Synthetic artificial “protein”

Ting Xu, a polymer scientist at UC Berkeley, developed a technique that mimics specific functions in natural proteins using only two, four or six different molecules that are currently used in plastics, she found These surrogate polymers are as effective as real proteins and are much easier to synthesize than trying to replicate nature’s design. She and her colleagues published their findings in the March 9 issue of the journal Nature.

She used a design approach based on machine learning, or artificial intelligence, to synthesize polymers that mimic blood plasma. The artificial biofluid preserved the integrity of the natural protein biomarkers without refrigeration and even made the natural proteins more resistant to high temperatures—an advantage over real blood plasma.

Protein substitutes, or random heteropolymers (RHP), could be a game-changer for biomedical applications, as much effort is spent today to engineer native proteins to acquire some new function, or to try to recreate the 3D of native proteins structure. Small molecule drug delivery that mimics native human proteins is an active area of research.

Instead, AI can select the correct number, type and arrangement of molecules to mimic the protein’s desired function, and can make it using simple polymer chemistry.

Using plasma as an example, artificial polymers are designed to solubilize and stabilize natural protein biomarkers in blood. Ting Xu and her team also created a mixture of synthetic polymers to replace the cell’s gut, the so-called cytosol. In a test tube filled with an artificial biological fluid, the cell’s nanomachine ribosomes continued to pump out natural proteins, as if they didn’t care whether the fluid was natural or artificial.

“Essentially, all the data shows that we can use this design framework to generate polymers in such a way that biological systems can’t tell whether it’s a polymer or a protein,” said Ting Xu, a professor of materials chemistry at UC Berkeley.

“If you really design it and inject plastics as part of an ecosystem, they should behave like proteins. If other proteins say, ‘Well, you’re part of us,’ that’s fine.”

This design framework also opens the door to the design of hybrid biological systems in which plastic polymers interact smoothly with native proteins to improve systems such as photosynthesis. The polymers can degrade naturally, making the system recyclable and sustainable.

“A Perfect Combination of Biopolymers and Non-Biopolymers”


Xu Ting views living tissue as a complex polymer of proteins that have evolved to flexibly work together, paying less attention to the actual amino acid sequence of each protein than to the protein’s functional subunits, the key to how these proteins interact. place. As in lock and key mechanisms, it doesn’t make much difference whether the key is aluminum or steel, so the actual composition of the functional subunits is less important than what they do.

Since these natural protein polymers have evolved randomly over millions of years, if you use the right principles to design and select them, it should be possible to create similar mixtures at random, with different building blocks, allowing scientists There is no need to recreate the exact protein mix in living tissue.

“Nature doesn’t do a lot of bottom-up, molecular formula, precision-driven design like we do in the lab,” Xu Ting said. “Nature needs flexibility to get there. Nature doesn’t say, let’s study the structure of this virus and make an antigen to attack it. It expresses a repertoire of antigens and then picks the one that works.”

Xu says this randomness could be used to design synthetic polymers that mix well with natural proteins, making it easier to create biocompatible plastics than today’s targeted techniques.

Working with UC Berkeley professor and applied statistician Haiyan Huang, the researchers developed a deep learning method to match the functions of natural proteins with those of plastic polymers to design a functional A man-made polymer that is similar but not identical to a natural protein.

For example, when trying to design a fluid that stabilizes a particular natural protein, the most important functions of the fluid are the charge of the polymer subunits and whether these subunits like to interact with water, that is, whether they are hydrophilic or hydrophobic. Synthetic polymers are designed to conform to these properties, but not to other properties of natural proteins in fluids.

Huang and graduate student Shuni Li trained the deep learning technique on a database of about 60,000 natural proteins, a hybrid of classical artificial intelligence (AI), which Huang calls a modified variational autoencoder (VAE). The proteins were broken down into 50 amino acid fragments, and the properties of the fragments were compared to man-made polymers composed of only four building blocks.

With experimental feedback from Zhiyuan Ruan, a graduate student in Xu’s lab, the team was able to chemically synthesize a random set of polymers, RHP, that mimic natural proteins in terms of charge and hydrophobicity.

“We look at the sequence space that nature has designed, analyze it, make the polymers match what nature has evolved, and they work,” Xu Ting said.

“How well you understand protein sequences determines the properties of the polymers you get. Extracting information from an established system, such as a naturally occurring protein, is the easiest shortcut that allows us to figure out how to make biocompatible polymers. correct standard of things.”

Colleagues in the lab of Carlos Bustamante, professor of cell biology and chemical physics at UC Berkeley, performed single-molecule optical tweezers studies that clearly show that RHP can mimic the behavior of proteins. Xu, Huang and their colleagues are now trying to mimic other protein features, replicating many other functions of natural amino acid polymers in plastics.

“Right now, our goal is just to stabilize the protein and mimic the most basic protein functions,” Huang said. “But as the RHP system is more elaborately designed, I think it’s natural for us to explore enhancing other functions. We’re trying to figure out what sequence components can inform the function or behavior of proteins that RHP might carry.”

The design platform opens the door to hybrid systems of natural and synthetic polymers, but also suggests easier ways to create biocompatible materials, from artificial tears or cartilage to coatings that can be used to deliver drugs.

“If you want to develop biomaterials that interact with the body, for tissue engineering or drug delivery, or if you want to do stent coating, you have to be compatible with biological systems,” Xu Ting said.

Her ultimate goal is to completely rethink how biomaterials are currently engineered.