GSK3β, CREB, and BDNF in peripheral blood of patients with Alzheimer’s disease and depression.

Prog Neuropsychopharmacol Biol Psychiatry. 2014 Apr 3;50:83-93. (Impact factor 4.03)

GSK3β, CREB, and BDNF in peripheral blood of patients with Alzheimer’s disease and depression.


Pláteník J1, Fišar Z2, Buchal R1, Jirák R3, Kitzlerová E3, Zvěřová M3, Raboch J3.
  • 1Institute of Medical Biochemistry and Laboratory Diagnostics, First Faculty of Medicine, Charles University in Prague and General University Hospital in Prague, Kateřinská 32, 121 08 Prague 2, Czech Republic.
  • 2Department of Psychiatry, First Faculty of Medicine, Charles University in Prague and General University Hospital in Prague, Ke Karlovu 11, 120 00 Prague 2, Czech Republic. Electronic address:
  • 3Department of Psychiatry, First Faculty of Medicine, Charles University in Prague and General University Hospital in Prague, Ke Karlovu 11, 120 00 Prague 2, Czech Republic.



BACKGROUND: Glycogen synthase kinase-3β (GSK3β), cAMP-response element-binding protein (CREB) and brain-derived neurotrophic factor (BDNF) play critical roles in neuronal survival, synaptic plasticity and memory and participate in the pathophysiology of both depressive disorder and Alzheimer’s disease (AD).

METHODS: This study was designed to determine the association of GSK3β activity, CREB activity and BDNF concentration in peripheral blood of patients with AD with or without depressive symptoms and in depressive patients without AD. GSK3β activity in platelets, CREB activity in lymphocytes and BDNF concentration in plasma, platelet-rich plasma or platelets were measured in 85 AD patients (36 of whom displayed co-morbid depressive symptoms), 65 non-AD patients with depressive disorder and 96 healthy controls. AD patients were clinically assessed for stage of dementia, cognitive impairment and severity of depressive symptoms. Depressive patients were clinically assessed for severity of depression.

RESULTS: We observed increased CREB activity and GSK3β activity in AD with depressive symptoms or in AD at mild stage of dementia. Decreased BDNF concentration was found in platelet-rich plasma of AD patients at moderate to severe stages of dementia or in AD without depressive symptoms. An association was revealed of the severity of cognitive impairment with the increase of GSK3β in the platelets of AD patients with mild dementia. In depressive patients, a lower concentration of phosphorylated GSK3β was associated with a higher severity of depression. Association was confirmed between severity of depression, CREB activation, and BDNF concentration in drug-naïve depressive patients.

CONCLUSION: Our data demonstrated that AD is accompanied by increased CREB activity in lymphocytes and a decreased concentration of BDNF in platelet-rich plasma. The decreased BDNF concentration appears to correlate with moderate to severe stages of dementia in AD. Observation of decreased phosphorylation of GSK3β in platelets of both AD patients with depressive symptoms and depressive patients after treatment confirms the role of increased GSK3β activity in the pathophysiology of both AD and depressive disorder. Associations were confirmed between AD and platelet GSK3β activity, lymphocyte CREB activity and plasma BDNF. CREB activity and platelet BDNF concentration seems to be related to depressive disorder.

KEYWORDS: Alzheimer’s disease; BDNF; CREB; Depressive disorder; GSK3β

PMID: 24334212



Alzheimer’s disease (AD) and depressive disorder are major neuropsychiatric diseases. Depression occurs as co-morbidity in up to half of AD patients and undoubtedly adds to their suffering. Both conditions may have some pathophysiological mechanisms in common, such as disturbance of signaling pathways involved in neurogenesis and neuronal plasticity/survival. Many components of these signaling pathways also perform other functions outside brain. In this study we chose three signaling proteins CREB, GSK3β, and BDNF, previously widely implicated in pathogenesis of AD and depression, and measured them in peripheral blood of patients with AD and/or depression. We wondered 1) whether the mean levels of these markers would reflect the supposed changes of these proteins in the patients‘ brains, 2) whether they would correlate with the severity of cognitive impairment and depressive symptoms, and 3) whether depression would act additively with AD.

The 85 AD patients (36 with co-morbid depression), 65 non-AD depressive patients, and 96 healthy volunteers were recruited during a period of six years (2007-2013). The AD patients were sub-grouped according to presence/absence of co-morbid depression, and also according to stage of dementia (mild or moderate to severe). The depressive patients were examined at two different time points: upon admission and then again after about one month of treatment. 75% of the depressive patients responded to the treatment. From the beginning of the study we were aware that the antidepressant pharmacotherapy can affect the measured parameters more than the disease itself. However, the drug naïve (untreated for at least 6 weeks before admission) patients were rare – only 16 such patients could be recruited.

CREB (cAMP-response element-binding protein) is an ubiquitous, constitutively expressed transcription factor that has been widely studied in synaptic plasticity underlying learning and memory, as well as in the neuronal survival pathways. Its role has been implicated in both neurodegenerative and affective diseases, and also in the mechanism of action of antidepressant drugs.

We measured the CREB activity by Western immunoblotting in peripheral mononucleated cells. The blots were probed with antibody against the Ser133-phoshorylated CREB (pCREB, the active, transcriptionally competent form), then stripped and reprobed with antibody against total CREB. The results were reported as pCREB/CREB ratio.

The CREB protein activation is generally supposed to promote neuronal plasticity and survival; therefore, in AD or depression a lower-than-normal activity would be expected. Surprisingly, however, we found rather increased CREB activity in AD patients. The increase was limited to those with mild, but not moderate to severe dementia, and was more pronounced in patients with co-morbid depression. We believe the increased CREB activity represents a compensatory activation of cellular defence mechanisms in the early stages of the disease.

In non-AD depressed patients the dominant finding was an increased CREB activity after treatment, most likely reflecting the effects of medication.

GSK3β (Glycogen synthase kinase-3β) is a widespread, often rather pro-apoptotic protein kinase involved in variety of cellular processes. It participates in many core aspects of AD pathophysiology, such as Aβ neurotoxicity, regulation of synaptic plasticity, neuronal survival, neurogenesis and neuroinflammation. This is the kinase held responsible for the hyperphosphorylation of tau protein.

We employed Western immunoblotting to measure the enzyme in blood platelets, where it is remarkably abundant and likely participates in the regulation of platelet activation during blood clotting cascade. The levels of total GSK3β and Ser9-phosphorylated GSK3β (pGSK, inactive form of the enzyme) were determined separately and combined to the pGSK/GSK3β ratio as a measure of GSK3β activity.

In the AD patients a rather low pGSK/GSK3β ratio (i.e., high GSK3β activity) was observed; only in the patients with co-morbid depression was this alteration pronounced enough to reach a statistical significance. Factor analysis further indicated that presence of depressive symptomes is responsible for the association of GSK3β activity with AD. In addition, a total GSK3β in platelets correlated with disease severity in AD patients with mild dementia.

An increased mean GSK3β activity in comparison to controls was also found in the subgroup of non-AD depressive patients after treatment. Before treatment, the levels of inactive pGSK (supposedly reflecting the activity of upstream kinase Akt) inversely correlated with severity of depression. Interestingly, only in the non-responders a significant association between GSK3β activity and disease severity was found before treatment, suggesting that further investigation of platelet GSK3β could lead to a predictor of response to therapy.

BDNF (brain-derived neurotrophic factor) is essential for brain development, synaptic plasticity and neuronal survival. This neurotrophin is also stored in platelets and altered ability of platelets to store and/or release BDNF was described in depressive patients. Information in the literature on the serum BDNF levels in AD patients is controversial.

We used ELISA to estimate BDNF separately in blood plasma and in lysate of platelets with plasma (platelet-rich plasma, PRP); from these values we calculated the amount of BDNF in platelets.

BDNF concentration in PRP or platelets was found to decline with age even in healthy controls (Fig. 1). A significant decrease of mean BDNF in PRP was observed in AD patients with moderate to severe dementia; the difference disappeared after calculation of BDNF in platelets and could best be explained by a lower number of platelets that was seen in AD patients compared to controls.


Fig. 1: BDNF concentration in platelets plotted against age for healthy controls. Decline of BDNF may possibly reflect a decreased neuroprotective action of BDNF and higher susceptibility to neurodegenerative disorders that comes with age.


No changes in mean BDNF concentrations were observed in the non-AD depressed patients. Higher BDNF in plasma was associated with lower severity of residual depression in the subgroup of responders after several weeks of treatment. Only in the small group of 16 drug-naïve patients we could observe an inverse correlation between severity of depression and BDNF in PRP or platelets.

Overall, we could confirm our introductory hypothesis that many features of the pathophysiological processes occurring in the brains of patients with AD and/or depression are reflected in the peripheral blood. Association of at least some biochemical markers with disease severity could be seen in both AD and depression. CREB and GSK3β data indeed indicated additive effects of AD and depression.