Understanding the 15-PGDH study: methods and results

Adapted from Kim and colleagues, Redox Biology (2026), under CC BY 4.0. Published plots from Figures 2B, 4B and 4L are reproduced as cropped panels with explanatory labels alongside them. Other illustrations and narration are by Parkinsons Library; not endorsed by the authors.
AI-authored; source quotations checked; no expert scientific review

Who, where and why
Young Kwang Kim, Yun Jae Cha and colleagues published this study in Redox Biology in twenty twenty six. The animal work was carried out at Seoul National University in South Korea. They wanted to find out whether blocking an enzyme could protect the nerve cells damaged in Parkinson's. They tested this in mice and laboratory cells, and also examined human brain tissue donated after death.
Visual: Young-Kwang Kim, Yun Jae Cha and colleagues; Institution building, named researchers, and separate mouse, cell-dish and human-tissue symbols.. Procedure illustration; mouse icons represent groups, not sample counts.

What were they testing?
Parkinson's involves the loss of nerve cells that produce dopamine, a chemical involved in controlling movement. The researchers tested a drug called S W zero three three two nine one. It blocks fifteen P G D H. This is an enzyme: a protein that helps chemical reactions happen. The question was whether blocking it would reduce damage and preserve movement. In the following experiments, the drug was started before the experimental injury.
Visual: Dopamine-producing cells help control movement; Dopamine nerve cell and enzyme/signaling diagram, with the drug blocking the enzyme.. Procedure illustration; mouse icons represent groups, not sample counts.

Human tissue comparison
The team compared donated brain tissue from people with Parkinson's with tissue from comparison subjects. They examined the substantia nigra, the brain region containing the dopamine-producing cells affected by Parkinson's. There were five to eight samples per group. They measured messenger R N A: instructions cells use to make proteins. Instructions for making this enzyme were higher in the Parkinson's samples. This comparison measured a difference; it did not test a treatment.
Visual: Brain tissue donated after death; Two labeled human tissue sample dishes leading to an RNA-message comparison; reported 5–8 samples per group.. Procedure illustration; mouse icons represent groups, not sample counts.

The mouse comparison
For the first mouse experiment, the team used M P T P, a toxin that damages dopamine-producing nerve cells. There were three kinds of comparison: mice without toxin exposure; mice given the toxin and the drug's carrier liquid; and mice given the toxin plus the active drug. The carrier is the liquid used to deliver the drug. Comparing the last two shows what changes when the active drug is added. The treated mice received one of two doses.
Visual: No toxin exposure; Four labeled mouse-group lanes: baseline, toxin and carrier, toxin with lower drug dose, toxin with higher drug dose. Colored syringes show assigned exposures.. Procedure illustration; mouse icons represent groups, not sample counts.

First: prepare and pretreat
First, the mice had one week to adjust to their surroundings. Next came two days of drug treatment before any toxin was given. The drug was injected into the abdominal cavity twice a day. The two doses were zero point five and five milligrams per kilogram of body weight. That unit adjusts the amount of drug for the animal's weight. These are the doses used in the mouse experiment.
Visual: 1 week to adjust before treatment; A mouse acclimates, receives labeled abdominal drug injections, and a zero-based input chart compares the reported 0.5 and 5 mg/kg per-injection doses.. Procedure illustration; mouse icons represent groups, not sample counts.

Then: seven days of toxin exposure
The mice then received the toxin once a day for seven days. The toxin dose was thirty milligrams per kilogram. Drug treatment continued alongside it. Finally, two to three hours after the last toxin injection, the researchers tested movement. The order matters: the drug was already being given when injury began. This experiment tested protection during toxin exposure, rather than starting treatment after Parkinson's had developed.
Visual: MPTP: 30 mg/kg once daily for 7 days; Seven-day treatment calendar, labeled toxin and drug syringes, and the 2–3 hour wait before movement testing.. Procedure illustration; mouse icons represent groups, not sample counts.

Three movement tests
They used three movement tests, with ten to eleven mice per group. In the rotating-rod test, they measured how long a mouse stayed on a turning rod before falling. They scored hind-limb clasping: how much the back legs drew together. In the pole test, they measured how long the mouse took to turn downward and descend. Better performance meant longer on the rod, less clasping, and shorter times on the pole.
Visual: Longer time before falling = better performance; Animated rotating rod, hind-limb clasping schematic and descending pole test; direction of better performance labeled.. Procedure illustration; mouse icons represent groups, not sample counts.

Movement results
The toxin worsened all three movement measures. Mice receiving the drug had less impairment, with stronger protection at the higher dose. The researchers describe the result as, preserving motor performance to near non M P T P exposed levels. In plain language, the treated mice performed close to mice that had not received the toxin. That was a positive result for protection in this experiment.
Visual: Toxin exposure worsened movement tests; Published Figure 2B with intact measurement scale and data points, mouse-group color key and plain-language axis explanation.. Published result plots reproduced from Kim et al. (2026), CC BY 4.0.

What did the brain tissue show?
The team also examined brain tissue using a protein called tyrosine hydroxylase as a marker of dopamine-producing nerve cells. They used staining to locate and measure this marker. Toxin exposure reduced the staining signal. The drug preserved it, with greater protection at the higher dose. The toxin-model staining comparison used four mice per group. This tissue measurement supported the movement findings.
Visual: Tyrosine hydroxylase: a nerve-cell marker; Brain tissue slice, staining, microscope and the reported direction of nerve-cell-marker preservation; no fabricated signal values.. Procedure illustration; mouse icons represent groups, not sample counts.

Check the enzyme a second way
In a separate experiment, the researchers compared mice with one copy of the enzyme's gene removed with their littermates. A gene contains instructions for making a protein. Removing one copy reduced the enzyme genetically, without using the drug. Eight-week-old male and female mice received toxin daily for seven days, followed by movement tests two to three hours later. The genetically altered mice were better protected. Movement tests used fifteen to twenty two mice per group.
Visual: One Hpgd gene copy removed versus normal littermates; Two gene-copy symbols versus one copy, mice receiving toxin, seven-day calendar and movement-test outcome.. Procedure illustration; mouse icons represent groups, not sample counts.

Inflammation experiment
The team also used L P S, a bacterial substance that triggers inflammation. Mice received drug or carrier before L P S was injected into both sides of the substantia nigra. Drug treatment continued at five milligrams per kilogram, twice daily. One week later, the team used the same movement tests. The drug prevented the movement deficits seen with L P S. This tested a different way of injuring the brain from the toxin experiment.
Visual: SW033291 or carrier; drug dose 5 mg/kg twice daily; Drug pretreatment, localized LPS brain injection schematic, one-week wait and movement-test sequence.. Procedure illustration; mouse icons represent groups, not sample counts.

The protein experiment
A further experiment used alpha synuclein, a protein that can form abnormal clumps in Parkinson's. Seven-week-old mice had one week to acclimate, followed by two days of drug or carrier before surgery. The researchers used a viral delivery system to increase production of this protein in the brain, then injected prepared protein clumps. They assessed movement and collected brain tissue thirty three days after injection. The drug dose was five milligrams per kilogram, twice daily.
Visual: 1 week acclimation; 2 days drug or carrier before surgery; Pretreatment, viral delivery and protein clumps in the brain, followed by the reported day-33 tissue collection.. Procedure illustration; mouse icons represent groups, not sample counts.

Protein-model results
The drug protected movement and preserved the nerve-cell staining marker in this model too. These measurements used five to six mice per group. Markers of chemical damage were also lower. However, the measured buildup of an abnormal form of alpha synuclein did not significantly decrease. So the positive result was protection of movement and tissue markers. The study did not demonstrate that the drug cleared this protein buildup.
Visual: Movement protected; nerve-cell marker preserved; Published Figure 4B movement plot beside Figure 4L protein-signal plot; group key and explanation of ns.. Published result plots reproduced from Kim et al. (2026), CC BY 4.0.

Look for changes inside the tissue
To investigate why the drug helped, the team compared gene activity in brain tissue from drug-treated and carrier-treated mice. They used R N A sequencing, which reads the messages made from many genes at once. The results pointed toward reduced inflammation and reduced production of damaging reactive chemicals. Some individual gene results did not pass the stricter statistical check used when testing many genes together. The researchers therefore tested candidate changes in new, independent groups of animals.
Visual: Compare RNA messages in treated and comparison tissue; RNA messages from two tissue samples to sequencing, pathway candidates, and independent validation mice.. Procedure illustration; mouse icons represent groups, not sample counts.

Check damage and blood vessels
Follow-up tissue tests found lower markers of chemical damage with enzyme inhibition. The researchers also examined the blood-brain barrier: the blood-vessel boundary that controls passage into brain tissue. Using an electron microscope, they saw less structural damage in drug-treated mice exposed to the toxin. They also measured a blood antibody in the brain as a sign of leakage. The drug prevented the toxin-related increase in that leakage marker.
Visual: Lower markers of reactive-chemical damage; Brain blood-vessel boundary, leakage-marker schematic and microscope; direction-only outcome labels.. Procedure illustration; mouse icons represent groups, not sample counts.

Test a possible chemical pathway
The team then examined microglia, the brain's immune cells, grown in the laboratory. They added prostaglandin E two, a chemical signal normally broken down by the enzyme, before exposing the cells to L P S. This reduced inflammatory and reactive-chemical responses. One result was less lipocalin two, a protein linked to nerve-cell damage. Blocking E P four, a receptor that receives the chemical signal, interfered with that effect. These tests help explain a possible pathway.
Visual: Microglia: immune cells of the brain; Microglial cells, PGE2 and LPS added in order, and the EP4 receptor-blocking experiment.. Procedure illustration; mouse icons represent groups, not sample counts.

What do the results mean?
The main result was positive: blocking this enzyme protected movement and nerve-cell markers across three mouse injury models. Reducing the enzyme genetically supported the result in the toxin model. But the treatment began before the experimental injury. The human tissue comparison was not a treatment trial. This paper therefore supports further testing of the target; it does not establish that the drug treats Parkinson's in people. The protein buildup measurement also remained unresolved.
Visual: Protection in three drug-tested mouse models; Mouse, movement-test and nerve-cell symbols summarize protection; protein buildup and human-treatment questions separated.. Procedure illustration; mouse icons represent groups, not sample counts.

What comes next?
The authors call for further experiments to test the chemical pathway in living animals, including blocking specific parts of it. They also call for additional human tissue samples. These are the next steps proposed in this publication. The result to take from this study is specific: reducing the activity of fifteen P G D H protected mice against several experimental injuries. Further work is needed to find out how that protection might apply to Parkinson's treatment.
Visual: Additional experiments in living animals; Animal pathway experiments and additional human tissue samples shown as the authors’ proposed next studies.. Procedure illustration; mouse icons represent groups, not sample counts.