September 15, 2018 (Vol. 38, No. 16)

Researchers Adapt Existing Techniques to Deal with Challenging New Monoclonal Antibodies

Monoclonal antibody (mAb)-based therapeutics are emerging as one of the fastest-growing categories of drugs being developed today. In fact, according to a recent analysis,1 the market for mAbs has doubled in the last five years. Much of this growth is due to the emergence of new mAb-derived products, such as antibody-drug conjugates (ADCs) and bispecific antibodies (bsAbs).

These new biotherapeutics have a wide range of applications, including applications in cancer treatments and in antiviral treatments for acute and chronic diseases. Essentially, mAbs are multiple copies of the same lab-produced molecule. They work as substitutes for antibodies within the body, binding to antigens on the surface of a cell. By doing this, they can flag cancer cells for the immune system or eliminate infected cells displaying viral antigens on their surface. bsAbs, which constitute a class of newer and more complex mAb-related biotherapeutics, have two mAbs within the same drug to target two types of antigen.

Getting Competitive

As with most therapeutics, mAbs are developed through extensive selection of lead candidates, and the subsequent characterization of their size, shape, stability, activity, potency, developability, behavior in formulation, and other factors.

“For therapeutic antibodies, our goal is always to make sure the antibody is ‘fit for purpose,’” says Liusong Yin, Ph.D., director of antibody services, GenScript. “We need to characterize the mAbs and mitigate the risk at every step because, if you accumulate risk, you will face bigger and bigger challenges, and it will cost you more and more.”

According to Dr. Yin, the development of mAb therapeutics is becoming increasingly competitive as a fast-growing collection of drug candidates chases a relatively slow-growing collection of drug targets. This dynamic is hardly new, notes Dr. Yin, who is wont to cite a 12-year-old paper2 indicating that drug targets are surprisingly scarce. This paper reported that only 5.3 new drug targets are identified each year and that “all current drugs with a known mode of action act through [just] 324 distinct molecular drug targets.”

Yes, these figures are from 2006, but the dynamic persists. “Each year,” Dr. Yin points out, “there are numerous clinical trials with several hundred therapeutic antibody campaigns going on, and most of them are against existing targets.”

A similar view is expressed by Jennifer F. Nemeth, Ph.D., director of biophysics, structural characterization, Janssen Research & Development. “The antibody space has become highly competitive,” she says. “You can’t just be another company hitting a target of choice.”

Part of the rise in competitive pressure, she observes, is due to regulatory trends, such as those occurring in the European Union: “The European Medicines Agency may not necessarily accept a molecule that’s only as good as another product that is already approved. If you’re going into an already populated space, your product will most likely need to show improved efficacy and/or safety.”

Bigger Panels, Faster Robots

According to Dr. Nemeth, companies need to do characterization earlier and use bigger antibody panels to find that “needle in the haystack”—the result that will lead to a product with the desired characteristics. She explored this idea during her keynote talk at the 10th Annual Bioprocessing Summit, in Boston. There, Dr. Nemeth described how assays are being adapted for high-throughput applications.

“If you look back 10 years ago, many assays were lower throughput, accommodating just 2–10 antibodies,” she says, revisiting one of her keynote’s main points. “Now we’re pushing multiple panels of 10s to 100s of antibodies in an assay.” These higher-throughput assays, she adds, have been pushed earlier into the drug discovery process.

She gives the example of a high-throughput mass profiling workflow Janssen R&D is developing with Genedata. The companies are using Genedata’s Biologics Refiner® software platform for the batch processing of “intact” mass spectrometry data that accumulate during screens that incorporate sizable antibody panels (10–100 different antibodies). The output consists of the masses identified in each chromatographic peak derived from deconvoluted mass spectra.

“In discovery, you’re looking at lots of different antibodies once or twice, unlike the development space that focuses on multiple lots of the same antibody,” she remarks. The co-developed workflow will batch process multiple antibodies with multiple sequences and masses. “We commissioned this work through Genedata, and to our knowledge, ours is the only lab using the Genedata platform for this application at the moment,” she asserts. “However, once the code is complete and the beta testing is done, the application will be available through the vendor in a future release.”

Other pharmaceutical companies are also looking at automating their mAb characterization processes for higher throughput. For example, at AbbVie, senior scientist Bo Yan, Ph.D., is adamant that mAb characterization needs to integrate high-throughput operations, automation, and a “deep understanding” of critical quality attributes (CQAs).

Dr. Yan notes that when he was invited to speak at a recent BioProcess International event, he decided to discuss how biotherapeutics can be rapidly characterized even if only a very limited amount of material is available. According to Dr. Yan, the key is using microfluidic capillary electrophoresis (CE) mass spectrometry. “The enhanced sensitivity of this new method enables deep understanding of low-abundant species of mAb,” he insists. “The fast separation will definitely help increase the throughput of analysis.”

This schematic from Janssen Research & Development shows assay flows for all stages of biopharmaceutical development. The company notes that in early discovery, fast assessments of limited material are necessary. In later development, studies are routine, validated, and robust, and they present minimal material constraints.

Growing Complexity

The emergence of mAb-related biotherapeutics, such as bsAbs or even multispecifics, has increased the complexity of the drug-characterization process. “Bispecifics are becoming a bigger part of the entities pushed into Phase I and II trials, specifically in the oncology space,” notes Dr. Nemeth. “And the degree of complexity in the data, for some bispecific class forms, means there’s a need for more rigorous analysis early on.”

According to Dr. Nemeth, characterizing bsAbs takes longer than running assays for a single mAb. “If you combine binding specificities for two or more targets into one therapeutic, you have to interrogate each drug component separately at the screening stage; then make it into a multispecific, and then test again,” she details. “You end up doing two to three times the amount of work, analyzing and repeating analyses, and that means more resources to get one therapeutic to the clinic than was the case 10–15 years ago.”

For ADCs, for example, it’s necessary to characterize all three parts of the therapeutic—the mAb, the linker, and the payload. “You have to characterize the half-life of the mAb, the stability of your linker, how many copies of the payload there are, how stable they are, and—if there’s a toxic payload [such as a cytotoxic small molecule in a chemotherapy drug]—the way those toxins are released,” explains Dr. Yin. He adds that it is necessary to determine if the payload targets only the desired cell type, or if it targets normal, healthy cells, too.

In the next few years, the complexity of characterization will become even greater, Dr. Yin predicts. “We are characterizing different modalities,” he states. “But even for the simplest mAb, people are using mAbs for different therapeutic applications.” Different applications require various kinds of characterization, he explains. The focus of mAbs targeted against late-stage cancer, for example, is on improving survival rates, while those selected to treat chronic disease need to have lower toxicity.

More Relevant Data, More Quickly

In general, the growing complexity of characterization means researchers are looking to do more characterization earlier in the drug discovery process. At GenScript, for example, Dr. Yin says that scientists might validate an antibody with more than one animal model, or use several different biochemical, biophysical, in vitro assays, and in vivo assays. Also, if the scientists determine that the antibody doesn’t perform, they might stop the development process and look for alternatives.

For MaxCyte, a provider of cell-engineering solutions, a key trend is producing antibodies in manufacturing host cell lines at earlier stages of discovery and development. “There are so many choices for expression systems,” observes Joan Foster, a senior field applications scientist at MaxCyte. “It is important to work with cells and production media that will generate data that are indicative of a molecule’s manufacturability.” (Like Dr. Nemeth, Foster spoke at the 10th Annual Bioprocessing Summit.)

According to Foster, MaxCyte technology allows customers to create a stable cell line or transient transfections within a few weeks to carry out a proof-of-concept study early in development. “By making transient materials early, you can evaluate the efficacy of your antibody and potentially stop your project before spending millions of dollars,” she asserts. “Or you can start the processes to get to market quicker.”

Adapting Existing Methods

Other researchers are improving mAb characterization by adapting existing techniques. “If you want to build up and purify a promising therapeutic, you want the best information analytically as early as possible,” says Jeff Beckman, Ph.D., senior scientist, biologics development, Bristol-Myers Squibb. (At the 10th Annual Bioprocessing Summit, Dr. Beckman gave a talk on improving a popular method for the purity testing of therapeutic proteins such as mAbs.)

Purity testing often relies on CE procedures that incorporate a protein denaturing step, and this step typically uses the detergent sodium dodecyl sulfate (SDS). After this step, proteins and impurities can be separated by molecular weight. “SDS became popular back in the 1960s in the lab of Harry Eagle, M.D., and his colleagues at the Albert Einstein College of Medicine,” notes Dr. Beckman. “They needed to separate proteins on a gel, and as part of this they needed a good denaturant. SDS came up in the literature as a candidate, and it turned out that it worked beautifully for most proteins.”

As a result, Dr. Beckman explains, the process was optimized around the detergent that Dr. Eagle’s colleagues originally picked—and it endured. Unfortunately, some proteins appear to be resistant to being denatured by SDS, making it difficult to isolate impurities.

Dr. Beckman decided to experiment with different detergents, and he eventually found that sodium hexadecyl sulfate (SHS) performs better than SDS for some proteins—including some mAbs. “So far, what we’ve found is that some of our thermophilic proteins, really rigid proteins, tend to have resistance to SDS,” Dr. Beckman points out. SHS, he continues, is relatively hydrophobic and hence is better able to overcome this resistance.

For rigid proteins, CE-SHS gives a clearer separation by molecular weight as the mixture of protein and detergent moves down a capillary filled with a hydrophilic gel buffer solution. A graph of migration time down the capillary versus absorbance by an ultraviolet detector located on the capillary has a stronger, sharper peak, making it easier to measure the purity of the protein. “It’s an improvement on an existing method used commonly,” he maintains. According to Dr. Beckman, CE-SHS will become more important as companies devote more attention to designing mAbs that have higher thermophilicity. Such mAbs, he notes, are more stable and have longer expiration dates.

At Bristol-Myers Squibb, researchers showed that detergents other than sodium dodecyl sulfate (SDS) can improve separations based on capillary electrophoresis (CE). When the researchers subjected an Fc-fusion therapeutic to capillary sieving electrophoresis, they used buffer containing SDS only or buffer containing sodium hexadecyl sulfate (SHS). When they compared CE-SHS and CE-SDS traces, they found that the CE-SHS trace had a more highly resolved protein peak. Similar CE-SHS improvements were observed for some mAbs. [Reprinted with permission from Anal. Chem. 2018; 90(4): 2542–2547. (Image supplied by Jeff Beckman, Ph.D.)]

Characterization of Critical Reagents for Ligand Binding Assays

Ligand binding assays (LBAs) are instrumental in the drug development process to measure the immunogenicity of biotherapeutic molecules, as well as to determine drug concentrations for pharmacokinetic analyses. The robustness, accuracy, and reproducibility of these assays depends on the quality of critical reagents. Hence, careful characterization is essential for ensuring optimal assay performance is maintained over time.

Best practices for identifying the physicochemical attributes of critical reagents are outlined in white paper that appeared in the AAPS Journal.1 Bio-Rad incorporates several of these in the product development and manufacturing of its anti-biotherapeutic antibody portfolio, which comprises fully human, monoclonal recombinant antibodies. As such, every new antibody is well characterized, and subsequent lots are subject to strict quality control to assess batch-to batch variability.

Each new antibody is produced in three independent batches, and the activity is compared in an LBA. The batch with activity closest to the average curve is then chosen as the reference batch. Before each new production, the antibody gene is re-sequenced to ensure product identity, and prior to release, new batches are tested for specificity and purity, including size-exclusion chromatography for assessing monodispersity of full-length antibodies.

Knowing the affinity of an antibody can give an indication of the sensitivity of an assay. Bio-Rad determines the affinity of all antibiotherapeutic antibodies in monovalent Fab format, which truly reflects the intrinsic antigen-binding affinity.

Is It Possible to Monitor Multiple Biologic CQAs within a Single Analysis?

Kyle D’Silva, Ph.D.

The multi-attribute method (MAM) for structural confirmation of protein therapeutics involves a targeted search of peptide-mapping data using advanced, GMP compliant–ready software, for predetermined components that are indicative of numerous critical quality attributes (CQAs).

Protein sample preparation for MAM demands reproducible, fast, and low artifact–inducing proteolytic digestion into peptides; automatable methods are favored. Peptide separations must be rapid, consistent, and also free of artifacts induced through detection techniques involving intense, broad spectrum UV light.

MAM enables highly accurate relative quantification (% difference) of post-translational and process-induced modifications by comparison to a reference sample. High-resolution accurate mass (HRAM) mass spectrometry (MS) delivers both the required sensitivity and specificity for increased confidence in the detection and quantification of post-translational modifications. Moreover, HRAM-MS also brings the capability to detect additional components in parallel. This critical additional data-processing capability, called new peak detection,
automatically detects and flags new chromatographic components in a sample once compared to a reference.

This capability not only allows MAM to deliver both quantification of known differences, it also allows the flagging of new, unknown impurities in the sample that are present above preset detection limits. Low-resolution MS techniques have lower specificity and are thus unable to determine some critical modifications, especially where there are no chromatographic differences in modified and unmodified peptides. The robustness of modern HRAM-MS benchtop systems, combined
with simple and familiar compliance-ready chromatography data system software, makes MAM both highly practical and desirable for GMP lab use.


1. Grilo AL, Mantalaris A. The increasingly human and profitable monoclonal antibody market. Trends Biotechnol. 2018 June 23.
2. Overington J, Al-Lazikani B, Hopkins AL. How many drug targets are there? Nat Rev Drug Discov. 2007; 5: 993–6. 10.1038/nrd2199.
Previous articleAltered Levels of Neonatal Inflammatory Markers Associated with Childhood Leukemia
Next articleMicrobiome Response Can Serve as Early Warning to Nerve Gas Exposure