SINGLE MOLECULE DNA DETECTION, MAPPING &
SEQUENCING USING MAGNETIC TWEEZERS
Jimmy Ouellet1, Saurabh Raj1, Gordon Hamilton2, Charles André1,2, Jean-Baptiste Boulé3, Jean-Francois Allemand1, David Bensimon1, Vincent Croquette1
1
Laboratoire de Physique Statistique, ENS, 24 rue Lhomond, 75005 Paris;
CNRS 7196 / INSERM U565
2
PicoSeq SAS, 82 rue Fondary, 75015, Paris; 3Régulations et dynamique des génomes - MNHN -
ABSTRACT
PicoSeq and the lab of Vincent Croquette are developing a sequencing technology called SIMDEQTM (short for SIngle-molecule Magnetic DEtection & Quantification). In our
novel approach, individual DNA molecules are tethered to micron-sized paramagnetic beads and then manipulated using a ‘magnetic tweezers’ instrument. We have
demonstrated that a wide variety of different genetic assays can be performed using SIMDEQTM. The approach is simple, accurate, and has the potential to be run at very high
throughput. Here we present preliminary data showing partial and full re-sequencing of individual DNA molecules of up to 10kb in length. We also demonstrate the high
resolution mapping of DNA modifications (such as 5-methylcytosine) on unamplified DNA fragments, achieved by analysing the binding of modification-specific antibodies.
INTRODUCTION
The SIMDEQTM bench-top
prototype
RESULTS
Interrogating long fragments of DNA
LED
Sequencing CEB25, a 286 bps, 72% GC-rich, repetitive fragment
Our sequence detection process revolves around the ability to determine the 3dimensional position of a paramagnetic bead to an extremely high resolution. This is
achieved by shining a red LED light onto the beads (left panel) which generates a
diffraction pattern that can be imaged on a video camera (centre panel). Moving the
detector by precise increments while keeping the bead immobile allows the generation of
a 3 dimensional calibration image (a lateral section of which is shown on the right panel).
This system can accurately detect bead movement of 2-3 nm
•100
•Extension nM
•Open
Hairpin
•50
•Closed
Hairpin
•Time
Unzipping the hairpin. Lowering the magnet increases the force pulling on the
paramagnetic beads and the double stranded portion of the DNA is “unzipped,” typically
at a force of 10-15 pN. This reaction is reversible, and can be repeated many thousands of
times. The opening and closing of the hairpin can be monitored by tracking the Z position
of the bead in real time (as shown in panel B). The bead starts in the closed position
(extension = 0nM) and is subsequently opened (extension ~80nM) and allowed to close
again. The length of the open hairpin allows the total length of the DNA molecule to be
determined (in this case about 80 nucleotides).
SIMDEQTM can be used for a variety of applications. Three examples of raw
data are shown, demonstrating three applications for SIMDEQTM technology.
The top panel shows an 83 bp hairpin analyzed with 2 different oligos (note
that in addition to the 2 blocking sites, there is an additional slight pause
(Zunzip) as the hairpin first begins to rezip). The bottom left panel corresponds
to a 2.5 kb fragment of unamplified human genomic DNA with the blocking
pattern produced by an antibody against 5-methylcytosine; any DNA
modification for which a suitable antibody exists can be analyzed with this
technique. The bottom right panel illustrates the ability to analyze long DNA
fragments (in this case larger than 10 kb).
16s Bacterial Fingerprint
Rapid identification and quantification of complex bacterial populations using a unique barcode. The bacterial 16s DNA can be easily
characterised by using our fingerprint technique, typically using only 15 carefully selected oligos. This set of oligos, typically between 9 to 12
nucleotides, provides very specific hybridisation patterns along the 1.6kb of the PCR-amplified 16s DNA molecule. The hybridisation
positions have been chosen to be nearly equally spaced along the hairpin molecule. The resulting hybridisation pattern produces a unique
barcode profile for each DNA molecule, which is ideal for fast identification. This technique can also be used when the sample contains a
mixture of 16s molecules derived from different bacteria, in which case we can deduce the relative abundance of each species.
CONCLUSIONS
•100
•Extension (nM)
SIMDEQTM allows the sequencing of DNA fragments that are difficult or impossible with standard NGS techniques. GC-rich, repetitive
sequences like the CEB25 mini-satellites are notoriously difficult to sequence. SIMDEQTM provides a way to sequence these DNAs at the single
molecule level. Hybridisation of a single oligo produces a blockage for each repeat. Using a set of overlapping oligos we have resequenced a
286bp fragment containing more than 5 repeats with a 72% GC content. The hybridisation positions (measured in nm) are used to assemble
the full sequence using a Monte-Carlo algorithm. The overlap between the oligos allows correction of small errors in the calculated
hybridisation positions . This process was used to assemble the complete 286bp DNA sequence of a number of identical individual DNA
molecules without error.
•50
}
SIMDEQTM technology allows:
40nM
•Time
Blocking the rezipping with a hybridising oligo. In its unzipped state, the hairpin can bind
to complementary oligos. A bound oligo will then block the rezipping of the hairpin when
the force is reduced, which will be detectable as a pause in the retraction of the bead
position from open to closed. The length of this pause is largely dependent on the size of
the oligo which can be designed to block for just a few seconds before falling off and
allowing the hairpin to re-zip completely. The binding position of the oligo can be
determined precisely based on the position of the bead (in the figure above, at 40 nm) and
from this the sequence of that region of the hairpin can be inferred.
REFERENCES
Single-molecule mechanical identification and sequencing (2012) Ding F,
Manosas M, Spiering MM, Benkovic SJ, Bensimon D, Allemand JF, Croquette V.
Nat Methods. Mar 11;9(4):367-72
 Interrogation of DNA fragments ranging from 100 bp to more than 10 kb
 Sequencing repetitive, GC-rich regions such as the human minisatellite CEB25,
without error
 The rapid characterisation and identification of bacterial species, for example
through 16S- or operon-based fingerprinting
 The detection of modified bases on genomic DNA, with amplification-free
hairpin library preparation and antibody-assisted detection
Follow us on LinkedIn: http://www.linkedin.com/company/picoseq