Glucocorticoid may influence amyloid β metabolism in patients with depression

Highlights

• The biological link between depression and Alzheimer's disease is still unclear.

• Altered serum cortisol and amyloid β levels were found in patients with depression.

• Cortisol levels at admission correlated with serum amyloid β levels 1 year later.

• Cortisol may affect amyloid metabolism over prolonged periods of time.

Abstract

Epidemiological studies have demonstrated that depression may be a risk factor for Alzheimer's disease (AD); however, the biological mechanisms of the transition from depression to AD are still not clear. Changes of amyloid β protein (Aβ) metabolism and increased glucocorticoid (GC) levels have been found in both depression and AD. Moreover, several studies in animal models have demonstrated that GC administration changes Aβ metabolism. To reveal whether GC affects amyloid metabolism in patients with depression, we evaluated serum levels of Aβ40, Aβ42 and cortisol at admission in 187 inpatients with major depressive disorder (MDD) and 224 healthy comparisons. Additionally, we re-evaluated the serum levels of Aβs in 27 patients with MDD 1 year later. The results of multiple regression analyses revealed that serum cortisol and Aβ levels are not correlated at the time of admission. However, serum cortisol levels at admission correlated with serum Aβ42 levels and Aβ40/Aβ42 ratio 1 year later. These findings suggest that increased cortisol in patients with MDD may influence the metabolism of Aβ over prolonged periods of time.

Introduction

Epidemiological studies have demonstrated that depression may be a risk factor for Alzheimer's disease (AD) (Devanand et al., 1996, Wilson et al., 2002, Ownby et al., 2006, Irie et al., 2008, Byers and Yaffe, 2011, Barnes et al., 2012, da Silva et al., 2013, Diniz et al., 2013). This association has been shown even in cases where depression occurs long before the onset of AD (Green et al., 2003, Ownby et al., 2006). Recent studies have supported the idea that early-onset depression also increases the risk of developing AD (Geerlings et al., 2008, Byers and Yaffe, 2011, da Silva et al., 2013). Moreover, the number of depressive episodes (Kessing and Andersen, 2004, Dotson et al., 2010) and severity of depressive symptoms (Wilson et al., 2002, Saczynski et al., 2010) may increase the risk of dementia including AD. Whether depression is a prodromal symptom of AD or risk factor that contributes to its onset is still debated, although previous reports suggest the latter.

Although the biological mechanism underlying the transition from depression to AD is still not clear, alterations of amyloid β protein (Aβ) metabolism in depression may be the possible background (Kita et al., 2009, Byers and Yaffe, 2011, Baba et al., 2012, Namekawa et al., 2013).

Aβ is the major component of senile plaques in AD and is a 40- or 42-amino acid peptide cleaved from the amyloid precursor protein (APP) by β- and γ-secretase. Cerebrospinal fluid (CSF) levels of Aβ42 are reduced in patients with AD (Schroder et al., 1997, Andreasen et al., 1999), but increase at the early stage of the disease (Jensen et al., 1999). These variations may be associated with the selective deposition of Aβ42 in the brain. However, studies of plasma Aβ levels in subjects with AD have been contradictory (Fukumoto et al., 2003, Sobow et al., 2005, Ertekin-Taner et al., 2008, Lopez et al., 2008). The reason for this discrepancy remains unclear, although blood or CSF Aβ levels could alter according to AD stage (Kawarabayashi et al., 2001). Although the results have not been consistent, several epidemiological studies have indicated that changes in peripheral levels of Aβ, especially higher Aβ40/Aβ42 ratios, represent a risk for future onset of AD (Mayeux et al., 1999, Mayeux et al., 2003, van Oijen et al., 2006, Graff-Radford et al., 2007, Schupf et al., 2008, Lambert et al., 2009).

Peripheral Aβ levels in elderly patients with depression have been found to be inconsistent (Pomara et al., 2006, Sun et al., 2007, Sun et al., 2008). We previously reported that serum Aβ40/Aβ42 ratios were significantly higher in patients with major depressive disorder (MDD) than in healthy comparisons, and this difference is seen in both elderly and younger subjects (Kita et al., 2009, Baba et al., 2012). We suggest that Aβ metabolism may be affected in depression, and this may indicate why even early-onset depression is a risk factor for developing AD. However, the mechanism of the association between depression and Aβ is still unclear.

Hyperactivity of the hypothalamus-pituitary-adrenal (HPA) axis, which is well described in MDD (Pariante and Lightman, 2008, Marques et al., 2009), is also observed in AD and results in increased glucocorticoid (GC, cortisol in primates) levels in blood and CSF (Davis et al., 1986, Martignoni et al., 1990, Rasmuson et al., 2001, Rasmuson et al., 2002, Crochemore et al., 2005, Krishnan and Nestler, 2008, Popp et al., 2009, Popp et al., 2015, Caraci et al., 2010, Chi et al., 2014).

Stressful stimuli induce hyperactivity of the HPA-axis and result in an increase of GC secretion from the adrenal cortex. Normally, the GC increase is transient due to a negative-feedback system (Sapolsky et al., 1986). In patients with depression and AD, impairment of the negative-feedback system has been suggested to occur. Excessively increased GC may damage neural cells in the hippocampus, inducing hippocampal volume loss and memory impairment (Butters et al., 2008, Sierksma et al., 2010, Byers and Yaffe, 2011).

Interestingly, some studies have demonstrated that GC administration increases Aβ and tau pathology in a mouse model of AD (Green et al., 2006), and that chronic GC administration decreases plasma Aβ42 levels in a macaque (Kulstad et al., 2005).

Combining these previous results, we hypothesized that increased levels of GC may lead to pathological changes in amyloid metabolism in patients with depression. We evaluated serum Aβ40, Aβ42, and cortisol levels in patients with MDD and healthy comparisons, and analyzed the correlations between Aβ and cortisol levels. Additionally, patients were followed and their Aβ re-evaluated 1-year later to reveal any long-term influence of cortisol on Aβ. This study is a part of the Juntendo University Mood Disorder Project (JUMP).