Re-assessing the catecholamine hypothesis of depression: the case of melancholic depression

Abstract

The hypothesis that major depression represents a deficiency of norepinephrine (NE) at critical synapses in the central nervous system has been extremely influential since the mid-1960s. First espoused by Schildkraut [1] and Bunney and Davis [2], the hypothesis was based on the following pharmacological evidence: (i) that imipramine acutely increased NE neurotransmission; (ii) that MAO (monoamine oxidase) inhibitors blocked the metabolism of NE, and hence increased NE neurotransmission; (iii) reserpine, which produced depression, depleted NE from noradrenergic neurons [1, 2]. Though several have questioned the validity of the hypothesis, it has never been definitively challenged. We report here that noradrenergic neurotransmission and locus caeruleus activity is increased rather than decreased in the specific subtype of major depression, melancholia [3–5]. The symptom complex of melancholic depression is compatible with NE hypersecretion in multiple ways [5]. Melancholia contradicts the term depression, in that it is not a state of behavioral and physiological suppression, but rather, often a state of hyperarousal and a tortured sense of worthlessness. NE promotes anxiety and hyperarousal, stimulates corticotropin-releasing hormone (CRH)-mediated hypercortisolism, is neurotoxic promotes insulin resistance, and is an important mediator of bone loss, all of which are relevant to melancholic depression (reviewed in [5]). We found that patients with melancholic depression had increased CSF (cerebrospinal fluid) and plasma levels of NE in samples taken hourly through indwelling lumbar and venous cannula sampled hourly for 30 h [3, 4] (Fig. 1). Plasma cortisol and epinephrine levels were also elevated during the entire 30 h (Fig. 1) [3]. We also found that in both patients and controls, CSF and plasma NE had virtually identical diurnal variations and their diurnal variations were also superimposable with that of plasma cortisol (Fig. 1) [3, 4]. Detrended lag correlations revealed that around-the-clock CSF and plasma NE and plasma cortisol levels were significantly correlated independent of the diurnal rhythm [4]. Not only are CSF and plasma NE levels elevated around-the-clock in depressed patients, but their linkage with the hypercortisolism of melancholia makes it virtually inconceivable that plasma and CSF NE are reduced in depressed patients in the setting sustained elevations in plasma cortisol. This is further supported by the well-known close linkage between the CRH and locus caeruleus/NE systems. Valentino has shown that CRH stimulates the locus caeruleus, while we have shown that NE stimulates CRH release [6]. In addition, it has been found that differential blockade of CRF-evoked behaviors occurs by depletion of NE in rats. Finally, data confirm that CRH engagement of the locus caeruleus-NE system mediates stress-induced anxiety [7]. We also found NE spillover into arterial plasma is also significantly elevated in patients with depression compared to controls at baseline and in response to video-game stress and yohimbine, an alpha-2 noradrenergic antagonist that increases the secretion of plasma NE [3] (Fig. 2). NE spillover into arterial plasma represents the levels of arterial NE corrected for their half-lives via the clearance of NE. This study clearly supports the results of the 30 h studies and strongly support the premise that NE is elevated in patients with melancholic depression. This premise is supported by multiple lines of data showing that multiple tricyclic antidepressant drugs, which are preferentially effective in melancholia. decrease the basal and sensory-evoked firing rates of locus caeruleus * Philip W. Gold [email protected]