🧬 Aging Wiki

Regeneration of neuromuscular synapses after acute and chronic denervation by inhibiting the gerozyme 15-prostaglandin dehydrogenase

15 min read

Regeneration of neuromuscular synapses after acute and chronic denervation by inhibiting the gerozyme 15-prostaglandin dehydrogenase

TL;DR

Bakooshli et al. (2023) extend the 15-PGDH gerozyme story established by Palla et al. (2021) from muscle-fiber rejuvenation to neuromuscular junction (NMJ) regeneration. The paper shows that skeletal muscle 15-PGDH — an enzyme that degrades the pro-regenerative prostaglandin PGE2 — is strongly upregulated not just by aging but also acutely by sciatic nerve transection (denervation). Pharmacological inhibition of 15-PGDH (PGDHi, using the SW033291 compound class) in mice: (1) promotes regrowth of motor axons and formation of new NMJs after acute nerve crush; (2) rescues motor neuron viability, NMJ morphology, and functional muscle force in aged mice with chronic aging-related denervation. The pre-synaptic (neuronal) recovery synergizes with the post-synaptic (muscle fiber) recovery previously shown by Palla 2021, yielding a marked combined gain in aged-muscle strength. A translational signal: 15-PGDH protein aggregates co-localize with target fibers — the histopathologic hallmarks of human neurogenic myopathies — suggesting gerozyme dysregulation may be part of human disease etiology, not just a mouse aging phenomenon.

Aging context. NMJ degeneration is a primary driver of age-related sarcopenia, but no clinical intervention currently regenerates lost NMJs. This paper positions 15-PGDH inhibition as a candidate for both acute (trauma, surgical denervation) and chronic (aging, ALS, post-polio) NMJ restoration.

Background

Neuromuscular junctions degenerate with aging in both mice and humans: pre-synaptic motor axon terminals retract, acetylcholine receptor (AChR) clusters at the post-synaptic endplate become fragmented (“pretzel-like” to “simple” morphology), and the ratio of innervated to denervated fibers declines 1. In aged EDL muscles, 20.2 ± 2.9% of myofibers are denervated (assessed by absence of neurofilament and SV2 staining at postsynaptic AChRs), compared with 4.4 ± 1.5% in young EDL muscles (data reported within Bakooshli 2023, consistent with prior reports cited therein). This progressive denervation is considered a major proximate cause of the sarcopenic loss of force-generating capacity in aged skeletal-muscle, distinct from muscle-fiber atrophy itself 2. needs-human-replication — the quantitative contribution of NMJ loss vs. fiber atrophy to human sarcopenia is not cleanly separated in any clinical trial.

15-Prostaglandin dehydrogenase (15-PGDH; gene HPGD) degrades PGE2 and PGF2α, the prostaglandins that drive muscle satellite-cell proliferation, Schwann-cell survival, and motor-neuron trophic support. Its discovery as a “gerozyme” — an enzyme whose aging-associated upregulation impairs tissue regeneration — was established by Palla et al. 2021 in the context of muscle-fiber regeneration 3. The Bakooshli 2023 paper asks whether the same gerozyme logic extends to the neural side of the neuromuscular unit.

Methods

Acute denervation model (sciatic nerve crush)

Male C57BL/6 mice (2–4 months old, from Jackson Laboratory) underwent sciatic nerve crush (SNC) as a model of acute nerve injury with regeneration potential (crush, unlike transection, preserves the endoneurial tubes as a scaffold for axon regrowth). Spontaneous recovery without treatment occurs over ~50 days. PGDHi (SW033291; 5 mg/kg i.p. once daily, starting day of injury, for 14 days) or vehicle (10% ethanol, 5% Cremophor EL, 85% D5W) was administered following crush. Primary outcome: plantar flexion peak tetanic force (millinewton) measured by percutaneous electrical stimulation of the tibial nerve at 3, 7, and 14 days post-injury (dpi). Secondary outcomes: histological NMJ morphology by bungarotoxin (BTX) + neurofilament/SV2 co-staining in EDL whole-mount preparations; motor axon counts in sciatic nerve cross-sections by immunofluorescence for NF and ChAT (neural-network-quantified); innervation percentage of NMJs.

Chronic denervation model (sciatic nerve transection + aged mice)

Male C57BL/6 mice (2–4 months old) underwent unilateral sciatic nerve transection (SNT) — removing 5 mm of sciatic nerve of one leg (non-regenerating model) while the contralateral leg served as control — to characterize 15-PGDH upregulation dynamics (assessment at 14 days post-SNT). Separately, aged male C57BL/6 mice (24 to 26 months old, obtained from the NIA aged rodent colony) with chronic denervation from aging received SW033291 (5 mg/kg i.p. once daily for 1 month) or vehicle. Mice were randomized by body weight to treatment groups. Outcomes: NMJ morphology (BTX/NF/SV2 co-staining in EDL whole mounts), plantar flexion tetanic force (in vivo), motor neuron viability (cleaved caspase-3+ and ChAT+ cells in lumbar spinal cord sections), and CREB phosphorylation in ChAT+ motor neurons.

PGE2 quantification

LC-MS/MS mass spectrometry (Xevo TQ-XS mass spectrometer, Waters; Vincent Coates Foundation Mass Spectrometry Laboratory, Stanford) of muscle lysates to directly measure the PGE2 breakdown product PGEM (13,14-dihydro-15-keto-PGE2) as a proxy for 15-PGDH enzymatic activity in control vs. denervated conditions. 15-PGDH enzyme-specific activity was also measured directly from protein lysates of GA muscles. This distinguishes on-target PGE2 restoration from off-target effects.

Human tissue analysis

Immunofluorescence and histological analysis for 15-PGDH protein, autophagy marker LC3A, and mitochondrial marker PDHA in human skeletal muscle biopsy specimens (n=10 patients with neurogenic myopathies, encompassing axonal neuropathies, myositis, and motor neuron disease). Classic NADH staining was performed to identify target fibers (central bullseye of NADH-reduced staining, indicating mitochondrial absence). Serial sections allowed co-localization of 15-PGDH aggregates with NADH-defined target fibers. Healthy muscle biopsies served as negative controls.

Key findings

1. Denervation induces 15-PGDH upregulation in muscle

Sciatic nerve transection (SNT) in young mice produced a marked increase in 15-PGDH abundance and activity in the denervated TA and GA muscle at 14 days post-injury 4. Quantification: ~20-fold increase in Hpgd mRNA (from published RNA-seq time-course data, ref 20 in paper) and fourfold increase in protein abundance by Western blot in denervated vs. contralateral control GA muscles. Enzymatic activity increased proportionally. CODEX multiplex proteomics confirmed the protein localization to denervated myofibers (not Schwann cells or endothelial cells). The Hpgd mRNA increase was early and persistent: marked at 3 days post-SNT, reaching ~10-fold at day 14 and ~20-fold at day 90 vs. contralateral controls — distinguishing it from transient atrogene/myostatin expression. A 2.5-fold increase in Hpgd mRNA was also observed in the SMA mouse model (SMNA7), establishing that motor neuron disease drives the same gerozyme induction. This is mechanistically significant: any cause of muscle denervation — acute trauma, surgical nerve damage, progressive motor neuron disease — engages the PGE2-degrading arm that limits regeneration. needs-replication — single-lab finding; independent quantification of 15-PGDH induction magnitude across denervation models is needed.

2. PGDHi accelerates motor axon regeneration and NMJ formation after acute crush

Following sciatic nerve crush (SNC), SW033291-treated mice showed accelerated regrowth of motor axons and more rapid formation of mature NMJ morphology compared to vehicle controls 4. Key quantitative outcomes at 14 dpi:

  • Force recovery: 37.2 ± 4.9% increase in plantar flexion peak tetanic force in PGDHi-treated vs. vehicle-treated mice (measured by in vivo tibial nerve stimulation). Note: no significant difference between groups at 3 or 7 dpi — the divergence emerged in the regenerative phase (days 7–14).
  • Specific force: 32.2 ± 4.1% increase in PGDHi vs. vehicle (exceeding the 6.1 ± 1.7% increase in muscle mass alone, confirming enhanced innervation rather than hypertrophy as the driver).
  • Motor axon counts: Sciatic nerves of PGDHi mice contained 797 ± 69.3 ChAT+NF+ motor axons vs. 415.2 ± 52.5 in vehicle mice (1.9-fold more); uninjured nerves contained 1045.1 ± 55 motor axons. PGDHi specificity was confirmed: other axon types were unaffected.
  • NMJ innervation: 97.4 ± 2.1% of EDL NMJs innervated in PGDHi mice vs. 84.5 ± 2.2% in vehicle mice at 14 dpi.

Not yet replicated in humans. needs-human-replication

3. PGDHi restores motor neuron viability and NMJ morphology in aged mice

In aged mice (24–26 months old) with chronic denervation-type NMJ degradation, SW033291 (5 mg/kg i.p. daily for 1 month) reduced NMJ structural abnormalities and increased motor neuron viability compared to vehicle 4. Key quantitative outcomes:

  • Baseline NMJ pathology in aged vehicle mice: denervation 21.6 ± 2.3%; axonal swelling 19.6 ± 1.4%; postsynaptic AChR fragmentation 28.9 ± 2.8%. PGDHi reduced all three.
  • NMJ morphology: PGDHi restored AChR cluster morphology toward the complex “pretzel” pattern found in young EDL muscles, and reduced endolysosomal BTX+ AChR vesicles at NMJs.
  • Motor neuron viability: Aged mice had 11.6 ± 2.5% cleaved caspase-3+ ChAT+ apoptotic motor neurons in lumbar ventral horn. PGDHi reduced this to 4.2 ± 0.6% (compared to 0% in young mice).
  • CREB phosphorylation (mechanism): After SNC, PGDHi increased nuclear p-CREB in ChAT+ motor neurons to 37 ± 2% vs. 22 ± 3% in vehicle. Direct intramuscular injection of non-degradable dmPGE2 into aged GA muscles elicited p-CREB in 75 ± 4% of lumbar ChAT+ neurons vs. 30 ± 4% in PBS controls — confirming muscle-derived PGE2 acts retrogradely on motor neurons via EP4→cAMP→CREB signaling.

The mechanistic pathway identified is: PGDHi → elevated PGE2 → EP4 receptor (highly expressed in cholinergic neurons) → cAMP → CREB phosphorylation → axon regeneration and anti-apoptotic signaling in motor neurons. The relative contribution of Schwann cells has not been dissected. no-mechanism

4. Functional force recovery in aged muscle

Aged mice (24–26 months old) treated with PGDHi showed increased muscle mass and plantar flexion force compared to vehicle-treated aged controls (Fig. 6F and G in paper) 4. Specific force values are reported in figures; the paper’s discussion frames the PGDHi-mediated pre-synaptic (neuronal/axonal) recovery as synergizing with the post-synaptic (muscle-fiber) rescue demonstrated in Palla 2021 3. The abstract states directly: “These presynaptic changes synergized with previously reported muscle tissue remodeling to result in a marked increase in the strength of aged muscles.” This is the paper’s own framing, not an interpolation. needs-replication

5. 15-PGDH aggregates in human neurogenic myopathy target fibers

In human muscle biopsy specimens (n=10 patients with neurogenic myopathies encompassing axonal neuropathies, myositis, and motor neuron disease), 15-PGDH protein aggregates composed the central bullseye “target” region of target fibers — the NADH-staining-negative, mitochondria-depleted foci that are the histopathological hallmark of denervation and reinnervation in neurogenic disease 4. Key details:

  • 9 of 10 biopsies: 15-PGDH aggregates composed at least part of the target fiber bullseye. The one exception was a biopsy from a patient with drug toxicity-induced neuropathy in which NADH staining patterns were complex and targets could not be confidently distinguished.
  • LC3A co-localization: LC3A (autophagosome marker) co-localized with 15-PGDH in the target region; PDHA (mitochondrial marker) was excluded from these regions — consistent with the snRNA-seq data from denervated mouse muscle showing DN myonuclei upregulate Hpgd and autophagy genes simultaneously.
  • Target fiber diseases: associated with SMA type 3, ALS, nerve root compression (lumbar radiculopathy), neurogenic amyloidosis, and axonal neuropathies — a broader disease spectrum than post-polio syndrome alone.
  • Controls: 15-PGDH aggregates were absent in muscle biopsies from healthy patients.

This is the most translationally important finding: it establishes the same gerozyme mechanism in human neurogenic disease, and that PGE2 deficiency in target fibers may impair the natural reinnervation response.

Extrapolation table — human neurogenic myopathy findings:

DimensionStatusNotes
Pathway conserved in humans?yes15-PGDH / PGE2 / EP receptors are conserved
Phenotype conserved in humans?partialTarget-fiber 15-PGDH co-localization shown in human biopsies; interventional response not tested in humans
Replicated in humans?noHuman data is cross-sectional correlative IHC only; no interventional human data

Mechanism

The proposed mechanism is an extension of the Palla 2021 gerozyme model to the neural compartment:

  1. Denervation or aging → upregulates 15-PGDH in denervated myofibers (snRNA-seq confirms myofibers as the main source — specifically a new “DN myonuclei” population that reprograms to express Hpgd alongside autophagy and stress genes; tenocytes also express Hpgd irrespective of innervation; Schwann cells were not identified as a major source in this study).
  2. Elevated 15-PGDH → accelerated PGE2 → PGEM (13,14-dihydro-15-keto-PGE2) conversion → local PGE2 deficiency in the muscle-nerve interface (confirmed by LC-MS/MS measurement of PGEM and direct enzyme-specific activity assay in GA muscle lysates).
  3. PGE2 deficiency → reduced activation of EP4 receptor (highly expressed in spinal cord cholinergic neurons per single-cell transcriptomic atlas) → impaired cAMP→CREB signaling in motor neurons → reduced axon regeneration and increased apoptosis.
  4. PGDHi (SW033291, 5 mg/kg i.p.) → blocks 15-PGDH enzymatic activity → PGE2 restored → EP4→cAMP→CREB activation in motor neuron soma → axon sprouting and anti-apoptotic survival signaling. Post-synaptic effect: PGDHi also reduces MuRF1 (muscle RING-finger protein 1) levels and AChR endolysosomal degradation, stabilizing the post-synaptic apparatus.

The mechanistic data in this paper establishes motor neurons (CREB phosphorylation) as the primary neural target; the role of Schwann cells was not directly tested. The dual pre-synaptic (motor neuron axon regeneration/survival) and post-synaptic (AChR stability + muscle fiber health from Palla 2021) substrate means a single small-molecule may address both sides of the NMJ simultaneously.

Open question: The relative contributions of Schwann cells, motor neuron soma, and post-synaptic muscle to the observed recovery are not fully disentangled. PGDHi is delivered systemically (i.p.), so CNS-vs-PNS partitioning cannot be ruled out. Conditional Hpgd or EP4 knockout in specific cell types would clarify the mechanism hierarchy. no-mechanism

Translational signal: human neurogenic myopathies

The histological finding linking 15-PGDH aggregation to human target fibers is the paper’s strongest translational evidence. Target fibers (also termed “targetoid” fibers) are a well-established pathological diagnosis associated with denervation and signs of reinnervation, observed in a range of motor nerve injuries including axonal neuropathies, nerve root compression (lumbar radiculopathy), neurogenic amyloidosis, myositis, ALS, and SMA — not limited to post-polio syndrome (which is not mentioned by the paper as a specific diagnosis). Target fibers are characterized histochemically by the central bullseye of NADH-reduced staining (absence of mitochondria). If 15-PGDH accumulation impairs the natural reinnervation response in human disease — not merely in mice — then PGDHi would have a therapeutic rationale in:

  • Acute traumatic/surgical nerve injury (peripheral nerve repair context)
  • Aging-related NMJ decline contributing to sarcopenia
  • ALS / motor neuron diseases (adjunct to slow denervation progression)
  • Post-polio syndrome / PPMA

The IP estate for 15-PGDH inhibition is held jointly by Stanford and licensed to Myoforte Therapeutics (subsequently to Epirium Bio as of approximately 2025). SW033291 is a research compound; the clinical development candidate may differ.

Conflicts of interest

Bakooshli MA, Wang YX, and Blau HM are inventors on patent application 63/257,264 (“A method to restore neuromuscular junction morphology”) assigned to Stanford University. Blau HM is also a named inventor on patent applications 62/860,180 (“Targeting PGE2 degrading enzyme to ameliorate muscle wasting and augment strength”) and 62/883,025 (“Rejuvenation of aged tissues by inhibition of the PGE2 degrading enzyme”). The 15-PGDH IP estate was licensed to Myoforte Therapeutics at the time of publication. These conflicts should be weighted when evaluating the positive framing of the data.

Limitations and gaps

  • Mouse-primary evidence: All functional and mechanistic claims derive from rodent models (sciatic nerve transection and crush in C57BL/6 mice). Human data is limited to cross-sectional IHC of biopsy specimens — no interventional human data. needs-human-replication
  • n values not confirmed: Subject numbers per group require extraction from the PDF methods section (frontmatter field left null pending verification). needs-verification
  • Long-term safety not characterized: Systemic PGDHi elevates PGE2 in multiple tissues. PGE2 has pleiotropic effects including modulation of cancer surveillance (PGE2 is immunosuppressive and tumorigenic at high levels). Long-term safety profile in aged subjects is not characterized. long-term-unknown
  • SW033291 dose-response: The dose and schedule of SW033291 used in this study may not be translatable directly to human dosing. The compound class requires further pharmacokinetic optimization. dose-response-unclear
  • Cell-type mechanistic dissection: Whether PGE2 acts primarily on Schwann cells, motor axons, or post-synaptic muscle to restore NMJ is not determined. Conditional genetic approaches have not been published. no-mechanism
  • Conflict-of-interest risk: Strong patent incentives exist for positive framing; independent replication by groups without IP stakes is essential before clinical translation claims are made.
  • Single-lab: To date, the 15-PGDH gerozyme story (Palla 2021 + this paper) comes primarily from the Blau laboratory at Stanford. Independent replication is needed. needs-replication

Cross-references

  • sw033291 (implicit stub — compound page not yet seeded; see also: molecules/compounds/) — the small-molecule 15-PGDH inhibitor used in this study
  • 15-pgdh — the gerozyme target (gene HPGD; UniProt P15428 in humans)
  • sarcopenia — the phenotype this intervention addresses; NMJ loss is a proximate driver
  • skeletal-muscle — tissue page for muscle biology
  • stem-cell-exhaustion — satellite cell deficits in aged muscle; context for post-synaptic recovery
  • altered-intercellular-communication — Schwann cell / motor neuron / muscle crosstalk is the communication axis disrupted by denervation
  • palla-2021-15pgdh-muscle-rejuvenation (implicit stub — study page not yet seeded; doi:10.1126/science.abc8059) — the prior Blau lab paper establishing 15-PGDH as a gerozyme in muscle-fiber regeneration; this paper is a direct extension

Footnotes

Footnotes

  1. doi:10.1016/j.exger.2010.03.007 · Deschenes MR, Roby MA, Eason MK, Harris MB · Experimental Gerontology 45(5):389–93 (2010) · PMID: 20226849 · observational · model: aged rodents · NMJ remodeling precedes sarcopenia-related myofiber alterations — corrected DOI: seeder had cited 10.1152/japplphysiol.00773.2010 (invalid; returns HTTP 404). Confirmed correct DOI via PubMed search (PMID 20226849). ↩

  2. doi:10.3389/fnagi.2014.00208 · Gonzalez-Freire M, de Cabo R, Studenski SA, Ferrucci L · Frontiers in Aging Neuroscience 6:208 (2014) · PMID: 25157231 · PMCID: PMC4127816 · review · “The Neuromuscular Junction: Aging at the Crossroad between Nerves and Muscle” — corrected DOI and journal: seeder had cited 10.1093/gerona/glu072 which resolves to an unrelated paper on co-trimoxazole and sulfonylurea hypoglycemia (Tan et al., Journals of Gerontology 2014). Confirmed correct DOI via PubMed (PMID 25157231). ↩

  3. doi:10.1126/science.abc8059 · Palla AR et al. · Science 2021 (published online 2020) · n=TBD · in-vivo · model: aged C57BL/6 mice · Established 15-PGDH as a gerozyme that limits muscle satellite-cell function and muscle regeneration in aging; PGDHi rejuvenated aged muscle mass and strength; foundational paper that Bakooshli 2023 extends to the neural (NMJ) compartment · PDF not locally available (download failed in a local paper archive) ↩ ↩2

  4. bakooshli-2023-15pgdh-nmj-regeneration · multiple cohorts: young male C57BL/6 (2–4 mo, SNC/SNT experiments); aged male C57BL/6 (24–26 mo, NIA colony, aging experiments); human biopsy series n=10 patients with neurogenic myopathies (IHC) · in-vivo + human correlative IHC · model: male C57BL/6 mice (sciatic nerve crush 14 dpi, nerve transection 14 days, aged chronic denervation 1-month PGDHi) · journal: Science Translational Medicine 15(717):eadg1485 · doi:10.1126/scitranslmed.adg1485 · PMID: 37820010 · PMC10763629 · citation percentile: 100th (fwci 5.76 per a local paper archive) ↩ ↩2 ↩3 ↩4 ↩5

Graph View

Backlinks