Name: Jens Gundlach
Position: Professor of biophysics and gravitational physics, University of Washington, since 2003
Experience and Education:
In 2006, Jens Gundlach, a professor of biophysics and gravitational physics at the University of Washington, won a two-year, $650,000 grant under the National Human Genome Research Institute's $1,000 genome program to develop MspA, a protein pore from Mycobacterium smegmatis, for DNA nanopore sequencing.
Last month, he and his colleagues published a paper in PNAS in which they show that they can detect single molecules of DNA passing through engineered versions of the protein.
In Sequence spoke with Gundlach last week about his work and the prospects of nanopore sequencing.
Why did you decide to explore MspA for DNA sequencing, rather than alpha-hemolysin like most other protein nanopore researchers? What are the advantages of this protein?
This protein is a very stable protein that has an ideal geometry for sequencing. While alpha-hemolysin has a long, skinny barrel that is about 2 or 2.5 nanometers in diameter, MspA has a single, narrow section that, at the inner constriction, is somewhere between 1 and 1.4 nanometers wide, and also very short. So the specificity to individual bases of this protein is much higher.
It is also very robust to mutations. So we can change single amino acids in this protein, and it seems to fold up just the same way and forms stable pores. We are currently experimenting with all kinds of mutants of this protein to optimize it for sequencing. And the results are, at this point, very promising, to say the least.
How did you first become aware of this protein?
It was by accident. We have a friend here [at the University of Washington], Bertil Hille. He plunked a Science magazine onto my desk and said, 'Look at this,' and I went, 'Wow, this is fabulous.'
Michael Niederweiss, our collaborator [at the University of Alabama], is the guy who was involved in isolating this protein. I think he discovered it in some research that's related to tuberculosis. When we saw this protein in Science magazine, we called him up and asked him if we could have some of this stuff, and he sent us some protein, and nothing would go through it. So initially, it was actually very disappointing, because we were hoping to see events similar to what we had studied before with alpha-hemolysin, but we were not getting any DNA through there. Together, we figured out that we could probably modify this protein enough so that DNA would go through there.
[The crystal structure] had been done before we started using the protein. The structure is totally known from the crystal structure, and we believe that it very much reflects the in vivo shape.
In the paper, we showed we can translocate DNA through the pore. One of our proofs to show that DNA went through there is with this what we call 'fishing experiments,' where we anchor DNA segments to a streptavidin molecule that cannot pass through the constriction. And when the DNA single-stranded overhang dips into the inner constriction of the MspA pore, we see that the current gets blocked, and when we reverse the voltage, it diffuses out. However, we modified the system with a little trick: We put complementary DNA in the trans volume, so DNA that reaches through from the cis volume to the trans volume hybridizes with those segments in the trans volume and cannot easily diffuse out. And it requires a reverse voltage in order to pull the double-stranded section apart and have the streptavidin-anchored piece of DNA retract into the cis volume. That was very clear proof for us that the DNA actually went to the other side.
How have you improved this further since you wrote the paper?
The next really important thing is to show that MspA is actually specific to individual nucleotides, and we found excellent resolving power of MspA. We put sections of DNA that are homopolymers through MspA, and we observed very different current levels. And then we anchored heterogeneous sections of DNA with a hairpin - so there is a little section that folds over onto itself that holds the DNA in there for on the order of milliseconds or so. And when we do that, then we can see, very clearly, what nucleotide is at, for example, position two following the hairpin. That is actually a very nice result that we will be coming out with rather soon.
Are you planning to exploit this protein commercially for DNA sequencing?
Yes, of course we have patent applications in the process. And ultimately, we hope we can use this in a commercial fashion. One of the key points that is still elusive is to slow down the DNA as it passes through there. That is common to all nanopore sequencing methods, even in alpha-hemolysin: things pass through way too fast to be sequenced at the single-nucleotide level. We have several strategies that we are trying out right now to slow down the DNA passage. Those range from mutations to environmental conditions to little sections of double-stranded DNA interspersed, as well as anchoring the DNA to magnetic beads and so on — there is a lot of little tricks. I think just about everybody in the field of nanopore sequencing is working on how to reduce the velocity of DNA through the pore. But I think MspA forms a superior basis to alpha-hemolysin because of its excellent specificity.
Are you already working with a specific company, like Oxford Nanopore Technologies?
We have not made any industrial contacts at this point. Hagan Bayley and Oxford Nano are certainly people that we would be interested to talk to.
Generally speaking, when do you think nanopore sequencing will become a reality?
It's hard to say. We will have to solve that one issue, so that we can have free-streaming nanopore sequencing. As you know, Oxford Nano has successfully shown that they can look at single nucleotides going through a modified alpha-hemolysin, and that same thing, of course, can be done with MspA, and MspA might even be a bit more straightforward than alpha-hemolysin. The key point that has to be solved is to slow the DNA down - the electronics is basically not fast enough. We are also working on faster electronics and those kind of things. It could be just a few years and we will have a working system. What we have is what we believe, at least, is a significant development in the right direction. We have the most integral part of the sequencing process, this engineerable pore that we believe we understand quite well, and that is robust, and that has the intrinsic geometry that everybody wanted for sequencing.
How does this project fit in with your overall research interests? How did you first become interested in protein nanopores?
It's a weird story. I have another research program in a totally different area of physics, in gravitational physics. One day I went for a job interview and made it to the list of the three final candidates, and before I went I read the job description one more time, and it said 'We are actually quite interested in hiring a biophysicist.' I had heard about nanopore sequencing and thought that would be an interesting thing to look into. So I presented this idea, and I almost got the job. But it wasn't quite enough. And then I came back to my home institution here at the University of Washington, and I thought, that would be an interesting thing to try out. And there is actually very good and broad support for new, interesting ideas here, so the University of Washington helped me to get this started, and I had a very good graduate student, Tom Butler.
Initially, we were investigating alpha-hemolysin, and then we learned about MspA, and with some initial data on MspA, we applied for an NIH grant under the $1,000 genome program, and we were able to develop MspA to the point where we are right now and showed that this is quite a good workhorse for making the sequencing work.
I still have my other research group in gravitational physics. That's actually very interesting, because a lot of techniques and broad skills that I have learned in the other field are actually applicable to this also. For example, we are working on some amplifiers, which I learned about in nuclear physics, of all fields, and we modified those amplifiers, and they are actually superior to the commercial amplifiers, so we are using those also. I think it is actually a fruitful thing to bring in other disciplines and experiences from seemingly orthogonal fields, which, as I do this research, discover aren't all that orthogonal.
Is this the first paper that came out of your $1,000 genome grant?
This is pretty much the first major paper. We have published a few other ones before, but those were on alpha-hemolysin. This one is certainly the one with the biggest impact, I am pretty sure. It's just simply because the field of nanopore sequencing is actually quite big, and the field splits into two categories. One of them is solid-state nanopores, and the other one is protein nanopores. And it's on the protein side where actually most of the technology has been demonstrated, to show the biggest specificity to the DNA that goes through there. Everybody has been using alpha-hemolysin, though. So this is the first time that a group has shown that another protein pore is very suitable for sequencing, and I believe actually quite a bit more suitable.