Childhood insulin resistance and neural stem cell dysfunction in psychiatric disorders: Role of de novo lipogenesis and treatment perspectives

Abstract

Insulin plays a crucial role in the metabolic priming and proliferation of neural stem cells (NSCs). However, insulin resistance (IR) is associated with impaired NSC proliferation and cognitive dysfunction, which are the hallmarks of psychiatric disorders (PDs). In addition to insulin, de novo lipogenesis (DNL) also plays an essential role in NSC proliferation and function as it supplies fatty acids for membrane phospholipid synthesis and cell signaling. However, enhanced DNL is associated with lipid/fatty acid accumulation, IR, and impaired NSC proliferation. Intriguingly, data from lipidomic studies suggest that DNL could be enhanced before the onset of classical symptoms in patients with PDs. Further, evidence suggests that patients with PDs may develop IR during childhood or before adolescence; therefore, DNL could be enhanced preceding the development of IR. Regarding treatment, while most antidepressants and antipsychotic drugs have been shown to further deteriorate IR and stimulate DNL, various adjunctive drugs/therapies, including chemical, physical, and stem cell therapy, which have shown promising success in treating PDs, reduce DNL while enhancing insulin sensitivity, NSC proliferation, and cognitive function in laboratory animals. Preliminary clinical outcomes and future prospects of these adjunctive drugs/therapies, especially stem cell therapy in treating PDs including schizophrenia and depression, are discussed.

Key Words: Childhood insulin resistance; Neural stem cell dysfunction; De novo lipogenesis; Psychiatric disorders; Anti-inflammatory agents; Stem cell therapy

Core Tip: Childhood insulin resistance (IR) is a potential risk factor for developing psychiatric disorders (PDs). Although insulin regulates neural stem cell (NSC) proliferation and function, IR is associated with impaired NSC proliferation and cognitive dysfunction, which are prominent features of PDs. While the mechanisms underlying IR and NSC dysfunction in PDs remain unclear, intracellular lipids/fatty acids synthesized via de novo lipogenesis could be the primary mediators. Since psychotropic drugs further deteriorate IR and stimulate de novo lipogenesis, prospects of various adjunctive therapies, especially stem cell therapy, in treating IR in schizophrenia and depression are discussed.

INTRODUCTION

Psychiatric disorders (PDs) are complex brain disorders with heterogenous etiology[1]. Among various PDs, a high prevalence rate has been reported for schizophrenia, depression, bipolar disorder, anxiety, and attention-deficit hyperactivity disorder, among others[2]. Although PDs are diagnosed based on the appearance of psychiatric symptoms, brain abnormalities, including impaired neural stem cell (NSC) proliferation, synaptic dysfunction, and reduced grey and white matter volume of the cortex and hippocampus, usually develop before the onset of psychiatric symptoms[3-5]. In addition, cognitive dysfunction and metabolic abnormalities have been reported to precede the onset of psychiatric symptoms in patients with PDs[6]. Although various neurological disorders such as Alzheimer’s disease (AD) and Parkinson’s disease also display the aforementioned brain abnormalities and metabolic dysfunction[7,8], the extent of severity is high due to neuronal degeneration, and the illness onset is late; whereas, PDs could be diagnosed early, usually in adolescence or before, and show less severe brain pathologies[3-5].

Recent evidence suggests that childhood insulin resistance (IR) could be a potential risk factor for developing PDs[9-11]. Although animal studies have shown that insulin regulates NSC proliferation, cognitive behavior, and energy homeostasis[12-15], IR is associated with impaired NSC proliferation, cognitive dysfunction, and reduced energy production[15-18], which are the hallmarks of patients with PDs[9,10,19-21]. These findings suggest that reducing IR could be an effective strategy in improving cognitive function and global outcomes in patients with PDs.

Although several mechanisms have been shown to induce IR, they seem to be triggered by endogenous lipids/fatty acids synthesized via de novo lipogenesis (DNL)[22-26]. As shown in Figure 1, DNL produces mainly saturated fatty acids (SFAs) and monounsaturated fatty acids (MUFAs), which play an essential role in membrane phospholipid synthesis and cell signaling. However, enhanced DNL has been associated with elevated levels of SFAs and MUFAs and their increased incorporation into membrane phospholipids, and consequently, altered membrane fluidity, which has been reported in patients with PDs[25-29]. Several lines of evidence suggest that reduced membrane fluidity substantially increases the risk of developing IR and diabetes, which are the characteristic features of patients with PDs[6,9,10,24-26]. Further evidence suggests that the effect of SFAs and their ceramide derivatives (SFAs stimulate de novo ceramide biosynthesis as shown in Figure 2) on IR could be detrimental because they increase oxidative stress and inflammation by disrupting calcium (Ca²⁺) homeostasis, the endoplasmic reticulum (ER), and mitochondrial function[22,23,25,30,31]. The disrupted Ca²⁺ homeostasis, ER stress, mitochondrial dysfunction, and elevated oxidative stress coincide with the onset of cognitive and psychiatric symptoms in patients with PDs[32-35]. It should be mentioned that several studies, including our own, have reported that membrane polyunsaturated fatty acids are reduced, whereas SFAs and MUFAs and their ceramide derivatives are mostly increased and show a strong association with cognitive and psychiatric symptoms in patients with PDs[32,36-42].

Figure 1 Scheme of de novo lipogenesis and its role in endoplasmic reticulum stress, mitochondrial dysfunction, and pro-inflammatory cytokine production in psychiatric disorders. (1) Various non-carbohydrate metabolic intermediates could be used as substrates in gluconeogenesis to produce glucose, causing hyperglycemia in psychiatric disorders; (2) Excess glucose is oxidized by glycolysis to produce pyruvate, which is converted to acetyl-CoA; (3) Acetyl-CoA is converted to malonyl-CoA, and both are used together for fatty acid biosynthesis via de novo lipogenesis; (4) Excess saturated fatty acids and monounsaturated fatty acids synthesized via de novo lipogenesis can be converted to ceramides and triglycerides and can also be incorporated into membrane phospholipids, reducing membrane fluidity as well as inducing endoplasmic reticulum stress; (5) Mitochondrial dysfunction increasing the production of reactive oxygen species; and (6) Elevated reactive oxygen species can activate the nuclear factor kappa B pathway, which is the main pathway involved in stimulating the production of pro-inflammatory cytokines (such as interleukin 1 beta, interleukin 6, and tumor necrosis factor alpha). These pro-inflammatory cytokines have been shown to activate various cell signaling pathways involved in insulin resistance, as shown in Figure 2. TRP: Transporter; SOC: Store-operated calcium; TG: Triglycerides; Cer: Ceramides; MUFAs: Monounsaturated fatty acids; GLN: Gluconeogenesis; GLY: Glycolysis; DNL: De novo lipogenesis; SFAs: Saturated fatty acids; ER: Endoplasmic reticulum; ROS: Reactive oxygen species; NF-κB: Nuclear factor kappa B; IL: Interleukin; TNF: Tumor necrosis factor; NSC: Neural stem cells.

Figure 2 Role of pro-inflammatory cytokines and saturated fatty acids in insulin resistance and neural stem cell dysfunction in psychiatric disorders. A: Insulin activates insulin receptor (IR) A/B, IR substrate 1-5 (IRS1-5), and downstream rat sarcoma-extracellular signal-regulated kinase 1/2 and phosphoinositide 3 kinase-Akt/protein kinase B pathways to regulate various metabolic functions including glucose uptake (↑), glucose oxidation/glycolysis (↑), gluconeogenesis (↓), glycogen synthesis (↑), fatty acid synthesis (↓), fatty acid oxidation (↓), and protein synthesis (↑); B: Enhanced de novo lipogenesis can reduce membrane fluidity, which has been shown to reduce IR A/B activity and downstream phosphorylation of IRSs and associated metabolic functions; C and D: Elevated pro-inflammatory cytokines, including tumor necrosis factor alpha (C) and interleukin-6 (D), can activate c-Jun N-terminal kinase and Janus kinase-signal transducer and activator of transcription pathways, respectively, which reduce IRS activation and downstream metabolic functions, indicating insulin resistance; E: Elevated saturated fatty acids can directly induce insulin resistance by activating the membrane toll-like receptor 2/4 receptor and inducing endoplasmic reticulum stress or by increasing de novo biosynthesis of ceramides. Elevated ceramides can induce insulin resistance by inhibiting Akt, which increases glucose uptake by increasing membrane expression of glucose transporter 4, and also by increasing endoplasmic reticulum stress. IR A/B: Insulin receptor A/B; TNF: Tumor necrosis factor; IL: Interleukin; SFAs: Saturated fatty acids; TNFR: Tumor necrosis factor receptor; Rec: Receiver; TLR4: Toll-like receptor 4; Glut4: Glucose transporter 4; IRS1-5: Insulin receptor substrate 1-5; JNK: c-Jun N-terminal kinase; JAK: Janus kinase; STAT: Signal transducer and activator of transcription; SOCS: Suppressor of cytokine signaling; IKKβ: Inhibitory kappa B kinase beta; Ras: Rat sarcoma; ERK1/2: Extracellular signal-regulated kinase 1/2; PI3K: Phosphoinositide 3 kinase; PKB: Protein kinase B; ER: Endoplasmic reticulum; NSC: Neural stem cell.

DNL is an essential process, as genetic ablation of fatty acid synthase (FAS), the main enzyme of DNL, has been shown to induce embryonic and stem cell lethality[43-45]. However, enhanced DNL in adult animals has been associated with lipid/fatty acid accumulation, NSC dysfunction, metabolic abnormalities, and cognitive dysfunction[43,44,46]. Intriguingly, several lines of evidence suggest that DNL could be enhanced before the onset or diagnosis of psychiatric symptoms in patients with PDs[33,37-41], which, along with IR, could be another potential confounder associated with impaired NSC proliferation in PDs.

In this review, recent findings from basic and clinical research are discussed, which suggest that IR in PDs is primarily triggered by endogenous lipids/fatty acids synthesized via DNL. Once IR develops, it may further stimulate DNL (Figures 1 and 2). Therefore, IR and enhanced DNL seem to be the two faces of the same coin and together could be the most likely cause of impaired NSC proliferation, cognitive dysfunction, and impaired energy homeostasis in patients with PDs. Since psychotropic drugs further deteriorate IR and stimulate DNL[32,33], prospects of various adjunctive drugs/therapies, including chemical, physical, and especially stem cell therapy, in treating IR and NSC dysfunction in patients with PDs, particularly schizophrenia and depression, are also discussed. The literature cited in this article was searched using PubMed, Scopus, and Google, and the articles published primarily in English and occasionally in French or German were included. References were selected mostly from those published within the last 5 years, while the older references were included only when deemed necessary.

EVIDENCE FOR INSULIN IN PDS

Childhood or early life adversities play a major role in disposing individuals to develop metabolic, cognitive, and psychiatric diseases[1,10,11]. As mentioned above, longitudinal studies have reported that the great majority of patients with PDs can develop IR long before the onset of psychiatric symptoms, usually during childhood or before adolescence[10]. This suggests that IR could be an intrinsic trait in patients with PDs. In adult patients with schizophrenia, Steiner et al[47], while assessing IR and stress hormone levels, observed that IR and disrupted glucose homeostasis were present in a significantly high proportion of drug-naive patients with recent-onset psychosis. The authors suggested that this could be illness-related and not arise redundantly, due to pharmacotherapy, obesity, or hypothalamic-pituitary-adrenal axis activation; although, serum stress hormone levels could be elevated. Chouinard et al[9] reported that siblings or first-degree relatives of patients with psychosis may also display IR, which could be an intrinsic risk factor for psychosis. These findings have been confirmed in recent meta-analyses, which suggest that hormonal stress axis activation and lifestyle factors could also be the potential confounders associated with IR in PDs[48]. In support of this, reports suggest that plasma cortisol could be elevated in patients with psychosis[49], a condition that is associated with IR and ectopic fat deposition as a result of enhanced DNL[31,42,50].

Likewise, evidence suggests that a high proportion of patients with depression may develop IR before the diagnosis of classical symptoms[10,51]. In a longitudinal study, Perry et al[10] reported that insulin signaling is dysregulated from childhood or early adolescence in patients with depression. They also observed a positive association between IR and obesity at the age of 24 years, which is the time of diagnosis. Additionally, several studies have shown that IR is mostly deteriorated both in children and adolescents after treatment with both antidepressants and antipsychotic drugs[52]. This suggests that IR in PD could be an irreparable risk factor and requires more effective therapies/drugs.

MECHANISM OF IR IN PDS: ROLE OF DNL

Glucose is the main substrate for DNL, and evidence suggests that increased gluconeogenesis could be the major source of glucose for DNL in patients with PDs[24,31,33,53]. In support of this, several non-carbohydrate metabolic precursors, including citrate, lactate, glutamate, glutamine, serine, and others, which are used in gluconeogenesis, have been reported to be elevated in patients with PDs[54-57]. Since patients with PDs cannot utilize dietary glucose efficiently due to IR, elevated levels of these metabolic intermediates could be one of the potential causative factors associated with enhanced gluconeogenesis, hyperglycemia, and consequently increased DNL in PDs. An overview of gluconeogenesis, DNL, and other downstream pathways is given in Figure 1. Here, it is worth differentiating between lipogenesis and DNL; while lipogenesis is a general term used for lipid synthesis from fatty acids that are obtained either through the diet or synthesized via the de novo pathway from glucose, the term DNL is used for lipid synthesis from fatty acids produced only via the de novo pathway and is essential for the development and function of NSCs[24,25,46].

Although several mechanisms have been shown to induce IR, they appear to be triggered mostly by elevated endogenous lipids/fatty acids synthesized via DNL (Figure 1). Even high-fat diet-induced IR is triggered by lipids/fatty acids synthesized via DNL[22-25,31,33]. Since patients with PDs may develop IR during childhood or before adolescence, DNL could also be enhanced preceding the development of IR[9-11,24,51,53]. In support of this, several studies, including our own, have shown that the levels of SFAs, MUFAs, and their ceramide derivatives are increased in patients with recent onset PDs, which could be due to enhanced DNL[32,33,36-41]. In addition, the levels of ceramides are also elevated in the brain phospholipid from patients with PDs[42]. These findings, together with increased brain lactate levels, suggest that DNL could also be enhanced in the brain of patients with PDs[56,57].

While several studies have shown that both SFAs and MUFAs can induce IR, the effect of SFAs could be detrimental because they increase reactive oxygen species (ROS) and pro-inflammatory cytokine production[23,25,30,58]. On the other hand, elevated MUFAs usually do not increase the production of ROS and pro-inflammatory cytokines but can induce IR[22,59]. Therefore, one way or the other, patients with PDs will remain at risk of developing IR either due to elevated levels of SFAs or MUFAs. In addition, triglycerides (TG) and ceramide derivatives of SFAs and MUFAs, which are elevated in patients with PDs[42,48,55,60], can further increase oxidative stress, inflammation, and the severity of IR[31,61-63].

Regarding the mechanism(s) underlying lipid/fatty acid-induced IR, it has been shown that extracellular SFAs (especially palmitic acid) can trigger IR directly or indirectly by increasing pro-inflammatory cytokine production via activating the nuclear factor kappa B pathway and plasma membrane toll-like receptor 2/4. In addition, elevated intracellular SFAs (palmitic acid) can also induce IR by inhibiting sarcoplasmic ER Ca²⁺ pump and blocking Ca²⁺ release from the ER causing ER stress[30,31,33,63,64]. This can dramatically increase cytosolic Ca²⁺via entry through plasma membrane-bound store-operated Ca²⁺ channels (Figure 1). Although excess cytosolic Ca²⁺ is mainly sequestered and stored in the ER and mitochondria, during ER stress, cytosolic Ca²⁺ is diverted mostly towards mitochondria, causing mitochondrial dysfunction, a condition that leads to increased ROS production[30,31,64]. As mentioned before, recent studies have implicated ER stress and mitochondrial dysfunction in the pathophysiology of PDs, likely due to their well-established role in ROS and proinflammatory cytokine production. Elevated ROS increase the production of pro-inflammatory cytokines in various cells, including microglia, monocytes-macrophages, adipocytes, and liver, by activating the nuclear factor kappa B pathway[30,63,64]. While activated microglia play a key role in brain inflammation, activated monocytes-macrophages and adipocytes are the major players involved in the peripheral tissue inflammation through their secretion of various pro-inflammatory cytokines, including interleukin 1 beta (IL-1β), IL-6, IL-8, and tumor necrosis factor alpha (TNF-α). These pro-inflammatory cytokines are reportedly elevated in patients with PDs[65,66]. Although treatment with antidepressants and antipsychotic drugs has been shown to reduce the production of some of these cytokines, others, including IL-1β, IL-6, IL-8, and TNF-α, are only transiently and marginally affected[65,66].

Figure 2 shows various signaling pathways activated by pro-inflammatory cytokines and SFAs and their role in IR and NSC dysfunction. As shown in Figure 2A, insulin performs its action by activating plasma membrane-bound insulin receptor A/B (IR A/B), IR substrate 1-5 (IRS1-5), and downstream rat sarcoma-extracellular-regulated kinase 1/2 and phosphoinositide 3 kinase-Akt/protein kinase B pathways, which have been shown to regulate various metabolic functions including glucose uptake, oxidation/glycolysis, gluconeogenesis, glycogen synthesis, fatty acid synthesis/oxidation, and protein synthesis; whereas, IR has been shown to have opposite effects[12,15,18]. Evidence suggests that reduced membrane fluidity (Figure 2B) can itself reduce IR A/B activity and function, leading to reduced insulin sensitivity[33,67]. Pro-inflammatory cytokines TNF-α (Figure 2C) and IL-6 (Figure 2D) induce IR by activating the c-Jun terminal kinase and Janus kinase-signal transducer and activator of transcription pathways, which inhibit IRS2 and IRS1 activation (phosphorylation), respectively. IL-6 increases the expression of suppressor of cytokine signaling proteins, which mediates inhibition of IRS1 and also IR. In addition to pro-inflammatory cytokines, SFAs can also induce IR indirectly by increasing de novo ceramide biosynthesis (Figure 2E). Ceramides can induce IR in two ways, namely by inducing ER stress and by inhibiting Akt/protein kinase B, which facilitates glucose uptake by increasing translocation of glucose transporter 4 to the plasma membrane[31,58,62].

In Table 1, we have summarized various markers/products of DNL, precursors of gluconeogenesis, and various metabolic parameters (abnormalities) associated with the mechanism(s) discussed above, which are also illustrated in Figures 1 and 2. Although IR may affect the development and function of stem cells residing in distinct tissues, including bone marrow, liver, adipose tissue, and the brain, below we discuss the influence of IR on NSCs only.

Table 1 Markers of de novo lipogenesis, substrates of gluconeogenesis, and metabolic parameters/abnormalities in patients with psychiatric disorders.

Parameters	Psychosis	Depression
DNL markers
SFAs	Increased[32,36-39]	Increased[40,41]
MUFAs	Increased[32,36-39]	Increased[40,41]
PUFAs1	Reduced[32,36-39]	Reduced[40,41]
Ceramides	Increased[42]	Increased[42]
Triglycerides	Increased[48,60]	Increased[55,61]
GLN substrates
Lactate	Increased[56]	Increased[57]
Citrate	Increased[33]	Unknown
Pyruvate	Increased[54]	Unknown
GluAA	Increased[4,54]	Increased[55]
Metabolic parameters
Blood glucose	Increased[9,10,47,48]	Increased[10,51]
IR/insulin level	Increased[9,10,47,48]	Increased[10,51]
Obesity (BMI)	Increased[6,47]	Increased[10,51]
Diabetes	Increased[9,10,47]	Increased[10,51]

¹Obtained through diet and not synthesized via de novo lipogenesis in the body.

DNL: De novo lipogenesis; SFAs: Saturated fatty acids; MUFAs: Monounsaturated fatty acids; PUFAs: Polyunsaturated fatty acids; GLN: Gluconeogenesis; GluAA: Glucogenic amino acids; IR: Insulin resistance; BMI: Body mass index.

ROLE OF IR IN NSC DYSFUNCTION IN PDS

Over the years, it has been established that NSCs exist in several functionally distinct areas/neurogenic niches of the adult brain, including the subventricular zone around the lateral ventricle and subgranular in the dentate gyrus of the hippocampus[68]. In the adult brain, NSCs differentiate into three types of cells: Including neurons (via a process called neurogenesis), astrocytes, and oligodendrocytes (via a process known as gliogenesis). These three brain cell types are the major players that maintain structural organization and regulate functional plasticity of the brain throughout adult life[68-71]. IR affects both aspects of NSCs, that is, neurogenesis and gliogenesis.

Impaired neurogenesis

Neurogenesis is a dynamic process in which NSCs produce an enormous number of new neurons daily in the adult brain[69]. New neurons mature primarily into interneurons and participate in cellular and synapse circuit formation and in adaptation to various behavioral and environmental cues[70,71]. Patients with PDs display learning and memory deficits, which suggest that neurogenesis could be impaired[72,73]. Indeed, several studies have reported reduced neurogenesis in functionally distinct areas of the brain of patients with PDs.

In schizophrenia, postmortem studies have documented reduced expression of the cell proliferation marker Ki67 in the anterior dentate gyrus of the hippocampus, suggesting that neurogenesis could be reduced[74,75]. Another study found a significant reduction in polysialylated neural cell adhesion molecule-positive staining in the hilus regions of the hippocampus but not in the dentate gyrus, suggesting that the density of immature neurons could be reduced in schizophrenia[76]. Furthermore, reduced density of mature neurons has also been reported in the left dentate gyrus and in the cingulate cortices in schizophrenia, but not in the cornu ammonis 1-4 regions or subiculum of the anterior hippocampus[77-79]. A recent study also reported reduced cortical neuron number and neuron density in the Brodmann area 24 of the medial prefrontal cortex in the schizophrenic brain[80]. Weissleder et al[81] reported reduced NSC proliferation in the subependymal zone, which supplies new neurons to cortical and subcortical areas. This suggests that cortical neurogenesis could be reduced in patients with schizophrenia. Finally, while several studies have reported reduced neuronal size in patients with depression, neuronal density was also either reduced or unchanged[82-84].

Insulin plays a crucial role in neurogenesis by stimulating the proliferation and maturation of NSCs with a concomitant improvement in cognitive behaviors and a decrease in the inflammatory cues in laboratory animals[12-14,16]. Insulin has been shown to mediate these effects both in vitro and in vivo by activating specific IR A/B and IRSs, which themselves have been reported to be highly expressed in the neurogenic niches of the adult brain[12,85]. In addition, insulin signaling also plays a crucial role in the self-renewal, proliferation, and differentiation of NSCs during development[12,85]. Conversely, IR or impaired insulin signaling has been shown to have a detrimental effect on neurogenesis, learning, and memory in laboratory animals[16,86], which could be the most likely scenario in patients with PDs[72,73].

Impaired gliogenesis

Glial cells, especially astrocytes, act like ‘glue’ in the brain. They provide chemical and biological support to the neurons and regulate the formation of both inhibitory and excitatory synapses[87]. Astrocytes are identified in the brain by their high expression of a protein called glial fibrillary acid protein (GFAP). A significant proportion of NSCs express GFAP in the neurogenic niches of the brain[68,70]. This suggests that astrocytes develop from the same neuronal cell lineage as neurons. Over the years, several studies have analyzed astrocyte density and expression of astrocyte-specific markers in the brain of patients with PDs. The findings are briefly discussed below.

In schizophrenia, reduced astrocyte density has been observed in various anatomically and functionally distinct regions of the brain including the dentate gyrus, cingulate cortex, motor cortex, nucleus accumbens, basal nuclei, and substantia nigra[88,89]. It should be noted that a few studies have reported no change or an increase in astrocyte density in other regions, including the temporal and frontal cortex and amygdala[89]. Altogether, while it appears that astrocyte proliferation could be affected differentially in a region-dependent manner in schizophrenia, the findings in patients with depression are more consistent with regard to reduced astrocyte density. Using histological/immunohistological methods, several authors have reported reduced proliferation or loss of astrocytes in the prefrontal cortex, hippocampus, and other regions of the brain in patients with depression[82,84,88-92].

The density of oligodendrocytes is also significantly affected in the brain of patients with PDs[79,90,93,94]. Oligodendrocytes are the major component of the white matter and provide support and insulation to the axons of certain types of pyramidal neurons in the central nervous system[95]. Oligodendrocytes wrap around the axons of pyramidal neurons, forming a myelin sheath, a process called myelination, which increases the axon’s diameter and is periodically interrupted. Intriguingly, myelinated axons show substantially increased pulse propagation velocity compared to that of unmyelinated axons[96]. Thus, it is believed that myelination is crucial for information processing and retrieval speed within the neural circuits that facilitate sensory, motor, and higher-order cognitive functions[95,96]. Reduced myelination is well established and is considered a potential contributing factor in the pathophysiology of PDs. In this regard, several studies have reported reduced density of oligodendrocytes in the hippocampus and cortex in patients with both schizophrenia and depression[79,82,90,93,92], with concomitant disruption in information processing and retrieval speed[93,94,97-99].

The evidence discussed above suggests that the proliferation and differentiation of NSCs into neurons and glial cells could be impaired in patients with PDs; however, the underlying mechanism(s) remain unclear. IR could be a potential mediator, as insulin alone or in association with growth factors (discussed below) has been shown to regulate the proliferation of NSCs, and IR itself is associated with reduced proliferation[100,101]. Regarding the role of DNL in NSC dysfunction, IR coincides with enhanced DNL, and evidence suggests that enhanced DNL leads to lipid/fatty acid accumulation, which could be associated with reduced neurogenesis and learning-memory deficit[43,44,46,102]. In support of this, mice expressing mutant FAS (the main enzyme of DNL) with increased activity had impaired proliferation of NSCs in the hippocampus along with cognitive deficit, most likely due to fatty acid accumulation in NSCs and development of lipogenic ER stress[43]. In another experiment, human embryonic stem cell-derived NSCs expressing mutant FAS showed reduced proliferation[43,46]. Further, it was shown that tissue-specific deletion of FAS in mice significantly reduced NSC proliferation and altered the polarity of apical and radial glia cell progenitors[102]. Together, the above findings suggest that while DNL is essential for neurogenesis, fatty acid accumulation as a result of enhanced DNL could be associated with reduced neurogenesis, reduced oligodendrocyte density, and cognitive abnormalities in patients with PDs.

Growth factor deficit

Growth factors are master regulators of NSC proliferation, maturation, and survival. In addition, growth factors also regulate metabolic functions, including glucose and lipid metabolism and energy homeostasis, directly or in association with insulin[103,104]. In the last few years, several studies have reported reduced expression of over a dozen growth factors in patients with PDs. Among the notable growth factors, which are reduced, are brain-derived neurotrophic factor, insulin growth factor, fibroblast growth factor, epidermal growth factor, nerve growth factor, and others[105-108]. While these studies provide further evidence for impaired NSC proliferation and/or function in patients with PDs, IR/diabetes is reportedly associated with the reduced expression of many of these growth factors in otherwise mentally healthy individuals[109,110]. In cell culture studies, growth factor treatment has been shown to improve insulin sensitivity, hyperglycemia, and glucose tolerance[111-113].

Although psychotropic drugs have been shown to significantly restore circulating levels of various growth factors, levels of some growth factors either remain unaltered or only slightly increased. For instance, Zhang et al[114] reported that serum epidermal growth factor was significantly reduced in drug-naïve patients with psychosis and was not increased by antipsychotic treatment alone or by a combined treatment with electroconvulsive therapy. Moreover, as mentioned above, almost all antidepressants and antipsychotic drugs induce IR, obesity, and nonalcoholic fatty liver disease (NAFLD)[51,52,115-118], which are associated with the reduced growth factor expression[109,110]. Thus, additional research is needed to identify new effective treatments for reducing IR and increasing NSC proliferation and function in patients with PDs.

PROSPECTS OF CHEMICAL, PHYSICAL, AND STEM CELL THERAPY IN PDS

Psychotropic drugs are the first line of treatment for PDs; however, their long-term use has been shown to induce or deteriorate pre-existing IR, leading to the development of obesity, diabetes, and NAFLD[115-117]. The mechanism(s) underlying these metabolic abnormalities remain unclear; however, evidence suggests that enhanced DNL leading to lipid/fatty acid accumulation could be the most likely mechanism(s) involved[33,118]. In support of this, several studies, including our own, have shown that antipsychotic treatment increases the levels of erythrocyte SFAs, MUFAs, and plasma TG[32,38,119,120] and triggers the development of obesity and NAFLD in patients with psychosis[115-117,121]. Since SFA and MUFA composition of erythrocytes has been used to assess the extent of DNL in health and diseases, an increase in erythrocyte SFA and MUFA levels by treatment with psychotropic drugs suggests that DNL could be enhanced. Intriguingly, while no clinical trials have been conducted to directly target IR in PDs, several adjunctive drugs/therapies, which have shown promising success in reducing psychiatric symptoms in patients with PDs, also reduce IR and enhance NSC proliferation and function. Among these are chemical therapy, including anti-inflammatory agents and antioxidants, aerobic/physical therapy, and stem cell therapy, which are discussed below.

Chemical therapy

In the last two decades, several anti-inflammatory agents and antioxidants have been used to reduce inflammation and improve the therapeutic efficacy of psychotropic drugs in patients with PDs. As shown in Table 2, these agents include aspirin, N-acetylcysteine, minocycline, pregnanolone, estrogens, raloxifene, curcumin, pioglitazone, celecoxib, and w-3 polyunsaturated fatty acids. Addition of these agents to the clinically approved doses of antidepressants or antipsychotic drugs has been shown to reduce symptoms of psychosis and depression in patients with PDs[122-141]. However, in patients with depression, evidence suggests that some of these agents, including minocycline, estrogens, and raloxifene, may not be effective or may even worsen the symptoms; thus, further studies are needed to evaluate their safety in treating PDs.

Table 2 Effects of selective adjunctive drugs/therapies on symptoms of psychosis and depression in patients with psychiatric disorders.

Drugs/therapies	Psychosis1	Depression2
Aspirin	Reduced[122]	Reduced[130]
N-acetylcysteine	Reduced[123]	Reduced[131]
Minocycline	Reduced[122]	No effect[132]
Pregnenolone	Reduced[122]	Reduced[133]
Estrogens	Reduced[122,124]	Deteriorated[134]
Raloxifene	Reduced[122]	No effect[135]
Curcumin	Reduced[125]	Reduced[136]
Pioglitazone	Reduced[126]	Reduced[137]
Celecoxib	Reduced[127]	Reduced[138]
ω3-PUFAs	Reduced[128]	Reduced[139]
Aerobic exercise	Reduced[129]	Reduced[140]
Resistance exercise	May reduce[129]	Reduced [141]
MSC therapy	Unknown	Unknown

¹Measure of Positive and Negative Syndrome Scale score.

²Measure of Hamilton Depression Rating Scale total scores.

ω3-PUFAs: ω-3 polyunsaturated fatty acids; MSC: Mesenchymal stem cell.

On the other hand, as shown in Table 3, the chemical agents discussed above also have potent antidiabetic properties as they reduce IR while improving cognitive behavior and increasing NSC proliferation/neurogenesis in laboratory animals[142-184]. Together, the above evidence suggests that therapeutic agents that regulate DNL and/or IR could be effective in enhancing the proliferation and function of NSCs in patients with PDs. Since patients with PDs may develop IR from the childhood or adolescence stage, early intervention with an appropriate adjunctive drug could be more effective in increasing NSC proliferation and treatment outcomes in patients with PDs.

Table 3 Effects of selected adjunctive drugs/therapies on insulin resistance, neural stem cell proliferation/neurogenesis, and cognition in animals and/or human subjects.

Drugs/therapies	Insulin resistance	Neurogenesis	Cognition
Aspirin	Reduced[142]	Increased[157]	Improved[157]
N-acetylcysteine	Reduced[143]	Increased[158]	Improved[172]
Minocycline	Reduced[144]	Increased[159]	Improved[173]
Pregnenolone	May reduce[145]	Increased[160]	Improved[174]
Estrogens	Reduced[146]	Increased[161]	Improved[175]
Raloxifene	Reduced[147]	Increased[162]	Improved[176]
Tamoxifen	Increased[148]	Increased[163]	Improved[177]
Curcumin	Reduced[149]	Increased[164]	Improved[164]
Pioglitazone	Reduced[150]	Increased[165]	Improved[178]
Celecoxib	Reduced[151]	May reduce[166]	Improved[179]
ω3-PUFAs	Reduced[152]	Increased[167]	Improved[180]
SIRT1-A	Reduced[153]	Increased[168]	Improved[181]
Aerobic exercise	Reduced[154]	Increased[169]	Improved[182]
Resistance exercise	May reduce[155]	Increased[170]	Improved[183]
MSC therapy	Reduced[156]	Increased[171]	Improved[184]

ω3-PUFAs: ω-3 polyunsaturated fatty acids; SIRT1-A: Sirtuin 1-agonist; MSC: Mesenchymal stem cell.

Apart from the chemical agents discussed above, sirtuin 1 (SIRT1) agonists/activators (Table 3) have shown tremendous therapeutic potential in laboratory animals[153,168,181]. SIRT1 is an NAD⁺-dependent deacetylase that regulates gene expression by deacetylating proteins such as histones and various transcription factors. SIRT1 possesses potent anti-inflammatory properties, which are influenced by factors like the NAD⁺/NADH ratio, binding partners, and post-translational modifications. SIRT1 deficiency has been associated with the development of metabolic syndrome and cognitive and neurodegenerative diseases. Recent evidence suggests that SIRT1 is reduced in patients with PDs; therefore, it could be a potential target for developing effective therapies/drugs for treating various ailments in PDs[185,186]. Indeed, in the last one to two decades, several SIRT1 agonists have been developed and tested in laboratory animals. These agonists have been shown to increase SIRT1 expression, reduce IR, and reduce neurodegeneration while increasing neurogenesis and cognitive outcomes in model animals, as shown in Table 3[153,168,181]. These properties make them worthy of conducting clinical trials in PDs.

Physical/aerobic therapy

Emerging evidence suggests that aerobic exercise as an add-on treatment can significantly increase the effectiveness of psychotropic drugs to alleviate the symptoms of depression and psychosis in patients with PDs[129,140,141]. In patients with depression, aerobic exercise improves global cognitive function and reduces depression in older adults with mild cognitive impairment[182]. The mean effect of global cognitive function is increased with higher exercise frequency. In patients with schizophrenia, aerobic exercise has also been shown to improve various domains of cognition, including global cognition, working memory, social cognition, and attention/vigilance[183]. These cognition-enhancing effects of aerobic exercise could be due to increased neurogenesis/NSC proliferation, as well as reduced IR and DNL in patients with PDs[154,169,187-189].

In addition, recent evidence suggests that resistance exercise could also improve cognitive behavior, increase neurogenesis, and reduce IR in laboratory animals and/or human volunteers[155,170,183,189]. These pleiotropic effects of aerobic and resistance exercise are in support of the findings suggesting that patients with PDs may develop tissue hypoxia on or before the onset of psychiatric symptoms[190]. Intriguingly, chronic hypoxia has been shown to disrupt NSC proliferation and differentiation in laboratory animals, likely due to lipid/fatty acid accumulation as a consequence of enhanced DNL[191,192]. In conclusion, increased efficacy of psychotropic drugs when combined with resistance exercise in PDs could be a result of the cumulative effect of reduced IR, reduced DNL, and enhanced NSC proliferation.

Stem cell therapy

Stem cells, especially those isolated from mesenchyme of bone marrow or human umbilical cord, have been used extensively in model animals and human subjects for ameliorating inflammation, IR/diabetes, and cognitive symptoms[193]. Mesenchymal stem cells (MSCs) are also referred to as multipotential stromal cells, mesenchymal stromal cells, or mesenchymal progenitor cells. They have been shown to differentiate into specific cell types under in vitro-specified conditions and in vivo after implantation[193]. MSCs can be isolated from distinct tissues, including umbilical cord, endometrial polyps, menstrual blood, bone marrow, and adipose tissue. At present, MSCs are being used in numerous clinical trials for various diseases[193]. Here, the outcomes of MSC therapy in model animals of diabetes and PDs are discussed.

Stem cell therapy reduces IR and lipid abnormalities: Emerging data from recent animal model and clinical studies suggest that MSC therapy can have multiple effects in a diabetic environment[193-195]. Along with regulating immune cell proliferation and function, MSCs can reduce peripheral IR, halt beta-cell destruction, preserve residual beta-cell mass, promote beta-cell regeneration and insulin production, support islet grafts, and reduce lipid accumulation and DNL[194,195].

In mice, MSC therapy has been shown to reduce IR and enhance glucose uptake by peripheral tissues such as skeletal muscle, liver, and adipose tissue, while restoring glycemic control and beta-cell function and reducing the risk of type 2 diabetes-related complications[195,196]. MSC therapy reduces peripheral tissue IR via phosphoinositide 3 kinase-dependent phosphorylation of IRS-1, which in turn increases glucose transporter 4 and IR expression on the cell membrane and reduces activation of stress-induced serine kinases, such as c-Jun terminal kinase 1 and extracellular-regulated kinase 1[197-199]. Also, several studies have shown that MSCs from different sources can differentiate into glucose-responsive insulin-producing beta cells in vitro and in vivo[194]. Further, MSCs secrete various trophic/growth factors such as vascular endothelial growth factor, fibroblast growth factor, angiopoietin-1, and hepatocyte growth factor, which can profoundly improve beta-cell function[200].

Regarding the influence of MSC therapy on lipid abnormalities, several studies have shown that infusion of MSCs, especially from umbilical cords, can significantly reduce hyperglycemia and elevate hepatic transaminases and lipid contents, including TGs, total cholesterol, and low-density lipoprotein cholesterol[195,201]. It can also significantly reduce liver injury, as suggested by reduced lipid accumulation and decreased hepatic steatosis. Regarding the mechanism, MSC therapy has been found to exert dual effects on lipid metabolism, with increased expression of fatty acid oxidation-related genes and reduced expression of lipogenesis-related genes mediated by the upregulated hepatocyte nuclear factor 4 alpha-carboxylesterase 2 pathway[195,201].

Stem cell therapy enhances NSC proliferation and cognitive outcome: Stem cells, including both MSCs and NSCs, have been extensively used in various animal model studies for improving cognition and reducing neurodegeneration[202-206]. In the amyloid-β (Aβ)-related animal model of AD, the Wnt signaling pathway has been suggested to impair neurogenesis. MSC infusion in Aβ-treated animals significantly increases NSC proliferation/neurogenesis in the dentate gyrus of the hippocampus at 2 weeks and 4 weeks[202]. Likewise, in another study, MSC infusion significantly reduced neurodegeneration and increased neurogenesis and synaptic function[203]. Similar results have been obtained with both human umbilical cord and adipose-derived MSCs, suggesting that intravenous infusion of human MSCs could be an effective approach for increasing neurogenesis and synaptic function in human patients. In another study, McGinley et al[204] implanted human NSCs in a murine model of AD and observed significant improvement in cognitive function along with reduced load of Aβ peptide/plaques[204].

Aging is a potential risk factor for developing PDs and is accompanied by a significant decline in NSC proliferation (neurogenesis) and cognitive function. Recent studies showed that MSCs and exosome implantation increased neurogenesis and improved cognitive function while reducing inflammation in aged animals[205]. Another study showed that MSC-derived exosome implantation significantly increased neurogenesis and improved cognitive function in mice treated with repeated injections of methamphetamine, a chemical that is used to induce psychosis in humans and animals[206]. These results suggest that MSC therapy could be effective in improving metabolic, neurogenic, and cognitive function in patients with PDs.

Outcome of stem cell therapy in an animal model of PDs: In a recent study, Gobshtis et al[171] transplanted bone marrow-derived MSCs in the intracerebroventricular region of a ketamine-induced murine model of schizophrenia. The authors observed that MSCs successfully engrafted and survived for up to 3 months following transplantation. The animals showed significant improvement in social novelty preference and pre-pulse inhibition with a concomitant increase in hippocampal neurogenesis. They also observed an independent aging effect on behavior and neurogenesis, which was attenuated by MSC treatment. These collective findings suggest that long-term effects during aging could be dependent on the self-renewal potential of NSCs.

In another study, You et al[207] intravenously infused human umbilical cord-derived MSCs (hUC-MSCs) with a potent immunomodulatory effect on an animal model of schizophrenia. The authors observed that neuroinflammation along with peripheral TNF-α elevation was associated with schizophrenia-relevant behaviors in amphetamine-sensitized mice, and hUC-MSC infusion significantly reduced schizophrenia-relevant behaviors and neuroinflammatory cues. They concluded that a single hUC-MSC infusion had long-term beneficial effects via regulatory T-cell induction and secretion of IL-10 in this mouse model of schizophrenia.

Tfilin et al[208] analyzed the therapeutic potential of MSCs in the rat Flinders sensitive line (FSL), an animal model for depression. The authors gave an intracerebroventricular injection of culture-expanded and 1,1’-dioctadecyl-3,3,3’,3’-tetramethylindocarbocyanine perchlorate-labeled bone marrow-derived MSCs to FSL rats. They observed that MSC-transplanted FSL rats had significant improvement in behavioral performance. They also observed that neurogenesis was increased in the ipsilateral dentate gyrus and hippocampus, and correlated with behavioral performance. Although these findings suggest that MSCs may serve as a novel option for treating depression, clinical trials with MSCs in patients with depression remain to be conducted.

CONCLUSION

Insulin signaling is required for priming NSCs for glucose uptake, metabolism, and energy production. However, IR is associated with detrimental effects on NSC proliferation and function. Since patients with PDs may develop IR before adolescence, impaired insulin signaling could be a potential causative factor associated with reduced NSC proliferation, cognitive dysfunction, and reduced energy production in patients with PDs.

In addition to insulin, DNL also plays an essential role in NSC proliferation and function as it supplies SFAs and MUFAs for membrane phospholipid synthesis and cell signaling. Intriguingly, insulin signaling regulates DNL, whereas IR is associated with enhanced DNL. Evidence suggests that IR is primarily triggered by excess endogenous lipids/fatty acids (SFAs, MUFAs, TG, ceramides) synthesized via DNL. However, once IR develops, it can further stimulate DNL, leading to lipid/fatty acid accumulation, impaired NSC proliferation, cognitive dysfunction, and reduced energy production, which appears to be the most likely scenario in patients with PDs. Therefore, reducing IR could be a promising therapeutic option in PDs.

Although no specific clinical trials targeting IR have been performed in PDs, various adjunctive drugs/therapies, including chemical, physical/aerobic, and MSC therapy, which have been shown to improve cognitive and psychiatric symptoms in patients with PDs, also improve insulin sensitivity and metabolic profile in animal models of diabetes and PDs. Moreover, beneficial effects of these agents/therapies in model animals were correlated with increased NSC proliferation and improvement in cognitive behavior. Evidence suggests that these agents/therapies also reduce DNL and lipid/fatty acid accumulation, a leading cause of IR in PDs.

Intriguingly, similar to chemical/physical therapies, treatment with MSCs in rodents has been reported to reduce IR, enhance neurogenesis, and improve cognitive behavior with concomitant improvement in metabolic parameters without inducing any serious side effects. Thus, MSC therapy in PDs could be highly beneficial and further investigation is warranted. Since IR could be diagnosed during childhood or in the adolescent stage in patients with PDs, early intervention with an appropriate adjunctive therapy/drug alone or in combination may normalize cellular signaling(s) that trigger the development of IR and lipid abnormalities while disrupting neurogenesis and cognitive function in patients with PDs. For example, combining MSC therapy with aerobic/resistance exercise may be worth exploring. In this regard, animal model studies as well as human clinical trials are urgently warranted.

ACKNOWLEDGEMENTS

We sincerely acknowledge the facilities provided by the Department of Biotechnology, Era’s Lucknow Medical College and Hospital (Lucknow, India) and Faculty of Science, Era University (Lucknow, India). We also gratefully acknowledge the expert comments and language editing of this manuscript by Professor Dr. Darrell W Brann, Augusta University (Augusta, GA, United States).