Start with the chain

PDRN means polydeoxyribonucleotide. An early analytical study characterised a particular placenta-derived preparation as a mixture of DNA fragments, measuring lengths of roughly 50–2,000 base pairs. That historical material is not a specification for modern cosmetics, but it demonstrates why the acronym cannot universally mean salmon-derived DNA. [1]

A polymer is a molecule built from many repeating units. Cutting a polymer into shorter pieces changes the distribution of chain lengths without turning every piece into the same small molecule. In double-stranded DNA, base pairs describe length; kilodaltons describe molecular mass. An average value and a maximum value tell different stories about a mixture.

What the cosmetic ingredient names describe

The Korea Cosmetic Association defines Sodium DNA as the sodium salt of DNA and lists a skin-conditioning function. The name identifies a chemical form, not a species or a measured fragment-size distribution. [2]

Hydrolyzed DNA is separately defined as DNA broken down using acid, enzymes or another hydrolysis method. This describes processing. It does not establish that two hydrolysed ingredients have identical sizes or composition. [3]

Salt form, biological source and processing are different pieces of identity. “Polynucleotide” describes a chain of nucleotide units; commercial PN and PDRN terminology should not be converted into a universal size cutoff without the material’s actual definition. The label is a useful starting point, but a named study preparation provides more detail.

Source and processing can change together

Researchers prepared DNA from Paeonia lactiflora and used ultrasound to shorten it. Their low-size fraction was predominantly below approximately 200 base pairs. This was an isolated DNA-derived preparation, not an ordinary peony extract containing the plant’s broader mixture of compounds. [4]

Extraction separates material from its original tissue or cells; purification reduces other components; fragmentation changes chain length. These steps answer different questions. A plant source does not make the ingredient equivalent to a whole botanical extract, just as a fish source does not describe every property of the final DNA preparation.

Another primary study extracted a microbial PDRN preparation from Lactobacillus rhamnosus. Experiments in skin-cell systems investigated A2A-receptor-related signalling and downstream responses involving FAK and AKT, proteins involved in cellular communication. The work offers a defined alternative source and mechanistic research, not a human topical equivalence trial against salmon-derived material. [5]

A receptor is a molecule that receives a signal. A pathway is a linked series of cellular events. Researchers can use inhibitors to ask whether blocking one part changes the response. That helps investigate a mechanism, while leaving separate questions about delivery and visible outcomes in people.

A controlled topical study, with a deliberately narrow conclusion

Ye and colleagues tested a salmon-derived PDRN preparation averaging about 1,200 base pairs. Thirty-one Chinese women aged 35–55 used a 0.1% water-in-silicone eye cream on one side and 0.1% retinol in the same base on the other, twice daily for 28 days. The randomised, double-blind comparison favoured PDRN on several wrinkle and ultrasound measures. [6]

In cultured fibroblasts, the same research investigated signalling involved in matrix production. Inhibiting selected pathways reduced the measured gene responses. Those cellular findings help formulate a hypothesis, rather than proving the clinical changes arose through that exact mechanism. [6]

Smaller exploratory studies add different pieces

The peony study tested a 1% cream in ten healthy Korean adults aged 20–59. Eye-area elasticity improved from baseline after four weeks, while the control cream showed little change. A separate detergent-challenge experiment examined barrier recovery on the arm. The small sample, incompletely described randomisation and blinding, and company-employed authors make these preliminary findings. [4]

A plasma-processing study changed the measured size and charge behaviour of PDRN, then examined cell responses and transport through an artificial membrane. A related cream study followed 21 Korean women for two weeks and reported lower roughness and greater gloss relative to baseline. Without a matching vehicle-control group, those changes cannot isolate the processed DNA contribution. [7]

An appearance outcome is not a gene-editing claim

DNA’s familiar role in heredity can make the cosmetic terminology sound more dramatic than the experiments. These studies measure physical skin properties, distribution signals and cellular responses; they do not demonstrate replacement or rewriting of a person’s genes. Nor does a topical result reproduce the exposure from an injected preparation.

The useful way to read this emerging evidence is to keep the preparation visible: its source, fragment profile, vehicle and route. That allows genuinely relevant topical findings to contribute without making all DNA-containing skincare seem interchangeable.

Sources to follow

1. Tonello and colleagues, 1996. Characterisation of a historical PDRN preparation; DOI 10.1016/0731-7085(96)01788-8.
2. Korea Cosmetic Association: Sodium DNA. Salt identity and cosmetic function.
3. Korea Cosmetic Association: Hydrolyzed DNA. Processing-based identity.
4. Bak and colleagues, 2025. Peony-derived preparation and small topical experiments; PMID 41516097.
5. Microbial-derived PDRN study, 2025. Source-specific cell research, not a human cosmetic trial.
6. Ye and colleagues, 2026. Defined material, mechanistic experiments and short topical comparison.
7. Plasma-engineered PDRN study, 2025. Processing, laboratory models and uncontrolled cosmetic study; PMID 41012473.