The Bench to Clinic Blog Series by Lonza brings the latest from our team of experts on progressing your therapeutic candidates from discovery to clinic.
Monoclonal antibodies (mAbs) have high specificity and low toxicity, which when combined with their versatility, makes them very attractive for disease treatment, whether it’s a cancer or an infection. Cancer cells are targeted by mAbs through specific recognition of tumour-associated antigens. The binding of a mAb with its cognate target results in recruitment of immune effector functions. This occurs through the Fc region binding to receptors on a range of immune cells and on components of the complement pathway. In an infectious disease, binding of the mAb to target can also trigger responses through cell and complement based pathways.
For many years use of recombinant mAbs was focussed on anti-cancer and anti-immune therapies. Recombinant mAbs against pathogens were first approved in 1998 (Synagis®), with another 3 approved 2016. The increasing frequency of infections caused by multidrug resistant bacteria or viruses and the recent viral disease epidemics feeds the growing need for new prophylatic or therapeutic approaches that include anti-infective mAbs. A hurdle to successful use is the need for a better understanding of how the Fc receptor, isotype, and other structural regions mediate protection.
Modes of effector function (recruitment of immune cells, complement dependent cytoxicity, and phagocytosis) for mAbs in these three disease areas involve binding of the Fc region to Fcγ receptors. Efficient interaction of a mAb with relevant FcγRs is central to effector function efficacy. Modulating effector function can improve safety by reducing off-target cytotoxicity, unwanted cytokine secretion, and (possibly) reduce the frequency of escape variant appearance.
There are three general approaches to modulating effector function:
Protein engineering of the Fc region can enhance effector function. For example, margetuximab was mutated to reduce FcγRIIb binding resulting in enhanced ADCC activity. Protein engineering can create new combinations of effector function. IgG1 is the most potent ADCC activator whilst IgG3 is most potent for CDC. Antibodies with chimeric CH regions from IgG1 and IgG3 show 25-60% increase in ADCC and CDC activities. In addition to enhancing activity, effector function can also be reduced. IgG2 and IgG4 have very limited ability to elicit ADCC activity, alongside lower CDC activity, compared to IgG1 and IgG3. Soliris® combines sequences from IgG2 and IgG4 to eliminate effector function. Mutations that enhance low pH binding to FcRn translate to increased half-life in humans.
With regard to manipulation of Fc glycans, evidence is strongest for the involvement of fucose with possibly a supportive role for other sugars. The absence of fucose leads to stronger binding between Fc glycans and FcγRIII receptors, as stearic hindrance by fucose inhibits interactions between the mAb and the FcγRIII receptor, which enhances ADCC activity. A number of approaches to reducing the fucose content have been developed ranging from: knockout of FUT8 (such as in Potelligent® CHOK1SV® and other CHO host cell lines); insertion of a bisecting GlcNAc that stops the oligosaccharide being a substrate for α-1,6-fucosyltransferase; redirecting fucose biosynthesis through use of a heterologous enzyme; use of small chemical inhibitors of glycosyltransferases e.g. 2-deoxy-2-fluoro-L-fucose. These approaches result in a varying reduction in fucose content, enabling tuning of ADCC function to help meet the needs of an indication.
Increasing the terminal galactose potentially enhances both CDC and ADCC, but the benefit may be antibody specific. Terminal galactose introduces conformational changes in the CH2 domain enhancing FcγR binding. Some highly sialylated mAbs have reduced ADCC effector function as bulky Neu5Ac groups may reduce flexibility in the mAb’s hinge region, thereby reducing affinity for RcγRIIIa.
Due to the potential immunogenicity of some glycan structures and their role in clinical efficacy, glycosylation is now considered a critical quality attribute that must be within an appropriate range to ensure the desired product quality, safety and efficacy is achieved reproducibly. There is an intricate relationship between mAb glycosylation and culture conditions used to express the protein. This complexity creates many opportunities to affect glycan structures – for good or bad! The importance of process in modulating effector function will be explored in a future blog article.
Antibodies with engineered effector function are not some future idyll; they have been here for a while – Soliris® was approved in 2007 and by 2018 at least 3 glycoengineered mAbs were approved. Historically, the primary use of mAbs was in anti-cancer and anti-immune therapies. In such therapies, there is a strong link between effector function and clinical efficacy, and a good understanding of how effector function can be manipulated is needed. With the appearance of multidrug resistant pathogens, and the need for rapid responses to the appearance of novel pathogens (e.g. the current coronavirus outbreak), there is the need for new prophylactic and therapeutic approaches - mAbs will play a role. If mAbs are to be successful in such circumstances, a better understanding of how effector function mediates protection is needed. However, there is a large body of knowledge available that will enhance the rapid exploitation of that new understanding.
Register to the series and to download Lonza’s new whitepaper: Modulating IgG Effector Function by FC Engineering and Glycoengineering.
In the early years, development of biotherapeutics using mammalian expression systems was primarily focused on monoclonal antibodies (mAbs). Although mAbs still dominate in the clinic, we are seeing a very clear shift in molecule type in the pipeline of many companies.
You’ll hear different terms to describe this new wave of molecules for example ‘Next Generation Biologics’ (NGBs), or ‘complex proteins’. These NGBs or complex proteins’ can include bispecifics, other novel format mAbs, and novel glycoproteins. They can encompass domains from various molecules along with engineered elements often resulting in proteins not seen in nature before. This is creating challenges to those trying to express them by established means.
Why are they causing challenges? Some immediate thoughts on this from an ‘upstream’ viewpoint are that they may require more than the standard two genes of a mAb to be expressed. When these proteins have not been seen in nature before, they are often difficult to express. And as there is such a wide array of different molecules being brought to clinic, solutions can no longer focus on the needs of one specific molecule type.
As part of the investigation for solutions to these challenges, we recognized we would need to address general needs arising from such engineered proteins rather than the needs of a specific molecule type. For this, we would need to develop a toolbox of solutions and a more tailored approach to cell line and process development for scientists to get the best out of their particular molecule. One expression tool Lonza developed was our GS Xceed® Multigene Vectors (MGVs).
The GS Xceed® MGVs provide an easy way to assemble a GS™ vector for the simultaneous expression of 3 or 4 genes in a streamlined, efficient, one-vessel assembly method. The genes expressed can be all product genes or a mix of product and helper genes and via GS™ MGVs, the genes can be inserted into the expression vector in any order. The use of our MGVs remove many of the disadvantages associated with available alternatives. These solutions include a “mix and match” approach, where portions of a product are made in different cell lines and later chemically combined; or co-transfection, where protein genes are split out over several vectors and all transfected into a host cell line at the same time. Both of these approaches have been used successfully by others, but there are downsides. For example, the time and cost associated with having to make several cell lines with the mix and match approach; or the enhanced screening required to find a cell line that has taken up all vectors and is adequately expressing the protein with the co-transfection route.
We believe the use of the MGVs removes these disadvantages as, for example, all the genes are on the one vector, you only need to create one cell line, and standard CLC workflows can be used.
As is typical with any new technology Lonza develops, we first wanted to verify that it behaved as expected when applied to some model proteins. For example, when trying to express a complex antibody consisting of a common light chain and two different heavy chains, we were able to confirm successful assembly of a single GS™ vector for the three genes. Then, by LC-ESI-MS, we confirmed that the predominant species expressed when this vector was transfected into Lonza’s CHOK1SV GS-KO® host cell line was the correctly assembled heterodimer. We observed similar successes for an MGV consisting of four genes as well.
To demonstrate the efficacy of a multigene vector system in constructing an expression vector for a relevant triple chain protein, a triple gene vector was assembled and transfected into CHOK1SV GS-KO® cells. A vector assembly success rate of 89% to 100% for this product using these vectors was achieved (left image). Following transfection, it was observed from LC-ESI-MS analysis of purified cell culture supernatant that the correctly assembled heterodimer was the predominant species with a low level of mispaired homodimer (right image).
Lonza’s MGVs is one of many we felt we needed to develop to be able to efficiently and optimally produce NGBs. Development of the MGV technology allows us to now apply this to client molecules with our Cell Line Construction services, as well as provide our GS™ technology users with access to it.