Document DDbzr7gKJ8x96n0BrqEVxLzJQ

gea-lhpid chromatography. In the fish removed within 24 hr of tr# aln.ont; endothaU concentration in the flesh ranged from 0.02 to 0:1 ppro. After 30-min exposuro to water containing 2 ppm of |,4C}rndthal), ,4C-Ial>eled residues (expressed as endothall) ranging from 0.02 to 0.04 ppm were detected in the blood. After 60 min, the concentration of >4C in the blood increased up to 0.0ft ppm. Longer exposure up to 4 hr did not significantly increase the level of radioactivity in the blood. The concentration of 14C in the various tissues, 48 hr after the fish were fed {,4C]endolhall through the digestive tract, it shown in Table III. The results show that the her bicide !a absorbed by the intestinal tract though the fish were observed to eliminate 73% of the administered herbi cide during this period. In the fish fed t4C-labeled herbi cide through the digestive tract, lika the bath-expoaed fish, the concentration of radioactivity was highest in the viscera and lowest in the flesh. The pattern of distribution of ra dioactivity in the various tissues in the feed-exposed fish was similar to that observed for the bath-exposed fish, with the exception that the scales had a relatively lower propor tion of the total radioactivity. These findings have demonstrated that bluegilis absorb endothaU directly from water as well as through the intesti nal tract. Adsorption of endothaU on the scales also con tributes to the herbicide residues in the fish. This is sup MM-nni'ijam cnbcmvoiriuA i 1 A0U1.IOIVI ported by the observation that the proportion of ,4C in the scales decreased when the fish were fed endothaU through the digestive tract. Thin-layer chromatographic analysis of the methanol ex tract from the fish treated with the herbicide for 48 hr showed that all the 14C in the alcohol extractable fraction was present in the form of unchanged endothaU (Rf 0.48 and 0.75 on silica gel and cellulose plates, respectively). The UC in the extract co-chromatographed with authentic [uC|endothall. In contrast to aquatic microorganisms which were found to readily degrade endothaU (Sikka and Saxena, 1973), bluegilis do not appear to be capable of me tabolizing the herbicide. . ACKNOWLEDGMENT [14C]EndothaIl was furnished by the Pennwalt Corpora tion. LITERATURE CITED Sanborn, J. R., EPA Report No. 660-13-74-025,1974. Schultz, D. P., J. Agric. Food Chem. 21,166 (1973). Sikka. H. C-, and Rice, C. P., J. Agric. Food Chem. 21, 642 (1973). Sikka, H. C., and Saxena, J.,J. Agric. Food Chem. 21,402 (1973). Welker, C. R., Weeds 11,226 (1963). Received for review October 3,1974. Accepted April 29,1975. Thie investigation was eupported by a contract from the Office of the Chief of Engineers end by a grant from Pennwalt Corporation. The Mechanism of Chlorobiphenyl Metabolism Stephen Safe,* Otto Hutxinger, and Dan Jones 4-Chlorobiphenyl was administered to rabbits and the major urinary metabolites were identified as 4'-chloro-4*biphenylol and 4'-chloro*3,4-biphenyl- diol. It was also shown that 4'-chloro-4-biphenylol was also converted into the diol as well as its mo nomethyl derivatives. The biohydroxylation path way was further investigated using 4'-|2H)-4-chIorobipheny! as a substrate. The 4'-chloro-4-biphenyiol metabolite retained 79% of the deuterium and the mulls are consistent with the intermediacy of an arena oxide in the first hydroxylation reaction. The diol metabolite retained ca. one-half the deu terium found in the phenol (39%) and it is there fore not formed directly from the arene oxide but by direct hydroxylation of the phenolic metabo lite. The sequence of hydroxylation of chlorobi phenyl is, therefore, analogous to the stepwise hy droxylation of phenylalanine to give 3,4-dihydroxyphenylalanine and butamoxane to give 6.7-dihydroxybutamoxane. Polychlorinated biphenyls (PCB) are industrial com pound* which are now recognized as among the most wide spread pollutants in the global ecosystem (Hutsinger et a)., 1974b; Risebrough et al., 1968; Pishbein, 1973). Residues hnve bean identified in both the terrestrial and equatic en vironments (Koeman et al., 1969; Harvey et al., 1974; Jenten et al., 1969; Horn et el., 1974) as well as in animals (Baglay et al., 1970; Butler, 1973; Presit et al., 1970) and humans (Biros et al., 1970; Jamieson et al., 1973). Recent work has shown that both commercial PCB mixtures and pure isomeric chlorobiphenyls are metabolized by animals (Hutxinger et al., 1972, 1974a; Safe et al., 1974, 1975a,b; Burse et al., 1974; Greb et al., 1973; Yoehimura and Yamomoto, 1973; Block and Cornish, 1959; Goto et al., 1974; Department of Chemistry (S.S.) and Department of Bio medical Sciences (D.J.), University of Guelph, Guelph, On tario, Canada, and the Department of Environmental Chemistry, University of Amsterdam, Nieuwe Achtergrachl 166, Amsterdam, The Netherlands (O.H.). Gardner et al., 1973), plants (Moza et aL, 1973,1974), and microorganisms (Wailnofer et al., 1973; Ahmed and Focht, 1973; Maas et al., 1975) to give a range of hydroxylated me tabolites. The major rabbit urinary metabolites of 2,2',5,6'-tetrachIorobiphenyl were 2,2',5,6'-tetracMoro-3- biphenylol, 2,2',5,5'-tetrachloro-4-bipheny!ol, and Irons- 3,4-dihydro-2,2',5,6'-tetrachloro3,4-biphenyldio] hnd these all can conceivably be formed from the 3,4-epoxy interme diate. NIH rearrangement (Jerina, 1974; Jerina and Daly, 1974; Daly et al., 1972) of this intermediate (i.e., lrone-3,4- dihydro-2,2',5,5'-tetrechloro-3,4-epoxybiphenyl) could yield the two phenolic metabolites and hydrolysis of the epoxide would yield the dihydrodiol product. The experi ments described in this paper are concerned with the mechanism of PCB hydroxylation in the rabbit using 4- chlorobiphenyl, 4'-chloro-4-biphenyIol, and 4'-J2H|-4-chlorokiphenyl as model substrates. MATERIALS AND METHODS Chlorobiphenyl Substrates. 4'-ChIoro-4-biphenylol was synthesized as described (Savoy and Abernethy, 1942). J. Agric. Food Chem., Vol. 23, No. 5, 1975 B51 MOMS 0 8 1 5 3 1 SAFE. HUTZ1NCER, JONES Table I. Identification and Recovery of the Rabbit Urinary Metabolites of 4-Chlorobipheny) and 4'-Chloro-4-biphenylol Substrate0 Compd (M*)* 4-Chloroblphenyl 4'-Chloro-4 -biphenylol ------------------------------------ -------------------------------Free, mg Bound, mg Free, mg Bound, mg NMR data (or the metabolites, 8, ppm 4'-Chloro-4 -biphenylol (m/e 204) 80.4 15.0 72.6 12.9 7.55 (d,J = 8.2 Hz), 7.48(d, J = 8.2 Hz), 7.42 (d,Js 8.2 Hz), 7.18 (d, J = 8.2 Hz) 4' -Chloro -3 -methoxy -4 -biphenylol* 13.2 3.0 8.0-7.0(m), 3.94(s),3.91(a) 4' -Chloro -4 -methoxy -3 -biphenylol (m/e 234) 4' -Chloro-3,4 -blphenyldtol (m/e 10.5 1,5 45.6 7.5 7.43 (d,J = 8.2 Hz), 7.37 (d, 220) J= 6.2 Hz), 7.08 (d, J = 1.9 Hz), 7.01 (q, J a 8.2, 1.9 Hz), 6.93(d, J * 8.2 Hz) M* molecular ion. * The two methoxy iaomer* could not ba separated by TLC and were analyzed by gas-liquid chromatography (GL.C) as described (Safe et at., 1975b); the relative percentages of the methoxy isomers were determined by the GLC peak areas and the values given in the table were calculated from the yield of diol obtained after demethylation of the mixture.f The figures given are baaed on milligrams of compound recovered from tbe urine of a single rabbit; each rabbit (average wt, 5 kg) was given 300 mg of substrate by intraperitoneat injection in a vegetable oil solution. 4-Chloro*4'*iodobiphenyl (1.0 g, prepared from the corre sponding 4-araino-4'-chlorobiphenyl) wu dissolved in dry tetrahydrofuran (26 ml) and lithium aluminum deuteride (0.20 g) was carefully added. The mixture was then re fluxed for 8 hr, cooled, and quenched by the addition of deuterium oiide (6 ml). The products were isolated with ether extrection and purified by preparative thin-layer chromatography (TLC). The major product was 4'-[2H]-4ehiorobiphenyi (0.260 g, 65% purity by mass spectrometry) and some 4,4'-|2Ha)biphenyl (0.05 g) was also isolated. Un labeled 4-chlorobipheny! was obtained from commercial eources (Eastman) and used directly at a metabolic sub strate. Animal Feeding Experiments. The chlorobiphenyl substrate (300 mg) was dissolved Jo vegetable oil (5-10 ml) and administered intraperitoneally to a male rabbit (ca. 5 kg) which was housed in a metabolic cage. Urine and feces were collected for 10 days after edministration of the xenobiotic. Extraction and Analyeie. The urine samples were acid ified to pH 4-5 with the addition of glacial acetic acid and extracted with ether to give the free metabolite extract. The urine samples were then carefully acidified with con centrated sulfuric acid to give a 2 N sulfuric acid solution which was then heated bn a steam bath for 2 hr. The solu tion was cooled and then extracted with ether to give the bound metabolite extract The 4-chlorobiphenyl and 4*-f*H}-4-ch)orobipheny) me tabolites were isolated by preparative TLC as described (Safe et a!., 1975b) and the 4'-chloro-4-biphenylol metabo lites were also obtained by TLC (Safe et a)., 1975a). The crude hydroxy fractions were purified via their acetate de rivatives and the corresponding hydroxylated metabolites were obtained by subsequent basic hydrolyses of the ace tates. Demethylations ware carried out by treatment of the metabolite with boron tribromide (0.1 ml) in methylene chloride (2 ml) for 9 hr at 25". The solution was diluted with water (15 ml) and methylene chloride (20 ml). The de methylation product was then obtained by evaporation of the dried methylene chloride extract. Spectroscopic Methods. Mass spectra were obtained on a Varian MAT CH-7 spectrometer and nuclear magnetic resonance (NMR) spectra were recorded on a Varian HR220 spectrometer using deuteriochloroform as solvent. RESULTS A summary of tha major rabbit urinary metabolitaa of 4* chlorobiphenyl and 4'-chloro-4-biphenylol is given in Tabla 1. The former substrata waa converted into two metabo lites, 4'-chloro-4-biphenyIol and 4/-chloro-3,4-biphenyldiot; 4'-chloro-4-biphenylol waa metabolized to give V-chloro3,4-biphenyldiol as the mejor product and smaller amounts of a mixture of 4,-chloro-3-methoxy-4-biphenylol and 4'chloro-4-methoxy-3-biphenylol. The latter two compounds could not be separated by TLC; however, the NMR spec trum of this fraction gave two methoxy) signals at 4 3.94 and 3.91 ppm. Treatment of the mixture with boron tribro mide in methylene chloride geve 4'-chIoro-3,4-biphenyldiol as the sole product. These metabolism results with 4-chlo robipheny] and 4/-chloro-4-biphenylol were similar to those obtained from the metabolism of these substrates in tha rat (Safe et al., 1974; Safe et a)., 1975a). Administration of 4/-(2H)-4-chlorobiphenyl alao gave 4'chloro-4-biphenyiol and 4'-ch!oro-3,4-biphenyldiol ae ma tsbolitas and these were examined by mass spectrometry. The phenol gave molecular ion specias at m/ 205 and 204 and 79% of the deuterium was retained in this metabolite. Examination of tha NMR spectrum showed that the signal at 7.18 ppm was diminished in its expected intensity and confirmed the shift of the 4-2H atom to the 3 position. The mass spectrum of the 4'-chloro-3,4-biphtnyldiol isolated from this experiment gave molecular ion peaks at m/c 221 and 220 and 39% of the deuterium waa retained in this me tabolite. The mass spectral data were obtained ueing the metabolites isolated from the free urine extract. Treatment of the deuterated chlorobiphenyl# with 2 N sulfuric acid re sults in some H-2H exchange and for this reaeon the maae spectral data for the metabolites isolated from the bound fraction were not used for calculation of the 9H retention. DISCUSSION MGNS 08153.: The results obtained tor the metabolism ot 4-chlorobi phenyl and 4'-chIoro-4-bipheny)ol in tha rabbit were simi lar to the experiments already described for rats (Safe et al., 1974, 1975a). In addition, the metabolism of a series of chlorobiphenyls including 2,3-dichiorobiphenyl, 2.4.6-trichlorobiphenyl, 2,3,5,8-tetrachlorobiphenyl, and 2,3,4,5,6pentachlorobiphenyl also gave the corresponding phenols. 852 J. Aorlc. Food Chem Vnl M Nr* a iora catechols, and methoxyphenola as the major metabolites (Goto et a)7 1974). It is conceivable that the formation of these pfoductw'could occur via an arcne oxide which on re arrangement would give tho phenol and on hydrolysis fol lowed by dehydrogenat ion would give the catechol (Daly et al., 1972). The methoxyphenola could then be formed by methylation of tho eatechole. The conversion of 4'-chloro4-biphenyloI into the corresponding diol presents a second alternative pathway for the formation of this metabolite. Analysis of the two urinary metabolites of 4'-{aH)-4-chlorobiphenyl cloerly resolves this problem. The mass spectrum of 4'-chloro*4-biphenylol shows that 79% of the deuterium is retained in the metabolite and this is consistent with an arene oxide intermediate in which the hydrogen haa mi grated from the site of hydroxylation to a neighboring car bon atom (the N1H shift). This result is supported by the NMR spectrum of the metabolite in which the signal at 7.18 ppm is of diminished intensity due to the presence of deuterium. The N1H shift has also previously been re ported for the oxidation of biphenyl to give biphenylol (Daly at al., 1968) end 4-iluorobiphenyl to give 4'-fluoro-4biphenylol (Daly et al., 1969). The mass spectrum of the 4'-chloro-3,4-biphenyldio] me tabolite gave molecular iona at m/e 220 and 22! and 39% of the deuterium from the original 4''(aH]-4-chJorobiphenyl substrate was retained in the metabolite. This result sup ports the formation of the diol by two independent hydrox ylation reactions in which the first step occurs via an arene oxide followed by a second biological hydroxylation reac tion. The monooxygenase enzymes responsible for the ineertlon of the second hydroxyl group thus remove one-half of the deuterium present at the 3 poeition in 4'-ch!oro-4biphenylol. This sequence of events is analogous to the stepwise hydroxylation of phenylalanine to give tyrosine whieh is in turn hydroxylated yielding 3,4-dihydroxyphenylalenine (Guroff et al., 1966). Similarly, it has also been shown that the oxidation of butamoxane to give 6,7-dihydroxybutamoxane proceeds via two consecutive hydroxyl ation reactions rather then by tho more direct route through an arene oxide intermediate (Murphy et al.t 1974). The arena oxides of severe! polycyclic aromatic hydro carbons are known to be more active as carcinogens than either the parent hydrocarbon, the corresponding phenols, or dihydrodioli (Grover et el., 1971). The results reported In this paper show that a lowsr chlorinated PCB isomer, 4chlorobiphenyl, is also metabolised via an arene oxide in termediate. Since the biological propertiaa of PCB and their metabolites ere not fully understood this is, therefore, an area of environmental concern. The poeaiblc toxicologi cal properties of PCBs and their metabolites are currently under inveetSgation in our laboratory. ACKNOWLEDGMENT The assistance of A. Grey and V. Robinson of the Cana dian 220-MHZ Nmr Centre is gratefully acknowledged. LITERATURE CITED Ahmed, M., Foeht, D. D., Bull. Environ. Contam. Toxicol. 10, 70 0973). MtfCHANISM OF CHLOR08IPHKNVL MKTAUOUSM Baxley, G. E., Reichel, W. L., Cromartie, E., J. Assoc. Of/. Anal CAmR 251 (1970). "' Biroe, F. J., Welker, A. C., Medbury, A., Bull. Environ. Contain Toxicol. 6.317 (1970). Block, W. D., Corninh, H. H , J Biol. Chem. 334, 3301 (19S9). Burse, V. W., Kimbrough, R. D., Villanueva, E. C., Jenning, R. W. Linder, R. E., Sovocool, C. W., Arch. Environ. Health 39, 301 (1974). Butler, P. A., Pestic. Monit. J. 6,238 (1973). Daly, J., Jerins. D., Farnsworth, J., Guroff, G., Arch. Biochem Biophyt. 131,238(1969). DaIy^J.j Jenna, D., Witkop, B., Arch. Biochem Biophys. 128, 617 Daly, J. W., Jenna, D. M., Witkop, W. T., Experientio 38. 1129 (1972). Fiehbein, L., Sci. Total Environ. 4,304 (1973). Gardner, A. M., Chen, J. T., Roach, J. A. G.t Rsgelis, E. P.. Bio- chem. Biophys. Res. Common. 65,1377 (1973). Goto, M., Sugiurs, K., Hatton, M., Miyaeawa, T., Okareura. M Chemosphere 227,233 (1974). Grab, W., Klein, W., Coulston, F., Goldberg, L., Korte, F., Chem- sphere, 143 (1973). Grover, P. A., bIdii, P., Huberman, E., Marquerdl, H., Kuroki, T. Heidalberger. C., Proc. Natl. Acad. Sci. U.S.A. (6,1098 (1971). Guroff, G., Reifsnyder, C. A., Daly, J., Biochem. Biophys. Rss. Commun. 24,720 (1966). . Harvey, G. R, Steinhaver, W. G., Miklas, H. P., Naturs (London) 262,387 (1974). Horaij W.^RUebrough, R. W., Soutar, A., Young, D. R., Science Hutxlnger, 0., Jamieson, W. D., Safe, S , Paulman, L., Ammon. R.. Nature (London) 2S2,698 (1974a). Hutzinger, O., Nash, D. M,, Safe, 6., DeFrsitaa, A. 8. W., Noratrom, ft. J., Wildish, D. J., Zitko, V., Seisnce 178,312 (1972). Hutzinger, O., Safe, S.. Zitko, V., "The Chemistry of PCBe", Chemical Rubber Publishing Co.. Cleveland, Ohio. 1974b. Jamieson, W. D., Hutzinger, 0., Safa, S., Crocker, J. P. 8., Zitko, V., Proc. Am. Soc. Moss. Spectrom. Allied Top., 486 (1973). Jensen, S., Johnels, A. G., Olsson, M., Otterlina, G., Nature (Lon don) 224,247 (1969). Jerina, D. M., Lloydia 37,212 (1974). Jerina, D. M., Daly, J. W., Science 185,673 (1974a). Koeman, J. H., Ten Noevr de Brsuw, M. C., deVoe, R. H.. Nature (London) 221,1126(1969). Maas, W. a. G., Safa, S., Hutzinger, 0., Arch. Environ. Contam. Toxica/., in preea (1976). Moza, P., Weisgerber, I., Klein, W., Korte, F., Chemoephere, 217 (1973). Moza, R. Weisgerber, I., Klein, W., Korte, F., Bull. Environ. Con tarn. Toxicol. 12,641 0974). Murphy, P. J., Bernstein, J. R., McMahon, R. E., Mol. Pharmacol. 10,634(1974). Preatt, 1., Jefferies, D. J., Moore, N. W., Environ. Pollut. 1, 3 (1970). Risebrough, R. W., Rieche, P., Peakall, D. B.. Harman, 8. G , Kir- van, M. N., Nature (London) 220,1098 (1968). Safe, S., Hutzinger, O., Ecobichon, D., Experientio tl, 720 (1974). Safe, S., Hutainger, O., Ecobichon, D., Grey, A. A., Can. J. Bio chem., in preaa (1976a). Safe, 9., Platonow, N., Hutzinger. O.. J. Agric. Food Chem. 22.369 (1976b). Savoy, C. M. S., Abernathy, J. L., J. Am. Chem. Soc. 64, 2219 (1943). Wallnofer, P. R., Engelhardt, G., Safe, S., Hutzinger, 0., Chemo sphere, 69 (1973). Yoehimura, H., Yeroomoto, H., Chem. Phorm. Bull. 21, 1168 (1973). Received for review March 18, 1976. Accepted May 27,1976. We acknowledge the asaiitance of the National Research Council of Canada. MOWS 081533 J. Aprlc Food Chem.. Vol. 23. No. 5. 1975 852 . ------- . ik'ttiuvccior tf) (2) ^h<r< the numeric*! coefficients arc taken from equation ii;. Similarly, the tree-growth anomaly pattern correspond ing to the year following a high percentage of albacore caught north of San Francisco can be represented as map 68,56 (eigenvector 4)-6.17 (eigenvector 9) + 11.54 eigenvector 10) (3) These maps arc presented in Fig. 4. The ring-width data were mostly from trees sited in arid localities, so that a ide ring would generally be associated with anomalously cool, cloudy weather and above normal precipitation whereas a narrow ring would reflect warm, sunny and dry conditions. Below normal tree growth in the Pacific North-west (Fig, 4) is indicative of dry conditions associated with below normal cyclonic activity during the Ashing season. Sunny and mild weather would favour albaoore Ashing ' in adjacent waters, as would above normal insolation regard less of weather. The resulting excess of stored heat in the ocean would be given up through evaporation during the following autumn and winter and lead to increased cyclonic activity and precipitation along the coast north of San Francisco. These conditions would lead to increased tree growth during the following growing season (Fig. 4). Autumn and winter climatic anomaly features, com bined with spring climate and the year-to-year autocorrela tion of tree-ring widths, produce the other ring-width anomaly features in Fig. 4 for the following growing season. Narrow ring widths south of San Francisco, for example, imply below norma) precipitation--an expected feature since winter precipitation in the Pacific North-west is negatively correlated with winter precipitation in southern California*. The reconstructed values of albacore catch distribution data (Fig. 3) and inferred population distribution also seem to exhibit long term changes over intervals of 100 yr or more, which suggest the possibility that long term fluctua tions in the occan-atmosphere system may be involved. The success of (he calibration of tre rings with albacore catch indicates the possibility of relating tree-ring varia tions to any type of biological variations which are affected by large scale climatic fluctuations. Such relationships may he quantified and used to reconstruct objectively other climatically-caused biotic variations in the past. N. E. Clark Notional Oceanic anti Atmospheric Administration, NationU Marine Fisheries Service, Southwest Fisheries Center. La Jolla. California 92037 T. J. Blasiko H. C. Fritts Laboratory of Tree-Ring Research. University of Arizona, Tucson. Arizona 8572/ Df.'tntber 3. (974: ampieil May 6. 1975. ' laMarche. V. C,. Svkmr, IIJ. I04.M04I (1974). J . M,m. IV, 01*. Mr*.. U S. P>p. 4*U, . 91, I7J-I9J (19691. * l.avtfc. R. M,, KifiOtl Jut'll A"I'.mut Mu'tor ri*rrlrt SrrvUv-Amrrtcun Oi'itwiA f'tHHirfulntn dNimwc Siwtiri CnmJuitnt Diirhif /P7J (National Marine Fulxiiet Service. Southwest l Mvttie-. Center. La Jolla, I97)|, * Clemen*. H. R. am) Cialj. W. L., Co/If. Dr/ii t'tih umt Gtimr. Hi* Hull., 13$ 11965). * Sell*. O. L . I'iih/. Co/. Oiruuii ri\h. Mirif Arpl, 7. Ill-194 (1960). * C H . -4m lr><< Ittnull"* of torn? prtripiliilion fiitert-t In amt ail/.iirnl r-rwt lO'tivvrtily u< California Wafer Rfinunn Center, 1966). umy grow Defter in water depleted in oxygen 18 and deuterium While growing oats at different temperatures in w*icr of ditTeiem **C> and deuterium (D) abundances, we noticed th.u oats grown in Antarctic water in which is depleted in "O ami D by - 49%,and -400", relative to standard mean ocean w,ner (SMOW used as u comparative reference in hydrogen and oxygen isotope studies), showed initial growth 1-2 weeks sooner than did oats grown in water containing greater "O and D concentrations. The oats seemed to grow better in water which was most depleted in the stable isotopes throughout the grow th period. The oats were grown from the same batch of seeds in two sealed glass-covered glass jars (approximately 10 )). Twenty- five oat seeds were added to each jar, containing the same amount of vermiculite and 500 ml water to which 5.0 g JUpid- Cro, a commercial fertiliser, had been added. One jar contained melted glacial ice from the Antarctic with isotope concentrations of -49%* 6'*0 (SMOW) and -400%, 6D (SMOW). The other jar contained distilled ocean water with + 1.0%, 5'`0 (SMOW; and -f J 7%, 6D (SMOW). Both jars were placed in the chamber at the same time. The experiment was repeated three times with new materials: once (hegrowth chamber was maintained between 1.7 and 3.3 C; once between 24 and 26.6 'C; and once the temperature fluctuated between 1.7 and 26.6 `C. Each time the oats in (he jar containing water depleted in the heavy isotopes showed germination 1-2 weeks earlier and seemed to grow better throughout the growth period, than oats grown in distilled ocean water. Using oats grown a t J 5 :C\ the first sign of germination in the jar containing water depleted in the heavy isotopes was 4 d after planting. On the day 6, eight plants (out of 25) had attained a height of 6 cni. The first sign of germination in the jar with water containing the heavier isotope concentration, was after 17 d. By the time five plants had attained a height of 6 cm in this jar, in that with water depleted in the isotopes, 23 plants that had reached the top of the jar (approximately 25 cm). Kashutin* observed that snow-water depleted in D increases the yield of cucumbers, radishes and spring wheat compared with controls grown in ordinary water of unspecified isotopic composition. He cites experiments on the egg productivity of hens and the weight gain of suckling pigs. In both cases water depleted in D was especially efficient in promoting productivity. Although much has been done on the effect of D-enriched water on biological systems, we suggest that research on the effect D-depleted water on plant and animal growth may prove fruitful. A major source of water depleted in D by over 400 V, (40%) compared with SMOW is snow and ice from the Antarctic polar plateau. Water depleted by ISO-180%, is readily available in the USA from Rocky Mountain snow precipitating above 10.C00 feet elevaI ion. Jim D. Glcason US Geological Survey, Irving Frjidman Denver, Colorado 80225 Received February 10, accepted June ), 197). I Ka-Tiulift, K.. frhotiti tUSSK). SI, 107(1969). Identification of chlorinated dibenzofuraus in American polychlorinated biphenyls Mortality of embryos has contributed to the reproductive failures of several bird species, including the sparrowluiwls {Acclpiter nistts) of southern Scotland1, the white-tailed eagles [HaliurvUts a/bici/ln) of Schleswig Holstein*, and the herring '. ?* -V HONS 08 gulls ((.oruj argcntaiux) of Luke Ontario*, Suspected causes include ;>,/>'!)Dfc (2,2-/>/.y-(/*chlorophenyl)-l,l-clichJoroeibykne), oilier chlorinated biocides and/or their derivatives, and (lie polychlorinated biphenyls (PCBs), all of which are present as contaminants in the eggs1"*. PCBs are present in 'high concentrations in the bird populations which suffer embryonic mortality1*3. Other orgunochlorine compounds which may be present in food webs include the chlorinated dibctiz.odioxins and the chlorinated dibenzofurans (Fig. 1), which are toxic to embryos in amounts4** less than I pg. They arc therefore among the most toxic substances known and are possible causes of (he observed mortality. The chlorinated dibenzodioxins and chlorinated dibenzofornns, however, have proved exceedingly difficult to detect in environmental samples in the concentrations at which they arc expected to be cmbryotoxic*',#. The chlorinated diben zodioxins enter the environment as contaminants in preparations of the herbicide 2,4,5-1' (refs S and II) and the fungicide pcntachlorophenol1*-13. Chlorinated dibenzofurans have been found In a French (Phenoclor DP6) and a German fClophen A60) PCB and were shown to be the active cmbryotoxic agent in these preparations4. The techniques used, however, did not detect chlorinated dibenzofurans in an American PCB, Aroclor 1260. We report here the presence of chlorinated dibenzofurans in Aroclor PCB, widely used in North America and Great Britain, and in the same Aroclor 1260 preparation examined previously with negative findings*. Samples of PC B examined include: Aroclor 1248, 1254, and 1260 (1969); Aroclor 1254 (1970); Aroclor 1016 (1972); and the same three preparations studied by Vos et a!.*: Aroclor 1260, lot No. AK-3; Clophen A-60, lot No. 912434; and Phenoclor DP-6, lot not specified. The latter three PCBs were obtained from Or Vos, the others from the Monsanto Company in the years indicated in parentheses. PCBs extracted from environmental samples most often have gas chromatographic profiles similar to those of PCB formula tions containing approximately 48, 54 or 60% chlorine. In the Aroclor series, the former (wo PCBs are equivalent to Aroclor 1248 and Aroclor 1254, respectively. Aroclor 1260, Phenoclor DP6, and Clophen A60 all contain approximately 60% chlorine. Chlorinated dibenzofurans were identified in ail Aroclor preparations except Aroclor 1016, as well as in Clophen A60 and Phenoclor DP6. Aroclor 1016 is a PCB mixture containing Pig* 1 Skeletal structures of: a, chlorinated biphenyl. x~y -- 1-10; b, chlorinated dibenzofurans, a--I y * 1-8; c, chlorinated diberuodiuxius, x-\-y ^ 1-8. av Cl* o C1.X b m m r-4 S3 O <A Z. Cl.x 0` r Nature Vol. 256 July 24 IV75 Table 1 Chlorinated dibenzofwan concentrations* in ArtKlor. Clophen and Phenoclor PCB 4 - Cl 5-CI 6-CI 1 otal Aroclor 1248 (1969) Aroclor 1254 (1969) Aroclor 1254 (1970) Aroclor 1260(1969) Aroclor 1260 (lot. AK3) Aroclor 1016 (1972) Clophen A-60 Phenoclor DP-6 0 5 (25) 0.1 (6) 0.2(13) 0 1 (JO) 0.2 (25) ND 1.4 (17) 0.7 (5) 1.2(60) 0.3 (15) 0.2{J2) 1.4 (82) 0.4 (27) 0.9(60) 0.4 (40) 0.5 (50) 0.3 (.18) 0.3(38) NO ND 5.0(59) 2 2 (26) 10.0 (74) 2.9(21) 20 1.7 15 1.0 08 ._ 84 13.6 Expressed as pgg-' PCB. Values in parentheses represent quantity as percentage total dibenzofuran. ND, not delected (<0.001 mr g-1). Amounts of PCB ranging from 1.0 to 2.0 g were dissolved in 400ml hexane, placed on a Elorisil column (180 p. internal diameter 31.5 mm), and eluted with: an additional 1,600 ml hexane, and successive 800 ml volumes each of 5% diethyl-clher-hcxanc, 25% diclhyl-ether-hexanc and acetone, at a rate of approximately 7 ml min-1.Thcmajor portion of thePCBwas elutedm the hexane fraction, which was discarded. On addition of the 5% mixture, the eluatev were collected in six successive 4(<0 ml volumes. To eliminate the polar solvents, each cluate was evaporated twice just to dryness and taken up each time in n minimal amount of hexane. Each fraction, in 1 ml hexane, was placed on a microalumina column'* and eluted with 10 ml each of 1 % and 20% methylene chloride in hexane. These were also taken twice just to dryness and made to un a volume of I ml in hexane to eliminate the methylene chloride before gas chromato graphic analysis. Aliauots of ail fractions obtained before and after partitioning on alumina were injected into a six foot glass column containing 3% OVI on 100-120 mesh Supelcoport in Trocor MT220 and Hewlett-Packard 5700 gas chromatographs equipped with **Ni electron-capture detectors. PCBs were found to be present in each fraction eluted from the Elorisil column in amounts sufficient to interfere with the detection of trace contaminants. Partitioning on the alumina columns separated most of the PCB interference into the I % methylene chloride fractions. On removal of this interference, different peak patterns appeared in the chromatograms of the 20% methylene chloride fractions. Compounds eluting in the 20% methy lene chloride fraction were collected for mass spcctromelric analysis using a 20:1 cfTluent splitter, and a trap consisting of a capillary lube (I mm internal diameter, 100 mm long) bent to a U shape, immersed in a liquid nitrogen bath. Methylene chloride (20%: 4 pi) in hexane was injected into the capillary as a rinse, removed with a 1.0 pi micropipelte, and placed directly on the mass spectrometer probe. The probe was inserted into a GEC AE.I MS902 high resolution mass spectrometer and the solvent removed by the force pump. 1 he probe was rapidly inserted into the ion source and multiple scans were recorded in the un-line high resolution mode11. approximately 42% chlorine and has replaced Aroclor 1242 in many applications, principally as the dielectric fluid in capacitors'*. Values reported in Table 1 represent the total of those compounds found in 400 ml Fiorisil fractions 2-6. A total of 10-12 isomers was identified in each PCB. Two chlorin ated dibenzofuran contaminants have been reported for the Clophen and Phenoclor previously*; our first analyses of the Clophen revealed an additional five chlorinated dibenzofurans'. The structures contained four to six chlorine atoms. Other diben zofurans including those chlorinated to a lessor extent may have been present in the first 400 ml fraction but this was not examined in detail as it contained substantial PCB interference. Recently synthesised 2,3,7,8-telra-, 2,3,4,7,8-penta- and 2,3,4,6,7.8hexachlorodibenzofuran were used to quantify tetra-, penta-, and hexachlorodibcnzofurans, respectively. The former two authentic standards had retention times on (lie OVI column the same as those of two dibenzofurans isolated from the PCB Vos et /.* detected no chlorinated dibenzofurans in an Aroclor 1260 preparation at a detection limit of I p.p.m. Fractionation and examination of the identical Aroctor 1260 in our study confirm their findings based on the slated limit, but reveal the presence of II chlorinated dibenzofurans in the preparation, having a total concentration of 0.8 pg g '' PCB (Table 1). The same workers also found diethyl ether extracts of the Clophen A60 and Phenoclor DP6 to be much more toxic to chick embryos than diethyl ether extracts of Aroclor 1260 Our study confirms those findings on the basis of chlorinated Suture Yol. 2!6 July 24 1975 Ufl" LJ & ' 10 l ime (min) 20~ T,i.rir.T-- 200 250 Fig. 2 a, Gas chromatogram of a fraction of Aroclor I2S4 comaining a mixtuic of chlorinated biphenyls, dibenzofurans. amt naphthalene*. Identities of peaks arc given in the text, ft, Mass spccuum of peak. 2, a letrachlorodibeiuofuran. diben/ofuran content: the identical Clophcn and Phenoclor contain 11 and 17 times more total chlorinated dibenzofurans, respectively, than Ihc Aroclor 1260. A gas chromatogram showing components derived from the Arnclnr 1254 obtained in 1969 is represented in Fig. 2. The components were eluted in the Second 400 ml Florisil fraction and receive!cd from the alumina column in 20% methylene chloride-hexane. Fach of the numbered peaks was trapped as described here, and identified by mass spectrometric analysis. A nominal mass plot of the high resolution mass spectrum of peak 2 is shown in Fig. 2. The plot includes all the ions with elemental compositions ranging to the maximum empirical formula CiH*0,*Clk,,C1*,*Cv The molecular ion duster at nominal m/e 304- 310 fragments by successive losses of Cl to yield the ions at w/e 269-275 and CO to the ions at m{c 241-245. A minor loss of Cl from the peaks at m/e 269 -275 also occurs to yield the ions at m/r 234 -238, followed by CO clinmiHlion to m/c 206-210. The group of peaks at m/c 152-154 arc the doubly charged molecular ions. An identical spectrum was obtained from an authentic standard of 2,3,7,8-lclrachlorodibenzofuran. This latter compound has a retention time identical to that of peak 4. Peak 2 is, therefore, a positional isomer. The accurate mass mcxMiicmcnts for the characteristic ions arc within 2 p.p.m. oft he calculated exact masses. Peaks identified on this chromato gram and their retention times relative todicldrin areas follows: h mix lure of tetra- and pcntachlorobiphcnyl (1.02); tctrachlorodihcn/ofuiaw (1.30); pcntachlorobiphcnyl (l .46); tclrachlorodihenzofurun (I.S7); hcxachloronaphthalcnc (1.75); penlachlorobiphcnyI (1.86); hcxachioronaphthalenc (2.00); and hcphictdoronapluUtdcnc (3.46). An aliquot of combined fractions derived from the same Aroclor 1254 was treated with di;i/omcth;mc to assess whether any chlorinated oriho-hydroxybiphcnyls (ptc-furans) were present. Gas chromatographic analysis of the suniplc before and after Meatmen) resulted in identical chromatograms. As large Quantities of I'Clls have entered the global environmay Ik*assumed that (hecontaminant dibenzofurans 307 also have been released in proportional amounts. Their per sistence, effects, and significance rcrmiin to be determined. We thank .1. A. Buikc, M. L. Porter and J. (\. Vos for discussions; A. S. Kendc for standards of chlorinated Qi- benzofurans; and F. C. Walls for assistance with the mass spectrometry. This work was supported by the Canadian Wildlife Service, National Science Foundation, and NASA. Gikaid W. Bowls* Canadian Wildlife Service, Michah. J. Mui.vihii l Toxic Chemicals Section, Ottawa, Canada KIA OH3 Space Sciences Laboratory, BtRND R. T. StMONltl A. L. BurlinOamc University of California, Berkeley, California 94720 Bodega Marine Laboratory, University of California, R. W. RiSFBwouon Bodega Buy, California 94923 Received February IS; accepted May 2*. 1975. Prevent iddrett: California W'ur Resource* Control Board. Oiviuon of Planning and Research, 14)6 Ninth Street. Sacramemo, California 95114, ' Newton. V. and Bogan. J.. Salute, 24*. *92-59.) (1974). i Koantan. J. H.. Hadderinah, R, H.. and RiilevtlU. M. F I. J.. glut. Cw>r'a. J7J-377 (1972). * Oilbertton. M.. anti Hale. R .Cort. rid Set.. IS. 354 ISC (1974). * Hi'eclnbothant, O. R., ei .>/. Katutr. 270, 102-70) (1961). 1 Sparsebu, G. L.. Dunn, F. L.. and Roe. V. K., Fomi Cittmet. Taxie., 9,*05 412 0971). * Voa, J. O.,Koentan J. >1.. Van der Maav. H. L . ten Noewr de llrauw, M. C.. nd de Vo. R. H.. Food CoimH. Turn-.. I, (,2.1-OJJ (1970). 7 Vo, J. C.. Fjrtito*. Hllh Peng., 1. 105-117 (1972). * Bowel, G. W.. ftintonclt, 8. R.. Burlinj*ni, A. L., de lappe, B- W., and Ritabrough, R. W., L'mlt'.n. Hllh Perip.. f. 141-191 11971). * Baushman. R. and MeteJvnn, M ./-">. Hhh Pena.. S, 27-J5 (I97J). * Baaihman, R .and Mevtken. M., Adv. CA.m- 120. 92-104 097)). ' Repotton 2.4.5-T (ecu;>>e Office of ih P-evidrnt. Science Advitory Committer. Office of Science and 7echnolocy. March. t?71), *7 Jettten, S., and Renters. l.. Ambio. I. 62-'>5 11972). >) Firestone, D.. Ren, J.. Bton. N. L., Barton. R. P., and Damico, J. N., J. A\. Off. Anafyi. Chtm., 55, 95-92 (I9'2,. rottei, M.L., and Burke. J. A..J- Ait.Or. them., 14. 1426-1420(197)). Burlingame. A. 1... Olaen. R. W. and MePherton, R. V., Adr. Matt Spent.. 6. 1051-1059 (19? t>. I* Ntvbcl, I. C. T., and Sarof.m. A. F.. F.mhnit. Hhh Pmp. I, 21-3R (1972). 11 Bowes. G. W., and Jnnkcl, C. J.. J. fhh Pn. lid Can. (in the press). ** Jenven. S.. Johnels. A. G.. Olvvnn. M. and Otirrlind, G . Nature. 124. 247- 250 0969).. >* Koeman, J. H.. ten Noe^er de Brauw, M C . and de Vo>. R. H., Salute. 221, 1)26-1)29 0 969). * Rltrbtouch. K- W.. Keiche. P. Peakall. D B.. Herman. S. G . and Knven. M. N., Nature. 220. 1096-1102 (1969). Niche breadth in Bryozoa as a test of competition theory CoMt'timoN theory predicts that intraspecilic and inter specific competition should often have opposite effects on the use of resources by a population, the former increas ing, the latter decreasing, the range of resource actually used''*, Field data supporting these predictions are well kn.own for the interspecific case' ` but arc scarce for the inlraspecific condition, and we have been unable to find any reference demonstrating both effects within a single species. We therefore report here the verification of both predictions In respect of competition for space by the epiphytic bryozoan Alcyonidium hirsutum; less extensive data suggesting the same effects within other hryozoans arc also reported. Intraspccifie competition should result in an increase in the range of a resource spectrum used by a species, as at high population levels the advantages to any individual of being at the competition-free optimum of a resource gradient arc offset h> the intense inlraspecific competition found there (Fig. lo). this is the 'principle of equal oppor tunity' of MacArthur1. Interspecific competition, on the other hand, should tend to restrict the range of the resource spectrum used by a species, as individuals attempting to exploit marginal resources cannot do so as efficient!) as HOMS 0 6 1 5 3 6