www.technologyreview.com May/June 2018 MIT News 21
strains glided through the channel while others stopped at the
bottleneck, where the electrical field was most intense.
Precisely where a cell stopped—and how intense the electrical
field was at that point—showed researchers how polarized the
surface of the bacterium was, providing clues about its patho-genicity. More pathogenic strains of Pseudomonas aeruginosa
are more likely to polarize at lower voltages, which led Buie to
think that the research could be applied to help diagnose bacterial conditions like sepsis in time to save lives.
More practical methods already existed to determine whether
bacteria are pathogenic, however. Now, Buie is focusing on using
dielectrophoresis for connecting genetic information to the physical properties of bacteria. If, for example, genes are knocked out
of a little-studied bacterial strain, any resulting changes in polarization can offer “some information
on what might be the utility of those
genes or what area of the cell they
might be affecting,” he says.
FROM THE OUTSIDE IN
By 2013, Buie had turned his atten-
tion to finding a way to dramati-
cally speed up genetic engineering.
Researchers had long been inserting
different kinds of DNA into cells to
try to get them to, say, fight pathogens or metabolize carbon dioxide
to help mitigate climate change. But
tools for delivering the foreign DNA were not advancing as fast
as the strategies for harnessing the manipulated cells, he learned
after meeting a representative of the Defense Advanced Research
Projects Agency at a talk about synthetic biology.
To get DNA into cells, many researchers rely on electroporation—a method of using finely tuned electrical pulses to temporarily open pores in cell membranes. But the process requires
researchers to know the exact electrical field that will open the
pores without killing the cell. Finding that specific field, and the
right growth medium for any given bacterial strain, can take
years. And once that’s done, the process of preparing, pipetting,
and electrically zapping each sample is painstakingly slow. Buie
estimates that one skilled researcher can electroporate only 20
to 50 samples per hour, significantly restricting the number of
experiments a lab can conduct.
So with the help of a DARPA grant, Buie began working on a
faster way to electroporate cells—and eventually, he would also
tackle automating the process. His team—which included LEMI
postdocs Paulo Garcia and Jeffrey Moran as well as graduate
student Zhifei Ge, PhD ’ 16—used a setup similar to the one in
Buie’s previous dielectrophoresis experiments, but they added to
the microfluidic channel a fluorescent marker that would glow in
the presence of DNA. Channels were filled with bacteria, and as
electrical fields increased around the bottleneck, membrane pores
opened, letting in the marker. Once inside, it reacted with the bac-
teria’s DNA and made the cell glow, providing a visible indicator
of the electrical field needed to open a particular strain’s pores.
Once researchers know that, they still face the laborious task
of manually zapping each cell to insert the desired DNA. So
Buie, Garcia, and LEMI graduate research assistant Rameech
McCormack, SM ’ 17, also designed a pipette that applies the cor-
rect electrical field as the cells flow through electrically charged
microfluidic channels built into its tip. Through their startup,
called Kytopen, Buie and Garcia (who serves as Kytopen’s CEO) are
engineering an automated system
equipped with 96 or more simul-
taneously electroporating pipettes,
each of which can tackle one sam-
ple every eight to 10 seconds. By
making it possible to insert DNA
into bacterial cells up to 10,000
times faster, the device could
allow researchers to rapidly churn
through millions of variations on an
experiment in genetic engineering.
(Kytopen is among the first com-
panies supported by The Engine,
MIT’s venture fund/accelerator for
early-stage startups working on technologies with big potential but
long development time lines. See “Investing in Tech That’s Worth
the Wait,” March/April 2018.)
Buie’s next step is to test-drive his electroporation device on
organisms that scientists aren’t yet able to genetically engineer,
and he’s starting with bacteria in the mouth. The Forsyth Institute, a biotechnology research nonprofit, has isolated hundreds
of varieties of human oral bacteria. Buie has teamed up with
Forsyth’s Christopher Johnston, a microbiology researcher who’s
developing methods of dodging cell defense systems that reject
foreign DNA. Together, they aim to make at least 200 bacterial
strains available for genetic engineering.
If they succeed at significantly expanding the number of
organisms scientists can manipulate, the research could one day
be used to engineer bacteria for important applications in health
care, energy, agriculture, and environmental science.
Buie hopes his work will empower other researchers to take
on more difficult questions. “People will stop saying ‘I can’t work
on that bug’ because they can’t do genetics,” he says. “They’ll say,
‘Why don’t we try this?’” n
Surface properties determine whether bacteria
flowing through a pinched microfluidic channel
are immobilized by certain electrical fields.
“THERE’S JUST SO MUCH WE DON’T KNOW. ALL OF
THE WORK IS BEING DONE ON 1 PERCENT OF 1 PERCENT.
WE’RE LITERALLY JUST SCRATCHING THE SURFACE.”