September 1, 2006 (Vol. 26, No. 15)
PTMs Help Address Protein Characterization and Validation
One of the main goals of proteomics is to unravel the many modifications important for biological activity. Post-translational modifications (PTMs) extend the range of protein function by attaching it to other biochemical functional groups like phosphates and carbohydrates. Many companies are developing new ways to address the hurdles of protein characterization and validation by focusing on PTMs. Several of these companies will be presenting information on how they address PTMs at the upcoming IBC conference, “Well Characterized Biologicals and Post-Translational Modifications.”
A hot topic at the moment is glycosylation because of concerns regarding immunogenicity and whether adding a glycol causes safety problems, according to Barry Rosenblatt, Ph.D., director of technical services at Charles River Laboratories (www.criver.com). Dr. Rosenblatt will be providing an overview of post-translational modifications and analytical technology at the upcoming conference. Glycosylation rarely has a direct effect on the efficacy of a protein but it does affect its half-life. “So, the main concern is more with clinical efficacy,” he explains.
“For the most part, PTMs are going to occur. So, the questions are can you determine what they are and can you control it?”
Some analytical techniques, like PAT, or process analytical technology, are designed to monitor glycosylation pattern changes and the conditions causing them in real time, explains Dr. Rosenblatt. A turn-key system, however, doesn’t currently exist, he adds. “The holy grail is to eliminate having to wait until the end of the experiment to test the protein.”
Companies engineering PTMs are still in the early stages, according to Dr. Rosenblatt. It is important to consider what the modifications are, what they mean, and how to correlate data from analysis with in vivo applications. “I think it comes down to who has a predictive system that works in real-time,” he concludes.
Engineering post-translation modifications is becoming increasingly important in drug development research. Recent advances have made enzymes more efficient in creating a desired biocatalytic conversion. Neose Technologies (www.neose.com) developed GlycoPEGylation™, which enables selective addition of PEG (polyethylene glycol) to sugar chains using the company’s enzymes.
GlycoPEGlyation can extend the half-life of a protein by linking various sized PEG polymers to glycans, remote from the protein’s active site. The company believes that adding PEG to the glycan, rather than the protein backbone, may have advantages over traditional chemical pegylation of amino acid residues.
The company’s GlycoAdvance® technology serves as the starting point for GlycoPEGylation. This uses enzymes to modify or initiate glycan structures on proteins. GlycoAdvance allows the company to work in a variety of expression systems, including CHO cells, E. coli, and insect cells.
Both GlycoAdvance and GlycoPEGylation can be performed in the context of a therapeutic protein manufacturing process at a cost similar to other PEGylation processes. Neose Technologies is currently developing a variety of GlycoPEGylated protein therapeutics, alone and with its partners. One is a protein expressed in insect cells, called GlycoPEG-EPO, a long-lasting version of erythropoietin. Another, GlycoPEG-GCSF, is prepared using GlycoAdvance to initiate glycosylation of the E. coli-expressed protein and employs GlycoPEGylation to extend the added glycan with a sugar-PEG chain.
Engineering Proteins for Better Performance
Ambrx (www.ambrx.com) will be presenting information about its ReCODE™ technology, which incorporates chemically specified side chains into biosynthetic proteins via a reconstituted system that includes tRNA synthetase and tRNA.
ReCODE (reconstituting chemically orthogonal directed engineering) has enabled incorporation of more than 30 new amino acids into biosynthetic proteins in E.coli, yeast, and mammalian cells. The tRNA incorporates a nonnative amino acid any place an amber codon is positioned, converting the amber from a stop codon to one that signals incorporation of the chemically specified amino acid. “The power of this is that one can add any side chain adduct explicitly to the protein in a given position,” states Thomas Daniel, M.D., CSO.
ReCODE’s first application has been to attach a polyethylene glycol side chain to the substituted amino acid residue. The position of the PEG affects two key functions of the optimized protein lead, Dr. Daniel explains. These include preservation of biological activity and improved longevity of action. The technology allows the lead optimization of proteins—similar to the optimization of small molecules conducted through medicinal chemistry—in what the company is calling protein medicinal chemistry.
“We’ve spent a lot of time optimizing amino acids for their chemical reactivity. Our focus has been on coupling efficiency and stability of the linkage.”
Dr. Daniel says the technology’s limitations are defined by the chemistry used to couple activities onto the protein. Although currently a nonnative amino acid can only be incorporated into one position with high efficiency, the company is working to expand this.
Ambrx has two products under development: a long-acting human growth hormone that will be entering the clinic by early next year and a long-acting interferon.
Researchers at Genzyme (www.genzyme.com) have developed a method to produce human and humanized antibodies with inhibitors for enzymes involved in the glycosylation pathway. These antibodies have specific oligosaccharides with improved ADCC activity. “We’re interested in this because there have been several excellent reports in the literature demonstrating the enhancement of ADCC activity of no- or low-fucose containing antibodies,” explains Qun Zhou, Ph.D., senior scientist. He adds that most of these carbohydrate-modification methods use genetic engineering techniques, such as knocking in or out certain glycosyltranferases in the production cell lines or expressing antibodies by using different systems.
“Our process is very simple. We have a method that can metabolically modulate the glycosylation and produce an antibody with specific types of oligosaccharides without fucose,” says Dr. Zhou. An inhibitor molecule is added to the medium of cells expressing the antibody in culture. The antibody or the cell line doesn’t need to be engineered, and since there is no protein engineering involved, the researchers can quickly evaluate the antibodies for the benefit of ADCC activity.
The group has been successful in using the inhibitor to produce different humanized Mabs in research with specific types of oligosaccharides without fucose. “The antibodies produced using our method have higher FC-receptor-gamma binding and higher ADCC activity,” reports Dr. Zhou.
This process is currently limited to in vitro models, as there are limited in vivo animal models to evaluate ADCC activity. Yet, the antibodies that Dr. Zhou’s group have studied using inhibitors are currently in research for various diseases.
Humanized Yeast Enhances GlycoProteins
Proteins that are post-translationally modified make up 70% of the market, with the most important modification being glycosylation, according to Tillman Gerngross, Ph.D., CSO at GlycoFi (www.glycofi.com). “All commercially approved drugs that are glycoproteins are heterogeneous products. What’s been appreciative in the past five years is that as those sugars change, the actual protein changes its properties in terms of half-life and how it binds to receptors. So there has been an increasing interest to control glycosylation. Some are trying to do this in mammalian or insect cells, we’ve been doing this in yeast.”
The company developed a protein-expression technology to optimize glycan structures of a therapeutic protein. This is based on a library of yeast strains engineered to perform specific human glycosylation. Expressing a protein in different glycosylated yeast strains enables the generation of a library of glycoproteins, all with the same peptide backbone but with different attached sugars. This allows the identification of a glycoform with the highest therapeutic potency.
Dr. Gerngross explains that his group will take a commercial antibody, like Rituxan®, and show that it has 12 different glycoforms in it. Then they will make the individual glycoforms and see how the pure glycoform behaves in a pure binding assay to a receptor and mediates an immune response.
Some glycoforms are 100 times more potent than others in eliciting a reaction, he adds. “When we talk about an optimization platform, we say that, because when we put on different glycoforms, we can modify function over almost two orders of magnitude (100-fold) that we know are therapeutically relevant.”
Once an antibody binds to an antigen, it recruits the immune system to the FC domain. This is the function optimized by GlycoFi’s approach for changing the sugars. “The power of our technology is that we can modulate this particular interaction in a very significant way.”
There are several benefits to using yeast. It saves time, about five months from gene to final, purified product; it works in the existing infrastructure of stainless steel tanks; and it secretes proteins well. “Yeast has been used for a long time to make therapeutic proteins but not glycoproteins. We’ve discovered how to do that,” says Dr. Gerngross.
Antibody Production Cell Lines
Pablo Umana, Ph.D., CSO of GlycArt Biotechnology (www.glycart.com), will be discussing the use of GlycoMab® for engineering antibody production cell lines to produce therapeutic antibody glycoforms with modified glycosylation patterns. This modification is associated with higher biological activity and, hopefully, higher efficacy in the clinic.
Using CHO cells, the researchers add genes coding for an antibody along with a gene that encodes for a sugar-processing enzyme. “This works by over-expressing the enzyme (beta-1, 4-N-acetylglucosaminyltransferase-III) that causes production of bisected nonfucosylated sugars,” says Dr. Umana. Sugars normally produced by CHO cells are nonbisected fucosylated sugars. “Lack of this core fucose residue increases binding affinity to antibody receptors present on the immune system cells that recognize the antibody and are important for mediating the biological activity of the antibody,” he explains. Since the technology is the same used to develop antibodies, it’s easily applied.
It’s only been in the past few years that companies have begun to manipulate post-translational modifications, especially glycosylation, to create new forms of therapeutic proteins. “In the case of antibodies, alternative approaches to reduce the content of glycosylated sugars are a really unique example of how to modify the sugars and increase specific biological activity to make it more potent. We were the first to use recombinant DNA technology to manipulate and control the glycosylation profile of antibodies produced in mammalian cells to increase biological activity,” Dr. Omana states. The company currently has several antibodies modified using this technique in preclinical stages
Biacore (www.biacore.com) says that its partners and clients are increasingly using the company’s biosensors, based on real-time, label-free protein interaction analysis technology, in a process analytical setting. The goal is to assess the comparability of biological products between batches, or detect even slight protein alterations, such as post-translational modifications, following scale-up.
Biacore’s systems use a target immobilized on a sensor surface to reveal the specificity and rates of molecular binding interactions. The technology can analyze reaction rate kinetics (and thereby define affinity) and specificity, as well as concentration, points out Fredrik Sundberg Ph.D., the company’s director of global pharma operations.
“Using our technology, the interaction of an analyte with its binding partner, such as antibody to antigen or recombinant protein to target, is determined in terms of the rates of association and dissociation, presented as a graphical ‘sensorgram,’ the curve of which provides a unique ‘fingerprint,’” he notes.
“Any change in rate kinetics due, for example, to an increase in the off-rate—i.e., faster dissociation of the binding complex—will result in a change in the sensorgram fingerprint, even if the overall affinity of the protein determined by other kinds of end-point assay, remains the same. As a result, even minor modifications in the protein due, for example, to post-translational modifications or aggregation, will be detected, in real time. Once flagged, the nature of any modification to the protein can be identified using other technologies, such as size-exclusion HPLC, LC/MS and amino acid sequencing.”
Biacore believes this type of analysis could play an important role in all stages of biomanufacturing processes, as well as in batch release testing or lot-to-lot consistency testing of therapeutic protein batches.
“By quantifying the on-rate and off-rate constants for a particular batch, and setting an acceptable variable, it becomes fast and straightforward to compare data between process batches,” continues Dr. Sundberg. “A number of our major pharmaceutical and vaccine clients are now regularly using our technology in this area, and it is an application we will be actively promoting for our systems.”