We've updated our Privacy Policy to make it clearer how we use your personal data. We use cookies to provide you with a better experience. You can read our Cookie Policy here.


Advances in Protein Engineering

Nanobodies attached to a cell.
Credit: iStock
Listen with
Register for free to listen to this article
Thank you. Listen to this article using the player above.

Want to listen to this article for FREE?

Complete the form below to unlock access to ALL audio articles.

Read time: 5 minutes

The ability to engineer proteins is a fundamental part of the R&D process for several industries – from manufacturing enzymes to designing the next generation of therapeutics. The concept of engineering a protein to introduce new properties or optimize function is not novel, but a sophisticated new toolkit is rapidly emerging. In this article, we look at two of the hottest areas in protein engineering for therapeutic applications and how machine learning and artificial intelligence (AI) is set to transform the field.

Antibody engineering

Proteins account for a vast area of therapeutic space – from target enzymes such as kinases, phosphatases and proteases, through to engineered versions of naturally occurring human proteins used as therapeutic anticoagulants, hormones and growth factors. But perhaps the most exciting and fastest moving group of protein drugs are antibodies. The antibody therapeutic field has rapidly grown from the early days of monoclonal antibodies to a field that encompasses many different engineered derivatives. Today, antibodies are being engineered into a range of different novel drug classes, from antibody-drug conjugates, where the antibody delivers a drug payload to a target site, to multivalent antibodies that bind different antigens simultaneously.

Dr. Zhiquiang An is the director of the Texas Therapeutics Institute at the University of Texas Health Science Center at Houston and specializes in engineering antibodies for a range of therapeutic indications. During the COVID-19 pandemic, his lab quickly mobilized to produce a neutralizing IgM antibody nasal spray that provided potent and broad protection against SARS-CoV-2.1 This was achieved by engineering a parental IgG antibody against the virus, which had 2 binding sites, to an IgM version that had 10 binding sites, making it 230 times more potent. Now, as well as working on the next generation of antibodies for COVID-19 and future pandemics, his team are at the forefront of engineering antibodies for other indications such as neurodegenerative diseases and cancer.

“One of our priority areas is to engineer antibodies that can cross the blood-brain barrier, so that they can reach a CNS target such as a brain tumor or Alzheimer’s disease (AD),” says An. “Our other focus is to increase the valency of antibodies, so that we can achieve greater potency.”

In a recent study, An’s team combined both these approaches to design an antibody-based drug for AD.2 They engineered an antibody that targets TREM2 (triggering receptor on myeloid cells 2), a receptor that directs microglia to swallow up amyloid plaques. By engineering a bivalent IgG1 antibody to a tetra-variable domain antibody they improved the potency. Next, they engineered a bispecific antibody targeting both TREM2 and the transferrin receptor to improve brain entry.

“The transferrin receptor is responsible for shuttling ferric iron from the circulation to the brain, and it’s very efficient,” explained An. “When an antibody binds to the transferrin receptor, the receptor can flip the antibody across the blood brain barrier.” The transferrin transport concept is not new – but when An combined this technology with the team’s TREM2 antibody, they increased the antibody’s potency by 100-fold and its ability to reach its target in the brain by 10-fold. The bispecific antibody is now in preclinical development.

An’s lab is also interested in the design of nanobodies – antibodies that have a heavy chain and no light chain. “Because antibodies are large molecules with poor tissue penetration, we are looking at the potential to engineer nanobodies derived from camels and some other animal species that have a smaller binding site,” says An. “These are interesting for targeting complex membrane proteins – such as GPCRs and transporters – which are challenging to drug with larger molecules.”

Determining Real-Time Kinetics and Affinity

This compendium serves as a user-friendly reference for assay design on both bio-layer interferometry (BLI) and surface plasmon resonance (SPR) systems. Download this compendium to learn more about the advantages of real-time, label-free analyses, a variety of assay applications that meet GxP compliance standards and what to consider before starting experimental design.

View Compendium

Engineering post-translational modifications

Another hot topic in protein engineering is manipulating protein post-translational modifications. “Traditionally, protein engineering has focused on site-directed mutagenesis of a protein’s amino acid sequence,” says Professor Matthew DeLisa, William L. Lewis Professor of Engineering and director of the Cornell Institute of Biotechnology, “but nearly all proteins are post-translationally modified in some way or form, and so this has become an important part of protein design and engineering.”  

One of the most abundant modifications is glycosylation, the addition of a glycan group to the protein backbone. A glycan is a complex carbohydrate comprising one or more monosaccharides that are covalently assembled into unique structures. Glycans play important roles in the folding, stability, interaction and biological activity of proteins, making the ability to engineer glycan-mediated properties highly desirable.

There are two main approaches to glycan engineering: introducing a new glycan or changing an existing one. “In the first scenario, you’re actually changing the sequence of the protein by introducing new amino acid motifs that provide a signal to specific glycosylation enzymes to add a glycan group at those particular motifs,” says DeLisa. “In the second approach you’re remodeling a natural glycan structure by engineering the biosynthetic pathways that produces the glycan.”

Again, one of the major applications of these methods right now is in developing antibodies. This is because glycosylation of the Fc domain of antibodies is critical for their ability to engage with downstream effectors.

The first antibody to be produced with an engineered glycan was obinutuzumab (Gazyvaro), which is approved for the treatment of chronic lymphocytic leukaemia and follicular lymphoma. This antibody was produced by engineering the glycan biosynthesis pathway: the cells in which the antibody is produced are engineered to overexpress two glycosylation enzymes, MGAT3 and Golgi mannosidase. These overexpressed enzymes reduce the amount of fucose attached to the antibody, increasing its ability to activate natural killer cells.3

“By manipulating or changing the glycosylation that happens in the Fc domain you can have profound effects on antibody structure and function,” says DeLisa. “This allows you to create antibodies with customized properties such as being either highly pro- or anti-inflammatory.”

AI-driven protein engineering

Although the technology for producing protein-based therapeutics such as antibodies is improving, it still involves time-consuming experimental workflows. But AI could soon accelerate this process, says An: “AI-based programmes like AlphaFold are already being used to design more specific and potent antibodies. In the future, these will allow us to design antibodies with better efficacy based on the antigen structure.”

It’s an approach that is already changing the way protein engineering is done, according to DeLisa: “We can now easily obtain reasonably good predictions of three-dimensional protein structure all from a computer, without X-ray crystallography. Such predictions can be used to guide rational engineering campaigns that seek to improve a particular property of a protein, such as how tightly an antibody binds to a target antigen. This approach will be particularly valuable for certain subclasses of proteins like membrane proteins, for which very few crystal structures exist.”

Computational tools are also now starting to be used to create protein therapeutics from scratch. “This approach could be used to create designer enzymes or antibodies that don’t currently exist among the vast collection of nature’s enzymes,” says DeLisa, “It means you can come up with entirely new solutions binding a pathogenic target or achieving an enzymatic activity, without being constrained by how nature solved these problems.”

Once an antibody or protein is designed, it can be synthesized in the lab and tested for its desired properties. Even though not every design works, the tool gets better at design by learning from its successes and failures, through an iterative process.

“These tools are going to play a huge role in the next decade and beyond, in their ability to help us design better protein-based medicines, or entirely new ones from scratch,” says DeLisa. “But I don’t see this as limited to therapeutics, these tools are going to transform many fields – from potent enzyme catalysts that enable production of clean energy and biobased materials to sustainably sourced proteins that help meet the world’s food needs.”


1. Ku Z, Xie X, Hinton PR, et al. Nasal delivery of an IgM offers broad protection from SARS-CoV-2 variants. Nature. 2021;595(7869):718-723. doi:10.1038/s41586-021-03673-2

2. Zhao P, Xu Y, Jiang L, et al. A tetravalent TREM2 agonistic antibody reduced amyloid pathology in a mouse model of Alzheimer's disease. Sci Transl Med. 2022;14(661):eabq0095. doi:10.1126/scitranslmed.abq0095

3. Umaña P, Jean-Mairet J, Moudry R, Amstutz H, Bailey JE. Engineered glycoforms of an antineuroblastoma IgG1 with optimized antibody-dependent cellular cytotoxic activity. Nat Biotechnol. 1999;17(2):176-180. doi:10.1038/6179