The endogenous neurotransmitter noradrenaline (NA) plays several important roles in maintaining normal CNS physiology. NA acting at β2-adrenergic receptors can suppress glial as well as neuronal inflammatory responses (Frohman et al. 1988; Minghetti et al. 1997; Szabo et al. 1997; Galea and Feinstein 1999), and in vivo selective NA reuptake inhibitors can reduce CNS cytokine and chemokine expression (O’Sullivan et al. 2010), and increase anti-inflammatory cytokine expression (McNamee et al. 2010). NA acting directly on neurons can reduce neurotoxicity elicited by inflammatory (Madrigal et al. 2005) or excitotoxic (Madrigal et al. 2007) stimuli, both in vitro and in vivo (Marien et al. 2004; Traver et al. 2005). NA can increase neurotrophin expression in glial cells (Debeir et al. 2004), including that of brain derived neurotrophic factor (BDNF) and glial cell derived neurotrophic factor (GDNF) (Remy et al. 2001), important for maintenance of neuronal integrity and synaptic plasticity. NA also has beneficial effects on neural progenitor cells (Ghiani et al. 1999), including on the maturation of oligodendrocyte progenitor cells which is of particular relevance within the context of demyelinating diseases. Disturbances in NA homeostasis could therefore influence the course of neurological diseases such as MS.
Evidence that NA plays a role in MS is suggested by studies showing alterations in CNS NA levels. In experimental autoimmune encephalomyelitis (EAE), CSF and white matter NA levels were found increased before clinical signs appeared, but were reduced at later times (Khoruzhaia and Saakov 1975). Similarly, NA levels were found reduced in the brainstem and spinal cords of EAE rats (White et al. 1983; Krenger et al. 1986). There is also evidence that dysregulation of CNS NA occurs in MS, for example NA levels were higher in CSF samples of MS patients compared with controls (Barkhatova et al. 1998); and CSF levels of the NA metabolite methoxyhydroxyphenylglycol were inversely correlated to the duration of illness and number of relapses (Markianos et al. 2009).
The major source of NA in the CNS are large tyrosine hydroxylase (TH)-expressing neurons of the Locus coeruleus (LC) located in the rostral portion of the pons. The LC provides noradrenergic input to most regions of the brain including cortex, cerebellum, brainstem, and spinal cord. Substantial loss of LC neurons (Mann et al. 1980; Tomlinson et al. 1981; Bondareff et al. 1987) and reduced levels of NA (Adolfsson et al. 1979; Mann and Yates 1983) are well known hallmarks of several diseases including Alzheimer's disease (AD) and Parkinson's disease (PD). The consequences of LC loss are not fully understood, but increasing reports demonstrate the experimental lesion of the LC exacerbates inflammatory responses and the progression of neurodegeneration in mouse models of AD (Heneka et al. 2006; Kalinin et al. 2007; Jardanhazi-Kurutz et al. 2010). In view of reports suggesting that NA levels are disturbed in EAE and MS, we examined LC physiology in MOG peptide immunized mice and in sections of the LC from MS patients and healthy controls (Polak et al. 2011). These findings demonstrated the presence of LC stress and damage in both EAE and MS, that was associated with increased glial inflammation and reduced levels of NA. A recent study showing perturbation of the LC:Noradrenergic system in a mouse model of Down syndrome (Salehi et al. 2009), and of LC damage in veterans with post-traumatic stress disorder (Bracha et al. 2005), suggests that LC damage may be a commonality to several neurological conditions and diseases.
To address LC damage and the perturbation of NA levels that occurs, several methods to increase CNS NA levels have been tested and shown to provide benefit. Treatment with α2-antagonists, which increase NA release from LC neurons, reduced damage caused by intraparenchymal injection of Αβ1-42 (Kalinin et al. 2006), and that treatment with droxidopa, a synthetic precursor of NA, reduced pathology in a robust mouse model of AD (Kalinin et al. 2011), and reduced behavioral deficits in a mouse model of Down’s syndrome (Salehi et al. 2009). More recently, we showed that treatment of EAE mice with droxidopa stabilized disease progression in MOG peptide induced EAE while co-treatment with a NA reuptake inhibitor led to clinical improvement (Simonini et al. 2010).
While methods to increase CNS NA levels may provide some benefit in neurological diseases including MS, they do not address a primary cause of NA dysfunction, namely LC stress and damage. In this regard, drugs which selectively increase LC neuronal viability or activity might be of benefit. Vindeburnol (14,15-dihydro-20,21-dinoreburnamenine-14-ol, Fig. 1a) is a semi-synthetic derivative of the plant alkaloid vincamine (Fig. 1b), and structurally related to vinpocetine (Fig. 1c), recently shown to inhibit the NFkB signaling system (Jeon et al. 2010). Vincamine is a peripheral vasodilator isolated from the plant Vinca minor; originally characterized for possible effects on cognition and memory caused by its ability to increase blood flow to the brain (Vas and Gulyas 2005). Observations that vincamine could activate noradrenergic firing in the LC (Olpe and Steinmann 1982) suggested vindeburnol might influence LC physiology. Vindeburnol injection caused a rapid depletion of brain NA levels in rats (Euvrard and Boissier 1981), and increased TH activity and protein expression in the LC, but not in other dopaminergic nuclei (Weissmann et al. 1988). Although mechanistic studies have not been done, immunohistochemical studies following treatment in adult rats suggest that vindeburnol induces re-activation of TH in a set of LC neurons that transiently express TH during early development (Marcel et al. 1998; Bezin et al. 2000). In this study, we demonstrate that vindeburnol reduces clinical scores in mice immunized with MOG peptide to develop chronic disease, and is associated with improvements in LC physiology and function.
Figure 1. Chemical structures of related vincamine derivatives: (a) vindeburnol; (b) vincamine; and (c) vinpocetine.
Download figure to PowerPoint
- Top of page
- Materials and methods
In this study, we show that vindeburnol provides benefit in a chronic mouse model of MS, associated with improvements in LC physiology. Early studies using vindeburnol, also known as RU-24722, were based on observations that the structurally related compound vincamine increases cerebral blood flow and metabolism (Tesseris et al. 1975) and activates noradrenergic neurons in the LC of the rat (Olpe and Steinmann 1982). Vindeburnol was originally shown to enhance post-ischemic recovery in gerbils and rats (Barzaghi et al. 1985), and found to selectively deplete CNS NA levels (Euvrard and Boissier 1981). Subsequently, it was found that vindeburnol (a single dose of 10 mg/kg, i.p.) caused a rapid (within 1 h) decrease of NA levels in mouse brain at the same time increasing levels of several NA metabolites, indicating an increase in NA turnover (Takeda et al. 1984). Vindeburnol’s mechanism of action appears to differ from that of vincamine which had little effect on brain monoamine levels when used at higher doses, and differs from that of reserpine, an inhibitor of the vesicular monoamine transporter which blocks storage of monoamines into vesicles.
The above studies led to the hypothesis that vinderburnol influenced the LC: noradrenergic signaling system. This was confirmed in studies using adult rats, where single injections of vindeburnol were shown to selectively increase TH protein levels in the LC (Labatut et al. 1988) but not in other dopaminergic nuclei (Weissmann et al. 1988). Within the LC, the specific cells that are induced to increase TH expression are not located throughout, but are present in certain sub-regions. Using quantitative immunostaining methods for TH (Debure et al. 1992), it was shown that in normal adult rat LC, there is a heterogeneous distribution of TH+ cells and TH protein expression with a significantly greater number and greater expression in the posterior compared with the anterior region of the LC. Following treatment with vindeburnol, the total numbers of TH+ neurons and levels of TH expression were significantly increased across the entire LC, due to an increase of TH expression in the anterior located cells, and to de novo appearance of TH expression in the posterior region. These cells were subsequently shown to be Phox2A+/TH− neurons that are capable of re-expressing TH (Bezin et al. 2000). Our data showing that vindeburnol increased levels of several mRNAs that are involved in LC maturation (Mash1), NA synthesis (DBH, dopamine decarboxylase), or NA signaling (NET1, α2AR) could therefore be due to reduced damage occurring to existing TH+ neurons, or to the appearance of new TH+ cells.
The mechanisms of action of vindeburnol are not yet well known. In vivo studies showed that its effects on TH maturation could be prevented by co-treatment with clonidine, an α2-adrenergic receptor antagonist (Labatut et al. 1988), suggesting that vindeburnol actions involve binding to these receptors. As LC neurons express α2-adrenergic receptors, binding of vindeburnol to those autoreceptors could lead to a sustained increase in NA release and ultimately a requirement for de novo NA synthesis and increased TH expression. The related compound vinpocetine has been tested in several models of neurological diseases, is used in Europe to treat age related memory deficits, and has been considered for treatment of AD and Parkinson's disease (Szatmari and Whitehouse 2003). Vinpocetine was shown to inhibit certain phosphodiesterases and modulate calcium channels (Vas and Gulyas 2005), and recently to inhibit the NFkB signaling system (Jeon et al. 2010). Whether vindeburnol has any of these properties remains to be determined.
Our results using splenic T cells show no reduction because of vindeburnol treatment on pro-inflammatory cytokine production; and is consistent with the premise that the actions of vindeburnol occur within the CNS. However, it is unclear why the effects of vindeburnol on clinical scores in EAE were not observed in all the treated mice. After 3 weeks treatment with vindeburnol given 3× per week, we observed significant improvement of clinical signs in about 50% of the vindeburnol-treated mice. When the dosing regimen was increased to 6× per week (only given once on weekends), the number of responders increased to 75% (6/8). The observations that IFNγ and IL17 levels were greater in splenic T cells prepared from non-responder mice offers a possible explanation to help account for the apparent lack of clinical improvement, because larger peripheral inflammatory responses could have caused increased inflammation within the LC, thereby reducing the efficacy of vindeburnol to prevent LC neuronal damage. That these mice showed similar clinical signs as the mice who responded to vindeburnol could be due to the fact that treatment was started at day 10, the time when clinical signs first appear and before the consequences of increased immune responses could be detected by the scoring methods used. Other possible reasons for lack of response to low dosing of vindeburnol, for example differences in expression levels of enzymes that metabolize NA, could also contribute to this observation and are currently being explored.
Regardless of the precise reasons why some mice did not respond, our findings that increases in LC TH and DBH mRNA levels, and in spinal cord NA levels, occurred in responders supports the premise that beneficial effects of vindeburnol are due, at least in part, to improvement of LC function. In those mice, measurements of spinal cord NA showed an approximate 50% increase over levels measured in vehicle-treated mice. However, these values (300 pg/mg wet weight) are still significantly reduced as compared with levels measured in non-immunized mice. One possible explanation is that during EAE, there is not only reduced production of NA, but increased metabolism. NA (as well as other catecholamines) are metabolized by a variety of enzymes including catecholamine-O-methyl transferase (COMT) and monoamine oxidase (MAO). Both MAO and COMT are ubiquitous enzymes expressed throughout the CNS including in the spinal cord (Hong et al. 1998; Samantaray et al. 2008), and both have been shown to be increased in neurological diseases including amyotrophic lateral sclerosis (Ekblom et al. 1993). MAO inhibitors have shown efficacy in MOG peptide induced EAE (Musgrave et al. 2011), while COMT inhibitors are now used in Parkinson’s disease as adjuncts to treatment with levodopa (Marin and Obeso 2010). Treatments to increase CNS NA levels, either by directly targeting the LC, or by use of NA precursors such as droxidopa (Simonini et al. 2010) may therefore benefit by co-administration of MAO or COMT inhibitors.
In conclusion, the current data suggest that treatment with vindeburnol can provide benefit in EAE, and is associated with positive effects on LC physiology. Vindeburnol is now being tested in phase I clinical trials in France for treatment of clinical resistant depression (http://www.biocortech.com/newEvents.php?news=2010), based on its ability to increase blood flow to the brain. Positive findings from those studies together with our current findings could provide the basis for testing of vindeburnol in MS patients.