NOTE Canmo and Caldarone. Fluorometnc techniques for determining nucleic acid contents of fish larvae 



163 



likely. The consistently lower estimates of nucleic 

 acids with the CFA method, relative to the FIA 

 method, implies a lower ratio of fluorescence yield of 

 samples to standards. One possibility is that the ef- 

 fective sample concentration at the detector may be 

 greater in CFA compared with FIA. In FIA, the time 

 between the mixing of dye and sample is precisely 

 controlled but the ratio of dye to sample is unknown. 

 The concentrations of the fluorochrome working re- 

 agents used in FIA were increased ( 14x and lOOx for 

 EB and Hoechst, respectively) in order to saturate 

 the standards for the CFA modification. A higher ef- 

 fective sample concentration at the detector (possi- 

 bly making the samples less available to the dye), 

 coupled with a longer pathlength, may explain the 

 lower fluorescence yield of the CFA samples relative 

 to the standards. The higher correlation between the 

 methods for RNA estimates than for DNA was an 

 unexpected result; RNA content is calculated indi- 

 rectly (total sample fluorescence minus estimated 

 fluorescence due to DNA) and, presumably, is more 

 subject to measurement error than is direct fluoro- 

 metric determination of DNA. 



Fluorescence by compounds other than nucleic ac- 

 ids represents a potential source of interference in 

 any fluorometric assay. The level of endogenous 

 sample fluorescence should be determined by pre- 

 liminary analyses when a new fish species or devel- 

 opmental stage is being investigated. Crude tissue 

 homogenates exhibit fluorescence characteristics not 

 found in commercial preparations of nucleic acids 

 that appear to be related to tissue type and develop- 

 mental stage of the fish. Caldarone and Buckley 



(1991) found that winter flounder and American sand 

 lance, Ammodytes americanus, larvae and the nucleic 

 acids used as standards exhibited negligible amounts 

 of endogenous or residual fluorescence, whereas win- 

 ter flounder postlarvae and juvenile muscle and liver 

 had levels ranging from approximately 5 to 55% of 

 total fluorescence in EB or Hoechst dyes. Clemmesen 

 ( 1993) reported endogenous fluorescence values rang- 

 ing up to 40% in Hoechst determinations of the DNA 

 content of herring, Clupea harengus, larvae. In this 

 study, only homogenates of the juvenile inland sil- 

 verside and juvenile winter flounder exhibited high 

 levels of endogenous fluorescence in Hoechst dye or 

 EB (Table 3), providing further evidence of an onto- 

 genetic effect. Gadid larvae, of approximately the 

 same age as the juvenile winter flounder, displayed 

 negligible amounts of endogenous fluorescence, sug- 

 gesting that systematic differences may also exist. 



Interference in nucleic acid estimation from con- 

 taminants of crude homogenate preparations has 

 been reported previously (Brunk et al., 1979; Mordy 

 and Carlson, 1991; Clemmesen, 1993). McGurk and 

 Kusser ( 1992) compared three fluorescence methods 

 for quantitating nucleic acids in Pacific herring, 

 Clupea pallasi, larvae. Of the three, the method in- 

 corporating the most extensive purification steps 

 (Clemmesen, 1988) resulted in higher estimates of 

 RNA content and RNA/DNA ratios than the other 

 two methods, suggesting that higher yields may have 

 resulted from the elimination of substances interfer- 

 ing with accurate fluorometric quantitation. How- 

 ever, quenching of nucleic acid fluorescence by con- 

 taminants in the samples does not appear to be sig- 



