Chemogenetic activation of locus coeruleus neurons ameliorates the severity of multiple sclerosis

Background Most current disease-modifying therapies approved for multiple sclerosis (MS) are immunomodulatory drugs that counteract the aberrant activity of the immune system. Hence, new pharmacological interventions that drive anti-inflammatory activity and neuroprotection would represent interesting alternative therapeutic approaches or complementary strategies to treat progressive forms of MS. There is evidence of reduced noradrenaline levels and alterations to locus coeruleus (LC) noradrenergic neurons in MS patients, as well as in animal models of this disease, potentially factors contributing to the pathophysiology. Drugs that enhance noradrenaline appear to have some beneficial effects in MS, suggesting their potential to dampen the underlying pathology and disease progression. Methods Therefore, we explored the consequences of chronic LC noradrenergic neurons activation by chemogenetics in experimental autoimmune encephalomyelitis (EAE) mice, the most widely used experimental model of MS. LC activation from the onset or the peak of motor symptoms was explored as two different therapeutic approaches, assessing the motor and non-motor behavioral changes as EAE progresses, and studying demyelination, inflammation and glial activation in the spinal cord and cerebral cortex during the chronic phase of EAE. Results LC activation from the onset of motor symptoms markedly alleviated the motor deficits in EAE mice, as well as their anxiety-like behavior and sickness, in conjunction with reduced demyelination and perivascular infiltration in the spinal cord and glial activation in the spinal cord and prefrontal cortex (PFC). When animals exhibited severe paralysis, LC activation produced a modest alleviation of EAE motor symptoms and it enhanced animal well-being, in association with an improvement of the EAE pathology at the spinal cord and PFC level. Interestingly, the reduced dopamine beta-hydroxylase expression associated with EAE in the spinal cord and PFC was reversed through chemogenetic LC activation. Conclusion Therefore, clear anti-inflammatory and neuroprotective effects were produced by the selective activation of LC noradrenergic neurons in EAE mice, having greater benefits when LC activation commenced earlier. Overall, these data suggest noradrenergic LC neurons may be targets to potentially alleviate some of the motor and non-motor symptoms in MS. Supplementary Information The online version contains supplementary material available at 10.1186/s12974-023-02865-z.


Animals
Female C57BL/6J TH:Cre mice, 10 weeks old, were maintained under standard laboratory conditions (22 ºC, 12 h light/dark cycle, lights on at 08:00 am, food and water ad libitum) at the University of Cadiz.The TH:Cre founders were provided by INFRAFRONTIER/EMMA (EM:00254) [1] (kindly donated by T. Ebendal, Karolinska Institutet, Stockholm, Sweden).All animal handling and the procedures were carried out in accordance with the European Commission guidelines (2010/63/EU) and Spanish Law (RD 53/2013) regulating animal research.Moreover, the protocols were approved by the Ethics Committee for Animal Experimentation at the School of Medicine of the University of Cadiz (Spain).

Experimental design and groups
The experimental designs are shown in Fig. 1A and Fig. 5A.Briefly, 10-week-old mice were administered with Designer Receptor Exclusively Activated by Designer Drugs (DREADD) virus bilaterally into the LC.After a 10 day recovery, EAE was induced and the clinical symptoms were monitored from day 7 post EAE induction (dpi) from 09:00 am to 02:00 pm.Chronic chemogenetic LC activation was achieved by orally gavage of clozapine-N-oxide (CNO, 3 mg/kg daily) from the onset of EAE (~12 dpi) or the peak of the motor symptoms (peak phase: clinical score ≥ 3, ~17 dpi).Finally, the mice were sacrificed in the chronic phase of EAE (from ~20 dpi), collecting the spinal cord and brain tissues (27-29 dpi) for neurobiological assessment.The experimental groups studied were healthy control animals (naïve group) and EAE animals previously administered control-DREADD or rM3D(Gs)-DREADD virus into LC (EAE and EAE-rM3D groups, respectively).All the EAE mice were treated chronically with CNO (from the onset or the peak of the motor symptoms) and the naïve animals received the vehicle alone (drinking water).Mice were assigned to different treatment groups in a randomized manner.The sample size was calculated using the G*Power tool, based on similar published studies.
Experimenters were blind to the treatment.

EAE induction and evaluation
After DREADD administration, mice (12-weeks-old) were immunized subcutaneously in the upper and lower back areas with an emulsion containing 200 μg of mouse myelin oligodendrocyte glycoprotein peptide (MOG35-55) in complete Freund's adjuvant (CFA).The mice were also administered 100 ng of pertussis toxin in PBS intraperitoneally (PTX) 2 and 24 hours later (#EK-2110, Hooke Laboratories Inc, USA).Immunization was carried out under isoflurane anesthesia and the control animals were subjected to the same procedure but administered PBS alone.As soon as the first signs of paralysis appeared, the mice were provided with easily accessible water, wet food pellets and jelly with sucrose on the floor of the cage to ensure proper nutrition and hydration, even to animals with severe paralysis.
Clinical signs: All mice were monitored daily for clinical signs of EAE and every two days their body weight was controlled blind to the treatment.The mice were scored using the following scale (clinical score): 0, no clinical signs; 0.5, partially limp tail; 1, limp tail; 1.5, limp tail and imbalance; 2, limp tail and moderate hind limb weakness; 2.5, limp tail and partial paralysis in at least one hind limb; 3, limp tail and complete paralysis of both hind limbs; 3.5, limp tail, complete hind limb paralysis and weakness in the forelimbs (able to right itself when placed on its side); 4, limp tail, complete hind limb paralysis and weakness in the forelimbs (unable to right itself when placed on its side); 4.5, limp tail and paralysis in all the limbs (quadriplegia, end-point criteria).The area under the curve (AUC) values of the clinical score, the maximum score achieved and the relative change in clinical score at the peak (17 dpi) and in the chronic (25 dpi) phase of EAE (relativized to EAE group at the peak phase) were also analyzed.
Open field test: Mice were placed individually in a square arena (45 × 45 × 35 cm) that they were allowed to explore freely for 5 min.The sessions were videotaped and analyzed offline, using SMART (Spontaneous Motor Activity Recording and Tracking) video 3.0 software (Panlab, Spain).Spontaneous locomotor activity was measured as the total distance travelled, expressed in arbitrary units (AU) and through the AUC values of the total distance travelled.The time spent in the central area of the arena was also measured as a readout of anxiety-like behavior.Animals were placed in the center of the arena and those that failed to move out of the central region were not considered for the trial.Finally, the activity/attention score was evaluated over the last 4 min of free exploration period.This score was related to the "sickness" of the mice and it was assessed through the time spent in a specific sedentary posture that is characterized by the placement of both forepaws on the ground and maintaining the head relatively still with a steady gaze directed below horizontal [3].The activity/attention score was based on the following criteria regarding the time spent in this position in each minute of the 4 minutes test period: 0 (45-60 s), 1 (30-45 s), 2 (15-30 s) and 3 (0-15 s).

Histology, immunohistochemistry and immunofluorescence assays
Animals were perfused with paraformaldehyde (4%) at the end of the experiments, and their brain and spinal cord were collected, post-fixed (2 hours) and preserved at 4°C in sucrose (30% in 0.1 M phosphate-buffered saline, PBS).Coronal sections were obtained on a Sliding Microtome (Microm HM 450: Fisher Scientific SL, Spain) coupled to BFS-MP freezing stage for microtome (Physitemp Instruments, USA), that contained the lumbar spinal cord (50 µm, 1 in every 5 sequential sections), the prefrontal cortex (PFC): prelimbic (PL) and infralimbic (IL), motor cortices: secondary (M2) and primary (M1) and LC, A5 and A7 (40 µm, 1 in every 6 sequential sections).In addition, coronal sections obtained on a Microm HM 525 cryostat (Thermo-Scientific, Germany) and containing 1 cm of the thoracic spinal cord region (20 µm, 1 in every 4 sequential sections) were also collected on glass slides.For immunohistochemistry and immunofluorescence assays, detailed information about the primary and secondary antibodies used are included in Table S1, and the assays were quantified using Fiji Image software (USA).
Noradrenergic cells and astrocytes in the LC: Sections containing the LC, from other set of naïve and EAE animals (27-29 dpi), were incubated for 2 nights at 4ºC with a mouse antibody against glial acidic fibrillary protein (GFAP, 1:1000; Millipore, Spain) and a sheep antibody against tyrosine hydroxylase (TH, 1:1000, Abcam, Spain).The binding of the antibodies was then detected with an Alexa Fluor 568 conjugated donkey anti-mouse (1:1000; Invitrogen TM , USA) and an Alexa Fluor 647 conjugated donkey anti-sheep (1:1000; Invitrogen TM , USA) antibodies.
After washing and mounting in fluoro-gel aqueous medium, images were acquired on a Zeiss LSM 880 Confocal microscope using FAST Airyscan (Carl Zeiss Microscopy GmbH, Germany), and the number of TH + cells in the LC per field and the area occupied by GFAP (%) were assessed.
Perivascular infiltration in the spinal cord: Perivascular infiltration of the CNS parenchyma can be analyzed by hematoxylin-eosin staining.Thus, sections of the thoracic spinal cord were collected on glass slides (5 animals per group) and stained with hematoxylin-eosin following standard procedures.The images were visualized on an Olympus BX60 microscope and acquired with an Olympus DP74 camera (Spain).Perivascular infiltration in spinal cord sections was scored using the following scale: 0, no infiltration; 1, a few scattered inflammatory cells; 2, perivascular infiltrates; 3, perivascular extension into the parenchyma; and 4, extensive cell infiltration in the white matter [4].Moreover, perivascular cuffs were counted in the spinal cord discriminating the number of those in which infiltrates did not penetrate into the parenchyma (preclinical cuffs) [5,6].

Demyelination in the spinal cord and cortex:
In order to evaluate the demyelination in the lumbar spinal cord white matter and cortex, sections (4-5 animals per group) were subjected to antigen retrieval to unmask the epitopes.Thus, prior to probing with the antibodies, the sections were treated with a 2N HCl solution at 37°C (20 min) and then left at room temperature for 5 min.The sections were then probed for 2 nights at 4ºC with a mouse antibody against myelin basic protein (MBP, 1:1000; Biolegend, USA).Antibody binding was then detected with a biotinylated donkey anti-mouse IgG (1:200; Jackson ImmunoResearch Europe, UK) followed by Alexa Fluor 568 streptavidin (1:1000; Invitrogen TM , USA).After washing and mounting in fluoro-gel aqueous medium, images were acquired on a fluorescent Olympus BX60 microscope equipped with an Olympus DP74 camera (Spain).The absence of MBP with respect to the total area was assessed in the spinal cord white matter and expressed as a percentage.Meanwhile, MBP immunoreactivity in the PL/IL and M2/M1 was quantified in a region of interest defined in each structure by obtaining the mean intensity after subtracting the background noise from each section (expressed in arbitrary units, AU).

Astrocyte activation in the spinal cord and cortex:
As a marker of astrocytes, GFAP immunoreactivity was assessed in the dorsal and ventral horns of the lumbar spinal cord and cortex (4-5 animals per group).Antigen retrieval was first performed to unmask the epitopes, incubating sections in 2N HCl at 37°C (20 min) and then accommodating them for 5 min at room temperature.The sections were then probed for 2 nights at 4ºC with a rabbit antibody against GFAP (1:1000; Dako, USA), the binding of which was detected with a biotinylated donkey antirabbit antibody (1:200; Jackson ImmunoResearch Europe, UK), and visualized with an ultrasensitive ABC peroxidase staining kit (1:1000; Thermo Scientific, Spain) and 3,3'diaminobenzidine tetrahydrochloride (DAB) [7].The sections were then mounted on slides, cleared in xylene and coverslipped with DPX, and the images were acquired at the same exposure and illumination settings on an Olympus BX60 microscope equipped with an Olympus DP74 camera.The optical density of GFAP expression was quantified in the grey matter of the dorsal and ventral horns of the spinal cord, and in the PL/IL and M2/M1 cortices, defining a region of interest in each structure and obtaining the mean intensity by subtracting the background noise from each section (expressed in AU).

Microglial activation in the spinal cord and cortex:
To evaluate microglial activation, we analyzed the expression of the ionized calcium-binding adapter molecule 1 (Iba1), and that of arginase-1 (Arg1) and the inducible isoform of nitric oxide synthase (iNOS) in these Iba1 + cells.Sections (3-5 animals per group) containing the lumbar spinal cord and cortex were probed for 2 nights at 4ºC with a rabbit (1:1000, Wako, USA) or goat antibodies (1:500, Abcam, UK) against Iba1, combined with a mouse antibody against Arg1 (1:500, BD Biosciences, USA) or a rabbit antibody against iNOS (1:500, Abcam, UK).The distribution of Arg1 and iNOS was visualized with a biotinylated donkey anti-mouse or anti-rabbit IgG (1:200; Jackson ImmunoResearch Europe, UK) in conjunction with Alexa Fluor 488 streptavidin (1:1000; Invitrogen TM , USA).Iba1 was detected with an Alexa Fluor 568 conjugated donkey anti-rabbit or anti-goat antibody

Noradrenergic projections in the spinal cord and cortex:
To evaluate the DBH fibers in the dorsal and ventral horns of the lumbar spinal cord and cortex, sections (4-5 animals per group) were probed for 2 nights at 4ºC with a rabbit antibody against DBH (1:500, Abcam, UK).Subsequently, the sections were incubated with biotinylated donkey anti-rabbit antibody (1:200; Jackson ImmunoResearch Europe, UK), and visualized with an ultra-sensitive ABC peroxidase staining kit (1:1000; Thermo Scientific, Spain) and DAB [7].The sections were then mounted on slides, cleared in xylene and coverslipped with DPX, and the images were acquired at the same exposure and illumination settings on an Olympus BX60 microscope equipped with an Olympus DP74 camera.The optical density of DBH expression was calculated in the grey matter of the dorsal and ventral horns of the spinal cord, and the PL/IL and M2/M1 cortices, defining a region of interest in each region and obtaining the mean intensity by subtracting the background noise from each section (expressed in AU).

DREADD virus expression:
To verify the mCherry expression in the LC, and the A5 and A7 noradrenergic areas, sections containing these regions were probed for 2 nights at 4ºC with a rat anti-red fluorescent protein (RFP, 1:500; Chromotek, Germany) and a rabbit anti-DBH (1:500; Abcam, UK) antibodies.The binding of these antibodies was detected with a biotinylated donkey anti-rat antibody (1:200; Jackson ImmunoResearch Europe, UK), which was visualized with Alexa Fluor 568 streptavidin or an Alexa Fluor 488 conjugated donkey anti-rabbit antibody (1:1000; Invitrogen TM , USA).Nuclear counterstaining with DAPI (#D9542-10MG, 1:5000, Sigma-Aldrich, USA) was performed and sections were then washed and coverslipped in fluorogel aqueous mounting medium and the images were acquired on a Zeiss LSM 880 Confocal microscope using FAST Airyscan (Carl Zeiss Microscopy GmbH, Germany).The relative mCherry expression in DBH + neurons of the LC was quantified and expressed in percentage.The selective expression of DREADD was also assessed in the A5 and A7 noradrenergic nuclei [7].

Statistical analysis
The data are represented as the means + standard error of the mean (SEM) and the results were analyzed using GraphPad Prism version 9.4.0 (GraphPad Software, USA), applying a Student's t-test (unpaired, two-tailed), or a one or two-way analysis of variance (ANOVA), with or without repeated measures (RM), followed by a Bonferroni or Dunnett's post hoc test.For RM ANOVA Geisser-Greenhouse correction was applied.For non-parametric data, a Mann-Whitney U or Kruskal-Wallis test was used, followed by Dunn's post hoc test.The differences were considered significant at p<0.05 (Tables S2-11).Normality was assessed using the Shapiro-Wilk test and statistical outliers were identified using Grubbs' test.Correlations were determined using a Pearson's or the non-parametric Spearman correlation coefficients.S11).Pearson or Spearman correlation analysis.DBH, dopamine beta-hydroxylase; GFAP, glial fibrillary acidic protein; IL, infralimbic cortex; M1, primary motor cortex; M2, secondary motor cortex; PL, prelimbic cortex

( 1 :
1000; Invitrogen TM , USA).For the quantification of the relative area occupied by Iba1 in the spinal cord and cortex, Alexa Fluor 488 conjugated donkey anti-rabbit (1:1000; Invitrogen TM , USA) was used.After washing and mounting in fluoro-gel aqueous medium, the Iba1 and Arg1/Iba1 images were acquired on a fluorescent Olympus BX60 microscope equipped with an Olympus DP74 camera (Spain), while iNOS/Iba1 images were acquired on a Zeiss LSM 880Confocal microscope (Carl Zeiss Microscopy GmbH, Germany).The relative area occupied by Iba1 in lumbar spinal cord white matter and in the cortex (PL/IL, M2/M1) was quantified.Arg1/Iba1 expression was analyzed as a percentage of the relative Arg1/Iba1 co-localization with respect to total Iba1 area in lumbar spinal cord.The presence of iNOS expression was also evaluated in lumbar spinal cord.

Fig. S2
Fig. S2 Assessment of noradrenergic cells and astrocyte status in the LC.Representative images of TH + neurons (green) and GFAP (red) staining in the LC of (A) naïve and (B) EAE animals.(C) Quantification of the number of TH + neurons and (D) the relative area occupied by GFAP in this nucleus.The data represent the mean + SEM and each point corresponds to an individual mouse (n=4-5; TableS9).Scale bars: 20 µm.GFAP, glial fibrillary acidic protein; TH, tyrosine hydroxylase

Table S3 . Summary of statistical analysis of Fig. 2 Shapiro-Wilk normality test
(W, P value, Passed normality test?)

Table S4 . Summary of statistical analysis of Fig. 3 Shapiro-Wilk normality test
(W, P value, Passed normality test?)

Table S5 . Summary of statistical analysis of Fig. 4 Shapiro-Wilk normality test
(W, P value, Passed normality test?) for DBH

Table S6 . Summary of statistical analysis of Fig. 5 Shapiro-Wilk normality test
(W, P value, Passed normality test?)

Table S7 . Summary of statistical analysis of Fig. 6 Shapiro-Wilk normality test
(W, P value, Passed normality test?)

Table S8 . Summary of statistical analysis of Fig. 7 Shapiro-Wilk normality test
(W, P value, Passed normality test?)

Table S10 . Summary of statistical analysis of Fig. S4
Shapiro-Wilk normality test (W, P value, Passed normality test?) for MBP

Table S11 . Summary of statistical analysis of Fig. S5 Shapiro-Wilk normality test
(W, P value, Passed normality test?) for GFAP and DBH expression