Simultaneous determination of selenite and trimethylselenonium ions

(10) Azzam, R. M. A.; Bashara, N. M. Elllpsometry and Polarized Light·,. North Holland Publ.: Amsterdam, 1977. (11) McCrackln, F. L. A Fortran Progra...
0 downloads 0 Views 487KB Size
Anal. Chem. 1987, 5 9 , 2063-2066 (10) Azzam, R. M. A.; Bashara, N. M. €illpsometry and Pokrized Light; North Holland Publ.: Amsterdam, 1977. (1 1) McCrackin. F. L. A Fortran Progrem for Analysis of Elllpsometer Mea surements; National Bureau of Standards Technical Note 479, Government Printing Office: Washington, D.C., 1969. (12) Neal, W. E. J. Appl. Surf. Sci. 1979, 2 , 445. (13) Sundqvist, 6.; Hhkansson, P.; Kamensky, I.; Kjellberg, J.; Saiehpour, M.; Widdiyasekera, S.; Fohiman J.; Peterson, P.; Roepstorff, P. Biomed. Mass Spectrom. 1984, 7 7 , 242. (14) McCrackin, F. L.; Colson, J. P. I n €/iipsometfyin the Measurement of Surfaces and Thin Films; Passaglia, E., Stromberg, R. R., Kruger, J., National Bureau of Standard Miscellaneous Publication 256, U S . Government Printing Office: Washington, D.C. 1964; pp 61-82. (15) Yoriume, Y. J. Opt. SOC.Am. 1983, 73, 888.

2063

(16) Vedam, K.; Ral. R.; Lukes, F.; Srinivasan, R. J. Opt. SOC.Am. 1968, 58, 526. (17) McMeekin, T. L.; Groves, M. T.; Hipp, N. J. I n Amino Acids and Serum Proteins; Advances in Chemistry Series No 44; American Chemical Society Washington, DC, 1964; pp 54-66. (18) Putzeys, P.; Brosteaux, J. Bull SOCChim. Biol. 1936, 18, 1681. (19) Doty, P.; Geiiuschek, E. P. I n The Proteins; Neurath, H.. Bailey, K., Eds; Academic Press: New York, 1953; Vol. 1. Part A, pp 393-460. (20) Bateman, J. 6.; Adams, E. D. J. Phys. Chem. 1957, 67,1039.

RECE~VED for review November 12,1985. Resubmitted March 23, 1987. Accepted April 30, 1987.

Simultaneous Determination of Selenite and Trimethylselenonium Ions in Urine by Anion Exchange Chromatography and Molecular Neutron Activation Analysis Alan J. Blotcky and Gregory T. Hansen

Medical Research, Veterans Administration Medical Center, Omaha, Nebraska 68105 Nitin Borkar, Alireza Ebrahim, and Edward P. Rack*

Department of Chemistry, University of Nebraska, Lincoln, Nebraska 68588-0304

A method has been developed for the determination of trlmethylselenonium (TMSe) ion and selenite (SeOt-) Ion in urine by anion exchange chromatography, selectlvely elutlng TMSe and SeOt-. Since the recoveries are quantitative, a method of addtthre Spwres is employed lo determine TMSe and SeOt- lon concentratlonsIn urlne specimens. The TMSe and SeOt- chromatographic elutlons were collected In vlals, irradiated with neutrons, and radioassayed for 77mSeactivity. The limit of detectlon is 10 ng of Se as TMSe or SeOz-/mL of urlne. Thlrteen urine specimens from normal subjects were analyzed for TMSe, SeOt-, and total selenium concentratlon.

Because of the biological importance of selenium in living biological systems (1-5) various analytical procedures have been developed for analysis of microquantities of elemental selenium in urine (6-9) and in serum and tissue (6,9,10). For urine selenium these include atomic absorption spectrometry, solution absorption spectrometry, solution fluorescence spectrometry, volumetry, and neutron activation analysis (6-9). Of equal or greater importance is the determination of selenium metabolites present in urine for the purpose of describing the biological pathways for the metabolism of selenium in living organisms. It is known from previous studies that trimethylselenonium ion (TMSe) is a major metabolite in urine, probably the result of reduction and methylation reactions (11-13). Early techniques to measure TMSe levels in urine involved the use of the radiotracer '%e (11,1.317). Because of the long biological half-life of selenium and issues of radiation exposure, its use in humans has been limited. Recently TMSe has been determined in urine by atomic absorption spectrometry (18)and neutron activation analysis (I 7,19). While these reported techniques are reliable, they are rather time consuming in the number of chemical manipulations that are required and as a result do not lend themselves readily to routine analysis. The purpose of the 0003-2700/67/0359-2063$01 SO10

present work was to develop a molecular neutron activation analysis (MoNAA) approach employing a single step anion exchange chromatography procedure for the separation of TMSe and inorganic selenium from organic selenium compounds such as selenoamino acids. This would allow routine analysis of these ions by neutron activation analysis, which is potentially the most sensitive technique for the detection of trace selenium in the parts per billion range. EXPERIMENTAL SECTION Reagents and Solvents. ACS reagent grade lithium hydroxide (anhydrous) 99.3% from Alpha Products (Danvers, MA), concentrated ammonium hydroxide from J. T. Baker (Phillipsburg, NJ) and formaldehydesolution (37.9%)from Mallinckrodt (Paris, KY) were used without further purification. Selenium (IV) oxide, Puratomic (99.999%)from Johnson Mathey (Seabrook,NH) was used as the selenium standard. Sodium selenite, SelenO-DL-CyStine and seleno-DL-methioninewere obtained from Sigma Chemical Co. (St. Louis, MO). Dimethyl selenide, obtained from Columbia Organic Chemical Co., Inc. (Camden, SC), was used to prepare trimethylselenonium according to the procedure described by Palmer et al. (15). Triply distilled water was used in all phases of this work. Resin Columns. The anion exchange chromatography columns for separation of the selenium were prepared by loading a 0.7 X 10 cm borosilicate column with 6 g of Bio-Rad (Richmond, CA) AG-l-X8 resin in the acetate form (100-200 mesh) or AG2-X8 resin in the chloride form (200-400 mesh). Neutron Irradiation. Clear polystyrene 5-mL sample tubes and polyethylene caps from Sarstedt, Inc. (Princeton, NJ), were used for all samples. Samples were irradiated for 20 s at a thermal neutron flux of 3.1 X 10" n cm-2 s d in the Omaha Veterans Administration Medical Center TRIGA nuclear reactor by means of a pneumatic transfer tube. Radioassay. All irradiated chromatography samples were allowed to decay for 20 s and counted for a live time of 20 s. Unseparated total urine samples were counted for 20 s of live time after allowing the sample to decay for 20 and 140 s. The 140-5 spectrum was then subtracted from the 20-5 spectrum so as to subtract out the interfering radionuclides of long half-life. 7Spectra analyses were accomplished by using a Nuclear Data 0 1987 American Chemical Society

2064

ANALYTICAL CHEMISTRY, VOL. 59, NO. 17, SEPTEMBER 1, 1987

Table I. Elution Characteristics of Se Metabolites in Urine by Anion Exchange Chromatography Employing Bio-Rad AG-2-X8 (Chloride) Resin and 0.5 M LiOH Elution

metabolite" TMSe (0.71 pg of Se) Se032-(0.59 p g of Se) Se0,2- (1.1pug of Se) elution resin selenomethionine (34 pg of Se) elution resin selenocystine (14 pg of Se) elution resin

selenocysteine (1.3 pg of Se) resin

recovery, 7% 101

Table 11. Elution Characteristics of Se Metabolites in Urine by Anion Exchange Chromatography Employing Bio-Rad AG-1-X8 (Acetate) Resin and 0.5 M LiOH Elution

elution

elution vials, mL

100

3-7 11-14

21

80-100

79

80 20

20-40

5-15 85-95

20-120

metabolite"

recovery, %

vial, mL

TMSe (35 p g of Se) Se0:- (31 pg of Se) selenomethionine (34 pg of Se)

94 95

3-10 17-19

68 23

55-98

elution resin

selenocystine (14 pg of Se) elution resin

100

The amount of selenium in the metabolite loaded on the anion resin is given in Darentheses. (Schaumburg,IL) ND680 4096-channelanalyzer and peak search software. Preparation of TMSe. Trimethylselenonium chloride (Se(CH3)3+Cl-)was synthesized according to the method of Palmer et al. (14). Analysis of the synthesized compound by mass spectrometry indicated its purity as >98%. Pretreatment of Urine Samples. In order to preserve the urine, it was treated with 0.5% by volume of the 37% formaldehyde solution upon receipt of the sample from the donor (20). Total Selenium Determination in Urine. Total selenium in urine was determined by instrumental neutron activation analysis (INAA) employing 3-mL quantities of urine. Neutron irradiation and radioassay procedures were as previously described (19). These determinations are important for establishing a selenium material balance in urine samples. Determination of TMSe and Se0;- Ions in Urine. All urine samples were alkalized to a pH of 10-11 in order to dissociate selenoamino acids from possible protein-binding sites (21). The denaturing step is necessary in that it renders the protein matter soluble allowing ready passage through the resin columns. One milliliter of concentrated ammonium hydroxide was added to each 5-mL urine sample and the solution was allowed to stand for 30 min. As will be seen later in the Results and Discussion section, the most efficacious resin allowing separation of TMSe and Se0:ions from the selenoamino acids is Bio-Rad AG-2-X8 (chloride). The concentrations of TMSe and Se032-ions in the urine were determined by the method of standard additions. For each simultaneous determination of TMSe and Se032-in urine, 1 mL of standard aqueous TMSe and Se032-is added to the denatured urine where the total volume is now 7 mL. For the determination of the amount of TMSe and SeOa2-in a subjects urine, three separate urine samples each containing a spike of different concentrations of TMSe and Se0:- are required. Three milliliters of the spiked urine were then placed on the resin. After the level of urine solution reached the top of the resin bed, the column was eluted with 0.5 M LiOH, and 1-mL aliquot portions were collected by means of a Model 328 ISCO automatic sample changer (Lincoln, NE). Each vial was irradiated for 20 s and assayed for selenium. Elution characteristics are reported in Table I. A complete separation was always obtained between ions and any selenoamino acids present in the urine. The amount of TMSe and Se03*-in the subjects urine was found by plotting the micrograms of Se added vs. the micrograms of Se assayed and calculating the y intercept of the line fitted t o the data points by using the least-squares linear regression technique (22). RESULTS A N D DISCUSSION Development of Anion Exchange Chromatography Separation Procedure. The chemical form of selenium ingested has a strong influence on its metabolic products. Dietary selenium derived from plants is mainly selenomethionine while from animal sources it is largely selenocysteine and selenomethionine in various proportions (12).

0 100

1-120

"The amount of selenium in the metabolite loaded on the anion resin is eiven in uarentheses. In addition, selenite in the form of vitamin supplements and selenate can be administered to human and animal subjects. At this time all the chemical forms of selenium in urine have not been fully identified, but TMSe is a major metabolite (11-13,15). Little is known about the pathways by which the ingested forms of selenium are metabolized to TMSe (13). In a previous study (16) by Nahapetian et al., utilizing selenium-75 labeled selenomethionine, selenocystine, and selenite, no detectable free selenite was found in urine samples. This may be due to a binding of selenite to sulfur-containing amino acids as reported by others (23, 24). For a procedure affecting a separation among TMSe, Se02-, Se02-, and selenoamino acid metabolites, it is first necessary to denature urine protein as described in the Experimental Section. The two anion exchange chromatographic resins evaluated were Bio-Rad AG-2-X8 (chloride) and Bio-Rad AG-l-X8 (acetate). The elution solution of choice was 0.5 M LiOH, which in addition to its elution characteristics facilitated continuing denaturing of the urine. Presented in Tables I and I1 are the elution characteristics of TMSe, Se02-, SeOZ-, and various selenoamino acids employing the two resins. One-milliliter elutions were collected with an automatic fraction collector. I t can be seen from Tables I and I1 that the most efficacious separations with the highest recovery are obtaired employing the AG-2-X8 (chloride) resin. While there is excellent separation among the ions, the elution of selenomethionine occurs with a broad peak and with a significant fraction remaining on the resin after elution of 100 mL of 0.5 M LiOH. This vitiates any quantitative measurement of the various selenoamino acids even though selenocystine and selenocysteine are completely retained on the resin. By use of urine samples spiked with various concentrations of TMSe or Se032-,it was found that not only were the recoveries quantative but as the spiked concentrations approached the parts per billion range the percent recovery increased progressively over 100%. This indicated the contribution of the TMSe or selenite ions in the urine blank. Depicted in Figures 1and 2 are the elution characteristics and recoveries of TMSe and Se032-in urine a t various concentrations. Determination of TMSe and Se032-in Urine by the Method of Additive Spikes. In order to determine TMSe and Se032-concentrations in a urine blank, it is necessary to take three 5-mL urine samples. For each urine sample both TMSe and Se0;- can be determined. Presented in Table I11 is a description of the calculations for three samples of the same urine spiked with different concentrations of TMSe and Se0,2-. It should be noted that the concentrations of TMSe and Se032-are given in micrograms of Se. This is necessary for a material balance since total Se determinations in urine are expressed as micrograms of Se. Depicted in Figure 4 is a regression analysis plot of micrograms of Se analyte (TMSe

2065

ANALYTICAL CHEMISTRY, VOL. 59, NO. 17, SEPTEMBER 1, 1987

Table 111. Determination of TMSe and SeOa2-Concentration in Urine Samples by the Method of Additive Spikes TMSe

Se0:-

D

B sample

pg of Se added

sum of iimSe counts in TMSe elutionn

1 2 3

0.7111 0.474 0.356

3580 2413 1904

A

std counts/ pg of Se

pg of Se in TMSe elutionb

5222 5222 5303

0.686 0.462 0.359

C

F

G

pg of Se added

sum of IimSe counts in Se032elutionn

pg of Se in Se0:elution'

0.436 0.297 0.237

2132 1660 1483

0.408 0.318 0.280

E

OSum of TMSe counts in elution vials in chromatographic peak. See Figure 4. bColumn B/C. cColumn F/C.

0.3

0.3

0.2

u

0

z

>

00

==I

m

m

600

W

0. I

v)

. 0

9;

y

A\

0.1

cu h

E

.

?z

6

400

3

0 0

200 -

V I A L (ml)

Y2 L0.i

Figure 1. Anion exchange chromatogram of trimethyiseionium ion in urine at various solute concentrations: (A) Se concentration 0.71 1 pg, recovery 101 %; (B) Sa concentration 0.474 pg, recovery 1 0 3 % ; (C) Se concentration 0.356 pg, recovery 1 1 8 % ; (D)Se concentration 0.285 pg, recovery 121 %.

4

16

12

8 V I A L (ml)

Figure 3. Typical determination of TMSe and SeO,*- concentration in urine, employing anion exchange chromatography, by the method of additive spikes. TMSe: (A) 0.71 p g of Se; (B) 0.47 p g of Se; (C) 0.36 p g of Se. SeOt-: (D)0.40 mg of Se; (E) 0.30 pug of Se; (F) 0.24 p g of Se. 0.7,

A

n W

0.5

> Q co v ,

-

Q a l

v,

a

01 a

I

1

I

0.2

Figure 2. Anion exchange chromatogram of Se0,'- ion in urine at various solute concentrations. (A) Se concentration 0.59 pg, recovery 100%; (e) Se concentration 0.40 pg, recovery 1 0 2 % ; (C) Se concentration 0.30 pg. recovery 1 1 0 % ; (D) Se concentration 0.24 pg, recovery 1 12 YO.

or Se0& vs. micrograms of Se added (spike). The y intercepts for TMSe and Se032- are 0.027 and 0.081 pg, respectively. The concentrations of these ions in the urine blank are found by dividing the y intercept of Figure 4 by l5l7since 3 mL of a 7-mL sample (5 mL of urine + 1 mL of NHIOH 1mL of spike) is placed on the resin. The concentrations of TMSe and SeOS2-in the urine blank are 0.0126 and 0.0378 pg of

+

I

I

I

0.4

I

0.6

I

1

0.8

ug Se ( A D D E D )

Figure 4. Plot of micrograms of Se added to resin vs. micrograms of Se assayed in elution, showing least-squares fitted curve, as used in method of additive spikes: SeOt-, 0; TMSe, A.

Se/mL, respectively. These correspond to sample 7 in Table IV. Employing the method of additive spikes and the INAA procedure for total selenium, urine specimens from 13 normal subjects were analyzed for total Se, TMSe, and SeOS2-. The urine samples represented a 6-h collection from the first elimination in the morning. Presented in Table IV are these data. The experimentally determined limit of detection is 10

Anal. Chem. 1987, 59,2066-2069

2066

Table IV. Determination of Total Selenium, Trimethylselonium Ion, and Selenite Ion in Urine of Normal Subjects (concentrations in ng of Se/mL; limit of detection 10 ng of Se/mL) TMSe"

sample

Se032-

1 male 2 male 3 male 9 male 10 male 11 male 1 3 male

23 56 22 63 21 ND 20

ND' 13 ND ND 41 ND ND

4 female

ND 33 ND 13 ND

62 ND ND 38 ND ND

5 female 6 female 7 female 8 female 12 female

11

total Seb 97 f 12 70 f 10 50 10 160 f 10 89 f 13 14 f 3 51 f 6 65 f 7 37 f 5 32 f 8 56 f 7 54 f 9 50 f 20

*

"Determined by the method of additive spikes. bDetermined by INAA. Nondetectable.

ng of Se/mL. While there appears to be no correlation in the concentration of the ions and total selenium between male and female subjects, the data suggest that urine samples with increasing total selenium have an increasing amount of TMSe. Our data support the results of Nahapetian et al. (16),which found no detectable free selenite in urine samples of rats ingesting 75Selabeled selenoamino acids or selenite. It appears that the MoNAA approach is sensitive enough to detect four selenite concentration values in a sample pool of 13 subjects. LITERATURE CITED (1) McConnell, K. P.; Broghamer, W. L., Jr.; Blotcky, A. J.; Hart, 0. J. J . Nutr. 1975, 705,206. (2) Broghamer, W. L., Jr.; McConnell, K. P.; Blotcky. A. J. Cancer (Philadelphia) 1978, 4 7 , 1462.

(3) Broghamer, W. L., Jr.; McConnell, K. P.; Blotcky, A. J. Cancer(Philadelphia) 1976, 37, 1384. (4) Shamberger, R. J.; Rukovena, E.; Longfeld, S. A,; Tylko, S.; Deodhar, C. E.; Willis, C. E. J . Natl. Cancer I n s t ( U . S . )1973, 50, 863. (5) Willett, W. C.; Polk, B. F.; Morris, J. E.; et al. Lancet 1983, July 16, 130. (6) Iyengar, G. V.; Kollmer, W. E.; Bowen, H. J. M. The Hemental Composition of Human Tissues and Body Fluids; Verlag Chemie: New York, 1978. (7) Cornelis, R.; Speecke, A.; Hoste, J. Anal. Chim. Acta 1975, 78, 317. (8) Valentine, J. L.; Kang, H. K.; Spiney, G. H. Environ. Res. 1978, 77, 347. (9) Ishizaki, M. Talanta 1978, 25, 167. (IO) Versieck, J.; Cornelis, R. Anal. Chim. Acta 1980, 776, 217. (11) Sunde, R. A.; Hoekstra, W. G. Biochem. Biophys. Res. Commun. 1980, 93, 1181. (12) Burk, R. F. J . N u h . 1988, 776, 1564. (13) Foster, S. J.; Ganther, H. E. Anal. Biochem. 1984, 737,205. (14) Palmer, L. s.; Gunsalus, R. P.; Halverson, A. W.; Olson, 0. E. Biochim. Biophys. Acta 1970, 208, 260. (15) Palmer, L. S.; Fischer, D.D.; Halverson. A. W.; Olson, 0. E. Biochim. Biophys. Acta 1989, 777,336. (16) Nahapetian, A. T.; Janghorbani, M.; Young, V. R. J . Nutr. 1983, 773, 401. (17) Nahapetian, A. T.; Young, V. R.; Janghorbani, M. Anal. Biochem. 1984, 740, 56. (18) Oyamada, N.; Ishizaki, M. Jpn. J . Ind. Health 1983, 25. 319. (19) Blotcky, A. J.; Hansen, G. T.; Opelanio-Buencamino,L. R.; Rack, E. P. Anal. Chem. 1985, 57, 137. (20) Blotcky, A. J.; Rack, E. P. J . Res. Natl. Bur. Stand. (U.S.) 1986, 9 7 , 93. (21) Kessler, G.; Pileggi, V. J. Clin. Chem. (Winston-Salem, N . C . ) 1968, 14, 811. (22) Bancroft, H. Introduction to Biostatistics; Hoeber 8, Harper: New York, 1959; p 161. (23) Schwarz, K.; Sweeney, E. Fed. Proc. Fed. Am. SOC. Exp. Bioi. 1964, 23, 421. (24) Cummins. L. M.; Martin, J. L. Biochemistry 1967, 6, 3162.

RECEIVED for review March 12,1987. Accepted May 12,1987. This research was supported by the US. Department of Energy, Division of Chemical Sciences, Fundamental Interaction Branch, under Contract DE-FG02-84ER13231 and a University of Nebraska Research Council NIH Biomedical Research Support Grant No. RR-07055.

Extraction of Low Molecular Weight Polynuclear Aromatic Hydrocarbons from Ashes of Coal-Operated Power Plants Filippo Mangani, Achille Cappiello, Giancarlo Crescentini, and Fabrizio Bruner*

Istituto di Scienze Chimiche, Universitd di Urbino, Piazza Rinascimento, 6, 61029 Urbino, Italy Loretta Bonfanti

ENEL-CRTN,Via Cesare Battisti, 69, 56100 Pisa, Italy

A new procedure based on llquld-solld chromatography for the extraction of polynuclear aromatic hydrocarbons has been implemented. This ylelds results analogous to those of Soxhlet extractlon for low molecular weight compounds up to the four-membered ring compounds. A very Important reductlon In the time required for sample preparatlon prlor to gas chromatography/mass spectrometry analyds Is obtalned.

The large amount of ashes produced by coal-operated power plants and the necessity for their disposal make it necessary to frequently carry out careful measurements of the polynuclear aromatic hydrocarbon (PAH) content of these materials. This implies extraction from the matrix and separation by chromatographic procedures. The most widely used methods for recovering PAHs from solid matrices involve Soxhlet ex-

traction with methylene chloride (1-3). The use of ultrasonic extraction and of a rotary shaker has also been recently evaluated (4). Vacuum sublimination has also been used (5). The effectiveness of the extraction methods, expressed as recovery of PAH from spiked ashes, decreases with decreasing concentration of these compounds. It is well-known (6) that toluene shows a higher extraction efficiency with respect to methylene chloride because its structure is similar to that of the compounds of interest. Moreover, methylene chloride tends to extract more unwanted polar compounds, being a more polar solvent than toluene. However, concentration of large volumes of solvent is needed in the traditional Soxhlet extraction procedures so that methylene chloride is usually the solvent of choice because of the lower boiling point. By the use of alternative methods of extraction that require small volumes of solvent, the superior extraction efficiency of toluene could be more conveniently exploited.

0003-2700/87/0359-2066$01.50/0 0 1987 American Chemical Society