The novel HS-mimetic, Tet-29, regulates immune cell trafficking across barriers of the CNS during inflammation

Tet-29 (Fig. 1), BODIPY-labelled Tet-29 (BDP-Tet-29) and unsulfated BDP-Tet-29 were provided by Dr Olga Zubkova and Sam Spijkers-Shaw from the Ferrier Research Institute (Victoria University of Wellington (VUW), Wellington, NZ). PI-88 (Additional file 1: Fig S1) was kindly donated by Professor Chris Parish (John Curtin School of Medical Research, Canberra, Australia). The anti-mouse/human VLA-4 (PS/2; CD49d) and rat IgG2b anti-keyhole limpet hemocyanin (LTF-2) antibodies were purchased from Bio X Cell (Lebanon, NH, USA). All compounds were reconstituted in sterile phosphate-buffered saline (PBS; i.e. vehicle) and administered to mice via intraperitoneal (i.p.) injection.

Female C57Bl/6J mice were bred and held in a temperature- and humidity-controlled environment at the VUW Animal Research Facility or the Biomedical Research Unit at the Malaghan Institute of Medical Research (Wellington, NZ). All housing and experimental procedures were approved by the VUW Animal Ethics Committee under the approvals AEC25295 and AEC29122.

EAE was induced in 8–12-week old mice. Briefly, mice were immunised with 50 µg of MOG_35–55 (Genscript, Piscataway, NJ, USA) in complete Freund’s adjuvant (Sigma, St. Louis, MO, USA) containing 500 µg/mouse heat-killed Mycobacterium (Difco Laboratories, Franklin Lakes, NJ, USA) by subcutaneous (s.c.) injection in the rear flanks. Concurrently, mice received a 200 ng dose of pertussis toxin (PTX; List Labs, Campbell, CA, USA) by i.p. injection. Two days later, mice were treated with a second 200 ng PTX dose. Following immunisation, mice were weighed and scored daily by a single unblinded observer. EAE severity was scored as follows; 0 = not affected, 0.5 = tail weakness or distal tail paralysis, 1 = significant tail weakness or half tail paralysis, 2 = full tail paralysis, 3 = one hind limb paralysis or severe weakness in both hindlimbs, 4 = full hindlimb paralysis, 5 = moribund. Mice were considered recovered from disease when they reached a score of ≤ 0.05 after their initial peak in disease. Relapse was defined as an increase in disease score by ≥ 1 for ≥ 2 days after the initial peak in disease.

Tissues were collected after CO₂ asphyxiation and processed into single-cell suspensions. Animals were perfused with 20 mL of PBS before brains and spinal cords were collected. Spinal cords were minced with a scalpel then digested in a 2.4 mg/mL type II collagenase solution (Gibco, Thermo Fisher Scientific, Waltham, MA, USA) for 30 min at 37 °C under constant agitation. Both brains and digested spinal cords were processed into single-cell suspensions by forced passage through a 70 µm cell strainer. CNS tissues were centrifuged for 5 min at 760 × g before resuspension in a 37% Percoll (Sigma-Aldrich) solution. A Percoll gradient was created by centrifuging samples for 30 min at a low acceleration and no deceleration. Myelin was removed from the top of the gradient before cells were washed, pelleted by centrifugation, and resuspended in flow cytometry staining (FACs) buffer containing 2% foetal calf serum (FCS, Gibco) and 0.1% sodium azide (1 M).

Blood was collected in ethylenediaminetetraacetic acid (EDTA)-coated tubes to prevent coagulation. Samples were incubated in red blood cell lysis buffer (Sigma-Aldrich) for 2 min before being rinsed with wash buffer and centrifuged for 5 min at 760 × g. Spleens were collected and mashed through a 70 µm cell strainer before the cell lysis step was carried out. After satisfactory lysis, splenocytes and blood cells were rinsed, centrifuged, and resuspended in FACs buffer.

Single-cell suspensions in FACs buffer were incubated in Fc block (2.4G2, BD Biosciences, Franklin Lakes, NJ, USA) for 15 min then samples were washed in FACs buffer and centrifuged at 400 × g for 4 min. Cells were then incubated in a solution containing fluorescently labelled antibodies for 30 min in the dark on ice. Following staining, flow cytometry was performed on a BD FACS Canto II (BD Biosciences) or a LSRFortessa™ (BD Biosciences). Analysis was performed on FlowJo software (BD Biosciences) using the gating strategy outlined in Additional file 2: Fig S2.

The following antibodies were used for the detection and phenotyping of immune cells: CD45-BV510 (30-F11; Biolegend, San Diego, CA, USA), CD3-APC (17A2; Biolegend), B220-APC-Cy7 (RA3-6B2; Biolegend), CD4-BV421 (RM4-5; Biolegend), CD8a-PerCP-Cy5.5 (53–6.7; Biolegend), CD62L-FITC (MEL-14; Biolegend), CD44-PE (IM7; Biolegend), CD11b-PE-Cy7 (M1/70; Biolegend), Ly6G-APC (1A8; Biolegend), Ly6C-PE (HK1.4; Biolegend), and CD49d-AF488 (R1-2; Biolegend).

Brains were fixed in 4% paraformaldehyde then cryoprotected in a 30% sucrose (Sigma-Aldrich) solution for 48 h. Samples were snap frozen in PolyFreeze (Sigma-Aldrich) and cryo-sectioned into 20 µm thick sections on a Leica CM3050 cryotome (Leica Microsystem, Wetzlar, Germany). Sections were heated to 70 °C in Tris–EDTA (Sigma-Aldrich) buffer for 10 min and then incubated in 1 mg/mL sodium borohydride (Sigma-Aldrich) quench for 30 min. Following quenching, sections were blocked using 5% donkey serum (Sigma-Aldrich) in PBS with 0.05% Tween20 (Sigma-Aldrich) for 2 h at room temperature. Primary antibodies: rabbit anti-mouse CD31 (1:200, cat. # ab182981, Abcam, Cambridge, UK), goat anti-mouse Albumin (1:200, cat. # ab19194, Abcam), rat anti-mouse VCAM-1/CD106 (1:200, cat. # 105702, Biolegend), rat anti-mouse ICAM-1 (1:200, cat. # 116102, Biolegend), and rabbit anti-mouse Collagen IV (1:200, cat. # ab19808, Abcam) were added to blocking buffer (PBST and 5% donkey serum; Sigma-Aldrich) and then applied to sections overnight at 4 °C. Slides were then washed in PBST and secondary antibodies: anti-goat Alexa Fluor 488 (1:1000, cat. # ab150129, Abcam), anti-rabbit Alexa Fluor 647 (1:800, cat. # ab150075, Abcam), anti-rat Alexa Fluor 568 (1:800, cat. # ab175475, Abcam) were added to PBS and incubated for 2 h at room temperature. After washing off the secondary antibodies with PBS-Tween20, ethanol washed coverslips were mounted using anti-fade glycerol DAPI containing mounting medium (Abcam).

Images were obtained using an Olympus FV3000 laser scanning confocal microscope using the 20× and 40× air objectives (Olympus, Lower Hutt, New Zealand) by z-series imaging with a 0.60 µm step size. Images were processed using ImageJ [22] features “Remove Bright Outliers” and Median filter. Cell Profiler [23] was used to obtain measurements of area using the following modules: “ColourToGray”, “Threshold”, and “MeasureImageAreaOccupied.”

The transformed C57BL/6 Mus musculus epithelial choroid plexus cell (ECPC-4) cell line (RRID: CVCL_4836) was provided by the RIKEN Bioresource Centre through the National BioResource Project of the AMED, Japan. ECPC-4 cells were cultured in Dulbecco’s modified Eagle medium (DMEM) supplemented with 10% FCS, 100 U/mL penicillin and 100 µg/mL streptomycin (all from Gibco) at 37 °C/5% CO₂.

The murine cerebral brain endothelioma bEnd.3 cell line was obtained from the American Type Tissue Culture Collections (ATCC^® CRL™-2299™, Manassas, VA, USA). bEnd.3 cells were cultured in DMEM (ATCC 30-2002) supplemented with 10% FCS, 2 mM l-glutamine, 100 U/mL penicillin and 100 µg/mL streptomycin at 37 °C/5% CO₂.

Splenocytes were isolated and seeded at 1 × 10⁶ cells/well in growth medium (DMEM plus 10% FCS, 100 U/mL penicillin, 100 µg/mL streptomycin, 10 mM HEPES, 2 mM l-glutamine, 50 μM 2-mercaptoethanol, and non-essential amino acids; Gibco) in a 96-well plate and incubated for 24 h at 37 °C with 1 µg/mL Concanavalin A (ConA, Sigma-Aldrich) or growth media alone. Treatment with Tet-29 was carried out in both conditions by supplementing the media with 60 µg/mL Tet-29.

bEnd.3 and ECPC-4 cells were seeded at 5 × 10⁴ cells per insert (CLS3422 for bEnd.3 and CLS3415 for ECPC-4; Coring Costar, Kennebunk, ME, USA) in growth media on the collagen-coated membranes. ECPC-4 cells were seeded on the basolateral side of the membrane and left to adhere for 12 h in the inverted position, and then reoriented in wells containing growth media. bEnd.3 cells were seeded on the apical side of the insert in wells containing growth media. Both cultures were left to form a confluent monolayer by incubating at 37 °C/5% CO₂ for 48 h.

Growth media from the upper chamber of the transwell inserts was aspirated, and 1 × 10⁶ splenocytes were added to the upper chamber of confluent ECPC-4 or bEnd.3 transwell models. For Tet-29 treated conditions, monolayers were treated with 60 μg/mL Tet-29 for 24 h prior to the addition of splenocytes, and further treated with Tet-29 upon addition of splenocytes. Cultures were incubated at 37 °C/5% CO₂ for 24 h. To assess the extent of migration across the monolayer, the upper and lower fractions were collected and prepared for flow cytometry. A migration index for each sample was calculated by dividing the number of migrating cells in the sample by the combined number of sample migrating cells plus the number of migrating cells in an unstimulated vehicle control.

Statistical analysis was carried out using GraphPad Prism version 9.3.1 (GraphPad Software Inc., La Jolla, CA, USA). For parametric datasets, differences between means were established using unpaired Student’s t test, one-way, or two-way ANOVA, as indicated in figure captions. Multiple comparisons testing was carried out using the recommended post hoc test. All data are displayed as mean ± SEM, and differences between groups were considered statistically significant when the p value was ≤ 0.05.

First, we investigated the disease-modifying effect of Tet-29 delivered before the onset of disease as the compound was expected to inhibit the initial influx of immune cells into the CNS during the onset of EAE. A daily dose of 30 mg/kg/day (i.e. 600 μg/mouse/day) was used based upon a previous cancer therapy study from our group [20]. Vehicle (i.e. PBS)-treated EAE mice exhibited severe and unresolving disease whereas mice treated with daily Tet-29, from day 5 post EAE induction, had a significantly lower average disease score over the course of treatment (Fig. 2a). Additionally, Tet-29 treatment caused a significant reduction in disease burden as assessed by area under the curve (AUC; Fig. 2b), peak disease scores (Fig. 2c) and disease incidence (24/30 [80%] Veh compared to 13/24 [54%] Tet-29; * < 0.05 by Fisher’s exact test). Flow cytometry was used to investigate the impact of Tet-29 on the infiltration of lymphocytes and inflammatory monocytes into the CNS as these immune cells are commonly associated with early disease in EAE [24]. To compare cellular infiltration into the CNS between animals, cell counts in the brain and spinal cord were normalised to microglia as microglia counts did not vary between disease or treatment groups (Additional file 2: Fig S2b). Across both CNS compartments, EAE induced a significant increase in the infiltration of T cells, B cells and Ly6C^high monocytes (Fig. 2d–k). In contrast, Tet-29 significantly inhibited the infiltration of CD4⁺ and CD8⁺ T cells into the spinal cord (Fig. 2d, e) and brain (Fig. 2h, i). Tet-29 additionally ameliorated invasion of B cells (Fig. 2j) and inflammatory monocytes (Fig. 2k) into the brain. These effects were not reflected in the periphery as Tet-29 did not alter immune cell counts in the spleen (Fig. 2l–o), although both EAE groups did have slightly reduced CD8⁺ T cell and B cells counts (Fig. 2m, n). Together, these data demonstrates that Tet-29 specifically inhibits CNS infiltration of immune cells in EAE, supporting the amelioration of disease.

Most previous publications on HS-mimetics in EAE report on their use as prophylactic treatments [17] since rescue of disease-induced BBB breakdown is not expected to produce a significant disease-modifying effect after the initial immune cell infiltration. We investigated a more clinically relevant regimen that initiated Tet-29 treatment after disease onset to determine whether therapeutic delivery of an HS-mimetic could reverse CNS infiltration during established disease. Both Tet-29- and vehicle-treated animals followed a similar course of disease for the first 14 days of treatment (Fig. 3a). Like prophylactic treatment, a significant interaction between treatment and time emerged, and Tet-29 significantly reduced the average daily disease scores in the last 10 days of treatment and reduced disease burden as assessed by AUC (Fig. 3b). As therapeutic treatment experiments were carried out over a longer period, the number of animals that recovered from paralysis or experienced a relapse in disease could be evaluated. Tet-29 enhanced disease recovery—both the number (11/29 Tet-29 compared to 4/28 Vehicle; p = 0.07 by Fisher’s exact test) and duration of recovery (Fig. 3c, left) while it significantly decreased the number (4/29 Tet-29 compared to 12/28 Vehicle; p < 0.05) and duration of relapses (Fig. 3c, right). As expected, EAE induced a significant increase in the infiltration of CD4⁺ and CD8+ T cells, as well as their respective CD62L^-CD44⁺ effector memory subsets, into the spinal cord and brain (Fig. 3d–g; Additional file 3: Fig S3a). Tet-29 treatment did not significantly reduce this EAE-mediated infiltration. However, accumulation of CD4⁺ and CD8⁺ effector memory T cells into the spinal cord of EAE mice treated with Tet-29 appeared to be reduced although the effect did not reach statistical significance. This trend was not observed in the brain (Additional file 4: Fig S4a). Like prophylactic treatment, the effects of Tet-29 on CNS infiltration were not associated with changes in immune cell subsets in the periphery (Additional file 3: Fig S3b). Overall, Tet-29 treatment initiated after established disease could blunt, but not reverse, EAE-induced T cell infiltration into the spinal cord, resulting in a significant reduction in disease symptoms.

To identify whether Tet-29 reduces CNS infiltration in EAE by modulating BBB integrity, the brain vasculature was investigated by confocal microscopy. Albumin staining within the CNS parenchyma is an established method of quantifying long-term changes in blood vessel permeability [25, 26]. In healthy mice, staining for albumin was contained within blood vessels (i.e. co-localisation of albumin and CD31 expression), indicating an intact BBB (Fig. 4b). EAE induced a significant increase in albumin staining outside of blood vessels, as quantified by a ratio of albumin/CD31 coverage, indicating a significant loss of blood vessel integrity. Therapeutic treatment with Tet-29 treatment reversed disease-induced blood vessel permeabilisation, evidenced by a reduction in albumin leakage comparable to healthy controls (Fig. 4a). A similar effect of Tet-29 on BBB integrity was validated by analysing the co-localisation of albumin and Collagen IV expression (Additional file 4: Fig S4a-b).

We next assessed the impact of Tet-29 treatment on the expression of the adhesion molecules VCAM-1 and ICAM-1. Under steady-state conditions, the brain endothelium significantly restricts expression of cell adhesion molecules (CAMs) to regulate cell migration into the parenchyma, whereas upregulation of CAM expression promotes migration across inflamed endothelium [27]. Indeed, we observed a significant increase in VCAM-1 (Fig. 4c, d) and ICAM-1 (Fig. 4e, f) expression in the cerebellum due to EAE, which was reversed in mice treated therapeutically with Tet-29 (Fig. 4c, e). Additionally, when assessing all animals, the expression of VCAM-1 but not ICAM-1 was significantly correlated to the final disease score (p < 0.01 by Sprearman’s r; Additional file 4: Fig S4c). Taken together, analysis of changes at the BBB found that Tet-29 reverses EAE-induced blood vessel permeability and expression of CAMs, suggesting that therapeutic treatment with Tet-29 returns the BBB to a healthy-like phenotype.

By reducing the expression of CAMs on brain endothelium, Tet-29 treatment altered the ability of immune cells to interact with the BBB. Next, we investigated whether Tet-29 interacted with immune cells directly, and if this interaction was via the same mechanism as endogenous HS. While Tet-29 is known to bind directly to HPSE [20], it is also likely that Tet-29, as a HS-mimetic, can bind cell surface proteoglycans and other proteins as does endogenous HS [28]. To determine the binding capacity and specificity of Tet-29 in vivo, we administered BODIPY (BDP)-Tet-29, an unsulfated BDP-Tet-29, or vehicle to healthy mice and mice with chronic EAE. Because the sulfate groups on endogenous HS molecules are known to be crucial for its binding capacity and diversity [28], the non-sulfated BDP-Tet-29 served as a negative control. We found that BDP-Tet-29 was able to bind multiple immune cell populations including T cells, monocytes, and neutrophils in the blood of EAE mice (Fig. 5a–d), and BDP-positive cells were also detected in the spleen of healthy and EAE mice (Fig. 5e–h). Unsulfated BDP-Tet-29 could not be detected on immune cells above the level of control animals that received no mimetic (Fig. 5i). These findings indicate that Tet-29 has a strong binding affinity for immune cells in vivo, and this interaction is dependent on its sulfate groups.

To investigate the barrier specific effects of Tet-29 on immune cell migration, we utilised an in vitro monolayer model of the BBB (inflammatory migration) and the choroid plexus (ChP; inflammatory and homeostatic trafficking.) Across both the ChP and BBB models, ConA pre-stimulation induced a significant increase in the number of CD4⁺ and CD8⁺T cells trafficking into the “CNS” compartment although CD8⁺ trafficking across the bEnd.3 monolayer was unaffected (Fig. 6a–d). This effect was more prominent across the epithelial versus endothelial barrier, which is consistent with the in vivo characteristics of these barriers, whereby the ChP is more permeable to immune cell diapedesis than the BBB [29]. An increase in T cell trafficking is characteristic of MS, in which relapses are predominantly mediated by aberrant and unregulated pro-inflammatory T cell trafficking into the neural parenchyma [30]. In our endothelial model of the BBB, Tet-29 abolished ConA mediated trafficking of CD4⁺ (Fig. 6a), whereas Tet-29 significantly reduced inflammatory trafficking of both CD4⁺ and CD8⁺ T cells across the epithelial ChP model (Fig. 6c, d). Interestingly, Tet-29 did not significantly alter trafficking across either barrier in unstimulated conditions. Together, these results suggest that Tet-29 inhibits inflammatory but not homeostatic T cell trafficking into the CNS across the ChP and BBB.

To assess the impact of Tet-29 on homeostatic CNS trafficking, healthy mice were treated with daily Tet-29 (30 mg/kg), PI-88 (10 mg/kg), or vehicle, or every 4 days with an anti-VLA-4 monoclonal antibody (αVLA-4; PS/2; 5 mg/kg) for a minimum of 10 days. PI-88 (Muparfostat; Additional file 1: Fig S1) is an HS-mimetic developed as an angiogenesis and tumour metastasis inhibitor [16]. Like Tet-29, PI-88 exhibits a beneficial effect in EAE and demyelinating models [14, 31]. The 10 mg/kg/day dose used in this study has previously been used in a chronic autoimmune model of diabetes in mice, where it significantly delayed the onset of disease [12]. The effect of Tet-29 was also compared to αVLA-4, a murine version of natalizumab, as this is a known inhibitor of homeostatic trafficking into the CNS of humans [32]. By binding VLA-4 (Additional file 5: Fig S5a-b), αVLA-4 prevents the formation of an adhesion complex with VCAM-1 and effectively reduces leukocyte accumulation in the CNS [33]. The 5 mg/kg dose every 4 days was selected based on previous literature reporting a beneficial effect of αVLA-4 treatment in EAE when treatment is initiated before the onset of disease [34, 35].

Mice treated with αVLA-4 exhibited significantly lower numbers of CD8⁺ T cells and Ly6C^high monocytes in the spinal cord and brain compared to Tet-29-treated animals while CD4⁺ T cell trafficking was reduced in the brain but unchanged in the spinal cord (Fig. 7a–f). Compared to vehicle, αVLA-4 also significantly lowed the number of CD8+ T cells and Ly6C^high monocytes in the spinal cord and CD8⁺ and CD4⁺ T cells in the brain, and hese findings are consistent with previous reports [36, 37]. Neither the Tet-29 nor the PI-88 treatment groups deviated significantly from vehicle alone when the individual cell types were compared. However, we found that Tet-29 treatment had a significant overall affect with increased CNS infiltration in the brain and spinal cord of Tet-29-treated animals compared to vehicle, αVLA-4, and PI-88 (2-way ANOVA; Fig. 7). In comparison to EAE-induced migration where there was an increase in immune cells (e.g. 5.8-fold for CD8⁺ T cells in the brain; Fig. 2i), Tet-29 treatment in healthy animals increased homeostatic trafficking compared to vehicle (e.g. 1.38 fold for CD8⁺ T cells in the brain; Fig. 7e). Interestingly, the effect of Tet-29 on homeostatic migration appeared specific, as PI-88 did not follow the same trend and was significantly different from Tet-29 in the brain and spinal cord (Fig. 7). As expected, the impact of these treatments on CNS invasion was not reflected in the periphery as no treatment group exhibited significant differences in cell frequencies in the spleen (Fig. 7g–i). Taken together, although the effects on homeostatic CNS trafficking are modest, these data indicate that Tet-29 may be less disruptive to this trafficking pathway than αVLA-4 and is distinctly different from PI-88.