A study of young moon jellyfish has linked arm regeneration with changes in the animals’ resident bacteria, or microbiome, and with added arginine. In laboratory comparisons, penicillin-treated animals regenerated more often than untreated controls, while added arginine was also associated with regeneration activation. The findings point to a connection between nutrients, the microbiome and regrowth, but they do not establish the biochemical chain that links them.
The work used two-day-old Aurelia coerulea ephyrae, a young jellyfish stage. Researchers amputated three arms from each animal and assessed regeneration 14 days later by checking for a newly regenerated arm and measuring its length.
The test conditions
To combine results from independent experiments, the researchers used a risk ratio for the share of animals that regenerated and a response ratio for average regenerate-arm length. They then used a restricted maximum likelihood meta-analysis, a method that accounts for differences among experiments.
The microbiome comparison included control feeding, increased food at approximately eight times the control amount, and control food plus penicillin at 100 U/mL.
Two bacteria dominated the animals
The ephyrae carried a bacterial community dominated by Mariplasma and Marinirickettsia. Together, the two groups made up about 97% of the measured bacterial abundance and appeared at roughly a one-to-one ratio.
More food did not significantly change either bacterium on its own. But the reported Mariplasma-to-Marinirickettsia ratio rose by a mean of two-fold. The ratio also shifted in the penicillin comparison, where the reported p-value was below 0.001.
Penicillin left Mariplasma abundance unchanged but reduced Marinirickettsia abundance five-fold. In paired laboratory comparisons, 65% to 75% of penicillin-treated animals activated regeneration even when controls showed none. The pattern was additive with increased nutrients, and treated animals regenerated longer arms. The reported p-value for that regeneration comparison was below 0.0001.
Models offered clues, not measurements
The researchers then used genome-scale metabolic models, or computer models of metabolism, to predict what the bacteria might be doing. The Marinirickettsia model predicted no functioning glycolysis, pyruvate as the bacterium’s top uptake, and host-metabolite sequestration. It assigned 33% of total flux to the tricarboxylic acid cycle and 39% to electron transport.
Mariplasma showed a different model pattern. The ADI pathway accounted for 71% of its model flux. The bacterium was predicted to take up host arginine, send 98% of that uptake through ADI and use 2% for growth, while secreting ornithine and ammonium. These outputs are predictions rather than direct measurements of metabolite uptake or secretion.
The jellyfish’s genes also pointed to arginine
A separate RNA-sequencing experiment, which surveys gene activity, compared control food, food at three to four times the control amount, and control food plus penicillin. Samples were collected five days after amputation, and the methods reported five to seven animals per collection. The analysis identified 5,324 differentially expressed genes with increased food, 1,830 with penicillin and 1,416 shared by both conditions.
The shared genes were enriched, or overrepresented, in arginine metabolism and in the arginine biosynthetic or urea-cycle pathway. Carbamoyl phosphate synthase and argininosuccinate synthase were upregulated. This is a molecular association with the regeneration-promoting conditions, not direct evidence of metabolic flux.
Animals given 200 micromolar arginine showed 34% plus or minus 1% regeneration activation, while controls showed none. The pattern was additive with increased food, and the reported p-value was below 0.001. Arginine also markedly increased relative Mariplasma abundance and had much less effect on Marinirickettsia; the reported p-value for that comparison was below 0.0001.
A missing link remains untested
Comparative genome analysis reported that Aurelia lacks arginase and that arginase was absent from all surveyed scyphozoan jellyfish genomes but present in other cnidarian classes. The Mariplasma model suggested a possible biochemical complement, with its ADI pathway potentially helping fill that missing link through predicted ornithine production.
That explanation remains untested. The model outputs do not directly show that Mariplasma supplies ornithine or that Marinirickettsia’s predicted metabolite uptake changes the host’s nutrient pool. The authors could not test ammonium or ornithine effects because even slight increases of these metabolites in the water made the animals unhealthy.
The study’s boundaries
The bacterial flux predictions and gene-expression findings identify possible links, but they do not directly establish metabolite flow or the intracellular mechanism behind regeneration. The main regeneration outcome was measured 14 days after amputation by recording whether an arm appeared and how long it was, leaving the proposed biochemical steps unresolved.
The article front reports acceptance on July 27, 2026 and publication on August 21, 2026. RNA-sequencing reads were reported in NCBI BioProject PRJNA1248490, while the two metabolic models, raw images and analysis scripts were reported in CaltechDATA.
Paper data and sources
Original title: Microbiome modulation in regeneration in jellyfish.
Authors: Aki Ohdera, Matthew Wang, Maille Mansbridge et al.
Journal/Repository: PloS one
Status: Peer-reviewed
First online: 2026-08-21
DOI: 10.1371/journal.pone.0355863
Original paper · Full text