— Dmitry, how did you find your way into biology?
— Actually, I was more into physics and mathematics than biology in secondary school. My physics teacher moved me to the math and physics class after one of the school Olympiads. In my graduate year, she talked me into applying to the Moscow Institute of Physics and Technology, and she convinced my parents that it was the right thing for me to do. I am deeply grateful for this to her and my parents. I also got accepted into the Department of Physics at Moscow State University, but I chose the MIPT and enrolled in the department I had my mind set on — Radio Engineering and Cybernetics. I joined the Chair of Systems Programming and found myself studying under the tutelage of the very team that had developed the Russian supercomputer.
Around this time, a friend told me about the Institute of Protein Research of the Russian Academy of Sciences. He told me that Alexei Finkelstein — my first research supervisor-to-be, a corresponding member of the Russian Academy of Sciences, and simply an outstanding person and scientist at the helm of the Protein Physics Lab in Pushchino — was looking for a student to work on a project. I promised my friend that I would go check it out, but I thought to myself, "Nonsense! What do I have to do with proteins?" I didn't even know what proteins were! But I went anyway.
The project sounded interesting, and Alexei Vitalievich took me on board right away. Later, it transpired that I didn't even have to switch departments. And so I continued my studies at the Chair of Systems Programming, writing some protein folding software, which was my degree project. At my viva voce, I held forth on what proteins are and how they fold. Some people were impressed, others resented the fact that they were taken away from their work on microprocessor design and system programming to listen to such nonsense. I joined the Institute of Protein Research first as a research intern and then became a graduate student. I worked there until the end of 2009, whereupon I continued my work overseas.

— Did you see better opportunities there?
— Not just that. In an academic environment, there is some pressure to change employers and take on new projects — they believe this is good for a researcher. The more employers you can show on your CV, the more diverse work experience you boast. Plus I figured there were no more big questions left in my field, so I might as well travel to see how things are in the neighboring fields of science. I spent my first four years abroad doing postdoctoral studies under Professor Dmitry Frishman at Technical University of Munich in Germany. Then I went to Barcelona, Spain for another four years under Professor Fyodor Kondrashov. When Fyodor's entire lab moved to Austria, I moved with it.

— How did all this diverse experience with so many colleagues benefit you?
— My lab's work branched out. My lab is active in two fields. One is study of proteins’ structure, the other — evolution, and specifically the study of epistasis (a type of gene interaction. — Author's note). I owe the latter line of work to my work at Fyodor Kondrashov's Evolutionary Genomics Lab. I might as well add that international experience significantly broadens one's interests, merging old knowledge with new insights. The older the researcher gets, the broader their perspective becomes, and the more mature and holistic view of science they develop.

— OK, now let's imagine we are attending your viva voce with microprocessor people in the audience. Explain what proteins are.
— Formally speaking, proteins are heteropolymers. Polymers are long molecules composed of some units. There are homopolymers, consisting of identical units (we all remember polystyrene or polyethylene from school), and then there are heteropolymers, made from different units. Proteins consist of 20 types of amino acids, with another two used in specific contexts. In reality, there are many more amino acids, but nature chose to use only these 20 in proteins. However, if we count in the so-called posttranslational modifications (covalent chemical modifications of a protein after its synthesis on the ribosome. — Author's note), their number will significantly increase.

— Why this number?
— It's the genetic code. It was decoded in the 1960s, and found to contain 20 amino acids. This remarkable universality is observed in all living organisms — they all share the same genetic code and set of amino acids. There are some variations in the genetic code, but they are all minor and they only number about 15 in total. The number of polypeptides, however, is astronomical. But not all of them become proteins, as proteins have the unique ability to perform various functions in the body.

— How is a protein different from a random polypeptide chain?
— Evolution has worked out an arrangement of amino acids wherein a specific protein amino acid sequence defines the protein's three-dimensional shape. This shape is stable and rigid, that's why it is able to perform certain functions. Let me clarify: take a pair of scissors, for example. Their function is to cut. Why are they capable of this? Firstly, because they are hard, and secondly, because they have sharp edges. If we had a piece of jelly shaped like a pair of scissors, we wouldn't be able to cut anything with it. It is the same with proteins: there are proteins, for instance, that cut other molecules (DNA, RNA, other proteins). They have to have a specific rigid shape to do it. This is what sets proteins apart from random polypeptides.

— What other roles do proteins have?
— Proteins perform 99.9% of all functions in our body, practically all there are. One exception are ribosomes, which produce protein chains as per instructions encoded in messenger RNA. The same holds for ribozymes. These are RNA molecules that can also fold into a strictly defined three-dimensional shape and have their own functions to perform.
Overall, protein functions can be divided into several categories. One category is hormonal. A certain well-known protein, insulin, is responsible for it. Insulin was the first protein to have its amino acid sequence determined. Frederick Sanger did it, and won his first Nobel Prize for it.
Another functional category is transportational. This work is performed, say, by hemoglobin, which is the red blood cell component that binds oxygen. This oxygen is then carried to the muscles and handed over to another protein — myoglobin. Hemoglobin and myoglobin were the first two proteins to have their three-dimensional structures determined using X-ray crystallography, earning Perutz and Kendrew the Nobel Prize.
There's also the catalytic function, executed with the aid of special proteins known as enzymes. These enzymes catalyze certain reactions. For instance, the enzyme alcoholdehydrogenase, as suggested by its name, breaks down alcohol. Collagen, the protein that forms the basis of the body's connective tissue (tendons, bones, cartilage), performs a connective function. Proteins also have other roles. Let me put it this way: there are over 20,000 genes in our body. And nearly all of them are genes that code for proteins. This speaks volumes.

— What are some fascinating developments in protein science there days?
— A very timely question. Had you asked it before the summer of 2018, my answer would have been entirely different. But today, I can confidently say: a game-changer has arrived! But first, a bit of history. In 1961, it was experimentally demonstrated that the three-dimensional shape of a protein, which we discussed earlier, is determined solely by its amino acid sequence (it was previously believed that ribosomes had something to do with it). This was conclusively reaffirmed in 1969 when a fully functional protein was chemically synthesized without the involvement of a ribosome. Then scientists reflected: if we know the amino acid sequence, we can predict the three-dimensional structure of a protein. But it wasn't so simple on the ground.
Scientists struggled with it continually for half a century with mixed success and many disappointments.
In 2018, DeepMind, a Google-owned company, introduced AlphaFold, a unique program for predicting the three-dimensional structure of proteins. Interestingly, they had honed their knowhow developing chess and go programs, both of which operate on a similar principle — learning from good and bad moves leading, respectively, to a win or a lose. The chess program got so advanced that, given a certain number of processors and six hours of training, it emerged capable of outplaying any human chess player in the world. In 2020, AlphaFold achieved another milestone through training: it learned to predict protein structures. The accuracy of its predictions ranges from 88% to 90%, nearly on par with scientific experiment.
This is an impressive achievement. Now anyone can install this program and gain the ability to predict three-dimensional structures with near-experimental precision. Alternatively, anyone can look up the structure of any known protein on the AlphaFold model database.

— What's in it for us?
— Quite a lot, actually. When we know the three-dimensional structure of a protein, we can find out what its function is. Take coronavirus, for instance: it's known to attach to the host cell using the so-called spike protein, which sparked much controversy and has been amply written about. Now that we know its structure, we have the means to combat the virus. And not just this virus, but other diseases as well. With this game-changing achievement, it will be easier to design effective drugs, develop new vaccines and antibiotics.
It will also drive progress in designing proteins with preset properties. The applications scope for those will extend beyond biology and medicine. The chemical industry already uses proteins as bio-additives in laundry detergents. They are the selfsame "smart molecules" touted in commercials. Some proteins taste sweet. They can serve as safer alternatives to sugar.

— What are you working on at this time?
— The topic of protein structure and its modifications sounds very interesting to me. The thing is, we are now equipped to introduce "mutations" to proteins and thus enhance the properties we need. For example, we could engineer proteins that are more resistant to high temperatures or pH levels. Our lab is working on that at the moment. Ultimately, my ambition is to reach the expertise level of David Baker's team. Baker directs the Institute for Protein Design, which was established specifically to build on his accomplishments. He takes on serious challenges, like, for example, creating a protein from scratch that emits a certain color. In the late 2000s, his team engineered a protein that catalyzes a reaction not found in nature. My dream is to get similar results.