Lipidsof Human Urine, withCharacterizationof a NativeEmulsifer

In 1954 a project was started to systematically investigate the nondialyzable substances in human urine. Four reports1 have been presented describing the procedure for removing the insoluble material from urine preserved with toluene (10 ml/specimen) by high-speed continuous-flow centrifugation, after which a soft, buttery layer called “floating lipid fraction” (FF)2 by Smith (1) and Winkler (2) was skimmed off the top of all normal specimens; the yields varied from 10 to 20 mg of dry material per 24 h. Urine from diabetic patients, also preserved with 10 ml of toluene per 24-h specimen, contained little or none of this material, suggesting that this fraction might be clinically significant. However, Winkler similarly observed very little FF when only 1 to 3 ml of toluene was used as preservative for normal 24-h specimens, while none was found in urine preserved with phenylmercuric nitrate. In addition, the FF he obtained contained nearly the same proportion of urea as the urine itself. This suggested that the FF was emulsified toluene and urine, and the smaller the volume of toluene or the larger the volume of urine, the less FF obtained. When this FF was centrifuged at about 30,000 g for about 30 mm to break the emul-