Determination of platinum in rat dorsal root ganglion ... - Springer Link

9 1999 by Humana Press Inc. All rights of any nature, whatsoever, reserved. 0163-4984/99/6701--0001 $12.75

Determination of Platinum in Rat Dorsal Root Ganglion Using ICP-MS HONG DING,1 MARGARET M. GOLDBERG,1 JAMES H. RAYMER,*,1 JENNIFER HOLMES,2 J A S O N STANKO, 2 AND S T E P H E N G . CHANEY 2

'Research Triangle Institute, Analytical and Chemical Sciences, Research Triangle Park, NC 27709-2194; and 2Department of Biochemistry and Biophysics, School of Medicine, University of North Carolina-Chapel Hill, Chapel Hill, NC 27599-7260 Received J a n u a r y 27, 1998; Accepted April 27, 1998

ABSTRACT This study evaluated the performance of inductively coupled plasma mass spectrometry for the determination of platinum (Pt) in rat dorsal root ganglion. The method detection limit was found to be 0.008 n g / m L of Pt, which corresponds to 4 pg of Pt per milligram of ganglia. The standard deviations in the tissue matrix were 5.7% or better and minimum matrix effect was observed. Compared to indium, the use of iridium or a combination of iridium and bismuth as internal standard(s) provided more accurate measurement. The Pt in the tissue digestate was stable for a minimum of 46 d at levels above 0.05 ng/mL. Flow injection analysis using undiluted digestates resulted in approximately 20% signal enhancement. Internal standard correction was necessary to obtain accurate results. The method was used in initial studies in which rats were dosed with cisplatin and has shown that Pt accumulates and persists in dorsal rat ganglion following treatment. Index Entries: Platinum; cisplatin; neurotoxicity; inductively coupled plasma mass spectrometry; flow injection analysis.

INTRODUCTION P l a t i n u m (Pt) c o m p o u n d s h a v e b e e n u s e d as anticancer agents for almost three decades. T h e y play a m o s t i m p o r t a n t role in treating diseases *Author to whom all correspondence and reprint requests should be addressed. ~Paper presented at the Pittsburgh Conference, Atlanta, GA, March 1997. Biological Trace Element Research

1

Vol. 67, 1999

2

Ding et al.

such as ovarian and bladder cancers. However, because of their nephrotoxicity and neurotoxicity (1,2), the use of Pt drugs during the treatment is dose limiting. There has been substantial interest in searching for the second-generation Pt agents with both high antitumor activity and low or no toxicity. Even though the organs found with the highest Pt accumulation are the kidneys and liver, dorsal root ganglion is the target organ of interest because of the neurotoxicity of Pt. Very low concentrations of Pt in ganglia were expected and this required a very sensitive and robust method to ensure accurate measurement. Inductively coupled plasma mass spectrometry (ICP-MS) is a state-ofthe-art technique widely employed in trace-level elemental analysis. This technique has excellent sensitivity with parts-per-trillion level limits of detection (LODs), multielement and isotope capabilities, and fewer spectral interferences compared to atomic absorption (AA) or atomic emission spectroscopy (AES), rapid sample analysis, and it is applicable to a wide range of sample matrices including those of geological, biological, and environmental origin. The biological applications of ICP-MS cover a wide range of materials from animal tissues (3,4) and body fluids (5) to aquatic plants and humic materials (6-8) and food products (4,6,9,10). Platinum determination in biological media has been investigated using adsorptive voltammetry (11), neutron activation (12), atomic absorption spectroscopy (13,14), ICP-atomic emission spectrometry (13), and ICP-MS (13-19). This article deals with the determination of Pt in rat dorsal root ganglion matrix using ICP-MS. Given that the mass of ganglia that could be collected was minute, a very sensitive detection technique combined with contaminant-free sample-handling procedures were critical to the success of this analysis. In this study, a method for the determination of Pt in rat ganglion was developed and evaluated with regard to method detection limits, matrix effects, and extract stability with respect to storage time.

EXPERIMENTAL The ICP-MS was a VG PlasmaQuad (PQ)-XR unit (Franklin, MA, USA), equipped with a Gilson 222 autosampler (Middleton, WI, USA). A concentric nebulizer and a Scott double-pass glass spray chamber were used (VG, Franklin, MA, USA). Typical operating conditions are listed in Table 1. For experiments utilizing flow injection analysis as the mode of sample introduction, a Waters nonmetallic HPLC 626 system (Milford, MA, USA) fitted with a Rheodyne 9275i injector (Cotati, CA, USA) and a 50-gL sample loop was used. The Pt compound cisplatin [cis-diamminedichloroplatinum(II), CAS registry number 15663-27-1; Sigma, St. Louis, MO, USA] was administered to male Wister rats (Charles River Laboratories, Raleigh, NC, USA) each weighing approximately 250 g at the start of the study. A total of nine injections of 2 m g / k g of cisplatin in 0.9% NaC1 with 20 n g / k g Biological Trace Element Research

Vol. 67, 1999

Pt in Rat Dorsal Root Ganglion

3

Table 1 Typical Operating Conditions for ICP-MS Parameter

Value

Forward Power

1350 W

Reflected Power

< 1W

Coolant Flow

12.5 L/rain

Auxiliary Flow

1.3 L/min

Nebulizer Flow

0.840 L/rain

Spray Chamber Temperature

4.50 C

mannitol (Sigma) were given to each rat twice a week (Mondays and Thursdays). This represented the maximum tolerated dose for cisplatin. The control group was treated in the same manner except that injections were 0.9% NaC1. Three dorsal ganglia (L4-L6, total weight of approximately 10 mg) were removed for analysis from each animal in parallel groups of animals at 24 h, 48 h, or 8 wk following the treatment. Ganglia were placed into 5-mL test tubes with solid-glass flat head stoppers (Kontes, Vineland, NJ, USA) and 500 mL of trace-metal-grade, concentrated (70%) HNO3 (Fisher, Pittsburgh, PA, USA) were added. Samples were digested at 90~ in a water bath for 2 h. At the end of this digestion, samples were cooled to room temperature, 500 mL of trace-metalgrade 30% H202 (J.T. Baker, Phillipsburg, NJ, USA) were added to each digestion tube, and samples were subjected to digestion overnight at 50~ After the completion of the digestion, samples were transferred to clean 10-mL polypropylene autosampler tubes (VWR, Morrisville, NC, USA) or 15-mL polystyrene centrifuge tubes and stored in a refrigerator. The polymeric tubes were cleaned by soaking overnight in 20% HNO3 to leach out any potentially interfering metal ions. Immediately before analysis by ICP-MS, each sample was spiked with 0.25 mL of 1 mg/mL iridium (Ir) and bismuth (Bi) internal standard solution and diluted to 5 mL with de-ionized water (18 Mfl resistance quality). The resulting concentrations of Ir and Bi were 5 ng/mL each. For studies designed to define the best internal standard, different concentrations of Pt, Ir, Bi, and indium (In) were prepared by serial dilution from 1000 mg/mL stock solutions (High-Purity Standards, Charleston, SC, USA). Aqueous calibration standards of 0.01, 0.05, 0.1, 0.5, 1, and 10 ng Pt/mL were prepared in 7% HNO3 and 3% H 2 0 2 prepared from 70% HNO3 and 30% H202. Initial method validation studies were conducted using brain tissues from untreated male Sprague-Dawley rats (Charles River Laboratories) because of the limited availability of ganglia. Brains were homogenized for 10 s without the addition of water or saline. Aliquots of approximately 10 mg of brain homogenate were taken and subjected to the digestion and sample preparation procedures described above. Platinum was added to aliquots of the brain digestates to prepare the matrix calibration standards. All sample handling after the digestion Biological Trace Element Research

Vol. 67, 1999

4

Ding et al.

was performed in a class 100 hood and sample analysis was carried out in class 10,000 cleanroom where the ICP-MS was located.

RESULTS AND DISCUSSIONS

Method Development and Method Performance Selection of Internal Standard There are several criteria to be considered in choosing an internal standard in plasma spectrometry. The ideal internal standard should not be present in the sample; it should be monotopic if possible, stable in the matrix, and have a mass-to-charge ratio similar to the analyte. The internal standard should also undergo ionization processes in the plasma similar to those of the analyte. Elements such as In and yttrium (Y) are the most commonly used internal standards in ICP analysis. However, these elements are more easily ionized than Pt (Table 2) and behave differently than Pt during the atomization, excitation, and ionization (20). Thus, they may not serve as appropriate internal standards for Pt analysis. The potential internal standards were evaluated through analysis of Pt-spiked brain samples containing the different internal standards. The Pt concentrations were calculated based on a calibration curve generated using aqueous Pt standards and adjustment of the concentrations based on the responses of the various internal standards (In, Ir, Bi, Ir/Bi, or no internal standard). The resulting accuracies (bias relative to known, aqueous Pt concentrations) are given in Table 3. At any given Pt concentration, the accuracy for Pt determination was poorest with In internal standard correction. Ir and Bi were better choices for the internal standard than In for Pt determination, although the most accurate results were obtained using no internal standard correction. This suggests that there was a negligible matrix effect (see the following subsection). At a Pt concentration of 0.01 ng/mL, the accuracy (recovery) of Pt measurement was 77.2% at best. This indicates that Pt concentrations of 0.01 n g / m L in the matrix will not be reliably measured. However, at the higher concentrations tested (0.05 n g / m L and greater), recoveries were essentially quantitative.

Matrix Effect The calibration curves (linear regression) obtained using standards of 0.01, 0.05, 0.1, 1, and 10 n g / m L of Pt prepared in aqueous and brain matrices are given below. The instrument responses for Pt in the two matrices differed by less than 3%; this indicates minimum matrix interference, as suggested above. Therefore, the aqueous calibration was used for all subsequent analyses. Aqueous regression: y = 5816x - 6.50, r = 1.00 Matrix regression: y -- 5961x - 32.3, r -- 1.00 Slope matrix/slope aqueous = 102.5%

Biological Trace Element Research

Vol. 67, 1999


5

Table 2 Properties of Potential Internal Standards for Use in Pt Determination Ionization Potential (eV)

Element

Mass / Charse

Abundance

Y

89

100.0%

In

115

95.7%

5.8

Ir

193

62.7%

9.0

Pt

195

33.8%

9.0

Bi

209

100.0%

7.3

6A

Source: Ref. 21

Table 3 Accuracy of Pt Determination Using Different Internal Standard(s) Correction Percent Recovery Relative to Nominal for Different Elements Nominal [Pt] Conc. (ns/mL)

In

0.01

55.3

0.05

87.8

101

104

101

0.10

90.6

104

107

104

1.00

91.0

103

106

104

10.00

92.3

105

108

105

Ir

Bi

70.3

73.4

Ir/Bi 70.7

no internal standard 77.2 97.1 99.8 101 99.2

Note: Each standard present at 5 n g / m L .

where y is the instrument response (counts/s) and x is the concentration (ng Pt/mL).

Detection Limit The peak jumping (PJ) detection mode is often used because it allows more integration time on each ion and this results in better counting statistics, especially for lowqntensity peaks, compared to those obtained when the instrument scans the mass region of interest. These two modes of detection were compared and the precisions obtained at each concentration are shown in Table 4. Table 4 clearly shows that PJ provided better precision than scan mode, especially at Pt concentrations less than 2.0 ng/mL. The peak jumping mode was used for the detection of Pt in the remainder of this work. The most abundant Pt isotope (195pt, 33.8% natural abundance) was used for quantification. However, Table 4 shows that there was no appreciable difference in instrument sensitivity between 194pt (32.9% natural abundance) and 195pt. The method precision over the calibration range (from 0.05 to 10 n g / m L of Pt) was evaluated using triplicate preparations of standards in the brain matrix at three concentrations. Precision, expressed as percent relative standard deviation (% RSD), was 5.7% or better (Table 5). The calculated limit of detection (three times the standard deviation of the


VoL 67, 1999

Ding et al.

6

Table 4 Precision for Pt Determination by Different Detection Modes Nominal[Pt]

%RSDfor194pt

%RSDfor195pt

(ng/mL)

Scan

PJ

Scan

PJ

0.01 0.~ 0.1

105.4 10.1 15.0

7.8 2.4 3.6

23.2 17.9 12.4

4.8 3.2 1.2

0.2

12.5

2.2

8.7

0.6

2.0

4.5

0.2

3.5

0.8

5.0

0.7

0.5

5.2

0.1

10.0

0.9

0.3

1.8

0.5

Note: Scan = scan m o d e of detection; PJ = peak jumping. Percent relative standard deviation (% RSD) calculated from triplicate analyses.

Table 5 Determination of Pt in Spiked Brain Matrix Spiked Pt Concentration 0.05 ng/mL 0.10 ng/mL 10 ng/mL

% Recovery

% RSD~

97.0 96.2 95.7

5.7 3.6 2.9

aFor triplicate preparations.

0.05-ng/mL spike level divided by the slope of the calibration curve) was found to be 0.008 ng Pt/mL or 4 pg P t / m g of brain tissue. This limit of detection is superior to that reported by Rietz et al. (12), who used neutron activation for the determination of Pt in rat ganglia. The LOD reported for that work was 0.01 gg per 0.01-0.08 g of neural tissue. The method described here is substantially faster. The LOD calculated for the current work also compares favorably with ICP-MS methods for Pt in plasma (17), tissue (13), and cerebrospinal fluid (19), for which reported LODs were 0.05 ng/mL, 0.05 ng/mL, and 0.02 ng/mL, respectively.

Solution Stability The Pt standards prepared in the digested brain matrix were stored in the refrigerator and analyzed at 0, 3, 8, 18, and 46 d to assess the stability of trace amounts of Pt in the matrix with respect to storage time. Figure 1 shows that Pt at concentrations of 0.05 n g / m L and higher were stable over a storage period of at least 46 d.

Memory Effect The adsorption of Pt within the sample introduction lines and nebulizer would create errors in trace-level Pt analyses should the metal be released and transported to the plasma during analysis of an unknown. The potential for sample memory was evaluated in this sysBiological Trace Element Research

Vol. 67, 1999


7

Fig. 1. Stability of Pt in the brain matrix.

tern. A 10 n g / m L Pt solution was introduced and this was followed by the introduction of the sample diluent (a solution of 7% HNO3 and 3% H202). This solution is used as the rinse solution between each sample analysis. The washout time (time necessary for the signal to be reduced to 10% of maximum peak height) for the 10-ng/mL Pt solution was less than 30 s. Nebulization of 5% aqua regia solution did not indicate any residual Pt accumulation in the sample introduction system. Thus, retention of Pt within the ICP-MS system would not be expected to impact subsequent analyses.

Sample Analysis Data obtained from the analysis of dorsal ganglia from dosed and control rats are shown in Table 6. Accumulation of Pt in the dorsal root ganglia is evident. The average Pt concentration in dorsal root ganglia measured 24-48 h after completion of dosing was 0.800 ng/mg; this value was significantly different from control animals (Student's t-test, p -- 0.001). The concentrations are consistent with those measured by Rietz and coworkers (12). In their work, the rats were treated once a week for 7-12 wk at the same dose level as used in this study (2 mg/kg). Dorsal root ganglia were dissected at the end of treatment and subjected to neutron activation analysis. The average Pt level in rat dorsal root ganglia ranged from 0.342 to 1.45 n g / m g ganglia with an average of 0.817 ng/mg. In the current study, elevated concentrations of Pt were still detected 8 wk after treatment termination and were approximately 40% of the Pt levels measured 24-48 h following treatment. A study to determine the neurotoxicity Biological Trace Element Research

Vol. 67, 1999

8

Ding et al. Table 6 Measured Pt Concentrations in Rat Dorsal Root Ganglia Following Dosing with Cisplatin Dose Group Saline control

24-48 hrs following treatment

8 weeks following treatment

Rat No.

Measured [Pt] (n$/m$ tissue)

1

0.127

2

0.123

3

0.141

4

0.134

5

0.124

6

0.107

Average

0.127 + 0.005

1 2

0.553 1.255

3 4

0.646 0.507

5

0.861

6

0.975

Average

0.800 • 0.118

1

0.386

2

0.349

3

0.338

4

0.154

5 6

0.387 0.289

Average

0.317 • 0.036

Note: The Pt concentration reported for treated groups were corrected for background Pt concentration in saline control groups.

and persistence of cisplatin in the dorsal root ganglia is currently being conducted.

Flow Injection Analysis Compared to the conventional continuous sample introduction, flow injection analyses (FIA) requires only microliter volumes of sample and provides increased speed of analysis. Given low sample volume and the dilution needed for the continuous sample nebulization, FIA could allow the use of undiluted samples and thus lower the LOD. Figure 2 shows a calibration curve using Pt spikes (0.01, 0.05, 0.1, 0.5, 1, and 10 ng/mL) prepared in pooled brain digestates without dilution. Signal enhancement from 179% to 147% relative to the aqueous calibration curve was observed (Fig. 3) and could be related to the higher concentrations of both nitric acid and hydrogen peroxide in the undiluted samples as compared to the diluted samples. In general, concentrated acids cause a


Vol. 67, 1999


9

200000.

150000 .o .ar

100000-

a. 50000-

i 2

i

i 4

r

i

i

6

i 8

i

10

Concentration (ng/mL)

Fig. 2. Matrix calibration using FIA. Calibration range: 0.01-10 ng/mL of Pt; r = 0.9997.

Fig. 3. Accuracy of Pt determination in tissue matrix using aqueous calibration curve (FIA). decrease in sensitivity for most elements (22,23), whereas salt ions can lead to both signal enhancement and suppression (24,25). In this case, the higher concentration of h y d r o g e n peroxide might have caused changes in the ionization environment in the plasma that resulted in the enhanced signal. Internal standard correction with Ir was necessary to improve the accuracy to 98-119% of the nominal concentrations (Fig. 3). The limit of detection was calculated to be 0.003 n g / m L (three times the standard deviation of the 0.01-ng/mL spike level divided by the slope of the calibration curve) or an absolute detection limit of 0.3 pg P t / m g brain tissue. This represents a modest improvement over the continuous nebulization (LOD of 4 pg P t / m g tissue) as sample introduction.


Vol. 67, 1999

10

Ding et al.

CONCLUSIONS An ICP-MS method for the determination of Pt in dorsal root ganglia from rats was developed and shown to be more sensitive than methods previously described. The LOD was calculated to be 0.008 ng/mL (4 p g / m g ganglia). The method was applicable to the study of Pt in dorsal root ganglia from rats dosed with cisplatin where accumulation was shown to occur. Elevated concentrations of Pt were still measurable 8 wk after the completion of dosing suggesting the utility of this method in studies designed to determine the persistence of Pt in this or other tissues. The use of flow injection analysis resulted in an improvement of the LOD (0.003 ng/mL, 0.3 p g / m g tissue).

ACKNOWLEDGMENTS The authors thank the Research Triangle Institute and the National Institute of Health through UNC grant number CA 55326 for their financial support.

REFERENCES 1. S. W. Thompson, L. E. Davis, M. Kornfield, R. D. Hilgers, and J. C. Standefer, Cisplatin neuropathy, Cancer 54, 1269-1275 (1984). 2. R. W. Gregg, J. M. Molepo, V. J. A. Monpetit, M. Z. Mikael, D. Redmond, M. Gadia, et al., Cisplatin neurotoxicity: the relationship between dosage, time, and platinum concentration in neurologic tissues, and morphologic evidence of neurotoxicity. J. Clin. Oncol. 10, 795-803 (1992). 3. S. Munro, L. Ebdon, and D. J. McWeeny, Applications of inductively coupled plasma mass spectrometry for trace metal determination in foods, J. Anal. Atom. Spectrom. 1, 211-219 (1986). 4. R. C. Hutton, Technical information, PQ704, VG, Isotopes Ltd., Winsford, Cheshire, UK (1986). 5. S. Branceh, L. Ebdon, M. Ford, M. Foulkers, and P. O'Neill, Determination of arsenic in sample with high chloride content by inductively coupled plasma spectrometry, J. Anal. Atom. Spectrom. 6, 151-154 (1991). 6. C. J. Pickeford and R. M. Brown, Comparison of ICP-MS and ICP-ES: detection power and interference effects experienced with complex matrices, Spectrochim. Acta 41B(1-2), 183-187 (1986). 7. E. M. Thurman and R. L. Malcohn, Preparative isolation of aquatic humic substances, Environ. Sci. Technol. 15, 463-466 (1981). 8. H. E. Taylor and J. R. Garbarino, Occurrence and distribution of selected trace metals in the international humic substances society's standard and reference fulvic and humic acids isolated from the Suwannee River, U.S. Geological Survey Open File No. 87-557 (1990). 9. A. W. Boom, J. E. Fulford, and W. Wegscheider, Determination of trace elements in organic materials by inductively coupled plasma spectrometry, Mikrochim. Acta 2, 171-178 (1985). 10. J. R. Dean, R. C. Massey, and L. Ebdon, Selection of mode for the measurement of lead isotope ratios by ICP-MS and its application to milk powder analysis, ]. Anal. Atom. Spectrom. 2, 369-374 (1987).


Vol. 67, 1999


11

11. O. Nygren, G. T. Vaughan, T. M. Florence, G. M. P. Morrison, L. M. Warner, and L. S. Dale, Determination of platinum in blood by adsorptive voltammetry, Anal. Chem. 62, 1637-1640 (1990). 12. B. Rietz, J. Heydorn, and A. Krarup-Hansen, Determination of platinum by neutron activation analysis in nerve tissue from rats treated with cisplatin, Biol. Trace Element

Res. 47, 343-350 (1994). 13. T. Minami, M. Ichii, and Y. Okazaki, Comparison of three different methods for measurement of tissue platinum level, Biol. Trace Element Res. 48, 37-44 (1995). 14. P. Tothill, L. M. Matheson, J. E Smyth, and K. McKay, Inductively coupled plasma mass spectrometry for the determination of platinum in animal tissues and a comparison with atomic absorption spectrometry, J. Anal. Atom. Spectrom. 5, 619-622 (1990). 15. D. C. Gregoire, Determination of platinum, palladium, ruthenium and iridium geological materials by inductively coupled plasma mass spectrometry with sample introduction by electrothermal vaporization, J. Anal. Atom. Spectrom. 3, 309-314 (1988). 16. B. J. Perry, R. R. Barefoot, and J. C. Van Loon, Inductively coupled plasma spectrometry for the determination of platinum group elements and gold, Trends Anal. Chem. 14, 388-397 (1995). 17. P. Allain, S. Berre, Y. Mauras, and A. Le Bouil, Evaluation of inductively coupled plasma spectrometry for the determination of platinum in plasma, Biol. Mass Spectrom. 21, 141-143 (1992). 18. B. J. Perry and R. E. Balazs, ICP-MS methods for the determination of platinum in suspensions of cells exposed to cisplatin, Anal. Proc. lnclud. Anal. Commun. 31, 260-271 (1994). 19. B. Casetta, M. Roncadin, G. Montanari, and M. Furlanut, Determination of platinum in biological fluids by inductively coupled plasma spectrometry, Spectroscopy 12, 81-86 (1991). 20. A. Montaser and D. W. Golightly, Inductively Coupled Plasmas in Analytical Atomic Spectrometry, 2nd ed., VCH, New York, Chap. 13 (1992). 21. D. R. Lide, CRC Handbook of Chemistry and Physics, 72nd ed., CRC, Boca Raton, FL (1991). 22. S. Greenfield, H. M. McGreachin, and P. B. Smith, Nebulization effects with acid solutions in ICP spectrometry, Anal. Chim. Acta 84, 67-78 (1976). 23. E. G. Chudinoz, I. I. Ostroukhova, and G. V. Varvanina, Acid effects in ICP-AES, Fresenius Z. Anal. Chem. 335, 25-33 (1989). 24. M. Thompson and M. H. Ramsey, Matrix effects due to calcium in inductively coupled plasma-atomic emission spectrometry, Analyst 110, 1413-1422 (1985). 25. S. R. Koirtyohann, J. S. Jones, C. P. Jester, and D. A. Yates, Use of spatial emission profiles and a nomenclature system as aids in interpreting matrix effects in the low power argon inductively coupled plasma, Spectrochim. Acta 36B, 49-59 (1981).


Vol. 67, 1999