Currently, there are more than 300,000 commercially available antibodies that researchers can chose from to detect different proteins, protein forms, molecular components, and structures within cells. But in order to use these antibodies to visualize the cellular world a cell must be fixed and permeabilized to first immobilize the cellular components and then break up the membrane to allow the antibody to access the antigen. Although effective, an issue with this approach is that a researcher will always be looking at dead cells, and biological processes such as protein folding and protein interactions cannot be studied in dead cells in real-time.
Conventional antibodies are too large and chemically unstable to be effective inside living cells. “The chemistry is really too complex,” says Ulrich Rothbauer, research group leader in the department of biology at Ludwig Maximilians University, who is working with colleagues to develop tools to study cellular processes in living cells. “These antibodies have to assemble four different chains, two heavy and two light, and they’re assembled by disulfide bonds that cannot be correctly formed in the reducing environment of the cytoplasm. You cannot express such a huge complex molecule in living cells. You can [introduce] them by microinjection, for example, but it’s not applicable for high-throughput cell imaging.” Antigen-binding sites in these antibodies are created by two structural domains, one on a light chain and one on a heavy chain.
Although some genetically engineered recombinant antibody fragments are smaller and more stable, their binding sites still rely on the combination of a heavy-chain domain and a light-chain domain, which impairs their binding properties in living cells.
An alternative to these traditional antibody fragments has been found in a set of unique antibodies produced by the immune systems of camels, llamas, and sharks. Unlike other antibodies, these affinity reagents are composed of only two heavy chains; better yet, a single domain forms the antigen-binding sites for these heavy-chain antibodies. The domains can even be genetically engineered to produce extremely small, very stable single-domain recombinant antibody fragments, called “nanobodies.”
Small, stable, and easy to engineer
Conventional antibodies possess an atomic mass of about 150 kDa, while smaller recombinant antibodies have a mass range of about 25–50 kDa. Camel, llama, and shark nanobodies, on the other hand, weigh in at 12–13 kDa, which is half the mass of smaller recombinant antibodies and less than a tenth of conventional antibodies. At these small sizes, nanobodies can easily enter cells without the need for fixation and permeabilization.
“They are suitable for many applications where normal antibodies will never be applicable, including therapeutics,” says Sergey Sikora, vice president of business development at GenWay Biotech, a San Diego, CA–based firm offering customized recombinant shark antibodies.
Single domain nanobodies are also simpler and more stable than conventional antibodies. “You can boil shark antibodies, put them in extreme temperatures; they are also active in urea. So, altogether, that puts them in a different category,” says Sikora. These antibodies are so stable that, unlike conventional antibodies, they do not need to be refrigerated to be preserved.
The source of this thermostability is a mystery, however. As a postdoctoral fellow in the lab of Adam Godzik at the Burnham Institute, Sikora was involved in a thermostability study to understand why Thermotoga maritime proteins are more thermostable than their human protein counterparts, even though they are structurally similar. In the end, his team could not find any structural features that explained the difference in thermostability. “After all this research, after all these grants, nothing came out of it,” he says. The same question surrounds nanobody thermostability. “There’s something in the sequence of these antibodies, the same as in the sequences in Thermotoga maritime—which is still undetermined—that makes them thermostable. But that is still a big black hole.”
Because nanobodies are recombinant, genetic engineering tools used with recombinant proteins can be applied to these fragments as well, enabling modifications to affinity as well as expression in Escherichia coli. “You can produce tons of them,” says Sikora. “For example, some antibodies have produced hundreds of milligrams per liter of culture. They are extremely cost-effective once they are developed.”
Llamas help with cellular processes studies
While Sikora and colleagues make advances with shark nanobodies, Rothbauer and his colleagues at Ludwig have been utilizing llama antibody fragments for their in vivo studies. “Sharks in Munich are not very convenient,” says Rothbauer. “Camels in Munich are also not very reasonable. But the Alpaca llamas that we use basically are like sheep. You can keep them like sheep. The only difference is that the wool is much better than sheep. So that’s an added value.”
Rothbauer’s llama antibodies have helped address one issue in particular when it comes to using antibodies in vivo. There is a gap that exists between the experiments that use fusion proteins and those that use antibodies, according to Rothbauer. Using fusion proteins, such as GFP, researchers cannot detect proteins in their natural state but only through the artificial construct introduced into the cell. Combined with the fact that conventional antibodies can only be used after cells have been fixed and permeabilized, no probe would allow researchers to study cellular components as they function naturally.
To fill this gap, Rothbauer and colleague Heinrich Leonhardt, biology professor at Ludwig Maximilians, fused their llama nanobodies with fluorescent proteins to visualize antigens in living cells. They call these hybrid affinity reagents “chromobodies,” and by using these reagents in living cells, they were able to visualize changes in the cell cycle in real time (1). “It is a new genetic construct comprised of a binding fragment on one terminal and a recombinant protein on the other. Now it’s possible to bind to the target structure and follow it as it changes,” says Leonhardt.
Manipulating protein folds and functions
Beyond visualization, nanobodies also have the capacity to control the structure and function of proteins of interest. In many cases, the conformation of a protein dictates its function in the cell; indeed, several genetic disorders, including Creutzfeldt-Jakob disease and mad cow disease, are the direct result of misfolded proteins.
To demonstrate this facet of nanobody capability, Rothbauer, Leonhardt, and colleagues generated llama-derived antibody fragments with affinity for GFP that actually changed the brightness of the protein by altering its conformation in a living cell (2). One of the fragments minimized the fluorescence, while the other maximized the fluorescence. “It was a very dramatic switch on and off,” explained Rothbauer. They believe that these nanobodies will one day be used to cure diseases caused by misfolded proteins.
“Beforehand, there was no way to detect protein folding in living cells,” says Rothbauer. “To our knowledge, this was the first demonstration where anybody detected or controlled conformation in living cells.”
Rothbauer and Leonhardt have founded a company called Chomtek, which provides researchers with llama-derived antibody fragments for a variety of bioimaging applications. “We want to turn these binders into valuable tools for high-content screening that output biomarkers,” says Rothbauer. The company currently offers products including the GFP-Trap and GFP-Booster nanobodies. GFP-Trap identifies and pulls down interaction partners of proteins tagged with GFP, reducing background noise in the experiment when compared to conventional antibodies. GFP-Booster allows GFPs to be used in super-resolution microscopy by changing the protein’s conformation to restore or increase the fluorescent signal. Rothbauer says they hope to add apoptosis and cell cycle assays in the coming year.
Cost-prohibitive and limiting factors
As Rothbauer and Leonhart have demonstrated, these nanobodies could revolutionize proteomic studies. But the high cost of developing nanobodies can make them cost-prohibitive for many basic research applications, according to Genway’s Sikora. The cost of GenWay’s custom shark antibody service can reach $25,000.
“The [development] process is long and laborious,” says Sikora. “If you look at competitive antibodies, for example, rabbit antibody monoclonals also go for $25,000. It’s much more expensive than $10,000 for mice monoclonals.” And so GenWay primarily markets their custom antibody services to the diagnostic and therapeutic clients, who, unlike basic researchers, can afford to pay these high prices.
However, once developed, these nanobodies can be easily amplified in E. coli. For example, 100 µg of Chromotek’s GFP-Booster sells for under $350 dollars, and enough GFP-Trap for 20 reactions sells for under $450, making Chromotek’s tools much more affordable for basic researchers.
Although the development process is straightforward (immunize the animal and then isolate the specific binders produced by the animal’s immune system), it takes nearly 2 months of immunization for the animal to mount a good immune response. This process is called the affinity maturation. “It’s like red wine, it just takes some time to mature,” says Rothbauer. Caring for these larger animals is also much more expensive and time-consuming than caring for the average lab mouse.
An alternative method, which takes two to four weeks, is to select a specific binder from a recombinant library. But that is a complicated selection process of its own. “You’re not only looking for high-affinity binders, but for special binders that not only visualize the target but also do not interfere with the function, ” notes Rothbauer.
“It’s easier said than done,” says Leonhardt. “There are several groups worldwide using such approaches. But at the moment, one has to admit that the animals are doing a better job at this.”
- Rothbauer, U., K. Zolghadr, S. Tillib, D. Nowak, L. Schermelleh, A. Gahl, N. Backmann, K. Conrath, et al. 2006. Targeting and tracing antigens in live cells with fluorescent nanobodies. Nat. Methods 3: 887–889.
- Kirchhofer, A., J. Helma, K. Schmidthals, C. Frauer, S. Cui, A. Karcher, M. Pellis, S. Muyldermans, et al. 2010. Modulation of protein properties in living cells using nanobodies. Nat. Struct. Mol. Biol. 17:133–138.