American Scientist

A molecular engineering lab at Cornell University made headlines this spring when the research team led by Matthew DeLisa, the William L. Lewis professor of engineering, announced they had developed a new form of antibody by inserting an antibody fragment into an existing cellular pathway. DeLisa (below) recently talked with Sandra Ackerman, senior editor of American Scientist, about how he and his colleagues came up with this innovation and what makes it so exciting.

Why did your lab name the new form of antibody a “ubiquibody?”

We gave it that name because we combined an antibody with an existing quality control pathway known as ubiquitination. In the end, what we created was a way of directly targeting proteins for destruction by the ubiquitin pathway.

When people go to the website of your lab for more information, they’re likely to have a few questions. In an overview of your work, you say, “One approach in our laboratory is to exploit untapped mechanistic features of existing cellular machinery.“ What are these mechanistic features, and what’s the machinery in the cells?

What that phrase is trying to capture is the fact that all cells—including the types we work with in the lab, which are bacterial cells—carry out a host of specific tasks. One of the most important is to produce proteins, through an elaborate process that we now know involves a very complicated array of cellular machines. One example would be a machine known as the ribosome, which is the center of all protein synthesis. What the ribosome does is decode messenger RNA, and it does that by linking together various amino acids. So these cell features are mechanistic in the sense that they’re machines that contribute to building proteins.

Were these mechanisms unclear until recently, or were they known but because of technological reasons they couldn’t be harnessed?

Both, actually. We work on a multitude of different systems. In some cases it’s a system we’ve known about for a long time but the research was focused on understanding how the system works, so nobody has leveraged it yet for a new technology. In other cases, some of the mechanisms we work on are very newly discovered; five or ten years ago we didn’t know they existed. And now we see an opportunity to leverage these with new technologies.

Many of the things we ask the cells to make are things the cells wouldn’t normally do. For example, we ask a cell to produce a protein that might be used as a therapeutic, so we give the bacterial cell a sequence of DNA that encodes for a human protein and we ask the cell to make that protein for us.

And of course we’re not alone in doing this. Since the invention of recombinant DNA technology in the late 1970s, many scientists and engineers have leveraged this approach, as well as many very well known biotech and pharma companies.

Here’s another short question that probably has a long answer: a lot of your work draws on the idea of a quality-control system for proteins within cells. Is this a fairly new idea in your field, or is it old hat?

That one’s not so new for us, although there are so many types of quality control that we’re still discovering new mechanisms all the time. And because the idea has been around for a long time, we have begun to look for these kinds of mechanisms. It’s something we’re always looking for, given how important it is, when a cell goes about making a protein: new forms of quality control.

One of the best analogies is an assembly line at automobile plant. Some of the quality-control steps may involve a human being looking at various parts and pulling something off the line when it doesn’t look right. Well, the cell is very much the same way. We started this conversation talking about protein “machinery,” a concept borrowed from our own lives, but cells have protein machinery that’s making the proteins, linking together the amino acids, decorating them with complicated sugars, and making sure they fold into the proper formation. And at each one of those steps there are quality control systems that inspect the end product. When everything is fine, then the protein can proceed with its life. But when things go wrong, then hopefully, in a healthy cell, the mechanisms can deal with an improperly assembled or improperly folded protein and get rid of it when necessary—pull it off the assembly line, so to speak.

That leads me to ask, Do you have a sense there are more mechanisms in there?

I sure do, and I can give you a great example of why I think this is the case. One important type of biological function that cells perform is attaching complex sugars to their proteins. All cells, from yeast all the way up to humans, perform that function. It’s a process generally referred to as glycosylation. For decades, and perhaps longer, it was accepted as scientific fact that bacterial cells—one of the simplest of all forms of cellular life on our planet—did not perform this modification: they did not attach complex sugars to their proteins. But just in the last few years there have been a number of seminal studies showing that some bacteria in nature do glycosylate their proteins. To me that’s just one example of why I would never say “never” at this point, because there have been too many cases where “never” was proven to be untrue.

Now, in the case of E. coli bacteria ,which is what we and the biotech industry have used for decades, that still holds true, as far as we know: E.coli do not glycosylate their proteins to any great extent. And in light of that, one of the things my lab has worked on has been to introduce glycosylation systems into E. coli. Since we know some bacteria can do it, we thought perhaps we could engineer E. coli cells to do the same reactions, and in fact we’ve been very successful in making that happen.

But I think there are still many more mechanisms in bacterial cells—in any kind of cells, human cells included—that are yet to be discovered.

Let’s look at sequenced genomes. In all the sequences that are out there, from bacteria all the way up to humans, there’s an incredible number of genes that encode for proteins whose function is completely unknown. And even for the best-studied organisms, like E coli, we still only know about 50 percent of what their proteins do. This fact tells me that, while some of those proteins may be part of biological processes that we already know about, there could still be a whole host of proteins whose functions and activities have not yet been discovered.

I guess I’m a glass-is-half-full kind of person: the way I look at it, there are still a lot of things to be turned over. And what I hope my lab will do, if there’s technological potential in any of those biological mechanisms, we would like to be able to leverage those as well.

I think it’s really helpful to hear someone in your field say this. From outside, it can sometimes look like the attitude in research labs is, “We pretty much know what’s there and are figuring out better ways to use it.” But you’re saying that is far from the case.

I would frame that by saying that perhaps the rate at which we find new things has slowed, because many of the major things have been discovered. But we continue to find, and are amazed to find, new activities all the time. That’s true in simple bacteria, where we know a lot—and I think it extends in an even more relevant way to higher organisms, where our understanding is perhaps not quite as refined.

So yes, I still think there are many new things to be found. I would venture a guess that that’s one reason a lot of scientists and engineers continue to do what they do—the idea that there are still many undiscovered processes. These will be important not only for our understanding of life on this planet, but also because a lot of the greatest biological discoveries also have very powerful therapeutic potential.