AIM OF THE STUDY: Smooth muscle cells build up the normal media and stabilize atherosclerotic lesions whereas an inflammatory component is determinant for unstable angina. Smooth muscle cells, currently identified by alpha-actin, present a phenotypic heterogeneity and alpha-actin can be reduced in pathology. We tried to characterize vascular cell types, particularly smooth muscle cells, and coronary atherosclerotic tissues, by random genes expression fingerprints. MATERIALS AND METHODS: Expression fingerprints (cDNA electrophoresis) were performed by differential display reverse transcriptase-polymerase chain reaction. Variability of fingerprints was studied for a panel of arterial muscle cell phenotypes and comparisons were made with fingerprints from other cell types (endothelial cells and macrophages). The technique was then applied to human coronary atherectomy samples compared to control human arterial (mammary) smooth muscle. RESULTS: Arterial smooth muscle cells fingerprints were overall similar whatever the cell phenotype (native contractile, dedifferentiated in culture or epithelioid). Moreover, with two primer pairs, the muscular fingerprints markedly differed from the endothelial and the monocytic fingerprints. Application of differential display to coronary atherectomy samples was feasible. Interestingly, the pathological tissues exhibited either smooth muscle-like or smooth muscle-divergent fingerprints. CONCLUSIONS: Smooth muscle cells and inflammatory cells exhibited distinct differential display fingerprint patterns. Thus, a simple expression profile of arbitrary genes provides a molecular bar code tool (pattern signature) useful to characterize vascular cell cultures or tissues. The present work proposes a method to analyze coronary atherectomy samples which estimates their whole quality, muscular versus non muscular (inflammatory), this is of interest for clinical research.