Difficulties with the purification of high-molecular-weight genomic DNA has hampered progress in identifying restriction fragment length polymorphisms, developing genomic maps, and constructing genomic libraries for cotton. We have developed a protocol for purifying genomic DNA that has several advantages: first, the method is reproducible; second, the purified DNA is larger than 50 kb; third, restriction endonucleases digest this DNA; and fourth, this procedure can be used to isolate DNA from a large number of samples, as required by genetic studies.

DNA was purified from cotyledons and leaves of the Gossypium hirsutum (L.) cultivar 'DPL 50'. Cotyledons were harvested from either three day old seedlings grown in darkness, or 12 day old light-grown seedlings. Leaves were harvested 3 to 5 nodes below the shoot apex of field grown plants. All plant material was frozen in liquid N₂ and ground into a fine powder before addition of grinding buffer. Grinding buffer (5 ml/g tissue) consisted of 5X SSPE, 2.5% (w/v) SDS, 1 mM EDTA and various concentrations of urea, as determined by the tissue used for extraction. For the isolation of DNA from cotyledons, all concentrations of urea greater than or equal to 1 M had similar effects. For leaf material, 9 M urea gave maximum yields. The data for cotyledons indicates that urea is acting as a chaotropic agent or protectant, and not solely as a protein denaturant. Homogenates were extracted with 1.25 volumes of 2:1 chloroform:hexane solution for 30 min at 24 C. The quality of DNA from cotyledons was improved by organic extraction of homogenates; the mixture of chloroform and hexane gave superior results. Purification was completed by the standard protocol for ultracentrifugation in CsCl gradients.

We have compared DNA purified by our method and that prepared by the methods of Walbot and Dure (J. Mol. Biol. 101: 503-536) and Baker et al. (Bio Techniques 9: 268-272). Cotyledonary DNA extracted by our method and that of Baker et al. are comparable in quantity and molecular weight; the method of Walbot and Dure purifies cotyledonary DNA of smaller molecular weight. However, our procedure handles eight-fold more samples in half the time than the method of Baker et al. Neither our procedure, nor the two cited, purified genomic DNA of superior quality from leaves: the leaf DNA isolated by our procedure was of lower molecular weight than the DNA from cotyledons, the method of Walbot and Dure failed to isolate leaf DNA, and the DNA isolated by Baker et al. is coffee colored. We are currently making modifications in our procedure to improve its performance in PCR.