544
Chem. Res. Toxicol. 1998, 11, 544-549
Carbon Disulfide and N,N-Diethyldithiocarbamate Generate Thiourea Cross-Links on Erythrocyte Spectrin in Vivo John C. L. Erve,*,† Venkataraman Amarnath,† Doyle G. Graham,† Robert C. Sills,‡ A. L. Morgan,‡ and William M. Valentine† Department of Pathology and Center in Molecular Toxicology, Vanderbilt University Medical Center, Nashville, Tennessee 37232-2561, and Environmental Toxicology Program, NIEHS, Research Triangle Park, North Carolina Received January 8, 1998
CS2, a known neurotoxicant, is used in the viscose production of rayon and is also a decomposition product of N,N-diethyldithiocarbamate, a metabolic product of the drug disulfiram used in alcohol aversion therapy. Previous in vitro investigations have demonstrated the ability of CS2 to cross-link proteins through thiourea, dithiocarbamate ester, and disulfide structures. Although in vivo studies have supported protein cross-linking as both a mechanism of neurotoxicity and a potential biomarker of effect, the chemical structures responsible for CS2-mediated protein cross-linking in vivo have not been elucidated. In the present study, the structure of one type of stable protein cross-link produced on erythrocyte spectrin by CS2 in vivo is determined. Rats were exposed to 50, 500, and 800 ppm CS2 for 13 weeks by inhalation or to 3 mmol/kg N,N-diethyldithiocarbamate administered orally on alternating days for 8 weeks. Erythrocyte spectrin preparations from control and exposed rats were hydrolyzed using 6 N HCl and separated by size-exclusion chromatography. The fraction that coeluted with the synthetic deuterated lysine-lysine thiourea internal standard was derivatized with 3-[4′-[(N,N,N-trimethylamino)ethylene]phenyl] 2-isothiocyanate and analyzed by liquid chromatography tandem mass spectrometry using selected reaction monitoring detection. Lysine-lysine thiourea was detected in spectrin preparations obtained from CS2-treated rats at 500 and 800 ppm and N,N-diethyldithiocarbamate-treated rats, but not from controls. These results establish that CS2-mediated protein cross-linking occurs in vivo through the generation of Lys-Lys thiourea and that diethyldithiocarbamate can, through in vivo release of CS2, produce the same cross-linking structure. This observation supports the utility of cross-linking of peripheral proteins as a specific dosimeter of internal exposure for CS2 and provides a mechanistic explanation to account for the high-molecular-weight neurofilament protein species isolated from rats exposed to CS2 or N,N-diethyldithiocarbamate.
Introduction Today CS2 is used in the viscose production of rayon and cellophane during which approximately 76 million pounds are released into the atmosphere annually in the United States. A second major source of CS2 results from the class of chemicals known as dithiocarbamates that can decompose to the parent amine with the generation of CS2. Therefore, in addition to rayon workers, people involved with the production or use of dithiocarbamates and dithiocarbamate disulfides as pesticides, or for medical purposes, may also be exposed to CS2. Exposure to CS2 can result in damage to the liver (1), kidney, and nervous system (2, 3), and epidemiological evidence suggests that workers chronically exposed to CS2 are also at increased risk of developing cardiovascular disease (4).
* To whom correspondence should be addressed. † Vanderbilt University Medical Center. ‡ NIEHS.
N,N-Diethyldithiocarbamate (DEDC)1 is a known neurotoxicant, and previous investigations have supported release of CS2 as contributing to its toxicity (5). In light of these facts, development of an accurate and specific biomarker to assess exposure to CS2 would be useful in protecting workers from adverse consequences. Development of a biomarker of effect has been hindered by a lack of understanding regarding the mechanism of CS2 toxicity. CS2-mediated covalent cross-linking of neurofilament proteins has been proposed as a mechanism of neurotoxicity, producing a central peripheral distal axonopathy, identical to that produced by n-hexane (6). Investigations using NMR spectroscopy have characterized several protein cross-linking structures formed in vitro by CS2 including dithiocarbamate ester, thiourea, and disulfide bridges (7), and in vivo studies have 1 Abbreviations: ACN, acetonitrile; BSA, bovine serum albumin; DEDC, N,N-diethyldithiocarbamate; ESI, electrospray ionization; LysLys thiourea, di--lysylthiourea; PETAPITC, 3-[4′-[(N,N,N-trimethylamino)ethylene]phenyl] 2-isothiocyanate; SRM, selected reaction monitoring; TFA, trifluoroacetic acid.
S0893-228x(98)00007-1 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/21/1998
Thiourea Cross-Linking of Spectrin in Vivo
detected high-molecular-weight species of erythrocyte spectrin and neurofilament proteins following exposure to CS2 or DEDC. Evidence has been presented that the extent of cross-linking of both erythrocyte spectrin and globin reflects the degree of cross-linking occurring within the axon (8). One advantage of measuring cross-linked spectrin over current techniques used to assess CS2 exposure is that it appears applicable to chronic exposure scenarios with the ability to integrate exposure over the life of the erythrocyte (120 days in humans, 60 days in rats). Previous studies on CS2 (9) and DEDC (5) have demonstrated a cumulative dose response for CS2-mediated intermolecular cross-linking of spectrin, and intramolecular covalent modification of globin, that correlated with the amount of neurofilament protein crosslinking occurring in the spinal cord. Moreover, detection of spectrin and globin cross-linked species is possible before manifestations of neurotoxicity and thus appears suitable as a preneurotoxic marker of exposure. Although the cumulative dose response and temporal relationship of protein cross-linking to morphologic changes in the nervous system are consistent with protein cross-linking being a direct effect of CS2 that contributes to the formation of axonal swellings, identification of the CS2-derived protein cross-linking structures formed in vivo has not been performed. Identification of the protein cross-linking structures produced by CS2 in vivo would support the relevance of protein cross-linking chemistry previously elaborated in vitro and could help define the relationship between modification of peripheral proteins and those occurring within the nervous system. The purpose of the present work was to detect the proposed major cross-linking structure present on spectrin from both CS2 (inhalation)and DEDC (po)-exposed rats. To this end, we used a combination of liquid chromatography and electrospray ionization (ESI) mass spectrometry to detect the presence of Lys-Lys thiourea cross-links on erythrocyte spectrin.
Materials and Methods Chemicals. Constant boiling hydrochloric acid solution (6 N), trifluoroacetic acid (TFA), and amino acid standard mixture were from Sigma (St. Louis, MO). HPLC grade acetonitrile (ACN) and water were purchased from EM Science (Gibbstown, NJ). Triethylamine was distilled before use. 3-[4′-[(N,N,NTrimethylamino)ethylene]phenyl] 2-isothiocyanate (PETAPITC) was a gift from Ruedi Aebersold, Department of Molecular Biotechnology, University of Washington. DL-[4,4,5,5-d4]Lysine was purchased from Isotec (Miamisburg, OH) and 13CS2 from Cambridge Isotope (Cambridge, MA). [4,4,5,5-d4]-Nr-Acetyllysine. Sanger’s original synthesis was scaled down by carrying out the first four reaction steps in 12-mL centrifuge tubes (10). DL-d4-Lysine dihydrochloride (500 mg, 2.25 mmol) and CuCO3‚Cu(OH)2 (275 mg, 1.25 mmol) were heated in water (3 mL) at 60 °C for 1 h and cooled. The mixture was centrifuged, and the residue was washed with water (2 × 1 mL) and centrifuged. The supernatants were combined, concentrated to 1 mL, added to absolute ethanol (5 mL), and cooled to obtain the copper-lysine complex. The dried complex was dissolved in 2 N NaOH (1.2 mL), cooled on ice, and treated with benzyl chloroformate (10 × 45 µL) and 2 N NaOH (10 × 225 µL) over 40 min. The precipitated solid was centrifuged, washed with water (2 × 20 mL), and dried. It was suspended in water (7 mL) and heated at 65 °C. Disodium EDTA (1 g) was added over 15 min, and heating was continued for 10 min more. After cooling the DL-[4,4,5,5-d4]--(benzyloxycarbonyl)lysine was collected by filtration, washed with water until it
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 545 was white, and crystallized from water; 490 mg; m/z 285.0 (285.2). It was dissolved in 1 N NaOH (1.8 mL) and acetylated with acetic anhydride (10 × 30 µL) and 1 N NaOH (10 × 400 µL) over 30 min. The reaction mixture was extracted with ether (2 × 4 mL), cooled in ice, acidified to pH 3 with 2 N HCl, and extracted with ethyl acetate (3 × 5 mL). The extracts were combined, dried, and evaporated. The residue in ethanol (15 mL) was stirred under hydrogen in the presence of 10% Pd/C (35 mg). After 16 h, the catalyst was removed by filtration through a bed of Celite, and the filtrate was evaporated to obtain dl-[4,4,5,5-d4]-NR-acetyllysine (145 mg, 35%); m/z 193.0 (193.2). [d8,13CdS]Bis(Nr-acetyllysyl)thiourea. DL-[4,4,5,5-d4]-NRAcetyllysine (75 mg, 0.4 mmol) was dissolved in water (4 mL) and ethanol (1 mL), and 13CS2 (50 µL, 0.8 mmol) and 1 N NaOH (0.5 mL) were added. The reaction mixture was stirred for 8 h at room temperature, heated in an oil bath at 110 °C for 12 h, and cooled. The pH was adjusted to 3.5 with 1 N HCl and concentrated to 1 mL before purifying on a column of silica gel (20% water in ACN). The fractions were checked by UV scan (300-210 nm). Those exhibiting a maximum at 235 nm were combined and lyophilized (40 mg); m/z 428.4 (428.3). Animals and Exposures. Animal studies were conducted in accordance with the NIH Guide For Care and Use of Animals and approved by the institutional animal use and care committee. For the CS2 exposures, male Fisher 344 rats (Charles River Breeding Laboratories, Raleigh, NC) were used. The exposure levels were control, 50, 500, and 800 ppm for a duration of 13 weeks. Carbon disulfide, purity > 99%, was mixed with conditioned air and introduced into the inhalation chambers 5 days/week; daily exposure time was 6 h. Actual chamber CS2 concentrations were monitored continuously by infrared spectroscopy so that mean daily exposures were within 3% of desired concentrations. Further details can be found elsewhere (11). For the DEDC exposures, male Sprague-Dawley rats (Harlan, Sprague-Dawley, Indianapolis, IN) were used. The rats were administered 3 mmol/kg every other day orally for 8 weeks. DEDC concentrations were checked spectrophotometrically; purity was >99%. Dosing was accomplished by dissolving DEDC in 0.1 M phosphate buffer, pH 7.5, and introducing the solution into the rat by intragastric gavage once a day on alternate days throughout the exposure period. Further details can be found elsewhere (12). Cross-Linking of Bovine Serum Albumin (BSA) by [13CdS]DEDC. A 4% (w/v) solution of BSA was incubated at 37 °C with 25 mM [13C]DEDC for 11 days as described previously (7). At intervals, the protein was monitored by 13C NMR spectroscopy which revealed the presence of thiourea, as deduced from the isotopically enriched resonance at 180 ppm. Hydrolysis of Spectrin and Cross-Linked BSA. Isolation of erythrocyte spectrin was performed as described previously (9). Aliquots of 20 µL of BSA, or from 0.6 to 2.2 mg of spectrin, were placed in a hydrolysis tube (6 × 50 mm) and hydrolyzed for approximately 18 h at 110 °C under vacuum on a Picotag workstation (Waters, Milford, MA). Following removal of residual HCl, samples were reconstituted with 95 µL of 10% ACN, 0.1% TFA and filtered through a disposable Centrex MF 0.4 microcentrifuge filter with 0.2-µm pores (Schleicher and Schuell, Keene, NH). Derivatization of Lys-Lys Thiourea with PETAPITC. Enrichment of Lys-Lys thiourea was accomplished by sizeexclusion HPLC with a G2500PWXL column (300 × 7.8 mm i.d., 6 µm; TosoHaas, Montgomeryville, PA). The elution conditions were isocratic with 10% ACN, 0.1% TFA at a flow of 1 mL/min. Eluant was monitored at 232 nm which is near the UV maximum of the thiourea bond. For each analysis, 4 µL of [d8,13C]Lys-Lys thiourea (5 mM) internal standard was spiked into the hydrolysate (1200 nmol/µL), which was then injected onto the column. This amount of internal standard was detected easily and collected, along with any coeluting Lys-Lys thiourea present in the treated spectrin or BSA. The liquid chromatography system included a Waters 996 PDA detector and 2690 liquid chromatograph (Waters, Milford, MA) equipped with an
546 Chem. Res. Toxicol., Vol. 11, No. 5, 1998 autosampler coupled to a fraction collector (Foxy Jr., Isco, Lincoln, NE) which was programmed to collect the fraction containing Lys-Lys thiourea. Fractions containing Lys-Lys thiourea were derivatized in 1.5mL Eppendorf tubes. Briefly, samples were reconstituted with 25 µL of solvent [50:49:1 (v:v:v) methanol-water-ethyl acetate], and the pH was adjusted to approximately 9 by the addition of 3 µL of triethylamine. Next, 3 µL of a 100 mM solution of PETAPITC was added, followed by heating at 60-65 °C for 15 min. Samples were dried by rotoevaporation with a Speed-vac (Savant Instruments, Holbrook, NY) before the addition of 5 µL of 50:50 TFA-water and further heating at 45 °C for 30 min. The samples were dried and then reconstituted with 55 µL of 5% ACN, 0.1% acetic acid, 0.02% TFA. Mass Spectrometry. Liquid chromatography tandem mass spectrometry (LC/MS/MS) analysis of derivatized Lys-Lys thiourea with selected reaction monitoring (SRM) detection was performed with an HP liquid chromatography system (Waters, Milford, MA) in conjunction with a Finnigan TSQ 7000 triplequadrupole ESI-MS (Finnigan, San Jose, CA). SRM provides a method to specifically detect a compound of interest with high sensitivity. Briefly, the first quadrupole (Q1) selects for the mass-to-charge ratio corresponding to the precursor ion of interest and allows it to enter the second quadrupole (Q2, RFonly) which serves as a collision cell. The third quadrupole (Q3) monitors only for selected fragment ions formed in the collision cell and thus provides high sensitivity and specificity. Forty microliters of sample was chromatographed on a reverse-phase C18 column (1 × 50 mm i.d., 5 µm; Monitor, Ontario, CA) with a linear elution gradient from 17% B to 35.6% B in 5.5 min with a flow of 50 µL/min [solvent A ) ACN/H2O/acetic acid/TFA (5: 95:0.1:0.01, v/v/v/v); solvent B ) ACN/H2O/acetic acid/TFA (80: 20:0.1:0.01, v/v/v/v)] and the eluant directed into the mass spectrometer. The electrospray voltage of the atmospheric pressure ionization positive ion source was maintained at 4.2 kV and the capillary temperature fixed at 200 °C; nebulizer pressure was 80 psi. Collision-induced dissociation occurred in Q2, with argon as a collision gas (2.8 mT) at a collision energy of 22 eV (laboratory frame of reference). SRM experiments were conducted by monitoring the m/z 370.3 f 391.3 transition (m/z 374.8 f 396.3 for the internal standard). For the Lys-Lys thiourea cross-link, the doubly charged molecular ion at m/z 370.3 (m/z 374.8 for the internal standard) was selected in Q1 and the product ion at m/z 391.3 (m/z 396.8 for the internal standard) was monitored in Q3. The total scan time was 1.4 s. To increase sensitivity, resolution was decreased by raising the voltages on Q1 (parent resolution, 9.4 V) and Q3 (daughter resolution, 4.8 V).
Results Size-Exclusion Analysis of Amino Acid Hydrolysates and Lys-Lys Thiourea. Since size-exclusion chromatography with a Biorad P2 gel had been used previously to separate a variety of Lys-norleucine crosslinks following nonenzymatic hydrolysis of collagen (13), we attempted to use this technique in our investigation. Size exclusion takes advantage of the larger size of the cross-linked amino acids relative to the other amino acids, resulting in the cross-linked amino acids eluting first. Our initial attempts to utilize this technique failed to separate adequately Lys-Lys thiourea from the other amino acids. Use of another size-exclusion column (see Materials and Methods) compatible with HPLC provided greater resolution and did allow Lys-Lys thiourea to be separated from the other amino acids. Chromatography of an amino acid standard mixture resulted in two major peaks at 8 and 17 min when monitored at 232 nm. Interestingly, although Lys-Lys thiourea has the greater molecular weight, it eluted at 11.2 min, between these
Erve et al.
Figure 1. Size-exclusion HPLC analysis of an amino acid hydrolysate obtained from a CS2-treated rat without internal standard (A) or with [d8,13C]Lys-Lys thiourea internal standard that elutes at 11.2 min (B). The first peak (8 min) and last peak (17 min) represent amino acids, while the identity of the peak at 10 min is unknown. The chromatogram represents elution of 232-nm-absorbing material from the size-exclusion column as described in Materials and Methods.
two peaks, and was sufficiently resolved to allow isolation. This observation might be explained by additional column-analyte interactions, such as charge, so that elution order is not determined by size alone. Spectrin from either control or exposed rats displayed an additional peak not seen in the amino acid standard mixture eluting at approximately 10 min that was not characterized (Figure 1A). Nevertheless, the [d8,13C]LysLys thiourea internal standard was still well-resolved in the spectrin hydrolysates enabling its collection along with coeluting Lys-Lys thiourea originating from the spectrin (Figure 1B). Liquid Chromatography Mass Spectrometric Analysis of Derivatized Lys-Lys Thiourea. Purified Lys-Lys thiourea was derivatized with PETAPITC, a first-generation novel Edman reagent containing a quaternary amine group allowing for efficient ionization and sensitive detection by ESI-MS (14). Authentic derivatized Lys-Lys thiourea (calculated mass 738.6 Da) produced a doubly charged molecular ion at m/z 370.3, which when fragmented by collision with argon produced several characteristic ions (Figure 2). The major ions at m/z 391.3 and 349.3 result from cleavage at the thiourea bond. The internal standard had a molecular weight 9 Da higher resulting in fragment ions at m/z 396.3 and 353.3. In our SRM analysis, the most prominent fragment ion at m/z 391.3 and the corresponding fragment at m/z 396.3 from the internal standard were measured. The retention time of PETAPITC-derivatized Lys-Lys thiourea was approximately 3 min when chromatographed on a C18 reverse-phase column. The lower limit of detection was approximately 1 pmol. The analytical strategy was verified using BSA containing Lys-Lys thiourea cross-links generated by 13CS2 that were identified previously by 13C NMR (7). Figure 3 shows representative reconstructed ion chromatograms
Thiourea Cross-Linking of Spectrin in Vivo
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 547
Figure 2. Tandem mass spectrum of the [M]2+ parent (m/z 370.3) of Lys-Lys thiourea derivatized with PETAPITC. The fragment ions at m/z 391.3 and 349.3 arise from cleavage at the thiourea bond. The spectrum was obtained on a Finnigan TSQ7000 triplequadrupole mass spectrometer.
Figure 3. LC/MS/MS with SRM of BSA and BSA incubated with CS2. The upper panel in each set of ion chromatograms shows the monitored fragment ion m/z 396.3 arising from the derivatized internal standard, [d8,13C]Lys-Lys thiourea (m/z 374.8). The lower traces show the absence of Lys-Lys thiourea in the control BSA sample (A) and the presence of derivatized Lys-Lys thiourea in the treated sample (B).
of control (Figure 3A) and cross-linked (Figure 3B) BSA which reveal a signal in the CS2 cross-linked BSA having the same retention time as the [d8,13C]Lys-Lys thiourea internal standard. Reconstructed ion chromatograms obtained from representative control rats produced a negligible signal for Lys-Lys thiourea (Figure 4A). In contrast, a reconstructed ion chromatogram from a representative 800 ppm CS2-treated rat shown in Figure 4B revealed a peak with signal-to-noise ratio greater than 10:1 having the same retention time as the internal standard. Spectrin samples obtained from rats exposed to CS2 by inhalation at control, 50, 500, or 800 ppm for 13 weeks
Figure 4. LC/MS/MS with SRM of a typical control rat and CS2-treated rat. The internal standard, [d8,13C]Lys-Lys thiourea (m/z 374.8), is shown in the upper trace of each panel. The lower traces show the absence of Lys-Lys thiourea in the control sample (A) and the presence of Lys-Lys thiourea in the sample obtained from the CS2-exposed rat (B).
(Figure 5) or to DEDC at 3 mmol/kg every other day for 8 weeks were analyzed for the presence of Lys-Lys thiourea. The response for Lys-Lys was normalized to both the internal standard and amount of protein hydrolyzed and multipled by an arbitrary factor of 10 000. Although there was no statistical difference between the adjacent groups, there was a statistical difference (p < 0.05) between the controls and the two highest groups (500 and 800 ppm). The three DEDC-exposed rats also showed the presence of Lys-Lys thiourea in quantities significantly greater than control (X ( SE; 17.9 ( 6.1 vs 3.3 ( 0.9; p < 0.05).
Discussion The mechanism of protein cross-linking produced by CS2 was investigated previously in vitro using BSA and
548 Chem. Res. Toxicol., Vol. 11, No. 5, 1998
Erve et al. Scheme 1
results in an electrophilic protein-bound isothiocyanate (3) that may react with protein nucleophiles. Although dithiocarbamate formation can occur on both the R terminal amino group and -lysyl amino groups, only the latter site of reaction appears capable of proceeding to give the isothiocyanate (7). Reaction between the proteinbound isothiocyanate and a sulfhydryl group produces a thioester cross-link (4), whereas reaction with the -amino group of a second Lys generates a thiourea cross-link (5). Among the three possible cross-linking structures produced by CS2, the disulfide and dithiocarbamate ester are reversible under physiological conditions, while the thiourea cross-link between two Lys residues is not reversible and is expected to be the most stable in vivo. Figure 5. Dose response relating quantity of Lys-Lys thiourea detected versus exposure level at 13-week time point. Lys-Lys thiourea signal was normalized to both the amount of protein hydrolyzed and the internal standard intensity. Error bars represent standard errors (n ) 3).
NMR spectroscopy (7, 15) and showed the generation of both inter- and intramolecular cross-links. CS2-mediated protein cross-linking was subsequently reinvestigated (16), also using BSA, although in that investigation it was concluded that only intramolecular cross-links were produced. Both previous studies demonstrated that cross-linking chemistry begins with the derivatization of an amino group by CS2 to form an -monoalkyldithiocarbamate (1) (Scheme 1). Subsequently, disulfide formation can occur quickly through oxidative coupling of two dithiocarbamates, or a dithiocarbamate and cysteine, to generate a bis(thiocarbamoyl) disulfide (2) crosslinking structure. Alternatively, the release of hydrogen sulfide ion from the dithiocarbamate occurs slowly and
CS2 has been shown to produce high-molecular-weight species of spectrin and neurofilament proteins in vivo (11). These high-molecular-weight proteins were interpreted to be the result of intermolecular cross-links, although the structure of the proposed cross-link was not determined. Previous work has shown that the highmolecular-weight bands of spectrin consist of R,β heterodimers formed by covalent bond formation between two spectrin subunit monomers and that these heterodimers display a cumulative dose response. In vitro experiments using model proteins have suggested that among the identified CS2-generated cross-linking structures, a thiourea cross-link between two Lys residues would seem the most likely to accumulate in vivo due to the irreversibility of the thiourea bond. The stability of the thiourea cross-linking structure is demonstrated in the present study by its ability to withstand standard protein hydrolysis conditions (6 N HCl, 18 h at 110 °C). Accordingly, using LC/MS/MS, the present work provides structural evidence for the production of Lys-Lys thiourea
Thiourea Cross-Linking of Spectrin in Vivo
cross-links in vivo on spectrin from rats exposed to CS2 or DEDC. It is expected that the close proximity of spectrin subunits on the inner surface of the cell membrane facilitates cross-linking of Lys residues positioned on opposite strands resulting in the formation of heterodimers, although the possibility that additional intramolecular thiourea cross-links also are formed cannot be excluded. There is evidence presented elsewhere that thioester cross-links are also formed on spectrin (9) in addition to the Lys-Lys thiourea cross-links detected in this work. Examination of the dose-response curve for Lys-Lys thiourea indicates a positive correlation (Figure 5) between dose and amount of Lys-Lys thiourea, with a statistically significant difference in quantity of Lys-Lys thiourea detected between either 500 and 800 ppm exposures and controls. However, differences between adjacent groups, such as controls and 50 ppm or 500 and 800 ppm, were not statistically significant. This lack of statistical power may be overcome by analyzing a greater number of samples. However, our primary objective was the detection of Lys-Lys thiourea cross-links in treated rats in order to support their mechanistic pertinence to the neurotoxicity of CS2. The analytical methods developed in the present investigation may further advance studies into the mechanism of CS2 neurotoxicity by allowing evaluation of cross-linking in neurofilament proteins. The identification of Lys-Lys thiourea cross-linking structures produced in vivo by CS2 supports the relevance of the protein cross-linking chemistry previously delineated in vitro for CS2. These data suggest that the highmolecular-weight species of spectrin and neurofilament proteins observed by SDS-PAGE after exposure to CS2 are at least in part due to CS2-mediated Lys-Lys thiourea intermolecular cross-links. The ability of DEDC to produce thiourea protein cross-links in vivo is also consistent with previous work (5) and supports the release of CS2 as a contributing mechanism in the neurotoxicity associated with subchronic exposures to DEDC. This latter finding is relevant considering the widespread use of dithiocarbamates in industry, agriculture, and medicine.
Acknowledgment. We would like to thank Dr. Ruedi Aebersold (Department of Molecular Biotechnology, University of Washington) for the generous gift of PETAPITC used in these experiments. Deadre Johnson and Holly Valentine provided the spectrin from the DEDC-treated rats. We would also like to thank the Vanderbilt Mass Spectrometry Resource for use of the mass spectrometer.
Chem. Res. Toxicol., Vol. 11, No. 5, 1998 549
These studies were supported by Grant ES06387 and Center in Molecular Toxicology Grant P30 ES00267. J.C.L.E. was supported in part by NRSA Grant ES05764.
References (1) Beauchamp, R. O., Bus, J. S., Popp, J. A., Boreiko, C. J., and Golberg, L. (1983) A critical review of the literature on carbon disulfide toxicity. CRC Crit. Rev. Toxicol. 11, 169-278. (2) Chu, C. C., Huang, C. C., Chen, R. S., and Shih, T. S. (1995) Polyneuropathy induced by carbon disulfide in viscose rayon workers. Occup. Environ. Med. 52, 404-407. (3) Ruijten, M. W., Salle, H. J., and Verberk, M. M. (1993) Verification of effects on the nervous system of low level occupational exposure to CS2. Br. J. Ind. Med. 50, 301-307. (4) Egeland, G. M., Burkhart, G. A., Schnorr, T. M., Hornung, R. W., Fajen, J. M., and Lee, S. T. (1992) Effects of exposure to carbon disulfide on low-density lipoprotein cholesterol concentration and diastolic blood pressure. Br. J. Ind. Med. 49, 287-293. (5) Johnson, D. J., Graham, D. G., Amarnath, V., Amarnath, K., and Valentine, W. M. (1998) Release of carbon disulfide is a contributing mechanism in the axonopathy produced by N,N-diethyldithiocarbamate. Toxicol. Appl. Pharmacol. 148, 288-296. (6) Graham, D. G., Amarnath, V., Valentine, W. M., Pyle, S. J., and Anthony, D. C. (1995) Pathogenetic studies of hexane and carbon disulfide neurotoxicity. Crit. Rev. Toxicol. 25, 91-112. (7) Valentine, W. M., Amarnath, V., Graham, D. G., and Anthony, D. C. (1992) Covalent cross-linking of proteins by carbon disulfide. Chem. Res. Toxicol. 5, 254-262. (8) Valentine, W. M., Amarnath, V., Amarnath, K., Erve, J. C. L., Graham, D. G., Morgan, D. L., and Sills, R. C. (1998) Covalent modification of hemoglobin by carbon disulfide: A potential biomarker of effect. Neurotoxicology 19, 99-108. (9) Valentine, W. M., Graham, D. G., and Anthony, D. C. (1993) Covalent cross-linking of erythrocyte spectrin by carbon disulfide in vivo. Toxicol. Appl. Pharmacol. 121, 71-77. (10) Neuberger, A., and Sanger, F. (1943) The availability of the acetyl derivatives of lysine for growth. Biochem. J. 37, 515-518. (11) Valentine, W. M., Amarnath, V., Graham, D. G., Morgan, D. L., and Sills, R. C. (1997) CS2-mediated cross-linking of erythrocyte spectrin and neurofilament protein: Dose response and temporal relationship to the formation of axonal swellings. Toxicol. Appl. Pharmacol. 142, 95-105. (12) Johnson, D. J., Graham, D. G., Amarnath, V., Amarnath, K., and Valentine, W. M. (1996) The measurement of 2-thiothiazolidine4-carboxylic acid as an index of the in vivo release of CS2 by dithiocarbamates. Chem. Res. Toxicol. 9, 910-916. (13) Light, N. D., and Bailey, A. J. (1982) Covalent cross-links in collagen. Methods Enzymol. 82, 360-372. (14) Aebersold, R., Bures, E. J., Namchuk, M., Goghari, M. H., Shushan, B., and Covey, T. C. (1992) Design, synthesis, and characterization of a protein sequencing reagent yielding amino acid derivatives with enhanced detectability by mass spectrometry. Protein Sci. 1, 494-503. (15) Amarnath, V., Anthony, D. C., Valentine, W. M., and Graham, D. G. (1991) The molecular mechanism of the carbon disulfide mediated cross-linking of proteins. Chem. Res. Toxicol. 4, 148150. (16) DeCaprio, A. P., Spink, D. C., Chen, X., Fowke, J. H., Zhu, M. S., and Bank, S. (1992) Characterization of isothiocyanates, thioureas, and other lysine adduction products in carbon disulfidetreated peptides and proteins. Chem. Res. Toxicol. 5, 496-504.
TX980007I
