Journal of Neuroimmune Pharmacology

, Volume 5, Issue 2, pp 252–259

Increasing CNS Noradrenaline Reduces EAE Severity

  • Maria Vittoria Simonini
  • Paul E. Polak
  • Anthony Sharp
  • Susan McGuire
  • Elena Galea
  • Douglas L. Feinstein
Original Article

DOI: 10.1007/s11481-009-9182-2

Cite this article as:
Simonini, M.V., Polak, P.E., Sharp, A. et al. J Neuroimmune Pharmacol (2010) 5: 252. doi:10.1007/s11481-009-9182-2


The endogenous neurotransmitter noradrenaline (NA) is known to exert potent anti-inflammatory effects in glial cells, as well as provide neuroprotection against excitatory and inflammatory stimuli. These properties raise the possibility that increasing levels of NA in the central nervous system (CNS) could provide benefit in neurological diseases and conditions containing an inflammatory component. In the current study, we tested this possibility by examining the consequences of selectively modulating CNS NA levels on the development of clinical signs in experimental autoimmune encephalomyelitis (EAE). In mice immunized with myelin oligodendrocyte glycoprotein peptide to develop a chronic disease, pretreatment to selectively deplete CNS NA levels exacerbated clinical scores. Elevation of NA levels using the selective NA reuptake inhibitor atomoxetine did not affect clinical scores, while treatment of immunized mice with the synthetic NA precursor l-threo-3,4-dihydroxyphenylserine (l-DOPS) prevented further worsening. In contrast, treatment of mice with a combination of atomoxetine and l-DOPS led to significant improvement in clinical scores as compared to the control group. The combined treatment reduced astrocyte activation in the molecular layer of the cerebellum as assessed by staining for glial fibrillary protein but did not affect Th1 or Th17 type cytokine production from splenic T cells. These data suggest that selective elevation of CNS NA levels could provide benefit in EAE and multiple sclerosis without influencing peripheral immune responses.


multiple sclerosis locus coeruleus catecholamines astrocyte demyelination antidepressant 


Numerous studies have shown that the endogenous neurotransmitter noradrenaline (NA) can attenuate inflammatory events in astrocytes (Galea et al. 2003), microglia (Carnevale et al. 2007), and neurons (Madrigal et al. 2006). NA is also neuroprotective in vitro (Madrigal et al. 2007) and in vivo (Marien et al. 2004), involving induction of neurotrophin expression (Debeir et al. 2004) and increase of antioxidant systems (Lewerenz et al. 2003). Experimental depletion of NA exacerbates brain inflammation and neuronal damage following injection of pro-inflammatory cytokines or accumulation of aggregated forms of beta-amyloid in transgenic mice (Heneka et al. 2002), while elevating brain NA attenuates those responses (Hashioka et al. 2007; Kalinin et al. 2006). Increasing NA levels could, therefore, offer benefit in diseases such as multiple sclerosis (MS) which include an inflammatory component.

Evidence exists that perturbation of endogenous NA or NA signaling plays a role in MS and its animal model experimental autoimmune encephalomyelitis (EAE). Several early studies reported reduced levels of NA in spinal cords of EAE rats (White et al. 1983; Honegger and Isler 1984), altered levels of NA in CSF of MS patients (Barkhatova et al. 1998), and, more recently, reduced expression of β-adrenergic receptors (βARs) on white matter (WM) astrocytes in MS patients (Zeinstra et al. 2000). Beneficial effects of NA are suggested by studies showing that activating βARs (Chelmicka-Schorr et al. 1989) or increasing cAMP levels with phosphodiesterase inhibitors (Genain et al. 1995; Sommer et al. 1997) can ameliorate EAE symptoms, while depletion of sympathetic nervous system catecholamines exacerbated EAE (Chelmicka-Schorr et al. 1988). However, in general, the majority of these studies were not restricted to CNS but primarily involved modulation of peripheral lymphocyte and macrophage activation. In contrast, relatively few studies have examined effects of selectively increasing NA in the CNS. Findings that increasing central NA can provide benefit in EAE, with limited effects on peripheral NA and associated sympathetic regulation blood pressure, would make such treatments desirable.

In view of the known anti-inflammatory, neuroprotective, and neurogenic properties of NA, we, therefore, tested if selectively raising CNS NA levels would attenuate clinical symptoms of EAE in MOG-immunized C57BL/6 mice. NA levels can be increased by use of selective NA reuptake inhibitors (NARIs). Classical tricylic antidepressants such as imipramine or desipramine block NA as well as serotonin reuptake; in contrast, more recent non-tricyclic antidepressants show highly selective effects on NA reuptake and have been shown to have anti-inflammatory effects on microglial cells (Hashioka et al. 2007) as well as increase BDNF expression in hippocampus (Russo-Neustadt et al. 1999). In the current study, we tested the effects of atomoxetine which is currently used as non-stimulant treatment in the USA for attention deficit hyperactivity disorder (Prince 2006).

NA levels can also be increased by treatment with l-threo-3,4-dihydroxyphenylserine (l-DOPS, droxidopa) a synthetic catecholamino acid, which is given orally, and is converted to NA via decarboxylation by the ubiquitous enzyme l-aromatic-amino-acid decarboxylase (l-AAAD) (Goldstein 2006). Peripheral conversion of l-DOPS to NA is blocked by co-treatment with a selective l-AAAD inhibitor such as carbidopa or benserazide which do not pass the blood brain barrier. Since conversion of l-DOPS to NA is independent of noradrenergic activity, it is expected to show effects even if brain NA levels are perturbed. The effects of l-DOPS can be further increased by co-treatment with a NARI (Kato et al. 1986).

Our results suggest that a combination of a selective NA reuptake inhibitor together with a synthetic precursor of NA can afford benefit in this model, suggesting that similar therapeutic interventions may be of value in the treatment of MS.

Materials and methods


l-DOPS, benserazide, atomoxetine HCl, and general reagents were from Sigma (St. Louis, MO). Secondary antibodies were from Vector Labs, Burlingame, CA. The MOG35–55 peptide (MEVGWYRSPFSRVVHLYRNGK) was synthesized by Anaspec (San Jose, CA).


Female C57BL/6 mice, aged 6–8 weeks, were from Charles River Breeding (Cambridge, MA). Mice were maintained in a controlled 12:12 h light/dark environment and provided food ad libitum. All experiments were approved by the local IACUC committee.

Induction of EAE

EAE was actively induced in 6–8-week mice using synthetic myelin oligodendrocyte glycoprotein peptide 35–55 (MOG35–55) as described (Feinstein et al. 2002a). Mice were injected subcutaneously (s.c., two 100 μL injections into adjacent areas in one hind limb) with an emulsion of 300 μg MOG35–55 dissolved in 100 μL PBS, mixed with 100 μL complete Freund’s adjuvant containing 500 μg of Mycobacterium tuberculosis (Difco, Detroit, MI). Immediately after MOG35–55 injection, the animals received an intraperitoneal (i.p.) injection of pertussis toxin (PT, 200 ng in 200 μL PBS). Two days later, the mice received a second PT injection, and 1 week later, they received a booster injection of MOG35–55.

Clinical signs were scored on a five-point scale: grade 0, no clinical signs; 1, limp tail; 2, impaired righting; 3, paresis of one hind limb; 4; paresis of two hind limbs; 5, death. Scoring was done at the same time each day by a blinded investigator.


Brain NA levels were reduced by lesion of locus ceruleus (LC) noradrenergic neurons as previously described (Heneka et al. 2003). In brief, mice were treated with the selective neurotoxin N-(2-chloroethyl)-N-ethyl-2 bromobenzylamine (DSP4; Fritschy and Grzanna 1991) at 50 mg/kg, i.p., twice 1 week apart. Brain NA levels were raised by treating mice three times in a week with the non-tricyclic NA selective reuptake inhibitor Atomoxetine (Gould et al. 2005) at 20 mg/kg i.p.; the synthetic NA precursor l-threo-3,4-dihydroxyphenylserine (l-DOPS) at 400 mg/kg s.c. (Goldstein 2006) together with benserazide (125 mg/kg i.p.) which inhibits l-aromatic amino acid decarboxylase (l-AAAD) but does not pass the blood brain barrier and, therefore, prevents peripheral conversion to NA; or the combination of atomoxetine plus l-DOPS.

Tissue preparation and immunohistochemistry

Mouse brains were fixed in 4% paraformaldehyde in 0.1 M phosphate buffer pH 7.6 overnight at 4°C. Dehydration, embedding, paraffin removal, and sectioning were done using standard protocols as described (Sharp et al. 2008). Following paraffin removal, antigen retrieval was accomplished by boiling in 10 mM citrate buffer for 10 min, blocking with 5% normal donkey serum, permeabilization with 0.1% Triton X-100 for 30 min. Sections were incubated 4°C overnight with primary antibodies diluted in 1% normal donkey serum/rabbit polyclonal anti-MBP 1:300 (Zymed Laboratories, San Francisco, CA); and rat mAb anti-human GFAP B2.210 1:300 (Trojanowski et al. 1986). After washing, sections were incubated 1 h at 37°C with donkey anti-rabbit RRX- conjugated and donkey anti-rat conjugated with FITC secondary antibodies. Sections were washed, quenched in 50 mM ammonium chloride in PBS for 15 min, and final washes done in PBS with 400 ng/ml DAPI. Sections were coverslipped using VECTASHIELD mounting fluid (Vector Laboratories Inc., Burlingame, CA) and images obtained on a Zeiss Axioplan 2 fluorescence microscope.

Data analysis

Comparison of clinical signs over time in one group was done by one-way, nonparametric ANOVA (Kruskal–Wallis test) followed by Dunn's multiple comparison tests.

Comparison of the effect of treatment versus control on the development of clinical signs was done by two-way repeated measures ANOVA.

Comparison of effects of treatment versus control on Tcell cytokine production was done by Mann–Whitney nonparametric unpaired t test. In all cases, significance was taken at P < 0.05.


Reducing CNS NA exacerbates EAE

We first tested if depleting brain NA using the neurotoxin DSP4 to selectively lesion LC noradrenergic neurons influenced EAE. We previously showed that two i.p. injections (once per week) of DSP4 at 50 mg/kg cause significant loss of LC noradrenergic neurons (Heneka et al 2002). Two weeks after the first DSP4 injection, mice were immunized with MOG35–55 peptide using the protocol described above in the “Materials and methods” section. Results from two independent experiments show that in the control group, 11/15 (73%) mice developed clinical symptoms, and the average day of onset was 8.9 ± 0.8 (mean ± SE) days. In the DSP4-treated group, the incidence was 14/15 (87%), and the average day of onset (8.6 ± 0.6) was not significantly different from the control group (Fig. 1b). However, average daily clinical scores were significantly different than those in the control group (two-way ANOVA, [22,1] = 1.6, P = 0.04; Fig. 1a). Since DSP4 causes long-lasting reduction of brain NA content with limited effects on peripheral NA levels (Jonsson et al. 1981), this suggests that selectively reducing NA levels in the CNS exacerbates EAE. The fact that disease incidence and onset was not affected further suggests that LC loss did not influence peripheral T-cell activation.
Fig. 1

Lesion of locus coeruleus exacerbates EAE. C57BL/6 female mice were treated with DSP4 (50 mg/kg; i.p., twice 1 week apart) then 1 week later (14 days after the first DSP4 injection) were immunized with MOG35–55 peptide. The data shown is mean ± SE and is derived from two separate studies. a Average daily clinical scores; b average daily incidence of disease, for vehicle (open circles, n = 15) and DSP-4-treated (filled circles, n = 23) mice. In both groups, scores were significantly different from baseline values by day 9 post-MOG booster (one-way ANOVA), and there was no difference in the average day of onset (8.6 ± 0.6 vs. 8.9 ± 0.8, DSP4 and vehicle groups). However, there was a statistically significant difference between development of clinical scores in the DSP4 versus the vehicle groups (two-way ANOVA, F[22,1] = 1.6, P = 0.04)

Increasing NA improves EAE

We tested if treatment with a selective NARI could influence the course of EAE. Mice were given the non-tricyclic NARI atomoxetine beginning at day 17 after the booster MOG immunization and clinical scores monitored for 3 weeks; however, no changes were observed (Fig. 2a). We then tested if treatment with a synthetic precursor of NA, l-DOPS, which is converted to NA by the ubiquitous enzyme l-AAAD, would attenuate symptoms of EAE when given to already-immunized mice (Fig. 2b). Results from two independent studies show that in the control group, 3 of 15 mice worsened during treatment, and one improved; the average score increased from 1.9 ± 0.25 to 2.1 ± 0.10 at the end of the study (P < 0.05, one-way repeated measures ANOVA). In contrast, in the l-DOPS-treated group, 5 of 13 mice improved and two worsened during treatment period; the score decreased from 1.8 ± 0.2 to 1.5 ± 0.2 during that time, and there was a statistically significant effect of l-DOPS on development of clinical scores compared to controls (F[42,1] = 1.8, P < 0.005 time × treatment, two-way ANOVA). However, the final scores in the l-DOPS group were not significantly different from the score at the beginning of treatment, suggesting that treatment with l-DOPS stabilized but did not ameliorate clinical severity.
Fig. 2

Increasing brain NA attenuates EAE. C57BL/6 female mice were immunized with MOG35–55 peptide to develop EAE. At the indicated times, the mice were treated with a vehicle (n = 6) or the NA reuptake inhibitor atomoxetine (n = 6) 20 mg/kg i.p. three times in a week; bl-DOPS (n = 13) at 400 mg/kg s.c., three times in a week together with the l-AAAD inhibitor bensazeride (125 mg/kg s.c.) or with bensazeride only (control, n = 15 total); or c with the combination of l-DOPS/benserazide and atomoxetine (n = 9) or benserazide alone (n = 10). Combined treatment significantly decreased clinical scores combined to control group (F[16,1] = 3.6, P < .001, two-way ANOVA, time × treatment); *P < 0.05; **P < 0.01 versus day 17, one-way nonparametric ANOVA (Kruskal–Wallis; Dunn's multiple comparison test). In b, the data is derived from two independent studies combined and is, therefore, plotted relative to the day treatment was started

We then tested if the combination of l-DOPS together with atomoxetine could provide further benefit (Fig. 2c). There was a statistically significant effect of l-DOPS plus atomoxetine treatment on development of scores compared to control mice (P < 0.001, two-way ANOVA), and those scores significantly improved as compared to day 17 scores (one-way ANOVA and Dunn's multiple comparison test). Together, the data support that increasing CNS NA can have beneficial effects on EAE clinical signs, although significant improvements may require a combination treatment.

Immunohistochemical staining of brains isolated from control and l-DOPS-treated mice on day 19 after treatment was started revealed astrocyte activation, as assessed by staining for GFAP, throughout the brains of control mice at the peak of clinical disease. Strongest staining was observed in the cerebellum (Fig. 3) in white matter (WM) as well as in Bergman radial glial fibers in the molecular layer (ML) and little or no staining in the granule cell layer (GL) or Purkinje cell layer (PL). In contrast, in the treated mice, GFAP staining was observed only in cerebellar WM, suggesting astrocyte activation is reduced by raising brain NA levels.
Fig. 3

l-DOPS reduces astrocyte activation. At the end of the study shown in Fig. 2b (day 19 after treatment was started), mice were sacrificed, serial cerebellar sections from a vehicle-treated mice and bl-DOPS-treated mice were stained for the astrocyte marker GFAP. Representative images from one mouse in each group show a reduction in GFAP staining in ML of the l-DOPS-treated sample but similar levels of staining in the WM and little detectable staining in the PL or GL. ML molecular layer, WM white matter, PL Purkinje cell layer, GL granular layer

To test if treatment with l-DOPS and atomoxetine affected Tcell functionality, we measured cytokine production from splenic T cells isolated from control and l-DOPS plus atomoxetine-treated mice at the end of the study shown in Fig. 3 on day 42 (Fig. 4). After a 24-h restimulation with MOG peptide or upon activation using antibodies to Tcell receptor CD3 and co-stimulatory receptor CD28, there were no statistically significant differences in either IFNγ and IL-17 production between the two groups (Mann–Whitney nonparametric unpaired t test) suggesting that, at least at this time point, neither l-DOPS nor atomoxetine significantly reduced splenic Tcell inflammatory responses.
Fig. 4

Effects of l-DOPS and atomoxetine on T-cell activation. At day 42, at the end of the study shown in Fig. 2c, splenic T cells were isolated from control and l-DOPS/atomoxetine-treated mice and cultured in RPMI media alone or in the presence of MOG35–55 peptide (20 μg/ml), antibody to CD3, or antibodies to CD3 and CD28. After 24 h, levels of IFNγ (a) and IL-17 (b) were measured in aliquots of the media. Data are mean ± SD of n = 3 samples per group. There were no statistically significant differences between cytokine production from control or treated cells (nonparametric unpaired t test)


In the current study, we show that treatment of MOG-immunized mice with the NA precursor l-DOPS stabilized clinical scores, while using l-DOPS together with the NARI atomoxetine significantly decreased clinical scores. We did not see any effects following treatment with atomoxetine alone, suggesting that endogenous levels of NA are not sufficient to provide anti-inflammatory or neuroprotective effects. Since in some EAE studies, CNS levels of NA were shown to be reduced (Krenger et al. 1986; White et al. 1983), the lack of effect of atomoxetine could be due in part to sub-physiological NA levels during EAE, thereby explaining the need for further augmentation of NA using l-DOPS. Co-treatment of mice with l-DOPS and a NARI has previously been shown to increase brain NA levels over treatment with l-DOPS alone (Kato et al. 1986). Since l-DOPS is converted into NA by the ubiquitous enzyme l-AAAD which is present in glial cells as well as neurons (Nakamura et al. 2000), the combined effect may reflect increased availability of NA to glial adrenergic receptors whose activation is known to reduce inflammation (Feinstein et al. 2002b).

Several studies have shown that changes in peripheral NA levels or signaling occur during EAE and MS. For example, peripheral NA levels were decreased (Rajda et al. 2006) or increased (Cosentino et al. 2002) in MS patients versus healthy controls, possibly depending upon the state of disease, while polymorphic bone marrow cells from MS patients were shown by several groups to express increased βARs compared to controls (Zoukos et al. 1992). It has also been shown that treatments which modify peripheral NA levels influence the course of EAE. For example, lesion of the sympathetic nervous system augmented symptoms of acute monophasic and adoptively transferred EAE (Chelmicka-Schorr et al. 1988), as did peripheral depletion of splenic NA using 6-OHDA (Leonard et al. 1991). Similarly, increasing peripheral adrenergic activity can attenuate EAE, either by use of βAR agonists (Chelmicka-Schorr et al. 1993; Wiegmann et al. 1995) or treatment with type IV phosphodiesterase inhibitors such as rolipram (Genain et al. 1995; Dinter et al. 2000). However, the above studies focused primarily on suppression of Tcell activation or infiltration into the brain and did not address the consequences of modifying NA levels in the CNS.

In the current study, we examined if treatment with l-DOPS and atomoxetine reduced Tcell activities and found that Tcell Th1 (IFNγ) and Th17 (IL-17) type production was comparable in splenic cells isolated from control or treated mice at the end of treatment. This is consistent with the fact that the primary effects of these drugs are central as well as that Tcell priming, proliferation, and infiltration into the CNS occur at relatively early times during EAE progression before treatments were started. While changes in peripheral NA levels and signaling are known to occur in EAE and MS, our data suggest that this is not the primary means by which these drugs are influencing EAE.

In contrast to a role for peripheral NA in EAE and MS, the role of CNS NA is not well characterized. Several studies point to perturbation of CNS NA or NA signaling in EAE or MS. In one EAE study in dogs, CSF and WM NA levels were found increased during early time points but reduced during the clinical period (Khoruzhaia and Saakov 1975); while in rodents, brainstem and spinal cord NA levels were reduced (Krenger et al. 1986; White et al. 1983). In contrast, in MS patients, it was reported that NA levels are increased in CSF (Barkhatova et al. 1998) which could represent a compensatory response to reduced adrenergic activation. It has also been reported that astrocytes in WM of MS brain have reduced levels of β2-ARs compared to controls (Zeinstra et al. 2000) which would be expected to reduce the anti-inflammatory actions of NA in these cells.

Likewise, few studies have examined effects of selectively modulating CNS NA in EAE or MS. In one study, it was reported that electrolytic destruction of noradrenergic neurons attenuated the EAE in rats (Jovanova-Nesic et al. 1993), and a few studies reported that central depletion of NA using 6-OHDA (Konkol et al. 1990) or stereotaxic electrolytic destruction of the anterior hypothalamus (Abramsky et al. 1987) reduced the severity of EAE suggesting an exacerbating role for central NA. More recently, it was shown that chronic treatment with venlafaxine, a bicyclic, serotonin/noradrenaline reuptake inhibitor antidepressant ameliorated clinical disease in adoptive transfer EAE in mice, both preventatively as well as therapeutically (Vollmar et al. 2008). Although this was associated with an ability to reduce Tcell activation in vitro, effects on glial cell activation, remyelination, and oligodendrocyte progenitor cell proliferation or maturation were not examined.

In contrast to EAE studies, most reports of modulating NA levels in MS patients point to beneficial effects. An initial suggestion came from a clinical trial carried out in Sweden in 1985 (Berne-Fromell et al. 1987) in which researchers treated 300 MS patients with a combination of tri- or tetra-cyclic antidepressants together with l-DOPA (which, after conversion to dopamine by l-AAAD, leads to an increase in NA upon metabolism by dopamine beta-hydroxylase). The authors reported that within 1 to 2 months, approximately 75% of the patients showed substantial improvements in sensory, motor, and autonomic systems, which led them to propose that damage to the LC system and accompanying NA deficiency causes axonal damage and demyelination. More recently, it was reported that treatment with l-phenylalanine (an upstream precursor in the NA synthesis pathway) together with the NARI lofepramine (which is metabolized to desipramine in vivo) improved clinical symptoms (Loder et al. 2002) as well as reduced lesion volume (Puri et al. 2001) in a relatively large cohort (n = 69 drug-treated and n = 69 placebo-treated) of MS patients. Our current results confirm that increasing central NA levels pharmacologically can provide benefit in a chronic model of MS and, therefore, suggest that further testing of these or similar interventions may be of value in treatment of MS.

Copyright information

© Springer Science+Business Media, LLC 2009

Authors and Affiliations

  • Maria Vittoria Simonini
    • 1
    • 2
  • Paul E. Polak
    • 1
    • 2
  • Anthony Sharp
    • 1
    • 2
  • Susan McGuire
    • 1
    • 2
  • Elena Galea
    • 1
    • 2
  • Douglas L. Feinstein
    • 1
    • 2
    • 3
  1. 1.Department of AnesthesiologyUniversity of Illinois at ChicagoChicagoUSA
  2. 2.Department of Veterans AffairsJesse Brown VA HospitalChicagoUSA
  3. 3.ChicagoUSA

Personalised recommendations