high in the post-acute phase of EAE due to the presence of permanent axon injury or loss, even despite the resolution of the inflammatory response9. There have been a variety of attempts at developing more refined scoring systems, behavioral tests, measures of hindlimb and grip strength, and infrared monitoring systems to better capture differences in clinical deficits in EAE9,16,17,18; however, these more intricate scoring measures do not distinguish the contribution of inflammation versus tissue injury to the underlying neurological deficits. Thus, the gold standard approach to score the severity of EAE is to conduct both clinical and histological scoring.
Here, a protocol is described for how to dissect and embed mouse spinal cord and brain specimens in paraffin in a way that captures the stochastic process of lesion formation that occurs in EAE. A protocol is also presented of how to stain sections with Luxol fast blue (LFB), originally created by Kluver and Barrera19, which detects myelin in the CNS. Sections are either stained with LFB alone (for demyelination analysis) or are counter-stained with hematoxylin and eosin (H&E) to help visualize and score inflammatory lesions. Protocols are also provided to quantify the presence of total leukocytes (CD45), the loss of myelin, and the number of injured axons (SMI-32) in the spinal cord using commercially available antibodies, immunohistochemical (IHC) techniques, and publicly accessible software. The protocol used to quantify leukocytes in the spinal cord can also be applied to quantify leukocytes in the brain.
Histological evaluation of axonal loss and injury in the brain is comparatively more difficult than in the spinal cord since brain white matter tracts do not run in parallel to one another. The measurement of serum neurofilament light (sNF-L) has emerged as a promising biomarker for neuronal injury in MS20,21. Recent studies have extended this technology to EAE22,23,24. Here, a method is presented to measure serum neurofilament light (sNF-L) in living mice using a Small Molecule Assay (SIMOA) assay. This method requires only a small quantity of serum and can be done in live mice in just half a day, providing rapid feedback on how a tested therapy is affecting overall CNS injury. All of the methods described here can be applied to mice of any sex or strain.
All experiments conducted with mice were performed under animal use protocols approved by the Unity Health Toronto Animal Care Committee, following the guidelines set forth by the Canadian Council on Animal Care. Ensure to wear a lab coat, protective gloves, and eyewear throughout the laboratory procedures.
1. Harvesting and fixing the brain and spinal cord
2. Grossing and processing of the spinal cord and brain
NOTE: The following steps take place in a fume hood. Before starting, prepare 2 x 10 cm clean Petri dishes, an Erlenmeyer flask fitted with a funnel lined with filter paper, two scalpels (one for cutting bone and one for cutting CNS tissues), lens paper, a pencil, embedding cassettes, and specimen jars pre-filled with 10% formalin.
3. Embedding and cutting brain and spinal cord sections
4. De-paraffinization and rehydration of sections in preparation for staining
NOTE: Steps are performed in a fume hood. Before starting, prepare baths of solvents. Prepare 5 L of 1x PBS (1 L ddH2O, 8 g NaCl, 0.2 g KCl, 1.44 g Na2HPO4, 0.24 g KH2PO4; pH = 7.4) with 0.05% Tween-20 (PBS-T) for all washing steps.
5. LFB for myelin with H&E
6. LFB for myelin without H&E
7. Antigen retrieval and peroxidase quenching for immunohistochemical (IHC) stains
NOTE: Before starting, prepare 100 mL of hydrogen peroxide in methanol (1 part 30% hydrogen peroxide solution in 9 parts 100% methanol, in a fume hood). Prepare 1 L of 10 mM citrate buffer with Tween-20 (2.94 g of trisodium citrate, dissolved in 1 L of ddH20 in a beaker on stir plate, bring pH to 6.0, and add 500 µl of Tween-20). Prepare PBS-T (see step 4). All washes are done in baths of PBS-T with gentle agitation (on a shaker) unless otherwise indicated.
8. CD45 immunohistochemistry
NOTE: This IHC method is used to visualize infiltrating leukocytes. The avidin/biotin blocking steps are combined with the blocking and primary antibody incubation steps.
9. SMI-32 IHC for axonal damage
NOTE: This protocol uses a mouse SMI-32 antibody, which reacts against non-phosphorylated neurofilament heavy, which can accumulate in injured axons25. Since this antibody has been raised in mouse and detects a mouse antigen, it is recommended that a Mouse on Mouse (MOM) kit be utilized. In this procedure, the avidin/biotin blocking step is done as a separate step from the primary antibody incubation. Before starting this protocol, de-paraffinize, rehydrate, quench endogenous peroxidase activity, and conduct antigen retrieval as described in step 4 and step 7.
10. LFB and H&E scoring for the presence of demyelinating lesions
NOTE: The following is an analysis approach that can be applied to gain rapid insights into the severity of inflammatory demyelination. This analysis is conducted in sections of the cord sampled at different levels (cervical, thoracic, and lumbar, at least 3 sections per/level). Refer to the Allen Brain Atlas for Mouse spinal cord26 to help identify the anatomical level of the spinal cord. This analysis requires TIFF files. If scanned images are in .czi format, follow instructions in Supplementary Table 4 to convert czi files to TIFF files.
11. Calculating the area fraction of LFB-staining in spinal cord white matter
NOTE: This analysis measures the percent area fraction of spinal cord white matter that is stained with LFB.
12. Analysis of the number of CD45+ cells and SMI-32+ axon ovoids
13. Measurement of sNF-L using a SIMOA assay
Figure 3 shows representative IHC and histochemical staining, with examples of both acute (left) and older EAE lesions (right). Representative CD45 staining with hematoxylin counterstaining is shown in Figure 3A,B. Figure 3C–F show examples of LFB staining with (Figure 3C,D) or without (Figure 3E,F) the H&E counterstain. Though hematoxylin is not specific to immune cells, the nuclei of the immune cells stain more darkly and can be distinguished from CNS resident cells. Figure 3G,H show representative staining of SMI-32+ axons, counterstained with hematoxylin. Notice the increased appearance of this stain in older EAE lesions.
Damage of myelinated tracks is most prevalent in the spinal cord in active murine EAE and this is the main driver of paralysis in this disease7,9. Thus, scoring for the presence of inflammation and tissue damage in the spinal cord is prioritized in histological analyses. EAE lesions occur sporadically at different regions (anterior, lateral or dorsal) (Figure 2A,B) and at different levels (sacral, lumbar, thoracic, cervical) of the spinal cord. The described embedding method ensures good sampling of lesions throughout the cord. More sections are embedded than analyzed, since some sections can become damaged in the processing or sectioning process. To ensure representative sampling, a minimum of 3 representative sections are analyzed at cervical, thoracic, and lumbar levels of the spinal cord for each mouse. The identity of each specimen is blinded in order that the person conducting the analysis is not biased when selecting representative sections for analysis.
To gain quick insights into differences in the histological severity of EAE, one can score for the presence of sub-meningeal demyelinating lesions in spinal cord quadrants in selected sections (Figure 2A,B). This is a rapid method that can be performed on scanned images or using a light microscope. This analysis is sensitive enough to detect differences in histological EAE severity between groups when EAE is severe in one group and mild in another. For example, in the experiment in Figure 4, EAE was induced in female wild type (WT) and mice with a deletion in OGR1 (OGR1 KO) using MOG p35-55/CFA plus PTX. Mice in the WT group developed severe EAE with complete paralysis, whereas the OGR1 knockout group developed mild disease. This difference in clinical score corresponded to a difference in the fraction of quadrants that had sub-meningeal lesions (Figure 4C).
It is important to complement the scoring of demyelinating lesions with percent area fraction of myelin staining to capture the extent of loss of myelin and/or myelinated axons during the autoimmune attack. In the example in Figure 4, the percent myelin fraction also differed signficiantly between the OGR1 and WT mice (Figure 4D). The percent myelin fraction also significantly correlates with the cumulative EAE score in mice with EAE (Figure 4E) and therefore serves as a good measure of overall tissue damage in this disease. Note that this protocol does not distinguish the intensity of myelin stain. If this is the desired outcome, one should conduct immunofluorescence staining for myelin proteins such as proteolipid protein or myelin basic protein and measure the intensity of this staining.
In the case where EAE is severe in both comparator groups, a higher fraction of spinal cord quadrants will contain inflammatory/demyelinating lesions. In this case, a more sensitive approach to score inflammation is to count the number of CD45+ leukocytes per mm2 white matter (see representative staining in Figure 3A). The CD45 antibody clone described here detects all infiltrating leukocytes, and only stains occasional microglia that upregulate CD45 expression in EAE (see open arrow in Figure 3B) and therefore is useful at capturing peripheral immune cell infiltration.
In longer-term EAE studies (>20 days), it is recommended that one also carry out an analysis of axon injury. SMI-32 staining in spinal cord sections is a sensitive method to detect damaged axons. Though inflammation in the spinal cord subsides with time and spared axons can re-myelinate, surviving axons exhibit a differential extent of residual injury9 (Figure 3G,H). For example, in the MOG p35-55-induced model of EAE in C57BL6/J mice, the extent of axonal injury and loss is a driver of clinical scores after the inflammatory process has subsided9. Figure 5 shows an example of this in an EAE experiment in male and female mice WT mice and mice that are deficient in a gene called peroxisome proliferator-activated receptor-delta (PPAR-delta) in the myeloid compartment (LysMCre: Ppardfl/fl). In the males, the WT mice regained hindlimb function, yet the clinical scores remained high in male LysMCre: Ppardfl/fl group. By contrast, in the experiment in the females, both experimental groups had high scores throughout. At first glance, this result suggested that PPAR-delta had a sex-specific effect in EAE; however, pathological scoring of the spinal cord revealed that mice of both sexes in the LysMCre: Ppardfl/fl group had increased axonal injury compared to WT counterparts (Figure 5B). A genotype effect on clinical scores was likely not observed in females because WT female mice tended to exhibit increased axonal injury, which manifested into chronic neurological deficits.
In this same experiment, female LysMCre: Ppardfl/fl female mice were found to have more extensive T cell infiltration in the cerebellum, providing an example of how scoring brain inflammation may be useful in EAE. In EAE, inflammation in the brain is predominantly found in the cerebellum and brain stem (Figure 6A,D,G), but can also be found in the meninges (seen under hippocampus in Figure 6C), near the ventricles (Figure 6F), and other white matter tracts including the optic nerve and corpus collosum (Figure 6B,E). Scoring brain inflammation is done in a specific brain region (e.g., cerebellar white matter) by counting the number of CD45 cells per mm2 tissue region using the same methodology as outlined for the spinal cord protocol. In the grossing method outlined here, a cut is in the middle of the cerebellum, which provides the perspective of the cerebellum and brain stem as shown in Figure 6A.
Measuring sNF-L using a SIMOA assay has become a useful biomarker for assessing ongoing axonal injury and responses to therapy in relapsing-remitting MS20,21,27,28. The same SIMOA assay kit used to measure of sNF-L humans can be applied to measure mouse sNFL22,23,24. To explore how well this assay performs is detecting axonal injury in EAE, sNF-L was measured in female C57BL6/J mice at the end-point of an EAE experiment and levels were compared to those in sex-matched healthy control mice that did not have EAE. It was found that mice with EAE had much higher levels of sNF-L than in healthy mice (Figure 7A) and these levels correlated with the density of SMI-32+ axons in the spinal cord (Figure 7B). Compared to histological scoring of axonal injury, the SIMOA assay is faster (from bleeding mice to results can be achieved in just over half a day) and therefore provides rapid feedback of how a treatment is working in living mice. This assay also has the advantage that it reflects axonal injury in both the spinal cord and the brain.
Figure 1: Representative paraffin block of brain and spinal cord sections. The 5 coronal brain sections and spinal cord cross sections (1.5–2 mm thick) are embedded in the same block in order that they can be cut in one section. At least 15 sections of spinal cord should be embedded, allowing for adequate selection of sections for analysis. Please click here to view a larger version of this figure.
Figure 2: Scoring meningeal inflammation and percent myelin area at the level of the thoracic spinal cord. (A,B) show images of the thoracic spinal cord from a female C57BL6/J mouse with MOG p35-55-induced EAE stained with LFB/H&E. Shown is the approach used to visualize quadrants and examples of demyelinating lesions (traced in dotted line). The mouse in A has 4 of 4 quadrants with confluent demyelinating lesions, while the mouse in B has 1 of 4 quadrants affected. The mouse in B does have some inflammation in other quadrants, but this has not manifested into a confluent lesion and therefore is not scored. (C–E) Example of LFB image, and the greyscale and thresholded image in imageJ. Please click here to view a larger version of this figure.
Figure 3: Spinal cord sections stained with CD45, LFB H&E, LFB and SMI-32. Examples of an early (A,C,E,G) and late (B,D,F,H) sub-meningeal lesion in the spinal cord stained for CD45 antibody (A,B), LFB/H&E (C,D), LFB alone (E,F), and SMI-32 antibody (G,H). Black arrows show examples of cells stained with each respective antibody. White arrows show putative microglia that have been stained as CD45+. Scale bar = 50 µm. This figure shows representative staining of lesions in the spinal cord of a female C57BL6/J mouse during EAE and highlights how pathology can be different across different spinal cord sections. Please click here to view a larger version of this figure.
Figure 4: Application of scoring for lesions and percent demyelination in EAE. Shown is an example of an EAE experiment where female mice deficient in Ovarian cancer G-protein coupled receptor 1 (OGR1) gene on the C57BL6/J background developed less severe clinical EAE than wildtype (WT) female C57BL6/J mice. EAE was induced by immunization with MOG p35-55/CFA plus PTX and mice were scored according to the following clinical scale: 1 = tail paralysis. 2 = hindlimb and foot weakness, 3 = hindlimb paralysis, 4 = forelimb weakness, 5 = moribund. (A) Mean + SEM clinical scores of mice over time. (B) Shown is an example of LFB/H&E staining in the ventral spinal cord. Scale bar = 50 µm. (C) Mean + SEM percent quadrants that contained demyelinating lesions. (D) Mean + SEM percent demyelination in each group. (E) shows result from another experiment in MOG p35-55-induced EAE in C57BL6/J mice where EAE scores of individual mice were summed over the 30 days of observation and were correlated with the percent demyelination in the spinal cord. Correlations were performed using a Spearman test. Panels in (A–D) are adapted from Souza C et al.29. Data in (E) are original data. *P<0.05, **P<0.01, ***P<0.001. Please click here to view a larger version of this figure.
Figure 5: Application of SMI-32 staining to understand the effect of a genotype on clinical EAE phenotype. This figure shows an example of an EAE experiment where male and female wildtype (carry floxed allele of Ppard) and myeloid specific Ppard mutant mice (LysMCre: Ppardfl/fl) on the C57BL6/J background were immunized with MOG p35-55/CFA and PTX and were followed for 45 days. (A) shows the mean + SEM clinical scores of mice. (B) shows mean + SEM results of histological scoring of the number of SMI-32+ axons in the spinal cord, %quadrants with submeningeal lesions, percent quadrants with perivascular cuffs, and #CD3 lesions in the cerebellum per mm2 tissue. This experiment showed a genotype effect on SMI-32 staining. This figure is adapted from Drohomyrecky. et al.15. Please click here to view a larger version of this figure.
Figure 6: Examples of CD45+/hematoxylin staining in brain coronal sections in MOG p35-55-induced EAE in female C57BL6/J mice. CD45+ lesions are shown in brown. (A) CD45+ lesions in the brain stem of coronal sections. Scale bar = 150 μm. (B–G) Examples of CD45+ lesions in the optic nerves (B), meningeal extensions under the hippocampus (C), the brain stem (D), the corpus collosum (E), the medial habenula near the ventricle (F), and the cerebellum (G). Scale bar: (B–G) = 50 µm. Please click here to view a larger version of this figure.
Figure 7: sNF-L levels in serum in MOG p35-55-induced EAE. (A) Serum NFL levels collected from female control and EAE mice at end-point of one experiment. Data was analysed using a two-tailed Mann Whitney test. (****p value < 0.0001). (B) Spinal cord sections were harvested at end-point and stained with SMI-32. The number of positive cells per white matter tissue area was determined and correlated with serum NF-L at endpoint using a Spearman test. Please click here to view a larger version of this figure.
Supplementary Table 1: Description of baths used in tissue processing. Cassettes are automatically moved through these series of baths using an automated processor. Please click here to download this File.
Supplementary Table 2: Steps in Luxol Fast Blue and Hematoxylin and Eosin staining. This table outlines the order of steps in the Luxol Fast Blue and Hematoxylin and Eosin staining protocol. Please click here to download this File.
Supplementary Table 3: Antibodies used for immunohistochemical staining. Described are the antibodies that are used in this protocol as well as those that can be used to further explore inflammation, microgliosis, and astrogliosis. Please click here to download this File.
Supplementary Table 4: How to convert .czi to TIFF files. Note that it is optimal to use a high-resolution image, but medium-resolution images can be saved instead if the working memory of the computer is limiting. It is imperative to use images of the same resolution across analyses. Also, note that the last image of the series is the slide label. Avoid reading the label to ensure that the analysis is blinded.30,31 Please click here to download this File.
Histological staining of the spinal cord is an important tool in assessing EAE disease severity, particularly in instances where there are differences between treatment groups in the extent of disease recovery in the post-acute phase of disease. Staining for immune cell infiltration (CD45), myelin (LFB) and axonal injury (SMI-32) helps characterize the underlying cause of the altered clinical scores in mice. The histological staining protocol described here provides a perspective of inflammation as well as the extent of myelin and axonal injury. Furthermore, the results shown validate sNF-L measurement as a method to assess the extent of overall neuronal damage in EAE.
The critical parameters for this analysis are to ensure that investigators are blinded to the identity of the sections and that there is equivalent sampling at each level of the spinal cord across the different mice. This is because the severity of inflammation can be greater at lower levels of the cord. Another critical parameter is the size of the experimental groups. Spinal cords and brains are typically harvested from 6–8 mice per group at endpoint to see significant differences between groups with treatments or genotypes having modest effect sizes. It is also important to ensure that selected mice, when averaged, have representative mean scores of the entire group. Regarding trouble shooting, a common problem encountered by those who are inexperienced with the protocol is that the spinal cord is fixed for an insufficient length of time and it is not easily extruded from the spinal column. If this is the case, the spinal cord can be manually dissected from the column by clipping along the spinous processes using fine scissors and opening the column to reveal the spinal cord. Alternatively, tissues can be fixed for a few additional days without interfering with the success of antibody staining. The antibody clones that are described here work in tissue fixed up to 2 weeks in formalin.
Embedding the spinal cord pieces requires skill and practice. It is recommended that eye loupes be worn and a lamp be directed over the embedding station to better visualize whether the sections are falling in cross-section or in longitudinal section. Keeping the lengths of the spinal cord pieces at less than 2 mm during grossing will help them fall in cross-section. Another common problem encountered for less-experienced users is that the LFB evaporates during the overnight incubation, leaving half of the slide stained and half unstained. To avoid evaporation, the glass staining dish should be sealed with thermoplastic film and then plastic wrap. If evaporation occurs and sections are unevenly stained, it is recommended to completely de-blue the slides with lithium carbonate and re-stain them again in LFB overnight. Another common issue is that users do not fully de-blue the grey matter after LFB. It is critical to examine individual sections under the microscope to ensure that a sufficient amount of de-bluing has been reached before proceeding with other steps in the protocol. In addition, though the CD45 and SMI-32 IHC stains perform robustly, it is still important to trouble shoot antibody concentrations in preliminary experiments for each new antibody lot received. This can be done by testing a variety of concentrations of the antibody on a positive control section (EAE spinal cord). First-time staining should also include a negative control that consists of secondary antibody alone without primary antibody added. Finally, it is critical in the image analysis to threshold individual images as staining can be uneven across slides or sections.
This protocol uses freely available software. If one does not have access to a processor, an embedder, or microtome, these steps can be sourced to a hospital-based pathology core that offers these services. Also, if one does not have access to a slide scanner, one can use a light microscope that is fitted with a video camera to save TIFF images of the spinal cord or brain regions. For a microscope-based workflow, capture LFB or LFB/H&E sections at low power (40x magnification) and for CD45 and SMI-32 staining, image at least four windows that are centered in the ventral, dorsal, and lateral parts of the spinal cord (200x magnification for CD45 and 400x magnification for SMI-32). Image analysis can be performed on these images to quantify staining in a similar manner as described.
The decision of what histological approach to take to score EAE is dependent on how much the clinical scores differ between groups. For example, if there are drastic differences in EAE clinical score (one group got EAE and one did not), this usually relates to differences in peripheral-mediated inflammation. In this case, scoring for the presence of demyelinating lesions on LFB/H&E-stained sections is sufficient and will reveal differences between groups. If groups are more similar in clinical score at onset and there are instead differences in the extent of clinical recovery (e.g. experiment in Figure 5A), it is best to apply the full histological workflow that is outlined here, including scoring of brain inflammation in the brain stem and cerebellum, to distinguish whether the differences the disease chronicity relate to differences in inflammation or tissue damage. If differences in inflammation are found as assessed by CD45 counting, further IHC studies can be done to stain for T cells (anti-CD3), infiltrating monocyte/macrophages (Mac3) and microglia (Iba-1/TMEM119) (recommended antibody clones are in Supplementary Table 3). Microglia activation is reflected by an increase in the intensity of Iba-1 staining on double-labeled Iba-1+TMEM-19+ microglia and an increased retraction of microglia processes that can be assessed by Sholl analysis on sections32. Furthermore, techniques like flow cytometry or single cell RNA sequencing can be applied to conduct a deeper characterization of the frequency and phenotype of immune populations in the brain and spinal cord.
The counting of SMI-32+ axons is a sensitive method to detect axon injury in EAE32,33 and in MS34. SMI-32, which detects the non-phosphorylated form of neurofilament heavy or medium accumulates in end-bulbs of transected neurons. An alternative to detect injured axons is to stain with amyloid precursor protein (APP) that can accumulate in axons as a result of disrupted axon transport33. The pattern of staining for SMI-32 and APP though both both reflective of axon injury, do not typically overlap, indicating they they are detecting different pathologies33. One can also complement histological measures of axon injury by measuring sNF-L, which is a rapid and sensitive measure of ongoing axonal injury in both the spinal cord and the brain. It offers the advantage that it can be done in half a day in living mice. A drawback of this method is that the kits are expensive and the machine is highly specialized. The company that sells the sNF-L kit does offer a fee for service for those who do not have access to a SIMOA machine. An alternative to assessing axon injury is to score for axon loss by either counting axons in toluidine blue-stained sections of the spinal cord12 or counting neurofilament bundles detected by SMI-31 in areas of the spinal cord white matter32. Both of these are more laborious approaches than SMI-32 or sNF-L measurement.
If EAE clinical scores differ between groups, but scoring for inflammation, demyelination and axonal injury does not reveal differences between groups, it may be useful to stain for astrocyte activation using GFAP (see Supplementary Table 3 for recommended antibody clone). Astrocyte activation is associated with an increase in GFAP staining and this has been shown to correlate with EAE progression in some EAE models including chronic EAE in the DA rat35.
In conclusion, this protocol describes methods and provides an analysis workflow to conduct histological scoring of EAE.
The authors have nothing to disclose.
We thank Dr. Raymond Sobel (Stanford University) for showing us his method of grossing and fix brain and spinal cord sections. We thank Kyle Roberton and Milan Ganguly from the Toronto Centre for Phenogenomics for learning the embedding method and for cutting so many of our brain and spinal cord sections. We thank Dr. Matthew Cussick and Dr. Robert Fujinami (University of Utah) for sharing their protocols for scoring submeningeal and perivascular inflammation in the spinal cord. We thank Shalina Ousman for sharing the clone of the CD45 antibody. We thank Xiofang Lu for training on the tissue processor and tissue embedding station and maintaining this equipment at the Keenan Research Centre of Biomedical Research at St. Michael’s Hospital. This work was supported by a Biomedical grant from MS Canada (to SED). Carmen Ucciferri is supported by a studentship from the Government of Canada. Nuria Alvarez-Sanchez is supported by a Keenan post-doctoral fellowship.
10% Neutral Buffered Formalin | Sigma Aldrich | HT501128-4L | Used to fix spinal cord and brain specimens |
1000 mL Glass Beaker | Pyrex | 1000 | |
15 mL Falcon Tube | Starstedt | 62.554.100 | Fixing and storing spinal cord and brain |
250 mL Erlenmeyer Flask | Pyrex | 4980 | |
500 mL Glass Beaker | Pyrex | 1003 | |
92 mm x 16 mm Petri Dishes | Starstedt | 82-1473-001 | Used in the tissue grossing procedure |
95% Ethyl Alcohol | Commercial Alcohols | P016EA95 | Dehydration and rehydration steps |
ABC Elite Kit | Vector Labratories | PK6100 | Used for immunohistochemistry labeling |
Aqua Hold 2 PAP Pen | Cole Parmer | UZ-75955-53 | Used for drawing around tissue sections in Immunohistochemical Staining |
Avidin/Biotin Blocking Kit | Vector Labratories | SP-2001 | |
Biosafety Cabinet | Any | ||
Biotinylated rabbit anti-rat IgG | Vector Labratories | BA-4000 | Used for CD45 staining |
C57BL6/J Mice | Jackson Laboratory | Stock # 664 | These mice were used in experiments shown in paper. |
Centrifuge | Thermo Fisher Scientific | Sorvall ST Plus | |
CitriSolv | Fisher Scientific | 04-355-121 | Used for de-waxing. Is an alternative to xylene |
DAB Kit | Vector Labratories | SK-4100 | Used for developing in immunohistochemistry |
ddH2O | – | – | |
Disposable Scalpel | Magna | M92-10 | Used for grossing spinal cord and brain |
DWK Life Sciences (Wheaton) glass staining dish | Cole Parmer | UZ-48585-60 | Used for histochemical staining and washes |
DWK Life Sciences (Wheaton) glass staining rack | Cole Parmer | 10061392 | Used for immunohistochemistry and histochemistry |
Eosin Y | Bioshop | 173772-87-1 | Stains cytoplasm |
Feather Microtome Blades | Fisher Scientific | 12-634-1C | Used for sectioning paraffin |
Filter Paper | Whatman | 1001110 | Used to filter the formalin (during grossing) and the luxol fast blue |
Fine Surgical Scissors | Fine Science Tools | 14160-10 | Used to snip brain and the skull |
Fumehood | Any | ||
Gibco DPBS | Fisher Scientific | 14190944 | |
Glacial Acetic Acid | BioShop | ACE333.4 | Used in the luxol fast blue staining procedure |
Histoplex Histology Containers | Starplex Scientific | 565-060-26 | Fixing spinal cord and brain |
Hydrogen Peroxide | Fisher Chemicals | H325-500 | Used to remove endogenous peroxidase in the tissue |
ImageJ | NIH | https://imagej.nih.gov/ij/download.html | |
Kimtech Science Kimwipes | Kimberly Clark Professional | 34155 | Used for immunohistochemistry |
Lens paper | VWR | 52846-001 | Used for trapping spinal cord species in cassette during processing |
Light microscope | Any | ||
Lithium carbonate | Sigma Aldrich | 554-13-3 | De-blueing after luxol fast blue staining |
Luxol blue | Sigma Aldrich | 1328-51-4 | Stains CNS myelin |
M.O.M Immunodetection Kit | Vector Labratories | BMK-2202 | Used to stain SMI-32 |
Methanol | Fisher Chemicals | A454.2 | Used for fixation |
Mayer's Hematoxylin | Electron Microscopy Sciences | 26381-02 | Stains nuclei |
Micro-Adson Forceps with Teeth | Fine Science Tools | 11027-12 | Used for reflecting the skull during dissections |
Microcentrifuge | Eppendorf | Model 5417R | |
Microvette Capillary Tubes CB 300 Z | Starstedt | 16.440.100 | Used for blood collection |
Micrscope Cover Glass | Fisher Scientific | 12545A | Used for coverslipping |
Mini Shaker | VWR | 12620-938 | Used for making buffers |
NF light kit | Quanterix | 103186 | This kit can be used for detection of mouse or human soluble neurofilament in serum |
Nitrile Gloves | VWR | 76307-462 | Safety |
Normal Goat Serum | Vector Labratories | S-1000 | Blocking reagent |
Normal Rabbit Serum | Vector Labratories | S-5000 | Blocking reagent |
OmniSette Tissue Cassettes | Fisher Scientific | M4935FS | Used for embedding spinal cord and brain |
p1000 Pipette and Tips | various | ||
p200 Pipette and Tips | various | ||
Paraffin Embedding station | Leica Biosystems | Model EG1160 | |
Paraplast Tissue Infiltration/Embedding Medium | Leica Biosystems | 39601006 | Used for embedding spinal cord and brain |
Permount Mounting Medium | Fisher Chemicals | SP15-100 | Used for mounting coverslips on slides |
pH meter | Fisher Scientific | 13636AB315B | Used for pHing buffers |
Plastic Transfer Pipettes | Fisher Scientific | 13-711-20 | Used for pHing buffers |
Potassium Chloride | BioShop | 7447-40-7 | Used for making PBS |
Potassium Phosphate Monobasic | BioShop | 7778-77-0 | Used for making PBS |
Pressure Cooker | Nordic Ware | Tender Cooker | |
Purified rat anti-mouse CD45 | Vector Labratories | 553076 | Detects leukocytes |
Reagent grade alcohol 100% | VWR | 89370-084 | Dehydration and rehydration steps |
Reagent grade alcohol 70% | VWR | 64-17-5 | Dehydration and rehydration steps |
Rotary Microtome | Leica Biosystems | Model RM2235 | |
Simoa Machine | Quanterix | HD-X | |
Slide Scanner | Zeiss | AxioScan.Z1 | |
SMI-32 mouse IgG1 antibody | Biolegend | 801701 | Detects damaged axons |
Sodium Chloride | BioShop | 7647-14-5 | Used for making PBS |
Sodium Phosphate Dibasic | Bioshop | 7558-79-4 | Used for making PBS |
Standard Adson Forceps | Fine Science Tools | 11150-10 | Used for dissection steps |
Superfrost Plus Microscope slides | Fisher Scientific | 12-550-15 | Used to collect sections |
Surgical Tough Cuts | Fine Science Tools | 14110-15 | Used to cut through the spine, body wall, and skin |
Tissue Processor | Leica Biosystems | Model TP1020 | |
Tri-soldium citrate | Thermo Fisher Scientific | 03-04-6132 | Used for antigen retrieval |
Tween-20 | BioBasic | 9005-64-5 | Used for washing sections |
X-P Pierce XP-100 plate seal | Excel Scientific | 12-140 | Used for the sNF-L Assay |
Xylene | Fisher Chemicals | 1330-20-7 | Used for de-waxing and clearing sections |
Funnel | Cole Parmer | RK-63100-64 | Used to filter formalin before grossing tissue |
Stir Plate | Any | Used to make solutions | |
Oven | Any | Used to bake tissue sections after cutting | |
Parafilm | Bemis | 13-374-10 | Used to seal LFB staining dish |
Microwave | Any | Timing may vary depending on the microwave model | |
Bovine Serum Albumin (BSA) | Sigma Aldrich | 9048-46-8 | Used to make blocking buffer |
1.5 mL Microcentrifuge Tubes | Fisher Scientific | 05-408–129 | Used to store mouse serum samples |
Vortex | Any | Used to prepare samples for sNF-L assay | |
Waterbath | Any | Used to warm enzyme substrate for sNF-L assay |