chatGPT(5.5paid)
Summary
This paper studies how age changes the biological response to isoflurane anesthesia plus laparotomy surgery in mice. The authors used young adult, late middle-aged, and geriatric male C57BL/6 mice, then profiled olfactory bulb, hippocampus, lung, spleen, and circulating extracellular vesicles 24 hours after the procedure. The central claim is that perioperative stress produces a much larger and more maladaptive molecular response in older animals than in young adults.
The experimental design was: 2 hours of 2% isoflurane in oxygen plus a midline laparotomy, with sham animals kept in home cages. Tissues and blood were collected 24 hours later. RNA-seq was performed on olfactory bulb, hippocampus, lung, and spleen; extracellular vesicles were analysed by nano-flow cytometry and Olink proteomics. Group sizes were small: generally n=4–5 for RNA-seq and n=5–6 for EV proteomics, with no formal prospective power calculation.
The main finding is an age-dependent shift from adaptive plasticity to stress/inflammatory dominance. In young mice, the response is limited and often looks like synaptic or metabolic adaptation. In 17-month mice, the response is stronger: brain tissues show ER stress, proteostatic stress, altered metabolism, and suppression of synaptic/lipid/structural programmes. Lung and spleen show amplified immune and inflammatory signalling, with suppression of structural, regulatory, and metabolic homeostatic pathways.
In the olfactory bulb, 17-month ISO/OP mice showed clear transcriptomic separation from young ISO/OP mice. Upregulated pathways included synaptic plasticity and synaptic transport, while downregulated pathways included gliogenesis and extracellular matrix organisation. The authors interpret this as acute age-specific remodeling in a region relevant to postoperative olfactory dysfunction and possibly postoperative cognitive disorder.
In the hippocampus, the 17-month mice showed upregulation of ER stress/unfolded protein response genes such as Chac1, Hspa5, Manf, Pdia6, and Abca7, while genes involved in dendritic and synaptic architecture, including Shank3, Slitrk1, Nrxn2, and others, were reduced. This is important because the hippocampus is central to cognition and postoperative neurocognitive disorder models.
In the lung and spleen, the pattern was strongly inflammatory. The lung showed increased leukocyte migration, chemotaxis, lymphocyte-mediated immunity, wound healing, and inflammatory response pathways, while structural and repair programmes were suppressed. The spleen showed increased adaptive/humoral immune programmes, including immunoglobulin-related genes, but reduced lymphocyte differentiation and regulatory pathways.
The extracellular vesicle data are used as a systemic readout. EV abundance and cargo changed after ISO/OP in an age-dependent way. In 17-month mice, circulating EVs reflected the transition toward immune/stress signalling. In 27-month geriatric mice, ISO/OP produced strong EV protein cargo remodeling, including inflammatory proteins such as Cxcl1, Cxcl9, Il1a, Tnf, Ccl5, growth factor/developmental proteins such as Erbb4, Hgf, Pdgfb, Notch3, and stress/cell-death proteins such as Fas.
Novelty
The main novelty is not that anesthesia and surgery can provoke inflammation, but that the paper maps this response across multiple organs and circulating EVs at different ages. It tries to connect brain vulnerability, peripheral organ inflammation, and EV-mediated systemic signalling in one experimental framework.
The second novel point is the emphasis on late middle age rather than only old age. The authors argue that 17-month mice already show amplified peri-anesthetic biological reactivity, making late middle age a potentially under-recognised window for perioperative risk stratification and prevention.
A third novel feature is the inclusion of olfactory bulb transcriptomics. Since olfactory dysfunction may precede or accompany postoperative cognitive dysfunction, looking at the olfactory bulb alongside hippocampus is a useful extension beyond standard hippocampal-only models.
A fourth novelty is the EV angle: the paper suggests that circulating EV cargo may act as a biomarker-like integrator of perioperative stress across brain and peripheral organs. This is plausible and useful, although the study does not prove EVs are causal.
Critique
The paper is useful as a systems-level descriptive study, but it is much weaker as a mechanistic study. It shows strong associations between age, ISO/OP, transcriptomic changes, and EV cargo changes, but does not prove that the EVs cause the tissue changes, or that the transcriptomic changes cause postoperative dysfunction.
The biggest limitation is the absence of functional outcome data. The paper discusses postoperative neurocognitive disorder, olfactory dysfunction, lung vulnerability, and immune dysregulation, but the core data are molecular at 24 hours. It would be stronger if paired with behavioural cognition tests, olfactory testing, pulmonary function, infection susceptibility, wound healing, or survival/recovery outcomes.
The second limitation is that the ISO/OP model combines isoflurane, 100% oxygen, laparotomy, wound stress, analgesia, and recovery stress. This is clinically relevant as a combined perioperative insult, but it makes causality difficult. Separate arms for isoflurane alone, surgery alone, oxygen alone, and different anesthetics would help identify what drives the molecular phenotype.
The third limitation is the small sample size. The authors acknowledge n=4–5 for RNA-seq and n=5–6 for EV proteomics, with no formal prospective power calculation. That is not unusual for omics studies, but it increases the risk that some pathway-level findings are unstable, especially for subgroup and correlation analyses.
The fourth limitation is that only male C57BL/6 mice were used. This matters because both immune aging and EV biology are sex-sensitive. The results may not generalise well to female mice, other strains, or humans.
The fifth issue is that the paper relies heavily on GO enrichment. GO terms can be broad, overlapping, and sometimes biologically vague. Terms such as “synaptic plasticity,” “wound healing,” “stress response,” and “immune response” are useful pointers, but they do not substitute for cell-type-resolved analysis or direct validation of key pathways.
The sixth issue is tissue heterogeneity. Bulk RNA-seq in hippocampus, olfactory bulb, lung, and spleen cannot easily distinguish whether changes reflect altered gene expression within cells or altered cell composition, such as immune cell infiltration in lung or spleen. Single-cell or spatial transcriptomics would materially strengthen the conclusions.
The seventh issue is EV interpretation. Plasma EVs are a mixed population from many tissues. Changes in EV protein cargo may reflect altered release from immune cells, endothelium, platelets, injured tissues, or other sources. Without cell-of-origin tracing or EV depletion/transfer experiments, EVs should be treated as candidate biomarkers, not proven mediators.
Overall assessment
This is a strong hypothesis-generating paper. Its most important claim is that late middle age already creates a qualitatively different perioperative biology, with exaggerated stress, immune, and proteostatic responses across brain and peripheral tissues. That is an interesting and clinically relevant framing.
The evidence supports the descriptive conclusion that ISO/OP causes age-dependent systemic molecular remodeling. It does not yet prove that these signatures cause postoperative neurocognitive disorder or other complications. The next step should be functional validation, sex-balanced cohorts, cell-type resolution, separated anesthesia/surgery arms, and intervention studies targeting ER stress, inflammation, circadian disruption, or EV signalling.