DHA showed the strongest pattern
In differentiated SH-SY5Y cells, DHA showed the strongest overall molecular response, including changes in PSD-95, a protein scaffold at the receiving side of a neural connection, and in several proteins involved in glutamatergic signaling. EPA showed the weakest or a distinct response, while mixed DHA:EPA compositions generally fell between the two. The DHA-dominant 2:1 mixture most closely resembled DHA alone.
What the experiment measured
The researchers asked whether defined DHA:EPA ratios were associated with differences in PSD-95 organization and related glutamatergic protein expression. After a 14-day differentiation period, they exposed the cells for 24 hours to DHA, EPA, or mixtures at a fixed total fatty-acid concentration of 50 μM. The mixtures used DHA:EPA ratios of 2:1, 1:1 and 1:2. Vehicle-treated cells received matched ethanol, and untreated cells were included as an additional control.
Each treatment condition used in Western blotting was evaluated in three independent biological replicates from separate cell passages differentiated and treated on different days. Immunofluorescence and Western blotting measured PSD-95 abundance, its puncta density and its overlap with synaptophysin, referred to as colocalization, along with NMDAR, AMPAR and CaMKII protein abundance. MTT assessed metabolic activity. Group differences were tested with one-way ANOVA followed by Tukey multiple-comparisons testing, after normality was assessed with the Shapiro-Wilk test.
On the imaging measures, DHA-treated cells showed the largest PSD-95 puncta-density response and the strongest total colocalization with synaptophysin. EPA showed little change in puncta density or total colocalization, and a lower proportion of PSD-95 colocalized with synaptophysin. The reported adjusted p-values were 0.0005 for puncta density, less than 0.0001 for total colocalization, 0.0214 for the DHA colocalized proportion and 0.0079 for the EPA colocalized-proportion comparison.
Looking at MAP2-positive neurites, the cell projections included in the analysis, the study found a more qualified pattern. All lipid-treated conditions had higher PSD-95 intensity in these neurites than control. But neurite-length-normalized puncta density was higher for DHA, the 2:1 DHA:EPA mixture and the 1:1 mixture, not for the 1:2 mixture or EPA. The reported adjusted p-values for the DHA, 2:1 and 1:1 comparisons were less than 0.0001, 0.0002 and 0.0076, respectively.
The pattern extended to proteins linked to glutamatergic signaling. DHA-treated cells showed the strongest NMDAR, AMPAR and CaMKII responses. EPA-treated cells had lower NMDAR than control and no significant AMPAR change, while all lipid-treated conditions had higher CaMKII than control. Reported p-values across these comparisons ranged from less than 0.0001 to 0.024, including less than 0.0001 for the reported DHA NMDAR and CaMKII comparisons, 0.0056 for the EPA NMDAR comparison and 0.0008 for the DHA AMPAR comparison.
The study reports direction and significance, but not complete effect sizes or confidence intervals for the comparisons. That shows which conditions differed in the reported tests, but not the size of every molecular change.
The signal had important limits
MTT-derived metabolic activity did not differ significantly across treatment groups after the 24-hour exposure. That finding is narrower than a test of synaptic function: the study did not directly measure receptor phosphorylation, synaptic transmission or activity-dependent plasticity. Its protein changes therefore remain molecular evidence, not direct evidence of enhanced synaptic function.
One design feature makes the ratios hard to interpret. Total fatty-acid concentration was held at 50 μM, so a higher EPA proportion also meant lower absolute DHA exposure. The design therefore cannot distinguish an effect specific to the DHA:EPA balance from the effect of giving cells less DHA.
Taken together, the measured endpoints showed a composition pattern: DHA had the strongest overall molecular response, while EPA and mixed compositions showed distinct, generally attenuated profiles without detectable reduction in MTT-derived metabolic activity. The result is evidence about molecular changes in this cell experiment, not direct evidence of enhanced synaptic function. The supplied metadata identifies the work as a peer-reviewed research article, and publication support was provided by the DLG4 SHINE Foundation.
Paper data and sources
Original title: DHA-dominant Omega-3 compositions enhance PSD-95 organization and glutamatergic protein expression
Authors: Medina G, Baird S, Rodriguez-Alcocer L, MacKenzie A
Journal/Repository: Frontiers in synaptic neuroscience
Status: Peer-reviewed
First online: 2026-10-07
DOI: Not available
Original paper · Full text