Critical Evaluation of “Psilocybin triggers an activity-dependent rewiring of large-scale cortical networks” (Jiang et al., 2026)
Scientific Rigour
Hypothesis & Rationale: The study is grounded in a clear hypothesis about lasting neuroplastic changes induced by psilocybin. Prior work showed that a single psilocybin dose can spur new dendritic spine formation in mouse frontal cortex. Jiang et al. build on this by asking which presynaptic neurons form those new connections. This question is well-justified given evidence that depressed patients have synaptic deficits and psychedelics may reverse them . The rationale connects clinical observations (e.g. long-lasting antidepressant effects of one psilocybin dose) to a mechanistic inquiry about brain circuit rewiring.
Methodological Approach: The experimental design is appropriately comprehensive and state-of-the-art. The authors used monosynaptic rabies virus tracing to map input neurons throughout the brain that connect to defined frontal cortical pyramidal cells. This technique is highly suitable for identifying connectivity changes, as it labels only first-order presynaptic partners of “starter” neurons . By employing a Fezf2-CreER transgenic mouse, they specifically targeted PT-type pyramidal neurons (projection neurons) in medial frontal cortex as starter cells . This cell-type specificity is a strength, since PT neurons are hypothesised to be key players in subcortical output and plasticity. The study didn’t rely on just one method: they corroborated the tracing results with two-photon imaging of axonal boutons and both in vitro and in vivo electrophysiology. This multi-modal approach greatly enhances reliability. For example, two-photon microscopy revealed an increase in retrosplenial cortex (RSP) axonal boutons in frontal cortex after psilocybin , validating the rabies-tracing finding of strengthened RSP→frontal connectivity. Likewise, slice recordings confirmed a functional enhancement of RSP synapses onto PT neurons (increased optogenetic EPSC amplitude) after psilocybin , and in vivo Neuropixels recordings showed heightened firing of RSP→frontal neurons during psilocybin exposure . Using chemogenetics to silence RSP activity during drug administration was an innovative addition to test causality. In sum, the methodology is cutting-edge and well-matched to the questions, combining anatomical mapping with functional assays for a rigorous systems-neuroscience investigation.
Controls & Statistical Analysis: The study design included appropriate controls at every step. Mice were given either psilocybin or saline, with identical timing for virus injections, to isolate drug effects . Importantly, the rabies tracing had no significant baseline differences between groups: the number of starter cells and total input cells labelled were virtually the same for psilocybin vs. saline mice (e.g. ~2.5×10^3 starters and ~5.6×10^5 inputs in each, p≫0.05). This indicates the virus transduction and tracing efficiency were equivalent, lending confidence that any differences in input patterns are due to psilocybin, not technical artifacts. The authors also ran multiple control injections: e.g. helper virus lacking rabies glycoprotein, AAV alone, or rabies alone, none of which produced artifactual labelling. Such controls demonstrate the tracing method’s specificity and integrity. Statistical analysis was rigorous: rather than simple t-tests, they employed linear mixed-effects models to account for paired measurements and multiple factors (e.g. drug vs. saline, cell-type, time). This is appropriate given the data structure (for instance, EPSCs were measured in PT vs. neighbouring IT cells as pairs in each slice). They report exact p-values and effect sizes, and use nonparametric tests (Wilcoxon) when needed for counts. They also performed a network-level χ² analysis to test if connectivity changes were randomly distributed or network-specific, finding a highly significant selective pattern. These analyses go beyond superficial stats and strengthen the validity of the conclusions. Overall, the use of proper controls (saline injections, within-slice pairing, etc.) and advanced statistics reflects strong experimental rigor.
Reproducibility & Clarity of Experimental Design: The experiments are described with sufficient detail and appear reproducible. Key steps (drug timing, virus injection sites, doses) are clearly outlined in text and figures. The timeline (virus injection → wait 2 weeks → drug or saline → 1 day later rabies virus → 1 week later analyse) is logical and ensures that labelled connections represent those present after the drug’s effects. The sample sizes are reasonable given the complexity (e.g. N=9 psilocybin, 8 saline for tracing; ~6–7 mice per group for physiology). While some subgroup analyses (chemogenetic conditions) had as few as 3–4 mice, these were treated as exploratory and still yielded clear interactions. The data presentation (see Figures 1 and 2) makes it relatively easy to assess variability: they plot individual animals or cell pairs as points alongside means, which is transparent. Additionally, the authors acknowledge methodological limitations candidly. For example, they note it’s unknown if rabies virus might favour certain synapses (strength-biased spread), which drugs could potentially influence. By highlighting such uncertainties, the authors demonstrate caution in interpretation. Minor points: The rabies tracing technique inherently involves substantial analysis (counting many labelled neurons across 65 regions), but they normalised inputs as proportions and used automated whole-brain imaging, which improves consistency. All key experiments (tracing, imaging, electrophysiology) were likely done with experimenters blinded to treatment when quantifying results (not explicitly stated, but standard practice). In summary, the study’s design is meticulous and clearly communicated, supporting confidence that the findings are reliable and could be reproduced by other laboratories with similar tools.
Interpretation of Findings
Accuracy of Authors’ Conclusions: The conclusions drawn by Jiang et al. are well-supported by their data and are logically consistent. They conclude that psilocybin induces a network-specific reorganisation of cortical connectivity, specifically strengthening certain long-range inputs and weakening others. This is directly substantiated by their monosynaptic tracing results: psilocybin-treated mice showed a significant increase in the fraction of inputs from perceptual and retrosplenial areas into frontal PT neurons, coupled with a decrease in inputs from some frontal/insular cortices, relative to controls. The authors also report that these connectivity changes differ between neuron subtypes – inputs strengthened in PT-type cells tended to be concurrently weakened in IT-type pyramidal cells, and vice versa. This nuanced interpretation (a redistribution of synaptic inputs between cell classes) is backed by both the rabies data and follow-up analyses. For instance, many of the regions that gained inputs onto PT neurons (like RSP and sensory areas) had reduced connectivity onto IT neurons after psilocybin. The internal consistency here is strong: it suggests psilocybin doesn’t just add synapses indiscriminately, but rewires circuits by reallocating connections (strengthening some pathways at the expense of others). The authors’ summary statements are accurate to the evidence – e.g. “psilocybin strengthens routing of inputs from perceptual and medial (default-mode) regions to subcortical targets via PT neurons, while weakening cortico-cortical recurrent loops.” Each element of that claim is demonstrably shown in their figures. They also conclude that this pattern of synaptic change is activity-dependent, which is convincingly supported by the chemogenetic experiment: silencing RSP during drug administration abolished the increase of RSP→frontal inputs that would normally occur with psilocybin. In control mice, psilocybin raised RSP input fraction from ~4.7% to 5.9%, but under RSP inhibition, psilocybin produced no increase (3.3% vs 3.2% with saline). The authors appropriately interpret this as RSP neural activity being required for that synaptic strengthening to manifest. This causal test strongly supports their conclusion about “activity-dependent rewiring,” rather than some activity-independent mechanism.
Strength of Evidence Linking Psilocybin to Network-Specific Plasticity: The evidence presented is robust and multifaceted, making a compelling case that psilocybin truly causes network-specific synaptic changes. The shift in input sources is not only statistically significant but also substantial in magnitude for key networks. For example, inputs from the RSP (retrosplenial cortex) to frontal PT neurons increased by ~25% relative to saline, while inputs from a lateral cortical network (including agranular insular cortex) decreased by ~20%. These are sizable changes given the brain-wide baseline. Moreover, the effects were confirmed at structural and functional levels: increased anatomical connectivity was corroborated by an observed rise in axonal bouton density in frontal cortex for RSP neurons post-psilocybin, and by enhanced synaptic transmission efficacy from RSP axons (larger EPSCs and higher release probability onto PT cells). Such convergence across techniques greatly strengthens the link between psilocybin and synaptic rewiring. The network-selectivity is further bolstered by a global analysis showing the distribution of changes was non-random across five defined cortical networks (p = 6×10^−5 for non-random pattern). The authors also provide evidence that regions heavily engaged during the psychedelic experience are those that get rewired: an exploratory analysis found that brain areas with high psilocybin-evoked c-Fos activation tended to be the ones gaining inputs onto PT cells (and losing inputs onto IT cells). Since c-Fos marks neuronal activity, this finding aligns with the idea that active circuits are preferentially strengthened – an internally consistent narrative reinforcing causality. The only relatively weaker link in evidence might be the lack of direct receptor-level manipulation in this study – e.g. they did not pharmacologically confirm that a 5-HT₂A antagonist blocks these plasticity changes. However, they did examine correlations with serotonin receptor expression (Htr2a, Htr2c mRNA) across regions; while 5-HT₂A levels did not clearly predict which inputs changed, 5-HT₂C expression showed some correlation with psilocybin-induced input changes. This suggests the network effects are not trivially explained by where 5-HT₂A receptors are most abundant, but rather by circuit engagement and downstream plasticity cascades. Overall, the evidence for network-specific synaptic reorganisation due to psilocybin is very strong – it is anchored by multiple independent lines of data and a causal intervention.
Consideration of Alternative Explanations and Limitations: The authors are careful to consider and address alternative interpretations. One potential concern is whether the rabies tracing could be biased by psilocybin in ways unrelated to true connectivity changes (for example, if psilocybin somehow made certain neurons more susceptible to rabies infection or altered trans-synaptic spread efficiency). The design mitigates this by equalising the timing (virus injected after drug in all cases) and showing that total starter and input counts were unchanged by the drug. In the text, they acknowledge it’s unknown if rabies might favour crossing “strong” synapses or synapses with certain proteins (which a drug could theoretically influence), but they had no evidence this occurred. Another alternative explanation could be that psilocybin’s acute neurochemical effects (e.g. increased serotonin) might generally elevate synaptic markers without true rewiring. However, the persistence of changes at 7 days post-drug and the requirement of concurrent neural firing argue for genuine structural plasticity rather than a transient modulation. The study’s limitations are mostly inherent to its scope and models. The authors explicitly note that they only tested psilocybin, so it remains to be seen if these rewiring principles generalise to other psychedelics (though prior studies suggest LSD, DMT, etc. also promote plasticity). They also discuss that presynaptic activity alone isn’t the whole story – increased firing in a pathway is necessary but likely not sufficient for strengthening; postsynaptic factors (like dendritic excitability via 5-HT₂A activation) are also proposed as needed. This balanced consideration shows the authors are not over-simplifying the mechanism. They stop short of claiming that the observed circuit changes directly explain antidepressant effects – instead they propose it as a plausible mechanism (“could underlie therapeutic effects”) but acknowledge it’s a correlative link in need of further study (since they did not measure behaviour or therapeutic outcomes in these mice). Additionally, they recognise that weakening intracortical loops and enhancing sensory-to-subcortical pathways might reflect a shift toward bottom-up information processing, which aligns with some theories of psychedelic action, but this too is discussed cautiously rather than as proven fact. In summary, alternative explanations (technical or biological) are considered and largely addressed through controls and discussion, and the authors appropriately delineate the limits of inference (e.g. noting that they did not test multiple doses, disease models, or longer timepoints beyond a week). This careful approach lends credibility to their interpretations.
Clinical Relevance
Implications for Psychiatric Conditions: Although this is a mechanistic mouse study, the findings carry significant implications for understanding and treating psychiatric disorders like depression or PTSD. The authors’ observations suggest that psilocybin can durably remodel brain circuits that underlie how information is processed. In particular, they found a weakening of cortico-cortical “feedback” loops – essentially the recurrent connections within and between cortical areas. In a clinical context, such loops are thought to maintain rigid, self-referential thinking (for example, the rumination and negative thought spirals in depression). A Cornell press release about this work explicitly notes that psilocybin “weakens the cortico-cortical feedback loops that can lock people into negative thinking”, while strengthening pathways linking perception to action. This points to a mechanism by which psilocybin therapy might “unlock” pathological patterns: by loosening over-connected internal networks (perhaps analogous to the overactive default-mode network in depression) and simultaneously enhancing engagement with the external world and new stimuli. Furthermore, the strengthening of inputs from sensory and retrosplenial (memory/contextual) regions to subcortical action centres hints that the brain may become more sensitive to environmental inputs and emotional learning after psilocybin – a window of opportunity where therapeutic interventions (psychotherapy, positive experiences) could be more impactful. This aligns with clinical reports that patients often experience a heightened capacity to process and integrate new perspectives during psychedelic-assisted therapy. Importantly, the long-lasting nature of the synaptic changes (observed at least a week out, and by analogy to dendritic spine data, likely weeks) parallels the sustained symptom relief seen in patients for weeks or months after a single dose. The study therefore provides a biological substrate for those clinical observations: enduring synaptic rewiring might maintain the therapeutic gains even after the drug has left the body.
Translational Value (Mice to Humans): While direct extrapolation from mice to humans must be cautious, the translational insights here are notable. Many of the affected regions in mice have clear analogues in human brain networks. For instance, the mouse retrosplenial cortex is often considered analogous to the human posterior cingulate/precuneus, a core node of the Default Mode Network implicated in self-referential thinking and disrupted connectivity in depression. Psilocybin’s effect in mice of increasing RSP→frontal connectivity and simultaneously disrupting intracortical loops mirrors human neuroimaging studies where psilocybin acutely reduces default-mode network connectivity and increases global integration of sensory networks. This convergence suggests the mouse findings are capturing phenomena relevant to the human psychedelic experience and its therapeutic potential. Moreover, the demonstration that neural activity during the drug experience dictates the pattern of plasticity has important translational ramifications. It provides scientific support for the clinical emphasis on “set and setting” – i.e. the mental and environmental context during psychedelic therapy. The fact that experimentally silencing one region (RSP) in mice altered the wiring outcomes implies that in humans, engaging or not engaging specific cognitive/emotional circuits during the psychedelic session could similarly influence which synaptic connections get strengthened. This opens the door to guidedinterventions: for example, therapies might aim to activate positive emotional networks or expose patients to certain stimuli during or after psilocybin to promote beneficial rewiring, while minimising activation of potentially negative circuits. As the senior author noted, this approach could help “avoid some of the plasticity that’s negative and enhance specifically those that are positive.” It’s an exciting translational idea that therapeutic outcomes might be improved by coupling psychedelics with targeted neural activity (through behavioural exercises, neurofeedback, or even neuromodulation). Of course, translating to humans comes with challenges: the dose used in mice (1 mg/kg) is high in absolute terms, and individuals vary in receptor distributions and network organisation. The mouse study was in healthy animals; clinical populations might have different baseline connectivity (e.g. depression models show prefrontal hypo-connectivity or hyper-connectivity in certain loops). Thus, one should be careful not to overgeneralise. Still, this work provides a strong proof-of-concept for circuit-level changes underlying psilocybin’s lasting effects, thereby bridging the gap between molecular pharmacology and systems-level therapy in translation.
Risks or Overextensions in Extrapolation: The authors and the community should be mindful of a few potential overextensions when linking these findings to therapy. First, therapeutic efficacy is not proven by this study – it shows mechanism, not that these specific synaptic changes are either necessary or sufficient for alleviating depression (no behaviour was tested in the mice). It is tempting to assume weakening “negative-thinking loops” will cure depression, but mental illness involves complex network dynamics and neurochemical factors beyond just the structural connectivity. There’s a risk of oversimplification: for example, calling the affected recurrent circuits “pathological” is speculative – recurrent cortical loops are also responsible for normal cognitive functions (working memory, attention, etc.), so broadly dampening them could have downsides. Psychedelics acutely cause cognitive disorganisation, which might be related to these loops weakening; the therapeutic model assumes the brain retains useful flexibility afterwards without impairing normal function. While current data from human trials are promising, long-term risks (e.g. potential memory or executive function changes from persistent circuit rewiring) should be monitored. Another potential overreach would be assuming all the plasticity induced by psychedelics is beneficial. The study hints that some changes could be “negative plasticity” if occurring in maladaptive circuits, and that manipulating activity can steer it. This is a double-edged sword – it suggests we must ensure the patient’s brain is engaged in healthy patterns during treatment. In practice, this could be difficult to control; uncontrolled environments or adverse experiences on the drug might reinforce the wrong circuits. Thus, while the results justify optimism that psilocybin can “reset” networks, they also underscore the importance of the therapeutic container (to encourage positive rewiring). Lastly, species differences mean we should be careful: mice do not have the complex prefrontal cortical subdivisions that humans do (the mouse “medial frontal” encompasses regions analogous to human anterior cingulate and medial prefrontal cortex, but mice lack a granular prefrontal cortex). So, the exact circuits that were “rewired” are not identical to a human brain’s, even if conceptually similar networks exist. The authors’ comparisons to the default mode network and salience network are reasonable, but one should not over-interpret region names. In conclusion, the extrapolations to therapy are plausible and exciting, but should be viewed as hypotheses to be clinically tested, rather than confirmed therapeutic mechanisms at this stage. The study provides a guiding framework, not a final answer, for how to harness psilocybin’s plasticity safely and effectively in psychiatric treatment.
Figures and Data Presentation
Clarity and Integrity of Rabies Tracing Maps: The figures related to the monosynaptic tracing are data-rich yet thoughtfully organised. Figure 1, for example, includes a brain schematic and timeline (panel B) clearly illustrating the experimental design. Panel 1E shows actual images of the whole-brain labelling with starter cells and input cells colour-coded, giving readers a tangible sense of the data. The numbers of labelled neurons are presented in panel 1F with individual animal data points, which is excellent for transparency (one can see variability and group differences at a glance). Crucially, the heart of the tracing result is conveyed in panel 1I: a sorted list of all 65 presynaptic regions’ input fractions, expressed as the drug-evoked percentage change. This could have been an overwhelming wall of acronyms, but the authors improved clarity by colour-coding each region according to its network membership and highlighting those above a certain threshold of change. Thus, one can immediately pick out that, for instance, RSP and visual areas (medial and perceptual networks) have large positive changes (coloured to one side), whereas ventral insular (AIv) and related regions (lateral network) have large negative changes, without needing to decipher every label. They even include a cartoon schematic (panel 1J) marking those regions in a simplified brain diagram, which is very helpful for visual learners. The integrity of these figures appears high: the data are shown with appropriate statistics (confidence intervals in panel 1I, significance asterisks or p-values reported for panel 1F) and nothing seems selectively hidden. Supplementary figures further validate that labelling was confined to intended neurons (controls in Fig. S1) and categorise PT vs IT neurons (Fig. S2), indicating the authors took care to ensure the maps are accurate. In summary, the rabies tracing data is presented in a manner that is comprehensive but still interpretable, thanks to network grouping and visual aids. This supports the authors’ claims by literally mapping those claims onto the brain for the reader.
Visualisation of Electrophysiology and Other Data: The figures depicting functional experiments (electrophysiology and imaging) are similarly well-designed to support the conclusions. In Figure 4, which covers the slice electrophysiology, the authors pair example illustrations with quantitative summaries. Panels 4E–F likely show the experimental setup (e.g. a histological image of PT neurons labelled with tdTomato and ChR2-expressing RSP axons, to prove targeting). Panel 4H then schematises the key comparison – recording from a PTFezf2 neuron versus a neighbouring IT neuron in the same slice. By arranging recordings as PT/IT pairs, they controlled for slice-to-slice variability, and the figure reflects this by plotting relative EPSC amplitude of PT vs IT under each condition. The graphs (panels 4J–4O) show clear differences: after psilocybin, PT cells have larger evoked EPSCs relative to IT cells, whereas in controls they were similar. The inclusion of multiple timepoints (24 h and 3 d) in the figure demonstrates that the effect persisted, which visually reinforces the text’s statement that potentiation lasted at least 3 days. Error bars and individual data points (if shown as in text description) indicate variability and seem appropriate (the stats in text show very low p-values, implying the figure likely shows a noticeable separation between groups). The electrophysiology figures also examine mechanisms: for instance, panel 4P and 4Q might display paired-pulse ratio (PPR) changes, where psilocybin-treated slices showed a lower PPR in PT cells (signifying increased presynaptic release probability). These subtle points are communicated with clarity – e.g. using bar graphs or line plots comparing PPR across conditions, with asterisks to denote significance. The fact that they even include supplemental figures (Fig. S6) for inhibitory currents and TTX/4-AP tests shows a high level of data transparency; while not all of that can be shown in the main figures, the main text directs interested readers to it.
For the in vivo Neuropixels recordings (Figure 5), the figure provides a step-by-step visual story: Panel 5A–B outline the experiment timeline and design (including an illustration of the head-fixed mouse, drug infusion, and opto-tagging protocol). Panel 5C–D show verification of electrode placement (a fluorescent probe track in RSP), confirming they recorded from the intended region. Example data are given in rasters (panel 5E–F) differentiating an opto-tagged neuron vs an untagged neuron during laser flashes – this visually substantiates how they identify RSP→frontal projection neurons among all recorded cells. The firing rate results are likely depicted in a before-after plot or bar graph (panel 5G–I), which in text showed a ~39% firing rate increase in tagged neurons after psilocybin, versus no change in saline. The figure apparently uses scatterplots with crosshairs to show mean ± SEM and individual neuron changes, aiding interpretation of variability. They even illustrate local field potential spectral changes in a separate panel (with heatmaps or spectra, panel 5M–N) to note the increase in low-frequency oscillations unique to psilocybin. All these visual elements are directly supportive of claims: e.g., one can seethat only under psilocybin do the opto-tagged cells ramp up firing, which aligns with the idea that presynaptic RSP activity is specifically boosted by the drug.
The chemogenetics experiment in Figure 6 also appears well-presented. Panel 6A–C likely depict the approach (showing an image with mCherry-labeled RSP neurons expressing the inhibitory DREADD, and EGFP-labelled input cells in the same region). This visually confirms that they successfully silenced the targeted region during the experiment. Panels 6D–E plot the key outcome: with vehicle, psilocybin increases RSP input fraction (~5.9% vs ~4.7% in saline), but with DCZ (DREADD agonist), psilocybin’s effect is nullified. The figure presumably shows this as bar graphs for each condition with error bars and significance annotations, which makes it straightforward to appreciate the interaction (indeed the text notes a significant drug×chemogenetic interaction p=0.046). They also examine other regions in panels 6F–G, demonstrating that silencing RSP did not alter psilocybin’s effects on, say, thalamic inputs (VPM) or on insular cortex inputs, reinforcing specificity. The visual contrast between panels (some bars changing with drug, others flat) helps convey that RSP activity was selectively required for RSP-related rewiring but not for all changes.
Overall, the data visualisation is excellent and supports the claims without distortion. The figures are somewhat dense (understandably, given the breadth of experiments), but each is well-labelled and accompanied by clear legends. Important trends are highlighted through colour or schematic cartoons (for instance, the final Figure 7 provides a summary cartoon of how networks are rewired: thicker lines for strengthened inputs from RSP/sensory, thinner lines for weakened cortical loops). This helps distil complex results into an intuitive image, reinforcing the take-home message graphically. There is no indication of any data manipulation beyond standard presentation; if anything, the authors err on the side of showing more data (e.g. individual data points, multiple timepoints) rather than less, which speaks to the integrity of figure presentation. In conclusion, the figures are well-crafted to bolster the paper’s conclusions and are presented with a high level of clarity and honesty, making it easy for readers to follow the evidence for each claim.
Overall Contribution and Significance
Novelty and Significance in Psychedelic Neuroscience: This study makes a highly novel contribution by mapping how a psychedelic drug rewires specific brain networks at the synaptic level. Prior to this work, scientists knew that psychedelics like psilocybin can induce synaptic growth (e.g. dendritic spines) and cause acute changes in brain activity, but these phenomena had not been connected into a coherent, network-level picture. Jiang et al. are the first to provide a brain-wide circuit map of psilocybin’s lasting effects, revealing that the drug’s impact is not a uniform increase in connectivity, but a targeted reshaping of pathways. This finding is significant because it moves beyond generic statements that “psychedelics cause neuroplasticity” into identifying which circuits are strengthened and which are weakened. The pattern they uncovered – enhanced sensory/association inputs to subcortical-projecting neurons, diminished recurrent cortical interactions – is striking and offers a concrete neural mechanism that could explain both the subjective effects of psychedelics and their therapeutic outcomes. It introduces the concept that psychedelic-induced plasticity is not random, but rather depends on neuronal activity and circuit usage during the drug experience. This is a novel insight with broad implications for neuroscience: it suggests a parallel to “fire together, wire together” principles of developmental plasticity, now applied in an adult psychedelic context. In fact, the authors explicitly draw an analogy to experienced-dependent synapse formation in development (like visual cortex changes with sensory deprivation or stimulation). By showing that a modern psychiatric drug can recapitulate activity-dependent circuit reorganisation, the paper opens up new lines of research into harnessing critical-period-like plasticity in adult therapy. The multidisciplinary approach (viral tracing, imaging, physiology, chemogenetics) is itself a valuable model for integrative neuroscience research, setting a high standard for future studies.
The significance also lies in bridging different scales of investigation: it connects molecular-level action (5-HT receptor activation, immediate-early genes) to synapse-level changes (new inputs), to network-level outcomes (altered information flow), all the way to potential behavioural effects (e.g. reduced negative thinking loops). This kind of multi-level understanding is crucial in neuropsychiatry, and this paper provides one of the clearest examples to date of linking synaptic plasticity to large-scale brain network reconfiguration in the context of a therapeutic compound. Given the current renaissance in psychedelic research, these findings are likely to be highly influential. They not only validate the idea that psychedelics are “psychoplastogens” (substances that enhance neural plasticity), but refine it by showing the plastic changes follow an orderly pattern shaped by neural activity. This injects a new level of sophistication into theories of how psychedelic therapy works, supporting frameworks like the REBUS model (which posits relaxed high-level priors and heightened sensory influence) with concrete anatomical evidence. In summary, the study’s novelty lies in demonstrating activity-dependent circuit rewiring by a psychedelic, and its significance is amplified by the clinical relevance of those circuits to mental health. It stands to substantially impact both basic neuroscience (e.g. models of learning and plasticity) and clinical science (guiding more effective therapeutic protocols).
Context and Comparison with Prior Studies: This work stands out in comparison to prior research on rapid-acting antidepressants and psychedelics, while also building upon their foundations. In the realm of ketamine (a fast-acting antidepressant), earlier studies had shown that ketamine causes new spine formation in prefrontal neurons and elevates synaptic connectivity, correlating with depressive behaviour reversal. However, ketamine studies largely focused on local changes in the prefrontal cortex and did not map long-range network rewiring. Jiang et al. push the envelope by revealing that psilocybin’s synaptic effects span distributed networks involving posterior cortical and thalamic regions, not just frontal cortex. Interestingly, they note overlap with ketamine’s functional targets: for example, both psilocybin and ketamine implicate the deep retrosplenial/posterior cingulate area (ketamine is known to disrupt the posterior DMN leading to dissociative experiences, similar to psilocybin’s effect on RSP connectivity). An earlier study by the same group compared immediate-early gene (IEG) activation by ketamine vs psilocybin and found some shared regions and some distinct. The present paper extends those observations from transient IEG activation to lasting structural changes. It suggests that while both compounds induce plasticity, psilocybin may have a broader or different network impact (engaging sensory/association inputs), whereas ketamine might preferentially act on frontal-limbic circuitry (this study did not map ketamine, but future comparisons can be made). In any case, the concept of network-guided plasticity is a novel contribution that complements prior findings of synaptogenesis by psychedelics (e.g. Ly et al., 2018, who showed psychedelics grow neurons in vitro). It’s also worth noting the comparison to studies that mapped brain activity under psychedelics: human fMRI studies (Carhart-Harris et al., 2017; Preller et al., 2020) reported increased global connectivity and decreased integrity of certain networks under psilocybin. The Jiang et al. results provide a structural connectivity basis for those functional changes, especially the weakening of frontal “hub” connections (ventromedial PFC and insular inputs were reduced here) which resonates with fMRI findings of disintegrated high-level networks. Additionally, past c-Fos mapping in rodents (e.g. Inserra et al., 2021 or the authors’ own ACS Chem Neurosci 2023 paper) identified regions like RSP and prefrontal cortex as activated by psychedelics – Jiang et al. tie this to synaptic changes, showing activation likely foreshadows where new connections form or old ones retract.
In terms of serotonergic mechanisms, previous pharmacological studies established that 5-HT₂A receptors are critical for both the acute psychedelic experience and its therapeutic effects (e.g. blocking 5-HT₂A prevents the subjective effects and the long-term antidepressant-like effects in animal models). Jiang et al.’s findings are consistent with that: they propose that postsynaptic 5-HT₂A-driven excitability increase in PT neurons is one condition enabling the observed synaptic strengthening. Their data indirectly support this by showing PT neurons (which responded with calcium increases to psilocybin in prior studies) underwent input gain, whereas IT neurons (which do not show such dendritic excitation) did not gain inputs. This aligns with the idea that 5-HT₂A activation in frontal PT neurons is a key trigger for growth, dovetailing with prior cellular studies (e.g. Stimpson et al., 2020, found DOI (a 5-HT₂A agonist) required 5-HT₂A to induce plasticity-related gene expression). Moreover, the hint that 5-HT₂C receptor distribution correlated with some input changes adds a new layer; 5-HT₂C hasn’t been the focus of classic psychedelic action, but this result may inspire exploration of how 5-HT₂C (perhaps in visual or sensory areas) might modulate plasticity.
Compared to earlier studies of structural plasticity from psychedelics (e.g. the authors’ own 2021 Neuron paper showing spine growth, or a recent 2023 study on 5-MeO-DMT inducing plasticity), the present work significantly advances the field by moving from microscopic changes in isolated neurons to macroscopic reorganisation of networks. It doesn’t refute prior findings; rather, it provides a unifying framework: yes, psychedelics cause new synapses (as others showed), and here is where those new synapses are forming in the brain’s circuitry and how they alter information flow. This addresses a major gap in the literature and will likely spur new research – for instance, investigating if therapeutic outcomes in patients correlate with specific network connectivity changes (perhaps measurable by neuroimaging or EEG connectivity proxies). Additionally, this study’s methodology can be applied to other interventions: it sets a precedent for using viral circuit-tracing to map drug-induced plasticity, which could be used for other drugs (e.g. MDMA or even non-psychedelic treatments) to see if they also produce activity-dependent circuit reorganisation.
Overall Assessment of Accuracy and Impact: In sum, “Psilocybin triggers an activity-dependent rewiring of large-scale cortical networks” is a rigorously executed and highly illuminating study. The data strongly support the authors’ claims, and they avoid over-claiming beyond their evidence. The paper convincingly identifies a pattern of circuit changes caused by psilocybin and ties it to both neural activity and potential therapeutic mechanisms. Any minor weaknesses (like lacking direct proof of behavioural effect or 5-HT antagonist experiments) do not undermine the core findings; rather they mark areas for future investigation. The impact of this work is likely to be substantial: it provides a concrete biological explanation for how a brief psychedelic experience can lead to lasting changes in brain function, which has been a mystery in the field. By doing so, it validates the idea that treating mental illness may require altering entrenched neural networks, and that psychedelics are one promising tool to achieve that. This paper will be of great interest not only to neuroscientists studying psychedelics, but also to the broader fields of neuropsychopharmacology, neural plasticity, and therapy development. It stands as an excellent example of translationally relevant basic science – precise in its experimental detail and profound in its implications. Overall, the study’s accuracy in data and interpretation is commendable, and its impact is likely to shape both our scientific understanding of psychedelic drugs and the clinical strategies by which we might safely harness their brain-rewiring capacity.