Rolling circle amplification has been useful for detecting point mutations in isolated nucleic acids, but its application in cytological preparations has been problematic. By pretreating cells with a combination of restriction enzymes and exonucleases, we demonstrate that rolling circle amplification in situ can detect gene copy number and single base mutations in fixed cells with efficiencies up to 90%. It can also detect and quantify transcribed RNA in individual cells, making it a versatile tool for cell-based assays. R olling circle amplification (RCA) is a molecular cytogenetic technique used with a padlock oligonucleotide probe to detect single base changes in isolated nucleic acids (1-5). Padlock probes are composed of Ϸ100 nucleotides that hybridize to targets of Ϸ30 bases. The 30-base target-binding region of the probe is split into two 15-base segments placed in opposite orientation at each end of the linear probe so that a circle must be formed for hybridization to occur (6, 7). At 10 bases per helical turn, the hybridized probe wraps around its target three times, and the remaining 70 bases form an unhybridized singlestranded loop. Posthybridization DNA ligation connects the two ends of the probe in the middle of the 30-base binding region. The unbound 70-base loop facilitates probe circularization and permits Ϸ20 bases to serve as a primer recognition site for DNA polymerase to replicate the circle. RCA is an isothermal process in which the polymerase progresses continuously around the loop until the 100 bases have been replicated hundreds or thousands of times. Incorporating a labeled nucleotide during the RCA reaction produces sufficient signal for easy visualization of the target.Application of RCA to in situ targets in fixed or permeabilized cells has not been uniformly successful to date. Whereas recent work has demonstrated that the concept is viable (8), DNA detection efficiencies of 20-30% lessen the utility of RCA as an assay. Lack of success has been attributed to possible blocking of the polymerase by the target strand, and it was suggested that this problem might be overcome by cutting the target DNA strand near the RCA probe's hybridization site (5). Under these conditions, DNA polymerase could free the probe from the target, in effect spinning the probe away from the target, keeping the polymerase from being blocked during the amplification process. Here, we report that in addition to restriction enzyme digestion of DNA, additional steps were required to achieve consistent and satisfactory results for RCA in situ. Whereas heat denaturation is typically used to render the target DNA single stranded, we found that complete removal of the nontargeted DNA strand by digestion with exonuclease III significantly increased the efficiency of the process.We also demonstrate the use of RCA to detect mRNA in cytological preparations. Using appropriate image analysis techniques, the RCA assay is sufficiently quantitative to enable transcriptionally mediated dose-response curves to be generated. Inc...