Advances in Linker Technology: Improving the Safety and Efficacy of Antibody Drug Conjugates

With the approval of several antibody drug conjugates (ADCs) providing targeted delivery of cytotoxic APIs to specific sites of action, generally cancer cells at this point, their potential has been confirmed, and further investment in R&D has resulted.

ADCs consist of an antibody, a pharmaceutically active small molecule drug or toxin (i.e., the payload), and a linker to connect the two. The linker, typically a peptide derivative, joins the small-molecule, highly potent drug to the large-molecule antibody, which is selected or engineered to target the antigens on a specific type of cell. “ADCs can, through the linkage of cytotoxic drugs to antibodies, safely deliver cytotoxic drugs that cannot be administered systemically because they are only released at the site of action, and it is the linker that makes this approach possible,” states Jonathan Drachman, chief medical officer and executive vice-president of R&D with Seattle Genetics.

Why use a linker?
An ADC is effective only if it delivers the cytotoxic payload to the target cell. If the drug is released too soon, it can cause harm to normal cells. That means it must be stable until the ADC enters the target cell, but then allows release of the payload in a manner that maintains the reactivity of the cytotoxic therapeutic. “Antibodies have very long half-lives; they circulate in the blood stream for some time before they bind to and enter the target cell and then get into the lysosome, which generally is the target,” explains Clay Siegall, CEO of Seattle Genetics. The attachment of payloads directly to antibodies using a stable linker has been investigated, but the performance seems to vary depending on the cell surface target selected. Stable linkers are dependent on lysosomal trafficking to release the cytotoxic payloads, and a certain percentage of the antibody never reaches the lysosome of the target cell, limiting the efficacy of stable linker compounds,” notes Hanspeter Gerber, vice-president of Pfizer. The linker, therefore, is necessary to control the release of the cytotoxic within the target cells outside of the lysosome without being released in non-targeted tissues or during circulation, to ensure both increased safety and efficacy.

Matching chemistry and biology
Most importantly, according to Bob Hollingsworth, senior director of oncology research with MedImmune, the global biologics R&D arm of AstraZeneca, the linker makes it possible to match the chemistry of the payload with the biology of the antibody. Most linkers developed to date have been peptide-based. Initially, according to Gerber, the linker and antibody were bound together through hydrazone or disulfide bonds, which are sensitive to oxidative and reductive environments, respectively. However, there are reductive and oxidative environments in places other than inside of cells. Therefore, most linkers being developed today are designed to be selectively cleaved from the antibody by enzymes that are only present inside the lysosome where there is a low-pH environment, typically proteases or esterases, thus imparting intracellular biodegradability. “The challenge then, is choosing a linker chemistry that is stable chemically but sensitive to enzymatic cleavage, but only for these specific enzymes that are present in tumor cells,” says Thomas H. Pillow, ADC chemistry team leader with Genentech, a member of the Roche Group.

For the ADC technology used in ADCETRIS, Seattle Genetics has also found that it is necessary to include a “spacer” between the bond to the antibody and the peptide portion of the linker, according to Drachman. “This structural spacer is needed to provide room for the enzyme to recognize and bind to the cleavable portion of the linker,” he explains. The linker that the company developed for ADCETRIS, its ADC linker commercialized in collaboration with Takeda Pharmaceuticals and approved in more than 35 countries for the treatment of relapsed Hodgkin lymphoma and relapsed systemic anaplastic large cell lymphoma. Large cell lymphoma is comprised of valine and citrulline with structural spacers on each end. Once the linker is cleaved from the antibody, the spacer undergoes self-immolation, resulting in the release of the small molecule drug without attachment to any
linker components.

Linker chemistry selection
The chemistry that is used to attach the linker to the antibody and the sites within the antibody to which the linker are bound both affect the performance of an ADC. It is most common to connect the linker to the antibody via reaction with amino acids, with cysteine the most common, followed by lysine.

For ADCETRIS, Seattle Genetics uses maleimide chemistry to connect its linker to the natural cysteines in the antibody, resulting in ADCs that have approximately four drugs bound per antibody. “This approach provides a stable and highly reproducible product. Although there is some heterogeneity, it has not been an issue with ADCETRIS or any of the other ADCs we have in clinical development,” Drachman states. The linker in Kadcyla, Roche’s ADC approved for the treatment of HER2 (human epidermal growth factor receptor 2)-positive advanced breast cancer in the US and EU, is connected to the antibody through an amide bond formed with lysine. “We considered several different linkers, including one based on a disulfide bond, but found that a noncleavable linker gave improved pharmacokinetics and safety,” Pillow observes. He also notes that the same linker may not always work with different payload/antibody combinations, and therefore screening of the payload, linker, and antibody combinations is necessary to determine the ADC with the highest efficacy and safety.

The location of the site of linkage to the antibody is also important. “It is necessary to consider the location of the amino acids that are serving as the sites for connection of the linker,” notes Timothy Lowinger, chief scientific officer for Mersana Therapeutics. “In non-engineered antibodies, cysteines for conjugation are typically in the hinge region, whereas lysines can be located throughout the antibody sequence including near the site on the antibody for binding to the antigen on the target cell, and thus attachment of linkers and drugs at these locations can potentially impact the binding ability,” he explains.

Moving toward site specificity
While heterogeneity is not necessarily an issue for ADCs, there is growing interest in developing linker technologies that are site-specific. “Site-specific approaches avoid random conjugation of the ADC payload, which can result in multiple positional isomers with potentially different bioactivity, pharmacokinetics, distribution, immunogenicity, safety, and stability profiles,” states John Burt, CEO of PolyTherics. “Site-specific conjugation yields more consistent ADCs and also leads to increased safety and efficacy,” Hollingsworth adds.

One solution, according to Pillow, is to engineer antibodies with incorporated non-natural amino acids placed in specific locations, and then use click chemistry, particularly cycloaddition reactions, to form stable linkages. Seattle Genetics has taken a similar approach, carefully placing cysteines into an engineered antibody called an EC-mAb. “This site-specific technology was developed in response to the need to attach just two payloads to an antibody for a specific cytotoxic agent,” Drachman comments. “We now have two options for attaching our linkers with antibodies, and we continue to explore others as well,” he adds. Genentech, meanwhile, is investigating the formation of haloacetamides and the selective reaction of other electrophiles to connect its linkers with the cysteines in amino acids.

PolyTherics’ ThioBridge linker technology achieves site-specific conjugation via a two-step process in which the disulfide bond between two cysteines in the antibody is reduced and then bis-alkylation enables the introduction of the three-carbon bridge of the ThioBridge linker, already carrying the cytotoxic payload, via covalent attachment between the cysteine thiols. The reconstruction of the disulfide bond differentiates ThioBridge technology from others, according to Burt. “With our technology, the tertiary structure of the protein is not disrupted, and as a result, the structural integrity is retained, unlike with maleimide chemistry, for example, where the disulfide bond remains broken. We recently have been able to show that, as a consequence, ADCs prepared using maleimide chemistry are not as stable as those obtained using ThioBridge linkers,” he explains.

Polymeric approach
With direct ADCs, one linker is used to attach the drug to the antibody. That linker must be sufficiently stable not to release the drug prematurely, but also sufficiently labile to release the drug within the cancer cell. With its Fleximer polymer approach, Mersana has chosen to use two linkers; one between the drug and the polymer, which controls drug stability and release, and a second between the polymer and the antibody, ensuring stability in circulation. Its Fleximer polymer is a polyacetal that has been shown to be clinically safe and biocompatible, according to Lowinger. It is extremely water soluble and very stable at the neutral and basic pH levels seen in the bloodstream, but is slowly degraded via hydrolysis of glycal and glycolate groups under the acidic conditions found in the endosome/lysosome of the target cell. Release of the drug from the polymer occurs quickly inside the cell via enzymatic processes.

“One of the advantages of the Fleximer technology is the ability to create ADCs with payloads that are either too hydrophobic or have insufficient potency due to the fact that only 3-4 drug molecules can be bound with typical linkers. Due to the very high water solubility of Fleximer, it is possible to load up to 30 drug molecules, including hydrophobic therapeutics, while maintaining excellent aqueous solubility and pharmacokinetics,” Lowinger says.

The Fleximer polymer can be conjugated to antibodies through lysines or cysteines, according to Lowinger. In the case of cysteine conjugation, electrophilic groups on the polymer react with thiol groups produced via reduction of disulfide bonds, and then both bonds are reformed, resulting in increased stability. Mersana is also working on site-specific approaches using non-natural amino acids and other engineered residues.

The Fleximer linker can also be used with antibody fragments, which are easier to manufacture and may achieve better and deeper cell penetration than whole antibodies. “Because of the smaller protein size in antibody fragments, however, they have poorer pharmacokinetics. By careful selection of the size of the Fleximer polymer, one can achieve pharmacokinetics similar to that observed for whole antibodies,” Lowinger explains.

Scale-up challenges
Conjugation of the linker to the antibody at large scale without affecting the structure of the antibody can be challenging. Proteins are very sensitive to changes in the environment, and changes in the structure of the antibody can affect the performance of the ADC in terms of both efficacy and safety, according to Drachman. Gerber adds, “Because the linker serves as the interface between chemistry and biology, there is always the opportunity for unexpected reactivity during scale up of the conjugation process. Reactivity can be affected by many factors, just as with the scale-up of any process.” As a result, it is necessary to develop a lab-scale conjugation process that can be scaled up to provide the desired product consistently from batch-to-batch.

Both Gerber and Siegall note, however, that fairly standard methods are used in the production of ADCs, and it is possible to develop robust, scalable processes. “Even though manufacturing ADCETRIS is a complex process, we have been able to establish the large-scale production of this ADC with consistently high quality. In addition, we can now apply the experience gained in the scale-up of ADCETRIS to the production of other ADCs in our pipeline,” Siegall asserts.

Future expectations
Given the potential for ADCs to provide the targeted delivery of highly potent drugs, it is not surprising that many pharmaceutical companies are working with biopharmaceutical companies focused on linker and conjugation technologies to build a pipeline of ADC candidates.

Pfizer is collaborating with CytomX Therapeutics to develop and commercialize ADCs based on CytomX’s masked probodies that remain inert in healthy tissue but are activated specificallyin the tumor microenvironment by dysregulated proteases that are associated with many disease states, and therefore can target very disease-specific tissues. “We believe that there are opportunities to identify new linkers with different mechanisms of action to target different types of cells,” Gerber comments. In October 2013, MedImmune acquired Spirogen, which has developed pyrrolobenzodiazepine (PBD) dimer technology for the attachment of highly potent cytotoxic agents to specific cancer-targeting antibodies using biodegradable linkers, according to Hollingsworth. “We are very excited about the PBD technology, as well as our armed antibody portfolio in general, including moxetumomab, our first-in-class drug in development for hairy cell leukemia, which is in phase III clinical trials,” he adds. Roche, meanwhile, has eight ADCs under development and is investing $208.9 million to construct a new ADC production facility in Basel, Switzerland, according to Pillow.

ADC technology companies are also developing their own portfolios of drug candidates. Seattle Genetics has a total of 20 ADCs (including the eight Roche ADCs mentioned previously) in various stages of development, either on its own or through collaborations with 12 different companies, not including additional applications for ADCETRIS. “With the success of ADCETRIS, we have stepped up our investment and have committed to putting two or three ADCs into clinical testing each year,” Siegall says. The investment is paying off, with three new candidates developed in 2013, and more expected in 2014 and 2105. Mersana is developing ADCs in partnership with Endo Pharmaceuticals and collaborating with a number of other undisclosed pharma and biotech firms in addition to developing its own pipeline of ADCs using its technology and through its collaboration with Adimab. PolyTherics has also established several collaborations, including payload collaborations with Spirogen for the site-specific conjugation of PBDs and Tube Pharma for the generation of ThioBridge cytolysin reagents, as well as with Macrogenics to produce cytotoxic conjugates with its Dual Affinity Re-Targeting proteins.

“At the end of the day we are all working to design systems that will enable the development of innovative therapies that meet the unmet needs of patients with improved safety and efficacy,” Siegall concludes.

About the Author
Cynthia A. Challener is a contributing editor to BioPharm International.