New directions for the treatment of depression: Targeting the photic regulation of arousal and mood (PRAM) pathway

Abstract

Both preclinical and clinical studies demonstrate that depression is strongly associated with reduced light availability, which in turn contributes to decreased function of brain regions that control mood. Here, we review findings that support a critical pathway for the control of mood that depends upon ambient light. We put forward a novel hypothesis, functionally linking retina to locus coeruleus (LC) in depression, and discuss the role of norepinephrine in affective disease. Finally, we discuss how utilizing the chemogenetic tool Designer Receptors Exclusively Activated by Designer Drugs (DREADDs) to precisely control this retina–LC circuit may be used as a novel therapeutic to treat depression.

Depression is a complex mood disorder with a lifetime prevalence of 16% (Kessler et al., 2005), and a 12-month prevalence of over 9% in the United States (Demyttenaere et al., 2004). It is characterized by persistent low mood, loss of pleasure or interest, disrupted sleep and appetite, and feelings of guilt or low self-worth (Serretti, Mandelli, Lattuada, & Smeraldi, 2004). Depression is associated with impaired cognitive (Hammar & Ardal, 2009) and social function (Rubin & Burgess, 2001; Setterfield, Walsh, Frey, & McCabe, 2016), and is predicted to be the leading cause of disease burden by 2030 (Lépine & Briley, 2011), possibly due to its comorbidity with other psychiatric disorders such as anxiety (Fava et al., 2008), drug addiction (Volkow, 2004), and attention-deficit hyperactivity disorder (Waxmonsky, 2003).

Despite extensive research efforts, the precise etiology of depression remains elusive. The conceptualization of depression as a physical illness was prompted by the serendipitous discovery of the first psychopharmacological treatments in the 1950s. Since then, the formation of the monoamine (Prange, 1964; Schildkraut, 1965) and neurohormonal (Nemeroff, 1988) theories of depression have undoubtedly enhanced our understanding of the mechanisms underlying the disease, and indeed, have arguably led to relative enhancements of treatment outcomes. Yet despite these and other credible theories underlying the biology of depression (for review see Ordway, Klimek, & Mann, 2002), many important questions remain unanswered: no consistent genetic associations between loci and depression diagnosis or treatment responsivity have been identified Fried, 2015 (Lewis et al., 2010; Ripke et al., 2013; Shi et al., 2011; Tansey et al., 2012; Wray et al., 2012), pharmacological treatments remain suboptimal, and there has been a stagnation in the development of new pharmacological treatments (Khan, Khan, & Brown, 2002; Pigott, Leventhal, Alter, & Boren, 2010). These deficits may reflect lack of understanding of disease mechanisms at both the molecular and circuit levels. Therefore, there is a clear need to better understand the precise brain circuitry associated with depression.

1 CIRCADIAN RHYTHMS AND DEPRESSION: A PROPOSED CRITICAL ROLE FOR THE PRAM PATHWAY

Decades of research indicates that depression is associated with disrupted circadian rhythms. Indeed, depression is associated with a blunted amplitude and phase delay of circadian rhythms (Duncan, 1996), increased core temperature (Avery, Wildschiødtz, & Rafaelsen, 1982), and phase-advanced oscillations of noradrenaline and cortisol plasma concentrations (Koenigsberg et al., 2004). These findings are further supported by well-documented alterations of sleep in depressed patients (Perlis, Giles, Buysse, Tu, & Kupfer, 1997; Posmontier, 2008; Stewart et al., 2006), and are reflected in the DSM-V that lists insomnia or hypersomnia as criteria for major depressive disorder (MDD; American Psychiatric Association, 2013). Given the close association between light and the entrainment of circadian rhythms, significant effort has been directed toward understanding the role of light in depressive disorders. Particular attention has focused on seasonal affective disorder (SAD), the etiology of which is typically associated with decreased light availability (American Psychiatric Association, 2013). The onset of SAD symptoms often coincides with the decreased day lengths during fall and winter, and SAD is more prevalent in extreme latitudes where natural light is very limited during the colder months (Rosenthal et al., 1984). These patterns are supported by preclinical literature, as short day-length lighting schedules induce depression and anxiety-like behavior in rats (Einat, Kronfeld-Schor, & Eilam, 2006), and mice housed under irregular light conditions display depression-like behavior that can be alleviated by the antidepressant fluoxetine (LeGates et al., 2012). Despite a clear role for light-mediated disruption of circadian rhythms in depression, the precise neural mechanisms underlying this phenomenon are not well understood.

Fifteen years ago, our laboratory described a circuit for the circadian regulation of arousal. We demonstrated that the brain nucleus locus coeruleus (LC) receives an indirect circadian input from suprachiasmatic nucleus (SCN) via a relay in dorsomedial hypothalamus (DMH), thus revealing a circuit for the regulation of sleep and waking (Aston-Jones, Chen, Zhu, & Oshinsky, 2001). Subsequently, we discovered that light deprivation over 6 weeks in Sprague-Dawley rats led to depression-like behavior and decreased innervation of frontal cortex by noradrenergic (NA) LC fibers and terminals (Gonzalez & Aston-Jones, 2008); the latter is a physiological trait of animals with depression-like behavior (Kitayama et al., 1994).

SCN receives light information via a major retinal input, and works to coordinate the circadian clocks of the brain and body. It is therefore in a key position to integrate information about light with circuitry that regulates mood, via the pathway: retina → SCN → DMH → LC. We refer to this pathway as the photic regulation of arousal and mood (PRAM) pathway.

Here, we propose the hypothesis that some forms of depression are associated with disruptions to the normal physiological activity of the PRAM pathway. This hypothesis is based upon three key points: (1) the SCN → DMH → LC circuit plays a critical role in the circadian regulation of arousal (Aston-Jones et al., 2001), (2) light has a profound effect on mood and depressive behavior (Gonzalez & Aston-Jones, 2008) and SCN receives light information from a major retinal input, and (3) decreased light availability leads to degeneration of NA-LC cortical fibers (Gonzalez & Aston-Jones, 2008), which is associated with depression (Kitayama et al., 1994). This review will provide evidence for this pathway in depression, and will discuss how altering this pathway may eventually lead to functional alterations in mood. First, we focus on both clinical and preclinical evidence implicating the NA–LC system in the development of depression and depression-like symptomology. We then provide anatomical and functional evidence implicating SCN as a key regulator of the NA–LC system in depression, and how this region is modulated by photic input from the retina. Finally, we propose using a targeted chemogenetic approach to normalize activity of the PRAM pathway in depressed individuals, thus ameliorating symptomology associated with hypoarousal and disrupted circadian rhythms.

2 THE NA–LC SYSTEM AND DEPRESSION

LC (catecholaminergic cell group A6) is the principal site of norepinephrine (NE) synthesis within the brain, and provides the main source of ascending and descending NA brain fibers. The NA–LC system projects widely throughout the neuraxis, and has critical roles in a range of physiological functions, including the circadian regulation of arousal. NA–LC neurons fire tonically faster during waking (Aston-Jones & Bloom, 1981) or during the active period in anesthetized animals (Aston-Jones et al., 2001) than during sleep or the inactive period. Also, plasma NE levels are higher during waking (Linsell, Lightman, Mullen, Brown, & Causon, 1985). Unsurprisingly therefore, the NA–LC system plays a critical role in regulating the sleep-wake cycle; reducing LC activity promotes sleep, and increasing its activity facilitates alertness and attention (Berridge & Waterhouse, 2003) as well as emergence from anesthesia (Vazey & Aston-Jones, 2014). Additionally, the loss of NE cortical fibers induced by either light deprivation or the neurotoxin N-(2-chloroethyl)-N-ethyl-2-bromobenzylamine (DSP-4) decreases the amplitude of the sleep-wake rhythm (Gonzalez & Aston-Jones, 2006; Gonzalez & Aston-Jones, 2008). As we outline below, there is considerable evidence from both clinical studies and animal models that activity of the NA–LC system is profoundly dysregulated in depression. We propose that this disruption of normal NA–LC function in depression results in reduced NE signaling in key arousal regions in the brain, thus contributing to the dysregulation of mood observed in depressed patients.

Clinical studies generally report a deficit of NA–LC function in depressed patients. For example, Chan-Palay and Asan (1989) described an extreme loss of catecholaminergic LC cells in the postmortem brains of patients with MDD. Similarly, Arango, Underwood, and Mann (1996) reported that suicide patients have a lower density of LC neurons compared with psychiatrically normal patients, whereas others have reported reduced concentrations of 3-methoxy, 4-hydroxphenylglycol (MHPG), the principal metabolite of NE, in the cerebrospinal fluid (CSF) of suicide attempters (Agren, 1980, 1982). Moreover, reduced NE levels are associated with reduced mood in persons with a family history of depression (Ruhe, Mason, & Schene, 2007), and growth factor signaling and glial functioning are altered in LC of patients with depression (Bernard et al., 2011).

Further evidence of NE dysregulation in depression comes from the relative effectiveness of antidepressant treatments that normalize NE in ameliorating depressive symptomology in clinical populations. Serotonin-NA reuptake inhibitors (SNRIs), NE-dopamine reuptake inhibitors (NDRIs), NE reuptake inhibitors (NRIs), and tricyclic antidepressants (TCAs) all act to increase NE by preventing its reuptake by the presynaptic neuron. These therapies have shown some success as pharmacological treatments for depression, and have been suggested to offer a greater therapeutic advantage over selective serotonin reuptake inhibitors (SSRIs; Moret & Briley, 2011). Additionally, a meta-analysis of 193 trials found that SNRIs may be more effective than SSRIs in attaining remission from depression (Papakostas, Thase, Fava, Nelson, & Shelton, 2007). Taken together, depression is associated with a loss of NA-LC cells and a subsequent deficit of NE signaling, and pharmacotherapies that restore NE levels can be effective treatments for depression symptomology.

These findings from clinical populations are supported by evidence arising from animal studies of depression-like behavior. Kitayama et al. (1994) demonstrated that rats exposed to stress-induced depression exhibited profound degeneration of cortical NA axons, and that chronic treatment with the antidepressant imipramine restored cortical NA axons. Other studies have reported that chemical ablation of NA-LC fibers using the selective LC neurotoxin DSP-4 is associated with the expression of a range of depressive-like behaviors, including increased despair on the forced swim test (FST; Harro et al., 1999), increased neophobia (Delini-Stula, Mogilnicka, Hunn, & Dooley, 1984), reduced social interaction (Cornwell-Jones et al., 1992), and impaired exploratory behavior (Harro, Oreland, Vasar, & Bradwejn, 1995; van den Buuse, Lambert, Fluttert, & Eikelis, 2001). Finally, the Wistar-Kyoto (WKY) rat, a strain that expresses hormonal and behavioral profiles that resemble those observed in depressive patients, shows altered expression of several genes that are related to NE synthesis in LC (Pearson, Stephen, Beck, & Valentino, 2006; Sands, Strong, Corbitt, & Morilak, 2000).

Given the association between NA–LC system deficits and depression, and the relationship between the NA-LC system and circadian rhythms, we asked whether manipulating the availability of light, a critical zeitgeber, can induce changes in NA–LC functioning and depression-like behavior in the rat (Gonzalez & Aston-Jones, 2006). We subjected rats to 6 weeks of light deprivation, a paradigm known to decrease the amplitude of the sleep-wake rhythm, and analyzed depressive-like behavior as well as cell viability of several systems associated with the pathophysiology of depression, including LC. We found that animals maintained in constant darkness showed significant depression-like behavior, as well as greater apoptosis in LC, compared to those maintained in standard 12:12 lighting conditions. The NA–LC neuronal apoptosis was associated with decreased NA boutons in frontal cortex. These findings were the first demonstration that light deprivation leads to NA cell apoptosis in LC and reduced NA innervation in cortex, and provided a possible mechanism through which light may influence mood.

3 THE SUPRACHIASMATIC NUCLEUS-DORSOMEDIAL HYPOTHALAMUS PATHWAY AS A RELAY OF LIGHT INFORMATION TO THE NA–LC SYSTEM

How might photic information influence the NA-LC system? As no known direct projections exist from retina to LC, light information must be relayed to LC via an indirect pathway. The SCN is a strong candidate in such a network, as this structure receives light information from retina (Gooley, Lu, Chou, Scammell, & Saper, 2001; Hannibal, Hindersson, Knudsen, Georg, & Fahrenkrug, 2002) and projects indirectly to LC (as reviewed above; Aston-Jones et al., 2001). SCN synchronizes many biological circadian rhythms to produce internal homeostasis with the solar day (Antle & Silver, 2005), and is considered the master pacemaker of the brain (van Esseveldt, Lehman, & Boer, 2000). Similar to the NA-LC system, there is strong evidence of dysregulation of SCN function in mood disorders. For example, Sprouse, Braselton, and Reynolds (2006) demonstrated that both the synthesis and release of arginine vasopressin (AVP), a major peptide of SCN, is reduced in depressed patients. Bernstein et al. (2005) showed that patients with depression have fewer nitric oxide synthase (NOS) neurons in SCN than their nondepressed counterparts, while others have reported decreased levels of NO metabolites in plasma of depressed patients (Chrapko et al., 2006). Additionally, melatonin receptor 1 (MT1) is specifically increased in SCN of patients with depression, and the abundance of MT1 receptors is positively correlated with disease duration, indicating that SCN MT1 receptors increase with the duration of depression (Wu et al., 2013). These clinical findings are supported by animal studies that have investigated SCN function in models of depression. Using chronic unpredictable stress (CUS), Jiang et al. (2011) demonstrated that expression of SCN PER2, an important clock gene that regulates circadian cycles (Panda et al., 2002), was dampened in animals that exhibited depression-like behavior. A related study demonstrated that other SCN circadian proteins, CLOCK and BMAL1, are also reduced following CUS (Jiang et al., 2013). Vagell, McGinnis, Possidente, Narasimhan, and Lumia (1991) showed that depression-like behavior in rats, induced by olfactory bulbectomy, is associated with an increase in cyclic- 3′,5′-adenosine monophosphate (cAMP) levels, specifically in SCN. Together, both clinical and preclinical findings point to a dysregulation in SCN function in depression, likely resulting in aberrant signaling in downstream structures and systems, including the NA–LC system. It should be noted that dysregulation of SCN may also affect mood via its other projection targets, for example its direct connection with lateral hebenula (Buijs, 1978), which is known to influence DA, 5HT, and NA systems (Christoph, Leonzio, & Wilcox, 1986; Herkenham & Nauta, 1979; Ji & Shepard, 2007).

However, here we restrict our focus to the SCN–LC/NA hypothesis. How is dysregulated SCN function conveyed to the NA–LC system? The DMH is anatomically well-positioned to act as a relay in the SCN– LC pathway, as DMH receives input from SCN (Leak & Moore, 2001) and projects directly to LC (Aston-Jones et al., 2001; Gompf & Aston-Jones, 2008). Our laboratory previously found that the same DMH neurons that receive input from SCN project to LC, and that disruption of this pathway results in dysregulated circadian rhythm and arousal (Aston-Jones et al., 2001; Gonzalez & Aston-Jones, 2006).

The phenotype of these DMH “relay cells” is not yet known. It is interesting to note, however, that DMH contains a discrete set of neurons that produce the arousal-related neuropeptide orexin/hypocretin, and orexin/hypocretin cell function has been reliably shown to be dysregulated in depression. For example, levels of orexin-A in CSF are significantly lower in MDD patients compared to healthy controls (Brundin, Björkqvist, Petersén, & Träskman-Bendz, 2007), and MDD patients exhibit blunted diurnal variation of orexin CSF levels (Salomon et al., 2003). Among patients with MDD, the severity of depressive symptomology is negatively correlated with orexin levels in CSF (Brundin et al., 2007) as well as orexin mRNA in the blood (Rotter, 2011). These findings are largely reflected in the preclinical literature, with reduced orexin levels observed in genetic models of depression (Allard, Tizabi, Shaffery, Trouth, & Manaye, 2004; Taheri et al., 2001) as well as in animals exposed to various stress protocols that induce depression-like behavior (James et al., 2014; Lutter et al., 2008; Nocjar, Zhang, Feng, & Panksepp, 2012; for full review, see James, Campbell, & Dayas, 2016, Yeoh, Campbell, James, Graham, & Dayas, 2014). Together, these findings point to a possible role for DMH orexin/hypocretin cells in relaying information in the SCN-LC pathway, a hypothesis that ongoing studies in our laboratory are seeking to confirm.

We posit that under normal circumstances, the retina provides photic input to the SCN-DMH-LC pathway (collectively, the PRAM pathway; see Fig. 1), which in turn regulates brain circuits that support normal mood. In some forms of depression, photic input onto the PRAM network is disrupted, resulting in pathological dysregulation of sleep, arousal, and mood. Disruption of PRAM pathway signaling likely has numerous etiologies: in the case of SAD, PRAM signaling is likely to be reduced owing to environmental conditions (low light levels). Other forms of depression may be associated with decreased light exposure due to social withdrawal or irregular sleep patterns, thereby augmenting depressive symptomology in patients. Other forms of depression may arise from dysfunction in LC, with normal functioning of upstream parts of the PRAM circuit (SCN and DMH). However, even in those cases we propose that manipulations of the PRAM pathway might be effective as a treatment to offset such dysfunctions in downstream circuit elements and normalize mood. Thus, we argue that by restoring normal signaling in the PRAM pathway, or manipulating this circuit to compensate for dysfunctions downstream from the circuit, it may be possible to ameliorate some of the symptoms that characterize depression. Indeed, there are some reports that “light therapy” has some therapeutic efficacy for some forms of depression, particularly SAD. However, therapies designed to restore normal signaling in the PRAM pathway would ideally seek to selectively enhance input from those cells in the retina that provide specific input onto SCN, without affecting other visual pathways. Such an approach, we believe, would provide a more nuanced and targeted approach for the restoration of normal circadian and arousal states in depressed individuals.

FIGURE 1.

The photic regulation of arousal and mood (PRAM) pathway. A midsaggital section of a rat brain illustrating loci involved in PRAM-associated depressive-like behavior. Structures of PRAM pathway are indicated by blue boxes. SCN, suprachiasmatic nucleus; DMH, dorsomedial hypothalamus; LC, locus coeruleus. Melanopsin-containing retinal ganglion cells (RGCs) are putative the origin of the PRAM pathway, as they project directly onto SCN, their major target, which then projects indirectly onto LC via DMH. Dysregulation of the PRAM pathway may lead to the decreased noradrenergic-LC signaling (red arrows) that is associated with depression

4 MELANOPSIN-EXPRESSING, INTRINSICALLY PHOTOSENSITIVE RETINAL GANGLION CELLS: A DIRECT ROUTE TO THE PRAM PATHWAY?

In recent years there has been great interest in a subset of ∼4–5% of retinal ganglion cells (RGCs) that express the photopigment melanopsin (Schmidt, Chen, & Hattar, 2011). These specialized cells intrinsically respond to light independently of traditional photoreceptor signaling, have robust effects on the regulation of circadian rhythms, and are a prominent candidate cell-type for the nonimage forming functions of light such as mood regulation. Known as intrinsically photosensitive RGCs (ipRGCs), these cells project directly to their major target, SCN (Baver, Pickard, Sollars, & Pickard, 2008; Guler et al., 2008; Hattar et al., 2006; Sollars et al., 2003; Hattar et al., 2002; Provencio et al., 2000), and thus are likely to directly influence PRAM pathway function. Indeed, human fMRI studies indicate that 480 nm blue light, which maximally activates ipRGCs, increases the bilateral activity in the area consistent with LC (Vandewalle et al., 2007). Further, a number of studies show striking effects of melanopsin on mood. For example, in humans, two gene-association studies demonstrated a link between melanopsin and SAD (Roecklein et al., 2009), and between melanopsin and seasonal alterations in sleep timing (Roecklein et al., 2012). Four hundred eighty nanometers blue light, which maximally activates ipRGCs, is superior to green light in modulating emotional brain responses (Vandewalle et al., 2010). Further, depression-like behavior in mice induced by aberrant light cycle conditions is dependent on melanopsin-containing cells (LeGates et al., 2012). In this way, melanopsin-containing ipRGCs represent an attractive candidate cell type involved in the photic regulation of circadian entrainment, arousal, and regulation of mood.

5 A CHEMOGENETIC APPROACH TO TARGETING MELANOPSIN IPRGCS FOR THE TREATMENT OF DEPRESSION

DREADDs are a powerful new chemogenetic tool that is transforming neuropsychiatric research. Allowing for the direct and selective control of neuronal populations, DREADDs modulate neuronal function by targeting specific intracellular signaling pathways. DREADDs are modified human GPCRs that are solely activated by a specific orthogonal ligand, clozapine-N-oxide (CNO), which itself is otherwise pharmacologically inert (Armbruster, Li, Pausch, Herlitze, & Roth, 2007). Three DREADDs commonly used are hM3Dq, hM4Di, and rM3Ds that activate Gq, Gi, and Gs signaling, respectively. Effectively, DREADDs create a powerful “lock and key” approach for targeted neuronal regulation (for example, see Mahler et al., 2014; Vazey and Aston-Jones (2014)). Different from other genetic approaches, such as optogenetics, DREADDs are based upon human proteins and therefore have minimal potential for immunogenic reactivity and require no optical fiber implant, and thus arguably provide greater translational potential for psychiatric disorders. However, despite their translational appeal, current technology requires the direct injection of a viral-vector into the brain, creating multiple surgical risks and potential prohibitive expense for the patient. The eye presents an alternate route of direct administration of the viral-vector, as the retina is the only part of the central nervous system that can be accessed without the risk of invasive brain surgery.

Intravitreal injection (IVI), which bypasses the blood–retinal barrier, allows for highly targeted vector delivery. With the increased use of injectable drugs for the treatment of blinding diseases, IVIs have become one of the most common ophthalmic procedures (Ho & Kuo, 2007). IVIs are rapid, the overall risk for serious complications is low, and only local anesthesia is required. As such, it is feasible to conceive of a therapy whereby DREADDs are transfected specifically into melanopsin ipRGCs via IVI, allowing for regulation of the PRAM pathway with local (eye drops) or systemic (oral dosing) CNO. Clearly further preclinical work is required to demonstrate the therapeutic utility of such an approach, a line of work our laboratory is currently actively pursuing.

6 SUMMARY

The PRAM pathway presents a novel circuit for the regulation of mood in both humans and other species. Evidence indicates that melanopsin-containing ipRGCs are the origin of the PRAM pathway, as they project directly onto SCN, their major target, which then projects indirectly onto LC via DMH. Because access to the retina is simple and carries little risk, IVI-injected DREADDs provide an exciting opportunity for manipulating the PRAM pathway in humans, without the need for brain surgery. While further research is required to understand the precise mechanism(s) undying the PRAM pathway-mediated regulation of mood, we have nonetheless demonstrated a potentially critical role for this pathway in such processes. The recent development of DREADDs, allowing for reversible control of the PRAM pathway, creates exciting new directions for the treatment of depression in humans.