CTAB Extraction Method to obtain DNA for Genomic Library Preparation

Written by Michael Frohlich, based on methods from the Doyle lab and on advice from Mark Hershkovitz
Elliot Meyerowitz Lab, Caltech, and Frohlich lab, University of Michigan

In Brief (numbers correspond to items in the more detailed instructions that follow):

CTAB Instructions in more detail (without the above instructions for CsCl purification or spectrophotometry):

  1. Wipe off bench top to remove potential contaminating DNA.
  2. Seize a water bath, setting it to 65°, and putting a label on requesting that it not be turned off, if you plan to extract today.
  3. Collect 0.5 to 3 g or more of plant material. I try to use 2 or 3 g if available. I weigh it and temporarily store it in the weighing boat.
  4. Wash and dry a mortar and pestle, and a white plastic scoop. The scoop is officially called a "stir rod" in the Fisher Catalog (6168-0010; p. 1682); it is by far the most effective scoop I've found.
  5. Get liquid nitrogen.
  6. Label the sides and tops of large orange cap tubes.
  7. Set up orange cap tubes in liquid nitrogen, (for transfer of samples). To hold the LN2 I use two nested styrofoam cups with a piece of plastic wrap inside the inner cup to make it liquid tight. I put the cups into a mortar, to prevent them from falling over. This is extreemly effective, as well as cheap, simple, etc. Jam two large orange cap tubes into the cups. They won't quite fit; the friction with the cup sides will prevent tubes from floating on the liquid nitrogen in the cups. Fill the inner cup with liquid nitrogen, but don't put LN2 inside the orange cap tubes. Interior of tubes will stay very cold up to the level of the top of Styrofoam cups. Put white plastic scoop in liquid nitrogen briefly, then put it in the orange cap tube that will receive ground sample. This keeps the scoop cold. The scoop is described in item 4 above.
  8. Begin grinding. Put some plant material into the mortar and add LN2 to ca 1/4 the depth of mortar. Put in pestle. Push down with pestle to break up plant, but effective grinding can't be done until the liquid has nearly evaporated, leaving a paste of plant particles and LN2 in the bottom. Grind vigorously briefly, then add more LN2 before plant thaws. Grind again. Add a little bit of glass beads only if necessary. Comment: It is difficult to grind plant material very finely. It is impossible if there is more than a fraction of a gram of material present, so grind a little bit at a time. Some plants are too tough to grind without glass beads, so, if necessary, add a few microliters (poured from container). Larger amounts of sand impede grinding, and the finest grinding can be done without using glass beads, if the plant is not too tough. During the grind I periodically use the pestle to scrape compacted powder away from the walls of mortar, so it can be ground more effectively.
  9. Use the white plastic scoop to transfer ground plant into orange cap tube. This is very effective. The plant does not thaw during transfer. I temporarily store pestle in a hole in an orange-cap-tube rack. It is cold enough that particles on pestle do not thaw during transfer. One can scrape almost all the material off the sides of the mortar, and the plastic does not leave dark marks on the mortar as metal spatulas do. Such metal powder might contaminate samples with enzyme-killing metals. After transferring put scoop into the orange cap tube with the DNA. See items 4 & 7 above for more about the white scoop. The plant powder on the upper sides of the pestle may begin thawing during grinding or transfer. If so, I avoid transfering it, and I wipe it off with a paper towel to avoid contaminating the sample with oxidized phenolic compounds.
  10. Grind another bit of plant and transfer it to the orange cap tube. Wipe scoop with paper towel if particles on it have thawed.
  11. When all material is ground, cap the tube loosely, so cap does not pop off. Tubes may be stored at -80. I chill them in LN2 when bringing them up for extraction, either in Styrofoam cups, as above, or in a larger Styrofoam container with the tubes erect in place by on of the Styrofoam racks they are sold in.
  12. Measure out 10 ml of 2x CTAB buffer (see recipes at end of this text) for each sample to be extracted. Shortly before use add 20µl of beta-mercaptoethanol (BME) for each 10 ml of solution. The BME is in the refrigerator near balances, on middle shelf, in second row of bottles, near center of shelf. I add BME in the hood, but use the solution at my bench. Close tube and mix CTAB with BME.
  13. Draw 10 ml of 2x CTAB with BME into pipette, then take tube out of LN2, and wait for contents to warm slightly. I wait until the frost on outside top half of tube begins to thaw. Remove cap, hold tube 2/3 of way toward top and put on vortex (powder gets shaken). Promptly start adding CTAB while continuing to vortex for ca. 30 seconds. Put cap back on. Comment: This seems to be very effective at separating the particles and suspending them in the liquid. Previously I put the CTAB into the mortar and "ground" it in order to suspend the particles. The CTAB would freeze, and it was a pain to get it out- I had to wait for it to thaw, and then spoon it out. The particles were not as well suspended as they are by vortexing. Presumably the DNA has not come out of the particles during this period, so it doesn't get sheared.
  14. Put in water bath at 65° for about 1 hr, mixing occasionally.
  15. Dump into a 30 ml high speed centrifuge tube and cap tube.
  16. Spin at 9000 rpm 10 min. in high speed centrifuge. Spinning faster may make tube break. Comment: This spin, before adding chloroform, greatly reduces the semi-soluble gunk that comes down in the first precipitation. It is a great improvement in the extraction method, suggested by Mark Hershkovitz. Without it one can't effectively process more than 1 g of tissue, because the first precipitate won't redissolve sufficiently. Apparently the chloroform allows something in the particulate material to dissolve which can't dissolve during the 65° incubation when no chloroform is present.
  17. Pipette (or dump) liquid into another high speed centrifuge tube. I do this at the centrifuge, to avoid stirring up the sediment while carrying it.
  18. Add 10 ml of Chloroform - isoamyl alcohol. Cap with same cap as in previous spin. Mix by gentle inversion for a few minutes.
  19. Spin at 9000 rpm in high speed centrifuge for 10 -15 min. After spin carefully pipette upper layer to a large orange cap tube. Try not to get the fine particles just above the interface, though they are easily sucked into the pipette. This should be done at the centrifuge, without carrying the tubes to the bench.
  20. Add 2/3 volume isopropanol from the freezer. Mix by inversion several times. If precipitate forms immediately, (the trapping of air bubbles under the liquid surface is the give away) then I go to the next step immediately. Otherwise, put in the refrigerator for 1/2 hr.
  21. Spin at 3500 rpm for 5 min. in bench top centrifuge. Precipitate forms a plug in bottom of orange cap tube. (Alternative for Heliotropium:: spin 500 rpm for ca 3-5 min, decant supernatant, add 70% ETOH in a forcefull stream to break up pellet.)
  22. Dump supernatant, observing pellet so it does not get lost.
  23. Wash in 70% ETOH from freezer for 30 min. to overnight, in refrigerator.
  24. Spin 3500 rpm for 5 min. in bench top centrifuge.
  25. Dump supernatant.
  26. Get pellet off bottom of tube and onto side of tube, with a Pasteur pipette, if necessary, so pellet won't fly out of tube in drying step
  27. Dry under vacuum for 1-4 minutes, until pellet begins to look shrunken, but not dry.
  28. Resuspend in 2 ml TE. Add 1 µl ribonuclease solution (get this from Leonard). Gently mix to encourage pellet to dissolve. This usually takes only a few minutes.
  29. Put at 37°, in bacteria incubator or a water bath for ca 30 minutes.
  30. Dump into 30 ml high speed centrifuge tube. I rinse orange cap tube with 1/2 ml TE.
  31. Reprecipitate with 1/2 volume 7.5M ammonium acetate and 2.5x total volume frozen ETOH. Precipitate forms immediately, but I put in refrigerator for 1/2-1 hr.
  32. Spin 9000 rpm for 10-15 min.
  33. Dump supernatant, checking that pellet doesn't come out. Sometimes a part of the pellet won't sink, because of trapped air bubbles. I use the edge of the cap held against the tube opening to keep any part of the pellet from escaping.
  34. Wash pellet in 5 ml 70% ETOH from freezer for 30 min. to overnight.
  35. Spin 9000 for 10 min.
  36. Dump supernatant, checking that pellet doesn't come out.
  37. Get pellet on to side of tube, and dry in vacuum for 2-5 minutes, until pellet is shrunken but not dry.
  38. Resuspend in 200-300µl TE. It may dissolve slowly.

The CsCl spin does not need a more detailed description.

Spectrophotometry instructions in more detail:

I think a common reason for incorrect quantitation of DNA is failure of the DNA to dissolve completely. I have seen trapped air bubbles remain under the top of the solution for a very long time. Sometimes I've detected this by noticing a group of air bubbles that move in concert as the tube is inverted.

39) Quartz cuvettes are stored under water in a jar. Remove it with forceps and dry with Kimwipe. I put the quartz cuvette in the spec, and I don't move or remove it until all measurements are finished. That means that:

40) one must wash out the previous samples and blanks without moving the cuvette, which is, in fact, the critical step.

41) To wash it I use a jar of ddH2O, two 1 ml pipettors (that's important!), each with a blue tip, a single 1.5 ml eppindorf microfuge tube, and a pasteur pipette with a long narrow pipettor tip stuck on it. The long narrow tip is the special kind used to load sequencing gels- it is so narrow you can put it between the two glass plates. This tip is expensive and uncommon, but many labs have some around. Otherwise one could just use the pasteur pipette, but this tip works much better. Jam the tip on the end of the pasteur pipette and it will stick quite well. Occassionally it may fall off, so I then get another tip. To make measurements I use a 2 microliter pipettor to take aliquots of the DNA solution to be quantitated. NOTE that I use the same two blue tips, the same eppindorf and the same pasteur pipette with (I hope) the same fine tip for the whole session. If you must replace any of these you have to start over with stabilization to be sure you get reproducible results. (I use a fresh tip for the 2 µl pipettor for each sample; these tips seem to lack UV absorbing contaminants.)

42) First I stabilize and wash the tips and tubes. There is UV absorbing gunk on most of these materials, which gradually washes off and distorts measurement. I take water from the jar (1 ml to start, when I'm cleaning everything), put it in the eppindorf, then use the other pipettor to take it out of the eppindorf and into the cuvette. I do a spectrum (I do about 230 to 300 nm). Then I suck the liquid out with the pasteur pipette, and get another ml of water as described above and put it in the cuvette and do another spectrum, and another, and another. Gradually the absorbance will go down, as the gunk in the tubes and tips washes out. If it looks like the absorbance is stabilizing I do a reference, and then continue to read absorbances of additional 1 ml samples. If the gunk is really washed out, and things really are stable, then the extreme range of coordinates on the absorbance graph will have a decimal point and three zeros before the first significant figure (a "9"). (Our machine expands the scale to show detail on the absorbance curve.) The curve then represents only the noise of measurement inherent in the machine. I try to get this sort of graph several times (maybe 6 in a row, if I'm being especially careful) before I do a real measurement. In the latter of these I switch to a lower amount of liquid (0.45 ml in our machine), such that the beam is just covered by the liquid in the cuvette.

43) I then continue putting water in the cuvette as above, except I sometimes put 1 or 1/2 microliter of DNA solution into the water as it sits in the eppindorf, before I pick it up with the second pipettor to put it in the cuvette. It gets well mixed by the sucking and squirting of the second pipettor tip. when I do the measurement I see the spectrum (I hope) of DNA, and do a numerical reading at 260 nm. It is important to look at the spectrum to be sure you're measuring DNA and not gunk.

44) Then I wash the tube twice with water, as described above, (still using the same tips and same eppindorf tube), taking measurements with each filling of the cuvette, to observe the DNA being diluted to insignificance. Usually after two washings I can put in another real DNA sample for measurement. As needed, (often because the instrument drifts) I re-zero, but then I check stability a couple of times before making a measurement. If in any doubt it is very wise to measure each sample twice. You can get a clump of DNA gel when you suck out a 1 microliter sample, but won't happen exactly the same way both times, so you'll get different values.

In this way I can get reliable measurements at .01 absorbance units, and pretty good measurements down to absorbance .004. To get the concentration of DNA, I use 420 microliters of water to transfer; I add 1 microliter of DNA solution. I take the absorbance at 260, multiply by 1000, then multiply by 23.1 and the result is the concentration of DNA in the sample from which the alliquot came in nanograms per microliter. This is for long double stranded DNA, and assumes 10% of the volume measure is solution from the previous filling of the cuvette. If the sample is too dilute to get a reasonable measurement I use 2 or more microliters and do it again. In this way you can quantify DNA without risking contamination or loss by putting lots of it in the cuvette. I never use DNA that has been in the cuvette. Even 0.1% contamination is unacceptible for my work; as the UV cuvettes aren't ' disposible you can never be sure about them.

This method takes persistence and time. Occasionally you get a new blob of gunk from somewhere, and you have to wash it out and verify stability before continuing with measurements. The gunk has its absorbance peak at the shortest wavelengths, so it is readily distinguishable from DNA. All large tips, and/or eppindorfs, and especially the Pasteur pipetter have lots of UV absorbing gunk on them. I have never been able to clean a batch of any of these sufficiently to avoid the laborious washing at the spectrophotometer, so I just use ordinary tips and eppindorf and wash as described above.

A smaller-volume cuvette can allow a measurement of even smaller amounts of DNA, but in many spectrophotometers the smallest cuvettes block much of the beam, limiting the accuracy of measurements made with them.


Concentration Ingredient For 200 ml:


  • CTAB is Hexadecyltrimethylammonium bromide. Dissolve it before adding salt, with stirring and a little warmth, if necessary.
  • When the salt is dissolved lots of tiny bubbles come out of solution; they rise to the surface very slowly, simulating undissolved material.
  • Turn off the stirrer 5 minutes after adding the salt, to see if salt is dissolved and only bubbles are present.
  • PVP of 40,000 average molecular weight makes the solution slightly translucent, but no large particles should be present after dissolving
  • Beta-mercaptoethanol should be kept in the refrigerator in a dry box.