Yin-Yang 1, a player regulated systematic inflammatory involved in cognitive impairment of depression

1. Introduction

Major depressive disorder (MDD), is a severe affective disorder characterized by a group of symptoms including persistent low mood, worthlessness or guilt, anhedonia, inattention, changes in appetite, mental slow movement or agitation, sleep disturbance, or suicidal thoughts, imposing a heavy social burden(Karrouri et al., 2021). Depression is often accompanied by the clinical symptom of cognitive impairment, which resulting in a decline in quality of life and psychosocial function(Frampton, 2016). The DSM-5 has defined cognitive impairment as one of nine diagnostic criteria for depression, including impaired thinking, memory, attention and decision making. Cognitive dysfunction in MDD mainly displays in learning, attention, memory, processing speed and executive function (Varghese et al., 2022, Bortolato et al., 2016). Because of its potential contribution to the full recovery of function in MDD patients, the remission of cognitive dysfunction is increasingly being promoted as an important goal in the treatment of major depression(Lam et al., 2014). Additionally, increasing studies have shown that cognitive disorders may have complex interactions with emotional symptoms. For example, the aggravation of depressive symptoms may negatively affect cognitive function(McDermott and Ebmeier, 2009)and the changes in cognitive function may be related to the response of patients to antidepresant treatment(Culang et al., 2009). Therefore, the study of the pathophysiological mechanism of cognitive impairment in depression is conducive to further understanding the pathogenesis of depression, and may find new therapeutic targets for clinical diagnosis and treatment.

Transcription factor Yin Yang 1 (YY1), firstly identified in 1991, is widespread expression in mammals. Previous studies showed that YY1 plays an important role in early neurodevelopment(Shiu et al., 2016),(Dong and Kwan, 2020),(Zambrano et al., 1997). Some recent studies revealed that YY1 also plays a role in inflammation regulation (Darcy et al., 2020). Zhang et al. found that YY1 was gradually upregulated in lipopolysaccharide (LPS)-stimulated BV2 microglia, which was dependent on the transactivation function of nuclear factor kappa B (NF-κB). Furthermore, LPS-induced activation of NF-κB signaling pathway and increase of interleukin 6 (IL-6) expression were significantly inhibited in microglia of YY1 knockout mice. Meanwhile, YY1 enhanced the binding of NF-κB to the IL-6 promoter, reducing the H3K27ac modification of the IL-6 promoter, and finally increasing IL-6 transcription(Zhang et al., 2018). Our previous study found that YY1 expression was increased in the peripheral blood mononuclear cell (PBMC) from MDD patients in single-cell RNA sequencing analysis compared with healthy volunteers(Lu et al., 2021). These studies suggested that YY1 is closely associated with both central and peripheral inflammation. Inflammation was proved to be associated with depression and cognitive impairment. Reichenbergde et al. demonstrated that cytokines could mediate affective and cognitive disturbances in humans, intravenous administration of S. equinum endotoxin significantly increased circulated TNF-α, soluble TNF receptors, IL-6 and IL-1 receptor antagonists in healthy volunteers, which elevated the level of depressed mood and impaired cognitive functions such as verbal memory and non-verbal memory(Reichenberg et al., 2001). Many other studies have also reported that inflammation may be attributed to depression and its cognitive impairment(Zhao et al., 2022, Jin et al., 2020) Thus, to investigate the role of YY1 in depression and cognitive impairment is of great importance.

Examination of the brain using positron emission tomography (PET) may lead to a better understanding of the underlying pathophysiology of MDD, which makes it possible to demonstrate changes in the regional brain metabolic rate (rCMRGlu) of glucose in vivo(Hendler et al., 1997). Recent studies have shown that brain hypometabolic pattern is closely related to the type and severity of cognitive deficits(Laforce et al., 2018). The roles of peripheral inflammation on glucose metabolism in the parahippocampal and rhinal directly mediate the effects of inflammation on spatial memory impairment(Harrison et al., 2014). However, little is known about the changes of brain glucose metabolism before and after treatment in depression animal model and its relationship with inflammation and behaviour.

Vortioxetine is a novel multimodal antidepressant with 5-HT receptor modulator and serotonin transporter (SERT) inhibitor properties(Pehrson et al., 2018). A number of preclinical and clinical data have demonstrated the antidepressant effects of Vortioxetine(Alvarez et al., 2012, Mork et al., 2012), and many preclinical studies have suggested that vortioxetine can improve certain aspects of MDD related cognitive impairment, such as processing speed, executive function, and memory(du Jardin et al., 2014, Li et al., 2015). Baune et al. also reported that vortioxetine had beneficial effects on mood and cognitive function of patients with major depression(Baune et al., 2018). It is noteworthy that vortioxetine may have anti-inflammatory effects due to its structural similarity to non-steroidal anti-inflammatory drugs(Bayram et al., 2018). Furthermore, in one study of human monocytes/macrophages, vortioxetine showed antioxidant activity and anti-inflammatory effects, polarizing macrophages into other phenotypes(Talmon et al., 2018). In addition, the antidepressant effects of vortioxetine were enhanced by the anti-inflammatory celecoxib(Fourrier et al., 2018). Vortioxetine may, contrary to what was previously believed, play an antidepressant and improve cognitive function in part through its anti-inflammatory effects, Therefore, it is necessary to study the immunomodulatory mechanism of Vortioxetine in depressed individuals.

Therefore, this study aims to elucidate the changes of YY1-related inflammatory pathways before and after treatment of depression and its relationship with cognitive function. First, we measured the level changes of proinflammatory factors, YY1 and cognitive function in peripheral blood of MDD patients before and after treatment, and explored their correlations. Second, three groups were constructed, including control, CUMS and VOR groups, and the changes of behavior, glucose metabolism and mRNA level of YY1 and proinflammatory factors were detected. In conclusion, our study reveals the role of YY1-related inflammatory pathway in MDD and cognitive impairment, and provides theoretical support for seeking new targets for clinical diagnosis and treatment.

2. Material and Methods

3. Result

3.2 Cytokine characteristics of subjects

To investigate level of plasma YY1 and inflammatory factors, peripheral plasma of MDD patients and healthy control was collected to perfrom ELISA assay. The result indicated that plasma YY1 level increased in MDD patients (0.4982 ± 0.2836 ng/ml) compared with CTR(0.0274 ± 0.0215 ng/ml, p<0.001) and reduced after treated with vortioxetine (0.2279 ± 0.1969 ng/ml, p<0.001) (Figure 1A). Plasma level of IL-1β was increased in MDD (565.4724 ± 405.3974 pg/ml) than CTR (371.5978 ± 153.6342 pg/ml, p<0.001) and reduced after treatment ((187.9717±104.7408 pg/ml, p<0.001, Figure 1B). Plasma level of IL-6 was increased in MDD (1287.4829±789.9515 pg/ml) than CTR (608.2849 ± 506.9919 pg/ml, p<0.0001) while showed no significant change after treatment (912.2897 ± 707.8910 pg/ml Figure 1C, p=0.547).

• Download figure
• Open in new tab

Figure 1.

Changes of plasma level of Yin Yang 1 (YY1), IL-1β and IL-6 levels in patients with major depressive disorder (MDD, n = 40) and healthy controls (CTR, n = 40). (A) Comparison of YY1 in MDD and CTR. (B) Comparison of IL-1β in MDD and CTR; (C) Comparison of IL-6 in MDD and CTR; (D) Correlations between plasma YY1 levels and score of visual learning; (E) Correlations between plasma YY1 levels and score of reasoning and problem solving in MDD; (F) Correlations between plasma YY1 levels and plasma level of IL-1β in the patients with MDD. (*p < 0.05, **p < 0.01, ***p < 0.001, **** p<0.0001)

In order to determine whether the plasma level of YY1 was associated with cognition and cytokines, Pearson correlation analysis was conducted in MDD patients. The result showed that plasma YY1 was negatively correlated with the score of visual learning (p = 0.008, rho = -0.416, Fig. 1D) and the score of reasoning problem solving (p = 0.046, rho = -0.318, Fig. 1E), Plasma YY1 level was positively correlated with plasma IL-1β level (p < 0.0001, rho = 0.541, Fig. 1F). while these correlations were absent in control (all p > 0.05).

3.3 Behaviour and Cytokine characteristics of rats

The CUMS rats showed significant anxiety-like behavior in open field test (Fig. 2B-D), which showed significant decrease total distance in CUMS group in comparison to CTR and VOR group (p=0.0005, p=0.1777, respectively) and in center zone entries(p<0.0001 and p=0.0247, respectively). Depression-like behavior also performed, Sucrose preference test (CUMS vs. CTR p=0.0004, CUMS vs. VOR p=0.0034, Figure 2M) and forced swim test (CUMS vs. CTR p=0.0001, CUMS vs. VOR p=0.0020, Figure 2L) and and learning and memory behavior test (Elevated plus maze, CUMS vs. CTR p=0.0095, CUMS vs. VOR p=0.0079, Figure 2H-I; Y-maze, CUMS vs. CTR p=0.0398, CUMS vs. VOR p=0.0033, Figure 2J-K; Morris water maze, CUMS vs. CTR p=0.0001, CUMS vs. VOR p=0.0008, Figure 2E-G), and VOR treatment could partly reverse these performance.

• Download figure
• Open in new tab

Figure 2.

Vortioxetine treatment ameliorates depression-like and cognitive dysfunction behaviors in CUMS rat. (A) Diagram of experimental schedule; (B-D) Open field test. Trajectories of open field test (B); Total distance in open field test (C); Center zone entries in open field test (D). (E-G) Morris water maze test. Representative swimming paths during the spatial probe test (E); The latency to the platform of 5-day acquisition training in Morris water maze test (F, * means CTR vs. CUMS; # means VOR vs. CUMS); (G) The time in target quarter. (H and I) Elevated plus maze. Representative swimming paths during elevated plus maze (H). The time spent in open arm (I); (J and K) Y maze test. The representative swimming paths during Y maze test (J); the time spent in the new arm during the test (K); (L) Forced swim test; (M) Sucrose preference test. (n = 7 per group, ^# p<0.05, *p < 0.05, **p < 0.01, ***p < 0.001, n.s.: non-significant difference)

3.5 Glucose metabolism in brain regions of CUMS rats

Altered glucose metabolism in mPFC, striatum, amygdala, hippocampus and entorhinal cortex were observed (Fig. 3A). Tilted frontal-side view showed the location of altered brain regions (Fig. 3B). The results indicated that glucose metabolism of left mPFC (p=0.0102, Fig. 3C), striatum (p=0.0001, Fig. 3D), amygdala (p=0.019, Fig. 3E), hippocampus (p=0.0045, Fig. 3F) and entorhinal cortex (p=0.0006, Fig. 3G) was significantly decreased in CUMS compared with that of CTR group, and decreased in amygdala (p=0.0003, Fig. 3E), hippocampus (p=0.001, Fig. 3F) and entorhinal cortex (p=0.0262, Fig. 3G) after VOR treatment, while no significant changes in left mPFC and left striatum after VOR treatment (p=0.2569, p=0.088, respectively). CUMS rats showed significantly decrease glucose metabolism than controls and elevated after VOR treatment in the right mPFC (p<0.0001, p=0.0122, respectively, Fig. 3C), striatum (p=0.0245, p=0.0510, respectively, Fig. 3D), amygdala (p=0.012, p=0.01, respectively, Fig. 3E), hippocampus (p=0.0029, p=0.0030, respectively, Fig. 3F) and entorhinal cortex (p=0.003, p=0.006, respectively, Fig. 3G)

• Download figure
• Open in new tab

Figure 3.

The glucose metabolic changes of hippocampus (HIP), amygdala (AMY), striatum (STR), median prefrontal cortex (mPFC) and entorhinal cortex (EC) in control group (CTR), depression model (CUMS group) and antidepressant (vortioxetine) treated group (VOR group). (A) Representative ¹⁸F-FDG PET images of HIP, AMY, STR, mPFC and EC from CTR, CUMS and VOR group; (B) Tilted frontal-side view of brain regions whose glucose metabolism altered; (C-G) Glucose metabolism was markedly decreased in CUMS group compared with CTR in bilateral mPFC (C), STR (D), AMY (E), HIP (F) and EC (G), while vortioxetine treatment restored glucose metabolism of right mPFC (C), AMY (E), HIP (F) and EC (G) in CUMS group. Results are showed as means ± S.D. (n = 7 per group, *p < 0.05, **p < 0.01, ***p < 0.001)

3.4 mRNA level of Cytokine in hippocampus, frontal cortex and striatum

Hippocampus, frontal cortex and striatum mRNA levels of YY1, IL-6, IL-1β and NF-kB-mRNA levels were measured in different groups by qPCR. Firstly, in frontal cortex area, we did detected significant increase of IL-1β (p=0.0033, p=0.0026, respectively) and YY1-mRNA (p=0.0321, p=0.0177, respectively) changes in CUMS than CTR or VOR group (Fig 4A-D). As for striatum, CUMS groups showed siginificantly increased of IL-1β (p=0.0079, p=0.0051, respectively, Fig 4E), YY1-mRNA expression compared to CTR or VOR group (p=0.0389, p=0.0002, respectively, Fig 4H), NF-kB-mRNA expression was elevated in CUMS than CTR, whereas do not descend after VOR treatment (p=0.0148, p=0.4170, respectively, Fig 4G). In hippocampus the IL-1β (p=0.0057, p=0.0017, respectively), NF-kB (p=0.0170, p=0.0362, respectively) and YY1-mRNA(p=0.0102, p=0.0017, respectively) was significant increase in CUMS in comparison with control or VOR group (Figure 4I, K, L), IL-6-mRNA expression level was also elevated in CUMS than control, while Vortioxetine treated showed no significant changes (p=0.0442, p=0.0697, respectively, Figure 4J).

• Download figure
• Open in new tab

Figure 4.

Changes of IL-1β, IL-6, NF-kB and YY1-mRNA expression in Frontal cortex (A-D), Striatum (E-H), hippocampus (I-L) among CTR (n = 8), CUMS (n = 8) and VOR group (n = 8); (M) Representative western blot image of NF-kB in nuclear level and cytosolic in the hippocampus among CTR, CUMS and VOR groups; (N) The ratio of nuclear NF-kB/nuclear TBP; (O) The ratio of cytosolic NF-kB/cytosolic β-actin) (n = 8 per group, *p < 0.05, **p < 0.01, ***p < 0.001, n.s.: non-significant difference)

4. Discussion

In the present study, the role of YY1 in depression associated inflammation and cognitive impairment was investigated in MDD patients and CUMS rodent model. The results showed increased plasma level of YY1 combined with cognitive impairment and elevated IL-1β and IL-6 level in MDD patients, which could be restored by vortioxetine treatment. Meanwhile, the result of rodents study is consistent with clinical findings that CUMS rats showed depressive and cognitive impairment which were reversed by vortioxetine administration. Among related depressive and connition brain area ie: frontal cortex, hippocamus, striatum, Increased hippocampus YY1, NF-κB, IL-6 and IL-1β in CUMS rats, which were suppressed by treated with vortioxetine. Protein level of NF-κB was increased in the nuclear level of CUMS than control and VOR group in hippocampus, while cytosolic level changes in the opposite way.¹⁸F-FDG microPET/CT scanning revealed decreased glucose uptake in hippocampus, amygdala, striatum and entorhinal cortex of CUMS group which reserved by vortioxetine.

Notably, both MDD patients and CUMS model rats showed recovery in mood and cognitive function after treatment with vortioxetine, which suggests that there may be common underlying mechanisms between depression and cognitive impairment. In fact, the structures of the medial temporal lobe, including the amygdala, hippocampus and entorhinal cortex, are involved in a variety of tasks, such as the production of emotional behaviors, the formation and storage of emotional memories, the consolidation of explicit memories of emotional arousal events(Bombardi, 2014, LeDoux, 2007, Eichenbaum and Lipton, 2008), and spatial information processing and memory, which is consistent with the patient’s clinical symptoms and our findings in animals. In our study, ¹⁸F-FDG scanning indicated decreased glucose uptake in hippocampus, amygdala and entorhinal cortex of CUMS group rats, and they increased after subsequent treatment. Many previous studies have shown significant changes in glucose metabolism in several brain regions in MDD patients, but the results have been inconsistent due to confounding factors such as age, medication and course of disease. Mayberg et al. reported reduced glucose metabolism in the frontal, hypothalamus, and medial frontal cortex(Hendler et al., 1997). Meanwhile, decreased brain metabolism in cingulate gyrus, superior frontal gyrus, rectum gyrus and orbital gyrus was also observed in MDD patients(Wei et al., 2016). The antidepressant effects of electric shock therapy were associated with increased metabolism in the left subgenicular anterior cingulate gyrus and hippocampus(McCormick et al., 2007). Recent studies have shown that brain hypometabolic pattern is closely related to the type and severity of cognitive deficits(Laforce et al., 2018). Previous study also demonstrated that chronic stress could lead to decreased glucose metabolism in brain regions of rats, such as amygdala and hippocampus (Wei et al., 2018, Van Laeken et al., 2018). Typhoid vaccination caused inflammation in healthy volunteers, which led to a sharp drop in MTL glucose metabolism and selectively impaired human spatial memory, which is similar to what we found in animals.

We also found changes in the levels of IL-1β and YY1 in peripheral blood of depressed patients before and after treatment, and positively correlation between YY1 and IL-1β, whereas, the same changes were found only in the hippocampus of the three available animal brain regions both in the mRNA changes and glucose metabolism level, therefore, the inflammatory mediated by the hippocampal YY1-NF-κB-IL-1β pathway is at least partially involved in the improvement of cognition and mood in depressed individuals by vortioxetine. A previous study has found the level of TNF-α, IL-17, IL-6, and IL-1β were markedly decreased in YY1-deficient mice with collagen-induced arthritis(Kwon et al., 2018). YY1 was upregulated in lipopolysaccharide (LPS) -stimulated BV2 microglia dependent on NF-κB activity. Furthermore, YY1 knockdown also inhibited LPS-induced the activation of NF-κB signaling BV-2□cells. Common binding sequences of NF-κB have been identified in promoter regions of several cytokine genes, including TNF-α and IL-1β(Muller et al., 1993), and inhibition of NF-κB inhibits the synthesis and release of IL-1β and TNF-α(Kuo et al., 2000). Though in vitro experiments on BV-2 cells showed that YY1 knockdown significantly inhibited LPS-induced IL-6 levels but not IL-1β(Zhang et al., 2018), and YY1 positively regulated IL-6 transcription by binding to the promoter region of the IL-6 gene as well(Lin et al., 2017), nonetheless, our results indicated that IL-1β appears to play a more prominent role in the treatment of depressed individuals with vortioxetine. Actually, brain IL-1β is necessary for depressive symptoms induced by chronic mild stress, and chronic IL-1β exposure also induces depressive behavior and reduces hippocampal neurogenesis(Goshen et al., 2008). Menin deficiency increases NF-κB-induced IL-1β levels in astrocytes, contributing to depression-like behavior(Leng et al., 2018). Il-1β knockdown in the hippocampus significantly alleviated IPS-induced memory deficits and anxiety- and depression-like behaviors in mice(Li et al., 2017). In clinical studies, the use of endotoxin in healthy volunteers resulted in increased levels of anxiety and depression in subjects, which is related to the transcription factor NF-κB(Cho et al., 2019).

This is a multi-dimensional study that comprehensively illustrates the role of YY1/NF-κB mediated inflammatory pathway in depression. But the study still has some limitations. Firstly, this study is an aggregated crossover study, but we did not verify the findings in rats with depression in human brain. Additional evidence may be needed to support the role of YY1/NF-κB -mediated inflammatory pathway in cognitive function and mood changes in depression, which may affect the overall interpretation of our study results. Secondly, the representative sample size of clinical samples is unknown, and our patients are mainly outpatients and inpatients, which may limit the universality of our study results. Thirdly, the relatively small sample size of this study may result in some potential relationships not being discovered. In the future, multi-center studies should be established to test the generality of these findings by recruiting more patients with different severity levels for long-term follow-up. Forthly, we have not further explored how inflammation causes depression-like behavior and cognitive impairment, and though studies have shown that depression patients have increased inflammation in both central and peripheral locations, the specific mechanism and link between them may need to be explored further in subsequent experiments. Finally, depression is a very heterogeneous disease. We did not classify patients with depression according to the characteristics of depression (including anhedonia and suicidal ideation), but only focused on the characteristics of cognitive impairment of these patients, and we included patients with moderate or severe depression. The lack of a more specific grouping may make the statistical results of our study less convincing to a certain extent. In addition, we only constructed the CUMS model, and did not verify such changes in other depression-like models such as chronic immobilization stress model and chronic social defeat stress.

Overall, our study suggests that YY1-NF-κB -IL-1β inflammatory pathway may play an essential part in vortioxetine’s effect on mood change and cognitive improvement in depression, and may be associated with changes in glucose metabolism in the mood regulation and cognitive related brain regions. These findings provide new evidence for the inflammatory mechanisms of depression and cognition repairment, and this multi-dimensional study may be beneficial to the development of diagnostic or therapeutic tools for depression in the future.

Competing interests

The authors declare no competing interests.