Bruton Tyrosine Kinase in Lesions of Multiple Sclerosis and 3 of Its Models

Abstract

Background and Objectives

The pathophysiology of multiple sclerosis (MS) is contributed by B lymphocytes, macrophages, and microglia. Bruton tyrosine kinase (BTK) is an intracellular enzyme within these cells that modulates their inflammatory properties. Thus, central nervous system–penetrant inhibitors of BTK may counter immune dysregulation, and this aspiration is highlighted by 11 phase 3 clinical trials in MS to inhibit this enzyme. Despite the keen interest, the spatial and temporal elevation of BTK in lesions of MS and its models is not well characterized.

Methods

We used quantitative fluorescence immunohistology to assess the expression of BTK and a phosphorylated activated form in different lesion types of MS and 3 of its models: inflammatory experimental autoimmune encephalomyelitis (EAE), toxin-induced demyelination of lysolecithin, and oxidized phosphatidylcholine injuries. GDC-0853 (fenebrutinib), a BTK inhibitor in phase 3 clinical trials in MS, was evaluated in EAE for its capacity to alter disease course.

Results

We observed low expression of BTK and a phosphorylated form (pBTK) in murine spinal cord but significant upregulation in white matter lesions inflicted by oxidized phosphatidylcholine, lysolecithin, and EAE. Expression predominantly localized to microglia/macrophages shown through colocalization analysis by Imaris 3-dimensional rendering. GDC-0853 (fenebrutinib) significantly reduced clinical severity of EAE when administered prophylactically and marginally ameliorated disability when initiated from onset of clinical disability. Finally, we report the increase in BTK expression in microglia/macrophages in active plaques and in the hypercellular rim of chronic active lesions of MS. In the inactive core of chronic active MS lesions, the few remaining HLA-DR⁺ myeloid cells were still BTK immunoreactive.

Discussion

Our results demonstrate that BTK immunoreactivity is normally undetectable in uninjured areas or normal-appearing white matter of human and murine CNS, but that expression becomes prominent in lesions with hypercellular aggregates of microglia and macrophages. Staining for pBTK reveals that its upregulation declines in the later stage of lysolecithin and chronic stage of EAE injury while BTK upregulation is maintained. Our collective results support the rationale of using brain-penetrant BTK inhibitors to modulate the elevation of this enzyme in microglia/macrophages within inflamed plaques of MS.

Introduction

Bruton tyrosine kinase (BTK) is a cytoplasmic enzyme that belongs to the Tec family of kinases. Genetic sequence variant^1-49 of BTK presents clinically as agammaglobulinemia^1-3 due to its importance in regulating the proliferation, maturation, and activity of B lymphocytes.^4,5 BTK expression is not required for the survival of mature B cells,⁶ so pharmacologic inhibitors of BTK may regulate aberrant B-cell functions in MS without overtly depleting their numbers; the latter is a concerning feature of anti-CD20 monoclonal antibodies used in MS.⁷ Interest in BTK inhibition has escalated with the appreciation that the enzyme also modulates activity of macrophages and microglia,⁸ which are myeloid cells that aggregate in active and chronic active lesions of MS.⁹ Persistent proinflammatory and degeneration-associated microglia are proposed to exacerbate progression of disability in MS by inducing axonal injury and loss, and demyelination.^10,11 In tissue culture, BTK inhibitors such as fenebrutinib and evobrutinib attenuate the proinflammatory activity of microglia that is stimulated by immune complexes that signal through Fc receptors on myeloid cells.^12-14 Thus, CNS-penetrant BTK inhibitors have the potential to reduce ‘compartmentalized’ inflammation ‘smoldering’ within the CNS and slow the accumulation of disability.^7,10,15

The interest of BTK inhibitors in MS is manifested by 11 phase 3 trials involving 4 drugs: evobrutinib, tolebrutinib, fenebrutinib, and remibrutinib.^15-18 In phase 2 trials in relapsing MS, evobrutinib, tolebrutinib, and fenebrutinib reduced annualized relapse rates when compared with placebo.^19-21 All 4 medications have/had 2 simultaneously running phase 3 trials in relapsing MS while tolebrutinib has an additional trial each in nonrelapsing secondary progressive MS (SPMS) and primary progressive MS (PPMS), and fenebrutinib with a trial in PPMS.

There are differences to the inhibitors including their selectivity to BTK and pharmacokinetic properties such as their capacity to reach the CNS parenchyma.^18,22-24 Moreover, evobrutinib, tolebrutinib, and remibrutinib are irreversible covalent inhibitors²⁵ while fenebrutinib is a reversible noncovalent inhibitor of BTK.²²

Relative to the enthusiastic pursuit of BTK inhibitors in clinical trials in MS, the basic science studies of BTK in the CNS have lagged behind. There is scarce analysis on the expression of BTK in lesions of MS, with most of the articles^12,14,26-28 containing one panel of BTK staining of an MS lesion or reporting only on transcript elevation. In mouse models, microglia have been shown to express the highest level of BTK expression, but neurons and oligodendrocytes do not express BTK.^29-32 In addition, BTK staining overlaps with CD19⁺ B cells and Iba1⁺ microglia and macrophages in the meninges of a progressive EAE model.³³ Sparse literature is available on the activated phosphorylated form of BTK at tyrosine residue 223, henceforth called pBTK.

Given the scarcity of information on BTK and pBTK in the CNS, we set out in this study to examine the spatiotemporal expression of BTK and pBTK in MS and in 3 of its animal models: inflammatory EAE, toxin-induced lysolecithin, and oxidized phosphatidylcholine injury. We also evaluated the effect of GDC-0853²², fenebrutinib, in EAE. This article describes our findings.

Methods

Mice

All animal experiments were approved by the University of Calgary Animal Care Committee and conducted following regulations established by the Canadian Council of Animal Care. Lysolecithin (LPC) and oxidized phosphatidylcholine (OxPC) samples were generated using 6–8-week-old C57BL/6 female mice from Charles River (Montreal, Quebec), and transgenic CX3CR1^CreER-Eyfp/+:Rosa26^tdTomato/+ mice were 8–12-week-old female. CX3CR1^creER-Eyfp (JAX 021160) and Rosa26^tdTomato (JAX 007909) strains from Jackson Laboratory (Bar Harbor, ME) were bred at the University of Calgary to produce CX3CR1^CreER-Eyfp/+:Rosa26^tdTomato/+ mice. Mice immunized for EAE were C57BL/6 female mice that were 9–10-week-old from Jackson Laboratory. All mice were housed between 23°C and 25°C with 12-hour light and dark cycles with free access to food and water.

EAE Model and Treatment With GDC-0853

Myelin oligodendrocyte glycoprotein (MOG) 35–55 peptide (50 μg/mouse) was emulsified in complete Freund adjuvant containing 11 mg/mL of heat-inactivated Mycobacterium tuberculosis H38RA, and it was placed on ice until ready to use. Mice were anesthetized with an intraperitoneal injection of ketamine (100 mg/kg) and xylazine (10 mg/kg) solution, and emulsified MOG (50 μL) was injected subcutaneously into each hind flank. Each mouse also received 300 ng of pertussis toxin i.p. on the day of MOG immunization and day 2 after immunization. Mice were monitored for clinical signs of EAE using a 15-point scoring scale.³⁴ Mice were sacrificed during the acute peak of disease severity (day 18) and during the chronic phase of the disease (day 35).

Mice were either treated from day 0 (EAE induction) or from the clinical onset of disease (around day 11), with twice a day (b.i.d.) 10 mg/kg of GDC-0853 (Genentech) suspended in 200 μL of 1% hydroxypropyl methylcellulose, 0.2% Tween-80, and 100 mM sodium citrate adjusted to pH 3.0 (HPMCT). Vehicle controls received equivolume of 1% HPMCT. Clinical onset was defined as having an EAE score of at least 1, which presents as partial tail paralysis.

LPC and OxPC Models

LPC and OxPC surgeries were completed as described.^35,36 Specifically for modeling OxPC-associated oxidative injury in MS, 0.5 µL of 10 mg/mL 1-palmitoyl-2-(5′-oxo-valeroyl)-sn-glycero-3-phosphocholine (POVPC), a purified OxPC species highly elevated in MS lesions,³⁶ was stereotaxically deposited into the ventral spinal cord white matter of mice, between the T3 and T4 vertebrae. Tissues were then collected for analysis after 7 days. The CX3CR1^CreER-Eyfp/+: Rosa26^tdTomato/+ transgenic mice aged 8–12 weeks were injected intraperitoneally with 2 mg/20 g mouse of tamoxifen (Sigma, T-5648) in corn oil (Sigma, C-8267) for 3 consecutive days and received surgery 3 weeks after tamoxifen injections.

Spinal Cord Harvest and Sectioning

Mice were sacrificed through an i.p. overdose of ketamine/xylazine and then transcardially perfused with approximately 12 mL of cold phosphate-buffered saline (PBS). The spinal cord was removed, and the thoracic section of the spinal cord was kept for histologic analysis. The thoracic spinal cord was placed into 4% paraformaldehyde (PFA) solution for 24 hours at 4°C and then switched to 30% sucrose solution for 72 hours for cryopreservation. Spinal cords were frozen in Frozen Section Compound 22 Clear (Leica, 3801480) over dry ice with 2-methylbutane (Sigma, 78-78-4). Frozen tissues were kept at −80°C before cryosectioning coronally for LPC and OxPC tissues and longitudinally for EAE tissues into 20-μm sections. Tissue sections were collected on slides and stored at −20°C.

Murine Tissue Immunofluorescence Staining and Microscopy

Tissue slides were thawed at room temperature for 10 minutes and rehydrated with PBS. If slides were stained for myelin basic protein (MBP), samples were delipidated with a series of ascending 50%, 70%, 90%, 95%, and 100% ethanol washes and then descending concentration washes for one minute each. Slides were then fixed with 4% PFA for 20 minutes at room temperature and washed with PBS. Tissues were permeabilized with 0.01% Triton X-100 (Sigma, X100) PBS for 5 minutes before blocking with horse serum solution (0.01 M PBS, 10% horse serum, 1% bovine serum albumin (BSA), 0.1% cold fish skin gelatin, 0.1% Triton X-100, and 0.05% Tween-20) for 1 hour at room temperature. Slides were then incubated with primary antibodies diluted in antibody dilution buffer (0.01 M PBS, 1% BSA, 0.1% cold fish skin gelatin, 0.5% Triton X-100) at 4°C overnight. The next day, slides were washed 3 times with 0.1% Tween-20 in PBS for 5 minutes and then incubated with fluorophore-conjugated secondary antibodies (1:400) and 1 μg/mL of DAPI in antibody dilution buffer for 1 hour at room temperature. The slides were washed 3 times again with PBS-Tween and then mounted using Fluoromount G (SouthernBiotech, 0100-01). The stained slides were stored at 4°C until ready for imaging with the Leica TCS SP8 confocal laser scanning microscope. For overview images, slides were imaged using the Olympus VS110-S5 Slide Scanner.

Human MS Brain Tissue Immunofluorescence Staining

Human postmortem brain tissue from 5 patients with MS, all secondary progressive at death, was obtained from the University of Montreal brain bank. Diagnosis of MS was according to the revised 2010 McDonald criteria,³⁷ and brains were processed immediately after the legal medical assistance in dying program. Donors provided written informed consent for their brains to be used in research. The use of these human tissues for research was also approved by the Conjoint Health Research Ethics Board at the University of Calgary (Ethics ID REB15-0444). Autopsy samples were preserved, and lesions were classified³⁸ using Luxol fast blue/hematoxylin and eosin staining. Staining and analyses were conducted as mentioned above for mouse tissues.

Fluorescence Confocal Microscopy Image Analysis

Fluorescent images acquired using confocal microscopy were analyzed using ImageJ (National Institute of Health). Maximum intensity projections were created for each image from the z-stack and then split into their respective color channels. The region of interest (ROI)—defined by the area of demyelination or elevation of immune cell density—was manually traced on the channel containing MBP or CD68 staining and pasted onto each channel of interest from the same image series. The outside of the selection was cleared to limit analysis to the ROI. The fluorescent intensity threshold was set and held constant across each set of experiments to determine the positive signal. The size and circularity settings of the analyze particle function in ImageJ were held constant across all samples within each experiment. The positive signal area was divided by the area of the ROI to obtain the percent staining area of a particular marker. For counting the number of OPCs, Olig2 and PDGFRα double-positive cells were counted manually within the lesion ROI and then calculated to yield the number of cells per mm².

Antibodies

The following primary antibodies were used for immunofluorescent staining: rabbit anti-mouse arginase-1 (1:200; Cell Signaling Technology, 93668), mouse anti-human BTK (1:100; BD Biosciences, 611117), rabbit anti-mouse BTK (1:200; Invitrogen, PA5-20085), rabbit anti-mouse/human pBTK (1:200; Novus Biologicals, NBP1-78295), rat anti-mouse CD68 (1:500; BioLegend, 137002), rat anti-mouse dectin-1 (1:100; InvivoGen, mabg-mdect), mouse anti-human HLA-DR (1:500; Invitrogen, 14-9956-82, clone: LN3), rabbit anti-mouse/human Iba1 (1:500; Wako, 019-19741), goat anti-human Iba1 (1:200; Novus Biologicals, NB100-1028), goat anti-mouse IL-1β (1:200; R&D Systems, AF-401-NA), rat anti-mouse iNOS (1:100; Invitrogen, 14-5920-82), chicken anti-mouse MBP (1:500; Invitrogen, PA1-10008), rat anti-mouse MHC-II (1:500; Invitrogen, 14-5321-82), mouse-anti-HLA-DR (1:500, Invitrogen, 14-9956-82), mouse-anti-PLP (1:500, Bio-Rad, MCA839G), rabbit anti-mouse NF-H (1:1000; EnCor Biotechnology, RPCA-NF-H), rabbit anti-mouse Olig2 (1:200; EMD Millipore, ab9610), and goat anti-mouse PDGFRα (1:100; R&D Systems, AF1062).

The following secondary antibodies were used: Alexa Fluor 488 donkey anti-rat IgG, donkey anti-rabbit IgG, donkey anti-mouse IgG, and donkey anti-chicken IgY; Alexa Fluor 594 donkey anti-goat IgG, donkey anti-rabbit IgG, and donkey anti-rat IgG; Alexa Fluor 647 donkey anti-goat IgG and donkey anti-rabbit IgG; and HRP-conjugated goat-anti-mouse IgG (1:1000, Abcam, ab205719). All secondary antibodies were purchased from Jackson ImmunoResearch and used at a concentration of 1:400, except for Alexa Fluor 546 goat anti-mouse IgG2b (Invitrogen, A-21143). DAPI (1:1000; Thermo Fisher Scientific, 62248) was used to counterstain the nucleus.

Imaris Image Analysis

Fluorescent confocal microscopy images of the spinal cord and MS brain tissue were used for Imaris 3D rendering. For each marker of interest, a surface was created with seed points for each cell. The signal threshold and threshold volume were set to be consistent within a set of stained tissues. For colocalization analyses, the overlap distance between 2 stains was set to zero and the number of double-positive cells was recorded for subsequent statistical analysis.

Statistical Analysis

All statistical analyses were completed using GraphPad Prism 9. The percent staining area for each spinal cord was averaged from 4–5 images in the EAE model. Images from the LPC and OxPC models were analyzed using the Student t test, Mann-Whitney test, or analysis of variance (ANOVA) with Sidak post hoc comparison. EAE images were analyzed using one-way ANOVA with Sidak post hoc multiple comparisons between groups. EAE mean clinical scores of mice were compared using 2-way ANOVA with Bonferroni post hoc for comparisons between groups or with the Mann-Whitney test for area under the curve. Graphs include the mean value with error bars that represent the SD for all graphs except for the EAE mean clinical score graph in eFigure 1A in which error bars represent the standard error of the mean.

Data Availability

All data will be provided by the corresponding author, V.W. Yong, on reasonable request.

Results

Elevation of BTK and pBTK in the Spinal Cord of EAE-Afflicted Mice

We started by evaluating the expression of BTK and pBTK in EAE. The mean disability scores of EAE-afflicted animals on a 15-point scale³⁴ are displayed in eFigure 1A. The spinal cord was harvested at day 18 and day 35, referred to as acute and chronic time points, respectively. Accumulation of CD68⁺ or Iba1+ microglia/macrophages was used to define white matter lesions in the thoracic spinal cord. Figure 1A displays a low-magnification slide scanner image of a longitudinal section of a day 18 EAE spinal cord, which contained numerous CD68⁺ lesions (green) along the lateral column white matter (top and bottom of the image). BTK expression (red) was readily detected in lesional areas but not in parenchymal regions without microglia/macrophage accumulation. Higher magnification images of a day 18 spinal cord lesion showed BTK elevation to correspond closely to Iba1+CD45⁺ microglia/macrophage immunoreactivity (Figure 1B). There was no signal when the BTK antibody was omitted (data not shown).

Figure 1. Immunofluorescence Staining and Quantitation of BTK and pBTK in Thoracic EAE Spinal Cord With Acute (Day 18) and Chronic (Day 35) Disability.

(A) Low-magnification slide scanner image of a day 18 EAE thoracic cord depicting multiple hypercellular (DAPI) lesions containing CD68⁺ microglia/macrophages that overlap with BTK immunoreactivity. In this representative longitudinal section, the white matter is on the lateral top and bottom edges of the spinal cord with the central canal in the middle. (B) A day 18 EAE lesion is shown with CD45 representing immune cells, Iba1 for microglia/macrophages, and BTK. Analyses of BTK (c) and pBTK (D) ROI (region of interest) in the respective areas and time points. Statistical analysis used one-way ANOVA with Sidak post hoc comparisons between groups. Error bars represent SD; ***p < 0.001, ****p < 0.0001. BTK = Bruton tyrosine kinase; EAE = experimental autoimmune encephalomyelitis; ns = not significant; ROI = region of interest.

Next, we evaluated BTK protein expression in lesions vs normal-appearing white matter (NAWM) of mice with EAE in the acute vs chronic phase. While BTK elevation was obvious in hypercellular lesions, low BTK expression was found in NAWM with sparse density of CD68⁺ cells and also in the spinal cord of naïve mice that were not immunized (eFigure 1B). At day 35 of EAE, CD68⁺ density in hypercellular lesional areas of the spinal cord was still prominent with elevated BTK immunoreactivity. Quantitation of BTK expression in the ROI affirmed the elevation of BTK in acute lesions, which was maintained in the chronic period. BTK expression in NAWM was similar to the naïve spinal cord at both time points (Figure 1C).

We assessed the expression of activated BTK using an antibody to BTK phosphorylated at the Tyr223 residue, henceforth called pBTK. pBTK was readily detected in acute EAE lesions but not in NAWM or naïve spinal cord white matter (eFigure 1C). At day 35, pBTK expression in chronic lesions was approaching naïve levels (Figure 1D).

These findings show that BTK expression is upregulated in EAE spinal cord lesions at acute and chronic time points, but despite the presence of high immunoreactivity of BTK in CD68⁺ microglia and macrophages at day 35, pBTK elevation was no longer maintained.

Upregulation of BTK and pBTK in Lysolecithin Lesions

The lysolecithin (lysophosphatidylcholine, LPC) model of demyelination and remyelination of MS was next used to observe the temporal expression of BTK and pBTK. In this model,^35,39 day 3 represents a period of demyelination, day 7 as time of repopulation of oligodendrocyte lineage cells, day 14 as onset of remyelination, and day 21 as a period of active remyelination (eFigure 2A). Coronal sections of the focal lesion were stained with CD68 to define the lesional area, which was localized to one side of the ventral white matter of the spinal cord and did not affect the contralateral side, which was used as comparative NAWM (Figure 2A). BTK expression was evident in the lesional area but can also be observed along the meninges, whereas pBTK expression was more localized to the lesion (Figure 2B).

Figure 2. Immunofluorescence and Quantitation of BTK and pBTK Expression in Lysolecithin Injury.

Whole spinal cord coronal image showing the expression of BTK (A) and pBTK (B) with CD68, indicating the lesion site and the NAWM in single panels and merged images at day 7 of injury. The adjacent ventral white matter is used as NAWM. Temporal expression of BTK (C) and pBTK (D) in the lesion at days 3–21 after injury. Statistical analysis with one-way ANOVA with Sidak post hoc for comparisons between groups. Error bars represent SD; *p < 0.05, ****p < 0.001. BTK = Bruton tyrosine kinase; NAWM = normal-appearing white matter.

At day 3 of LPC demyelination, BTK expression was observed but was not significantly upregulated until day 7. Elevated immunoreactivity of BTK was sustained at day 21 in comparison with the contralateral NAWM (Figure 2C, eFigure 2B). The expression of pBTK was increased at all time points observed compared with that in the NAWM, but on day 21, there was relatively lower expression of pBTK in comparison with day 14 (Figure 2D, eFigure 2C).

Overall, similar to observations in the EAE model, BTK is elevated in lesions for prolonged periods after injury, but there was a relative reduction in the activation state of BTK at the later time point (day 21). In addition, BTK and pBTK expressions were closely associated with areas of CD68⁺ microglia/macrophages in lesional areas.

Increase in BTK and pBTK After OxPC Injury

We evaluated another toxin-induced injury model for BTK expression inflicted by OxPC formed from oxidation of membrane lipids, which is prevalent in active MS lesions.³⁶ Specifically, we used purified POVPC to model oxidative injury in MS because it is one of the most abundant species of toxic oxidized phosphatidylcholines elevated within MS lesions.³⁶ Lesions from mice 7 days after POVPC injection in the ventral spinal cord white matter were examined. We observed elevation of CD68⁺ microglia and macrophages at the focal lesion and minimal immunoreactivity of CD68⁺ cells in the contralateral NAWM (Figure 3, A and B, eFigure 3A). BTK and pBTK expressions were apparent in the lesion at the one time point measured (day 7) and were seen in close association with CD68⁺ cells (Figure 3, C and D, eFigure 3B). Figure 3C illustrates an Imaris 3D reconstruction of pBTK (red) with CD68 (green) showing their close overlap.

Figure 3. Immunofluorescence Analysis of BTK and pBTK Expression in Day 7 OxPC Injury.

Representative image of BTK in lesion inflicted in the ventral white matter of the thoracic spinal cord, with CD68 in merged or single-panel images (A) and its quantitation in lesional ROI (B). (C) Imaris rendition of day 7 OxPC lesion of pBTK (red) and CD68 (green). (D) The quantification of % pBTK in lesional ROI is displayed. Scale bars represent 100 μm. Statistical analysis with the two-tailed unpaired Mann-Whitney test for comparisons between groups. Error bars represent SD; **p < 0.01, ***p < 0.001. BTK = Bruton tyrosine kinase; OxPC = oxidized phosphatidylcholine; ROI = region of interest.

Both Microglia and Macrophages Express BTK and pBTK After Injury

To determine whether the expression of BTK and pBTK was preferentially induced by microglia vs monocyte-derived macrophages, we used the tamoxifen-inducible CX3CR1^CreER-Eyfp/+:Rosa26^tdTomato/+ transgenic mouse line. Intraperitoneal administration of tamoxifen induced tdTomato expression in CX3CR1⁺ cells, which included both microglia and macrophages. OxPC surgery was performed 3 weeks after tamoxifen injection to allow fast repopulation of bone marrow–derived monocytes in the circulation to renew as wild-type cells while slow turnover microglia remained tdTom⁺; CD68⁺tdTom⁺ cells in lesions were designated as microglia, whereas CD68⁺TdTom^- cells were considered macrophages (eFigure 4A). At day 3 after surgery, we observed extensive representation of CD68⁺ and tdTom⁺ cells within the lesion, along with BTK or pBTK expression (Figure 4, A and B, eFigure 4, B and C). At higher magnification, it was observed that CD68⁺tdT^- (macrophages) and CD68⁺tdT⁺ (microglia) cells were both present within lesions, although microglia outnumbered macrophages (Figure 4, C and D). Both microglia and macrophages in lesions expressed BTK and pBTK, which was verified using Imaris 3D image rendering that showed overlapping surfaces of CD68, tdTomato, and BTK or pBTK (Figure 4C). Quantification showed that both cell types had similar proportions of cells that were BTK⁺ and pBTK⁺ (Figure 4D). Altogether, these results demonstrate that both microglia and macrophages upregulate the expression of BTK and pBTK in lesions.

Figure 4. BTK and pBTK Are Expressed in Both Microglia and Monocyte-Derived Macrophages.

Spinal cord coronal image showing the expression of BTK (A) and pBTK (B) in CX3CR1^CreER-Eyfp/+:Rosa26^tdTomato/+ transgenic mice at 3 days after OxPC injury. (C) Imaris rendering shows that BTK and pBTK are found within macrophages (CD68+tdTom-) and microglia (CD68+tdTom+). (D) Microglia are more abundant within lesions than macrophages, and most of the macrophages and microglia within lesions express BTK and pBTK. Scale bars represent 100 μm. Statistical analysis with the two-tailed unpaired Mann-Whitney test for comparisons between groups. Error bars represent SD; *p < 0.05, ****p < 0.0001. BTK = Bruton tyrosine kinase; OxPC = oxidized phosphatidylcholine.

GDC-0853 Treatment in MOG_35-55 EAE

To evaluate the impact of BTK upregulation and activation by microglia and macrophages on CNS inflammation, we assessed the effect of the BTK inhibitor, GDC-0853, on EAE disease. GDC-0853, fenebrutinib, is being evaluated in phase 3 trials of relapsing MS and PPMS. The MOG_35-55 EAE model was used for treatment evaluation because it has both adaptive and innate immune system involvement like MS while the LPC and OxPC models only engage the innate immune system.

In a preventative treatment paradigm, GDC-0853 (10 mg/kg) administration was initiated from the time of MOG immunization twice a day by oral gavage until day 23. Mice that received GDC-0853 had significantly reduced EAE clinical severity and cumulative EAE scores compared with 1% HPMCT vehicle controls (Figure 5, A and B); EAE disability was scored blind. Staining of the longitudinal thoracic sections of the spinal cord revealed improved preservation of myelin and axons in the GDC-0853 group, indicated by higher percent areas of MBP and NF-H expression, respectively, compared with the vehicle EAE group (Figure 5C, eFigure 5, A–C). However, no significant differences in CD68 expression were observed within lesion sites (eFigure 5B), likely because of the high variability of data across samples in each group.

Figure 5. Effect of GDC-0853 Initiated at MOG_35-55 Immunization in Mice With EAE.

(A) The mean clinical score of mice with EAE treated with vehicle (n = 4) or GDC-0853 (n = 8) (10 mg/kg b.i.d.) orally from day 0. (B) The cumulative EAE score of mice depicting the burden of disability per mouse over the course of the experiment (mean ± SD). Two-way ANOVA with Bonferroni post hoc for comparisons between treatment groups for longitudinal scoring and the two-tailed unpaired t test for cumulative EAE scoring; *p < 0.05, **p < 0.01, ***p < 0.001. (C) NF-H density in the spinal cord ROI. Two-tailed unpaired Mann-Whitney test; *p < 0.05. ANOVA = analysis of variance; EAE = experimental autoimmune encephalomyelitis; ROI = region of interest.

Next, we evaluated GDC-0853 at 10 mg/kg twice a day by oral gavage when treatment was initiated after clinical signs (manifesting as a limp tail) have become apparent. Over 20 days of treatment and where disability was scored blind, there was a marginal but statistically significant reduction in the clinical scores of mice treated with GDC-0853 compared with vehicle controls (Figure 6, A and B). Analyses of the longitudinal thoracic spinal cord of these mice showed no significant differences in myelin and axonal preservation (Figure 6, C and D), CD68 density (eFigure 6A), and the number of olig2+PDGFRα+ oligodendrocyte precursor cells (OPCs) within lesions (eFigure 6, B and C). Further analyses of microglia and macrophage properties, including MHC-II and dectin-1, showed no significant differences between vehicle controls and drug-treated mice (eFigure 6, D–F).

Figure 6. Effect of GDC-0853 Initiated at Onset of Clinical Signs in Mice With EAE.

(A) Longitudinal mean clinical score of mice with EAE treated with vehicle (n = 19) or GDC-0853 (10 mg/kg, n = 16) b.i.d. orally from onset of disability, with treatment initiation depicted as day 1 in the graph. The overall difference (Mann-Whitney area under the curve) between the 2 groups is p = 0.012. (B) The cumulative EAE score of vehicle and GDC-0853–treated mice. Percent MBP (C), NF-H (D), and CD68 (E) in ROI (mean ± SD). EAE = experimental autoimmune encephalomyelitis; ROI = region of interest.

Overall, we found that early treatment initiation (day 0) of BTK inhibitor GDC-0853 before EAE clinical onset significantly reduced EAE severity and promoted myelin and axonal preservation. However, treatment initiation after clinical onset was marginally effective in reducing EAE disability, which was reflected in the absence of significant changes to myelin, axonal, OPC, and microglia markers.

Expression of BTK in Lesions of MS Brain

Finally, we determined whether BTK expression is also present in MS autopsy tissues. Frozen human MS brain sections from 5 patients with SPMS were stained using Luxol fast blue (LFB), hematoxylin and eosin (H&E), PLP, and HLA-DR (or Iba1) to determine areas of demyelination and immune activity; an example of a large, demyelinated lesion can be observed in the middle of the section displayed in Figure 7A. Staining for Iba1+ microglia/macrophages in the same sample showed a hypercellular region within the lesion in which Iba1⁺ microglia/macrophages were closely associated with BTK staining (Figure 7B), and it was corroborated through Imaris 3D image rendering (Figure 7C).

Figure 7. Analyses of BTK Expression in MS Autopsy Cases.

(A) Low-magnification image of an MS brain section showing demyelination evaluated through LFB and H&E staining with loss of LFB in the middle. The box in (A) is magnified in (B) as a merged image and its individual panels show staining for nuclei with DAPI, microglia/macrophages with Iba1, immune cells with CD45, and BTK. The colocalization of BTK in Iba1⁺ microglia/macrophages is ascertained by Imaris 3D rendering in (C). BTK = Bruton tyrosine kinase; LFB = Luxol fast blue.

We sought to evaluate the expression of BTK throughout various lesion types of MS. Iba1 staining was substituted for HLA-DR because of inconsistent Iba1 intensity across MS specimens. We stained for tissue sections that contained active, chronic active, and inactive lesions, classified using LFB, H&E, and oil red O staining.⁴⁰ Active and chronic active lesion edges showed hypercellularity of amoeboid HLA-DR⁺ cells, which were often associated with BTK staining that was speckled in appearance (eFigure 7). Chronic active lesion center and inactive lesions presented with few HLA-DR⁺ cells, but these cells also expressed BTK. The NAWM contained HLA-DR⁺ cells that appeared to be more ramified than the ones found within lesions, but few cells were associated with BTK. These qualitative results were observed across the MS specimens. We could not assess pBTK in human specimens because the antibodies produced high background and nonspecific signals in the context of the human autopsy samples.

Overall, our observations showed that BTK is expressed by HLA-DR⁺ microglia and macrophages in MS lesions, which can be observed most abundantly in active lesions and the rim of chronic active lesions with high HLA-DR⁺ density. BTK expression is also found at the center of chronic active lesions and inactive lesions, within remnant HLA-DR⁺ cells that were present.

Discussion

Our collective results demonstrate that BTK immunoreactivity is normally undetectable in uninjured areas or NAWM of human and murine CNSs, but that expression becomes prominent in lesions with hypercellular aggregates of microglia and macrophages. In the inactive core of chronic active MS lesions, we find that the few remaining HLA-DR⁺ myeloid cells are still BTK immunoreactive. Staining for pBTK informs that its upregulation declines in the later stage of lysolecithin and chronic stage of EAE injury while BTK upregulation is maintained. Although this suggests that the activation of BTK is not chronically persistent in MS models despite the elevation of BTK protein, we do not know whether this is the case in MS as the pBTK antibodies that we tested on human specimens proved to be unreliable. Overall, these results provide a comprehensive map of the substantial elevation of BTK in microglia and macrophages in MS models and in active lesions of MS and suggest residual expression of BTK in the few myeloid cells that remain in inactive MS lesions.

On reviewing results presented by others on BTK expression in MS specimens, we noted that 2 groups^14,28 showed overlapping Iba1⁺BTK⁺ cells in the rim of chronic active MS lesions and another²⁷ displayed BTK expression in CD68⁺ cells in active lesions and in the rim of a chronic active lesion. Previously, we provided an example of BTK immunoreactivity on HLA-DR+ microglia/macrophages in an active MS lesion but not in NAWM.¹² Our current analyses across MS cases extend and affirm the substantial elevation of BTK immunoreactivity in MS lesions that contain high-density or remnant microglia/macrophages, and we also provide previously undocumented maps of the expression of BTK and pBTK in the OxPC, lysolecithin, and EAE animal models of MS.

An important question to consider is whether BTK expression in microglia/macrophages in lesions is of pathologic significance, which can be partially elucidated with preclinical studies in mice and in tissue culture. Conditional BTK deletion in CX3CR1⁺ myeloid cells using BTK-floxed mice during the chronic phase of EAE lowers disability scores⁴¹ while the knock-in of a constitutive active form of BTK in myeloid cells promotes microglia proliferation in the murine CNS in vivo.¹³ In tissue culture, microglia activated by immune complexes elevate proinflammatory CD86, CCL3, and TNF-α and microglia activated with lipopolysaccharide upregulate iNOS—both of which can be reduced with the addition of evobrutinib.¹⁴ In tricultures of human astrocytes, neurons, and microglia formed from inducible pluripotent stem cells, the elevation of proinflammatory features of microglia (numerous cytokines, chemokines, and metalloproteinases) and microglia toxicity to neurons (neurofilament release) are lowered by fenebrutinib.¹² Reduced TNF-α and neurofilament release in activated microglia by treatment with fenebrutinib is also observed in human brain organoids.¹² These results support the contention that elevation and activation of BTK in microglia/macrophages are undesirable in MS and should be prevented. A cautionary note regarding the abovementioned tissue culture experiments is that high concentrations of BTK inhibitors, well above their IC₅₀ for BTK in kinase assays, are often used to achieve significant reduction of proinflammatory markers.

The use of BTK inhibitors in mouse models of MS may also reveal the pathophysiologic significance of BTK elevation. Several studies have used a preventative BTK treatment paradigm with drug initiated at the time of immunization or earlier; in this approach, evobrutinib,⁴² tolebrutinib,⁴³ and remibrutinib⁴⁴ prevent or reduce subsequent EAE severity similar to that observed for GDC-0853 in our study. These latter preclinical results are in line with the current available clinical data that fenebrutinib lowers annualized relapse rates and has substantial impact on lowering gadolinium enhancements in a phase 2 trial in relapsing MS,²⁰ which is maintained in the open-label extension phase.⁴⁵

In the therapeutic treatment approach, where drug is initiated at onset or at chronic disability in relapsing-remitting EAE using SJL/J PLP_139-151–immunized mice, evobrutinib reduces the number of MRI-detected meningeal lesions although clinical scores of disability are only marginally lowered (p = 0.03).⁴⁶ Another study using the same relapsing-remitting EAE model found that evobrutinib administered from day 15 after immunization lowers clinical scores in a dose-dependent manner over a 30-day period, where the doses of 3 mg/kg and 10 mg/kg provided significant reduction in disability.¹⁴ These results are similar to our observations with GDC-0853, where there is only modest reduction of disability scores when administered from onset of clinical signs. However, the neuropathologic correlates in our study using microglia, axonal, and myelin markers did not reveal significant benefits for GDC-0853; we attribute this to the relative insensitivity of neuropathologic outcomes in differentiating between treatment groups when the clinical scores were not highly divergent. In addition, in the evobrutinib study mentioned above, when the drug is initiated from day 40 after prolonged disability, the subsequent disease course over 20 days is not improved compared with vehicle.¹⁴ Thus, at least in EAE, the early initiation of drug treatment provides a better outcome than later into the disease course after onset of severe clinical disability.

It is reasonable to consider whether the current results of fenebrutinib in ameliorating EAE severity could be due to an effect on B cells because B cells express BTK^4,5 and are known targets for BTK inhibitors.^7,15 Moreover, there is cross-talk between B cells and microglia; for example, the conditioned media of proinflammatory B cells increase the inflammatory activity of microglia in culture, and vice versa.⁴⁷ However, B cells are not known to be contributors to MOG_35-55–induced EAE disease.⁴⁸ Another consideration is the infiltration of Schwann cells into areas of spinal cord demyelination⁴⁹; it is currently unknown whether Schwann cells express BTK in the demyelinated spinal cord and whether BTK inhibitors may influence this cell population.

While BTK activation is indeed observed in the CNS across all the animal models examined and its inhibition has demonstrated physiologic benefits, caution is needed in translating these findings to clinical practice. These models lack significant cognate antibody response or immune complex presence, which are prominent in MS and can lead to Fc receptor and BTK activation. Therefore, the precise mechanism of BTK activation in these models and how it recapitulates the pathologic setting in MS remain to be elucidated. Indeed, it would be instructive to conduct a mechanistic investigation into the drivers of BTK activation and related signaling pathways in microglia across the different models, a subject of future studies. Another future direction would be to examine BTK expression in other consequential areas of disease activity in MS, such as the choroid plexus and meninges, at different disease stages in the various animal models.

In summary, we describe our comprehensive analyses of BTK and pBTK expression in lesions of MS and its models. The prominent expression of BTK by microglia/macrophages in active and the rim of chronic active lesions where substantial cytotoxicity is manifesting, along with several lines of evidence that heightened BTK activity in myeloid cells drives proinflammatory and potentially neurotoxic consequences, justifies the intense ongoing efforts on clinical trials of BTK inhibitors particularly to reduce the progression of disability in MS.

Glossary

BSA

bovine serum albumin

BTK

Bruton tyrosine kinase

EAE

experimental autoimmune encephalomyelitis

LFB

Luxol fast blue

LPC

lysolecithin

MBP

myelin basic protein

MOG

myelin oligodendrocyte glycoprotein

MS

multiple sclerosis

NAWM

normal-appearing white matter

OxPC

oxidized phosphatidylcholine

PBS

phosphate-buffered saline

PFA

paraformaldehyde

PPMS

primary progressive MS

ROI

region of interest

SPMS

secondary progressive MS

Author Contributions

C. Li: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data. M.T. Morch: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. R. Gorter: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. B. Lozinski: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. S. Ghorbani: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. Y. Dong: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; analysis or interpretation of data. Y.-A. Shen: drafting/revision of the manuscript for content, including medical writing for content. C. Harp: drafting/revision of the manuscript for content, including medical writing for content. S. Zandee: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data. W. Klement: major role in the acquisition of data. A. Prat: major role in the acquisition of data. V.W. Yong: drafting/revision of the manuscript for content, including medical writing for content; major role in the acquisition of data; study concept or design; analysis or interpretation of data.

Study Funding

This research is funded by operating grants from the Canadian Institutes of Health Research and by Genentech to V.W. Yong. C. Li was funded by the Hotchkiss Brain Institute graduate recruitment scholarship, M.T. Morch by an Internationalisation fellowship from the Carlsberg Foundation Denmark, R. Gorter by fellowships from MS Canada and the Gemmy and Mibeth Tichelaar Award from the Dutch MS Society, B. Lozinski by a studentship from MS Canada, and S. Ghorbani and Y. Dong by fellowships from the Canadian Institutes of Health Research.

Disclosure

Y.-A. Shen is an employee of Genentech and a shareholder in F. Hoffmann-La Roche. C. Harp is an employee of Genentech and a shareholder in F. Hoffmann-La Roche. V.W. Yong has received consultant fees from EMD Serono, Novartis, Roche (where Genentech is a subsidiary) and Sanofi, which are companies with proprietary BTK inhibitors being tested or completed in MS. The other authors report no relevant disclosures. Go to Neurology.org/NN for full disclosures.