NOTE Canino and Caldarone: Fluorometric techniques for determining nucleic acid contents of fish larvae 



159 



ized, distilled water in an Ultra-Turrax homogenizer 

 (Tekmar Co.) with three 15-second pulses at maximum 

 power (Caldarone and Buckley, 1991). Previously fro- 

 zen walleye pollock larvae were homogenized in 

 prechilled glass homogenizers. Six aliquots of homo- 

 genate were pipetted into 1.5-mL polyethylene vials 

 and immediately frozen at -80°C until analysis of three 

 aliquots by FIA and three aliquots by CFA. 



Sample extraction procedures followed those out- 

 lined by Caldarone and Buckley ( 1991). Homogenates 

 were extracted in 1% sarcosine (N-lauroylsarcosine), 

 TRIS-EDTA (5.0 mM TRIS-HC1, 0.5 mM EDTA, pH 

 7.5) buffer for 30 minutes at room temperature, 

 mixed vigorously in a sample vortexer for 10—15 sec- 

 onds, then extracted for another 30 minutes. Aliquots 

 were diluted with TRIS-EDTA buffer to yield a final 

 concentration of 0.1% sarcosine, then centrifuged at 

 room temperature for 5 minutes at 2,500 times grav- 

 ity (x g) for CFA and 12,000 x g for FIA. The super- 

 natant was recovered for estimation of nucleic acid 

 concentrations by FIA or CFA. Sarcosine, like other 

 anionic detergents, fluoresces at excitation and emis- 

 sion wavelengths used in the analyses. To reduce this 

 effect, samples extracted in 1% sarcosine required a 

 10-fold dilution with TRIS-EDTA buffer before FIA 

 (Caldarone and Buckley, 1991). In the modified CFA 

 procedure, a final sample concentration of 0.0125% 

 sarcosine produced an acceptable background fluo- 

 rescence (blank) value of approximately 5% of the 

 highest nucleic acid standard. 



Differences in the total sample volumes required 

 by FIA and CFA procedures required modifications 

 to the concentrations of nucleic acid standards, fluo- 

 rochrome reagents, and sample volumes of fish 

 homogenates. Nucleic acid standard ranges and 

 homogenate sample volumes were chosen to encom- 

 pass the typical range of concentrations encountered 

 by each method during routine analysis of individal 



fish larvae. Working standards of RNA and DNA 

 were prepared as described by Caldarone and 

 Buckley (1991) by serial dilution of previously fro- 

 zen stock solutions with 0.1% sarcosine in TRIS- 

 EDTA buffer. RNA standards (Sigma Chemical Co., 

 St. Louis, MO, Type IV, calf liver) ranged from 1.97 

 to 17.68 /ig-mL" 1 for FIA and from 0.41 to 13.14 

 /ig-mLr 1 for CFA. DNA standards (Boehringer-Mann- 

 heim Corp., Indianapolis, IN, high molecular weight, 

 calf thymus) ranged from 0.16 to 1.52 /ig-mLr 1 for 

 FIA and from 0.07 to 2.37 jig-mL" 1 for CFA. Estimates 

 of contamination of the calf liver RNA standard by 

 DNA, determined by fluoresecence in Hoechst, was 

 less than 1% by weight. A 100-//L "spike" of homogenate 

 from walleye pollock larvae was added to serial dilu- 

 tions of RNA and DNA standards in order to estimate 

 the recovery efficiencies using the CFA protocol. 



Fluorochrome working reagents of EB (137.5 

 ng-mLr 1 ) and Hoechst (25 ng-mLr 1 ) prepared for FIA 

 in TRIS-EDTA buffer according to Caldarone and 

 Buckley ( 1991) were modified by reducing the sodium 

 chloride (NaCl) concentration in the Hoechst reagent 

 from 0.2 N to 0. 1 N and the pH from 7.5 to 7.0. Work- 

 ing dye solutions of 2.0 pg-mL" 1 EB and 5.0 /ig-mLr 1 

 Hoechst prepared for CFA at the same pH and NaCl 

 concentration as for FIA provided a sensitive linear 

 response to the standards while maintaining a low 

 background fluorescence. 



The RNA and DNA concentrations were estimated 

 for each aliquot. In addition, the amount of endog- 

 enous fluorescence (sample fluorescence in the ab- 

 sence of fluorochrome dye) was determined for most 

 homogenate samples (21 for FIA, 17 for CFA) by sub- 

 stituting an equal volume of TRIS-EDTA buffer for the 

 fluorochrome working reagent in the assay procedure. 



The FIA system for nucleic acid determination is 

 fully described by Caldarone and Buckley (1991). A 

 50-/iL sample is injected into a reagent stream con- 



