160 



Fishery Bulletin 93(1). 1995 



taining one of the fluorochrome dyes. The injected 

 sample is mixed and transported with the reagent 

 to the fluorescence detector which continuously 

 records the fluorescence at 525 nm excitation and 

 600 nm emission for EB, or 356 nm excitation and 

 458 emission for Hoechst. The sample fluorescence 

 is displayed as a peak, whose area is proportional to 

 the concentration (Caldarone and Buckley, 1991). 

 Modification of this procedure for CFA required a 

 minimal total volume (sample plus fluorochrome re- 

 agent) of 2 mL in order to be measured accurately by 

 the fluorometer (Shimadzu RF-540 spectro- 

 fluorophotometer, Shimadzu Corp., Kyoto, Japan), 

 which was adapted to use 12x75 mm borosilicate 

 glass test tubes as cuvettes. For all samples, except 

 the homogenates of larval pollock, a 0.1-mL aliquot 

 of extracted sample was combined with 0.9 mL of 

 TRIS-EDTA buffer and 1.0 mL of fluorochrome work- 

 ing reagent (EB or Hoechst). For larval pollock 

 homogenates, a 0.03-mL aliquot was combined with 

 0.97 mL of TRIS-EDTA buffer prior to addition of 

 1.0 mL of the fluorochrome reagents. The sample- 

 fluorochrome mixture was incubated at room tem- 

 perature for 15-30 minutes before fluorescence was 

 measured at the same excitation and emission wave- 

 lengths as in the FIA procedure. Initial trials indi- 

 cated that maximum sample fluorescence was ob- 

 tained within 15 minutes and was stable for more 

 than 4 hours at room temperature. 



Calculations of nucleic acid concentrations were 

 identical for both methods. First, endogenous sample 

 fluorescence was subtracted from total sample EB 

 or Hoechst dye fluorescence. Sample DNA concen- 

 trations were estimated directly from fluorescence 

 in Hoechst dye by using a DNA-Hoechst standard 

 curve. The computed DNA concentration was used 

 to estimate the fluorescence contribution by DNA to 

 the total sample EB-fluorescence by using a DNA- 

 EB standard curve. Fluorescence due to DNA-EB was 

 subtracted from the total sample fluorescence and 

 the remaining fluorescence was assumed to be due 

 to RNA. The RNA concentration was then estimated 

 by using an RNA-EB standard curve. 



The relationships between mean RNA and DNA con- 

 centration and RNA/DNA ratios of the fish homo- 

 genates predicted by FIA and CFA methods were ana- 

 lyzed by using a geometric mean regression procedure 

 (Ricker, 1984) that describes the linear central trend 

 between two independent estimates of the variate. 



Results 



Standard calibration curves indicated that detection 

 limits for CFA are about 0.07 /ig-mL -1 for RNA and 



0.03 /igmL -1 for DNA, similar to the values detect- 

 able with automated FIA (Caldarone and Buckley, 

 1991). The precision of both methods was comparable; 

 mean coefficients of variation, V (standard devia- 

 tion as a percentage of the mean), for triplicate de- 

 terminations from each homogenate averaged 5 to 

 7% for RNA and 3 to 4% for DNA over a broad range of 

 estimated sample concentrations. Recovery of DNA 

 standards from six replicate "spikes" of larval pollock 

 homogenate with the CFA method averaged 99.5 + 0.9% 

 in Hoechst and 99.2 ± 2.9% in EB, and recovery of 

 "spiked" RNA standards averaged 94.8 ± 6.0% in EB. 



Mean nucleic acid concentrations and RNA/DNA 

 ratios offish homogenates were generally lower when 

 estimated by CFA relative to FIA (Fig. 1). RNA con- 

 centration was most strongly correlated between the 

 two methods and DNA concentration less so (Table 

 2). The ratio of RNA to DNA was only moderately 

 correlated between the two methods and provided 

 the poorest basis for comparison. Intermethodological 

 calibration between FIA and CFA results was 

 achieved by the application of regression coefficients 

 (Table 2) to mean nucleic acid concentrations and 

 RNA/DNA ratios estimated by CFA (Fig. 2). 



Homogenates prepared from larval stages of the 

 gadid species (Gadus morhua, Melanogrammus 

 aeglefinus, and Theragra chalcogramma), tautog, 

 Tautoga onitis, and winter flounder, Pleuronectes 

 americanus, exhibited negligible endogenous fluores- 

 cence, regardless of fluorochrome, over a 3- to 4-fold 

 range of nucleic acid concentrations (Table 3). En- 

 dogenous fluorescence was highest for juvenile in- 

 land silverside and juvenile winter flounder. 



Discussion 



Modification of the FIA method described by 

 Caldarone and Buckley (1991) to conventional fluo- 

 rometry produced an assay protocol with comparable 



Table 2 



Geometric mean functional regression coefficients describ- 

 ing mean RNA and DNA concentrations (/jgmL 1 

 homogenate I and RNA/DNA ratios of 24 fish homogenates 

 determined by conventional fluourometric analysis (CFA) 

 regressed upon estimates obtained by flow injection analy- 

 sis (FIA). 



Variate 



Y-intercept 



Slope 



RNA 

 DNA 

 RNA/DNA 



-5.318 

 -1.795 

 -0.309 



0.652 

 0.991 

 0.733 



0.972 

 0.808 

 0.569 



