Major interference from leukocytes in reverse transcription PCR identifiedas neurotoxin ribonuclease from eosinophils: Detection of residual chronicmyelogenous leukemia from cell lysates by use of an eosinophil-depleted cell preparation

Background: The extraction of RNA from leukocytes for reverse transcription -PCR (RT-PCR) is time-consuming and contributes to variation in analysis of the Philadelphia (Phl) chromosome of chronic myelogenous leukemia (CML) by RT-PCR. To detect residual CML after bone marrow transplantation, mRNA fro m at least 10(5) leukocytes should be analyzed, but the RNase activity of t he cells precludes simple leukocytes lysis as an alternative to RNA extract ion. We sought to identify the main source of RNase activity of leukocytes. Methods: We used a three-step chromatographic process and amino acid sequen ce analysis. We selected eosinophil-free granulocytes by using a biotinylat ed CD16 antibody and Selected mononuclear cells by fractionating the leukoc ytes with a Ficoll-Paque(R) density gradient. Results: Chromatography and amino acid sequencing identified eosinophil-der ived neurotoxin (EDN) as the main source of leukocyte RNase. Depletion of e osinophils reduced the EDN content of cell lysates by similar to 90%, allow ing a signal from a lysate of 50 K562 Ph-1-positive cells mixed with 10(5) CD16(+) granulocytes that was equivalent to 77% of the signal in the absenc e of leukocytes. A similar lysate with mononuclear cells gave a signal equi valent to 53% of that without mononuclear cells. RNA extraction gave a sign al equivalent to only 24% of the leukocyte-free control. Conclusion: The depletion of eosinophils during the preparation of leukocyt e samples for-RT-PCR efficiently reduces the risk of mRNA degradation by ri bonucleases, enabling RT-PCR analysis directly from cell lysates with a bet ter signal than can be obtained by RNA extraction. (C) 1999 American Associ ation for Clinical Chemistry.