Biology of Autism — Molecular Pathways Explained

Section 1

Cytokine Surge and Cortical Patterning

Maternal Immune Activation designates a state in which infections, autoimmune activity, or sustained physiological stress during pregnancy drive a significant rise in inflammatory cytokines — particularly interleukin‑6 (IL‑6) and interleukin‑17A (IL‑17A) — within the maternal–fetal milieu. These cytokines act not merely as markers of illness but as developmental morphogens capable of reshaping cortical organization, glial programming, and inhibitory circuit assembly.

Elevated IL‑6 and IL‑17A during critical gestational windows can influence:

• Radial migration and cortical lamination, leading to subtle forms of cortical dysplasia
• Regional patterning, particularly in frontal and temporal regions implicated in social communication and executive function

Rather than invoking gross malformations, this model emphasizes micro-architectural deviations: altered positioning and density of projection neurons and interneurons that may be invisible on routine imaging but have large effects on circuit dynamics. Animal MIA models in which IL‑6/IL‑17A are experimentally manipulated reproduce autism-relevant behavioral profiles — including social deficits, repetitive behaviors, and sensory abnormalities — supporting the plausibility of this pathway.

Section 2

Long-Term Glial Programming: Microglia as "Primed Responders"

MIA functions as a glial programming event. Microglia exposed to elevated IL‑6/IL‑17A during development do not mount an acute response and return to baseline; they undergo transcriptional and epigenetic changes that shift their default state toward heightened reactivity. Key features of this programming include:

• Lower activation thresholds in response to subsequent signals such as LPS, toxins, or neuronal stress
• A bias toward pro-inflammatory (M1-like) phenotypes when activated, with robust production of IL‑6, TNF‑α, and IL‑1β

In the context of the cascade, this means that subsequent insults — gut-derived LPS, environmental toxins, metabolic stress — are more likely to produce sustained microglial activation rather than transient, self-limited responses. MIA thus supplies an upstream explanation for why some individuals show chronic neuroinflammation after relatively modest later exposures.

Section 4

Selective Vulnerability of PV and SST Interneuron Development

MIA specifically impairs the development of two critical inhibitory interneuron classes:

• Parvalbumin-positive (PV) interneurons, which orchestrate high-frequency oscillations and temporal precision in cortical networks
• Somatostatin-positive (SST) interneurons, which regulate dendritic integration, local gain control, and plasticity within cortical columns

The proposed mechanisms include disruptions of progenitor proliferation, migration, and differentiation in interneuron lineages exposed to elevated IL‑6/IL‑17A and related inflammatory mediators. The net result is a cortex with reduced or dysregulated inhibitory control, predisposing to excitation–inhibition imbalance, impaired synchrony, and aberrant sensory gating.

This early interneuron fragility is later compounded by SST elevation in response to chronic postnatal stress and inflammation, and by microglia-mediated synapse pruning and A1 astrocyte conversion. MIA thus establishes a foundational inhibitory circuit weakness that later cascade elements amplify.

Section 6

Position within the Overall ASD Cascade

Positioning MIA at Level 1A makes several broad claims:

• A subset of autism risk is rooted in prenatal programming of cortical, glial, and gut systems, not solely in postnatal exposures or single genes
• The later central nodes — SIRT1 deficiency, SST vulnerability, microglial over-reactivity, and gut-mediated inflammation — are not independent events but partially reflect a shared early developmental history
• Animal MIA models recapitulating autism-relevant behaviors support the view that early cytokine exposure can set long-term trajectories for both neural architecture and immune set-points

Section 1

Dysbiosis and Barrier Breakdown: Creating the "Open Gate"

The Level 1B sequence is not framed as "bad bacteria" in a generic sense, but as a structural problem: dysbiosis plus impaired barrier integrity allows LPS from Gram-negative commensals to enter the bloodstream in the absence of overt infection. This state is conceptualized as metabolic endotoxemia without active infection — circulating innate immune ligands with no classical infectious focus.

• Short-chain fatty acids such as butyrate, normally produced by commensals, support epithelial integrity and regulatory immune tone; their reduction predisposes to tight junction loosening
• Local inflammation, dietary factors, and stress hormones further weaken epithelial cohesion and mucosal defenses

Section 3

Conversion to Neuroinflammation: Peripheral–Central Coupling

Peripheral LPS-driven inflammation is efficiently coupled to the central nervous system:

• Circulating cytokines and LPS alter blood–brain barrier permeability and activate brain endothelial cells, encouraging microglial activation
• LPS and inflammatory mediators can signal via humoral routes and via neural pathways (e.g., vagal afferents) to brainstem and limbic structures, further modulating stress and immune tone

An intestinal barrier problem is thereby rapidly translated into microglial priming and activation, embedding gut-origin signals into the central branch of the cascade. The gut node functionally sits upstream of the Level 6A reactive glia layer.

Section 5

Chronicity: Why "Metabolic Endotoxemia" Is Self-Sustaining

Once dysbiosis and barrier failure are established, baseline LPS flux remains elevated, continually re-stimulating TLR4. The immune system receives a constant low-grade "danger" signal, maintaining NF‑κB activity and cytokine production. Microglia, previously primed by factors such as MIA, respond more vigorously and remain activated longer, reinforcing the neuroinflammatory arm. No new external infection is required to sustain the loop.

Section 1

ROS and mtDNA as Triggers of the NLRP3 Inflammasome

When the electron transport chain is dysfunctional, electron leakage at complexes I and III increases ROS formation. Oxidative damage to mitochondrial membranes and genomes promotes release of mtDNA into the cytosol and extracellular space. Both ROS and mtDNA serve as signals activating the NLRP3 inflammasome in microglia and other myeloid cells:

• ROS act as upstream stress signals that favor NLRP3 assembly
• Oxidized mtDNA acts as a DAMP, binding to and stabilizing inflammasome complexes

Once NLRP3 is activated, caspase‑1 processes pro-IL‑1β and pro-IL‑18 into their mature, secreted forms — translating mitochondrial injury into a specific cytokine signature that reinforces microglial activation.

Section 3

Integration with the Broader Cascade

• Upstream: SIRT1 and PGC‑1α deficits (established by MIA, IDO1-driven NAD⁺ strain, and chronic inflammation) impair mitochondrial biogenesis and repair, increasing the likelihood of persistent mitochondrial injury
• Lateral: Environmental/oxidative stressors further damage mitochondrial membranes and respiratory chain components, amplifying ROS and mtDNA release
• Downstream: Elevated IL‑1β and IL‑18 reinforce microglial M1 activation, promote A1 astrocyte conversion, and feed back into NF‑κB and NLRP3 activity

Positioning mitochondria at Level 1 emphasizes that bioenergetic failure and inflammatory amplification begin early and co-evolve, rather than appearing as late secondary complications.

Section 1

Redox Disruption: Glutathione Depletion and Oxidative Burden

Many environmental toxicants relevant to ASD risk place stress on the GSH system — either by direct conjugation during phase II detoxification, by increasing ROS generation, or by impairing enzymes involved in glutathione synthesis and recycling. As reduced GSH declines and oxidized forms accumulate:

• The capacity to neutralize ROS diminishes
• Oxidative damage to lipids, proteins, and nucleic acids increases, reflected in markers such as 8-hydroxy-2′-deoxyguanosine (8-OHdG)

In the wider cascade, this redox imbalance interacts with the FOXO/Nrf2 axis and further constrains SIRT1/PGC‑1α-mediated resilience, as mitochondrial and nuclear damage accumulate under insufficient antioxidant defense.

Section 2

Innate Immune Activation and Microglial Priming

Environmental toxicants can act as direct or indirect activators of innate immune pathways:

• Some metals and organic pollutants stimulate pattern-recognition receptors, inflammasomes, or complement pathways, leading to increased pro-inflammatory cytokine production
• Chronic low-level exposure appears to prime microglia, shifting them toward a state in which subsequent insults elicit stronger and more prolonged M1-like responses

Toxin load thus lowers the activation threshold of the neuroimmune system. In combination with MIA-induced glial programming and gut/LPS-mediated TLR4 stimulation, environmental toxicants contribute to a milieu where NF‑κB, NLRP3, and related inflammatory pathways are more easily and persistently engaged.

Section 4

Evidence Grading and Role in the Cascade

Environmental toxins function within the cascade as load-modulating channel inputs — not as universal or singular causes of autism. They reduce redox and detoxification reserves, enhance immune reactivity, and magnify the impact of other channels (MIA, gut/LPS, mitochondrial dysfunction, HPA/SST dysregulation), reinforcing the multi-factorial nature of the ASD biological state.

Section 2

SST Interneuron Development as a Pre-Set Vulnerability

MIA disrupts the development of somatostatin-expressing (SST) interneurons, which are crucial for dendritic inhibition, gain control within cortical columns, and regulation of local plasticity. Perturbations in progenitor proliferation, migration, or differentiation can yield:

• Quantitative loss of SST interneurons in key regions
• Qualitative alterations in their intrinsic properties or connectivity

SST is later treated as a co-equal central node — a system-wide plasticity brake suppressing AC/cAMP/PKA/CREB signaling and altering gut and glial function. Developmental priming of SST circuitry implies that the system enters postnatal life with both a structural deficit in SST interneuron networks and a heightened susceptibility to SST-mediated suppression when chronic stress and inflammation emerge. The SST arm of the spiral is therefore not purely reactive; it is partially pre-wired.

Section 3

Independence from Postnatal Triggers

The core regulatory architecture (SIRT1/PGC‑1α and SST circuits) may already be shifted toward a low-resilience, low-plasticity regime at birth. Postnatal triggers then act on a system that is pre-biased toward chronic inflammation and impaired repair, so smaller insults produce larger, more persistent effects.

For such individuals, apparent "first hits" — early GI issues or environmental exposures — are actually second-order events superimposed on a prenatally established vulnerability. This explains heterogeneity in response to postnatal exposures: individuals with strong L2A priming require less additional stress to enter the self-sustaining state.

Section 1

Upstream Activation: Coupling of SST to Stress and Inflammatory Load

SST is positioned downstream of chronic stress signaling and inflammatory tone:

• HPA axis dysregulation, sustained cortisol abnormalities, and metabolic stress (e.g., hyperinsulinemia) increase SST release from hypothalamic and extrahypothalamic sources
• Pro-inflammatory cytokines and ongoing microglial activation further modulate SST expression and release across brain regions and in the gut

This coupling means that the same conditions driving IDO1 activation and functional NAD⁺ insufficiency also drive chronic SST elevation — shifting the system into a state where both the metabolic arm (SIRT1 fuel) and the neuropeptide arm (SST signaling) act together to suppress plasticity-enabling pathways.

Section 2

Receptor Distribution and Multi-Compartment Effects

Somatostatin acts primarily via SSTR2 and SSTR5, Gi-coupled receptors expressed on multiple cell types:

• Hippocampal and cortical neurons: SSTR2/5 activation reduces excitability and dampens activity-dependent plasticity
• Astrocytes: SST signaling modulates synaptogenic support functions, including hevin and glypican expression, favoring a more restrictive synaptic environment
• Microglia: SST influences surveillance and activation states, biasing toward lower plasticity and higher threat sensitivity
• Thalamic relay cells: SST may alter thalamocortical transmission, impacting sensory gating and relay fidelity

Section 5

System-Wide Consequences: Multi-Organ Impact

Because SST acts in both central and peripheral systems, its chronic elevation has multi-organ implications aligned with ASD-relevant phenotypes:

• In the brain: reduced CREB/BDNF and diminished synaptogenesis contribute to rigid connectivity patterns, impaired learning, and limited capacity to integrate new experiences
• In the gut: SST suppresses gastric acid secretion, pancreatic enzymes, bile release, and motility — worsening nutrient absorption and promoting dysbiosis, feeding back into gut/LPS and IDO1 arms
• In glial compartments: SST signaling inhibits astrocytic and microglial support roles for circuit remodeling

Level 2B positions SST not as a local inhibitory neurotransmitter but as a global homeostatic shift — under chronic stress and inflammatory load, SST drives the system into a low-plasticity, high-conservation mode that simultaneously touches neuronal plasticity, glial function, and gut physiology.

Section 1

Astrocytic SSTR2/5: Suppression of Synaptogenic Proteins

SSTR2/5 activation on astrocytes reduces transcription and/or secretion of hevin and glypicans, while increasing SPARC. The functional implications:

• Hevin ↓: fewer excitatory synapses are correctly bridged and assembled, particularly thalamocortical synapses relying on hevin as a Neurexin–Neuroligin bridging molecule
• Glypicans ↓: fewer silent synapses are converted into functionally mature AMPA-receptor-containing synapses — leading to a higher proportion of structurally present but functionally weak connections
• SPARC ↑: active antagonism of hevin-mediated synapse formation; SPARC interferes with hevin's bridging function and promotes synapse destabilization

SST signaling thereby converts astrocytes from synaptogenic supporters into synapse-restricting regulators, directly worsening the synapse protein imbalance that manifests at Level 7.

Section 2

Astrocytic Glutamate Clearance: Metabolic Stress on Synapses

SSTR2/5 activation is associated with reduced glutamate uptake by astrocytes via excitatory amino acid transporters. Elevated extracellular glutamate increases the risk of excitotoxic stress and contributes to excitation–inhibition imbalance, adding a metabolic and excitotoxic component to synaptic stress — particularly in regions already energy-compromised by mitochondrial dysfunction.

Section 1

Suppression of Gastric Acid, Pancreatic Enzymes, and Bile

SST exerts potent inhibitory effects along the upper GI tract: inhibition of parietal cells reduces luminal acidity, impairing protein denaturation and peptide breakdown; reduced exocrine pancreatic secretion compromises digestion of proteins, fats, and carbohydrates; inhibition of cholecystokinin-mediated gallbladder contraction reduces lipid absorption and fat-soluble nutrient uptake. Together, these effects create a functional hypodigestive state, in which macronutrients and cofactors are incompletely processed before reaching the small intestine.

Section 3–4

Microbiome Shifts, Dysbiosis, and the Bidirectional Stress–Gut–Brain Loop

Lower gastric acidity permits increased survival of ingested organisms; reduced pancreatic enzymes and bile change nutrient profiles available to colonic bacteria; slowed transit promotes SIBO-like states. These changes promote dysbiosis, which increases harmful metabolite production and raises the likelihood of barrier disruption and LPS translocation — feeding back into the gut/LPS → TLR4 → NF‑κB inflammatory channel.

The bidirectional loop runs:

Level 2D positions SST on the gut as a key coupling point: it translates chronic stress and inflammation into persistent gut–brain disruption that reinforces the IDO1/kynurenine pathway, mitochondrial stress, and microglial activation.

Section 1

IDO1 Induction: Inflammatory Control of Tryptophan Fate

Pro-inflammatory cytokines (IFN‑γ, TNF‑α, IL‑6) and innate immune signals (e.g., TLR ligands such as LPS) induce IDO1 transcription and activity in dendritic cells, macrophages, and intestinal and hepatic cells. IDO1 converts tryptophan to N-formylkynurenine, initiating the kynurenine pathway. Two immediate consequences follow: reduced tryptophan availability for serotonin synthesis, and increased flux into downstream kynurenine metabolites.

Core principle

Substrate Dependence: Why NAD⁺, Not Just SIRT1 Expression, Is Central

SIRT1 requires NAD⁺ as an obligate cofactor. During each catalytic cycle, SIRT1 consumes NAD⁺ and generates nicotinamide and O-acetyl-ADP-ribose. When NAD⁺ concentrations fall below a functional threshold, SIRT1's deacetylation capacity decreases irrespective of how much SIRT1 protein is present. In the cascade, this frames SIRT1 failure as primarily fuel-limited: the enzyme is present but under-fed, leading to de facto loss of its regulatory influence.

NAD⁺ is further depleted by increased PARP activity during DNA repair and elevated CD38 activity during immune activation — both consequences of the chronic inflammatory state driven by upstream cascade levels. The defining insight of Level 3B is that four major regulatory domains fail together once SIRT1 is under-fueled:

• Inflammation: NF‑κB restraint lost → p65 remains hyper-acetylated → chronic cytokine elevation
• Mitochondria: PGC‑1α under-activated → reduced biogenesis and higher ROS
• Antioxidants: FOXO-linked defenses blunted → glutathione and related systems less adaptive
• Plasticity: CREB/BDNF support weakened → impaired synaptic consolidation and circuit updating

Section 1–2

SIRT1–AMPK Coordination and mTOR Overactivity

Under physiological conditions, SIRT1 and AMPK form a coordinated energy-sensing module: AMPK is activated by increased AMP/ATP and ADP/ATP ratios; SIRT1, fueled by NAD⁺, deacetylates substrates promoting catabolic pathways and stress resistance. Together, they suppress anabolic signaling when energy is scarce and promote autophagy and mitophagy.

Under NAD⁺ insufficiency and chronic metabolic strain, SIRT1 activity declines, weakening AMPK-supporting pathways. mTOR, a key growth and autophagy-suppressing kinase, remains relatively overactive — suppressing macroautophagy (formation and maturation of autophagosomes) and mitophagy (targeted autophagic removal of dysfunctional mitochondria). The cell is locked into a non-resetting growth mode while damage accumulates.

Section 5

Genetic Evidence: PTEN/TSC/Fragile X mTOR Axis as Proof-of-Concept

This AMPK/mTOR/autophagy axis is well-documented in genetically defined ASD subtypes:

• PTEN mutations: alter PI3K/Akt/mTOR signaling; associated with macrocephaly and ASD
• TSC1/TSC2 mutations (Tuberous Sclerosis Complex): mTOR is dysinhibited due to loss of TSC-mediated suppression
• Fragile X syndrome: FMRP loss affects synaptic translation and mTOR-linked pathways

The cascade generalizes this mechanistic principle to idiopathic ASD, positing that inflammatory and metabolic stress can engage the same axis even in the absence of canonical mTOR-pathway mutations.

Core argument

Why Damage Becomes Effectively Permanent

Autophagy failure explains three critical observations: (1) damaged mitochondria continue to accumulate even after upstream triggers are partially reduced; (2) excess dendritic spines and local hyper-connectivity persist instead of normalizing; and (3) inflammatory signaling remains self-sustaining because its endogenous substrates — dysfunctional organelles and debris — are never adequately cleared.

• Damaged mitochondria retained → ROS and mtDNA continuously released → NLRP3 repeatedly re-activated, sustaining caspase‑1, IL‑1β, and IL‑18 independently of new external triggers
• Dendritic spine over-retention → excess spine density in certain cortical regions; synapses that are mislocalized, weak, or maladaptive are not efficiently removed → local over-connectivity with reduced long-range integration
• Protein aggregate and debris accumulation → extracellular material activates pattern-recognition receptors on microglia and astrocytes → continuous NF‑κB activation independent of new external insults

Mechanism

SIRT1–FOXO and the High-Output, Low-Clearance Oxidative State

Deacetylated FOXO (e.g., FOXO3a) drives transcription of catalase, superoxide dismutases, peroxiredoxins, and glutathione system components. When SIRT1 activity falls, FOXO remains acetylated and less transcriptionally effective. GSH levels fall; the balance shifts toward GSSG; the cell's capacity to detoxify ROS and reactive metabolites is reduced.

Astrocytes are particularly sensitive to redox imbalance. Under GSH depletion and ROS excess, astrocytes experience oxidative damage to membranes, organelles, and secretory machinery. For hevin (SPARCL1) specifically, oxidative and ER stress can promote misfolding and impaired secretion — reducing functional hevin available to support thalamocortical synapse formation and directly contributing to the synaptogenic protein deficit at Level 7A.

The overall picture is a high-output, low-clearance state for oxidative species: PGC‑1α under-activation increases ROS sources, FOXO under-activation reduces ROS sinks, and the resulting imbalance amplifies mitochondrial injury, glial reactivity, and synaptic instability.

Section 1

NF‑κB-Driven Transition from Homeostatic to M1-Like State

Under physiological conditions, microglia maintain a homeostatic surveillance profile — continuously monitoring the microenvironment and supporting synaptic maturation and pruning in a tightly regulated manner. Chronic NF‑κB activation — driven by SIRT1 failure, LPS/TLR4 signaling, mitochondrial distress, and toxin load — alters microglial gene expression, upregulating surface receptors, cytokines, and enzymes associated with M1-like activation while reducing the threshold for activation by additional stimuli (debris, DAMPs).

Section 1–2

A1 Markers and the Reversal of the Synaptogenic Secretome

A1 astrocytes exhibit a characteristic molecular signature: GFAP ↑ (reactive gliosis morphology), S100β ↑ (heightened astrocytic activation), and complement C3 ↑ (marking engagement of complement pathways that contribute to synapse tagging and elimination). These markers signal a state less focused on metabolic and synaptogenic support, and more involved in inflammatory and complement-mediated processes.

Beyond reactivity markers, A1 astrocytes undergo a functional inversion of their synaptogenic secretome:

• Under supportive (A2-like) conditions: astrocytes secrete hevin (SPARCL1) and glypicans, promoting synapse formation and maturation
• In the A1 state: hevin and glypican expression and secretion decline, while synapse-antagonistic factors such as SPARC increase

The extracellular environment becomes actively unfavorable for stable excitatory synapse formation and maturation. Existing synapses receive less trophic support and more destabilizing signals, predisposing to synaptic weakening and loss — particularly where complement tagging by microglia is also active.

Section 1–2

Hevin as Neurexin–Neuroligin Bridge and Its Suppression in A1 Astrocytes

Hevin is an astrocyte-derived matricellular protein that acts as a critical organizer of excitatory synapse formation. It physically bridges Neurexin‑1α on the presynaptic terminal and Neuroligin‑1 on the postsynaptic site, facilitating the initial assembly and stabilization of thalamocortical synapses — which are central to sensory relay and integration.

In the A1 reactive astrocyte state, hevin production is markedly reduced: A1 astrocytes downregulate SPARCL1 transcription and/or secretion, reducing the amount of functional hevin available in the extracellular space. The initiation of new thalamocortical synapses is less efficient and less frequent, particularly in regions already stressed by inflammation and mitochondrial dysfunction.

Section 1–2

SPARC as Hevin Antagonist and Pathological Pruning Agent

SPARC directly interferes with hevin-mediated synaptogenesis: it binds hevin, preventing hevin from effectively connecting Neurexin‑1α to Neuroligin‑1. In the presence of elevated SPARC, even when hevin is present, its synaptogenic capacity is functionally neutralized — the system shifts from synapse-building to synapse-blocking mode.

In A1 reactive astrocytes, SPARC is massively over-expressed in response to TNF‑α, IL‑1β, and microglial activation — entering a pathological range where it not only modulates but actively removes synapses that developmental circuits required. Synapses in critical circuits (thalamocortical sensory relays, social processing networks) are inappropriately eliminated or destabilized, contributing to under-connectivity in key long-range and relay pathways.

Mechanism

Glypicans as Synapse Maturers and the Silent Synapse Problem

Glypicans 4 and 6 stimulate insertion and stabilization of AMPA receptors (GluA1-containing) at postsynaptic densities, converting "silent" synapses — those with NMDA receptors but minimal AMPA-mediated current — into fully functional synapses participating in normal excitatory transmission.

A1 reactive astrocytes exhibit irregular glypican 4/6 secretion: expression levels, timing, or release patterns become dysregulated, and local extracellular concentrations no longer reliably support the conversion of silent synapses into fully AMPA-competent synapses. Many structurally formed synapses consequently remain weak or functionally silent, contributing little to effective circuit throughput despite their morphological presence.

The net effect is a decoupling between anatomical and functional connectivity: imaging or histological measures may show preserved or even excessive spine density, while electrophysiologically many synapses are inefficient or silent. Networks appear over-connected structurally but under-effective functionally — a configuration supporting rigid, low-flexibility dynamics.

Mechanism

Hevin Dependence, Relay Failure, and Unfiltered Cortical Input

Thalamocortical synapses are particularly hevin-dependent: hevin's bridging of Neurexin‑1α (thalamic presynaptic) and Neuroligin‑1 (cortical postsynaptic) is a key molecular step in assembling these long-range excitatory synapses. When hevin is suppressed, the number of successfully formed thalamocortical synapses declines, and those that do form may be weaker, less stable, and more vulnerable to SPARC-mediated elimination.

The result is an under-connected sensory relay: fewer effective synaptic contacts, reduced capacity for precise gating and state-dependent modulation of sensory input. Thalamus normally provides a gated, pre-processed stream of sensory data to cortex; with fewer and weaker synapses, cortical regions receive input that is less modulated by normal thalamic gating mechanisms, less synchronized, and less reliably timed.

This mismatch between weak relay structure and locally over-connected, low-plasticity cortical circuits can produce heightened sensitivity to some stimuli (due to local amplification of poorly gated input) and fragmented or incoherent cross-modal integration — cortex is asked to interpret sensory signals that arrive without the usual thalamic scaffolding.

Mechanism

Coexistence of Over- and Under-Connectivity

Several upstream factors promote excess local connectivity: autophagy/mitophagy impairment reduces synaptic pruning, allowing excess dendritic spines and local synapses to persist; A1 astrocyte conversion and glypican irregularity interfere with normal activity-dependent refinement; SPARC and complement activity may selectively remove some synapses while leaving clusters of others intact.

In contrast, several mechanisms selectively weaken long-range and relay connectivity: hevin suppression reduces formation of thalamocortical and hevin-dependent excitatory synapses critical for inter-regional communication; SPARC over-expression destabilizes association connections; glypican irregularity leaves many formed synapses functionally immature or silent.

The key conceptual point is that both over- and under-connectivity are present simultaneously, but at different scales and locations:

• Local connectivity ↑ — excess insufficiently-pruned synapses within cortical regions; high clustering of short-range connections contributing to intense, narrowly focused processing
• Long-range integration ↓ — reduced effective connectivity between distant hubs and between thalamus and cortex; weak coordination across networks necessary for global integration of sensory, social, and cognitive information

This pattern aligns with neuroimaging findings in ASD — increased local functional and structural connectivity in certain cortical regions alongside decreased long-range connectivity between major networks (default mode, salience, fronto-temporal, sensory). It provides a network-level substrate for coexisting strengths (hyper-specialized local functions) and weaknesses (impaired cross-network coordination and flexible adaptation) in the ASD profile.

Mechanism → Expression

Under-Gated Relay, Cortical Amplification, and Clinical Presentation

Reduced hevin-dependent thalamocortical synapse formation and weakened long-range relay connectivity mean that sensory signals reach cortex with insufficient thalamic gating and preprocessing. Rather than being filtered, prioritized, and integrated before entering conscious awareness, sensory input arrives in a relatively raw and unmodulated form.

Cortical regions receiving this input operate within a network characterized by local hyper-connectivity, immature synapses, and reduced long-range integration. Sensory signals are less synchronized and less context-modulated, but may be locally amplified by dense, rigid microcircuits — experienced as intense, poorly organized, and difficult to predict, rather than smoothly graded and contextually framed.

At the behavioral level, this network configuration manifests as sensory over-reactivity: heightened or atypical responses to sound, touch, light, and texture; over-responsivity reported in a large majority of autistic individuals. These are not isolated sensory quirks but expected downstream expressions of the thalamocortical and connectivity alterations described in earlier cascade levels.

Mechanism → Expression

Social Circuit Under-Development as a Network Integration Problem

Complex social understanding depends on long-range integration among medial prefrontal cortex (mPFC) (mental state inference, value and intention attribution), superior temporal sulcus (STS) (perception of biological motion, gaze, prosody), and amygdala (emotional salience and reward in social signals) — along with temporoparietal junction, posterior cingulate, and associated hubs. Effective social cognition requires synchronized, bidirectional communication across these regions.

Hevin suppression, SPARC over-expression, glypican irregularity, and autophagy/pruning dysregulation reduce formation and maintenance of long-range excitatory synapses. Structural and functional connectivity among mPFC, STS, and amygdala falls below the level needed for efficient joint processing of social information.

The behavioral manifestations follow directly from this network-integration failure:

• Theory of mind — impaired because information from perceptual (STS), emotional (amygdala), and evaluative (mPFC) nodes is not reliably combined into cohesive representations
• Pragmatic language — compromised, as it depends on integrating linguistic content with social and emotional context across distributed regions
• Social reciprocity — reduced because the system lacks fast, coordinated updates across the network in response to dynamic social input

Mechanism → Expression

Why Circuits Fail to Update and Remain Structurally "Stuck"

Two upstream mechanisms converge: (1) CREB/BDNF impairment — SIRT1 deficiency reduces CREB co-activation and BDNF expression, while SST-mediated AC/cAMP/PKA suppression further reduces CREB phosphorylation; (2) persistence of silent synapses — glypican 4/6 irregularity leaves many synapses in NMDA-only (silent) states with poor AMPA-receptor recruitment. Together, these factors prevent circuits from updating their synaptic weights effectively in response to experience.

Activity-induced synaptic changes are less likely to be consolidated; circuits have reduced capacity to encode new contingencies, rules, or environmental regularities. Many synapses are incapable of supporting robust potentiation due to poor AMPA incorporation. Networks carry a significant load of ineffective connections that cannot reliably support propagation of new activity patterns — traffic repeatedly traverses the same limited subset of strong connections, while many potential alternative pathways remain under-engaged.

At the behavioral level:

• Cognitive rigidity — difficulty shifting mental sets, strategies, or interpretations in response to new information
• Difficulty with transitions — trouble disengaging from one activity or routine; an inability to reconfigure circuit dynamics smoothly
• Restricted interests — intense, narrow focus consistent with strong, well-established local circuits and weak flexibility to broaden engagement
• Perseverative patterns — repetitive thoughts or behaviors reflecting the dominance of well-worn neural routes that are not easily modified

Mechanism → Expression

Three Converging Mechanisms; One Shared Inflammatory Origin

Three mechanisms converge: gut dysbiosis and barrier failure, serotonin depletion via the kynurenine pathway, and vagal nerve-mediated inflammatory signaling.

• Gut dysbiosis and barrier failure (L1B): maintains LPS flux and chronic TLR4–NF‑κB activation; locally reinforces epithelial tight junction damage and increased permeability ("leaky gut")
• Serotonin depletion (L3A): IDO1-driven tryptophan diversion depletes enteric serotonin (5-HT), impairing normal regulation of motility, secretion, and sensory signaling. Dysmotility manifests as constipation, irregular transit, or alternating patterns; altered serotonin signaling also influences visceral sensitivity and pain perception
• Vagal inflammatory signaling: inflamed and dysbiotic gut tissue activates vagal sensory fibers projecting to brainstem nuclei, modulating autonomic balance, stress pathways, and microglial activation — providing a rapid, bidirectional gut–brain communication channel

The clinical GI phenotype — dysmotility, constipation, visceral pain, and leaky gut — is not an incidental comorbidity but a direct consequence of the same processes (LPS/IDO1, NAD⁺ insufficiency, SST elevation, glial activation) that shape brain development and connectivity. GI symptoms provide a peripheral window into the activity of the central inflammatory and metabolic cascade.

Mechanism → Expression

Two Converging Pathways to Sleep and Mood Disruption

Pathway 1 — Serotonin/melatonin: IDO1-driven tryptophan diversion reduces substrate availability for serotonin synthesis. Reduced serotonin in the pineal and related tissues impairs synthesis of melatonin (derived from serotonin via N-acetylation and O-methylation). Melatonin insufficiency disrupts circadian signaling, causing impaired sleep onset, fragmented sleep, and altered timing of sleep-wake cycles. Circadian dysregulation then feeds back into stress and metabolic systems, further destabilizing the SIRT1/SST balance and autophagy/cleanup processes.

Pathway 2 — HPA axis: Chronic stress and inflammatory load contribute to HPA axis dysregulation — initially elevated or flattened cortisol curves with loss of normal diurnal variation, and over time cortisol blunting or exhaustion patterns. Abnormal cortisol dynamics are linked to emotional dysregulation, including rapid mood shifts, irritability, and difficulty recovering from stress. Anxiety and low frustration tolerance arise as circuits involved in threat detection, reward, and executive control (amygdala, prefrontal cortex, hippocampus) are repeatedly activated under dysregulated hormonal conditions. HPA dysregulation also maintains elevated SST tone, creating an additional feedback that further suppresses plasticity and gut function.

Sleep dysregulation and mood instability are not isolated psychiatric comorbidities but predictable downstream outputs of inflammation-driven tryptophan diversion, chronic stress, and HPA/SST activation — integral expressions of the same inflammatory, metabolic, and stress-regulatory framework driving the core neural and behavioral features of autism.

Intervention axes

Coordinated Upstream Levers — Not a Single Target