Anal. Chem. 1991, 63,283-287
LITERATURE CITED (1) Mount, D. I.; Anderson-Carnahan, L. Methods for Aquatic Toxlcity Identification Evaluations: Phase I Toxicity Characterization Promdures; EPAi600i3-88-34 US. Environmental Protection Agency: Duluth. MN, 1988. (2) Mount. D. I.: Anderson-Carnahan, L. Methods for Aquatic Toxicity Identification Evaluations: Phase X I Toxicity Identification Promdures : EPA/600/3-88-35; US. Environmental Protection Agency: Duluth, MN, 1988. (3) Mount, D. I . Methods for Aqu8tic Toxlcity Identification Evaluations: Phase 111 Toxicity Confirmation Procedures ; EPA/600/3-86-36; US. Environmental Protection Agency: Duluth, MN, 1988. (4) Mast, T. J.; Hsieh, D. P. H.; Selber, J. N. €nviron. Sci. Techno/. 1984, 18, 338-346. ( 5 ) Nishioka. M. G.; Chuang, C. C.: Peterson, B. A.; Austin, A,; Lewtas, J. Environ. Int. 1985, 1 1 , 137-146. (6) Schuetzle, D.; Jensen. T. E.; Ball, J. C. Environ. Int. 1985, 1 1 , 169- 18 1. (7) Alsberg, T.; Strandell, M.; Westerholm, R.; Stenberg, U. Environ. Int. 1985, 1 1 , 249-257. (8) Schuetzle, D.; Lewtas, J. Anal. Chem. 1986, 58, 1061A-1075A. (9) Nishioka, M. G.; Howard, C. C.; Contos, D. A.; Bail, L. M.; Lewtas, J. Environ. Sci. Technol. 1988, 22, 908-915. (10) Hoimbom, B.; Voss, R. H.; Mortimer, R. D.; Wong, A. Environ. Sci. Technol. 1904, 18, 333-337. (11) Samoiloff, M. R.; Bell, J.; Birkholz. D. A.; Webster, G. R. B.; Arnott, E. G.; Pulak, R.; Madrid, A. Environ. Sei. Techno/. 1983, 17, 329-334. (12) West, W. R.: Smtth, P. A.; Booth, G. M.; Lee, M. L. Environ. Sci. Technol. 1908. 22, 224-228. (13) Wlemer, D. F. Rev. Latinoam. Quim. 1985, 16, 98-102. (14) Quilllam, M. A.: Wright, J. L. C. Anal. Chem. 1989, 6 1 , 1053A1060A. (15) Wright. J. L. C.; Boyd, R. K.; defrettas, A. S. W.; Falk, M.; Foxali, R. A.; Jamieson, W. D.; Laycock, M. V.; McCulloch, A. W.; McInnes, A. 0.; Odense, P.; Pathak, V. P.;Quilllam, M. A.; Ragan, M. A.; Sim, P. G.; Thibault. P.; Walter, J. A,; Gllgan, M.; Richard, D. J. A,; Dewar, D. Can. J. Chem. 1989, 67, 481-490.
283
(16) Ampofo, S. A.; Roussis, V.; Weimer, D. F. Phyfochem/sby 1987, 26, 2367-2370. (17) Ampofo, S. A.; Roussis. V.; Weimer, D. F. Phyfmhemktry 1987, 26, 2371-2375. (18) Reece, C. H.; Burks, S. L. I n Aquatic Toxicology and Hazard Assessm n t : Seventh Symposium; ASTM STP 854; Carwell, R. D.. Purdy. R., Bahner, R. C., Eds.; Amerlcan Society for Testing and Materials: Philadelphia, PA, 1985; pp 319-322. (19) Parkhurst, B. R.; Gehrs, C. W.; Rubln, I. B. I n Aquatic Toxico/o#'; ASTM STP 667; Marking, L. L., Kimerie, R. A., Eds.; American Society for Testing and Materials: Philadelphia, PA, 1979; pp 122-130. (20) Lopez-Avila, V.: McKenzie, W. D.; Sutton, W. W.; Kaminsky, R.; Spanage4 U.; Oisson, T. A,; Taylor, J. H. Application of Chemlal Fractionation /Aquatic Bioassay Procedure to Hazardow Waste Sne Mnitoring; EPA/600/S4-85/059; US. Environmental Protection Agency: Las Vegas, NV, 1986. (21) Galassi, S.; Battaglia, C.; Vlgano, L. Chemosphere 1988, 17, 783-787. (22) Durhan, E. J.; Lukasewycz, M. T.; Amato, J. R. Environ. Toxicoi. Chem. 1990, 9 , 463-486. (23) Hamilton, M. A,; Russo, R. C.; Thurston, R. V. Environ. Sei. Techno/. 1977. 1 1 . 714-719. (24) Hamiiton,'M. A.; Russo, R. C.; Thurston, R. V. Environ. Sei. Technol. 1978, 12, 417. (25) Junk, G. A.; Richard, J. J. Anal. Chem. 1988, 60, 451-454. (26) Hinckley. D. A.; Bidieman, T. F. Environ. Sci. Techno/. 1989. 23, 995-1000. (27) 'Baker'-10 S E Applications Gukfe; Baker Chemical: Phillipsburg, NJ, 1982; Voi. 1. (28) Landum, P. F.; Hihart, S. R.; Eadie, B. J.; Gardner, W. S. Environ. Scl. Technol. 1904, 18, 187-192.
RECEIVED for review June 7,1990. Accepted October 10,1990. Funds for this work have been provided by the Office of Research and Development and the Office of Water, Enforcement, and Permits of EPA.
Determination of Lead and Other Trace Element Species in Blood by Size Exclusion Chromatography and Inductively Coupled PlasmaIMass Spectrometry Bertold Gercken a n d Ramon M. Barnes* Department of Chemistry, University of Massachusetts, GRC Towers, Amherst, Massachusetts 01003-0035 The comblnatlon of liquid chromatography and lnductlvely coupled plasmalmass spectrometry (ICP/MS) was employed In an explorltory study to determlne lead and other trace element specles In blood components. I n human blood serum, lead was found In at least three molecular weight fractions at >600 000, 260 000, and 140 000. The major part of lead was colncldent wlth the maln copper slgnal at a molecular welgM of 140000. This fractlon, bindlng both copper and lead, was proven to be ceruloplasmlnby the appllcatlon of an Immunological reaction prlor to chromatographlc separatlon. I n rat serum, lead could be detected In four fractlons wlth molecular weights of >600 000,400 000,145 000, and 11000. I n human red blood cell hemolysate, the major fraction of lead was found at 250 000, wlth mlnor fractlons at 140 000 and at 30000 together with Iron In hemoglobin and rlnc In carbonic anhydrase. I n rat red blood cell hemolysate, lead was detected at >600000, 145000, 30000, and 11000. Lead Isotope ratlor were determlned In lead blndlng proteln fractlons wlth a precision of & I O % . The detectlon limit for lead In protein fractlons was 0.15 ~.lgL-'.
INTRODUCTION Coupling an inductively coupled plasma mass spectrometer as the element detector for a high-performance liquid chromatography (HPLC) system in order to achieve element 0003-2700/91/0363-0283$02.50/0
species determination in biological samples offers two major advantages compared to ICP atomic emission spectrometry (ICP-AES). ICP/MS enables sensitive element determination especially of the heavy elements, and single isotopes can be determined and isotopic ratios can be measured. Application of tracer experiments using stable isotopes for specific molecules becomes practical with HPLC-ICP/MS. Although others described the use of HPLC-ICP/MS techniques (I+), none have examined lead or other trace elements species in blood compounds. For determination of trace element containing species in blood compounds, various techniques for element detection have been used in combination with size exclusion chromatography (SEC). These include atomic absorption spectrometry (AAS) (9-13), neutron activation analysis (NAA) (14), and ICP-AES (15,16). Only Lyon and Fell have used SEC prior to ICP/MS measurements of 63Cu and 65Cu in blood serum for separating copper from interfering sodium and phosphate ions (17). Lead speciation studies in blood compounds have been performed by using SEC off-line with graphite furnace AAS. The major lead binding site in erythrocyte was identified as hemoglobin (18-20). Lead appeared in high molecular weight and in low molecular weight (10) fractions. Chromatographic protein separations involving lead-210 as the tracer isotope also were performed by measuring the radioactivity of lead-210 with a yspectrometer (21). 0 1991 American Chemical Society
284
ANALYTICAL CHEMISTRY, VOL. 63, NO. 3, FEBRUARY 1, 1991
Table I. SEC and ICP/MS Instrumental and Operational Parameters SEC
column buffer pH 7.2 flow rate, mL min-' sample volume, pL UV absorbance, nm
ICP/MS rf generator frequency, MHz rf forward power, W reflected power, W distance load coil to sampler orifice, mm end of torch to sampler orifice, mm torch Sciex, "long", mm spray chamber nebulizer argon gas flow rates nebulizer (L min-') auxiliary (L min-') plasma (L min-') Ni sampler and skimmer cones (Perkin-Elmer) orifice diameter, mm ion lens settings, V bessel box barrel (B) einzel (El) plates (P) stop (SI resolution measurements per peak scanning mode measurement mode measurement time. s
Table 11. Isotopes Determined by On-Line ICP/MS after Size Exclusion Chromatographic Protein Separation"
TSK G 3000 SW, 300
mm long, 7 mm diameter 0.1 M Tris/HCl 0.5
100 280 27.12 1300 80000 was degraded by the longer ( 3 4 measurement time. A 1-s measurement time and application of a three-point moving average smoothing process appeared to retain resolution and reduce signal noise. A measurement time of 2 s was also a reasonable compromise in order to reduce the number of collected data points from 320 to 170 for a chromatographic run of 32 min. The chromatogram in Figure 1 shows the lead, copper, and zinc distribution in rat blood serum. Lead could be detected in at least four molecular weight fractions. The total lead concentration measured by isotope dilution ICP/MS (22,27) for the sample was 12.8 f 1.2 pgL-*. Separation of rat red blood cell (RBC) hemolysate also demonstrated that lead could be observed in at least four molecular weight fractions (Figure 2). The total lead concentration measured by isotope dilution (ID) ICP/MS (22,27)for the sample was 45.2 f 4.0
15
Elution volume (mL)
Elution volume (mL)
Figure 1. Chromatogram of 2wPb (solid line), 83Cu (- -), and 64Zn
10
Flgure 2. Chromatogram of '08Pb (solid line), 54Fe and e4Zn (dotted line) distribution in rat RBC hemolysate. 54Fecount rate has (-e-),
been divided by 300 and MZn by 50. Evaluated molecular weights of lead fractions are labeled.
Table 111. Molecular Weight Fractions of Lead-Containing Species and Coeluting Metals Determined in Human and Rat Serum and RBC Hemolysate MW of Pb binding species
human >600 000
trace elements coeluting with Pb
Serum
260000 f 20000 140000 f 15000 rat
>600 000
Zn, Cu
cu Zn, Cu
400000 f 40000
145000 f 15000 11000 f 2000
human
cu Zn
Red Blood Cell Hemolysate
250000 f 20000 140000 f 15000 30000 f 3000
cu cu
Fe, Zn, Cu
rat
>600 000
Zn
140000 f 20000 30000 f 3000 11000 f 2000
Fe, Zn
pgL-'. These results are summarized in Table I11 for only
the major lead peaks. Serum samples were analyzed in quadruplicate and RBC hemolysate samples in duplicate. The calculated molecular weights in Table 111represent the mean values f the standard deviation or the mean deviation, respectively. The column calibration regression equation was y = 7.3636 - 0.3059~with r = 0.9983, and the standard error of the slope was 0.0079. In both rat blood serum and RBC hemolysate, lead was found in the high molecular weight range (>600 000) corresponding to the void volume, in the 140 000-145 000 range, and in a low molecular weight fraction a t 11000. In serum, lead occurred in the void volume together with copper and zinc, alone at 400000, and coeluting with copper and zinc at 145000 and 11000, respectively. Previously, lead in the high molecular weight fraction had not been considered in detail (18, 19). A minor portion of lead also could be detected in the albumin fraction located by the zinc peak at 85000 f 5000. With a standard albumin solution, albumin appeared under the chromatographic conditions used at a higher molecular weight than expected for its molecular weight of 67 000. In RBC hemolysate, lead was detected a t a MW of 30 000 3000 in coeluting fractions with Fe and Zn. These elements are functional components of hemoglobin (Hb) and carbonic
*
286
ANALYTICAL CHEMISTRY, VOL. 63,NO. 3, FEBRUARY 1, 1991
2000
1
l4OkDa '-"
n
250kDa
3OkDa
l4OkDa
0
5
10
0
15
5
$,
10
Elution volume (mL)
Elutlon volume (mL)
15
-
Flgure 4. Chromatogram of 2oBPb(solidline), %u (- -), *e (dashed lines), and 84Zn (dotted lines) distribution in human RBC hemolysate (injected sample volume: 25 pL). "Fe count rate has been divided by 1000, 84Zn by 100, and 03Cu by 10. Evaluated molecular weights of lead fractions are labeled.
1.o
0.5
0.0 5
10
Elution volume (mL),
15
-
Figure 3. (A, top) Chromatogram of *"Pb (solid line), 63Cu (- -), and 'UZn (dashed lines) distribution in human blood serum. %u count rate has been divided by 20 and 'UZn by 60. Evaluated molecular weights of lead fractions are labeled. (B, bottom) Chromatogram of human blood serum with UV absorbance at 280 nm. anhydrase (CAR), respectively. Raghavan et al. (18,19)found lead in a low molecular weight fraction together with Hb in RBC hemolysate. In the chromatogram of human blood serum (Figure 3), lead was observed in three molecular weight fractions (Table 111). The total lead concentration measured by isotope dilution ICP/MS (22,27)was 4.0 f 0.5 pgL-'. The major part of the lead was found coeluting with copper in the 140000 molecular weight peak. In this molecular weight fraction, copper is known to be a functional element of ceruloplasmin (molecular weight between 130000 and 143000) (28). In the chromatogram of human RBC hemolysate (Figure 4), lead appeared in a t least three fractions also listed in Table 111. The total lead concentration measured by ID ICP/MS (22, 27) was 95 200 f 2000 pcg.L-'. The major peak was oberved at 250OOq. and minor fractions appeared at 140000 and 30000. The latter corresponded to Hb and carbonic anhydrase as indicated by the coeluting iron and zinc signals. In none of the human blood compounds was lead found in a low molecular weight fraction as in rat blood serum and RBC hemolysate. The coelution of the major lead fraction in the chromatograms of human and rat serum together with the copper peak of ceruloplasmin at MWs of 140CMX)-145000 may indicate that ceruloplasmin also has a binding site for lead. This might be one possible explanation for the observed interference of lead with the copper and iron metabolism as indicated by lowered copper concentrations and ceruloplasmin activities and inhibited heme synthesis (29, 30). Further evidence for the ceruloplasmin binding site for both copper and lead could be obtained by specific precipitation of the ceruloplasmin from serum with an immunological reaction. The distributions of lead, copper, and zinc (%Zn)in
Y
.
0
5
10
15
Elutlon volume (mL)
400
1
n Elutlon volume (mL)
Figure 5. Chromatogram of *'*Pb (solid line), 83Cu (dashed line), and BsZn (dotted line) in (A, top) human blood serum and (B, bottom) human blood serum treated with human ceruloplamin antiserum. %u count rate was divided by 20 and wZn by 10. human serum before (Figure 5a) and after (Figure 5b) treatment with the human ceruloplasmin antiserum indicate the removal of ceruloplasmin. After precipitation of ceruloplasmin, the copper and lead signals in the 140000 MW fraction disappeared. The zinc distribution was unaffected, since zinc is not bound to ceruloplasmin. The remaining copper and lead in Figure 5b are associated with the zincalbumin fraction a t MW 85 000. This peak was observed as a shoulder of major copper and lead peaks at 140000 in Figure 5a. These results might corroborate the speculation that the lead-sensitive component of copper metabolism involved in
ANALYTICAL CHEMISTRY, VOL.
63,NO. 3,FEBRUARY 1, 1991
287
utilization of iron in hemoglobin synthesis is ceruloplasmin. However, further studies are necessary to test this hypothesis. In all chromatograms of human and rat blood Compounds, magnesium appeared exclusively in the fraction of the electrolytes at an elution volume of 12.6 mL corresponding to a molecular weight of