Rapid cycle DNA amplification was continuously monitored by three diff
erent fluorescence techniques. Fluorescence was monitored by (i) the d
ouble-strand-specific dye SYBR(R) Green I, (ii) a decrease in fluoresc
ein quenching by rhodamine after exonuclease cleavage of a dual-labele
d hydrolysis probe and (iii) resonance energy transfer of fluorescein
to Cy5(TM) by adjacent hybridization probes. Fluorescence data acquire
d once per cycle provides rapid absolute quantification of initial tem
plate copy number. The sensitivity of SYBR Green I detection is limite
d by nonspecific product formation. Use of a single exonuclease hydrol
ysis probe or two adjacent hybridization probes offers increasing leve
ls of specificity. In contrast to fluorescence measurement once per cy
cle, continuous monitoring throughout each cycle monitors the temperat
ure dependence of fluorescence. The cumulative, irreversible signal of
hydrolysis probes can be distinguished easily from the temperature-de
pendent, reversible signal of hybridization probes. By using SYBR Gree
n I, product denaturation, annealing and extension can be followed wit
hin each cycle. Substantial product-to-product annealing occurs during
later amplification cycles, suggesting that product annealing is a ma
jor cause of the plateau effect. Continuous within-cycle monitoring al
lows rapid optimization of amplification conditions and should be part
icularly useful in developing new, standardized clinical assays.