1 ml. of distilled water followed by centrifugation) in a 100-ml. Erlenmeyer flask. Determine the bromide present according t o the modified van der lleulen method. Perform a blank test simultaneously.
Table I1 gives the results obtained for this procedure in four parallel determinations (A) compared with three parallel determinations, starting from 4 ml. of serum (B). The blank titration values for these determinations n-ere constantly 1.3 ml. of sodium thiosulfate. A maximum 1.8% was obtained with deviation of this procedure for five determinations using a solution known to contain 0.038 nig. of bromide instead of serum. Duplicate bromide determinations n-ere carried out on several samples of urine and serum (Table 111).
+
OTHER PHYSIOLOGICAL FLUIDS
Bromide can also be evaluated in sweat, saliva, or cerebrospinal fluid. Cerebrospinal fluid is treated in the same way as serum (using 2 ml. for protein precipitation). Because sweat and saliva are almost protein free, 0.5 to 2 ml. may be treated directly with silver nitrate, after adding 0.1 gram of sodium chloride. ACKNOWLEDGMENT
The authors are indebted t o hI. R. Bloch, ~ h osuggested this work, for his help and advice throughout this investigation.
(2) Hunter, G., Goldspink, A. A, Analyst 79, 467 (1954). (3), Xason, hl. F., J . Biol. Chenz. 113, 61-73 (1936). (4'1 Meulen. J. H. van der. Chem. Weekblad 28,' 82, 238 (1931); 31, 558 (1934). (5) Sagv, hl., Straub, J., Z. S e u r o l . Psych." 153, 215-21 (1935). 16) Seufeld. -4.H.. Can. J . Research 14 B, 160-94 (1936): (7'1 Pincussen. L.. Roman. W..Biochem. Z. 216, 361 (1929). (8)Weir, E. G., Hastinge, A. B., J . Biol. C'hem. 129, 547-58 (1939). (9) Kikoff, H. L., Bame, E., Brandt, AI.. J Lab. Clin. M e d . 24. 427 (1939). (10) Zondek, H., Tvipi, G., I . P i s a n i , Giorn. patol. neruosa e mentale 58 (1938). \
,
~
\
,
,
I
LITERATURE CITED
(1) Hunter, G., Biochem.
J . 60, 261
(1955).
RECEIVEDfor reviev August 1, 1957. Accepted December 7, 1957.
Microestimation of Intact Phenylmercury Compounds in Animal Tissue V. L. MILLER, DONNA LILLIS, and ELIZABETH CSONKA Western Washington Experiment Station, Puyallup, Wash. F A procedure for the estimation of approximately 5 to 20 y of phenylmercury acetate per gram of animal tissue or 5 ml. of urine i s presented. The sample is saponified and oxidized with alkaline permanganate. Excess oxidant i s destroyed b y hydroxylamine and ammonium sulfamate. Following acidification with hydrochloric acid, the phenylmercury i s extracted as chloride with chloroform. The analysis is completed b y the dithizone reaction for RHgX compounds.
I
and agriculture, organic compounds of mercury are now used more than inorganic. However, almost all biochemical knowledge of mercury, whether organic or inorganic, is based on analysis of the material as inorganic mercury. Because of the many types of organic compounds of mercury with varying biological properties, knowledge of the intact organic mercurial compound in the body is desirable. I n the field of toxicology, Schoeller and Schrauth (6) reported that some organic mercurials are sufficiently stable to be excreted unchanged. More recently Sivennson (7) and Hagen ( 2 ) investigated ethylmercury, niethj-lmercury, and phenylmercury compounds used as fungicides. They each reported that the symptoms produced by injection or feeding of these materials are different from the symptoms from mercuric chloride or acetate, and that each produces distinguishable N MEDICINE
symptoms. Fitzhugh et al. (1) reported that larger amounts of mercury are stored in the liver following feeding of phenylmercury acetate, than when mercuric acetate is fed. Lundgren and Swennson (3) reported that alkyl mercury compounds are stored to a much greater extent in nervous tissue than inorganic mercury, folloI\-ing exposure of workmen to these materials. All the results except those of Schoeller and Schrauth (6) are based on analysis of tissue or excretion products for inorganic mercury. Because inorganic mercury and phenylmercury react differently with the dithizone reagent (4, 5 ) , it appeared feasible to determine this organic mercurial in its original form in urine or tissue. This paper reports a method suitable for the determination of from 5 to 20 y of phenylrnercury acetate in urine, liver, kidney, spleen, muscle, or brain. REAGENTS
Although micro amounts of inorganic mercury are a common contaminant of reagents, phenylmercury is an unlikely one. Therefore rigorous purification of reagents, except for chloroform, is unnecessary. All reagents are S.C.S. unless otherwise indicated. Extract chloroform, U.S.P. or A.C.S. grade, with 2% of its volume of sulfuric acid. Then extract three times with 30% sodium chloride. Shake the chloroform with lime and anhydrous calcium chloride, filter, and distill. Add 1%ethyl alcohol as preservative. Used chloroform may be recovered by first distill-
ing over lime, then following the procedure outlined. Test all chloroform for satisfactory performance by preparing a standard curve before use. Dissolve hydroxylamine sulfate, Commercial Solvents technical grade, 30% weight per volume in water, and filter. Dissolve ammonium sulfamate, Eastman Kodak. yellow label grade, 30% weight per volume in water, and filter. Dissolve Eastman Kodak white label dithizone in purified chloroform a t the rate of 1 mg. per ml. For daily use, dilute a t the rate of 1 t o 30 ml. of chloroform. This concentration gives a working range of from 0 to 30 y of phenylmercury acetate. Absorption Columns. Prepare the columns for all tissue except brain by sealing 7 5 mm. of 0.5-mm. inside diameter glass capillary tubing t o 85 m m . of 10-mm. inside diameter tubing with a reservoir of 7 5 mm. of 18-mm. inside diameter tubing a t the top. Prepare the columns for brain by sealing 7 5 mm. of 0.5-mm, inside diameter capillary tubing to 200 mm. of 13-mm. inside diameter tubing. The adsorbant is Hyflo Supercel, which has been treated with 1N hydrochloric acid 3 to 4 times on the steam bath until the acid gives only a faint test for iron with thiocyanate. TTash the Supercel free from acid and oven dry a t approximately looo c. Prepare the small columns for use by placing a piece of glass fiber filter paper in the bottom of the column, and add and pack the Supercel t o a depth of 55 mm. Pack 5 mm. of calcium carbonate on the top. Pack the large columns for brain analysis to a depth of VOL. 30, NO. 10, OCTOBER 1958
1705
65 t o 70 mm. with Supercel with a n added 5 mm. of calcium carbonate. PROCEDURE
Urine. Place 5 or 10 ml. of r a t urine in a 250-ml. standard taper flask with 20 ml. of 1N sodium hydroxide. Connect t h e flask t o a reflux condenser a n d place on a boiling water bath for 20 t o 30 minutes. After cooling, add 12.5 ml. of 5y0potassium permanganate solution for each 5 ml. of urine with thorough mixing. Mix a volume of hydroxylamine sulfate equal to half the volume of permanganate used, with half of its volume of ammonia. iidd the mixture to the urine and permanganate with swirling. Last, add 5 ml. of the ammonium sulfamate solution. Cool the reaction mixture with running water. Add sufficient 1 2 N hydrochloric acid under the surface of the liquid with vigorous swirling to lower the pH to 1 or less. If necessary, cool the sample not over 5 minutes. Transfer the solution to a separatory funnel containing 11 ml. of chloroform and shake the mixture vigorously for 1 minute. Transfer the chloroform laver to a second sepnratory funnel containing 20 ml. of 1N hydrochloric acid. Shake this funnel for 45 seconds. Transfer the lower layer to a funnel containing 25 ml. of 0.3.V acetic acid, add 1.00 ml. of the dilute dithizone, and shake for 30 seconds. Transfer the chloroform to a volumetric test tubc and make up to 11.0 ml. Determine traqsmittance with the 620-mp filter in an Evelyn photoelectric colorimeter. The percentage transmittance of the green unreacted dithizone is determined, rather than the yellow phenylmercury dithizonate. Compare the percentage transmittance with a standard curve prepared in the mme way. Kidney, Liver, Muscle, or Spleen. Weigh 1 gram of tissue into a 250-ml. fla& wits 20 ml. of 1.V sodium hydroxide, as with urine, and heat on the steam b a t h for 10 minutes after t h e tissue is dissolved. The procedure is t h e same for tissue samples as for urine through the 1S hydrochloric acid extraction. However, the chloroform will be cloudy. For clarification, put the chloroform layer on the small Celite column, using about 8 p.s.i.g. pressure to force the chloroform through. Wash the column with a small amount of chloroform. Collect not more than 10 nil. of chloroform in a separatory funnel containing 25 ml. of 0.3.V acetic acid. The procedure is then the same as for urine. Brain. Apparently because of large amounts of lipide material, brain requires more vigorous treatment than other tissues. Weigh 1 gram of brain into a 250-mi. s t a n d d taper flask. After saponification with 20 ml. of 1.Y sodium hydroxide, cool the flask by immersing in cold water a few minutes. Add 25 ml. of 5% potassium permangenate and place the flask again under reflux on a steam bath. After 15 to 20 minutes, cool the flask or allow to cool. Destroy the excess permanganate with hydroxylamine as before, but use 10 mi. of ammonium sulfamate. Acidifi1706
ANALYTICAL CHEMISTRY
cation and chloroform extraction are done as with other tissue samples. The chloroform layer may be slow in separating and very cloudy. Purify by shaking the chloroform layer with 20 ml. of N hydrochloric acid. followed by passing the chloroform phase through the large Celite column. Then complete the analysis as for the other tissue samples. DISCUSSION A N D RESULTS
This procedure depends on the quantitative extraction of phenylmercury chloride from a urine or tissue digest into a clear and colorless chloroform solution. Undigested urine gave a yellow extract with much emulsion. I n an investigation of the amount of sodium hydroxide necessary for saponification, half of the recommended amount resulted in a more severe emulsion during extraction. Somewhat larger amounts caused no difficulty, except that more heat mas formed during neutralization. Three times as much sodium hydroxide weakened the action of the permanganate, and extracts tended to be ypllowish. Smaller amounts of permanganate may result in a yellon. chloroform extract. Much larger amounts apparently slowly attack phenylmercurg compounds causing low results. The optimum amount of permanganate is the smallest amount that will destroy the color, thus giving a colorless chloroform extract. The digest must be made acid with chloride present to give a q u a n t i t a t i v extract. However, the destruction of the excess permanganate must begin under alkaline conditions to prevent the formation of an oxidizing material. This oxidant will bleach the dithizone. causing high results. Therefore, sufficient ammonia is added to render the hydroxylamine sulfate solution strongly alkaline. This mixture is prepared as used to prevent decomposition of the hydro-q-lamine base. The ammoniim sulfamate is added to destroy the oxidizing material formed during the alkaline reduction of the excess permanganate by hydrovylamine. The oxidizing material is probably an oxide of nitrogen, reported by Yost and Russel (8) to be formed from miving the chemicals used. Occasional samples of urine or tissue may form difficultly separable emulsions under conditions of the procedure. Separation may result from subjecting the separatory funnel containing the emulsion to vacuum long enough t o cause air bubbles to separate. This will aid in breaking the emulsion, with resulting separation of the chloroform. Sometimes it is necessary t o let the separatory funnel stand for some time to allom the maximum separation of the chloroform phase. The dithizone procedure described for organic mercurials is free from inter-
Table I. Average Recovery of Phenylmercury Acetate Added to Tissue Samples
Added,. Y. 0 2 5
Found, y KidUrine Liver ney 0.6 1.9 4~. 6
10
...
20
19.7
Std. dev. at20
0.85
...
0.9 2.6
6. . 0
5 1
1.5
Brain 0.7
...
...
97 1’96 19.9
9.5 18.8
0 82
0.89
1.08
ference. Samples of tissue saponified with 1000 y of mercury as chloride showed black mercuric sulfide formation. However, this amount of mercury, many times more than could ever be found in animal tissue, did not interfere in the analysis. Attempts to separate ethylmercury from tissue by the method described have been only partially successful. Losses occur during saponification and oxidation, indicating that a major modification of the technique will be necessary to analyze tissue for intact ethylmercury. The data in Table I record results of adding phenylmercury to tissue, followed by analysis by the described procedure. The values reported indicate that results will average within approximately 1 y of the true value. Values less than 1 y ( 2 in the case of liver) should be considered as blanks. Some chicken liver samples gave a pale yellowish chloroform extract, which may account for slightly higher blanks. ACKNOWLEDGMENT
The authors Ivish to thank P. A. Klavano and -4.C. Jerstad of the Department of Veterinary Medicine for continued interest in this m r k . LITERATURE CITED
Fitzhugh, 0. G., Kelson, -4.A, Laug, E. P., Kunze, F. M., Arch. I n d . Hug. __
(1)
OccuGational M e d . 2 , 433 (1950). (2) Hagen, U.,Arch. erptl. Pathol. Pharmakol. 224, 193 (1955). (3) Lundgren, K. D.,Swennson, A , J . I n d . Hug. Tosicol. 31, 190 (1949). (4) hliller, V.L.,Polley, D., ANAL.CHEM. 26, 1333 (1954). (5) Miller, V. L., Polley, D., Gould, C. J., Ibid., 23, 1286 (1951). (6) Schoeller. W..Schrauth. W..Chem. ‘ Z t g . 36, Ill2 (1912). (7) Swennson, A., Acta M e d . Scand. 143, 365 (1952). ( 8 ) Yost, D. M., RusseI, H., Jf;, “Systematic Inorganic Chemistry, PrenticeHall, New York, 1944. I
;
RECEIVED for review February 18, 1958. Accepted June 16, 1958. Northwest Regional Meeting, ACS, Spokane, Wash., June 13 and 14, 1957. Approved by Washington Agricultural Experiment &ation for publication as scientific paper No. 1696. Project 1316 (181). Financed in part by Medical and Biological Research Funds, Initiative 171, State of Washington.
