Visualizing Single-molecule DNA Replication with Fluorescence Microscopy



This protocol demonstrates a simple single-molecule fluorescence microscopy technique for visualizing DNA replication by individual replisomes in real time.

Cite this Article

Copy Citation | Download Citations | Reprints and Permissions

Tanner, N. A., Loparo, J. J., van Oijen, A. M. Visualizing Single-molecule DNA Replication with Fluorescence Microscopy. J. Vis. Exp. (32), e1529, doi:10.3791/1529 (2009).


We describe a simple fluorescence microscopy-based real-time method for observing DNA replication at the single-molecule level. A circular, forked DNA template is attached to a functionalized glass coverslip and replicated extensively after introduction of replication proteins and nucleotides (Figure 1). The growing product double-strand DNA (dsDNA) is extended with laminar flow and visualized by using an intercalating dye. Measuring the position of the growing DNA end in real time allows precise determination of replication rate (Figure 2). Furthermore, the length of completed DNA products reports on the processivity of replication. This experiment can be performed very easily and rapidly and requires only a fluorescence microscope with a reasonably sensitive camera.


Before performing a single-molecule replication experiment, a few materials need to be prepared in advance.

1. DNA Replication Template

The substrate for the replication reaction is a biotinylated, tailed M13 rolling circle prepared using standard molecular biology techniques1,2.

Materials: M13mp18 single-stranded DNA, Biotinylated tail oligonucleotide primer (5'-Biotin-T36AATTCGTAATCATGGTCATAGCTGTTTCCT-3'), T7 DNA polymerase, Heat block, Phenol/Isoamyl Alcohol/Chloroform

  1. Anneal the biotinylated tail oligo to the M13 by adding a 10-fold excess of oligo in TBS buffer (330 nM tail oligo, 33 nM M13).
  2. Heat the mixture to 65 °C in the heat block. Once heated, turn the heat block off and allow slow cooling to properly anneal the primer.
  3. Add the primed M13 to a mixture of 64 nM T7 DNA polymerase and T7 replication buffer containing dNTPs and MgCl2.
  4. Incubate the reaction at 37 °C for 12 minutes and quench with 100 mM EDTA.
  5. Purify the filled-in product DNA with phenol/chloroform/isoamyl alcohol extraction and dialyze into TE buffer or other suitable storage buffer and determine DNA concentration ( Back-extract each organic phase to recover any residual primed M13).
  6. One iteration of template preparation can easily make several milliliters of nanomolar concentration template, enough for hundreds to thousands of single-molecule experiments.

2. Functionalized Coverslips

For attaching the DNA to the glass coverslip, the glass is first functionalized with an aminosilane, which is then coupled to biotinylated PEG molecules. This coating helps to reduce the nonspecific interactions of DNA and replication proteins with the surface1,3.

Materials: Glass coverslips, Staining jars, 3-aminopropyltriethoxysilane, Methoxy-PEG5k-NHS ester, Biotin-PEG5k-NHSester, Acetone, 1M KOH, Ethanol, Oven, Bath sonicator

  1. Thoroughly clean the glass coverslips by placing them in staining jars and filling the jars with ethanol. Sonicate for 30 minutes and rinse out the jars and fill with 1 M KOH. Sonicate for 30 minutes Repeat both washes.
  2. For the first functionalization reaction, all traces of water need to be removed. Fill the jars with acetone and sonicate for 10 min. Rinse the jars again with acetone.
  3. Prepare a 2 % v/v solution of the silane reagent in acetone. Add to the staining jars and agitate for 2 minutes. This reaction couples the alkoxy group of the aminosilane to the glass, leaving a reactive amine for the next coupling step. Quench the reaction with a large excess of water.
  4. Dry the coverslips by baking them at 110 °C in the oven for 30 minutes.
  5. Prepare a 50:1 methoxy:biotin PEG mixture in 100 mM NaHCO3 pH 8.2. Aim for around 0.2 % w/v biotinylated PEG.
  6. Pipette 100 μL of the PEG mixture onto a dry silanized coverslip and place another coverslip on top. Including a glass coverslip spacer will allow separation of the coverslips.
  7. Incubate the coverslips in the PEG solution for 3 hours, then separate each pair of coverslips and wash extensively with water. Be careful to keep the coverslips functionalized side up as only once side will be coated with PEG.
  8. Dry the coverslips and store under vacuum. The surfaces remain stable for at least a month, so dozens of coverslips can be made in a batch and used as needed.

These materials should be made ahead of time, and then used for each experiment. To start every single-molecule experiment, begin by assembling the flow chamber.

3. Flow Chamber Preparation

The experiment is performed using a simple flow chamber constructed with a functionalized coverslip, double-sided tape, a quartz slide and tubing. One flow chamber is prepared for each single-molecule experiment1,4.

Materials: Double-sided tape, Razor blade, Quartz slide with holes for tubing, Quick-dry epoxy, Functionalized coverslip (see above), Streptavidin solution (25 μL of 1 mg/mL in PBS), Tubing, Blocking buffer (20 mM Tris pH 7.5, 2 mM EDTA, 50 mM NaCl, 0.2 mg/mL BSA, 0.005% Tween-20)

  1. Begin by placing 20-25 mL of blocking buffer in a desiccator to remove any air bubbles for later steps.
  2. Take a PEG-functionalized coverslip and spread 125 of PBS-diluted streptavidin solution over the surface. Leave this to incubate for 20 minutes while preparing other parts of the chamber, allowing streptavidin to bind surface biotins.
  3. Cut a piece of double-sided tape to match the size of the quartz slide. Mark the position of the tubing holes on the tape with a pencil so that an outline of the flow chamber can be drawn on the tape (1 mg/mL).
  4. Make a 2 mm-wide rectangle around the holes. This will serve as the flow channel, so be sure to make straight sides and leave room around the tubing inlet and outlet. If desired, multiple channels or different sizes can be used.
  5. Cut along the drawn outline, making straight, clean cuts to make sure no adhesive protrudes into the channel.
  6. Clean the quartz thoroughly using acetone or ethanol to remove adhesive from the previous flow cell construction.
  7. Find the best alignment of the channel outline, remove one side of the adhesive backing and carefully place the tape onto the quartz slide. Be careful to align the tape properly, as the inlet and outlet holes need to remain unblocked.
  8. Cut lengths of tubing for the inlet and outlet of the flow cell. It helps to cut the end of the tube at about a 30° angle so that the tube will not press flat against the chamber bottom. Place the tubes on some sort of support to suspend them for easy attachment to the flow cell in the next steps.
  9. Rinse the streptavidin-coated coverslip thoroughly with water and dry using compressed air. Remember that only one side is functionalized; be careful not to turn the coverslip over. Remove the other side of the tape backing and place the quartz slide onto the coverslip.
  10. Lightly press on the coverslip to push out any air trapped in the adhesive. This will help avoid any air bubbles getting into the flow channel.
  11. Seal the sides of the chamber with epoxy. Insert the cut tubing into the holes of the quartz and seal in place with epoxy. This needs to dry for a few minutes.
  12. Once dry, begin blocking the surface by pulling some of the degassed blocking buffer through one of tubes. A 21-gauge needle fits perfectly inside the 0.76 mm tubing. Flush out a few times to remove air bubbles, and allow the chamber to incubate for at least half an hour.

4. Single-Molecule Replication Experiment

Materials: Prepared flow chamber, SYTOX Orange, TIRF microscope, 532 nm laser, CCD camera, Computer with image acquisition software, Rolling-circle DNA substrate, Replication proteins

  1. After the flow cell has been blocked, everything is ready to begin the single-molecule experiment.
  2. Take the flow cell and place it on the microscope stage. Hold the chamber in place with stage clips and be sure that the flow channel is positioned straight along the y-axis.
  3. Connect the flow cell outlet tube to the syringe pump using a larger diameter connector tubing or a needle. Place the inlet in blocking buffer and pull back on the syringe to remove any air in the tube. A gentle flick of the outlet tube will help clear any air bubbles trapped in the flow cell.
  4. Dilute the stock DNA template to 25 pM in 1 mL of blocking buffer. Flow into the chamber at moderate flow rate to allow good surface coverage of DNA. 0.025 mL/min for 30 minutes works well. This can be varied based on how many DNA molecules are on the surface for each batch of coverslips or how many are desired for the experiment.
  5. Once DNA is in, wash out excess DNA using blocking buffer. Wash for at least 200 flow-cell volumes to get rid of all the free DNA.
  6. Begin degassing the replication buffer. For the E. coli replication experiments, do this at 37 °C to reduce the risk of bubbles in the heated flow chamber.
  7. Turn on the laser and camera. The cooled CCD needs to reach temperature before it can be used.
  8. If performing a 37 °C experiment, turn the heater on. Be sure the objective is in contact with the flow cell or it will be a heat sink later.
  9. After washing and degassing, the reaction can be prepared. Mix nucleotides, DTT, and SYTOX in the replication buffer. Then add proteins and mix gently.
  10. Flow the reaction mixture into the flow cell. This flow speed needs to be sufficient to stretch dsDNA. For a 2-mm wide, 100-μm tall flow channel 0.04 mL/min works well.
  11. Allow time for the mixture to enter the chamber and begin imaging. A good tip is to move the field of view to the side of the channel and focus on the adhesive material. This will ensure you are near the surface for focusing.
  12. Adjust laser power, microscope focus, and TIRF angle. Keep the power low to avoid any photocleavage of the stained DNA. Acquire at 1-5 frames/second for several minutes depending on the rate of DNA replication. If desired, repeat to get longer replication trajectories or move to a new field to see more molecules.
  13. After the reaction mixture is almost empty, it is a good idea to flow in more buffer with SYTOX to see other fields of view. Taking multiple images of different replicated molecules provides statistics for processivity determination.

5. Representative Results

Actively replicating molecules are easily seen as growing lines of stained DNA. In our experiments, flowing 25 pM of the M13 substrate gives 100-1000 molecules in a 60x, 125 μm x 125 μm field of view (Figure 1). Performing replication experiments using the E. coli replisome proteins at 37 °C (see Materials for details) gives 5-50 replicating molecules per field of view. At equivalent concentrations of the T7 replisome, which initiates on the substrate much more efficiently, we observe >70% of the molecules in a field replicate. These conditions yield a product density so high that individual molecules are difficult to resolve and analyze, so the T7 experiments are performed at lower protein concentrations.

Data Analysis: To obtain rate data, simply plot the endpoint of the DNA vs. time and calculate the slope (Figure 2). For processivity, determine the total length of the DNA. Both of these numbers will need to be converted to base pairs. Before performing the experiment, you should know the pixel size of the camera at the proper magnification. For basepair conversion, a good estimation is simply converting based on the crystallographic length of DNA, 2.9 bp/nm. However, laminar flow will not completely stretch the DNA, so a DNA length calibration should be performed by taking a known length DNA, e.g. 48,502 bp λ DNA, attaching it to the flow cell with a biotinylated oligo and measuring its SYTOX-stained length at the flow rate used for replication experiment. Determining the number of base pairs/pixel will allow accurate calculation of replication rates and processivities. Taking numerous images and movies will provide large numbers of product molecule, allowing plotting of rate and processivity distributions (Figure 2)1.

Figure 1
Figure 1: (From Ref. 1) a) Cartoon of the rolling-circle replication assay. (SA, streptavidin). Leading-strand synthesis extends the tail linking the circle and surface. The tail is converted to dsDNA by the lagging strand polymerase and stretched by laminar flow.
b) Example field of view with SYTOX Orange and 532 nm excitation. Small fluorescent spots are tethered, non-replicated substrates. Note the extreme length of the products and the large number of products per field (each flow cell contains >5000 such fields).

Figure 2
Figure 2: (adapted from Ref. 1) a,b) Kymographs of typical replicating molecules from T7 (a) and E. coli (b) experiments. c) Endpoint trajectories of molecules from a) and b) plotted vs. time. Trajectories are fitted with linear regression to obtain replication rates. Shown examples are 99.4 bp s-1 and 467.1 bp s-1. d) Length distributions of replication products, fit with single exponential decay to obtain processivity: 25.3 ± 1.7 kbp (T7), 85.3 ± 6.1 kbp (E. coli). e) Rate distributions of single molecule trajectories, fit with single Gaussians. Means: 75.9 ± 4.8 bp s-1 (T7), 535.5 ± 39 bp s-1 (E. coli).


One critical control is examining the effect of the stain, SYTOX Orange, on the replication proteins of interest. A simple way to do this is to perform the replication experiment in the flow cell as described but with the omission of SYTOX. After the reaction mixture has flowed through the chamber, add buffer with SYTOX to stain DNA and examine the length distribution of replicated molecules. Alternatively, standard bulk reactions can be used to check any effect of SYTOX on replication rate and efficiency.

The experiment described here uses only a single flow channel. This can be changed easily by creating multi-channel flow chambers or using PDMS or similar microfluidic devices. Increasing the number of channels greatly facilitates screening of protein concentrations, mutants, or inhibitor molecules and increases the rapidity of replication data collection.

As mentioned, we perform the E. coli replication experiments at 37 °C using a home-built aluminum flow cell heater, a resistive heating element (cartridge heater) and variable power supply. This gives good temperature stability and avoids the purchase of an objective heater. To calibrate the heater, we simply drilled a hole in the center of a quartz flow cell top, inserted a thermocouple into the flow channel and flowed buffer as normal. Measuring the flow cell buffer temperature at increasing voltages allows accurate heating.


Samir Hamdan aided in the development of this technique. E. coli proteins are from the lab of Prof. Nick Dixon, University of Wollongong, and T7 proteins are from Prof. Charles Richardson, Harvard Medical School. Work is supported by the National Institutes of Health (GM077248 to A.M.v.O.) and the Jane Coffin Childs Foundation (J.J.L.).


Name Company Catalog Number Comments
M13mp18 ssDNA New England Biolabs N4040
Biotinylated Tail Oligo Integrated DNA Technologies
T7 DNA Polymerase New England Biolabs M0274 Use T7 replication buffer for substrate preparation
Phenol/ Isoamyl Alcohol/ Chloroform Roche Group 03117987001 24:24:1 v/v
3-aminopropyl-triethoxysilane Sigma-Aldrich A3648 Other aminosilanes can be used or mixed with non-amine reactive silanes for sparser surfaces
Succinimidyl propionate PEG Nektar Similar PEGs can be purchased from Nanocs, CreativePEGWorks, etc.
Biotin-PEG-NHS Nektar
Double-sided tape Grace Bio-Lab Inc. SA-S-1L 100 μm thickness
Quartz slide Technical Glass 20 mm (W)x 50 mm (L)x 1mm (H) Size to fit on coverslips. Drill holes with diamond-tip drill bits (
Polyethylene tubing BD Biosciences 427416 0.76 mm ID, 1.22 OD
Other size tubing can be substituted.
Streptavidin Sigma-Aldrich S4762 Make 1 mg/mL solution, 25 μL aliquots in PBS pH 7.3
Deoxyribonucleotide triphosphate solution mix New England Biolabs N0447
Ribonucleotide triphosphate solutions Amersham 272056 272066 272076 272086
SYTOX Orange Invitrogen S11368 Other dsDNA stain can be used
Fluorescence Microscope with 60x TIRF objective Olympus Corporation IX-71 Microscope, camera, etc. can be substituted for similar equipment
Syringe Pump Harvard Apparatus 11 Plus Operate in refill mode to facilitate solution changes
532 nm laser Coherent Inc. Compass 215M-75 Select wavelength to correspond to stain of choice
EMCCD Camera Hamamatsu Corp. ImagEM
Emission filter Chroma Technology Corp. HQ600/75m

T7 Replication: 40 mM Tris pH 7.5, 50 mM potassium glutamate, 10 mM magnesium chloride, 100 μg/mL BSA, with 5 mM dithiothreitol, 600 μM dNTPs, 300 μM ATP, 300 μM CTP, and 15 nM SYTOX Orange added immediately before use. Proteins added as: 5 nM gp4 (hexamer), 40 nM polymerase (1:1 gp5: thioredoxin), 360 nm gp2.51,5.

E. coli Replication: 50 mM HEPES pH 7.9, 12 mM magnesium acetate, 80 mM potassium chloride, 100 μg/mL BSA with 10 mM dithiothreitol, 40 μM dNTPs, 200 μM rNTPs, and 15 nM SYTOX Orange added immediately before use. Proteins added as: 30 nM DnaB (hexamer), 180 nM DnaC (monomer), 30 nM αεθ, 15 nM τ2γ1δδ’χψ or τ3δδ’χψ, 30 nM β (dimer), 300 nM DnaG, 250 nM SSB (tetramer), 20 nM PriA, 40 nM PriB, 320 nM PriC, 480 nM DnaT1,6

E. coli Replication: 50 mM HEPES pH 7.9, 12 mM magnesium acetate, 80 mM potassium chloride, 100 μg/mL BSA with 10 mM dithiothreitol, 40 μM dNTPs, 200 μM rNTPs, and 15 nM SYTOX Orange added immediately before use. Proteins added as: 30 nM DnaB (hexamer), 180 nM DnaC (monomer), 30 nM αεθ, 15 nM τ2γ1δδ’χψ or τ3δδ’χψ, 30 nM β (dimer), 300 nM DnaG, 250 nM SSB (tetramer), 20 nM PriA, 40 nM PriB, 320 nM PriC, 480 nM DnaT1,6



  1. Tanner, N. A., et al. Real-time single-molecule observation of rolling-circle DNA replication. Nucleic Acids Res. 37, e27 (2009).
  2. Tabor, S., Huber, H. E., Richardson, C. C. Escherichia coli thioredoxin confers processivity on the DNA polymerase activity of the gene 5 protein of bacteriophage t7. J Biol Chem. 262, 16212-16223 (1987).
  3. Sofia, S. J., Premnath, V. V., Merrill, E. W. Poly(ethylene oxide) grafted to silicon surfaces: Grafting density and protein adsorption. Macromolecules. 31, 5059-5070 (1998).
  4. Tanner, N. A., van Oijen, A. M. Single-molecule observation of prokaryotic DNA replication. DNA replication: Methods and protocols. Vengrova, S., Dalgaard, J. Z. 521, Humana Press. New York. 397-410 (2009).
  5. Hamdan, S. M., Loparo, J. J., Takahashi, M., Richardson, C. C., van Oijen, A. M. Dynamics of DNA replication loops reveal temporal control of lagging-strand synthesis. Nature. 457, 336-339 (2009).
  6. Heller, R. C., Marians, K. J. Replication fork reactivation downstream of a blocked nascent leading strand. Nature. 439, 557-562 (2006).



  1. Nice job!

    Posted by: Gloria K.
    March 4, 2010 - 10:35 AM
  2. 1-Her voice makes me nervous.
    ²- What is the brand/manufacturer of the double sided tape?

    Posted by: Anonymous
    May 23, 2011 - 10:28 PM
  3. Double-sided tape is from Grace BioLabs, SecureSeal SA-S-1L

    Posted by: Anonymous
    May 23, 2011 - 10:38 PM
  4. Thanks a lot for the response. One last question, s the tape waterproof too?
    ps. Her voice still bothers me!

    Posted by: Anonymous
    May 24, 2011 - 5:10 PM
  5. Random questions, where did you get your air gun from?

    Posted by: Anonymous
    January 30, 2012 - 6:59 PM
  6. I honestly don't remember, and another guy in the lab had ordered them, but it's a clean room product, so maybe some clean room supplier? Sorry.

    Posted by: Anonymous
    February 9, 2012 - 9:01 AM
  7. DŒs any part of the work need to be done under a fume hood? Or everything can be done in open air.

    Posted by: Anonymous
    February 8, 2012 - 9:45 PM
  8. No, none of it requires a fume hood. If you don't like the smell of acetone you could do the glass drying/functionalization steps in a hood, but it's not necessary.

    Posted by: Anonymous
    February 9, 2012 - 9:02 AM
  9. Thanks for the answer. Really nice work!

    Posted by: Anonymous
    February 9, 2012 - 9:19 AM
  10. Did you face any problem with the laser focus drifting because of the flow? If that happens how do you handle it.

    Posted by: Anonymous
    March 2, 2012 - 5:34 PM
  11. No, the flow in the chamber is sufficiently independent of the microscope optics. The only thing that can affect focus is sudden, large pressure changes which could cause the glass to flex, and these would be only temporary. Flow dŒs affect TIRF, as molecules are stretched close to the surface and therefore in the TIRF evanescent wave, but that's the only way the optics and flow are connected.

    Posted by: Anonymous
    March 5, 2012 - 5:00 PM
  12. I am doing single molecule FRET experiments. My slide surfaces get very dirty (at single molecule level) after the silanisation step. I am using Piranha solution for surface cleaning and the surfaces are very clean just before silanisation. Is there something I should be careful about in the silanisation step, do you have a cleaning step in addition to what you have described?
    thanks and best regards

    Posted by: Tunc Y.
    June 20, 2012 - 8:20 PM
  13. Hi! I have exactly the same problem! Before the silanization everything is completely clean, without any inpurity. Do you filter your aminosilane or do you treat it somehow (e.g centrifuging before applying?)
    Thanks for your help

    Posted by: Ulf H.
    May 29, 2013 - 5:36 AM
  14. We certainly saw some surface variability, i.e. clean spots and dirty spots, but it was never really a big issue (though we're looking stained dsDNA here vs. single fluorophores). We didn't filter our silane solutions, but from the two comments it sounds like some sort of impurity is conferred in that step so perhaps some cleanup would help. We did notice that the fresher the silane, the better the surfaces, so I'd recommend the smallest bottle size and limited re-use. Sorry for not being more helpful, just never spent time dealing with that step. I'm not currently doing fluorescence experiments, though, so there may be improved methods out there.

    Posted by: Anonymous
    May 29, 2013 - 8:07 AM
  15. Thank you very much!

    Posted by: Ulf H.
    May 30, 2013 - 4:35 AM
  16. Hi, I should say first it is a very nice technique that you developed. I was wondering if I can analyze long DNA chains with this tecnique. So, my question is; does this technique have a size limitation of DNA?

    Posted by: Fabio C.
    November 20, 2015 - 11:40 PM

Post a Question / Comment / Request

You must be signed in to post a comment. Please or create an account.

Usage Statistics