Compost Enzyme CCPUR1 Engineered to Break Down Polyurethane Shoe Foam and Nylon

Scientists hunting for microbes that eat plastic have long relied on genome databases, searching for sequences that resemble enzymes already known to break down synthetic polymers. That approach has a well-known limitation: it only finds more of what has already been found. A group of researchers in Denmark decided to look somewhere else, namely warm environments where plastic waste had been sitting for years.

Their search took them to a landfill in Kenya, a tropical zoo, the guts of wax moth larvae and a compost heap near Aarhus. The payoff was a set of 12 enzymes capable of attacking polyurethane and nylon, two families of plastic that are notoriously hard to recycle. The most promising candidate came from a compost bacterium named Chelatococcus composti, and it was given the name CCPUR1, with CC standing for the bacterium and PUR1 for polyurethane hydrolase.

Now the same team has pushed that enzyme further. Working with the University of Porto, researchers at Aarhus University and the Danish Technological Institute used molecular-dynamics simulation and targeted DNA edits to build a variant, designated N308F:V312F, that degrades about 1.4 percent of the polyurethane in real shoe-sole foam within three days, with no pre-treatment of the plastic.

The work was published on September 11, 2026 in the journal Chem Catalysis, following an earlier discovery paper in Angewandte Chemie. The engineered enzyme is far from a commercial recycling process, but it is an unusually concrete demonstration that a laboratory-designed biocatalyst can bite into a commercial thermoset plastic in its finished form.

Key Facts

EurekAlert! reported on September 11, 2026 that the bioprospecting effort collected millions of bacteria from four sources: a landfill in Kenya, Randers Regnskov Tropical Zoo, the guts of wax moth larvae, and a compost heap near Aarhus. Using fluorescent molecules that mimic polyurethane and nylon chemical bonds together with fluorescence-activated cell sorting, the team screened active cells at up to 10,000 per second. The result was 29 plastic-degrading bacteria and 12 enzymes, of which 10 act on polyurethane and 2 on nylon.

CCPUR1 has a wide binding cleft that can accommodate large polymers, along with unusually high thermostability, traits that may reflect its origin. Compost heaps contain tough plant polymers such as wax and lignin and can reach up to 80 degrees Celsius. In three days the wild-type enzyme broke down just under 1 percent of powdered polyurethane mattress foam. A second enzyme, PYNYL1, was the most effective candidate against nylon.

Phys.org reported on September 11, 2026 that the conventional route to finding plastic-degrading enzymes amounts to looking for a lost key under a streetlamp rather than where it was dropped. Postdoc Andreas Møllebjerg proposed searching warm environments where plastic had been present for a long time, and traveled to the landfill in Kenya to collect samples. Co-authors traveled 34 kilometers to Randers Regnskov Tropical Zoo, bought a bag of wax moth larvae to study their gut enzymes, and fetched a shovelful of compost from a recycling centre about 10 kilometers from Aarhus University.

The engineering study analyzed a 1.65 angstrom crystal structure of CCPUR1 and ran molecular dynamics simulations to identify persistent substrate-contact hotspots. The team then engineered 29 variants across seven residues, achieving a 5-fold increase in hydrolytic activity while preserving or improving thermostability. The best variant, N308F:V312F, showed a 20-fold improvement in catalytic efficiency (kcat/KM) toward a polyether-polyurethane analog, consistent with enhanced pi-pi stacking and deeper substrate burial seen in the simulations. Postdoc Rosie Graham tested the enzymes, and Malthe Kjær Bendtsen was a co-author.

Chem Catalysis reported on September 11, 2026 that N308F:V312F hydrolyzed commercial thermoset polyurethane shoe foam, releasing detectable monomers and causing surface erosion, outperforming the reference PURases ABPURase and UMG-SP2. Globally, companies produce roughly 22 million tons (20 million metric tonnes) of polyurethane every year, more than 5 percent of all plastics reaching the market, and most polyurethane waste is landfilled or incinerated. A special class called thermosets appears in shoe soles and heels.

Analysis

What this really means is that the bottleneck in enzymatic plastic recycling is shifting from discovery to engineering. Finding CCPUR1 was the hard part of the first study; making it work better on a real, cross-linked commercial material is the subject of the second. The jump from just under 1 percent degradation of mattress foam powder to about 1.4 percent on shoe-sole foam looks modest in absolute terms, but the substrates are not equivalent, and the shoe foam was processed without any pre-treatment at all.

Cell Press reported on September 11, 2026 that the authors frame CCPUR1 as a tunable scaffold and their structure-and-dynamics-guided modeling as a general strategy for enzyme engineering. That framing is deliberate. Polyurethane thermosets are cross-linked networks, which is precisely why mechanical and chemical recycling struggle with them, and existing biocatalysts have lacked the activity, stability and substrate compatibility such materials demand.

Co-first author Pedro Paiva of the University of Porto described the result as an early demonstration that polyurethane can be degraded with enzymes without processing it beforehand, adding that it is quite significant. Co-first author Rosie Graham of Aarhus University noted that the advantage of enzymes is that they work under mild conditions, at lower temperatures and pressures than most current recycling technologies. That is the economic argument: energy and pressure are where chemical recycling routes spend their money.

The counterargument is equally clear, and the team makes it themselves. Roughly 1.4 percent in three days is not a process. Scaling that to a waste stream of more than 20 million metric tonnes a year would require rates orders of magnitude higher, or reactor designs that keep the enzyme in contact with an enormous surface area of shredded polymer. The researchers say industrial-scale recycling remains distant and may require combining mechanical, chemical and enzymatic steps.

Why It Matters

Polyurethane is a special case among plastics because so much of it is thermoset, meaning it cannot simply be melted and remolded. Shoe soles, heels, mattresses and kitchen sponges are named in the research as candidate waste streams. Today most of that material is landfilled or burned, and burning it recovers energy but loses the chemical building blocks that could be reused.

The broader significance is the screening method. Searching warm, plastic-exposed environments and using fluorescent bond mimics to sort up to 10,000 cells per second produced 12 working enzymes in a single campaign. The team says this is the first time bioprospecting has been done at this scale and breadth for polyurethane and nylon. If that approach generalizes, it becomes a template for finding biocatalysts against other stubborn polymers.

There is also a stability story that matters for industry. CCPUR1 tolerates higher temperatures and solvents, which is unusual for an enzyme recovered from a compost heap and useful for any real recycling line where heat, pH swings and organic solvents are routine. A biocatalyst that denatures in an industrial stream is not a biocatalyst at all.

Next Up

The authors point to continued design advances combined with complementary process strategies before scalable biocatalytic valorization of polyurethane waste becomes realistic. That means more rounds of simulation-guided mutation, more testing on real post-consumer items rather than on analogs, and process engineering to concentrate and reuse the enzyme across repeated cycles.

The discovery paper in Angewandte Chemie and the engineering paper in Chem Catalysis, published online September 11, 2026 with DOI 10.1016/j.checat.2026.101844, will now be judged on whether the 5-fold activity gain and 20-fold catalytic-efficiency gain can be compounded further. Funding for the work included the Novo Nordisk Foundation, and the enzyme work was carried out through the EnZync Center at Aarhus University.