The first amino acid, discovered in 1806, had a disarmingly ordinary source: asparagus juice. Hence the name asparagine. The next amino acids to be discovered had sources that were, if anything, even more humble. By 1820, cystine had been isolated from urinary stones, glycine from gelatin, and leucine from muscle and wool. All of the common amino acids were discovered by 1935, when threonine emerged from studies of animal nutrition involving growth stimulation assays.

Although threonine was the last of the 20 common amino acids, it was not the last of the proteinogenic amino acids. There were two more: selenocysteine, which was discovered in 1986, and pyrrolysine, which was discovered in 2002. These more recently discovered (and less commonly found) amino acids opened questions about the adaptability and utility of amino acids. Are there others that remain undiscovered? Could new amino acids be engineered? If so, what applications could they enable?

These questions prompted researchers to venture beyond their textbooks and initiate the study of noncanonical amino acids (ncAAs). Indeed, ncAAs have been inspiring academic work for over two decades. They have also been the focus of commercial activity, albeit only in recent years.

Much of this activity occurs at startup companies. One such company is Constructive Bio, which was launched to apply research initiated by Jason Chin, PhD, at the MRC Laboratory of Molecular Biology (MRC-LMB). Another such company is GRO Biosciences (GRObio), which is commercializing research initiated by George Church, PhD, at Harvard Medical School. The academics may have pioneered the biology underlying ncAAs and how they can influence protein engineering. Now, it is the companies’ work to turn that research into a new class of improved drugs.

Starting from scratch

When describing the work going on at Constructive Bio, Ola Wlodek, PhD, the company’s CEO, refers to noncanonical amino acids, or ncAAs. Other scientists working in this space may refer to nonstandard amino acids, or nsAAs. This is a “to-may-to/to-mah-to” issue, she explains. Everyone is talking about the same thing: amino acids outside of the natural code of 20 (or 22).

Ola Wlodek
Ola Wlodek, PhD
Contructive Bio

Constructive Bio, Wlodek continues, is interested in writing genomes from scratch to “open the chemical space of cells into many new possibilities.” Because the company relies on large-scale genome synthesis technology to write genomes from scratch, it can modify the existing alphabet of ncAAs. It can even create completely new blueprints.

According to Wlodek, Chin decided that Constructive Bio could write things better, “from de novo, from the get-go.” This is backed by decades of work and multiple papers from Chin’s academic group. His laboratory showed, in high-profile papers published in 2019 and 2021, that it is possible to write a genome from scratch with little redundancy in the genetic code.1,2 The lack of redundancy when it comes to genetic code (or, as they say, reimagining the genetic codes of the cells) allows the cells to produce designer biomolecules with novel functionalities.

Why do this? To build cellular biofactories, Wlodek answers. ncAAs allow the researchers at Constructive Bio to bring novel functionalities to peptides, proteins, and polymers. The goal is making biomolecules—designer proteins of the future—through cellular biofactories that will skip over many of the steps found in current bioprocessing technologies.

Wlodek points to the specific example of semaglutide, the active ingredient in Ozempic (approved for type 2 diabetes) and Wegovy (approved for weight loss). Semaglutide is a peptide that requires an ncAA at its N-terminus. To manufacture Ozempic, Novo Nordisk relies on solid-phase synthesis—a method that comes with certain supply chain challenges. The goal at Constructive Bio is to avoid such challenges. Specifically, the company aims to make ncAA-containing drugs (and endow them with designer properties) through bacterial synthesis, a common means of producing biologics.

How do you start to think about making proteins with new amino acids? With the end goal in mind, Wlodek answers. First, Constructive Bio identifies a problem with a biologic. For example, cytokine drugs are unstable. Indeed, they may be so vulnerable to the rigors of administration that their ability to carry out the functions of (structurally identical) natural cytokines is compromised. Second, the company develops a solution. For example, a redesign solution could enhance the stability of cytokine drugs so that they could be taken less frequently or in lower doses. (Redesigns typically involve chemical modifications that introduce novel and pharmacologically beneficial properties.) Third, the team at Constructive Bio links its solutions, on the ncAA level, with discoveries around tRNA synthase, the enzyme responsible for loading an ncAA onto a tRNA.

Constructive Bio focuses on using orthogonal elements—that is, elements that react with the novel ncAA and its codon but nothing else—to minimize cross-reactivity. Orthogonality can depend on choosing the right starting point. For example, in synthetase discovery, starting with pyrrolysyl-tRNA synthetase (PylRS) can be advantageous. Still, the enzymes must be evolved so that they start recognizing only the ncAA (and tRNA) of interest through their active sites. Then, the reaction is specific and has high fidelity.

Like other synthetic biology companies, Constructive Bio must face scaling challenges. It was launched last August, and it currently employs 15 people at a site in Cambridge, U.K. Soon, the company will move to a larger facility, also in Cambridge. Another move the company plans is one from small-scale to large-scale fermentation.

Constructive Bio maintains that its decision to use Escherichia coli may facilitate scaling. The company’s E. coli strain is not an average laboratory strain. It’s called Syn61 because it uses only 61 codons in its genome, not 64 like other bacteria. The three “excess” codons were replaced in every single coding sequence in the Syn61 genome, meaning that 18,000 codons across the genome were changed in one go. But even with its new genetic code, Syn61 behaves like standard E. coli.

Wlodek indicates that Constructive Bio is pragmatic. It looks at synthetic biology as a tool and not as a business model. It leaves fundamental research tasks, such as determining the degree to which the genetic code can be reduced, to the Chin laboratory. Indeed, the company is content to turn discoveries from the Chin laboratory into practical applications. In Wlodek’s words, the company is interested in the “fruit of the technology.”

The proteins that Constructive Bio is starting with are complex biologics with new functionalities. But Wlodek says that the company won’t stop at drugs. Instead, it will pursue the “completely programmable polymers of the future.”

Turning immunity around

Like Constructive Bio, Cambridge, MA-based GRObio develops ncAA-based technology that has academic roots. In GRObio’s case, the research started in the Church laboratory at Harvard Medical School.

When Dan Mandell, PhD, was a postdoctoral researcher in the laboratory, he fully recoded the UAG codon and repurposed it to incorporate an nsAA. One implication of this work was that other codons could be recoded. Another was that a one-codon reassignment could permit an nsAA to be installed, as Mandell says, “anywhere you want, as many times as you want.” Eventually, this work led to a genomically recoded organism (GRO).

The organism that GRObio is working on now, together with colleagues, will have seven codons reassigned. According to GRObio, it will be a “radically recoded organism.” The microbe cannot be made through genome editing, like the first one. Rather, it will be created through whole genome synthesis.

The goal is to get rid of the tRNA to truly repurpose selected codons so that no competition remains in the organism. Recoding seven codons will yield three tRNAs and one stop codon. In that case, four nsAAs can be different at the same time, in the same protein. But whether it’s one or seven, these bacteria are production organisms that make proteins with nsAAs. (The nsAAs are made somewhere else, and then fed to the organism like standard amino acids.)

GRObio’s ProGly nsAAs
GRObio’s ProGly nsAAs are designed to reprogram the immune system by inducing tolerance and blocking the induction of stimulatory T and B cells. The lynchpin in the process is the dendritic cell. Typically, an antigen induces an immunogenic state in a dendritic cell, setting off an immunogenic reaction. ProGly flips the switch. After recognizing a ProGly nsAA, the dendritic cell becomes tolerogenic and sets off a tolerogenic immune response. The technology, GRObio hopes, will lead to treatments for autoimmune disorders.

GRObio has two platform chemistries: DuraLogic and ProGly. DuraLogic creates proteins that are more stable. For example, it can extend the half-lives of proteins, enabling more convenient administration regimens—lower and less frequent doses—through the flattening of pharmacodynamic profiles. One can lose the peaks and troughs that come with having too much or too little of a drug.

ProGly (programmable glycosylation) uses preglycosylated nsAAs to direct how an immune system responds to a protein. Mandell refers to this as antigen-specific tolerization, which could be useful as a treatment for autoimmune disease.

Using ProGly, GRObio modifies a known antigen, creating a version that has tolerogenic and glycosylated amino acids. When the ProGly version is administered to patients, it can tolerize them. The ProGly version is the same protein that the person would be reacting to, but with a tolerogenic glycan signature.

A ProGly drug, GRObio asserts, can induce antigen-specific T regulatory cells to reeducate the immune system and, in doing so, alter the autoimmune disease progression. Key to this process are the dendritic cells, which the ProGly drug changes from stimulatory to tolerogenic. When the dendritic cells present the antigen in a tolerogenic context, naïve T cells become T regulatory cells rather than T helper or T effector cells.

GRObio’s first program is for the auto-immune disorder myasthenia gravis. A ProGly drug, the company hopes, will tolerize patients to the acetylcholine receptor, which is what 85% of myasthenia gravis patients react to. Mandell reports that GRObio has promising preclinical data using the myasthenia gravis rat model.

GRObio can also use this platform to tolerize a patient to a protein from an outside source, such as a protein therapy causing an antibody response, which will lead to a decrease in the drug’s activity over time. By making a ProGly version of the marketed enzyme replacement therapy, GRObio can create a version that will not induce antibodies in the patient. The enzyme will not be blocked or cleared, and the glycans will not interfere with its activity.

Because GRObio has developed a system that can tamp down immune responses, the company is pushing into the gene therapy space. The Achilles’ heel of gene therapy is the inability to control the body’s immune response to the delivery vehicle—most notably adeno-associated virus (AAV). A mechanism to control the immune response would allow therapies for people who have previously seroconverted to AAV (and have thus become ineligible for additional AAV-based treatments). It would also make gene therapy safer, given that many adverse events are due to an immune response to the AAV vector carrying the genetic payload.

When a similar strategy is used to treat autoimmune disease, it may be possible, GRObio asserts, to tolerize humans to key proteins in the AAV capsule by using a ProGly version of the empty virus-like particle. GRObio says that it aims to partner with gene therapy companies in order to tolerize patients to its capsid.

From Cambridge, U.K., to Cambridge, MA, and from creating better cytokines to aiding gene therapy, the development of new amino acids is likely to open up a world of possibilities in drug development and in many other applications. As Wlodek declares, now that the chemical alphabet of bacteria can be expanded, “The sky is the limit.”


1. Fredens J, Wang K, de la Torre D, et al. Total synthesis of Escherichia coli with a recoded genome. Nature. 2019; 569(7757): 514–518.
2. Robertson WE, Funke LFH, de la Torre D, et al. Sense codon reassignment enables viral resistance and encoded polymer synthesis. Science. 2021; 372(6546): 1057–1062.

Previous articleSetting the Stage for Digital Biomanufacturing
Next articleU.S. Sets Bioeconomy Goals and Confronts Data Challenges