Comments on the spectrophotometric procedure for angiotensin-converting enzyme.

I have been using the spectrophoto-metric assay of Cushman and Cheung (1) as modified by Lieberman (2) to assay for serum angiotensin-converting enzyme (ACE) in diagnosis of possible sarcoidosis. However, replicate determinations produced results that varied considerably, creating some doubt about its analytical usefulness. Taylor and Freeman (3) identified a cause of this variability as cloudiness in the cuvet, from extracted lipid material , which develops when the extracted residue is reconstituted with 1 mollL sodium chloride. They eliminated the cloudiness by substituting aceto-mtrile for saline solutions and further improved the method by doubling the extraction volume of ethyl acetate to 3.0 mL, which eliminated troublesome emulsions. For the past two years, I have been reconstituting the extracts with " spec-tro-grade " methanol. No cloudiness develops and the odor is not as objectionable as that of acetonitrile. Ethanol will work equally well but is not available to commercial laboratories in a spectro.-grade quality. I determined the molar absorptivity of hippuric acid in methanol at 228 nm to be 10.3 L mmol' cm' compared with 9.8 L ' mmoY' cm1 reported by Cush-man and Cheung for aqueous solutions (1). Cushman and Cheung also determined that 91% of hippuric acid was extracted with 1.5 mL of ethyl acetate. Given that the solvent volume was not doubled in relation to the aqueous phase, I rechecked the amount of hip-puric acid extracted from standards containing 1.0-4.0 mmol/L (16.7-66.7 UIL). Analytical recovery of hippuric acid was 89% when 1.5 mL of ethyl acetate was used, 99% for 3.0 mL. Therefore, the factor of 0.91 in Lieber-man's formula can be eliminated, changing his final factor from 56.1 to 50.8 to calculate the enzyme activity at 1-h incubation, or to 101.6 for 30-mm incubation. The standards are linear over the range listed and can also be used to calculate ACE units. ACE units are defined as nanomoles of hip-puric acid formed per minute per milliliter of serum (2). To compare results with those by Lieberman's method, I tested 12 samples in duplicate by both methods. To assess cloudiness in the cuvets, I used the " nephlo-scale " of a Perkin-Elmer Amylase/Lipase analyzer to measure the cloudiness before taking the spec-trophotometer readings. Both saline and methanol solutions gave a reading of 6 to 7 " nephios, " as did all of the blanks and samples that were reconstituted with methanol. Samples reconstituted with saline, however, produced " nephlo …