Researchers from the University of Alabama at Birmingham and the University of Washington have shown that they can tweak the way the Mycobacterium smegmatis porin A, or MspA, protein is produced — a step that is expected to lead to more refined methods for tailoring the protein in the nanopore sequencing arena.
MspA is a homo-octomer protein produced from eight identical subunits encoded by the same gene. At the moment, that means that any mutation that is introduced into the gene — for instance, to try to optimize the protein's potential for nanopore sequencing or to customize it for other applications — ends up in all eight MspA subunits.
"Because MspA consists of eight monomers, when we mutate the MspA gene, each mutation appears eight times in the pore," University of Alabama at Birmingham microbiology researcher Michael Niederweis told In Sequence.
"We want to create an MspA nanopore that consists of only one protein domain," he said.
If and when it is possible to produce all of the MspA protein subunits from one large gene rather than eight identical subunits from one small gene, researchers would be able to be more precise about the mutations they make in the mspA gene and their location in the resulting pore, Niederweis explained.
For instance, they could theoretically target mutations to occur just once in the entire protein or add mutations affecting only one or a few of the subunits that comprise the pore, depending on the type of improvement they are after.
In PLoS One this June, Niederweis and his colleagues published a proof-of-principle study suggesting it should be possible to produce such single-chain versions of MspA. In particular, the researchers reported that they have taken the first step toward that goal: producing functional MspA from four subunit dimers rather than from eight independent subunits. Within these proteins, each dimer is encoded by a pair of mspA genes that are expressed together with a linker sequence in between.
"We were very encouraged by the result that we just published," Niederweis said.
"We made, basically, a gene that consists of two mspA genes," he added. "We connected these mspA genes to create dimeric MspA and that assembles with three other dimers to form the MspA pore."
The resulting protein proved capable of forming functional channels in M. smegmatis cells. And, the researchers reported, the purified dimer-derived MspA protein formed pores in their lipid bilayer studies and translocated DNA with similar current profiles as conventional MspA pores.
"These results represent a key step in altering the subunit assembly of MspA to increase the control of the chemical and biophysical properties of the MspA channel for nanotechnological applications such as DNA sequencing," the study's authors wrote.
The team now plans to try making MspA from even longer strings of genes, starting with pairs of tetramer subunits encoded by four linked copies of the mspA gene.
"The next step would be to make a gene that consists of not two individual mspA genes, but four. Then we would only need two of those tetramers to form an MspA pore," Niederweis said. "The final step would be to create a very long gene that consists of eight individual mspA genes so that the protein made from this gene is the complete MspA pore."
Once that type of construct has been created, researchers can pursue more targeted potential improvements to the pore protein, such as mutations leading to even more distinguished signals for each nucleotide and/or mutations that slow DNA transit through the pore.
The latter property could be particularly important for coming up with a streamlined nanopore sequencing device down the road, since the team is currently relying on a base detection strategy that uses a second protein, phi29 DNA polymerase, to slow DNA transit through the pore.
Niederweis noted that the team is also interested in exploring the consequences of changing the size of the MspA constriction zone to see if it's possible to improve on the protein's natural pore diameter.
Their current subunit dimer production strategy is similar to that outlined in a Journal of Biological Chemistry article last year by University of Oxford chemist and Oxford Nanopore Technologies founder Hagan Bayley and his colleagues. They showed that it is possible to put together the alpha hemolysin pore protein, which contains seven subunits, from subunit dimers in combination with monomers.
That team has been developing its own nanopore sequencing system based on modified versions of alpha hemolysin (IS 10/12/2010), a protein that originates in Staphylococcus bacteria. Oxford Nanopore's device is reportedly set to hit the market sometime this year (IS 2/7/2012).
While that group has been hitching its wagon to alpha hemolysin, Niederweis, University of Washington biophysicist Jens Gundlach, and their colleagues have been working on a nanopore sequencing strategy built around MspA, which their research suggests could be inherently better suited for such applications (IS 8/24/2012).
And earlier this year, authors of the current study collaborated with other University of Alabama and University of Washington researchers on a Nature Biotechnology study demonstrating that they could get single nucleotide resolution using a nanopore based on a mutant version of MspA while controlling DNA movement through this pore (IS 3/27/2012).
For the current study, researchers started with a series of preliminary experiments involving proteins produced using MspA dimers and dimers of a closely related protein, MspB.
After showing that it was feasible to produce proteins from those dimers, the team turned its attention to proteins produced from MspA subunits alone, creating fusions between two mspA genes that were joined by linker sequences and cloned into an expression plasmid.
To test the functionality of MspA proteins produced from MspA dimers, the researchers expressed them in M. smegmatis bacteria that were missing a wild type version of the gene. The presence of MspA produced from the dimers they had designed bumped up the cells' ability to take up glucose, they reported, suggesting the pore protein is functional and capable of proper folding and localization.
They also purified the resulting MspA proteins and characterized a range of properties such as pore conductance and voltage gating. "What we saw was that the MspA that was made from MspA dimers had properties that were very similar to those of wild type MspA," Niederweis said.
Similarly, when collaborators in Gundlach's University of Washington lab tested purified MspA proteins produced from dimer subunits, they found that these pores effectively translocated DNA.
Again, properties of the pore were very similar to wild type MspA despite the presence of linker regions joining the two parts of each dimer.
"We concluded that the linker does not influence the pore properties of MspA," Niederweis said.
"This was important to show because we want to use the single chain MspA for nanopore sequencing," he explained. "If this linker had a drastic influence on the pore properties, this would have meant that we could not have used this approach for DNA sequencing."
So far, the researchers have not done a systematic analysis to see if there is an optimal linker length, though Niederweis said that that may be an avenue to explore in the future, particularly since it might influence MspA expression levels in M. smegmatis.
Expression of single-chain MspA proteins is, indeed, a potential concern. In the current study, for instance, researchers found that the linked proteins were trickier to produce than their wild type counterparts: their expression in M. smegmatis was one-tenth of that found for the wild-type MspA protein.
Though these expression levels are still high enough that researchers can work with the protein, Niederweis noted that the team might turn to an alternative organism such as Escherichia coli to express linked monomers down the road — for instance, if the expression of increasingly longer constructs drops off too far.
There are other considerations in that system, he added, with studies by other groups suggesting that it can be tough to refold the MspA protein in E. coli.
The team has patented its MspA production and other nanopore sequencing techniques, Niederwies noted, and is aiming to produce a commercial nanopore sequencing system based on MspA, though the timeline for that device is still uncertain.
"We still have a lot of work to do to make MspA into such a device, but when you look at the published data, MspA is by far the best protein for that application," he said, "so I think there is a lot of commercial interest in achieving this."
On the nanopore sequencing front, Niederweis, Gundlach, and their collaborators are continuing to take advantage of the DNA ratcheting method developed by members of Mark Akeson's University of California at Santa Cruz lab (IS 3/29/2011).
The strategy relies on the processive phi29 DNA polymerase at the nanopore periphery and on the presence of a blocking oligomer incorporated onto the template DNA that gets removed during DNA's first trip through the pore. Once it's gone, the polymerase begins replication and the strand passes through the pore in the other direction (IS 2/21/2012).
"This is currently the best approach, but it is a complicated setup because you have two proteins that both need to be controlled and that need to work together," Niederweis noted.
Consequently, as their ability to specifically modify MspA improves, the researchers hope to come up with tweaks to MspA that would delay DNA translocation without the additional phi29 component.
But, Niederweis cautioned, it remains to be seen whether it is possible to introduce mutations to MspA that would allow properties of the pore itself to sufficiently slow DNA transit. And, he said, it's possible that a first-generation MspA nanopore sequencing device could include both the phi29 polymerase and MspA proteins.
"Ultimately we'd love to get rid of the DNA polymerase and just use the pore because this would be a much simpler mechanism, but whether this is feasible or not we don't know yet," he said.