Electron microscopy DNA sequencing

Electron microscopy DNA sequencing is the name for a family of methods, proposed from 1959 and pursued by two companies until 2012, that would read the order of bases in a DNA molecule by attaching heavy atoms to the bases and photographing the molecule in an electron microscope. The sequence would be read off the picture, in the way a line of text is read off a page. No such method produced the sequence of a genome. The genomes of the 2000s were read by other means, in short fragments.^[1]

The idea has a single obstacle at its center. The atoms in a DNA strand are carbon, nitrogen, oxygen, hydrogen and phosphorus, which scatter electrons weakly and look alike in a micrograph, and the electron beam that images them also destroys them.

The idea was stated in There's Plenty of Room at the Bottom, Richard Feynman's lecture of December 29, 1959, which listed what biologists would learn if the electron microscope were a hundred times better. "It is very easy to answer many of these fundamental biological questions; you just look at the thing," he said. "You will see the order of bases in the chain."^[2] By March 1962 Michael Beer and Evangelos Moudrianakis had published, in the Proceedings of the National Academy of Sciences, a paper titled "Determination of base sequence in nucleic acids with the electron microscope: visibility of a marker".^[3] The company's quiz game JeoPOREdy! credits Feynman's lecture with "the first ever proposed method to sequence DNA" and notes that Beer and Moudrianakis started on the experiment "around the same year".^[4]

For an unscreened nucleus the cross-section for elastic scattering of electrons grows as the square of the atomic number. Mercury, with atomic number 80, would scatter about 180 times as strongly as carbon, with atomic number 6, and a heavy atom on a light strand should be conspicuous. The difficulty was seeing one atom at all.

In 1970 A. V. Crewe, J. Wall and J. Langmore reported that with a high-resolution scanning electron microscope "it is now possible to obtain pictures of a single heavy atom resting on a thin carbon substrate".^[5] A year later R. M. Henkelman and F. P. Ottensmeyer published dark-field micrographs of individual atoms of palladium, iodine, platinum, osmium and uranium, and found that their scattering matched the cross-sections calculated from a Thomas-Fermi-Dirac model of the atom.^[6] Ottensmeyer's group went on to nucleic acids. In 1972 R. F. Whiting and Ottensmeyer imaged heavy atoms in model compounds and in nucleic acids, and in 1977 they reported that a platinum-dimethylsulfoxide complex binds the bases unevenly: one platinum atom to each pyrimidine, two to adenine, and one or two to guanine, depending on pH. It could therefore mark the adenine and guanine sequence of a single strand, and the authors thought that with a complementary strand all four bases should be readable.^[7][8]

Two companies took up the method in the late 2000s. Halcyon Molecular, in Redwood City, California, labeled the bases with heavy atoms, stretched long strands flat on films a few nanometers thick with a process called Molecular Threading, and planned to read them in a scanning transmission electron microscope. The company is described in its own article, and its founding team from 2009 to 2012 included Kent Kemmish.

ZS Genetics, a startup in North Reading, Massachusetts, was described in 2008 as developing a technology that allowed scientists "to take photographs of the sequence of a DNA molecule".^[9] In 2012 it and its collaborators reported what they called the first identification of DNA bases within intact molecules by electron microscopy. They used a nucleotide carrying a single mercury atom, which the polymerase placed opposite each adenine in a synthetic DNA of 3,272 base pairs and in a viral genome of 7,249, and they imaged the labels by annular dark-field scanning transmission electron microscopy. Only one of the four bases was labeled. Further refinements to identify additional base types, they wrote, "would allow full sequencing of long, intact DNA molecules".^[10]

The method asked for a long strand with a label on every base, so the scale is a problem. The bases of DNA sit about 0.34 nm apart along the strand, so a read of 150,000 bases, the length that the NIH grant to George Church's laboratory sought, is a thread 51 µm long that has to be resolved at better than the spacing between bases.^[11] Radiation damage is the second problem. Richard Henderson wrote in 1995 that it "is the main problem which prevents the determination of the structure of a single biological macromolecule at atomic resolution using any kind of microscopy", although for biological material electrons "provide the most information for a given amount of radiation damage".^[12] Labels with a heavy atom each had also to be put on the bases without disturbing the strand, and by 2012 one base of four had been labeled.

Meanwhile, sequencing of fragments became cheap. The National Human Genome Research Institute notes "the sudden and profound out-pacing of Moore's Law beginning in January 2008", when sequencing centers moved from the Sanger method to second-generation machines.^[1] In August 2012 GigaOm reported that Halcyon had closed, writing that Oxford Nanopore had "beaten Halcyon to the punch".^[13] Nanopore sequencing reads a long strand by the changes it makes in the current through a pore, with no image and no heavy-atom label.

• Halcyon Molecular
• Molecular Threading
• There's Plenty of Room at the Bottom
• History of the microscope
• History of nanopore sequencing