Deciphering Mental Illness Authors: Arif Khan, Rashmi Prakash, Anshu Arora, Aishwarya Prasad Published: 2026 ABSTRACT Although more than 60 psychotropic drugs are currently approved for mental illness, their therapeutic index remains modest, with limited efficacy over placebo and substantial adverse effects. Initial concerns that glucagon-like peptide-1 (GLP-1) analogs might produce starvation-related central nervous system (CNS) effects, like those of anorexia nervosa, have been subverted by their largely positive outcomes. Clinical observations of reduced “food noise”, altered reward behaviors, and improved mood have prompted trials evaluating these agents for their potential as psychotropics. This serendipitous paradox compels re-evaluation of prevailing psychiatric disease models. The focus on small-molecule neurotransmitters operating at synapses appears insufficient to explain complex psychiatric illness or the effects of larger molecules like GLP-1. Notably, psychiatric disorders, though familial, do not follow Mendelian inheritance and show considerable heterogeneity in symptoms, suggesting that DNA sequence variation alone is inadequate as a unifying explanation. Alan Turing’s pre-DNA morphogenesis model, emphasizing chemical gradients and developmental divergence, offers an alternative framework. While Virchow and Cajal established neuron-centered paradigms, later reinforced by neurochemical and monoaminergic discoveries, emerging evidence increasingly implicates neuroglia, particularly astrocytes, in higher mental functions, and their disruption in psychiatric diseases. This nascent and fragmented field warrants integration. Herein, we review astrocyte phylogenetics across vertebrates, their regional specialization within the human brain, and their synthesis of diverse small and large molecules. We propose adopting a model in which neurons and neuroglia act together in syncytial constellations that underlie complex mental functions, mediated by their molecular products. Finally, we outline strategies for drug-development and therapeutics harnessing syncytial constellations and astrocyte-derived molecules. MODERN SERENDIPITY Having long grown accustomed to the effect sizes of about 0.3 in psychopharmacological agents,1 we were pleasantly surprised by the impact of glucagon like peptide-1 (GLP-1) analogs in clinical trials for obesity. The effect sizes among these agents in weight management are greater than 1 as observed from published data2 and FDA archives.3 Not only does the weight fall off, but trial participants seem to be mildly euphoric and nonchalant, noting the absence of “food noise”, as well as changes in other behaviors around alcohol consumption, smoking and even gambling.4 Specifically, despite dramatic weight loss and appetite suppression, the adverse mental and behavioral effects of chronic starvation that occur in eating disorders like anorexia nervosa are not observed with GLP-1 and other incretin analogs. Thus, there can be little doubt that the central nervous system (CNS) effects of these agents are remarkable. These findings raise questions about the longstanding assumptions about the biological basis of psychiatric disease. Neurotransmitters that are largely transformed amino acids, lipids and gaseous compounds have been the focus of our understanding and any attempts at treating psychiatric diseases. Glycine, choline, histidine, tyrosine, and tryptophan, with their low molecular weight derivatives, namely glutamate, Gamma-Aminobutyric Acid (GABA), acetylcholine, histamine, catecholamines and indoleamines have been extensively studied and implicated, as also the lipid-derived endocannabinoids like 2-arachidonoylglycerol and anandamide, and the gaseous nitric oxide. However, we take into consideration that the above molecules that provide rapid neural communications, are end products of a complex system, subject to various other factors.5 We thus derive that manipulating these neurotransmitters provides relatively modest relief of acute symptoms but does not seem to significantly alter the pathological course of mental illnesses.6 Further, certain medications like lithium seem to provide benefit without a definitive known target.7 Though current pharmacotherapy does provide considerable relief and quality of life improvement for many patients, those who do not respond, or respond partially, have few alternatives.6 Additionally, the adverse effects of many of these drugs are often an unfair trade off, with many lasting beyond drug cessation and resisting treatment; some examples are tardive dyskinesia, metabolic syndrome and sexual dysfunction.8 There is inadequate data for long term safety and efficacy. For instance, pharmacotherapy for major depressive disorder is often required for longer durations than the recommended 6 months to 1 year, with symptoms returning after drug discontinuation.9 Drug cessation is also known to produce distressing withdrawal syndromes.10 Furthermore, randomized control trials of drugs in psychiatric illnesses seem to have several unaddressed issues—insufficient trial duration not reflective of pathology, lack of uniform use of scales and metrics used as endpoints, and the notorious placebo effect.1 The observed CNS effects of incretin molecules, though under ongoing study, open an avenue for exploration into the neuropathological bases of psychiatric disorders. Specifically, they demonstrate that therapeutically relevant CNS effects can be achieved from mechanisms beyond modulating neurotransmitter-based mechanisms. In this review, we aim to address the underlying mechanisms and factors at play by revisiting the fundamental concepts of human development. We examine how they translate into the development of mental illness. Our paper is intended to provide a conceptual model to better understand these complex factors, specifically the multi-cellular and molecular dysregulation that extend beyond the neuron-centric and monoamine hypotheses of psychiatric disorders. While not intended as a comprehensive solution, we seek to stimulate re-evaluation of existing paradigms, encourage hypothesis generation and broader exploration of pathogenesis and translation in psychiatry. DNA AND ORGANOGENESIS Psychiatric diseases are complex, multifactorial, chronic diseases that cannot be attributed entirely to genetic makeup. To start with, we note that the Watson–Crick model of DNA, though foundational, offers limited insight into why most psychiatric disorders show strong familial patterns, yet incomplete penetrance. For example, the prima facies evidence consists of a lack of 100% heritability of schizophrenia in monozygotic twins, and less than expected frequency among dizygotic twins and other first-degree relatives. Further, there is significant inter-patient variability of symptoms and disease course of any given psychiatric disorder.11,12 Thus, DNA variants and polygenic disease models do not fully account for these patterns, and risk implying a deterministic view of genetic vulnerability. Moreover, life and reproduction are more than mere cloning of DNA; gene expression is dynamic across the lifespan, regulated by environmental influences without alterations to the inherent DNA structure—a domain of regulation known as epigenetics.11,12 Thus, to understand this heterogeneity, embryonic perspectives become particularly informative. The pre-DNA research on morphogenesis by Alan Turing is worth noting.13 In his seminal paper, he describes the physical and chemical factors behind cellular organization and orientation required for morphogenesis. He suggests that the variations of chemicals called morphogens, specifically in their diffusion and concentrations, form gradients that break symmetry, leading to variations arising from an initially homogenous system. To illustrate this, we note that twins, though genetically identical, do not develop identical states of health and disease. Monozygotic twins share a uterus but not an identical intrauterine environment. Differences in placental structure, vascular sharing, and localized morphogen gradients, consistent with Turing’s principles, allow very small initial asymmetries to be amplified into meaningful developmental differences, and thereby, diverge in phenotype from the very beginning. Thus, even slight perturbations in these factors can later influence susceptibility to psychiatric illness differentially within twin pairs. Here, we also consider the work of Dr. Barker and colleagues on the Dutch Hunger Winter, and the subsequent concept that evolved, namely the Developmental Origins of Health and Disease (DOHaD).14 Their data show that even subtle variations in the intrauterine environment can leave lasting marks on physiology and disease risk. From the earliest stages, the embryo is not simply unfolding DNA’s blueprint but adapting to its micro- and macro-environment, with consequences that extend across the lifespan. Putting this into context, the germinal cell has the DNA life script, but once it starts to divide after fertilization, external factors quickly step in and influence development at every stage. After about three divisions, the zygotic cells start to become autonomous.15 Subsequently, the blastocyst undergoes implantation into the endometrium. Chemical factors mainly include the availability of nutrients, such as amino acids and sugars, for the synthesis of proteins and nucleic acids, and efficient cellular metabolism; exposure to toxins and infectious agents, and adequate blood supply are other important factors. Further, physical factors such as site of implantation and orientation of cells within the blastocyst also affect nutrient availability and utilization. Needless to mention, the “morphogens” and their concentration gradients are also acting here to direct cell migration and organogenesis, leading to diverging phenotypes from identical or similar genes. Beyond genetics, it is necessary to consider more immediately accessible etiopathological factors in psychiatry, including different developmental, molecular and cellular influences. Having discussed the developmental levels, it is important to note that many common psychiatric diseases do not develop until adolescence and adulthood, suggesting that other later influences come into play, affecting processes involving various cells and molecules. At a larger scale, or the organ level of psychiatric illness is, of course, the brain. However, in recent years, increasing scientific attention is being directed at the brain-gut axis and its role in various diseases. This is relevant, as observed in the effects of GLP-1 medications. To bridge these developmental and organ-level perspectives, it is instructive to examine these molecules that mediate communication within and between brain and gut. This occurs through the rapid neural route, and the slower chemical or endocrine route.16 Considering the above, we focus on the observation that psychiatric diseases may involve dysfunction at various molecular levels upstream of classical monoamine neurotransmitters.17 We would suggest a classification of these molecules based on their size and duration of function. We hypothesize that the larger molecules that act on longer or sustained timescales are likely to provide larger effects as therapeutic targets. Herein, we examine the various types of molecules that facilitate communication between emotions, thoughts, and behaviors within the brain, and between the brain and gut. Firstly, we have the intracellular molecules, namely the nucleic acids RNA and DNA. Here we also include enzymatic proteins that modify these molecules, such as methyltransferases and histone modifying proteins. Other enzymatic proteins include those involved in the synthesis and transport of various neurotransmitters and peptides that act extracellularly. Now, we consider the molecules that perform extracellular communication. The smallest of the molecules are the neurotransmitters,18 specifically those derived from amino acids, such as dopamine, serotonin, glutamate, and GABA. They are pre-made and stored within vesicles, available immediately to convey messages into synaptic clefts. They act rapidly and are cleared rapidly, achieving real-time communication. Non canonical neurotransmitters, such as the lipid derived endocannabinoids and gaseous nitric oxide, are of low molecular-weight and may also be considered in this category.18 The medium-sized molecules, such as the gastrointestinal tract (GIT) peptides like GLP-1, insulin, glucagon, somatostatin and ghrelin, and the neuropeptides like orexin and endogenous opioids are usually under fifty amino acids long.16,19 They are larger and hence require a longer time to be transcribed from their genes, processed, released and to enact their effects at near and distant receptors. Though they are also pre-made and stored, they are released in response to stimuli that reflect need. They also condition the neurons and affect the degree of neurotransmitter release, and the excitability of the neurons by neurotransmitters. In short, while the smaller neurotransmitters convey messages, the medium-sized peptides determine the message that needs to be conveyed.16,19 Herein, it is notable that small molecules like dopamine, serotonin and norepinephrine may also act as “neuromodulators”, affecting the state (e.g., arousal), tone and excitability of neurons, but these effects seem to occur on a slower timescale in contrast to their effects as rapid neurotransmitters, which are also more immediately observable, modifiable and reproducible.18 Now we consider the larger proteins, such as leptin from adipose tissue, and brain derived neurotropic factor (BDNF) and Nerve Growth Factor (NGF) within the brain.20,21 These molecules seem to sustain or maintain the signals that are conveyed by smaller and medium messengers. Unlike the fleeting neurotransmitters or the condition-setting peptides, these larger proteins operate more slowly, reinforcing long-term tone and establishing a baseline state upon which other signals are superimposed. They influence cellular metabolism, gene expression, and receptor availability. Further, BDNF and other neurotropic factors like are involved in synaptic remodeling, learning, memory and stress adaptability, all key processes in neuroplasticity.18 Therefore, alongside communicating signals, they define the enduring “climate” within which those signals are interpreted.20,21 Indeed, in alignment with the above classification, the work of Dr. Schatzberg has demonstrated that ketamine, now studied for its rapid anti-depressant and anti-suicidality effects, fails to produce these effects in the presence of naltrexone, which blocks opioid receptors.22,23 In other words, the synthetic amine molecule produces its therapeutic effects through the release of medium-sized endogenous opioids, which are essential for their permissive action. Though these findings have limitations, convergent evidence24 is somewhat supportive, indicating that anti-depressant effects are not explained solely by their immediate impact on neurotransmitter levels, but rather by their indirect effect on deeper cellular changes, including gene regulation and plastic changes in various neural circuits. However, these findings, with significant implications for understanding the pathological bases of many psychiatric diseases have so far not been translated into more efficacious pharmacotherapy in a broad sense. In light of this complexity, it is evident that a narrow focus on transformed amino acids offers only partial insight into psychiatric disorders. These neurotransmitters, while critical for rapid signaling, cannot alone account for the persistence or depth of psychiatric illness. Chronic non-communicable disorders reveal heritability that is incompletely understood, as we pointed out in our earlier example of schizophrenia. Another exemplifying condition is post-traumatic stress disorder, which shows how external traumatic events can become deep-rooted beyond the reach of serotonin or dopamine modulation in many cases. Therefore, we hypothesize that targeting larger molecules and their mechanisms may produce longer-lasting therapeutic benefit to larger effects. Testing this hypothesis would involve drug development, as outlined in later sections, as well as randomized controlled trials (RCTs) to compare their safety and efficacy to existing drugs. IF NOT NEURONS, WHAT ELSE IS THERE? While neurotransmitters are fundamental to rapid neuronal communication at the synapse, psychiatric disorders are characterized by enduring alterations in mood, cognition, and behavior that persist far beyond the timescale of synaptic events. It is therefore reasonable to ask whether molecules of varying sizes operating at the level of only neurons and synapses fully account for these temporal and structural dimensions. In other words, wouldn’t the complexity at the molecular level reflect or necessitate complexity and heterogeneity at the cellular level? This inherent weakness in exclusively focusing on neurons and neurotransmitters for understanding mental illness made us go back to earliest descriptions of neuropathology. The original histological work of Vesalius noted that blood vessels and nerve cells were closely associated and Dutrochet observed that there were many types of cells in the brain. In this context, Virchow named the non-neuronal cells as ‘nervenkitt’ and became generally known as neuroglia or cells that held the neurons together.25 Cajal’s discovery of neurons as the cellular unit of the brain in the late 19th century promulgated neuronal primacy, which has held sway until recent decades.25 The knowledge of electrical excitability of neurons, along with the subsequent discovery of neurochemical and neurohormonal transmission starting in the late 19th to early 20th century,26 seemed to have made study of neurons easier than that of neuroglia. Specifically, the discovery of monoamines, their presence in neuronal vesicles and their effects on the various autonomic and behavioral aspects consolidated the role of neurons and neurotransmitters.26,27 The findings that amphetamines can enhance dopamine release and the resultant occurrence of psychosis with excessive use of these further supported the role of neurons and neurotransmitters in mental illness. Blocking dopamine effects appeared to reduce psychotic symptoms.28 Even mood symptoms could be changed by modulating indolamines and monoamines at the synaptic level by various agents such as Monoamine Oxidase (MAO) Inhibitors, tricyclic antidepressants (TCA) and selective serotonin receptor uptake inhibitors (SSRI’s) further lent credence to the pre-eminent role of neurotransmission as the basis of mental illness.27 Paradoxically, seminal work by Kandel and colleagues established that complex mental functions such as long-term memory consolidation depend on new protein synthesis and sustained synaptic remodeling.29 But this insight does not fully consider the potential contributions of non-neuronal cells necessary for these functions and therefore has not challenged the prevailing neuron-centric view of mental illness.29 From the above timeline of developments, along with the various shortcomings of modern psychopharmacology including their modest therapeutic benefits, we are led to acknowledge that current understanding of etiopathogenesis of psychiatric illnesses is incomplete, and this seems to be the major blockade for attempts at “curing” or effectively managing severe mental illness. ALTERNATE MODELS FOR MENTAL ILLNESS Given the above factors of missing heritability, developmental origins, inter-individual variations in symptom manifestation in psychiatric illnesses, and the gaps uncovered by neuron-centric and neurotransmitter-based models, what are possible alternatives? In this context, we are suggesting that mental functions, as also psychiatric diseases emerge primarily from not only neurons but the larger network of neurons and non-neuronal brain parenchymal cells acting in concert. Specifically, the complex of glial cells, with the associated neurons acting as the main communicators may provide an alternate model of learning, memory, recall, emotions, and behavior. So, are all the neuroglial cells involved? This critical question can only be partially addressed given the state of knowledge, which at best is scattered. The research into neuroglial cells has been extensive but focused on specific individual aspects. For example, neuropathologists have focused on morphology, embryology and histology. Neurologists have focused on the degenerative aspects of neurons and neuroglia. Geneticists have focused on gene expressions, in particular protein synthesis by the various neural cells.30 In order to synthesize a coherent picture, we have attempted to obtain relevant data from various fields. First, we have reviewed the nature and numbers of neural cells across vertebrate species (see Table 1). Next, we have reviewed the role of various neuroglial cells in human CNS. Following this, we have focused on the nature and potential significance of astrocytes in mental illness (see Fig. 1). TABLE 1Maximal Reported Cell Counts of Neurons and Neuroglia Across Six Vertebrate SpeciesSpeciesTotal Brain Neurons (Millions)Total Brain Non-Neuronal CELLS (Millions)Total Brain Cell Counts (Non-neuronal Cell Percentage)Adult Madagascar Ground Gecko31 5611 (55%)Songbirds32 2,1701,2403,410 (36%)Parrots32 3,1401,8004,940 (36%)Mouse33 8,1304,45012,580 (35%)Yellow Babboon34 20,1209,18929,291 (31%)HUMANS35 94,12094,400188,520 (50%)Numbers obtained from datasets and calculated numbers (column 4) have been rounded up for clarity. As averages were reported in the source datasets, the highest numerical value was calculated from the mean + 2 (standard deviation). Maximal Reported Cell Counts of Neurons and Neuroglia Across Six Vertebrate Species Numbers obtained from datasets and calculated numbers (column 4) have been rounded up for clarity. As averages were reported in the source datasets, the highest numerical value was calculated from the mean + 2 (standard deviation). The complex of pre- and post-synaptic neurons, the associated astrocyte and oligodendrocyte.36–38 These complexes may form constellations across various brain regions involved in higher mental functions, such as amygdala, hippocampus, hypothalamus and cortex. Astrocytes possess primary processes which branch and ultimately end in peri-synaptic astrocytic processes (PAP’s). The prevailing concept of the tripartite synapse includes pre-synaptic and post-synaptic neuronal terminals, along with the PAPs. It is considered that the messages conducted between neurons modulate and are modulated by the PAP, which has been shown to play a crucial role in maintaining synaptic environment, preventing glutamate-induced toxicity, and providing metabolic substrates essential for neuronal activity. Each astrocyte is associated with hundreds of thousands of synapses through their numerous PAPs. The orientation of astrocytes with neuron may vary based on the type of astrocyte. Astrocytes interact with one another through connections between their processes, i.e., gap junctions. It is considered that activation of one astrocyte by synaptic activity or neurotransmitters, is communicated to the neighboring astrocytes through the gap junctions. In that way, a single astrocyte can affect the activity of several neurons over larger distances. Oligodendrocytes are known to be involved in myelination, and act to ensheathe axons in myelin as they grow or regenerate. Microglia would ensure CNS immune homeostasis and be mediate inflammatory processes as needed. This would warrant their communication with astrocytes, which regulate synaptic development and neuronal recruitment during activity, such as learning, memory acquisition, long term memory consolidation, neuronal repair and other essential functions, such that myelination may occur in a timely fashion. Finally, astrocytic processes that are in contact with blood vessels are called perivascular processes. These processes modulate local blood flow through various vasoactive molecules, thus altering blood flow to match neuronal activity. They are also considered to be central to the blood brain barrier, which can be a significant roadblock in pharmacotherapy, making astrocytes an important consideration in the development of drug delivery systems. Created in BioRender by Rashmi Prakash, 2026. https://BioRender.com/8octxro. Lastly, we have reviewed the relevant literature on potential interventions. This last section is focused on the presence of appropriate peptides, proteins, lipoproteins and glycoproteins that may have relevance to mental illnesses (see Table 2). Following this we propose models for therapeutic strategies based on available data. TABLE 2Neurotransmitters, Peptides, Large Proteins, Lipoproteins and Glycoproteins Present Among Neural Cells, Including Neurons, Astrocytes And Other Neuroglial CellsName of MoleculeMolecular Weight (Da)Amino acid lengthSection A. Small Neurotransmitters Produced Predominantly by Neurons18,39,40 Histamine1110 (Diamine) Acetylcholine1460 (Quaternary ammonium compound) Dopamine1530 (Monoamine) Norepinephrine1690- (Monoamine) Serotonin1760-(Monoamine)Section B. Small Neurotransmitters Produced Predominantly by Astrocytes and Glial Cells18,37,39,40 Nitric oxide (NO)300 (Gaseous neurotransmiter) L-lactate900 (Monocarboxylic acid) GABA (γ-aminobutyric acid)1031 D-Serine1051 L-glutamate1471 Kynurenic acid1890 (Tryptophan metabolite) Prostaglandin E2 (PGE₂)* 3520 (Lipid compound- eicosanoid) Cholesterol* 3870 (Lipid molecules) ATP (adenosine triphosphate)/ Adenosine† 507/2670 (Nucleotide/ nucleoside)Section C. Medium and Large Molecules Including Peptides, Large Proteins, Glycoproteins and Lipoproteins Present Among Neural Cells, Including Neurons, Astrocytes and Other Neuroglial Cells18–20,37,39–59 Substance P (Tachykinin-1–derived peptide)1,34811# Somatostatin‡ 1,60014# Endothelin-12,50021# Hepcidin‡ 2,80025# Ghrelin‡ 3,24428# Glucagon-like peptide-1 (GLP-1)‡ 3,30030# Neuropeptide Y (NPY)4,27036# Corticotropin-releasing hormone (CRH)4,75041# Insulin§ 5,80051# Metallothionein-16,09461 S100B (S100 calcium-binding protein B)10,71392 Cystatin C (Cystatin-3)16,445141 Fibroblast growth factor-2 (FGF-2)18,000155 Leptin§ 18,641167 Vascular endothelial growth factor A (VEGF-A)22,314191 Insulin-like growth factor 1 (IGF-1)21,841195 Lipocalin-2 (LCN2)22,588198 Ciliary neurotrophic factor (CNTF)22,931200 Glial cell line-derived neurotrophic factor (GDNF)23,720211 Interleukin-6 (IL-6)23,718212 Tumor necrosis factor (TNF; TNF-α)25,644233 Brain-derived neurotrophic factor (BDNF)27,818247 Pro-enkephalin A30,787267 Interleukin-33 (IL-33)30,759270 SPARC (Secreted protein acidic and cysteine-rich)34,632303 Apolipoprotein E (ApoE)∥ 36,154317 Aquaporin-4 (AQP4; Aquaporin 4)34,830323 Purinergic P2Y1 receptor42,027373 Connexin-43 (GJA1; Gap junction alpha-1 protein)43,008382 Transforming growth factor-β1 (TGF-β1)44,325390 Indoleamine 2,3-dioxygenase (IDO1)45,326403 Adenosine A2A receptor44,707412 Clusterin (Apolipoprotein J)∥ 52,495449 Chordin-like 1 (CHRDL1; Chordin-like protein 1)52,026456 GLP-1 receptor (GLP1R)53,026463 GPR37L1 (G protein-coupled receptor 37-like 1)52,771481 MCT1 (SLC16A1; Monocarboxylate transporter 1)53,944500 EAAT1 (SLC1A3; Excitatory amino acid transporter 1)59,572542 Glypican-6 (GPC6; Glypican 6)¶ 62,736555 Glypican-4 (GPC4; Glypican 4)¶ 62,586557 EAAT2 (SLC1A2; Excitatory amino acid transporter 2)62,104574 Estrogen receptor-α (ESR1)66,216595 GAT-3 (SLC6A11; Sodium- and chloride-dependent GABA transporter 3)70,606632 SPARCL1 (Hevin)¶ 75,208664 Toll-like receptor 4 (TLR4)95,680839 mGluR3 (GRM3; Metabotropic glutamate receptor 3)98,879879 Leptin receptor (LEPR)132,4941,165 Thrombospondin-1 (TSP-1; THBS1)¶ 129,3831,170 Thrombospondin-2 (TSP-2; THBS2)¶ 130,0001,172 mGluR5 (GRM5; Metabotropic glutamate receptor 5)132,4691,212 Insulin receptor (INSR)156,3331,382 Complement C3187,1481,663 Tenascin-C¶ 240,8532,201 ABCA1 (ATP-binding cassette transporter A1)254,3022,261The above is a non-exhaustive list of molecules involved in astrocytic function and signaling, with molecular astrocytic diversity under ongoing study. Growing evidence shows the implication of the above molecules in neurological and psychiatric disorders.*Lipid-based molecules and compounds†Astrocytes can produce and transport ATP to neurons, whereas adenosine is largely non-astrocytic in origin.‡Gut brain molecules§Peripheral molecules other than gut-derived∥Lipoproteins- no. of amino acids denoted for the protein constituent of the molecule¶Glycoprotein molecule; no. of amino acids denoted for the protein constituent of the molecule; molecular weight is not inclusive of glycosaminoglycan chains.#Mature form of protein with approximate molecular weight. Neurotransmitters, Peptides, Large Proteins, Lipoproteins and Glycoproteins Present Among Neural Cells, Including Neurons, Astrocytes And Other Neuroglial Cells The above is a non-exhaustive list of molecules involved in astrocytic function and signaling, with molecular astrocytic diversity under ongoing study. Growing evidence shows the implication of the above molecules in neurological and psychiatric disorders. Lipid-based molecules and compounds Astrocytes can produce and transport ATP to neurons, whereas adenosine is largely non-astrocytic in origin. Gut brain molecules Peripheral molecules other than gut-derived Lipoproteins- no. of amino acids denoted for the protein constituent of the molecule Glycoprotein molecule; no. of amino acids denoted for the protein constituent of the molecule; molecular weight is not inclusive of glycosaminoglycan chains. Mature form of protein with approximate molecular weight. As shown in Table 1,31–35 the numbers of the neural cells, both neuronal and neuroglial, are listed across various vertebrate species. Simply put, mammals have a significantly large number of CNS cells, with adult humans thought to have the highest numbers of both neuronal and neuroglial cells per body mass. With increasing body and brain size, mammals tend to exhibit higher absolute and relative numbers of neural cells; however, the distribution of these cells varies markedly across species. For example, African elephants possess more total neurons than humans, yet they are more concentrated within their cerebelli, whereas in humans a greater proportion of neurons are noted in the neocortex.60 Similarly, birds have primate-like numbers of forebrain neurons densely packed into a compact cortical (pallial) structure, while maintaining lower numbers of non-neuronal cells, likely due to overall brain size constraints.32 Together, these observations indicate that neural cell numbers alone do not fully account for interspecies differences in cognition and highlight the importance of considering cellular composition, distribution, and organizational differences.31–35 In this context, it is important to note that human infants at birth have approximately 20 billion neurons and approximately 6 billion neuroglial cells.36 By the time humans reach adulthood, the number of neuroglial cells match the number of neurons. In other words, the number of neuroglial cells increase by 5 to 6-fold from birth to adulthood,36 indicating extensive postnatal glial development not reported in other vertebrates. Interestingly, the number of neurons and closely associated oligodendrocytes seem to decline with age, whereas cells such as astrocytes and ependymal cell numbers may not decline with age.36 The significance of these observations is enormous. For example, the newborn chick doesn’t acquire its capacity for additional mental abilities as it reaches adulthood as a mature bird.61 On the other hand, humans have enormous capacity to learn multiple new concepts including language, memory, thoughts, emotions and behavior. Although several mammals do acquire knowledge, humans are exceptional in the magnitude of learning from experience, knowledge and emotions.62,63 The drawback obviously is that it makes humans much more susceptible to mental illnesses.64,65 In this context, there is considerable evidence to suggest that of the four types of neuroglial cells namely oligodendrocytes, astrocytes, microglia and ependymal cells, astrocytes may have an exceptional role in mental functions and need further review, which we will examine in the following sections. THE ROLES OF NEUROGLIA IN PSYCHIATRY Although oligodendrocytes may constitute up to 75% of neuroglial cells, their primary role in humans has been known as developing and maintaining the myelinated neurons throughout life.66 Microglia are known to be primarily involved in the immune homeostasis of neural tissue.66 Ependymal cells line the ventricular system and are involved in CSF dynamics and nourishing neural stem cells in the adult brain.67 Astrocytes have been known at the “support cells”, assisting neurons with metabolic support necessary for rapid and efficient electrical communication.66 However, emergent evidence has brought to the fore the extended roles of these non-neuronal cells in neuropsychiatric health and disease. For example, oligodendrocytes ensheathe neuronal axons in myelin in an activity-dependent manner. A landmark study on mice showed that stimulation of the premotor cortex neurons was associated with an increase in myelination, which also manifested in functional improvement in the represented limbs.68 Oligodendrocytes also possess various neurotransmitter receptors and ion channels, support axonal metabolism, and have been shown to maintain axonal and myelin sheath integrity, particularly under stress.69 Stress-related disruption of oligodendrocyte precursors during critical developmental windows, e.g., postnatal period or adolescence have been shown to contribute to the risk of depression over the lifespan.70 Psychiatric disorders have been associated with brain white matter changes of varying degrees in humans; however, an example of a causal association is in the 2’,3’-cyclic nucleotide 3’-phosphodiesterase (CNP) gene. This gene encodes a specific oligodendrocyte-derived protein found in the myelin sheath, which is necessary for axonal survival and integrity. In a mouse model, mice which were heterozygous for a loss-of-function mutation of CNP gene displayed catatonic depression-like behaviors, which were unmasked by older age (i.e., second hit).71 Reduced expression of oligodendrocyte-related genes have also been observed post-mortem in brains of people with schizophrenia.72 White matter changes observed during learning further suggest a role of myelin and oligodendrocytes in higher cognitive function.73 Needless to mention, myelination is necessary for rapid impulse conduction and therefore, connectivity between different brain regions, disruption of which produces significant cognitive and neuropsychiatric abnormalities, as evidenced by demyelinating disorders. Whereas microglia have been known as the resident CNS macrophages, being largely quiescent, performing surveillance functions, and being activated by injury and inflammation, through complex mechanisms. The activated microglia may be either pro-inflammatory, or regulatory, i.e., limiting excessive activation and subsequent neuronal damage by inflammatory mediators.74 Microglia possess several membrane receptors that are immune-specific (e.g., cytokine receptors), as well as for neurotransmitters and neuropeptides. They are capable of modulating BBB permeability through various inflammatory mediators, regulate cerebral blood flow, as well as eliminate excessive synapses, particularly in post-natal development, which is crucial for normal brain connectivity.75,76 Microglial phagocytosis is also essential for clearance of apoptotic cellular debris, particularly in periods of rapid neuronal development.74–76 Moreover, neuroinflammation is associated with several psychiatric disorders, including depression and anxiety.77 It was found that gene expression changes in primates with depression-like symptoms were most notable in microglia of the frontal cortex.78 Some of the central effects of peripheral immune and inflammatory processes seem to be mediated by microglia, as well as those of peripheral molecules including GLP-1 and adiponectin.79–81 Various disruptions in microglial physiology, particularly resulting in excess activity have been linked to schizophrenia autism spectrum disorders and various other psychiatric disorders in post-mortem studies and animal models.82,83 Ependymal cells are a type of neuroglia that sit at the interface of the ventricles and brain parenchyma. They are traditionally known as ciliated cells which assist in CSF flow and clearance. Furthermore, they develop from precursor cells which also give rise to neurons and astrocytes, and have been shown to promote neurogenesis, or transform into neurogenic stem cells themselves, in certain populations of the brain and spinal cord.84 While their structural roles may be more obvious, such as CSF flow and lymphatic clearance, they perform various roles in transport of metabolites,85 water and metal ions.86 Structural and functional alterations in ependymal cells have been shown to underlie altered CSF dynamics and stress-induced depression, in human patients and mouse models.87 There is more abundant evidence regarding the role of glial cells in psychiatric disorders, most notably from post-mortem and animal models.88 While microglia, oligodendrocytes and ependymal cells play important roles in the structural and functional integrity of the CNS, we now shift our focus on to astrocytes, which are the largest of the glial cells, and seem to have significant real-time contributions in neural processes. THE ROLES OF ASTROCYTES IN PSYCHIATRY Astrocytes are the stellate (star-like) cells that originate from ectodermal neural stem cells and are generated from the radial glial cells in the immediate post-natal period in humans.30,37,66 Conventionally, the astrocytes are thought to maintain CNS homeostasis by various mechanisms. They are considered to regulate pH, regulate glutamate homeostasis, morph into neurons on demand, maintain synaptic plasticity, control over blood-brain barrier (BBB) and glycogen synthesis and storage.37,66 Histologically, numerous morphological types of astrocytes have been noted across brain regions, a degree of heterogeneity on par with that of neurons.37,41 Some of this variety, such as varicose projections, is uniquely more well-defined and abundant in humans compared to other vertebrates, including primates; astrocytes are thought to show disproportionate evolutionary scaling in humans compared to other cell types across species, in terms of morphological complexity and heterogeneity.38 Even more important is the likelihood that the astrocytes not only have varying histological pictures, but their functional capabilities may also vary based on where these cells are present in the brain.37,38,41–44 This is seen as differences in gene expression patterns and proteins produced based on brain region, which is still under ongoing study. Therefore, we hypothesize that astrocytes as such may primarily underlie complex mental functions and disorders unique to humans. Astrocytic morphological and functional heterogeneity may underlie localized brain functions through production of different sets of proteins. Astrocytes and other neuroglia form syncytial complexes with one another and are closely associated with neurons at the synaptic junctions through their processes, as shown in Figure 1.37 The detailed study and molecular assays for understanding the intricate nature of these syncytia are difficult as current methods for such an effort are tedious and not fully developed. Furthermore, these studies may mostly be feasible in animal models or post-mortem studies. Given this, it is important to note several key factors that are relatively well established about the astrocytes. Transplanting human astrocytes leads to a definite and measurable increase in cognitive abilities in other animals such as rodents.89 Rodents who were raised with light deprivation from birth developed fewer astrocytes in the optical cortex and conversely develop a larger number of astrocytes in the auditory cortex.90 Human astrocytes are larger and more complex in nature compared to other mammals.38 Last, but not least, a simple increase in astrocyte numbers compared to neurons does not result in positive effect. Specifically, Down’s syndrome is associated with large increases in astrocytes throughout human brain.91 More recent experimental evidence shows the effects of separately altering neuronal and astrocytic activity. A 2018 study showed that disrupting astrocytic activity in rodent hippocampus impairs contextual memory acquisition even when neurons remained functional. Furthermore, it showed that directly increasing neuronal activity without astrocytic activation led to severe memory impairment.92 Similarly, stimulating amygdalar neurons without astrocytic activation led to maladaptive fear responses in animal models, as detailed in this comprehensive review.93 The observations of the above studies suggest that astrocytes contribute more directly to higher mental functions than simply supporting the neurons performing them.45 Furthermore, there is growing body of evidence suggesting that hypothalamic astrocytes play a key role in the gut-brain interactions, with high calorific diet triggering gliosis as well as changes in gene expression.46,47 Beyond the emerging data about their extended physiological roles, various animal models have demonstrated the role of astrocytic dysfunctions underlying mental illness. For instance, a study by Banasr and Duman showed that pharmacologically induced loss of astrocytic function in prefrontal cortex was sufficient to induce depression-like symptoms in rats without neuronal loss.94 Furthermore, calcium signaling within astrocytes has been shown as crucial for astrocytic reactivity and neural inflammation-induced depressive behaviors in mice.95 Chronic stress appears to disrupt astrocytic functions through various molecular and gene-expression alterations, contributing to imbalances in excitatory-inhibitory neurotransmission, as detailed in a recent comprehensive review.96 Whereas human studies are mostly restricted to post-mortem and human-derived stem cell analyses, which provides insufficient clarity or indirect data on causality. However, a 2022 study examined gene expression in astrocytes of mice, as well as in astrocytes derived from induced stem cells from human schizophrenic patients and healthy controls.97 They found that genes associated with specific astrocyte subtypes were disproportionately represented in schizophrenia compared to healthy controls, supporting a potential role of astrocytes in schizophrenia pathology. Further analyses also demonstrated that commonly used psychotropics have little to no effect on astrocytic gene expression, which suggests that these medications may not primarily target astrocytes.97 Furthermore, evidence from human post-mortem analyses98 and pre-clinical animal models99 show astrocytosis, possibly maladaptive, as well as astrocyte gene expression changes in various psychiatric disorders, including depression, substance abuse and schizophrenia. In summary, the presence of large numbers and high degree of specialization of neuroglial cells seems to be the hallmark of evolution in mammals and especially in humans38,48 and seems to contribute significantly to psychiatric pathology and vulnerability. As of now, it is reasonable to consider that the primary cell types that may distinguish humans from other species are astrocytes, based on evolutionary studies.37,38,41–48,89–99 However, it is premature to rule out the role of other neuroglial cells in mental functions and thus mental illnesses, as the area of study is difficult and complex.49 Specifically, brain biopsies for evaluating mental functions and illnesses are not acceptable; the BBB is an additional roadblock, and functional radiological studies are inadequate to measure the full spectrum and cellular details of mental functions.100 Lastly, post-mortem examinations are of limited value. In this context, it is interesting to note that chimeric techniques of implanting neural cells into rodents to understand disease processes are underway.101 While diversity exists in varying degrees among all neuroglia, astrocytes display remarkable region-specific diversity while performing a large variety of functions overall, with disproportionate evolutionary scaling in humans, and therefore may serve as a reliable entry point for expanding therapeutic avenues. WORKING MODEL These findings lead us to propose a working model of human mind, its potential and vulnerabilities. From conception, the fertilized human egg can divide about three times and develop on its own, and further development is subject to its survivability and adaptability to the outside environment. This is like the development of other mammals. However, the major differences start to emerge after birth. Specifically, the human brain develops by a large production of glial cells, in particular astrocytes, starting postnatally. In addition, the presence of many oligodendrocytes appears to provide additional support to neurons. Microglia appear to regulate development, and ependymal cells assist through CSF flow and metabolite clearance. Thus, constellations of a complex of neurons and neuroglia appear to take over human mental functioning, as modeled in Figure 1. The production of a diverse profile of molecules such as transmitters, peptides, medium to large size proteins, lipoproteins and glycoproteins, particularly in the astrocytes (see Table 2) appears to underlie the astrocytic modulation of neuronal function in humans. Such a complexity is unlikely to be seen in other vertebrate species, particularly non-mammalian species, given the unique features of human astrocytes.38 To illustrate with an example, the constellations of syncytia in the amygdala communicating with those of the cortex may allow for the various emotions to be learned, recalled and used as needed, to a greater extent than in any other species. Similarly, syncytial constellations in the hypothalamus underlie the inherent and coherent knowledge of feeding and survival, a skill highly developed in humans, allows for phenotypic, metabolic and behavioral heterogeneity, and the underlying mechanisms of communication between the gut and brain, both neural and chemical. Another specialized area of such syncytial constellation may be based in hippocampus with projections into the various perceptual and association areas in other parts of the brain, which may be the basis of the unique ability to learn, retain and utilize information gathered from sensory inputs. These considerations could offer a framework for better understanding the human brain, and allow us to develop newer lines of inquiry, rather than a sense impotence in explaining, preventing and managing mental illness, resilience and vulnerability. It would be worthwhile to explore syncytial constellations as a unit rather than individual cells or synapses, to obtain a deeper understanding of the functional nature of syncytial constellations to this end. For example, identifying specific large molecules in the astrocytes and/or syncytia, and applying targeted therapies using medium to large proteins to prevent auditory hallucinations may be more useful than applying non-specific dopamine blockers with a wide range of adverse effects. Similarly, specific monoclonal antibodies particularly targeting molecules in amygdalar constellations may have a significant effect on PTSD symptoms. This could be achieved, for instance, by culturing amygdalar or auditory lobe astrocytes, identifying or inducing the expression of specific proteins and related compounds that may be implicated, and thereby preparing monoclonal antibodies or other targeted therapies such as receptor analogs or antagonists, that could neutralize the identified target molecules and/or their effects. These compounds could then be tested for therapeutic benefit using established animal models of specific illnesses such as schizophrenia or PTSD. WHAT IS POSSIBLE NOW? Although this may sound like a daunting task, we do consider that there is sufficient knowledge to start using some of it therapeutically as of now. Herein, we consider the example of GLP-1. Importantly, GLP-1 analogs such as semaglutide appear to cross the BBB.102 Though exact mechanisms of central action are under ongoing study, weight loss as well as various CNS effects (appetite suppression and reward-related behaviors) have been observed with both subcutaneous (0.25 mg weekly) and oral (14 mg daily) dosing. This difference in dosages indicates that oral bioavailability of semaglutide is over 400 times lower than subcutaneous.103 Notably, GLP-1 precursors are also produced in certain brain neurons, specifically in the brainstem.104 In either case, is it possible that the astrocytes at the BBB can recognize brain-derived molecules and/or their sequences of amino acids, sugars and lipids, allowing them to cross through? If such a possibility exists, then it is worth considering other centrally present or centrally derived peptides, proteins, lipoproteins and glycoproteins as therapeutic tools.50,51 In order to explore such an option, we compiled some of the known molecules in CNS, of both astrocytic and neuronal origin in Table 2.37,39–59 As such, these molecules can range from nitric oxide (molecular weight of 30 Daltons) to complex proteins that may contain over twenty-two hundred amino acids with molecular weight greater than two hundred thousand Dalton.18–20,37,39–59 The nomenclature and descriptions of proteins seems to be without a standardized or universally accepted organizational system.105 Thus, we have categorized the molecules as per their size and functions, with a consideration of the neural cell of origin. In Section A, we have listed the small neurotransmitters with molecular weight that are synthesized predominantly by the neurons, and extensively utilized by both neurons and neuroglial cells. As discussed earlier, these small molecules have known CNS effects, but provide inadequate benefit in managing mental illnesses. Section B, of Table 2 lists molecules that are mostly produced by neuroglial cells, especially astrocytes, that are essential for neuronal function but may not be produced by neurons themselves. Section C, of Table 2, may have the most relevance to formulating newer agents with potential therapeutic benefits in mental illnesses. It is likely that these molecules, play a critical role in the functions of neuron-neuroglial syncytia (see Fig. 1),37 given their relatively larger size and longer timescale of operation than neurotransmitters, as previously discussed. A number of these molecules are also known for their roles in the gut-brain axis.102 It is also worthwhile to eventually explore the larger proteins, lipoproteins and glycoproteins, some of which exist in the extracellular space, for potential therapeutic benefit. Obviously, there is no ready answer for finding the perfect molecular target and its application at the level of syncytia and syncytial constellations. This area of research would involve the study of genes, protein design and synthesis, and application in animal models, which is time- and cost-intensive. POTENTIAL METHODS TO DEVELOP LARGE MOLECULE THERAPIES FOR MENTAL ILLNESS Essentially, there are two commonly used methods to develop larger compounds, particularly proteins and peptides, for therapeutic effects. The simple one is to understand the sequence of amino acids and produce smaller chains and then combine them into a peptide or a small protein, as in the case of semaglutide. Furthermore, it is possible to synthesize smaller ligands and test them for desired activity in vitro as well in animal models.106 On the other hand, recombinant DNA techniques, whereby microorganisms can be made to express complex molecules by altering their DNA, may be an alternative method to develop these larger therapeutic compounds. The classical example of this technique is humulin, the synthetic insulin. As of now, such experiments are underway at many of the pharmaceutical companies and results can be expected in five to ten years.106 Moreover, considering that humans as a species possess unique cognitive potential, abilities and vulnerabilities, as also the individual cell-specific evolutionary differences discussed above, a more accurate and reproducible study would be ideally done in humans, but this is constrained due to various reasons outlined in earlier sections. However, methods like functional magnetic resonance imaging (fMRI) using cell-specific molecular tags, chimeric models and electroencephalograms, would enable in-vivo studies in humans by offering useful guides with which to identify target brain regions and direct the necessary study in animal models.44,100,101 WILL THIS BE ENOUGH? Although we are optimistic about the potential role of larger molecules and non-neuronal targets as therapeutic avenues, caution is warranted. The etiopathology of psychiatric illness remains incompletely understood, and emerging observations, such as the central effects of gut–brain peptides, suggest that regulatory systems beyond classical synaptic neurotransmission may contribute to disease expression. These findings do not negate the role of neurons, or genetic vulnerability, but rather demonstrate that psychiatric disease risk unfolds within a complex developmental and environmental context beginning early in life. As with many chronic medical conditions, psychiatric illnesses appear to reflect multi-factorial processes that evolve over time and require longitudinal management. The therapeutic strategies of other chronic disorders may offer a conceptual parallel. In autoimmune diseases, broad anti-inflammatory strategies like glucocorticoids were once the only available approach, whereas subsequent advances led to more targeted biologic agents, such as monoclonal antibodies, designed to target specific components of disease biology and maintain remission over longer periods.107,108 While corticosteroids are still essential and effective in treating acute flares and inducing remission in many autoimmune disorders, disease-modifying agents are needed to offset their wider range of adverse effects of glucocorticoids and to achieve higher precision while maintaining remission. An improved understanding of pathology thus permits greater precision and, in some cases, improves long-term disease control. With metabolic disorders such as hypertension and diabetes, multiple medications targeting various molecules and pathways (e.g., calcium-channel blockade and angiotensin receptor blockade in hypertension, insulin secretagogues vs. sensitizers in diabetes) are available, with multiple medications needed for optimal control.109,110 By analogy, as our understanding of cellular regulation in the brain expands, including glial, metabolic, and peptide-mediated mechanisms, it may become possible to identify more targeted strategies that complement existing therapies and potentially influence illness trajectory. Whether such approaches will ultimately prove disease-modifying in psychiatry remains an empirical question, but the conceptual shift toward greater biological specificity warrants exploration. LIMITATIONS Hereby we acknowledge some of the limitations of our paper. First and foremost, this manuscript presents a conceptual synthesis rather than original experimental data. Although we draw upon data from a wider scientific scope and emerging clinical observations, direct causal evidence linking astrocytic dysfunction to specific psychiatric syndromes in humans is so far limited. Much of the data derives from rodent and post-mortem models, which may not fully capture the structural and molecular complexity of human astrocytes. Secondly, our proposed model of syncytial constellations is necessarily inferential. As previously discussed, current in vivo methods do not permit detailed functional mapping of astrocytic networks in the living human brain. Brain biopsies are not feasible or acceptable for psychiatric conditions, functional imaging does not allow for cellular-level detail, and post-mortem studies are limited in terms of temporal causality. Thus, the framework offered here must be regarded as hypothesis-generating rather than definitive. Third, while GLP-1 receptor agonists and other peptides suggest the therapeutic relevance of larger molecules, extrapolating from metabolic and behavioral observations to broad psychiatric applicability is possibly premature. Our speculations regarding central bioavailability, and the subsequent events at receptor and intracellular levels in psychiatric populations require systematic, rigorous study. Not all medium- or large-molecule interventions are likely to engage astrocytic mechanisms in a therapeutically meaningful way. Fourth, by highlighting astrocytes, we would emphasize that we do not intend to undermine the already established, critical role of neurons. Mental functions emerge from dynamic interactions among neuronal and non-neuronal cells of the brain parenchyma, vascular elements, and systemic influences. Though the relative contribution of each remains incompletely defined, we would reiterate that psychiatric disorders are unlikely to be explained by a single cellular class, neuronal or non-neuronal. Finally, development of large-molecule therapies targeting astrocytic pathways faces practical translational challenges similar to those seen in other chronic diseases, including blood–brain barrier permeability, cellular delivery systems, immunogenicity, long-term safety, and cost. Despite these limitations, the strengths of this work lie in its integrative scope and conceptual synthesis. By drawing together developmental biology, evolutionary neuroanatomy, molecular signaling hierarchies, and emerging clinical observations, we offer a structured framework that moves beyond neurotransmitter- and neuron-centric explanations without dismissing their relevance. With our model, we attempt to provide possible hypotheses, identify translational directions, and reframe existing findings within a broader biological context. In that sense, this work is not intended to provide definitive solutions, but to widen the field of inquiry.